diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15079 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.998837479655894,
+  "eval_steps": 500,
+  "global_step": 2148,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0013950244129272262,
+      "grad_norm": 9.061795042102073,
+      "learning_rate": 0.0,
+      "loss": 1.2425,
+      "step": 1
+    },
+    {
+      "epoch": 0.0027900488258544524,
+      "grad_norm": 7.74265471964034,
+      "learning_rate": 4.6511627906976744e-08,
+      "loss": 1.1023,
+      "step": 2
+    },
+    {
+      "epoch": 0.0041850732387816785,
+      "grad_norm": 6.389720437917294,
+      "learning_rate": 9.302325581395349e-08,
+      "loss": 0.9673,
+      "step": 3
+    },
+    {
+      "epoch": 0.005580097651708905,
+      "grad_norm": 8.421817121439247,
+      "learning_rate": 1.3953488372093024e-07,
+      "loss": 1.1736,
+      "step": 4
+    },
+    {
+      "epoch": 0.006975122064636131,
+      "grad_norm": 7.580574098658975,
+      "learning_rate": 1.8604651162790698e-07,
+      "loss": 1.0903,
+      "step": 5
+    },
+    {
+      "epoch": 0.008370146477563357,
+      "grad_norm": 8.104772232891781,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 1.1323,
+      "step": 6
+    },
+    {
+      "epoch": 0.009765170890490584,
+      "grad_norm": 8.156214672673578,
+      "learning_rate": 2.790697674418605e-07,
+      "loss": 1.218,
+      "step": 7
+    },
+    {
+      "epoch": 0.01116019530341781,
+      "grad_norm": 8.526741313250225,
+      "learning_rate": 3.2558139534883724e-07,
+      "loss": 1.231,
+      "step": 8
+    },
+    {
+      "epoch": 0.012555219716345037,
+      "grad_norm": 6.989791978173938,
+      "learning_rate": 3.7209302325581396e-07,
+      "loss": 1.0199,
+      "step": 9
+    },
+    {
+      "epoch": 0.013950244129272262,
+      "grad_norm": 7.652571631562263,
+      "learning_rate": 4.186046511627907e-07,
+      "loss": 1.0509,
+      "step": 10
+    },
+    {
+      "epoch": 0.015345268542199489,
+      "grad_norm": 8.029656913285312,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 1.1003,
+      "step": 11
+    },
+    {
+      "epoch": 0.016740292955126714,
+      "grad_norm": 6.822970986263167,
+      "learning_rate": 5.116279069767442e-07,
+      "loss": 1.0076,
+      "step": 12
+    },
+    {
+      "epoch": 0.01813531736805394,
+      "grad_norm": 7.446027562629797,
+      "learning_rate": 5.58139534883721e-07,
+      "loss": 1.1319,
+      "step": 13
+    },
+    {
+      "epoch": 0.01953034178098117,
+      "grad_norm": 8.32595611951422,
+      "learning_rate": 6.046511627906977e-07,
+      "loss": 1.1838,
+      "step": 14
+    },
+    {
+      "epoch": 0.020925366193908394,
+      "grad_norm": 6.223433909698243,
+      "learning_rate": 6.511627906976745e-07,
+      "loss": 0.9731,
+      "step": 15
+    },
+    {
+      "epoch": 0.02232039060683562,
+      "grad_norm": 7.0235311966869585,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 1.0504,
+      "step": 16
+    },
+    {
+      "epoch": 0.023715415019762844,
+      "grad_norm": 5.211705448894807,
+      "learning_rate": 7.441860465116279e-07,
+      "loss": 0.8548,
+      "step": 17
+    },
+    {
+      "epoch": 0.025110439432690073,
+      "grad_norm": 7.825683956735391,
+      "learning_rate": 7.906976744186047e-07,
+      "loss": 1.1983,
+      "step": 18
+    },
+    {
+      "epoch": 0.0265054638456173,
+      "grad_norm": 5.0106067786805575,
+      "learning_rate": 8.372093023255814e-07,
+      "loss": 0.9202,
+      "step": 19
+    },
+    {
+      "epoch": 0.027900488258544524,
+      "grad_norm": 6.830548485045446,
+      "learning_rate": 8.837209302325582e-07,
+      "loss": 1.0445,
+      "step": 20
+    },
+    {
+      "epoch": 0.029295512671471752,
+      "grad_norm": 5.726619544487671,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.9837,
+      "step": 21
+    },
+    {
+      "epoch": 0.030690537084398978,
+      "grad_norm": 5.8000958536016265,
+      "learning_rate": 9.767441860465117e-07,
+      "loss": 1.0272,
+      "step": 22
+    },
+    {
+      "epoch": 0.03208556149732621,
+      "grad_norm": 5.703395850340747,
+      "learning_rate": 1.0232558139534884e-06,
+      "loss": 1.0211,
+      "step": 23
+    },
+    {
+      "epoch": 0.03348058591025343,
+      "grad_norm": 3.586220906876636,
+      "learning_rate": 1.0697674418604653e-06,
+      "loss": 0.834,
+      "step": 24
+    },
+    {
+      "epoch": 0.03487561032318066,
+      "grad_norm": 4.9748168215918,
+      "learning_rate": 1.116279069767442e-06,
+      "loss": 0.9711,
+      "step": 25
+    },
+    {
+      "epoch": 0.03627063473610788,
+      "grad_norm": 4.018635863811052,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.9558,
+      "step": 26
+    },
+    {
+      "epoch": 0.03766565914903511,
+      "grad_norm": 5.03749104808541,
+      "learning_rate": 1.2093023255813954e-06,
+      "loss": 1.0273,
+      "step": 27
+    },
+    {
+      "epoch": 0.03906068356196234,
+      "grad_norm": 3.9324782587964053,
+      "learning_rate": 1.2558139534883723e-06,
+      "loss": 0.9129,
+      "step": 28
+    },
+    {
+      "epoch": 0.04045570797488956,
+      "grad_norm": 3.550424744823837,
+      "learning_rate": 1.302325581395349e-06,
+      "loss": 0.8896,
+      "step": 29
+    },
+    {
+      "epoch": 0.04185073238781679,
+      "grad_norm": 3.446948058404404,
+      "learning_rate": 1.3488372093023258e-06,
+      "loss": 0.8867,
+      "step": 30
+    },
+    {
+      "epoch": 0.043245756800744016,
+      "grad_norm": 4.032262199203236,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.8187,
+      "step": 31
+    },
+    {
+      "epoch": 0.04464078121367124,
+      "grad_norm": 4.125895000157052,
+      "learning_rate": 1.4418604651162794e-06,
+      "loss": 0.899,
+      "step": 32
+    },
+    {
+      "epoch": 0.04603580562659847,
+      "grad_norm": 3.8704377788889506,
+      "learning_rate": 1.4883720930232558e-06,
+      "loss": 0.9593,
+      "step": 33
+    },
+    {
+      "epoch": 0.04743083003952569,
+      "grad_norm": 3.9800203887919126,
+      "learning_rate": 1.534883720930233e-06,
+      "loss": 0.9171,
+      "step": 34
+    },
+    {
+      "epoch": 0.04882585445245292,
+      "grad_norm": 3.6904834931603605,
+      "learning_rate": 1.5813953488372093e-06,
+      "loss": 0.913,
+      "step": 35
+    },
+    {
+      "epoch": 0.050220878865380146,
+      "grad_norm": 3.63034156122338,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 0.9465,
+      "step": 36
+    },
+    {
+      "epoch": 0.05161590327830737,
+      "grad_norm": 3.1225625019387353,
+      "learning_rate": 1.6744186046511629e-06,
+      "loss": 0.8543,
+      "step": 37
+    },
+    {
+      "epoch": 0.0530109276912346,
+      "grad_norm": 3.0282628625387824,
+      "learning_rate": 1.7209302325581397e-06,
+      "loss": 0.7939,
+      "step": 38
+    },
+    {
+      "epoch": 0.054405952104161825,
+      "grad_norm": 3.1020850997672964,
+      "learning_rate": 1.7674418604651164e-06,
+      "loss": 0.9034,
+      "step": 39
+    },
+    {
+      "epoch": 0.05580097651708905,
+      "grad_norm": 3.352063702118774,
+      "learning_rate": 1.8139534883720933e-06,
+      "loss": 0.8522,
+      "step": 40
+    },
+    {
+      "epoch": 0.057196000930016276,
+      "grad_norm": 3.1699146178254916,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.8169,
+      "step": 41
+    },
+    {
+      "epoch": 0.058591025342943505,
+      "grad_norm": 3.158963197691139,
+      "learning_rate": 1.9069767441860468e-06,
+      "loss": 0.8102,
+      "step": 42
+    },
+    {
+      "epoch": 0.05998604975587073,
+      "grad_norm": 3.2617484927812606,
+      "learning_rate": 1.9534883720930235e-06,
+      "loss": 0.792,
+      "step": 43
+    },
+    {
+      "epoch": 0.061381074168797956,
+      "grad_norm": 3.087539810163175,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.826,
+      "step": 44
+    },
+    {
+      "epoch": 0.06277609858172518,
+      "grad_norm": 3.3949003321789726,
+      "learning_rate": 2.0465116279069768e-06,
+      "loss": 0.8794,
+      "step": 45
+    },
+    {
+      "epoch": 0.06417112299465241,
+      "grad_norm": 2.9635898587108325,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.7587,
+      "step": 46
+    },
+    {
+      "epoch": 0.06556614740757963,
+      "grad_norm": 3.089235385596214,
+      "learning_rate": 2.1395348837209305e-06,
+      "loss": 0.8536,
+      "step": 47
+    },
+    {
+      "epoch": 0.06696117182050686,
+      "grad_norm": 2.800805655960869,
+      "learning_rate": 2.1860465116279074e-06,
+      "loss": 0.8456,
+      "step": 48
+    },
+    {
+      "epoch": 0.06835619623343409,
+      "grad_norm": 2.6203629396683437,
+      "learning_rate": 2.232558139534884e-06,
+      "loss": 0.8775,
+      "step": 49
+    },
+    {
+      "epoch": 0.06975122064636131,
+      "grad_norm": 2.8227128567921405,
+      "learning_rate": 2.2790697674418607e-06,
+      "loss": 0.8134,
+      "step": 50
+    },
+    {
+      "epoch": 0.07114624505928854,
+      "grad_norm": 2.7691797695343614,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.8683,
+      "step": 51
+    },
+    {
+      "epoch": 0.07254126947221576,
+      "grad_norm": 2.5774573516975074,
+      "learning_rate": 2.3720930232558144e-06,
+      "loss": 0.8091,
+      "step": 52
+    },
+    {
+      "epoch": 0.07393629388514299,
+      "grad_norm": 2.813557786400301,
+      "learning_rate": 2.418604651162791e-06,
+      "loss": 0.8208,
+      "step": 53
+    },
+    {
+      "epoch": 0.07533131829807022,
+      "grad_norm": 2.9437387416240823,
+      "learning_rate": 2.4651162790697678e-06,
+      "loss": 0.8564,
+      "step": 54
+    },
+    {
+      "epoch": 0.07672634271099744,
+      "grad_norm": 2.6877272303391324,
+      "learning_rate": 2.5116279069767446e-06,
+      "loss": 0.7985,
+      "step": 55
+    },
+    {
+      "epoch": 0.07812136712392467,
+      "grad_norm": 2.6939943804651314,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.8415,
+      "step": 56
+    },
+    {
+      "epoch": 0.07951639153685189,
+      "grad_norm": 2.680875520170419,
+      "learning_rate": 2.604651162790698e-06,
+      "loss": 0.8192,
+      "step": 57
+    },
+    {
+      "epoch": 0.08091141594977912,
+      "grad_norm": 2.5440723987372666,
+      "learning_rate": 2.6511627906976744e-06,
+      "loss": 0.7869,
+      "step": 58
+    },
+    {
+      "epoch": 0.08230644036270635,
+      "grad_norm": 2.7357452934000164,
+      "learning_rate": 2.6976744186046517e-06,
+      "loss": 0.8903,
+      "step": 59
+    },
+    {
+      "epoch": 0.08370146477563357,
+      "grad_norm": 2.4887440165658283,
+      "learning_rate": 2.744186046511628e-06,
+      "loss": 0.7487,
+      "step": 60
+    },
+    {
+      "epoch": 0.0850964891885608,
+      "grad_norm": 2.606394012411958,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.7484,
+      "step": 61
+    },
+    {
+      "epoch": 0.08649151360148803,
+      "grad_norm": 2.551608507364557,
+      "learning_rate": 2.8372093023255815e-06,
+      "loss": 0.8419,
+      "step": 62
+    },
+    {
+      "epoch": 0.08788653801441525,
+      "grad_norm": 2.457113149714823,
+      "learning_rate": 2.8837209302325587e-06,
+      "loss": 0.7387,
+      "step": 63
+    },
+    {
+      "epoch": 0.08928156242734248,
+      "grad_norm": 2.6907615175482293,
+      "learning_rate": 2.930232558139535e-06,
+      "loss": 0.8337,
+      "step": 64
+    },
+    {
+      "epoch": 0.0906765868402697,
+      "grad_norm": 2.394218938845241,
+      "learning_rate": 2.9767441860465116e-06,
+      "loss": 0.7526,
+      "step": 65
+    },
+    {
+      "epoch": 0.09207161125319693,
+      "grad_norm": 2.5987172501099605,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.8057,
+      "step": 66
+    },
+    {
+      "epoch": 0.09346663566612416,
+      "grad_norm": 2.5273926089912058,
+      "learning_rate": 3.069767441860466e-06,
+      "loss": 0.8628,
+      "step": 67
+    },
+    {
+      "epoch": 0.09486166007905138,
+      "grad_norm": 2.539576955247472,
+      "learning_rate": 3.1162790697674423e-06,
+      "loss": 0.8137,
+      "step": 68
+    },
+    {
+      "epoch": 0.0962566844919786,
+      "grad_norm": 2.4735207026102652,
+      "learning_rate": 3.1627906976744187e-06,
+      "loss": 0.7567,
+      "step": 69
+    },
+    {
+      "epoch": 0.09765170890490583,
+      "grad_norm": 2.329402847723333,
+      "learning_rate": 3.2093023255813956e-06,
+      "loss": 0.7229,
+      "step": 70
+    },
+    {
+      "epoch": 0.09904673331783306,
+      "grad_norm": 2.51709324412196,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.7585,
+      "step": 71
+    },
+    {
+      "epoch": 0.10044175773076029,
+      "grad_norm": 2.5097007626658403,
+      "learning_rate": 3.3023255813953493e-06,
+      "loss": 0.7643,
+      "step": 72
+    },
+    {
+      "epoch": 0.10183678214368752,
+      "grad_norm": 2.516406479315615,
+      "learning_rate": 3.3488372093023258e-06,
+      "loss": 0.7743,
+      "step": 73
+    },
+    {
+      "epoch": 0.10323180655661474,
+      "grad_norm": 2.39487772273171,
+      "learning_rate": 3.3953488372093026e-06,
+      "loss": 0.7836,
+      "step": 74
+    },
+    {
+      "epoch": 0.10462683096954196,
+      "grad_norm": 2.5591412684982573,
+      "learning_rate": 3.4418604651162795e-06,
+      "loss": 0.7613,
+      "step": 75
+    },
+    {
+      "epoch": 0.1060218553824692,
+      "grad_norm": 2.4784750402218925,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.6626,
+      "step": 76
+    },
+    {
+      "epoch": 0.10741687979539642,
+      "grad_norm": 2.1253121107818203,
+      "learning_rate": 3.534883720930233e-06,
+      "loss": 0.6501,
+      "step": 77
+    },
+    {
+      "epoch": 0.10881190420832365,
+      "grad_norm": 2.4808216335670537,
+      "learning_rate": 3.5813953488372093e-06,
+      "loss": 0.794,
+      "step": 78
+    },
+    {
+      "epoch": 0.11020692862125087,
+      "grad_norm": 2.326427938400137,
+      "learning_rate": 3.6279069767441866e-06,
+      "loss": 0.7634,
+      "step": 79
+    },
+    {
+      "epoch": 0.1116019530341781,
+      "grad_norm": 2.6060325903250603,
+      "learning_rate": 3.674418604651163e-06,
+      "loss": 0.8042,
+      "step": 80
+    },
+    {
+      "epoch": 0.11299697744710532,
+      "grad_norm": 2.4821373099188615,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.762,
+      "step": 81
+    },
+    {
+      "epoch": 0.11439200186003255,
+      "grad_norm": 2.898157554318415,
+      "learning_rate": 3.7674418604651163e-06,
+      "loss": 0.8571,
+      "step": 82
+    },
+    {
+      "epoch": 0.11578702627295978,
+      "grad_norm": 2.4843334688361374,
+      "learning_rate": 3.8139534883720936e-06,
+      "loss": 0.7576,
+      "step": 83
+    },
+    {
+      "epoch": 0.11718205068588701,
+      "grad_norm": 2.387738096347522,
+      "learning_rate": 3.86046511627907e-06,
+      "loss": 0.7543,
+      "step": 84
+    },
+    {
+      "epoch": 0.11857707509881422,
+      "grad_norm": 2.335434248204514,
+      "learning_rate": 3.906976744186047e-06,
+      "loss": 0.7395,
+      "step": 85
+    },
+    {
+      "epoch": 0.11997209951174145,
+      "grad_norm": 2.3375745829156584,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.6685,
+      "step": 86
+    },
+    {
+      "epoch": 0.12136712392466868,
+      "grad_norm": 2.5136417014675074,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8063,
+      "step": 87
+    },
+    {
+      "epoch": 0.12276214833759591,
+      "grad_norm": 2.5442316547663584,
+      "learning_rate": 4.0465116279069775e-06,
+      "loss": 0.8206,
+      "step": 88
+    },
+    {
+      "epoch": 0.12415717275052314,
+      "grad_norm": 2.555591985292735,
+      "learning_rate": 4.0930232558139536e-06,
+      "loss": 0.8606,
+      "step": 89
+    },
+    {
+      "epoch": 0.12555219716345037,
+      "grad_norm": 2.4405315538013808,
+      "learning_rate": 4.1395348837209304e-06,
+      "loss": 0.7434,
+      "step": 90
+    },
+    {
+      "epoch": 0.12694722157637758,
+      "grad_norm": 2.398256087733658,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.7421,
+      "step": 91
+    },
+    {
+      "epoch": 0.12834224598930483,
+      "grad_norm": 2.4927763742524203,
+      "learning_rate": 4.232558139534884e-06,
+      "loss": 0.7935,
+      "step": 92
+    },
+    {
+      "epoch": 0.12973727040223204,
+      "grad_norm": 2.5428069514715994,
+      "learning_rate": 4.279069767441861e-06,
+      "loss": 0.7443,
+      "step": 93
+    },
+    {
+      "epoch": 0.13113229481515926,
+      "grad_norm": 2.364288381058622,
+      "learning_rate": 4.325581395348837e-06,
+      "loss": 0.7112,
+      "step": 94
+    },
+    {
+      "epoch": 0.1325273192280865,
+      "grad_norm": 2.011169128440702,
+      "learning_rate": 4.372093023255815e-06,
+      "loss": 0.6569,
+      "step": 95
+    },
+    {
+      "epoch": 0.1339223436410137,
+      "grad_norm": 2.349408114611445,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.7492,
+      "step": 96
+    },
+    {
+      "epoch": 0.13531736805394096,
+      "grad_norm": 2.4540877828585677,
+      "learning_rate": 4.465116279069768e-06,
+      "loss": 0.7292,
+      "step": 97
+    },
+    {
+      "epoch": 0.13671239246686817,
+      "grad_norm": 2.3254215830382963,
+      "learning_rate": 4.5116279069767445e-06,
+      "loss": 0.6947,
+      "step": 98
+    },
+    {
+      "epoch": 0.13810741687979539,
+      "grad_norm": 2.2929641391200986,
+      "learning_rate": 4.558139534883721e-06,
+      "loss": 0.7623,
+      "step": 99
+    },
+    {
+      "epoch": 0.13950244129272263,
+      "grad_norm": 2.34717403373499,
+      "learning_rate": 4.604651162790698e-06,
+      "loss": 0.6874,
+      "step": 100
+    },
+    {
+      "epoch": 0.14089746570564984,
+      "grad_norm": 2.4456769972692065,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.771,
+      "step": 101
+    },
+    {
+      "epoch": 0.1422924901185771,
+      "grad_norm": 2.418110442277691,
+      "learning_rate": 4.697674418604651e-06,
+      "loss": 0.7647,
+      "step": 102
+    },
+    {
+      "epoch": 0.1436875145315043,
+      "grad_norm": 2.574913622817416,
+      "learning_rate": 4.744186046511629e-06,
+      "loss": 0.7203,
+      "step": 103
+    },
+    {
+      "epoch": 0.14508253894443152,
+      "grad_norm": 2.389331829708997,
+      "learning_rate": 4.790697674418605e-06,
+      "loss": 0.7176,
+      "step": 104
+    },
+    {
+      "epoch": 0.14647756335735876,
+      "grad_norm": 2.422530481112432,
+      "learning_rate": 4.837209302325582e-06,
+      "loss": 0.7032,
+      "step": 105
+    },
+    {
+      "epoch": 0.14787258777028597,
+      "grad_norm": 2.4343657094034574,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.6991,
+      "step": 106
+    },
+    {
+      "epoch": 0.14926761218321322,
+      "grad_norm": 2.3962869391793884,
+      "learning_rate": 4.9302325581395355e-06,
+      "loss": 0.7668,
+      "step": 107
+    },
+    {
+      "epoch": 0.15066263659614043,
+      "grad_norm": 2.3514460901020304,
+      "learning_rate": 4.976744186046512e-06,
+      "loss": 0.738,
+      "step": 108
+    },
+    {
+      "epoch": 0.15205766100906765,
+      "grad_norm": 2.7488872789310093,
+      "learning_rate": 5.023255813953489e-06,
+      "loss": 0.9257,
+      "step": 109
+    },
+    {
+      "epoch": 0.1534526854219949,
+      "grad_norm": 2.1877704055112397,
+      "learning_rate": 5.069767441860466e-06,
+      "loss": 0.6922,
+      "step": 110
+    },
+    {
+      "epoch": 0.1548477098349221,
+      "grad_norm": 2.292046125466671,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.7168,
+      "step": 111
+    },
+    {
+      "epoch": 0.15624273424784935,
+      "grad_norm": 2.39558344544093,
+      "learning_rate": 5.162790697674419e-06,
+      "loss": 0.7668,
+      "step": 112
+    },
+    {
+      "epoch": 0.15763775866077656,
+      "grad_norm": 2.37532566612421,
+      "learning_rate": 5.209302325581396e-06,
+      "loss": 0.7413,
+      "step": 113
+    },
+    {
+      "epoch": 0.15903278307370378,
+      "grad_norm": 2.570621775062793,
+      "learning_rate": 5.255813953488372e-06,
+      "loss": 0.7662,
+      "step": 114
+    },
+    {
+      "epoch": 0.16042780748663102,
+      "grad_norm": 2.437707261804497,
+      "learning_rate": 5.302325581395349e-06,
+      "loss": 0.7503,
+      "step": 115
+    },
+    {
+      "epoch": 0.16182283189955823,
+      "grad_norm": 2.3490732373961993,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.7234,
+      "step": 116
+    },
+    {
+      "epoch": 0.16321785631248548,
+      "grad_norm": 2.553878394018376,
+      "learning_rate": 5.395348837209303e-06,
+      "loss": 0.7924,
+      "step": 117
+    },
+    {
+      "epoch": 0.1646128807254127,
+      "grad_norm": 2.53998019160818,
+      "learning_rate": 5.44186046511628e-06,
+      "loss": 0.7456,
+      "step": 118
+    },
+    {
+      "epoch": 0.16600790513833993,
+      "grad_norm": 2.5752545753938216,
+      "learning_rate": 5.488372093023256e-06,
+      "loss": 0.7171,
+      "step": 119
+    },
+    {
+      "epoch": 0.16740292955126715,
+      "grad_norm": 2.5025495310223866,
+      "learning_rate": 5.534883720930233e-06,
+      "loss": 0.7204,
+      "step": 120
+    },
+    {
+      "epoch": 0.16879795396419436,
+      "grad_norm": 2.465879917317966,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.7511,
+      "step": 121
+    },
+    {
+      "epoch": 0.1701929783771216,
+      "grad_norm": 2.4842892653706583,
+      "learning_rate": 5.627906976744186e-06,
+      "loss": 0.7705,
+      "step": 122
+    },
+    {
+      "epoch": 0.17158800279004882,
+      "grad_norm": 2.656958791529201,
+      "learning_rate": 5.674418604651163e-06,
+      "loss": 0.7901,
+      "step": 123
+    },
+    {
+      "epoch": 0.17298302720297606,
+      "grad_norm": 2.5105437832639,
+      "learning_rate": 5.72093023255814e-06,
+      "loss": 0.731,
+      "step": 124
+    },
+    {
+      "epoch": 0.17437805161590328,
+      "grad_norm": 2.307050253420812,
+      "learning_rate": 5.7674418604651175e-06,
+      "loss": 0.7704,
+      "step": 125
+    },
+    {
+      "epoch": 0.1757730760288305,
+      "grad_norm": 2.4819230064286706,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.7553,
+      "step": 126
+    },
+    {
+      "epoch": 0.17716810044175774,
+      "grad_norm": 2.6590280236699106,
+      "learning_rate": 5.86046511627907e-06,
+      "loss": 0.7641,
+      "step": 127
+    },
+    {
+      "epoch": 0.17856312485468495,
+      "grad_norm": 2.3396747269388696,
+      "learning_rate": 5.906976744186047e-06,
+      "loss": 0.76,
+      "step": 128
+    },
+    {
+      "epoch": 0.1799581492676122,
+      "grad_norm": 2.3959239000906263,
+      "learning_rate": 5.953488372093023e-06,
+      "loss": 0.83,
+      "step": 129
+    },
+    {
+      "epoch": 0.1813531736805394,
+      "grad_norm": 2.4068741447922277,
+      "learning_rate": 6e-06,
+      "loss": 0.7451,
+      "step": 130
+    },
+    {
+      "epoch": 0.18274819809346662,
+      "grad_norm": 2.5306239349683977,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.6759,
+      "step": 131
+    },
+    {
+      "epoch": 0.18414322250639387,
+      "grad_norm": 2.445947719768049,
+      "learning_rate": 6.093023255813954e-06,
+      "loss": 0.769,
+      "step": 132
+    },
+    {
+      "epoch": 0.18553824691932108,
+      "grad_norm": 2.5399331456936274,
+      "learning_rate": 6.139534883720932e-06,
+      "loss": 0.7435,
+      "step": 133
+    },
+    {
+      "epoch": 0.18693327133224832,
+      "grad_norm": 2.2974985090948845,
+      "learning_rate": 6.186046511627908e-06,
+      "loss": 0.7345,
+      "step": 134
+    },
+    {
+      "epoch": 0.18832829574517554,
+      "grad_norm": 2.6237877338647873,
+      "learning_rate": 6.2325581395348845e-06,
+      "loss": 0.8024,
+      "step": 135
+    },
+    {
+      "epoch": 0.18972332015810275,
+      "grad_norm": 2.5524074818038485,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.7748,
+      "step": 136
+    },
+    {
+      "epoch": 0.19111834457103,
+      "grad_norm": 2.562006189602683,
+      "learning_rate": 6.325581395348837e-06,
+      "loss": 0.753,
+      "step": 137
+    },
+    {
+      "epoch": 0.1925133689839572,
+      "grad_norm": 2.4958479653255545,
+      "learning_rate": 6.372093023255814e-06,
+      "loss": 0.7657,
+      "step": 138
+    },
+    {
+      "epoch": 0.19390839339688445,
+      "grad_norm": 2.717321817819931,
+      "learning_rate": 6.418604651162791e-06,
+      "loss": 0.71,
+      "step": 139
+    },
+    {
+      "epoch": 0.19530341780981167,
+      "grad_norm": 2.4089710684455983,
+      "learning_rate": 6.465116279069767e-06,
+      "loss": 0.7847,
+      "step": 140
+    },
+    {
+      "epoch": 0.1966984422227389,
+      "grad_norm": 2.477539261318324,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.7988,
+      "step": 141
+    },
+    {
+      "epoch": 0.19809346663566613,
+      "grad_norm": 2.6298344335984685,
+      "learning_rate": 6.558139534883722e-06,
+      "loss": 0.7494,
+      "step": 142
+    },
+    {
+      "epoch": 0.19948849104859334,
+      "grad_norm": 2.37624775861681,
+      "learning_rate": 6.604651162790699e-06,
+      "loss": 0.7409,
+      "step": 143
+    },
+    {
+      "epoch": 0.20088351546152058,
+      "grad_norm": 2.3379677755783597,
+      "learning_rate": 6.651162790697675e-06,
+      "loss": 0.7019,
+      "step": 144
+    },
+    {
+      "epoch": 0.2022785398744478,
+      "grad_norm": 2.404331847659135,
+      "learning_rate": 6.6976744186046515e-06,
+      "loss": 0.7986,
+      "step": 145
+    },
+    {
+      "epoch": 0.20367356428737504,
+      "grad_norm": 2.329041308240456,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.7384,
+      "step": 146
+    },
+    {
+      "epoch": 0.20506858870030226,
+      "grad_norm": 2.554766751041686,
+      "learning_rate": 6.790697674418605e-06,
+      "loss": 0.8329,
+      "step": 147
+    },
+    {
+      "epoch": 0.20646361311322947,
+      "grad_norm": 2.380435295817839,
+      "learning_rate": 6.837209302325581e-06,
+      "loss": 0.7275,
+      "step": 148
+    },
+    {
+      "epoch": 0.20785863752615671,
+      "grad_norm": 2.542245175246868,
+      "learning_rate": 6.883720930232559e-06,
+      "loss": 0.8281,
+      "step": 149
+    },
+    {
+      "epoch": 0.20925366193908393,
+      "grad_norm": 2.436384871263683,
+      "learning_rate": 6.930232558139536e-06,
+      "loss": 0.763,
+      "step": 150
+    },
+    {
+      "epoch": 0.21064868635201117,
+      "grad_norm": 2.577321451306135,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.7587,
+      "step": 151
+    },
+    {
+      "epoch": 0.2120437107649384,
+      "grad_norm": 2.1824061916227544,
+      "learning_rate": 7.023255813953489e-06,
+      "loss": 0.6997,
+      "step": 152
+    },
+    {
+      "epoch": 0.2134387351778656,
+      "grad_norm": 2.5006587847408013,
+      "learning_rate": 7.069767441860466e-06,
+      "loss": 0.7555,
+      "step": 153
+    },
+    {
+      "epoch": 0.21483375959079284,
+      "grad_norm": 2.357971605450965,
+      "learning_rate": 7.1162790697674425e-06,
+      "loss": 0.7273,
+      "step": 154
+    },
+    {
+      "epoch": 0.21622878400372006,
+      "grad_norm": 2.4104364154469096,
+      "learning_rate": 7.1627906976744185e-06,
+      "loss": 0.7338,
+      "step": 155
+    },
+    {
+      "epoch": 0.2176238084166473,
+      "grad_norm": 2.331567384611702,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.755,
+      "step": 156
+    },
+    {
+      "epoch": 0.21901883282957452,
+      "grad_norm": 2.199020523596816,
+      "learning_rate": 7.255813953488373e-06,
+      "loss": 0.6787,
+      "step": 157
+    },
+    {
+      "epoch": 0.22041385724250173,
+      "grad_norm": 2.5951329934583787,
+      "learning_rate": 7.30232558139535e-06,
+      "loss": 0.7364,
+      "step": 158
+    },
+    {
+      "epoch": 0.22180888165542897,
+      "grad_norm": 2.2810633560589597,
+      "learning_rate": 7.348837209302326e-06,
+      "loss": 0.7253,
+      "step": 159
+    },
+    {
+      "epoch": 0.2232039060683562,
+      "grad_norm": 2.4924517088895652,
+      "learning_rate": 7.395348837209303e-06,
+      "loss": 0.748,
+      "step": 160
+    },
+    {
+      "epoch": 0.22459893048128343,
+      "grad_norm": 2.4519591046477682,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.741,
+      "step": 161
+    },
+    {
+      "epoch": 0.22599395489421065,
+      "grad_norm": 2.427349032223749,
+      "learning_rate": 7.488372093023256e-06,
+      "loss": 0.7369,
+      "step": 162
+    },
+    {
+      "epoch": 0.22738897930713786,
+      "grad_norm": 2.4574734880892235,
+      "learning_rate": 7.534883720930233e-06,
+      "loss": 0.7681,
+      "step": 163
+    },
+    {
+      "epoch": 0.2287840037200651,
+      "grad_norm": 2.3629173763703175,
+      "learning_rate": 7.5813953488372095e-06,
+      "loss": 0.6676,
+      "step": 164
+    },
+    {
+      "epoch": 0.23017902813299232,
+      "grad_norm": 2.499777704465563,
+      "learning_rate": 7.627906976744187e-06,
+      "loss": 0.7475,
+      "step": 165
+    },
+    {
+      "epoch": 0.23157405254591956,
+      "grad_norm": 2.124247898175583,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.6732,
+      "step": 166
+    },
+    {
+      "epoch": 0.23296907695884678,
+      "grad_norm": 2.4833385152209417,
+      "learning_rate": 7.72093023255814e-06,
+      "loss": 0.7008,
+      "step": 167
+    },
+    {
+      "epoch": 0.23436410137177402,
+      "grad_norm": 2.440899551714562,
+      "learning_rate": 7.767441860465116e-06,
+      "loss": 0.725,
+      "step": 168
+    },
+    {
+      "epoch": 0.23575912578470123,
+      "grad_norm": 2.3266389679720123,
+      "learning_rate": 7.813953488372094e-06,
+      "loss": 0.7323,
+      "step": 169
+    },
+    {
+      "epoch": 0.23715415019762845,
+      "grad_norm": 2.1958822444892916,
+      "learning_rate": 7.86046511627907e-06,
+      "loss": 0.6617,
+      "step": 170
+    },
+    {
+      "epoch": 0.2385491746105557,
+      "grad_norm": 2.1838589739129355,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.7279,
+      "step": 171
+    },
+    {
+      "epoch": 0.2399441990234829,
+      "grad_norm": 2.429658187245016,
+      "learning_rate": 7.953488372093024e-06,
+      "loss": 0.7089,
+      "step": 172
+    },
+    {
+      "epoch": 0.24133922343641015,
+      "grad_norm": 2.5141806643559423,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7744,
+      "step": 173
+    },
+    {
+      "epoch": 0.24273424784933736,
+      "grad_norm": 2.3532093516479073,
+      "learning_rate": 8.046511627906977e-06,
+      "loss": 0.738,
+      "step": 174
+    },
+    {
+      "epoch": 0.24412927226226458,
+      "grad_norm": 2.3567949350894892,
+      "learning_rate": 8.093023255813955e-06,
+      "loss": 0.7661,
+      "step": 175
+    },
+    {
+      "epoch": 0.24552429667519182,
+      "grad_norm": 2.422487493668288,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.7738,
+      "step": 176
+    },
+    {
+      "epoch": 0.24691932108811904,
+      "grad_norm": 2.533137019052819,
+      "learning_rate": 8.186046511627907e-06,
+      "loss": 0.7684,
+      "step": 177
+    },
+    {
+      "epoch": 0.24831434550104628,
+      "grad_norm": 2.5053972465039376,
+      "learning_rate": 8.232558139534885e-06,
+      "loss": 0.7735,
+      "step": 178
+    },
+    {
+      "epoch": 0.2497093699139735,
+      "grad_norm": 2.2993122862473534,
+      "learning_rate": 8.279069767441861e-06,
+      "loss": 0.7121,
+      "step": 179
+    },
+    {
+      "epoch": 0.25110439432690074,
+      "grad_norm": 2.4034617971465946,
+      "learning_rate": 8.325581395348837e-06,
+      "loss": 0.758,
+      "step": 180
+    },
+    {
+      "epoch": 0.25249941873982795,
+      "grad_norm": 2.5374194516013566,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.75,
+      "step": 181
+    },
+    {
+      "epoch": 0.25389444315275517,
+      "grad_norm": 2.321732284911669,
+      "learning_rate": 8.418604651162792e-06,
+      "loss": 0.7085,
+      "step": 182
+    },
+    {
+      "epoch": 0.2552894675656824,
+      "grad_norm": 2.4488456028756462,
+      "learning_rate": 8.465116279069768e-06,
+      "loss": 0.7467,
+      "step": 183
+    },
+    {
+      "epoch": 0.25668449197860965,
+      "grad_norm": 2.452894565948684,
+      "learning_rate": 8.511627906976744e-06,
+      "loss": 0.7142,
+      "step": 184
+    },
+    {
+      "epoch": 0.25807951639153687,
+      "grad_norm": 2.369770045849115,
+      "learning_rate": 8.558139534883722e-06,
+      "loss": 0.7071,
+      "step": 185
+    },
+    {
+      "epoch": 0.2594745408044641,
+      "grad_norm": 2.0567126918314877,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.6288,
+      "step": 186
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 2.350291216305672,
+      "learning_rate": 8.651162790697674e-06,
+      "loss": 0.7848,
+      "step": 187
+    },
+    {
+      "epoch": 0.2622645896303185,
+      "grad_norm": 2.314951381733544,
+      "learning_rate": 8.697674418604652e-06,
+      "loss": 0.7262,
+      "step": 188
+    },
+    {
+      "epoch": 0.2636596140432458,
+      "grad_norm": 2.304013798793699,
+      "learning_rate": 8.74418604651163e-06,
+      "loss": 0.6982,
+      "step": 189
+    },
+    {
+      "epoch": 0.265054638456173,
+      "grad_norm": 2.331062126039812,
+      "learning_rate": 8.790697674418606e-06,
+      "loss": 0.7191,
+      "step": 190
+    },
+    {
+      "epoch": 0.2664496628691002,
+      "grad_norm": 2.3981290812587024,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.7599,
+      "step": 191
+    },
+    {
+      "epoch": 0.2678446872820274,
+      "grad_norm": 2.3469258787402123,
+      "learning_rate": 8.88372093023256e-06,
+      "loss": 0.7005,
+      "step": 192
+    },
+    {
+      "epoch": 0.26923971169495464,
+      "grad_norm": 2.322650241728307,
+      "learning_rate": 8.930232558139535e-06,
+      "loss": 0.7553,
+      "step": 193
+    },
+    {
+      "epoch": 0.2706347361078819,
+      "grad_norm": 2.390850393607927,
+      "learning_rate": 8.976744186046511e-06,
+      "loss": 0.7011,
+      "step": 194
+    },
+    {
+      "epoch": 0.2720297605208091,
+      "grad_norm": 2.3611482677828004,
+      "learning_rate": 9.023255813953489e-06,
+      "loss": 0.715,
+      "step": 195
+    },
+    {
+      "epoch": 0.27342478493373634,
+      "grad_norm": 2.251847626785549,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.7798,
+      "step": 196
+    },
+    {
+      "epoch": 0.27481980934666356,
+      "grad_norm": 2.2203131238743374,
+      "learning_rate": 9.116279069767443e-06,
+      "loss": 0.7094,
+      "step": 197
+    },
+    {
+      "epoch": 0.27621483375959077,
+      "grad_norm": 2.3743505262739992,
+      "learning_rate": 9.162790697674419e-06,
+      "loss": 0.7082,
+      "step": 198
+    },
+    {
+      "epoch": 0.27760985817251804,
+      "grad_norm": 2.382614063698063,
+      "learning_rate": 9.209302325581397e-06,
+      "loss": 0.7115,
+      "step": 199
+    },
+    {
+      "epoch": 0.27900488258544526,
+      "grad_norm": 2.4087918128165775,
+      "learning_rate": 9.255813953488373e-06,
+      "loss": 0.7255,
+      "step": 200
+    },
+    {
+      "epoch": 0.2803999069983725,
+      "grad_norm": 2.2494923080443168,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.7212,
+      "step": 201
+    },
+    {
+      "epoch": 0.2817949314112997,
+      "grad_norm": 2.293386790821147,
+      "learning_rate": 9.348837209302326e-06,
+      "loss": 0.7484,
+      "step": 202
+    },
+    {
+      "epoch": 0.2831899558242269,
+      "grad_norm": 2.3471917637572477,
+      "learning_rate": 9.395348837209302e-06,
+      "loss": 0.7688,
+      "step": 203
+    },
+    {
+      "epoch": 0.2845849802371542,
+      "grad_norm": 2.400058271047396,
+      "learning_rate": 9.44186046511628e-06,
+      "loss": 0.7279,
+      "step": 204
+    },
+    {
+      "epoch": 0.2859800046500814,
+      "grad_norm": 2.4522386019787725,
+      "learning_rate": 9.488372093023258e-06,
+      "loss": 0.7716,
+      "step": 205
+    },
+    {
+      "epoch": 0.2873750290630086,
+      "grad_norm": 2.6239510123204477,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.743,
+      "step": 206
+    },
+    {
+      "epoch": 0.2887700534759358,
+      "grad_norm": 2.136688074275227,
+      "learning_rate": 9.58139534883721e-06,
+      "loss": 0.6951,
+      "step": 207
+    },
+    {
+      "epoch": 0.29016507788886303,
+      "grad_norm": 2.4603362182924076,
+      "learning_rate": 9.627906976744188e-06,
+      "loss": 0.7707,
+      "step": 208
+    },
+    {
+      "epoch": 0.2915601023017903,
+      "grad_norm": 2.3184884483119643,
+      "learning_rate": 9.674418604651164e-06,
+      "loss": 0.7577,
+      "step": 209
+    },
+    {
+      "epoch": 0.2929551267147175,
+      "grad_norm": 2.3897163298829716,
+      "learning_rate": 9.72093023255814e-06,
+      "loss": 0.767,
+      "step": 210
+    },
+    {
+      "epoch": 0.29435015112764473,
+      "grad_norm": 2.290611569859189,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.7195,
+      "step": 211
+    },
+    {
+      "epoch": 0.29574517554057195,
+      "grad_norm": 2.614865650081256,
+      "learning_rate": 9.813953488372093e-06,
+      "loss": 0.7677,
+      "step": 212
+    },
+    {
+      "epoch": 0.29714019995349916,
+      "grad_norm": 2.409329394194762,
+      "learning_rate": 9.860465116279071e-06,
+      "loss": 0.7396,
+      "step": 213
+    },
+    {
+      "epoch": 0.29853522436642643,
+      "grad_norm": 2.327197479223712,
+      "learning_rate": 9.906976744186047e-06,
+      "loss": 0.7102,
+      "step": 214
+    },
+    {
+      "epoch": 0.29993024877935365,
+      "grad_norm": 2.4339401095238045,
+      "learning_rate": 9.953488372093025e-06,
+      "loss": 0.7671,
+      "step": 215
+    },
+    {
+      "epoch": 0.30132527319228086,
+      "grad_norm": 2.406077578159272,
+      "learning_rate": 1e-05,
+      "loss": 0.7447,
+      "step": 216
+    },
+    {
+      "epoch": 0.3027202976052081,
+      "grad_norm": 2.4625625301213647,
+      "learning_rate": 9.999993396473114e-06,
+      "loss": 0.7896,
+      "step": 217
+    },
+    {
+      "epoch": 0.3041153220181353,
+      "grad_norm": 2.2307465016265513,
+      "learning_rate": 9.999973585909898e-06,
+      "loss": 0.6554,
+      "step": 218
+    },
+    {
+      "epoch": 0.30551034643106256,
+      "grad_norm": 2.3558678943771962,
+      "learning_rate": 9.99994056836268e-06,
+      "loss": 0.7142,
+      "step": 219
+    },
+    {
+      "epoch": 0.3069053708439898,
+      "grad_norm": 2.2935965111443273,
+      "learning_rate": 9.999894343918674e-06,
+      "loss": 0.7234,
+      "step": 220
+    },
+    {
+      "epoch": 0.308300395256917,
+      "grad_norm": 2.2050114399403333,
+      "learning_rate": 9.999834912699974e-06,
+      "loss": 0.7748,
+      "step": 221
+    },
+    {
+      "epoch": 0.3096954196698442,
+      "grad_norm": 2.2482433300501734,
+      "learning_rate": 9.999762274863567e-06,
+      "loss": 0.7253,
+      "step": 222
+    },
+    {
+      "epoch": 0.3110904440827714,
+      "grad_norm": 2.399635963689245,
+      "learning_rate": 9.999676430601318e-06,
+      "loss": 0.7806,
+      "step": 223
+    },
+    {
+      "epoch": 0.3124854684956987,
+      "grad_norm": 2.2011507392317187,
+      "learning_rate": 9.999577380139976e-06,
+      "loss": 0.7026,
+      "step": 224
+    },
+    {
+      "epoch": 0.3138804929086259,
+      "grad_norm": 2.3655833768745977,
+      "learning_rate": 9.999465123741172e-06,
+      "loss": 0.7078,
+      "step": 225
+    },
+    {
+      "epoch": 0.3152755173215531,
+      "grad_norm": 2.387043969086896,
+      "learning_rate": 9.999339661701424e-06,
+      "loss": 0.7031,
+      "step": 226
+    },
+    {
+      "epoch": 0.31667054173448034,
+      "grad_norm": 2.2360354228157666,
+      "learning_rate": 9.99920099435213e-06,
+      "loss": 0.7505,
+      "step": 227
+    },
+    {
+      "epoch": 0.31806556614740755,
+      "grad_norm": 2.4994174229622836,
+      "learning_rate": 9.999049122059565e-06,
+      "loss": 0.7708,
+      "step": 228
+    },
+    {
+      "epoch": 0.3194605905603348,
+      "grad_norm": 2.133696380023235,
+      "learning_rate": 9.998884045224886e-06,
+      "loss": 0.7241,
+      "step": 229
+    },
+    {
+      "epoch": 0.32085561497326204,
+      "grad_norm": 2.241964560461335,
+      "learning_rate": 9.998705764284132e-06,
+      "loss": 0.6949,
+      "step": 230
+    },
+    {
+      "epoch": 0.32225063938618925,
+      "grad_norm": 2.475034321702833,
+      "learning_rate": 9.998514279708212e-06,
+      "loss": 0.7826,
+      "step": 231
+    },
+    {
+      "epoch": 0.32364566379911647,
+      "grad_norm": 2.36402035615075,
+      "learning_rate": 9.998309592002914e-06,
+      "loss": 0.7745,
+      "step": 232
+    },
+    {
+      "epoch": 0.32504068821204374,
+      "grad_norm": 2.3658257454389475,
+      "learning_rate": 9.99809170170891e-06,
+      "loss": 0.6859,
+      "step": 233
+    },
+    {
+      "epoch": 0.32643571262497095,
+      "grad_norm": 2.273680448775777,
+      "learning_rate": 9.997860609401732e-06,
+      "loss": 0.7775,
+      "step": 234
+    },
+    {
+      "epoch": 0.32783073703789817,
+      "grad_norm": 2.201487231731571,
+      "learning_rate": 9.99761631569179e-06,
+      "loss": 0.7803,
+      "step": 235
+    },
+    {
+      "epoch": 0.3292257614508254,
+      "grad_norm": 2.4155650660144348,
+      "learning_rate": 9.997358821224365e-06,
+      "loss": 0.8011,
+      "step": 236
+    },
+    {
+      "epoch": 0.3306207858637526,
+      "grad_norm": 2.359732428185931,
+      "learning_rate": 9.997088126679607e-06,
+      "loss": 0.7299,
+      "step": 237
+    },
+    {
+      "epoch": 0.33201581027667987,
+      "grad_norm": 2.3093662561481367,
+      "learning_rate": 9.996804232772528e-06,
+      "loss": 0.7319,
+      "step": 238
+    },
+    {
+      "epoch": 0.3334108346896071,
+      "grad_norm": 2.272053877068559,
+      "learning_rate": 9.996507140253012e-06,
+      "loss": 0.7559,
+      "step": 239
+    },
+    {
+      "epoch": 0.3348058591025343,
+      "grad_norm": 2.1570286455643575,
+      "learning_rate": 9.9961968499058e-06,
+      "loss": 0.7401,
+      "step": 240
+    },
+    {
+      "epoch": 0.3362008835154615,
+      "grad_norm": 2.4238684062299205,
+      "learning_rate": 9.9958733625505e-06,
+      "loss": 0.7662,
+      "step": 241
+    },
+    {
+      "epoch": 0.3375959079283887,
+      "grad_norm": 2.146480418424315,
+      "learning_rate": 9.995536679041568e-06,
+      "loss": 0.7078,
+      "step": 242
+    },
+    {
+      "epoch": 0.338990932341316,
+      "grad_norm": 2.320463871074251,
+      "learning_rate": 9.99518680026833e-06,
+      "loss": 0.7589,
+      "step": 243
+    },
+    {
+      "epoch": 0.3403859567542432,
+      "grad_norm": 2.2760580968152144,
+      "learning_rate": 9.994823727154957e-06,
+      "loss": 0.7427,
+      "step": 244
+    },
+    {
+      "epoch": 0.34178098116717043,
+      "grad_norm": 2.5684664291414294,
+      "learning_rate": 9.994447460660473e-06,
+      "loss": 0.7679,
+      "step": 245
+    },
+    {
+      "epoch": 0.34317600558009764,
+      "grad_norm": 2.1635703031164977,
+      "learning_rate": 9.994058001778754e-06,
+      "loss": 0.7536,
+      "step": 246
+    },
+    {
+      "epoch": 0.34457102999302486,
+      "grad_norm": 2.282114366302871,
+      "learning_rate": 9.99365535153852e-06,
+      "loss": 0.7267,
+      "step": 247
+    },
+    {
+      "epoch": 0.34596605440595213,
+      "grad_norm": 2.231237877658349,
+      "learning_rate": 9.993239511003338e-06,
+      "loss": 0.7372,
+      "step": 248
+    },
+    {
+      "epoch": 0.34736107881887934,
+      "grad_norm": 2.481316957977788,
+      "learning_rate": 9.992810481271611e-06,
+      "loss": 0.7335,
+      "step": 249
+    },
+    {
+      "epoch": 0.34875610323180656,
+      "grad_norm": 2.302521453924701,
+      "learning_rate": 9.992368263476585e-06,
+      "loss": 0.7511,
+      "step": 250
+    },
+    {
+      "epoch": 0.3501511276447338,
+      "grad_norm": 2.2387510321179875,
+      "learning_rate": 9.991912858786335e-06,
+      "loss": 0.7613,
+      "step": 251
+    },
+    {
+      "epoch": 0.351546152057661,
+      "grad_norm": 2.361564774571516,
+      "learning_rate": 9.991444268403776e-06,
+      "loss": 0.7642,
+      "step": 252
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 2.343319789459909,
+      "learning_rate": 9.990962493566645e-06,
+      "loss": 0.8018,
+      "step": 253
+    },
+    {
+      "epoch": 0.3543362008835155,
+      "grad_norm": 2.444856016657922,
+      "learning_rate": 9.99046753554751e-06,
+      "loss": 0.7034,
+      "step": 254
+    },
+    {
+      "epoch": 0.3557312252964427,
+      "grad_norm": 2.295542656637622,
+      "learning_rate": 9.989959395653756e-06,
+      "loss": 0.7357,
+      "step": 255
+    },
+    {
+      "epoch": 0.3571262497093699,
+      "grad_norm": 2.1172219443470057,
+      "learning_rate": 9.989438075227588e-06,
+      "loss": 0.6318,
+      "step": 256
+    },
+    {
+      "epoch": 0.3585212741222971,
+      "grad_norm": 2.317934296315436,
+      "learning_rate": 9.988903575646032e-06,
+      "loss": 0.6826,
+      "step": 257
+    },
+    {
+      "epoch": 0.3599162985352244,
+      "grad_norm": 2.249565309628914,
+      "learning_rate": 9.988355898320917e-06,
+      "loss": 0.7302,
+      "step": 258
+    },
+    {
+      "epoch": 0.3613113229481516,
+      "grad_norm": 2.61405603733907,
+      "learning_rate": 9.987795044698885e-06,
+      "loss": 0.7599,
+      "step": 259
+    },
+    {
+      "epoch": 0.3627063473610788,
+      "grad_norm": 2.393309752919588,
+      "learning_rate": 9.98722101626138e-06,
+      "loss": 0.7434,
+      "step": 260
+    },
+    {
+      "epoch": 0.36410137177400603,
+      "grad_norm": 2.255993754480847,
+      "learning_rate": 9.986633814524648e-06,
+      "loss": 0.7009,
+      "step": 261
+    },
+    {
+      "epoch": 0.36549639618693325,
+      "grad_norm": 2.1899632672810796,
+      "learning_rate": 9.986033441039731e-06,
+      "loss": 0.7352,
+      "step": 262
+    },
+    {
+      "epoch": 0.3668914205998605,
+      "grad_norm": 2.4758174903524837,
+      "learning_rate": 9.985419897392459e-06,
+      "loss": 0.7922,
+      "step": 263
+    },
+    {
+      "epoch": 0.36828644501278773,
+      "grad_norm": 1.9484666270746862,
+      "learning_rate": 9.984793185203456e-06,
+      "loss": 0.692,
+      "step": 264
+    },
+    {
+      "epoch": 0.36968146942571495,
+      "grad_norm": 2.231671078245673,
+      "learning_rate": 9.984153306128124e-06,
+      "loss": 0.7602,
+      "step": 265
+    },
+    {
+      "epoch": 0.37107649383864216,
+      "grad_norm": 2.5406824511193933,
+      "learning_rate": 9.983500261856646e-06,
+      "loss": 0.7612,
+      "step": 266
+    },
+    {
+      "epoch": 0.3724715182515694,
+      "grad_norm": 2.215219371603928,
+      "learning_rate": 9.982834054113982e-06,
+      "loss": 0.6931,
+      "step": 267
+    },
+    {
+      "epoch": 0.37386654266449665,
+      "grad_norm": 2.4375255794403103,
+      "learning_rate": 9.98215468465986e-06,
+      "loss": 0.7192,
+      "step": 268
+    },
+    {
+      "epoch": 0.37526156707742386,
+      "grad_norm": 2.168943298360225,
+      "learning_rate": 9.981462155288773e-06,
+      "loss": 0.7385,
+      "step": 269
+    },
+    {
+      "epoch": 0.3766565914903511,
+      "grad_norm": 2.2784901454453554,
+      "learning_rate": 9.980756467829977e-06,
+      "loss": 0.6834,
+      "step": 270
+    },
+    {
+      "epoch": 0.3780516159032783,
+      "grad_norm": 2.335696005944563,
+      "learning_rate": 9.98003762414748e-06,
