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": 9.06177574123373,
+      "learning_rate": 0.0,
+      "loss": 1.2425,
+      "step": 1
+    },
+    {
+      "epoch": 0.0027900488258544524,
+      "grad_norm": 7.74250980936599,
+      "learning_rate": 4.6511627906976744e-08,
+      "loss": 1.1023,
+      "step": 2
+    },
+    {
+      "epoch": 0.0041850732387816785,
+      "grad_norm": 6.388418368134404,
+      "learning_rate": 9.302325581395349e-08,
+      "loss": 0.9672,
+      "step": 3
+    },
+    {
+      "epoch": 0.005580097651708905,
+      "grad_norm": 8.412567905846219,
+      "learning_rate": 1.3953488372093024e-07,
+      "loss": 1.1729,
+      "step": 4
+    },
+    {
+      "epoch": 0.006975122064636131,
+      "grad_norm": 7.573918067233279,
+      "learning_rate": 1.8604651162790698e-07,
+      "loss": 1.0899,
+      "step": 5
+    },
+    {
+      "epoch": 0.008370146477563357,
+      "grad_norm": 8.091442848700519,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 1.1316,
+      "step": 6
+    },
+    {
+      "epoch": 0.009765170890490584,
+      "grad_norm": 8.167879080412323,
+      "learning_rate": 2.790697674418605e-07,
+      "loss": 1.218,
+      "step": 7
+    },
+    {
+      "epoch": 0.01116019530341781,
+      "grad_norm": 8.519414634553403,
+      "learning_rate": 3.2558139534883724e-07,
+      "loss": 1.2305,
+      "step": 8
+    },
+    {
+      "epoch": 0.012555219716345037,
+      "grad_norm": 7.003865916552692,
+      "learning_rate": 3.7209302325581396e-07,
+      "loss": 1.0201,
+      "step": 9
+    },
+    {
+      "epoch": 0.013950244129272262,
+      "grad_norm": 7.644602533034688,
+      "learning_rate": 4.186046511627907e-07,
+      "loss": 1.0511,
+      "step": 10
+    },
+    {
+      "epoch": 0.015345268542199489,
+      "grad_norm": 8.054509366220413,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 1.1006,
+      "step": 11
+    },
+    {
+      "epoch": 0.016740292955126714,
+      "grad_norm": 6.828467947829633,
+      "learning_rate": 5.116279069767442e-07,
+      "loss": 1.0077,
+      "step": 12
+    },
+    {
+      "epoch": 0.01813531736805394,
+      "grad_norm": 7.451234628274601,
+      "learning_rate": 5.58139534883721e-07,
+      "loss": 1.1322,
+      "step": 13
+    },
+    {
+      "epoch": 0.01953034178098117,
+      "grad_norm": 8.34154278843471,
+      "learning_rate": 6.046511627906977e-07,
+      "loss": 1.185,
+      "step": 14
+    },
+    {
+      "epoch": 0.020925366193908394,
+      "grad_norm": 6.204510873867492,
+      "learning_rate": 6.511627906976745e-07,
+      "loss": 0.9731,
+      "step": 15
+    },
+    {
+      "epoch": 0.02232039060683562,
+      "grad_norm": 7.03015750664511,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 1.0501,
+      "step": 16
+    },
+    {
+      "epoch": 0.023715415019762844,
+      "grad_norm": 5.239488986570632,
+      "learning_rate": 7.441860465116279e-07,
+      "loss": 0.8553,
+      "step": 17
+    },
+    {
+      "epoch": 0.025110439432690073,
+      "grad_norm": 7.842221666246864,
+      "learning_rate": 7.906976744186047e-07,
+      "loss": 1.1987,
+      "step": 18
+    },
+    {
+      "epoch": 0.0265054638456173,
+      "grad_norm": 5.010590976687375,
+      "learning_rate": 8.372093023255814e-07,
+      "loss": 0.9201,
+      "step": 19
+    },
+    {
+      "epoch": 0.027900488258544524,
+      "grad_norm": 6.810234105335915,
+      "learning_rate": 8.837209302325582e-07,
+      "loss": 1.044,
+      "step": 20
+    },
+    {
+      "epoch": 0.029295512671471752,
+      "grad_norm": 5.719852480088899,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.9848,
+      "step": 21
+    },
+    {
+      "epoch": 0.030690537084398978,
+      "grad_norm": 5.804104402736001,
+      "learning_rate": 9.767441860465117e-07,
+      "loss": 1.0273,
+      "step": 22
+    },
+    {
+      "epoch": 0.03208556149732621,
+      "grad_norm": 5.698265140721901,
+      "learning_rate": 1.0232558139534884e-06,
+      "loss": 1.0212,
+      "step": 23
+    },
+    {
+      "epoch": 0.03348058591025343,
+      "grad_norm": 3.5836689034500155,
+      "learning_rate": 1.0697674418604653e-06,
+      "loss": 0.8341,
+      "step": 24
+    },
+    {
+      "epoch": 0.03487561032318066,
+      "grad_norm": 4.9843301829941495,
+      "learning_rate": 1.116279069767442e-06,
+      "loss": 0.9715,
+      "step": 25
+    },
+    {
+      "epoch": 0.03627063473610788,
+      "grad_norm": 4.0192024122191174,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.9561,
+      "step": 26
+    },
+    {
+      "epoch": 0.03766565914903511,
+      "grad_norm": 5.0428069070438815,
+      "learning_rate": 1.2093023255813954e-06,
+      "loss": 1.0269,
+      "step": 27
+    },
+    {
+      "epoch": 0.03906068356196234,
+      "grad_norm": 3.9250973474293027,
+      "learning_rate": 1.2558139534883723e-06,
+      "loss": 0.9129,
+      "step": 28
+    },
+    {
+      "epoch": 0.04045570797488956,
+      "grad_norm": 3.5444590391314015,
+      "learning_rate": 1.302325581395349e-06,
+      "loss": 0.889,
+      "step": 29
+    },
+    {
+      "epoch": 0.04185073238781679,
+      "grad_norm": 3.4307236283710103,
+      "learning_rate": 1.3488372093023258e-06,
+      "loss": 0.8861,
+      "step": 30
+    },
+    {
+      "epoch": 0.043245756800744016,
+      "grad_norm": 4.025994100913768,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.8176,
+      "step": 31
+    },
+    {
+      "epoch": 0.04464078121367124,
+      "grad_norm": 4.1408375259554075,
+      "learning_rate": 1.4418604651162794e-06,
+      "loss": 0.8987,
+      "step": 32
+    },
+    {
+      "epoch": 0.04603580562659847,
+      "grad_norm": 3.8731866291196444,
+      "learning_rate": 1.4883720930232558e-06,
+      "loss": 0.9591,
+      "step": 33
+    },
+    {
+      "epoch": 0.04743083003952569,
+      "grad_norm": 3.989912476418152,
+      "learning_rate": 1.534883720930233e-06,
+      "loss": 0.9171,
+      "step": 34
+    },
+    {
+      "epoch": 0.04882585445245292,
+      "grad_norm": 3.7064707308962936,
+      "learning_rate": 1.5813953488372093e-06,
+      "loss": 0.9138,
+      "step": 35
+    },
+    {
+      "epoch": 0.050220878865380146,
+      "grad_norm": 3.6202184488771993,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 0.9461,
+      "step": 36
+    },
+    {
+      "epoch": 0.05161590327830737,
+      "grad_norm": 3.126403206575003,
+      "learning_rate": 1.6744186046511629e-06,
+      "loss": 0.8543,
+      "step": 37
+    },
+    {
+      "epoch": 0.0530109276912346,
+      "grad_norm": 3.02742858939562,
+      "learning_rate": 1.7209302325581397e-06,
+      "loss": 0.7944,
+      "step": 38
+    },
+    {
+      "epoch": 0.054405952104161825,
+      "grad_norm": 3.0983156242512324,
+      "learning_rate": 1.7674418604651164e-06,
+      "loss": 0.9039,
+      "step": 39
+    },
+    {
+      "epoch": 0.05580097651708905,
+      "grad_norm": 3.3601546255070347,
+      "learning_rate": 1.8139534883720933e-06,
+      "loss": 0.8522,
+      "step": 40
+    },
+    {
+      "epoch": 0.057196000930016276,
+      "grad_norm": 3.170617165847687,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.8171,
+      "step": 41
+    },
+    {
+      "epoch": 0.058591025342943505,
+      "grad_norm": 3.163047232937274,
+      "learning_rate": 1.9069767441860468e-06,
+      "loss": 0.8097,
+      "step": 42
+    },
+    {
+      "epoch": 0.05998604975587073,
+      "grad_norm": 3.2539566987966526,
+      "learning_rate": 1.9534883720930235e-06,
+      "loss": 0.7917,
+      "step": 43
+    },
+    {
+      "epoch": 0.061381074168797956,
+      "grad_norm": 3.0851698620768935,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.8263,
+      "step": 44
+    },
+    {
+      "epoch": 0.06277609858172518,
+      "grad_norm": 3.3988761007126183,
+      "learning_rate": 2.0465116279069768e-06,
+      "loss": 0.8794,
+      "step": 45
+    },
+    {
+      "epoch": 0.06417112299465241,
+      "grad_norm": 2.9613711343959785,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.7589,
+      "step": 46
+    },
+    {
+      "epoch": 0.06556614740757963,
+      "grad_norm": 3.093943745703619,
+      "learning_rate": 2.1395348837209305e-06,
+      "loss": 0.8537,
+      "step": 47
+    },
+    {
+      "epoch": 0.06696117182050686,
+      "grad_norm": 2.7983992330490532,
+      "learning_rate": 2.1860465116279074e-06,
+      "loss": 0.8453,
+      "step": 48
+    },
+    {
+      "epoch": 0.06835619623343409,
+      "grad_norm": 2.6204585607252375,
+      "learning_rate": 2.232558139534884e-06,
+      "loss": 0.8777,
+      "step": 49
+    },
+    {
+      "epoch": 0.06975122064636131,
+      "grad_norm": 2.81413032613818,
+      "learning_rate": 2.2790697674418607e-06,
+      "loss": 0.8138,
+      "step": 50
+    },
+    {
+      "epoch": 0.07114624505928854,
+      "grad_norm": 2.7779036842721703,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.8686,
+      "step": 51
+    },
+    {
+      "epoch": 0.07254126947221576,
+      "grad_norm": 2.5802389971258557,
+      "learning_rate": 2.3720930232558144e-06,
+      "loss": 0.8089,
+      "step": 52
+    },
+    {
+      "epoch": 0.07393629388514299,
+      "grad_norm": 2.8171256077152527,
+      "learning_rate": 2.418604651162791e-06,
+      "loss": 0.8212,
+      "step": 53
+    },
+    {
+      "epoch": 0.07533131829807022,
+      "grad_norm": 2.9444479391684997,
+      "learning_rate": 2.4651162790697678e-06,
+      "loss": 0.8565,
+      "step": 54
+    },
+    {
+      "epoch": 0.07672634271099744,
+      "grad_norm": 2.6944161956139596,
+      "learning_rate": 2.5116279069767446e-06,
+      "loss": 0.7982,
+      "step": 55
+    },
+    {
+      "epoch": 0.07812136712392467,
+      "grad_norm": 2.7051824317676005,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.8422,
+      "step": 56
+    },
+    {
+      "epoch": 0.07951639153685189,
+      "grad_norm": 2.6720226291628637,
+      "learning_rate": 2.604651162790698e-06,
+      "loss": 0.8189,
+      "step": 57
+    },
+    {
+      "epoch": 0.08091141594977912,
+      "grad_norm": 2.5400408658615565,
+      "learning_rate": 2.6511627906976744e-06,
+      "loss": 0.7868,
+      "step": 58
+    },
+    {
+      "epoch": 0.08230644036270635,
+      "grad_norm": 2.72174082098465,
+      "learning_rate": 2.6976744186046517e-06,
+      "loss": 0.8908,
+      "step": 59
+    },
+    {
+      "epoch": 0.08370146477563357,
+      "grad_norm": 2.493197204847241,
+      "learning_rate": 2.744186046511628e-06,
+      "loss": 0.7486,
+      "step": 60
+    },
+    {
+      "epoch": 0.0850964891885608,
+      "grad_norm": 2.590372903491486,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.7484,
+      "step": 61
+    },
+    {
+      "epoch": 0.08649151360148803,
+      "grad_norm": 2.5390193429571877,
+      "learning_rate": 2.8372093023255815e-06,
+      "loss": 0.8414,
+      "step": 62
+    },
+    {
+      "epoch": 0.08788653801441525,
+      "grad_norm": 2.455465710423539,
+      "learning_rate": 2.8837209302325587e-06,
+      "loss": 0.7389,
+      "step": 63
+    },
+    {
+      "epoch": 0.08928156242734248,
+      "grad_norm": 2.6820415416133208,
+      "learning_rate": 2.930232558139535e-06,
+      "loss": 0.8335,
+      "step": 64
+    },
+    {
+      "epoch": 0.0906765868402697,
+      "grad_norm": 2.3815607596683543,
+      "learning_rate": 2.9767441860465116e-06,
+      "loss": 0.7524,
+      "step": 65
+    },
+    {
+      "epoch": 0.09207161125319693,
+      "grad_norm": 2.5996024857047537,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.8056,
+      "step": 66
+    },
+    {
+      "epoch": 0.09346663566612416,
+      "grad_norm": 2.5309785387428927,
+      "learning_rate": 3.069767441860466e-06,
+      "loss": 0.8627,
+      "step": 67
+    },
+    {
+      "epoch": 0.09486166007905138,
+      "grad_norm": 2.539797914962022,
+      "learning_rate": 3.1162790697674423e-06,
+      "loss": 0.8127,
+      "step": 68
+    },
+    {
+      "epoch": 0.0962566844919786,
+      "grad_norm": 2.472187333798063,
+      "learning_rate": 3.1627906976744187e-06,
+      "loss": 0.757,
+      "step": 69
+    },
+    {
+      "epoch": 0.09765170890490583,
+      "grad_norm": 2.326102202172242,
+      "learning_rate": 3.2093023255813956e-06,
+      "loss": 0.722,
+      "step": 70
+    },
+    {
+      "epoch": 0.09904673331783306,
+      "grad_norm": 2.515731147365297,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.7583,
+      "step": 71
+    },
+    {
+      "epoch": 0.10044175773076029,
+      "grad_norm": 2.509293997344943,
+      "learning_rate": 3.3023255813953493e-06,
+      "loss": 0.7642,
+      "step": 72
+    },
+    {
+      "epoch": 0.10183678214368752,
+      "grad_norm": 2.507853588356,
+      "learning_rate": 3.3488372093023258e-06,
+      "loss": 0.7746,
+      "step": 73
+    },
+    {
+      "epoch": 0.10323180655661474,
+      "grad_norm": 2.3923574692972136,
+      "learning_rate": 3.3953488372093026e-06,
+      "loss": 0.7831,
+      "step": 74
+    },
+    {
+      "epoch": 0.10462683096954196,
+      "grad_norm": 2.5608099260959634,
+      "learning_rate": 3.4418604651162795e-06,
+      "loss": 0.7607,
+      "step": 75
+    },
+    {
+      "epoch": 0.1060218553824692,
+      "grad_norm": 2.487504578194893,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.6625,
+      "step": 76
+    },
+    {
+      "epoch": 0.10741687979539642,
+      "grad_norm": 2.1209002756919726,
+      "learning_rate": 3.534883720930233e-06,
+      "loss": 0.6498,
+      "step": 77
+    },
+    {
+      "epoch": 0.10881190420832365,
+      "grad_norm": 2.4857824461829163,
+      "learning_rate": 3.5813953488372093e-06,
+      "loss": 0.7939,
+      "step": 78
+    },
+    {
+      "epoch": 0.11020692862125087,
+      "grad_norm": 2.3220344228660745,
+      "learning_rate": 3.6279069767441866e-06,
+      "loss": 0.763,
+      "step": 79
+    },
+    {
+      "epoch": 0.1116019530341781,
+      "grad_norm": 2.605269469509474,
+      "learning_rate": 3.674418604651163e-06,
+      "loss": 0.8045,
+      "step": 80
+    },
+    {
+      "epoch": 0.11299697744710532,
+      "grad_norm": 2.4819518552638486,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.7621,
+      "step": 81
+    },
+    {
+      "epoch": 0.11439200186003255,
+      "grad_norm": 2.8984252770783128,
+      "learning_rate": 3.7674418604651163e-06,
+      "loss": 0.8571,
+      "step": 82
+    },
+    {
+      "epoch": 0.11578702627295978,
+      "grad_norm": 2.473293325148463,
+      "learning_rate": 3.8139534883720936e-06,
+      "loss": 0.7575,
+      "step": 83
+    },
+    {
+      "epoch": 0.11718205068588701,
+      "grad_norm": 2.3794798368298493,
+      "learning_rate": 3.86046511627907e-06,
+      "loss": 0.7546,
+      "step": 84
+    },
+    {
+      "epoch": 0.11857707509881422,
+      "grad_norm": 2.337615082729649,
+      "learning_rate": 3.906976744186047e-06,
+      "loss": 0.7393,
+      "step": 85
+    },
+    {
+      "epoch": 0.11997209951174145,
+      "grad_norm": 2.336365843765245,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.6678,
+      "step": 86
+    },
+    {
+      "epoch": 0.12136712392466868,
+      "grad_norm": 2.5123928536454434,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.8062,
+      "step": 87
+    },
+    {
+      "epoch": 0.12276214833759591,
+      "grad_norm": 2.5461458978648115,
+      "learning_rate": 4.0465116279069775e-06,
+      "loss": 0.821,
+      "step": 88
+    },
+    {
+      "epoch": 0.12415717275052314,
+      "grad_norm": 2.546220602321451,
+      "learning_rate": 4.0930232558139536e-06,
+      "loss": 0.8603,
+      "step": 89
+    },
+    {
+      "epoch": 0.12555219716345037,
+      "grad_norm": 2.4356199052550744,
+      "learning_rate": 4.1395348837209304e-06,
+      "loss": 0.7435,
+      "step": 90
+    },
+    {
+      "epoch": 0.12694722157637758,
+      "grad_norm": 2.4024492143380054,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.7425,
+      "step": 91
+    },
+    {
+      "epoch": 0.12834224598930483,
+      "grad_norm": 2.4927376817528017,
+      "learning_rate": 4.232558139534884e-06,
+      "loss": 0.7943,
+      "step": 92
+    },
+    {
+      "epoch": 0.12973727040223204,
+      "grad_norm": 2.546222385189929,
+      "learning_rate": 4.279069767441861e-06,
+      "loss": 0.7444,
+      "step": 93
+    },
+    {
+      "epoch": 0.13113229481515926,
+      "grad_norm": 2.3638464894566,
+      "learning_rate": 4.325581395348837e-06,
+      "loss": 0.7103,
+      "step": 94
+    },
+    {
+      "epoch": 0.1325273192280865,
+      "grad_norm": 2.0116912405977367,
+      "learning_rate": 4.372093023255815e-06,
+      "loss": 0.6561,
+      "step": 95
+    },
+    {
+      "epoch": 0.1339223436410137,
+      "grad_norm": 2.3569671450914402,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.7494,
+      "step": 96
+    },
+    {
+      "epoch": 0.13531736805394096,
+      "grad_norm": 2.455174006078999,
+      "learning_rate": 4.465116279069768e-06,
+      "loss": 0.7291,
+      "step": 97
+    },
+    {
+      "epoch": 0.13671239246686817,
+      "grad_norm": 2.3259679224878895,
+      "learning_rate": 4.5116279069767445e-06,
+      "loss": 0.6947,
+      "step": 98
+    },
+    {
+      "epoch": 0.13810741687979539,
+      "grad_norm": 2.2889925038801358,
+      "learning_rate": 4.558139534883721e-06,
+      "loss": 0.7622,
+      "step": 99
+    },
+    {
+      "epoch": 0.13950244129272263,
+      "grad_norm": 2.34754065529055,
+      "learning_rate": 4.604651162790698e-06,
+      "loss": 0.6875,
+      "step": 100
+    },
+    {
+      "epoch": 0.14089746570564984,
+      "grad_norm": 2.4553141008398893,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.7712,
+      "step": 101
+    },
+    {
+      "epoch": 0.1422924901185771,
+      "grad_norm": 2.4054475492234153,
+      "learning_rate": 4.697674418604651e-06,
+      "loss": 0.7644,
+      "step": 102
+    },
+    {
+      "epoch": 0.1436875145315043,
+      "grad_norm": 2.5744411230055153,
+      "learning_rate": 4.744186046511629e-06,
+      "loss": 0.7202,
+      "step": 103
+    },
+    {
+      "epoch": 0.14508253894443152,
+      "grad_norm": 2.3847788121371774,
+      "learning_rate": 4.790697674418605e-06,
+      "loss": 0.7174,
+      "step": 104
+    },
+    {
+      "epoch": 0.14647756335735876,
+      "grad_norm": 2.3377822690454444,
+      "learning_rate": 4.837209302325582e-06,
+      "loss": 0.7038,
+      "step": 105
+    },
+    {
+      "epoch": 0.14787258777028597,
+      "grad_norm": 2.428457278015599,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.6988,
+      "step": 106
+    },
+    {
+      "epoch": 0.14926761218321322,
+      "grad_norm": 2.3799317230323846,
+      "learning_rate": 4.9302325581395355e-06,
+      "loss": 0.7673,
+      "step": 107
+    },
+    {
+      "epoch": 0.15066263659614043,
+      "grad_norm": 2.35083503319128,
+      "learning_rate": 4.976744186046512e-06,
+      "loss": 0.7376,
+      "step": 108
+    },
+    {
+      "epoch": 0.15205766100906765,
+      "grad_norm": 2.805407049871798,
+      "learning_rate": 5.023255813953489e-06,
+      "loss": 0.9254,
+      "step": 109
+    },
+    {
+      "epoch": 0.1534526854219949,
+      "grad_norm": 2.1905803113537403,
+      "learning_rate": 5.069767441860466e-06,
+      "loss": 0.6927,
+      "step": 110
+    },
+    {
+      "epoch": 0.1548477098349221,
+      "grad_norm": 2.2807494953381706,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.7171,
+      "step": 111
+    },
+    {
+      "epoch": 0.15624273424784935,
+      "grad_norm": 2.3977975510102922,
+      "learning_rate": 5.162790697674419e-06,
+      "loss": 0.7676,
+      "step": 112
+    },
+    {
+      "epoch": 0.15763775866077656,
+      "grad_norm": 2.374163440061089,
+      "learning_rate": 5.209302325581396e-06,
+      "loss": 0.7407,
+      "step": 113
+    },
+    {
+      "epoch": 0.15903278307370378,
+      "grad_norm": 2.5800289116181085,
+      "learning_rate": 5.255813953488372e-06,
+      "loss": 0.7664,
+      "step": 114
+    },
+    {
+      "epoch": 0.16042780748663102,
+      "grad_norm": 2.430187403796108,
+      "learning_rate": 5.302325581395349e-06,
+      "loss": 0.751,
+      "step": 115
+    },
+    {
+      "epoch": 0.16182283189955823,
+      "grad_norm": 2.343909179556326,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.7237,
+      "step": 116
+    },
+    {
+      "epoch": 0.16321785631248548,
+      "grad_norm": 2.548797829296356,
+      "learning_rate": 5.395348837209303e-06,
+      "loss": 0.7932,
+      "step": 117
+    },
+    {
+      "epoch": 0.1646128807254127,
+      "grad_norm": 2.5320929256806823,
+      "learning_rate": 5.44186046511628e-06,
+      "loss": 0.7457,
+      "step": 118
+    },
+    {
+      "epoch": 0.16600790513833993,
+      "grad_norm": 2.5449094558951266,
+      "learning_rate": 5.488372093023256e-06,
+      "loss": 0.7163,
+      "step": 119
+    },
+    {
+      "epoch": 0.16740292955126715,
+      "grad_norm": 2.504891081408036,
+      "learning_rate": 5.534883720930233e-06,
+      "loss": 0.7197,
+      "step": 120
+    },
+    {
+      "epoch": 0.16879795396419436,
+      "grad_norm": 2.475849061902109,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.7522,
+      "step": 121
+    },
+    {
+      "epoch": 0.1701929783771216,
+      "grad_norm": 2.4821935283018584,
+      "learning_rate": 5.627906976744186e-06,
+      "loss": 0.7701,
+      "step": 122
+    },
+    {
+      "epoch": 0.17158800279004882,
+      "grad_norm": 2.642205698934932,
+      "learning_rate": 5.674418604651163e-06,
+      "loss": 0.7893,
+      "step": 123
+    },
+    {
+      "epoch": 0.17298302720297606,
+      "grad_norm": 2.5207552171233187,
+      "learning_rate": 5.72093023255814e-06,
+      "loss": 0.7317,
+      "step": 124
+    },
+    {
+      "epoch": 0.17437805161590328,
+      "grad_norm": 2.3092477828749645,
+      "learning_rate": 5.7674418604651175e-06,
+      "loss": 0.7706,
+      "step": 125
+    },
+    {
+      "epoch": 0.1757730760288305,
+      "grad_norm": 2.489926710299257,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.7555,
+      "step": 126
+    },
+    {
+      "epoch": 0.17716810044175774,
+      "grad_norm": 2.656527553681413,
+      "learning_rate": 5.86046511627907e-06,
+      "loss": 0.7651,
+      "step": 127
+    },
+    {
+      "epoch": 0.17856312485468495,
+      "grad_norm": 2.3563815058683923,
+      "learning_rate": 5.906976744186047e-06,
+      "loss": 0.7595,
+      "step": 128
+    },
+    {
+      "epoch": 0.1799581492676122,
+      "grad_norm": 2.3679438646515596,
+      "learning_rate": 5.953488372093023e-06,
+      "loss": 0.8296,
+      "step": 129
+    },
+    {
+      "epoch": 0.1813531736805394,
+      "grad_norm": 2.4057010731210307,
+      "learning_rate": 6e-06,
+      "loss": 0.745,
+      "step": 130
+    },
+    {
+      "epoch": 0.18274819809346662,
+      "grad_norm": 2.522120968034962,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.6755,
+      "step": 131
+    },
+    {
+      "epoch": 0.18414322250639387,
+      "grad_norm": 2.4421735038455683,
+      "learning_rate": 6.093023255813954e-06,
+      "loss": 0.7682,
+      "step": 132
+    },
+    {
+      "epoch": 0.18553824691932108,
+      "grad_norm": 2.549138792539662,
+      "learning_rate": 6.139534883720932e-06,
+      "loss": 0.7427,
+      "step": 133
+    },
+    {
+      "epoch": 0.18693327133224832,
+      "grad_norm": 2.2976453414137628,
+      "learning_rate": 6.186046511627908e-06,
+      "loss": 0.7355,
+      "step": 134
+    },
+    {
+      "epoch": 0.18832829574517554,
+      "grad_norm": 2.612827929707925,
+      "learning_rate": 6.2325581395348845e-06,
+      "loss": 0.8027,
+      "step": 135
+    },
+    {
+      "epoch": 0.18972332015810275,
+      "grad_norm": 2.5538000384173296,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.7745,
+      "step": 136
+    },
+    {
+      "epoch": 0.19111834457103,
+      "grad_norm": 2.5739544308712565,
+      "learning_rate": 6.325581395348837e-06,
+      "loss": 0.7536,
+      "step": 137
+    },
+    {
+      "epoch": 0.1925133689839572,
+      "grad_norm": 2.4900615363384473,
+      "learning_rate": 6.372093023255814e-06,
+      "loss": 0.7661,
+      "step": 138
+    },
+    {
+      "epoch": 0.19390839339688445,
+      "grad_norm": 2.72347356864473,
+      "learning_rate": 6.418604651162791e-06,
+      "loss": 0.7113,
+      "step": 139
+    },
+    {
+      "epoch": 0.19530341780981167,
+      "grad_norm": 2.4124895218072173,
+      "learning_rate": 6.465116279069767e-06,
+      "loss": 0.7844,
+      "step": 140
+    },
+    {
+      "epoch": 0.1966984422227389,
+      "grad_norm": 2.470332588836169,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.7984,
+      "step": 141
+    },
+    {
+      "epoch": 0.19809346663566613,
+      "grad_norm": 2.634184887029152,
+      "learning_rate": 6.558139534883722e-06,
+      "loss": 0.749,
+      "step": 142
+    },
+    {
+      "epoch": 0.19948849104859334,
+      "grad_norm": 2.3551296639045742,
+      "learning_rate": 6.604651162790699e-06,
+      "loss": 0.7404,
+      "step": 143
+    },
+    {
+      "epoch": 0.20088351546152058,
+      "grad_norm": 2.318054989604665,
+      "learning_rate": 6.651162790697675e-06,
+      "loss": 0.7011,
+      "step": 144
+    },
+    {
+      "epoch": 0.2022785398744478,
+      "grad_norm": 2.394650549678866,
+      "learning_rate": 6.6976744186046515e-06,
+      "loss": 0.7989,
+      "step": 145
+    },
+    {
+      "epoch": 0.20367356428737504,
+      "grad_norm": 2.3289567515126226,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.7381,
+      "step": 146
+    },
+    {
+      "epoch": 0.20506858870030226,
+      "grad_norm": 2.533330784474734,
+      "learning_rate": 6.790697674418605e-06,
+      "loss": 0.8333,
+      "step": 147
+    },
+    {
+      "epoch": 0.20646361311322947,
+      "grad_norm": 2.3789567467158474,
+      "learning_rate": 6.837209302325581e-06,
+      "loss": 0.7275,
+      "step": 148
+    },
+    {
+      "epoch": 0.20785863752615671,
+      "grad_norm": 2.5367630078041827,
+      "learning_rate": 6.883720930232559e-06,
+      "loss": 0.8271,
+      "step": 149
+    },
+    {
+      "epoch": 0.20925366193908393,
+      "grad_norm": 2.4354594911133263,
+      "learning_rate": 6.930232558139536e-06,
+      "loss": 0.7624,
+      "step": 150
+    },
+    {
+      "epoch": 0.21064868635201117,
+      "grad_norm": 2.5775059164500758,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.7584,
+      "step": 151
+    },
+    {
+      "epoch": 0.2120437107649384,
+      "grad_norm": 2.190325083282746,
+      "learning_rate": 7.023255813953489e-06,
+      "loss": 0.7009,
+      "step": 152
+    },
+    {
+      "epoch": 0.2134387351778656,
+      "grad_norm": 2.4921469453291345,
+      "learning_rate": 7.069767441860466e-06,
+      "loss": 0.7554,
+      "step": 153
+    },
+    {
+      "epoch": 0.21483375959079284,
+      "grad_norm": 2.3569726016132635,
+      "learning_rate": 7.1162790697674425e-06,
+      "loss": 0.727,
+      "step": 154
+    },
+    {
+      "epoch": 0.21622878400372006,
+      "grad_norm": 2.3994767087861897,
+      "learning_rate": 7.1627906976744185e-06,
+      "loss": 0.7333,
+      "step": 155
+    },
+    {
+      "epoch": 0.2176238084166473,
+      "grad_norm": 2.316378530911968,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.755,
+      "step": 156
+    },
+    {
+      "epoch": 0.21901883282957452,
+      "grad_norm": 2.1898592032828033,
+      "learning_rate": 7.255813953488373e-06,
+      "loss": 0.6771,
+      "step": 157
+    },
+    {
+      "epoch": 0.22041385724250173,
+      "grad_norm": 2.6149357411219154,
+      "learning_rate": 7.30232558139535e-06,
+      "loss": 0.7357,
+      "step": 158
+    },
+    {
+      "epoch": 0.22180888165542897,
+      "grad_norm": 2.2792049067557216,
+      "learning_rate": 7.348837209302326e-06,
+      "loss": 0.7254,
+      "step": 159
+    },
+    {
+      "epoch": 0.2232039060683562,
+      "grad_norm": 2.4877232933622775,
+      "learning_rate": 7.395348837209303e-06,
+      "loss": 0.7488,
+      "step": 160
+    },
+    {
+      "epoch": 0.22459893048128343,
+      "grad_norm": 2.4560817023680097,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.7413,
+      "step": 161
+    },
+    {
+      "epoch": 0.22599395489421065,
+      "grad_norm": 2.4221544352169975,
+      "learning_rate": 7.488372093023256e-06,
+      "loss": 0.7363,
+      "step": 162
+    },
+    {
+      "epoch": 0.22738897930713786,
+      "grad_norm": 2.4563224434742805,
+      "learning_rate": 7.534883720930233e-06,
+      "loss": 0.7673,
+      "step": 163
+    },
+    {
+      "epoch": 0.2287840037200651,
+      "grad_norm": 2.366320987925906,
+      "learning_rate": 7.5813953488372095e-06,
+      "loss": 0.6675,
+      "step": 164
+    },
+    {
+      "epoch": 0.23017902813299232,
+      "grad_norm": 2.5227544509508517,
+      "learning_rate": 7.627906976744187e-06,
+      "loss": 0.7483,
+      "step": 165
+    },
+    {
+      "epoch": 0.23157405254591956,
+      "grad_norm": 2.1154831131584975,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.6732,
+      "step": 166
+    },
+    {
+      "epoch": 0.23296907695884678,
+      "grad_norm": 2.4673285889919927,
+      "learning_rate": 7.72093023255814e-06,
+      "loss": 0.7009,
+      "step": 167
+    },
+    {
+      "epoch": 0.23436410137177402,
+      "grad_norm": 2.43878962840609,
+      "learning_rate": 7.767441860465116e-06,
+      "loss": 0.7252,
+      "step": 168
+    },
+    {
+      "epoch": 0.23575912578470123,
+      "grad_norm": 2.31935934137348,
+      "learning_rate": 7.813953488372094e-06,
+      "loss": 0.733,
+      "step": 169
+    },
+    {
+      "epoch": 0.23715415019762845,
+      "grad_norm": 2.1995960807975603,
+      "learning_rate": 7.86046511627907e-06,
+      "loss": 0.6615,
+      "step": 170
+    },
+    {
+      "epoch": 0.2385491746105557,
+      "grad_norm": 2.168278254449867,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.7284,
+      "step": 171
+    },
+    {
+      "epoch": 0.2399441990234829,
+      "grad_norm": 2.4173940989857847,
+      "learning_rate": 7.953488372093024e-06,
+      "loss": 0.7081,
+      "step": 172
+    },
+    {
+      "epoch": 0.24133922343641015,
+      "grad_norm": 2.51400863859905,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.7747,
+      "step": 173
+    },
+    {
+      "epoch": 0.24273424784933736,
+      "grad_norm": 2.3463603446571857,
+      "learning_rate": 8.046511627906977e-06,
+      "loss": 0.7386,
+      "step": 174
+    },
+    {
+      "epoch": 0.24412927226226458,
+      "grad_norm": 2.3501001279570897,
+      "learning_rate": 8.093023255813955e-06,
+      "loss": 0.7658,
+      "step": 175
+    },
+    {
+      "epoch": 0.24552429667519182,
+      "grad_norm": 2.390612949278623,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.7733,
+      "step": 176
+    },
+    {
+      "epoch": 0.24691932108811904,
+      "grad_norm": 2.5310428833750325,
+      "learning_rate": 8.186046511627907e-06,
+      "loss": 0.7699,
+      "step": 177
+    },
+    {
+      "epoch": 0.24831434550104628,
+      "grad_norm": 2.5096094301774383,
+      "learning_rate": 8.232558139534885e-06,
+      "loss": 0.774,
+      "step": 178
+    },
+    {
+      "epoch": 0.2497093699139735,
+      "grad_norm": 2.3076005947523646,
+      "learning_rate": 8.279069767441861e-06,
+      "loss": 0.7115,
+      "step": 179
+    },
+    {
+      "epoch": 0.25110439432690074,
+      "grad_norm": 2.407603733719916,
+      "learning_rate": 8.325581395348837e-06,
+      "loss": 0.7587,
+      "step": 180
+    },
+    {
+      "epoch": 0.25249941873982795,
+      "grad_norm": 2.5252107383278233,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.75,
+      "step": 181
+    },
+    {
+      "epoch": 0.25389444315275517,
+      "grad_norm": 2.309734616227639,
+      "learning_rate": 8.418604651162792e-06,
+      "loss": 0.7088,
+      "step": 182
+    },
+    {
+      "epoch": 0.2552894675656824,
+      "grad_norm": 2.426978371777092,
+      "learning_rate": 8.465116279069768e-06,
+      "loss": 0.7468,
+      "step": 183
+    },
+    {
+      "epoch": 0.25668449197860965,
+      "grad_norm": 2.4432053380067065,
+      "learning_rate": 8.511627906976744e-06,
+      "loss": 0.7142,
+      "step": 184
+    },
+    {
+      "epoch": 0.25807951639153687,
+      "grad_norm": 2.3740041197703023,
+      "learning_rate": 8.558139534883722e-06,
+      "loss": 0.7071,
+      "step": 185
+    },
+    {
+      "epoch": 0.2594745408044641,
+      "grad_norm": 2.0601901871327377,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.6276,
+      "step": 186
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 2.364861360993871,
+      "learning_rate": 8.651162790697674e-06,
+      "loss": 0.7856,
+      "step": 187
+    },
+    {
+      "epoch": 0.2622645896303185,
+      "grad_norm": 2.3561545953781597,
+      "learning_rate": 8.697674418604652e-06,
+      "loss": 0.7252,
+      "step": 188
+    },
+    {
+      "epoch": 0.2636596140432458,
+      "grad_norm": 2.282838512331675,
+      "learning_rate": 8.74418604651163e-06,
+      "loss": 0.6964,
+      "step": 189
+    },
+    {
+      "epoch": 0.265054638456173,
+      "grad_norm": 2.38101153955705,
+      "learning_rate": 8.790697674418606e-06,
+      "loss": 0.7203,
+      "step": 190
+    },
+    {
+      "epoch": 0.2664496628691002,
+      "grad_norm": 2.4046315413267605,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.7592,
+      "step": 191
+    },
+    {
+      "epoch": 0.2678446872820274,
+      "grad_norm": 2.3462917230346885,
+      "learning_rate": 8.88372093023256e-06,
+      "loss": 0.701,
+      "step": 192
+    },
+    {
+      "epoch": 0.26923971169495464,
+      "grad_norm": 2.36595710790288,
+      "learning_rate": 8.930232558139535e-06,
+      "loss": 0.7542,
+      "step": 193
+    },
+    {
+      "epoch": 0.2706347361078819,
+      "grad_norm": 2.3866511967235944,
+      "learning_rate": 8.976744186046511e-06,
+      "loss": 0.7001,
+      "step": 194
+    },
+    {
+      "epoch": 0.2720297605208091,
+      "grad_norm": 2.3552326990004517,
+      "learning_rate": 9.023255813953489e-06,
+      "loss": 0.7146,
+      "step": 195
+    },
+    {
+      "epoch": 0.27342478493373634,
+      "grad_norm": 2.234288729152599,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.7804,
+      "step": 196
+    },
+    {
+      "epoch": 0.27481980934666356,
+      "grad_norm": 2.2137571036217722,
+      "learning_rate": 9.116279069767443e-06,
+      "loss": 0.709,
+      "step": 197
+    },
+    {
+      "epoch": 0.27621483375959077,
+      "grad_norm": 2.362054981862146,
+      "learning_rate": 9.162790697674419e-06,
+      "loss": 0.7079,
+      "step": 198
+    },
+    {
+      "epoch": 0.27760985817251804,
+      "grad_norm": 2.3776725722324925,
+      "learning_rate": 9.209302325581397e-06,
+      "loss": 0.7115,
+      "step": 199
+    },
+    {
+      "epoch": 0.27900488258544526,
+      "grad_norm": 2.360369888299576,
+      "learning_rate": 9.255813953488373e-06,
+      "loss": 0.7238,
+      "step": 200
+    },
+    {
+      "epoch": 0.2803999069983725,
+      "grad_norm": 2.2507389312676462,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.7214,
+      "step": 201
+    },
+    {
+      "epoch": 0.2817949314112997,