+      "loss": 0.7263,
+      "step": 271
+    },
+    {
+      "epoch": 0.3794466403162055,
+      "grad_norm": 2.181003569392429,
+      "learning_rate": 9.979305626140046e-06,
+      "loss": 0.6903,
+      "step": 272
+    },
+    {
+      "epoch": 0.3808416647291328,
+      "grad_norm": 2.1412825487141087,
+      "learning_rate": 9.978560475741181e-06,
+      "loss": 0.712,
+      "step": 273
+    },
+    {
+      "epoch": 0.38223668914206,
+      "grad_norm": 2.2673165084490687,
+      "learning_rate": 9.977802174919134e-06,
+      "loss": 0.7921,
+      "step": 274
+    },
+    {
+      "epoch": 0.3836317135549872,
+      "grad_norm": 2.142363448667009,
+      "learning_rate": 9.977030725676887e-06,
+      "loss": 0.6938,
+      "step": 275
+    },
+    {
+      "epoch": 0.3850267379679144,
+      "grad_norm": 2.2876816347837434,
+      "learning_rate": 9.976246130052157e-06,
+      "loss": 0.7565,
+      "step": 276
+    },
+    {
+      "epoch": 0.38642176238084164,
+      "grad_norm": 2.2736845275915187,
+      "learning_rate": 9.97544839011738e-06,
+      "loss": 0.8085,
+      "step": 277
+    },
+    {
+      "epoch": 0.3878167867937689,
+      "grad_norm": 2.1051620059141247,
+      "learning_rate": 9.974637507979721e-06,
+      "loss": 0.6934,
+      "step": 278
+    },
+    {
+      "epoch": 0.3892118112066961,
+      "grad_norm": 2.3005570106943587,
+      "learning_rate": 9.973813485781045e-06,
+      "loss": 0.7029,
+      "step": 279
+    },
+    {
+      "epoch": 0.39060683561962334,
+      "grad_norm": 2.060997419935413,
+      "learning_rate": 9.972976325697938e-06,
+      "loss": 0.6991,
+      "step": 280
+    },
+    {
+      "epoch": 0.39200186003255055,
+      "grad_norm": 2.049588415049291,
+      "learning_rate": 9.972126029941685e-06,
+      "loss": 0.7229,
+      "step": 281
+    },
+    {
+      "epoch": 0.3933968844454778,
+      "grad_norm": 2.2293739933043155,
+      "learning_rate": 9.97126260075826e-06,
+      "loss": 0.7045,
+      "step": 282
+    },
+    {
+      "epoch": 0.39479190885840504,
+      "grad_norm": 2.3316135043046327,
+      "learning_rate": 9.97038604042834e-06,
+      "loss": 0.7787,
+      "step": 283
+    },
+    {
+      "epoch": 0.39618693327133225,
+      "grad_norm": 2.2340910959336147,
+      "learning_rate": 9.969496351267278e-06,
+      "loss": 0.7174,
+      "step": 284
+    },
+    {
+      "epoch": 0.39758195768425947,
+      "grad_norm": 2.298045654117731,
+      "learning_rate": 9.96859353562511e-06,
+      "loss": 0.7462,
+      "step": 285
+    },
+    {
+      "epoch": 0.3989769820971867,
+      "grad_norm": 2.1837557728814123,
+      "learning_rate": 9.967677595886542e-06,
+      "loss": 0.7167,
+      "step": 286
+    },
+    {
+      "epoch": 0.40037200651011395,
+      "grad_norm": 2.27299808508551,
+      "learning_rate": 9.96674853447095e-06,
+      "loss": 0.7433,
+      "step": 287
+    },
+    {
+      "epoch": 0.40176703092304117,
+      "grad_norm": 2.137166870007726,
+      "learning_rate": 9.96580635383236e-06,
+      "loss": 0.6633,
+      "step": 288
+    },
+    {
+      "epoch": 0.4031620553359684,
+      "grad_norm": 1.9262822690731025,
+      "learning_rate": 9.964851056459465e-06,
+      "loss": 0.6312,
+      "step": 289
+    },
+    {
+      "epoch": 0.4045570797488956,
+      "grad_norm": 2.269824827715954,
+      "learning_rate": 9.963882644875594e-06,
+      "loss": 0.7178,
+      "step": 290
+    },
+    {
+      "epoch": 0.4059521041618228,
+      "grad_norm": 2.2048707014322,
+      "learning_rate": 9.96290112163872e-06,
+      "loss": 0.7068,
+      "step": 291
+    },
+    {
+      "epoch": 0.4073471285747501,
+      "grad_norm": 2.322124072727594,
+      "learning_rate": 9.961906489341452e-06,
+      "loss": 0.7339,
+      "step": 292
+    },
+    {
+      "epoch": 0.4087421529876773,
+      "grad_norm": 2.4571329521811367,
+      "learning_rate": 9.960898750611019e-06,
+      "loss": 0.7715,
+      "step": 293
+    },
+    {
+      "epoch": 0.4101371774006045,
+      "grad_norm": 2.5174823882242015,
+      "learning_rate": 9.959877908109274e-06,
+      "loss": 0.8336,
+      "step": 294
+    },
+    {
+      "epoch": 0.41153220181353173,
+      "grad_norm": 2.208634018316973,
+      "learning_rate": 9.958843964532683e-06,
+      "loss": 0.6829,
+      "step": 295
+    },
+    {
+      "epoch": 0.41292722622645894,
+      "grad_norm": 2.222290757312979,
+      "learning_rate": 9.957796922612314e-06,
+      "loss": 0.7138,
+      "step": 296
+    },
+    {
+      "epoch": 0.4143222506393862,
+      "grad_norm": 2.5274252464650755,
+      "learning_rate": 9.956736785113833e-06,
+      "loss": 0.7725,
+      "step": 297
+    },
+    {
+      "epoch": 0.41571727505231343,
+      "grad_norm": 2.0860385651911346,
+      "learning_rate": 9.955663554837503e-06,
+      "loss": 0.6504,
+      "step": 298
+    },
+    {
+      "epoch": 0.41711229946524064,
+      "grad_norm": 2.3408784898712787,
+      "learning_rate": 9.954577234618162e-06,
+      "loss": 0.7158,
+      "step": 299
+    },
+    {
+      "epoch": 0.41850732387816786,
+      "grad_norm": 2.2270712846507514,
+      "learning_rate": 9.953477827325229e-06,
+      "loss": 0.7358,
+      "step": 300
+    },
+    {
+      "epoch": 0.4199023482910951,
+      "grad_norm": 2.1586659413799127,
+      "learning_rate": 9.952365335862693e-06,
+      "loss": 0.7668,
+      "step": 301
+    },
+    {
+      "epoch": 0.42129737270402234,
+      "grad_norm": 2.234673416696326,
+      "learning_rate": 9.951239763169097e-06,
+      "loss": 0.7895,
+      "step": 302
+    },
+    {
+      "epoch": 0.42269239711694956,
+      "grad_norm": 2.195858274108098,
+      "learning_rate": 9.950101112217543e-06,
+      "loss": 0.7191,
+      "step": 303
+    },
+    {
+      "epoch": 0.4240874215298768,
+      "grad_norm": 2.1565944180125376,
+      "learning_rate": 9.948949386015677e-06,
+      "loss": 0.6953,
+      "step": 304
+    },
+    {
+      "epoch": 0.425482445942804,
+      "grad_norm": 2.2088440863358905,
+      "learning_rate": 9.947784587605678e-06,
+      "loss": 0.6961,
+      "step": 305
+    },
+    {
+      "epoch": 0.4268774703557312,
+      "grad_norm": 2.411909196100123,
+      "learning_rate": 9.946606720064257e-06,
+      "loss": 0.7571,
+      "step": 306
+    },
+    {
+      "epoch": 0.4282724947686585,
+      "grad_norm": 2.398065188628216,
+      "learning_rate": 9.945415786502649e-06,
+      "loss": 0.7381,
+      "step": 307
+    },
+    {
+      "epoch": 0.4296675191815857,
+      "grad_norm": 2.166520322060508,
+      "learning_rate": 9.944211790066597e-06,
+      "loss": 0.7238,
+      "step": 308
+    },
+    {
+      "epoch": 0.4310625435945129,
+      "grad_norm": 2.3660967755228746,
+      "learning_rate": 9.94299473393635e-06,
+      "loss": 0.7767,
+      "step": 309
+    },
+    {
+      "epoch": 0.4324575680074401,
+      "grad_norm": 2.239729559357496,
+      "learning_rate": 9.941764621326655e-06,
+      "loss": 0.7166,
+      "step": 310
+    },
+    {
+      "epoch": 0.43385259242036733,
+      "grad_norm": 2.118960181107485,
+      "learning_rate": 9.94052145548674e-06,
+      "loss": 0.754,
+      "step": 311
+    },
+    {
+      "epoch": 0.4352476168332946,
+      "grad_norm": 2.374426767740763,
+      "learning_rate": 9.939265239700321e-06,
+      "loss": 0.7863,
+      "step": 312
+    },
+    {
+      "epoch": 0.4366426412462218,
+      "grad_norm": 2.369429638423822,
+      "learning_rate": 9.93799597728558e-06,
+      "loss": 0.7489,
+      "step": 313
+    },
+    {
+      "epoch": 0.43803766565914903,
+      "grad_norm": 2.387324721804876,
+      "learning_rate": 9.936713671595158e-06,
+      "loss": 0.7778,
+      "step": 314
+    },
+    {
+      "epoch": 0.43943269007207625,
+      "grad_norm": 2.2609612055446093,
+      "learning_rate": 9.935418326016153e-06,
+      "loss": 0.7073,
+      "step": 315
+    },
+    {
+      "epoch": 0.44082771448500346,
+      "grad_norm": 2.194501412971286,
+      "learning_rate": 9.934109943970103e-06,
+      "loss": 0.7858,
+      "step": 316
+    },
+    {
+      "epoch": 0.44222273889793073,
+      "grad_norm": 2.2443511833058687,
+      "learning_rate": 9.932788528912983e-06,
+      "loss": 0.7556,
+      "step": 317
+    },
+    {
+      "epoch": 0.44361776331085795,
+      "grad_norm": 2.262002280386999,
+      "learning_rate": 9.931454084335192e-06,
+      "loss": 0.7557,
+      "step": 318
+    },
+    {
+      "epoch": 0.44501278772378516,
+      "grad_norm": 2.0267896100249034,
+      "learning_rate": 9.930106613761549e-06,
+      "loss": 0.6984,
+      "step": 319
+    },
+    {
+      "epoch": 0.4464078121367124,
+      "grad_norm": 2.192015174282251,
+      "learning_rate": 9.928746120751275e-06,
+      "loss": 0.7323,
+      "step": 320
+    },
+    {
+      "epoch": 0.4478028365496396,
+      "grad_norm": 2.2328778916714014,
+      "learning_rate": 9.927372608897992e-06,
+      "loss": 0.7685,
+      "step": 321
+    },
+    {
+      "epoch": 0.44919786096256686,
+      "grad_norm": 2.1611404588310323,
+      "learning_rate": 9.925986081829708e-06,
+      "loss": 0.6904,
+      "step": 322
+    },
+    {
+      "epoch": 0.4505928853754941,
+      "grad_norm": 2.3906038854954232,
+      "learning_rate": 9.924586543208812e-06,
+      "loss": 0.7651,
+      "step": 323
+    },
+    {
+      "epoch": 0.4519879097884213,
+      "grad_norm": 2.2150342718775384,
+      "learning_rate": 9.923173996732058e-06,
+      "loss": 0.7088,
+      "step": 324
+    },
+    {
+      "epoch": 0.4533829342013485,
+      "grad_norm": 2.257194983707292,
+      "learning_rate": 9.921748446130564e-06,
+      "loss": 0.7583,
+      "step": 325
+    },
+    {
+      "epoch": 0.4547779586142757,
+      "grad_norm": 2.049635371526883,
+      "learning_rate": 9.920309895169793e-06,
+      "loss": 0.7118,
+      "step": 326
+    },
+    {
+      "epoch": 0.456172983027203,
+      "grad_norm": 2.2226016906419734,
+      "learning_rate": 9.91885834764955e-06,
+      "loss": 0.7132,
+      "step": 327
+    },
+    {
+      "epoch": 0.4575680074401302,
+      "grad_norm": 2.2068467340854614,
+      "learning_rate": 9.917393807403965e-06,
+      "loss": 0.7676,
+      "step": 328
+    },
+    {
+      "epoch": 0.4589630318530574,
+      "grad_norm": 2.0506777316664775,
+      "learning_rate": 9.915916278301496e-06,
+      "loss": 0.7045,
+      "step": 329
+    },
+    {
+      "epoch": 0.46035805626598464,
+      "grad_norm": 2.24941878997737,
+      "learning_rate": 9.9144257642449e-06,
+      "loss": 0.6901,
+      "step": 330
+    },
+    {
+      "epoch": 0.46175308067891185,
+      "grad_norm": 1.9825175559087265,
+      "learning_rate": 9.91292226917124e-06,
+      "loss": 0.6572,
+      "step": 331
+    },
+    {
+      "epoch": 0.4631481050918391,
+      "grad_norm": 2.29410576307341,
+      "learning_rate": 9.91140579705186e-06,
+      "loss": 0.7534,
+      "step": 332
+    },
+    {
+      "epoch": 0.46454312950476634,
+      "grad_norm": 2.1703274382213507,
+      "learning_rate": 9.909876351892388e-06,
+      "loss": 0.769,
+      "step": 333
+    },
+    {
+      "epoch": 0.46593815391769355,
+      "grad_norm": 2.028231097517855,
+      "learning_rate": 9.908333937732718e-06,
+      "loss": 0.7213,
+      "step": 334
+    },
+    {
+      "epoch": 0.46733317833062077,
+      "grad_norm": 2.1183249761708995,
+      "learning_rate": 9.906778558647e-06,
+      "loss": 0.7089,
+      "step": 335
+    },
+    {
+      "epoch": 0.46872820274354804,
+      "grad_norm": 2.0941215519004683,
+      "learning_rate": 9.905210218743626e-06,
+      "loss": 0.7282,
+      "step": 336
+    },
+    {
+      "epoch": 0.47012322715647525,
+      "grad_norm": 2.093942221954034,
+      "learning_rate": 9.903628922165227e-06,
+      "loss": 0.7011,
+      "step": 337
+    },
+    {
+      "epoch": 0.47151825156940247,
+      "grad_norm": 2.3302808968769986,
+      "learning_rate": 9.902034673088656e-06,
+      "loss": 0.7477,
+      "step": 338
+    },
+    {
+      "epoch": 0.4729132759823297,
+      "grad_norm": 2.152602661906724,
+      "learning_rate": 9.90042747572498e-06,
+      "loss": 0.6833,
+      "step": 339
+    },
+    {
+      "epoch": 0.4743083003952569,
+      "grad_norm": 2.349715876121456,
+      "learning_rate": 9.898807334319471e-06,
+      "loss": 0.7693,
+      "step": 340
+    },
+    {
+      "epoch": 0.47570332480818417,
+      "grad_norm": 2.173833769245502,
+      "learning_rate": 9.897174253151583e-06,
+      "loss": 0.7206,
+      "step": 341
+    },
+    {
+      "epoch": 0.4770983492211114,
+      "grad_norm": 2.259839449442954,
+      "learning_rate": 9.895528236534957e-06,
+      "loss": 0.7403,
+      "step": 342
+    },
+    {
+      "epoch": 0.4784933736340386,
+      "grad_norm": 2.3578368736204207,
+      "learning_rate": 9.893869288817397e-06,
+      "loss": 0.7909,
+      "step": 343
+    },
+    {
+      "epoch": 0.4798883980469658,
+      "grad_norm": 2.009722455372278,
+      "learning_rate": 9.89219741438087e-06,
+      "loss": 0.7118,
+      "step": 344
+    },
+    {
+      "epoch": 0.48128342245989303,
+      "grad_norm": 2.2025524361064814,
+      "learning_rate": 9.890512617641474e-06,
+      "loss": 0.7331,
+      "step": 345
+    },
+    {
+      "epoch": 0.4826784468728203,
+      "grad_norm": 2.285563908098744,
+      "learning_rate": 9.888814903049458e-06,
+      "loss": 0.756,
+      "step": 346
+    },
+    {
+      "epoch": 0.4840734712857475,
+      "grad_norm": 2.2142506724911457,
+      "learning_rate": 9.88710427508918e-06,
+      "loss": 0.7708,
+      "step": 347
+    },
+    {
+      "epoch": 0.48546849569867473,
+      "grad_norm": 2.2408289991941195,
+      "learning_rate": 9.885380738279111e-06,
+      "loss": 0.6929,
+      "step": 348
+    },
+    {
+      "epoch": 0.48686352011160194,
+      "grad_norm": 2.1080662535889476,
+      "learning_rate": 9.883644297171821e-06,
+      "loss": 0.7265,
+      "step": 349
+    },
+    {
+      "epoch": 0.48825854452452916,
+      "grad_norm": 2.090284461635897,
+      "learning_rate": 9.881894956353963e-06,
+      "loss": 0.7531,
+      "step": 350
+    },
+    {
+      "epoch": 0.48965356893745643,
+      "grad_norm": 2.1569142034243027,
+      "learning_rate": 9.880132720446265e-06,
+      "loss": 0.7172,
+      "step": 351
+    },
+    {
+      "epoch": 0.49104859335038364,
+      "grad_norm": 2.217562182182237,
+      "learning_rate": 9.878357594103516e-06,
+      "loss": 0.7737,
+      "step": 352
+    },
+    {
+      "epoch": 0.49244361776331086,
+      "grad_norm": 2.136573317949441,
+      "learning_rate": 9.876569582014554e-06,
+      "loss": 0.7205,
+      "step": 353
+    },
+    {
+      "epoch": 0.4938386421762381,
+      "grad_norm": 2.170725336806224,
+      "learning_rate": 9.874768688902252e-06,
+      "loss": 0.7557,
+      "step": 354
+    },
+    {
+      "epoch": 0.4952336665891653,
+      "grad_norm": 1.9745473881575657,
+      "learning_rate": 9.87295491952351e-06,
+      "loss": 0.7539,
+      "step": 355
+    },
+    {
+      "epoch": 0.49662869100209256,
+      "grad_norm": 2.1969404437698485,
+      "learning_rate": 9.871128278669238e-06,
+      "loss": 0.7255,
+      "step": 356
+    },
+    {
+      "epoch": 0.4980237154150198,
+      "grad_norm": 2.1672166906798886,
+      "learning_rate": 9.869288771164344e-06,
+      "loss": 0.7185,
+      "step": 357
+    },
+    {
+      "epoch": 0.499418739827947,
+      "grad_norm": 2.260245290048278,
+      "learning_rate": 9.867436401867723e-06,
+      "loss": 0.7316,
+      "step": 358
+    },
+    {
+      "epoch": 0.5008137642408742,
+      "grad_norm": 2.3386260178155363,
+      "learning_rate": 9.865571175672245e-06,
+      "loss": 0.7634,
+      "step": 359
+    },
+    {
+      "epoch": 0.5022087886538015,
+      "grad_norm": 2.1433025976011373,
+      "learning_rate": 9.863693097504733e-06,
+      "loss": 0.7115,
+      "step": 360
+    },
+    {
+      "epoch": 0.5036038130667286,
+      "grad_norm": 2.130459491307474,
+      "learning_rate": 9.86180217232597e-06,
+      "loss": 0.7737,
+      "step": 361
+    },
+    {
+      "epoch": 0.5049988374796559,
+      "grad_norm": 2.419050275087982,
+      "learning_rate": 9.859898405130661e-06,
+      "loss": 0.792,
+      "step": 362
+    },
+    {
+      "epoch": 0.5063938618925832,
+      "grad_norm": 2.301514017797739,
+      "learning_rate": 9.85798180094744e-06,
+      "loss": 0.7134,
+      "step": 363
+    },
+    {
+      "epoch": 0.5077888863055103,
+      "grad_norm": 2.2579550766262795,
+      "learning_rate": 9.856052364838846e-06,
+      "loss": 0.8086,
+      "step": 364
+    },
+    {
+      "epoch": 0.5091839107184376,
+      "grad_norm": 2.1715387851235093,
+      "learning_rate": 9.854110101901308e-06,
+      "loss": 0.6743,
+      "step": 365
+    },
+    {
+      "epoch": 0.5105789351313648,
+      "grad_norm": 1.9745286906089936,
+      "learning_rate": 9.852155017265146e-06,
+      "loss": 0.6278,
+      "step": 366
+    },
+    {
+      "epoch": 0.511973959544292,
+      "grad_norm": 2.211330485942951,
+      "learning_rate": 9.850187116094538e-06,
+      "loss": 0.7201,
+      "step": 367
+    },
+    {
+      "epoch": 0.5133689839572193,
+      "grad_norm": 2.089703071837029,
+      "learning_rate": 9.848206403587521e-06,
+      "loss": 0.762,
+      "step": 368
+    },
+    {
+      "epoch": 0.5147640083701465,
+      "grad_norm": 2.291478063761164,
+      "learning_rate": 9.84621288497597e-06,
+      "loss": 0.7553,
+      "step": 369
+    },
+    {
+      "epoch": 0.5161590327830737,
+      "grad_norm": 2.1165384247792787,
+      "learning_rate": 9.844206565525585e-06,
+      "loss": 0.7479,
+      "step": 370
+    },
+    {
+      "epoch": 0.5175540571960009,
+      "grad_norm": 2.153127909928685,
+      "learning_rate": 9.842187450535881e-06,
+      "loss": 0.7315,
+      "step": 371
+    },
+    {
+      "epoch": 0.5189490816089282,
+      "grad_norm": 2.4142012063691767,
+      "learning_rate": 9.840155545340169e-06,
+      "loss": 0.7433,
+      "step": 372
+    },
+    {
+      "epoch": 0.5203441060218554,
+      "grad_norm": 2.248477383285279,
+      "learning_rate": 9.838110855305548e-06,
+      "loss": 0.71,
+      "step": 373
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 2.018915478548314,
+      "learning_rate": 9.836053385832881e-06,
+      "loss": 0.6496,
+      "step": 374
+    },
+    {
+      "epoch": 0.5231341548477099,
+      "grad_norm": 2.1418708589656763,
+      "learning_rate": 9.833983142356792e-06,
+      "loss": 0.7287,
+      "step": 375
+    },
+    {
+      "epoch": 0.524529179260637,
+      "grad_norm": 2.2765930572342206,
+      "learning_rate": 9.831900130345645e-06,
+      "loss": 0.702,
+      "step": 376
+    },
+    {
+      "epoch": 0.5259242036735643,
+      "grad_norm": 2.059446170325579,
+      "learning_rate": 9.829804355301527e-06,
+      "loss": 0.6754,
+      "step": 377
+    },
+    {
+      "epoch": 0.5273192280864916,
+      "grad_norm": 2.054278197112765,
+      "learning_rate": 9.827695822760245e-06,
+      "loss": 0.6766,
+      "step": 378
+    },
+    {
+      "epoch": 0.5287142524994187,
+      "grad_norm": 2.2795716839624522,
+      "learning_rate": 9.825574538291293e-06,
+      "loss": 0.7354,
+      "step": 379
+    },
+    {
+      "epoch": 0.530109276912346,
+      "grad_norm": 2.0454813982579445,
+      "learning_rate": 9.823440507497863e-06,
+      "loss": 0.6778,
+      "step": 380
+    },
+    {
+      "epoch": 0.5315043013252732,
+      "grad_norm": 2.2322626658913163,
+      "learning_rate": 9.821293736016802e-06,
+      "loss": 0.6926,
+      "step": 381
+    },
+    {
+      "epoch": 0.5328993257382004,
+      "grad_norm": 2.256035142767682,
+      "learning_rate": 9.819134229518617e-06,
+      "loss": 0.7456,
+      "step": 382
+    },
+    {
+      "epoch": 0.5342943501511277,
+      "grad_norm": 2.1446252720693253,
+      "learning_rate": 9.81696199370745e-06,
+      "loss": 0.7224,
+      "step": 383
+    },
+    {
+      "epoch": 0.5356893745640549,
+      "grad_norm": 2.122015405672622,
+      "learning_rate": 9.814777034321069e-06,
+      "loss": 0.7462,
+      "step": 384
+    },
+    {
+      "epoch": 0.5370843989769821,
+      "grad_norm": 2.215882487448917,
+      "learning_rate": 9.812579357130848e-06,
+      "loss": 0.7693,
+      "step": 385
+    },
+    {
+      "epoch": 0.5384794233899093,
+      "grad_norm": 2.0156378005077924,
+      "learning_rate": 9.810368967941757e-06,
+      "loss": 0.714,
+      "step": 386
+    },
+    {
+      "epoch": 0.5398744478028366,
+      "grad_norm": 2.175601216674339,
+      "learning_rate": 9.808145872592341e-06,
+      "loss": 0.7194,
+      "step": 387
+    },
+    {
+      "epoch": 0.5412694722157638,
+      "grad_norm": 2.1421467209857425,
+      "learning_rate": 9.80591007695471e-06,
+      "loss": 0.6899,
+      "step": 388
+    },
+    {
+      "epoch": 0.542664496628691,
+      "grad_norm": 2.168215447780632,
+      "learning_rate": 9.803661586934514e-06,
+      "loss": 0.7049,
+      "step": 389
+    },
+    {
+      "epoch": 0.5440595210416183,
+      "grad_norm": 2.2372275971583826,
+      "learning_rate": 9.801400408470943e-06,
+      "loss": 0.7563,
+      "step": 390
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 2.1698822817889765,
+      "learning_rate": 9.799126547536695e-06,
+      "loss": 0.7542,
+      "step": 391
+    },
+    {
+      "epoch": 0.5468495698674727,
+      "grad_norm": 2.1058241804935633,
+      "learning_rate": 9.796840010137972e-06,
+      "loss": 0.7029,
+      "step": 392
+    },
+    {
+      "epoch": 0.5482445942804,
+      "grad_norm": 2.107054044072128,
+      "learning_rate": 9.79454080231446e-06,
+      "loss": 0.8368,
+      "step": 393
+    },
+    {
+      "epoch": 0.5496396186933271,
+      "grad_norm": 2.0206549367281177,
+      "learning_rate": 9.79222893013931e-06,
+      "loss": 0.646,
+      "step": 394
+    },
+    {
+      "epoch": 0.5510346431062544,
+      "grad_norm": 2.2344446550776302,
+      "learning_rate": 9.789904399719124e-06,
+      "loss": 0.7744,
+      "step": 395
+    },
+    {
+      "epoch": 0.5524296675191815,
+      "grad_norm": 2.10009826145433,
+      "learning_rate": 9.787567217193944e-06,
+      "loss": 0.6684,
+      "step": 396
+    },
+    {
+      "epoch": 0.5538246919321088,
+      "grad_norm": 2.143962574748327,
+      "learning_rate": 9.785217388737232e-06,
+      "loss": 0.744,
+      "step": 397
+    },
+    {
+      "epoch": 0.5552197163450361,
+      "grad_norm": 2.061035666351773,
+      "learning_rate": 9.782854920555844e-06,
+      "loss": 0.6964,
+      "step": 398
+    },
+    {
+      "epoch": 0.5566147407579632,
+      "grad_norm": 1.9503185836027328,
+      "learning_rate": 9.780479818890032e-06,
+      "loss": 0.7649,
+      "step": 399
+    },
+    {
+      "epoch": 0.5580097651708905,
+      "grad_norm": 2.1497215485134293,
+      "learning_rate": 9.778092090013416e-06,
+      "loss": 0.7134,
+      "step": 400
+    },
+    {
+      "epoch": 0.5594047895838177,
+      "grad_norm": 1.9121136304128945,
+      "learning_rate": 9.775691740232966e-06,
+      "loss": 0.6743,
+      "step": 401
+    },
+    {
+      "epoch": 0.560799813996745,
+      "grad_norm": 2.0333296587405756,
+      "learning_rate": 9.773278775888995e-06,
+      "loss": 0.7461,
+      "step": 402
+    },
+    {
+      "epoch": 0.5621948384096722,
+      "grad_norm": 2.1416975619415672,
+      "learning_rate": 9.77085320335513e-06,
+      "loss": 0.706,
+      "step": 403
+    },
+    {
+      "epoch": 0.5635898628225994,
+      "grad_norm": 2.2072504287834893,
+      "learning_rate": 9.768415029038304e-06,
+      "loss": 0.7114,
+      "step": 404
+    },
+    {
+      "epoch": 0.5649848872355266,
+      "grad_norm": 2.0068520321883985,
+      "learning_rate": 9.76596425937874e-06,
+      "loss": 0.6889,
+      "step": 405
+    },
+    {
+      "epoch": 0.5663799116484538,
+      "grad_norm": 2.043527386927699,
+      "learning_rate": 9.763500900849926e-06,
+      "loss": 0.6655,
+      "step": 406
+    },
+    {
+      "epoch": 0.5677749360613811,
+      "grad_norm": 2.3563340564084028,
+      "learning_rate": 9.761024959958605e-06,
+      "loss": 0.7508,
+      "step": 407
+    },
+    {
+      "epoch": 0.5691699604743083,
+      "grad_norm": 2.318133765773976,
+      "learning_rate": 9.75853644324475e-06,
+      "loss": 0.7575,
+      "step": 408
+    },
+    {
+      "epoch": 0.5705649848872355,
+      "grad_norm": 2.336246379174633,
+      "learning_rate": 9.756035357281559e-06,
+      "loss": 0.7395,
+      "step": 409
+    },
+    {
+      "epoch": 0.5719600093001628,
+      "grad_norm": 2.0727771110869555,
+      "learning_rate": 9.753521708675426e-06,
+      "loss": 0.7023,
+      "step": 410
+    },
+    {
+      "epoch": 0.5733550337130899,
+      "grad_norm": 2.1368053878525615,
+      "learning_rate": 9.75099550406593e-06,
+      "loss": 0.7111,
+      "step": 411
+    },
+    {
+      "epoch": 0.5747500581260172,
+      "grad_norm": 2.0396312273143304,
+      "learning_rate": 9.748456750125817e-06,
+      "loss": 0.7549,
+      "step": 412
+    },
+    {
+      "epoch": 0.5761450825389445,
+      "grad_norm": 2.15230083012261,
+      "learning_rate": 9.745905453560976e-06,
+      "loss": 0.7068,
+      "step": 413
+    },
+    {
+      "epoch": 0.5775401069518716,
+      "grad_norm": 1.9402253989855511,
+      "learning_rate": 9.74334162111043e-06,
+      "loss": 0.7451,
+      "step": 414
+    },
+    {
+      "epoch": 0.5789351313647989,
+      "grad_norm": 2.0708294796332423,
+      "learning_rate": 9.740765259546312e-06,
+      "loss": 0.7244,
+      "step": 415
+    },
+    {
+      "epoch": 0.5803301557777261,
+      "grad_norm": 2.247404010227105,
+      "learning_rate": 9.738176375673856e-06,
+      "loss": 0.7391,
+      "step": 416
+    },
+    {
+      "epoch": 0.5817251801906533,
+      "grad_norm": 2.158120031551307,
+      "learning_rate": 9.735574976331362e-06,
+      "loss": 0.6581,
+      "step": 417
+    },
+    {
+      "epoch": 0.5831202046035806,
+      "grad_norm": 2.1433546708079123,
+      "learning_rate": 9.732961068390199e-06,
+      "loss": 0.7543,
+      "step": 418
+    },
+    {
+      "epoch": 0.5845152290165078,
+      "grad_norm": 2.340443501167355,
+      "learning_rate": 9.730334658754767e-06,
+      "loss": 0.8277,
+      "step": 419
+    },
+    {
+      "epoch": 0.585910253429435,
+      "grad_norm": 2.071223812481573,
+      "learning_rate": 9.727695754362498e-06,
+      "loss": 0.7457,
+      "step": 420
+    },
+    {
+      "epoch": 0.5873052778423622,
+      "grad_norm": 2.072387969114693,
+      "learning_rate": 9.725044362183817e-06,
+      "loss": 0.7159,
+      "step": 421
+    },
+    {
+      "epoch": 0.5887003022552895,
+      "grad_norm": 2.0362411373585934,
+      "learning_rate": 9.722380489222145e-06,
+      "loss": 0.6802,
+      "step": 422
+    },
+    {
+      "epoch": 0.5900953266682167,
+      "grad_norm": 1.9495420698612345,
+      "learning_rate": 9.71970414251386e-06,
+      "loss": 0.6581,
+      "step": 423
+    },
+    {
+      "epoch": 0.5914903510811439,
+      "grad_norm": 2.12038681669184,
+      "learning_rate": 9.717015329128294e-06,
+      "loss": 0.7466,
+      "step": 424
+    },
+    {
+      "epoch": 0.5928853754940712,
+      "grad_norm": 2.130743892578285,
+      "learning_rate": 9.714314056167711e-06,
+      "loss": 0.7127,
+      "step": 425
+    },
+    {
+      "epoch": 0.5942803999069983,
+      "grad_norm": 1.8415534741189161,
+      "learning_rate": 9.711600330767278e-06,
+      "loss": 0.6402,
+      "step": 426
+    },
+    {
+      "epoch": 0.5956754243199256,
+      "grad_norm": 2.1495516487854354,
+      "learning_rate": 9.708874160095061e-06,
+      "loss": 0.7193,
+      "step": 427
+    },
+    {
+      "epoch": 0.5970704487328529,
+      "grad_norm": 2.0662475547351504,
+      "learning_rate": 9.706135551351996e-06,
+      "loss": 0.7214,
+      "step": 428
+    },
+    {
+      "epoch": 0.59846547314578,
+      "grad_norm": 2.0147057581611603,
+      "learning_rate": 9.703384511771874e-06,
+      "loss": 0.6683,
+      "step": 429
+    },
+    {
+      "epoch": 0.5998604975587073,
+      "grad_norm": 2.0161707870474923,
+      "learning_rate": 9.700621048621322e-06,
+      "loss": 0.6625,
+      "step": 430
+    },
+    {
+      "epoch": 0.6012555219716345,
+      "grad_norm": 1.9250124900125618,
+      "learning_rate": 9.697845169199775e-06,
+      "loss": 0.6494,
+      "step": 431
+    },
+    {
+      "epoch": 0.6026505463845617,
+      "grad_norm": 2.2505039960336695,
+      "learning_rate": 9.69505688083948e-06,
+      "loss": 0.7627,
+      "step": 432
+    },
+    {
+      "epoch": 0.604045570797489,
+      "grad_norm": 1.9421507056472695,
+      "learning_rate": 9.692256190905444e-06,
+      "loss": 0.6568,
+      "step": 433
+    },
+    {
+      "epoch": 0.6054405952104162,
+      "grad_norm": 2.1917513336388263,
+      "learning_rate": 9.689443106795442e-06,
+      "loss": 0.7268,
+      "step": 434
+    },
+    {
+      "epoch": 0.6068356196233434,
+      "grad_norm": 2.177671026580378,
+      "learning_rate": 9.686617635939988e-06,
+      "loss": 0.6854,
+      "step": 435
+    },
+    {
+      "epoch": 0.6082306440362706,
+      "grad_norm": 2.1823436209408302,
+      "learning_rate": 9.683779785802306e-06,
+      "loss": 0.7476,
+      "step": 436
+    },
+    {
+      "epoch": 0.6096256684491979,
+      "grad_norm": 2.2350100854518016,
+      "learning_rate": 9.680929563878327e-06,
+      "loss": 0.7023,
+      "step": 437
+    },
+    {
+      "epoch": 0.6110206928621251,
+      "grad_norm": 2.210520947527818,
+      "learning_rate": 9.678066977696656e-06,
+      "loss": 0.7577,
+      "step": 438
+    },
+    {
+      "epoch": 0.6124157172750523,
+      "grad_norm": 2.096018066597859,
+      "learning_rate": 9.675192034818561e-06,
+      "loss": 0.7356,
+      "step": 439
+    },
+    {
+      "epoch": 0.6138107416879796,
+      "grad_norm": 2.1973984969358873,
+      "learning_rate": 9.672304742837945e-06,
+      "loss": 0.8018,
+      "step": 440
+    },
+    {
+      "epoch": 0.6152057661009067,
+      "grad_norm": 2.1078413941321847,
+      "learning_rate": 9.669405109381335e-06,
+      "loss": 0.7291,
+      "step": 441
+    },
+    {
+      "epoch": 0.616600790513834,
+      "grad_norm": 2.1083633709141676,
+      "learning_rate": 9.66649314210785e-06,
+      "loss": 0.7295,
+      "step": 442
+    },
+    {
+      "epoch": 0.6179958149267613,
+      "grad_norm": 1.971998147575059,
+      "learning_rate": 9.663568848709194e-06,
+      "loss": 0.7279,
+      "step": 443
+    },
+    {
+      "epoch": 0.6193908393396884,
+      "grad_norm": 2.12121490325952,
+      "learning_rate": 9.660632236909628e-06,
+      "loss": 0.7248,
+      "step": 444
+    },
+    {
+      "epoch": 0.6207858637526157,
+      "grad_norm": 1.91738211282543,
+      "learning_rate": 9.657683314465948e-06,
+      "loss": 0.6881,
+      "step": 445
+    },
+    {
+      "epoch": 0.6221808881655428,
+      "grad_norm": 2.1377260400944667,
+      "learning_rate": 9.65472208916747e-06,
+      "loss": 0.7835,
+      "step": 446
+    },
+    {
+      "epoch": 0.6235759125784701,
+      "grad_norm": 2.1471575379428147,
+      "learning_rate": 9.651748568836007e-06,
+      "loss": 0.7644,
+      "step": 447
+    },
+    {
+      "epoch": 0.6249709369913974,
+      "grad_norm": 2.0889010116447744,
+      "learning_rate": 9.648762761325847e-06,
+      "loss": 0.6891,
+      "step": 448
+    },
+    {
+      "epoch": 0.6263659614043245,
+      "grad_norm": 2.052766079746152,
+      "learning_rate": 9.645764674523732e-06,
+      "loss": 0.7842,
+      "step": 449
+    },
+    {
+      "epoch": 0.6277609858172518,
+      "grad_norm": 2.0717698815155754,
+      "learning_rate": 9.642754316348846e-06,
+      "loss": 0.7092,
+      "step": 450
+    },
+    {
+      "epoch": 0.629156010230179,
+      "grad_norm": 2.003583787136136,
+      "learning_rate": 9.639731694752776e-06,
+      "loss": 0.7502,
+      "step": 451
+    },
+    {
+      "epoch": 0.6305510346431062,
+      "grad_norm": 2.2022170633535905,
+      "learning_rate": 9.636696817719511e-06,
+      "loss": 0.7535,
+      "step": 452
+    },
+    {
+      "epoch": 0.6319460590560335,
+      "grad_norm": 2.2113641807562483,
+      "learning_rate": 9.633649693265406e-06,
+      "loss": 0.7492,
+      "step": 453
+    },
+    {
+      "epoch": 0.6333410834689607,
+      "grad_norm": 2.2244980414553717,
+      "learning_rate": 9.630590329439169e-06,
+      "loss": 0.7785,
+      "step": 454
+    },
+    {
+      "epoch": 0.634736107881888,
+      "grad_norm": 2.1937904123098346,
+      "learning_rate": 9.627518734321837e-06,
+      "loss": 0.7498,
+      "step": 455
+    },
+    {
+      "epoch": 0.6361311322948151,
+      "grad_norm": 2.1493347301981562,
+      "learning_rate": 9.624434916026752e-06,
+      "loss": 0.7419,
+      "step": 456
+    },
+    {
+      "epoch": 0.6375261567077424,
+      "grad_norm": 2.052486403142985,
+      "learning_rate": 9.621338882699547e-06,
+      "loss": 0.7394,
+      "step": 457
+    },
+    {
+      "epoch": 0.6389211811206696,
+      "grad_norm": 2.094585882639152,
+      "learning_rate": 9.618230642518117e-06,
+      "loss": 0.6976,
+      "step": 458
+    },
+    {
+      "epoch": 0.6403162055335968,
+      "grad_norm": 2.0524171839415466,
+      "learning_rate": 9.615110203692602e-06,
+      "loss": 0.7693,
+      "step": 459
+    },
+    {
+      "epoch": 0.6417112299465241,
+      "grad_norm": 2.1271109954195255,
+      "learning_rate": 9.61197757446536e-06,
+      "loss": 0.7251,
+      "step": 460
+    },
+    {
+      "epoch": 0.6431062543594513,
+      "grad_norm": 1.9893310410454985,
+      "learning_rate": 9.608832763110955e-06,
+      "loss": 0.6991,
+      "step": 461
+    },
+    {
+      "epoch": 0.6445012787723785,
+      "grad_norm": 2.1917383599900195,
+      "learning_rate": 9.605675777936123e-06,
+      "loss": 0.721,
+      "step": 462
+    },
+    {
+      "epoch": 0.6458963031853058,
+      "grad_norm": 2.065643026017452,
+      "learning_rate": 9.60250662727976e-06,
+      "loss": 0.7297,
+      "step": 463
+    },
+    {
+      "epoch": 0.6472913275982329,
+      "grad_norm": 2.110893858590918,
+      "learning_rate": 9.599325319512893e-06,
+      "loss": 0.7308,
+      "step": 464
+    },
+    {
+      "epoch": 0.6486863520111602,
+      "grad_norm": 2.0073869238725686,
+      "learning_rate": 9.596131863038664e-06,
+      "loss": 0.7128,
+      "step": 465
+    },
+    {
+      "epoch": 0.6500813764240875,
+      "grad_norm": 2.0487238199374116,
+      "learning_rate": 9.592926266292305e-06,
+      "loss": 0.7089,
+      "step": 466
+    },
+    {
+      "epoch": 0.6514764008370146,
+      "grad_norm": 2.256500523858216,
+      "learning_rate": 9.589708537741109e-06,
+      "loss": 0.7346,
+      "step": 467
+    },
+    {
+      "epoch": 0.6528714252499419,
+      "grad_norm": 2.02187405184828,
+      "learning_rate": 9.586478685884424e-06,
+      "loss": 0.6863,
+      "step": 468
+    },
+    {
+      "epoch": 0.6542664496628691,
+      "grad_norm": 2.079210108446843,
+      "learning_rate": 9.583236719253611e-06,
+      "loss": 0.7043,
+      "step": 469
+    },
+    {
+      "epoch": 0.6556614740757963,
+      "grad_norm": 2.1732934816756595,
+      "learning_rate": 9.579982646412039e-06,
+      "loss": 0.7179,
+      "step": 470
+    },
+    {
+      "epoch": 0.6570564984887236,
+      "grad_norm": 1.9394830539544319,
+      "learning_rate": 9.576716475955048e-06,
+      "loss": 0.6399,
+      "step": 471
+    },
+    {
+      "epoch": 0.6584515229016508,
+      "grad_norm": 2.056824279904441,
+      "learning_rate": 9.573438216509937e-06,
+      "loss": 0.6993,
+      "step": 472
+    },
+    {
+      "epoch": 0.659846547314578,
+      "grad_norm": 2.0143468165401277,
+      "learning_rate": 9.570147876735937e-06,
+      "loss": 0.7372,
+      "step": 473
+    },
+    {
+      "epoch": 0.6612415717275052,
+      "grad_norm": 1.8931355054634165,
+      "learning_rate": 9.566845465324185e-06,
+      "loss": 0.6776,
+      "step": 474
+    },
+    {
+      "epoch": 0.6626365961404325,
+      "grad_norm": 1.9966387912195087,
+      "learning_rate": 9.563530990997707e-06,
+      "loss": 0.7671,
+      "step": 475
+    },
+    {
+      "epoch": 0.6640316205533597,
+      "grad_norm": 2.0817025057585106,
+      "learning_rate": 9.560204462511392e-06,
+      "loss": 0.7347,
+      "step": 476
+    },
+    {
+      "epoch": 0.6654266449662869,
+      "grad_norm": 1.9834426105549026,
+      "learning_rate": 9.556865888651965e-06,
+      "loss": 0.7549,
+      "step": 477
+    },
+    {
+      "epoch": 0.6668216693792142,
+      "grad_norm": 1.9850422235674683,
+      "learning_rate": 9.553515278237975e-06,
+      "loss": 0.6782,
+      "step": 478
+    },
+    {
+      "epoch": 0.6682166937921413,
+      "grad_norm": 2.1734748685895076,
+      "learning_rate": 9.550152640119757e-06,
+      "loss": 0.7347,
+      "step": 479
+    },
+    {
+      "epoch": 0.6696117182050686,
+      "grad_norm": 2.141674611099673,
+      "learning_rate": 9.546777983179421e-06,
+      "loss": 0.7581,
+      "step": 480
+    },
+    {
+      "epoch": 0.6710067426179959,
+      "grad_norm": 1.8999376232525609,
+      "learning_rate": 9.543391316330822e-06,
+      "loss": 0.6363,
+      "step": 481
+    },
+    {
+      "epoch": 0.672401767030923,
+      "grad_norm": 2.0279797772457187,
+      "learning_rate": 9.539992648519538e-06,
+      "loss": 0.7463,
+      "step": 482
+    },
+    {
+      "epoch": 0.6737967914438503,
+      "grad_norm": 2.232464103440747,
+      "learning_rate": 9.536581988722848e-06,
+      "loss": 0.8279,
+      "step": 483
+    },
+    {
+      "epoch": 0.6751918158567775,
+      "grad_norm": 2.1498923126332934,
+      "learning_rate": 9.533159345949704e-06,
+      "loss": 0.7615,
+      "step": 484
+    },
+    {
+      "epoch": 0.6765868402697047,
+      "grad_norm": 1.9252842661024974,
+      "learning_rate": 9.529724729240712e-06,
+      "loss": 0.6805,
+      "step": 485
+    },
+    {
+      "epoch": 0.677981864682632,
+      "grad_norm": 2.0678336057421647,
+      "learning_rate": 9.526278147668104e-06,
+      "loss": 0.696,
+      "step": 486
+    },
+    {
+      "epoch": 0.6793768890955592,
+      "grad_norm": 2.075760114316058,
+      "learning_rate": 9.522819610335721e-06,
+      "loss": 0.7246,
+      "step": 487
+    },
+    {
+      "epoch": 0.6807719135084864,
+      "grad_norm": 1.8528598818636521,
+      "learning_rate": 9.519349126378975e-06,
+      "loss": 0.6839,
+      "step": 488
+    },
+    {
+      "epoch": 0.6821669379214136,
+      "grad_norm": 2.200019813823996,
+      "learning_rate": 9.515866704964846e-06,
+      "loss": 0.7495,
+      "step": 489
+    },
+    {
+      "epoch": 0.6835619623343409,
+      "grad_norm": 2.0916960362541683,
+      "learning_rate": 9.512372355291838e-06,
+      "loss": 0.7135,
+      "step": 490
+    },
+    {
+      "epoch": 0.6849569867472681,
+      "grad_norm": 1.8178206174973344,
+      "learning_rate": 9.50886608658996e-06,
+      "loss": 0.6405,
+      "step": 491
+    },
+    {
+      "epoch": 0.6863520111601953,
+      "grad_norm": 2.038956346042495,
+      "learning_rate": 9.505347908120712e-06,
+      "loss": 0.7408,
+      "step": 492
+    },
+    {
+      "epoch": 0.6877470355731226,
+      "grad_norm": 2.0715000507645147,
+      "learning_rate": 9.501817829177046e-06,
+      "loss": 0.7166,
+      "step": 493
+    },
+    {
+      "epoch": 0.6891420599860497,
+      "grad_norm": 2.1401200540716094,
+      "learning_rate": 9.498275859083353e-06,
+      "loss": 0.7301,
+      "step": 494
+    },
+    {
+      "epoch": 0.690537084398977,
+      "grad_norm": 2.042730741273458,
+      "learning_rate": 9.494722007195427e-06,
+      "loss": 0.7038,
+      "step": 495
+    },
+    {
+      "epoch": 0.6919321088119043,
+      "grad_norm": 2.0381664858059594,
+      "learning_rate": 9.491156282900454e-06,
+      "loss": 0.7126,
+      "step": 496
+    },
+    {
+      "epoch": 0.6933271332248314,
+      "grad_norm": 2.1216424421132616,
+      "learning_rate": 9.487578695616974e-06,
+      "loss": 0.7825,
+      "step": 497
+    },
+    {
+      "epoch": 0.6947221576377587,
+      "grad_norm": 2.1488526330119817,
+      "learning_rate": 9.483989254794865e-06,
+      "loss": 0.7303,
+      "step": 498
+    },
+    {
+      "epoch": 0.6961171820506858,
+      "grad_norm": 1.9356761977741135,
+      "learning_rate": 9.480387969915318e-06,
+      "loss": 0.6243,
+      "step": 499
+    },
+    {
+      "epoch": 0.6975122064636131,
+      "grad_norm": 2.0978054484537436,
+      "learning_rate": 9.476774850490803e-06,
+      "loss": 0.738,
+      "step": 500
+    },
+    {
+      "epoch": 0.6989072308765404,
+      "grad_norm": 2.055561455319892,
+      "learning_rate": 9.47314990606505e-06,
+      "loss": 0.7204,
+      "step": 501
+    },
+    {
+      "epoch": 0.7003022552894675,
+      "grad_norm": 2.0680562366639017,
+      "learning_rate": 9.46951314621303e-06,
+      "loss": 0.6993,
+      "step": 502
+    },
+    {
+      "epoch": 0.7016972797023948,
+      "grad_norm": 2.2996197722023473,
+      "learning_rate": 9.465864580540917e-06,
+      "loss": 0.8076,
+      "step": 503
+    },
+    {
+      "epoch": 0.703092304115322,
+      "grad_norm": 2.120383099992889,
+      "learning_rate": 9.462204218686075e-06,
+      "loss": 0.703,
+      "step": 504
+    },
+    {
+      "epoch": 0.7044873285282492,
+      "grad_norm": 2.089709579485426,
+      "learning_rate": 9.458532070317021e-06,
+      "loss": 0.727,
+      "step": 505
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 2.003557653106253,
+      "learning_rate": 9.454848145133406e-06,
+      "loss": 0.7068,
+      "step": 506
+    },
+    {
+      "epoch": 0.7072773773541037,
+      "grad_norm": 2.2817445962593568,
+      "learning_rate": 9.451152452865991e-06,
+      "loss": 0.7133,
+      "step": 507
+    },
+    {
+      "epoch": 0.708672401767031,
+      "grad_norm": 1.9539415691936202,
+      "learning_rate": 9.447445003276618e-06,
+      "loss": 0.6586,
+      "step": 508
+    },
+    {
+      "epoch": 0.7100674261799581,
+      "grad_norm": 2.1970670306323252,
+      "learning_rate": 9.443725806158182e-06,
+      "loss": 0.7327,
+      "step": 509
+    },
+    {
+      "epoch": 0.7114624505928854,
+      "grad_norm": 1.9539434604144215,
+      "learning_rate": 9.439994871334614e-06,
+      "loss": 0.716,
+      "step": 510
+    },
+    {
+      "epoch": 0.7128574750058126,
+      "grad_norm": 2.187186646260326,
+      "learning_rate": 9.43625220866084e-06,
+      "loss": 0.7124,
+      "step": 511
+    },
+    {
+      "epoch": 0.7142524994187398,
+      "grad_norm": 2.1115906329667467,
+      "learning_rate": 9.432497828022775e-06,
+      "loss": 0.7073,
+      "step": 512
+    },
+    {
+      "epoch": 0.7156475238316671,
+      "grad_norm": 2.162561189896867,
+      "learning_rate": 9.428731739337277e-06,
+      "loss": 0.8114,
+      "step": 513
+    },
+    {
+      "epoch": 0.7170425482445942,
+      "grad_norm": 2.229988057904462,
+      "learning_rate": 9.424953952552134e-06,
+      "loss": 0.7294,
+      "step": 514
+    },
+    {
+      "epoch": 0.7184375726575215,
+      "grad_norm": 2.0444286689625777,
+      "learning_rate": 9.421164477646031e-06,
+      "loss": 0.7036,
+      "step": 515
+    },
+    {
+      "epoch": 0.7198325970704488,
+      "grad_norm": 2.234165446135079,
+      "learning_rate": 9.41736332462853e-06,
+      "loss": 0.7559,
+      "step": 516
+    },
+    {
+      "epoch": 0.7212276214833759,
+      "grad_norm": 2.119512330275446,
+      "learning_rate": 9.413550503540039e-06,
+      "loss": 0.7381,
+      "step": 517
+    },
+    {
+      "epoch": 0.7226226458963032,
+      "grad_norm": 2.0076919464556537,
+      "learning_rate": 9.409726024451781e-06,
+      "loss": 0.7415,
+      "step": 518
+    },
+    {
+      "epoch": 0.7240176703092304,
+      "grad_norm": 1.9682427916024752,
+      "learning_rate": 9.40588989746578e-06,
+      "loss": 0.689,
+      "step": 519
+    },
+    {
+      "epoch": 0.7254126947221576,
+      "grad_norm": 2.018101001493541,
+      "learning_rate": 9.402042132714817e-06,
+      "loss": 0.701,
+      "step": 520
+    },
+    {
+      "epoch": 0.7268077191350849,
+      "grad_norm": 2.0465788846719306,
+      "learning_rate": 9.398182740362424e-06,
+      "loss": 0.7293,
+      "step": 521
+    },
+    {
+      "epoch": 0.7282027435480121,
+      "grad_norm": 2.0004289134814948,
+      "learning_rate": 9.39431173060284e-06,
+      "loss": 0.7275,
+      "step": 522
+    },
+    {
+      "epoch": 0.7295977679609393,
+      "grad_norm": 1.9897615420029091,
+      "learning_rate": 9.390429113660993e-06,
+      "loss": 0.7067,
+      "step": 523
+    },
+    {
+      "epoch": 0.7309927923738665,
+      "grad_norm": 2.108334666567339,
+      "learning_rate": 9.38653489979247e-06,
+      "loss": 0.7598,
+      "step": 524
+    },
+    {
+      "epoch": 0.7323878167867938,
+      "grad_norm": 2.0753873908464016,
+      "learning_rate": 9.382629099283486e-06,
+      "loss": 0.7073,
+      "step": 525
+    },
+    {
+      "epoch": 0.733782841199721,
+      "grad_norm": 2.0224236511779643,
+      "learning_rate": 9.378711722450866e-06,
+      "loss": 0.7318,
+      "step": 526
+    },
+    {
+      "epoch": 0.7351778656126482,
+      "grad_norm": 2.030128554628202,
+      "learning_rate": 9.374782779642013e-06,
+      "loss": 0.7123,
+      "step": 527
+    },
+    {
+      "epoch": 0.7365728900255755,
+      "grad_norm": 2.0236490043479325,
+      "learning_rate": 9.370842281234876e-06,
+      "loss": 0.7294,
+      "step": 528
+    },
+    {
+      "epoch": 0.7379679144385026,
+      "grad_norm": 1.9068088991344194,
+      "learning_rate": 9.366890237637932e-06,
+      "loss": 0.6811,
+      "step": 529
+    },
+    {
+      "epoch": 0.7393629388514299,
+      "grad_norm": 1.9480660854694714,
+      "learning_rate": 9.362926659290149e-06,
+      "loss": 0.6497,
+      "step": 530
+    },
+    {
+      "epoch": 0.7407579632643572,
+      "grad_norm": 2.0886372769291,
+      "learning_rate": 9.358951556660968e-06,
+      "loss": 0.7122,
+      "step": 531
+    },
+    {
+      "epoch": 0.7421529876772843,
+      "grad_norm": 2.0028547451656307,
+      "learning_rate": 9.354964940250269e-06,
+      "loss": 0.6725,
+      "step": 532
+    },
+    {
+      "epoch": 0.7435480120902116,
+      "grad_norm": 1.9028605868969077,
+      "learning_rate": 9.35096682058834e-06,
+      "loss": 0.6919,
+      "step": 533
+    },
+    {
+      "epoch": 0.7449430365031388,
+      "grad_norm": 2.0144375936051304,
+      "learning_rate": 9.346957208235857e-06,
+      "loss": 0.7434,
+      "step": 534
+    },
+    {
+      "epoch": 0.746338060916066,
+      "grad_norm": 2.0375245422409587,
+      "learning_rate": 9.342936113783855e-06,
+      "loss": 0.701,
+      "step": 535
+    },
+    {
+      "epoch": 0.7477330853289933,
+      "grad_norm": 2.1341077847797685,
+      "learning_rate": 9.338903547853698e-06,
+      "loss": 0.727,
+      "step": 536
+    },
+    {
+      "epoch": 0.7491281097419205,
+      "grad_norm": 2.1262987908622346,
+      "learning_rate": 9.334859521097046e-06,
+      "loss": 0.7128,
+      "step": 537
+    },
+    {
+      "epoch": 0.7505231341548477,
+      "grad_norm": 1.9324820289749198,
+      "learning_rate": 9.330804044195836e-06,
+      "loss": 0.6821,
+      "step": 538
+    },
+    {
+      "epoch": 0.7519181585677749,
+      "grad_norm": 2.1731776165453995,
+      "learning_rate": 9.326737127862249e-06,
+      "loss": 0.7122,
+      "step": 539
+    },
+    {
+      "epoch": 0.7533131829807022,
+      "grad_norm": 2.16764837125525,
+      "learning_rate": 9.32265878283868e-06,
+      "loss": 0.6994,
+      "step": 540
+    },
+    {
+      "epoch": 0.7547082073936294,