+      "grad_norm": 2.279748703497406,
+      "learning_rate": 9.348837209302326e-06,
+      "loss": 0.7474,
+      "step": 202
+    },
+    {
+      "epoch": 0.2831899558242269,
+      "grad_norm": 2.3429648273532937,
+      "learning_rate": 9.395348837209302e-06,
+      "loss": 0.7681,
+      "step": 203
+    },
+    {
+      "epoch": 0.2845849802371542,
+      "grad_norm": 2.418888049162119,
+      "learning_rate": 9.44186046511628e-06,
+      "loss": 0.727,
+      "step": 204
+    },
+    {
+      "epoch": 0.2859800046500814,
+      "grad_norm": 2.399989246612679,
+      "learning_rate": 9.488372093023258e-06,
+      "loss": 0.7719,
+      "step": 205
+    },
+    {
+      "epoch": 0.2873750290630086,
+      "grad_norm": 2.660020683455921,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.7441,
+      "step": 206
+    },
+    {
+      "epoch": 0.2887700534759358,
+      "grad_norm": 2.143497203846691,
+      "learning_rate": 9.58139534883721e-06,
+      "loss": 0.6955,
+      "step": 207
+    },
+    {
+      "epoch": 0.29016507788886303,
+      "grad_norm": 2.4890129433808843,
+      "learning_rate": 9.627906976744188e-06,
+      "loss": 0.7716,
+      "step": 208
+    },
+    {
+      "epoch": 0.2915601023017903,
+      "grad_norm": 2.320446723006289,
+      "learning_rate": 9.674418604651164e-06,
+      "loss": 0.7583,
+      "step": 209
+    },
+    {
+      "epoch": 0.2929551267147175,
+      "grad_norm": 2.411341999029131,
+      "learning_rate": 9.72093023255814e-06,
+      "loss": 0.7681,
+      "step": 210
+    },
+    {
+      "epoch": 0.29435015112764473,
+      "grad_norm": 2.301874213978857,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.7208,
+      "step": 211
+    },
+    {
+      "epoch": 0.29574517554057195,
+      "grad_norm": 2.640983189271861,
+      "learning_rate": 9.813953488372093e-06,
+      "loss": 0.7693,
+      "step": 212
+    },
+    {
+      "epoch": 0.29714019995349916,
+      "grad_norm": 2.4184498532898564,
+      "learning_rate": 9.860465116279071e-06,
+      "loss": 0.7405,
+      "step": 213
+    },
+    {
+      "epoch": 0.29853522436642643,
+      "grad_norm": 2.3297820871229202,
+      "learning_rate": 9.906976744186047e-06,
+      "loss": 0.7111,
+      "step": 214
+    },
+    {
+      "epoch": 0.29993024877935365,
+      "grad_norm": 2.4217191627013626,
+      "learning_rate": 9.953488372093025e-06,
+      "loss": 0.7665,
+      "step": 215
+    },
+    {
+      "epoch": 0.30132527319228086,
+      "grad_norm": 2.401188958751426,
+      "learning_rate": 1e-05,
+      "loss": 0.7454,
+      "step": 216
+    },
+    {
+      "epoch": 0.3027202976052081,
+      "grad_norm": 2.4499106462941094,
+      "learning_rate": 9.999993396473114e-06,
+      "loss": 0.7878,
+      "step": 217
+    },
+    {
+      "epoch": 0.3041153220181353,
+      "grad_norm": 2.2276901973554777,
+      "learning_rate": 9.999973585909898e-06,
+      "loss": 0.6551,
+      "step": 218
+    },
+    {
+      "epoch": 0.30551034643106256,
+      "grad_norm": 2.3733342447896417,
+      "learning_rate": 9.99994056836268e-06,
+      "loss": 0.7145,
+      "step": 219
+    },
+    {
+      "epoch": 0.3069053708439898,
+      "grad_norm": 2.29189829429208,
+      "learning_rate": 9.999894343918674e-06,
+      "loss": 0.7239,
+      "step": 220
+    },
+    {
+      "epoch": 0.308300395256917,
+      "grad_norm": 2.217573234027078,
+      "learning_rate": 9.999834912699974e-06,
+      "loss": 0.7756,
+      "step": 221
+    },
+    {
+      "epoch": 0.3096954196698442,
+      "grad_norm": 2.2514531625857677,
+      "learning_rate": 9.999762274863567e-06,
+      "loss": 0.7255,
+      "step": 222
+    },
+    {
+      "epoch": 0.3110904440827714,
+      "grad_norm": 2.4225676302200942,
+      "learning_rate": 9.999676430601318e-06,
+      "loss": 0.7809,
+      "step": 223
+    },
+    {
+      "epoch": 0.3124854684956987,
+      "grad_norm": 2.167234001414,
+      "learning_rate": 9.999577380139976e-06,
+      "loss": 0.702,
+      "step": 224
+    },
+    {
+      "epoch": 0.3138804929086259,
+      "grad_norm": 2.3885170041413004,
+      "learning_rate": 9.999465123741172e-06,
+      "loss": 0.7084,
+      "step": 225
+    },
+    {
+      "epoch": 0.3152755173215531,
+      "grad_norm": 2.3894680417925604,
+      "learning_rate": 9.999339661701424e-06,
+      "loss": 0.7033,
+      "step": 226
+    },
+    {
+      "epoch": 0.31667054173448034,
+      "grad_norm": 2.2270776846137013,
+      "learning_rate": 9.99920099435213e-06,
+      "loss": 0.7521,
+      "step": 227
+    },
+    {
+      "epoch": 0.31806556614740755,
+      "grad_norm": 2.517415559204921,
+      "learning_rate": 9.999049122059565e-06,
+      "loss": 0.7706,
+      "step": 228
+    },
+    {
+      "epoch": 0.3194605905603348,
+      "grad_norm": 2.1313262835401288,
+      "learning_rate": 9.998884045224886e-06,
+      "loss": 0.7236,
+      "step": 229
+    },
+    {
+      "epoch": 0.32085561497326204,
+      "grad_norm": 2.2375739677464965,
+      "learning_rate": 9.998705764284132e-06,
+      "loss": 0.6958,
+      "step": 230
+    },
+    {
+      "epoch": 0.32225063938618925,
+      "grad_norm": 2.4499599731198214,
+      "learning_rate": 9.998514279708212e-06,
+      "loss": 0.7827,
+      "step": 231
+    },
+    {
+      "epoch": 0.32364566379911647,
+      "grad_norm": 2.355065280389441,
+      "learning_rate": 9.998309592002914e-06,
+      "loss": 0.7762,
+      "step": 232
+    },
+    {
+      "epoch": 0.32504068821204374,
+      "grad_norm": 2.3578314293622573,
+      "learning_rate": 9.99809170170891e-06,
+      "loss": 0.6852,
+      "step": 233
+    },
+    {
+      "epoch": 0.32643571262497095,
+      "grad_norm": 2.259491531550109,
+      "learning_rate": 9.997860609401732e-06,
+      "loss": 0.7778,
+      "step": 234
+    },
+    {
+      "epoch": 0.32783073703789817,
+      "grad_norm": 2.2038320551296673,
+      "learning_rate": 9.99761631569179e-06,
+      "loss": 0.7799,
+      "step": 235
+    },
+    {
+      "epoch": 0.3292257614508254,
+      "grad_norm": 2.42655922351898,
+      "learning_rate": 9.997358821224365e-06,
+      "loss": 0.801,
+      "step": 236
+    },
+    {
+      "epoch": 0.3306207858637526,
+      "grad_norm": 2.3511248366461253,
+      "learning_rate": 9.997088126679607e-06,
+      "loss": 0.7303,
+      "step": 237
+    },
+    {
+      "epoch": 0.33201581027667987,
+      "grad_norm": 2.290728698460702,
+      "learning_rate": 9.996804232772528e-06,
+      "loss": 0.7332,
+      "step": 238
+    },
+    {
+      "epoch": 0.3334108346896071,
+      "grad_norm": 2.2585293725350226,
+      "learning_rate": 9.996507140253012e-06,
+      "loss": 0.7565,
+      "step": 239
+    },
+    {
+      "epoch": 0.3348058591025343,
+      "grad_norm": 2.162605836590699,
+      "learning_rate": 9.9961968499058e-06,
+      "loss": 0.7403,
+      "step": 240
+    },
+    {
+      "epoch": 0.3362008835154615,
+      "grad_norm": 2.44214305770412,
+      "learning_rate": 9.9958733625505e-06,
+      "loss": 0.7672,
+      "step": 241
+    },
+    {
+      "epoch": 0.3375959079283887,
+      "grad_norm": 2.1410609124245257,
+      "learning_rate": 9.995536679041568e-06,
+      "loss": 0.7083,
+      "step": 242
+    },
+    {
+      "epoch": 0.338990932341316,
+      "grad_norm": 2.3045558078430917,
+      "learning_rate": 9.99518680026833e-06,
+      "loss": 0.759,
+      "step": 243
+    },
+    {
+      "epoch": 0.3403859567542432,
+      "grad_norm": 2.1962856519735063,
+      "learning_rate": 9.994823727154957e-06,
+      "loss": 0.7431,
+      "step": 244
+    },
+    {
+      "epoch": 0.34178098116717043,
+      "grad_norm": 2.538182288531259,
+      "learning_rate": 9.994447460660473e-06,
+      "loss": 0.7658,
+      "step": 245
+    },
+    {
+      "epoch": 0.34317600558009764,
+      "grad_norm": 2.147870815921298,
+      "learning_rate": 9.994058001778754e-06,
+      "loss": 0.7533,
+      "step": 246
+    },
+    {
+      "epoch": 0.34457102999302486,
+      "grad_norm": 2.268535359430784,
+      "learning_rate": 9.99365535153852e-06,
+      "loss": 0.7256,
+      "step": 247
+    },
+    {
+      "epoch": 0.34596605440595213,
+      "grad_norm": 2.2131932435600126,
+      "learning_rate": 9.993239511003338e-06,
+      "loss": 0.7366,
+      "step": 248
+    },
+    {
+      "epoch": 0.34736107881887934,
+      "grad_norm": 2.4607861240746445,
+      "learning_rate": 9.992810481271611e-06,
+      "loss": 0.7347,
+      "step": 249
+    },
+    {
+      "epoch": 0.34875610323180656,
+      "grad_norm": 2.2713964429296007,
+      "learning_rate": 9.992368263476585e-06,
+      "loss": 0.7511,
+      "step": 250
+    },
+    {
+      "epoch": 0.3501511276447338,
+      "grad_norm": 2.227939017521476,
+      "learning_rate": 9.991912858786335e-06,
+      "loss": 0.7604,
+      "step": 251
+    },
+    {
+      "epoch": 0.351546152057661,
+      "grad_norm": 2.354239376083879,
+      "learning_rate": 9.991444268403776e-06,
+      "loss": 0.7632,
+      "step": 252
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 2.312214771993005,
+      "learning_rate": 9.990962493566645e-06,
+      "loss": 0.8016,
+      "step": 253
+    },
+    {
+      "epoch": 0.3543362008835155,
+      "grad_norm": 2.431880313811741,
+      "learning_rate": 9.99046753554751e-06,
+      "loss": 0.7021,
+      "step": 254
+    },
+    {
+      "epoch": 0.3557312252964427,
+      "grad_norm": 2.2846086141474573,
+      "learning_rate": 9.989959395653756e-06,
+      "loss": 0.7367,
+      "step": 255
+    },
+    {
+      "epoch": 0.3571262497093699,
+      "grad_norm": 2.1027600513299674,
+      "learning_rate": 9.989438075227588e-06,
+      "loss": 0.6324,
+      "step": 256
+    },
+    {
+      "epoch": 0.3585212741222971,
+      "grad_norm": 2.3013858529857756,
+      "learning_rate": 9.988903575646032e-06,
+      "loss": 0.6831,
+      "step": 257
+    },
+    {
+      "epoch": 0.3599162985352244,
+      "grad_norm": 2.232795724746918,
+      "learning_rate": 9.988355898320917e-06,
+      "loss": 0.7296,
+      "step": 258
+    },
+    {
+      "epoch": 0.3613113229481516,
+      "grad_norm": 2.5816820477428664,
+      "learning_rate": 9.987795044698885e-06,
+      "loss": 0.7603,
+      "step": 259
+    },
+    {
+      "epoch": 0.3627063473610788,
+      "grad_norm": 2.3931287274339623,
+      "learning_rate": 9.98722101626138e-06,
+      "loss": 0.7424,
+      "step": 260
+    },
+    {
+      "epoch": 0.36410137177400603,
+      "grad_norm": 2.236682063683957,
+      "learning_rate": 9.986633814524648e-06,
+      "loss": 0.7005,
+      "step": 261
+    },
+    {
+      "epoch": 0.36549639618693325,
+      "grad_norm": 2.1826925703002416,
+      "learning_rate": 9.986033441039731e-06,
+      "loss": 0.7353,
+      "step": 262
+    },
+    {
+      "epoch": 0.3668914205998605,
+      "grad_norm": 2.4709157488020397,
+      "learning_rate": 9.985419897392459e-06,
+      "loss": 0.7928,
+      "step": 263
+    },
+    {
+      "epoch": 0.36828644501278773,
+      "grad_norm": 1.981866149603293,
+      "learning_rate": 9.984793185203456e-06,
+      "loss": 0.6918,
+      "step": 264
+    },
+    {
+      "epoch": 0.36968146942571495,
+      "grad_norm": 2.2345445872108045,
+      "learning_rate": 9.984153306128124e-06,
+      "loss": 0.7581,
+      "step": 265
+    },
+    {
+      "epoch": 0.37107649383864216,
+      "grad_norm": 2.5403963282899644,
+      "learning_rate": 9.983500261856646e-06,
+      "loss": 0.7616,
+      "step": 266
+    },
+    {
+      "epoch": 0.3724715182515694,
+      "grad_norm": 2.2145873038291866,
+      "learning_rate": 9.982834054113982e-06,
+      "loss": 0.6927,
+      "step": 267
+    },
+    {
+      "epoch": 0.37386654266449665,
+      "grad_norm": 2.415059957789154,
+      "learning_rate": 9.98215468465986e-06,
+      "loss": 0.7197,
+      "step": 268
+    },
+    {
+      "epoch": 0.37526156707742386,
+      "grad_norm": 2.1560691347775913,
+      "learning_rate": 9.981462155288773e-06,
+      "loss": 0.7377,
+      "step": 269
+    },
+    {
+      "epoch": 0.3766565914903511,
+      "grad_norm": 2.258433532015071,
+      "learning_rate": 9.980756467829977e-06,
+      "loss": 0.6838,
+      "step": 270
+    },
+    {
+      "epoch": 0.3780516159032783,
+      "grad_norm": 2.3329270998443166,
+      "learning_rate": 9.98003762414748e-06,
+      "loss": 0.7265,
+      "step": 271
+    },
+    {
+      "epoch": 0.3794466403162055,
+      "grad_norm": 2.1635360885551598,
+      "learning_rate": 9.979305626140046e-06,
+      "loss": 0.6886,
+      "step": 272
+    },
+    {
+      "epoch": 0.3808416647291328,
+      "grad_norm": 2.127114059196922,
+      "learning_rate": 9.978560475741181e-06,
+      "loss": 0.7118,
+      "step": 273
+    },
+    {
+      "epoch": 0.38223668914206,
+      "grad_norm": 2.217596961579976,
+      "learning_rate": 9.977802174919134e-06,
+      "loss": 0.7936,
+      "step": 274
+    },
+    {
+      "epoch": 0.3836317135549872,
+      "grad_norm": 2.136075798958187,
+      "learning_rate": 9.977030725676887e-06,
+      "loss": 0.6947,
+      "step": 275
+    },
+    {
+      "epoch": 0.3850267379679144,
+      "grad_norm": 2.271555095939749,
+      "learning_rate": 9.976246130052157e-06,
+      "loss": 0.7561,
+      "step": 276
+    },
+    {
+      "epoch": 0.38642176238084164,
+      "grad_norm": 2.256449368924786,
+      "learning_rate": 9.97544839011738e-06,
+      "loss": 0.8082,
+      "step": 277
+    },
+    {
+      "epoch": 0.3878167867937689,
+      "grad_norm": 2.089485218424677,
+      "learning_rate": 9.974637507979721e-06,
+      "loss": 0.6938,
+      "step": 278
+    },
+    {
+      "epoch": 0.3892118112066961,
+      "grad_norm": 2.2379904394260084,
+      "learning_rate": 9.973813485781045e-06,
+      "loss": 0.7031,
+      "step": 279
+    },
+    {
+      "epoch": 0.39060683561962334,
+      "grad_norm": 2.065922207391741,
+      "learning_rate": 9.972976325697938e-06,
+      "loss": 0.6992,
+      "step": 280
+    },
+    {
+      "epoch": 0.39200186003255055,
+      "grad_norm": 2.0549800888949528,
+      "learning_rate": 9.972126029941685e-06,
+      "loss": 0.7237,
+      "step": 281
+    },
+    {
+      "epoch": 0.3933968844454778,
+      "grad_norm": 2.212763606020422,
+      "learning_rate": 9.97126260075826e-06,
+      "loss": 0.7034,
+      "step": 282
+    },
+    {
+      "epoch": 0.39479190885840504,
+      "grad_norm": 2.3190648204003015,
+      "learning_rate": 9.97038604042834e-06,
+      "loss": 0.7811,
+      "step": 283
+    },
+    {
+      "epoch": 0.39618693327133225,
+      "grad_norm": 2.2158858186853,
+      "learning_rate": 9.969496351267278e-06,
+      "loss": 0.7176,
+      "step": 284
+    },
+    {
+      "epoch": 0.39758195768425947,
+      "grad_norm": 2.3115256506252693,
+      "learning_rate": 9.96859353562511e-06,
+      "loss": 0.7466,
+      "step": 285
+    },
+    {
+      "epoch": 0.3989769820971867,
+      "grad_norm": 2.1844162221084864,
+      "learning_rate": 9.967677595886542e-06,
+      "loss": 0.7166,
+      "step": 286
+    },
+    {
+      "epoch": 0.40037200651011395,
+      "grad_norm": 2.2684119567671126,
+      "learning_rate": 9.96674853447095e-06,
+      "loss": 0.743,
+      "step": 287
+    },
+    {
+      "epoch": 0.40176703092304117,
+      "grad_norm": 2.1345672593832608,
+      "learning_rate": 9.96580635383236e-06,
+      "loss": 0.6622,
+      "step": 288
+    },
+    {
+      "epoch": 0.4031620553359684,
+      "grad_norm": 1.920613760073464,
+      "learning_rate": 9.964851056459465e-06,
+      "loss": 0.6307,
+      "step": 289
+    },
+    {
+      "epoch": 0.4045570797488956,
+      "grad_norm": 2.2670166040543753,
+      "learning_rate": 9.963882644875594e-06,
+      "loss": 0.7169,
+      "step": 290
+    },
+    {
+      "epoch": 0.4059521041618228,
+      "grad_norm": 2.2233142865091384,
+      "learning_rate": 9.96290112163872e-06,
+      "loss": 0.7076,
+      "step": 291
+    },
+    {
+      "epoch": 0.4073471285747501,
+      "grad_norm": 2.3333906520435455,
+      "learning_rate": 9.961906489341452e-06,
+      "loss": 0.7351,
+      "step": 292
+    },
+    {
+      "epoch": 0.4087421529876773,
+      "grad_norm": 2.458417913325325,
+      "learning_rate": 9.960898750611019e-06,
+      "loss": 0.7711,
+      "step": 293
+    },
+    {
+      "epoch": 0.4101371774006045,
+      "grad_norm": 2.479058663768207,
+      "learning_rate": 9.959877908109274e-06,
+      "loss": 0.8339,
+      "step": 294
+    },
+    {
+      "epoch": 0.41153220181353173,
+      "grad_norm": 2.1552249804400128,
+      "learning_rate": 9.958843964532683e-06,
+      "loss": 0.681,
+      "step": 295
+    },
+    {
+      "epoch": 0.41292722622645894,
+      "grad_norm": 2.2069137654327537,
+      "learning_rate": 9.957796922612314e-06,
+      "loss": 0.7126,
+      "step": 296
+    },
+    {
+      "epoch": 0.4143222506393862,
+      "grad_norm": 2.536473553585731,
+      "learning_rate": 9.956736785113833e-06,
+      "loss": 0.771,
+      "step": 297
+    },
+    {
+      "epoch": 0.41571727505231343,
+      "grad_norm": 2.093714464350585,
+      "learning_rate": 9.955663554837503e-06,
+      "loss": 0.6509,
+      "step": 298
+    },
+    {
+      "epoch": 0.41711229946524064,
+      "grad_norm": 2.342384161868706,
+      "learning_rate": 9.954577234618162e-06,
+      "loss": 0.7163,
+      "step": 299
+    },
+    {
+      "epoch": 0.41850732387816786,
+      "grad_norm": 2.2305186397795613,
+      "learning_rate": 9.953477827325229e-06,
+      "loss": 0.7352,
+      "step": 300
+    },
+    {
+      "epoch": 0.4199023482910951,
+      "grad_norm": 2.1442071959324287,
+      "learning_rate": 9.952365335862693e-06,
+      "loss": 0.7667,
+      "step": 301
+    },
+    {
+      "epoch": 0.42129737270402234,
+      "grad_norm": 2.214992581838815,
+      "learning_rate": 9.951239763169097e-06,
+      "loss": 0.787,
+      "step": 302
+    },
+    {
+      "epoch": 0.42269239711694956,
+      "grad_norm": 2.2020890794485215,
+      "learning_rate": 9.950101112217543e-06,
+      "loss": 0.7207,
+      "step": 303
+    },
+    {
+      "epoch": 0.4240874215298768,
+      "grad_norm": 2.1586849736161016,
+      "learning_rate": 9.948949386015677e-06,
+      "loss": 0.6965,
+      "step": 304
+    },
+    {
+      "epoch": 0.425482445942804,
+      "grad_norm": 2.223387319666449,
+      "learning_rate": 9.947784587605678e-06,
+      "loss": 0.697,
+      "step": 305
+    },
+    {
+      "epoch": 0.4268774703557312,
+      "grad_norm": 2.4008683326131863,
+      "learning_rate": 9.946606720064257e-06,
+      "loss": 0.7564,
+      "step": 306
+    },
+    {
+      "epoch": 0.4282724947686585,
+      "grad_norm": 2.388205130653307,
+      "learning_rate": 9.945415786502649e-06,
+      "loss": 0.7356,
+      "step": 307
+    },
+    {
+      "epoch": 0.4296675191815857,
+      "grad_norm": 2.167142794736497,
+      "learning_rate": 9.944211790066597e-06,
+      "loss": 0.7234,
+      "step": 308
+    },
+    {
+      "epoch": 0.4310625435945129,
+      "grad_norm": 2.355995507133336,
+      "learning_rate": 9.94299473393635e-06,
+      "loss": 0.7773,
+      "step": 309
+    },
+    {
+      "epoch": 0.4324575680074401,
+      "grad_norm": 2.255660206292414,
+      "learning_rate": 9.941764621326655e-06,
+      "loss": 0.7175,
+      "step": 310
+    },
+    {
+      "epoch": 0.43385259242036733,
+      "grad_norm": 2.123386649465731,
+      "learning_rate": 9.94052145548674e-06,
+      "loss": 0.7542,
+      "step": 311
+    },
+    {
+      "epoch": 0.4352476168332946,
+      "grad_norm": 2.371212398389211,
+      "learning_rate": 9.939265239700321e-06,
+      "loss": 0.7845,
+      "step": 312
+    },
+    {
+      "epoch": 0.4366426412462218,
+      "grad_norm": 2.375596398430479,
+      "learning_rate": 9.93799597728558e-06,
+      "loss": 0.7494,
+      "step": 313
+    },
+    {
+      "epoch": 0.43803766565914903,
+      "grad_norm": 2.3770091900621826,
+      "learning_rate": 9.936713671595158e-06,
+      "loss": 0.7801,
+      "step": 314
+    },
+    {
+      "epoch": 0.43943269007207625,
+      "grad_norm": 2.2444664500496674,
+      "learning_rate": 9.935418326016153e-06,
+      "loss": 0.7055,
+      "step": 315
+    },
+    {
+      "epoch": 0.44082771448500346,
+      "grad_norm": 2.1949523416601804,
+      "learning_rate": 9.934109943970103e-06,
+      "loss": 0.7864,
+      "step": 316
+    },
+    {
+      "epoch": 0.44222273889793073,
+      "grad_norm": 2.234576229734612,
+      "learning_rate": 9.932788528912983e-06,
+      "loss": 0.7558,
+      "step": 317
+    },
+    {
+      "epoch": 0.44361776331085795,
+      "grad_norm": 2.248104750900924,
+      "learning_rate": 9.931454084335192e-06,
+      "loss": 0.7549,
+      "step": 318
+    },
+    {
+      "epoch": 0.44501278772378516,
+      "grad_norm": 2.036656477248027,
+      "learning_rate": 9.930106613761549e-06,
+      "loss": 0.7004,
+      "step": 319
+    },
+    {
+      "epoch": 0.4464078121367124,
+      "grad_norm": 2.1759704388763423,
+      "learning_rate": 9.928746120751275e-06,
+      "loss": 0.7335,
+      "step": 320
+    },
+    {
+      "epoch": 0.4478028365496396,
+      "grad_norm": 2.242033270499044,
+      "learning_rate": 9.927372608897992e-06,
+      "loss": 0.7691,
+      "step": 321
+    },
+    {
+      "epoch": 0.44919786096256686,
+      "grad_norm": 2.1579786835187305,
+      "learning_rate": 9.925986081829708e-06,
+      "loss": 0.6896,
+      "step": 322
+    },
+    {
+      "epoch": 0.4505928853754941,
+      "grad_norm": 2.383688707431281,
+      "learning_rate": 9.924586543208812e-06,
+      "loss": 0.7663,
+      "step": 323
+    },
+    {
+      "epoch": 0.4519879097884213,
+      "grad_norm": 2.2137281683181538,
+      "learning_rate": 9.923173996732058e-06,
+      "loss": 0.7102,
+      "step": 324
+    },
+    {
+      "epoch": 0.4533829342013485,
+      "grad_norm": 2.2671301590285178,
+      "learning_rate": 9.921748446130564e-06,
+      "loss": 0.7588,
+      "step": 325
+    },
+    {
+      "epoch": 0.4547779586142757,
+      "grad_norm": 2.0498917185694925,
+      "learning_rate": 9.920309895169793e-06,
+      "loss": 0.7111,
+      "step": 326
+    },
+    {
+      "epoch": 0.456172983027203,
+      "grad_norm": 2.2354446947614623,
+      "learning_rate": 9.91885834764955e-06,
+      "loss": 0.7137,
+      "step": 327
+    },
+    {
+      "epoch": 0.4575680074401302,
+      "grad_norm": 2.210652660226588,
+      "learning_rate": 9.917393807403965e-06,
+      "loss": 0.7683,
+      "step": 328
+    },
+    {
+      "epoch": 0.4589630318530574,
+      "grad_norm": 2.0554417534099243,
+      "learning_rate": 9.915916278301496e-06,
+      "loss": 0.7056,
+      "step": 329
+    },
+    {
+      "epoch": 0.46035805626598464,
+      "grad_norm": 2.2562712079069955,
+      "learning_rate": 9.9144257642449e-06,
+      "loss": 0.6896,
+      "step": 330
+    },
+    {
+      "epoch": 0.46175308067891185,
+      "grad_norm": 1.9777979766584077,
+      "learning_rate": 9.91292226917124e-06,
+      "loss": 0.6562,
+      "step": 331
+    },
+    {
+      "epoch": 0.4631481050918391,
+      "grad_norm": 2.2861549002868955,
+      "learning_rate": 9.91140579705186e-06,
+      "loss": 0.7529,
+      "step": 332
+    },
+    {
+      "epoch": 0.46454312950476634,
+      "grad_norm": 2.1618172913510865,
+      "learning_rate": 9.909876351892388e-06,
+      "loss": 0.7675,
+      "step": 333
+    },
+    {
+      "epoch": 0.46593815391769355,
+      "grad_norm": 1.9997881559101407,
+      "learning_rate": 9.908333937732718e-06,
+      "loss": 0.7209,
+      "step": 334
+    },
+    {
+      "epoch": 0.46733317833062077,
+      "grad_norm": 2.110744936433282,
+      "learning_rate": 9.906778558647e-06,
+      "loss": 0.71,
+      "step": 335
+    },
+    {
+      "epoch": 0.46872820274354804,
+      "grad_norm": 2.0830747263370304,
+      "learning_rate": 9.905210218743626e-06,
+      "loss": 0.7266,
+      "step": 336
+    },
+    {
+      "epoch": 0.47012322715647525,
+      "grad_norm": 2.0876262800113374,
+      "learning_rate": 9.903628922165227e-06,
+      "loss": 0.7023,
+      "step": 337
+    },
+    {
+      "epoch": 0.47151825156940247,
+      "grad_norm": 2.3098228154162217,
+      "learning_rate": 9.902034673088656e-06,
+      "loss": 0.7472,
+      "step": 338
+    },
+    {
+      "epoch": 0.4729132759823297,
+      "grad_norm": 2.1523425641211618,
+      "learning_rate": 9.90042747572498e-06,
+      "loss": 0.6827,
+      "step": 339
+    },
+    {
+      "epoch": 0.4743083003952569,
+      "grad_norm": 2.3481769610491954,
+      "learning_rate": 9.898807334319471e-06,
+      "loss": 0.7682,
+      "step": 340
+    },
+    {
+      "epoch": 0.47570332480818417,
+      "grad_norm": 2.1587172017907155,
+      "learning_rate": 9.897174253151583e-06,
+      "loss": 0.7217,
+      "step": 341
+    },
+    {
+      "epoch": 0.4770983492211114,
+      "grad_norm": 2.2768842079801836,
+      "learning_rate": 9.895528236534957e-06,
+      "loss": 0.7414,
+      "step": 342
+    },
+    {
+      "epoch": 0.4784933736340386,
+      "grad_norm": 2.267882603603539,
+      "learning_rate": 9.893869288817397e-06,
+      "loss": 0.7902,
+      "step": 343
+    },
+    {
+      "epoch": 0.4798883980469658,
+      "grad_norm": 2.003426082657925,
+      "learning_rate": 9.89219741438087e-06,
+      "loss": 0.7133,
+      "step": 344
+    },
+    {
+      "epoch": 0.48128342245989303,
+      "grad_norm": 2.2017001776217704,
+      "learning_rate": 9.890512617641474e-06,
+      "loss": 0.7315,
+      "step": 345
+    },
+    {
+      "epoch": 0.4826784468728203,
+      "grad_norm": 2.2874305641494663,
+      "learning_rate": 9.888814903049458e-06,
+      "loss": 0.7551,
+      "step": 346
+    },
+    {
+      "epoch": 0.4840734712857475,
+      "grad_norm": 2.2146672202568896,
+      "learning_rate": 9.88710427508918e-06,
+      "loss": 0.7701,
+      "step": 347
+    },
+    {
+      "epoch": 0.48546849569867473,
+      "grad_norm": 2.2368298170629473,
+      "learning_rate": 9.885380738279111e-06,
+      "loss": 0.6928,
+      "step": 348
+    },
+    {
+      "epoch": 0.48686352011160194,
+      "grad_norm": 2.1128133334522565,
+      "learning_rate": 9.883644297171821e-06,
+      "loss": 0.7264,
+      "step": 349
+    },
+    {
+      "epoch": 0.48825854452452916,
+      "grad_norm": 2.0878462954407517,
+      "learning_rate": 9.881894956353963e-06,
+      "loss": 0.7533,
+      "step": 350
+    },
+    {
+      "epoch": 0.48965356893745643,
+      "grad_norm": 2.1787208125747384,
+      "learning_rate": 9.880132720446265e-06,
+      "loss": 0.7176,
+      "step": 351
+    },
+    {
+      "epoch": 0.49104859335038364,
+      "grad_norm": 2.2222401905237126,
+      "learning_rate": 9.878357594103516e-06,
+      "loss": 0.7732,
+      "step": 352
+    },
+    {
+      "epoch": 0.49244361776331086,
+      "grad_norm": 2.164885730565751,
+      "learning_rate": 9.876569582014554e-06,
+      "loss": 0.7192,
+      "step": 353
+    },
+    {
+      "epoch": 0.4938386421762381,
+      "grad_norm": 2.176298873741657,
+      "learning_rate": 9.874768688902252e-06,
+      "loss": 0.7583,
+      "step": 354
+    },
+    {
+      "epoch": 0.4952336665891653,
+      "grad_norm": 1.9635642084372211,
+      "learning_rate": 9.87295491952351e-06,
+      "loss": 0.7543,
+      "step": 355
+    },
+    {
+      "epoch": 0.49662869100209256,
+      "grad_norm": 2.2033913590623633,
+      "learning_rate": 9.871128278669238e-06,
+      "loss": 0.7254,
+      "step": 356
+    },
+    {
+      "epoch": 0.4980237154150198,
+      "grad_norm": 2.1462134327569014,
+      "learning_rate": 9.869288771164344e-06,
+      "loss": 0.7201,
+      "step": 357
+    },
+    {
+      "epoch": 0.499418739827947,
+      "grad_norm": 2.2380749870735177,
+      "learning_rate": 9.867436401867723e-06,
+      "loss": 0.7316,
+      "step": 358
+    },
+    {
+      "epoch": 0.5008137642408742,
+      "grad_norm": 2.3135966107377195,
+      "learning_rate": 9.865571175672245e-06,
+      "loss": 0.7659,
+      "step": 359
+    },
+    {
+      "epoch": 0.5022087886538015,
+      "grad_norm": 2.1533066959256155,
+      "learning_rate": 9.863693097504733e-06,
+      "loss": 0.7098,
+      "step": 360
+    },
+    {
+      "epoch": 0.5036038130667286,
+      "grad_norm": 2.1171023788743644,
+      "learning_rate": 9.86180217232597e-06,
+      "loss": 0.7726,
+      "step": 361
+    },
+    {
+      "epoch": 0.5049988374796559,
+      "grad_norm": 2.4377521628329313,
+      "learning_rate": 9.859898405130661e-06,
+      "loss": 0.7916,
+      "step": 362
+    },
+    {
+      "epoch": 0.5063938618925832,
+      "grad_norm": 2.2854519749010844,
+      "learning_rate": 9.85798180094744e-06,
+      "loss": 0.7118,
+      "step": 363
+    },
+    {
+      "epoch": 0.5077888863055103,
+      "grad_norm": 2.244618640647697,
+      "learning_rate": 9.856052364838846e-06,
+      "loss": 0.8073,
+      "step": 364
+    },
+    {
+      "epoch": 0.5091839107184376,
+      "grad_norm": 2.177378244026803,
+      "learning_rate": 9.854110101901308e-06,
+      "loss": 0.6749,
+      "step": 365
+    },
+    {
+      "epoch": 0.5105789351313648,
+      "grad_norm": 1.9603446721831779,
+      "learning_rate": 9.852155017265146e-06,
+      "loss": 0.6283,
+      "step": 366
+    },
+    {
+      "epoch": 0.511973959544292,
+      "grad_norm": 2.220937568124573,
+      "learning_rate": 9.850187116094538e-06,
+      "loss": 0.721,
+      "step": 367
+    },
+    {
+      "epoch": 0.5133689839572193,
+      "grad_norm": 2.1071011951034637,
+      "learning_rate": 9.848206403587521e-06,
+      "loss": 0.7634,
+      "step": 368
+    },
+    {
+      "epoch": 0.5147640083701465,
+      "grad_norm": 2.294947828525865,
+      "learning_rate": 9.84621288497597e-06,
+      "loss": 0.7562,
+      "step": 369
+    },
+    {
+      "epoch": 0.5161590327830737,
+      "grad_norm": 2.122179199455438,
+      "learning_rate": 9.844206565525585e-06,
+      "loss": 0.7495,
+      "step": 370
+    },
+    {
+      "epoch": 0.5175540571960009,
+      "grad_norm": 2.15699679179535,
+      "learning_rate": 9.842187450535881e-06,
+      "loss": 0.731,
+      "step": 371
+    },
+    {
+      "epoch": 0.5189490816089282,
+      "grad_norm": 2.43922784941149,
+      "learning_rate": 9.840155545340169e-06,
+      "loss": 0.7457,
+      "step": 372
+    },
+    {
+      "epoch": 0.5203441060218554,
+      "grad_norm": 2.251971046707549,
+      "learning_rate": 9.838110855305548e-06,
+      "loss": 0.7103,
+      "step": 373
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 2.0408680111240614,
+      "learning_rate": 9.836053385832881e-06,
+      "loss": 0.6503,
+      "step": 374
+    },
+    {
+      "epoch": 0.5231341548477099,
+      "grad_norm": 2.1603103563459105,
+      "learning_rate": 9.833983142356792e-06,
+      "loss": 0.7281,
+      "step": 375
+    },
+    {
+      "epoch": 0.524529179260637,
+      "grad_norm": 2.2773618533718563,
+      "learning_rate": 9.831900130345645e-06,
+      "loss": 0.7009,
+      "step": 376
+    },
+    {
+      "epoch": 0.5259242036735643,
+      "grad_norm": 2.099468862050019,
+      "learning_rate": 9.829804355301527e-06,
+      "loss": 0.6762,
+      "step": 377
+    },
+    {
+      "epoch": 0.5273192280864916,
+      "grad_norm": 2.064470755868956,
+      "learning_rate": 9.827695822760245e-06,
+      "loss": 0.6789,
+      "step": 378
+    },
+    {
+      "epoch": 0.5287142524994187,
+      "grad_norm": 2.2842388942976326,
+      "learning_rate": 9.825574538291293e-06,
+      "loss": 0.7384,
+      "step": 379
+    },
+    {
+      "epoch": 0.530109276912346,
+      "grad_norm": 2.0597543670014806,
+      "learning_rate": 9.823440507497863e-06,
+      "loss": 0.6774,
+      "step": 380
+    },
+    {
+      "epoch": 0.5315043013252732,
+      "grad_norm": 2.2151766652128573,
+      "learning_rate": 9.821293736016802e-06,
+      "loss": 0.6933,
+      "step": 381
+    },
+    {
+      "epoch": 0.5328993257382004,
+      "grad_norm": 2.281926946410573,
+      "learning_rate": 9.819134229518617e-06,
+      "loss": 0.7449,
+      "step": 382
+    },
+    {
+      "epoch": 0.5342943501511277,
+      "grad_norm": 2.1443995307296464,
+      "learning_rate": 9.81696199370745e-06,
+      "loss": 0.7213,
+      "step": 383
+    },
+    {
+      "epoch": 0.5356893745640549,
+      "grad_norm": 2.11871178228737,
+      "learning_rate": 9.814777034321069e-06,
+      "loss": 0.7453,
+      "step": 384
+    },
+    {
+      "epoch": 0.5370843989769821,
+      "grad_norm": 2.218758315080764,
+      "learning_rate": 9.812579357130848e-06,
+      "loss": 0.7702,
+      "step": 385
+    },
+    {
+      "epoch": 0.5384794233899093,
+      "grad_norm": 2.0131323463999444,
+      "learning_rate": 9.810368967941757e-06,
+      "loss": 0.7129,
+      "step": 386
+    },
+    {
+      "epoch": 0.5398744478028366,
+      "grad_norm": 2.18529376681278,
+      "learning_rate": 9.808145872592341e-06,
+      "loss": 0.7192,
+      "step": 387
+    },
+    {
+      "epoch": 0.5412694722157638,
+      "grad_norm": 2.1408294189371113,
+      "learning_rate": 9.80591007695471e-06,
+      "loss": 0.689,
+      "step": 388
+    },
+    {
+      "epoch": 0.542664496628691,
+      "grad_norm": 2.1875152665270075,
+      "learning_rate": 9.803661586934514e-06,
+      "loss": 0.7041,
+      "step": 389
+    },
+    {
+      "epoch": 0.5440595210416183,
+      "grad_norm": 2.2494421729469587,
+      "learning_rate": 9.801400408470943e-06,
+      "loss": 0.7558,
+      "step": 390
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 2.1718406110229314,
+      "learning_rate": 9.799126547536695e-06,
+      "loss": 0.7542,
+      "step": 391
+    },
+    {
+      "epoch": 0.5468495698674727,
+      "grad_norm": 2.1100570891152524,
+      "learning_rate": 9.796840010137972e-06,
+      "loss": 0.7036,
+      "step": 392
+    },
+    {
+      "epoch": 0.5482445942804,
+      "grad_norm": 2.115132254738484,
+      "learning_rate": 9.79454080231446e-06,
+      "loss": 0.8377,
+      "step": 393
+    },
+    {
+      "epoch": 0.5496396186933271,
+      "grad_norm": 2.014824175362774,
+      "learning_rate": 9.79222893013931e-06,
+      "loss": 0.6456,
+      "step": 394
+    },
+    {
+      "epoch": 0.5510346431062544,
+      "grad_norm": 2.227921249849429,
+      "learning_rate": 9.789904399719124e-06,
+      "loss": 0.7739,
+      "step": 395
+    },
+    {
+      "epoch": 0.5524296675191815,
+      "grad_norm": 2.112220428527919,
+      "learning_rate": 9.787567217193944e-06,
+      "loss": 0.6688,
+      "step": 396
+    },
+    {
+      "epoch": 0.5538246919321088,
+      "grad_norm": 2.1494591416793636,
+      "learning_rate": 9.785217388737232e-06,
+      "loss": 0.7427,
+      "step": 397
+    },
+    {
+      "epoch": 0.5552197163450361,
+      "grad_norm": 2.0642390183703796,
+      "learning_rate": 9.782854920555844e-06,
+      "loss": 0.6962,
+      "step": 398
+    },
+    {
+      "epoch": 0.5566147407579632,
+      "grad_norm": 1.9583040952188322,
+      "learning_rate": 9.780479818890032e-06,
+      "loss": 0.7644,
+      "step": 399
+    },
+    {
+      "epoch": 0.5580097651708905,
+      "grad_norm": 2.1534478133290844,
+      "learning_rate": 9.778092090013416e-06,
+      "loss": 0.7137,
+      "step": 400
+    },
+    {
+      "epoch": 0.5594047895838177,
+      "grad_norm": 1.914276540514964,
+      "learning_rate": 9.775691740232966e-06,