+      "grad_norm": 2.1002883470462534,
+      "learning_rate": 9.318569019897713e-06,
+      "loss": 0.7543,
+      "step": 541
+    },
+    {
+      "epoch": 0.7561032318065566,
+      "grad_norm": 1.9503441807929596,
+      "learning_rate": 9.314467849842093e-06,
+      "loss": 0.7119,
+      "step": 542
+    },
+    {
+      "epoch": 0.7574982562194839,
+      "grad_norm": 2.082793001639134,
+      "learning_rate": 9.310355283504696e-06,
+      "loss": 0.7421,
+      "step": 543
+    },
+    {
+      "epoch": 0.758893280632411,
+      "grad_norm": 1.9966160162515905,
+      "learning_rate": 9.306231331748496e-06,
+      "loss": 0.7221,
+      "step": 544
+    },
+    {
+      "epoch": 0.7602883050453383,
+      "grad_norm": 2.0065387403716475,
+      "learning_rate": 9.302096005466547e-06,
+      "loss": 0.6442,
+      "step": 545
+    },
+    {
+      "epoch": 0.7616833294582656,
+      "grad_norm": 2.2265608152032033,
+      "learning_rate": 9.29794931558194e-06,
+      "loss": 0.729,
+      "step": 546
+    },
+    {
+      "epoch": 0.7630783538711927,
+      "grad_norm": 2.1210857543143544,
+      "learning_rate": 9.29379127304779e-06,
+      "loss": 0.7336,
+      "step": 547
+    },
+    {
+      "epoch": 0.76447337828412,
+      "grad_norm": 2.013283139083976,
+      "learning_rate": 9.289621888847194e-06,
+      "loss": 0.7628,
+      "step": 548
+    },
+    {
+      "epoch": 0.7658684026970471,
+      "grad_norm": 1.8491716982292445,
+      "learning_rate": 9.285441173993207e-06,
+      "loss": 0.6365,
+      "step": 549
+    },
+    {
+      "epoch": 0.7672634271099744,
+      "grad_norm": 2.0841548133116783,
+      "learning_rate": 9.281249139528816e-06,
+      "loss": 0.7515,
+      "step": 550
+    },
+    {
+      "epoch": 0.7686584515229017,
+      "grad_norm": 2.0286961204733256,
+      "learning_rate": 9.277045796526904e-06,
+      "loss": 0.7287,
+      "step": 551
+    },
+    {
+      "epoch": 0.7700534759358288,
+      "grad_norm": 2.068176459964423,
+      "learning_rate": 9.272831156090229e-06,
+      "loss": 0.7283,
+      "step": 552
+    },
+    {
+      "epoch": 0.7714485003487561,
+      "grad_norm": 1.9633423625558333,
+      "learning_rate": 9.268605229351387e-06,
+      "loss": 0.7137,
+      "step": 553
+    },
+    {
+      "epoch": 0.7728435247616833,
+      "grad_norm": 2.100616722256089,
+      "learning_rate": 9.264368027472785e-06,
+      "loss": 0.723,
+      "step": 554
+    },
+    {
+      "epoch": 0.7742385491746105,
+      "grad_norm": 2.0743653211044357,
+      "learning_rate": 9.260119561646614e-06,
+      "loss": 0.6835,
+      "step": 555
+    },
+    {
+      "epoch": 0.7756335735875378,
+      "grad_norm": 2.0116199805624735,
+      "learning_rate": 9.255859843094817e-06,
+      "loss": 0.7277,
+      "step": 556
+    },
+    {
+      "epoch": 0.777028598000465,
+      "grad_norm": 2.1259690175231927,
+      "learning_rate": 9.25158888306906e-06,
+      "loss": 0.6814,
+      "step": 557
+    },
+    {
+      "epoch": 0.7784236224133922,
+      "grad_norm": 2.0748219863598023,
+      "learning_rate": 9.247306692850705e-06,
+      "loss": 0.7092,
+      "step": 558
+    },
+    {
+      "epoch": 0.7798186468263194,
+      "grad_norm": 2.114742076251193,
+      "learning_rate": 9.243013283750774e-06,
+      "loss": 0.7712,
+      "step": 559
+    },
+    {
+      "epoch": 0.7812136712392467,
+      "grad_norm": 1.9896360099254937,
+      "learning_rate": 9.238708667109924e-06,
+      "loss": 0.7242,
+      "step": 560
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 2.002612155228932,
+      "learning_rate": 9.234392854298414e-06,
+      "loss": 0.7389,
+      "step": 561
+    },
+    {
+      "epoch": 0.7840037200651011,
+      "grad_norm": 1.8844768635672766,
+      "learning_rate": 9.230065856716081e-06,
+      "loss": 0.6751,
+      "step": 562
+    },
+    {
+      "epoch": 0.7853987444780284,
+      "grad_norm": 1.9496677958310997,
+      "learning_rate": 9.225727685792302e-06,
+      "loss": 0.7079,
+      "step": 563
+    },
+    {
+      "epoch": 0.7867937688909556,
+      "grad_norm": 2.209067432956444,
+      "learning_rate": 9.221378352985967e-06,
+      "loss": 0.7749,
+      "step": 564
+    },
+    {
+      "epoch": 0.7881887933038828,
+      "grad_norm": 2.0722492473038674,
+      "learning_rate": 9.217017869785453e-06,
+      "loss": 0.7057,
+      "step": 565
+    },
+    {
+      "epoch": 0.7895838177168101,
+      "grad_norm": 2.089575443137789,
+      "learning_rate": 9.212646247708585e-06,
+      "loss": 0.7431,
+      "step": 566
+    },
+    {
+      "epoch": 0.7909788421297372,
+      "grad_norm": 2.063481302101242,
+      "learning_rate": 9.208263498302613e-06,
+      "loss": 0.7566,
+      "step": 567
+    },
+    {
+      "epoch": 0.7923738665426645,
+      "grad_norm": 1.954690180809161,
+      "learning_rate": 9.203869633144182e-06,
+      "loss": 0.7144,
+      "step": 568
+    },
+    {
+      "epoch": 0.7937688909555918,
+      "grad_norm": 2.0082313404704295,
+      "learning_rate": 9.19946466383929e-06,
+      "loss": 0.6873,
+      "step": 569
+    },
+    {
+      "epoch": 0.7951639153685189,
+      "grad_norm": 1.9452585653387595,
+      "learning_rate": 9.19504860202327e-06,
+      "loss": 0.6699,
+      "step": 570
+    },
+    {
+      "epoch": 0.7965589397814462,
+      "grad_norm": 1.9448395893952017,
+      "learning_rate": 9.19062145936076e-06,
+      "loss": 0.6453,
+      "step": 571
+    },
+    {
+      "epoch": 0.7979539641943734,
+      "grad_norm": 2.195833612126511,
+      "learning_rate": 9.186183247545657e-06,
+      "loss": 0.783,
+      "step": 572
+    },
+    {
+      "epoch": 0.7993489886073006,
+      "grad_norm": 2.094994509102352,
+      "learning_rate": 9.181733978301103e-06,
+      "loss": 0.715,
+      "step": 573
+    },
+    {
+      "epoch": 0.8007440130202279,
+      "grad_norm": 2.0540910795180705,
+      "learning_rate": 9.177273663379449e-06,
+      "loss": 0.6741,
+      "step": 574
+    },
+    {
+      "epoch": 0.8021390374331551,
+      "grad_norm": 2.2376315377141442,
+      "learning_rate": 9.172802314562214e-06,
+      "loss": 0.7622,
+      "step": 575
+    },
+    {
+      "epoch": 0.8035340618460823,
+      "grad_norm": 1.9630842252894045,
+      "learning_rate": 9.16831994366007e-06,
+      "loss": 0.6517,
+      "step": 576
+    },
+    {
+      "epoch": 0.8049290862590095,
+      "grad_norm": 2.030945005147177,
+      "learning_rate": 9.1638265625128e-06,
+      "loss": 0.6638,
+      "step": 577
+    },
+    {
+      "epoch": 0.8063241106719368,
+      "grad_norm": 2.2152569686775547,
+      "learning_rate": 9.159322182989265e-06,
+      "loss": 0.7079,
+      "step": 578
+    },
+    {
+      "epoch": 0.807719135084864,
+      "grad_norm": 2.1222069890221342,
+      "learning_rate": 9.154806816987386e-06,
+      "loss": 0.6598,
+      "step": 579
+    },
+    {
+      "epoch": 0.8091141594977912,
+      "grad_norm": 2.039956611874931,
+      "learning_rate": 9.150280476434098e-06,
+      "loss": 0.6929,
+      "step": 580
+    },
+    {
+      "epoch": 0.8105091839107185,
+      "grad_norm": 1.9595741185624305,
+      "learning_rate": 9.145743173285325e-06,
+      "loss": 0.6494,
+      "step": 581
+    },
+    {
+      "epoch": 0.8119042083236456,
+      "grad_norm": 1.914717122402942,
+      "learning_rate": 9.141194919525949e-06,
+      "loss": 0.7067,
+      "step": 582
+    },
+    {
+      "epoch": 0.8132992327365729,
+      "grad_norm": 2.073947853584459,
+      "learning_rate": 9.136635727169776e-06,
+      "loss": 0.7273,
+      "step": 583
+    },
+    {
+      "epoch": 0.8146942571495002,
+      "grad_norm": 2.0532623111872086,
+      "learning_rate": 9.132065608259505e-06,
+      "loss": 0.7517,
+      "step": 584
+    },
+    {
+      "epoch": 0.8160892815624273,
+      "grad_norm": 2.1186320230239937,
+      "learning_rate": 9.127484574866699e-06,
+      "loss": 0.7242,
+      "step": 585
+    },
+    {
+      "epoch": 0.8174843059753546,
+      "grad_norm": 2.0267276329942097,
+      "learning_rate": 9.122892639091748e-06,
+      "loss": 0.7265,
+      "step": 586
+    },
+    {
+      "epoch": 0.8188793303882818,
+      "grad_norm": 1.9639315452093837,
+      "learning_rate": 9.118289813063842e-06,
+      "loss": 0.6903,
+      "step": 587
+    },
+    {
+      "epoch": 0.820274354801209,
+      "grad_norm": 1.9810738468423978,
+      "learning_rate": 9.11367610894093e-06,
+      "loss": 0.7175,
+      "step": 588
+    },
+    {
+      "epoch": 0.8216693792141363,
+      "grad_norm": 2.012247865597074,
+      "learning_rate": 9.109051538909707e-06,
+      "loss": 0.6597,
+      "step": 589
+    },
+    {
+      "epoch": 0.8230644036270635,
+      "grad_norm": 2.075632269020032,
+      "learning_rate": 9.104416115185557e-06,
+      "loss": 0.7694,
+      "step": 590
+    },
+    {
+      "epoch": 0.8244594280399907,
+      "grad_norm": 2.143338879263666,
+      "learning_rate": 9.099769850012539e-06,
+      "loss": 0.7563,
+      "step": 591
+    },
+    {
+      "epoch": 0.8258544524529179,
+      "grad_norm": 1.970372364129679,
+      "learning_rate": 9.095112755663349e-06,
+      "loss": 0.6855,
+      "step": 592
+    },
+    {
+      "epoch": 0.8272494768658452,
+      "grad_norm": 1.9819891434355508,
+      "learning_rate": 9.090444844439284e-06,
+      "loss": 0.6982,
+      "step": 593
+    },
+    {
+      "epoch": 0.8286445012787724,
+      "grad_norm": 1.9951180654869123,
+      "learning_rate": 9.085766128670218e-06,
+      "loss": 0.7537,
+      "step": 594
+    },
+    {
+      "epoch": 0.8300395256916996,
+      "grad_norm": 2.136480645592744,
+      "learning_rate": 9.08107662071456e-06,
+      "loss": 0.7076,
+      "step": 595
+    },
+    {
+      "epoch": 0.8314345501046269,
+      "grad_norm": 2.01008074884727,
+      "learning_rate": 9.076376332959222e-06,
+      "loss": 0.724,
+      "step": 596
+    },
+    {
+      "epoch": 0.832829574517554,
+      "grad_norm": 2.083568256311646,
+      "learning_rate": 9.071665277819603e-06,
+      "loss": 0.6798,
+      "step": 597
+    },
+    {
+      "epoch": 0.8342245989304813,
+      "grad_norm": 2.0713110208220917,
+      "learning_rate": 9.066943467739529e-06,
+      "loss": 0.7138,
+      "step": 598
+    },
+    {
+      "epoch": 0.8356196233434086,
+      "grad_norm": 2.1995098528060653,
+      "learning_rate": 9.06221091519124e-06,
+      "loss": 0.7579,
+      "step": 599
+    },
+    {
+      "epoch": 0.8370146477563357,
+      "grad_norm": 2.1177635938489097,
+      "learning_rate": 9.057467632675357e-06,
+      "loss": 0.7709,
+      "step": 600
+    },
+    {
+      "epoch": 0.838409672169263,
+      "grad_norm": 1.9726857051363949,
+      "learning_rate": 9.05271363272083e-06,
+      "loss": 0.7283,
+      "step": 601
+    },
+    {
+      "epoch": 0.8398046965821901,
+      "grad_norm": 2.042816975781451,
+      "learning_rate": 9.047948927884927e-06,
+      "loss": 0.7155,
+      "step": 602
+    },
+    {
+      "epoch": 0.8411997209951174,
+      "grad_norm": 1.9911024298162479,
+      "learning_rate": 9.043173530753196e-06,
+      "loss": 0.7073,
+      "step": 603
+    },
+    {
+      "epoch": 0.8425947454080447,
+      "grad_norm": 2.188550491711223,
+      "learning_rate": 9.038387453939416e-06,
+      "loss": 0.7105,
+      "step": 604
+    },
+    {
+      "epoch": 0.8439897698209718,
+      "grad_norm": 2.1281020722530015,
+      "learning_rate": 9.033590710085584e-06,
+      "loss": 0.6798,
+      "step": 605
+    },
+    {
+      "epoch": 0.8453847942338991,
+      "grad_norm": 1.9560683283165319,
+      "learning_rate": 9.028783311861874e-06,
+      "loss": 0.6667,
+      "step": 606
+    },
+    {
+      "epoch": 0.8467798186468263,
+      "grad_norm": 1.9865019489409481,
+      "learning_rate": 9.023965271966595e-06,
+      "loss": 0.673,
+      "step": 607
+    },
+    {
+      "epoch": 0.8481748430597535,
+      "grad_norm": 1.975291998773914,
+      "learning_rate": 9.019136603126171e-06,
+      "loss": 0.7207,
+      "step": 608
+    },
+    {
+      "epoch": 0.8495698674726808,
+      "grad_norm": 1.9802488936538558,
+      "learning_rate": 9.0142973180951e-06,
+      "loss": 0.7083,
+      "step": 609
+    },
+    {
+      "epoch": 0.850964891885608,
+      "grad_norm": 2.0530113658928224,
+      "learning_rate": 9.00944742965592e-06,
+      "loss": 0.7348,
+      "step": 610
+    },
+    {
+      "epoch": 0.8523599162985352,
+      "grad_norm": 1.891855625943604,
+      "learning_rate": 9.004586950619182e-06,
+      "loss": 0.6832,
+      "step": 611
+    },
+    {
+      "epoch": 0.8537549407114624,
+      "grad_norm": 2.1036407225739033,
+      "learning_rate": 8.999715893823404e-06,
+      "loss": 0.7417,
+      "step": 612
+    },
+    {
+      "epoch": 0.8551499651243897,
+      "grad_norm": 1.9636866345666872,
+      "learning_rate": 8.994834272135049e-06,
+      "loss": 0.7189,
+      "step": 613
+    },
+    {
+      "epoch": 0.856544989537317,
+      "grad_norm": 1.9975444039165857,
+      "learning_rate": 8.989942098448485e-06,
+      "loss": 0.7115,
+      "step": 614
+    },
+    {
+      "epoch": 0.8579400139502441,
+      "grad_norm": 1.9947649682649553,
+      "learning_rate": 8.985039385685952e-06,
+      "loss": 0.698,
+      "step": 615
+    },
+    {
+      "epoch": 0.8593350383631714,
+      "grad_norm": 1.9552201916690322,
+      "learning_rate": 8.98012614679753e-06,
+      "loss": 0.6364,
+      "step": 616
+    },
+    {
+      "epoch": 0.8607300627760985,
+      "grad_norm": 2.2114912964614946,
+      "learning_rate": 8.975202394761098e-06,
+      "loss": 0.6814,
+      "step": 617
+    },
+    {
+      "epoch": 0.8621250871890258,
+      "grad_norm": 2.0505082371309817,
+      "learning_rate": 8.970268142582312e-06,
+      "loss": 0.7083,
+      "step": 618
+    },
+    {
+      "epoch": 0.8635201116019531,
+      "grad_norm": 2.0065112653612003,
+      "learning_rate": 8.965323403294553e-06,
+      "loss": 0.671,
+      "step": 619
+    },
+    {
+      "epoch": 0.8649151360148802,
+      "grad_norm": 2.050198240561233,
+      "learning_rate": 8.960368189958913e-06,
+      "loss": 0.7219,
+      "step": 620
+    },
+    {
+      "epoch": 0.8663101604278075,
+      "grad_norm": 2.0821632762041,
+      "learning_rate": 8.955402515664144e-06,
+      "loss": 0.7255,
+      "step": 621
+    },
+    {
+      "epoch": 0.8677051848407347,
+      "grad_norm": 2.2374361424470472,
+      "learning_rate": 8.950426393526633e-06,
+      "loss": 0.7127,
+      "step": 622
+    },
+    {
+      "epoch": 0.8691002092536619,
+      "grad_norm": 1.949975651680759,
+      "learning_rate": 8.945439836690359e-06,
+      "loss": 0.703,
+      "step": 623
+    },
+    {
+      "epoch": 0.8704952336665892,
+      "grad_norm": 2.018953506212755,
+      "learning_rate": 8.940442858326871e-06,
+      "loss": 0.6668,
+      "step": 624
+    },
+    {
+      "epoch": 0.8718902580795164,
+      "grad_norm": 1.936317481354868,
+      "learning_rate": 8.935435471635238e-06,
+      "loss": 0.7056,
+      "step": 625
+    },
+    {
+      "epoch": 0.8732852824924436,
+      "grad_norm": 2.0282916260482526,
+      "learning_rate": 8.93041768984203e-06,
+      "loss": 0.6652,
+      "step": 626
+    },
+    {
+      "epoch": 0.8746803069053708,
+      "grad_norm": 2.121255903905443,
+      "learning_rate": 8.925389526201264e-06,
+      "loss": 0.6941,
+      "step": 627
+    },
+    {
+      "epoch": 0.8760753313182981,
+      "grad_norm": 1.8461831360013177,
+      "learning_rate": 8.920350993994387e-06,
+      "loss": 0.6375,
+      "step": 628
+    },
+    {
+      "epoch": 0.8774703557312253,
+      "grad_norm": 2.0563491350367724,
+      "learning_rate": 8.915302106530234e-06,
+      "loss": 0.6932,
+      "step": 629
+    },
+    {
+      "epoch": 0.8788653801441525,
+      "grad_norm": 2.0350553302487695,
+      "learning_rate": 8.91024287714499e-06,
+      "loss": 0.6683,
+      "step": 630
+    },
+    {
+      "epoch": 0.8802604045570798,
+      "grad_norm": 1.9629287472963255,
+      "learning_rate": 8.905173319202159e-06,
+      "loss": 0.6976,
+      "step": 631
+    },
+    {
+      "epoch": 0.8816554289700069,
+      "grad_norm": 1.9555638782336193,
+      "learning_rate": 8.900093446092523e-06,
+      "loss": 0.6904,
+      "step": 632
+    },
+    {
+      "epoch": 0.8830504533829342,
+      "grad_norm": 2.010183514705677,
+      "learning_rate": 8.895003271234116e-06,
+      "loss": 0.7959,
+      "step": 633
+    },
+    {
+      "epoch": 0.8844454777958615,
+      "grad_norm": 2.038174167643802,
+      "learning_rate": 8.889902808072178e-06,
+      "loss": 0.6629,
+      "step": 634
+    },
+    {
+      "epoch": 0.8858405022087886,
+      "grad_norm": 2.0007565750548957,
+      "learning_rate": 8.884792070079128e-06,
+      "loss": 0.6973,
+      "step": 635
+    },
+    {
+      "epoch": 0.8872355266217159,
+      "grad_norm": 1.9422648258075415,
+      "learning_rate": 8.879671070754527e-06,
+      "loss": 0.6891,
+      "step": 636
+    },
+    {
+      "epoch": 0.8886305510346431,
+      "grad_norm": 2.0732698051235987,
+      "learning_rate": 8.874539823625037e-06,
+      "loss": 0.7315,
+      "step": 637
+    },
+    {
+      "epoch": 0.8900255754475703,
+      "grad_norm": 1.9047096265263974,
+      "learning_rate": 8.869398342244387e-06,
+      "loss": 0.6898,
+      "step": 638
+    },
+    {
+      "epoch": 0.8914205998604976,
+      "grad_norm": 2.367826429998185,
+      "learning_rate": 8.86424664019334e-06,
+      "loss": 0.7776,
+      "step": 639
+    },
+    {
+      "epoch": 0.8928156242734248,
+      "grad_norm": 1.9082787894649413,
+      "learning_rate": 8.859084731079664e-06,
+      "loss": 0.6917,
+      "step": 640
+    },
+    {
+      "epoch": 0.894210648686352,
+      "grad_norm": 2.0179079093688097,
+      "learning_rate": 8.853912628538072e-06,
+      "loss": 0.7115,
+      "step": 641
+    },
+    {
+      "epoch": 0.8956056730992792,
+      "grad_norm": 1.9593411702559214,
+      "learning_rate": 8.84873034623022e-06,
+      "loss": 0.6561,
+      "step": 642
+    },
+    {
+      "epoch": 0.8970006975122065,
+      "grad_norm": 1.9223616018745664,
+      "learning_rate": 8.84353789784464e-06,
+      "loss": 0.7232,
+      "step": 643
+    },
+    {
+      "epoch": 0.8983957219251337,
+      "grad_norm": 1.9916894711438682,
+      "learning_rate": 8.83833529709672e-06,
+      "loss": 0.7092,
+      "step": 644
+    },
+    {
+      "epoch": 0.8997907463380609,
+      "grad_norm": 1.8700641744681288,
+      "learning_rate": 8.833122557728667e-06,
+      "loss": 0.6773,
+      "step": 645
+    },
+    {
+      "epoch": 0.9011857707509882,
+      "grad_norm": 2.0854934713518376,
+      "learning_rate": 8.827899693509467e-06,
+      "loss": 0.7535,
+      "step": 646
+    },
+    {
+      "epoch": 0.9025807951639153,
+      "grad_norm": 1.9895333488216704,
+      "learning_rate": 8.82266671823485e-06,
+      "loss": 0.7206,
+      "step": 647
+    },
+    {
+      "epoch": 0.9039758195768426,
+      "grad_norm": 1.8896104921166434,
+      "learning_rate": 8.817423645727252e-06,
+      "loss": 0.6875,
+      "step": 648
+    },
+    {
+      "epoch": 0.9053708439897699,
+      "grad_norm": 2.266703090067289,
+      "learning_rate": 8.812170489835784e-06,
+      "loss": 0.7802,
+      "step": 649
+    },
+    {
+      "epoch": 0.906765868402697,
+      "grad_norm": 2.002845695676102,
+      "learning_rate": 8.806907264436183e-06,
+      "loss": 0.6868,
+      "step": 650
+    },
+    {
+      "epoch": 0.9081608928156243,
+      "grad_norm": 1.9373983885808481,
+      "learning_rate": 8.801633983430794e-06,
+      "loss": 0.6593,
+      "step": 651
+    },
+    {
+      "epoch": 0.9095559172285514,
+      "grad_norm": 1.8743487506726677,
+      "learning_rate": 8.796350660748516e-06,
+      "loss": 0.646,
+      "step": 652
+    },
+    {
+      "epoch": 0.9109509416414787,
+      "grad_norm": 2.0801545860927626,
+      "learning_rate": 8.791057310344775e-06,
+      "loss": 0.6965,
+      "step": 653
+    },
+    {
+      "epoch": 0.912345966054406,
+      "grad_norm": 2.037064963904586,
+      "learning_rate": 8.785753946201484e-06,
+      "loss": 0.7919,
+      "step": 654
+    },
+    {
+      "epoch": 0.9137409904673331,
+      "grad_norm": 1.9757786317526882,
+      "learning_rate": 8.780440582327005e-06,
+      "loss": 0.6847,
+      "step": 655
+    },
+    {
+      "epoch": 0.9151360148802604,
+      "grad_norm": 1.904008821134603,
+      "learning_rate": 8.775117232756116e-06,
+      "loss": 0.6505,
+      "step": 656
+    },
+    {
+      "epoch": 0.9165310392931876,
+      "grad_norm": 1.9990189247235826,
+      "learning_rate": 8.769783911549968e-06,
+      "loss": 0.7283,
+      "step": 657
+    },
+    {
+      "epoch": 0.9179260637061148,
+      "grad_norm": 1.9708975603644308,
+      "learning_rate": 8.764440632796055e-06,
+      "loss": 0.7027,
+      "step": 658
+    },
+    {
+      "epoch": 0.9193210881190421,
+      "grad_norm": 1.8662648197783636,
+      "learning_rate": 8.75908741060817e-06,
+      "loss": 0.7245,
+      "step": 659
+    },
+    {
+      "epoch": 0.9207161125319693,
+      "grad_norm": 1.920194446953376,
+      "learning_rate": 8.75372425912637e-06,
+      "loss": 0.7019,
+      "step": 660
+    },
+    {
+      "epoch": 0.9221111369448965,
+      "grad_norm": 2.2198974167640655,
+      "learning_rate": 8.748351192516943e-06,
+      "loss": 0.7775,
+      "step": 661
+    },
+    {
+      "epoch": 0.9235061613578237,
+      "grad_norm": 2.094121585262976,
+      "learning_rate": 8.742968224972366e-06,
+      "loss": 0.6971,
+      "step": 662
+    },
+    {
+      "epoch": 0.924901185770751,
+      "grad_norm": 1.9084243245926524,
+      "learning_rate": 8.737575370711265e-06,
+      "loss": 0.6826,
+      "step": 663
+    },
+    {
+      "epoch": 0.9262962101836782,
+      "grad_norm": 2.032697852643676,
+      "learning_rate": 8.732172643978383e-06,
+      "loss": 0.7119,
+      "step": 664
+    },
+    {
+      "epoch": 0.9276912345966054,
+      "grad_norm": 2.0461653810471416,
+      "learning_rate": 8.726760059044542e-06,
+      "loss": 0.6939,
+      "step": 665
+    },
+    {
+      "epoch": 0.9290862590095327,
+      "grad_norm": 2.130754432146634,
+      "learning_rate": 8.721337630206603e-06,
+      "loss": 0.7491,
+      "step": 666
+    },
+    {
+      "epoch": 0.93048128342246,
+      "grad_norm": 2.0904014494400265,
+      "learning_rate": 8.715905371787426e-06,
+      "loss": 0.7298,
+      "step": 667
+    },
+    {
+      "epoch": 0.9318763078353871,
+      "grad_norm": 1.9843213685121637,
+      "learning_rate": 8.710463298135836e-06,
+      "loss": 0.6429,
+      "step": 668
+    },
+    {
+      "epoch": 0.9332713322483144,
+      "grad_norm": 1.917630694546447,
+      "learning_rate": 8.705011423626589e-06,
+      "loss": 0.7401,
+      "step": 669
+    },
+    {
+      "epoch": 0.9346663566612415,
+      "grad_norm": 2.13375068139831,
+      "learning_rate": 8.699549762660318e-06,
+      "loss": 0.6956,
+      "step": 670
+    },
+    {
+      "epoch": 0.9360613810741688,
+      "grad_norm": 1.975503491309611,
+      "learning_rate": 8.69407832966352e-06,
+      "loss": 0.7035,
+      "step": 671
+    },
+    {
+      "epoch": 0.9374564054870961,
+      "grad_norm": 2.0791411874973518,
+      "learning_rate": 8.688597139088494e-06,
+      "loss": 0.7222,
+      "step": 672
+    },
+    {
+      "epoch": 0.9388514299000232,
+      "grad_norm": 1.8469076240138975,
+      "learning_rate": 8.683106205413316e-06,
+      "loss": 0.6668,
+      "step": 673
+    },
+    {
+      "epoch": 0.9402464543129505,
+      "grad_norm": 1.7928528011941791,
+      "learning_rate": 8.677605543141797e-06,
+      "loss": 0.6437,
+      "step": 674
+    },
+    {
+      "epoch": 0.9416414787258777,
+      "grad_norm": 2.2286342353729136,
+      "learning_rate": 8.672095166803445e-06,
+      "loss": 0.7352,
+      "step": 675
+    },
+    {
+      "epoch": 0.9430365031388049,
+      "grad_norm": 1.905774060183822,
+      "learning_rate": 8.666575090953426e-06,
+      "loss": 0.7106,
+      "step": 676
+    },
+    {
+      "epoch": 0.9444315275517322,
+      "grad_norm": 2.068546911494722,
+      "learning_rate": 8.661045330172533e-06,
+      "loss": 0.7773,
+      "step": 677
+    },
+    {
+      "epoch": 0.9458265519646594,
+      "grad_norm": 1.9281757004651763,
+      "learning_rate": 8.65550589906713e-06,
+      "loss": 0.662,
+      "step": 678
+    },
+    {
+      "epoch": 0.9472215763775866,
+      "grad_norm": 1.855641783248281,
+      "learning_rate": 8.649956812269134e-06,
+      "loss": 0.6778,
+      "step": 679
+    },
+    {
+      "epoch": 0.9486166007905138,
+      "grad_norm": 1.8442734995129473,
+      "learning_rate": 8.644398084435959e-06,
+      "loss": 0.6684,
+      "step": 680
+    },
+    {
+      "epoch": 0.9500116252034411,
+      "grad_norm": 1.9714376993675073,
+      "learning_rate": 8.63882973025049e-06,
+      "loss": 0.6837,
+      "step": 681
+    },
+    {
+      "epoch": 0.9514066496163683,
+      "grad_norm": 1.8876407415682392,
+      "learning_rate": 8.63325176442104e-06,
+      "loss": 0.7157,
+      "step": 682
+    },
+    {
+      "epoch": 0.9528016740292955,
+      "grad_norm": 1.9056546611112206,
+      "learning_rate": 8.627664201681305e-06,
+      "loss": 0.6769,
+      "step": 683
+    },
+    {
+      "epoch": 0.9541966984422228,
+      "grad_norm": 2.0721004187013614,
+      "learning_rate": 8.622067056790333e-06,
+      "loss": 0.7439,
+      "step": 684
+    },
+    {
+      "epoch": 0.9555917228551499,
+      "grad_norm": 1.8805956857202908,
+      "learning_rate": 8.616460344532483e-06,
+      "loss": 0.7115,
+      "step": 685
+    },
+    {
+      "epoch": 0.9569867472680772,
+      "grad_norm": 1.8952885482730308,
+      "learning_rate": 8.610844079717387e-06,
+      "loss": 0.6435,
+      "step": 686
+    },
+    {
+      "epoch": 0.9583817716810045,
+      "grad_norm": 1.9989059022509683,
+      "learning_rate": 8.605218277179907e-06,
+      "loss": 0.7322,
+      "step": 687
+    },
+    {
+      "epoch": 0.9597767960939316,
+      "grad_norm": 1.7616680538530942,
+      "learning_rate": 8.599582951780095e-06,
+      "loss": 0.6235,
+      "step": 688
+    },
+    {
+      "epoch": 0.9611718205068589,
+      "grad_norm": 1.9662690661857283,
+      "learning_rate": 8.593938118403164e-06,
+      "loss": 0.6739,
+      "step": 689
+    },
+    {
+      "epoch": 0.9625668449197861,
+      "grad_norm": 2.0362652708193356,
+      "learning_rate": 8.588283791959437e-06,
+      "loss": 0.7047,
+      "step": 690
+    },
+    {
+      "epoch": 0.9639618693327133,
+      "grad_norm": 2.015021257411785,
+      "learning_rate": 8.582619987384311e-06,
+      "loss": 0.6863,
+      "step": 691
+    },
+    {
+      "epoch": 0.9653568937456406,
+      "grad_norm": 1.8836047793904986,
+      "learning_rate": 8.57694671963822e-06,
+      "loss": 0.652,
+      "step": 692
+    },
+    {
+      "epoch": 0.9667519181585678,
+      "grad_norm": 2.0164471666869357,
+      "learning_rate": 8.571264003706596e-06,
+      "loss": 0.7219,
+      "step": 693
+    },
+    {
+      "epoch": 0.968146942571495,
+      "grad_norm": 1.9205579637723724,
+      "learning_rate": 8.565571854599825e-06,
+      "loss": 0.6737,
+      "step": 694
+    },
+    {
+      "epoch": 0.9695419669844222,
+      "grad_norm": 2.0544942254164766,
+      "learning_rate": 8.559870287353214e-06,
+      "loss": 0.7372,
+      "step": 695
+    },
+    {
+      "epoch": 0.9709369913973495,
+      "grad_norm": 1.966795085808272,
+      "learning_rate": 8.554159317026939e-06,
+      "loss": 0.653,
+      "step": 696
+    },
+    {
+      "epoch": 0.9723320158102767,
+      "grad_norm": 2.021147157715166,
+      "learning_rate": 8.548438958706022e-06,
+      "loss": 0.7458,
+      "step": 697
+    },
+    {
+      "epoch": 0.9737270402232039,
+      "grad_norm": 2.113696018431275,
+      "learning_rate": 8.542709227500276e-06,
+      "loss": 0.7645,
+      "step": 698
+    },
+    {
+      "epoch": 0.9751220646361312,
+      "grad_norm": 1.9206237217188158,
+      "learning_rate": 8.536970138544278e-06,
+      "loss": 0.6644,
+      "step": 699
+    },
+    {
+      "epoch": 0.9765170890490583,
+      "grad_norm": 1.9453633858006116,
+      "learning_rate": 8.531221706997316e-06,
+      "loss": 0.6547,
+      "step": 700
+    },
+    {
+      "epoch": 0.9779121134619856,
+      "grad_norm": 2.0426945048651937,
+      "learning_rate": 8.525463948043365e-06,
+      "loss": 0.7077,
+      "step": 701
+    },
+    {
+      "epoch": 0.9793071378749129,
+      "grad_norm": 1.931889958107307,
+      "learning_rate": 8.519696876891024e-06,
+      "loss": 0.6439,
+      "step": 702
+    },
+    {
+      "epoch": 0.98070216228784,
+      "grad_norm": 1.9159274278923777,
+      "learning_rate": 8.513920508773499e-06,
+      "loss": 0.6966,
+      "step": 703
+    },
+    {
+      "epoch": 0.9820971867007673,
+      "grad_norm": 1.9359622556425784,
+      "learning_rate": 8.508134858948553e-06,
+      "loss": 0.6921,
+      "step": 704
+    },
+    {
+      "epoch": 0.9834922111136944,
+      "grad_norm": 1.9696782797857875,
+      "learning_rate": 8.502339942698463e-06,
+      "loss": 0.6635,
+      "step": 705
+    },
+    {
+      "epoch": 0.9848872355266217,
+      "grad_norm": 1.9677260443909579,
+      "learning_rate": 8.496535775329982e-06,
+      "loss": 0.7245,
+      "step": 706
+    },
+    {
+      "epoch": 0.986282259939549,
+      "grad_norm": 1.9675680167275715,
+      "learning_rate": 8.4907223721743e-06,
+      "loss": 0.6777,
+      "step": 707
+    },
+    {
+      "epoch": 0.9876772843524761,
+      "grad_norm": 1.806078447306801,
+      "learning_rate": 8.484899748587003e-06,
+      "loss": 0.6661,
+      "step": 708
+    },
+    {
+      "epoch": 0.9890723087654034,
+      "grad_norm": 2.2490305816466964,
+      "learning_rate": 8.479067919948032e-06,
+      "loss": 0.6972,
+      "step": 709
+    },
+    {
+      "epoch": 0.9904673331783306,
+      "grad_norm": 2.128480955160666,
+      "learning_rate": 8.473226901661643e-06,
+      "loss": 0.786,
+      "step": 710
+    },
+    {
+      "epoch": 0.9918623575912578,
+      "grad_norm": 2.085090723256057,
+      "learning_rate": 8.46737670915636e-06,
+      "loss": 0.6869,
+      "step": 711
+    },
+    {
+      "epoch": 0.9932573820041851,
+      "grad_norm": 2.0381152772038,
+      "learning_rate": 8.46151735788495e-06,
+      "loss": 0.6968,
+      "step": 712
+    },
+    {
+      "epoch": 0.9946524064171123,
+      "grad_norm": 1.8660889839328152,
+      "learning_rate": 8.455648863324364e-06,
+      "loss": 0.6421,
+      "step": 713
+    },
+    {
+      "epoch": 0.9960474308300395,
+      "grad_norm": 1.8767449139152448,
+      "learning_rate": 8.449771240975707e-06,
+      "loss": 0.6685,
+      "step": 714
+    },
+    {
+      "epoch": 0.9974424552429667,
+      "grad_norm": 1.9246273464410195,
+      "learning_rate": 8.443884506364192e-06,
+      "loss": 0.6835,
+      "step": 715
+    },
+    {
+      "epoch": 0.998837479655894,
+      "grad_norm": 1.99350094104789,
+      "learning_rate": 8.437988675039108e-06,
+      "loss": 0.6762,
+      "step": 716
+    },
+    {
+      "epoch": 1.0013950244129273,
+      "grad_norm": 2.7852354658837086,
+      "learning_rate": 8.432083762573761e-06,
+      "loss": 1.1719,
+      "step": 717
+    },
+    {
+      "epoch": 1.0027900488258545,
+      "grad_norm": 1.7585462074635738,
+      "learning_rate": 8.426169784565452e-06,
+      "loss": 0.4956,
+      "step": 718
+    },
+    {
+      "epoch": 1.0041850732387816,
+      "grad_norm": 1.8518154743606603,
+      "learning_rate": 8.420246756635431e-06,
+      "loss": 0.4464,
+      "step": 719
+    },
+    {
+      "epoch": 1.0055800976517089,
+      "grad_norm": 1.8843110874735378,
+      "learning_rate": 8.414314694428842e-06,
+      "loss": 0.4761,
+      "step": 720
+    },
+    {
+      "epoch": 1.0069751220646361,
+      "grad_norm": 1.6047121473200383,
+      "learning_rate": 8.408373613614699e-06,
+      "loss": 0.4887,
+      "step": 721
+    },
+    {
+      "epoch": 1.0083701464775634,
+      "grad_norm": 1.8784731067028053,
+      "learning_rate": 8.40242352988584e-06,
+      "loss": 0.4422,
+      "step": 722
+    },
+    {
+      "epoch": 1.0097651708904907,
+      "grad_norm": 1.8159678540316857,
+      "learning_rate": 8.396464458958876e-06,
+      "loss": 0.5311,
+      "step": 723
+    },
+    {
+      "epoch": 1.0111601953034177,
+      "grad_norm": 2.027238459653953,
+      "learning_rate": 8.390496416574166e-06,
+      "loss": 0.4802,
+      "step": 724
+    },
+    {
+      "epoch": 1.012555219716345,
+      "grad_norm": 2.179302244779955,
+      "learning_rate": 8.384519418495755e-06,
+      "loss": 0.5205,
+      "step": 725
+    },
+    {
+      "epoch": 1.0139502441292723,
+      "grad_norm": 2.175655456291187,
+      "learning_rate": 8.378533480511355e-06,
+      "loss": 0.4955,
+      "step": 726
+    },
+    {
+      "epoch": 1.0153452685421995,
+      "grad_norm": 2.8562208735960826,
+      "learning_rate": 8.372538618432282e-06,
+      "loss": 0.5112,
+      "step": 727
+    },
+    {
+      "epoch": 1.0167402929551268,
+      "grad_norm": 2.2651246091600314,
+      "learning_rate": 8.366534848093434e-06,
+      "loss": 0.5169,
+      "step": 728
+    },
+    {
+      "epoch": 1.0181353173680538,
+      "grad_norm": 2.2772660173459665,
+      "learning_rate": 8.360522185353234e-06,
+      "loss": 0.4812,
+      "step": 729
+    },
+    {
+      "epoch": 1.0195303417809811,
+      "grad_norm": 1.9713967680804616,
+      "learning_rate": 8.354500646093592e-06,
+      "loss": 0.4655,
+      "step": 730
+    },
+    {
+      "epoch": 1.0209253661939084,
+      "grad_norm": 2.22363796857583,
+      "learning_rate": 8.348470246219872e-06,
+      "loss": 0.5245,
+      "step": 731
+    },
+    {
+      "epoch": 1.0223203906068357,
+      "grad_norm": 2.018998971676804,
+      "learning_rate": 8.342431001660826e-06,
+      "loss": 0.4552,
+      "step": 732
+    },
+    {
+      "epoch": 1.023715415019763,
+      "grad_norm": 1.9417878883494086,
+      "learning_rate": 8.33638292836859e-06,
+      "loss": 0.4881,
+      "step": 733
+    },
+    {
+      "epoch": 1.02511043943269,
+      "grad_norm": 1.9746946629683426,
+      "learning_rate": 8.330326042318605e-06,
+      "loss": 0.5036,
+      "step": 734
+    },
+    {
+      "epoch": 1.0265054638456172,
+      "grad_norm": 1.8367303689092216,
+      "learning_rate": 8.324260359509594e-06,
+      "loss": 0.4848,
+      "step": 735
+    },
+    {
+      "epoch": 1.0279004882585445,
+      "grad_norm": 1.9065855234784623,
+      "learning_rate": 8.31818589596352e-06,
+      "loss": 0.4561,
+      "step": 736
+    },
+    {
+      "epoch": 1.0292955126714718,
+      "grad_norm": 2.203951507850641,
+      "learning_rate": 8.312102667725534e-06,
+      "loss": 0.5201,
+      "step": 737
+    },
+    {
+      "epoch": 1.030690537084399,
+      "grad_norm": 1.86322512488284,
+      "learning_rate": 8.306010690863943e-06,
+      "loss": 0.4636,
+      "step": 738
+    },
+    {
+      "epoch": 1.032085561497326,
+      "grad_norm": 2.021085507857238,
+      "learning_rate": 8.299909981470159e-06,
+      "loss": 0.4868,
+      "step": 739
+    },
+    {
+      "epoch": 1.0334805859102534,
+      "grad_norm": 2.0336313733189715,
+      "learning_rate": 8.29380055565866e-06,
+      "loss": 0.5011,
+      "step": 740
+    },
+    {
+      "epoch": 1.0348756103231807,
+      "grad_norm": 1.9700200483932437,
+      "learning_rate": 8.28768242956695e-06,
+      "loss": 0.4794,
+      "step": 741
+    },
+    {
+      "epoch": 1.036270634736108,
+      "grad_norm": 1.9333816097421348,
+      "learning_rate": 8.281555619355515e-06,
+      "loss": 0.5558,
+      "step": 742
+    },
+    {
+      "epoch": 1.0376656591490352,
+      "grad_norm": 2.1097816531163454,
+      "learning_rate": 8.275420141207775e-06,
+      "loss": 0.5026,
+      "step": 743
+    },
+    {
+      "epoch": 1.0390606835619622,
+      "grad_norm": 2.066539214548412,
+      "learning_rate": 8.269276011330048e-06,
+      "loss": 0.4581,
+      "step": 744
+    },
+    {
+      "epoch": 1.0404557079748895,
+      "grad_norm": 1.8956395492111373,
+      "learning_rate": 8.263123245951504e-06,
+      "loss": 0.5454,
+      "step": 745
+    },
+    {
+      "epoch": 1.0418507323878168,
+      "grad_norm": 2.100666149512502,
+      "learning_rate": 8.256961861324127e-06,
+      "loss": 0.4801,
+      "step": 746
+    },
+    {
+      "epoch": 1.043245756800744,
+      "grad_norm": 2.149566004009623,
+      "learning_rate": 8.250791873722662e-06,
+      "loss": 0.4692,
+      "step": 747
+    },
+    {
+      "epoch": 1.0446407812136713,
+      "grad_norm": 2.054274149085442,
+      "learning_rate": 8.244613299444581e-06,
+      "loss": 0.5404,
+      "step": 748
+    },
+    {
+      "epoch": 1.0460358056265984,
+      "grad_norm": 2.009479396810494,
+      "learning_rate": 8.238426154810035e-06,
+      "loss": 0.4521,
+      "step": 749
+    },
+    {
+      "epoch": 1.0474308300395256,
+      "grad_norm": 2.0004470029627464,
+      "learning_rate": 8.232230456161819e-06,
+      "loss": 0.4466,
+      "step": 750
+    },
+    {
+      "epoch": 1.048825854452453,
+      "grad_norm": 2.154526582868792,
+      "learning_rate": 8.226026219865313e-06,
+      "loss": 0.5059,
+      "step": 751
+    },
+    {
+      "epoch": 1.0502208788653802,
+      "grad_norm": 2.088904126217079,
+      "learning_rate": 8.219813462308458e-06,
+      "loss": 0.4749,
+      "step": 752
+    },
+    {
+      "epoch": 1.0516159032783075,
+      "grad_norm": 2.0692774381070125,
+      "learning_rate": 8.213592199901692e-06,
+      "loss": 0.4313,
+      "step": 753
+    },
+    {
+      "epoch": 1.0530109276912345,
+      "grad_norm": 1.9215415829687428,
+      "learning_rate": 8.207362449077932e-06,
+      "loss": 0.4707,
+      "step": 754
+    },
+    {
+      "epoch": 1.0544059521041618,
+      "grad_norm": 2.0945380875354225,
+      "learning_rate": 8.201124226292505e-06,
+      "loss": 0.4781,
+      "step": 755
+    },
+    {
+      "epoch": 1.055800976517089,
+      "grad_norm": 2.000101834591612,
+      "learning_rate": 8.19487754802312e-06,
+      "loss": 0.4688,
+      "step": 756
+    },
+    {
+      "epoch": 1.0571960009300163,
+      "grad_norm": 2.2202484065922876,
+      "learning_rate": 8.18862243076982e-06,
+      "loss": 0.4761,
+      "step": 757
+    },
+    {
+      "epoch": 1.0585910253429436,
+      "grad_norm": 2.0625752659923187,
+      "learning_rate": 8.18235889105494e-06,
+      "loss": 0.4315,
+      "step": 758
+    },
+    {
+      "epoch": 1.0599860497558706,
+      "grad_norm": 2.045140697582116,
+      "learning_rate": 8.17608694542306e-06,
+      "loss": 0.5018,
+      "step": 759
+    },
+    {
+      "epoch": 1.061381074168798,
+      "grad_norm": 2.003649970486099,
+      "learning_rate": 8.169806610440966e-06,
+      "loss": 0.4967,
+      "step": 760
+    },
+    {
+      "epoch": 1.0627760985817252,
+      "grad_norm": 2.1830296634359616,
+      "learning_rate": 8.163517902697602e-06,
+      "loss": 0.5049,
+      "step": 761
+    },
+    {
+      "epoch": 1.0641711229946524,
+      "grad_norm": 2.0816421018361995,
+      "learning_rate": 8.157220838804026e-06,
+      "loss": 0.5132,
+      "step": 762
+    },
+    {
+      "epoch": 1.0655661474075797,
+      "grad_norm": 2.084476758712473,
+      "learning_rate": 8.150915435393371e-06,
+      "loss": 0.4771,
+      "step": 763
+    },
+    {
+      "epoch": 1.0669611718205068,
+      "grad_norm": 1.9998410644432016,
+      "learning_rate": 8.1446017091208e-06,
+      "loss": 0.4261,
+      "step": 764
+    },
+    {
+      "epoch": 1.068356196233434,
+      "grad_norm": 1.9495671767021174,
+      "learning_rate": 8.138279676663458e-06,
+      "loss": 0.5264,
+      "step": 765
+    },
+    {
+      "epoch": 1.0697512206463613,
+      "grad_norm": 2.195184767639989,
+      "learning_rate": 8.131949354720425e-06,
+      "loss": 0.4366,
+      "step": 766
+    },
+    {
+      "epoch": 1.0711462450592886,
+      "grad_norm": 1.9821878881590826,
+      "learning_rate": 8.125610760012685e-06,
+      "loss": 0.4736,
+      "step": 767
+    },
+    {
+      "epoch": 1.0725412694722158,
+      "grad_norm": 1.9988670353991758,
+      "learning_rate": 8.11926390928307e-06,
+      "loss": 0.5174,
+      "step": 768
+    },
+    {
+      "epoch": 1.073936293885143,
+      "grad_norm": 2.0566037377417365,
+      "learning_rate": 8.112908819296217e-06,
+      "loss": 0.4676,
+      "step": 769
+    },
+    {
+      "epoch": 1.0753313182980702,
+      "grad_norm": 1.987767203894708,
+      "learning_rate": 8.106545506838533e-06,
+      "loss": 0.4954,
+      "step": 770
+    },
+    {
+      "epoch": 1.0767263427109974,
+      "grad_norm": 1.9333413400597803,
+      "learning_rate": 8.100173988718136e-06,
+      "loss": 0.4687,
+      "step": 771
+    },
+    {
+      "epoch": 1.0781213671239247,
+      "grad_norm": 2.0208241483857874,
+      "learning_rate": 8.093794281764824e-06,
+      "loss": 0.4971,
+      "step": 772
+    },
+    {
+      "epoch": 1.079516391536852,
+      "grad_norm": 2.1850495091404234,
+      "learning_rate": 8.087406402830026e-06,
+      "loss": 0.469,
+      "step": 773
+    },
+    {
+      "epoch": 1.080911415949779,
+      "grad_norm": 2.0322699994558384,
+      "learning_rate": 8.081010368786751e-06,
+      "loss": 0.4778,
+      "step": 774
+    },
+    {
+      "epoch": 1.0823064403627063,
+      "grad_norm": 1.9091269108666133,
+      "learning_rate": 8.074606196529554e-06,
+      "loss": 0.4675,
+      "step": 775
+    },
+    {
+      "epoch": 1.0837014647756336,
+      "grad_norm": 2.035468646541553,
+      "learning_rate": 8.068193902974482e-06,
+      "loss": 0.4726,
+      "step": 776
+    },
+    {
+      "epoch": 1.0850964891885608,
+      "grad_norm": 2.052286438982202,
+      "learning_rate": 8.06177350505904e-06,
+      "loss": 0.437,
+      "step": 777
+    },
+    {
+      "epoch": 1.086491513601488,
+      "grad_norm": 2.186255322231635,
+      "learning_rate": 8.055345019742133e-06,
+      "loss": 0.4578,
+      "step": 778
+    },
+    {
+      "epoch": 1.0878865380144151,
+      "grad_norm": 2.0869198110476237,
+      "learning_rate": 8.048908464004032e-06,
+      "loss": 0.5158,
+      "step": 779
+    },
+    {
+      "epoch": 1.0892815624273424,
+      "grad_norm": 2.2422828894640374,
+      "learning_rate": 8.042463854846325e-06,
+      "loss": 0.4455,
+      "step": 780
+    },
+    {
+      "epoch": 1.0906765868402697,
+      "grad_norm": 2.020266529668748,
+      "learning_rate": 8.036011209291872e-06,
+      "loss": 0.4682,
+      "step": 781
+    },
+    {
+      "epoch": 1.092071611253197,
+      "grad_norm": 2.1762604344777174,
+      "learning_rate": 8.029550544384758e-06,
+      "loss": 0.4544,
+      "step": 782
+    },
+    {
+      "epoch": 1.0934666356661242,
+      "grad_norm": 2.075854485507843,
+      "learning_rate": 8.023081877190257e-06,
+      "loss": 0.5303,
+      "step": 783
+    },
+    {
+      "epoch": 1.0948616600790513,
+      "grad_norm": 2.1957708758825385,
+      "learning_rate": 8.016605224794773e-06,
+      "loss": 0.4799,
+      "step": 784
+    },
+    {
+      "epoch": 1.0962566844919786,
+      "grad_norm": 2.047398963820383,
+      "learning_rate": 8.010120604305806e-06,
+      "loss": 0.465,
+      "step": 785
+    },
+    {
+      "epoch": 1.0976517089049058,
+      "grad_norm": 2.1167412676353714,
+      "learning_rate": 8.003628032851904e-06,
+      "loss": 0.4773,
+      "step": 786
+    },
+    {
+      "epoch": 1.099046733317833,
+      "grad_norm": 2.05543114145942,
+      "learning_rate": 7.997127527582613e-06,
+      "loss": 0.4638,
+      "step": 787
+    },
+    {
+      "epoch": 1.1004417577307604,
+      "grad_norm": 2.5883330126395183,
+      "learning_rate": 7.990619105668437e-06,
+      "loss": 0.4429,
+      "step": 788
+    },
+    {
+      "epoch": 1.1018367821436876,
+      "grad_norm": 1.7471444573255699,
+      "learning_rate": 7.984102784300794e-06,
+      "loss": 0.4823,
+      "step": 789
+    },
+    {
+      "epoch": 1.1032318065566147,
+      "grad_norm": 2.0341688631816552,
+      "learning_rate": 7.977578580691963e-06,
+      "loss": 0.4912,
+      "step": 790
+    },
+    {
+      "epoch": 1.104626830969542,
+      "grad_norm": 1.9309193576466954,
+      "learning_rate": 7.971046512075047e-06,
+      "loss": 0.4465,
+      "step": 791
+    },
+    {
+      "epoch": 1.1060218553824692,
+      "grad_norm": 2.073755962027876,
+      "learning_rate": 7.964506595703923e-06,
+      "loss": 0.456,
+      "step": 792
+    },
+    {
+      "epoch": 1.1074168797953965,
+      "grad_norm": 1.9513764017619648,
+      "learning_rate": 7.957958848853193e-06,
+      "loss": 0.5207,
+      "step": 793
+    },
+    {
+      "epoch": 1.1088119042083235,
+      "grad_norm": 2.210555465361039,
+      "learning_rate": 7.95140328881815e-06,
+      "loss": 0.5029,
+      "step": 794
+    },
+    {
+      "epoch": 1.1102069286212508,
+      "grad_norm": 2.0145979377685475,
+      "learning_rate": 7.944839932914718e-06,
+      "loss": 0.4881,
+      "step": 795
+    },
+    {
+      "epoch": 1.111601953034178,
+      "grad_norm": 2.1886552216768784,
+      "learning_rate": 7.938268798479419e-06,
+      "loss": 0.4801,
+      "step": 796
+    },
+    {
+      "epoch": 1.1129969774471054,
+      "grad_norm": 1.9143808865966454,
+      "learning_rate": 7.931689902869314e-06,
+      "loss": 0.4987,
+      "step": 797
+    },
+    {
+      "epoch": 1.1143920018600326,
+      "grad_norm": 2.229623106539157,
+      "learning_rate": 7.925103263461971e-06,
+      "loss": 0.4978,
+      "step": 798
+    },
+    {
+      "epoch": 1.11578702627296,
+      "grad_norm": 1.9414732070488792,
+      "learning_rate": 7.91850889765541e-06,
+      "loss": 0.458,
+      "step": 799
+    },
+    {
+      "epoch": 1.117182050685887,
+      "grad_norm": 2.092745745628236,
+      "learning_rate": 7.91190682286806e-06,
+      "loss": 0.4389,
+      "step": 800
+    },
+    {
+      "epoch": 1.1185770750988142,
+      "grad_norm": 2.123397220956053,
+      "learning_rate": 7.905297056538713e-06,
+      "loss": 0.4635,
+      "step": 801
+    },
+    {
+      "epoch": 1.1199720995117415,
+      "grad_norm": 1.968518140063321,
+      "learning_rate": 7.898679616126474e-06,
+      "loss": 0.4799,
+      "step": 802
+    },
+    {
+      "epoch": 1.1213671239246688,
+      "grad_norm": 1.9781979017384825,
+      "learning_rate": 7.892054519110726e-06,
+      "loss": 0.4876,
+      "step": 803
+    },
+    {
+      "epoch": 1.1227621483375958,
+      "grad_norm": 2.1208452802156725,
+      "learning_rate": 7.885421782991064e-06,
+      "loss": 0.5135,
+      "step": 804
+    },
+    {
+      "epoch": 1.124157172750523,
+      "grad_norm": 2.0090102765020954,
+      "learning_rate": 7.878781425287277e-06,
+      "loss": 0.4445,
+      "step": 805
+    },
+    {
+      "epoch": 1.1255521971634503,
+      "grad_norm": 1.9927344582343813,
+      "learning_rate": 7.872133463539274e-06,
+      "loss": 0.4357,
+      "step": 806
+    },
+    {
+      "epoch": 1.1269472215763776,
+      "grad_norm": 2.0716300721802665,
+      "learning_rate": 7.86547791530705e-06,
+      "loss": 0.5389,
+      "step": 807
+    },
+    {
+      "epoch": 1.1283422459893049,
+      "grad_norm": 1.9819888932135203,
+      "learning_rate": 7.858814798170644e-06,
+      "loss": 0.492,
+      "step": 808
+    },
+    {
+      "epoch": 1.1297372704022322,
+      "grad_norm": 2.0463707519984218,
+      "learning_rate": 7.852144129730087e-06,
+      "loss": 0.4736,
+      "step": 809
+    },
+    {
+      "epoch": 1.1311322948151592,
+      "grad_norm": 2.1975286753620877,
+      "learning_rate": 7.84546592760535e-06,
+      "loss": 0.4879,
+      "step": 810