+      "loss": 0.6742,
+      "step": 401
+    },
+    {
+      "epoch": 0.560799813996745,
+      "grad_norm": 2.016867139681356,
+      "learning_rate": 9.773278775888995e-06,
+      "loss": 0.7475,
+      "step": 402
+    },
+    {
+      "epoch": 0.5621948384096722,
+      "grad_norm": 2.1849766754698416,
+      "learning_rate": 9.77085320335513e-06,
+      "loss": 0.7055,
+      "step": 403
+    },
+    {
+      "epoch": 0.5635898628225994,
+      "grad_norm": 2.223991761051566,
+      "learning_rate": 9.768415029038304e-06,
+      "loss": 0.7101,
+      "step": 404
+    },
+    {
+      "epoch": 0.5649848872355266,
+      "grad_norm": 2.0011991831554954,
+      "learning_rate": 9.76596425937874e-06,
+      "loss": 0.6877,
+      "step": 405
+    },
+    {
+      "epoch": 0.5663799116484538,
+      "grad_norm": 2.0542369776804166,
+      "learning_rate": 9.763500900849926e-06,
+      "loss": 0.6662,
+      "step": 406
+    },
+    {
+      "epoch": 0.5677749360613811,
+      "grad_norm": 2.3730906372976546,
+      "learning_rate": 9.761024959958605e-06,
+      "loss": 0.7489,
+      "step": 407
+    },
+    {
+      "epoch": 0.5691699604743083,
+      "grad_norm": 2.3196262507878775,
+      "learning_rate": 9.75853644324475e-06,
+      "loss": 0.7565,
+      "step": 408
+    },
+    {
+      "epoch": 0.5705649848872355,
+      "grad_norm": 2.341441695587702,
+      "learning_rate": 9.756035357281559e-06,
+      "loss": 0.7399,
+      "step": 409
+    },
+    {
+      "epoch": 0.5719600093001628,
+      "grad_norm": 2.090797895413644,
+      "learning_rate": 9.753521708675426e-06,
+      "loss": 0.7035,
+      "step": 410
+    },
+    {
+      "epoch": 0.5733550337130899,
+      "grad_norm": 2.137767654681621,
+      "learning_rate": 9.75099550406593e-06,
+      "loss": 0.7104,
+      "step": 411
+    },
+    {
+      "epoch": 0.5747500581260172,
+      "grad_norm": 2.0270348604623942,
+      "learning_rate": 9.748456750125817e-06,
+      "loss": 0.755,
+      "step": 412
+    },
+    {
+      "epoch": 0.5761450825389445,
+      "grad_norm": 2.15807439031817,
+      "learning_rate": 9.745905453560976e-06,
+      "loss": 0.7055,
+      "step": 413
+    },
+    {
+      "epoch": 0.5775401069518716,
+      "grad_norm": 1.941910111218879,
+      "learning_rate": 9.74334162111043e-06,
+      "loss": 0.7452,
+      "step": 414
+    },
+    {
+      "epoch": 0.5789351313647989,
+      "grad_norm": 2.0347312764060037,
+      "learning_rate": 9.740765259546312e-06,
+      "loss": 0.7233,
+      "step": 415
+    },
+    {
+      "epoch": 0.5803301557777261,
+      "grad_norm": 2.250098777709644,
+      "learning_rate": 9.738176375673856e-06,
+      "loss": 0.7386,
+      "step": 416
+    },
+    {
+      "epoch": 0.5817251801906533,
+      "grad_norm": 2.152752510392946,
+      "learning_rate": 9.735574976331362e-06,
+      "loss": 0.6596,
+      "step": 417
+    },
+    {
+      "epoch": 0.5831202046035806,
+      "grad_norm": 2.1555034248115446,
+      "learning_rate": 9.732961068390199e-06,
+      "loss": 0.7568,
+      "step": 418
+    },
+    {
+      "epoch": 0.5845152290165078,
+      "grad_norm": 2.3258624129809107,
+      "learning_rate": 9.730334658754767e-06,
+      "loss": 0.8289,
+      "step": 419
+    },
+    {
+      "epoch": 0.585910253429435,
+      "grad_norm": 2.0681332356052233,
+      "learning_rate": 9.727695754362498e-06,
+      "loss": 0.7454,
+      "step": 420
+    },
+    {
+      "epoch": 0.5873052778423622,
+      "grad_norm": 2.1429328741531877,
+      "learning_rate": 9.725044362183817e-06,
+      "loss": 0.7189,
+      "step": 421
+    },
+    {
+      "epoch": 0.5887003022552895,
+      "grad_norm": 2.035384052327714,
+      "learning_rate": 9.722380489222145e-06,
+      "loss": 0.6795,
+      "step": 422
+    },
+    {
+      "epoch": 0.5900953266682167,
+      "grad_norm": 1.9628409776914946,
+      "learning_rate": 9.71970414251386e-06,
+      "loss": 0.6584,
+      "step": 423
+    },
+    {
+      "epoch": 0.5914903510811439,
+      "grad_norm": 2.119850722980248,
+      "learning_rate": 9.717015329128294e-06,
+      "loss": 0.7466,
+      "step": 424
+    },
+    {
+      "epoch": 0.5928853754940712,
+      "grad_norm": 2.127385339997239,
+      "learning_rate": 9.714314056167711e-06,
+      "loss": 0.711,
+      "step": 425
+    },
+    {
+      "epoch": 0.5942803999069983,
+      "grad_norm": 1.8294466072056457,
+      "learning_rate": 9.711600330767278e-06,
+      "loss": 0.6382,
+      "step": 426
+    },
+    {
+      "epoch": 0.5956754243199256,
+      "grad_norm": 2.1442575532302737,
+      "learning_rate": 9.708874160095061e-06,
+      "loss": 0.7207,
+      "step": 427
+    },
+    {
+      "epoch": 0.5970704487328529,
+      "grad_norm": 2.0715902524872436,
+      "learning_rate": 9.706135551351996e-06,
+      "loss": 0.721,
+      "step": 428
+    },
+    {
+      "epoch": 0.59846547314578,
+      "grad_norm": 2.017273861098596,
+      "learning_rate": 9.703384511771874e-06,
+      "loss": 0.6685,
+      "step": 429
+    },
+    {
+      "epoch": 0.5998604975587073,
+      "grad_norm": 2.0128870162232326,
+      "learning_rate": 9.700621048621322e-06,
+      "loss": 0.6621,
+      "step": 430
+    },
+    {
+      "epoch": 0.6012555219716345,
+      "grad_norm": 1.9324919806562177,
+      "learning_rate": 9.697845169199775e-06,
+      "loss": 0.6495,
+      "step": 431
+    },
+    {
+      "epoch": 0.6026505463845617,
+      "grad_norm": 2.2593974726575023,
+      "learning_rate": 9.69505688083948e-06,
+      "loss": 0.7673,
+      "step": 432
+    },
+    {
+      "epoch": 0.604045570797489,
+      "grad_norm": 1.9629658473435176,
+      "learning_rate": 9.692256190905444e-06,
+      "loss": 0.6567,
+      "step": 433
+    },
+    {
+      "epoch": 0.6054405952104162,
+      "grad_norm": 2.1948743143431373,
+      "learning_rate": 9.689443106795442e-06,
+      "loss": 0.7275,
+      "step": 434
+    },
+    {
+      "epoch": 0.6068356196233434,
+      "grad_norm": 2.173895338644472,
+      "learning_rate": 9.686617635939988e-06,
+      "loss": 0.6843,
+      "step": 435
+    },
+    {
+      "epoch": 0.6082306440362706,
+      "grad_norm": 2.183862902137113,
+      "learning_rate": 9.683779785802306e-06,
+      "loss": 0.7488,
+      "step": 436
+    },
+    {
+      "epoch": 0.6096256684491979,
+      "grad_norm": 2.247224096170596,
+      "learning_rate": 9.680929563878327e-06,
+      "loss": 0.7009,
+      "step": 437
+    },
+    {
+      "epoch": 0.6110206928621251,
+      "grad_norm": 2.2198797071377308,
+      "learning_rate": 9.678066977696656e-06,
+      "loss": 0.7576,
+      "step": 438
+    },
+    {
+      "epoch": 0.6124157172750523,
+      "grad_norm": 2.099155911257389,
+      "learning_rate": 9.675192034818561e-06,
+      "loss": 0.7342,
+      "step": 439
+    },
+    {
+      "epoch": 0.6138107416879796,
+      "grad_norm": 2.1826988188780456,
+      "learning_rate": 9.672304742837945e-06,
+      "loss": 0.7992,
+      "step": 440
+    },
+    {
+      "epoch": 0.6152057661009067,
+      "grad_norm": 2.11108770679613,
+      "learning_rate": 9.669405109381335e-06,
+      "loss": 0.7284,
+      "step": 441
+    },
+    {
+      "epoch": 0.616600790513834,
+      "grad_norm": 2.1204342794359974,
+      "learning_rate": 9.66649314210785e-06,
+      "loss": 0.7281,
+      "step": 442
+    },
+    {
+      "epoch": 0.6179958149267613,
+      "grad_norm": 1.9367375234614417,
+      "learning_rate": 9.663568848709194e-06,
+      "loss": 0.7283,
+      "step": 443
+    },
+    {
+      "epoch": 0.6193908393396884,
+      "grad_norm": 2.1130289482285574,
+      "learning_rate": 9.660632236909628e-06,
+      "loss": 0.7242,
+      "step": 444
+    },
+    {
+      "epoch": 0.6207858637526157,
+      "grad_norm": 1.9102389109395275,
+      "learning_rate": 9.657683314465948e-06,
+      "loss": 0.6874,
+      "step": 445
+    },
+    {
+      "epoch": 0.6221808881655428,
+      "grad_norm": 2.1435520707988314,
+      "learning_rate": 9.65472208916747e-06,
+      "loss": 0.7858,
+      "step": 446
+    },
+    {
+      "epoch": 0.6235759125784701,
+      "grad_norm": 2.1624999372909106,
+      "learning_rate": 9.651748568836007e-06,
+      "loss": 0.7646,
+      "step": 447
+    },
+    {
+      "epoch": 0.6249709369913974,
+      "grad_norm": 2.065236722576021,
+      "learning_rate": 9.648762761325847e-06,
+      "loss": 0.6898,
+      "step": 448
+    },
+    {
+      "epoch": 0.6263659614043245,
+      "grad_norm": 2.0485232389558616,
+      "learning_rate": 9.645764674523732e-06,
+      "loss": 0.7839,
+      "step": 449
+    },
+    {
+      "epoch": 0.6277609858172518,
+      "grad_norm": 2.0655621784312235,
+      "learning_rate": 9.642754316348846e-06,
+      "loss": 0.7075,
+      "step": 450
+    },
+    {
+      "epoch": 0.629156010230179,
+      "grad_norm": 2.0060672739124152,
+      "learning_rate": 9.639731694752776e-06,
+      "loss": 0.7489,
+      "step": 451
+    },
+    {
+      "epoch": 0.6305510346431062,
+      "grad_norm": 2.208261364040988,
+      "learning_rate": 9.636696817719511e-06,
+      "loss": 0.7535,
+      "step": 452
+    },
+    {
+      "epoch": 0.6319460590560335,
+      "grad_norm": 2.2089796542819866,
+      "learning_rate": 9.633649693265406e-06,
+      "loss": 0.7482,
+      "step": 453
+    },
+    {
+      "epoch": 0.6333410834689607,
+      "grad_norm": 2.2384194618104125,
+      "learning_rate": 9.630590329439169e-06,
+      "loss": 0.7792,
+      "step": 454
+    },
+    {
+      "epoch": 0.634736107881888,
+      "grad_norm": 2.2060852158547264,
+      "learning_rate": 9.627518734321837e-06,
+      "loss": 0.7525,
+      "step": 455
+    },
+    {
+      "epoch": 0.6361311322948151,
+      "grad_norm": 2.149847852979536,
+      "learning_rate": 9.624434916026752e-06,
+      "loss": 0.7439,
+      "step": 456
+    },
+    {
+      "epoch": 0.6375261567077424,
+      "grad_norm": 2.062138416788092,
+      "learning_rate": 9.621338882699547e-06,
+      "loss": 0.7396,
+      "step": 457
+    },
+    {
+      "epoch": 0.6389211811206696,
+      "grad_norm": 2.093876836162674,
+      "learning_rate": 9.618230642518117e-06,
+      "loss": 0.6971,
+      "step": 458
+    },
+    {
+      "epoch": 0.6403162055335968,
+      "grad_norm": 2.052608597617671,
+      "learning_rate": 9.615110203692602e-06,
+      "loss": 0.7709,
+      "step": 459
+    },
+    {
+      "epoch": 0.6417112299465241,
+      "grad_norm": 2.1332375538387676,
+      "learning_rate": 9.61197757446536e-06,
+      "loss": 0.7243,
+      "step": 460
+    },
+    {
+      "epoch": 0.6431062543594513,
+      "grad_norm": 1.9725195227245578,
+      "learning_rate": 9.608832763110955e-06,
+      "loss": 0.6988,
+      "step": 461
+    },
+    {
+      "epoch": 0.6445012787723785,
+      "grad_norm": 2.1827535475246997,
+      "learning_rate": 9.605675777936123e-06,
+      "loss": 0.7206,
+      "step": 462
+    },
+    {
+      "epoch": 0.6458963031853058,
+      "grad_norm": 2.0617095984448017,
+      "learning_rate": 9.60250662727976e-06,
+      "loss": 0.7297,
+      "step": 463
+    },
+    {
+      "epoch": 0.6472913275982329,
+      "grad_norm": 2.104755014090587,
+      "learning_rate": 9.599325319512893e-06,
+      "loss": 0.7303,
+      "step": 464
+    },
+    {
+      "epoch": 0.6486863520111602,
+      "grad_norm": 2.0095406432497924,
+      "learning_rate": 9.596131863038664e-06,
+      "loss": 0.7136,
+      "step": 465
+    },
+    {
+      "epoch": 0.6500813764240875,
+      "grad_norm": 2.0521696580726267,
+      "learning_rate": 9.592926266292305e-06,
+      "loss": 0.7085,
+      "step": 466
+    },
+    {
+      "epoch": 0.6514764008370146,
+      "grad_norm": 2.216895328470358,
+      "learning_rate": 9.589708537741109e-06,
+      "loss": 0.7338,
+      "step": 467
+    },
+    {
+      "epoch": 0.6528714252499419,
+      "grad_norm": 2.0241813457119293,
+      "learning_rate": 9.586478685884424e-06,
+      "loss": 0.6855,
+      "step": 468
+    },
+    {
+      "epoch": 0.6542664496628691,
+      "grad_norm": 2.0778866351895293,
+      "learning_rate": 9.583236719253611e-06,
+      "loss": 0.7044,
+      "step": 469
+    },
+    {
+      "epoch": 0.6556614740757963,
+      "grad_norm": 2.1863136845278186,
+      "learning_rate": 9.579982646412039e-06,
+      "loss": 0.719,
+      "step": 470
+    },
+    {
+      "epoch": 0.6570564984887236,
+      "grad_norm": 1.9539661965522732,
+      "learning_rate": 9.576716475955048e-06,
+      "loss": 0.6416,
+      "step": 471
+    },
+    {
+      "epoch": 0.6584515229016508,
+      "grad_norm": 2.0452081518877603,
+      "learning_rate": 9.573438216509937e-06,
+      "loss": 0.6972,
+      "step": 472
+    },
+    {
+      "epoch": 0.659846547314578,
+      "grad_norm": 1.995830987986953,
+      "learning_rate": 9.570147876735937e-06,
+      "loss": 0.7378,
+      "step": 473
+    },
+    {
+      "epoch": 0.6612415717275052,
+      "grad_norm": 1.8915266992396211,
+      "learning_rate": 9.566845465324185e-06,
+      "loss": 0.6788,
+      "step": 474
+    },
+    {
+      "epoch": 0.6626365961404325,
+      "grad_norm": 2.003656922577594,
+      "learning_rate": 9.563530990997707e-06,
+      "loss": 0.7674,
+      "step": 475
+    },
+    {
+      "epoch": 0.6640316205533597,
+      "grad_norm": 2.088291822709115,
+      "learning_rate": 9.560204462511392e-06,
+      "loss": 0.7355,
+      "step": 476
+    },
+    {
+      "epoch": 0.6654266449662869,
+      "grad_norm": 1.984879297358519,
+      "learning_rate": 9.556865888651965e-06,
+      "loss": 0.7545,
+      "step": 477
+    },
+    {
+      "epoch": 0.6668216693792142,
+      "grad_norm": 1.9946933617029812,
+      "learning_rate": 9.553515278237975e-06,
+      "loss": 0.6782,
+      "step": 478
+    },
+    {
+      "epoch": 0.6682166937921413,
+      "grad_norm": 2.1899693736873607,
+      "learning_rate": 9.550152640119757e-06,
+      "loss": 0.7334,
+      "step": 479
+    },
+    {
+      "epoch": 0.6696117182050686,
+      "grad_norm": 2.1437476826926374,
+      "learning_rate": 9.546777983179421e-06,
+      "loss": 0.7577,
+      "step": 480
+    },
+    {
+      "epoch": 0.6710067426179959,
+      "grad_norm": 1.9104178259795124,
+      "learning_rate": 9.543391316330822e-06,
+      "loss": 0.636,
+      "step": 481
+    },
+    {
+      "epoch": 0.672401767030923,
+      "grad_norm": 2.0148752084647525,
+      "learning_rate": 9.539992648519538e-06,
+      "loss": 0.7451,
+      "step": 482
+    },
+    {
+      "epoch": 0.6737967914438503,
+      "grad_norm": 2.2385594955618826,
+      "learning_rate": 9.536581988722848e-06,
+      "loss": 0.8286,
+      "step": 483
+    },
+    {
+      "epoch": 0.6751918158567775,
+      "grad_norm": 2.1601886300481414,
+      "learning_rate": 9.533159345949704e-06,
+      "loss": 0.762,
+      "step": 484
+    },
+    {
+      "epoch": 0.6765868402697047,
+      "grad_norm": 1.9477636560481595,
+      "learning_rate": 9.529724729240712e-06,
+      "loss": 0.6817,
+      "step": 485
+    },
+    {
+      "epoch": 0.677981864682632,
+      "grad_norm": 2.0753277030485857,
+      "learning_rate": 9.526278147668104e-06,
+      "loss": 0.6958,
+      "step": 486
+    },
+    {
+      "epoch": 0.6793768890955592,
+      "grad_norm": 2.071829326274995,
+      "learning_rate": 9.522819610335721e-06,
+      "loss": 0.7237,
+      "step": 487
+    },
+    {
+      "epoch": 0.6807719135084864,
+      "grad_norm": 1.8341834360542066,
+      "learning_rate": 9.519349126378975e-06,
+      "loss": 0.6825,
+      "step": 488
+    },
+    {
+      "epoch": 0.6821669379214136,
+      "grad_norm": 2.2062085487991827,
+      "learning_rate": 9.515866704964846e-06,
+      "loss": 0.7498,
+      "step": 489
+    },
+    {
+      "epoch": 0.6835619623343409,
+      "grad_norm": 2.080856174420327,
+      "learning_rate": 9.512372355291838e-06,
+      "loss": 0.7133,
+      "step": 490
+    },
+    {
+      "epoch": 0.6849569867472681,
+      "grad_norm": 1.8162007262738658,
+      "learning_rate": 9.50886608658996e-06,
+      "loss": 0.6417,
+      "step": 491
+    },
+    {
+      "epoch": 0.6863520111601953,
+      "grad_norm": 2.0373986097463996,
+      "learning_rate": 9.505347908120712e-06,
+      "loss": 0.7397,
+      "step": 492
+    },
+    {
+      "epoch": 0.6877470355731226,
+      "grad_norm": 2.0546808459729324,
+      "learning_rate": 9.501817829177046e-06,
+      "loss": 0.7141,
+      "step": 493
+    },
+    {
+      "epoch": 0.6891420599860497,
+      "grad_norm": 2.133311384676604,
+      "learning_rate": 9.498275859083353e-06,
+      "loss": 0.7297,
+      "step": 494
+    },
+    {
+      "epoch": 0.690537084398977,
+      "grad_norm": 2.045445358883216,
+      "learning_rate": 9.494722007195427e-06,
+      "loss": 0.7056,
+      "step": 495
+    },
+    {
+      "epoch": 0.6919321088119043,
+      "grad_norm": 2.032312961897094,
+      "learning_rate": 9.491156282900454e-06,
+      "loss": 0.7126,
+      "step": 496
+    },
+    {
+      "epoch": 0.6933271332248314,
+      "grad_norm": 2.0855627156665597,
+      "learning_rate": 9.487578695616974e-06,
+      "loss": 0.7824,
+      "step": 497
+    },
+    {
+      "epoch": 0.6947221576377587,
+      "grad_norm": 2.0792422212264734,
+      "learning_rate": 9.483989254794865e-06,
+      "loss": 0.7313,
+      "step": 498
+    },
+    {
+      "epoch": 0.6961171820506858,
+      "grad_norm": 1.9424464522170488,
+      "learning_rate": 9.480387969915318e-06,
+      "loss": 0.6236,
+      "step": 499
+    },
+    {
+      "epoch": 0.6975122064636131,
+      "grad_norm": 2.101297942682129,
+      "learning_rate": 9.476774850490803e-06,
+      "loss": 0.7401,
+      "step": 500
+    },
+    {
+      "epoch": 0.6989072308765404,
+      "grad_norm": 2.045369779079581,
+      "learning_rate": 9.47314990606505e-06,
+      "loss": 0.72,
+      "step": 501
+    },
+    {
+      "epoch": 0.7003022552894675,
+      "grad_norm": 2.070518186746418,
+      "learning_rate": 9.46951314621303e-06,
+      "loss": 0.6995,
+      "step": 502
+    },
+    {
+      "epoch": 0.7016972797023948,
+      "grad_norm": 2.2999734032614234,
+      "learning_rate": 9.465864580540917e-06,
+      "loss": 0.8061,
+      "step": 503
+    },
+    {
+      "epoch": 0.703092304115322,
+      "grad_norm": 2.1443964408217626,
+      "learning_rate": 9.462204218686075e-06,
+      "loss": 0.7044,
+      "step": 504
+    },
+    {
+      "epoch": 0.7044873285282492,
+      "grad_norm": 2.096007475498279,
+      "learning_rate": 9.458532070317021e-06,
+      "loss": 0.7273,
+      "step": 505
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 2.0041204965854162,
+      "learning_rate": 9.454848145133406e-06,
+      "loss": 0.7055,
+      "step": 506
+    },
+    {
+      "epoch": 0.7072773773541037,
+      "grad_norm": 2.2491552759682665,
+      "learning_rate": 9.451152452865991e-06,
+      "loss": 0.7134,
+      "step": 507
+    },
+    {
+      "epoch": 0.708672401767031,
+      "grad_norm": 1.9405986617138786,
+      "learning_rate": 9.447445003276618e-06,
+      "loss": 0.6579,
+      "step": 508
+    },
+    {
+      "epoch": 0.7100674261799581,
+      "grad_norm": 2.195166493132121,
+      "learning_rate": 9.443725806158182e-06,
+      "loss": 0.7339,
+      "step": 509
+    },
+    {
+      "epoch": 0.7114624505928854,
+      "grad_norm": 1.9578915579027396,
+      "learning_rate": 9.439994871334614e-06,
+      "loss": 0.7155,
+      "step": 510
+    },
+    {
+      "epoch": 0.7128574750058126,
+      "grad_norm": 2.190807561711358,
+      "learning_rate": 9.43625220866084e-06,
+      "loss": 0.7126,
+      "step": 511
+    },
+    {
+      "epoch": 0.7142524994187398,
+      "grad_norm": 2.098986891994908,
+      "learning_rate": 9.432497828022775e-06,
+      "loss": 0.7075,
+      "step": 512
+    },
+    {
+      "epoch": 0.7156475238316671,
+      "grad_norm": 2.1368418565133664,
+      "learning_rate": 9.428731739337277e-06,
+      "loss": 0.8113,
+      "step": 513
+    },
+    {
+      "epoch": 0.7170425482445942,
+      "grad_norm": 2.2290156819613687,
+      "learning_rate": 9.424953952552134e-06,
+      "loss": 0.7307,
+      "step": 514
+    },
+    {
+      "epoch": 0.7184375726575215,
+      "grad_norm": 2.04618346548946,
+      "learning_rate": 9.421164477646031e-06,
+      "loss": 0.7031,
+      "step": 515
+    },
+    {
+      "epoch": 0.7198325970704488,
+      "grad_norm": 2.2362515222635513,
+      "learning_rate": 9.41736332462853e-06,
+      "loss": 0.7545,
+      "step": 516
+    },
+    {
+      "epoch": 0.7212276214833759,
+      "grad_norm": 2.125062136154135,
+      "learning_rate": 9.413550503540039e-06,
+      "loss": 0.7379,
+      "step": 517
+    },
+    {
+      "epoch": 0.7226226458963032,
+      "grad_norm": 2.016687835394818,
+      "learning_rate": 9.409726024451781e-06,
+      "loss": 0.7415,
+      "step": 518
+    },
+    {
+      "epoch": 0.7240176703092304,
+      "grad_norm": 1.977367133466997,
+      "learning_rate": 9.40588989746578e-06,
+      "loss": 0.6891,
+      "step": 519
+    },
+    {
+      "epoch": 0.7254126947221576,
+      "grad_norm": 2.026408972891636,
+      "learning_rate": 9.402042132714817e-06,
+      "loss": 0.7004,
+      "step": 520
+    },
+    {
+      "epoch": 0.7268077191350849,
+      "grad_norm": 2.0546938078372183,
+      "learning_rate": 9.398182740362424e-06,
+      "loss": 0.7289,
+      "step": 521
+    },
+    {
+      "epoch": 0.7282027435480121,
+      "grad_norm": 1.9953897499935436,
+      "learning_rate": 9.39431173060284e-06,
+      "loss": 0.7276,
+      "step": 522
+    },
+    {
+      "epoch": 0.7295977679609393,
+      "grad_norm": 1.986025182877031,
+      "learning_rate": 9.390429113660993e-06,
+      "loss": 0.706,
+      "step": 523
+    },
+    {
+      "epoch": 0.7309927923738665,
+      "grad_norm": 2.1046471412286145,
+      "learning_rate": 9.38653489979247e-06,
+      "loss": 0.7602,
+      "step": 524
+    },
+    {
+      "epoch": 0.7323878167867938,
+      "grad_norm": 2.068123474852749,
+      "learning_rate": 9.382629099283486e-06,
+      "loss": 0.7071,
+      "step": 525
+    },
+    {
+      "epoch": 0.733782841199721,
+      "grad_norm": 1.9994411687174558,
+      "learning_rate": 9.378711722450866e-06,
+      "loss": 0.7304,
+      "step": 526
+    },
+    {
+      "epoch": 0.7351778656126482,
+      "grad_norm": 2.033536588615635,
+      "learning_rate": 9.374782779642013e-06,
+      "loss": 0.711,
+      "step": 527
+    },
+    {
+      "epoch": 0.7365728900255755,
+      "grad_norm": 2.0264601325168563,
+      "learning_rate": 9.370842281234876e-06,
+      "loss": 0.7279,
+      "step": 528
+    },
+    {
+      "epoch": 0.7379679144385026,
+      "grad_norm": 1.9113702618810533,
+      "learning_rate": 9.366890237637932e-06,
+      "loss": 0.6807,
+      "step": 529
+    },
+    {
+      "epoch": 0.7393629388514299,
+      "grad_norm": 1.968450453362606,
+      "learning_rate": 9.362926659290149e-06,
+      "loss": 0.6492,
+      "step": 530
+    },
+    {
+      "epoch": 0.7407579632643572,
+      "grad_norm": 2.09710149964991,
+      "learning_rate": 9.358951556660968e-06,
+      "loss": 0.7139,
+      "step": 531
+    },
+    {
+      "epoch": 0.7421529876772843,
+      "grad_norm": 1.9988693594361486,
+      "learning_rate": 9.354964940250269e-06,
+      "loss": 0.6732,
+      "step": 532
+    },
+    {
+      "epoch": 0.7435480120902116,
+      "grad_norm": 1.8960612048379188,
+      "learning_rate": 9.35096682058834e-06,
+      "loss": 0.6896,
+      "step": 533
+    },
+    {
+      "epoch": 0.7449430365031388,
+      "grad_norm": 2.005833390469395,
+      "learning_rate": 9.346957208235857e-06,
+      "loss": 0.7432,
+      "step": 534
+    },
+    {
+      "epoch": 0.746338060916066,
+      "grad_norm": 2.0295928389351476,
+      "learning_rate": 9.342936113783855e-06,
+      "loss": 0.7013,
+      "step": 535
+    },
+    {
+      "epoch": 0.7477330853289933,
+      "grad_norm": 2.1433987934073007,
+      "learning_rate": 9.338903547853698e-06,
+      "loss": 0.7294,
+      "step": 536
+    },
+    {
+      "epoch": 0.7491281097419205,
+      "grad_norm": 2.1474994409045287,
+      "learning_rate": 9.334859521097046e-06,
+      "loss": 0.7127,
+      "step": 537
+    },
+    {
+      "epoch": 0.7505231341548477,
+      "grad_norm": 1.921265945141032,
+      "learning_rate": 9.330804044195836e-06,
+      "loss": 0.6821,
+      "step": 538
+    },
+    {
+      "epoch": 0.7519181585677749,
+      "grad_norm": 2.168114835407853,
+      "learning_rate": 9.326737127862249e-06,
+      "loss": 0.711,
+      "step": 539
+    },
+    {
+      "epoch": 0.7533131829807022,
+      "grad_norm": 2.1528342823048074,
+      "learning_rate": 9.32265878283868e-06,
+      "loss": 0.7002,
+      "step": 540
+    },
+    {
+      "epoch": 0.7547082073936294,
+      "grad_norm": 2.0987262205834645,
+      "learning_rate": 9.318569019897713e-06,
+      "loss": 0.7538,
+      "step": 541
+    },
+    {
+      "epoch": 0.7561032318065566,
+      "grad_norm": 1.9509575446420426,
+      "learning_rate": 9.314467849842093e-06,
+      "loss": 0.7122,
+      "step": 542
+    },
+    {
+      "epoch": 0.7574982562194839,
+      "grad_norm": 2.057793685763029,
+      "learning_rate": 9.310355283504696e-06,
+      "loss": 0.7413,
+      "step": 543
+    },
+    {
+      "epoch": 0.758893280632411,
+      "grad_norm": 1.9925598392787818,
+      "learning_rate": 9.306231331748496e-06,
+      "loss": 0.7215,
+      "step": 544
+    },
+    {
+      "epoch": 0.7602883050453383,
+      "grad_norm": 1.9985902931925121,
+      "learning_rate": 9.302096005466547e-06,
+      "loss": 0.6459,
+      "step": 545
+    },
+    {
+      "epoch": 0.7616833294582656,
+      "grad_norm": 2.2363161994946505,
+      "learning_rate": 9.29794931558194e-06,
+      "loss": 0.7306,
+      "step": 546
+    },
+    {
+      "epoch": 0.7630783538711927,
+      "grad_norm": 2.1134851830312016,
+      "learning_rate": 9.29379127304779e-06,
+      "loss": 0.7338,
+      "step": 547
+    },
+    {
+      "epoch": 0.76447337828412,
+      "grad_norm": 2.0091531287699778,
+      "learning_rate": 9.289621888847194e-06,
+      "loss": 0.7636,
+      "step": 548
+    },
+    {
+      "epoch": 0.7658684026970471,
+      "grad_norm": 1.8526029005583842,
+      "learning_rate": 9.285441173993207e-06,
+      "loss": 0.6376,
+      "step": 549
+    },
+    {
+      "epoch": 0.7672634271099744,
+      "grad_norm": 2.076723919864085,
+      "learning_rate": 9.281249139528816e-06,
+      "loss": 0.7514,
+      "step": 550
+    },
+    {
+      "epoch": 0.7686584515229017,
+      "grad_norm": 2.024522052995256,
+      "learning_rate": 9.277045796526904e-06,
+      "loss": 0.7281,
+      "step": 551
+    },
+    {
+      "epoch": 0.7700534759358288,
+      "grad_norm": 2.07702628967691,
+      "learning_rate": 9.272831156090229e-06,
+      "loss": 0.7294,
+      "step": 552
+    },
+    {
+      "epoch": 0.7714485003487561,
+      "grad_norm": 1.9676509130458903,
+      "learning_rate": 9.268605229351387e-06,
+      "loss": 0.7135,
+      "step": 553
+    },
+    {
+      "epoch": 0.7728435247616833,
+      "grad_norm": 2.0873320795012993,
+      "learning_rate": 9.264368027472785e-06,
+      "loss": 0.7233,
+      "step": 554
+    },
+    {
+      "epoch": 0.7742385491746105,
+      "grad_norm": 2.0744695861945406,
+      "learning_rate": 9.260119561646614e-06,
+      "loss": 0.6836,
+      "step": 555
+    },
+    {
+      "epoch": 0.7756335735875378,
+      "grad_norm": 2.021884868718414,
+      "learning_rate": 9.255859843094817e-06,
+      "loss": 0.7258,
+      "step": 556
+    },
+    {
+      "epoch": 0.777028598000465,
+      "grad_norm": 2.1099089458381117,
+      "learning_rate": 9.25158888306906e-06,
+      "loss": 0.6828,
+      "step": 557
+    },
+    {
+      "epoch": 0.7784236224133922,
+      "grad_norm": 2.0907108018280507,
+      "learning_rate": 9.247306692850705e-06,
+      "loss": 0.7106,
+      "step": 558
+    },
+    {
+      "epoch": 0.7798186468263194,
+      "grad_norm": 2.0887522620268766,
+      "learning_rate": 9.243013283750774e-06,
+      "loss": 0.7697,
+      "step": 559
+    },
+    {
+      "epoch": 0.7812136712392467,
+      "grad_norm": 1.9955155672685916,
+      "learning_rate": 9.238708667109924e-06,
+      "loss": 0.7257,
+      "step": 560
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 2.0014379807399756,
+      "learning_rate": 9.234392854298414e-06,
+      "loss": 0.7381,
+      "step": 561
+    },
+    {
+      "epoch": 0.7840037200651011,
+      "grad_norm": 1.9045153376629644,
+      "learning_rate": 9.230065856716081e-06,
+      "loss": 0.6771,
+      "step": 562
+    },
+    {
+      "epoch": 0.7853987444780284,
+      "grad_norm": 1.9469069768920204,
+      "learning_rate": 9.225727685792302e-06,
+      "loss": 0.7056,
+      "step": 563
+    },
+    {
+      "epoch": 0.7867937688909556,
+      "grad_norm": 2.2031190641694125,
+      "learning_rate": 9.221378352985967e-06,
+      "loss": 0.7734,
+      "step": 564
+    },
+    {
+      "epoch": 0.7881887933038828,
+      "grad_norm": 2.0748393924695203,
+      "learning_rate": 9.217017869785453e-06,
+      "loss": 0.7082,
+      "step": 565
+    },
+    {
+      "epoch": 0.7895838177168101,
+      "grad_norm": 2.0995502525717864,
+      "learning_rate": 9.212646247708585e-06,
+      "loss": 0.7439,
+      "step": 566
+    },
+    {
+      "epoch": 0.7909788421297372,
+      "grad_norm": 2.068581983139408,
+      "learning_rate": 9.208263498302613e-06,
+      "loss": 0.7559,
+      "step": 567
+    },
+    {
+      "epoch": 0.7923738665426645,
+      "grad_norm": 1.9531887516172675,
+      "learning_rate": 9.203869633144182e-06,
+      "loss": 0.7143,
+      "step": 568
+    },
+    {
+      "epoch": 0.7937688909555918,
+      "grad_norm": 2.0266830492686347,
+      "learning_rate": 9.19946466383929e-06,
+      "loss": 0.6901,
+      "step": 569
+    },
+    {
+      "epoch": 0.7951639153685189,
+      "grad_norm": 1.9362491161859336,
+      "learning_rate": 9.19504860202327e-06,
+      "loss": 0.6699,
+      "step": 570
+    },
+    {
+      "epoch": 0.7965589397814462,
+      "grad_norm": 1.986999275214,
+      "learning_rate": 9.19062145936076e-06,
+      "loss": 0.6474,
+      "step": 571
+    },
+    {
+      "epoch": 0.7979539641943734,
+      "grad_norm": 2.2000153236749402,
+      "learning_rate": 9.186183247545657e-06,
+      "loss": 0.7826,
+      "step": 572
+    },
+    {
+      "epoch": 0.7993489886073006,
+      "grad_norm": 2.0941023318573215,
+      "learning_rate": 9.181733978301103e-06,
+      "loss": 0.7143,
+      "step": 573
+    },
+    {
+      "epoch": 0.8007440130202279,
+      "grad_norm": 2.071666423970005,
+      "learning_rate": 9.177273663379449e-06,
+      "loss": 0.6763,
+      "step": 574
+    },
+    {
+      "epoch": 0.8021390374331551,
+      "grad_norm": 2.2615998191295255,
+      "learning_rate": 9.172802314562214e-06,
+      "loss": 0.7621,
+      "step": 575
+    },
+    {
+      "epoch": 0.8035340618460823,
+      "grad_norm": 1.9739199882286245,
+      "learning_rate": 9.16831994366007e-06,
+      "loss": 0.6538,
+      "step": 576
+    },
+    {
+      "epoch": 0.8049290862590095,
+      "grad_norm": 2.0341275898427593,
+      "learning_rate": 9.1638265625128e-06,
+      "loss": 0.6651,
+      "step": 577
+    },
+    {
+      "epoch": 0.8063241106719368,
+      "grad_norm": 2.212386148497631,
+      "learning_rate": 9.159322182989265e-06,
+      "loss": 0.7095,
+      "step": 578
+    },
+    {
+      "epoch": 0.807719135084864,
+      "grad_norm": 2.1309132243222244,
+      "learning_rate": 9.154806816987386e-06,
+      "loss": 0.6603,
+      "step": 579
+    },
+    {
+      "epoch": 0.8091141594977912,
+      "grad_norm": 2.054866199242643,
+      "learning_rate": 9.150280476434098e-06,
+      "loss": 0.6942,
+      "step": 580
+    },
+    {
+      "epoch": 0.8105091839107185,
+      "grad_norm": 1.9524458748633242,
+      "learning_rate": 9.145743173285325e-06,
+      "loss": 0.6488,
+      "step": 581
+    },
+    {
+      "epoch": 0.8119042083236456,
+      "grad_norm": 2.030621418231089,
+      "learning_rate": 9.141194919525949e-06,
+      "loss": 0.7108,
+      "step": 582
+    },
+    {
+      "epoch": 0.8132992327365729,
+      "grad_norm": 2.084991031526505,
+      "learning_rate": 9.136635727169776e-06,
+      "loss": 0.7265,
+      "step": 583
+    },
+    {
+      "epoch": 0.8146942571495002,
+      "grad_norm": 2.0622263794834224,
+      "learning_rate": 9.132065608259505e-06,
+      "loss": 0.7503,
+      "step": 584
+    },
+    {
+      "epoch": 0.8160892815624273,
+      "grad_norm": 2.0982571575326228,
+      "learning_rate": 9.127484574866699e-06,
+      "loss": 0.7235,
+      "step": 585
+    },
+    {
+      "epoch": 0.8174843059753546,
+      "grad_norm": 2.0514483972342834,
+      "learning_rate": 9.122892639091748e-06,
+      "loss": 0.7288,
+      "step": 586
+    },
+    {
+      "epoch": 0.8188793303882818,
+      "grad_norm": 1.973001994554894,
+      "learning_rate": 9.118289813063842e-06,
+      "loss": 0.6897,
+      "step": 587
+    },
+    {
+      "epoch": 0.820274354801209,
+      "grad_norm": 1.983523663982352,
+      "learning_rate": 9.11367610894093e-06,
+      "loss": 0.7193,
+      "step": 588
+    },
+    {
+      "epoch": 0.8216693792141363,
+      "grad_norm": 2.011324854017405,
+      "learning_rate": 9.109051538909707e-06,
+      "loss": 0.6593,
+      "step": 589
+    },
+    {
+      "epoch": 0.8230644036270635,
+      "grad_norm": 2.046660765022832,
+      "learning_rate": 9.104416115185557e-06,
+      "loss": 0.7679,
+      "step": 590
+    },
+    {
+      "epoch": 0.8244594280399907,
+      "grad_norm": 2.1499540624366635,
+      "learning_rate": 9.099769850012539e-06,
+      "loss": 0.757,
+      "step": 591
+    },
+    {
+      "epoch": 0.8258544524529179,
+      "grad_norm": 1.9870557541462837,
+      "learning_rate": 9.095112755663349e-06,
+      "loss": 0.6832,
+      "step": 592
+    },
+    {
+      "epoch": 0.8272494768658452,
+      "grad_norm": 1.9850961987043347,
+      "learning_rate": 9.090444844439284e-06,
+      "loss": 0.698,
+      "step": 593
+    },
+    {
+      "epoch": 0.8286445012787724,
+      "grad_norm": 1.9852851734683514,
+      "learning_rate": 9.085766128670218e-06,
+      "loss": 0.7543,
+      "step": 594
+    },
+    {
+      "epoch": 0.8300395256916996,
+      "grad_norm": 2.128351674301273,
+      "learning_rate": 9.08107662071456e-06,
+      "loss": 0.7076,
+      "step": 595
+    },
+    {
+      "epoch": 0.8314345501046269,
+      "grad_norm": 2.0178147387617122,
+      "learning_rate": 9.076376332959222e-06,
+      "loss": 0.7241,
+      "step": 596
+    },
+    {
+      "epoch": 0.832829574517554,
+      "grad_norm": 2.086053250244484,
+      "learning_rate": 9.071665277819603e-06,
+      "loss": 0.6777,
+      "step": 597
+    },
+    {
+      "epoch": 0.8342245989304813,
+      "grad_norm": 2.058823763171869,
+      "learning_rate": 9.066943467739529e-06,
+      "loss": 0.713,
+      "step": 598
+    },
+    {
+      "epoch": 0.8356196233434086,
+      "grad_norm": 2.1956205519625835,
+      "learning_rate": 9.06221091519124e-06,
+      "loss": 0.7586,
+      "step": 599
+    },
+    {
+      "epoch": 0.8370146477563357,
+      "grad_norm": 2.0804960970930537,
+      "learning_rate": 9.057467632675357e-06,
+      "loss": 0.7694,
+      "step": 600
+    },
+    {
+      "epoch": 0.838409672169263,
+      "grad_norm": 1.9765569639718268,
+      "learning_rate": 9.05271363272083e-06,
+      "loss": 0.7285,
+      "step": 601
+    },
+    {
+      "epoch": 0.8398046965821901,
+      "grad_norm": 2.046920137660995,
+      "learning_rate": 9.047948927884927e-06,
+      "loss": 0.7139,
+      "step": 602
+    },
+    {
+      "epoch": 0.8411997209951174,
+      "grad_norm": 2.0022644204067523,
+      "learning_rate": 9.043173530753196e-06,
+      "loss": 0.7078,
+      "step": 603
+    },
+    {
+      "epoch": 0.8425947454080447,
+      "grad_norm": 2.1957522908702862,
+      "learning_rate": 9.038387453939416e-06,
+      "loss": 0.7107,
+      "step": 604
+    },
+    {
+      "epoch": 0.8439897698209718,
+      "grad_norm": 2.130770907443561,
+      "learning_rate": 9.033590710085584e-06,
+      "loss": 0.6822,
+      "step": 605
+    },
+    {
+      "epoch": 0.8453847942338991,
+      "grad_norm": 1.966006179455826,
+      "learning_rate": 9.028783311861874e-06,
+      "loss": 0.6649,
+      "step": 606
+    },
+    {
+      "epoch": 0.8467798186468263,
+      "grad_norm": 2.0759903419882164,
+      "learning_rate": 9.023965271966595e-06,
+      "loss": 0.6753,
+      "step": 607
+    },
+    {
+      "epoch": 0.8481748430597535,
+      "grad_norm": 1.9940729760728366,
+      "learning_rate": 9.019136603126171e-06,
+      "loss": 0.7197,
+      "step": 608
+    },
+    {
+      "epoch": 0.8495698674726808,
+      "grad_norm": 1.988407232009193,
+      "learning_rate": 9.0142973180951e-06,
+      "loss": 0.7094,
+      "step": 609
+    },
+    {
+      "epoch": 0.850964891885608,
+      "grad_norm": 2.035677011161458,
+      "learning_rate": 9.00944742965592e-06,
+      "loss": 0.7328,
+      "step": 610
+    },
+    {
+      "epoch": 0.8523599162985352,
+      "grad_norm": 1.8950053977425145,
+      "learning_rate": 9.004586950619182e-06,
+      "loss": 0.6823,
+      "step": 611
+    },
+    {
+      "epoch": 0.8537549407114624,
+      "grad_norm": 2.1156907480266605,
+      "learning_rate": 8.999715893823404e-06,
+      "loss": 0.7423,
+      "step": 612
+    },
+    {
+      "epoch": 0.8551499651243897,
+      "grad_norm": 1.961252039158567,
+      "learning_rate": 8.994834272135049e-06,
+      "loss": 0.7171,
+      "step": 613
+    },
+    {
+      "epoch": 0.856544989537317,
+      "grad_norm": 2.0001585593137303,
+      "learning_rate": 8.989942098448485e-06,
+      "loss": 0.7121,
+      "step": 614
+    },
+    {
+      "epoch": 0.8579400139502441,
+      "grad_norm": 1.9949836027553451,
+      "learning_rate": 8.985039385685952e-06,
+      "loss": 0.6975,
+      "step": 615
+    },
+    {
+      "epoch": 0.8593350383631714,
+      "grad_norm": 1.9357238900964018,
+      "learning_rate": 8.98012614679753e-06,
+      "loss": 0.6363,
+      "step": 616
+    },
+    {
+      "epoch": 0.8607300627760985,
+      "grad_norm": 2.2080743341990576,
+      "learning_rate": 8.975202394761098e-06,
+      "loss": 0.6823,
+      "step": 617
+    },
+    {
+      "epoch": 0.8621250871890258,
+      "grad_norm": 2.0593736149801574,
+      "learning_rate": 8.970268142582312e-06,
+      "loss": 0.7101,
+      "step": 618
+    },
+    {
+      "epoch": 0.8635201116019531,
+      "grad_norm": 2.0081932189174174,
+      "learning_rate": 8.965323403294553e-06,
+      "loss": 0.672,
+      "step": 619
+    },
+    {
+      "epoch": 0.8649151360148802,
+      "grad_norm": 2.0431522588311037,
+      "learning_rate": 8.960368189958913e-06,
+      "loss": 0.7229,
+      "step": 620
+    },
+    {
+      "epoch": 0.8663101604278075,
+      "grad_norm": 2.0999018236018636,
+      "learning_rate": 8.955402515664144e-06,
+      "loss": 0.7264,
+      "step": 621
+    },
+    {
+      "epoch": 0.8677051848407347,
+      "grad_norm": 2.2421086786694304,
+      "learning_rate": 8.950426393526633e-06,
+      "loss": 0.7121,
+      "step": 622
+    },
+    {
+      "epoch": 0.8691002092536619,
+      "grad_norm": 1.9548355595783216,
+      "learning_rate": 8.945439836690359e-06,
+      "loss": 0.7026,
+      "step": 623
+    },
+    {
+      "epoch": 0.8704952336665892,
+      "grad_norm": 2.0065422789990333,
+      "learning_rate": 8.940442858326871e-06,
+      "loss": 0.6661,
+      "step": 624
+    },
+    {
+      "epoch": 0.8718902580795164,
+      "grad_norm": 1.9311985998592487,
+      "learning_rate": 8.935435471635238e-06,
+      "loss": 0.7059,
+      "step": 625
+    },
+    {
+      "epoch": 0.8732852824924436,
+      "grad_norm": 2.0030403022945245,
+      "learning_rate": 8.93041768984203e-06,
+      "loss": 0.6671,
+      "step": 626
+    },
+    {
+      "epoch": 0.8746803069053708,
+      "grad_norm": 2.1110226512909716,
+      "learning_rate": 8.925389526201264e-06,
+      "loss": 0.6935,
+      "step": 627
+    },
+    {
+      "epoch": 0.8760753313182981,
+      "grad_norm": 1.837637633874893,
+      "learning_rate": 8.920350993994387e-06,
+      "loss": 0.6376,
+      "step": 628
+    },
+    {
+      "epoch": 0.8774703557312253,
+      "grad_norm": 2.0605269428260113,
+      "learning_rate": 8.915302106530234e-06,
+      "loss": 0.693,
+      "step": 629
+    },
+    {
+      "epoch": 0.8788653801441525,
+      "grad_norm": 2.0408953219727155,
+      "learning_rate": 8.91024287714499e-06,
+      "loss": 0.6685,
+      "step": 630
+    },
+    {
+      "epoch": 0.8802604045570798,
+      "grad_norm": 1.9732412676724707,
+      "learning_rate": 8.905173319202159e-06,
+      "loss": 0.696,
+      "step": 631
+    },
+    {
+      "epoch": 0.8816554289700069,
+      "grad_norm": 1.9599933746722413,
+      "learning_rate": 8.900093446092523e-06,
+      "loss": 0.6919,
+      "step": 632
+    },
+    {
+      "epoch": 0.8830504533829342,
+      "grad_norm": 2.0071687490178736,
+      "learning_rate": 8.895003271234116e-06,
+      "loss": 0.7947,
+      "step": 633
+    },
+    {
+      "epoch": 0.8844454777958615,
+      "grad_norm": 2.0368651268041766,
+      "learning_rate": 8.889902808072178e-06,
+      "loss": 0.6618,
+      "step": 634
+    },
+    {
+      "epoch": 0.8858405022087886,
+      "grad_norm": 2.006821931119106,
+      "learning_rate": 8.884792070079128e-06,
+      "loss": 0.6986,
+      "step": 635
+    },
+    {
+      "epoch": 0.8872355266217159,
+      "grad_norm": 1.953819130025858,
+      "learning_rate": 8.879671070754527e-06,
+      "loss": 0.6885,
+      "step": 636
+    },
+    {
+      "epoch": 0.8886305510346431,
+      "grad_norm": 2.077254864377799,
+      "learning_rate": 8.874539823625037e-06,
+      "loss": 0.7307,
+      "step": 637
+    },
+    {
+      "epoch": 0.8900255754475703,
+      "grad_norm": 1.9249651603564881,
+      "learning_rate": 8.869398342244387e-06,
+      "loss": 0.6911,
+      "step": 638
+    },
+    {
+      "epoch": 0.8914205998604976,
+      "grad_norm": 2.3835538973236616,
+      "learning_rate": 8.86424664019334e-06,
+      "loss": 0.7769,
+      "step": 639
+    },
+    {
+      "epoch": 0.8928156242734248,
+      "grad_norm": 1.8989674380166495,
+      "learning_rate": 8.859084731079664e-06,
+      "loss": 0.6892,
+      "step": 640
+    },
+    {
+      "epoch": 0.894210648686352,
+      "grad_norm": 2.0401897832182647,
+      "learning_rate": 8.853912628538072e-06,
+      "loss": 0.7131,
+      "step": 641
+    },
+    {
+      "epoch": 0.8956056730992792,
+      "grad_norm": 1.9764252068678023,
+      "learning_rate": 8.84873034623022e-06,
+      "loss": 0.6564,
+      "step": 642
+    },
+    {
+      "epoch": 0.8970006975122065,
+      "grad_norm": 1.92924996628619,
+      "learning_rate": 8.84353789784464e-06,
+      "loss": 0.7245,
+      "step": 643
+    },
+    {
+      "epoch": 0.8983957219251337,
+      "grad_norm": 2.057502035838945,
+      "learning_rate": 8.83833529709672e-06,
+      "loss": 0.71,
+      "step": 644
+    },
+    {
+      "epoch": 0.8997907463380609,
+      "grad_norm": 1.886888629633787,
+      "learning_rate": 8.833122557728667e-06,
+      "loss": 0.6765,
+      "step": 645
+    },
+    {
+      "epoch": 0.9011857707509882,
+      "grad_norm": 2.105286098847447,
+      "learning_rate": 8.827899693509467e-06,
+      "loss": 0.7547,
+      "step": 646
+    },
+    {
+      "epoch": 0.9025807951639153,
+      "grad_norm": 1.9861957912372208,
+      "learning_rate": 8.82266671823485e-06,
+      "loss": 0.7219,
+      "step": 647
+    },
+    {
+      "epoch": 0.9039758195768426,
+      "grad_norm": 1.888915458083586,
+      "learning_rate": 8.817423645727252e-06,
+      "loss": 0.6884,
+      "step": 648
+    },
+    {
+      "epoch": 0.9053708439897699,
+      "grad_norm": 2.277569646085658,
+      "learning_rate": 8.812170489835784e-06,
+      "loss": 0.7798,
+      "step": 649
+    },
+    {
+      "epoch": 0.906765868402697,
+      "grad_norm": 1.9944963507674216,
+      "learning_rate": 8.806907264436183e-06,
+      "loss": 0.6861,
+      "step": 650
+    },
+    {
+      "epoch": 0.9081608928156243,
+      "grad_norm": 1.9298477509530116,
+      "learning_rate": 8.801633983430794e-06,
+      "loss": 0.6591,
+      "step": 651
+    },
+    {
+      "epoch": 0.9095559172285514,
+      "grad_norm": 1.8705807621214512,
+      "learning_rate": 8.796350660748516e-06,
+      "loss": 0.6462,
+      "step": 652
+    },
+    {
+      "epoch": 0.9109509416414787,
+      "grad_norm": 1.781871433076164,
+      "learning_rate": 8.791057310344775e-06,
+      "loss": 0.696,
+      "step": 653
+    },
+    {
+      "epoch": 0.912345966054406,
+      "grad_norm": 2.03708716140148,
+      "learning_rate": 8.785753946201484e-06,
+      "loss": 0.7908,
+      "step": 654
+    },
+    {
+      "epoch": 0.9137409904673331,
+      "grad_norm": 1.9723359587123608,
+      "learning_rate": 8.780440582327005e-06,
+      "loss": 0.6875,
+      "step": 655
+    },
+    {
+      "epoch": 0.9151360148802604,
+      "grad_norm": 1.9079177689320517,
+      "learning_rate": 8.775117232756116e-06,
+      "loss": 0.6515,
+      "step": 656
+    },
+    {
+      "epoch": 0.9165310392931876,
+      "grad_norm": 2.0101883192315175,
+      "learning_rate": 8.769783911549968e-06,
+      "loss": 0.7285,
+      "step": 657
+    },
+    {
+      "epoch": 0.9179260637061148,
+      "grad_norm": 1.9520878357376525,
+      "learning_rate": 8.764440632796055e-06,
+      "loss": 0.7019,
+      "step": 658
+    },
+    {
+      "epoch": 0.9193210881190421,
+      "grad_norm": 1.8645433147053474,
+      "learning_rate": 8.75908741060817e-06,
+      "loss": 0.723,
+      "step": 659
+    },
+    {
+      "epoch": 0.9207161125319693,
+      "grad_norm": 1.9232448575331198,
+      "learning_rate": 8.75372425912637e-06,
+      "loss": 0.7015,
+      "step": 660
+    },
+    {
+      "epoch": 0.9221111369448965,
+      "grad_norm": 2.2313621348056074,
+      "learning_rate": 8.748351192516943e-06,
+      "loss": 0.779,
+      "step": 661
+    },
+    {
+      "epoch": 0.9235061613578237,
+      "grad_norm": 2.0761748347303737,
+      "learning_rate": 8.742968224972366e-06,
+      "loss": 0.695,
+      "step": 662
+    },
+    {
+      "epoch": 0.924901185770751,
+      "grad_norm": 1.9030018961305923,
+      "learning_rate": 8.737575370711265e-06,
+      "loss": 0.682,
+      "step": 663
+    },
+    {
+      "epoch": 0.9262962101836782,
+      "grad_norm": 2.039629984371781,
+      "learning_rate": 8.732172643978383e-06,
+      "loss": 0.7116,
+      "step": 664
+    },
+    {
+      "epoch": 0.9276912345966054,
+      "grad_norm": 2.0754184443315475,
+      "learning_rate": 8.726760059044542e-06,
+      "loss": 0.6939,
+      "step": 665
+    },
+    {
+      "epoch": 0.9290862590095327,
+      "grad_norm": 2.138276597253035,
+      "learning_rate": 8.721337630206603e-06,
+      "loss": 0.7504,
+      "step": 666
+    },
+    {
+      "epoch": 0.93048128342246,
+      "grad_norm": 2.088090556558224,
+      "learning_rate": 8.715905371787426e-06,
+      "loss": 0.7307,
+      "step": 667
+    },
+    {
+      "epoch": 0.9318763078353871,
+      "grad_norm": 1.9914456952917818,
+      "learning_rate": 8.710463298135836e-06,
+      "loss": 0.6436,
+      "step": 668
+    },
+    {
+      "epoch": 0.9332713322483144,
+      "grad_norm": 1.930717236841167,
+      "learning_rate": 8.705011423626589e-06,
+      "loss": 0.7399,
+      "step": 669
+    },
+    {
+      "epoch": 0.9346663566612415,
+      "grad_norm": 2.1376988846220746,
+      "learning_rate": 8.699549762660318e-06,
+      "loss": 0.6954,
+      "step": 670
+    },
+    {
+      "epoch": 0.9360613810741688,
+      "grad_norm": 1.9710970809397323,
+      "learning_rate": 8.69407832966352e-06,
+      "loss": 0.7039,
+      "step": 671
+    },
+    {
+      "epoch": 0.9374564054870961,
+      "grad_norm": 2.0867427607781757,
+      "learning_rate": 8.688597139088494e-06,
+      "loss": 0.7234,
+      "step": 672
+    },
+    {
+      "epoch": 0.9388514299000232,
+      "grad_norm": 1.8578210034776752,
+      "learning_rate": 8.683106205413316e-06,
+      "loss": 0.6676,
+      "step": 673
+    },
+    {
+      "epoch": 0.9402464543129505,
+      "grad_norm": 1.7992418975827844,
+      "learning_rate": 8.677605543141797e-06,
+      "loss": 0.6473,
+      "step": 674
+    },
+    {
+      "epoch": 0.9416414787258777,
+      "grad_norm": 2.208549526088649,
+      "learning_rate": 8.672095166803445e-06,
+      "loss": 0.734,
+      "step": 675
+    },
+    {
+      "epoch": 0.9430365031388049,
+      "grad_norm": 1.9309424859366655,
+      "learning_rate": 8.666575090953426e-06,
+      "loss": 0.7099,
+      "step": 676
+    },
+    {
+      "epoch": 0.9444315275517322,
+      "grad_norm": 2.0689005805396006,
+      "learning_rate": 8.661045330172533e-06,
+      "loss": 0.7762,
+      "step": 677
+    },
+    {
+      "epoch": 0.9458265519646594,
+      "grad_norm": 1.9327832133972194,
+      "learning_rate": 8.65550589906713e-06,
+      "loss": 0.6634,
+      "step": 678
+    },
+    {
+      "epoch": 0.9472215763775866,
+      "grad_norm": 1.8654184391224133,
+      "learning_rate": 8.649956812269134e-06,
+      "loss": 0.6768,
+      "step": 679
+    },
+    {
+      "epoch": 0.9486166007905138,
+      "grad_norm": 1.8537317310543904,
+      "learning_rate": 8.644398084435959e-06,
+      "loss": 0.6687,
+      "step": 680
+    },
+    {
+      "epoch": 0.9500116252034411,
+      "grad_norm": 1.9620198811804095,
+      "learning_rate": 8.63882973025049e-06,
+      "loss": 0.6837,
+      "step": 681
+    },
+    {
+      "epoch": 0.9514066496163683,
+      "grad_norm": 1.8902874912708485,
+      "learning_rate": 8.63325176442104e-06,
+      "loss": 0.7166,
+      "step": 682
+    },
+    {
+      "epoch": 0.9528016740292955,
+      "grad_norm": 1.9012648753801518,
+      "learning_rate": 8.627664201681305e-06,
+      "loss": 0.6771,
+      "step": 683
+    },
+    {
+      "epoch": 0.9541966984422228,
+      "grad_norm": 2.0772673760923657,
+      "learning_rate": 8.622067056790333e-06,
+      "loss": 0.7442,
+      "step": 684
+    },
+    {
+      "epoch": 0.9555917228551499,
+      "grad_norm": 1.881931761326016,
+      "learning_rate": 8.616460344532483e-06,
+      "loss": 0.7081,
+      "step": 685
+    },
+    {
+      "epoch": 0.9569867472680772,
+      "grad_norm": 1.886919445097637,
+      "learning_rate": 8.610844079717387e-06,
+      "loss": 0.6433,
+      "step": 686
+    },
+    {
+      "epoch": 0.9583817716810045,
+      "grad_norm": 2.002169759368971,
+      "learning_rate": 8.605218277179907e-06,
+      "loss": 0.7323,
+      "step": 687
+    },
+    {
+      "epoch": 0.9597767960939316,
+      "grad_norm": 1.7540706653539795,
+      "learning_rate": 8.599582951780095e-06,
+      "loss": 0.6229,
+      "step": 688
+    },
+    {
+      "epoch": 0.9611718205068589,
+      "grad_norm": 1.9764352278914321,
+      "learning_rate": 8.593938118403164e-06,
+      "loss": 0.6744,
+      "step": 689
+    },
+    {
+      "epoch": 0.9625668449197861,
+      "grad_norm": 2.043953151411979,
+      "learning_rate": 8.588283791959437e-06,
+      "loss": 0.7051,
+      "step": 690
+    },
+    {
+      "epoch": 0.9639618693327133,
+      "grad_norm": 2.0060893469750303,
+      "learning_rate": 8.582619987384311e-06,
+      "loss": 0.6858,
+      "step": 691
+    },
+    {
+      "epoch": 0.9653568937456406,
+      "grad_norm": 1.885119316458161,
+      "learning_rate": 8.57694671963822e-06,
+      "loss": 0.6503,
+      "step": 692
+    },
+    {
+      "epoch": 0.9667519181585678,
+      "grad_norm": 2.027815652298139,
+      "learning_rate": 8.571264003706596e-06,
+      "loss": 0.7235,
+      "step": 693
+    },
+    {
+      "epoch": 0.968146942571495,
+      "grad_norm": 1.905998996142595,
+      "learning_rate": 8.565571854599825e-06,
+      "loss": 0.6732,
+      "step": 694
+    },
+    {
+      "epoch": 0.9695419669844222,
+      "grad_norm": 2.066485530703545,
+      "learning_rate": 8.559870287353214e-06,
+      "loss": 0.737,
+      "step": 695
+    },
+    {
+      "epoch": 0.9709369913973495,
+      "grad_norm": 1.9622091919063105,
+      "learning_rate": 8.554159317026939e-06,
+      "loss": 0.6509,
+      "step": 696
+    },
+    {
+      "epoch": 0.9723320158102767,
+      "grad_norm": 1.988578644158127,
+      "learning_rate": 8.548438958706022e-06,
+      "loss": 0.7456,
+      "step": 697
+    },
+    {
+      "epoch": 0.9737270402232039,
+      "grad_norm": 2.116243375506424,
+      "learning_rate": 8.542709227500276e-06,
+      "loss": 0.765,
+      "step": 698
+    },
+    {
+      "epoch": 0.9751220646361312,
+      "grad_norm": 1.9231890289170668,
+      "learning_rate": 8.536970138544278e-06,
+      "loss": 0.6652,
+      "step": 699
+    },
+    {
+      "epoch": 0.9765170890490583,
+      "grad_norm": 1.9552199841603932,
+      "learning_rate": 8.531221706997316e-06,
+      "loss": 0.6576,
+      "step": 700
+    },
+    {
+      "epoch": 0.9779121134619856,
+      "grad_norm": 2.031756235992153,
+      "learning_rate": 8.525463948043365e-06,
+      "loss": 0.7072,
+      "step": 701
+    },
+    {
+      "epoch": 0.9793071378749129,
+      "grad_norm": 1.9389046550623854,
+      "learning_rate": 8.519696876891024e-06,
+      "loss": 0.6442,
+      "step": 702
+    },
+    {
+      "epoch": 0.98070216228784,
+      "grad_norm": 1.919889704452434,
+      "learning_rate": 8.513920508773499e-06,
+      "loss": 0.6964,
+      "step": 703
+    },
+    {
+      "epoch": 0.9820971867007673,
+      "grad_norm": 1.9236646080238848,
+      "learning_rate": 8.508134858948553e-06,
+      "loss": 0.6919,
+      "step": 704
+    },
+    {
+      "epoch": 0.9834922111136944,
+      "grad_norm": 1.9493128891249456,
+      "learning_rate": 8.502339942698463e-06,
+      "loss": 0.6646,
+      "step": 705
+    },
+    {
+      "epoch": 0.9848872355266217,
+      "grad_norm": 1.9673672208067479,
+      "learning_rate": 8.496535775329982e-06,
+      "loss": 0.7233,
+      "step": 706
+    },
+    {
+      "epoch": 0.986282259939549,
+      "grad_norm": 1.9621954045331194,
+      "learning_rate": 8.4907223721743e-06,
+      "loss": 0.6778,
+      "step": 707
+    },
+    {
+      "epoch": 0.9876772843524761,
+      "grad_norm": 1.8046753114923484,
+      "learning_rate": 8.484899748587003e-06,
+      "loss": 0.6646,
+      "step": 708
+    },
+    {
+      "epoch": 0.9890723087654034,
+      "grad_norm": 2.2295174686287274,
+      "learning_rate": 8.479067919948032e-06,
+      "loss": 0.6989,
+      "step": 709
+    },
+    {
+      "epoch": 0.9904673331783306,
+      "grad_norm": 2.1070329009446156,
+      "learning_rate": 8.473226901661643e-06,
+      "loss": 0.7844,
+      "step": 710
+    },
+    {
+      "epoch": 0.9918623575912578,
+      "grad_norm": 2.1002202467765927,
+      "learning_rate": 8.46737670915636e-06,
+      "loss": 0.6869,
+      "step": 711
+    },
+    {
+      "epoch": 0.9932573820041851,
+      "grad_norm": 2.0464784128561084,
+      "learning_rate": 8.46151735788495e-06,
+      "loss": 0.6978,
+      "step": 712
+    },
+    {
+      "epoch": 0.9946524064171123,
+      "grad_norm": 1.8705732417938106,
+      "learning_rate": 8.455648863324364e-06,
+      "loss": 0.6423,
+      "step": 713
+    },
+    {
+      "epoch": 0.9960474308300395,
+      "grad_norm": 1.8660210155077552,
+      "learning_rate": 8.449771240975707e-06,
+      "loss": 0.6669,
+      "step": 714
+    },
+    {
+      "epoch": 0.9974424552429667,
+      "grad_norm": 1.9315548591495726,
+      "learning_rate": 8.443884506364192e-06,
+      "loss": 0.6853,
+      "step": 715
+    },
+    {
+      "epoch": 0.998837479655894,
+      "grad_norm": 1.9920543550527006,
+      "learning_rate": 8.437988675039108e-06,
+      "loss": 0.6772,
+      "step": 716
+    },
+    {
+      "epoch": 1.0013950244129273,
+      "grad_norm": 2.763874765719216,
+      "learning_rate": 8.432083762573761e-06,
+      "loss": 1.1699,
+      "step": 717
+    },
+    {
+      "epoch": 1.0027900488258545,
+      "grad_norm": 1.754525337447559,
+      "learning_rate": 8.426169784565452e-06,
+      "loss": 0.4932,
+      "step": 718
+    },
+    {
+      "epoch": 1.0041850732387816,
+      "grad_norm": 1.846326658857791,
+      "learning_rate": 8.420246756635431e-06,
+      "loss": 0.448,
+      "step": 719
+    },
+    {
+      "epoch": 1.0055800976517089,
+      "grad_norm": 1.8768804454261978,
+      "learning_rate": 8.414314694428842e-06,
+      "loss": 0.4791,
+      "step": 720
+    },
+    {
+      "epoch": 1.0069751220646361,
+      "grad_norm": 1.6152669499441352,
+      "learning_rate": 8.408373613614699e-06,
+      "loss": 0.491,
+      "step": 721
+    },
+    {
+      "epoch": 1.0083701464775634,
+      "grad_norm": 1.8596160204714411,
+      "learning_rate": 8.40242352988584e-06,
+      "loss": 0.4418,
+      "step": 722
+    },
+    {
+      "epoch": 1.0097651708904907,
+      "grad_norm": 1.8162139230993801,
+      "learning_rate": 8.396464458958876e-06,
+      "loss": 0.5305,
+      "step": 723
+    },
+    {
+      "epoch": 1.0111601953034177,
+      "grad_norm": 2.0254017362191536,
+      "learning_rate": 8.390496416574166e-06,
+      "loss": 0.481,
+      "step": 724
+    },
+    {
+      "epoch": 1.012555219716345,
+      "grad_norm": 2.1506267595528867,
+      "learning_rate": 8.384519418495755e-06,
+      "loss": 0.5195,
+      "step": 725
+    },
+    {
+      "epoch": 1.0139502441292723,
+      "grad_norm": 2.101119705098204,
+      "learning_rate": 8.378533480511355e-06,
+      "loss": 0.4906,
+      "step": 726
+    },
+    {
+      "epoch": 1.0153452685421995,
+      "grad_norm": 2.684686947501843,
+      "learning_rate": 8.372538618432282e-06,
+      "loss": 0.513,
+      "step": 727
+    },
+    {
+      "epoch": 1.0167402929551268,
+      "grad_norm": 2.2485542720043092,
+      "learning_rate": 8.366534848093434e-06,
+      "loss": 0.5166,
+      "step": 728
+    },
+    {
+      "epoch": 1.0181353173680538,
+      "grad_norm": 2.28734638846399,
+      "learning_rate": 8.360522185353234e-06,
+      "loss": 0.4826,
+      "step": 729
+    },
+    {
+      "epoch": 1.0195303417809811,
+      "grad_norm": 1.9701409436889183,
+      "learning_rate": 8.354500646093592e-06,
+      "loss": 0.4653,
+      "step": 730
+    },
+    {
+      "epoch": 1.0209253661939084,
+      "grad_norm": 2.2247454732523426,
+      "learning_rate": 8.348470246219872e-06,
+      "loss": 0.5245,
+      "step": 731
+    },
+    {
+      "epoch": 1.0223203906068357,
+      "grad_norm": 2.007192211701691,
+      "learning_rate": 8.342431001660826e-06,
+      "loss": 0.4565,
+      "step": 732
+    },
+    {
+      "epoch": 1.023715415019763,
+      "grad_norm": 1.9247320693390784,
+      "learning_rate": 8.33638292836859e-06,
+      "loss": 0.4874,
+      "step": 733
+    },
+    {
+      "epoch": 1.02511043943269,
+      "grad_norm": 1.9650014172175294,
+      "learning_rate": 8.330326042318605e-06,
+      "loss": 0.5046,
+      "step": 734
+    },
+    {
+      "epoch": 1.0265054638456172,
+      "grad_norm": 1.8423438549714501,
+      "learning_rate": 8.324260359509594e-06,
+      "loss": 0.4836,
+      "step": 735
+    },
+    {
+      "epoch": 1.0279004882585445,
+      "grad_norm": 1.9039412894123366,
+      "learning_rate": 8.31818589596352e-06,
+      "loss": 0.4582,
+      "step": 736
+    },
+    {
+      "epoch": 1.0292955126714718,
+      "grad_norm": 2.1936250318321453,
+      "learning_rate": 8.312102667725534e-06,
+      "loss": 0.5209,
+      "step": 737
+    },
+    {
+      "epoch": 1.030690537084399,
+      "grad_norm": 1.84976943631084,
+      "learning_rate": 8.306010690863943e-06,
+      "loss": 0.4619,
+      "step": 738
+    },
+    {
+      "epoch": 1.032085561497326,
+      "grad_norm": 2.0139559424062443,
+      "learning_rate": 8.299909981470159e-06,
+      "loss": 0.487,
+      "step": 739
+    },
+    {
+      "epoch": 1.0334805859102534,
+      "grad_norm": 2.0270755400609515,
+      "learning_rate": 8.29380055565866e-06,
+      "loss": 0.5021,
+      "step": 740
+    },
+    {
+      "epoch": 1.0348756103231807,
+      "grad_norm": 1.9790811313609495,
+      "learning_rate": 8.28768242956695e-06,
+      "loss": 0.4792,
+      "step": 741
+    },
+    {
+      "epoch": 1.036270634736108,
+      "grad_norm": 1.931174120623029,
+      "learning_rate": 8.281555619355515e-06,
+      "loss": 0.5543,
+      "step": 742
+    },
+    {
+      "epoch": 1.0376656591490352,
+      "grad_norm": 2.094412424706352,
+      "learning_rate": 8.275420141207775e-06,
+      "loss": 0.5025,
+      "step": 743
+    },
+    {
+      "epoch": 1.0390606835619622,
+      "grad_norm": 2.086934890928623,
+      "learning_rate": 8.269276011330048e-06,
+      "loss": 0.4627,
+      "step": 744
+    },
+    {
+      "epoch": 1.0404557079748895,
+      "grad_norm": 1.9002029258471083,
+      "learning_rate": 8.263123245951504e-06,
+      "loss": 0.5457,
+      "step": 745
+    },
+    {
+      "epoch": 1.0418507323878168,
+      "grad_norm": 2.114961818573959,
+      "learning_rate": 8.256961861324127e-06,
+      "loss": 0.479,
+      "step": 746
+    },
+    {
+      "epoch": 1.043245756800744,
+      "grad_norm": 2.153993003889234,
+      "learning_rate": 8.250791873722662e-06,
+      "loss": 0.4701,
+      "step": 747
+    },
+    {
+      "epoch": 1.0446407812136713,
+      "grad_norm": 2.0652134135665343,
+      "learning_rate": 8.244613299444581e-06,
+      "loss": 0.5379,
+      "step": 748
+    },
+    {
+      "epoch": 1.0460358056265984,
+      "grad_norm": 1.9967021838933243,
+      "learning_rate": 8.238426154810035e-06,
+      "loss": 0.4514,
+      "step": 749
+    },
+    {
+      "epoch": 1.0474308300395256,
+      "grad_norm": 2.0039576329601037,
+      "learning_rate": 8.232230456161819e-06,
+      "loss": 0.4457,
+      "step": 750
+    },
+    {
+      "epoch": 1.048825854452453,
+      "grad_norm": 2.12340999999131,
+      "learning_rate": 8.226026219865313e-06,
+      "loss": 0.509,
+      "step": 751
+    },
+    {
+      "epoch": 1.0502208788653802,
+      "grad_norm": 2.082417711494918,
+      "learning_rate": 8.219813462308458e-06,
+      "loss": 0.4752,
+      "step": 752
+    },
+    {
+      "epoch": 1.0516159032783075,
+      "grad_norm": 2.1077205247643156,
+      "learning_rate": 8.213592199901692e-06,
+      "loss": 0.4321,
+      "step": 753
+    },
+    {
+      "epoch": 1.0530109276912345,
+      "grad_norm": 1.919259142002624,
+      "learning_rate": 8.207362449077932e-06,
+      "loss": 0.4702,
+      "step": 754
+    },
+    {
+      "epoch": 1.0544059521041618,
+      "grad_norm": 2.0585541043039646,
+      "learning_rate": 8.201124226292505e-06,
+      "loss": 0.4796,
+      "step": 755
+    },
+    {
+      "epoch": 1.055800976517089,
+      "grad_norm": 1.981427974951629,
+      "learning_rate": 8.19487754802312e-06,
+      "loss": 0.4664,
+      "step": 756
+    },
+    {
+      "epoch": 1.0571960009300163,
+      "grad_norm": 2.2319241325523786,
+      "learning_rate": 8.18862243076982e-06,
+      "loss": 0.475,
+      "step": 757
+    },
+    {
+      "epoch": 1.0585910253429436,
+      "grad_norm": 2.039055467549461,
+      "learning_rate": 8.18235889105494e-06,
+      "loss": 0.4312,
+      "step": 758
+    },
+    {
+      "epoch": 1.0599860497558706,
+      "grad_norm": 2.0535980015454856,
+      "learning_rate": 8.17608694542306e-06,
+      "loss": 0.5021,
+      "step": 759
+    },
+    {
+      "epoch": 1.061381074168798,
+      "grad_norm": 2.0117098685514128,
+      "learning_rate": 8.169806610440966e-06,
+      "loss": 0.4974,
+      "step": 760
+    },
+    {
+      "epoch": 1.0627760985817252,
+      "grad_norm": 2.123763035186518,
+      "learning_rate": 8.163517902697602e-06,
+      "loss": 0.5005,
+      "step": 761
+    },
+    {
+      "epoch": 1.0641711229946524,
+      "grad_norm": 2.0459984276686622,
+      "learning_rate": 8.157220838804026e-06,
+      "loss": 0.5113,
+      "step": 762
+    },
+    {
+      "epoch": 1.0655661474075797,
+      "grad_norm": 2.086483588594984,
+      "learning_rate": 8.150915435393371e-06,
+      "loss": 0.475,
+      "step": 763
+    },
+    {
+      "epoch": 1.0669611718205068,
+      "grad_norm": 2.0259222979461917,
+      "learning_rate": 8.1446017091208e-06,
+      "loss": 0.4294,
+      "step": 764
+    },
+    {
+      "epoch": 1.068356196233434,
+      "grad_norm": 1.939016863108113,
+      "learning_rate": 8.138279676663458e-06,
+      "loss": 0.5281,
+      "step": 765
+    },
+    {
+      "epoch": 1.0697512206463613,
+      "grad_norm": 2.188159597757035,
+      "learning_rate": 8.131949354720425e-06,
+      "loss": 0.4351,
+      "step": 766
+    },
+    {
+      "epoch": 1.0711462450592886,
+      "grad_norm": 1.9805506282461658,
+      "learning_rate": 8.125610760012685e-06,
+      "loss": 0.4761,
+      "step": 767
+    },
+    {
+      "epoch": 1.0725412694722158,
+      "grad_norm": 1.993266574219508,
+      "learning_rate": 8.11926390928307e-06,
+      "loss": 0.5198,
+      "step": 768
+    },
+    {
+      "epoch": 1.073936293885143,
+      "grad_norm": 2.051047444269424,
+      "learning_rate": 8.112908819296217e-06,
+      "loss": 0.4673,
+      "step": 769
+    },
+    {
+      "epoch": 1.0753313182980702,
+      "grad_norm": 2.017405300453376,
+      "learning_rate": 8.106545506838533e-06,
+      "loss": 0.4968,
+      "step": 770
+    },
+    {
+      "epoch": 1.0767263427109974,
+      "grad_norm": 1.926246923060768,
+      "learning_rate": 8.100173988718136e-06,
+      "loss": 0.4685,
+      "step": 771
+    },
+    {
+      "epoch": 1.0781213671239247,
+      "grad_norm": 2.0051822197441838,
+      "learning_rate": 8.093794281764824e-06,
+      "loss": 0.4998,
+      "step": 772
+    },
+    {
+      "epoch": 1.079516391536852,
+      "grad_norm": 2.192605416056662,
+      "learning_rate": 8.087406402830026e-06,
+      "loss": 0.4681,
+      "step": 773
+    },
+    {
+      "epoch": 1.080911415949779,
+      "grad_norm": 2.03492190725647,
+      "learning_rate": 8.081010368786751e-06,
+      "loss": 0.4774,
+      "step": 774
+    },
+    {
+      "epoch": 1.0823064403627063,
+      "grad_norm": 1.8916433025293278,
+      "learning_rate": 8.074606196529554e-06,
+      "loss": 0.4679,
+      "step": 775
+    },
+    {
+      "epoch": 1.0837014647756336,
+      "grad_norm": 2.029167018400251,
+      "learning_rate": 8.068193902974482e-06,
+      "loss": 0.471,
+      "step": 776
+    },
+    {
+      "epoch": 1.0850964891885608,
+      "grad_norm": 2.0278286303865816,
+      "learning_rate": 8.06177350505904e-06,
+      "loss": 0.437,
+      "step": 777
+    },
+    {
+      "epoch": 1.086491513601488,
+      "grad_norm": 2.1953587700733186,
+      "learning_rate": 8.055345019742133e-06,
+      "loss": 0.4578,
+      "step": 778
+    },
+    {
+      "epoch": 1.0878865380144151,
+      "grad_norm": 2.0927543689762023,
+      "learning_rate": 8.048908464004032e-06,
+      "loss": 0.5168,
+      "step": 779
+    },
+    {
+      "epoch": 1.0892815624273424,
+      "grad_norm": 2.250143542210799,
+      "learning_rate": 8.042463854846325e-06,
+      "loss": 0.4474,
+      "step": 780
+    },
+    {
+      "epoch": 1.0906765868402697,
+      "grad_norm": 2.0315426491546056,
+      "learning_rate": 8.036011209291872e-06,
+      "loss": 0.4701,
+      "step": 781
+    },
+    {
+      "epoch": 1.092071611253197,
+      "grad_norm": 2.189955370940604,
+      "learning_rate": 8.029550544384758e-06,
+      "loss": 0.4567,
+      "step": 782
+    },
+    {
+      "epoch": 1.0934666356661242,
+      "grad_norm": 2.098770628449276,
+      "learning_rate": 8.023081877190257e-06,
+      "loss": 0.5319,
+      "step": 783
+    },
+    {
+      "epoch": 1.0948616600790513,
+      "grad_norm": 2.2036959427181393,
+      "learning_rate": 8.016605224794773e-06,
+      "loss": 0.4803,
+      "step": 784
+    },
+    {
+      "epoch": 1.0962566844919786,
+      "grad_norm": 2.061217417191479,
+      "learning_rate": 8.010120604305806e-06,
+      "loss": 0.464,
+      "step": 785
+    },
+    {
+      "epoch": 1.0976517089049058,
+      "grad_norm": 2.0770104851414857,
+      "learning_rate": 8.003628032851904e-06,
+      "loss": 0.477,
+      "step": 786
+    },
+    {
+      "epoch": 1.099046733317833,
+      "grad_norm": 2.0452867521427107,
+      "learning_rate": 7.997127527582613e-06,
+      "loss": 0.4644,
+      "step": 787
+    },
+    {
+      "epoch": 1.1004417577307604,
+      "grad_norm": 2.5829534069096747,
+      "learning_rate": 7.990619105668437e-06,
+      "loss": 0.4444,
+      "step": 788
+    },
+    {
+      "epoch": 1.1018367821436876,
+      "grad_norm": 1.7402633467783815,
+      "learning_rate": 7.984102784300794e-06,
+      "loss": 0.4822,
+      "step": 789
+    },
+    {
+      "epoch": 1.1032318065566147,
+      "grad_norm": 2.015841114887429,
+      "learning_rate": 7.977578580691963e-06,
+      "loss": 0.4916,
+      "step": 790
+    },
+    {
+      "epoch": 1.104626830969542,
+      "grad_norm": 1.9106009742536385,
+      "learning_rate": 7.971046512075047e-06,
+      "loss": 0.4409,
+      "step": 791
+    },
+    {
+      "epoch": 1.1060218553824692,
+      "grad_norm": 2.05674016861173,
+      "learning_rate": 7.964506595703923e-06,
+      "loss": 0.4565,
+      "step": 792
+    },
+    {
+      "epoch": 1.1074168797953965,
+      "grad_norm": 1.9348516245501983,
+      "learning_rate": 7.957958848853193e-06,
+      "loss": 0.5205,
+      "step": 793
+    },
+    {
+      "epoch": 1.1088119042083235,
+      "grad_norm": 2.199190699063569,
+      "learning_rate": 7.95140328881815e-06,
+      "loss": 0.5013,
+      "step": 794
+    },
+    {
+      "epoch": 1.1102069286212508,
+      "grad_norm": 2.0083712618435277,
+      "learning_rate": 7.944839932914718e-06,
+      "loss": 0.4876,
+      "step": 795
+    },
+    {
+      "epoch": 1.111601953034178,
+      "grad_norm": 2.1922894266968673,
+      "learning_rate": 7.938268798479419e-06,
+      "loss": 0.4816,
+      "step": 796
+    },
+    {
+      "epoch": 1.1129969774471054,
+      "grad_norm": 1.901000847965792,
+      "learning_rate": 7.931689902869314e-06,
+      "loss": 0.5018,
+      "step": 797
+    },
+    {
+      "epoch": 1.1143920018600326,
+      "grad_norm": 2.222074221286817,
+      "learning_rate": 7.925103263461971e-06,
+      "loss": 0.4994,
+      "step": 798
+    },
+    {
+      "epoch": 1.11578702627296,
+      "grad_norm": 1.940812874546972,
+      "learning_rate": 7.91850889765541e-06,
+      "loss": 0.4576,
+      "step": 799
+    },
+    {
+      "epoch": 1.117182050685887,
+      "grad_norm": 2.0817228977455953,