+    },
+    {
+      "epoch": 1.1325273192280865,
+      "grad_norm": 1.911921932483923,
+      "learning_rate": 7.83878020943631e-06,
+      "loss": 0.5166,
+      "step": 811
+    },
+    {
+      "epoch": 1.1339223436410137,
+      "grad_norm": 2.1256753078627764,
+      "learning_rate": 7.832086992882697e-06,
+      "loss": 0.4841,
+      "step": 812
+    },
+    {
+      "epoch": 1.135317368053941,
+      "grad_norm": 1.8043220542453693,
+      "learning_rate": 7.825386295624043e-06,
+      "loss": 0.4479,
+      "step": 813
+    },
+    {
+      "epoch": 1.136712392466868,
+      "grad_norm": 2.027151075919505,
+      "learning_rate": 7.818678135359641e-06,
+      "loss": 0.4979,
+      "step": 814
+    },
+    {
+      "epoch": 1.1381074168797953,
+      "grad_norm": 2.0578677546704607,
+      "learning_rate": 7.811962529808499e-06,
+      "loss": 0.4757,
+      "step": 815
+    },
+    {
+      "epoch": 1.1395024412927226,
+      "grad_norm": 2.0604192529558656,
+      "learning_rate": 7.805239496709291e-06,
+      "loss": 0.4531,
+      "step": 816
+    },
+    {
+      "epoch": 1.1408974657056499,
+      "grad_norm": 1.94855850047,
+      "learning_rate": 7.798509053820305e-06,
+      "loss": 0.4417,
+      "step": 817
+    },
+    {
+      "epoch": 1.1422924901185771,
+      "grad_norm": 2.023024363909115,
+      "learning_rate": 7.79177121891941e-06,
+      "loss": 0.4615,
+      "step": 818
+    },
+    {
+      "epoch": 1.1436875145315044,
+      "grad_norm": 2.106362451480915,
+      "learning_rate": 7.785026009803993e-06,
+      "loss": 0.503,
+      "step": 819
+    },
+    {
+      "epoch": 1.1450825389444315,
+      "grad_norm": 2.270663758773595,
+      "learning_rate": 7.778273444290921e-06,
+      "loss": 0.4791,
+      "step": 820
+    },
+    {
+      "epoch": 1.1464775633573587,
+      "grad_norm": 2.1367829318763647,
+      "learning_rate": 7.771513540216496e-06,
+      "loss": 0.4937,
+      "step": 821
+    },
+    {
+      "epoch": 1.147872587770286,
+      "grad_norm": 2.1680686104730755,
+      "learning_rate": 7.764746315436399e-06,
+      "loss": 0.4543,
+      "step": 822
+    },
+    {
+      "epoch": 1.1492676121832133,
+      "grad_norm": 2.0947572237591454,
+      "learning_rate": 7.75797178782565e-06,
+      "loss": 0.4529,
+      "step": 823
+    },
+    {
+      "epoch": 1.1506626365961403,
+      "grad_norm": 1.7934788611254668,
+      "learning_rate": 7.751189975278561e-06,
+      "loss": 0.4068,
+      "step": 824
+    },
+    {
+      "epoch": 1.1520576610090676,
+      "grad_norm": 2.0307708711760895,
+      "learning_rate": 7.744400895708683e-06,
+      "loss": 0.4659,
+      "step": 825
+    },
+    {
+      "epoch": 1.1534526854219949,
+      "grad_norm": 2.061646896391218,
+      "learning_rate": 7.737604567048766e-06,
+      "loss": 0.5493,
+      "step": 826
+    },
+    {
+      "epoch": 1.1548477098349221,
+      "grad_norm": 2.193247299615853,
+      "learning_rate": 7.730801007250704e-06,
+      "loss": 0.4671,
+      "step": 827
+    },
+    {
+      "epoch": 1.1562427342478494,
+      "grad_norm": 2.033168768238963,
+      "learning_rate": 7.72399023428549e-06,
+      "loss": 0.4642,
+      "step": 828
+    },
+    {
+      "epoch": 1.1576377586607767,
+      "grad_norm": 1.921507053967542,
+      "learning_rate": 7.717172266143178e-06,
+      "loss": 0.5365,
+      "step": 829
+    },
+    {
+      "epoch": 1.1590327830737037,
+      "grad_norm": 2.1786658189049835,
+      "learning_rate": 7.710347120832821e-06,
+      "loss": 0.4573,
+      "step": 830
+    },
+    {
+      "epoch": 1.160427807486631,
+      "grad_norm": 2.04271905830797,
+      "learning_rate": 7.703514816382432e-06,
+      "loss": 0.4397,
+      "step": 831
+    },
+    {
+      "epoch": 1.1618228318995583,
+      "grad_norm": 2.010020828781065,
+      "learning_rate": 7.696675370838929e-06,
+      "loss": 0.454,
+      "step": 832
+    },
+    {
+      "epoch": 1.1632178563124855,
+      "grad_norm": 2.0652609343361292,
+      "learning_rate": 7.689828802268102e-06,
+      "loss": 0.4565,
+      "step": 833
+    },
+    {
+      "epoch": 1.1646128807254126,
+      "grad_norm": 1.8081504823167087,
+      "learning_rate": 7.682975128754548e-06,
+      "loss": 0.4496,
+      "step": 834
+    },
+    {
+      "epoch": 1.1660079051383399,
+      "grad_norm": 2.073580143759261,
+      "learning_rate": 7.676114368401635e-06,
+      "loss": 0.5335,
+      "step": 835
+    },
+    {
+      "epoch": 1.1674029295512671,
+      "grad_norm": 2.146815116325766,
+      "learning_rate": 7.66924653933145e-06,
+      "loss": 0.4867,
+      "step": 836
+    },
+    {
+      "epoch": 1.1687979539641944,
+      "grad_norm": 2.0789485213664145,
+      "learning_rate": 7.662371659684749e-06,
+      "loss": 0.5105,
+      "step": 837
+    },
+    {
+      "epoch": 1.1701929783771217,
+      "grad_norm": 2.0786101027949915,
+      "learning_rate": 7.655489747620913e-06,
+      "loss": 0.4726,
+      "step": 838
+    },
+    {
+      "epoch": 1.171588002790049,
+      "grad_norm": 2.021038432128547,
+      "learning_rate": 7.648600821317901e-06,
+      "loss": 0.4993,
+      "step": 839
+    },
+    {
+      "epoch": 1.172983027202976,
+      "grad_norm": 2.1039185200269848,
+      "learning_rate": 7.641704898972194e-06,
+      "loss": 0.4922,
+      "step": 840
+    },
+    {
+      "epoch": 1.1743780516159033,
+      "grad_norm": 2.1495345339138328,
+      "learning_rate": 7.634801998798755e-06,
+      "loss": 0.5102,
+      "step": 841
+    },
+    {
+      "epoch": 1.1757730760288305,
+      "grad_norm": 2.0719811459725057,
+      "learning_rate": 7.6278921390309834e-06,
+      "loss": 0.4355,
+      "step": 842
+    },
+    {
+      "epoch": 1.1771681004417578,
+      "grad_norm": 1.8630425662831487,
+      "learning_rate": 7.620975337920653e-06,
+      "loss": 0.4703,
+      "step": 843
+    },
+    {
+      "epoch": 1.1785631248546848,
+      "grad_norm": 2.046791765467504,
+      "learning_rate": 7.6140516137378786e-06,
+      "loss": 0.4223,
+      "step": 844
+    },
+    {
+      "epoch": 1.179958149267612,
+      "grad_norm": 1.9132364567899665,
+      "learning_rate": 7.607120984771058e-06,
+      "loss": 0.4596,
+      "step": 845
+    },
+    {
+      "epoch": 1.1813531736805394,
+      "grad_norm": 1.9802117029437554,
+      "learning_rate": 7.600183469326829e-06,
+      "loss": 0.4659,
+      "step": 846
+    },
+    {
+      "epoch": 1.1827481980934667,
+      "grad_norm": 2.025431606531357,
+      "learning_rate": 7.593239085730022e-06,
+      "loss": 0.5016,
+      "step": 847
+    },
+    {
+      "epoch": 1.184143222506394,
+      "grad_norm": 2.368669248836562,
+      "learning_rate": 7.586287852323605e-06,
+      "loss": 0.5117,
+      "step": 848
+    },
+    {
+      "epoch": 1.1855382469193212,
+      "grad_norm": 2.100026020486569,
+      "learning_rate": 7.579329787468639e-06,
+      "loss": 0.4535,
+      "step": 849
+    },
+    {
+      "epoch": 1.1869332713322482,
+      "grad_norm": 2.0663388575312056,
+      "learning_rate": 7.572364909544235e-06,
+      "loss": 0.4993,
+      "step": 850
+    },
+    {
+      "epoch": 1.1883282957451755,
+      "grad_norm": 1.884056547062916,
+      "learning_rate": 7.565393236947494e-06,
+      "loss": 0.4346,
+      "step": 851
+    },
+    {
+      "epoch": 1.1897233201581028,
+      "grad_norm": 2.1776882352748728,
+      "learning_rate": 7.558414788093467e-06,
+      "loss": 0.4755,
+      "step": 852
+    },
+    {
+      "epoch": 1.19111834457103,
+      "grad_norm": 2.061273986301572,
+      "learning_rate": 7.551429581415104e-06,
+      "loss": 0.5118,
+      "step": 853
+    },
+    {
+      "epoch": 1.192513368983957,
+      "grad_norm": 2.1865289893443527,
+      "learning_rate": 7.5444376353632064e-06,
+      "loss": 0.4577,
+      "step": 854
+    },
+    {
+      "epoch": 1.1939083933968844,
+      "grad_norm": 1.8247742290026874,
+      "learning_rate": 7.537438968406372e-06,
+      "loss": 0.4904,
+      "step": 855
+    },
+    {
+      "epoch": 1.1953034178098116,
+      "grad_norm": 2.1078079523273887,
+      "learning_rate": 7.530433599030962e-06,
+      "loss": 0.4653,
+      "step": 856
+    },
+    {
+      "epoch": 1.196698442222739,
+      "grad_norm": 2.1391503263974045,
+      "learning_rate": 7.5234215457410255e-06,
+      "loss": 0.5067,
+      "step": 857
+    },
+    {
+      "epoch": 1.1980934666356662,
+      "grad_norm": 2.2368165670079354,
+      "learning_rate": 7.516402827058283e-06,
+      "loss": 0.5353,
+      "step": 858
+    },
+    {
+      "epoch": 1.1994884910485935,
+      "grad_norm": 1.931059913532765,
+      "learning_rate": 7.509377461522049e-06,
+      "loss": 0.4786,
+      "step": 859
+    },
+    {
+      "epoch": 1.2008835154615205,
+      "grad_norm": 2.330043518759979,
+      "learning_rate": 7.502345467689202e-06,
+      "loss": 0.5049,
+      "step": 860
+    },
+    {
+      "epoch": 1.2022785398744478,
+      "grad_norm": 2.1386669121489437,
+      "learning_rate": 7.4953068641341255e-06,
+      "loss": 0.4504,
+      "step": 861
+    },
+    {
+      "epoch": 1.203673564287375,
+      "grad_norm": 2.1135532365263847,
+      "learning_rate": 7.488261669448662e-06,
+      "loss": 0.5204,
+      "step": 862
+    },
+    {
+      "epoch": 1.2050685887003023,
+      "grad_norm": 2.0651469477865816,
+      "learning_rate": 7.4812099022420636e-06,
+      "loss": 0.4627,
+      "step": 863
+    },
+    {
+      "epoch": 1.2064636131132294,
+      "grad_norm": 2.034159070644421,
+      "learning_rate": 7.474151581140947e-06,
+      "loss": 0.4753,
+      "step": 864
+    },
+    {
+      "epoch": 1.2078586375261566,
+      "grad_norm": 2.145642949505401,
+      "learning_rate": 7.4670867247892346e-06,
+      "loss": 0.4714,
+      "step": 865
+    },
+    {
+      "epoch": 1.209253661939084,
+      "grad_norm": 1.8618981108225734,
+      "learning_rate": 7.460015351848115e-06,
+      "loss": 0.4818,
+      "step": 866
+    },
+    {
+      "epoch": 1.2106486863520112,
+      "grad_norm": 2.0821388794136215,
+      "learning_rate": 7.4529374809959895e-06,
+      "loss": 0.4179,
+      "step": 867
+    },
+    {
+      "epoch": 1.2120437107649384,
+      "grad_norm": 1.944351950216618,
+      "learning_rate": 7.445853130928422e-06,
+      "loss": 0.5009,
+      "step": 868
+    },
+    {
+      "epoch": 1.2134387351778657,
+      "grad_norm": 2.2608731884977824,
+      "learning_rate": 7.438762320358089e-06,
+      "loss": 0.5172,
+      "step": 869
+    },
+    {
+      "epoch": 1.2148337595907928,
+      "grad_norm": 2.046899982267453,
+      "learning_rate": 7.431665068014737e-06,
+      "loss": 0.4701,
+      "step": 870
+    },
+    {
+      "epoch": 1.21622878400372,
+      "grad_norm": 1.9959592229395409,
+      "learning_rate": 7.424561392645122e-06,
+      "loss": 0.4717,
+      "step": 871
+    },
+    {
+      "epoch": 1.2176238084166473,
+      "grad_norm": 2.0838483616097827,
+      "learning_rate": 7.417451313012971e-06,
+      "loss": 0.49,
+      "step": 872
+    },
+    {
+      "epoch": 1.2190188328295746,
+      "grad_norm": 2.19031892792901,
+      "learning_rate": 7.410334847898921e-06,
+      "loss": 0.5118,
+      "step": 873
+    },
+    {
+      "epoch": 1.2204138572425016,
+      "grad_norm": 2.0327900333900923,
+      "learning_rate": 7.403212016100484e-06,
+      "loss": 0.4625,
+      "step": 874
+    },
+    {
+      "epoch": 1.221808881655429,
+      "grad_norm": 2.0196003750349183,
+      "learning_rate": 7.396082836431981e-06,
+      "loss": 0.4624,
+      "step": 875
+    },
+    {
+      "epoch": 1.2232039060683562,
+      "grad_norm": 2.043291413010201,
+      "learning_rate": 7.388947327724506e-06,
+      "loss": 0.4511,
+      "step": 876
+    },
+    {
+      "epoch": 1.2245989304812834,
+      "grad_norm": 2.0166904136766854,
+      "learning_rate": 7.3818055088258676e-06,
+      "loss": 0.4653,
+      "step": 877
+    },
+    {
+      "epoch": 1.2259939548942107,
+      "grad_norm": 2.172035909388715,
+      "learning_rate": 7.374657398600542e-06,
+      "loss": 0.4502,
+      "step": 878
+    },
+    {
+      "epoch": 1.227388979307138,
+      "grad_norm": 2.0085633602200015,
+      "learning_rate": 7.367503015929627e-06,
+      "loss": 0.4665,
+      "step": 879
+    },
+    {
+      "epoch": 1.228784003720065,
+      "grad_norm": 2.0313790988567075,
+      "learning_rate": 7.3603423797107845e-06,
+      "loss": 0.4512,
+      "step": 880
+    },
+    {
+      "epoch": 1.2301790281329923,
+      "grad_norm": 2.1088602216689663,
+      "learning_rate": 7.353175508858195e-06,
+      "loss": 0.4963,
+      "step": 881
+    },
+    {
+      "epoch": 1.2315740525459196,
+      "grad_norm": 2.222624875679751,
+      "learning_rate": 7.3460024223025095e-06,
+      "loss": 0.509,
+      "step": 882
+    },
+    {
+      "epoch": 1.2329690769588468,
+      "grad_norm": 1.678805125515682,
+      "learning_rate": 7.338823138990796e-06,
+      "loss": 0.4363,
+      "step": 883
+    },
+    {
+      "epoch": 1.234364101371774,
+      "grad_norm": 2.0825313015032143,
+      "learning_rate": 7.33163767788649e-06,
+      "loss": 0.4606,
+      "step": 884
+    },
+    {
+      "epoch": 1.2357591257847012,
+      "grad_norm": 2.008078631188548,
+      "learning_rate": 7.324446057969346e-06,
+      "loss": 0.5088,
+      "step": 885
+    },
+    {
+      "epoch": 1.2371541501976284,
+      "grad_norm": 2.120130042894298,
+      "learning_rate": 7.317248298235387e-06,
+      "loss": 0.4908,
+      "step": 886
+    },
+    {
+      "epoch": 1.2385491746105557,
+      "grad_norm": 2.026566890552797,
+      "learning_rate": 7.3100444176968514e-06,
+      "loss": 0.4453,
+      "step": 887
+    },
+    {
+      "epoch": 1.239944199023483,
+      "grad_norm": 2.16601301998483,
+      "learning_rate": 7.302834435382147e-06,
+      "loss": 0.5485,
+      "step": 888
+    },
+    {
+      "epoch": 1.2413392234364102,
+      "grad_norm": 2.140553776744054,
+      "learning_rate": 7.2956183703358e-06,
+      "loss": 0.4321,
+      "step": 889
+    },
+    {
+      "epoch": 1.2427342478493373,
+      "grad_norm": 1.9128745400529434,
+      "learning_rate": 7.288396241618401e-06,
+      "loss": 0.4705,
+      "step": 890
+    },
+    {
+      "epoch": 1.2441292722622646,
+      "grad_norm": 2.1228545685375377,
+      "learning_rate": 7.281168068306559e-06,
+      "loss": 0.5109,
+      "step": 891
+    },
+    {
+      "epoch": 1.2455242966751918,
+      "grad_norm": 1.9958676358218619,
+      "learning_rate": 7.2739338694928485e-06,
+      "loss": 0.4264,
+      "step": 892
+    },
+    {
+      "epoch": 1.246919321088119,
+      "grad_norm": 2.017602542711396,
+      "learning_rate": 7.266693664285761e-06,
+      "loss": 0.4858,
+      "step": 893
+    },
+    {
+      "epoch": 1.2483143455010464,
+      "grad_norm": 2.1655539794588843,
+      "learning_rate": 7.259447471809651e-06,
+      "loss": 0.4718,
+      "step": 894
+    },
+    {
+      "epoch": 1.2497093699139734,
+      "grad_norm": 1.8961156639730357,
+      "learning_rate": 7.252195311204689e-06,
+      "loss": 0.4976,
+      "step": 895
+    },
+    {
+      "epoch": 1.2511043943269007,
+      "grad_norm": 2.2453477317683594,
+      "learning_rate": 7.244937201626812e-06,
+      "loss": 0.5222,
+      "step": 896
+    },
+    {
+      "epoch": 1.252499418739828,
+      "grad_norm": 2.127532832546729,
+      "learning_rate": 7.237673162247667e-06,
+      "loss": 0.5229,
+      "step": 897
+    },
+    {
+      "epoch": 1.2538944431527552,
+      "grad_norm": 2.1894133849734203,
+      "learning_rate": 7.230403212254566e-06,
+      "loss": 0.4418,
+      "step": 898
+    },
+    {
+      "epoch": 1.2552894675656825,
+      "grad_norm": 2.0080709918299755,
+      "learning_rate": 7.223127370850433e-06,
+      "loss": 0.4867,
+      "step": 899
+    },
+    {
+      "epoch": 1.2566844919786098,
+      "grad_norm": 2.08931797506218,
+      "learning_rate": 7.215845657253755e-06,
+      "loss": 0.4997,
+      "step": 900
+    },
+    {
+      "epoch": 1.2580795163915368,
+      "grad_norm": 2.119270394002083,
+      "learning_rate": 7.208558090698528e-06,
+      "loss": 0.4672,
+      "step": 901
+    },
+    {
+      "epoch": 1.259474540804464,
+      "grad_norm": 2.0460383698950912,
+      "learning_rate": 7.2012646904342065e-06,
+      "loss": 0.5123,
+      "step": 902
+    },
+    {
+      "epoch": 1.2608695652173914,
+      "grad_norm": 2.1384976321501177,
+      "learning_rate": 7.193965475725659e-06,
+      "loss": 0.4686,
+      "step": 903
+    },
+    {
+      "epoch": 1.2622645896303184,
+      "grad_norm": 1.9220601194054647,
+      "learning_rate": 7.186660465853111e-06,
+      "loss": 0.4803,
+      "step": 904
+    },
+    {
+      "epoch": 1.2636596140432457,
+      "grad_norm": 2.0262630040398917,
+      "learning_rate": 7.1793496801120885e-06,
+      "loss": 0.472,
+      "step": 905
+    },
+    {
+      "epoch": 1.265054638456173,
+      "grad_norm": 2.0585200478781482,
+      "learning_rate": 7.172033137813387e-06,
+      "loss": 0.4805,
+      "step": 906
+    },
+    {
+      "epoch": 1.2664496628691002,
+      "grad_norm": 2.2127511883039315,
+      "learning_rate": 7.1647108582829924e-06,
+      "loss": 0.5139,
+      "step": 907
+    },
+    {
+      "epoch": 1.2678446872820275,
+      "grad_norm": 2.1318837737050584,
+      "learning_rate": 7.157382860862059e-06,
+      "loss": 0.4948,
+      "step": 908
+    },
+    {
+      "epoch": 1.2692397116949548,
+      "grad_norm": 2.0404146831089727,
+      "learning_rate": 7.1500491649068345e-06,
+      "loss": 0.4759,
+      "step": 909
+    },
+    {
+      "epoch": 1.270634736107882,
+      "grad_norm": 2.16885361299071,
+      "learning_rate": 7.1427097897886225e-06,
+      "loss": 0.4651,
+      "step": 910
+    },
+    {
+      "epoch": 1.272029760520809,
+      "grad_norm": 1.8126562233953132,
+      "learning_rate": 7.135364754893729e-06,
+      "loss": 0.4827,
+      "step": 911
+    },
+    {
+      "epoch": 1.2734247849337363,
+      "grad_norm": 2.037530800408793,
+      "learning_rate": 7.128014079623408e-06,
+      "loss": 0.471,
+      "step": 912
+    },
+    {
+      "epoch": 1.2748198093466636,
+      "grad_norm": 2.113312788616271,
+      "learning_rate": 7.120657783393809e-06,
+      "loss": 0.4536,
+      "step": 913
+    },
+    {
+      "epoch": 1.2762148337595907,
+      "grad_norm": 1.9683652942609144,
+      "learning_rate": 7.113295885635936e-06,
+      "loss": 0.4603,
+      "step": 914
+    },
+    {
+      "epoch": 1.277609858172518,
+      "grad_norm": 2.0398678179326266,
+      "learning_rate": 7.105928405795584e-06,
+      "loss": 0.53,
+      "step": 915
+    },
+    {
+      "epoch": 1.2790048825854452,
+      "grad_norm": 2.0550203014189155,
+      "learning_rate": 7.098555363333289e-06,
+      "loss": 0.4635,
+      "step": 916
+    },
+    {
+      "epoch": 1.2803999069983725,
+      "grad_norm": 2.0420956994823833,
+      "learning_rate": 7.091176777724291e-06,
+      "loss": 0.4923,
+      "step": 917
+    },
+    {
+      "epoch": 1.2817949314112997,
+      "grad_norm": 2.122118367593062,
+      "learning_rate": 7.083792668458463e-06,
+      "loss": 0.4572,
+      "step": 918
+    },
+    {
+      "epoch": 1.283189955824227,
+      "grad_norm": 1.9605764589538413,
+      "learning_rate": 7.076403055040271e-06,
+      "loss": 0.4613,
+      "step": 919
+    },
+    {
+      "epoch": 1.2845849802371543,
+      "grad_norm": 1.994014351784954,
+      "learning_rate": 7.069007956988718e-06,
+      "loss": 0.4917,
+      "step": 920
+    },
+    {
+      "epoch": 1.2859800046500813,
+      "grad_norm": 2.0322178140926526,
+      "learning_rate": 7.061607393837295e-06,
+      "loss": 0.5037,
+      "step": 921
+    },
+    {
+      "epoch": 1.2873750290630086,
+      "grad_norm": 2.0501993691106892,
+      "learning_rate": 7.0542013851339316e-06,
+      "loss": 0.4111,
+      "step": 922
+    },
+    {
+      "epoch": 1.2887700534759359,
+      "grad_norm": 1.9055682007118293,
+      "learning_rate": 7.04678995044094e-06,
+      "loss": 0.4402,
+      "step": 923
+    },
+    {
+      "epoch": 1.290165077888863,
+      "grad_norm": 1.9513266676688439,
+      "learning_rate": 7.039373109334957e-06,
+      "loss": 0.4794,
+      "step": 924
+    },
+    {
+      "epoch": 1.2915601023017902,
+      "grad_norm": 2.0144597802151134,
+      "learning_rate": 7.031950881406913e-06,
+      "loss": 0.4823,
+      "step": 925
+    },
+    {
+      "epoch": 1.2929551267147175,
+      "grad_norm": 2.100982200223343,
+      "learning_rate": 7.024523286261959e-06,
+      "loss": 0.448,
+      "step": 926
+    },
+    {
+      "epoch": 1.2943501511276447,
+      "grad_norm": 1.9455540836634895,
+      "learning_rate": 7.017090343519421e-06,
+      "loss": 0.4824,
+      "step": 927
+    },
+    {
+      "epoch": 1.295745175540572,
+      "grad_norm": 2.1126530723502595,
+      "learning_rate": 7.009652072812758e-06,
+      "loss": 0.4473,
+      "step": 928
+    },
+    {
+      "epoch": 1.2971401999534993,
+      "grad_norm": 1.9516374387293445,
+      "learning_rate": 7.0022084937895e-06,
+      "loss": 0.4934,
+      "step": 929
+    },
+    {
+      "epoch": 1.2985352243664265,
+      "grad_norm": 2.160226598724437,
+      "learning_rate": 6.994759626111189e-06,
+      "loss": 0.5027,
+      "step": 930
+    },
+    {
+      "epoch": 1.2999302487793536,
+      "grad_norm": 2.149165455764662,
+      "learning_rate": 6.987305489453352e-06,
+      "loss": 0.532,
+      "step": 931
+    },
+    {
+      "epoch": 1.3013252731922809,
+      "grad_norm": 2.169198580033365,
+      "learning_rate": 6.979846103505423e-06,
+      "loss": 0.496,
+      "step": 932
+    },
+    {
+      "epoch": 1.3027202976052081,
+      "grad_norm": 1.9709006258399464,
+      "learning_rate": 6.972381487970702e-06,
+      "loss": 0.4476,
+      "step": 933
+    },
+    {
+      "epoch": 1.3041153220181352,
+      "grad_norm": 2.080456063544832,
+      "learning_rate": 6.964911662566309e-06,
+      "loss": 0.517,
+      "step": 934
+    },
+    {
+      "epoch": 1.3055103464310625,
+      "grad_norm": 2.1878295005029407,
+      "learning_rate": 6.957436647023117e-06,
+      "loss": 0.492,
+      "step": 935
+    },
+    {
+      "epoch": 1.3069053708439897,
+      "grad_norm": 1.9473985278522523,
+      "learning_rate": 6.949956461085714e-06,
+      "loss": 0.4913,
+      "step": 936
+    },
+    {
+      "epoch": 1.308300395256917,
+      "grad_norm": 2.085237018377905,
+      "learning_rate": 6.942471124512346e-06,
+      "loss": 0.4729,
+      "step": 937
+    },
+    {
+      "epoch": 1.3096954196698443,
+      "grad_norm": 2.0171640171405705,
+      "learning_rate": 6.934980657074859e-06,
+      "loss": 0.4186,
+      "step": 938
+    },
+    {
+      "epoch": 1.3110904440827715,
+      "grad_norm": 1.9628946990758591,
+      "learning_rate": 6.9274850785586526e-06,
+      "loss": 0.482,
+      "step": 939
+    },
+    {
+      "epoch": 1.3124854684956988,
+      "grad_norm": 1.9230193122254409,
+      "learning_rate": 6.919984408762632e-06,
+      "loss": 0.4612,
+      "step": 940
+    },
+    {
+      "epoch": 1.3138804929086259,
+      "grad_norm": 2.009734075109168,
+      "learning_rate": 6.9124786674991465e-06,
+      "loss": 0.4674,
+      "step": 941
+    },
+    {
+      "epoch": 1.3152755173215531,
+      "grad_norm": 1.9403798007263833,
+      "learning_rate": 6.90496787459394e-06,
+      "loss": 0.4751,
+      "step": 942
+    },
+    {
+      "epoch": 1.3166705417344804,
+      "grad_norm": 2.0729815313209063,
+      "learning_rate": 6.897452049886103e-06,
+      "loss": 0.4748,
+      "step": 943
+    },
+    {
+      "epoch": 1.3180655661474074,
+      "grad_norm": 1.8042882335697186,
+      "learning_rate": 6.889931213228015e-06,
+      "loss": 0.5374,
+      "step": 944
+    },
+    {
+      "epoch": 1.3194605905603347,
+      "grad_norm": 2.1582377635460706,
+      "learning_rate": 6.882405384485294e-06,
+      "loss": 0.4967,
+      "step": 945
+    },
+    {
+      "epoch": 1.320855614973262,
+      "grad_norm": 2.0774574203867053,
+      "learning_rate": 6.874874583536748e-06,
+      "loss": 0.4928,
+      "step": 946
+    },
+    {
+      "epoch": 1.3222506393861893,
+      "grad_norm": 2.1126118556498525,
+      "learning_rate": 6.867338830274312e-06,
+      "loss": 0.5296,
+      "step": 947
+    },
+    {
+      "epoch": 1.3236456637991165,
+      "grad_norm": 2.2327182347799956,
+      "learning_rate": 6.8597981446030095e-06,
+      "loss": 0.5252,
+      "step": 948
+    },
+    {
+      "epoch": 1.3250406882120438,
+      "grad_norm": 2.0860141510662964,
+      "learning_rate": 6.852252546440885e-06,
+      "loss": 0.4903,
+      "step": 949
+    },
+    {
+      "epoch": 1.326435712624971,
+      "grad_norm": 2.044140821899386,
+      "learning_rate": 6.844702055718964e-06,
+      "loss": 0.4587,
+      "step": 950
+    },
+    {
+      "epoch": 1.3278307370378981,
+      "grad_norm": 1.9786800376206937,
+      "learning_rate": 6.837146692381197e-06,
+      "loss": 0.4675,
+      "step": 951
+    },
+    {
+      "epoch": 1.3292257614508254,
+      "grad_norm": 1.9825281201744271,
+      "learning_rate": 6.8295864763843965e-06,
+      "loss": 0.5127,
+      "step": 952
+    },
+    {
+      "epoch": 1.3306207858637527,
+      "grad_norm": 2.0986306954363227,
+      "learning_rate": 6.822021427698201e-06,
+      "loss": 0.4522,
+      "step": 953
+    },
+    {
+      "epoch": 1.33201581027668,
+      "grad_norm": 2.1261538900308747,
+      "learning_rate": 6.814451566305014e-06,
+      "loss": 0.5263,
+      "step": 954
+    },
+    {
+      "epoch": 1.333410834689607,
+      "grad_norm": 2.2448289680883855,
+      "learning_rate": 6.806876912199945e-06,
+      "loss": 0.4962,
+      "step": 955
+    },
+    {
+      "epoch": 1.3348058591025342,
+      "grad_norm": 2.14005220926737,
+      "learning_rate": 6.7992974853907655e-06,
+      "loss": 0.4844,
+      "step": 956
+    },
+    {
+      "epoch": 1.3362008835154615,
+      "grad_norm": 2.094501275219526,
+      "learning_rate": 6.791713305897861e-06,
+      "loss": 0.4773,
+      "step": 957
+    },
+    {
+      "epoch": 1.3375959079283888,
+      "grad_norm": 1.8776892871093087,
+      "learning_rate": 6.78412439375416e-06,
+      "loss": 0.4769,
+      "step": 958
+    },
+    {
+      "epoch": 1.338990932341316,
+      "grad_norm": 2.0257109173056476,
+      "learning_rate": 6.776530769005099e-06,
+      "loss": 0.4574,
+      "step": 959
+    },
+    {
+      "epoch": 1.3403859567542433,
+      "grad_norm": 2.011593652147381,
+      "learning_rate": 6.768932451708557e-06,
+      "loss": 0.4733,
+      "step": 960
+    },
+    {
+      "epoch": 1.3417809811671704,
+      "grad_norm": 2.0880918905336867,
+      "learning_rate": 6.761329461934814e-06,
+      "loss": 0.4717,
+      "step": 961
+    },
+    {
+      "epoch": 1.3431760055800976,
+      "grad_norm": 2.0756463467401467,
+      "learning_rate": 6.753721819766489e-06,
+      "loss": 0.4956,
+      "step": 962
+    },
+    {
+      "epoch": 1.344571029993025,
+      "grad_norm": 2.1951630126685346,
+      "learning_rate": 6.746109545298488e-06,
+      "loss": 0.4821,
+      "step": 963
+    },
+    {
+      "epoch": 1.3459660544059522,
+      "grad_norm": 2.076217947734719,
+      "learning_rate": 6.738492658637957e-06,
+      "loss": 0.4408,
+      "step": 964
+    },
+    {
+      "epoch": 1.3473610788188792,
+      "grad_norm": 1.8012492184873017,
+      "learning_rate": 6.730871179904218e-06,
+      "loss": 0.5316,
+      "step": 965
+    },
+    {
+      "epoch": 1.3487561032318065,
+      "grad_norm": 2.0662235372305093,
+      "learning_rate": 6.723245129228732e-06,
+      "loss": 0.4425,
+      "step": 966
+    },
+    {
+      "epoch": 1.3501511276447338,
+      "grad_norm": 2.1510994544097723,
+      "learning_rate": 6.7156145267550275e-06,
+      "loss": 0.47,
+      "step": 967
+    },
+    {
+      "epoch": 1.351546152057661,
+      "grad_norm": 2.06829815044842,
+      "learning_rate": 6.707979392638663e-06,
+      "loss": 0.4589,
+      "step": 968
+    },
+    {
+      "epoch": 1.3529411764705883,
+      "grad_norm": 2.0138794871206858,
+      "learning_rate": 6.700339747047162e-06,
+      "loss": 0.489,
+      "step": 969
+    },
+    {
+      "epoch": 1.3543362008835156,
+      "grad_norm": 2.041938633849983,
+      "learning_rate": 6.692695610159966e-06,
+      "loss": 0.481,
+      "step": 970
+    },
+    {
+      "epoch": 1.3557312252964426,
+      "grad_norm": 1.9340098627616258,
+      "learning_rate": 6.685047002168382e-06,
+      "loss": 0.4518,
+      "step": 971
+    },
+    {
+      "epoch": 1.35712624970937,
+      "grad_norm": 2.0444945166428288,
+      "learning_rate": 6.677393943275525e-06,
+      "loss": 0.4652,
+      "step": 972
+    },
+    {
+      "epoch": 1.3585212741222972,
+      "grad_norm": 2.0057051447255527,
+      "learning_rate": 6.669736453696266e-06,
+      "loss": 0.4662,
+      "step": 973
+    },
+    {
+      "epoch": 1.3599162985352244,
+      "grad_norm": 2.099453331432414,
+      "learning_rate": 6.66207455365718e-06,
+      "loss": 0.4565,
+      "step": 974
+    },
+    {
+      "epoch": 1.3613113229481515,
+      "grad_norm": 1.9165876489537292,
+      "learning_rate": 6.6544082633964955e-06,
+      "loss": 0.4398,
+      "step": 975
+    },
+    {
+      "epoch": 1.3627063473610788,
+      "grad_norm": 2.135272106746183,
+      "learning_rate": 6.646737603164031e-06,
+      "loss": 0.5094,
+      "step": 976
+    },
+    {
+      "epoch": 1.364101371774006,
+      "grad_norm": 2.139613224778207,
+      "learning_rate": 6.639062593221152e-06,
+      "loss": 0.5341,
+      "step": 977
+    },
+    {
+      "epoch": 1.3654963961869333,
+      "grad_norm": 2.530233359548065,
+      "learning_rate": 6.6313832538407106e-06,
+      "loss": 0.4969,
+      "step": 978
+    },
+    {
+      "epoch": 1.3668914205998606,
+      "grad_norm": 1.974541177446713,
+      "learning_rate": 6.623699605306999e-06,
+      "loss": 0.4917,
+      "step": 979
+    },
+    {
+      "epoch": 1.3682864450127878,
+      "grad_norm": 2.1328867084448313,
+      "learning_rate": 6.6160116679156874e-06,
+      "loss": 0.4667,
+      "step": 980
+    },
+    {
+      "epoch": 1.369681469425715,
+      "grad_norm": 2.217976729899946,
+      "learning_rate": 6.608319461973778e-06,
+      "loss": 0.4602,
+      "step": 981
+    },
+    {
+      "epoch": 1.3710764938386422,
+      "grad_norm": 2.0033070485546705,
+      "learning_rate": 6.6006230077995424e-06,
+      "loss": 0.4614,
+      "step": 982
+    },
+    {
+      "epoch": 1.3724715182515694,
+      "grad_norm": 1.7949480230161823,
+      "learning_rate": 6.592922325722483e-06,
+      "loss": 0.4625,
+      "step": 983
+    },
+    {
+      "epoch": 1.3738665426644967,
+      "grad_norm": 2.3605371814580987,
+      "learning_rate": 6.58521743608326e-06,
+      "loss": 0.5013,
+      "step": 984
+    },
+    {
+      "epoch": 1.3752615670774238,
+      "grad_norm": 2.079920357785801,
+      "learning_rate": 6.577508359233653e-06,
+      "loss": 0.4887,
+      "step": 985
+    },
+    {
+      "epoch": 1.376656591490351,
+      "grad_norm": 2.168691229695169,
+      "learning_rate": 6.569795115536502e-06,
+      "loss": 0.4736,
+      "step": 986
+    },
+    {
+      "epoch": 1.3780516159032783,
+      "grad_norm": 2.041653674751869,
+      "learning_rate": 6.562077725365648e-06,
+      "loss": 0.5144,
+      "step": 987
+    },
+    {
+      "epoch": 1.3794466403162056,
+      "grad_norm": 2.018405822849198,
+      "learning_rate": 6.554356209105892e-06,
+      "loss": 0.4193,
+      "step": 988
+    },
+    {
+      "epoch": 1.3808416647291328,
+      "grad_norm": 2.0791572967119665,
+      "learning_rate": 6.54663058715293e-06,
+      "loss": 0.5241,
+      "step": 989
+    },
+    {
+      "epoch": 1.38223668914206,
+      "grad_norm": 2.0829602943845797,
+      "learning_rate": 6.538900879913301e-06,
+      "loss": 0.4861,
+      "step": 990
+    },
+    {
+      "epoch": 1.3836317135549872,
+      "grad_norm": 2.1547428115463094,
+      "learning_rate": 6.531167107804337e-06,
+      "loss": 0.4959,
+      "step": 991
+    },
+    {
+      "epoch": 1.3850267379679144,
+      "grad_norm": 2.0439903006152833,
+      "learning_rate": 6.523429291254109e-06,
+      "loss": 0.5031,
+      "step": 992
+    },
+    {
+      "epoch": 1.3864217623808417,
+      "grad_norm": 2.1425187821342435,
+      "learning_rate": 6.515687450701367e-06,
+      "loss": 0.476,
+      "step": 993
+    },
+    {
+      "epoch": 1.387816786793769,
+      "grad_norm": 2.043901736829992,
+      "learning_rate": 6.507941606595492e-06,
+      "loss": 0.4368,
+      "step": 994
+    },
+    {
+      "epoch": 1.389211811206696,
+      "grad_norm": 1.9144888000668951,
+      "learning_rate": 6.500191779396439e-06,
+      "loss": 0.4878,
+      "step": 995
+    },
+    {
+      "epoch": 1.3906068356196233,
+      "grad_norm": 2.136038237292774,
+      "learning_rate": 6.492437989574689e-06,
+      "loss": 0.4824,
+      "step": 996
+    },
+    {
+      "epoch": 1.3920018600325506,
+      "grad_norm": 2.123580451737323,
+      "learning_rate": 6.48468025761118e-06,
+      "loss": 0.4832,
+      "step": 997
+    },
+    {
+      "epoch": 1.3933968844454778,
+      "grad_norm": 2.030731921604436,
+      "learning_rate": 6.476918603997273e-06,
+      "loss": 0.4543,
+      "step": 998
+    },
+    {
+      "epoch": 1.394791908858405,
+      "grad_norm": 2.33185642715308,
+      "learning_rate": 6.469153049234683e-06,
+      "loss": 0.5085,
+      "step": 999
+    },
+    {
+      "epoch": 1.3961869332713324,
+      "grad_norm": 2.1984413784598393,
+      "learning_rate": 6.461383613835427e-06,
+      "loss": 0.4935,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3975819576842594,
+      "grad_norm": 1.9194445089831693,
+      "learning_rate": 6.453610318321777e-06,
+      "loss": 0.4486,
+      "step": 1001
+    },
+    {
+      "epoch": 1.3989769820971867,
+      "grad_norm": 1.9845065956042707,
+      "learning_rate": 6.445833183226201e-06,
+      "loss": 0.5038,
+      "step": 1002
+    },
+    {
+      "epoch": 1.400372006510114,
+      "grad_norm": 2.261131914609956,
+      "learning_rate": 6.438052229091303e-06,
+      "loss": 0.4836,
+      "step": 1003
+    },
+    {
+      "epoch": 1.4017670309230412,
+      "grad_norm": 2.1179514773362595,
+      "learning_rate": 6.430267476469783e-06,
+      "loss": 0.4887,
+      "step": 1004
+    },
+    {
+      "epoch": 1.4031620553359683,
+      "grad_norm": 2.0705617150447955,
+      "learning_rate": 6.4224789459243705e-06,
+      "loss": 0.485,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4045570797488955,
+      "grad_norm": 2.0525161666844114,
+      "learning_rate": 6.4146866580277686e-06,
+      "loss": 0.4524,
+      "step": 1006
+    },
+    {
+      "epoch": 1.4059521041618228,
+      "grad_norm": 2.009720205709189,
+      "learning_rate": 6.406890633362618e-06,
+      "loss": 0.4404,
+      "step": 1007
+    },
+    {
+      "epoch": 1.40734712857475,
+      "grad_norm": 1.9015907571149264,
+      "learning_rate": 6.3990908925214155e-06,
+      "loss": 0.4251,
+      "step": 1008
+    },
+    {
+      "epoch": 1.4087421529876774,
+      "grad_norm": 1.886712184163379,
+      "learning_rate": 6.391287456106483e-06,
+      "loss": 0.4772,
+      "step": 1009
+    },
+    {
+      "epoch": 1.4101371774006046,
+      "grad_norm": 2.1448546108736135,
+      "learning_rate": 6.383480344729903e-06,
+      "loss": 0.5044,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4115322018135317,
+      "grad_norm": 2.1232999504648546,
+      "learning_rate": 6.375669579013461e-06,
+      "loss": 0.5154,
+      "step": 1011
+    },
+    {
+      "epoch": 1.412927226226459,
+      "grad_norm": 2.0320440158568642,
+      "learning_rate": 6.367855179588597e-06,
+      "loss": 0.5162,
+      "step": 1012
+    },
+    {
+      "epoch": 1.4143222506393862,
+      "grad_norm": 2.082682281577738,
+      "learning_rate": 6.3600371670963525e-06,
+      "loss": 0.4703,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4157172750523135,
+      "grad_norm": 2.0549361603179057,
+      "learning_rate": 6.352215562187307e-06,
+      "loss": 0.4864,
+      "step": 1014
+    },
+    {
+      "epoch": 1.4171122994652405,
+      "grad_norm": 2.008308887519786,
+      "learning_rate": 6.344390385521534e-06,
+      "loss": 0.4803,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4185073238781678,
+      "grad_norm": 2.153291637026902,
+      "learning_rate": 6.33656165776854e-06,
+      "loss": 0.4873,
+      "step": 1016
+    },
+    {
+      "epoch": 1.419902348291095,
+      "grad_norm": 2.234058551814876,
+      "learning_rate": 6.328729399607206e-06,
+      "loss": 0.4814,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4212973727040223,
+      "grad_norm": 2.027510476745004,
+      "learning_rate": 6.320893631725748e-06,
+      "loss": 0.4638,
+      "step": 1018
+    },
+    {
+      "epoch": 1.4226923971169496,
+      "grad_norm": 2.032668407885514,
+      "learning_rate": 6.313054374821647e-06,
+      "loss": 0.4828,
+      "step": 1019
+    },
+    {
+      "epoch": 1.4240874215298769,
+      "grad_norm": 2.0022767352637283,
+      "learning_rate": 6.305211649601595e-06,
+      "loss": 0.4163,
+      "step": 1020
+    },
+    {
+      "epoch": 1.425482445942804,
+      "grad_norm": 1.8646715081707912,
+      "learning_rate": 6.29736547678146e-06,
+      "loss": 0.4599,
+      "step": 1021
+    },
+    {
+      "epoch": 1.4268774703557312,
+      "grad_norm": 2.104118263529368,
+      "learning_rate": 6.289515877086199e-06,
+      "loss": 0.5469,
+      "step": 1022
+    },
+    {
+      "epoch": 1.4282724947686585,
+      "grad_norm": 2.0535906765696468,
+      "learning_rate": 6.2816628712498315e-06,
+      "loss": 0.4806,
+      "step": 1023
+    },
+    {
+      "epoch": 1.4296675191815857,
+      "grad_norm": 1.9740833955901596,
+      "learning_rate": 6.273806480015374e-06,
+      "loss": 0.4685,
+      "step": 1024
+    },
+    {
+      "epoch": 1.4310625435945128,
+      "grad_norm": 2.068364016166393,
+      "learning_rate": 6.265946724134782e-06,
+      "loss": 0.4175,
+      "step": 1025
+    },
+    {
+      "epoch": 1.43245756800744,
+      "grad_norm": 1.93323563183723,
+      "learning_rate": 6.258083624368895e-06,
+      "loss": 0.4816,
+      "step": 1026
+    },
+    {
+      "epoch": 1.4338525924203673,
+      "grad_norm": 1.981160775740525,
+      "learning_rate": 6.250217201487395e-06,
+      "loss": 0.4893,
+      "step": 1027
+    },
+    {
+      "epoch": 1.4352476168332946,
+      "grad_norm": 1.99307431242008,
+      "learning_rate": 6.242347476268733e-06,
+      "loss": 0.4831,
+      "step": 1028
+    },
+    {
+      "epoch": 1.4366426412462219,
+      "grad_norm": 2.1199144029755566,
+      "learning_rate": 6.2344744695000855e-06,
+      "loss": 0.4926,
+      "step": 1029
+    },
+    {
+      "epoch": 1.4380376656591491,
+      "grad_norm": 2.1548929965564527,
+      "learning_rate": 6.226598201977299e-06,
+      "loss": 0.4912,
+      "step": 1030
+    },
+    {
+      "epoch": 1.4394326900720762,
+      "grad_norm": 2.138841310759008,
+      "learning_rate": 6.218718694504831e-06,
+      "loss": 0.5061,
+      "step": 1031
+    },
+    {
+      "epoch": 1.4408277144850035,
+      "grad_norm": 2.059139335954622,
+      "learning_rate": 6.2108359678956954e-06,
+      "loss": 0.4455,
+      "step": 1032
+    },
+    {
+      "epoch": 1.4422227388979307,
+      "grad_norm": 2.257526307269232,
+      "learning_rate": 6.202950042971414e-06,
+      "loss": 0.5006,
+      "step": 1033
+    },
+    {
+      "epoch": 1.443617763310858,
+      "grad_norm": 2.0602776029453627,
+      "learning_rate": 6.19506094056195e-06,
+      "loss": 0.5123,
+      "step": 1034
+    },
+    {
+      "epoch": 1.445012787723785,
+      "grad_norm": 2.3502378746003174,
+      "learning_rate": 6.187168681505666e-06,
+      "loss": 0.4879,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4464078121367123,
+      "grad_norm": 2.069457866799051,
+      "learning_rate": 6.17927328664926e-06,
+      "loss": 0.4667,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4478028365496396,
+      "grad_norm": 1.8813160122162929,
+      "learning_rate": 6.171374776847711e-06,
+      "loss": 0.4959,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4491978609625669,
+      "grad_norm": 2.146202652822659,
+      "learning_rate": 6.163473172964229e-06,
+      "loss": 0.4592,
+      "step": 1038
+    },
+    {
+      "epoch": 1.4505928853754941,
+      "grad_norm": 2.088457160035899,
+      "learning_rate": 6.1555684958701965e-06,
+      "loss": 0.4867,
+      "step": 1039
+    },
+    {
+      "epoch": 1.4519879097884214,
+      "grad_norm": 2.0634903279814125,
+      "learning_rate": 6.1476607664451105e-06,
+      "loss": 0.4948,
+      "step": 1040
+    },
+    {
+      "epoch": 1.4533829342013485,
+      "grad_norm": 2.0752059196250565,
+      "learning_rate": 6.1397500055765345e-06,
+      "loss": 0.5038,
+      "step": 1041
+    },
+    {
+      "epoch": 1.4547779586142757,
+      "grad_norm": 2.1416930775369236,
+      "learning_rate": 6.131836234160036e-06,
+      "loss": 0.4784,
+      "step": 1042
+    },
+    {
+      "epoch": 1.456172983027203,
+      "grad_norm": 2.019710198196397,
+      "learning_rate": 6.123919473099134e-06,
+      "loss": 0.4724,
+      "step": 1043
+    },
+    {
+      "epoch": 1.4575680074401303,
+      "grad_norm": 2.096288377676786,
+      "learning_rate": 6.115999743305252e-06,
+      "loss": 0.4567,
+      "step": 1044
+    },
+    {
+      "epoch": 1.4589630318530573,
+      "grad_norm": 2.0619796314944026,
+      "learning_rate": 6.1080770656976444e-06,
+      "loss": 0.4204,
+      "step": 1045
+    },
+    {
+      "epoch": 1.4603580562659846,
+      "grad_norm": 1.9981309556081335,
+      "learning_rate": 6.100151461203359e-06,
+      "loss": 0.4727,
+      "step": 1046
+    },
+    {
+      "epoch": 1.4617530806789119,
+      "grad_norm": 1.9543783730619007,
+      "learning_rate": 6.0922229507571716e-06,
+      "loss": 0.4923,
+      "step": 1047
+    },
+    {
+      "epoch": 1.4631481050918391,
+      "grad_norm": 2.0229372400285293,
+      "learning_rate": 6.084291555301537e-06,
+      "loss": 0.5037,
+      "step": 1048
+    },
+    {
+      "epoch": 1.4645431295047664,
+      "grad_norm": 2.142636652811151,
+      "learning_rate": 6.076357295786526e-06,
+      "loss": 0.4535,
+      "step": 1049
+    },
+    {
+      "epoch": 1.4659381539176937,
+      "grad_norm": 2.003318668809125,
+      "learning_rate": 6.068420193169779e-06,
+      "loss": 0.5209,
+      "step": 1050
+    },
+    {
+      "epoch": 1.4673331783306207,
+      "grad_norm": 2.1145852835254373,
+      "learning_rate": 6.0604802684164436e-06,
+      "loss": 0.5086,
+      "step": 1051
+    },
+    {
+      "epoch": 1.468728202743548,
+      "grad_norm": 2.114398404283612,
+      "learning_rate": 6.052537542499122e-06,
+      "loss": 0.455,
+      "step": 1052
+    },
+    {
+      "epoch": 1.4701232271564753,
+      "grad_norm": 2.005080980781578,
+      "learning_rate": 6.044592036397816e-06,
+      "loss": 0.4439,
+      "step": 1053
+    },
+    {
+      "epoch": 1.4715182515694025,
+      "grad_norm": 1.9455243320200077,
+      "learning_rate": 6.0366437710998715e-06,
+      "loss": 0.4855,
+      "step": 1054
+    },
+    {
+      "epoch": 1.4729132759823296,
+      "grad_norm": 2.252933665609378,
+      "learning_rate": 6.0286927675999205e-06,
+      "loss": 0.4724,
+      "step": 1055
+    },
+    {
+      "epoch": 1.4743083003952568,
+      "grad_norm": 1.919732919513372,
+      "learning_rate": 6.02073904689983e-06,
+      "loss": 0.5262,
+      "step": 1056
+    },
+    {
+      "epoch": 1.4757033248081841,
+      "grad_norm": 2.3597721797516726,
+      "learning_rate": 6.012782630008646e-06,
+      "loss": 0.5051,
+      "step": 1057
+    },
+    {
+      "epoch": 1.4770983492211114,
+      "grad_norm": 2.214881592520836,
+      "learning_rate": 6.004823537942528e-06,
+      "loss": 0.4667,
+      "step": 1058
+    },
+    {
+      "epoch": 1.4784933736340387,
+      "grad_norm": 1.99206372251337,
+      "learning_rate": 5.996861791724713e-06,
+      "loss": 0.503,
+      "step": 1059
+    },
+    {
+      "epoch": 1.479888398046966,
+      "grad_norm": 2.102693460428217,
+      "learning_rate": 5.98889741238544e-06,
+      "loss": 0.4772,
+      "step": 1060
+    },
+    {
+      "epoch": 1.481283422459893,
+      "grad_norm": 2.1595527269970267,
+      "learning_rate": 5.9809304209619054e-06,
+      "loss": 0.4647,
+      "step": 1061
+    },
+    {
+      "epoch": 1.4826784468728202,
+      "grad_norm": 1.8101345913472888,
+      "learning_rate": 5.9729608384982085e-06,
+      "loss": 0.4393,
+      "step": 1062
+    },
+    {
+      "epoch": 1.4840734712857475,
+      "grad_norm": 2.0572441260481837,
+      "learning_rate": 5.964988686045289e-06,
+      "loss": 0.507,
+      "step": 1063
+    },
+    {
+      "epoch": 1.4854684956986748,
+      "grad_norm": 2.1342940374448958,
+      "learning_rate": 5.957013984660875e-06,
+      "loss": 0.4949,
+      "step": 1064
+    },
+    {
+      "epoch": 1.4868635201116018,
+      "grad_norm": 2.141413426865797,
+      "learning_rate": 5.949036755409432e-06,
+      "loss": 0.4561,
+      "step": 1065
+    },
+    {
+      "epoch": 1.488258544524529,
+      "grad_norm": 2.102762243321115,
+      "learning_rate": 5.941057019362095e-06,
+      "loss": 0.4462,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4896535689374564,
+      "grad_norm": 1.96172978194436,
+      "learning_rate": 5.933074797596627e-06,
+      "loss": 0.4478,
+      "step": 1067
+    },
+    {
+      "epoch": 1.4910485933503836,
+      "grad_norm": 2.1140432852002427,
+      "learning_rate": 5.925090111197355e-06,
+      "loss": 0.4675,
+      "step": 1068
+    },
+    {
+      "epoch": 1.492443617763311,
+      "grad_norm": 2.0851675116564743,
+      "learning_rate": 5.917102981255114e-06,
+      "loss": 0.4895,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4938386421762382,
+      "grad_norm": 2.1064824290137913,
+      "learning_rate": 5.909113428867195e-06,
+      "loss": 0.5739,
+      "step": 1070
+    },
+    {
+      "epoch": 1.4952336665891652,
+      "grad_norm": 2.363919632384231,
+      "learning_rate": 5.901121475137287e-06,
+      "loss": 0.4841,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4966286910020925,
+      "grad_norm": 2.1554543460639515,
+      "learning_rate": 5.893127141175425e-06,
+      "loss": 0.5,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4980237154150198,
+      "grad_norm": 2.100025898464971,
+      "learning_rate": 5.885130448097926e-06,
+      "loss": 0.5064,
+      "step": 1073
+    },
+    {
+      "epoch": 1.499418739827947,
+      "grad_norm": 2.1902729566688226,
+      "learning_rate": 5.877131417027343e-06,
+      "loss": 0.5269,
+      "step": 1074
+    },
+    {
+      "epoch": 1.500813764240874,
+      "grad_norm": 2.1036882743228,
+      "learning_rate": 5.869130069092401e-06,
+      "loss": 0.5054,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5022087886538014,
+      "grad_norm": 2.0250157904579584,
+      "learning_rate": 5.861126425427949e-06,
+      "loss": 0.4583,
+      "step": 1076
+    },
+    {
+      "epoch": 1.5036038130667286,
+      "grad_norm": 2.1122885769868294,
+      "learning_rate": 5.853120507174894e-06,
+      "loss": 0.518,
+      "step": 1077
+    },
+    {
+      "epoch": 1.504998837479656,
+      "grad_norm": 2.0922750769271636,
+      "learning_rate": 5.845112335480159e-06,
+      "loss": 0.4901,
+      "step": 1078
+    },
+    {
+      "epoch": 1.5063938618925832,
+      "grad_norm": 2.0533938325641663,
+      "learning_rate": 5.83710193149661e-06,
+      "loss": 0.4318,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5077888863055104,