+      "learning_rate": 7.91190682286806e-06,
+      "loss": 0.4388,
+      "step": 800
+    },
+    {
+      "epoch": 1.1185770750988142,
+      "grad_norm": 2.1179633199248316,
+      "learning_rate": 7.905297056538713e-06,
+      "loss": 0.4652,
+      "step": 801
+    },
+    {
+      "epoch": 1.1199720995117415,
+      "grad_norm": 1.9607161065971666,
+      "learning_rate": 7.898679616126474e-06,
+      "loss": 0.4803,
+      "step": 802
+    },
+    {
+      "epoch": 1.1213671239246688,
+      "grad_norm": 1.9838680808121743,
+      "learning_rate": 7.892054519110726e-06,
+      "loss": 0.4878,
+      "step": 803
+    },
+    {
+      "epoch": 1.1227621483375958,
+      "grad_norm": 2.107089413069441,
+      "learning_rate": 7.885421782991064e-06,
+      "loss": 0.5175,
+      "step": 804
+    },
+    {
+      "epoch": 1.124157172750523,
+      "grad_norm": 2.0093683502992152,
+      "learning_rate": 7.878781425287277e-06,
+      "loss": 0.446,
+      "step": 805
+    },
+    {
+      "epoch": 1.1255521971634503,
+      "grad_norm": 1.9616515420321192,
+      "learning_rate": 7.872133463539274e-06,
+      "loss": 0.4333,
+      "step": 806
+    },
+    {
+      "epoch": 1.1269472215763776,
+      "grad_norm": 2.001100054062502,
+      "learning_rate": 7.86547791530705e-06,
+      "loss": 0.5393,
+      "step": 807
+    },
+    {
+      "epoch": 1.1283422459893049,
+      "grad_norm": 1.9580216968373765,
+      "learning_rate": 7.858814798170644e-06,
+      "loss": 0.492,
+      "step": 808
+    },
+    {
+      "epoch": 1.1297372704022322,
+      "grad_norm": 2.025423359557969,
+      "learning_rate": 7.852144129730087e-06,
+      "loss": 0.4743,
+      "step": 809
+    },
+    {
+      "epoch": 1.1311322948151592,
+      "grad_norm": 2.218171379188175,
+      "learning_rate": 7.84546592760535e-06,
+      "loss": 0.489,
+      "step": 810
+    },
+    {
+      "epoch": 1.1325273192280865,
+      "grad_norm": 1.9026069806809947,
+      "learning_rate": 7.83878020943631e-06,
+      "loss": 0.5161,
+      "step": 811
+    },
+    {
+      "epoch": 1.1339223436410137,
+      "grad_norm": 2.1794929058113146,
+      "learning_rate": 7.832086992882697e-06,
+      "loss": 0.4831,
+      "step": 812
+    },
+    {
+      "epoch": 1.135317368053941,
+      "grad_norm": 1.7960116856711474,
+      "learning_rate": 7.825386295624043e-06,
+      "loss": 0.4499,
+      "step": 813
+    },
+    {
+      "epoch": 1.136712392466868,
+      "grad_norm": 2.0247282315582376,
+      "learning_rate": 7.818678135359641e-06,
+      "loss": 0.4982,
+      "step": 814
+    },
+    {
+      "epoch": 1.1381074168797953,
+      "grad_norm": 2.0547296042290792,
+      "learning_rate": 7.811962529808499e-06,
+      "loss": 0.477,
+      "step": 815
+    },
+    {
+      "epoch": 1.1395024412927226,
+      "grad_norm": 2.0574834628921446,
+      "learning_rate": 7.805239496709291e-06,
+      "loss": 0.4546,
+      "step": 816
+    },
+    {
+      "epoch": 1.1408974657056499,
+      "grad_norm": 1.9394156746898406,
+      "learning_rate": 7.798509053820305e-06,
+      "loss": 0.4439,
+      "step": 817
+    },
+    {
+      "epoch": 1.1422924901185771,
+      "grad_norm": 2.0254153845417733,
+      "learning_rate": 7.79177121891941e-06,
+      "loss": 0.4619,
+      "step": 818
+    },
+    {
+      "epoch": 1.1436875145315044,
+      "grad_norm": 2.096601881606947,
+      "learning_rate": 7.785026009803993e-06,
+      "loss": 0.5061,
+      "step": 819
+    },
+    {
+      "epoch": 1.1450825389444315,
+      "grad_norm": 2.262505975665749,
+      "learning_rate": 7.778273444290921e-06,
+      "loss": 0.4799,
+      "step": 820
+    },
+    {
+      "epoch": 1.1464775633573587,
+      "grad_norm": 2.119118003205617,
+      "learning_rate": 7.771513540216496e-06,
+      "loss": 0.4928,
+      "step": 821
+    },
+    {
+      "epoch": 1.147872587770286,
+      "grad_norm": 2.133703419481106,
+      "learning_rate": 7.764746315436399e-06,
+      "loss": 0.4542,
+      "step": 822
+    },
+    {
+      "epoch": 1.1492676121832133,
+      "grad_norm": 2.070568845440914,
+      "learning_rate": 7.75797178782565e-06,
+      "loss": 0.4523,
+      "step": 823
+    },
+    {
+      "epoch": 1.1506626365961403,
+      "grad_norm": 1.7991652992094438,
+      "learning_rate": 7.751189975278561e-06,
+      "loss": 0.4077,
+      "step": 824
+    },
+    {
+      "epoch": 1.1520576610090676,
+      "grad_norm": 2.038280747930084,
+      "learning_rate": 7.744400895708683e-06,
+      "loss": 0.4647,
+      "step": 825
+    },
+    {
+      "epoch": 1.1534526854219949,
+      "grad_norm": 2.0602555390170316,
+      "learning_rate": 7.737604567048766e-06,
+      "loss": 0.5507,
+      "step": 826
+    },
+    {
+      "epoch": 1.1548477098349221,
+      "grad_norm": 2.180790828248436,
+      "learning_rate": 7.730801007250704e-06,
+      "loss": 0.4655,
+      "step": 827
+    },
+    {
+      "epoch": 1.1562427342478494,
+      "grad_norm": 2.0458938855615605,
+      "learning_rate": 7.72399023428549e-06,
+      "loss": 0.4643,
+      "step": 828
+    },
+    {
+      "epoch": 1.1576377586607767,
+      "grad_norm": 1.9127629534222181,
+      "learning_rate": 7.717172266143178e-06,
+      "loss": 0.5365,
+      "step": 829
+    },
+    {
+      "epoch": 1.1590327830737037,
+      "grad_norm": 2.209826641763816,
+      "learning_rate": 7.710347120832821e-06,
+      "loss": 0.4545,
+      "step": 830
+    },
+    {
+      "epoch": 1.160427807486631,
+      "grad_norm": 2.010738480822612,
+      "learning_rate": 7.703514816382432e-06,
+      "loss": 0.4423,
+      "step": 831
+    },
+    {
+      "epoch": 1.1618228318995583,
+      "grad_norm": 2.0297966468628457,
+      "learning_rate": 7.696675370838929e-06,
+      "loss": 0.4518,
+      "step": 832
+    },
+    {
+      "epoch": 1.1632178563124855,
+      "grad_norm": 2.0826482950087226,
+      "learning_rate": 7.689828802268102e-06,
+      "loss": 0.4533,
+      "step": 833
+    },
+    {
+      "epoch": 1.1646128807254126,
+      "grad_norm": 1.8008744406646637,
+      "learning_rate": 7.682975128754548e-06,
+      "loss": 0.451,
+      "step": 834
+    },
+    {
+      "epoch": 1.1660079051383399,
+      "grad_norm": 2.0565446537269194,
+      "learning_rate": 7.676114368401635e-06,
+      "loss": 0.5319,
+      "step": 835
+    },
+    {
+      "epoch": 1.1674029295512671,
+      "grad_norm": 2.1287025934347086,
+      "learning_rate": 7.66924653933145e-06,
+      "loss": 0.4858,
+      "step": 836
+    },
+    {
+      "epoch": 1.1687979539641944,
+      "grad_norm": 2.0809030253570944,
+      "learning_rate": 7.662371659684749e-06,
+      "loss": 0.508,
+      "step": 837
+    },
+    {
+      "epoch": 1.1701929783771217,
+      "grad_norm": 2.091119136946387,
+      "learning_rate": 7.655489747620913e-06,
+      "loss": 0.4702,
+      "step": 838
+    },
+    {
+      "epoch": 1.171588002790049,
+      "grad_norm": 2.0095956535575987,
+      "learning_rate": 7.648600821317901e-06,
+      "loss": 0.5005,
+      "step": 839
+    },
+    {
+      "epoch": 1.172983027202976,
+      "grad_norm": 2.108702760839066,
+      "learning_rate": 7.641704898972194e-06,
+      "loss": 0.4931,
+      "step": 840
+    },
+    {
+      "epoch": 1.1743780516159033,
+      "grad_norm": 2.1002645954174577,
+      "learning_rate": 7.634801998798755e-06,
+      "loss": 0.5118,
+      "step": 841
+    },
+    {
+      "epoch": 1.1757730760288305,
+      "grad_norm": 2.060993512560868,
+      "learning_rate": 7.6278921390309834e-06,
+      "loss": 0.4357,
+      "step": 842
+    },
+    {
+      "epoch": 1.1771681004417578,
+      "grad_norm": 1.8482022743666582,
+      "learning_rate": 7.620975337920653e-06,
+      "loss": 0.4732,
+      "step": 843
+    },
+    {
+      "epoch": 1.1785631248546848,
+      "grad_norm": 2.0436494232781435,
+      "learning_rate": 7.6140516137378786e-06,
+      "loss": 0.4204,
+      "step": 844
+    },
+    {
+      "epoch": 1.179958149267612,
+      "grad_norm": 1.916148888628701,
+      "learning_rate": 7.607120984771058e-06,
+      "loss": 0.4595,
+      "step": 845
+    },
+    {
+      "epoch": 1.1813531736805394,
+      "grad_norm": 1.971942182009117,
+      "learning_rate": 7.600183469326829e-06,
+      "loss": 0.464,
+      "step": 846
+    },
+    {
+      "epoch": 1.1827481980934667,
+      "grad_norm": 2.0099670391093576,
+      "learning_rate": 7.593239085730022e-06,
+      "loss": 0.4997,
+      "step": 847
+    },
+    {
+      "epoch": 1.184143222506394,
+      "grad_norm": 2.364610588530083,
+      "learning_rate": 7.586287852323605e-06,
+      "loss": 0.511,
+      "step": 848
+    },
+    {
+      "epoch": 1.1855382469193212,
+      "grad_norm": 2.0980427797891323,
+      "learning_rate": 7.579329787468639e-06,
+      "loss": 0.4545,
+      "step": 849
+    },
+    {
+      "epoch": 1.1869332713322482,
+      "grad_norm": 2.0688493550588403,
+      "learning_rate": 7.572364909544235e-06,
+      "loss": 0.5001,
+      "step": 850
+    },
+    {
+      "epoch": 1.1883282957451755,
+      "grad_norm": 1.8729866135500641,
+      "learning_rate": 7.565393236947494e-06,
+      "loss": 0.4338,
+      "step": 851
+    },
+    {
+      "epoch": 1.1897233201581028,
+      "grad_norm": 2.168283835304086,
+      "learning_rate": 7.558414788093467e-06,
+      "loss": 0.4762,
+      "step": 852
+    },
+    {
+      "epoch": 1.19111834457103,
+      "grad_norm": 2.066651188721128,
+      "learning_rate": 7.551429581415104e-06,
+      "loss": 0.5172,
+      "step": 853
+    },
+    {
+      "epoch": 1.192513368983957,
+      "grad_norm": 2.1920591445911364,
+      "learning_rate": 7.5444376353632064e-06,
+      "loss": 0.4567,
+      "step": 854
+    },
+    {
+      "epoch": 1.1939083933968844,
+      "grad_norm": 1.831321422970195,
+      "learning_rate": 7.537438968406372e-06,
+      "loss": 0.491,
+      "step": 855
+    },
+    {
+      "epoch": 1.1953034178098116,
+      "grad_norm": 2.0968609844910056,
+      "learning_rate": 7.530433599030962e-06,
+      "loss": 0.4656,
+      "step": 856
+    },
+    {
+      "epoch": 1.196698442222739,
+      "grad_norm": 2.1422512387969994,
+      "learning_rate": 7.5234215457410255e-06,
+      "loss": 0.5081,
+      "step": 857
+    },
+    {
+      "epoch": 1.1980934666356662,
+      "grad_norm": 2.2362083848047596,
+      "learning_rate": 7.516402827058283e-06,
+      "loss": 0.5361,
+      "step": 858
+    },
+    {
+      "epoch": 1.1994884910485935,
+      "grad_norm": 1.9286458341926276,
+      "learning_rate": 7.509377461522049e-06,
+      "loss": 0.4805,
+      "step": 859
+    },
+    {
+      "epoch": 1.2008835154615205,
+      "grad_norm": 2.3123734766029527,
+      "learning_rate": 7.502345467689202e-06,
+      "loss": 0.5048,
+      "step": 860
+    },
+    {
+      "epoch": 1.2022785398744478,
+      "grad_norm": 2.116964583062056,
+      "learning_rate": 7.4953068641341255e-06,
+      "loss": 0.4499,
+      "step": 861
+    },
+    {
+      "epoch": 1.203673564287375,
+      "grad_norm": 2.1029489962024144,
+      "learning_rate": 7.488261669448662e-06,
+      "loss": 0.5238,
+      "step": 862
+    },
+    {
+      "epoch": 1.2050685887003023,
+      "grad_norm": 2.0626515202597777,
+      "learning_rate": 7.4812099022420636e-06,
+      "loss": 0.4624,
+      "step": 863
+    },
+    {
+      "epoch": 1.2064636131132294,
+      "grad_norm": 2.004582769305792,
+      "learning_rate": 7.474151581140947e-06,
+      "loss": 0.4733,
+      "step": 864
+    },
+    {
+      "epoch": 1.2078586375261566,
+      "grad_norm": 2.1127969533998843,
+      "learning_rate": 7.4670867247892346e-06,
+      "loss": 0.4704,
+      "step": 865
+    },
+    {
+      "epoch": 1.209253661939084,
+      "grad_norm": 1.8514576184123828,
+      "learning_rate": 7.460015351848115e-06,
+      "loss": 0.4825,
+      "step": 866
+    },
+    {
+      "epoch": 1.2106486863520112,
+      "grad_norm": 2.078998051567728,
+      "learning_rate": 7.4529374809959895e-06,
+      "loss": 0.4166,
+      "step": 867
+    },
+    {
+      "epoch": 1.2120437107649384,
+      "grad_norm": 1.9641129542739804,
+      "learning_rate": 7.445853130928422e-06,
+      "loss": 0.5042,
+      "step": 868
+    },
+    {
+      "epoch": 1.2134387351778657,
+      "grad_norm": 2.2654120011022414,
+      "learning_rate": 7.438762320358089e-06,
+      "loss": 0.5175,
+      "step": 869
+    },
+    {
+      "epoch": 1.2148337595907928,
+      "grad_norm": 2.0429675987113134,
+      "learning_rate": 7.431665068014737e-06,
+      "loss": 0.4706,
+      "step": 870
+    },
+    {
+      "epoch": 1.21622878400372,
+      "grad_norm": 1.982257979787058,
+      "learning_rate": 7.424561392645122e-06,
+      "loss": 0.4706,
+      "step": 871
+    },
+    {
+      "epoch": 1.2176238084166473,
+      "grad_norm": 2.076586625001963,
+      "learning_rate": 7.417451313012971e-06,
+      "loss": 0.4895,
+      "step": 872
+    },
+    {
+      "epoch": 1.2190188328295746,
+      "grad_norm": 2.173072442479642,
+      "learning_rate": 7.410334847898921e-06,
+      "loss": 0.5118,
+      "step": 873
+    },
+    {
+      "epoch": 1.2204138572425016,
+      "grad_norm": 2.047041786534623,
+      "learning_rate": 7.403212016100484e-06,
+      "loss": 0.463,
+      "step": 874
+    },
+    {
+      "epoch": 1.221808881655429,
+      "grad_norm": 2.011805245547945,
+      "learning_rate": 7.396082836431981e-06,
+      "loss": 0.4624,
+      "step": 875
+    },
+    {
+      "epoch": 1.2232039060683562,
+      "grad_norm": 2.046453837182149,
+      "learning_rate": 7.388947327724506e-06,
+      "loss": 0.4503,
+      "step": 876
+    },
+    {
+      "epoch": 1.2245989304812834,
+      "grad_norm": 2.0074523324744233,
+      "learning_rate": 7.3818055088258676e-06,
+      "loss": 0.465,
+      "step": 877
+    },
+    {
+      "epoch": 1.2259939548942107,
+      "grad_norm": 2.15744694552415,
+      "learning_rate": 7.374657398600542e-06,
+      "loss": 0.4489,
+      "step": 878
+    },
+    {
+      "epoch": 1.227388979307138,
+      "grad_norm": 2.010685773385559,
+      "learning_rate": 7.367503015929627e-06,
+      "loss": 0.4692,
+      "step": 879
+    },
+    {
+      "epoch": 1.228784003720065,
+      "grad_norm": 2.0352429043175055,
+      "learning_rate": 7.3603423797107845e-06,
+      "loss": 0.4546,
+      "step": 880
+    },
+    {
+      "epoch": 1.2301790281329923,
+      "grad_norm": 2.1476461439581245,
+      "learning_rate": 7.353175508858195e-06,
+      "loss": 0.4962,
+      "step": 881
+    },
+    {
+      "epoch": 1.2315740525459196,
+      "grad_norm": 2.1997937220709405,
+      "learning_rate": 7.3460024223025095e-06,
+      "loss": 0.5091,
+      "step": 882
+    },
+    {
+      "epoch": 1.2329690769588468,
+      "grad_norm": 1.666080626807778,
+      "learning_rate": 7.338823138990796e-06,
+      "loss": 0.4352,
+      "step": 883
+    },
+    {
+      "epoch": 1.234364101371774,
+      "grad_norm": 2.06513419711823,
+      "learning_rate": 7.33163767788649e-06,
+      "loss": 0.4584,
+      "step": 884
+    },
+    {
+      "epoch": 1.2357591257847012,
+      "grad_norm": 2.0063814585695123,
+      "learning_rate": 7.324446057969346e-06,
+      "loss": 0.5118,
+      "step": 885
+    },
+    {
+      "epoch": 1.2371541501976284,
+      "grad_norm": 2.136466810820014,
+      "learning_rate": 7.317248298235387e-06,
+      "loss": 0.4905,
+      "step": 886
+    },
+    {
+      "epoch": 1.2385491746105557,
+      "grad_norm": 2.005738467998993,
+      "learning_rate": 7.3100444176968514e-06,
+      "loss": 0.4471,
+      "step": 887
+    },
+    {
+      "epoch": 1.239944199023483,
+      "grad_norm": 2.1451364094879155,
+      "learning_rate": 7.302834435382147e-06,
+      "loss": 0.5487,
+      "step": 888
+    },
+    {
+      "epoch": 1.2413392234364102,
+      "grad_norm": 2.121313221799734,
+      "learning_rate": 7.2956183703358e-06,
+      "loss": 0.431,
+      "step": 889
+    },
+    {
+      "epoch": 1.2427342478493373,
+      "grad_norm": 1.9010978340983133,
+      "learning_rate": 7.288396241618401e-06,
+      "loss": 0.4701,
+      "step": 890
+    },
+    {
+      "epoch": 1.2441292722622646,
+      "grad_norm": 2.1243540704544133,
+      "learning_rate": 7.281168068306559e-06,
+      "loss": 0.5113,
+      "step": 891
+    },
+    {
+      "epoch": 1.2455242966751918,
+      "grad_norm": 1.9858608436318579,
+      "learning_rate": 7.2739338694928485e-06,
+      "loss": 0.4278,
+      "step": 892
+    },
+    {
+      "epoch": 1.246919321088119,
+      "grad_norm": 2.0319961298094404,
+      "learning_rate": 7.266693664285761e-06,
+      "loss": 0.4874,
+      "step": 893
+    },
+    {
+      "epoch": 1.2483143455010464,
+      "grad_norm": 2.1329652725825294,
+      "learning_rate": 7.259447471809651e-06,
+      "loss": 0.4732,
+      "step": 894
+    },
+    {
+      "epoch": 1.2497093699139734,
+      "grad_norm": 1.8887015726470242,
+      "learning_rate": 7.252195311204689e-06,
+      "loss": 0.4973,
+      "step": 895
+    },
+    {
+      "epoch": 1.2511043943269007,
+      "grad_norm": 2.2465835446589484,
+      "learning_rate": 7.244937201626812e-06,
+      "loss": 0.5229,
+      "step": 896
+    },
+    {
+      "epoch": 1.252499418739828,
+      "grad_norm": 2.137057584376221,
+      "learning_rate": 7.237673162247667e-06,
+      "loss": 0.5224,
+      "step": 897
+    },
+    {
+      "epoch": 1.2538944431527552,
+      "grad_norm": 2.165616318347168,
+      "learning_rate": 7.230403212254566e-06,
+      "loss": 0.4415,
+      "step": 898
+    },
+    {
+      "epoch": 1.2552894675656825,
+      "grad_norm": 1.9723542482886345,
+      "learning_rate": 7.223127370850433e-06,
+      "loss": 0.4883,
+      "step": 899
+    },
+    {
+      "epoch": 1.2566844919786098,
+      "grad_norm": 2.0860634186365123,
+      "learning_rate": 7.215845657253755e-06,
+      "loss": 0.4966,
+      "step": 900
+    },
+    {
+      "epoch": 1.2580795163915368,
+      "grad_norm": 2.0929823854721645,
+      "learning_rate": 7.208558090698528e-06,
+      "loss": 0.4694,
+      "step": 901
+    },
+    {
+      "epoch": 1.259474540804464,
+      "grad_norm": 2.0668298861598893,
+      "learning_rate": 7.2012646904342065e-06,
+      "loss": 0.514,
+      "step": 902
+    },
+    {
+      "epoch": 1.2608695652173914,
+      "grad_norm": 2.15111282561836,
+      "learning_rate": 7.193965475725659e-06,
+      "loss": 0.4687,
+      "step": 903
+    },
+    {
+      "epoch": 1.2622645896303184,
+      "grad_norm": 1.9150017426541963,
+      "learning_rate": 7.186660465853111e-06,
+      "loss": 0.4807,
+      "step": 904
+    },
+    {
+      "epoch": 1.2636596140432457,
+      "grad_norm": 2.047097651427044,
+      "learning_rate": 7.1793496801120885e-06,
+      "loss": 0.4753,
+      "step": 905
+    },
+    {
+      "epoch": 1.265054638456173,
+      "grad_norm": 2.0657264775450495,
+      "learning_rate": 7.172033137813387e-06,
+      "loss": 0.4814,
+      "step": 906
+    },
+    {
+      "epoch": 1.2664496628691002,
+      "grad_norm": 2.191648239680036,
+      "learning_rate": 7.1647108582829924e-06,
+      "loss": 0.512,
+      "step": 907
+    },
+    {
+      "epoch": 1.2678446872820275,
+      "grad_norm": 2.1481439685301,
+      "learning_rate": 7.157382860862059e-06,
+      "loss": 0.4949,
+      "step": 908
+    },
+    {
+      "epoch": 1.2692397116949548,
+      "grad_norm": 2.0193130576657436,
+      "learning_rate": 7.1500491649068345e-06,
+      "loss": 0.4745,
+      "step": 909
+    },
+    {
+      "epoch": 1.270634736107882,
+      "grad_norm": 2.152032552208028,
+      "learning_rate": 7.1427097897886225e-06,
+      "loss": 0.4649,
+      "step": 910
+    },
+    {
+      "epoch": 1.272029760520809,
+      "grad_norm": 1.79491248450368,
+      "learning_rate": 7.135364754893729e-06,
+      "loss": 0.4824,
+      "step": 911
+    },
+    {
+      "epoch": 1.2734247849337363,
+      "grad_norm": 2.006219248252164,
+      "learning_rate": 7.128014079623408e-06,
+      "loss": 0.4683,
+      "step": 912
+    },
+    {
+      "epoch": 1.2748198093466636,
+      "grad_norm": 2.1167631241450833,
+      "learning_rate": 7.120657783393809e-06,
+      "loss": 0.4544,
+      "step": 913
+    },
+    {
+      "epoch": 1.2762148337595907,
+      "grad_norm": 1.9858989311000603,
+      "learning_rate": 7.113295885635936e-06,
+      "loss": 0.463,
+      "step": 914
+    },
+    {
+      "epoch": 1.277609858172518,
+      "grad_norm": 2.033452279845351,
+      "learning_rate": 7.105928405795584e-06,
+      "loss": 0.5281,
+      "step": 915
+    },
+    {
+      "epoch": 1.2790048825854452,
+      "grad_norm": 2.0872449768803123,
+      "learning_rate": 7.098555363333289e-06,
+      "loss": 0.4642,
+      "step": 916
+    },
+    {
+      "epoch": 1.2803999069983725,
+      "grad_norm": 2.04826654738253,
+      "learning_rate": 7.091176777724291e-06,
+      "loss": 0.4912,
+      "step": 917
+    },
+    {
+      "epoch": 1.2817949314112997,
+      "grad_norm": 2.067550326057046,
+      "learning_rate": 7.083792668458463e-06,
+      "loss": 0.4561,
+      "step": 918
+    },
+    {
+      "epoch": 1.283189955824227,
+      "grad_norm": 2.012007521941956,
+      "learning_rate": 7.076403055040271e-06,
+      "loss": 0.4612,
+      "step": 919
+    },
+    {
+      "epoch": 1.2845849802371543,
+      "grad_norm": 2.0098154320853534,
+      "learning_rate": 7.069007956988718e-06,
+      "loss": 0.4952,
+      "step": 920
+    },
+    {
+      "epoch": 1.2859800046500813,
+      "grad_norm": 2.0434296920508945,
+      "learning_rate": 7.061607393837295e-06,
+      "loss": 0.5048,
+      "step": 921
+    },
+    {
+      "epoch": 1.2873750290630086,
+      "grad_norm": 2.0366764578127543,
+      "learning_rate": 7.0542013851339316e-06,
+      "loss": 0.4098,
+      "step": 922
+    },
+    {
+      "epoch": 1.2887700534759359,
+      "grad_norm": 1.9173857646730377,
+      "learning_rate": 7.04678995044094e-06,
+      "loss": 0.4423,
+      "step": 923
+    },
+    {
+      "epoch": 1.290165077888863,
+      "grad_norm": 1.9666733803958376,
+      "learning_rate": 7.039373109334957e-06,
+      "loss": 0.4782,
+      "step": 924
+    },
+    {
+      "epoch": 1.2915601023017902,
+      "grad_norm": 2.0178632072627756,
+      "learning_rate": 7.031950881406913e-06,
+      "loss": 0.4829,
+      "step": 925
+    },
+    {
+      "epoch": 1.2929551267147175,
+      "grad_norm": 2.099833553417301,
+      "learning_rate": 7.024523286261959e-06,
+      "loss": 0.447,
+      "step": 926
+    },
+    {
+      "epoch": 1.2943501511276447,
+      "grad_norm": 1.953265508839639,
+      "learning_rate": 7.017090343519421e-06,
+      "loss": 0.4818,
+      "step": 927
+    },
+    {
+      "epoch": 1.295745175540572,
+      "grad_norm": 2.1071393237014484,
+      "learning_rate": 7.009652072812758e-06,
+      "loss": 0.4478,
+      "step": 928
+    },
+    {
+      "epoch": 1.2971401999534993,
+      "grad_norm": 1.9409046297212447,
+      "learning_rate": 7.0022084937895e-06,
+      "loss": 0.4926,
+      "step": 929
+    },
+    {
+      "epoch": 1.2985352243664265,
+      "grad_norm": 2.157406168167787,
+      "learning_rate": 6.994759626111189e-06,
+      "loss": 0.4971,
+      "step": 930
+    },
+    {
+      "epoch": 1.2999302487793536,
+      "grad_norm": 2.1435260313512723,
+      "learning_rate": 6.987305489453352e-06,
+      "loss": 0.5317,
+      "step": 931
+    },
+    {
+      "epoch": 1.3013252731922809,
+      "grad_norm": 2.178710662283696,
+      "learning_rate": 6.979846103505423e-06,
+      "loss": 0.4975,
+      "step": 932
+    },
+    {
+      "epoch": 1.3027202976052081,
+      "grad_norm": 1.9560199347527047,
+      "learning_rate": 6.972381487970702e-06,
+      "loss": 0.4461,
+      "step": 933
+    },
+    {
+      "epoch": 1.3041153220181352,
+      "grad_norm": 2.0653446781054283,
+      "learning_rate": 6.964911662566309e-06,
+      "loss": 0.5171,
+      "step": 934
+    },
+    {
+      "epoch": 1.3055103464310625,
+      "grad_norm": 2.186381545142342,
+      "learning_rate": 6.957436647023117e-06,
+      "loss": 0.494,
+      "step": 935
+    },
+    {
+      "epoch": 1.3069053708439897,
+      "grad_norm": 1.9412815445850002,
+      "learning_rate": 6.949956461085714e-06,
+      "loss": 0.4895,
+      "step": 936
+    },
+    {
+      "epoch": 1.308300395256917,
+      "grad_norm": 2.0736709472728188,
+      "learning_rate": 6.942471124512346e-06,
+      "loss": 0.4724,
+      "step": 937
+    },
+    {
+      "epoch": 1.3096954196698443,
+      "grad_norm": 2.0017211386286897,
+      "learning_rate": 6.934980657074859e-06,
+      "loss": 0.4184,
+      "step": 938
+    },
+    {
+      "epoch": 1.3110904440827715,
+      "grad_norm": 1.9545935061243371,
+      "learning_rate": 6.9274850785586526e-06,
+      "loss": 0.4803,
+      "step": 939
+    },
+    {
+      "epoch": 1.3124854684956988,
+      "grad_norm": 1.9307069252931495,
+      "learning_rate": 6.919984408762632e-06,
+      "loss": 0.4624,
+      "step": 940
+    },
+    {
+      "epoch": 1.3138804929086259,
+      "grad_norm": 1.994431506317055,
+      "learning_rate": 6.9124786674991465e-06,
+      "loss": 0.466,
+      "step": 941
+    },
+    {
+      "epoch": 1.3152755173215531,
+      "grad_norm": 1.9292964581080267,
+      "learning_rate": 6.90496787459394e-06,
+      "loss": 0.4776,
+      "step": 942
+    },
+    {
+      "epoch": 1.3166705417344804,
+      "grad_norm": 2.0814712850032198,
+      "learning_rate": 6.897452049886103e-06,
+      "loss": 0.4756,
+      "step": 943
+    },
+    {
+      "epoch": 1.3180655661474074,
+      "grad_norm": 1.8033829750621577,
+      "learning_rate": 6.889931213228015e-06,
+      "loss": 0.5377,
+      "step": 944
+    },
+    {
+      "epoch": 1.3194605905603347,
+      "grad_norm": 2.168509526523026,
+      "learning_rate": 6.882405384485294e-06,
+      "loss": 0.4954,
+      "step": 945
+    },
+    {
+      "epoch": 1.320855614973262,
+      "grad_norm": 2.0688957036575033,
+      "learning_rate": 6.874874583536748e-06,
+      "loss": 0.493,
+      "step": 946
+    },
+    {
+      "epoch": 1.3222506393861893,
+      "grad_norm": 2.087243202440419,
+      "learning_rate": 6.867338830274312e-06,
+      "loss": 0.5271,
+      "step": 947
+    },
+    {
+      "epoch": 1.3236456637991165,
+      "grad_norm": 2.213328060307952,
+      "learning_rate": 6.8597981446030095e-06,
+      "loss": 0.525,
+      "step": 948
+    },
+    {
+      "epoch": 1.3250406882120438,
+      "grad_norm": 2.0595937823990624,
+      "learning_rate": 6.852252546440885e-06,
+      "loss": 0.4905,
+      "step": 949
+    },
+    {
+      "epoch": 1.326435712624971,
+      "grad_norm": 2.0526208858028876,
+      "learning_rate": 6.844702055718964e-06,
+      "loss": 0.4614,
+      "step": 950
+    },
+    {
+      "epoch": 1.3278307370378981,
+      "grad_norm": 1.9783705959415367,
+      "learning_rate": 6.837146692381197e-06,
+      "loss": 0.4686,
+      "step": 951
+    },
+    {
+      "epoch": 1.3292257614508254,
+      "grad_norm": 1.9500237608372892,
+      "learning_rate": 6.8295864763843965e-06,
+      "loss": 0.5116,
+      "step": 952
+    },
+    {
+      "epoch": 1.3306207858637527,
+      "grad_norm": 2.072824515482108,
+      "learning_rate": 6.822021427698201e-06,
+      "loss": 0.4521,
+      "step": 953
+    },
+    {
+      "epoch": 1.33201581027668,
+      "grad_norm": 2.097209437746427,
+      "learning_rate": 6.814451566305014e-06,
+      "loss": 0.5228,
+      "step": 954
+    },
+    {
+      "epoch": 1.333410834689607,
+      "grad_norm": 2.2226534279386705,
+      "learning_rate": 6.806876912199945e-06,
+      "loss": 0.4926,
+      "step": 955
+    },
+    {
+      "epoch": 1.3348058591025342,
+      "grad_norm": 2.130421968111814,
+      "learning_rate": 6.7992974853907655e-06,
+      "loss": 0.4859,
+      "step": 956
+    },
+    {
+      "epoch": 1.3362008835154615,
+      "grad_norm": 2.0794274350627013,
+      "learning_rate": 6.791713305897861e-06,
+      "loss": 0.476,
+      "step": 957
+    },
+    {
+      "epoch": 1.3375959079283888,
+      "grad_norm": 1.892246451887343,
+      "learning_rate": 6.78412439375416e-06,
+      "loss": 0.4791,
+      "step": 958
+    },
+    {
+      "epoch": 1.338990932341316,
+      "grad_norm": 2.0209725081913894,
+      "learning_rate": 6.776530769005099e-06,
+      "loss": 0.4569,
+      "step": 959
+    },
+    {
+      "epoch": 1.3403859567542433,
+      "grad_norm": 1.9927839085181078,
+      "learning_rate": 6.768932451708557e-06,
+      "loss": 0.4678,
+      "step": 960
+    },
+    {
+      "epoch": 1.3417809811671704,
+      "grad_norm": 2.0939421385477477,
+      "learning_rate": 6.761329461934814e-06,
+      "loss": 0.4711,
+      "step": 961
+    },
+    {
+      "epoch": 1.3431760055800976,
+      "grad_norm": 2.081168989179342,
+      "learning_rate": 6.753721819766489e-06,
+      "loss": 0.4945,
+      "step": 962
+    },
+    {
+      "epoch": 1.344571029993025,
+      "grad_norm": 2.1984714091362876,
+      "learning_rate": 6.746109545298488e-06,
+      "loss": 0.486,
+      "step": 963
+    },
+    {
+      "epoch": 1.3459660544059522,
+      "grad_norm": 2.112630508666608,
+      "learning_rate": 6.738492658637957e-06,
+      "loss": 0.4402,
+      "step": 964
+    },
+    {
+      "epoch": 1.3473610788188792,
+      "grad_norm": 1.7973422226559084,
+      "learning_rate": 6.730871179904218e-06,
+      "loss": 0.5329,
+      "step": 965
+    },
+    {
+      "epoch": 1.3487561032318065,
+      "grad_norm": 2.0722305312694935,
+      "learning_rate": 6.723245129228732e-06,
+      "loss": 0.4406,
+      "step": 966
+    },
+    {
+      "epoch": 1.3501511276447338,
+      "grad_norm": 2.148255296747169,
+      "learning_rate": 6.7156145267550275e-06,
+      "loss": 0.4711,
+      "step": 967
+    },
+    {
+      "epoch": 1.351546152057661,
+      "grad_norm": 2.062135239126828,
+      "learning_rate": 6.707979392638663e-06,
+      "loss": 0.4591,
+      "step": 968
+    },
+    {
+      "epoch": 1.3529411764705883,
+      "grad_norm": 2.0077274982122835,
+      "learning_rate": 6.700339747047162e-06,
+      "loss": 0.4914,
+      "step": 969
+    },
+    {
+      "epoch": 1.3543362008835156,
+      "grad_norm": 2.0688184370721747,
+      "learning_rate": 6.692695610159966e-06,
+      "loss": 0.4816,
+      "step": 970
+    },
+    {
+      "epoch": 1.3557312252964426,
+      "grad_norm": 1.9158674884573788,
+      "learning_rate": 6.685047002168382e-06,
+      "loss": 0.4511,
+      "step": 971
+    },
+    {
+      "epoch": 1.35712624970937,
+      "grad_norm": 2.041299786373518,
+      "learning_rate": 6.677393943275525e-06,
+      "loss": 0.4621,
+      "step": 972
+    },
+    {
+      "epoch": 1.3585212741222972,
+      "grad_norm": 1.9859822832346596,
+      "learning_rate": 6.669736453696266e-06,
+      "loss": 0.4668,
+      "step": 973
+    },
+    {
+      "epoch": 1.3599162985352244,
+      "grad_norm": 2.085368871543757,
+      "learning_rate": 6.66207455365718e-06,
+      "loss": 0.4548,
+      "step": 974
+    },
+    {
+      "epoch": 1.3613113229481515,
+      "grad_norm": 2.016128984506122,
+      "learning_rate": 6.6544082633964955e-06,
+      "loss": 0.4429,
+      "step": 975
+    },
+    {
+      "epoch": 1.3627063473610788,
+      "grad_norm": 2.145371160224853,
+      "learning_rate": 6.646737603164031e-06,
+      "loss": 0.5083,
+      "step": 976
+    },
+    {
+      "epoch": 1.364101371774006,
+      "grad_norm": 2.1386222042642524,
+      "learning_rate": 6.639062593221152e-06,
+      "loss": 0.5335,
+      "step": 977
+    },
+    {
+      "epoch": 1.3654963961869333,
+      "grad_norm": 2.270440526076038,
+      "learning_rate": 6.6313832538407106e-06,
+      "loss": 0.496,
+      "step": 978
+    },
+    {
+      "epoch": 1.3668914205998606,
+      "grad_norm": 1.9889941520807715,
+      "learning_rate": 6.623699605306999e-06,
+      "loss": 0.4915,
+      "step": 979
+    },
+    {
+      "epoch": 1.3682864450127878,
+      "grad_norm": 2.1440931891884727,
+      "learning_rate": 6.6160116679156874e-06,
+      "loss": 0.4679,
+      "step": 980
+    },
+    {
+      "epoch": 1.369681469425715,
+      "grad_norm": 2.227537687856243,
+      "learning_rate": 6.608319461973778e-06,
+      "loss": 0.4598,
+      "step": 981
+    },
+    {
+      "epoch": 1.3710764938386422,
+      "grad_norm": 2.0287236808325884,
+      "learning_rate": 6.6006230077995424e-06,
+      "loss": 0.4601,
+      "step": 982
+    },
+    {
+      "epoch": 1.3724715182515694,
+      "grad_norm": 1.7888137177639865,
+      "learning_rate": 6.592922325722483e-06,
+      "loss": 0.4609,
+      "step": 983
+    },
+    {
+      "epoch": 1.3738665426644967,
+      "grad_norm": 2.3520604904332534,
+      "learning_rate": 6.58521743608326e-06,
+      "loss": 0.5031,
+      "step": 984
+    },
+    {
+      "epoch": 1.3752615670774238,
+      "grad_norm": 2.110792804041468,
+      "learning_rate": 6.577508359233653e-06,
+      "loss": 0.4888,
+      "step": 985
+    },
+    {
+      "epoch": 1.376656591490351,
+      "grad_norm": 2.3149781630133393,
+      "learning_rate": 6.569795115536502e-06,
+      "loss": 0.476,
+      "step": 986
+    },
+    {
+      "epoch": 1.3780516159032783,
+      "grad_norm": 2.0549954385732434,
+      "learning_rate": 6.562077725365648e-06,
+      "loss": 0.5159,
+      "step": 987
+    },
+    {
+      "epoch": 1.3794466403162056,
+      "grad_norm": 2.0125811654928283,
+      "learning_rate": 6.554356209105892e-06,
+      "loss": 0.4225,
+      "step": 988
+    },
+    {
+      "epoch": 1.3808416647291328,
+      "grad_norm": 2.0683754856323873,
+      "learning_rate": 6.54663058715293e-06,
+      "loss": 0.5251,
+      "step": 989
+    },
+    {
+      "epoch": 1.38223668914206,
+      "grad_norm": 2.084529371644239,
+      "learning_rate": 6.538900879913301e-06,
+      "loss": 0.4857,
+      "step": 990
+    },
+    {
+      "epoch": 1.3836317135549872,
+      "grad_norm": 2.1618474971322827,
+      "learning_rate": 6.531167107804337e-06,
+      "loss": 0.4976,
+      "step": 991
+    },
+    {
+      "epoch": 1.3850267379679144,
+      "grad_norm": 2.062302710914967,
+      "learning_rate": 6.523429291254109e-06,
+      "loss": 0.5046,
+      "step": 992
+    },
+    {
+      "epoch": 1.3864217623808417,
+      "grad_norm": 2.162972647685988,
+      "learning_rate": 6.515687450701367e-06,
+      "loss": 0.4762,
+      "step": 993
+    },
+    {
+      "epoch": 1.387816786793769,
+      "grad_norm": 2.018219733272469,
+      "learning_rate": 6.507941606595492e-06,
+      "loss": 0.4364,
+      "step": 994
+    },
+    {
+      "epoch": 1.389211811206696,
+      "grad_norm": 1.8942784577915137,
+      "learning_rate": 6.500191779396439e-06,
+      "loss": 0.4854,
+      "step": 995
+    },
+    {
+      "epoch": 1.3906068356196233,
+      "grad_norm": 2.1180507917189195,
+      "learning_rate": 6.492437989574689e-06,
+      "loss": 0.4829,
+      "step": 996
+    },
+    {
+      "epoch": 1.3920018600325506,
+      "grad_norm": 2.1192897908451758,
+      "learning_rate": 6.48468025761118e-06,
+      "loss": 0.4852,
+      "step": 997
+    },
+    {
+      "epoch": 1.3933968844454778,
+      "grad_norm": 2.0232470469302775,
+      "learning_rate": 6.476918603997273e-06,
+      "loss": 0.4559,
+      "step": 998
+    },
+    {
+      "epoch": 1.394791908858405,
+      "grad_norm": 2.2141742307018655,