+      "grad_norm": 2.038905596805902,
+      "learning_rate": 5.829089316383018e-06,
+      "loss": 0.4922,
+      "step": 1080
+    },
+    {
+      "epoch": 1.5091839107184377,
+      "grad_norm": 2.151329977005894,
+      "learning_rate": 5.821074511303988e-06,
+      "loss": 0.4657,
+      "step": 1081
+    },
+    {
+      "epoch": 1.5105789351313648,
+      "grad_norm": 1.92148106773121,
+      "learning_rate": 5.813057537429915e-06,
+      "loss": 0.4731,
+      "step": 1082
+    },
+    {
+      "epoch": 1.511973959544292,
+      "grad_norm": 2.091908495352274,
+      "learning_rate": 5.805038415936919e-06,
+      "loss": 0.4274,
+      "step": 1083
+    },
+    {
+      "epoch": 1.5133689839572193,
+      "grad_norm": 1.965352130042937,
+      "learning_rate": 5.797017168006791e-06,
+      "loss": 0.4769,
+      "step": 1084
+    },
+    {
+      "epoch": 1.5147640083701464,
+      "grad_norm": 2.1649638197222343,
+      "learning_rate": 5.7889938148269445e-06,
+      "loss": 0.5076,
+      "step": 1085
+    },
+    {
+      "epoch": 1.5161590327830736,
+      "grad_norm": 2.014733451344566,
+      "learning_rate": 5.7809683775903525e-06,
+      "loss": 0.4598,
+      "step": 1086
+    },
+    {
+      "epoch": 1.517554057196001,
+      "grad_norm": 2.2758162297619844,
+      "learning_rate": 5.7729408774954865e-06,
+      "loss": 0.4766,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5189490816089282,
+      "grad_norm": 2.1485122128659815,
+      "learning_rate": 5.764911335746275e-06,
+      "loss": 0.4787,
+      "step": 1088
+    },
+    {
+      "epoch": 1.5203441060218554,
+      "grad_norm": 2.1157777785516236,
+      "learning_rate": 5.756879773552037e-06,
+      "loss": 0.4864,
+      "step": 1089
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 2.049165688638103,
+      "learning_rate": 5.748846212127421e-06,
+      "loss": 0.5437,
+      "step": 1090
+    },
+    {
+      "epoch": 1.52313415484771,
+      "grad_norm": 2.1872948943308494,
+      "learning_rate": 5.74081067269237e-06,
+      "loss": 0.4513,
+      "step": 1091
+    },
+    {
+      "epoch": 1.524529179260637,
+      "grad_norm": 1.980706388022509,
+      "learning_rate": 5.732773176472042e-06,
+      "loss": 0.4243,
+      "step": 1092
+    },
+    {
+      "epoch": 1.5259242036735643,
+      "grad_norm": 1.8594637870370903,
+      "learning_rate": 5.7247337446967625e-06,
+      "loss": 0.4787,
+      "step": 1093
+    },
+    {
+      "epoch": 1.5273192280864916,
+      "grad_norm": 2.004778006837601,
+      "learning_rate": 5.716692398601975e-06,
+      "loss": 0.4724,
+      "step": 1094
+    },
+    {
+      "epoch": 1.5287142524994186,
+      "grad_norm": 1.9496047062065673,
+      "learning_rate": 5.708649159428181e-06,
+      "loss": 0.5118,
+      "step": 1095
+    },
+    {
+      "epoch": 1.5301092769123459,
+      "grad_norm": 2.323924983303737,
+      "learning_rate": 5.700604048420875e-06,
+      "loss": 0.5089,
+      "step": 1096
+    },
+    {
+      "epoch": 1.5315043013252732,
+      "grad_norm": 1.8596638029475623,
+      "learning_rate": 5.692557086830501e-06,
+      "loss": 0.4974,
+      "step": 1097
+    },
+    {
+      "epoch": 1.5328993257382004,
+      "grad_norm": 2.1299683325884056,
+      "learning_rate": 5.68450829591239e-06,
+      "loss": 0.4651,
+      "step": 1098
+    },
+    {
+      "epoch": 1.5342943501511277,
+      "grad_norm": 2.076100505238362,
+      "learning_rate": 5.676457696926703e-06,
+      "loss": 0.4883,
+      "step": 1099
+    },
+    {
+      "epoch": 1.535689374564055,
+      "grad_norm": 2.146952548510558,
+      "learning_rate": 5.668405311138382e-06,
+      "loss": 0.4918,
+      "step": 1100
+    },
+    {
+      "epoch": 1.5370843989769822,
+      "grad_norm": 2.0439577341140915,
+      "learning_rate": 5.660351159817083e-06,
+      "loss": 0.4709,
+      "step": 1101
+    },
+    {
+      "epoch": 1.5384794233899093,
+      "grad_norm": 2.188976123724856,
+      "learning_rate": 5.652295264237128e-06,
+      "loss": 0.4984,
+      "step": 1102
+    },
+    {
+      "epoch": 1.5398744478028366,
+      "grad_norm": 2.0659662235602942,
+      "learning_rate": 5.6442376456774495e-06,
+      "loss": 0.4316,
+      "step": 1103
+    },
+    {
+      "epoch": 1.5412694722157638,
+      "grad_norm": 1.9009883802453258,
+      "learning_rate": 5.636178325421524e-06,
+      "loss": 0.4689,
+      "step": 1104
+    },
+    {
+      "epoch": 1.5426644966286909,
+      "grad_norm": 2.072884089336916,
+      "learning_rate": 5.628117324757326e-06,
+      "loss": 0.4881,
+      "step": 1105
+    },
+    {
+      "epoch": 1.5440595210416181,
+      "grad_norm": 2.0809785200505466,
+      "learning_rate": 5.620054664977275e-06,
+      "loss": 0.3991,
+      "step": 1106
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 1.9217918849239009,
+      "learning_rate": 5.61199036737816e-06,
+      "loss": 0.4564,
+      "step": 1107
+    },
+    {
+      "epoch": 1.5468495698674727,
+      "grad_norm": 1.8909553621918078,
+      "learning_rate": 5.603924453261109e-06,
+      "loss": 0.4339,
+      "step": 1108
+    },
+    {
+      "epoch": 1.5482445942804,
+      "grad_norm": 2.128862572565254,
+      "learning_rate": 5.595856943931512e-06,
+      "loss": 0.4987,
+      "step": 1109
+    },
+    {
+      "epoch": 1.5496396186933272,
+      "grad_norm": 2.1719486428343457,
+      "learning_rate": 5.587787860698975e-06,
+      "loss": 0.4635,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5510346431062545,
+      "grad_norm": 1.9555844376182605,
+      "learning_rate": 5.579717224877261e-06,
+      "loss": 0.529,
+      "step": 1111
+    },
+    {
+      "epoch": 1.5524296675191815,
+      "grad_norm": 2.2688788762864927,
+      "learning_rate": 5.571645057784236e-06,
+      "loss": 0.5013,
+      "step": 1112
+    },
+    {
+      "epoch": 1.5538246919321088,
+      "grad_norm": 2.228130536027715,
+      "learning_rate": 5.5635713807418055e-06,
+      "loss": 0.4697,
+      "step": 1113
+    },
+    {
+      "epoch": 1.555219716345036,
+      "grad_norm": 2.0348706803599135,
+      "learning_rate": 5.55549621507587e-06,
+      "loss": 0.4896,
+      "step": 1114
+    },
+    {
+      "epoch": 1.5566147407579631,
+      "grad_norm": 1.985770851471805,
+      "learning_rate": 5.547419582116259e-06,
+      "loss": 0.5184,
+      "step": 1115
+    },
+    {
+      "epoch": 1.5580097651708904,
+      "grad_norm": 2.2260993059821073,
+      "learning_rate": 5.539341503196674e-06,
+      "loss": 0.4412,
+      "step": 1116
+    },
+    {
+      "epoch": 1.5594047895838177,
+      "grad_norm": 2.1526251050811336,
+      "learning_rate": 5.531261999654646e-06,
+      "loss": 0.4407,
+      "step": 1117
+    },
+    {
+      "epoch": 1.560799813996745,
+      "grad_norm": 2.021163422905953,
+      "learning_rate": 5.5231810928314555e-06,
+      "loss": 0.4614,
+      "step": 1118
+    },
+    {
+      "epoch": 1.5621948384096722,
+      "grad_norm": 1.974269573884685,
+      "learning_rate": 5.5150988040721e-06,
+      "loss": 0.4581,
+      "step": 1119
+    },
+    {
+      "epoch": 1.5635898628225995,
+      "grad_norm": 2.075129892774578,
+      "learning_rate": 5.507015154725226e-06,
+      "loss": 0.4614,
+      "step": 1120
+    },
+    {
+      "epoch": 1.5649848872355268,
+      "grad_norm": 1.8323848221545598,
+      "learning_rate": 5.4989301661430685e-06,
+      "loss": 0.451,
+      "step": 1121
+    },
+    {
+      "epoch": 1.5663799116484538,
+      "grad_norm": 2.0259660795350083,
+      "learning_rate": 5.490843859681404e-06,
+      "loss": 0.4601,
+      "step": 1122
+    },
+    {
+      "epoch": 1.567774936061381,
+      "grad_norm": 2.1270056899092076,
+      "learning_rate": 5.48275625669949e-06,
+      "loss": 0.458,
+      "step": 1123
+    },
+    {
+      "epoch": 1.5691699604743083,
+      "grad_norm": 1.9780355578127613,
+      "learning_rate": 5.474667378560007e-06,
+      "loss": 0.4608,
+      "step": 1124
+    },
+    {
+      "epoch": 1.5705649848872354,
+      "grad_norm": 1.9831337590058289,
+      "learning_rate": 5.466577246629006e-06,
+      "loss": 0.4808,
+      "step": 1125
+    },
+    {
+      "epoch": 1.5719600093001627,
+      "grad_norm": 1.964070734016315,
+      "learning_rate": 5.458485882275848e-06,
+      "loss": 0.4335,
+      "step": 1126
+    },
+    {
+      "epoch": 1.57335503371309,
+      "grad_norm": 1.8520479300798307,
+      "learning_rate": 5.45039330687315e-06,
+      "loss": 0.4896,
+      "step": 1127
+    },
+    {
+      "epoch": 1.5747500581260172,
+      "grad_norm": 1.987414988454785,
+      "learning_rate": 5.442299541796727e-06,
+      "loss": 0.4483,
+      "step": 1128
+    },
+    {
+      "epoch": 1.5761450825389445,
+      "grad_norm": 2.022178299234183,
+      "learning_rate": 5.4342046084255385e-06,
+      "loss": 0.4316,
+      "step": 1129
+    },
+    {
+      "epoch": 1.5775401069518717,
+      "grad_norm": 1.9494172415463071,
+      "learning_rate": 5.426108528141627e-06,
+      "loss": 0.4712,
+      "step": 1130
+    },
+    {
+      "epoch": 1.578935131364799,
+      "grad_norm": 1.9126927477747944,
+      "learning_rate": 5.4180113223300665e-06,
+      "loss": 0.443,
+      "step": 1131
+    },
+    {
+      "epoch": 1.580330155777726,
+      "grad_norm": 2.0412089862860614,
+      "learning_rate": 5.409913012378903e-06,
+      "loss": 0.4237,
+      "step": 1132
+    },
+    {
+      "epoch": 1.5817251801906533,
+      "grad_norm": 1.947246022834085,
+      "learning_rate": 5.401813619679102e-06,
+      "loss": 0.4951,
+      "step": 1133
+    },
+    {
+      "epoch": 1.5831202046035806,
+      "grad_norm": 2.0984893727921294,
+      "learning_rate": 5.3937131656244834e-06,
+      "loss": 0.4401,
+      "step": 1134
+    },
+    {
+      "epoch": 1.5845152290165077,
+      "grad_norm": 2.1947448476976437,
+      "learning_rate": 5.385611671611676e-06,
+      "loss": 0.4799,
+      "step": 1135
+    },
+    {
+      "epoch": 1.585910253429435,
+      "grad_norm": 2.026074719354588,
+      "learning_rate": 5.377509159040051e-06,
+      "loss": 0.4918,
+      "step": 1136
+    },
+    {
+      "epoch": 1.5873052778423622,
+      "grad_norm": 2.1061592247651486,
+      "learning_rate": 5.3694056493116745e-06,
+      "loss": 0.4472,
+      "step": 1137
+    },
+    {
+      "epoch": 1.5887003022552895,
+      "grad_norm": 2.119635832241502,
+      "learning_rate": 5.361301163831242e-06,
+      "loss": 0.4912,
+      "step": 1138
+    },
+    {
+      "epoch": 1.5900953266682167,
+      "grad_norm": 2.0514457851634504,
+      "learning_rate": 5.353195724006031e-06,
+      "loss": 0.4347,
+      "step": 1139
+    },
+    {
+      "epoch": 1.591490351081144,
+      "grad_norm": 1.9739425266282826,
+      "learning_rate": 5.345089351245834e-06,
+      "loss": 0.489,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5928853754940713,
+      "grad_norm": 2.097334194705438,
+      "learning_rate": 5.336982066962915e-06,
+      "loss": 0.4706,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5942803999069983,
+      "grad_norm": 2.0889563513652187,
+      "learning_rate": 5.328873892571941e-06,
+      "loss": 0.4324,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5956754243199256,
+      "grad_norm": 2.161914385832255,
+      "learning_rate": 5.320764849489929e-06,
+      "loss": 0.4957,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5970704487328529,
+      "grad_norm": 2.127711288940709,
+      "learning_rate": 5.312654959136194e-06,
+      "loss": 0.4044,
+      "step": 1144
+    },
+    {
+      "epoch": 1.59846547314578,
+      "grad_norm": 1.9955461760081252,
+      "learning_rate": 5.304544242932288e-06,
+      "loss": 0.5024,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5998604975587072,
+      "grad_norm": 2.146359711210145,
+      "learning_rate": 5.296432722301944e-06,
+      "loss": 0.5134,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6012555219716345,
+      "grad_norm": 2.056122492718699,
+      "learning_rate": 5.288320418671018e-06,
+      "loss": 0.4519,
+      "step": 1147
+    },
+    {
+      "epoch": 1.6026505463845617,
+      "grad_norm": 2.26187453297196,
+      "learning_rate": 5.280207353467438e-06,
+      "loss": 0.4829,
+      "step": 1148
+    },
+    {
+      "epoch": 1.604045570797489,
+      "grad_norm": 2.0985463489065426,
+      "learning_rate": 5.272093548121141e-06,
+      "loss": 0.4886,
+      "step": 1149
+    },
+    {
+      "epoch": 1.6054405952104163,
+      "grad_norm": 2.0830677262546002,
+      "learning_rate": 5.26397902406402e-06,
+      "loss": 0.4909,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6068356196233435,
+      "grad_norm": 2.2329507839509923,
+      "learning_rate": 5.255863802729866e-06,
+      "loss": 0.4694,
+      "step": 1151
+    },
+    {
+      "epoch": 1.6082306440362706,
+      "grad_norm": 2.1821069364143146,
+      "learning_rate": 5.247747905554311e-06,
+      "loss": 0.4388,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6096256684491979,
+      "grad_norm": 1.9642708773504776,
+      "learning_rate": 5.239631353974774e-06,
+      "loss": 0.5221,
+      "step": 1153
+    },
+    {
+      "epoch": 1.6110206928621251,
+      "grad_norm": 2.15241370034379,
+      "learning_rate": 5.231514169430403e-06,
+      "loss": 0.4742,
+      "step": 1154
+    },
+    {
+      "epoch": 1.6124157172750522,
+      "grad_norm": 2.1018420489764735,
+      "learning_rate": 5.223396373362013e-06,
+      "loss": 0.4391,
+      "step": 1155
+    },
+    {
+      "epoch": 1.6138107416879794,
+      "grad_norm": 1.8679671585704423,
+      "learning_rate": 5.215277987212041e-06,
+      "loss": 0.5281,
+      "step": 1156
+    },
+    {
+      "epoch": 1.6152057661009067,
+      "grad_norm": 2.1271535673490036,
+      "learning_rate": 5.207159032424478e-06,
+      "loss": 0.4999,
+      "step": 1157
+    },
+    {
+      "epoch": 1.616600790513834,
+      "grad_norm": 2.0550576490650565,
+      "learning_rate": 5.199039530444819e-06,
+      "loss": 0.452,
+      "step": 1158
+    },
+    {
+      "epoch": 1.6179958149267613,
+      "grad_norm": 2.086211359150292,
+      "learning_rate": 5.1909195027200055e-06,
+      "loss": 0.4656,
+      "step": 1159
+    },
+    {
+      "epoch": 1.6193908393396885,
+      "grad_norm": 1.9393017851532006,
+      "learning_rate": 5.182798970698361e-06,
+      "loss": 0.4661,
+      "step": 1160
+    },
+    {
+      "epoch": 1.6207858637526158,
+      "grad_norm": 2.11447885984754,
+      "learning_rate": 5.174677955829551e-06,
+      "loss": 0.4784,
+      "step": 1161
+    },
+    {
+      "epoch": 1.6221808881655428,
+      "grad_norm": 2.021527606784538,
+      "learning_rate": 5.166556479564511e-06,
+      "loss": 0.5031,
+      "step": 1162
+    },
+    {
+      "epoch": 1.6235759125784701,
+      "grad_norm": 1.989052449176325,
+      "learning_rate": 5.158434563355392e-06,
+      "loss": 0.5083,
+      "step": 1163
+    },
+    {
+      "epoch": 1.6249709369913974,
+      "grad_norm": 2.2313083148153625,
+      "learning_rate": 5.150312228655515e-06,
+      "loss": 0.4175,
+      "step": 1164
+    },
+    {
+      "epoch": 1.6263659614043244,
+      "grad_norm": 1.9280232439374705,
+      "learning_rate": 5.142189496919302e-06,
+      "loss": 0.4636,
+      "step": 1165
+    },
+    {
+      "epoch": 1.6277609858172517,
+      "grad_norm": 1.9246863328355193,
+      "learning_rate": 5.1340663896022206e-06,
+      "loss": 0.4467,
+      "step": 1166
+    },
+    {
+      "epoch": 1.629156010230179,
+      "grad_norm": 2.001675021541609,
+      "learning_rate": 5.125942928160736e-06,
+      "loss": 0.4713,
+      "step": 1167
+    },
+    {
+      "epoch": 1.6305510346431062,
+      "grad_norm": 2.1481197676528745,
+      "learning_rate": 5.117819134052246e-06,
+      "loss": 0.4849,
+      "step": 1168
+    },
+    {
+      "epoch": 1.6319460590560335,
+      "grad_norm": 2.022955901936903,
+      "learning_rate": 5.10969502873503e-06,
+      "loss": 0.4654,
+      "step": 1169
+    },
+    {
+      "epoch": 1.6333410834689608,
+      "grad_norm": 2.186736394556812,
+      "learning_rate": 5.101570633668185e-06,
+      "loss": 0.4674,
+      "step": 1170
+    },
+    {
+      "epoch": 1.634736107881888,
+      "grad_norm": 2.19635300793701,
+      "learning_rate": 5.093445970311576e-06,
+      "loss": 0.471,
+      "step": 1171
+    },
+    {
+      "epoch": 1.636131132294815,
+      "grad_norm": 2.0592350084511106,
+      "learning_rate": 5.085321060125775e-06,
+      "loss": 0.4522,
+      "step": 1172
+    },
+    {
+      "epoch": 1.6375261567077424,
+      "grad_norm": 2.0924620802741467,
+      "learning_rate": 5.07719592457201e-06,
+      "loss": 0.4646,
+      "step": 1173
+    },
+    {
+      "epoch": 1.6389211811206696,
+      "grad_norm": 2.1993464412256167,
+      "learning_rate": 5.069070585112097e-06,
+      "loss": 0.5289,
+      "step": 1174
+    },
+    {
+      "epoch": 1.6403162055335967,
+      "grad_norm": 1.9527466803709046,
+      "learning_rate": 5.060945063208399e-06,
+      "loss": 0.4692,
+      "step": 1175
+    },
+    {
+      "epoch": 1.641711229946524,
+      "grad_norm": 2.1525612142674833,
+      "learning_rate": 5.052819380323757e-06,
+      "loss": 0.4761,
+      "step": 1176
+    },
+    {
+      "epoch": 1.6431062543594512,
+      "grad_norm": 2.0655110259638207,
+      "learning_rate": 5.044693557921434e-06,
+      "loss": 0.4696,
+      "step": 1177
+    },
+    {
+      "epoch": 1.6445012787723785,
+      "grad_norm": 2.095404927614161,
+      "learning_rate": 5.036567617465067e-06,
+      "loss": 0.478,
+      "step": 1178
+    },
+    {
+      "epoch": 1.6458963031853058,
+      "grad_norm": 2.2636358073037557,
+      "learning_rate": 5.0284415804186025e-06,
+      "loss": 0.4676,
+      "step": 1179
+    },
+    {
+      "epoch": 1.647291327598233,
+      "grad_norm": 2.016150491128099,
+      "learning_rate": 5.02031546824624e-06,
+      "loss": 0.4707,
+      "step": 1180
+    },
+    {
+      "epoch": 1.6486863520111603,
+      "grad_norm": 2.1200548779498085,
+      "learning_rate": 5.012189302412383e-06,
+      "loss": 0.4903,
+      "step": 1181
+    },
+    {
+      "epoch": 1.6500813764240876,
+      "grad_norm": 2.154591495150212,
+      "learning_rate": 5.0040631043815715e-06,
+      "loss": 0.5041,
+      "step": 1182
+    },
+    {
+      "epoch": 1.6514764008370146,
+      "grad_norm": 2.072675257333494,
+      "learning_rate": 4.99593689561843e-06,
+      "loss": 0.4662,
+      "step": 1183
+    },
+    {
+      "epoch": 1.652871425249942,
+      "grad_norm": 1.9847287863038208,
+      "learning_rate": 4.987810697587618e-06,
+      "loss": 0.4988,
+      "step": 1184
+    },
+    {
+      "epoch": 1.654266449662869,
+      "grad_norm": 2.1882981619139477,
+      "learning_rate": 4.979684531753761e-06,
+      "loss": 0.4554,
+      "step": 1185
+    },
+    {
+      "epoch": 1.6556614740757962,
+      "grad_norm": 2.155215055693557,
+      "learning_rate": 4.971558419581398e-06,
+      "loss": 0.4495,
+      "step": 1186
+    },
+    {
+      "epoch": 1.6570564984887235,
+      "grad_norm": 1.9462286169261411,
+      "learning_rate": 4.963432382534933e-06,
+      "loss": 0.5166,
+      "step": 1187
+    },
+    {
+      "epoch": 1.6584515229016508,
+      "grad_norm": 2.319365094329712,
+      "learning_rate": 4.955306442078568e-06,
+      "loss": 0.5077,
+      "step": 1188
+    },
+    {
+      "epoch": 1.659846547314578,
+      "grad_norm": 2.013268429787615,
+      "learning_rate": 4.947180619676244e-06,
+      "loss": 0.4225,
+      "step": 1189
+    },
+    {
+      "epoch": 1.6612415717275053,
+      "grad_norm": 1.9030332809738684,
+      "learning_rate": 4.9390549367916004e-06,
+      "loss": 0.4474,
+      "step": 1190
+    },
+    {
+      "epoch": 1.6626365961404326,
+      "grad_norm": 1.9091704807756302,
+      "learning_rate": 4.930929414887904e-06,
+      "loss": 0.4471,
+      "step": 1191
+    },
+    {
+      "epoch": 1.6640316205533598,
+      "grad_norm": 2.131054390165848,
+      "learning_rate": 4.9228040754279915e-06,
+      "loss": 0.4812,
+      "step": 1192
+    },
+    {
+      "epoch": 1.665426644966287,
+      "grad_norm": 2.0380709327701862,
+      "learning_rate": 4.914678939874225e-06,
+      "loss": 0.5226,
+      "step": 1193
+    },
+    {
+      "epoch": 1.6668216693792142,
+      "grad_norm": 2.2253463416293786,
+      "learning_rate": 4.906554029688427e-06,
+      "loss": 0.4383,
+      "step": 1194
+    },
+    {
+      "epoch": 1.6682166937921412,
+      "grad_norm": 1.9151466529202,
+      "learning_rate": 4.898429366331815e-06,
+      "loss": 0.4818,
+      "step": 1195
+    },
+    {
+      "epoch": 1.6696117182050685,
+      "grad_norm": 2.103018852231921,
+      "learning_rate": 4.8903049712649705e-06,
+      "loss": 0.51,
+      "step": 1196
+    },
+    {
+      "epoch": 1.6710067426179958,
+      "grad_norm": 2.1336210474522033,
+      "learning_rate": 4.8821808659477544e-06,
+      "loss": 0.4482,
+      "step": 1197
+    },
+    {
+      "epoch": 1.672401767030923,
+      "grad_norm": 2.037967905605442,
+      "learning_rate": 4.874057071839265e-06,
+      "loss": 0.4624,
+      "step": 1198
+    },
+    {
+      "epoch": 1.6737967914438503,
+      "grad_norm": 1.9885465427196427,
+      "learning_rate": 4.86593361039778e-06,
+      "loss": 0.4824,
+      "step": 1199
+    },
+    {
+      "epoch": 1.6751918158567776,
+      "grad_norm": 2.1480503191792093,
+      "learning_rate": 4.857810503080701e-06,
+      "loss": 0.4892,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6765868402697048,
+      "grad_norm": 2.1285265643155182,
+      "learning_rate": 4.849687771344487e-06,
+      "loss": 0.518,
+      "step": 1201
+    },
+    {
+      "epoch": 1.677981864682632,
+      "grad_norm": 2.3375001661766217,
+      "learning_rate": 4.841565436644609e-06,
+      "loss": 0.514,
+      "step": 1202
+    },
+    {
+      "epoch": 1.6793768890955592,
+      "grad_norm": 2.1836516379544113,
+      "learning_rate": 4.8334435204354915e-06,
+      "loss": 0.4623,
+      "step": 1203
+    },
+    {
+      "epoch": 1.6807719135084864,
+      "grad_norm": 2.104463723618796,
+      "learning_rate": 4.825322044170451e-06,
+      "loss": 0.5108,
+      "step": 1204
+    },
+    {
+      "epoch": 1.6821669379214135,
+      "grad_norm": 1.9744889241875467,
+      "learning_rate": 4.81720102930164e-06,
+      "loss": 0.4725,
+      "step": 1205
+    },
+    {
+      "epoch": 1.6835619623343407,
+      "grad_norm": 2.1029141184734357,
+      "learning_rate": 4.809080497279998e-06,
+      "loss": 0.4917,
+      "step": 1206
+    },
+    {
+      "epoch": 1.684956986747268,
+      "grad_norm": 2.1916454273646315,
+      "learning_rate": 4.800960469555183e-06,
+      "loss": 0.4912,
+      "step": 1207
+    },
+    {
+      "epoch": 1.6863520111601953,
+      "grad_norm": 1.955939718054068,
+      "learning_rate": 4.792840967575523e-06,
+      "loss": 0.487,
+      "step": 1208
+    },
+    {
+      "epoch": 1.6877470355731226,
+      "grad_norm": 2.059844165344383,
+      "learning_rate": 4.784722012787961e-06,
+      "loss": 0.495,
+      "step": 1209
+    },
+    {
+      "epoch": 1.6891420599860498,
+      "grad_norm": 2.2676353256111823,
+      "learning_rate": 4.776603626637988e-06,
+      "loss": 0.4841,
+      "step": 1210
+    },
+    {
+      "epoch": 1.690537084398977,
+      "grad_norm": 2.024198676505603,
+      "learning_rate": 4.768485830569598e-06,
+      "loss": 0.4484,
+      "step": 1211
+    },
+    {
+      "epoch": 1.6919321088119044,
+      "grad_norm": 2.0004153672009513,
+      "learning_rate": 4.7603686460252265e-06,
+      "loss": 0.4911,
+      "step": 1212
+    },
+    {
+      "epoch": 1.6933271332248314,
+      "grad_norm": 2.1709204019397563,
+      "learning_rate": 4.75225209444569e-06,
+      "loss": 0.4627,
+      "step": 1213
+    },
+    {
+      "epoch": 1.6947221576377587,
+      "grad_norm": 2.1160225414488654,
+      "learning_rate": 4.744136197270135e-06,
+      "loss": 0.5304,
+      "step": 1214
+    },
+    {
+      "epoch": 1.6961171820506857,
+      "grad_norm": 2.2698535491172525,
+      "learning_rate": 4.736020975935981e-06,
+      "loss": 0.4793,
+      "step": 1215
+    },
+    {
+      "epoch": 1.697512206463613,
+      "grad_norm": 2.0436804954046,
+      "learning_rate": 4.72790645187886e-06,
+      "loss": 0.4407,
+      "step": 1216
+    },
+    {
+      "epoch": 1.6989072308765403,
+      "grad_norm": 1.9979146113097979,
+      "learning_rate": 4.7197926465325626e-06,
+      "loss": 0.4564,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7003022552894675,
+      "grad_norm": 2.0826310908062604,
+      "learning_rate": 4.711679581328983e-06,
+      "loss": 0.4837,
+      "step": 1218
+    },
+    {
+      "epoch": 1.7016972797023948,
+      "grad_norm": 2.221127633913511,
+      "learning_rate": 4.703567277698058e-06,
+      "loss": 0.5505,
+      "step": 1219
+    },
+    {
+      "epoch": 1.703092304115322,
+      "grad_norm": 2.3042281576812664,
+      "learning_rate": 4.695455757067712e-06,
+      "loss": 0.4513,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7044873285282494,
+      "grad_norm": 1.9139161970984668,
+      "learning_rate": 4.687345040863808e-06,
+      "loss": 0.4778,
+      "step": 1221
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 2.1441515179909656,
+      "learning_rate": 4.679235150510072e-06,
+      "loss": 0.4342,
+      "step": 1222
+    },
+    {
+      "epoch": 1.7072773773541037,
+      "grad_norm": 1.9784971632191246,
+      "learning_rate": 4.671126107428061e-06,
+      "loss": 0.4749,
+      "step": 1223
+    },
+    {
+      "epoch": 1.708672401767031,
+      "grad_norm": 1.9425240844023457,
+      "learning_rate": 4.663017933037087e-06,
+      "loss": 0.4606,
+      "step": 1224
+    },
+    {
+      "epoch": 1.710067426179958,
+      "grad_norm": 2.06859895894967,
+      "learning_rate": 4.6549106487541666e-06,
+      "loss": 0.491,
+      "step": 1225
+    },
+    {
+      "epoch": 1.7114624505928853,
+      "grad_norm": 2.1250812807049155,
+      "learning_rate": 4.646804275993971e-06,
+      "loss": 0.46,
+      "step": 1226
+    },
+    {
+      "epoch": 1.7128574750058125,
+      "grad_norm": 2.1870237837592357,
+      "learning_rate": 4.63869883616876e-06,
+      "loss": 0.4526,
+      "step": 1227
+    },
+    {
+      "epoch": 1.7142524994187398,
+      "grad_norm": 1.933173246666357,
+      "learning_rate": 4.630594350688327e-06,
+      "loss": 0.4965,
+      "step": 1228
+    },
+    {
+      "epoch": 1.715647523831667,
+      "grad_norm": 2.1720589382321935,
+      "learning_rate": 4.62249084095995e-06,
+      "loss": 0.4633,
+      "step": 1229
+    },
+    {
+      "epoch": 1.7170425482445943,
+      "grad_norm": 2.03257424247359,
+      "learning_rate": 4.614388328388327e-06,
+      "loss": 0.5121,
+      "step": 1230
+    },
+    {
+      "epoch": 1.7184375726575216,
+      "grad_norm": 2.1561072697752395,
+      "learning_rate": 4.606286834375517e-06,
+      "loss": 0.4976,
+      "step": 1231
+    },
+    {
+      "epoch": 1.7198325970704489,
+      "grad_norm": 2.053680058584616,
+      "learning_rate": 4.598186380320899e-06,
+      "loss": 0.433,
+      "step": 1232
+    },
+    {
+      "epoch": 1.721227621483376,
+      "grad_norm": 2.181642713396842,
+      "learning_rate": 4.5900869876210986e-06,
+      "loss": 0.4768,
+      "step": 1233
+    },
+    {
+      "epoch": 1.7226226458963032,
+      "grad_norm": 2.0522746791202318,
+      "learning_rate": 4.581988677669935e-06,
+      "loss": 0.5157,
+      "step": 1234
+    },
+    {
+      "epoch": 1.7240176703092303,
+      "grad_norm": 2.11368549262154,
+      "learning_rate": 4.573891471858375e-06,
+      "loss": 0.4832,
+      "step": 1235
+    },
+    {
+      "epoch": 1.7254126947221575,
+      "grad_norm": 2.0913027904457127,
+      "learning_rate": 4.565795391574465e-06,
+      "loss": 0.4737,
+      "step": 1236
+    },
+    {
+      "epoch": 1.7268077191350848,
+      "grad_norm": 2.0912014065923636,
+      "learning_rate": 4.5577004582032745e-06,
+      "loss": 0.4488,
+      "step": 1237
+    },
+    {
+      "epoch": 1.728202743548012,
+      "grad_norm": 2.050372387152731,
+      "learning_rate": 4.549606693126851e-06,
+      "loss": 0.4636,
+      "step": 1238
+    },
+    {
+      "epoch": 1.7295977679609393,
+      "grad_norm": 2.1154084542696587,
+      "learning_rate": 4.541514117724155e-06,
+      "loss": 0.5255,
+      "step": 1239
+    },
+    {
+      "epoch": 1.7309927923738666,
+      "grad_norm": 2.0748003125338546,
+      "learning_rate": 4.533422753370995e-06,
+      "loss": 0.4257,
+      "step": 1240
+    },
+    {
+      "epoch": 1.7323878167867939,
+      "grad_norm": 1.9621472789788037,
+      "learning_rate": 4.525332621439995e-06,
+      "loss": 0.4846,
+      "step": 1241
+    },
+    {
+      "epoch": 1.7337828411997211,
+      "grad_norm": 2.103237491557985,
+      "learning_rate": 4.517243743300513e-06,
+      "loss": 0.4228,
+      "step": 1242
+    },
+    {
+      "epoch": 1.7351778656126482,
+      "grad_norm": 2.177553621256369,
+      "learning_rate": 4.5091561403185976e-06,
+      "loss": 0.5123,
+      "step": 1243
+    },
+    {
+      "epoch": 1.7365728900255755,
+      "grad_norm": 2.091109158658429,
+      "learning_rate": 4.501069833856934e-06,
+      "loss": 0.4629,
+      "step": 1244
+    },
+    {
+      "epoch": 1.7379679144385025,
+      "grad_norm": 2.1467197858254408,
+      "learning_rate": 4.492984845274774e-06,
+      "loss": 0.4354,
+      "step": 1245
+    },
+    {
+      "epoch": 1.7393629388514298,
+      "grad_norm": 2.100552590196265,
+      "learning_rate": 4.484901195927901e-06,
+      "loss": 0.4618,
+      "step": 1246
+    },
+    {
+      "epoch": 1.740757963264357,
+      "grad_norm": 1.9713180215115427,
+      "learning_rate": 4.476818907168545e-06,
+      "loss": 0.4186,
+      "step": 1247
+    },
+    {
+      "epoch": 1.7421529876772843,
+      "grad_norm": 2.03751695096277,
+      "learning_rate": 4.4687380003453555e-06,
+      "loss": 0.4412,
+      "step": 1248
+    },
+    {
+      "epoch": 1.7435480120902116,
+      "grad_norm": 2.101186407375219,
+      "learning_rate": 4.460658496803327e-06,
+      "loss": 0.4753,
+      "step": 1249
+    },
+    {
+      "epoch": 1.7449430365031389,
+      "grad_norm": 2.1552915811427247,
+      "learning_rate": 4.4525804178837425e-06,
+      "loss": 0.4574,
+      "step": 1250
+    },
+    {
+      "epoch": 1.7463380609160661,
+      "grad_norm": 2.270007645721411,
+      "learning_rate": 4.4445037849241305e-06,
+      "loss": 0.5034,
+      "step": 1251
+    },
+    {
+      "epoch": 1.7477330853289934,
+      "grad_norm": 2.128789717967715,
+      "learning_rate": 4.436428619258196e-06,
+      "loss": 0.4722,
+      "step": 1252
+    },
+    {
+      "epoch": 1.7491281097419205,
+      "grad_norm": 2.0638528861792365,
+      "learning_rate": 4.428354942215766e-06,
+      "loss": 0.4406,
+      "step": 1253
+    },
+    {
+      "epoch": 1.7505231341548477,
+      "grad_norm": 2.1111439690461493,
+      "learning_rate": 4.42028277512274e-06,
+      "loss": 0.448,
+      "step": 1254
+    },
+    {
+      "epoch": 1.7519181585677748,
+      "grad_norm": 1.9221909816221243,
+      "learning_rate": 4.412212139301027e-06,
+      "loss": 0.4915,
+      "step": 1255
+    },
+    {
+      "epoch": 1.753313182980702,
+      "grad_norm": 2.116085778290425,
+      "learning_rate": 4.404143056068489e-06,
+      "loss": 0.4063,
+      "step": 1256
+    },
+    {
+      "epoch": 1.7547082073936293,
+      "grad_norm": 1.910803308672269,
+      "learning_rate": 4.3960755467388916e-06,
+      "loss": 0.4783,
+      "step": 1257
+    },
+    {
+      "epoch": 1.7561032318065566,
+      "grad_norm": 2.088441580996199,
+      "learning_rate": 4.388009632621841e-06,
+      "loss": 0.5135,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7574982562194839,
+      "grad_norm": 2.0746611561798503,
+      "learning_rate": 4.379945335022727e-06,
+      "loss": 0.4721,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7588932806324111,
+      "grad_norm": 2.2426717188517618,
+      "learning_rate": 4.371882675242674e-06,
+      "loss": 0.4971,
+      "step": 1260
+    },
+    {
+      "epoch": 1.7602883050453384,
+      "grad_norm": 2.2204709536429843,
+      "learning_rate": 4.363821674578479e-06,
+      "loss": 0.5132,
+      "step": 1261
+    },
+    {
+      "epoch": 1.7616833294582657,
+      "grad_norm": 2.0706691763772183,
+      "learning_rate": 4.355762354322552e-06,
+      "loss": 0.4508,
+      "step": 1262
+    },
+    {
+      "epoch": 1.7630783538711927,
+      "grad_norm": 2.0641312165343533,
+      "learning_rate": 4.347704735762872e-06,
+      "loss": 0.495,
+      "step": 1263
+    },
+    {
+      "epoch": 1.76447337828412,
+      "grad_norm": 2.0817958512499604,
+      "learning_rate": 4.339648840182919e-06,
+      "loss": 0.4815,
+      "step": 1264
+    },
+    {
+      "epoch": 1.765868402697047,
+      "grad_norm": 2.082454137764786,
+      "learning_rate": 4.331594688861619e-06,
+      "loss": 0.4674,
+      "step": 1265
+    },
+    {
+      "epoch": 1.7672634271099743,
+      "grad_norm": 2.1167567040967175,
+      "learning_rate": 4.323542303073297e-06,
+      "loss": 0.4455,
+      "step": 1266
+    },
+    {
+      "epoch": 1.7686584515229016,
+      "grad_norm": 2.1312357560431567,
+      "learning_rate": 4.315491704087613e-06,
+      "loss": 0.456,
+      "step": 1267
+    },
+    {
+      "epoch": 1.7700534759358288,
+      "grad_norm": 2.007748405961002,
+      "learning_rate": 4.3074429131695e-06,
+      "loss": 0.465,
+      "step": 1268
+    },
+    {
+      "epoch": 1.7714485003487561,
+      "grad_norm": 2.1649734551805366,
+      "learning_rate": 4.299395951579126e-06,
+      "loss": 0.4658,
+      "step": 1269
+    },
+    {
+      "epoch": 1.7728435247616834,
+      "grad_norm": 2.013706806375797,
+      "learning_rate": 4.291350840571821e-06,
+      "loss": 0.4307,
+      "step": 1270
+    },
+    {
+      "epoch": 1.7742385491746107,
+      "grad_norm": 2.081976239927056,
+      "learning_rate": 4.283307601398026e-06,
+      "loss": 0.4848,
+      "step": 1271
+    },
+    {
+      "epoch": 1.775633573587538,
+      "grad_norm": 2.2533206281427525,
+      "learning_rate": 4.275266255303238e-06,
+      "loss": 0.4909,
+      "step": 1272
+    },
+    {
+      "epoch": 1.777028598000465,
+      "grad_norm": 2.247778078118919,
+      "learning_rate": 4.2672268235279616e-06,
+      "loss": 0.4611,
+      "step": 1273
+    },
+    {
+      "epoch": 1.7784236224133922,
+      "grad_norm": 1.9320941214004883,
+      "learning_rate": 4.259189327307632e-06,
+      "loss": 0.4782,
+      "step": 1274
+    },
+    {
+      "epoch": 1.7798186468263193,
+      "grad_norm": 2.201372285618325,
+      "learning_rate": 4.251153787872579e-06,
+      "loss": 0.4793,
+      "step": 1275
+    },
+    {
+      "epoch": 1.7812136712392466,
+      "grad_norm": 1.952612762150762,
+      "learning_rate": 4.2431202264479665e-06,
+      "loss": 0.4583,
+      "step": 1276
+    },
+    {
+      "epoch": 1.7826086956521738,
+      "grad_norm": 2.1107717354597244,
+      "learning_rate": 4.235088664253726e-06,
+      "loss": 0.5093,
+      "step": 1277
+    },
+    {
+      "epoch": 1.784003720065101,
+      "grad_norm": 2.2135031912589853,
+      "learning_rate": 4.227059122504514e-06,
+      "loss": 0.485,
+      "step": 1278
+    },
+    {
+      "epoch": 1.7853987444780284,
+      "grad_norm": 1.8768740585602934,
+      "learning_rate": 4.21903162240965e-06,
+      "loss": 0.4444,
+      "step": 1279
+    },
+    {
+      "epoch": 1.7867937688909556,
+      "grad_norm": 2.083417129855745,
+      "learning_rate": 4.211006185173056e-06,
+      "loss": 0.5084,
+      "step": 1280
+    },
+    {
+      "epoch": 1.788188793303883,
+      "grad_norm": 2.0108554229556574,
+      "learning_rate": 4.20298283199321e-06,
+      "loss": 0.4198,
+      "step": 1281
+    },
+    {
+      "epoch": 1.7895838177168102,
+      "grad_norm": 2.2544475269254916,
+      "learning_rate": 4.1949615840630845e-06,
+      "loss": 0.4629,
+      "step": 1282
+    },
+    {
+      "epoch": 1.7909788421297372,
+      "grad_norm": 2.1051705170049235,
+      "learning_rate": 4.186942462570087e-06,
+      "loss": 0.5034,
+      "step": 1283
+    },
+    {
+      "epoch": 1.7923738665426645,
+      "grad_norm": 2.160525988076939,
+      "learning_rate": 4.178925488696012e-06,
+      "loss": 0.4845,
+      "step": 1284
+    },
+    {
+      "epoch": 1.7937688909555918,
+      "grad_norm": 2.095025653388082,
+      "learning_rate": 4.170910683616985e-06,
+      "loss": 0.4656,
+      "step": 1285
+    },
+    {
+      "epoch": 1.7951639153685188,
+      "grad_norm": 2.277468807868547,
+      "learning_rate": 4.1628980685033914e-06,
+      "loss": 0.5152,
+      "step": 1286
+    },
+    {
+      "epoch": 1.796558939781446,
+      "grad_norm": 2.056705613970272,
+      "learning_rate": 4.154887664519842e-06,
+      "loss": 0.4893,
+      "step": 1287
+    },
+    {
+      "epoch": 1.7979539641943734,
+      "grad_norm": 2.1349968125546437,
+      "learning_rate": 4.1468794928251064e-06,
+      "loss": 0.4686,
+      "step": 1288
+    },
+    {
+      "epoch": 1.7993489886073006,
+      "grad_norm": 2.1952902085733914,
+      "learning_rate": 4.138873574572053e-06,
+      "loss": 0.4743,
+      "step": 1289
+    },
+    {
+      "epoch": 1.800744013020228,
+      "grad_norm": 2.0284310881562826,
+      "learning_rate": 4.130869930907599e-06,
+      "loss": 0.4169,
+      "step": 1290
+    },
+    {
+      "epoch": 1.8021390374331552,
+      "grad_norm": 1.9048197342872124,
+      "learning_rate": 4.122868582972659e-06,
+      "loss": 0.4679,
+      "step": 1291
+    },
+    {
+      "epoch": 1.8035340618460824,
+      "grad_norm": 2.0323882452817896,
+      "learning_rate": 4.114869551902075e-06,
+      "loss": 0.4279,
+      "step": 1292
+    },
+    {
+      "epoch": 1.8049290862590095,
+      "grad_norm": 2.0139222221685618,
+      "learning_rate": 4.106872858824576e-06,
+      "loss": 0.4539,
+      "step": 1293
+    },
+    {
+      "epoch": 1.8063241106719368,
+      "grad_norm": 2.0047655257894874,
+      "learning_rate": 4.098878524862715e-06,
+      "loss": 0.4979,
+      "step": 1294
+    },
+    {
+      "epoch": 1.807719135084864,
+      "grad_norm": 2.2420685216305314,
+      "learning_rate": 4.090886571132807e-06,
+      "loss": 0.566,
+      "step": 1295
+    },
+    {
+      "epoch": 1.809114159497791,
+      "grad_norm": 2.219670568258558,
+      "learning_rate": 4.082897018744887e-06,
+      "loss": 0.4951,
+      "step": 1296
+    },
+    {
+      "epoch": 1.8105091839107184,
+      "grad_norm": 2.339784196893198,
+      "learning_rate": 4.074909888802648e-06,
+      "loss": 0.505,
+      "step": 1297
+    },
+    {
+      "epoch": 1.8119042083236456,
+      "grad_norm": 1.8525635944473475,
+      "learning_rate": 4.066925202403374e-06,
+      "loss": 0.418,
+      "step": 1298
+    },
+    {
+      "epoch": 1.813299232736573,
+      "grad_norm": 2.097308855709139,
+      "learning_rate": 4.058942980637906e-06,
+      "loss": 0.4854,
+      "step": 1299
+    },
+    {
+      "epoch": 1.8146942571495002,
+      "grad_norm": 1.9765348057867853,
+      "learning_rate": 4.050963244590571e-06,
+      "loss": 0.4356,
+      "step": 1300
+    },
+    {
+      "epoch": 1.8160892815624274,
+      "grad_norm": 1.9813369058762067,
+      "learning_rate": 4.042986015339126e-06,
+      "loss": 0.4718,
+      "step": 1301
+    },
+    {
+      "epoch": 1.8174843059753547,
+      "grad_norm": 1.9999920964885178,
+      "learning_rate": 4.035011313954713e-06,
+      "loss": 0.4839,
+      "step": 1302
+    },
+    {
+      "epoch": 1.8188793303882818,
+      "grad_norm": 2.1800679736320103,
+      "learning_rate": 4.027039161501795e-06,
+      "loss": 0.4071,
+      "step": 1303
+    },
+    {
+      "epoch": 1.820274354801209,
+      "grad_norm": 1.9433494165364085,
+      "learning_rate": 4.019069579038096e-06,
+      "loss": 0.4594,
+      "step": 1304
+    },
+    {
+      "epoch": 1.8216693792141363,
+      "grad_norm": 1.9112976678661902,
+      "learning_rate": 4.011102587614563e-06,
+      "loss": 0.4272,
+      "step": 1305
+    },
+    {
+      "epoch": 1.8230644036270633,
+      "grad_norm": 2.0895025081439487,
+      "learning_rate": 4.00313820827529e-06,
+      "loss": 0.4914,
+      "step": 1306
+    },
+    {
+      "epoch": 1.8244594280399906,
+      "grad_norm": 2.2075749414041135,
+      "learning_rate": 3.995176462057473e-06,
+      "loss": 0.5239,
+      "step": 1307
+    },
+    {
+      "epoch": 1.8258544524529179,
+      "grad_norm": 2.2550877735540555,
+      "learning_rate": 3.987217369991357e-06,
+      "loss": 0.468,
+      "step": 1308
+    },
+    {
+      "epoch": 1.8272494768658452,
+      "grad_norm": 2.050163159425238,
+      "learning_rate": 3.979260953100169e-06,
+      "loss": 0.4714,
+      "step": 1309
+    },
+    {
+      "epoch": 1.8286445012787724,
+      "grad_norm": 2.0819553331959315,
+      "learning_rate": 3.97130723240008e-06,
+      "loss": 0.4548,
+      "step": 1310
+    },
+    {
+      "epoch": 1.8300395256916997,
+      "grad_norm": 2.414283571515306,
+      "learning_rate": 3.96335622890013e-06,
+      "loss": 0.4951,
+      "step": 1311
+    },
+    {
+      "epoch": 1.831434550104627,
+      "grad_norm": 1.6184625533605201,
+      "learning_rate": 3.955407963602184e-06,
+      "loss": 0.4758,
+      "step": 1312
+    },
+    {
+      "epoch": 1.832829574517554,
+      "grad_norm": 2.0008120841892967,
+      "learning_rate": 3.94746245750088e-06,
+      "loss": 0.4571,
+      "step": 1313
+    },
+    {
+      "epoch": 1.8342245989304813,
+      "grad_norm": 2.1349958964625815,
+      "learning_rate": 3.939519731583557e-06,
+      "loss": 0.4837,
+      "step": 1314
+    },
+    {
+      "epoch": 1.8356196233434086,
+      "grad_norm": 1.837343945362903,
+      "learning_rate": 3.9315798068302214e-06,
+      "loss": 0.438,
+      "step": 1315
+    },
+    {
+      "epoch": 1.8370146477563356,
+      "grad_norm": 1.8766035582552343,
+      "learning_rate": 3.923642704213475e-06,
+      "loss": 0.4936,
+      "step": 1316
+    },
+    {
+      "epoch": 1.8384096721692629,
+      "grad_norm": 2.0632480061696117,
+      "learning_rate": 3.915708444698465e-06,
+      "loss": 0.4904,
+      "step": 1317
+    },
+    {
+      "epoch": 1.8398046965821901,
+      "grad_norm": 2.0966458414991576,
+      "learning_rate": 3.907777049242828e-06,
+      "loss": 0.4587,
+      "step": 1318
+    },
+    {
+      "epoch": 1.8411997209951174,
+      "grad_norm": 2.0322570194266056,
+      "learning_rate": 3.899848538796643e-06,
+      "loss": 0.4788,
+      "step": 1319
+    },
+    {
+      "epoch": 1.8425947454080447,
+      "grad_norm": 2.000161074661116,
+      "learning_rate": 3.891922934302356e-06,
+      "loss": 0.4924,
+      "step": 1320
+    },
+    {
+      "epoch": 1.843989769820972,
+      "grad_norm": 2.1998938474565213,
+      "learning_rate": 3.884000256694749e-06,
+      "loss": 0.4836,
+      "step": 1321
+    },
+    {
+      "epoch": 1.8453847942338992,
+      "grad_norm": 2.1050792021780027,
+      "learning_rate": 3.876080526900867e-06,
+      "loss": 0.5069,
+      "step": 1322
+    },
+    {
+      "epoch": 1.8467798186468263,
+      "grad_norm": 2.1401256807615727,
+      "learning_rate": 3.868163765839966e-06,
+      "loss": 0.4742,
+      "step": 1323
+    },
+    {
+      "epoch": 1.8481748430597535,
+      "grad_norm": 1.9999712616159633,
+      "learning_rate": 3.860249994423467e-06,
+      "loss": 0.4724,
+      "step": 1324
+    },
+    {
+      "epoch": 1.8495698674726808,
+      "grad_norm": 2.212420210483622,
+      "learning_rate": 3.852339233554891e-06,
+      "loss": 0.4594,
+      "step": 1325
+    },
+    {
+      "epoch": 1.8509648918856079,
+      "grad_norm": 2.1246908494803236,
+      "learning_rate": 3.844431504129804e-06,
+      "loss": 0.5114,
+      "step": 1326
+    },
+    {
+      "epoch": 1.8523599162985351,
+      "grad_norm": 2.092568223347476,
+      "learning_rate": 3.8365268270357715e-06,
+      "loss": 0.4775,
+      "step": 1327
+    },
+    {
+      "epoch": 1.8537549407114624,
+      "grad_norm": 1.9869337488454075,
+      "learning_rate": 3.828625223152291e-06,
+      "loss": 0.4637,
+      "step": 1328
+    },
+    {
+      "epoch": 1.8551499651243897,
+      "grad_norm": 2.055635422327693,
+      "learning_rate": 3.820726713350742e-06,
+      "loss": 0.4614,
+      "step": 1329
+    },
+    {
+      "epoch": 1.856544989537317,
+      "grad_norm": 2.1873731783351396,
+      "learning_rate": 3.812831318494335e-06,
+      "loss": 0.4888,
+      "step": 1330
+    },
+    {
+      "epoch": 1.8579400139502442,
+      "grad_norm": 2.0003013010135255,
+      "learning_rate": 3.804939059438052e-06,
+      "loss": 0.4689,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8593350383631715,
+      "grad_norm": 2.071605299747796,
+      "learning_rate": 3.797049957028588e-06,
+      "loss": 0.476,
+      "step": 1332
+    },
+    {
+      "epoch": 1.8607300627760985,
+      "grad_norm": 2.0816926086751155,
+      "learning_rate": 3.7891640321043054e-06,
+      "loss": 0.4216,
+      "step": 1333
+    },
+    {
+      "epoch": 1.8621250871890258,
+      "grad_norm": 2.1529782419951857,
+      "learning_rate": 3.781281305495171e-06,
+      "loss": 0.421,
+      "step": 1334
+    },
+    {
+      "epoch": 1.863520111601953,
+      "grad_norm": 1.9862903369732532,
+      "learning_rate": 3.773401798022701e-06,
+      "loss": 0.4328,
+      "step": 1335
+    },
+    {
+      "epoch": 1.8649151360148801,
+      "grad_norm": 2.1563537251967455,
+      "learning_rate": 3.765525530499915e-06,
+      "loss": 0.4637,
+      "step": 1336
+    },
+    {
+      "epoch": 1.8663101604278074,
+      "grad_norm": 2.091076414635607,
+      "learning_rate": 3.757652523731269e-06,
+      "loss": 0.4169,
+      "step": 1337
+    },
+    {
+      "epoch": 1.8677051848407347,
+      "grad_norm": 1.9448368258697788,
+      "learning_rate": 3.7497827985126054e-06,
+      "loss": 0.4475,
+      "step": 1338
+    },
+    {
+      "epoch": 1.869100209253662,
+      "grad_norm": 2.024573917452205,
+      "learning_rate": 3.741916375631105e-06,
+      "loss": 0.4323,
+      "step": 1339
+    },
+    {
+      "epoch": 1.8704952336665892,
+      "grad_norm": 2.314125789311409,
+      "learning_rate": 3.7340532758652217e-06,
+      "loss": 0.5298,
+      "step": 1340
+    },
+    {
+      "epoch": 1.8718902580795165,
+      "grad_norm": 2.1177367118528614,
+      "learning_rate": 3.7261935199846266e-06,
+      "loss": 0.4449,
+      "step": 1341
+    },
+    {
+      "epoch": 1.8732852824924437,
+      "grad_norm": 2.262864442565798,
+      "learning_rate": 3.7183371287501684e-06,
+      "loss": 0.4593,
+      "step": 1342
+    },
+    {
+      "epoch": 1.8746803069053708,
+      "grad_norm": 1.971116485194465,
+      "learning_rate": 3.7104841229138034e-06,
+      "loss": 0.4833,
+      "step": 1343
+    },
+    {
+      "epoch": 1.876075331318298,
+      "grad_norm": 2.10454369475845,
+      "learning_rate": 3.7026345232185416e-06,
+      "loss": 0.438,
+      "step": 1344
+    },
+    {
+      "epoch": 1.8774703557312253,
+      "grad_norm": 1.9595998318710819,
+      "learning_rate": 3.6947883503984037e-06,
+      "loss": 0.4438,
+      "step": 1345
+    },
+    {
+      "epoch": 1.8788653801441524,
+      "grad_norm": 1.9520125987169128,
+      "learning_rate": 3.686945625178356e-06,
+      "loss": 0.4469,
+      "step": 1346
+    },
+    {
+      "epoch": 1.8802604045570797,
+      "grad_norm": 2.026901639450632,
+      "learning_rate": 3.6791063682742535e-06,
+      "loss": 0.4895,
+      "step": 1347
+    },
+    {
+      "epoch": 1.881655428970007,