+      "learning_rate": 6.469153049234683e-06,
+      "loss": 0.5079,
+      "step": 999
+    },
+    {
+      "epoch": 1.3961869332713324,
+      "grad_norm": 2.225056470302364,
+      "learning_rate": 6.461383613835427e-06,
+      "loss": 0.494,
+      "step": 1000
+    },
+    {
+      "epoch": 1.3975819576842594,
+      "grad_norm": 1.9182609150895602,
+      "learning_rate": 6.453610318321777e-06,
+      "loss": 0.4497,
+      "step": 1001
+    },
+    {
+      "epoch": 1.3989769820971867,
+      "grad_norm": 1.964134487756821,
+      "learning_rate": 6.445833183226201e-06,
+      "loss": 0.5051,
+      "step": 1002
+    },
+    {
+      "epoch": 1.400372006510114,
+      "grad_norm": 2.2625631905462886,
+      "learning_rate": 6.438052229091303e-06,
+      "loss": 0.4843,
+      "step": 1003
+    },
+    {
+      "epoch": 1.4017670309230412,
+      "grad_norm": 2.095268057911824,
+      "learning_rate": 6.430267476469783e-06,
+      "loss": 0.4879,
+      "step": 1004
+    },
+    {
+      "epoch": 1.4031620553359683,
+      "grad_norm": 2.0824573756180316,
+      "learning_rate": 6.4224789459243705e-06,
+      "loss": 0.4845,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4045570797488955,
+      "grad_norm": 2.051949843595703,
+      "learning_rate": 6.4146866580277686e-06,
+      "loss": 0.454,
+      "step": 1006
+    },
+    {
+      "epoch": 1.4059521041618228,
+      "grad_norm": 2.01418184621082,
+      "learning_rate": 6.406890633362618e-06,
+      "loss": 0.4382,
+      "step": 1007
+    },
+    {
+      "epoch": 1.40734712857475,
+      "grad_norm": 1.9009970558050422,
+      "learning_rate": 6.3990908925214155e-06,
+      "loss": 0.4246,
+      "step": 1008
+    },
+    {
+      "epoch": 1.4087421529876774,
+      "grad_norm": 1.9077315815732818,
+      "learning_rate": 6.391287456106483e-06,
+      "loss": 0.477,
+      "step": 1009
+    },
+    {
+      "epoch": 1.4101371774006046,
+      "grad_norm": 2.1376510322845816,
+      "learning_rate": 6.383480344729903e-06,
+      "loss": 0.5019,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4115322018135317,
+      "grad_norm": 2.1232082872185845,
+      "learning_rate": 6.375669579013461e-06,
+      "loss": 0.5143,
+      "step": 1011
+    },
+    {
+      "epoch": 1.412927226226459,
+      "grad_norm": 2.0264092739997226,
+      "learning_rate": 6.367855179588597e-06,
+      "loss": 0.5137,
+      "step": 1012
+    },
+    {
+      "epoch": 1.4143222506393862,
+      "grad_norm": 2.0673469407609453,
+      "learning_rate": 6.3600371670963525e-06,
+      "loss": 0.469,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4157172750523135,
+      "grad_norm": 2.061303960217057,
+      "learning_rate": 6.352215562187307e-06,
+      "loss": 0.4851,
+      "step": 1014
+    },
+    {
+      "epoch": 1.4171122994652405,
+      "grad_norm": 2.0208655316631496,
+      "learning_rate": 6.344390385521534e-06,
+      "loss": 0.4796,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4185073238781678,
+      "grad_norm": 2.1338536003958506,
+      "learning_rate": 6.33656165776854e-06,
+      "loss": 0.4878,
+      "step": 1016
+    },
+    {
+      "epoch": 1.419902348291095,
+      "grad_norm": 2.2451798484385908,
+      "learning_rate": 6.328729399607206e-06,
+      "loss": 0.4825,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4212973727040223,
+      "grad_norm": 2.0158460176508672,
+      "learning_rate": 6.320893631725748e-06,
+      "loss": 0.4634,
+      "step": 1018
+    },
+    {
+      "epoch": 1.4226923971169496,
+      "grad_norm": 2.0286301190724734,
+      "learning_rate": 6.313054374821647e-06,
+      "loss": 0.4824,
+      "step": 1019
+    },
+    {
+      "epoch": 1.4240874215298769,
+      "grad_norm": 1.996772494073169,
+      "learning_rate": 6.305211649601595e-06,
+      "loss": 0.4183,
+      "step": 1020
+    },
+    {
+      "epoch": 1.425482445942804,
+      "grad_norm": 1.9064412054811033,
+      "learning_rate": 6.29736547678146e-06,
+      "loss": 0.4563,
+      "step": 1021
+    },
+    {
+      "epoch": 1.4268774703557312,
+      "grad_norm": 2.0971755867348434,
+      "learning_rate": 6.289515877086199e-06,
+      "loss": 0.5462,
+      "step": 1022
+    },
+    {
+      "epoch": 1.4282724947686585,
+      "grad_norm": 2.0723675284704175,
+      "learning_rate": 6.2816628712498315e-06,
+      "loss": 0.4796,
+      "step": 1023
+    },
+    {
+      "epoch": 1.4296675191815857,
+      "grad_norm": 1.9705008439394425,
+      "learning_rate": 6.273806480015374e-06,
+      "loss": 0.4649,
+      "step": 1024
+    },
+    {
+      "epoch": 1.4310625435945128,
+      "grad_norm": 2.0561169300756053,
+      "learning_rate": 6.265946724134782e-06,
+      "loss": 0.4159,
+      "step": 1025
+    },
+    {
+      "epoch": 1.43245756800744,
+      "grad_norm": 1.9343638344307528,
+      "learning_rate": 6.258083624368895e-06,
+      "loss": 0.4818,
+      "step": 1026
+    },
+    {
+      "epoch": 1.4338525924203673,
+      "grad_norm": 2.004281252907952,
+      "learning_rate": 6.250217201487395e-06,
+      "loss": 0.491,
+      "step": 1027
+    },
+    {
+      "epoch": 1.4352476168332946,
+      "grad_norm": 1.9996076845166801,
+      "learning_rate": 6.242347476268733e-06,
+      "loss": 0.4851,
+      "step": 1028
+    },
+    {
+      "epoch": 1.4366426412462219,
+      "grad_norm": 2.1397928370434998,
+      "learning_rate": 6.2344744695000855e-06,
+      "loss": 0.4941,
+      "step": 1029
+    },
+    {
+      "epoch": 1.4380376656591491,
+      "grad_norm": 2.1653211753488666,
+      "learning_rate": 6.226598201977299e-06,
+      "loss": 0.491,
+      "step": 1030
+    },
+    {
+      "epoch": 1.4394326900720762,
+      "grad_norm": 2.133936024923636,
+      "learning_rate": 6.218718694504831e-06,
+      "loss": 0.5052,
+      "step": 1031
+    },
+    {
+      "epoch": 1.4408277144850035,
+      "grad_norm": 2.1209125416279635,
+      "learning_rate": 6.2108359678956954e-06,
+      "loss": 0.4457,
+      "step": 1032
+    },
+    {
+      "epoch": 1.4422227388979307,
+      "grad_norm": 2.2685773861447633,
+      "learning_rate": 6.202950042971414e-06,
+      "loss": 0.5026,
+      "step": 1033
+    },
+    {
+      "epoch": 1.443617763310858,
+      "grad_norm": 2.0835264796801627,
+      "learning_rate": 6.19506094056195e-06,
+      "loss": 0.5125,
+      "step": 1034
+    },
+    {
+      "epoch": 1.445012787723785,
+      "grad_norm": 2.335657398503236,
+      "learning_rate": 6.187168681505666e-06,
+      "loss": 0.487,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4464078121367123,
+      "grad_norm": 2.0727595506282865,
+      "learning_rate": 6.17927328664926e-06,
+      "loss": 0.4688,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4478028365496396,
+      "grad_norm": 1.841772404318466,
+      "learning_rate": 6.171374776847711e-06,
+      "loss": 0.4967,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4491978609625669,
+      "grad_norm": 2.131993766169631,
+      "learning_rate": 6.163473172964229e-06,
+      "loss": 0.4573,
+      "step": 1038
+    },
+    {
+      "epoch": 1.4505928853754941,
+      "grad_norm": 2.0927552772654643,
+      "learning_rate": 6.1555684958701965e-06,
+      "loss": 0.4897,
+      "step": 1039
+    },
+    {
+      "epoch": 1.4519879097884214,
+      "grad_norm": 2.0569744283426923,
+      "learning_rate": 6.1476607664451105e-06,
+      "loss": 0.4963,
+      "step": 1040
+    },
+    {
+      "epoch": 1.4533829342013485,
+      "grad_norm": 2.079023172581386,
+      "learning_rate": 6.1397500055765345e-06,
+      "loss": 0.5044,
+      "step": 1041
+    },
+    {
+      "epoch": 1.4547779586142757,
+      "grad_norm": 2.154680029422567,
+      "learning_rate": 6.131836234160036e-06,
+      "loss": 0.4789,
+      "step": 1042
+    },
+    {
+      "epoch": 1.456172983027203,
+      "grad_norm": 2.0278215354462175,
+      "learning_rate": 6.123919473099134e-06,
+      "loss": 0.4712,
+      "step": 1043
+    },
+    {
+      "epoch": 1.4575680074401303,
+      "grad_norm": 2.0902931301788765,
+      "learning_rate": 6.115999743305252e-06,
+      "loss": 0.4583,
+      "step": 1044
+    },
+    {
+      "epoch": 1.4589630318530573,
+      "grad_norm": 2.051969856622332,
+      "learning_rate": 6.1080770656976444e-06,
+      "loss": 0.4222,
+      "step": 1045
+    },
+    {
+      "epoch": 1.4603580562659846,
+      "grad_norm": 2.019221977340812,
+      "learning_rate": 6.100151461203359e-06,
+      "loss": 0.474,
+      "step": 1046
+    },
+    {
+      "epoch": 1.4617530806789119,
+      "grad_norm": 1.953257537676229,
+      "learning_rate": 6.0922229507571716e-06,
+      "loss": 0.4927,
+      "step": 1047
+    },
+    {
+      "epoch": 1.4631481050918391,
+      "grad_norm": 2.035957959177665,
+      "learning_rate": 6.084291555301537e-06,
+      "loss": 0.506,
+      "step": 1048
+    },
+    {
+      "epoch": 1.4645431295047664,
+      "grad_norm": 2.171235089740684,
+      "learning_rate": 6.076357295786526e-06,
+      "loss": 0.4568,
+      "step": 1049
+    },
+    {
+      "epoch": 1.4659381539176937,
+      "grad_norm": 2.0290741267430636,
+      "learning_rate": 6.068420193169779e-06,
+      "loss": 0.519,
+      "step": 1050
+    },
+    {
+      "epoch": 1.4673331783306207,
+      "grad_norm": 2.119657460018465,
+      "learning_rate": 6.0604802684164436e-06,
+      "loss": 0.5092,
+      "step": 1051
+    },
+    {
+      "epoch": 1.468728202743548,
+      "grad_norm": 2.1165818757895956,
+      "learning_rate": 6.052537542499122e-06,
+      "loss": 0.4536,
+      "step": 1052
+    },
+    {
+      "epoch": 1.4701232271564753,
+      "grad_norm": 2.0023105481496457,
+      "learning_rate": 6.044592036397816e-06,
+      "loss": 0.4459,
+      "step": 1053
+    },
+    {
+      "epoch": 1.4715182515694025,
+      "grad_norm": 1.9714486621554073,
+      "learning_rate": 6.0366437710998715e-06,
+      "loss": 0.4845,
+      "step": 1054
+    },
+    {
+      "epoch": 1.4729132759823296,
+      "grad_norm": 2.237061577630844,
+      "learning_rate": 6.0286927675999205e-06,
+      "loss": 0.4729,
+      "step": 1055
+    },
+    {
+      "epoch": 1.4743083003952568,
+      "grad_norm": 1.92340136259166,
+      "learning_rate": 6.02073904689983e-06,
+      "loss": 0.5227,
+      "step": 1056
+    },
+    {
+      "epoch": 1.4757033248081841,
+      "grad_norm": 2.350503819784743,
+      "learning_rate": 6.012782630008646e-06,
+      "loss": 0.503,
+      "step": 1057
+    },
+    {
+      "epoch": 1.4770983492211114,
+      "grad_norm": 2.1697136056489934,
+      "learning_rate": 6.004823537942528e-06,
+      "loss": 0.466,
+      "step": 1058
+    },
+    {
+      "epoch": 1.4784933736340387,
+      "grad_norm": 1.9939215886063746,
+      "learning_rate": 5.996861791724713e-06,
+      "loss": 0.5017,
+      "step": 1059
+    },
+    {
+      "epoch": 1.479888398046966,
+      "grad_norm": 2.1106593906557896,
+      "learning_rate": 5.98889741238544e-06,
+      "loss": 0.4771,
+      "step": 1060
+    },
+    {
+      "epoch": 1.481283422459893,
+      "grad_norm": 2.1574303269272153,
+      "learning_rate": 5.9809304209619054e-06,
+      "loss": 0.4632,
+      "step": 1061
+    },
+    {
+      "epoch": 1.4826784468728202,
+      "grad_norm": 1.8051167313806595,
+      "learning_rate": 5.9729608384982085e-06,
+      "loss": 0.4388,
+      "step": 1062
+    },
+    {
+      "epoch": 1.4840734712857475,
+      "grad_norm": 2.0543453042457656,
+      "learning_rate": 5.964988686045289e-06,
+      "loss": 0.5055,
+      "step": 1063
+    },
+    {
+      "epoch": 1.4854684956986748,
+      "grad_norm": 2.1478980267908314,
+      "learning_rate": 5.957013984660875e-06,
+      "loss": 0.4961,
+      "step": 1064
+    },
+    {
+      "epoch": 1.4868635201116018,
+      "grad_norm": 2.14056081044156,
+      "learning_rate": 5.949036755409432e-06,
+      "loss": 0.4581,
+      "step": 1065
+    },
+    {
+      "epoch": 1.488258544524529,
+      "grad_norm": 2.1026984133609745,
+      "learning_rate": 5.941057019362095e-06,
+      "loss": 0.4443,
+      "step": 1066
+    },
+    {
+      "epoch": 1.4896535689374564,
+      "grad_norm": 1.9689741478177671,
+      "learning_rate": 5.933074797596627e-06,
+      "loss": 0.4511,
+      "step": 1067
+    },
+    {
+      "epoch": 1.4910485933503836,
+      "grad_norm": 2.1362113215367238,
+      "learning_rate": 5.925090111197355e-06,
+      "loss": 0.4693,
+      "step": 1068
+    },
+    {
+      "epoch": 1.492443617763311,
+      "grad_norm": 2.0910124749432537,
+      "learning_rate": 5.917102981255114e-06,
+      "loss": 0.4889,
+      "step": 1069
+    },
+    {
+      "epoch": 1.4938386421762382,
+      "grad_norm": 2.097915816754819,
+      "learning_rate": 5.909113428867195e-06,
+      "loss": 0.5733,
+      "step": 1070
+    },
+    {
+      "epoch": 1.4952336665891652,
+      "grad_norm": 2.329805560189044,
+      "learning_rate": 5.901121475137287e-06,
+      "loss": 0.483,
+      "step": 1071
+    },
+    {
+      "epoch": 1.4966286910020925,
+      "grad_norm": 2.1245862949546814,
+      "learning_rate": 5.893127141175425e-06,
+      "loss": 0.4991,
+      "step": 1072
+    },
+    {
+      "epoch": 1.4980237154150198,
+      "grad_norm": 2.090320969759168,
+      "learning_rate": 5.885130448097926e-06,
+      "loss": 0.5058,
+      "step": 1073
+    },
+    {
+      "epoch": 1.499418739827947,
+      "grad_norm": 2.191307478211539,
+      "learning_rate": 5.877131417027343e-06,
+      "loss": 0.5275,
+      "step": 1074
+    },
+    {
+      "epoch": 1.500813764240874,
+      "grad_norm": 2.1275043838983034,
+      "learning_rate": 5.869130069092401e-06,
+      "loss": 0.5048,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5022087886538014,
+      "grad_norm": 2.0220782985141055,
+      "learning_rate": 5.861126425427949e-06,
+      "loss": 0.4596,
+      "step": 1076
+    },
+    {
+      "epoch": 1.5036038130667286,
+      "grad_norm": 2.1080717931508643,
+      "learning_rate": 5.853120507174894e-06,
+      "loss": 0.5189,
+      "step": 1077
+    },
+    {
+      "epoch": 1.504998837479656,
+      "grad_norm": 2.091932259125407,
+      "learning_rate": 5.845112335480159e-06,
+      "loss": 0.4921,
+      "step": 1078
+    },
+    {
+      "epoch": 1.5063938618925832,
+      "grad_norm": 2.058742155033756,
+      "learning_rate": 5.83710193149661e-06,
+      "loss": 0.4337,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5077888863055104,
+      "grad_norm": 1.9879796252815132,
+      "learning_rate": 5.829089316383018e-06,
+      "loss": 0.4928,
+      "step": 1080
+    },
+    {
+      "epoch": 1.5091839107184377,
+      "grad_norm": 2.160851363843134,
+      "learning_rate": 5.821074511303988e-06,
+      "loss": 0.4643,
+      "step": 1081
+    },
+    {
+      "epoch": 1.5105789351313648,
+      "grad_norm": 1.913208554086219,
+      "learning_rate": 5.813057537429915e-06,
+      "loss": 0.4714,
+      "step": 1082
+    },
+    {
+      "epoch": 1.511973959544292,
+      "grad_norm": 2.077158539782065,
+      "learning_rate": 5.805038415936919e-06,
+      "loss": 0.4272,
+      "step": 1083
+    },
+    {
+      "epoch": 1.5133689839572193,
+      "grad_norm": 1.9639093713317004,
+      "learning_rate": 5.797017168006791e-06,
+      "loss": 0.4794,
+      "step": 1084
+    },
+    {
+      "epoch": 1.5147640083701464,
+      "grad_norm": 2.1781993886924744,
+      "learning_rate": 5.7889938148269445e-06,
+      "loss": 0.5108,
+      "step": 1085
+    },
+    {
+      "epoch": 1.5161590327830736,
+      "grad_norm": 2.021353634653322,
+      "learning_rate": 5.7809683775903525e-06,
+      "loss": 0.4582,
+      "step": 1086
+    },
+    {
+      "epoch": 1.517554057196001,
+      "grad_norm": 2.2318747817283078,
+      "learning_rate": 5.7729408774954865e-06,
+      "loss": 0.4792,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5189490816089282,
+      "grad_norm": 2.1777046381098795,
+      "learning_rate": 5.764911335746275e-06,
+      "loss": 0.4792,
+      "step": 1088
+    },
+    {
+      "epoch": 1.5203441060218554,
+      "grad_norm": 2.1281373438537208,
+      "learning_rate": 5.756879773552037e-06,
+      "loss": 0.4847,
+      "step": 1089
+    },
+    {
+      "epoch": 1.5217391304347827,
+      "grad_norm": 2.0558258117624724,
+      "learning_rate": 5.748846212127421e-06,
+      "loss": 0.5479,
+      "step": 1090
+    },
+    {
+      "epoch": 1.52313415484771,
+      "grad_norm": 2.207562158111901,
+      "learning_rate": 5.74081067269237e-06,
+      "loss": 0.4505,
+      "step": 1091
+    },
+    {
+      "epoch": 1.524529179260637,
+      "grad_norm": 1.9924679823320044,
+      "learning_rate": 5.732773176472042e-06,
+      "loss": 0.4275,
+      "step": 1092
+    },
+    {
+      "epoch": 1.5259242036735643,
+      "grad_norm": 1.8602848014371722,
+      "learning_rate": 5.7247337446967625e-06,
+      "loss": 0.4798,
+      "step": 1093
+    },
+    {
+      "epoch": 1.5273192280864916,
+      "grad_norm": 1.9906027660956394,
+      "learning_rate": 5.716692398601975e-06,
+      "loss": 0.4743,
+      "step": 1094
+    },
+    {
+      "epoch": 1.5287142524994186,
+      "grad_norm": 1.9544357110617374,
+      "learning_rate": 5.708649159428181e-06,
+      "loss": 0.5145,
+      "step": 1095
+    },
+    {
+      "epoch": 1.5301092769123459,
+      "grad_norm": 2.3355021057730796,
+      "learning_rate": 5.700604048420875e-06,
+      "loss": 0.5103,
+      "step": 1096
+    },
+    {
+      "epoch": 1.5315043013252732,
+      "grad_norm": 1.8568276485535198,
+      "learning_rate": 5.692557086830501e-06,
+      "loss": 0.4975,
+      "step": 1097
+    },
+    {
+      "epoch": 1.5328993257382004,
+      "grad_norm": 2.125493413956754,
+      "learning_rate": 5.68450829591239e-06,
+      "loss": 0.4663,
+      "step": 1098
+    },
+    {
+      "epoch": 1.5342943501511277,
+      "grad_norm": 2.071457449537922,
+      "learning_rate": 5.676457696926703e-06,
+      "loss": 0.4903,
+      "step": 1099
+    },
+    {
+      "epoch": 1.535689374564055,
+      "grad_norm": 2.1674492150404987,
+      "learning_rate": 5.668405311138382e-06,
+      "loss": 0.4943,
+      "step": 1100
+    },
+    {
+      "epoch": 1.5370843989769822,
+      "grad_norm": 2.023194559841172,
+      "learning_rate": 5.660351159817083e-06,
+      "loss": 0.4715,
+      "step": 1101
+    },
+    {
+      "epoch": 1.5384794233899093,
+      "grad_norm": 2.196574189719957,
+      "learning_rate": 5.652295264237128e-06,
+      "loss": 0.5002,
+      "step": 1102
+    },
+    {
+      "epoch": 1.5398744478028366,
+      "grad_norm": 2.0729450743671425,
+      "learning_rate": 5.6442376456774495e-06,
+      "loss": 0.4333,
+      "step": 1103
+    },
+    {
+      "epoch": 1.5412694722157638,
+      "grad_norm": 1.8939760093629232,
+      "learning_rate": 5.636178325421524e-06,
+      "loss": 0.4704,
+      "step": 1104
+    },
+    {
+      "epoch": 1.5426644966286909,
+      "grad_norm": 2.0676972299709955,
+      "learning_rate": 5.628117324757326e-06,
+      "loss": 0.4879,
+      "step": 1105
+    },
+    {
+      "epoch": 1.5440595210416181,
+      "grad_norm": 2.0776134536197457,
+      "learning_rate": 5.620054664977275e-06,
+      "loss": 0.3994,
+      "step": 1106
+    },
+    {
+      "epoch": 1.5454545454545454,
+      "grad_norm": 1.9149487033980843,
+      "learning_rate": 5.61199036737816e-06,
+      "loss": 0.4564,
+      "step": 1107
+    },
+    {
+      "epoch": 1.5468495698674727,
+      "grad_norm": 1.8789388381354126,
+      "learning_rate": 5.603924453261109e-06,
+      "loss": 0.4306,
+      "step": 1108
+    },
+    {
+      "epoch": 1.5482445942804,
+      "grad_norm": 2.1151583976969444,
+      "learning_rate": 5.595856943931512e-06,
+      "loss": 0.4965,
+      "step": 1109
+    },
+    {
+      "epoch": 1.5496396186933272,
+      "grad_norm": 2.168510370730916,
+      "learning_rate": 5.587787860698975e-06,
+      "loss": 0.4645,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5510346431062545,
+      "grad_norm": 1.9652877078434625,
+      "learning_rate": 5.579717224877261e-06,
+      "loss": 0.5291,
+      "step": 1111
+    },
+    {
+      "epoch": 1.5524296675191815,
+      "grad_norm": 2.2649679803534637,
+      "learning_rate": 5.571645057784236e-06,
+      "loss": 0.5016,
+      "step": 1112
+    },
+    {
+      "epoch": 1.5538246919321088,
+      "grad_norm": 2.1957127413656328,
+      "learning_rate": 5.5635713807418055e-06,
+      "loss": 0.4695,
+      "step": 1113
+    },
+    {
+      "epoch": 1.555219716345036,
+      "grad_norm": 2.0362707966220723,
+      "learning_rate": 5.55549621507587e-06,
+      "loss": 0.4885,
+      "step": 1114
+    },
+    {
+      "epoch": 1.5566147407579631,
+      "grad_norm": 1.9927839746067004,
+      "learning_rate": 5.547419582116259e-06,
+      "loss": 0.5161,
+      "step": 1115
+    },
+    {
+      "epoch": 1.5580097651708904,
+      "grad_norm": 2.196498341950231,
+      "learning_rate": 5.539341503196674e-06,
+      "loss": 0.4412,
+      "step": 1116
+    },
+    {
+      "epoch": 1.5594047895838177,
+      "grad_norm": 2.1458152403575768,
+      "learning_rate": 5.531261999654646e-06,
+      "loss": 0.442,
+      "step": 1117
+    },
+    {
+      "epoch": 1.560799813996745,
+      "grad_norm": 2.006850128740523,
+      "learning_rate": 5.5231810928314555e-06,
+      "loss": 0.4598,
+      "step": 1118
+    },
+    {
+      "epoch": 1.5621948384096722,
+      "grad_norm": 1.9493183184314022,
+      "learning_rate": 5.5150988040721e-06,
+      "loss": 0.457,
+      "step": 1119
+    },
+    {
+      "epoch": 1.5635898628225995,
+      "grad_norm": 2.0840902735769684,
+      "learning_rate": 5.507015154725226e-06,
+      "loss": 0.4632,
+      "step": 1120
+    },
+    {
+      "epoch": 1.5649848872355268,
+      "grad_norm": 1.8428305966849245,
+      "learning_rate": 5.4989301661430685e-06,
+      "loss": 0.4506,
+      "step": 1121
+    },
+    {
+      "epoch": 1.5663799116484538,
+      "grad_norm": 1.9920482502185506,
+      "learning_rate": 5.490843859681404e-06,
+      "loss": 0.46,
+      "step": 1122
+    },
+    {
+      "epoch": 1.567774936061381,
+      "grad_norm": 2.1161191567578324,
+      "learning_rate": 5.48275625669949e-06,
+      "loss": 0.4538,
+      "step": 1123
+    },
+    {
+      "epoch": 1.5691699604743083,
+      "grad_norm": 1.9746726770175769,
+      "learning_rate": 5.474667378560007e-06,
+      "loss": 0.4625,
+      "step": 1124
+    },
+    {
+      "epoch": 1.5705649848872354,
+      "grad_norm": 1.9991374998558773,
+      "learning_rate": 5.466577246629006e-06,
+      "loss": 0.4789,
+      "step": 1125
+    },
+    {
+      "epoch": 1.5719600093001627,
+      "grad_norm": 1.9636955625225274,
+      "learning_rate": 5.458485882275848e-06,
+      "loss": 0.433,
+      "step": 1126
+    },
+    {
+      "epoch": 1.57335503371309,
+      "grad_norm": 1.8591665976898522,
+      "learning_rate": 5.45039330687315e-06,
+      "loss": 0.4911,
+      "step": 1127
+    },
+    {
+      "epoch": 1.5747500581260172,
+      "grad_norm": 2.016208821382892,
+      "learning_rate": 5.442299541796727e-06,
+      "loss": 0.4491,
+      "step": 1128
+    },
+    {
+      "epoch": 1.5761450825389445,
+      "grad_norm": 2.043786587370076,
+      "learning_rate": 5.4342046084255385e-06,
+      "loss": 0.4343,
+      "step": 1129
+    },
+    {
+      "epoch": 1.5775401069518717,
+      "grad_norm": 1.9497412453804057,
+      "learning_rate": 5.426108528141627e-06,
+      "loss": 0.47,
+      "step": 1130
+    },
+    {
+      "epoch": 1.578935131364799,
+      "grad_norm": 1.9244444803664198,
+      "learning_rate": 5.4180113223300665e-06,
+      "loss": 0.4416,
+      "step": 1131
+    },
+    {
+      "epoch": 1.580330155777726,
+      "grad_norm": 2.0298017508341006,
+      "learning_rate": 5.409913012378903e-06,
+      "loss": 0.4235,
+      "step": 1132
+    },
+    {
+      "epoch": 1.5817251801906533,
+      "grad_norm": 1.9384109846754765,
+      "learning_rate": 5.401813619679102e-06,
+      "loss": 0.4938,
+      "step": 1133
+    },
+    {
+      "epoch": 1.5831202046035806,
+      "grad_norm": 2.1038176771719512,
+      "learning_rate": 5.3937131656244834e-06,
+      "loss": 0.4393,
+      "step": 1134
+    },
+    {
+      "epoch": 1.5845152290165077,
+      "grad_norm": 2.1838408097559086,
+      "learning_rate": 5.385611671611676e-06,
+      "loss": 0.4802,
+      "step": 1135
+    },
+    {
+      "epoch": 1.585910253429435,
+      "grad_norm": 2.055193473610198,
+      "learning_rate": 5.377509159040051e-06,
+      "loss": 0.4912,
+      "step": 1136
+    },
+    {
+      "epoch": 1.5873052778423622,
+      "grad_norm": 2.1117830596189258,
+      "learning_rate": 5.3694056493116745e-06,
+      "loss": 0.4495,
+      "step": 1137
+    },
+    {
+      "epoch": 1.5887003022552895,
+      "grad_norm": 2.126637279332641,
+      "learning_rate": 5.361301163831242e-06,
+      "loss": 0.4921,
+      "step": 1138
+    },
+    {
+      "epoch": 1.5900953266682167,
+      "grad_norm": 2.072596734971921,
+      "learning_rate": 5.353195724006031e-06,
+      "loss": 0.4346,
+      "step": 1139
+    },
+    {
+      "epoch": 1.591490351081144,
+      "grad_norm": 1.9778671288497442,
+      "learning_rate": 5.345089351245834e-06,
+      "loss": 0.4889,
+      "step": 1140
+    },
+    {
+      "epoch": 1.5928853754940713,
+      "grad_norm": 2.0619924037997315,
+      "learning_rate": 5.336982066962915e-06,
+      "loss": 0.4723,
+      "step": 1141
+    },
+    {
+      "epoch": 1.5942803999069983,
+      "grad_norm": 2.095968190026346,
+      "learning_rate": 5.328873892571941e-06,
+      "loss": 0.4294,
+      "step": 1142
+    },
+    {
+      "epoch": 1.5956754243199256,
+      "grad_norm": 2.1452385884936405,
+      "learning_rate": 5.320764849489929e-06,
+      "loss": 0.4964,
+      "step": 1143
+    },
+    {
+      "epoch": 1.5970704487328529,
+      "grad_norm": 2.1195901942810624,
+      "learning_rate": 5.312654959136194e-06,
+      "loss": 0.4024,
+      "step": 1144
+    },
+    {
+      "epoch": 1.59846547314578,
+      "grad_norm": 1.9898460201178139,
+      "learning_rate": 5.304544242932288e-06,
+      "loss": 0.502,
+      "step": 1145
+    },
+    {
+      "epoch": 1.5998604975587072,
+      "grad_norm": 2.1486403537619614,
+      "learning_rate": 5.296432722301944e-06,
+      "loss": 0.5122,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6012555219716345,
+      "grad_norm": 2.048752350661396,
+      "learning_rate": 5.288320418671018e-06,
+      "loss": 0.4522,
+      "step": 1147
+    },
+    {
+      "epoch": 1.6026505463845617,
+      "grad_norm": 2.242124204991126,
+      "learning_rate": 5.280207353467438e-06,
+      "loss": 0.4824,
+      "step": 1148
+    },
+    {
+      "epoch": 1.604045570797489,
+      "grad_norm": 2.1087909007924206,
+      "learning_rate": 5.272093548121141e-06,
+      "loss": 0.4882,
+      "step": 1149
+    },
+    {
+      "epoch": 1.6054405952104163,
+      "grad_norm": 2.0689612978514185,
+      "learning_rate": 5.26397902406402e-06,
+      "loss": 0.4895,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6068356196233435,
+      "grad_norm": 2.2404499749663596,
+      "learning_rate": 5.255863802729866e-06,
+      "loss": 0.4703,
+      "step": 1151
+    },
+    {
+      "epoch": 1.6082306440362706,
+      "grad_norm": 2.1796296336015004,
+      "learning_rate": 5.247747905554311e-06,
+      "loss": 0.4371,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6096256684491979,
+      "grad_norm": 1.83480870370753,
+      "learning_rate": 5.239631353974774e-06,
+      "loss": 0.5222,
+      "step": 1153
+    },
+    {
+      "epoch": 1.6110206928621251,
+      "grad_norm": 2.1557579183524207,
+      "learning_rate": 5.231514169430403e-06,
+      "loss": 0.4727,
+      "step": 1154
+    },
+    {
+      "epoch": 1.6124157172750522,
+      "grad_norm": 2.1001021273931504,
+      "learning_rate": 5.223396373362013e-06,
+      "loss": 0.4419,
+      "step": 1155
+    },
+    {
+      "epoch": 1.6138107416879794,
+      "grad_norm": 1.8542588865560108,
+      "learning_rate": 5.215277987212041e-06,
+      "loss": 0.5285,
+      "step": 1156
+    },
+    {
+      "epoch": 1.6152057661009067,
+      "grad_norm": 2.128214277180948,
+      "learning_rate": 5.207159032424478e-06,
+      "loss": 0.4982,
+      "step": 1157
+    },
+    {
+      "epoch": 1.616600790513834,
+      "grad_norm": 2.05081239685872,
+      "learning_rate": 5.199039530444819e-06,
+      "loss": 0.4488,
+      "step": 1158
+    },
+    {
+      "epoch": 1.6179958149267613,
+      "grad_norm": 2.0930206418500936,
+      "learning_rate": 5.1909195027200055e-06,
+      "loss": 0.4665,
+      "step": 1159
+    },
+    {
+      "epoch": 1.6193908393396885,
+      "grad_norm": 1.9511896188233324,
+      "learning_rate": 5.182798970698361e-06,
+      "loss": 0.4662,
+      "step": 1160
+    },
+    {
+      "epoch": 1.6207858637526158,
+      "grad_norm": 2.1151104009916377,
+      "learning_rate": 5.174677955829551e-06,
+      "loss": 0.479,
+      "step": 1161
+    },
+    {
+      "epoch": 1.6221808881655428,
+      "grad_norm": 2.0233329936544675,
+      "learning_rate": 5.166556479564511e-06,
+      "loss": 0.5033,
+      "step": 1162
+    },
+    {
+      "epoch": 1.6235759125784701,
+      "grad_norm": 2.005641777027395,
+      "learning_rate": 5.158434563355392e-06,
+      "loss": 0.5077,
+      "step": 1163
+    },
+    {
+      "epoch": 1.6249709369913974,
+      "grad_norm": 2.3289383835606796,
+      "learning_rate": 5.150312228655515e-06,
+      "loss": 0.4157,
+      "step": 1164
+    },
+    {
+      "epoch": 1.6263659614043244,
+      "grad_norm": 1.9091089129449355,
+      "learning_rate": 5.142189496919302e-06,
+      "loss": 0.4645,
+      "step": 1165
+    },
+    {
+      "epoch": 1.6277609858172517,
+      "grad_norm": 1.9509007916244274,
+      "learning_rate": 5.1340663896022206e-06,
+      "loss": 0.449,
+      "step": 1166
+    },
+    {
+      "epoch": 1.629156010230179,
+      "grad_norm": 1.9978926625299753,
+      "learning_rate": 5.125942928160736e-06,
+      "loss": 0.4693,
+      "step": 1167
+    },
+    {
+      "epoch": 1.6305510346431062,
+      "grad_norm": 2.1593985262641393,
+      "learning_rate": 5.117819134052246e-06,
+      "loss": 0.4853,
+      "step": 1168
+    },
+    {
+      "epoch": 1.6319460590560335,
+      "grad_norm": 2.0010212399413074,
+      "learning_rate": 5.10969502873503e-06,
+      "loss": 0.4646,
+      "step": 1169
+    },
+    {
+      "epoch": 1.6333410834689608,
+      "grad_norm": 2.179277406220846,
+      "learning_rate": 5.101570633668185e-06,
+      "loss": 0.4685,
+      "step": 1170
+    },
+    {
+      "epoch": 1.634736107881888,
+      "grad_norm": 2.19745037995779,
+      "learning_rate": 5.093445970311576e-06,
+      "loss": 0.4716,
+      "step": 1171
+    },
+    {
+      "epoch": 1.636131132294815,
+      "grad_norm": 2.0643787126674162,
+      "learning_rate": 5.085321060125775e-06,
+      "loss": 0.4534,
+      "step": 1172
+    },
+    {
+      "epoch": 1.6375261567077424,
+      "grad_norm": 2.0764902906142697,
+      "learning_rate": 5.07719592457201e-06,
+      "loss": 0.4648,
+      "step": 1173
+    },
+    {
+      "epoch": 1.6389211811206696,
+      "grad_norm": 2.1659245368856554,
+      "learning_rate": 5.069070585112097e-06,
+      "loss": 0.5295,
+      "step": 1174
+    },
+    {
+      "epoch": 1.6403162055335967,
+      "grad_norm": 2.3959883670923587,
+      "learning_rate": 5.060945063208399e-06,
+      "loss": 0.4686,
+      "step": 1175
+    },
+    {
+      "epoch": 1.641711229946524,
+      "grad_norm": 2.1564182143544675,
+      "learning_rate": 5.052819380323757e-06,
+      "loss": 0.4761,
+      "step": 1176
+    },
+    {
+      "epoch": 1.6431062543594512,
+      "grad_norm": 2.053086822481449,
+      "learning_rate": 5.044693557921434e-06,
+      "loss": 0.4694,
+      "step": 1177
+    },
+    {
+      "epoch": 1.6445012787723785,
+      "grad_norm": 2.067471373158249,
+      "learning_rate": 5.036567617465067e-06,
+      "loss": 0.4787,
+      "step": 1178
+    },
+    {
+      "epoch": 1.6458963031853058,
+      "grad_norm": 2.2385488484589433,
+      "learning_rate": 5.0284415804186025e-06,
+      "loss": 0.4661,
+      "step": 1179
+    },
+    {
+      "epoch": 1.647291327598233,
+      "grad_norm": 2.0037594738988678,
+      "learning_rate": 5.02031546824624e-06,
+      "loss": 0.4727,
+      "step": 1180
+    },
+    {
+      "epoch": 1.6486863520111603,
+      "grad_norm": 2.104281706724569,
+      "learning_rate": 5.012189302412383e-06,
+      "loss": 0.4894,
+      "step": 1181
+    },
+    {
+      "epoch": 1.6500813764240876,
+      "grad_norm": 2.1163907295817563,
+      "learning_rate": 5.0040631043815715e-06,
+      "loss": 0.5042,
+      "step": 1182
+    },
+    {
+      "epoch": 1.6514764008370146,
+      "grad_norm": 2.084686423119246,
+      "learning_rate": 4.99593689561843e-06,
+      "loss": 0.4659,
+      "step": 1183
+    },
+    {
+      "epoch": 1.652871425249942,
+      "grad_norm": 1.972423605655503,
+      "learning_rate": 4.987810697587618e-06,
+      "loss": 0.4969,
+      "step": 1184
+    },
+    {
+      "epoch": 1.654266449662869,
+      "grad_norm": 2.1644206111383255,
+      "learning_rate": 4.979684531753761e-06,
+      "loss": 0.4543,
+      "step": 1185
+    },
+    {
+      "epoch": 1.6556614740757962,
+      "grad_norm": 2.1503763455287808,
+      "learning_rate": 4.971558419581398e-06,
+      "loss": 0.4504,
+      "step": 1186
+    },
+    {
+      "epoch": 1.6570564984887235,
+      "grad_norm": 1.971907358337702,
+      "learning_rate": 4.963432382534933e-06,
+      "loss": 0.5174,
+      "step": 1187
+    },
+    {
+      "epoch": 1.6584515229016508,
+      "grad_norm": 2.311562963469942,
+      "learning_rate": 4.955306442078568e-06,
+      "loss": 0.5096,
+      "step": 1188
+    },
+    {
+      "epoch": 1.659846547314578,
+      "grad_norm": 2.0406738494866117,
+      "learning_rate": 4.947180619676244e-06,
+      "loss": 0.4221,
+      "step": 1189
+    },
+    {
+      "epoch": 1.6612415717275053,
+      "grad_norm": 1.8981730758330377,
+      "learning_rate": 4.9390549367916004e-06,
+      "loss": 0.4456,
+      "step": 1190
+    },
+    {
+      "epoch": 1.6626365961404326,
+      "grad_norm": 1.882840958072787,
+      "learning_rate": 4.930929414887904e-06,
+      "loss": 0.4484,
+      "step": 1191
+    },
+    {
+      "epoch": 1.6640316205533598,
+      "grad_norm": 2.1125846152280054,
+      "learning_rate": 4.9228040754279915e-06,
+      "loss": 0.4778,
+      "step": 1192
+    },
+    {
+      "epoch": 1.665426644966287,
+      "grad_norm": 2.0262565344195536,
+      "learning_rate": 4.914678939874225e-06,
+      "loss": 0.5206,
+      "step": 1193
+    },
+    {
+      "epoch": 1.6668216693792142,
+      "grad_norm": 2.2200965218244972,
+      "learning_rate": 4.906554029688427e-06,
+      "loss": 0.4383,
+      "step": 1194
+    },
+    {
+      "epoch": 1.6682166937921412,
+      "grad_norm": 1.9240667691854707,
+      "learning_rate": 4.898429366331815e-06,
+      "loss": 0.4802,
+      "step": 1195
+    },
+    {
+      "epoch": 1.6696117182050685,
+      "grad_norm": 2.1125221215706236,
+      "learning_rate": 4.8903049712649705e-06,
+      "loss": 0.5099,
+      "step": 1196
+    },
+    {
+      "epoch": 1.6710067426179958,
+      "grad_norm": 2.114831308671667,