+      "grad_norm": 2.048549645209667,
+      "learning_rate": 3.6712706003927937e-06,
+      "loss": 0.4631,
+      "step": 1348
+    },
+    {
+      "epoch": 1.8830504533829342,
+      "grad_norm": 2.0279995779597932,
+      "learning_rate": 3.6634383422314622e-06,
+      "loss": 0.442,
+      "step": 1349
+    },
+    {
+      "epoch": 1.8844454777958615,
+      "grad_norm": 2.1434290748836,
+      "learning_rate": 3.655609614478467e-06,
+      "loss": 0.5072,
+      "step": 1350
+    },
+    {
+      "epoch": 1.8858405022087887,
+      "grad_norm": 2.222887686762255,
+      "learning_rate": 3.647784437812693e-06,
+      "loss": 0.49,
+      "step": 1351
+    },
+    {
+      "epoch": 1.887235526621716,
+      "grad_norm": 2.2009445876073253,
+      "learning_rate": 3.6399628329036496e-06,
+      "loss": 0.4928,
+      "step": 1352
+    },
+    {
+      "epoch": 1.888630551034643,
+      "grad_norm": 2.174116284837636,
+      "learning_rate": 3.632144820411405e-06,
+      "loss": 0.3996,
+      "step": 1353
+    },
+    {
+      "epoch": 1.8900255754475703,
+      "grad_norm": 1.8585093779164292,
+      "learning_rate": 3.624330420986541e-06,
+      "loss": 0.4241,
+      "step": 1354
+    },
+    {
+      "epoch": 1.8914205998604976,
+      "grad_norm": 1.9457784843844574,
+      "learning_rate": 3.6165196552701e-06,
+      "loss": 0.4694,
+      "step": 1355
+    },
+    {
+      "epoch": 1.8928156242734246,
+      "grad_norm": 2.00270102931314,
+      "learning_rate": 3.6087125438935187e-06,
+      "loss": 0.4603,
+      "step": 1356
+    },
+    {
+      "epoch": 1.894210648686352,
+      "grad_norm": 2.0737253073991853,
+      "learning_rate": 3.6009091074785853e-06,
+      "loss": 0.461,
+      "step": 1357
+    },
+    {
+      "epoch": 1.8956056730992792,
+      "grad_norm": 2.064571146406589,
+      "learning_rate": 3.5931093666373845e-06,
+      "loss": 0.5147,
+      "step": 1358
+    },
+    {
+      "epoch": 1.8970006975122065,
+      "grad_norm": 2.1201036087247283,
+      "learning_rate": 3.585313341972232e-06,
+      "loss": 0.4543,
+      "step": 1359
+    },
+    {
+      "epoch": 1.8983957219251337,
+      "grad_norm": 2.080966829655726,
+      "learning_rate": 3.577521054075631e-06,
+      "loss": 0.4277,
+      "step": 1360
+    },
+    {
+      "epoch": 1.899790746338061,
+      "grad_norm": 2.0981462870140097,
+      "learning_rate": 3.5697325235302183e-06,
+      "loss": 0.4193,
+      "step": 1361
+    },
+    {
+      "epoch": 1.9011857707509883,
+      "grad_norm": 2.0868066804557257,
+      "learning_rate": 3.5619477709086982e-06,
+      "loss": 0.4302,
+      "step": 1362
+    },
+    {
+      "epoch": 1.9025807951639153,
+      "grad_norm": 2.1039057194297492,
+      "learning_rate": 3.5541668167738003e-06,
+      "loss": 0.5433,
+      "step": 1363
+    },
+    {
+      "epoch": 1.9039758195768426,
+      "grad_norm": 2.4297797676999,
+      "learning_rate": 3.546389681678224e-06,
+      "loss": 0.4555,
+      "step": 1364
+    },
+    {
+      "epoch": 1.9053708439897699,
+      "grad_norm": 1.9533632396939948,
+      "learning_rate": 3.538616386164575e-06,
+      "loss": 0.4645,
+      "step": 1365
+    },
+    {
+      "epoch": 1.906765868402697,
+      "grad_norm": 2.050877909307051,
+      "learning_rate": 3.530846950765318e-06,
+      "loss": 0.4768,
+      "step": 1366
+    },
+    {
+      "epoch": 1.9081608928156242,
+      "grad_norm": 2.1111581143101943,
+      "learning_rate": 3.5230813960027275e-06,
+      "loss": 0.4791,
+      "step": 1367
+    },
+    {
+      "epoch": 1.9095559172285514,
+      "grad_norm": 2.266748524269278,
+      "learning_rate": 3.5153197423888206e-06,
+      "loss": 0.4986,
+      "step": 1368
+    },
+    {
+      "epoch": 1.9109509416414787,
+      "grad_norm": 1.950535419944206,
+      "learning_rate": 3.5075620104253123e-06,
+      "loss": 0.4389,
+      "step": 1369
+    },
+    {
+      "epoch": 1.912345966054406,
+      "grad_norm": 2.0576812430591804,
+      "learning_rate": 3.4998082206035606e-06,
+      "loss": 0.4473,
+      "step": 1370
+    },
+    {
+      "epoch": 1.9137409904673333,
+      "grad_norm": 1.825998507257751,
+      "learning_rate": 3.492058393404509e-06,
+      "loss": 0.51,
+      "step": 1371
+    },
+    {
+      "epoch": 1.9151360148802605,
+      "grad_norm": 2.085578235101986,
+      "learning_rate": 3.4843125492986345e-06,
+      "loss": 0.4212,
+      "step": 1372
+    },
+    {
+      "epoch": 1.9165310392931876,
+      "grad_norm": 1.9189758160127117,
+      "learning_rate": 3.4765707087458912e-06,
+      "loss": 0.4944,
+      "step": 1373
+    },
+    {
+      "epoch": 1.9179260637061148,
+      "grad_norm": 2.113280326915886,
+      "learning_rate": 3.468832892195664e-06,
+      "loss": 0.4991,
+      "step": 1374
+    },
+    {
+      "epoch": 1.9193210881190421,
+      "grad_norm": 2.2202722381166025,
+      "learning_rate": 3.4610991200867006e-06,
+      "loss": 0.4734,
+      "step": 1375
+    },
+    {
+      "epoch": 1.9207161125319692,
+      "grad_norm": 1.9939976073920789,
+      "learning_rate": 3.453369412847071e-06,
+      "loss": 0.4053,
+      "step": 1376
+    },
+    {
+      "epoch": 1.9221111369448964,
+      "grad_norm": 2.0479827879447057,
+      "learning_rate": 3.445643790894109e-06,
+      "loss": 0.4436,
+      "step": 1377
+    },
+    {
+      "epoch": 1.9235061613578237,
+      "grad_norm": 2.1023765087199338,
+      "learning_rate": 3.4379222746343534e-06,
+      "loss": 0.4684,
+      "step": 1378
+    },
+    {
+      "epoch": 1.924901185770751,
+      "grad_norm": 2.0722694658703236,
+      "learning_rate": 3.4302048844634995e-06,
+      "loss": 0.4759,
+      "step": 1379
+    },
+    {
+      "epoch": 1.9262962101836782,
+      "grad_norm": 2.0161034662991693,
+      "learning_rate": 3.4224916407663484e-06,
+      "loss": 0.4283,
+      "step": 1380
+    },
+    {
+      "epoch": 1.9276912345966055,
+      "grad_norm": 2.114718900925972,
+      "learning_rate": 3.414782563916742e-06,
+      "loss": 0.5011,
+      "step": 1381
+    },
+    {
+      "epoch": 1.9290862590095328,
+      "grad_norm": 2.1968385042774283,
+      "learning_rate": 3.407077674277518e-06,
+      "loss": 0.4709,
+      "step": 1382
+    },
+    {
+      "epoch": 1.93048128342246,
+      "grad_norm": 2.2136573594525535,
+      "learning_rate": 3.3993769922004584e-06,
+      "loss": 0.5332,
+      "step": 1383
+    },
+    {
+      "epoch": 1.931876307835387,
+      "grad_norm": 2.143714300983868,
+      "learning_rate": 3.391680538026224e-06,
+      "loss": 0.4267,
+      "step": 1384
+    },
+    {
+      "epoch": 1.9332713322483144,
+      "grad_norm": 2.069489588319107,
+      "learning_rate": 3.3839883320843125e-06,
+      "loss": 0.4491,
+      "step": 1385
+    },
+    {
+      "epoch": 1.9346663566612414,
+      "grad_norm": 1.9452039632794036,
+      "learning_rate": 3.3763003946930023e-06,
+      "loss": 0.4664,
+      "step": 1386
+    },
+    {
+      "epoch": 1.9360613810741687,
+      "grad_norm": 2.189405225206946,
+      "learning_rate": 3.36861674615929e-06,
+      "loss": 0.4569,
+      "step": 1387
+    },
+    {
+      "epoch": 1.937456405487096,
+      "grad_norm": 2.220840979892466,
+      "learning_rate": 3.360937406778849e-06,
+      "loss": 0.5108,
+      "step": 1388
+    },
+    {
+      "epoch": 1.9388514299000232,
+      "grad_norm": 2.1295737957858423,
+      "learning_rate": 3.35326239683597e-06,
+      "loss": 0.4236,
+      "step": 1389
+    },
+    {
+      "epoch": 1.9402464543129505,
+      "grad_norm": 2.048196278694426,
+      "learning_rate": 3.3455917366035058e-06,
+      "loss": 0.4695,
+      "step": 1390
+    },
+    {
+      "epoch": 1.9416414787258778,
+      "grad_norm": 2.082360128476303,
+      "learning_rate": 3.337925446342819e-06,
+      "loss": 0.4389,
+      "step": 1391
+    },
+    {
+      "epoch": 1.943036503138805,
+      "grad_norm": 1.8747651704705666,
+      "learning_rate": 3.3302635463037352e-06,
+      "loss": 0.3972,
+      "step": 1392
+    },
+    {
+      "epoch": 1.9444315275517323,
+      "grad_norm": 2.048555484081868,
+      "learning_rate": 3.3226060567244767e-06,
+      "loss": 0.4803,
+      "step": 1393
+    },
+    {
+      "epoch": 1.9458265519646594,
+      "grad_norm": 2.136883881979085,
+      "learning_rate": 3.314952997831618e-06,
+      "loss": 0.4223,
+      "step": 1394
+    },
+    {
+      "epoch": 1.9472215763775866,
+      "grad_norm": 1.7874295552797317,
+      "learning_rate": 3.307304389840036e-06,
+      "loss": 0.4322,
+      "step": 1395
+    },
+    {
+      "epoch": 1.9486166007905137,
+      "grad_norm": 2.0438982119199465,
+      "learning_rate": 3.29966025295284e-06,
+      "loss": 0.4468,
+      "step": 1396
+    },
+    {
+      "epoch": 1.950011625203441,
+      "grad_norm": 2.1099373836087105,
+      "learning_rate": 3.292020607361337e-06,
+      "loss": 0.5667,
+      "step": 1397
+    },
+    {
+      "epoch": 1.9514066496163682,
+      "grad_norm": 2.1250544635117516,
+      "learning_rate": 3.284385473244974e-06,
+      "loss": 0.4201,
+      "step": 1398
+    },
+    {
+      "epoch": 1.9528016740292955,
+      "grad_norm": 2.0986308513899163,
+      "learning_rate": 3.2767548707712693e-06,
+      "loss": 0.4825,
+      "step": 1399
+    },
+    {
+      "epoch": 1.9541966984422228,
+      "grad_norm": 2.1463740035501977,
+      "learning_rate": 3.2691288200957826e-06,
+      "loss": 0.4821,
+      "step": 1400
+    },
+    {
+      "epoch": 1.95559172285515,
+      "grad_norm": 2.208153989617607,
+      "learning_rate": 3.2615073413620467e-06,
+      "loss": 0.5005,
+      "step": 1401
+    },
+    {
+      "epoch": 1.9569867472680773,
+      "grad_norm": 2.0747031690202524,
+      "learning_rate": 3.2538904547015137e-06,
+      "loss": 0.4718,
+      "step": 1402
+    },
+    {
+      "epoch": 1.9583817716810046,
+      "grad_norm": 2.058110546374091,
+      "learning_rate": 3.2462781802335124e-06,
+      "loss": 0.4699,
+      "step": 1403
+    },
+    {
+      "epoch": 1.9597767960939316,
+      "grad_norm": 2.187911351474957,
+      "learning_rate": 3.2386705380651877e-06,
+      "loss": 0.4686,
+      "step": 1404
+    },
+    {
+      "epoch": 1.961171820506859,
+      "grad_norm": 1.8656819381405738,
+      "learning_rate": 3.2310675482914444e-06,
+      "loss": 0.4688,
+      "step": 1405
+    },
+    {
+      "epoch": 1.962566844919786,
+      "grad_norm": 2.0876329165242047,
+      "learning_rate": 3.2234692309949034e-06,
+      "loss": 0.4658,
+      "step": 1406
+    },
+    {
+      "epoch": 1.9639618693327132,
+      "grad_norm": 2.0606274699365943,
+      "learning_rate": 3.2158756062458422e-06,
+      "loss": 0.4721,
+      "step": 1407
+    },
+    {
+      "epoch": 1.9653568937456405,
+      "grad_norm": 2.030107454192887,
+      "learning_rate": 3.208286694102141e-06,
+      "loss": 0.502,
+      "step": 1408
+    },
+    {
+      "epoch": 1.9667519181585678,
+      "grad_norm": 2.16972444150237,
+      "learning_rate": 3.2007025146092345e-06,
+      "loss": 0.4383,
+      "step": 1409
+    },
+    {
+      "epoch": 1.968146942571495,
+      "grad_norm": 1.8504090641131754,
+      "learning_rate": 3.1931230878000586e-06,
+      "loss": 0.4498,
+      "step": 1410
+    },
+    {
+      "epoch": 1.9695419669844223,
+      "grad_norm": 2.0621064144305143,
+      "learning_rate": 3.1855484336949876e-06,
+      "loss": 0.4547,
+      "step": 1411
+    },
+    {
+      "epoch": 1.9709369913973496,
+      "grad_norm": 1.9346127805447864,
+      "learning_rate": 3.1779785723017988e-06,
+      "loss": 0.5044,
+      "step": 1412
+    },
+    {
+      "epoch": 1.9723320158102768,
+      "grad_norm": 2.093246509824468,
+      "learning_rate": 3.170413523615605e-06,
+      "loss": 0.4427,
+      "step": 1413
+    },
+    {
+      "epoch": 1.9737270402232039,
+      "grad_norm": 2.1928383212100453,
+      "learning_rate": 3.162853307618805e-06,
+      "loss": 0.3999,
+      "step": 1414
+    },
+    {
+      "epoch": 1.9751220646361312,
+      "grad_norm": 1.8107419503015638,
+      "learning_rate": 3.155297944281036e-06,
+      "loss": 0.4503,
+      "step": 1415
+    },
+    {
+      "epoch": 1.9765170890490582,
+      "grad_norm": 2.08654083770065,
+      "learning_rate": 3.1477474535591167e-06,
+      "loss": 0.4615,
+      "step": 1416
+    },
+    {
+      "epoch": 1.9779121134619855,
+      "grad_norm": 2.0894026454200834,
+      "learning_rate": 3.1402018553969917e-06,
+      "loss": 0.5014,
+      "step": 1417
+    },
+    {
+      "epoch": 1.9793071378749127,
+      "grad_norm": 2.0296124204690487,
+      "learning_rate": 3.132661169725688e-06,
+      "loss": 0.4736,
+      "step": 1418
+    },
+    {
+      "epoch": 1.98070216228784,
+      "grad_norm": 1.9519281154170998,
+      "learning_rate": 3.125125416463254e-06,
+      "loss": 0.4329,
+      "step": 1419
+    },
+    {
+      "epoch": 1.9820971867007673,
+      "grad_norm": 2.1011005867616337,
+      "learning_rate": 3.1175946155147064e-06,
+      "loss": 0.4609,
+      "step": 1420
+    },
+    {
+      "epoch": 1.9834922111136946,
+      "grad_norm": 2.3096209237137524,
+      "learning_rate": 3.110068786771987e-06,
+      "loss": 0.4718,
+      "step": 1421
+    },
+    {
+      "epoch": 1.9848872355266218,
+      "grad_norm": 1.7175842413795206,
+      "learning_rate": 3.1025479501139e-06,
+      "loss": 0.4926,
+      "step": 1422
+    },
+    {
+      "epoch": 1.986282259939549,
+      "grad_norm": 2.206831702580605,
+      "learning_rate": 3.095032125406062e-06,
+      "loss": 0.4488,
+      "step": 1423
+    },
+    {
+      "epoch": 1.9876772843524761,
+      "grad_norm": 2.0869072404586597,
+      "learning_rate": 3.0875213325008548e-06,
+      "loss": 0.5062,
+      "step": 1424
+    },
+    {
+      "epoch": 1.9890723087654034,
+      "grad_norm": 2.1927222257464085,
+      "learning_rate": 3.0800155912373696e-06,
+      "loss": 0.4717,
+      "step": 1425
+    },
+    {
+      "epoch": 1.9904673331783305,
+      "grad_norm": 2.175734397835776,
+      "learning_rate": 3.0725149214413487e-06,
+      "loss": 0.4203,
+      "step": 1426
+    },
+    {
+      "epoch": 1.9918623575912577,
+      "grad_norm": 1.9894712393139962,
+      "learning_rate": 3.065019342925143e-06,
+      "loss": 0.4895,
+      "step": 1427
+    },
+    {
+      "epoch": 1.993257382004185,
+      "grad_norm": 2.2695006617555653,
+      "learning_rate": 3.0575288754876565e-06,
+      "loss": 0.478,
+      "step": 1428
+    },
+    {
+      "epoch": 1.9946524064171123,
+      "grad_norm": 1.8870689048595921,
+      "learning_rate": 3.0500435389142867e-06,
+      "loss": 0.4129,
+      "step": 1429
+    },
+    {
+      "epoch": 1.9960474308300395,
+      "grad_norm": 2.1967720177098946,
+      "learning_rate": 3.042563352976884e-06,
+      "loss": 0.456,
+      "step": 1430
+    },
+    {
+      "epoch": 1.9974424552429668,
+      "grad_norm": 2.0420373038544866,
+      "learning_rate": 3.035088337433694e-06,
+      "loss": 0.4766,
+      "step": 1431
+    },
+    {
+      "epoch": 1.998837479655894,
+      "grad_norm": 2.112347205511411,
+      "learning_rate": 3.0276185120292996e-06,
+      "loss": 0.4888,
+      "step": 1432
+    },
+    {
+      "epoch": 2.0013950244129273,
+      "grad_norm": 2.1052085745081897,
+      "learning_rate": 3.0201538964945787e-06,
+      "loss": 0.7158,
+      "step": 1433
+    },
+    {
+      "epoch": 2.0027900488258545,
+      "grad_norm": 1.9402636424171937,
+      "learning_rate": 3.0126945105466486e-06,
+      "loss": 0.2674,
+      "step": 1434
+    },
+    {
+      "epoch": 2.004185073238782,
+      "grad_norm": 1.8085458724472494,
+      "learning_rate": 3.005240373888812e-06,
+      "loss": 0.2916,
+      "step": 1435
+    },
+    {
+      "epoch": 2.005580097651709,
+      "grad_norm": 1.9029597036365433,
+      "learning_rate": 2.9977915062105023e-06,
+      "loss": 0.2669,
+      "step": 1436
+    },
+    {
+      "epoch": 2.0069751220646364,
+      "grad_norm": 1.7044125918783548,
+      "learning_rate": 2.9903479271872416e-06,
+      "loss": 0.2359,
+      "step": 1437
+    },
+    {
+      "epoch": 2.008370146477563,
+      "grad_norm": 1.9190122425840221,
+      "learning_rate": 2.9829096564805804e-06,
+      "loss": 0.2301,
+      "step": 1438
+    },
+    {
+      "epoch": 2.0097651708904904,
+      "grad_norm": 1.7303744404988668,
+      "learning_rate": 2.975476713738043e-06,
+      "loss": 0.2489,
+      "step": 1439
+    },
+    {
+      "epoch": 2.0111601953034177,
+      "grad_norm": 1.721857219696397,
+      "learning_rate": 2.9680491185930877e-06,
+      "loss": 0.2564,
+      "step": 1440
+    },
+    {
+      "epoch": 2.012555219716345,
+      "grad_norm": 1.894684087795828,
+      "learning_rate": 2.960626890665044e-06,
+      "loss": 0.2444,
+      "step": 1441
+    },
+    {
+      "epoch": 2.0139502441292723,
+      "grad_norm": 1.839486201760737,
+      "learning_rate": 2.953210049559062e-06,
+      "loss": 0.2157,
+      "step": 1442
+    },
+    {
+      "epoch": 2.0153452685421995,
+      "grad_norm": 2.4897922879317425,
+      "learning_rate": 2.945798614866068e-06,
+      "loss": 0.2716,
+      "step": 1443
+    },
+    {
+      "epoch": 2.016740292955127,
+      "grad_norm": 2.344531912366505,
+      "learning_rate": 2.9383926061627055e-06,
+      "loss": 0.239,
+      "step": 1444
+    },
+    {
+      "epoch": 2.018135317368054,
+      "grad_norm": 2.914955661673377,
+      "learning_rate": 2.9309920430112825e-06,
+      "loss": 0.2717,
+      "step": 1445
+    },
+    {
+      "epoch": 2.0195303417809813,
+      "grad_norm": 2.498308889226278,
+      "learning_rate": 2.92359694495973e-06,
+      "loss": 0.2606,
+      "step": 1446
+    },
+    {
+      "epoch": 2.0209253661939086,
+      "grad_norm": 2.1050902414844015,
+      "learning_rate": 2.9162073315415384e-06,
+      "loss": 0.2375,
+      "step": 1447
+    },
+    {
+      "epoch": 2.0223203906068354,
+      "grad_norm": 2.456539411273798,
+      "learning_rate": 2.9088232222757085e-06,
+      "loss": 0.2524,
+      "step": 1448
+    },
+    {
+      "epoch": 2.0237154150197627,
+      "grad_norm": 2.1469646592078653,
+      "learning_rate": 2.9014446366667115e-06,
+      "loss": 0.2684,
+      "step": 1449
+    },
+    {
+      "epoch": 2.02511043943269,
+      "grad_norm": 2.0918990363899552,
+      "learning_rate": 2.8940715942044204e-06,
+      "loss": 0.2578,
+      "step": 1450
+    },
+    {
+      "epoch": 2.0265054638456172,
+      "grad_norm": 1.9929495750968211,
+      "learning_rate": 2.8867041143640663e-06,
+      "loss": 0.2332,
+      "step": 1451
+    },
+    {
+      "epoch": 2.0279004882585445,
+      "grad_norm": 2.139820882014351,
+      "learning_rate": 2.8793422166061918e-06,
+      "loss": 0.2801,
+      "step": 1452
+    },
+    {
+      "epoch": 2.029295512671472,
+      "grad_norm": 1.9416468277806427,
+      "learning_rate": 2.8719859203765955e-06,
+      "loss": 0.227,
+      "step": 1453
+    },
+    {
+      "epoch": 2.030690537084399,
+      "grad_norm": 1.9592248880362728,
+      "learning_rate": 2.864635245106272e-06,
+      "loss": 0.2398,
+      "step": 1454
+    },
+    {
+      "epoch": 2.0320855614973263,
+      "grad_norm": 1.942022515983665,
+      "learning_rate": 2.8572902102113788e-06,
+      "loss": 0.2338,
+      "step": 1455
+    },
+    {
+      "epoch": 2.0334805859102536,
+      "grad_norm": 2.0314568308212535,
+      "learning_rate": 2.849950835093168e-06,
+      "loss": 0.2413,
+      "step": 1456
+    },
+    {
+      "epoch": 2.034875610323181,
+      "grad_norm": 1.8345231386774408,
+      "learning_rate": 2.8426171391379433e-06,
+      "loss": 0.2498,
+      "step": 1457
+    },
+    {
+      "epoch": 2.0362706347361077,
+      "grad_norm": 1.73045801685932,
+      "learning_rate": 2.835289141717008e-06,
+      "loss": 0.2173,
+      "step": 1458
+    },
+    {
+      "epoch": 2.037665659149035,
+      "grad_norm": 1.9676717354052715,
+      "learning_rate": 2.827966862186616e-06,
+      "loss": 0.2358,
+      "step": 1459
+    },
+    {
+      "epoch": 2.0390606835619622,
+      "grad_norm": 1.9796197987643156,
+      "learning_rate": 2.820650319887911e-06,
+      "loss": 0.2414,
+      "step": 1460
+    },
+    {
+      "epoch": 2.0404557079748895,
+      "grad_norm": 2.0928848682799384,
+      "learning_rate": 2.8133395341468915e-06,
+      "loss": 0.2527,
+      "step": 1461
+    },
+    {
+      "epoch": 2.041850732387817,
+      "grad_norm": 2.044176118101689,
+      "learning_rate": 2.8060345242743427e-06,
+      "loss": 0.2319,
+      "step": 1462
+    },
+    {
+      "epoch": 2.043245756800744,
+      "grad_norm": 1.8085526333410318,
+      "learning_rate": 2.7987353095657944e-06,
+      "loss": 0.2663,
+      "step": 1463
+    },
+    {
+      "epoch": 2.0446407812136713,
+      "grad_norm": 2.1890698281590772,
+      "learning_rate": 2.7914419093014734e-06,
+      "loss": 0.2424,
+      "step": 1464
+    },
+    {
+      "epoch": 2.0460358056265986,
+      "grad_norm": 1.8839951262325991,
+      "learning_rate": 2.784154342746246e-06,
+      "loss": 0.2453,
+      "step": 1465
+    },
+    {
+      "epoch": 2.047430830039526,
+      "grad_norm": 2.1446428182352024,
+      "learning_rate": 2.7768726291495667e-06,
+      "loss": 0.2616,
+      "step": 1466
+    },
+    {
+      "epoch": 2.048825854452453,
+      "grad_norm": 2.12053100151656,
+      "learning_rate": 2.7695967877454356e-06,
+      "loss": 0.2628,
+      "step": 1467
+    },
+    {
+      "epoch": 2.05022087886538,
+      "grad_norm": 2.010891601213425,
+      "learning_rate": 2.7623268377523356e-06,
+      "loss": 0.2103,
+      "step": 1468
+    },
+    {
+      "epoch": 2.0516159032783072,
+      "grad_norm": 1.9005154237356725,
+      "learning_rate": 2.755062798373189e-06,
+      "loss": 0.2456,
+      "step": 1469
+    },
+    {
+      "epoch": 2.0530109276912345,
+      "grad_norm": 2.119569304257289,
+      "learning_rate": 2.747804688795311e-06,
+      "loss": 0.2576,
+      "step": 1470
+    },
+    {
+      "epoch": 2.0544059521041618,
+      "grad_norm": 1.9846330923110018,
+      "learning_rate": 2.7405525281903506e-06,
+      "loss": 0.2108,
+      "step": 1471
+    },
+    {
+      "epoch": 2.055800976517089,
+      "grad_norm": 2.0242598137098358,
+      "learning_rate": 2.7333063357142414e-06,
+      "loss": 0.2376,
+      "step": 1472
+    },
+    {
+      "epoch": 2.0571960009300163,
+      "grad_norm": 2.065760746797269,
+      "learning_rate": 2.7260661305071523e-06,
+      "loss": 0.2202,
+      "step": 1473
+    },
+    {
+      "epoch": 2.0585910253429436,
+      "grad_norm": 1.9897327351088137,
+      "learning_rate": 2.718831931693443e-06,
+      "loss": 0.2508,
+      "step": 1474
+    },
+    {
+      "epoch": 2.059986049755871,
+      "grad_norm": 1.9011348442504956,
+      "learning_rate": 2.7116037583816e-06,
+      "loss": 0.2504,
+      "step": 1475
+    },
+    {
+      "epoch": 2.061381074168798,
+      "grad_norm": 2.054795671413816,
+      "learning_rate": 2.7043816296642005e-06,
+      "loss": 0.2314,
+      "step": 1476
+    },
+    {
+      "epoch": 2.0627760985817254,
+      "grad_norm": 2.1098702197687156,
+      "learning_rate": 2.6971655646178544e-06,
+      "loss": 0.2603,
+      "step": 1477
+    },
+    {
+      "epoch": 2.064171122994652,
+      "grad_norm": 1.88931900817361,
+      "learning_rate": 2.689955582303152e-06,
+      "loss": 0.2164,
+      "step": 1478
+    },
+    {
+      "epoch": 2.0655661474075795,
+      "grad_norm": 1.8274028969126306,
+      "learning_rate": 2.6827517017646154e-06,
+      "loss": 0.2246,
+      "step": 1479
+    },
+    {
+      "epoch": 2.0669611718205068,
+      "grad_norm": 2.097344575244166,
+      "learning_rate": 2.6755539420306565e-06,
+      "loss": 0.2263,
+      "step": 1480
+    },
+    {
+      "epoch": 2.068356196233434,
+      "grad_norm": 1.6940391063154205,
+      "learning_rate": 2.668362322113512e-06,
+      "loss": 0.2429,
+      "step": 1481
+    },
+    {
+      "epoch": 2.0697512206463613,
+      "grad_norm": 1.8672184430662262,
+      "learning_rate": 2.661176861009205e-06,
+      "loss": 0.2409,
+      "step": 1482
+    },
+    {
+      "epoch": 2.0711462450592886,
+      "grad_norm": 2.0268800259192523,
+      "learning_rate": 2.6539975776974926e-06,
+      "loss": 0.2332,
+      "step": 1483
+    },
+    {
+      "epoch": 2.072541269472216,
+      "grad_norm": 1.9571704604901023,
+      "learning_rate": 2.646824491141807e-06,
+      "loss": 0.2227,
+      "step": 1484
+    },
+    {
+      "epoch": 2.073936293885143,
+      "grad_norm": 1.9668513577356352,
+      "learning_rate": 2.6396576202892176e-06,
+      "loss": 0.2221,
+      "step": 1485
+    },
+    {
+      "epoch": 2.0753313182980704,
+      "grad_norm": 1.8267652821156959,
+      "learning_rate": 2.632496984070375e-06,
+      "loss": 0.2181,
+      "step": 1486
+    },
+    {
+      "epoch": 2.0767263427109977,
+      "grad_norm": 2.0560708153439533,
+      "learning_rate": 2.6253426013994586e-06,
+      "loss": 0.2557,
+      "step": 1487
+    },
+    {
+      "epoch": 2.0781213671239245,
+      "grad_norm": 2.17511504619303,
+      "learning_rate": 2.6181944911741333e-06,
+      "loss": 0.2532,
+      "step": 1488
+    },
+    {
+      "epoch": 2.0795163915368517,
+      "grad_norm": 2.094401377084581,
+      "learning_rate": 2.6110526722754955e-06,
+      "loss": 0.2683,
+      "step": 1489
+    },
+    {
+      "epoch": 2.080911415949779,
+      "grad_norm": 2.175178881238892,
+      "learning_rate": 2.603917163568021e-06,
+      "loss": 0.2466,
+      "step": 1490
+    },
+    {
+      "epoch": 2.0823064403627063,
+      "grad_norm": 1.882430762960756,
+      "learning_rate": 2.5967879838995176e-06,
+      "loss": 0.2261,
+      "step": 1491
+    },
+    {
+      "epoch": 2.0837014647756336,
+      "grad_norm": 2.138876118948161,
+      "learning_rate": 2.589665152101081e-06,
+      "loss": 0.2428,
+      "step": 1492
+    },
+    {
+      "epoch": 2.085096489188561,
+      "grad_norm": 1.9993127285312577,
+      "learning_rate": 2.582548686987031e-06,
+      "loss": 0.2349,
+      "step": 1493
+    },
+    {
+      "epoch": 2.086491513601488,
+      "grad_norm": 1.9705015030533575,
+      "learning_rate": 2.5754386073548775e-06,
+      "loss": 0.2198,
+      "step": 1494
+    },
+    {
+      "epoch": 2.0878865380144154,
+      "grad_norm": 2.032583776330532,
+      "learning_rate": 2.5683349319852647e-06,
+      "loss": 0.2423,
+      "step": 1495
+    },
+    {
+      "epoch": 2.0892815624273426,
+      "grad_norm": 1.9542879134817772,
+      "learning_rate": 2.5612376796419126e-06,
+      "loss": 0.242,
+      "step": 1496
+    },
+    {
+      "epoch": 2.09067658684027,
+      "grad_norm": 2.098429761287188,
+      "learning_rate": 2.5541468690715797e-06,
+      "loss": 0.2491,
+      "step": 1497
+    },
+    {
+      "epoch": 2.0920716112531967,
+      "grad_norm": 2.0326652718730287,
+      "learning_rate": 2.5470625190040105e-06,
+      "loss": 0.2499,
+      "step": 1498
+    },
+    {
+      "epoch": 2.093466635666124,
+      "grad_norm": 2.105130932551734,
+      "learning_rate": 2.5399846481518857e-06,
+      "loss": 0.2501,
+      "step": 1499
+    },
+    {
+      "epoch": 2.0948616600790513,
+      "grad_norm": 1.9278156061938734,
+      "learning_rate": 2.5329132752107675e-06,
+      "loss": 0.2606,
+      "step": 1500
+    },
+    {
+      "epoch": 2.0962566844919786,
+      "grad_norm": 2.0731891083929472,
+      "learning_rate": 2.525848418859055e-06,
+      "loss": 0.2497,
+      "step": 1501
+    },
+    {
+      "epoch": 2.097651708904906,
+      "grad_norm": 1.7753623857882452,
+      "learning_rate": 2.518790097757938e-06,
+      "loss": 0.2183,
+      "step": 1502
+    },
+    {
+      "epoch": 2.099046733317833,
+      "grad_norm": 1.8816805999317845,
+      "learning_rate": 2.51173833055134e-06,
+      "loss": 0.223,
+      "step": 1503
+    },
+    {
+      "epoch": 2.1004417577307604,
+      "grad_norm": 2.050127188663919,
+      "learning_rate": 2.504693135865875e-06,
+      "loss": 0.2604,
+      "step": 1504
+    },
+    {
+      "epoch": 2.1018367821436876,
+      "grad_norm": 2.068754384900015,
+      "learning_rate": 2.497654532310799e-06,
+      "loss": 0.2444,
+      "step": 1505
+    },
+    {
+      "epoch": 2.103231806556615,
+      "grad_norm": 1.8837169365636035,
+      "learning_rate": 2.490622538477952e-06,
+      "loss": 0.2445,
+      "step": 1506
+    },
+    {
+      "epoch": 2.104626830969542,
+      "grad_norm": 2.1180481981544097,
+      "learning_rate": 2.483597172941718e-06,
+      "loss": 0.2346,
+      "step": 1507
+    },
+    {
+      "epoch": 2.106021855382469,
+      "grad_norm": 2.0748715550357946,
+      "learning_rate": 2.4765784542589754e-06,
+      "loss": 0.2497,
+      "step": 1508
+    },
+    {
+      "epoch": 2.1074168797953963,
+      "grad_norm": 2.0079446467193995,
+      "learning_rate": 2.46956640096904e-06,
+      "loss": 0.2032,
+      "step": 1509
+    },
+    {
+      "epoch": 2.1088119042083235,
+      "grad_norm": 1.8790214788210842,
+      "learning_rate": 2.4625610315936267e-06,
+      "loss": 0.245,
+      "step": 1510
+    },
+    {
+      "epoch": 2.110206928621251,
+      "grad_norm": 2.1128877451193864,
+      "learning_rate": 2.4555623646367952e-06,
+      "loss": 0.2778,
+      "step": 1511
+    },
+    {
+      "epoch": 2.111601953034178,
+      "grad_norm": 2.1529575803923353,
+      "learning_rate": 2.448570418584898e-06,
+      "loss": 0.2477,
+      "step": 1512
+    },
+    {
+      "epoch": 2.1129969774471054,
+      "grad_norm": 1.9471977071121356,
+      "learning_rate": 2.4415852119065343e-06,
+      "loss": 0.2454,
+      "step": 1513
+    },
+    {
+      "epoch": 2.1143920018600326,
+      "grad_norm": 1.9889979101530386,
+      "learning_rate": 2.4346067630525084e-06,
+      "loss": 0.2442,
+      "step": 1514
+    },
+    {
+      "epoch": 2.11578702627296,
+      "grad_norm": 2.139859675456495,
+      "learning_rate": 2.427635090455766e-06,
+      "loss": 0.2289,
+      "step": 1515
+    },
+    {
+      "epoch": 2.117182050685887,
+      "grad_norm": 2.013707423934541,
+      "learning_rate": 2.42067021253136e-06,
+      "loss": 0.232,
+      "step": 1516
+    },
+    {
+      "epoch": 2.1185770750988144,
+      "grad_norm": 1.89760209157116,
+      "learning_rate": 2.4137121476763965e-06,
+      "loss": 0.2077,
+      "step": 1517
+    },
+    {
+      "epoch": 2.1199720995117413,
+      "grad_norm": 2.0816857703833143,
+      "learning_rate": 2.4067609142699798e-06,
+      "loss": 0.2193,
+      "step": 1518
+    },
+    {
+      "epoch": 2.1213671239246685,
+      "grad_norm": 2.1617449428356705,
+      "learning_rate": 2.3998165306731713e-06,
+      "loss": 0.2411,
+      "step": 1519
+    },
+    {
+      "epoch": 2.122762148337596,
+      "grad_norm": 1.985051182731335,
+      "learning_rate": 2.3928790152289443e-06,
+      "loss": 0.2066,
+      "step": 1520
+    },
+    {
+      "epoch": 2.124157172750523,
+      "grad_norm": 2.0332885897822917,
+      "learning_rate": 2.385948386262123e-06,
+      "loss": 0.2589,
+      "step": 1521
+    },
+    {
+      "epoch": 2.1255521971634503,
+      "grad_norm": 2.1609643142255477,
+      "learning_rate": 2.3790246620793466e-06,
+      "loss": 0.2453,
+      "step": 1522
+    },
+    {
+      "epoch": 2.1269472215763776,
+      "grad_norm": 2.1104388284929074,
+      "learning_rate": 2.372107860969019e-06,
+      "loss": 0.2327,
+      "step": 1523
+    },
+    {
+      "epoch": 2.128342245989305,
+      "grad_norm": 2.0050701731009664,
+      "learning_rate": 2.3651980012012454e-06,
+      "loss": 0.2183,
+      "step": 1524
+    },
+    {
+      "epoch": 2.129737270402232,
+      "grad_norm": 1.8374702594979928,
+      "learning_rate": 2.358295101027807e-06,
+      "loss": 0.2271,
+      "step": 1525
+    },
+    {
+      "epoch": 2.1311322948151594,
+      "grad_norm": 1.7744485016961058,
+      "learning_rate": 2.351399178682101e-06,
+      "loss": 0.2318,
+      "step": 1526
+    },
+    {
+      "epoch": 2.1325273192280867,
+      "grad_norm": 2.1707657393472557,
+      "learning_rate": 2.3445102523790876e-06,
+      "loss": 0.2192,
+      "step": 1527
+    },
+    {
+      "epoch": 2.1339223436410135,
+      "grad_norm": 1.9212023430718843,
+      "learning_rate": 2.3376283403152527e-06,
+      "loss": 0.2274,
+      "step": 1528
+    },
+    {
+      "epoch": 2.135317368053941,
+      "grad_norm": 1.9431330763254702,
+      "learning_rate": 2.330753460668553e-06,
+      "loss": 0.2363,
+      "step": 1529
+    },
+    {
+      "epoch": 2.136712392466868,
+      "grad_norm": 1.850992735000709,
+      "learning_rate": 2.323885631598366e-06,
+      "loss": 0.2398,
+      "step": 1530
+    },
+    {
+      "epoch": 2.1381074168797953,
+      "grad_norm": 1.9617578235530817,
+      "learning_rate": 2.3170248712454525e-06,
+      "loss": 0.247,
+      "step": 1531
+    },
+    {
+      "epoch": 2.1395024412927226,
+      "grad_norm": 2.0582404649051385,
+      "learning_rate": 2.3101711977318995e-06,
+      "loss": 0.2654,
+      "step": 1532
+    },
+    {
+      "epoch": 2.14089746570565,
+      "grad_norm": 2.0530723360930927,
+      "learning_rate": 2.3033246291610717e-06,
+      "loss": 0.2297,
+      "step": 1533
+    },
+    {
+      "epoch": 2.142292490118577,
+      "grad_norm": 2.0428682175339183,
+      "learning_rate": 2.2964851836175705e-06,
+      "loss": 0.2479,
+      "step": 1534
+    },
+    {
+      "epoch": 2.1436875145315044,
+      "grad_norm": 2.033140247859454,
+      "learning_rate": 2.2896528791671807e-06,
+      "loss": 0.2287,
+      "step": 1535
+    },
+    {
+      "epoch": 2.1450825389444317,
+      "grad_norm": 1.9706953955373776,
+      "learning_rate": 2.2828277338568226e-06,
+      "loss": 0.2309,
+      "step": 1536
+    },
+    {
+      "epoch": 2.146477563357359,
+      "grad_norm": 2.0657013312891763,
+      "learning_rate": 2.2760097657145096e-06,
+      "loss": 0.2376,
+      "step": 1537
+    },
+    {
+      "epoch": 2.147872587770286,
+      "grad_norm": 2.010341964625042,
+      "learning_rate": 2.2691989927492984e-06,
+      "loss": 0.2577,
+      "step": 1538
+    },
+    {
+      "epoch": 2.149267612183213,
+      "grad_norm": 2.2001973373784334,
+      "learning_rate": 2.262395432951235e-06,
+      "loss": 0.2596,
+      "step": 1539
+    },
+    {
+      "epoch": 2.1506626365961403,
+      "grad_norm": 1.9469948847454945,
+      "learning_rate": 2.2555991042913177e-06,
+      "loss": 0.2278,
+      "step": 1540
+    },
+    {
+      "epoch": 2.1520576610090676,
+      "grad_norm": 2.108148021681362,
+      "learning_rate": 2.248810024721441e-06,
+      "loss": 0.2251,
+      "step": 1541
+    },
+    {
+      "epoch": 2.153452685421995,
+      "grad_norm": 1.9836437529062159,
+      "learning_rate": 2.2420282121743513e-06,
+      "loss": 0.2367,
+      "step": 1542
+    },
+    {
+      "epoch": 2.154847709834922,
+      "grad_norm": 2.024544254750702,
+      "learning_rate": 2.235253684563602e-06,
+      "loss": 0.2259,
+      "step": 1543
+    },
+    {
+      "epoch": 2.1562427342478494,
+      "grad_norm": 2.016705862662843,
+      "learning_rate": 2.228486459783506e-06,
+      "loss": 0.2176,
+      "step": 1544
+    },
+    {
+      "epoch": 2.1576377586607767,
+      "grad_norm": 1.8277595229039698,
+      "learning_rate": 2.221726555709079e-06,
+      "loss": 0.2328,
+      "step": 1545
+    },
+    {
+      "epoch": 2.159032783073704,
+      "grad_norm": 2.0853945150495137,
+      "learning_rate": 2.2149739901960088e-06,
+      "loss": 0.2496,
+      "step": 1546
+    },
+    {
+      "epoch": 2.160427807486631,
+      "grad_norm": 1.933919574208661,
+      "learning_rate": 2.208228781080592e-06,
+      "loss": 0.2276,
+      "step": 1547
+    },
+    {
+      "epoch": 2.161822831899558,
+      "grad_norm": 1.9957596951759287,
+      "learning_rate": 2.201490946179696e-06,
+      "loss": 0.2202,
+      "step": 1548
+    },
+    {
+      "epoch": 2.1632178563124853,
+      "grad_norm": 1.8822509043654059,
+      "learning_rate": 2.19476050329071e-06,
+      "loss": 0.2122,
+      "step": 1549
+    },
+    {
+      "epoch": 2.1646128807254126,
+      "grad_norm": 1.984553964575256,
+      "learning_rate": 2.188037470191502e-06,
+      "loss": 0.2258,
+      "step": 1550
+    },
+    {
+      "epoch": 2.16600790513834,
+      "grad_norm": 1.7587934213104526,
+      "learning_rate": 2.181321864640362e-06,
+      "loss": 0.2286,
+      "step": 1551
+    },
+    {
+      "epoch": 2.167402929551267,
+      "grad_norm": 2.0834079951630495,
+      "learning_rate": 2.1746137043759594e-06,
+      "loss": 0.2099,
+      "step": 1552
+    },
+    {
+      "epoch": 2.1687979539641944,
+      "grad_norm": 1.7944200796734386,
+      "learning_rate": 2.167913007117306e-06,
+      "loss": 0.2262,
+      "step": 1553
+    },
+    {
+      "epoch": 2.1701929783771217,
+      "grad_norm": 2.0717320889653177,
+      "learning_rate": 2.1612197905636913e-06,
+      "loss": 0.2554,
+      "step": 1554
+    },
+    {
+      "epoch": 2.171588002790049,
+      "grad_norm": 1.9794269773744197,
+      "learning_rate": 2.154534072394651e-06,
+      "loss": 0.2332,
+      "step": 1555
+    },
+    {
+      "epoch": 2.172983027202976,
+      "grad_norm": 2.0786474656189524,
+      "learning_rate": 2.147855870269916e-06,
+      "loss": 0.2512,
+      "step": 1556
+    },
+    {
+      "epoch": 2.1743780516159035,
+      "grad_norm": 1.905587487947106,
+      "learning_rate": 2.1411852018293583e-06,
+      "loss": 0.2725,
+      "step": 1557
+    },
+    {
+      "epoch": 2.1757730760288303,
+      "grad_norm": 1.9686301802786672,
+      "learning_rate": 2.1345220846929514e-06,
+      "loss": 0.2291,
+      "step": 1558
+    },
+    {
+      "epoch": 2.1771681004417576,
+      "grad_norm": 2.0061029510599266,
+      "learning_rate": 2.127866536460727e-06,
+      "loss": 0.2323,
+      "step": 1559
+    },
+    {
+      "epoch": 2.178563124854685,
+      "grad_norm": 1.8392842837936532,
+      "learning_rate": 2.1212185747127235e-06,
+      "loss": 0.2392,
+      "step": 1560
+    },
+    {
+      "epoch": 2.179958149267612,
+      "grad_norm": 1.8963609241375565,
+      "learning_rate": 2.1145782170089346e-06,
+      "loss": 0.2241,
+      "step": 1561
+    },
+    {
+      "epoch": 2.1813531736805394,
+      "grad_norm": 2.020691216176965,
+      "learning_rate": 2.107945480889276e-06,
+      "loss": 0.2202,
+      "step": 1562
+    },
+    {
+      "epoch": 2.1827481980934667,
+      "grad_norm": 2.1222330186555034,
+      "learning_rate": 2.1013203838735273e-06,
+      "loss": 0.2338,
+      "step": 1563
+    },
+    {
+      "epoch": 2.184143222506394,
+      "grad_norm": 1.8451126874497825,
+      "learning_rate": 2.094702943461289e-06,
+      "loss": 0.2271,
+      "step": 1564
+    },
+    {
+      "epoch": 2.185538246919321,
+      "grad_norm": 2.0925263169112243,
+      "learning_rate": 2.0880931771319395e-06,
+      "loss": 0.261,
+      "step": 1565
+    },
+    {
+      "epoch": 2.1869332713322485,
+      "grad_norm": 2.0486486674355566,
+      "learning_rate": 2.0814911023445904e-06,
+      "loss": 0.2056,
+      "step": 1566
+    },
+    {
+      "epoch": 2.1883282957451757,
+      "grad_norm": 1.8973623339063204,
+      "learning_rate": 2.0748967365380292e-06,
+      "loss": 0.2288,
+      "step": 1567
+    },
+    {
+      "epoch": 2.1897233201581026,
+      "grad_norm": 1.9264105513759306,
+      "learning_rate": 2.0683100971306873e-06,
+      "loss": 0.2458,
+      "step": 1568
+    },
+    {
+      "epoch": 2.19111834457103,
+      "grad_norm": 2.1351016674119396,
+      "learning_rate": 2.0617312015205844e-06,
+      "loss": 0.2618,
+      "step": 1569
+    },
+    {
+      "epoch": 2.192513368983957,
+      "grad_norm": 2.0301845360513444,
+      "learning_rate": 2.055160067085283e-06,
+      "loss": 0.222,
+      "step": 1570
+    },
+    {
+      "epoch": 2.1939083933968844,
+      "grad_norm": 1.9366582361580502,
+      "learning_rate": 2.0485967111818506e-06,
+      "loss": 0.2464,
+      "step": 1571
+    },
+    {
+      "epoch": 2.1953034178098116,
+      "grad_norm": 1.8379955335152875,
+      "learning_rate": 2.0420411511468086e-06,
+      "loss": 0.2218,
+      "step": 1572
+    },
+    {
+      "epoch": 2.196698442222739,
+      "grad_norm": 1.924135371652443,
+      "learning_rate": 2.0354934042960804e-06,
+      "loss": 0.2684,
+      "step": 1573
+    },
+    {
+      "epoch": 2.198093466635666,
+      "grad_norm": 1.9427660648400258,
+      "learning_rate": 2.0289534879249544e-06,
+      "loss": 0.2195,
+      "step": 1574
+    },
+    {
+      "epoch": 2.1994884910485935,
+      "grad_norm": 2.117053729594471,
+      "learning_rate": 2.0224214193080394e-06,
+      "loss": 0.2363,
+      "step": 1575
+    },
+    {
+      "epoch": 2.2008835154615207,
+      "grad_norm": 2.0212314391425323,
+      "learning_rate": 2.015897215699208e-06,
+      "loss": 0.2407,
+      "step": 1576
+    },
+    {
+      "epoch": 2.202278539874448,
+      "grad_norm": 1.9744514677082492,
+      "learning_rate": 2.0093808943315636e-06,
+      "loss": 0.213,
+      "step": 1577
+    },
+    {
+      "epoch": 2.2036735642873753,
+      "grad_norm": 2.0360684189888976,
+      "learning_rate": 2.0028724724173886e-06,
+      "loss": 0.218,
+      "step": 1578
+    },
+    {
+      "epoch": 2.205068588700302,
+      "grad_norm": 1.990797315085629,
+      "learning_rate": 1.996371967148098e-06,
+      "loss": 0.2252,
+      "step": 1579
+    },
+    {
+      "epoch": 2.2064636131132294,
+      "grad_norm": 2.1560737572043225,
+      "learning_rate": 1.989879395694194e-06,
+      "loss": 0.2704,
+      "step": 1580
+    },
+    {
+      "epoch": 2.2078586375261566,
+      "grad_norm": 2.017853645994684,
+      "learning_rate": 1.9833947752052286e-06,
+      "loss": 0.2195,
+      "step": 1581
+    },
+    {
+      "epoch": 2.209253661939084,
+      "grad_norm": 2.041977727763059,
+      "learning_rate": 1.976918122809744e-06,
+      "loss": 0.2405,
+      "step": 1582
+    },
+    {
+      "epoch": 2.210648686352011,
+      "grad_norm": 1.9608446629611556,
+      "learning_rate": 1.9704494556152413e-06,
+      "loss": 0.2365,
+      "step": 1583
+    },
+    {
+      "epoch": 2.2120437107649384,
+      "grad_norm": 1.8806591549300393,
+      "learning_rate": 1.9639887907081297e-06,
+      "loss": 0.2084,
+      "step": 1584
+    },
+    {
+      "epoch": 2.2134387351778657,
+      "grad_norm": 2.1068982589953307,
+      "learning_rate": 1.9575361451536772e-06,
+      "loss": 0.2541,
+      "step": 1585
+    },
+    {
+      "epoch": 2.214833759590793,
+      "grad_norm": 1.9806962496339104,
+      "learning_rate": 1.9510915359959694e-06,
+      "loss": 0.2053,
+      "step": 1586
+    },
+    {
+      "epoch": 2.2162287840037203,
+      "grad_norm": 1.7041895265128661,
+      "learning_rate": 1.944654980257869e-06,
+      "loss": 0.2245,
+      "step": 1587
+    },
+    {
+      "epoch": 2.217623808416647,
+      "grad_norm": 1.9984605463110598,
+      "learning_rate": 1.9382264949409614e-06,
+      "loss": 0.2348,
+      "step": 1588
+    },
+    {
+      "epoch": 2.2190188328295744,
+      "grad_norm": 2.0330867230795286,
+      "learning_rate": 1.931806097025517e-06,
+      "loss": 0.2491,
+      "step": 1589
+    },
+    {
+      "epoch": 2.2204138572425016,
+      "grad_norm": 2.0364218516052133,
+      "learning_rate": 1.925393803470447e-06,
+      "loss": 0.2109,
+      "step": 1590
+    },
+    {
+      "epoch": 2.221808881655429,
+      "grad_norm": 2.452548285422924,
+      "learning_rate": 1.9189896312132506e-06,
+      "loss": 0.2388,
+      "step": 1591
+    },
+    {
+      "epoch": 2.223203906068356,
+      "grad_norm": 2.028391891671147,
+      "learning_rate": 1.912593597169975e-06,
+      "loss": 0.2078,
+      "step": 1592
+    },
+    {
+      "epoch": 2.2245989304812834,
+      "grad_norm": 2.0694649414957476,
+      "learning_rate": 1.9062057182351768e-06,
+      "loss": 0.2378,
+      "step": 1593
+    },
+    {
+      "epoch": 2.2259939548942107,
+      "grad_norm": 1.8977475386258975,
+      "learning_rate": 1.899826011281865e-06,
+      "loss": 0.2551,
+      "step": 1594
+    },
+    {
+      "epoch": 2.227388979307138,
+      "grad_norm": 1.8794144304149887,
+      "learning_rate": 1.893454493161468e-06,
+      "loss": 0.222,
+      "step": 1595
+    },
+    {
+      "epoch": 2.2287840037200652,
+      "grad_norm": 1.8421851383635428,
+      "learning_rate": 1.8870911807037856e-06,
+      "loss": 0.2292,
+      "step": 1596
+    },
+    {
+      "epoch": 2.2301790281329925,
+      "grad_norm": 2.1294035926487216,
+      "learning_rate": 1.8807360907169326e-06,
+      "loss": 0.2573,
+      "step": 1597
+    },
+    {
+      "epoch": 2.23157405254592,
+      "grad_norm": 1.9710536419619946,
+      "learning_rate": 1.8743892399873154e-06,
+      "loss": 0.2262,
+      "step": 1598
+    },
+    {
+      "epoch": 2.2329690769588466,
+      "grad_norm": 2.054320626994414,
+      "learning_rate": 1.868050645279576e-06,
+      "loss": 0.2363,
+      "step": 1599
+    },
+    {
+      "epoch": 2.234364101371774,
+      "grad_norm": 1.9376735935313658,
+      "learning_rate": 1.8617203233365427e-06,
+      "loss": 0.2277,
+      "step": 1600
+    },
+    {
+      "epoch": 2.235759125784701,
+      "grad_norm": 1.96875678147811,
+      "learning_rate": 1.8553982908792e-06,
+      "loss": 0.2387,
+      "step": 1601
+    },
+    {
+      "epoch": 2.2371541501976284,
+      "grad_norm": 2.183861599239592,
+      "learning_rate": 1.8490845646066303e-06,
+      "loss": 0.2772,
+      "step": 1602
+    },
+    {
+      "epoch": 2.2385491746105557,
+      "grad_norm": 2.3569650294971227,
+      "learning_rate": 1.8427791611959762e-06,
+      "loss": 0.2558,
+      "step": 1603
+    },
+    {
+      "epoch": 2.239944199023483,
+      "grad_norm": 1.907449594297111,
+      "learning_rate": 1.8364820973024e-06,
+      "loss": 0.2141,
+      "step": 1604
+    },
+    {
+      "epoch": 2.2413392234364102,
+      "grad_norm": 2.1933972049660477,
+      "learning_rate": 1.8301933895590362e-06,
+      "loss": 0.253,
+      "step": 1605
+    },
+    {
+      "epoch": 2.2427342478493375,
+      "grad_norm": 1.9857324925925934,
+      "learning_rate": 1.8239130545769408e-06,
+      "loss": 0.2216,
+      "step": 1606
+    },
+    {
+      "epoch": 2.2441292722622648,
+      "grad_norm": 2.199122492257547,
+      "learning_rate": 1.8176411089450618e-06,
+      "loss": 0.2547,
+      "step": 1607
+    },
+    {
+      "epoch": 2.2455242966751916,
+      "grad_norm": 2.007748115189075,
+      "learning_rate": 1.8113775692301822e-06,
+      "loss": 0.2354,
+      "step": 1608
+    },
+    {
+      "epoch": 2.246919321088119,
+      "grad_norm": 2.144681332284037,
+      "learning_rate": 1.8051224519768817e-06,
+      "loss": 0.2368,
+      "step": 1609
+    },
+    {
+      "epoch": 2.248314345501046,
+      "grad_norm": 1.9606369218220936,
+      "learning_rate": 1.7988757737074959e-06,
+      "loss": 0.216,
+      "step": 1610
+    },
+    {
+      "epoch": 2.2497093699139734,
+      "grad_norm": 2.0077910044386553,
+      "learning_rate": 1.7926375509220695e-06,
+      "loss": 0.2246,
+      "step": 1611
+    },
+    {
+      "epoch": 2.2511043943269007,
+      "grad_norm": 2.0150546849238355,
+      "learning_rate": 1.7864078000983076e-06,
+      "loss": 0.2102,
+      "step": 1612
+    },
+    {
+      "epoch": 2.252499418739828,
+      "grad_norm": 2.048720247263012,
+      "learning_rate": 1.7801865376915451e-06,
+      "loss": 0.2207,
+      "step": 1613
+    },
+    {
+      "epoch": 2.2538944431527552,
+      "grad_norm": 2.051461913853357,
+      "learning_rate": 1.7739737801346895e-06,
+      "loss": 0.2479,
+      "step": 1614
+    },
+    {
+      "epoch": 2.2552894675656825,
+      "grad_norm": 1.8388149137964327,
+      "learning_rate": 1.7677695438381831e-06,