+      "learning_rate": 4.8821808659477544e-06,
+      "loss": 0.4483,
+      "step": 1197
+    },
+    {
+      "epoch": 1.672401767030923,
+      "grad_norm": 2.0277403831475853,
+      "learning_rate": 4.874057071839265e-06,
+      "loss": 0.4633,
+      "step": 1198
+    },
+    {
+      "epoch": 1.6737967914438503,
+      "grad_norm": 1.9861660198892896,
+      "learning_rate": 4.86593361039778e-06,
+      "loss": 0.48,
+      "step": 1199
+    },
+    {
+      "epoch": 1.6751918158567776,
+      "grad_norm": 2.134110515376525,
+      "learning_rate": 4.857810503080701e-06,
+      "loss": 0.4877,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6765868402697048,
+      "grad_norm": 2.1201931398630824,
+      "learning_rate": 4.849687771344487e-06,
+      "loss": 0.5193,
+      "step": 1201
+    },
+    {
+      "epoch": 1.677981864682632,
+      "grad_norm": 2.3401336739315686,
+      "learning_rate": 4.841565436644609e-06,
+      "loss": 0.5108,
+      "step": 1202
+    },
+    {
+      "epoch": 1.6793768890955592,
+      "grad_norm": 2.173303767477864,
+      "learning_rate": 4.8334435204354915e-06,
+      "loss": 0.4609,
+      "step": 1203
+    },
+    {
+      "epoch": 1.6807719135084864,
+      "grad_norm": 2.1088182001036957,
+      "learning_rate": 4.825322044170451e-06,
+      "loss": 0.5097,
+      "step": 1204
+    },
+    {
+      "epoch": 1.6821669379214135,
+      "grad_norm": 1.9686905359024993,
+      "learning_rate": 4.81720102930164e-06,
+      "loss": 0.4725,
+      "step": 1205
+    },
+    {
+      "epoch": 1.6835619623343407,
+      "grad_norm": 2.1236567609317265,
+      "learning_rate": 4.809080497279998e-06,
+      "loss": 0.4915,
+      "step": 1206
+    },
+    {
+      "epoch": 1.684956986747268,
+      "grad_norm": 2.2136938683111986,
+      "learning_rate": 4.800960469555183e-06,
+      "loss": 0.4904,
+      "step": 1207
+    },
+    {
+      "epoch": 1.6863520111601953,
+      "grad_norm": 1.944578414169582,
+      "learning_rate": 4.792840967575523e-06,
+      "loss": 0.4837,
+      "step": 1208
+    },
+    {
+      "epoch": 1.6877470355731226,
+      "grad_norm": 2.068496086879585,
+      "learning_rate": 4.784722012787961e-06,
+      "loss": 0.4957,
+      "step": 1209
+    },
+    {
+      "epoch": 1.6891420599860498,
+      "grad_norm": 2.259138484792416,
+      "learning_rate": 4.776603626637988e-06,
+      "loss": 0.4853,
+      "step": 1210
+    },
+    {
+      "epoch": 1.690537084398977,
+      "grad_norm": 2.0361945607382017,
+      "learning_rate": 4.768485830569598e-06,
+      "loss": 0.4494,
+      "step": 1211
+    },
+    {
+      "epoch": 1.6919321088119044,
+      "grad_norm": 1.9985570020748773,
+      "learning_rate": 4.7603686460252265e-06,
+      "loss": 0.4912,
+      "step": 1212
+    },
+    {
+      "epoch": 1.6933271332248314,
+      "grad_norm": 2.1741124461981034,
+      "learning_rate": 4.75225209444569e-06,
+      "loss": 0.464,
+      "step": 1213
+    },
+    {
+      "epoch": 1.6947221576377587,
+      "grad_norm": 2.101109519431568,
+      "learning_rate": 4.744136197270135e-06,
+      "loss": 0.5325,
+      "step": 1214
+    },
+    {
+      "epoch": 1.6961171820506857,
+      "grad_norm": 2.2648499488417504,
+      "learning_rate": 4.736020975935981e-06,
+      "loss": 0.4799,
+      "step": 1215
+    },
+    {
+      "epoch": 1.697512206463613,
+      "grad_norm": 2.050720067395807,
+      "learning_rate": 4.72790645187886e-06,
+      "loss": 0.4379,
+      "step": 1216
+    },
+    {
+      "epoch": 1.6989072308765403,
+      "grad_norm": 1.9934244980053937,
+      "learning_rate": 4.7197926465325626e-06,
+      "loss": 0.4568,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7003022552894675,
+      "grad_norm": 2.0680803010766353,
+      "learning_rate": 4.711679581328983e-06,
+      "loss": 0.4844,
+      "step": 1218
+    },
+    {
+      "epoch": 1.7016972797023948,
+      "grad_norm": 2.1697148699026427,
+      "learning_rate": 4.703567277698058e-06,
+      "loss": 0.5517,
+      "step": 1219
+    },
+    {
+      "epoch": 1.703092304115322,
+      "grad_norm": 2.3093501303979056,
+      "learning_rate": 4.695455757067712e-06,
+      "loss": 0.4494,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7044873285282494,
+      "grad_norm": 1.9215030222404832,
+      "learning_rate": 4.687345040863808e-06,
+      "loss": 0.4841,
+      "step": 1221
+    },
+    {
+      "epoch": 1.7058823529411766,
+      "grad_norm": 2.194442976010545,
+      "learning_rate": 4.679235150510072e-06,
+      "loss": 0.4329,
+      "step": 1222
+    },
+    {
+      "epoch": 1.7072773773541037,
+      "grad_norm": 1.9682684824994168,
+      "learning_rate": 4.671126107428061e-06,
+      "loss": 0.4774,
+      "step": 1223
+    },
+    {
+      "epoch": 1.708672401767031,
+      "grad_norm": 1.9357493109786206,
+      "learning_rate": 4.663017933037087e-06,
+      "loss": 0.4623,
+      "step": 1224
+    },
+    {
+      "epoch": 1.710067426179958,
+      "grad_norm": 2.07959597800617,
+      "learning_rate": 4.6549106487541666e-06,
+      "loss": 0.4922,
+      "step": 1225
+    },
+    {
+      "epoch": 1.7114624505928853,
+      "grad_norm": 2.12106114838888,
+      "learning_rate": 4.646804275993971e-06,
+      "loss": 0.4598,
+      "step": 1226
+    },
+    {
+      "epoch": 1.7128574750058125,
+      "grad_norm": 2.193483910207871,
+      "learning_rate": 4.63869883616876e-06,
+      "loss": 0.4517,
+      "step": 1227
+    },
+    {
+      "epoch": 1.7142524994187398,
+      "grad_norm": 1.9267269677926486,
+      "learning_rate": 4.630594350688327e-06,
+      "loss": 0.4974,
+      "step": 1228
+    },
+    {
+      "epoch": 1.715647523831667,
+      "grad_norm": 2.159284260285703,
+      "learning_rate": 4.62249084095995e-06,
+      "loss": 0.464,
+      "step": 1229
+    },
+    {
+      "epoch": 1.7170425482445943,
+      "grad_norm": 2.04247259167603,
+      "learning_rate": 4.614388328388327e-06,
+      "loss": 0.5124,
+      "step": 1230
+    },
+    {
+      "epoch": 1.7184375726575216,
+      "grad_norm": 2.172072715605954,
+      "learning_rate": 4.606286834375517e-06,
+      "loss": 0.498,
+      "step": 1231
+    },
+    {
+      "epoch": 1.7198325970704489,
+      "grad_norm": 2.06551781626508,
+      "learning_rate": 4.598186380320899e-06,
+      "loss": 0.4357,
+      "step": 1232
+    },
+    {
+      "epoch": 1.721227621483376,
+      "grad_norm": 2.2163136697977404,
+      "learning_rate": 4.5900869876210986e-06,
+      "loss": 0.4807,
+      "step": 1233
+    },
+    {
+      "epoch": 1.7226226458963032,
+      "grad_norm": 2.0544727475631164,
+      "learning_rate": 4.581988677669935e-06,
+      "loss": 0.5156,
+      "step": 1234
+    },
+    {
+      "epoch": 1.7240176703092303,
+      "grad_norm": 2.1102785785058025,
+      "learning_rate": 4.573891471858375e-06,
+      "loss": 0.4841,
+      "step": 1235
+    },
+    {
+      "epoch": 1.7254126947221575,
+      "grad_norm": 2.091638327396252,
+      "learning_rate": 4.565795391574465e-06,
+      "loss": 0.4742,
+      "step": 1236
+    },
+    {
+      "epoch": 1.7268077191350848,
+      "grad_norm": 2.073708631501961,
+      "learning_rate": 4.5577004582032745e-06,
+      "loss": 0.4483,
+      "step": 1237
+    },
+    {
+      "epoch": 1.728202743548012,
+      "grad_norm": 2.0663949647182505,
+      "learning_rate": 4.549606693126851e-06,
+      "loss": 0.464,
+      "step": 1238
+    },
+    {
+      "epoch": 1.7295977679609393,
+      "grad_norm": 2.0950685671584885,
+      "learning_rate": 4.541514117724155e-06,
+      "loss": 0.5271,
+      "step": 1239
+    },
+    {
+      "epoch": 1.7309927923738666,
+      "grad_norm": 2.069764512552555,
+      "learning_rate": 4.533422753370995e-06,
+      "loss": 0.4246,
+      "step": 1240
+    },
+    {
+      "epoch": 1.7323878167867939,
+      "grad_norm": 1.9471175111782795,
+      "learning_rate": 4.525332621439995e-06,
+      "loss": 0.4876,
+      "step": 1241
+    },
+    {
+      "epoch": 1.7337828411997211,
+      "grad_norm": 2.106836696946677,
+      "learning_rate": 4.517243743300513e-06,
+      "loss": 0.4195,
+      "step": 1242
+    },
+    {
+      "epoch": 1.7351778656126482,
+      "grad_norm": 2.1428624250471917,
+      "learning_rate": 4.5091561403185976e-06,
+      "loss": 0.5121,
+      "step": 1243
+    },
+    {
+      "epoch": 1.7365728900255755,
+      "grad_norm": 2.1001384678126596,
+      "learning_rate": 4.501069833856934e-06,
+      "loss": 0.4635,
+      "step": 1244
+    },
+    {
+      "epoch": 1.7379679144385025,
+      "grad_norm": 2.124440573719448,
+      "learning_rate": 4.492984845274774e-06,
+      "loss": 0.4337,
+      "step": 1245
+    },
+    {
+      "epoch": 1.7393629388514298,
+      "grad_norm": 2.0913781241638056,
+      "learning_rate": 4.484901195927901e-06,
+      "loss": 0.4648,
+      "step": 1246
+    },
+    {
+      "epoch": 1.740757963264357,
+      "grad_norm": 1.988763311594427,
+      "learning_rate": 4.476818907168545e-06,
+      "loss": 0.4183,
+      "step": 1247
+    },
+    {
+      "epoch": 1.7421529876772843,
+      "grad_norm": 2.026709075336435,
+      "learning_rate": 4.4687380003453555e-06,
+      "loss": 0.4416,
+      "step": 1248
+    },
+    {
+      "epoch": 1.7435480120902116,
+      "grad_norm": 2.1037185629898834,
+      "learning_rate": 4.460658496803327e-06,
+      "loss": 0.4772,
+      "step": 1249
+    },
+    {
+      "epoch": 1.7449430365031389,
+      "grad_norm": 2.1544403103897563,
+      "learning_rate": 4.4525804178837425e-06,
+      "loss": 0.461,
+      "step": 1250
+    },
+    {
+      "epoch": 1.7463380609160661,
+      "grad_norm": 2.2928421209767595,
+      "learning_rate": 4.4445037849241305e-06,
+      "loss": 0.5024,
+      "step": 1251
+    },
+    {
+      "epoch": 1.7477330853289934,
+      "grad_norm": 2.135997919508182,
+      "learning_rate": 4.436428619258196e-06,
+      "loss": 0.4752,
+      "step": 1252
+    },
+    {
+      "epoch": 1.7491281097419205,
+      "grad_norm": 2.085493617102473,
+      "learning_rate": 4.428354942215766e-06,
+      "loss": 0.4422,
+      "step": 1253
+    },
+    {
+      "epoch": 1.7505231341548477,
+      "grad_norm": 2.120851175775019,
+      "learning_rate": 4.42028277512274e-06,
+      "loss": 0.4488,
+      "step": 1254
+    },
+    {
+      "epoch": 1.7519181585677748,
+      "grad_norm": 1.9217827847817792,
+      "learning_rate": 4.412212139301027e-06,
+      "loss": 0.4928,
+      "step": 1255
+    },
+    {
+      "epoch": 1.753313182980702,
+      "grad_norm": 2.128839111883042,
+      "learning_rate": 4.404143056068489e-06,
+      "loss": 0.4084,
+      "step": 1256
+    },
+    {
+      "epoch": 1.7547082073936293,
+      "grad_norm": 1.9027826190815191,
+      "learning_rate": 4.3960755467388916e-06,
+      "loss": 0.475,
+      "step": 1257
+    },
+    {
+      "epoch": 1.7561032318065566,
+      "grad_norm": 2.0787775348878195,
+      "learning_rate": 4.388009632621841e-06,
+      "loss": 0.5112,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7574982562194839,
+      "grad_norm": 2.0669238505154985,
+      "learning_rate": 4.379945335022727e-06,
+      "loss": 0.4698,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7588932806324111,
+      "grad_norm": 2.251534310250234,
+      "learning_rate": 4.371882675242674e-06,
+      "loss": 0.5002,
+      "step": 1260
+    },
+    {
+      "epoch": 1.7602883050453384,
+      "grad_norm": 2.2205227372571517,
+      "learning_rate": 4.363821674578479e-06,
+      "loss": 0.5113,
+      "step": 1261
+    },
+    {
+      "epoch": 1.7616833294582657,
+      "grad_norm": 2.0821257827137356,
+      "learning_rate": 4.355762354322552e-06,
+      "loss": 0.4511,
+      "step": 1262
+    },
+    {
+      "epoch": 1.7630783538711927,
+      "grad_norm": 2.0560505922750965,
+      "learning_rate": 4.347704735762872e-06,
+      "loss": 0.496,
+      "step": 1263
+    },
+    {
+      "epoch": 1.76447337828412,
+      "grad_norm": 2.0968400735874506,
+      "learning_rate": 4.339648840182919e-06,
+      "loss": 0.4827,
+      "step": 1264
+    },
+    {
+      "epoch": 1.765868402697047,
+      "grad_norm": 2.0915848890730655,
+      "learning_rate": 4.331594688861619e-06,
+      "loss": 0.4674,
+      "step": 1265
+    },
+    {
+      "epoch": 1.7672634271099743,
+      "grad_norm": 2.1096575286314523,
+      "learning_rate": 4.323542303073297e-06,
+      "loss": 0.4453,
+      "step": 1266
+    },
+    {
+      "epoch": 1.7686584515229016,
+      "grad_norm": 2.1182357112080714,
+      "learning_rate": 4.315491704087613e-06,
+      "loss": 0.4546,
+      "step": 1267
+    },
+    {
+      "epoch": 1.7700534759358288,
+      "grad_norm": 1.978572548445151,
+      "learning_rate": 4.3074429131695e-06,
+      "loss": 0.4644,
+      "step": 1268
+    },
+    {
+      "epoch": 1.7714485003487561,
+      "grad_norm": 2.1705214494874387,
+      "learning_rate": 4.299395951579126e-06,
+      "loss": 0.468,
+      "step": 1269
+    },
+    {
+      "epoch": 1.7728435247616834,
+      "grad_norm": 2.0255085627659146,
+      "learning_rate": 4.291350840571821e-06,
+      "loss": 0.4304,
+      "step": 1270
+    },
+    {
+      "epoch": 1.7742385491746107,
+      "grad_norm": 2.072433117751704,
+      "learning_rate": 4.283307601398026e-06,
+      "loss": 0.4824,
+      "step": 1271
+    },
+    {
+      "epoch": 1.775633573587538,
+      "grad_norm": 2.2629627765772677,
+      "learning_rate": 4.275266255303238e-06,
+      "loss": 0.4915,
+      "step": 1272
+    },
+    {
+      "epoch": 1.777028598000465,
+      "grad_norm": 2.234922030785417,
+      "learning_rate": 4.2672268235279616e-06,
+      "loss": 0.4629,
+      "step": 1273
+    },
+    {
+      "epoch": 1.7784236224133922,
+      "grad_norm": 1.9270261774066482,
+      "learning_rate": 4.259189327307632e-06,
+      "loss": 0.4784,
+      "step": 1274
+    },
+    {
+      "epoch": 1.7798186468263193,
+      "grad_norm": 2.2081059039595683,
+      "learning_rate": 4.251153787872579e-06,
+      "loss": 0.4803,
+      "step": 1275
+    },
+    {
+      "epoch": 1.7812136712392466,
+      "grad_norm": 1.939134834396615,
+      "learning_rate": 4.2431202264479665e-06,
+      "loss": 0.4586,
+      "step": 1276
+    },
+    {
+      "epoch": 1.7826086956521738,
+      "grad_norm": 2.095040560965543,
+      "learning_rate": 4.235088664253726e-06,
+      "loss": 0.5089,
+      "step": 1277
+    },
+    {
+      "epoch": 1.784003720065101,
+      "grad_norm": 2.1919774266759515,
+      "learning_rate": 4.227059122504514e-06,
+      "loss": 0.485,
+      "step": 1278
+    },
+    {
+      "epoch": 1.7853987444780284,
+      "grad_norm": 1.8782861952292524,
+      "learning_rate": 4.21903162240965e-06,
+      "loss": 0.4446,
+      "step": 1279
+    },
+    {
+      "epoch": 1.7867937688909556,
+      "grad_norm": 2.0890195629123385,
+      "learning_rate": 4.211006185173056e-06,
+      "loss": 0.5088,
+      "step": 1280
+    },
+    {
+      "epoch": 1.788188793303883,
+      "grad_norm": 2.026437008974195,
+      "learning_rate": 4.20298283199321e-06,
+      "loss": 0.425,
+      "step": 1281
+    },
+    {
+      "epoch": 1.7895838177168102,
+      "grad_norm": 2.2467401498211252,
+      "learning_rate": 4.1949615840630845e-06,
+      "loss": 0.4653,
+      "step": 1282
+    },
+    {
+      "epoch": 1.7909788421297372,
+      "grad_norm": 2.093166602036887,
+      "learning_rate": 4.186942462570087e-06,
+      "loss": 0.5048,
+      "step": 1283
+    },
+    {
+      "epoch": 1.7923738665426645,
+      "grad_norm": 2.156850489159853,
+      "learning_rate": 4.178925488696012e-06,
+      "loss": 0.4903,
+      "step": 1284
+    },
+    {
+      "epoch": 1.7937688909555918,
+      "grad_norm": 2.085389759207312,
+      "learning_rate": 4.170910683616985e-06,
+      "loss": 0.4688,
+      "step": 1285
+    },
+    {
+      "epoch": 1.7951639153685188,
+      "grad_norm": 2.213200329005782,
+      "learning_rate": 4.1628980685033914e-06,
+      "loss": 0.5185,
+      "step": 1286
+    },
+    {
+      "epoch": 1.796558939781446,
+      "grad_norm": 2.050195188781033,
+      "learning_rate": 4.154887664519842e-06,
+      "loss": 0.4881,
+      "step": 1287
+    },
+    {
+      "epoch": 1.7979539641943734,
+      "grad_norm": 2.1196860562270294,
+      "learning_rate": 4.1468794928251064e-06,
+      "loss": 0.4673,
+      "step": 1288
+    },
+    {
+      "epoch": 1.7993489886073006,
+      "grad_norm": 2.177343517348136,
+      "learning_rate": 4.138873574572053e-06,
+      "loss": 0.4747,
+      "step": 1289
+    },
+    {
+      "epoch": 1.800744013020228,
+      "grad_norm": 2.003472990127112,
+      "learning_rate": 4.130869930907599e-06,
+      "loss": 0.4159,
+      "step": 1290
+    },
+    {
+      "epoch": 1.8021390374331552,
+      "grad_norm": 1.8839471124092069,
+      "learning_rate": 4.122868582972659e-06,
+      "loss": 0.469,
+      "step": 1291
+    },
+    {
+      "epoch": 1.8035340618460824,
+      "grad_norm": 2.035907134161031,
+      "learning_rate": 4.114869551902075e-06,
+      "loss": 0.4282,
+      "step": 1292
+    },
+    {
+      "epoch": 1.8049290862590095,
+      "grad_norm": 2.001424449827008,
+      "learning_rate": 4.106872858824576e-06,
+      "loss": 0.4554,
+      "step": 1293
+    },
+    {
+      "epoch": 1.8063241106719368,
+      "grad_norm": 2.008750426771798,
+      "learning_rate": 4.098878524862715e-06,
+      "loss": 0.4981,
+      "step": 1294
+    },
+    {
+      "epoch": 1.807719135084864,
+      "grad_norm": 2.2520451839673012,
+      "learning_rate": 4.090886571132807e-06,
+      "loss": 0.5648,
+      "step": 1295
+    },
+    {
+      "epoch": 1.809114159497791,
+      "grad_norm": 2.216674336178155,
+      "learning_rate": 4.082897018744887e-06,
+      "loss": 0.4966,
+      "step": 1296
+    },
+    {
+      "epoch": 1.8105091839107184,
+      "grad_norm": 2.3579614946903833,
+      "learning_rate": 4.074909888802648e-06,
+      "loss": 0.5033,
+      "step": 1297
+    },
+    {
+      "epoch": 1.8119042083236456,
+      "grad_norm": 1.8522361974582706,
+      "learning_rate": 4.066925202403374e-06,
+      "loss": 0.4183,
+      "step": 1298
+    },
+    {
+      "epoch": 1.813299232736573,
+      "grad_norm": 2.0409109799975242,
+      "learning_rate": 4.058942980637906e-06,
+      "loss": 0.4835,
+      "step": 1299
+    },
+    {
+      "epoch": 1.8146942571495002,
+      "grad_norm": 1.9846733203898725,
+      "learning_rate": 4.050963244590571e-06,
+      "loss": 0.4369,
+      "step": 1300
+    },
+    {
+      "epoch": 1.8160892815624274,
+      "grad_norm": 1.9908168675190487,
+      "learning_rate": 4.042986015339126e-06,
+      "loss": 0.4705,
+      "step": 1301
+    },
+    {
+      "epoch": 1.8174843059753547,
+      "grad_norm": 1.97724651246631,
+      "learning_rate": 4.035011313954713e-06,
+      "loss": 0.4826,
+      "step": 1302
+    },
+    {
+      "epoch": 1.8188793303882818,
+      "grad_norm": 2.1982531234476284,
+      "learning_rate": 4.027039161501795e-06,
+      "loss": 0.4077,
+      "step": 1303
+    },
+    {
+      "epoch": 1.820274354801209,
+      "grad_norm": 1.9503370551181338,
+      "learning_rate": 4.019069579038096e-06,
+      "loss": 0.4615,
+      "step": 1304
+    },
+    {
+      "epoch": 1.8216693792141363,
+      "grad_norm": 1.9113061201404709,
+      "learning_rate": 4.011102587614563e-06,
+      "loss": 0.427,
+      "step": 1305
+    },
+    {
+      "epoch": 1.8230644036270633,
+      "grad_norm": 2.088228005988047,
+      "learning_rate": 4.00313820827529e-06,
+      "loss": 0.4925,
+      "step": 1306
+    },
+    {
+      "epoch": 1.8244594280399906,
+      "grad_norm": 2.265032952966295,
+      "learning_rate": 3.995176462057473e-06,
+      "loss": 0.5266,
+      "step": 1307
+    },
+    {
+      "epoch": 1.8258544524529179,
+      "grad_norm": 2.2633304605254065,
+      "learning_rate": 3.987217369991357e-06,
+      "loss": 0.4703,
+      "step": 1308
+    },
+    {
+      "epoch": 1.8272494768658452,
+      "grad_norm": 2.034863642322609,
+      "learning_rate": 3.979260953100169e-06,
+      "loss": 0.4721,
+      "step": 1309
+    },
+    {
+      "epoch": 1.8286445012787724,
+      "grad_norm": 2.0974244608448736,
+      "learning_rate": 3.97130723240008e-06,
+      "loss": 0.4539,
+      "step": 1310
+    },
+    {
+      "epoch": 1.8300395256916997,
+      "grad_norm": 2.414871430447381,
+      "learning_rate": 3.96335622890013e-06,
+      "loss": 0.4988,
+      "step": 1311
+    },
+    {
+      "epoch": 1.831434550104627,
+      "grad_norm": 1.6347770659827563,
+      "learning_rate": 3.955407963602184e-06,
+      "loss": 0.4765,
+      "step": 1312
+    },
+    {
+      "epoch": 1.832829574517554,
+      "grad_norm": 1.9715809826573552,
+      "learning_rate": 3.94746245750088e-06,
+      "loss": 0.4583,
+      "step": 1313
+    },
+    {
+      "epoch": 1.8342245989304813,
+      "grad_norm": 2.1527465543653252,
+      "learning_rate": 3.939519731583557e-06,
+      "loss": 0.4851,
+      "step": 1314
+    },
+    {
+      "epoch": 1.8356196233434086,
+      "grad_norm": 1.8555310432515144,
+      "learning_rate": 3.9315798068302214e-06,
+      "loss": 0.4367,
+      "step": 1315
+    },
+    {
+      "epoch": 1.8370146477563356,
+      "grad_norm": 1.871881038073264,
+      "learning_rate": 3.923642704213475e-06,
+      "loss": 0.4948,
+      "step": 1316
+    },
+    {
+      "epoch": 1.8384096721692629,
+      "grad_norm": 2.068493987616388,
+      "learning_rate": 3.915708444698465e-06,
+      "loss": 0.4911,
+      "step": 1317
+    },
+    {
+      "epoch": 1.8398046965821901,
+      "grad_norm": 2.108597697463182,
+      "learning_rate": 3.907777049242828e-06,
+      "loss": 0.4578,
+      "step": 1318
+    },
+    {
+      "epoch": 1.8411997209951174,
+      "grad_norm": 2.034896589340762,
+      "learning_rate": 3.899848538796643e-06,
+      "loss": 0.4772,
+      "step": 1319
+    },
+    {
+      "epoch": 1.8425947454080447,
+      "grad_norm": 2.0072300235614255,
+      "learning_rate": 3.891922934302356e-06,
+      "loss": 0.4897,
+      "step": 1320
+    },
+    {
+      "epoch": 1.843989769820972,
+      "grad_norm": 2.188051096125535,
+      "learning_rate": 3.884000256694749e-06,
+      "loss": 0.4872,
+      "step": 1321
+    },
+    {
+      "epoch": 1.8453847942338992,
+      "grad_norm": 2.118044774667855,
+      "learning_rate": 3.876080526900867e-06,
+      "loss": 0.5054,
+      "step": 1322
+    },
+    {
+      "epoch": 1.8467798186468263,
+      "grad_norm": 2.1344072938179344,
+      "learning_rate": 3.868163765839966e-06,
+      "loss": 0.4731,
+      "step": 1323
+    },
+    {
+      "epoch": 1.8481748430597535,
+      "grad_norm": 2.053672801159552,
+      "learning_rate": 3.860249994423467e-06,
+      "loss": 0.4723,
+      "step": 1324
+    },
+    {
+      "epoch": 1.8495698674726808,
+      "grad_norm": 2.230370059524332,
+      "learning_rate": 3.852339233554891e-06,
+      "loss": 0.4587,
+      "step": 1325
+    },
+    {
+      "epoch": 1.8509648918856079,
+      "grad_norm": 2.1250159637012827,
+      "learning_rate": 3.844431504129804e-06,
+      "loss": 0.5112,
+      "step": 1326
+    },
+    {
+      "epoch": 1.8523599162985351,
+      "grad_norm": 2.0955476438334806,
+      "learning_rate": 3.8365268270357715e-06,
+      "loss": 0.4793,
+      "step": 1327
+    },
+    {
+      "epoch": 1.8537549407114624,
+      "grad_norm": 1.9947397547326173,
+      "learning_rate": 3.828625223152291e-06,
+      "loss": 0.4653,
+      "step": 1328
+    },
+    {
+      "epoch": 1.8551499651243897,
+      "grad_norm": 2.0597481264501214,
+      "learning_rate": 3.820726713350742e-06,
+      "loss": 0.4625,
+      "step": 1329
+    },
+    {
+      "epoch": 1.856544989537317,
+      "grad_norm": 2.2135489734378755,
+      "learning_rate": 3.812831318494335e-06,
+      "loss": 0.4897,
+      "step": 1330
+    },
+    {
+      "epoch": 1.8579400139502442,
+      "grad_norm": 2.010088784483442,
+      "learning_rate": 3.804939059438052e-06,
+      "loss": 0.4686,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8593350383631715,
+      "grad_norm": 2.0704467732349485,
+      "learning_rate": 3.797049957028588e-06,
+      "loss": 0.4763,
+      "step": 1332
+    },
+    {
+      "epoch": 1.8607300627760985,
+      "grad_norm": 2.0974552233676036,
+      "learning_rate": 3.7891640321043054e-06,
+      "loss": 0.421,
+      "step": 1333
+    },
+    {
+      "epoch": 1.8621250871890258,
+      "grad_norm": 2.1435925715160145,
+      "learning_rate": 3.781281305495171e-06,
+      "loss": 0.419,
+      "step": 1334
+    },
+    {
+      "epoch": 1.863520111601953,
+      "grad_norm": 2.0082993030068437,
+      "learning_rate": 3.773401798022701e-06,
+      "loss": 0.4321,
+      "step": 1335
+    },
+    {
+      "epoch": 1.8649151360148801,
+      "grad_norm": 2.1583115545533347,
+      "learning_rate": 3.765525530499915e-06,
+      "loss": 0.4628,
+      "step": 1336
+    },
+    {
+      "epoch": 1.8663101604278074,
+      "grad_norm": 2.11597183573991,
+      "learning_rate": 3.757652523731269e-06,
+      "loss": 0.4195,
+      "step": 1337
+    },
+    {
+      "epoch": 1.8677051848407347,
+      "grad_norm": 1.949534543566651,
+      "learning_rate": 3.7497827985126054e-06,
+      "loss": 0.4481,
+      "step": 1338
+    },
+    {
+      "epoch": 1.869100209253662,
+      "grad_norm": 2.0253447244970335,
+      "learning_rate": 3.741916375631105e-06,
+      "loss": 0.4319,
+      "step": 1339
+    },
+    {
+      "epoch": 1.8704952336665892,
+      "grad_norm": 2.3285146478186247,
+      "learning_rate": 3.7340532758652217e-06,
+      "loss": 0.5306,
+      "step": 1340
+    },
+    {
+      "epoch": 1.8718902580795165,
+      "grad_norm": 2.152694903826016,
+      "learning_rate": 3.7261935199846266e-06,
+      "loss": 0.4423,
+      "step": 1341
+    },
+    {
+      "epoch": 1.8732852824924437,
+      "grad_norm": 2.2329276959444257,
+      "learning_rate": 3.7183371287501684e-06,
+      "loss": 0.4592,
+      "step": 1342
+    },
+    {
+      "epoch": 1.8746803069053708,
+      "grad_norm": 1.9615943761756212,
+      "learning_rate": 3.7104841229138034e-06,
+      "loss": 0.4831,
+      "step": 1343
+    },
+    {
+      "epoch": 1.876075331318298,
+      "grad_norm": 2.0983585004345557,
+      "learning_rate": 3.7026345232185416e-06,
+      "loss": 0.4368,
+      "step": 1344
+    },
+    {
+      "epoch": 1.8774703557312253,
+      "grad_norm": 1.9460400781991807,
+      "learning_rate": 3.6947883503984037e-06,
+      "loss": 0.4429,
+      "step": 1345
+    },
+    {
+      "epoch": 1.8788653801441524,
+      "grad_norm": 1.9426677248205293,
+      "learning_rate": 3.686945625178356e-06,
+      "loss": 0.4477,
+      "step": 1346
+    },
+    {
+      "epoch": 1.8802604045570797,
+      "grad_norm": 2.010880459472735,
+      "learning_rate": 3.6791063682742535e-06,
+      "loss": 0.4897,
+      "step": 1347
+    },
+    {
+      "epoch": 1.881655428970007,
+      "grad_norm": 2.032395468552332,
+      "learning_rate": 3.6712706003927937e-06,
+      "loss": 0.464,
+      "step": 1348
+    },
+    {
+      "epoch": 1.8830504533829342,
+      "grad_norm": 2.033909833330093,
+      "learning_rate": 3.6634383422314622e-06,
+      "loss": 0.4415,
+      "step": 1349
+    },
+    {
+      "epoch": 1.8844454777958615,
+      "grad_norm": 2.1329623060747607,
+      "learning_rate": 3.655609614478467e-06,
+      "loss": 0.5069,
+      "step": 1350
+    },
+    {
+      "epoch": 1.8858405022087887,
+      "grad_norm": 2.225055368929594,
+      "learning_rate": 3.647784437812693e-06,
+      "loss": 0.4889,
+      "step": 1351
+    },
+    {
+      "epoch": 1.887235526621716,
+      "grad_norm": 2.141063499651939,
+      "learning_rate": 3.6399628329036496e-06,
+      "loss": 0.4978,
+      "step": 1352
+    },
+    {
+      "epoch": 1.888630551034643,
+      "grad_norm": 2.2248929304810097,
+      "learning_rate": 3.632144820411405e-06,
+      "loss": 0.4004,
+      "step": 1353
+    },
+    {
+      "epoch": 1.8900255754475703,
+      "grad_norm": 1.8643692161399157,
+      "learning_rate": 3.624330420986541e-06,
+      "loss": 0.4244,
+      "step": 1354
+    },
+    {
+      "epoch": 1.8914205998604976,
+      "grad_norm": 1.9733735956698852,
+      "learning_rate": 3.6165196552701e-06,
+      "loss": 0.4697,
+      "step": 1355
+    },
+    {
+      "epoch": 1.8928156242734246,
+      "grad_norm": 1.9848401068488462,
+      "learning_rate": 3.6087125438935187e-06,
+      "loss": 0.4624,
+      "step": 1356
+    },
+    {
+      "epoch": 1.894210648686352,
+      "grad_norm": 2.073164703831999,
+      "learning_rate": 3.6009091074785853e-06,
+      "loss": 0.4629,
+      "step": 1357
+    },
+    {
+      "epoch": 1.8956056730992792,
+      "grad_norm": 2.0659109705444556,
+      "learning_rate": 3.5931093666373845e-06,
+      "loss": 0.513,
+      "step": 1358
+    },
+    {
+      "epoch": 1.8970006975122065,
+      "grad_norm": 2.1125582756734103,
+      "learning_rate": 3.585313341972232e-06,
+      "loss": 0.4522,
+      "step": 1359
+    },
+    {
+      "epoch": 1.8983957219251337,
+      "grad_norm": 2.0584832403532904,
+      "learning_rate": 3.577521054075631e-06,
+      "loss": 0.4274,
+      "step": 1360
+    },
+    {
+      "epoch": 1.899790746338061,
+      "grad_norm": 2.0843140708165815,
+      "learning_rate": 3.5697325235302183e-06,
+      "loss": 0.4201,
+      "step": 1361
+    },
+    {
+      "epoch": 1.9011857707509883,
+      "grad_norm": 2.0579271534464585,
+      "learning_rate": 3.5619477709086982e-06,
+      "loss": 0.4297,
+      "step": 1362
+    },
+    {
+      "epoch": 1.9025807951639153,
+      "grad_norm": 2.1115377455750535,
+      "learning_rate": 3.5541668167738003e-06,
+      "loss": 0.5433,
+      "step": 1363
+    },
+    {
+      "epoch": 1.9039758195768426,
+      "grad_norm": 2.2747225928648978,
+      "learning_rate": 3.546389681678224e-06,
+      "loss": 0.4562,
+      "step": 1364
+    },
+    {
+      "epoch": 1.9053708439897699,
+      "grad_norm": 1.9665911130251685,
+      "learning_rate": 3.538616386164575e-06,
+      "loss": 0.4618,
+      "step": 1365
+    },
+    {
+      "epoch": 1.906765868402697,
+      "grad_norm": 2.04359025515372,
+      "learning_rate": 3.530846950765318e-06,
+      "loss": 0.4771,
+      "step": 1366
+    },
+    {
+      "epoch": 1.9081608928156242,
+      "grad_norm": 2.105984355882839,
+      "learning_rate": 3.5230813960027275e-06,
+      "loss": 0.4773,
+      "step": 1367
+    },
+    {
+      "epoch": 1.9095559172285514,
+      "grad_norm": 2.256498118072638,
+      "learning_rate": 3.5153197423888206e-06,
+      "loss": 0.4975,
+      "step": 1368
+    },
+    {
+      "epoch": 1.9109509416414787,
+      "grad_norm": 1.9536343765489645,
+      "learning_rate": 3.5075620104253123e-06,
+      "loss": 0.4375,
+      "step": 1369
+    },
+    {
+      "epoch": 1.912345966054406,
+      "grad_norm": 2.0611008824205754,
+      "learning_rate": 3.4998082206035606e-06,
+      "loss": 0.4481,
+      "step": 1370
+    },
+    {
+      "epoch": 1.9137409904673333,
+      "grad_norm": 1.8294594643156625,
+      "learning_rate": 3.492058393404509e-06,
+      "loss": 0.5112,
+      "step": 1371
+    },
+    {
+      "epoch": 1.9151360148802605,
+      "grad_norm": 2.0956104344978823,
+      "learning_rate": 3.4843125492986345e-06,
+      "loss": 0.4213,
+      "step": 1372
+    },
+    {
+      "epoch": 1.9165310392931876,
+      "grad_norm": 1.911894442742041,
+      "learning_rate": 3.4765707087458912e-06,
+      "loss": 0.493,
+      "step": 1373
+    },
+    {
+      "epoch": 1.9179260637061148,
+      "grad_norm": 2.102284559161196,
+      "learning_rate": 3.468832892195664e-06,
+      "loss": 0.4985,
+      "step": 1374
+    },
+    {
+      "epoch": 1.9193210881190421,
+      "grad_norm": 2.5416534065559935,
+      "learning_rate": 3.4610991200867006e-06,
+      "loss": 0.4735,
+      "step": 1375
+    },
+    {
+      "epoch": 1.9207161125319692,
+      "grad_norm": 2.0019610291575805,
+      "learning_rate": 3.453369412847071e-06,
+      "loss": 0.4061,
+      "step": 1376
+    },
+    {
+      "epoch": 1.9221111369448964,
+      "grad_norm": 2.0730001274558947,
+      "learning_rate": 3.445643790894109e-06,
+      "loss": 0.4451,
+      "step": 1377
+    },
+    {
+      "epoch": 1.9235061613578237,
+      "grad_norm": 2.0984153028815538,
+      "learning_rate": 3.4379222746343534e-06,
+      "loss": 0.4679,
+      "step": 1378
+    },
+    {
+      "epoch": 1.924901185770751,
+      "grad_norm": 2.074651799672325,
+      "learning_rate": 3.4302048844634995e-06,
+      "loss": 0.4776,
+      "step": 1379
+    },
+    {
+      "epoch": 1.9262962101836782,
+      "grad_norm": 2.027386258813131,
+      "learning_rate": 3.4224916407663484e-06,
+      "loss": 0.427,
+      "step": 1380
+    },
+    {
+      "epoch": 1.9276912345966055,
+      "grad_norm": 2.1138275693841986,
+      "learning_rate": 3.414782563916742e-06,
+      "loss": 0.5011,
+      "step": 1381
+    },
+    {
+      "epoch": 1.9290862590095328,
+      "grad_norm": 2.192773328005699,
+      "learning_rate": 3.407077674277518e-06,
+      "loss": 0.4716,
+      "step": 1382
+    },
+    {
+      "epoch": 1.93048128342246,
+      "grad_norm": 2.213292858679933,
+      "learning_rate": 3.3993769922004584e-06,
+      "loss": 0.5317,
+      "step": 1383
+    },
+    {
+      "epoch": 1.931876307835387,
+      "grad_norm": 2.1487844965796308,
+      "learning_rate": 3.391680538026224e-06,
+      "loss": 0.4242,
+      "step": 1384
+    },
+    {
+      "epoch": 1.9332713322483144,
+      "grad_norm": 2.090023360978835,
+      "learning_rate": 3.3839883320843125e-06,
+      "loss": 0.4513,
+      "step": 1385
+    },
+    {
+      "epoch": 1.9346663566612414,
+      "grad_norm": 1.9959369737145263,
+      "learning_rate": 3.3763003946930023e-06,
+      "loss": 0.4656,
+      "step": 1386
+    },
+    {
+      "epoch": 1.9360613810741687,
+      "grad_norm": 2.1931110824630418,
+      "learning_rate": 3.36861674615929e-06,
+      "loss": 0.4548,
+      "step": 1387
+    },
+    {
+      "epoch": 1.937456405487096,
+      "grad_norm": 2.190386146798527,
+      "learning_rate": 3.360937406778849e-06,
+      "loss": 0.5095,
+      "step": 1388
+    },
+    {
+      "epoch": 1.9388514299000232,
+      "grad_norm": 2.135412483356473,
+      "learning_rate": 3.35326239683597e-06,
+      "loss": 0.4221,
+      "step": 1389
+    },
+    {
+      "epoch": 1.9402464543129505,
+      "grad_norm": 2.0515132595158616,
+      "learning_rate": 3.3455917366035058e-06,
+      "loss": 0.4702,
+      "step": 1390
+    },
+    {
+      "epoch": 1.9416414787258778,
+      "grad_norm": 2.073867022732308,
+      "learning_rate": 3.337925446342819e-06,
+      "loss": 0.4391,
+      "step": 1391
+    },
+    {
+      "epoch": 1.943036503138805,
+      "grad_norm": 1.8548990606837181,
+      "learning_rate": 3.3302635463037352e-06,
+      "loss": 0.3957,
+      "step": 1392
+    },
+    {
+      "epoch": 1.9444315275517323,
+      "grad_norm": 2.0591050886917666,
+      "learning_rate": 3.3226060567244767e-06,
+      "loss": 0.479,
+      "step": 1393
+    },
+    {
+      "epoch": 1.9458265519646594,
+      "grad_norm": 2.1031599275109993,
+      "learning_rate": 3.314952997831618e-06,
+      "loss": 0.4221,
+      "step": 1394
+    },
+    {
+      "epoch": 1.9472215763775866,
+      "grad_norm": 1.8174504828503302,