+      "loss": 0.2531,
+      "step": 1615
+    },
+    {
+      "epoch": 2.2566844919786098,
+      "grad_norm": 2.0495512505404823,
+      "learning_rate": 1.761573845189965e-06,
+      "loss": 0.2349,
+      "step": 1616
+    },
+    {
+      "epoch": 2.258079516391537,
+      "grad_norm": 1.8415416726079457,
+      "learning_rate": 1.7553867005554215e-06,
+      "loss": 0.2042,
+      "step": 1617
+    },
+    {
+      "epoch": 2.2594745408044643,
+      "grad_norm": 1.9768703900158413,
+      "learning_rate": 1.7492081262773397e-06,
+      "loss": 0.2323,
+      "step": 1618
+    },
+    {
+      "epoch": 2.260869565217391,
+      "grad_norm": 2.1606595752623923,
+      "learning_rate": 1.7430381386758748e-06,
+      "loss": 0.2465,
+      "step": 1619
+    },
+    {
+      "epoch": 2.2622645896303184,
+      "grad_norm": 1.9661673368498434,
+      "learning_rate": 1.7368767540484965e-06,
+      "loss": 0.2219,
+      "step": 1620
+    },
+    {
+      "epoch": 2.2636596140432457,
+      "grad_norm": 1.8878002265817848,
+      "learning_rate": 1.7307239886699546e-06,
+      "loss": 0.2537,
+      "step": 1621
+    },
+    {
+      "epoch": 2.265054638456173,
+      "grad_norm": 2.1599816560012632,
+      "learning_rate": 1.7245798587922263e-06,
+      "loss": 0.2409,
+      "step": 1622
+    },
+    {
+      "epoch": 2.2664496628691,
+      "grad_norm": 1.877718357363733,
+      "learning_rate": 1.7184443806444851e-06,
+      "loss": 0.2273,
+      "step": 1623
+    },
+    {
+      "epoch": 2.2678446872820275,
+      "grad_norm": 1.7823955384474417,
+      "learning_rate": 1.7123175704330514e-06,
+      "loss": 0.2189,
+      "step": 1624
+    },
+    {
+      "epoch": 2.2692397116949548,
+      "grad_norm": 1.9192171774637246,
+      "learning_rate": 1.706199444341341e-06,
+      "loss": 0.2403,
+      "step": 1625
+    },
+    {
+      "epoch": 2.270634736107882,
+      "grad_norm": 2.0764997514013617,
+      "learning_rate": 1.7000900185298418e-06,
+      "loss": 0.273,
+      "step": 1626
+    },
+    {
+      "epoch": 2.2720297605208093,
+      "grad_norm": 2.124933803249738,
+      "learning_rate": 1.6939893091360577e-06,
+      "loss": 0.2631,
+      "step": 1627
+    },
+    {
+      "epoch": 2.273424784933736,
+      "grad_norm": 2.0537090747806257,
+      "learning_rate": 1.6878973322744658e-06,
+      "loss": 0.2259,
+      "step": 1628
+    },
+    {
+      "epoch": 2.2748198093466634,
+      "grad_norm": 1.9496310576864384,
+      "learning_rate": 1.6818141040364816e-06,
+      "loss": 0.2286,
+      "step": 1629
+    },
+    {
+      "epoch": 2.2762148337595907,
+      "grad_norm": 1.9684283583462694,
+      "learning_rate": 1.6757396404904087e-06,
+      "loss": 0.2257,
+      "step": 1630
+    },
+    {
+      "epoch": 2.277609858172518,
+      "grad_norm": 1.9211403090843973,
+      "learning_rate": 1.6696739576813981e-06,
+      "loss": 0.24,
+      "step": 1631
+    },
+    {
+      "epoch": 2.279004882585445,
+      "grad_norm": 1.9601889627957176,
+      "learning_rate": 1.6636170716314114e-06,
+      "loss": 0.2641,
+      "step": 1632
+    },
+    {
+      "epoch": 2.2803999069983725,
+      "grad_norm": 2.2106724133570697,
+      "learning_rate": 1.657568998339175e-06,
+      "loss": 0.2624,
+      "step": 1633
+    },
+    {
+      "epoch": 2.2817949314112997,
+      "grad_norm": 2.3843544770866556,
+      "learning_rate": 1.6515297537801305e-06,
+      "loss": 0.2485,
+      "step": 1634
+    },
+    {
+      "epoch": 2.283189955824227,
+      "grad_norm": 1.9719774663966787,
+      "learning_rate": 1.6454993539064075e-06,
+      "loss": 0.2178,
+      "step": 1635
+    },
+    {
+      "epoch": 2.2845849802371543,
+      "grad_norm": 1.9373964616111852,
+      "learning_rate": 1.6394778146467672e-06,
+      "loss": 0.2687,
+      "step": 1636
+    },
+    {
+      "epoch": 2.2859800046500816,
+      "grad_norm": 2.184356785161183,
+      "learning_rate": 1.6334651519065658e-06,
+      "loss": 0.2558,
+      "step": 1637
+    },
+    {
+      "epoch": 2.287375029063009,
+      "grad_norm": 2.09734743364332,
+      "learning_rate": 1.6274613815677176e-06,
+      "loss": 0.2441,
+      "step": 1638
+    },
+    {
+      "epoch": 2.2887700534759357,
+      "grad_norm": 2.1459547086066872,
+      "learning_rate": 1.6214665194886474e-06,
+      "loss": 0.1963,
+      "step": 1639
+    },
+    {
+      "epoch": 2.290165077888863,
+      "grad_norm": 1.8154559511599928,
+      "learning_rate": 1.6154805815042457e-06,
+      "loss": 0.2092,
+      "step": 1640
+    },
+    {
+      "epoch": 2.29156010230179,
+      "grad_norm": 2.1323386307866543,
+      "learning_rate": 1.6095035834258365e-06,
+      "loss": 0.2485,
+      "step": 1641
+    },
+    {
+      "epoch": 2.2929551267147175,
+      "grad_norm": 1.8872892475373095,
+      "learning_rate": 1.6035355410411252e-06,
+      "loss": 0.1996,
+      "step": 1642
+    },
+    {
+      "epoch": 2.2943501511276447,
+      "grad_norm": 2.0382534164232684,
+      "learning_rate": 1.5975764701141611e-06,
+      "loss": 0.2359,
+      "step": 1643
+    },
+    {
+      "epoch": 2.295745175540572,
+      "grad_norm": 1.9572911333723713,
+      "learning_rate": 1.5916263863853e-06,
+      "loss": 0.2416,
+      "step": 1644
+    },
+    {
+      "epoch": 2.2971401999534993,
+      "grad_norm": 2.150341017189422,
+      "learning_rate": 1.585685305571159e-06,
+      "loss": 0.2625,
+      "step": 1645
+    },
+    {
+      "epoch": 2.2985352243664265,
+      "grad_norm": 2.1453971015429016,
+      "learning_rate": 1.5797532433645696e-06,
+      "loss": 0.2271,
+      "step": 1646
+    },
+    {
+      "epoch": 2.299930248779354,
+      "grad_norm": 2.096925731594485,
+      "learning_rate": 1.5738302154345475e-06,
+      "loss": 0.2363,
+      "step": 1647
+    },
+    {
+      "epoch": 2.3013252731922806,
+      "grad_norm": 2.0826325764632365,
+      "learning_rate": 1.5679162374262414e-06,
+      "loss": 0.2319,
+      "step": 1648
+    },
+    {
+      "epoch": 2.302720297605208,
+      "grad_norm": 1.9053150555031908,
+      "learning_rate": 1.5620113249608943e-06,
+      "loss": 0.2536,
+      "step": 1649
+    },
+    {
+      "epoch": 2.304115322018135,
+      "grad_norm": 2.013788051403245,
+      "learning_rate": 1.5561154936358069e-06,
+      "loss": 0.2379,
+      "step": 1650
+    },
+    {
+      "epoch": 2.3055103464310625,
+      "grad_norm": 2.0929701003948313,
+      "learning_rate": 1.5502287590242942e-06,
+      "loss": 0.2285,
+      "step": 1651
+    },
+    {
+      "epoch": 2.3069053708439897,
+      "grad_norm": 1.8618570907267726,
+      "learning_rate": 1.5443511366756375e-06,
+      "loss": 0.2168,
+      "step": 1652
+    },
+    {
+      "epoch": 2.308300395256917,
+      "grad_norm": 1.9443015801429164,
+      "learning_rate": 1.53848264211505e-06,
+      "loss": 0.2428,
+      "step": 1653
+    },
+    {
+      "epoch": 2.3096954196698443,
+      "grad_norm": 1.730889234878071,
+      "learning_rate": 1.5326232908436405e-06,
+      "loss": 0.2132,
+      "step": 1654
+    },
+    {
+      "epoch": 2.3110904440827715,
+      "grad_norm": 2.033663509452582,
+      "learning_rate": 1.526773098338359e-06,
+      "loss": 0.2326,
+      "step": 1655
+    },
+    {
+      "epoch": 2.312485468495699,
+      "grad_norm": 2.0086739819616684,
+      "learning_rate": 1.5209320800519683e-06,
+      "loss": 0.2299,
+      "step": 1656
+    },
+    {
+      "epoch": 2.313880492908626,
+      "grad_norm": 2.1150179443276693,
+      "learning_rate": 1.515100251412998e-06,
+      "loss": 0.2514,
+      "step": 1657
+    },
+    {
+      "epoch": 2.3152755173215533,
+      "grad_norm": 2.0580293221490744,
+      "learning_rate": 1.5092776278257027e-06,
+      "loss": 0.2269,
+      "step": 1658
+    },
+    {
+      "epoch": 2.31667054173448,
+      "grad_norm": 1.9664146763176193,
+      "learning_rate": 1.5034642246700203e-06,
+      "loss": 0.239,
+      "step": 1659
+    },
+    {
+      "epoch": 2.3180655661474074,
+      "grad_norm": 2.127726449836482,
+      "learning_rate": 1.4976600573015398e-06,
+      "loss": 0.2503,
+      "step": 1660
+    },
+    {
+      "epoch": 2.3194605905603347,
+      "grad_norm": 2.1049092968450216,
+      "learning_rate": 1.4918651410514479e-06,
+      "loss": 0.2497,
+      "step": 1661
+    },
+    {
+      "epoch": 2.320855614973262,
+      "grad_norm": 2.066851659126085,
+      "learning_rate": 1.486079491226501e-06,
+      "loss": 0.2209,
+      "step": 1662
+    },
+    {
+      "epoch": 2.3222506393861893,
+      "grad_norm": 2.0102872039801536,
+      "learning_rate": 1.4803031231089782e-06,
+      "loss": 0.232,
+      "step": 1663
+    },
+    {
+      "epoch": 2.3236456637991165,
+      "grad_norm": 1.8240849797174543,
+      "learning_rate": 1.4745360519566382e-06,
+      "loss": 0.2228,
+      "step": 1664
+    },
+    {
+      "epoch": 2.325040688212044,
+      "grad_norm": 2.0260207328677646,
+      "learning_rate": 1.4687782930026833e-06,
+      "loss": 0.2503,
+      "step": 1665
+    },
+    {
+      "epoch": 2.326435712624971,
+      "grad_norm": 1.9370440208183426,
+      "learning_rate": 1.4630298614557236e-06,
+      "loss": 0.2293,
+      "step": 1666
+    },
+    {
+      "epoch": 2.3278307370378983,
+      "grad_norm": 2.0408731656057126,
+      "learning_rate": 1.4572907724997249e-06,
+      "loss": 0.2745,
+      "step": 1667
+    },
+    {
+      "epoch": 2.329225761450825,
+      "grad_norm": 1.9501445042320373,
+      "learning_rate": 1.4515610412939791e-06,
+      "loss": 0.22,
+      "step": 1668
+    },
+    {
+      "epoch": 2.3306207858637524,
+      "grad_norm": 1.8239379510751645,
+      "learning_rate": 1.445840682973062e-06,
+      "loss": 0.2416,
+      "step": 1669
+    },
+    {
+      "epoch": 2.3320158102766797,
+      "grad_norm": 2.008062176566863,
+      "learning_rate": 1.4401297126467884e-06,
+      "loss": 0.2087,
+      "step": 1670
+    },
+    {
+      "epoch": 2.333410834689607,
+      "grad_norm": 1.9572423516680113,
+      "learning_rate": 1.4344281454001751e-06,
+      "loss": 0.2045,
+      "step": 1671
+    },
+    {
+      "epoch": 2.3348058591025342,
+      "grad_norm": 2.1525787216094128,
+      "learning_rate": 1.4287359962934055e-06,
+      "loss": 0.2466,
+      "step": 1672
+    },
+    {
+      "epoch": 2.3362008835154615,
+      "grad_norm": 2.1085528054352722,
+      "learning_rate": 1.4230532803617814e-06,
+      "loss": 0.249,
+      "step": 1673
+    },
+    {
+      "epoch": 2.337595907928389,
+      "grad_norm": 1.900064207744204,
+      "learning_rate": 1.4173800126156916e-06,
+      "loss": 0.2427,
+      "step": 1674
+    },
+    {
+      "epoch": 2.338990932341316,
+      "grad_norm": 1.8330769763213726,
+      "learning_rate": 1.411716208040566e-06,
+      "loss": 0.224,
+      "step": 1675
+    },
+    {
+      "epoch": 2.3403859567542433,
+      "grad_norm": 2.1005430164406387,
+      "learning_rate": 1.4060618815968375e-06,
+      "loss": 0.2487,
+      "step": 1676
+    },
+    {
+      "epoch": 2.3417809811671706,
+      "grad_norm": 2.0233119703276197,
+      "learning_rate": 1.4004170482199054e-06,
+      "loss": 0.2616,
+      "step": 1677
+    },
+    {
+      "epoch": 2.343176005580098,
+      "grad_norm": 1.9503887562468631,
+      "learning_rate": 1.3947817228200956e-06,
+      "loss": 0.2101,
+      "step": 1678
+    },
+    {
+      "epoch": 2.3445710299930247,
+      "grad_norm": 2.0266811706036796,
+      "learning_rate": 1.3891559202826133e-06,
+      "loss": 0.2634,
+      "step": 1679
+    },
+    {
+      "epoch": 2.345966054405952,
+      "grad_norm": 2.313818281195364,
+      "learning_rate": 1.3835396554675179e-06,
+      "loss": 0.2358,
+      "step": 1680
+    },
+    {
+      "epoch": 2.3473610788188792,
+      "grad_norm": 2.0086635288310677,
+      "learning_rate": 1.37793294320967e-06,
+      "loss": 0.261,
+      "step": 1681
+    },
+    {
+      "epoch": 2.3487561032318065,
+      "grad_norm": 1.9456694874174518,
+      "learning_rate": 1.3723357983186974e-06,
+      "loss": 0.2166,
+      "step": 1682
+    },
+    {
+      "epoch": 2.3501511276447338,
+      "grad_norm": 2.0159574470737236,
+      "learning_rate": 1.3667482355789607e-06,
+      "loss": 0.2259,
+      "step": 1683
+    },
+    {
+      "epoch": 2.351546152057661,
+      "grad_norm": 2.0882553918627504,
+      "learning_rate": 1.3611702697495088e-06,
+      "loss": 0.2127,
+      "step": 1684
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 1.9536375676927595,
+      "learning_rate": 1.3556019155640416e-06,
+      "loss": 0.2327,
+      "step": 1685
+    },
+    {
+      "epoch": 2.3543362008835156,
+      "grad_norm": 1.8321246562049889,
+      "learning_rate": 1.350043187730868e-06,
+      "loss": 0.2382,
+      "step": 1686
+    },
+    {
+      "epoch": 2.355731225296443,
+      "grad_norm": 2.4564159839079625,
+      "learning_rate": 1.34449410093287e-06,
+      "loss": 0.234,
+      "step": 1687
+    },
+    {
+      "epoch": 2.3571262497093697,
+      "grad_norm": 1.8735083700021802,
+      "learning_rate": 1.3389546698274686e-06,
+      "loss": 0.2153,
+      "step": 1688
+    },
+    {
+      "epoch": 2.358521274122297,
+      "grad_norm": 2.0889076665454063,
+      "learning_rate": 1.333424909046574e-06,
+      "loss": 0.2486,
+      "step": 1689
+    },
+    {
+      "epoch": 2.359916298535224,
+      "grad_norm": 2.152373586743707,
+      "learning_rate": 1.327904833196556e-06,
+      "loss": 0.2581,
+      "step": 1690
+    },
+    {
+      "epoch": 2.3613113229481515,
+      "grad_norm": 1.9192239036423904,
+      "learning_rate": 1.3223944568582047e-06,
+      "loss": 0.2151,
+      "step": 1691
+    },
+    {
+      "epoch": 2.3627063473610788,
+      "grad_norm": 1.9426489688983704,
+      "learning_rate": 1.3168937945866861e-06,
+      "loss": 0.2289,
+      "step": 1692
+    },
+    {
+      "epoch": 2.364101371774006,
+      "grad_norm": 2.0708618477896703,
+      "learning_rate": 1.311402860911507e-06,
+      "loss": 0.2416,
+      "step": 1693
+    },
+    {
+      "epoch": 2.3654963961869333,
+      "grad_norm": 1.9852614398809703,
+      "learning_rate": 1.3059216703364814e-06,
+      "loss": 0.2189,
+      "step": 1694
+    },
+    {
+      "epoch": 2.3668914205998606,
+      "grad_norm": 3.1422525655244855,
+      "learning_rate": 1.3004502373396821e-06,
+      "loss": 0.2364,
+      "step": 1695
+    },
+    {
+      "epoch": 2.368286445012788,
+      "grad_norm": 2.002483705304144,
+      "learning_rate": 1.2949885763734127e-06,
+      "loss": 0.2428,
+      "step": 1696
+    },
+    {
+      "epoch": 2.369681469425715,
+      "grad_norm": 1.9611827941041333,
+      "learning_rate": 1.2895367018641658e-06,
+      "loss": 0.2219,
+      "step": 1697
+    },
+    {
+      "epoch": 2.3710764938386424,
+      "grad_norm": 1.9929935498243903,
+      "learning_rate": 1.284094628212576e-06,
+      "loss": 0.2582,
+      "step": 1698
+    },
+    {
+      "epoch": 2.372471518251569,
+      "grad_norm": 1.9163017033696872,
+      "learning_rate": 1.278662369793398e-06,
+      "loss": 0.2196,
+      "step": 1699
+    },
+    {
+      "epoch": 2.3738665426644965,
+      "grad_norm": 2.1264033667590874,
+      "learning_rate": 1.273239940955459e-06,
+      "loss": 0.2552,
+      "step": 1700
+    },
+    {
+      "epoch": 2.3752615670774238,
+      "grad_norm": 2.029148427130633,
+      "learning_rate": 1.267827356021618e-06,
+      "loss": 0.2621,
+      "step": 1701
+    },
+    {
+      "epoch": 2.376656591490351,
+      "grad_norm": 2.141536132697765,
+      "learning_rate": 1.2624246292887377e-06,
+      "loss": 0.2347,
+      "step": 1702
+    },
+    {
+      "epoch": 2.3780516159032783,
+      "grad_norm": 2.0105615317777,
+      "learning_rate": 1.2570317750276374e-06,
+      "loss": 0.2582,
+      "step": 1703
+    },
+    {
+      "epoch": 2.3794466403162056,
+      "grad_norm": 2.062581019953467,
+      "learning_rate": 1.2516488074830586e-06,
+      "loss": 0.2473,
+      "step": 1704
+    },
+    {
+      "epoch": 2.380841664729133,
+      "grad_norm": 1.80393769257362,
+      "learning_rate": 1.246275740873631e-06,
+      "loss": 0.2301,
+      "step": 1705
+    },
+    {
+      "epoch": 2.38223668914206,
+      "grad_norm": 2.231699482955148,
+      "learning_rate": 1.2409125893918329e-06,
+      "loss": 0.2533,
+      "step": 1706
+    },
+    {
+      "epoch": 2.3836317135549874,
+      "grad_norm": 2.028453535194021,
+      "learning_rate": 1.2355593672039462e-06,
+      "loss": 0.1988,
+      "step": 1707
+    },
+    {
+      "epoch": 2.385026737967914,
+      "grad_norm": 1.917076340098864,
+      "learning_rate": 1.2302160884500337e-06,
+      "loss": 0.2623,
+      "step": 1708
+    },
+    {
+      "epoch": 2.3864217623808415,
+      "grad_norm": 2.0507149118696932,
+      "learning_rate": 1.2248827672438868e-06,
+      "loss": 0.2035,
+      "step": 1709
+    },
+    {
+      "epoch": 2.3878167867937687,
+      "grad_norm": 2.010253697172704,
+      "learning_rate": 1.2195594176729963e-06,
+      "loss": 0.236,
+      "step": 1710
+    },
+    {
+      "epoch": 2.389211811206696,
+      "grad_norm": 2.11324341731085,
+      "learning_rate": 1.2142460537985168e-06,
+      "loss": 0.2325,
+      "step": 1711
+    },
+    {
+      "epoch": 2.3906068356196233,
+      "grad_norm": 1.9697450740429523,
+      "learning_rate": 1.2089426896552265e-06,
+      "loss": 0.2245,
+      "step": 1712
+    },
+    {
+      "epoch": 2.3920018600325506,
+      "grad_norm": 2.0186158624900177,
+      "learning_rate": 1.2036493392514847e-06,
+      "loss": 0.2654,
+      "step": 1713
+    },
+    {
+      "epoch": 2.393396884445478,
+      "grad_norm": 2.086842619934469,
+      "learning_rate": 1.1983660165692078e-06,
+      "loss": 0.2071,
+      "step": 1714
+    },
+    {
+      "epoch": 2.394791908858405,
+      "grad_norm": 2.280522318427615,
+      "learning_rate": 1.1930927355638189e-06,
+      "loss": 0.2438,
+      "step": 1715
+    },
+    {
+      "epoch": 2.3961869332713324,
+      "grad_norm": 1.850134137584549,
+      "learning_rate": 1.1878295101642185e-06,
+      "loss": 0.2314,
+      "step": 1716
+    },
+    {
+      "epoch": 2.3975819576842596,
+      "grad_norm": 2.0987488388488744,
+      "learning_rate": 1.182576354272748e-06,
+      "loss": 0.2304,
+      "step": 1717
+    },
+    {
+      "epoch": 2.398976982097187,
+      "grad_norm": 1.9418839835561703,
+      "learning_rate": 1.1773332817651512e-06,
+      "loss": 0.226,
+      "step": 1718
+    },
+    {
+      "epoch": 2.400372006510114,
+      "grad_norm": 1.9467470286378656,
+      "learning_rate": 1.1721003064905329e-06,
+      "loss": 0.2132,
+      "step": 1719
+    },
+    {
+      "epoch": 2.401767030923041,
+      "grad_norm": 1.9135163938253588,
+      "learning_rate": 1.1668774422713336e-06,
+      "loss": 0.2419,
+      "step": 1720
+    },
+    {
+      "epoch": 2.4031620553359683,
+      "grad_norm": 1.9477074001689785,
+      "learning_rate": 1.1616647029032818e-06,
+      "loss": 0.24,
+      "step": 1721
+    },
+    {
+      "epoch": 2.4045570797488955,
+      "grad_norm": 2.0907606430864023,
+      "learning_rate": 1.1564621021553617e-06,
+      "loss": 0.2309,
+      "step": 1722
+    },
+    {
+      "epoch": 2.405952104161823,
+      "grad_norm": 2.0151485407481213,
+      "learning_rate": 1.1512696537697804e-06,
+      "loss": 0.2641,
+      "step": 1723
+    },
+    {
+      "epoch": 2.40734712857475,
+      "grad_norm": 2.3088467147809233,
+      "learning_rate": 1.1460873714619275e-06,
+      "loss": 0.2649,
+      "step": 1724
+    },
+    {
+      "epoch": 2.4087421529876774,
+      "grad_norm": 2.072519389501527,
+      "learning_rate": 1.140915268920339e-06,
+      "loss": 0.2409,
+      "step": 1725
+    },
+    {
+      "epoch": 2.4101371774006046,
+      "grad_norm": 1.9307910533793868,
+      "learning_rate": 1.13575335980666e-06,
+      "loss": 0.2245,
+      "step": 1726
+    },
+    {
+      "epoch": 2.411532201813532,
+      "grad_norm": 2.1765573168660355,
+      "learning_rate": 1.130601657755616e-06,
+      "loss": 0.2241,
+      "step": 1727
+    },
+    {
+      "epoch": 2.4129272262264587,
+      "grad_norm": 1.994483581997574,
+      "learning_rate": 1.125460176374965e-06,
+      "loss": 0.2287,
+      "step": 1728
+    },
+    {
+      "epoch": 2.414322250639386,
+      "grad_norm": 1.9387870571187331,
+      "learning_rate": 1.1203289292454728e-06,
+      "loss": 0.2544,
+      "step": 1729
+    },
+    {
+      "epoch": 2.4157172750523133,
+      "grad_norm": 2.1825510130358188,
+      "learning_rate": 1.1152079299208724e-06,
+      "loss": 0.2478,
+      "step": 1730
+    },
+    {
+      "epoch": 2.4171122994652405,
+      "grad_norm": 2.119670431535453,
+      "learning_rate": 1.1100971919278247e-06,
+      "loss": 0.2302,
+      "step": 1731
+    },
+    {
+      "epoch": 2.418507323878168,
+      "grad_norm": 1.9383266869156575,
+      "learning_rate": 1.104996728765887e-06,
+      "loss": 0.2428,
+      "step": 1732
+    },
+    {
+      "epoch": 2.419902348291095,
+      "grad_norm": 2.110592186403715,
+      "learning_rate": 1.0999065539074793e-06,
+      "loss": 0.2413,
+      "step": 1733
+    },
+    {
+      "epoch": 2.4212973727040223,
+      "grad_norm": 2.067613222890333,
+      "learning_rate": 1.094826680797843e-06,
+      "loss": 0.231,
+      "step": 1734
+    },
+    {
+      "epoch": 2.4226923971169496,
+      "grad_norm": 1.984326692206471,
+      "learning_rate": 1.0897571228550097e-06,
+      "loss": 0.2467,
+      "step": 1735
+    },
+    {
+      "epoch": 2.424087421529877,
+      "grad_norm": 2.0167088904753783,
+      "learning_rate": 1.0846978934697666e-06,
+      "loss": 0.2423,
+      "step": 1736
+    },
+    {
+      "epoch": 2.425482445942804,
+      "grad_norm": 1.8768002950793947,
+      "learning_rate": 1.0796490060056142e-06,
+      "loss": 0.2341,
+      "step": 1737
+    },
+    {
+      "epoch": 2.4268774703557314,
+      "grad_norm": 2.00793087438797,
+      "learning_rate": 1.074610473798738e-06,
+      "loss": 0.2293,
+      "step": 1738
+    },
+    {
+      "epoch": 2.4282724947686587,
+      "grad_norm": 2.2281107440159906,
+      "learning_rate": 1.0695823101579728e-06,
+      "loss": 0.2223,
+      "step": 1739
+    },
+    {
+      "epoch": 2.4296675191815855,
+      "grad_norm": 1.971731384171419,
+      "learning_rate": 1.0645645283647616e-06,
+      "loss": 0.2377,
+      "step": 1740
+    },
+    {
+      "epoch": 2.431062543594513,
+      "grad_norm": 2.0683911759403846,
+      "learning_rate": 1.0595571416731293e-06,
+      "loss": 0.2311,
+      "step": 1741
+    },
+    {
+      "epoch": 2.43245756800744,
+      "grad_norm": 1.9440966369219763,
+      "learning_rate": 1.0545601633096414e-06,
+      "loss": 0.2199,
+      "step": 1742
+    },
+    {
+      "epoch": 2.4338525924203673,
+      "grad_norm": 2.0991571579373662,
+      "learning_rate": 1.049573606473369e-06,
+      "loss": 0.2502,
+      "step": 1743
+    },
+    {
+      "epoch": 2.4352476168332946,
+      "grad_norm": 1.9757218408408208,
+      "learning_rate": 1.0445974843358563e-06,
+      "loss": 0.2072,
+      "step": 1744
+    },
+    {
+      "epoch": 2.436642641246222,
+      "grad_norm": 1.9940009794376974,
+      "learning_rate": 1.0396318100410868e-06,
+      "loss": 0.2224,
+      "step": 1745
+    },
+    {
+      "epoch": 2.438037665659149,
+      "grad_norm": 2.0684497758647638,
+      "learning_rate": 1.0346765967054472e-06,
+      "loss": 0.2581,
+      "step": 1746
+    },
+    {
+      "epoch": 2.4394326900720764,
+      "grad_norm": 2.277181118110738,
+      "learning_rate": 1.029731857417689e-06,
+      "loss": 0.2598,
+      "step": 1747
+    },
+    {
+      "epoch": 2.4408277144850032,
+      "grad_norm": 1.9937563449803872,
+      "learning_rate": 1.0247976052389018e-06,
+      "loss": 0.2369,
+      "step": 1748
+    },
+    {
+      "epoch": 2.4422227388979305,
+      "grad_norm": 1.9892653023975233,
+      "learning_rate": 1.0198738532024715e-06,
+      "loss": 0.2181,
+      "step": 1749
+    },
+    {
+      "epoch": 2.443617763310858,
+      "grad_norm": 1.9123071022144358,
+      "learning_rate": 1.0149606143140484e-06,
+      "loss": 0.2466,
+      "step": 1750
+    },
+    {
+      "epoch": 2.445012787723785,
+      "grad_norm": 1.9682942803900485,
+      "learning_rate": 1.0100579015515156e-06,
+      "loss": 0.2392,
+      "step": 1751
+    },
+    {
+      "epoch": 2.4464078121367123,
+      "grad_norm": 2.0981882014591786,
+      "learning_rate": 1.005165727864953e-06,
+      "loss": 0.2393,
+      "step": 1752
+    },
+    {
+      "epoch": 2.4478028365496396,
+      "grad_norm": 2.0362345779504105,
+      "learning_rate": 1.0002841061765989e-06,
+      "loss": 0.2249,
+      "step": 1753
+    },
+    {
+      "epoch": 2.449197860962567,
+      "grad_norm": 1.8793934193224626,
+      "learning_rate": 9.954130493808201e-07,
+      "loss": 0.2337,
+      "step": 1754
+    },
+    {
+      "epoch": 2.450592885375494,
+      "grad_norm": 2.165450490603712,
+      "learning_rate": 9.905525703440815e-07,
+      "loss": 0.2291,
+      "step": 1755
+    },
+    {
+      "epoch": 2.4519879097884214,
+      "grad_norm": 2.142970644431852,
+      "learning_rate": 9.85702681904902e-07,
+      "loss": 0.2133,
+      "step": 1756
+    },
+    {
+      "epoch": 2.4533829342013487,
+      "grad_norm": 2.0095221047598595,
+      "learning_rate": 9.808633968738297e-07,
+      "loss": 0.2371,
+      "step": 1757
+    },
+    {
+      "epoch": 2.454777958614276,
+      "grad_norm": 2.026353074856489,
+      "learning_rate": 9.760347280334064e-07,
+      "loss": 0.2537,
+      "step": 1758
+    },
+    {
+      "epoch": 2.456172983027203,
+      "grad_norm": 1.9325330175950954,
+      "learning_rate": 9.712166881381279e-07,
+      "loss": 0.2427,
+      "step": 1759
+    },
+    {
+      "epoch": 2.45756800744013,
+      "grad_norm": 2.0937852825424224,
+      "learning_rate": 9.664092899144156e-07,
+      "loss": 0.2497,
+      "step": 1760
+    },
+    {
+      "epoch": 2.4589630318530573,
+      "grad_norm": 1.8882699979737334,
+      "learning_rate": 9.616125460605857e-07,
+      "loss": 0.2189,
+      "step": 1761
+    },
+    {
+      "epoch": 2.4603580562659846,
+      "grad_norm": 1.9641098748743273,
+      "learning_rate": 9.56826469246806e-07,
+      "loss": 0.2163,
+      "step": 1762
+    },
+    {
+      "epoch": 2.461753080678912,
+      "grad_norm": 1.9865195042240242,
+      "learning_rate": 9.520510721150722e-07,
+      "loss": 0.2188,
+      "step": 1763
+    },
+    {
+      "epoch": 2.463148105091839,
+      "grad_norm": 2.1443424582013026,
+      "learning_rate": 9.472863672791721e-07,
+      "loss": 0.2784,
+      "step": 1764
+    },
+    {
+      "epoch": 2.4645431295047664,
+      "grad_norm": 2.1137717827169915,
+      "learning_rate": 9.425323673246461e-07,
+      "loss": 0.2322,
+      "step": 1765
+    },
+    {
+      "epoch": 2.4659381539176937,
+      "grad_norm": 2.131351078053459,
+      "learning_rate": 9.377890848087595e-07,
+      "loss": 0.2427,
+      "step": 1766
+    },
+    {
+      "epoch": 2.467333178330621,
+      "grad_norm": 2.140790053123637,
+      "learning_rate": 9.330565322604729e-07,
+      "loss": 0.2544,
+      "step": 1767
+    },
+    {
+      "epoch": 2.468728202743548,
+      "grad_norm": 2.152964895533497,
+      "learning_rate": 9.283347221803985e-07,
+      "loss": 0.2343,
+      "step": 1768
+    },
+    {
+      "epoch": 2.470123227156475,
+      "grad_norm": 2.0861415574061484,
+      "learning_rate": 9.236236670407772e-07,
+      "loss": 0.261,
+      "step": 1769
+    },
+    {
+      "epoch": 2.4715182515694023,
+      "grad_norm": 2.364724254514781,
+      "learning_rate": 9.189233792854424e-07,
+      "loss": 0.2511,
+      "step": 1770
+    },
+    {
+      "epoch": 2.4729132759823296,
+      "grad_norm": 2.0717700877473035,
+      "learning_rate": 9.142338713297838e-07,
+      "loss": 0.2462,
+      "step": 1771
+    },
+    {
+      "epoch": 2.474308300395257,
+      "grad_norm": 1.9399398996755268,
+      "learning_rate": 9.095551555607169e-07,
+      "loss": 0.2422,
+      "step": 1772
+    },
+    {
+      "epoch": 2.475703324808184,
+      "grad_norm": 2.097444199435732,
+      "learning_rate": 9.048872443366529e-07,
+      "loss": 0.2473,
+      "step": 1773
+    },
+    {
+      "epoch": 2.4770983492211114,
+      "grad_norm": 1.9711078649095015,
+      "learning_rate": 9.002301499874622e-07,
+      "loss": 0.2346,
+      "step": 1774
+    },
+    {
+      "epoch": 2.4784933736340387,
+      "grad_norm": 1.992416087567518,
+      "learning_rate": 8.955838848144449e-07,
+      "loss": 0.2225,
+      "step": 1775
+    },
+    {
+      "epoch": 2.479888398046966,
+      "grad_norm": 2.2371561835066065,
+      "learning_rate": 8.909484610902958e-07,
+      "loss": 0.2554,
+      "step": 1776
+    },
+    {
+      "epoch": 2.481283422459893,
+      "grad_norm": 1.9975667952863843,
+      "learning_rate": 8.863238910590704e-07,
+      "loss": 0.2093,
+      "step": 1777
+    },
+    {
+      "epoch": 2.4826784468728205,
+      "grad_norm": 1.986118930839998,
+      "learning_rate": 8.817101869361599e-07,
+      "loss": 0.2481,
+      "step": 1778
+    },
+    {
+      "epoch": 2.4840734712857477,
+      "grad_norm": 2.142304590119852,
+      "learning_rate": 8.77107360908253e-07,
+      "loss": 0.264,
+      "step": 1779
+    },
+    {
+      "epoch": 2.4854684956986746,
+      "grad_norm": 2.1182415329232014,
+      "learning_rate": 8.725154251333012e-07,
+      "loss": 0.2328,
+      "step": 1780
+    },
+    {
+      "epoch": 2.486863520111602,
+      "grad_norm": 2.2640669840530507,
+      "learning_rate": 8.679343917404959e-07,
+      "loss": 0.2952,
+      "step": 1781
+    },
+    {
+      "epoch": 2.488258544524529,
+      "grad_norm": 1.913135749354688,
+      "learning_rate": 8.633642728302266e-07,
+      "loss": 0.239,
+      "step": 1782
+    },
+    {
+      "epoch": 2.4896535689374564,
+      "grad_norm": 2.1042684303606634,
+      "learning_rate": 8.588050804740527e-07,
+      "loss": 0.2282,
+      "step": 1783
+    },
+    {
+      "epoch": 2.4910485933503836,
+      "grad_norm": 2.0527372099519385,
+      "learning_rate": 8.542568267146761e-07,
+      "loss": 0.2184,
+      "step": 1784
+    },
+    {
+      "epoch": 2.492443617763311,
+      "grad_norm": 1.8290280740715956,
+      "learning_rate": 8.49719523565904e-07,
+      "loss": 0.2139,
+      "step": 1785
+    },
+    {
+      "epoch": 2.493838642176238,
+      "grad_norm": 2.142700588699554,
+      "learning_rate": 8.451931830126148e-07,
+      "loss": 0.2395,
+      "step": 1786
+    },
+    {
+      "epoch": 2.4952336665891655,
+      "grad_norm": 1.9887460560020362,
+      "learning_rate": 8.40677817010736e-07,
+      "loss": 0.2276,
+      "step": 1787
+    },
+    {
+      "epoch": 2.4966286910020927,
+      "grad_norm": 2.149857725209497,
+      "learning_rate": 8.361734374872032e-07,
+      "loss": 0.2396,
+      "step": 1788
+    },
+    {
+      "epoch": 2.4980237154150196,
+      "grad_norm": 1.8153829921701001,
+      "learning_rate": 8.316800563399307e-07,
+      "loss": 0.2243,
+      "step": 1789
+    },
+    {
+      "epoch": 2.499418739827947,
+      "grad_norm": 2.0873477016865003,
+      "learning_rate": 8.271976854377861e-07,
+      "loss": 0.2184,
+      "step": 1790
+    },
+    {
+      "epoch": 2.500813764240874,
+      "grad_norm": 1.9058032255841029,
+      "learning_rate": 8.227263366205523e-07,
+      "loss": 0.2617,
+      "step": 1791
+    },
+    {
+      "epoch": 2.5022087886538014,
+      "grad_norm": 2.03985002067712,
+      "learning_rate": 8.182660216988964e-07,
+      "loss": 0.3254,
+      "step": 1792
+    },
+    {
+      "epoch": 2.5036038130667286,
+      "grad_norm": 2.4462443382032104,
+      "learning_rate": 8.138167524543445e-07,
+      "loss": 0.2494,
+      "step": 1793
+    },
+    {
+      "epoch": 2.504998837479656,
+      "grad_norm": 2.1759864601852534,
+      "learning_rate": 8.09378540639243e-07,
+      "loss": 0.2518,
+      "step": 1794
+    },
+    {
+      "epoch": 2.506393861892583,
+      "grad_norm": 2.0809827238097336,
+      "learning_rate": 8.049513979767304e-07,
+      "loss": 0.2461,
+      "step": 1795
+    },
+    {
+      "epoch": 2.5077888863055104,
+      "grad_norm": 1.9241245111870922,
+      "learning_rate": 8.00535336160711e-07,
+      "loss": 0.1999,
+      "step": 1796
+    },
+    {
+      "epoch": 2.5091839107184377,
+      "grad_norm": 2.1413620625370475,
+      "learning_rate": 7.96130366855819e-07,
+      "loss": 0.2423,
+      "step": 1797
+    },
+    {
+      "epoch": 2.510578935131365,
+      "grad_norm": 2.1005332347594354,
+      "learning_rate": 7.917365016973866e-07,
+      "loss": 0.2203,
+      "step": 1798
+    },
+    {
+      "epoch": 2.5119739595442923,
+      "grad_norm": 2.146855579080123,
+      "learning_rate": 7.873537522914155e-07,
+      "loss": 0.2518,
+      "step": 1799
+    },
+    {
+      "epoch": 2.5133689839572195,
+      "grad_norm": 2.084590198938967,
+      "learning_rate": 7.829821302145485e-07,
+      "loss": 0.2359,
+      "step": 1800
+    },
+    {
+      "epoch": 2.5147640083701464,
+      "grad_norm": 1.9631135365144416,
+      "learning_rate": 7.786216470140334e-07,
+      "loss": 0.2333,
+      "step": 1801
+    },
+    {
+      "epoch": 2.5161590327830736,
+      "grad_norm": 2.0533634112072474,
+      "learning_rate": 7.742723142076991e-07,
+      "loss": 0.2298,
+      "step": 1802
+    },
+    {
+      "epoch": 2.517554057196001,
+      "grad_norm": 1.9899219985356666,
+      "learning_rate": 7.699341432839203e-07,
+      "loss": 0.245,
+      "step": 1803
+    },
+    {
+      "epoch": 2.518949081608928,
+      "grad_norm": 1.9831540992143863,
+      "learning_rate": 7.656071457015879e-07,
+      "loss": 0.2082,
+      "step": 1804
+    },
+    {
+      "epoch": 2.5203441060218554,
+      "grad_norm": 2.0248350609157786,
+      "learning_rate": 7.612913328900784e-07,
+      "loss": 0.2297,
+      "step": 1805
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 2.0684064839827028,
+      "learning_rate": 7.569867162492283e-07,
+      "loss": 0.2006,
+      "step": 1806
+    },
+    {
+      "epoch": 2.52313415484771,
+      "grad_norm": 1.9978028339758431,
+      "learning_rate": 7.526933071492959e-07,
+      "loss": 0.2649,
+      "step": 1807
+    },
+    {
+      "epoch": 2.524529179260637,
+      "grad_norm": 2.2154732847751744,
+      "learning_rate": 7.484111169309399e-07,
+      "loss": 0.2274,
+      "step": 1808
+    },
+    {
+      "epoch": 2.525924203673564,
+      "grad_norm": 2.035660314013249,
+      "learning_rate": 7.441401569051848e-07,
+      "loss": 0.2304,
+      "step": 1809
+    },
+    {
+      "epoch": 2.5273192280864913,
+      "grad_norm": 2.1081064276269754,
+      "learning_rate": 7.398804383533886e-07,
+      "loss": 0.2153,
+      "step": 1810
+    },
+    {
+      "epoch": 2.5287142524994186,
+      "grad_norm": 2.103371622818518,
+      "learning_rate": 7.356319725272165e-07,
+      "loss": 0.2349,
+      "step": 1811
+    },
+    {
+      "epoch": 2.530109276912346,
+      "grad_norm": 1.9604309959647959,
+      "learning_rate": 7.313947706486136e-07,
+      "loss": 0.2438,
+      "step": 1812
+    },
+    {
+      "epoch": 2.531504301325273,
+      "grad_norm": 2.178669534413003,
+      "learning_rate": 7.271688439097713e-07,
+      "loss": 0.2594,
+      "step": 1813
+    },
+    {
+      "epoch": 2.5328993257382004,
+      "grad_norm": 1.9811957969660663,
+      "learning_rate": 7.229542034730952e-07,
+      "loss": 0.1927,
+      "step": 1814
+    },
+    {
+      "epoch": 2.5342943501511277,
+      "grad_norm": 2.202042500358503,
+      "learning_rate": 7.187508604711851e-07,
+      "loss": 0.2412,
+      "step": 1815
+    },
+    {
+      "epoch": 2.535689374564055,
+      "grad_norm": 2.027289212836733,
+      "learning_rate": 7.145588260067943e-07,
+      "loss": 0.2534,
+      "step": 1816
+    },
+    {
+      "epoch": 2.5370843989769822,
+      "grad_norm": 2.000708288343631,
+      "learning_rate": 7.103781111528074e-07,
+      "loss": 0.239,
+      "step": 1817
+    },
+    {
+      "epoch": 2.5384794233899095,
+      "grad_norm": 2.1008206829818907,
+      "learning_rate": 7.062087269522105e-07,
+      "loss": 0.208,
+      "step": 1818
+    },
+    {
+      "epoch": 2.5398744478028368,
+      "grad_norm": 1.7956015898367612,
+      "learning_rate": 7.020506844180608e-07,
+      "loss": 0.2397,
+      "step": 1819
+    },
+    {
+      "epoch": 2.541269472215764,
+      "grad_norm": 2.3670874992202298,
+      "learning_rate": 6.979039945334543e-07,
+      "loss": 0.245,
+      "step": 1820
+    },
+    {
+      "epoch": 2.542664496628691,
+      "grad_norm": 2.0068429751335835,
+      "learning_rate": 6.937686682515044e-07,
+      "loss": 0.2358,
+      "step": 1821
+    },
+    {
+      "epoch": 2.544059521041618,
+      "grad_norm": 2.1746011102665554,
+      "learning_rate": 6.896447164953057e-07,
+      "loss": 0.2563,
+      "step": 1822
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 2.0644119660815936,
+      "learning_rate": 6.855321501579077e-07,
+      "loss": 0.2371,
+      "step": 1823
+    },
+    {
+      "epoch": 2.5468495698674727,
+      "grad_norm": 2.116923220111308,
+      "learning_rate": 6.814309801022873e-07,
+      "loss": 0.2388,
+      "step": 1824
+    },
+    {
+      "epoch": 2.5482445942804,
+      "grad_norm": 1.983451872015146,
+      "learning_rate": 6.77341217161322e-07,
+      "loss": 0.2625,
+      "step": 1825
+    },
+    {
+      "epoch": 2.5496396186933272,
+      "grad_norm": 2.0390077022283326,
+      "learning_rate": 6.732628721377533e-07,
+      "loss": 0.2172,
+      "step": 1826
+    },
+    {
+      "epoch": 2.5510346431062545,
+      "grad_norm": 2.0429464708845404,
+      "learning_rate": 6.69195955804165e-07,
+      "loss": 0.289,
+      "step": 1827
+    },
+    {
+      "epoch": 2.5524296675191813,
+      "grad_norm": 2.1879865095437,
+      "learning_rate": 6.651404789029553e-07,
+      "loss": 0.2345,
+      "step": 1828
+    },
+    {
+      "epoch": 2.5538246919321086,
+      "grad_norm": 1.772535273422064,
+      "learning_rate": 6.610964521463032e-07,
+      "loss": 0.2353,
+      "step": 1829
+    },
+    {
+      "epoch": 2.555219716345036,
+      "grad_norm": 2.041075828824093,
+      "learning_rate": 6.570638862161449e-07,
+      "loss": 0.2114,
+      "step": 1830
+    },
+    {
+      "epoch": 2.556614740757963,
+      "grad_norm": 1.945830106436581,
+      "learning_rate": 6.530427917641447e-07,
+      "loss": 0.2164,
+      "step": 1831
+    },
+    {
+      "epoch": 2.5580097651708904,
+      "grad_norm": 2.0753411827360915,
+      "learning_rate": 6.490331794116633e-07,
+      "loss": 0.2533,
+      "step": 1832
+    },
+    {
+      "epoch": 2.5594047895838177,
+      "grad_norm": 1.7446405117787227,
+      "learning_rate": 6.450350597497335e-07,
+      "loss": 0.2094,
+      "step": 1833
+    },
+    {
+      "epoch": 2.560799813996745,
+      "grad_norm": 2.0417418372007456,
+      "learning_rate": 6.410484433390335e-07,
+      "loss": 0.2552,
+      "step": 1834
+    },
+    {
+      "epoch": 2.562194838409672,
+      "grad_norm": 1.9407852162679993,
+      "learning_rate": 6.370733407098517e-07,
+      "loss": 0.2034,
+      "step": 1835
+    },
+    {
+      "epoch": 2.5635898628225995,
+      "grad_norm": 2.0514788706278275,
+      "learning_rate": 6.331097623620697e-07,
+      "loss": 0.2379,
+      "step": 1836
+    },
+    {
+      "epoch": 2.5649848872355268,
+      "grad_norm": 1.8667937445523195,
+      "learning_rate": 6.291577187651255e-07,
+      "loss": 0.223,
+      "step": 1837
+    },
+    {
+      "epoch": 2.566379911648454,
+      "grad_norm": 2.0783967507217445,
+      "learning_rate": 6.252172203579892e-07,
+      "loss": 0.2411,
+      "step": 1838
+    },
+    {
+      "epoch": 2.5677749360613813,
+      "grad_norm": 2.0226917795020602,
+      "learning_rate": 6.212882775491352e-07,
+      "loss": 0.269,
+      "step": 1839
+    },
+    {
+      "epoch": 2.5691699604743086,
+      "grad_norm": 1.977576382502622,
+      "learning_rate": 6.173709007165158e-07,
+      "loss": 0.2432,
+      "step": 1840
+    },
+    {
+      "epoch": 2.5705649848872354,
+      "grad_norm": 2.113865208840359,
+      "learning_rate": 6.134651002075315e-07,
+      "loss": 0.2335,
+      "step": 1841
+    },
+    {
+      "epoch": 2.5719600093001627,
+      "grad_norm": 2.111195344110925,
+      "learning_rate": 6.095708863390065e-07,
+      "loss": 0.2428,
+      "step": 1842
+    },
+    {
+      "epoch": 2.57335503371309,
+      "grad_norm": 1.8766206380320936,
+      "learning_rate": 6.056882693971605e-07,
+      "loss": 0.2201,
+      "step": 1843
+    },
+    {
+      "epoch": 2.574750058126017,
+      "grad_norm": 2.09593489201861,
+      "learning_rate": 6.018172596375776e-07,
+      "loss": 0.2652,
+      "step": 1844
+    },
+    {
+      "epoch": 2.5761450825389445,
+      "grad_norm": 1.971523019642963,
+      "learning_rate": 5.979578672851843e-07,
+      "loss": 0.2481,
+      "step": 1845
+    },
+    {
+      "epoch": 2.5775401069518717,
+      "grad_norm": 2.0922975506755646,
+      "learning_rate": 5.941101025342239e-07,
+      "loss": 0.2221,
+      "step": 1846
+    },
+    {
+      "epoch": 2.578935131364799,
+      "grad_norm": 2.2444677637837653,
+      "learning_rate": 5.902739755482201e-07,
+      "loss": 0.29,
+      "step": 1847
+    },
+    {
+      "epoch": 2.580330155777726,
+      "grad_norm": 1.963684451073659,
+      "learning_rate": 5.864494964599615e-07,
+      "loss": 0.2135,
+      "step": 1848
+    },
+    {
+      "epoch": 2.581725180190653,
+      "grad_norm": 2.008525153651005,
+      "learning_rate": 5.826366753714707e-07,
+      "loss": 0.2359,
+      "step": 1849
+    },
+    {
+      "epoch": 2.5831202046035804,
+      "grad_norm": 1.8327259871668735,
+      "learning_rate": 5.788355223539698e-07,
+      "loss": 0.2244,
+      "step": 1850
+    },
+    {
+      "epoch": 2.5845152290165077,
+      "grad_norm": 2.1182315396159876,
+      "learning_rate": 5.750460474478675e-07,
+      "loss": 0.2228,
+      "step": 1851
+    },
+    {
+      "epoch": 2.585910253429435,
+      "grad_norm": 1.8049971008688241,
+      "learning_rate": 5.712682606627251e-07,
+      "loss": 0.2294,
+      "step": 1852
+    },
+    {
+      "epoch": 2.587305277842362,
+      "grad_norm": 2.23181809180191,
+      "learning_rate": 5.675021719772262e-07,
+      "loss": 0.2463,
+      "step": 1853
+    },
+    {
+      "epoch": 2.5887003022552895,
+      "grad_norm": 1.9925138661937265,
+      "learning_rate": 5.637477913391604e-07,
+      "loss": 0.2242,
+      "step": 1854
+    },
+    {
+      "epoch": 2.5900953266682167,
+      "grad_norm": 2.010379835853784,
+      "learning_rate": 5.600051286653884e-07,
+      "loss": 0.2568,
+      "step": 1855
+    },
+    {
+      "epoch": 2.591490351081144,
+      "grad_norm": 2.1668861502665884,
+      "learning_rate": 5.562741938418187e-07,
+      "loss": 0.2281,
+      "step": 1856
+    },
+    {
+      "epoch": 2.5928853754940713,
+      "grad_norm": 2.0561471815296843,
+      "learning_rate": 5.525549967233829e-07,
+      "loss": 0.2368,
+      "step": 1857
+    },
+    {
+      "epoch": 2.5942803999069985,
+      "grad_norm": 2.0978158108861567,
+      "learning_rate": 5.488475471340099e-07,
+      "loss": 0.2341,
+      "step": 1858
+    },
+    {
+      "epoch": 2.595675424319926,
+      "grad_norm": 2.085462389890947,
+      "learning_rate": 5.451518548665946e-07,
+      "loss": 0.2429,
+      "step": 1859
+    },
+    {
+      "epoch": 2.597070448732853,
+      "grad_norm": 1.806559468372376,
+      "learning_rate": 5.414679296829806e-07,
+      "loss": 0.2287,
+      "step": 1860
+    },
+    {
+      "epoch": 2.59846547314578,
+      "grad_norm": 1.7938452572055765,
+      "learning_rate": 5.377957813139262e-07,
+      "loss": 0.1991,
+      "step": 1861
+    },
+    {
+      "epoch": 2.599860497558707,
+      "grad_norm": 2.1163670152268446,
+      "learning_rate": 5.341354194590831e-07,
+      "loss": 0.3004,
+      "step": 1862
+    },
+    {
+      "epoch": 2.6012555219716345,
+      "grad_norm": 2.15912048187847,
+      "learning_rate": 5.304868537869706e-07,
+      "loss": 0.2219,
+      "step": 1863
+    },
+    {
+      "epoch": 2.6026505463845617,
+      "grad_norm": 1.843409231643337,
+      "learning_rate": 5.268500939349514e-07,
+      "loss": 0.2283,
+      "step": 1864
+    },
+    {
+      "epoch": 2.604045570797489,
+      "grad_norm": 1.9210220245156822,
+      "learning_rate": 5.232251495091989e-07,
+      "loss": 0.2133,
+      "step": 1865
+    },
+    {
+      "epoch": 2.6054405952104163,
+      "grad_norm": 1.8978056842019908,
+      "learning_rate": 5.196120300846835e-07,
+      "loss": 0.221,
+      "step": 1866
+    },
+    {
+      "epoch": 2.6068356196233435,
+      "grad_norm": 1.9173007014525827,
+      "learning_rate": 5.160107452051361e-07,
+      "loss": 0.2478,
+      "step": 1867
+    },
+    {
+      "epoch": 2.6082306440362704,
+      "grad_norm": 2.063458389194393,
+      "learning_rate": 5.124213043830278e-07,
+      "loss": 0.2033,
+      "step": 1868
+    },
+    {
+      "epoch": 2.6096256684491976,
+      "grad_norm": 1.964758365774162,
+      "learning_rate": 5.088437170995481e-07,
+      "loss": 0.2132,
+      "step": 1869
+    },
+    {
+      "epoch": 2.611020692862125,
+      "grad_norm": 1.8757381329933205,
+      "learning_rate": 5.052779928045737e-07,
+      "loss": 0.2257,
+      "step": 1870
+    },
+    {
+      "epoch": 2.612415717275052,
+      "grad_norm": 1.9938237138566777,
+      "learning_rate": 5.01724140916649e-07,
+      "loss": 0.2365,
+      "step": 1871
+    },
+    {
+      "epoch": 2.6138107416879794,
+      "grad_norm": 2.057328367380359,
+      "learning_rate": 4.981821708229545e-07,
+      "loss": 0.2523,
+      "step": 1872
+    },
+    {
+      "epoch": 2.6152057661009067,
+      "grad_norm": 2.093011615403081,
+      "learning_rate": 4.946520918792886e-07,
+      "loss": 0.22,
+      "step": 1873
+    },
+    {
+      "epoch": 2.616600790513834,
+      "grad_norm": 2.0856658826067878,
+      "learning_rate": 4.911339134100401e-07,
+      "loss": 0.2504,
+      "step": 1874
+    },
+    {
+      "epoch": 2.6179958149267613,
+      "grad_norm": 2.2370140372218805,
+      "learning_rate": 4.87627644708163e-07,
+      "loss": 0.2391,
+      "step": 1875
+    },
+    {
+      "epoch": 2.6193908393396885,
+      "grad_norm": 1.992843597428801,
+      "learning_rate": 4.841332950351535e-07,
+      "loss": 0.2452,
+      "step": 1876
+    },
+    {
+      "epoch": 2.620785863752616,
+      "grad_norm": 2.036241454180426,
+      "learning_rate": 4.806508736210253e-07,
+      "loss": 0.2378,
+      "step": 1877
+    },
+    {
+      "epoch": 2.622180888165543,
+      "grad_norm": 1.8430704939397233,
+      "learning_rate": 4.771803896642812e-07,
+      "loss": 0.218,
+      "step": 1878
+    },
+    {
+      "epoch": 2.6235759125784703,
+      "grad_norm": 1.9168731440964641,
+      "learning_rate": 4.737218523318965e-07,
+      "loss": 0.2575,
+      "step": 1879
+    },
+    {
+      "epoch": 2.6249709369913976,
+      "grad_norm": 1.9668663396982646,
+      "learning_rate": 4.7027527075929e-07,
+      "loss": 0.2195,
+      "step": 1880
+    },
+    {
+      "epoch": 2.6263659614043244,
+      "grad_norm": 2.1482072780264247,
+      "learning_rate": 4.6684065405029677e-07,
+      "loss": 0.2384,
+      "step": 1881
+    },
+    {
+      "epoch": 2.6277609858172517,
+      "grad_norm": 1.936941683232777,
+      "learning_rate": 4.6341801127715303e-07,
+      "loss": 0.2247,
+      "step": 1882
+    },
+    {
+      "epoch": 2.629156010230179,
+      "grad_norm": 2.010168782321282,
+      "learning_rate": 4.6000735148046316e-07,
+      "loss": 0.2536,
+      "step": 1883
+    },
+    {
+      "epoch": 2.6305510346431062,
+      "grad_norm": 2.0449854159937937,
+      "learning_rate": 4.566086836691791e-07,
+      "loss": 0.2277,
+      "step": 1884
+    },
+    {
+      "epoch": 2.6319460590560335,
+      "grad_norm": 1.9233425119697793,
+      "learning_rate": 4.532220168205798e-07,
+      "loss": 0.2421,
+      "step": 1885
+    },
+    {
+      "epoch": 2.633341083468961,
+      "grad_norm": 2.0816496032711003,
+      "learning_rate": 4.498473598802444e-07,
+      "loss": 0.2053,
+      "step": 1886
+    },
+    {
+      "epoch": 2.634736107881888,
+      "grad_norm": 2.013648412197882,
+      "learning_rate": 4.464847217620266e-07,
+      "loss": 0.2315,
+      "step": 1887
+    },
+    {
+      "epoch": 2.636131132294815,
+      "grad_norm": 1.9879522597479942,
+      "learning_rate": 4.4313411134803584e-07,
+      "loss": 0.2463,
+      "step": 1888
+    },
+    {
+      "epoch": 2.637526156707742,
+      "grad_norm": 2.0669142465872024,
+      "learning_rate": 4.397955374886104e-07,
+      "loss": 0.2174,
+      "step": 1889
+    },
+    {
+      "epoch": 2.6389211811206694,
+      "grad_norm": 2.048055292937162,
+      "learning_rate": 4.364690090022938e-07,
+      "loss": 0.2393,
+      "step": 1890
+    },
+    {
+      "epoch": 2.6403162055335967,
+      "grad_norm": 1.9212520880953143,
+      "learning_rate": 4.331545346758159e-07,
+      "loss": 0.2322,
+      "step": 1891
+    },
+    {
+      "epoch": 2.641711229946524,
+      "grad_norm": 1.9897143304092708,
+      "learning_rate": 4.2985212326406456e-07,
+      "loss": 0.2112,
+      "step": 1892
+    },
+    {
+      "epoch": 2.6431062543594512,
+      "grad_norm": 1.994917321745064,
+      "learning_rate": 4.265617834900637e-07,
+      "loss": 0.2139,
+      "step": 1893
+    },
+    {
+      "epoch": 2.6445012787723785,
+      "grad_norm": 1.9275707655087257,
+      "learning_rate": 4.2328352404495346e-07,
+      "loss": 0.2031,
+      "step": 1894
+    },
+    {
+      "epoch": 2.6458963031853058,
+      "grad_norm": 2.114603415128938,
+      "learning_rate": 4.2001735358796316e-07,
+      "loss": 0.2053,
+      "step": 1895
+    },
+    {
+      "epoch": 2.647291327598233,
+      "grad_norm": 1.7540496628185678,
+      "learning_rate": 4.167632807463895e-07,
+      "loss": 0.2058,
+      "step": 1896
+    },
+    {
+      "epoch": 2.6486863520111603,
+      "grad_norm": 1.9772733263814442,
+      "learning_rate": 4.135213141155769e-07,
+      "loss": 0.227,
+      "step": 1897
+    },
+    {
+      "epoch": 2.6500813764240876,
+      "grad_norm": 2.122744145357744,
+      "learning_rate": 4.1029146225889103e-07,
+      "loss": 0.2367,
+      "step": 1898
+    },
+    {
+      "epoch": 2.651476400837015,
+      "grad_norm": 2.0606193799135797,
+      "learning_rate": 4.0707373370769634e-07,
+      "loss": 0.2345,
+      "step": 1899
+    },
+    {
+      "epoch": 2.652871425249942,
+      "grad_norm": 2.037156489409722,
+      "learning_rate": 4.0386813696133564e-07,
+      "loss": 0.235,
+      "step": 1900
+    },
+    {
+      "epoch": 2.654266449662869,
+      "grad_norm": 2.025359743039822,
+      "learning_rate": 4.0067468048710756e-07,
+      "loss": 0.2194,
+      "step": 1901
+    },
+    {
+      "epoch": 2.6556614740757962,
+      "grad_norm": 1.9308904352486622,
+      "learning_rate": 3.974933727202412e-07,
+      "loss": 0.228,
+      "step": 1902
+    },
+    {
+      "epoch": 2.6570564984887235,
+      "grad_norm": 1.8104587056449903,
+      "learning_rate": 3.943242220638777e-07,
+      "loss": 0.2135,
+      "step": 1903
+    },
+    {
+      "epoch": 2.6584515229016508,
+      "grad_norm": 1.991585390255207,
+      "learning_rate": 3.911672368890462e-07,
+      "loss": 0.2403,
+      "step": 1904
+    },
+    {
+      "epoch": 2.659846547314578,
+      "grad_norm": 2.1170487598745242,
+      "learning_rate": 3.8802242553464096e-07,
+      "loss": 0.2307,
+      "step": 1905
+    },
+    {
+      "epoch": 2.6612415717275053,
+      "grad_norm": 2.0282853922616075,
+      "learning_rate": 3.8488979630739996e-07,
+      "loss": 0.2272,
+      "step": 1906
+    },
+    {
+      "epoch": 2.6626365961404326,
+      "grad_norm": 2.0265308519331855,
+      "learning_rate": 3.8176935748188425e-07,
+      "loss": 0.1872,
+      "step": 1907
+    },
+    {
+      "epoch": 2.66403162055336,
+      "grad_norm": 1.7643254590081145,
+      "learning_rate": 3.78661117300454e-07,
+      "loss": 0.2069,
+      "step": 1908
+    },
+    {
+      "epoch": 2.6654266449662867,
+      "grad_norm": 1.986084570077434,
+      "learning_rate": 3.755650839732489e-07,
+      "loss": 0.2139,
+      "step": 1909
+    },
+    {
+      "epoch": 2.666821669379214,
+      "grad_norm": 1.8960562363541258,
+      "learning_rate": 3.7248126567816454e-07,
+      "loss": 0.2557,
+      "step": 1910
+    },
+    {
+      "epoch": 2.668216693792141,
+      "grad_norm": 2.0653357761405995,
+      "learning_rate": 3.694096705608319e-07,
+      "loss": 0.2375,
+      "step": 1911
+    },
+    {
+      "epoch": 2.6696117182050685,
+      "grad_norm": 2.1077724754663407,
+      "learning_rate": 3.6635030673459413e-07,
+      "loss": 0.2215,
+      "step": 1912
+    },
+    {
+      "epoch": 2.6710067426179958,
+      "grad_norm": 2.0513233906590247,
+      "learning_rate": 3.6330318228049e-07,
+      "loss": 0.248,
+      "step": 1913
+    },
+    {
+      "epoch": 2.672401767030923,
+      "grad_norm": 1.9265363937989877,
+      "learning_rate": 3.6026830524722443e-07,
+      "loss": 0.2407,
+      "step": 1914
+    },
+    {
+      "epoch": 2.6737967914438503,
+      "grad_norm": 1.9526513264278522,
+      "learning_rate": 3.572456836511551e-07,
+      "loss": 0.2443,
+      "step": 1915
+    },
+    {
+      "epoch": 2.6751918158567776,
+      "grad_norm": 1.9303284847693432,
+      "learning_rate": 3.5423532547626816e-07,
+      "loss": 0.2205,
+      "step": 1916
+    },
+    {
+      "epoch": 2.676586840269705,
+      "grad_norm": 1.8811607733432787,
+      "learning_rate": 3.5123723867415527e-07,
+      "loss": 0.1948,
+      "step": 1917
+    },
+    {
+      "epoch": 2.677981864682632,
+      "grad_norm": 2.03022918053246,
+      "learning_rate": 3.4825143116399454e-07,
+      "loss": 0.2197,
+      "step": 1918
+    },
+    {
+      "epoch": 2.6793768890955594,
+      "grad_norm": 2.2929035930250397,
+      "learning_rate": 3.452779108325316e-07,
+      "loss": 0.2173,
+      "step": 1919
+    },
+    {
+      "epoch": 2.6807719135084866,
+      "grad_norm": 2.001488086541109,
+      "learning_rate": 3.4231668553405316e-07,
+      "loss": 0.2216,
+      "step": 1920
+    },
+    {
+      "epoch": 2.6821669379214135,
+      "grad_norm": 2.0989052899619187,
+      "learning_rate": 3.39367763090373e-07,
+      "loss": 0.2404,
+      "step": 1921
+    },
+    {
+      "epoch": 2.6835619623343407,
+      "grad_norm": 2.053071087875118,
+      "learning_rate": 3.3643115129080695e-07,
+      "loss": 0.2371,
+      "step": 1922
+    },
+    {
+      "epoch": 2.684956986747268,
+      "grad_norm": 2.003778312547755,
+      "learning_rate": 3.3350685789215133e-07,
+      "loss": 0.2249,
+      "step": 1923
+    },
+    {
+      "epoch": 2.6863520111601953,
+      "grad_norm": 2.091480355902784,
+      "learning_rate": 3.3059489061866625e-07,
+      "loss": 0.2453,
+      "step": 1924
+    },
+    {
+      "epoch": 2.6877470355731226,
+      "grad_norm": 2.1456977629987772,
+      "learning_rate": 3.276952571620556e-07,
+      "loss": 0.2521,
+      "step": 1925
+    },
+    {
+      "epoch": 2.68914205998605,
+      "grad_norm": 2.075417810454784,
+      "learning_rate": 3.248079651814395e-07,
+      "loss": 0.231,
+      "step": 1926
+    },
+    {
+      "epoch": 2.690537084398977,
+      "grad_norm": 1.8386109201307006,
+      "learning_rate": 3.2193302230334455e-07,
+      "loss": 0.196,
+      "step": 1927
+    },
+    {
+      "epoch": 2.6919321088119044,
+      "grad_norm": 1.8455305977705043,
+      "learning_rate": 3.190704361216751e-07,
+      "loss": 0.2119,
+      "step": 1928
+    },
+    {
+      "epoch": 2.693327133224831,
+      "grad_norm": 2.139309615920741,
+      "learning_rate": 3.162202141976956e-07,
+      "loss": 0.2647,
+      "step": 1929
+    },
+    {
+      "epoch": 2.6947221576377585,
+      "grad_norm": 2.186763765790992,
+      "learning_rate": 3.133823640600137e-07,
+      "loss": 0.2589,
+      "step": 1930
+    },
+    {
+      "epoch": 2.6961171820506857,
+      "grad_norm": 1.9436673585389133,
+      "learning_rate": 3.105568932045577e-07,
+      "loss": 0.2499,
+      "step": 1931
+    },
+    {
+      "epoch": 2.697512206463613,
+      "grad_norm": 2.048763078121567,
+      "learning_rate": 3.077438090945573e-07,
+      "loss": 0.2188,
+      "step": 1932
+    },
+    {
+      "epoch": 2.6989072308765403,
+      "grad_norm": 2.1053994973981487,
+      "learning_rate": 3.0494311916052234e-07,
+      "loss": 0.2464,
+      "step": 1933
+    },
+    {
+      "epoch": 2.7003022552894675,
+      "grad_norm": 2.0192397420316506,
+      "learning_rate": 3.021548308002248e-07,
+      "loss": 0.2486,
+      "step": 1934
+    },
+    {
+      "epoch": 2.701697279702395,
+      "grad_norm": 1.9945367846333557,
+      "learning_rate": 2.9937895137868046e-07,
+      "loss": 0.2329,
+      "step": 1935
+    },
+    {
+      "epoch": 2.703092304115322,
+      "grad_norm": 2.155115974239163,
+      "learning_rate": 2.9661548822812636e-07,
+      "loss": 0.2162,
+      "step": 1936
+    },
+    {
+      "epoch": 2.7044873285282494,
+      "grad_norm": 1.844769199810483,
+      "learning_rate": 2.9386444864800355e-07,
+      "loss": 0.2554,
+      "step": 1937
+    },
+    {
+      "epoch": 2.7058823529411766,
+      "grad_norm": 1.9952354992173393,
+      "learning_rate": 2.911258399049394e-07,
+      "loss": 0.2677,
+      "step": 1938
+    },
+    {
+      "epoch": 2.707277377354104,
+      "grad_norm": 2.136313468913573,
+      "learning_rate": 2.8839966923272286e-07,
+      "loss": 0.2281,
+      "step": 1939
+    },
+    {
+      "epoch": 2.708672401767031,
+      "grad_norm": 2.037309497070798,
+      "learning_rate": 2.8568594383229067e-07,
+      "loss": 0.2527,
+      "step": 1940
+    },
+    {
+      "epoch": 2.710067426179958,
+      "grad_norm": 2.177748542840591,
+      "learning_rate": 2.8298467087170655e-07,
+      "loss": 0.2567,
+      "step": 1941
+    },
+    {
+      "epoch": 2.7114624505928853,
+      "grad_norm": 2.0205312029912563,
+      "learning_rate": 2.8029585748614196e-07,
+      "loss": 0.2419,
+      "step": 1942
+    },
+    {
+      "epoch": 2.7128574750058125,
+      "grad_norm": 2.0495681104433046,
+      "learning_rate": 2.7761951077785676e-07,
+      "loss": 0.2418,
+      "step": 1943
+    },
+    {
+      "epoch": 2.71425249941874,
+      "grad_norm": 2.0238763267167275,
+      "learning_rate": 2.749556378161833e-07,
+      "loss": 0.2247,
+      "step": 1944
+    },
+    {
+      "epoch": 2.715647523831667,
+      "grad_norm": 2.0799212973766887,
+      "learning_rate": 2.723042456375036e-07,
+      "loss": 0.2214,
+      "step": 1945
+    },
+    {
+      "epoch": 2.7170425482445943,
+      "grad_norm": 1.9828707876406386,
+      "learning_rate": 2.696653412452327e-07,
+      "loss": 0.2254,
+      "step": 1946
+    },
+    {
+      "epoch": 2.7184375726575216,
+      "grad_norm": 1.9574822824704652,
+      "learning_rate": 2.6703893160980266e-07,
+      "loss": 0.2315,
+      "step": 1947
+    },
+    {
+      "epoch": 2.719832597070449,
+      "grad_norm": 2.00140036288395,
+      "learning_rate": 2.6442502366863854e-07,
+      "loss": 0.2433,
+      "step": 1948
+    },
+    {
+      "epoch": 2.7212276214833757,
+      "grad_norm": 2.0476945336094827,
+      "learning_rate": 2.618236243261452e-07,
+      "loss": 0.2334,
+      "step": 1949
+    },
+    {
+      "epoch": 2.722622645896303,
+      "grad_norm": 2.1346972742135892,
+      "learning_rate": 2.592347404536888e-07,
+      "loss": 0.2349,
+      "step": 1950
+    },
+    {
+      "epoch": 2.7240176703092303,
+      "grad_norm": 2.1642854854702773,
+      "learning_rate": 2.566583788895721e-07,
+      "loss": 0.2348,
+      "step": 1951
+    },
+    {
+      "epoch": 2.7254126947221575,
+      "grad_norm": 1.9708668257935886,
+      "learning_rate": 2.5409454643902543e-07,
+      "loss": 0.2381,
+      "step": 1952
+    },
+    {
+      "epoch": 2.726807719135085,
+      "grad_norm": 2.059735784249374,
+      "learning_rate": 2.5154324987418434e-07,
+      "loss": 0.2414,
+      "step": 1953
+    },
+    {
+      "epoch": 2.728202743548012,
+      "grad_norm": 2.004488455511273,
+      "learning_rate": 2.4900449593406984e-07,
+      "loss": 0.2245,
+      "step": 1954
+    },
+    {
+      "epoch": 2.7295977679609393,
+      "grad_norm": 1.9994468174505557,
+      "learning_rate": 2.4647829132457446e-07,
+      "loss": 0.2259,
+      "step": 1955
+    },
+    {
+      "epoch": 2.7309927923738666,
+      "grad_norm": 1.843524154811651,
+      "learning_rate": 2.439646427184428e-07,
+      "loss": 0.2207,
+      "step": 1956
+    },
+    {
+      "epoch": 2.732387816786794,
+      "grad_norm": 2.1830527721245594,
+      "learning_rate": 2.4146355675525145e-07,
+      "loss": 0.2292,
+      "step": 1957
+    },
+    {
+      "epoch": 2.733782841199721,
+      "grad_norm": 1.9550636161481592,
+      "learning_rate": 2.389750400413965e-07,
+      "loss": 0.2481,
+      "step": 1958
+    },
+    {
+      "epoch": 2.7351778656126484,
+      "grad_norm": 2.009516811101446,
+      "learning_rate": 2.364990991500743e-07,
+      "loss": 0.2264,
+      "step": 1959
+    },
+    {
+      "epoch": 2.7365728900255757,
+      "grad_norm": 2.185037327848579,
+      "learning_rate": 2.340357406212601e-07,
+      "loss": 0.2055,
+      "step": 1960
+    },
+    {
+      "epoch": 2.7379679144385025,
+      "grad_norm": 1.8805293460013788,
+      "learning_rate": 2.315849709616963e-07,
+      "loss": 0.1983,
+      "step": 1961
+    },
+    {
+      "epoch": 2.73936293885143,
+      "grad_norm": 1.8960709072653619,
+      "learning_rate": 2.2914679664487237e-07,
+      "loss": 0.1951,
+      "step": 1962
+    },
+    {
+      "epoch": 2.740757963264357,
+      "grad_norm": 2.0602378849111247,
+      "learning_rate": 2.2672122411100727e-07,
+      "loss": 0.2368,
+      "step": 1963
+    },
+    {
+      "epoch": 2.7421529876772843,
+      "grad_norm": 1.960524247241766,
+      "learning_rate": 2.2430825976703485e-07,
+      "loss": 0.221,
+      "step": 1964
+    },
+    {
+      "epoch": 2.7435480120902116,
+      "grad_norm": 2.1133291760383375,
+      "learning_rate": 2.2190790998658561e-07,
+      "loss": 0.2667,
+      "step": 1965
+    },
+    {
+      "epoch": 2.744943036503139,
+      "grad_norm": 2.0557807010861917,
+      "learning_rate": 2.1952018110996843e-07,
+      "loss": 0.2194,
+      "step": 1966
+    },
+    {
+      "epoch": 2.746338060916066,
+      "grad_norm": 2.10649236905764,
+      "learning_rate": 2.1714507944415708e-07,
+      "loss": 0.2541,
+      "step": 1967
+    },
+    {
+      "epoch": 2.7477330853289934,
+      "grad_norm": 1.9442564367344377,
+      "learning_rate": 2.1478261126276989e-07,
+      "loss": 0.2181,
+      "step": 1968
+    },
+    {
+      "epoch": 2.7491281097419202,
+      "grad_norm": 2.0582235498172055,
+      "learning_rate": 2.1243278280605517e-07,
+      "loss": 0.2414,
+      "step": 1969
+    },
+    {
+      "epoch": 2.7505231341548475,
+      "grad_norm": 2.0367916482967803,
+      "learning_rate": 2.1009560028087627e-07,
+      "loss": 0.2198,
+      "step": 1970
+    },
+    {
+      "epoch": 2.7519181585677748,
+      "grad_norm": 2.0136363532537978,
+      "learning_rate": 2.0777106986069162e-07,
+      "loss": 0.2335,
+      "step": 1971
+    },
+    {
+      "epoch": 2.753313182980702,
+      "grad_norm": 1.9593075020307442,
+      "learning_rate": 2.0545919768554078e-07,
+      "loss": 0.2151,
+      "step": 1972
+    },
+    {
+      "epoch": 2.7547082073936293,
+      "grad_norm": 2.068783498395829,
+      "learning_rate": 2.0315998986202902e-07,
+      "loss": 0.2484,
+      "step": 1973
+    },
+    {
+      "epoch": 2.7561032318065566,
+      "grad_norm": 2.0417988017428095,
+      "learning_rate": 2.0087345246330714e-07,
+      "loss": 0.229,
+      "step": 1974
+    },
+    {
+      "epoch": 2.757498256219484,
+      "grad_norm": 2.107556273533938,
+      "learning_rate": 1.985995915290595e-07,
+      "loss": 0.2346,
+      "step": 1975
+    },
+    {
+      "epoch": 2.758893280632411,
+      "grad_norm": 2.0145221028032694,
+      "learning_rate": 1.9633841306548717e-07,
+      "loss": 0.213,
+      "step": 1976
+    },
+    {
+      "epoch": 2.7602883050453384,
+      "grad_norm": 1.995502113277562,
+      "learning_rate": 1.9408992304529252e-07,
+      "loss": 0.2505,
+      "step": 1977
+    },
+    {
+      "epoch": 2.7616833294582657,
+      "grad_norm": 1.974801471687715,
+      "learning_rate": 1.9185412740765962e-07,
+      "loss": 0.1984,
+      "step": 1978
+    },
+    {
+      "epoch": 2.763078353871193,
+      "grad_norm": 1.9678999879577588,
+      "learning_rate": 1.8963103205824397e-07,
+      "loss": 0.2222,
+      "step": 1979
+    },
+    {
+      "epoch": 2.76447337828412,
+      "grad_norm": 2.044892899381473,
+      "learning_rate": 1.8742064286915329e-07,
+      "loss": 0.233,
+      "step": 1980
+    },
+    {
+      "epoch": 2.765868402697047,
+      "grad_norm": 1.930009643778441,
+      "learning_rate": 1.8522296567893282e-07,
+      "loss": 0.25,
+      "step": 1981
+    },
+    {
+      "epoch": 2.7672634271099743,
+      "grad_norm": 2.189950482821562,
+      "learning_rate": 1.830380062925513e-07,
+      "loss": 0.223,
+      "step": 1982
+    },
+    {
+      "epoch": 2.7686584515229016,
+      "grad_norm": 2.1618985560864883,
+      "learning_rate": 1.8086577048138432e-07,
+      "loss": 0.2372,
+      "step": 1983
+    },
+    {
+      "epoch": 2.770053475935829,
+      "grad_norm": 1.8935735279008907,
+      "learning_rate": 1.787062639831988e-07,
+      "loss": 0.2201,
+      "step": 1984
+    },
+    {
+      "epoch": 2.771448500348756,
+      "grad_norm": 2.1258001318663116,
+      "learning_rate": 1.7655949250213743e-07,
+      "loss": 0.2152,
+      "step": 1985
+    },
+    {
+      "epoch": 2.7728435247616834,
+      "grad_norm": 1.9586861209264248,
+      "learning_rate": 1.7442546170870654e-07,
+      "loss": 0.2247,
+      "step": 1986
+    },
+    {
+      "epoch": 2.7742385491746107,
+      "grad_norm": 2.042832703188186,
+      "learning_rate": 1.7230417723975766e-07,
+      "loss": 0.2408,
+      "step": 1987
+    },
+    {
+      "epoch": 2.775633573587538,
+      "grad_norm": 1.9728731477852715,
+      "learning_rate": 1.7019564469847372e-07,
+      "loss": 0.2128,
+      "step": 1988
+    },
+    {
+      "epoch": 2.7770285980004648,
+      "grad_norm": 1.915287465847493,
+      "learning_rate": 1.6809986965435675e-07,
+      "loss": 0.205,
+      "step": 1989
+    },
+    {
+      "epoch": 2.778423622413392,
+      "grad_norm": 2.0763097266218726,
+      "learning_rate": 1.660168576432092e-07,
+      "loss": 0.2432,
+      "step": 1990
+    },
+    {
+      "epoch": 2.7798186468263193,
+      "grad_norm": 2.0818890704572643,
+      "learning_rate": 1.6394661416711977e-07,
+      "loss": 0.2215,
+      "step": 1991
+    },
+    {
+      "epoch": 2.7812136712392466,
+      "grad_norm": 1.9996995437677372,
+      "learning_rate": 1.6188914469445372e-07,
+      "loss": 0.2231,
+      "step": 1992
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 2.0338022002237546,
+      "learning_rate": 1.5984445465983156e-07,
+      "loss": 0.2332,
+      "step": 1993
+    },
+    {
+      "epoch": 2.784003720065101,
+      "grad_norm": 2.1000818079003896,
+      "learning_rate": 1.5781254946412029e-07,
+      "loss": 0.2169,
+      "step": 1994
+    },
+    {
+      "epoch": 2.7853987444780284,
+      "grad_norm": 1.9706118972006013,
+      "learning_rate": 1.5579343447441663e-07,
+      "loss": 0.2156,
+      "step": 1995
+    },
+    {
+      "epoch": 2.7867937688909556,
+      "grad_norm": 1.825179731518908,
+      "learning_rate": 1.5378711502403164e-07,
+      "loss": 0.1899,
+      "step": 1996
+    },
+    {
+      "epoch": 2.788188793303883,
+      "grad_norm": 1.9856171305408068,
+      "learning_rate": 1.5179359641247948e-07,
+      "loss": 0.2378,
+      "step": 1997
+    },
+    {
+      "epoch": 2.78958381771681,
+      "grad_norm": 2.0723262091623758,
+      "learning_rate": 1.4981288390546188e-07,
+      "loss": 0.2406,
+      "step": 1998
+    },
+    {
+      "epoch": 2.7909788421297375,
+      "grad_norm": 2.0926030580473984,
+      "learning_rate": 1.4784498273485436e-07,
+      "loss": 0.2271,
+      "step": 1999
+    },
+    {
+      "epoch": 2.7923738665426647,
+      "grad_norm": 2.0940586610006826,
+      "learning_rate": 1.458898980986917e-07,
+      "loss": 0.2356,
+      "step": 2000
+    },
+    {
+      "epoch": 2.793768890955592,
+      "grad_norm": 2.136395647396599,
+      "learning_rate": 1.4394763516115573e-07,
+      "loss": 0.2287,
+      "step": 2001
+    },
+    {
+      "epoch": 2.795163915368519,
+      "grad_norm": 2.0999404737694722,
+      "learning_rate": 1.4201819905256043e-07,
+      "loss": 0.2258,
+      "step": 2002
+    },
+    {
+      "epoch": 2.796558939781446,
+      "grad_norm": 2.0712201828778847,
+      "learning_rate": 1.4010159486933906e-07,
+      "loss": 0.2514,
+      "step": 2003
+    },
+    {
+      "epoch": 2.7979539641943734,
+      "grad_norm": 2.053549791837711,
+      "learning_rate": 1.3819782767403034e-07,
+      "loss": 0.2193,
+      "step": 2004
+    },
+    {
+      "epoch": 2.7993489886073006,
+      "grad_norm": 1.8855245314887703,
+      "learning_rate": 1.363069024952668e-07,
+      "loss": 0.2356,
+      "step": 2005
+    },
+    {
+      "epoch": 2.800744013020228,
+      "grad_norm": 1.923053478031501,
+      "learning_rate": 1.344288243277575e-07,
+      "loss": 0.245,
+      "step": 2006
+    },
+    {
+      "epoch": 2.802139037433155,
+      "grad_norm": 2.013116779776721,
+      "learning_rate": 1.3256359813227758e-07,
+      "loss": 0.2242,
+      "step": 2007
+    },
+    {
+      "epoch": 2.8035340618460824,
+      "grad_norm": 1.8577386165706877,
+      "learning_rate": 1.3071122883565657e-07,
+      "loss": 0.246,
+      "step": 2008
+    },
+    {
+      "epoch": 2.8049290862590093,
+      "grad_norm": 2.073816182656257,
+      "learning_rate": 1.288717213307622e-07,
+      "loss": 0.2448,
+      "step": 2009
+    },
+    {
+      "epoch": 2.8063241106719365,
+      "grad_norm": 2.1025633652970894,
+      "learning_rate": 1.2704508047649e-07,
+      "loss": 0.2524,
+      "step": 2010
+    },
+    {
+      "epoch": 2.807719135084864,
+      "grad_norm": 1.9930940045825634,
+      "learning_rate": 1.2523131109774822e-07,
+      "loss": 0.2007,
+      "step": 2011
+    },
+    {
+      "epoch": 2.809114159497791,
+      "grad_norm": 1.7898062762710325,
+      "learning_rate": 1.234304179854473e-07,
+      "loss": 0.2162,
+      "step": 2012
+    },
+    {
+      "epoch": 2.8105091839107184,
+      "grad_norm": 2.0747699633734653,
+      "learning_rate": 1.2164240589648436e-07,
+      "loss": 0.2266,
+      "step": 2013
+    },
+    {
+      "epoch": 2.8119042083236456,
+      "grad_norm": 1.9999733625821126,
+      "learning_rate": 1.1986727955373588e-07,
+      "loss": 0.2281,
+      "step": 2014
+    },
+    {
+      "epoch": 2.813299232736573,
+      "grad_norm": 2.0375106273671855,
+      "learning_rate": 1.1810504364603737e-07,
+      "loss": 0.24,
+      "step": 2015
+    },
+    {
+      "epoch": 2.8146942571495,
+      "grad_norm": 2.0844670518449693,
+      "learning_rate": 1.163557028281792e-07,
+      "loss": 0.2058,
+      "step": 2016
+    },
+    {
+      "epoch": 2.8160892815624274,
+      "grad_norm": 1.9962373585540047,
+      "learning_rate": 1.146192617208891e-07,
+      "loss": 0.1974,
+      "step": 2017
+    },
+    {
+      "epoch": 2.8174843059753547,
+      "grad_norm": 1.8386401423765901,
+      "learning_rate": 1.128957249108209e-07,
+      "loss": 0.2076,
+      "step": 2018
+    },
+    {
+      "epoch": 2.818879330388282,
+      "grad_norm": 2.0468415910931874,
+      "learning_rate": 1.1118509695054236e-07,
+      "loss": 0.2085,
+      "step": 2019
+    },
+    {
+      "epoch": 2.8202743548012092,
+      "grad_norm": 1.9826632032740978,
+      "learning_rate": 1.094873823585263e-07,
+      "loss": 0.2078,
+      "step": 2020
+    },
+    {
+      "epoch": 2.8216693792141365,
+      "grad_norm": 1.8783716162476043,
+      "learning_rate": 1.0780258561913281e-07,
+      "loss": 0.2367,
+      "step": 2021
+    },
+    {
+      "epoch": 2.8230644036270633,
+      "grad_norm": 2.217419255609164,
+      "learning_rate": 1.0613071118260321e-07,
+      "loss": 0.2332,
+      "step": 2022
+    },
+    {
+      "epoch": 2.8244594280399906,
+      "grad_norm": 2.200154723898179,
+      "learning_rate": 1.0447176346504439e-07,
+      "loss": 0.2497,
+      "step": 2023
+    },
+    {
+      "epoch": 2.825854452452918,
+      "grad_norm": 2.0172069553573193,
+      "learning_rate": 1.0282574684841784e-07,
+      "loss": 0.2507,
+      "step": 2024
+    },
+    {
+      "epoch": 2.827249476865845,
+      "grad_norm": 2.0893341853420506,
+      "learning_rate": 1.011926656805301e-07,
+      "loss": 0.213,
+      "step": 2025
+    },
+    {
+      "epoch": 2.8286445012787724,
+      "grad_norm": 1.814949208944079,
+      "learning_rate": 9.957252427501951e-08,
+      "loss": 0.2115,
+      "step": 2026
+    },
+    {
+      "epoch": 2.8300395256916997,
+      "grad_norm": 1.9078100366735615,
+      "learning_rate": 9.796532691134453e-08,
+      "loss": 0.2388,
+      "step": 2027
+    },
+    {
+      "epoch": 2.831434550104627,
+      "grad_norm": 1.9158681074801742,
+      "learning_rate": 9.637107783477484e-08,
+      "loss": 0.2171,
+      "step": 2028
+    },
+    {
+      "epoch": 2.832829574517554,
+      "grad_norm": 2.034413782195638,
+      "learning_rate": 9.478978125637583e-08,
+      "loss": 0.2061,
+      "step": 2029
+    },
+    {
+      "epoch": 2.834224598930481,
+      "grad_norm": 2.0640999913344595,
+      "learning_rate": 9.322144135300137e-08,
+      "loss": 0.2329,
+      "step": 2030
+    },
+    {
+      "epoch": 2.8356196233434083,
+      "grad_norm": 2.0272822842665774,
+      "learning_rate": 9.166606226728103e-08,
+      "loss": 0.2074,
+      "step": 2031
+    },
+    {
+      "epoch": 2.8370146477563356,
+      "grad_norm": 1.7729057129292922,
+      "learning_rate": 9.012364810761121e-08,
+      "loss": 0.2384,
+      "step": 2032
+    },
+    {
+      "epoch": 2.838409672169263,
+      "grad_norm": 2.336513429393473,
+      "learning_rate": 8.859420294814014e-08,
+      "loss": 0.2352,
+      "step": 2033
+    },
+    {
+      "epoch": 2.83980469658219,
+      "grad_norm": 1.9596699673808875,
+      "learning_rate": 8.70777308287618e-08,
+      "loss": 0.253,
+      "step": 2034
+    },
+    {
+      "epoch": 2.8411997209951174,
+      "grad_norm": 2.216580635688953,
+      "learning_rate": 8.557423575510037e-08,
+      "loss": 0.2535,
+      "step": 2035
+    },
+    {
+      "epoch": 2.8425947454080447,
+      "grad_norm": 2.169592549290016,
+      "learning_rate": 8.408372169850521e-08,
+      "loss": 0.2498,
+      "step": 2036
+    },
+    {
+      "epoch": 2.843989769820972,
+      "grad_norm": 2.1695073881811306,
+      "learning_rate": 8.26061925960353e-08,
+      "loss": 0.2648,
+      "step": 2037
+    },
+    {
+      "epoch": 2.8453847942338992,
+      "grad_norm": 2.197398628568355,
+      "learning_rate": 8.114165235045268e-08,
+      "loss": 0.2425,
+      "step": 2038
+    },
+    {
+      "epoch": 2.8467798186468265,
+      "grad_norm": 2.0971862433821427,
+      "learning_rate": 7.969010483020845e-08,
+      "loss": 0.2298,
+      "step": 2039
+    },
+    {
+      "epoch": 2.8481748430597538,
+      "grad_norm": 1.8349596523090903,
+      "learning_rate": 7.825155386943784e-08,
+      "loss": 0.2355,
+      "step": 2040
+    },
+    {
+      "epoch": 2.849569867472681,
+      "grad_norm": 2.0915909568177797,
+      "learning_rate": 7.682600326794353e-08,
+      "loss": 0.2419,
+      "step": 2041
+    },
+    {
+      "epoch": 2.850964891885608,
+      "grad_norm": 1.9667085866119112,
+      "learning_rate": 7.541345679118961e-08,
+      "loss": 0.2325,
+      "step": 2042
+    },
+    {
+      "epoch": 2.852359916298535,
+      "grad_norm": 2.144176674881882,
+      "learning_rate": 7.401391817029257e-08,
+      "loss": 0.2259,
+      "step": 2043
+    },
+    {
+      "epoch": 2.8537549407114624,
+      "grad_norm": 2.0336094691866,
+      "learning_rate": 7.262739110200923e-08,
+      "loss": 0.2272,
+      "step": 2044
+    },
+    {
+      "epoch": 2.8551499651243897,
+      "grad_norm": 2.132962040695552,
+      "learning_rate": 7.125387924872552e-08,
+      "loss": 0.2544,
+      "step": 2045
+    },
+    {
+      "epoch": 2.856544989537317,
+      "grad_norm": 1.955162224372117,
+      "learning_rate": 6.98933862384521e-08,
+      "loss": 0.2132,
+      "step": 2046
+    },
+    {
+      "epoch": 2.857940013950244,
+      "grad_norm": 2.095095213879681,
+      "learning_rate": 6.854591566480884e-08,
+      "loss": 0.1864,
+      "step": 2047
+    },
+    {
+      "epoch": 2.8593350383631715,
+      "grad_norm": 1.9264801577988864,
+      "learning_rate": 6.721147108701864e-08,
+      "loss": 0.2338,
+      "step": 2048
+    },
+    {
+      "epoch": 2.8607300627760983,
+      "grad_norm": 2.030106096946145,
+      "learning_rate": 6.589005602989862e-08,
+      "loss": 0.2299,
+      "step": 2049
+    },
+    {
+      "epoch": 2.8621250871890256,
+      "grad_norm": 2.312046798351557,
+      "learning_rate": 6.458167398384896e-08,
+      "loss": 0.2375,
+      "step": 2050
+    },
+    {
+      "epoch": 2.863520111601953,
+      "grad_norm": 2.0184696688819206,
+      "learning_rate": 6.328632840484294e-08,
+      "loss": 0.1987,
+      "step": 2051
+    },
+    {
+      "epoch": 2.86491513601488,
+      "grad_norm": 1.8251619065621412,
+      "learning_rate": 6.200402271442085e-08,
+      "loss": 0.2252,
+      "step": 2052
+    },
+    {
+      "epoch": 2.8663101604278074,
+      "grad_norm": 2.0497067870912575,
+      "learning_rate": 6.073476029967884e-08,
+      "loss": 0.2308,
+      "step": 2053
+    },
+    {
+      "epoch": 2.8677051848407347,
+      "grad_norm": 2.0486210575713097,
+      "learning_rate": 5.947854451326007e-08,
+      "loss": 0.2437,
+      "step": 2054
+    },
+    {
+      "epoch": 2.869100209253662,
+      "grad_norm": 1.9864897653983566,
+      "learning_rate": 5.823537867334694e-08,
+      "loss": 0.2112,
+      "step": 2055
+    },
+    {
+      "epoch": 2.870495233666589,
+      "grad_norm": 1.9230197033304526,
+      "learning_rate": 5.7005266063650534e-08,
+      "loss": 0.2251,
+      "step": 2056
+    },
+    {
+      "epoch": 2.8718902580795165,
+      "grad_norm": 2.151992679895416,
+      "learning_rate": 5.5788209933403944e-08,
+      "loss": 0.2792,
+      "step": 2057
+    },
+    {
+      "epoch": 2.8732852824924437,
+      "grad_norm": 1.920997012465794,
+      "learning_rate": 5.4584213497351766e-08,
+      "loss": 0.2251,
+      "step": 2058
+    },
+    {
+      "epoch": 2.874680306905371,
+      "grad_norm": 1.9992012597394688,
+      "learning_rate": 5.339327993574339e-08,
+      "loss": 0.2103,
+      "step": 2059
+    },
+    {
+      "epoch": 2.8760753313182983,
+      "grad_norm": 1.9393989081162242,
+      "learning_rate": 5.221541239432415e-08,
+      "loss": 0.1968,
+      "step": 2060
+    },
+    {
+      "epoch": 2.8774703557312256,
+      "grad_norm": 2.063840366940395,
+      "learning_rate": 5.1050613984324756e-08,
+      "loss": 0.2034,
+      "step": 2061
+    },
+    {
+      "epoch": 2.8788653801441524,
+      "grad_norm": 2.2590391916977346,
+      "learning_rate": 4.989888778245744e-08,
+      "loss": 0.2559,
+      "step": 2062
+    },
+    {
+      "epoch": 2.8802604045570797,
+      "grad_norm": 1.9797057677091427,
+      "learning_rate": 4.8760236830903697e-08,
+      "loss": 0.2391,
+      "step": 2063
+    },
+    {
+      "epoch": 2.881655428970007,
+      "grad_norm": 2.0780206968147117,
+      "learning_rate": 4.763466413730822e-08,
+      "loss": 0.2426,
+      "step": 2064
+    },
+    {
+      "epoch": 2.883050453382934,
+      "grad_norm": 2.036645937963387,
+      "learning_rate": 4.65221726747711e-08,
+      "loss": 0.237,
+      "step": 2065
+    },
+    {
+      "epoch": 2.8844454777958615,
+      "grad_norm": 2.1721874584487995,
+      "learning_rate": 4.542276538183954e-08,
+      "loss": 0.2231,
+      "step": 2066
+    },
+    {
+      "epoch": 2.8858405022087887,
+      "grad_norm": 2.042288111806101,
+      "learning_rate": 4.433644516249891e-08,
+      "loss": 0.2465,
+      "step": 2067
+    },
+    {
+      "epoch": 2.887235526621716,
+      "grad_norm": 2.142452821109536,
+      "learning_rate": 4.326321488616836e-08,
+      "loss": 0.251,
+      "step": 2068
+    },
+    {
+      "epoch": 2.888630551034643,
+      "grad_norm": 2.0700122700221915,
+      "learning_rate": 4.220307738768859e-08,
+      "loss": 0.2239,
+      "step": 2069
+    },
+    {
+      "epoch": 2.89002557544757,
+      "grad_norm": 1.9911765497561993,
+      "learning_rate": 4.11560354673185e-08,
+      "loss": 0.2451,
+      "step": 2070
+    },
+    {
+      "epoch": 2.8914205998604974,
+      "grad_norm": 1.8703876336869296,
+      "learning_rate": 4.0122091890726354e-08,
+      "loss": 0.2296,
+      "step": 2071
+    },
+    {
+      "epoch": 2.8928156242734246,
+      "grad_norm": 1.9321327299338205,
+      "learning_rate": 3.9101249388981965e-08,
+      "loss": 0.2019,
+      "step": 2072
+    },
+    {
+      "epoch": 2.894210648686352,
+      "grad_norm": 2.012664815357327,
+      "learning_rate": 3.809351065854894e-08,
+      "loss": 0.2576,
+      "step": 2073
+    },
+    {
+      "epoch": 2.895605673099279,
+      "grad_norm": 2.087285982253065,
+      "learning_rate": 3.709887836128023e-08,
+      "loss": 0.2611,
+      "step": 2074
+    },
+    {
+      "epoch": 2.8970006975122065,
+      "grad_norm": 2.0487586659165093,
+      "learning_rate": 3.611735512440706e-08,
+      "loss": 0.251,
+      "step": 2075
+    },
+    {
+      "epoch": 2.8983957219251337,
+      "grad_norm": 2.114982811705642,
+      "learning_rate": 3.5148943540536105e-08,
+      "loss": 0.2257,
+      "step": 2076
+    },
+    {
+      "epoch": 2.899790746338061,
+      "grad_norm": 1.9590167118799493,
+      "learning_rate": 3.4193646167640646e-08,
+      "loss": 0.253,
+      "step": 2077
+    },
+    {
+      "epoch": 2.9011857707509883,
+      "grad_norm": 2.1822450790273953,
+      "learning_rate": 3.325146552905223e-08,
+      "loss": 0.2262,
+      "step": 2078
+    },
+    {
+      "epoch": 2.9025807951639155,
+      "grad_norm": 1.9201480800951032,
+      "learning_rate": 3.2322404113457886e-08,
+      "loss": 0.2547,
+      "step": 2079
+    },
+    {
+      "epoch": 2.903975819576843,
+      "grad_norm": 2.0972666441542107,
+      "learning_rate": 3.1406464374890144e-08,
+      "loss": 0.2294,
+      "step": 2080
+    },
+    {
+      "epoch": 2.90537084398977,
+      "grad_norm": 1.991224686802119,
+      "learning_rate": 3.0503648732722046e-08,
+      "loss": 0.25,
+      "step": 2081
+    },
+    {
+      "epoch": 2.906765868402697,
+      "grad_norm": 2.0853791989901547,
+      "learning_rate": 2.9613959571660468e-08,
+      "loss": 0.2189,
+      "step": 2082
+    },
+    {
+      "epoch": 2.908160892815624,
+      "grad_norm": 2.0370360728725916,
+      "learning_rate": 2.8737399241740016e-08,
+      "loss": 0.2187,
+      "step": 2083
+    },
+    {
+      "epoch": 2.9095559172285514,
+      "grad_norm": 2.014096101500848,
+      "learning_rate": 2.7873970058316934e-08,
+      "loss": 0.224,
+      "step": 2084
+    },
+    {
+      "epoch": 2.9109509416414787,
+      "grad_norm": 1.986275091522161,
+      "learning_rate": 2.7023674302061875e-08,
+      "loss": 0.225,
+      "step": 2085
+    },
+    {
+      "epoch": 2.912345966054406,
+      "grad_norm": 1.9278901568730784,
+      "learning_rate": 2.6186514218954905e-08,
+      "loss": 0.2088,
+      "step": 2086
+    },
+    {
+      "epoch": 2.9137409904673333,
+      "grad_norm": 2.0213138891500924,
+      "learning_rate": 2.5362492020280517e-08,
+      "loss": 0.216,
+      "step": 2087
+    },
+    {
+      "epoch": 2.9151360148802605,
+      "grad_norm": 1.9502452930485683,
+      "learning_rate": 2.4551609882619288e-08,
+      "loss": 0.2513,
+      "step": 2088
+    },
+    {
+      "epoch": 2.9165310392931874,
+      "grad_norm": 2.2370632409827684,
+      "learning_rate": 2.3753869947843457e-08,
+      "loss": 0.2172,
+      "step": 2089
+    },
+    {
+      "epoch": 2.9179260637061146,
+      "grad_norm": 2.015550383922713,
+      "learning_rate": 2.296927432311358e-08,
+      "loss": 0.2448,
+      "step": 2090
+    },
+    {
+      "epoch": 2.919321088119042,
+      "grad_norm": 2.03055889711366,
+      "learning_rate": 2.2197825080867432e-08,
+      "loss": 0.2196,
+      "step": 2091
+    },
+    {
+      "epoch": 2.920716112531969,
+      "grad_norm": 2.1961513763780176,
+      "learning_rate": 2.1439524258819456e-08,
+      "loss": 0.2498,
+      "step": 2092
+    },
+    {
+      "epoch": 2.9221111369448964,
+      "grad_norm": 1.9256101469230513,
+      "learning_rate": 2.0694373859954653e-08,
+      "loss": 0.22,
+      "step": 2093
+    },
+    {
+      "epoch": 2.9235061613578237,
+      "grad_norm": 1.9815134324395363,
+      "learning_rate": 1.99623758525197e-08,
+      "loss": 0.237,
+      "step": 2094
+    },
+    {
+      "epoch": 2.924901185770751,
+      "grad_norm": 1.907106053723731,
+      "learning_rate": 1.9243532170023504e-08,
+      "loss": 0.2211,
+      "step": 2095
+    },
+    {
+      "epoch": 2.9262962101836782,
+      "grad_norm": 1.9455456724863225,
+      "learning_rate": 1.8537844711227215e-08,
+      "loss": 0.2026,
+      "step": 2096
+    },
+    {
+      "epoch": 2.9276912345966055,
+      "grad_norm": 1.9285325866093366,
+      "learning_rate": 1.7845315340140334e-08,
+      "loss": 0.2107,
+      "step": 2097
+    },
+    {
+      "epoch": 2.929086259009533,
+      "grad_norm": 1.7732551833753323,
+      "learning_rate": 1.7165945886018498e-08,
+      "loss": 0.1909,
+      "step": 2098
+    },
+    {
+      "epoch": 2.93048128342246,
+      "grad_norm": 2.0531915086165236,
+      "learning_rate": 1.6499738143354594e-08,
+      "loss": 0.2141,
+      "step": 2099
+    },
+    {
+      "epoch": 2.9318763078353873,
+      "grad_norm": 2.1691009679826094,
+      "learning_rate": 1.584669387187765e-08,
+      "loss": 0.2288,
+      "step": 2100
+    },
+    {
+      "epoch": 2.9332713322483146,
+      "grad_norm": 1.879402148888189,
+      "learning_rate": 1.520681479654562e-08,
+      "loss": 0.2178,
+      "step": 2101
+    },
+    {
+      "epoch": 2.9346663566612414,
+      "grad_norm": 1.953785706612728,
+      "learning_rate": 1.4580102607541502e-08,
+      "loss": 0.2275,
+      "step": 2102
+    },
+    {
+      "epoch": 2.9360613810741687,
+      "grad_norm": 1.9696383581667096,
+      "learning_rate": 1.3966558960269994e-08,
+      "loss": 0.2084,
+      "step": 2103
+    },
+    {
+      "epoch": 2.937456405487096,
+      "grad_norm": 1.6864007194487332,
+      "learning_rate": 1.3366185475351957e-08,
+      "loss": 0.2093,
+      "step": 2104
+    },
+    {
+      "epoch": 2.9388514299000232,
+      "grad_norm": 1.7996177568725729,
+      "learning_rate": 1.2778983738620521e-08,
+      "loss": 0.2305,
+      "step": 2105
+    },
+    {
+      "epoch": 2.9402464543129505,
+      "grad_norm": 2.088014390966735,
+      "learning_rate": 1.2204955301116095e-08,
+      "loss": 0.2105,
+      "step": 2106
+    },
+    {
+      "epoch": 2.941641478725878,
+      "grad_norm": 1.884717455818504,
+      "learning_rate": 1.164410167908414e-08,
+      "loss": 0.2199,
+      "step": 2107
+    },
+    {
+      "epoch": 2.943036503138805,
+      "grad_norm": 2.105817440939439,
+      "learning_rate": 1.109642435396907e-08,
+      "loss": 0.2152,
+      "step": 2108
+    },
+    {
+      "epoch": 2.9444315275517323,
+      "grad_norm": 1.9437942158112311,
+      "learning_rate": 1.0561924772412024e-08,
+      "loss": 0.2228,
+      "step": 2109
+    },
+    {
+      "epoch": 2.945826551964659,
+      "grad_norm": 2.376506125566622,
+      "learning_rate": 1.0040604346245319e-08,
+      "loss": 0.27,
+      "step": 2110
+    },
+    {
+      "epoch": 2.9472215763775864,
+      "grad_norm": 1.7465087784159485,
+      "learning_rate": 9.532464452491341e-09,
+      "loss": 0.2221,
+      "step": 2111
+    },
+    {
+      "epoch": 2.9486166007905137,
+      "grad_norm": 2.105100907970145,
+      "learning_rate": 9.037506433355325e-09,
+      "loss": 0.228,
+      "step": 2112
+    },
+    {
+      "epoch": 2.950011625203441,
+      "grad_norm": 2.057081162802249,
+      "learning_rate": 8.555731596224803e-09,
+      "loss": 0.2096,
+      "step": 2113
+    },
+    {
+      "epoch": 2.9514066496163682,
+      "grad_norm": 1.9893991376538172,
+      "learning_rate": 8.087141213665717e-09,
+      "loss": 0.2394,
+      "step": 2114
+    },
+    {
+      "epoch": 2.9528016740292955,
+      "grad_norm": 1.953608082819364,
+      "learning_rate": 7.631736523416867e-09,
+      "loss": 0.2302,
+      "step": 2115
+    },
+    {
+      "epoch": 2.9541966984422228,
+      "grad_norm": 1.9960133893896974,
+      "learning_rate": 7.1895187283899104e-09,
+      "loss": 0.1825,
+      "step": 2116
+    },
+    {
+      "epoch": 2.95559172285515,
+      "grad_norm": 2.0829277277844622,
+      "learning_rate": 6.760488996662706e-09,
+      "loss": 0.2352,
+      "step": 2117
+    },
+    {
+      "epoch": 2.9569867472680773,
+      "grad_norm": 1.9399119855696576,
+      "learning_rate": 6.3446484614798635e-09,
+      "loss": 0.1935,
+      "step": 2118
+    },
+    {
+      "epoch": 2.9583817716810046,
+      "grad_norm": 1.9161326288025806,
+      "learning_rate": 5.941998221247192e-09,
+      "loss": 0.236,
+      "step": 2119
+    },
+    {
+      "epoch": 2.959776796093932,
+      "grad_norm": 2.089159927313854,
+      "learning_rate": 5.552539339528373e-09,
+      "loss": 0.2755,
+      "step": 2120
+    },
+    {
+      "epoch": 2.961171820506859,
+      "grad_norm": 1.8493546378825017,
+      "learning_rate": 5.176272845045516e-09,
+      "loss": 0.2264,
+      "step": 2121
+    },
+    {
+      "epoch": 2.962566844919786,
+      "grad_norm": 2.079176415927749,
+      "learning_rate": 4.813199731671381e-09,
+      "loss": 0.2042,
+      "step": 2122
+    },
+    {
+      "epoch": 2.963961869332713,
+      "grad_norm": 1.885503099661438,
+      "learning_rate": 4.463320958432716e-09,
+      "loss": 0.1929,
+      "step": 2123
+    },
+    {
+      "epoch": 2.9653568937456405,
+      "grad_norm": 1.8483044364688408,
+      "learning_rate": 4.1266374495024795e-09,
+      "loss": 0.2185,
+      "step": 2124
+    },
+    {
+      "epoch": 2.9667519181585678,
+      "grad_norm": 1.9885490703185613,
+      "learning_rate": 3.803150094200403e-09,
+      "loss": 0.2181,
+      "step": 2125
+    },
+    {
+      "epoch": 2.968146942571495,
+      "grad_norm": 1.9818060311208294,
+      "learning_rate": 3.4928597469885416e-09,
+      "loss": 0.2124,
+      "step": 2126
+    },
+    {
+      "epoch": 2.9695419669844223,
+      "grad_norm": 1.970519772764202,
+      "learning_rate": 3.1957672274723907e-09,
+      "loss": 0.2311,
+      "step": 2127
+    },
+    {
+      "epoch": 2.9709369913973496,
+      "grad_norm": 2.3609131575161273,
+      "learning_rate": 2.9118733203942207e-09,
+      "loss": 0.2369,
+      "step": 2128
+    },
+    {
+      "epoch": 2.972332015810277,
+      "grad_norm": 2.091411714396494,
+      "learning_rate": 2.6411787756353e-09,
+      "loss": 0.237,
+      "step": 2129
+    },
+    {
+      "epoch": 2.9737270402232037,
+      "grad_norm": 1.9145913000006876,
+      "learning_rate": 2.3836843082108987e-09,
+      "loss": 0.2468,
+      "step": 2130
+    },
+    {
+      "epoch": 2.975122064636131,
+      "grad_norm": 1.9045923373896427,
+      "learning_rate": 2.1393905982691752e-09,
+      "loss": 0.2164,
+      "step": 2131
+    },
+    {
+      "epoch": 2.976517089049058,
+      "grad_norm": 2.0044107224748284,
+      "learning_rate": 1.9082982910911817e-09,
+      "loss": 0.2264,
+      "step": 2132
+    },
+    {
+      "epoch": 2.9779121134619855,
+      "grad_norm": 1.961477665662354,
+      "learning_rate": 1.6904079970853083e-09,
+      "loss": 0.2187,
+      "step": 2133
+    },
+    {
+      "epoch": 2.9793071378749127,
+      "grad_norm": 1.8835567782670801,
+      "learning_rate": 1.4857202917900604e-09,
+      "loss": 0.2224,
+      "step": 2134
+    },
+    {
+      "epoch": 2.98070216228784,
+      "grad_norm": 2.1519901375116066,
+      "learning_rate": 1.2942357158701734e-09,
+      "loss": 0.2136,
+      "step": 2135
+    },
+    {
+      "epoch": 2.9820971867007673,
+      "grad_norm": 2.0606473023148135,
+      "learning_rate": 1.1159547751143918e-09,
+      "loss": 0.2397,
+      "step": 2136
+    },
+    {
+      "epoch": 2.9834922111136946,
+      "grad_norm": 1.9019884303178762,
+      "learning_rate": 9.508779404360235e-10,
+      "loss": 0.2472,
+      "step": 2137
+    },
+    {
+      "epoch": 2.984887235526622,
+      "grad_norm": 2.083765199600487,
+      "learning_rate": 7.990056478707209e-10,
+      "loss": 0.222,
+      "step": 2138
+    },
+    {
+      "epoch": 2.986282259939549,
+      "grad_norm": 1.9020968587306906,
+      "learning_rate": 6.603382985759244e-10,
+      "loss": 0.225,
+      "step": 2139
+    },
+    {
+      "epoch": 2.9876772843524764,
+      "grad_norm": 2.2043377624661127,
+      "learning_rate": 5.348762588286427e-10,
+      "loss": 0.2621,
+      "step": 2140
+    },
+    {
+      "epoch": 2.9890723087654036,
+      "grad_norm": 2.1382878635243983,
+      "learning_rate": 4.2261986002600783e-10,
+      "loss": 0.251,
+      "step": 2141
+    },
+    {
+      "epoch": 2.9904673331783305,
+      "grad_norm": 2.144632398744335,
+      "learning_rate": 3.235693986830546e-10,
+      "loss": 0.2618,
+      "step": 2142
+    },
+    {
+      "epoch": 2.9918623575912577,
+      "grad_norm": 2.136193399423466,
+      "learning_rate": 2.3772513643327555e-10,
+      "loss": 0.2406,
+      "step": 2143
+    },
+    {
+      "epoch": 2.993257382004185,
+      "grad_norm": 1.9404996403940398,
+      "learning_rate": 1.650873000258457e-10,
+      "loss": 0.2249,
+      "step": 2144
+    },
+    {
+      "epoch": 2.9946524064171123,
+      "grad_norm": 2.0262928151798407,
+      "learning_rate": 1.0565608132728778e-10,
+      "loss": 0.2308,
+      "step": 2145
+    },
+    {
+      "epoch": 2.9960474308300395,
+      "grad_norm": 1.9808655214657436,
+      "learning_rate": 5.943163732036183e-11,
+      "loss": 0.2503,
+      "step": 2146
+    },
+    {
+      "epoch": 2.997442455242967,
+      "grad_norm": 2.1179871403188932,
+      "learning_rate": 2.6414090102400147e-11,
+      "loss": 0.2435,
+      "step": 2147
+    },
+    {
+      "epoch": 2.998837479655894,
+      "grad_norm": 2.124758681637207,
+      "learning_rate": 6.6035268864173e-12,
+      "loss": 0.2391,
+      "step": 2148
+    },
+    {
+      "epoch": 2.998837479655894,
+      "step": 2148,
+      "total_flos": 3.265949006466908e+17,
+      "train_loss": 0.48496823534508215,
+      "train_runtime": 18125.988,
+      "train_samples_per_second": 2.847,
+      "train_steps_per_second": 0.119
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2148,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.265949006466908e+17,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}