+      "learning_rate": 3.307304389840036e-06,
+      "loss": 0.4339,
+      "step": 1395
+    },
+    {
+      "epoch": 1.9486166007905137,
+      "grad_norm": 2.0361355052531147,
+      "learning_rate": 3.29966025295284e-06,
+      "loss": 0.447,
+      "step": 1396
+    },
+    {
+      "epoch": 1.950011625203441,
+      "grad_norm": 2.116288896112645,
+      "learning_rate": 3.292020607361337e-06,
+      "loss": 0.5655,
+      "step": 1397
+    },
+    {
+      "epoch": 1.9514066496163682,
+      "grad_norm": 2.1226615577582777,
+      "learning_rate": 3.284385473244974e-06,
+      "loss": 0.4208,
+      "step": 1398
+    },
+    {
+      "epoch": 1.9528016740292955,
+      "grad_norm": 2.0903634352822804,
+      "learning_rate": 3.2767548707712693e-06,
+      "loss": 0.4854,
+      "step": 1399
+    },
+    {
+      "epoch": 1.9541966984422228,
+      "grad_norm": 2.1629936996115458,
+      "learning_rate": 3.2691288200957826e-06,
+      "loss": 0.4792,
+      "step": 1400
+    },
+    {
+      "epoch": 1.95559172285515,
+      "grad_norm": 2.186118380372086,
+      "learning_rate": 3.2615073413620467e-06,
+      "loss": 0.4976,
+      "step": 1401
+    },
+    {
+      "epoch": 1.9569867472680773,
+      "grad_norm": 2.0815648698913485,
+      "learning_rate": 3.2538904547015137e-06,
+      "loss": 0.4719,
+      "step": 1402
+    },
+    {
+      "epoch": 1.9583817716810046,
+      "grad_norm": 2.0373008076972,
+      "learning_rate": 3.2462781802335124e-06,
+      "loss": 0.4655,
+      "step": 1403
+    },
+    {
+      "epoch": 1.9597767960939316,
+      "grad_norm": 2.187393729560026,
+      "learning_rate": 3.2386705380651877e-06,
+      "loss": 0.4686,
+      "step": 1404
+    },
+    {
+      "epoch": 1.961171820506859,
+      "grad_norm": 1.857328655696565,
+      "learning_rate": 3.2310675482914444e-06,
+      "loss": 0.4681,
+      "step": 1405
+    },
+    {
+      "epoch": 1.962566844919786,
+      "grad_norm": 2.1023104143665847,
+      "learning_rate": 3.2234692309949034e-06,
+      "loss": 0.4678,
+      "step": 1406
+    },
+    {
+      "epoch": 1.9639618693327132,
+      "grad_norm": 2.036361172428041,
+      "learning_rate": 3.2158756062458422e-06,
+      "loss": 0.4722,
+      "step": 1407
+    },
+    {
+      "epoch": 1.9653568937456405,
+      "grad_norm": 2.0089004500677756,
+      "learning_rate": 3.208286694102141e-06,
+      "loss": 0.4994,
+      "step": 1408
+    },
+    {
+      "epoch": 1.9667519181585678,
+      "grad_norm": 2.1615010459513075,
+      "learning_rate": 3.2007025146092345e-06,
+      "loss": 0.4398,
+      "step": 1409
+    },
+    {
+      "epoch": 1.968146942571495,
+      "grad_norm": 1.8494820793896445,
+      "learning_rate": 3.1931230878000586e-06,
+      "loss": 0.45,
+      "step": 1410
+    },
+    {
+      "epoch": 1.9695419669844223,
+      "grad_norm": 2.0839549919773552,
+      "learning_rate": 3.1855484336949876e-06,
+      "loss": 0.4568,
+      "step": 1411
+    },
+    {
+      "epoch": 1.9709369913973496,
+      "grad_norm": 1.9340841864438192,
+      "learning_rate": 3.1779785723017988e-06,
+      "loss": 0.5065,
+      "step": 1412
+    },
+    {
+      "epoch": 1.9723320158102768,
+      "grad_norm": 2.0965032360845637,
+      "learning_rate": 3.170413523615605e-06,
+      "loss": 0.441,
+      "step": 1413
+    },
+    {
+      "epoch": 1.9737270402232039,
+      "grad_norm": 2.1238493162876466,
+      "learning_rate": 3.162853307618805e-06,
+      "loss": 0.402,
+      "step": 1414
+    },
+    {
+      "epoch": 1.9751220646361312,
+      "grad_norm": 1.8144332716416323,
+      "learning_rate": 3.155297944281036e-06,
+      "loss": 0.4518,
+      "step": 1415
+    },
+    {
+      "epoch": 1.9765170890490582,
+      "grad_norm": 2.0794891682927177,
+      "learning_rate": 3.1477474535591167e-06,
+      "loss": 0.4618,
+      "step": 1416
+    },
+    {
+      "epoch": 1.9779121134619855,
+      "grad_norm": 2.092341448966828,
+      "learning_rate": 3.1402018553969917e-06,
+      "loss": 0.5035,
+      "step": 1417
+    },
+    {
+      "epoch": 1.9793071378749127,
+      "grad_norm": 2.033890245909392,
+      "learning_rate": 3.132661169725688e-06,
+      "loss": 0.4757,
+      "step": 1418
+    },
+    {
+      "epoch": 1.98070216228784,
+      "grad_norm": 1.9352972803319826,
+      "learning_rate": 3.125125416463254e-06,
+      "loss": 0.4332,
+      "step": 1419
+    },
+    {
+      "epoch": 1.9820971867007673,
+      "grad_norm": 2.105289598294007,
+      "learning_rate": 3.1175946155147064e-06,
+      "loss": 0.4633,
+      "step": 1420
+    },
+    {
+      "epoch": 1.9834922111136946,
+      "grad_norm": 2.323601540036809,
+      "learning_rate": 3.110068786771987e-06,
+      "loss": 0.4709,
+      "step": 1421
+    },
+    {
+      "epoch": 1.9848872355266218,
+      "grad_norm": 1.7129060190000591,
+      "learning_rate": 3.1025479501139e-06,
+      "loss": 0.4934,
+      "step": 1422
+    },
+    {
+      "epoch": 1.986282259939549,
+      "grad_norm": 2.2024747583092115,
+      "learning_rate": 3.095032125406062e-06,
+      "loss": 0.4469,
+      "step": 1423
+    },
+    {
+      "epoch": 1.9876772843524761,
+      "grad_norm": 2.2101104273083885,
+      "learning_rate": 3.0875213325008548e-06,
+      "loss": 0.5061,
+      "step": 1424
+    },
+    {
+      "epoch": 1.9890723087654034,
+      "grad_norm": 2.1708027049479957,
+      "learning_rate": 3.0800155912373696e-06,
+      "loss": 0.4703,
+      "step": 1425
+    },
+    {
+      "epoch": 1.9904673331783305,
+      "grad_norm": 2.155953890818061,
+      "learning_rate": 3.0725149214413487e-06,
+      "loss": 0.4196,
+      "step": 1426
+    },
+    {
+      "epoch": 1.9918623575912577,
+      "grad_norm": 2.005245900668136,
+      "learning_rate": 3.065019342925143e-06,
+      "loss": 0.4886,
+      "step": 1427
+    },
+    {
+      "epoch": 1.993257382004185,
+      "grad_norm": 2.2555355821584158,
+      "learning_rate": 3.0575288754876565e-06,
+      "loss": 0.4786,
+      "step": 1428
+    },
+    {
+      "epoch": 1.9946524064171123,
+      "grad_norm": 1.868282129579876,
+      "learning_rate": 3.0500435389142867e-06,
+      "loss": 0.4108,
+      "step": 1429
+    },
+    {
+      "epoch": 1.9960474308300395,
+      "grad_norm": 2.1879958454321655,
+      "learning_rate": 3.042563352976884e-06,
+      "loss": 0.4564,
+      "step": 1430
+    },
+    {
+      "epoch": 1.9974424552429668,
+      "grad_norm": 2.036612583842871,
+      "learning_rate": 3.035088337433694e-06,
+      "loss": 0.4773,
+      "step": 1431
+    },
+    {
+      "epoch": 1.998837479655894,
+      "grad_norm": 2.0987197531994193,
+      "learning_rate": 3.0276185120292996e-06,
+      "loss": 0.4866,
+      "step": 1432
+    },
+    {
+      "epoch": 2.0013950244129273,
+      "grad_norm": 2.1060512775299687,
+      "learning_rate": 3.0201538964945787e-06,
+      "loss": 0.7126,
+      "step": 1433
+    },
+    {
+      "epoch": 2.0027900488258545,
+      "grad_norm": 1.932749210999598,
+      "learning_rate": 3.0126945105466486e-06,
+      "loss": 0.2665,
+      "step": 1434
+    },
+    {
+      "epoch": 2.004185073238782,
+      "grad_norm": 1.803959519110432,
+      "learning_rate": 3.005240373888812e-06,
+      "loss": 0.2909,
+      "step": 1435
+    },
+    {
+      "epoch": 2.005580097651709,
+      "grad_norm": 1.882695279541564,
+      "learning_rate": 2.9977915062105023e-06,
+      "loss": 0.2668,
+      "step": 1436
+    },
+    {
+      "epoch": 2.0069751220646364,
+      "grad_norm": 1.7001162613390293,
+      "learning_rate": 2.9903479271872416e-06,
+      "loss": 0.2382,
+      "step": 1437
+    },
+    {
+      "epoch": 2.008370146477563,
+      "grad_norm": 1.9030283807992499,
+      "learning_rate": 2.9829096564805804e-06,
+      "loss": 0.2321,
+      "step": 1438
+    },
+    {
+      "epoch": 2.0097651708904904,
+      "grad_norm": 1.7597971492802806,
+      "learning_rate": 2.975476713738043e-06,
+      "loss": 0.2496,
+      "step": 1439
+    },
+    {
+      "epoch": 2.0111601953034177,
+      "grad_norm": 1.7225059934269706,
+      "learning_rate": 2.9680491185930877e-06,
+      "loss": 0.2556,
+      "step": 1440
+    },
+    {
+      "epoch": 2.012555219716345,
+      "grad_norm": 1.8760479786303343,
+      "learning_rate": 2.960626890665044e-06,
+      "loss": 0.2429,
+      "step": 1441
+    },
+    {
+      "epoch": 2.0139502441292723,
+      "grad_norm": 1.8962652973945915,
+      "learning_rate": 2.953210049559062e-06,
+      "loss": 0.2149,
+      "step": 1442
+    },
+    {
+      "epoch": 2.0153452685421995,
+      "grad_norm": 2.384087090505903,
+      "learning_rate": 2.945798614866068e-06,
+      "loss": 0.2684,
+      "step": 1443
+    },
+    {
+      "epoch": 2.016740292955127,
+      "grad_norm": 2.373148278542001,
+      "learning_rate": 2.9383926061627055e-06,
+      "loss": 0.2407,
+      "step": 1444
+    },
+    {
+      "epoch": 2.018135317368054,
+      "grad_norm": 2.994047668095242,
+      "learning_rate": 2.9309920430112825e-06,
+      "loss": 0.2713,
+      "step": 1445
+    },
+    {
+      "epoch": 2.0195303417809813,
+      "grad_norm": 2.41812388777858,
+      "learning_rate": 2.92359694495973e-06,
+      "loss": 0.2561,
+      "step": 1446
+    },
+    {
+      "epoch": 2.0209253661939086,
+      "grad_norm": 2.064667983995233,
+      "learning_rate": 2.9162073315415384e-06,
+      "loss": 0.2361,
+      "step": 1447
+    },
+    {
+      "epoch": 2.0223203906068354,
+      "grad_norm": 2.4543601055709003,
+      "learning_rate": 2.9088232222757085e-06,
+      "loss": 0.2539,
+      "step": 1448
+    },
+    {
+      "epoch": 2.0237154150197627,
+      "grad_norm": 2.098982722115911,
+      "learning_rate": 2.9014446366667115e-06,
+      "loss": 0.2677,
+      "step": 1449
+    },
+    {
+      "epoch": 2.02511043943269,
+      "grad_norm": 2.098306519762135,
+      "learning_rate": 2.8940715942044204e-06,
+      "loss": 0.2583,
+      "step": 1450
+    },
+    {
+      "epoch": 2.0265054638456172,
+      "grad_norm": 2.000843801466561,
+      "learning_rate": 2.8867041143640663e-06,
+      "loss": 0.2395,
+      "step": 1451
+    },
+    {
+      "epoch": 2.0279004882585445,
+      "grad_norm": 2.145854321461008,
+      "learning_rate": 2.8793422166061918e-06,
+      "loss": 0.2821,
+      "step": 1452
+    },
+    {
+      "epoch": 2.029295512671472,
+      "grad_norm": 1.9525784509980673,
+      "learning_rate": 2.8719859203765955e-06,
+      "loss": 0.2263,
+      "step": 1453
+    },
+    {
+      "epoch": 2.030690537084399,
+      "grad_norm": 1.970474839624812,
+      "learning_rate": 2.864635245106272e-06,
+      "loss": 0.2427,
+      "step": 1454
+    },
+    {
+      "epoch": 2.0320855614973263,
+      "grad_norm": 1.9323635610264698,
+      "learning_rate": 2.8572902102113788e-06,
+      "loss": 0.2339,
+      "step": 1455
+    },
+    {
+      "epoch": 2.0334805859102536,
+      "grad_norm": 2.0460140070756276,
+      "learning_rate": 2.849950835093168e-06,
+      "loss": 0.2417,
+      "step": 1456
+    },
+    {
+      "epoch": 2.034875610323181,
+      "grad_norm": 1.8600960559305393,
+      "learning_rate": 2.8426171391379433e-06,
+      "loss": 0.2492,
+      "step": 1457
+    },
+    {
+      "epoch": 2.0362706347361077,
+      "grad_norm": 1.6992216414634733,
+      "learning_rate": 2.835289141717008e-06,
+      "loss": 0.2158,
+      "step": 1458
+    },
+    {
+      "epoch": 2.037665659149035,
+      "grad_norm": 1.9085343998278854,
+      "learning_rate": 2.827966862186616e-06,
+      "loss": 0.2362,
+      "step": 1459
+    },
+    {
+      "epoch": 2.0390606835619622,
+      "grad_norm": 1.9838552421241669,
+      "learning_rate": 2.820650319887911e-06,
+      "loss": 0.2375,
+      "step": 1460
+    },
+    {
+      "epoch": 2.0404557079748895,
+      "grad_norm": 2.0510000137002513,
+      "learning_rate": 2.8133395341468915e-06,
+      "loss": 0.2487,
+      "step": 1461
+    },
+    {
+      "epoch": 2.041850732387817,
+      "grad_norm": 2.0254719491785065,
+      "learning_rate": 2.8060345242743427e-06,
+      "loss": 0.2321,
+      "step": 1462
+    },
+    {
+      "epoch": 2.043245756800744,
+      "grad_norm": 1.7698260233356893,
+      "learning_rate": 2.7987353095657944e-06,
+      "loss": 0.2666,
+      "step": 1463
+    },
+    {
+      "epoch": 2.0446407812136713,
+      "grad_norm": 2.1844227717957545,
+      "learning_rate": 2.7914419093014734e-06,
+      "loss": 0.2418,
+      "step": 1464
+    },
+    {
+      "epoch": 2.0460358056265986,
+      "grad_norm": 1.8986321251662384,
+      "learning_rate": 2.784154342746246e-06,
+      "loss": 0.246,
+      "step": 1465
+    },
+    {
+      "epoch": 2.047430830039526,
+      "grad_norm": 2.1557796026506453,
+      "learning_rate": 2.7768726291495667e-06,
+      "loss": 0.2616,
+      "step": 1466
+    },
+    {
+      "epoch": 2.048825854452453,
+      "grad_norm": 2.1460228298343242,
+      "learning_rate": 2.7695967877454356e-06,
+      "loss": 0.2644,
+      "step": 1467
+    },
+    {
+      "epoch": 2.05022087886538,
+      "grad_norm": 2.183952950578596,
+      "learning_rate": 2.7623268377523356e-06,
+      "loss": 0.2136,
+      "step": 1468
+    },
+    {
+      "epoch": 2.0516159032783072,
+      "grad_norm": 1.8822992647896113,
+      "learning_rate": 2.755062798373189e-06,
+      "loss": 0.2454,
+      "step": 1469
+    },
+    {
+      "epoch": 2.0530109276912345,
+      "grad_norm": 2.0587391825233907,
+      "learning_rate": 2.747804688795311e-06,
+      "loss": 0.2572,
+      "step": 1470
+    },
+    {
+      "epoch": 2.0544059521041618,
+      "grad_norm": 2.0337932576382194,
+      "learning_rate": 2.7405525281903506e-06,
+      "loss": 0.2095,
+      "step": 1471
+    },
+    {
+      "epoch": 2.055800976517089,
+      "grad_norm": 2.180806793212479,
+      "learning_rate": 2.7333063357142414e-06,
+      "loss": 0.2398,
+      "step": 1472
+    },
+    {
+      "epoch": 2.0571960009300163,
+      "grad_norm": 2.1427495369403773,
+      "learning_rate": 2.7260661305071523e-06,
+      "loss": 0.2217,
+      "step": 1473
+    },
+    {
+      "epoch": 2.0585910253429436,
+      "grad_norm": 2.038226675373655,
+      "learning_rate": 2.718831931693443e-06,
+      "loss": 0.2492,
+      "step": 1474
+    },
+    {
+      "epoch": 2.059986049755871,
+      "grad_norm": 1.9218578361086367,
+      "learning_rate": 2.7116037583816e-06,
+      "loss": 0.2498,
+      "step": 1475
+    },
+    {
+      "epoch": 2.061381074168798,
+      "grad_norm": 2.0508632152717494,
+      "learning_rate": 2.7043816296642005e-06,
+      "loss": 0.2331,
+      "step": 1476
+    },
+    {
+      "epoch": 2.0627760985817254,
+      "grad_norm": 2.1232429074513224,
+      "learning_rate": 2.6971655646178544e-06,
+      "loss": 0.2612,
+      "step": 1477
+    },
+    {
+      "epoch": 2.064171122994652,
+      "grad_norm": 1.8524908410533427,
+      "learning_rate": 2.689955582303152e-06,
+      "loss": 0.2165,
+      "step": 1478
+    },
+    {
+      "epoch": 2.0655661474075795,
+      "grad_norm": 1.8152481510782883,
+      "learning_rate": 2.6827517017646154e-06,
+      "loss": 0.2245,
+      "step": 1479
+    },
+    {
+      "epoch": 2.0669611718205068,
+      "grad_norm": 2.0764635863149516,
+      "learning_rate": 2.6755539420306565e-06,
+      "loss": 0.2258,
+      "step": 1480
+    },
+    {
+      "epoch": 2.068356196233434,
+      "grad_norm": 1.720048095036099,
+      "learning_rate": 2.668362322113512e-06,
+      "loss": 0.2451,
+      "step": 1481
+    },
+    {
+      "epoch": 2.0697512206463613,
+      "grad_norm": 1.8692404797467597,
+      "learning_rate": 2.661176861009205e-06,
+      "loss": 0.2406,
+      "step": 1482
+    },
+    {
+      "epoch": 2.0711462450592886,
+      "grad_norm": 2.0026676185323313,
+      "learning_rate": 2.6539975776974926e-06,
+      "loss": 0.2306,
+      "step": 1483
+    },
+    {
+      "epoch": 2.072541269472216,
+      "grad_norm": 1.9167887054562096,
+      "learning_rate": 2.646824491141807e-06,
+      "loss": 0.2211,
+      "step": 1484
+    },
+    {
+      "epoch": 2.073936293885143,
+      "grad_norm": 1.94450012415561,
+      "learning_rate": 2.6396576202892176e-06,
+      "loss": 0.2207,
+      "step": 1485
+    },
+    {
+      "epoch": 2.0753313182980704,
+      "grad_norm": 1.8160393968583468,
+      "learning_rate": 2.632496984070375e-06,
+      "loss": 0.2144,
+      "step": 1486
+    },
+    {
+      "epoch": 2.0767263427109977,
+      "grad_norm": 2.041628010035314,
+      "learning_rate": 2.6253426013994586e-06,
+      "loss": 0.2523,
+      "step": 1487
+    },
+    {
+      "epoch": 2.0781213671239245,
+      "grad_norm": 2.134712711238633,
+      "learning_rate": 2.6181944911741333e-06,
+      "loss": 0.2525,
+      "step": 1488
+    },
+    {
+      "epoch": 2.0795163915368517,
+      "grad_norm": 2.113789784111181,
+      "learning_rate": 2.6110526722754955e-06,
+      "loss": 0.2679,
+      "step": 1489
+    },
+    {
+      "epoch": 2.080911415949779,
+      "grad_norm": 2.169975003010366,
+      "learning_rate": 2.603917163568021e-06,
+      "loss": 0.2487,
+      "step": 1490
+    },
+    {
+      "epoch": 2.0823064403627063,
+      "grad_norm": 1.8802285332237876,
+      "learning_rate": 2.5967879838995176e-06,
+      "loss": 0.2268,
+      "step": 1491
+    },
+    {
+      "epoch": 2.0837014647756336,
+      "grad_norm": 2.1346446458657375,
+      "learning_rate": 2.589665152101081e-06,
+      "loss": 0.2444,
+      "step": 1492
+    },
+    {
+      "epoch": 2.085096489188561,
+      "grad_norm": 1.995965282834599,
+      "learning_rate": 2.582548686987031e-06,
+      "loss": 0.233,
+      "step": 1493
+    },
+    {
+      "epoch": 2.086491513601488,
+      "grad_norm": 1.9479032032819366,
+      "learning_rate": 2.5754386073548775e-06,
+      "loss": 0.2245,
+      "step": 1494
+    },
+    {
+      "epoch": 2.0878865380144154,
+      "grad_norm": 1.981745320689665,
+      "learning_rate": 2.5683349319852647e-06,
+      "loss": 0.2411,
+      "step": 1495
+    },
+    {
+      "epoch": 2.0892815624273426,
+      "grad_norm": 1.9301685232993138,
+      "learning_rate": 2.5612376796419126e-06,
+      "loss": 0.2382,
+      "step": 1496
+    },
+    {
+      "epoch": 2.09067658684027,
+      "grad_norm": 2.1167901194901884,
+      "learning_rate": 2.5541468690715797e-06,
+      "loss": 0.2461,
+      "step": 1497
+    },
+    {
+      "epoch": 2.0920716112531967,
+      "grad_norm": 1.9947152398221217,
+      "learning_rate": 2.5470625190040105e-06,
+      "loss": 0.249,
+      "step": 1498
+    },
+    {
+      "epoch": 2.093466635666124,
+      "grad_norm": 2.0494633364154646,
+      "learning_rate": 2.5399846481518857e-06,
+      "loss": 0.2499,
+      "step": 1499
+    },
+    {
+      "epoch": 2.0948616600790513,
+      "grad_norm": 1.950090286363791,
+      "learning_rate": 2.5329132752107675e-06,
+      "loss": 0.2621,
+      "step": 1500
+    },
+    {
+      "epoch": 2.0962566844919786,
+      "grad_norm": 2.1301861314715924,
+      "learning_rate": 2.525848418859055e-06,
+      "loss": 0.2521,
+      "step": 1501
+    },
+    {
+      "epoch": 2.097651708904906,
+      "grad_norm": 1.771798927570182,
+      "learning_rate": 2.518790097757938e-06,
+      "loss": 0.2204,
+      "step": 1502
+    },
+    {
+      "epoch": 2.099046733317833,
+      "grad_norm": 1.8590300077093187,
+      "learning_rate": 2.51173833055134e-06,
+      "loss": 0.2212,
+      "step": 1503
+    },
+    {
+      "epoch": 2.1004417577307604,
+      "grad_norm": 2.063306202589427,
+      "learning_rate": 2.504693135865875e-06,
+      "loss": 0.2621,
+      "step": 1504
+    },
+    {
+      "epoch": 2.1018367821436876,
+      "grad_norm": 2.02342129140672,
+      "learning_rate": 2.497654532310799e-06,
+      "loss": 0.2454,
+      "step": 1505
+    },
+    {
+      "epoch": 2.103231806556615,
+      "grad_norm": 1.8990466972330668,
+      "learning_rate": 2.490622538477952e-06,
+      "loss": 0.2406,
+      "step": 1506
+    },
+    {
+      "epoch": 2.104626830969542,
+      "grad_norm": 2.1590362411607464,
+      "learning_rate": 2.483597172941718e-06,
+      "loss": 0.2375,
+      "step": 1507
+    },
+    {
+      "epoch": 2.106021855382469,
+      "grad_norm": 2.0545993680320596,
+      "learning_rate": 2.4765784542589754e-06,
+      "loss": 0.2492,
+      "step": 1508
+    },
+    {
+      "epoch": 2.1074168797953963,
+      "grad_norm": 2.038598849419707,
+      "learning_rate": 2.46956640096904e-06,
+      "loss": 0.2073,
+      "step": 1509
+    },
+    {
+      "epoch": 2.1088119042083235,
+      "grad_norm": 1.8605747435871027,
+      "learning_rate": 2.4625610315936267e-06,
+      "loss": 0.2454,
+      "step": 1510
+    },
+    {
+      "epoch": 2.110206928621251,
+      "grad_norm": 2.1562619973708057,
+      "learning_rate": 2.4555623646367952e-06,
+      "loss": 0.2756,
+      "step": 1511
+    },
+    {
+      "epoch": 2.111601953034178,
+      "grad_norm": 2.0329623515500246,
+      "learning_rate": 2.448570418584898e-06,
+      "loss": 0.2461,
+      "step": 1512
+    },
+    {
+      "epoch": 2.1129969774471054,
+      "grad_norm": 1.953214188504134,
+      "learning_rate": 2.4415852119065343e-06,
+      "loss": 0.247,
+      "step": 1513
+    },
+    {
+      "epoch": 2.1143920018600326,
+      "grad_norm": 2.0529609432367235,
+      "learning_rate": 2.4346067630525084e-06,
+      "loss": 0.24,
+      "step": 1514
+    },
+    {
+      "epoch": 2.11578702627296,
+      "grad_norm": 2.0799839572137695,
+      "learning_rate": 2.427635090455766e-06,
+      "loss": 0.229,
+      "step": 1515
+    },
+    {
+      "epoch": 2.117182050685887,
+      "grad_norm": 2.0126307201591014,
+      "learning_rate": 2.42067021253136e-06,
+      "loss": 0.2339,
+      "step": 1516
+    },
+    {
+      "epoch": 2.1185770750988144,
+      "grad_norm": 1.8308783448148316,
+      "learning_rate": 2.4137121476763965e-06,
+      "loss": 0.2071,
+      "step": 1517
+    },
+    {
+      "epoch": 2.1199720995117413,
+      "grad_norm": 1.9553265656138026,
+      "learning_rate": 2.4067609142699798e-06,
+      "loss": 0.2215,
+      "step": 1518
+    },
+    {
+      "epoch": 2.1213671239246685,
+      "grad_norm": 2.258180433222967,
+      "learning_rate": 2.3998165306731713e-06,
+      "loss": 0.2444,
+      "step": 1519
+    },
+    {
+      "epoch": 2.122762148337596,
+      "grad_norm": 2.028159380935348,
+      "learning_rate": 2.3928790152289443e-06,
+      "loss": 0.2116,
+      "step": 1520
+    },
+    {
+      "epoch": 2.124157172750523,
+      "grad_norm": 1.9801498809540945,
+      "learning_rate": 2.385948386262123e-06,
+      "loss": 0.2602,
+      "step": 1521
+    },
+    {
+      "epoch": 2.1255521971634503,
+      "grad_norm": 2.2663540920146863,
+      "learning_rate": 2.3790246620793466e-06,
+      "loss": 0.2457,
+      "step": 1522
+    },
+    {
+      "epoch": 2.1269472215763776,
+      "grad_norm": 2.101327168537436,
+      "learning_rate": 2.372107860969019e-06,
+      "loss": 0.2306,
+      "step": 1523
+    },
+    {
+      "epoch": 2.128342245989305,
+      "grad_norm": 1.9895205668500822,
+      "learning_rate": 2.3651980012012454e-06,
+      "loss": 0.2166,
+      "step": 1524
+    },
+    {
+      "epoch": 2.129737270402232,
+      "grad_norm": 1.8300310484062896,
+      "learning_rate": 2.358295101027807e-06,
+      "loss": 0.2249,
+      "step": 1525
+    },
+    {
+      "epoch": 2.1311322948151594,
+      "grad_norm": 1.765470113307758,
+      "learning_rate": 2.351399178682101e-06,
+      "loss": 0.2284,
+      "step": 1526
+    },
+    {
+      "epoch": 2.1325273192280867,
+      "grad_norm": 2.13174824831439,
+      "learning_rate": 2.3445102523790876e-06,
+      "loss": 0.2198,
+      "step": 1527
+    },
+    {
+      "epoch": 2.1339223436410135,
+      "grad_norm": 1.9155367612459846,
+      "learning_rate": 2.3376283403152527e-06,
+      "loss": 0.2256,
+      "step": 1528
+    },
+    {
+      "epoch": 2.135317368053941,
+      "grad_norm": 1.9875172700280743,
+      "learning_rate": 2.330753460668553e-06,
+      "loss": 0.2393,
+      "step": 1529
+    },
+    {
+      "epoch": 2.136712392466868,
+      "grad_norm": 1.8406312091494403,
+      "learning_rate": 2.323885631598366e-06,
+      "loss": 0.2396,
+      "step": 1530
+    },
+    {
+      "epoch": 2.1381074168797953,
+      "grad_norm": 1.9374184950557762,
+      "learning_rate": 2.3170248712454525e-06,
+      "loss": 0.2461,
+      "step": 1531
+    },
+    {
+      "epoch": 2.1395024412927226,
+      "grad_norm": 1.963431102509043,
+      "learning_rate": 2.3101711977318995e-06,
+      "loss": 0.2642,
+      "step": 1532
+    },
+    {
+      "epoch": 2.14089746570565,
+      "grad_norm": 2.0376501014600716,
+      "learning_rate": 2.3033246291610717e-06,
+      "loss": 0.2291,
+      "step": 1533
+    },
+    {
+      "epoch": 2.142292490118577,
+      "grad_norm": 2.074177144588648,
+      "learning_rate": 2.2964851836175705e-06,
+      "loss": 0.2508,
+      "step": 1534
+    },
+    {
+      "epoch": 2.1436875145315044,
+      "grad_norm": 2.033089929760175,
+      "learning_rate": 2.2896528791671807e-06,
+      "loss": 0.2301,
+      "step": 1535
+    },
+    {
+      "epoch": 2.1450825389444317,
+      "grad_norm": 1.9262718887772488,
+      "learning_rate": 2.2828277338568226e-06,
+      "loss": 0.2316,
+      "step": 1536
+    },
+    {
+      "epoch": 2.146477563357359,
+      "grad_norm": 2.0757309335688165,
+      "learning_rate": 2.2760097657145096e-06,
+      "loss": 0.2404,
+      "step": 1537
+    },
+    {
+      "epoch": 2.147872587770286,
+      "grad_norm": 1.9444214641725062,
+      "learning_rate": 2.2691989927492984e-06,
+      "loss": 0.2594,
+      "step": 1538
+    },
+    {
+      "epoch": 2.149267612183213,
+      "grad_norm": 2.198150007883586,
+      "learning_rate": 2.262395432951235e-06,
+      "loss": 0.264,
+      "step": 1539
+    },
+    {
+      "epoch": 2.1506626365961403,
+      "grad_norm": 1.9616462283529785,
+      "learning_rate": 2.2555991042913177e-06,
+      "loss": 0.2277,
+      "step": 1540
+    },
+    {
+      "epoch": 2.1520576610090676,
+      "grad_norm": 2.090474806693938,
+      "learning_rate": 2.248810024721441e-06,
+      "loss": 0.2245,
+      "step": 1541
+    },
+    {
+      "epoch": 2.153452685421995,
+      "grad_norm": 1.9854907398753259,
+      "learning_rate": 2.2420282121743513e-06,
+      "loss": 0.2384,
+      "step": 1542
+    },
+    {
+      "epoch": 2.154847709834922,
+      "grad_norm": 1.978976806270666,
+      "learning_rate": 2.235253684563602e-06,
+      "loss": 0.225,
+      "step": 1543
+    },
+    {
+      "epoch": 2.1562427342478494,
+      "grad_norm": 2.0143674773479088,
+      "learning_rate": 2.228486459783506e-06,
+      "loss": 0.2204,
+      "step": 1544
+    },
+    {
+      "epoch": 2.1576377586607767,
+      "grad_norm": 1.8497292458468502,
+      "learning_rate": 2.221726555709079e-06,
+      "loss": 0.2332,
+      "step": 1545
+    },
+    {
+      "epoch": 2.159032783073704,
+      "grad_norm": 2.041422679794062,
+      "learning_rate": 2.2149739901960088e-06,
+      "loss": 0.2502,
+      "step": 1546
+    },
+    {
+      "epoch": 2.160427807486631,
+      "grad_norm": 1.9189040453437212,
+      "learning_rate": 2.208228781080592e-06,
+      "loss": 0.2278,
+      "step": 1547
+    },
+    {
+      "epoch": 2.161822831899558,
+      "grad_norm": 1.9937984946523888,
+      "learning_rate": 2.201490946179696e-06,
+      "loss": 0.2222,
+      "step": 1548
+    },
+    {
+      "epoch": 2.1632178563124853,
+      "grad_norm": 1.901075385011242,
+      "learning_rate": 2.19476050329071e-06,
+      "loss": 0.2152,
+      "step": 1549
+    },
+    {
+      "epoch": 2.1646128807254126,
+      "grad_norm": 2.0137231979004286,
+      "learning_rate": 2.188037470191502e-06,
+      "loss": 0.2288,
+      "step": 1550
+    },
+    {
+      "epoch": 2.16600790513834,
+      "grad_norm": 1.7888962568094866,
+      "learning_rate": 2.181321864640362e-06,
+      "loss": 0.229,
+      "step": 1551
+    },
+    {
+      "epoch": 2.167402929551267,
+      "grad_norm": 2.0707458853980563,
+      "learning_rate": 2.1746137043759594e-06,
+      "loss": 0.2098,
+      "step": 1552
+    },
+    {
+      "epoch": 2.1687979539641944,
+      "grad_norm": 1.8413823655100272,
+      "learning_rate": 2.167913007117306e-06,
+      "loss": 0.2245,
+      "step": 1553
+    },
+    {
+      "epoch": 2.1701929783771217,
+      "grad_norm": 2.0529948691138524,
+      "learning_rate": 2.1612197905636913e-06,
+      "loss": 0.2544,
+      "step": 1554
+    },
+    {
+      "epoch": 2.171588002790049,
+      "grad_norm": 2.0489852227371586,
+      "learning_rate": 2.154534072394651e-06,
+      "loss": 0.235,
+      "step": 1555
+    },
+    {
+      "epoch": 2.172983027202976,
+      "grad_norm": 2.1022639361713216,
+      "learning_rate": 2.147855870269916e-06,
+      "loss": 0.2493,
+      "step": 1556
+    },
+    {
+      "epoch": 2.1743780516159035,
+      "grad_norm": 1.9016592054329882,
+      "learning_rate": 2.1411852018293583e-06,
+      "loss": 0.2737,
+      "step": 1557
+    },
+    {
+      "epoch": 2.1757730760288303,
+      "grad_norm": 2.017634751042281,
+      "learning_rate": 2.1345220846929514e-06,
+      "loss": 0.2296,
+      "step": 1558
+    },
+    {
+      "epoch": 2.1771681004417576,
+      "grad_norm": 2.0588142168838957,
+      "learning_rate": 2.127866536460727e-06,
+      "loss": 0.2306,
+      "step": 1559
+    },
+    {
+      "epoch": 2.178563124854685,
+      "grad_norm": 1.844060249249721,
+      "learning_rate": 2.1212185747127235e-06,
+      "loss": 0.2397,
+      "step": 1560
+    },
+    {
+      "epoch": 2.179958149267612,
+      "grad_norm": 1.8836100804054643,
+      "learning_rate": 2.1145782170089346e-06,
+      "loss": 0.226,
+      "step": 1561
+    },
+    {
+      "epoch": 2.1813531736805394,
+      "grad_norm": 2.027342005832641,
+      "learning_rate": 2.107945480889276e-06,
+      "loss": 0.2211,
+      "step": 1562
+    },
+    {
+      "epoch": 2.1827481980934667,
+      "grad_norm": 2.0716969692058336,
+      "learning_rate": 2.1013203838735273e-06,
+      "loss": 0.2313,
+      "step": 1563
+    },
+    {
+      "epoch": 2.184143222506394,
+      "grad_norm": 1.87998419305253,
+      "learning_rate": 2.094702943461289e-06,
+      "loss": 0.2265,
+      "step": 1564
+    },
+    {
+      "epoch": 2.185538246919321,
+      "grad_norm": 2.083535070952125,
+      "learning_rate": 2.0880931771319395e-06,
+      "loss": 0.2606,
+      "step": 1565
+    },
+    {
+      "epoch": 2.1869332713322485,
+      "grad_norm": 2.0437831993419833,
+      "learning_rate": 2.0814911023445904e-06,
+      "loss": 0.2067,
+      "step": 1566
+    },
+    {
+      "epoch": 2.1883282957451757,
+      "grad_norm": 1.927662318132159,
+      "learning_rate": 2.0748967365380292e-06,
+      "loss": 0.229,
+      "step": 1567
+    },
+    {
+      "epoch": 2.1897233201581026,
+      "grad_norm": 1.9753615761556702,
+      "learning_rate": 2.0683100971306873e-06,
+      "loss": 0.2482,
+      "step": 1568
+    },
+    {
+      "epoch": 2.19111834457103,
+      "grad_norm": 2.1292063771141496,
+      "learning_rate": 2.0617312015205844e-06,
+      "loss": 0.2629,
+      "step": 1569
+    },
+    {
+      "epoch": 2.192513368983957,
+      "grad_norm": 2.03833931308385,
+      "learning_rate": 2.055160067085283e-06,
+      "loss": 0.2224,
+      "step": 1570
+    },
+    {
+      "epoch": 2.1939083933968844,
+      "grad_norm": 1.99345008089848,
+      "learning_rate": 2.0485967111818506e-06,
+      "loss": 0.2465,
+      "step": 1571
+    },
+    {
+      "epoch": 2.1953034178098116,
+      "grad_norm": 1.8416073370699406,
+      "learning_rate": 2.0420411511468086e-06,
+      "loss": 0.2231,
+      "step": 1572
+    },
+    {
+      "epoch": 2.196698442222739,
+      "grad_norm": 2.022634849681668,
+      "learning_rate": 2.0354934042960804e-06,
+      "loss": 0.266,
+      "step": 1573
+    },
+    {
+      "epoch": 2.198093466635666,
+      "grad_norm": 1.9543607699785175,
+      "learning_rate": 2.0289534879249544e-06,
+      "loss": 0.2215,
+      "step": 1574
+    },
+    {
+      "epoch": 2.1994884910485935,
+      "grad_norm": 2.120691875324055,
+      "learning_rate": 2.0224214193080394e-06,
+      "loss": 0.2411,
+      "step": 1575
+    },
+    {
+      "epoch": 2.2008835154615207,
+      "grad_norm": 1.9968502256620426,
+      "learning_rate": 2.015897215699208e-06,
+      "loss": 0.2407,
+      "step": 1576
+    },
+    {
+      "epoch": 2.202278539874448,
+      "grad_norm": 1.972527239657243,
+      "learning_rate": 2.0093808943315636e-06,
+      "loss": 0.213,
+      "step": 1577
+    },
+    {
+      "epoch": 2.2036735642873753,
+      "grad_norm": 1.894788749194984,
+      "learning_rate": 2.0028724724173886e-06,
+      "loss": 0.2199,
+      "step": 1578
+    },
+    {
+      "epoch": 2.205068588700302,
+      "grad_norm": 1.9941924804567333,
+      "learning_rate": 1.996371967148098e-06,
+      "loss": 0.2255,
+      "step": 1579
+    },
+    {
+      "epoch": 2.2064636131132294,
+      "grad_norm": 2.1213230860539154,
+      "learning_rate": 1.989879395694194e-06,
+      "loss": 0.2711,
+      "step": 1580
+    },
+    {
+      "epoch": 2.2078586375261566,
+      "grad_norm": 2.005074782180638,
+      "learning_rate": 1.9833947752052286e-06,
+      "loss": 0.2191,
+      "step": 1581
+    },
+    {
+      "epoch": 2.209253661939084,
+      "grad_norm": 2.0916112373693663,
+      "learning_rate": 1.976918122809744e-06,
+      "loss": 0.2423,
+      "step": 1582
+    },
+    {
+      "epoch": 2.210648686352011,
+      "grad_norm": 1.943662048921364,
+      "learning_rate": 1.9704494556152413e-06,
+      "loss": 0.2364,
+      "step": 1583
+    },
+    {
+      "epoch": 2.2120437107649384,
+      "grad_norm": 1.8072918947965364,
+      "learning_rate": 1.9639887907081297e-06,
+      "loss": 0.2101,
+      "step": 1584
+    },
+    {
+      "epoch": 2.2134387351778657,
+      "grad_norm": 2.074058894664055,
+      "learning_rate": 1.9575361451536772e-06,
+      "loss": 0.2528,
+      "step": 1585
+    },
+    {
+      "epoch": 2.214833759590793,
+      "grad_norm": 1.948267902890602,
+      "learning_rate": 1.9510915359959694e-06,
+      "loss": 0.2052,
+      "step": 1586
+    },
+    {
+      "epoch": 2.2162287840037203,
+      "grad_norm": 1.6977361690821569,
+      "learning_rate": 1.944654980257869e-06,
+      "loss": 0.2248,
+      "step": 1587
+    },
+    {
+      "epoch": 2.217623808416647,
+      "grad_norm": 1.9851481647242695,
+      "learning_rate": 1.9382264949409614e-06,
+      "loss": 0.2334,
+      "step": 1588
+    },
+    {
+      "epoch": 2.2190188328295744,
+      "grad_norm": 2.0409126097908628,
+      "learning_rate": 1.931806097025517e-06,
+      "loss": 0.2512,
+      "step": 1589
+    },
+    {
+      "epoch": 2.2204138572425016,
+      "grad_norm": 2.193027559366708,
+      "learning_rate": 1.925393803470447e-06,
+      "loss": 0.2132,
+      "step": 1590
+    },
+    {
+      "epoch": 2.221808881655429,
+      "grad_norm": 1.962723403198113,
+      "learning_rate": 1.9189896312132506e-06,
+      "loss": 0.2341,
+      "step": 1591
+    },
+    {
+      "epoch": 2.223203906068356,
+      "grad_norm": 2.025113111819309,
+      "learning_rate": 1.912593597169975e-06,
+      "loss": 0.2081,
+      "step": 1592
+    },
+    {
+      "epoch": 2.2245989304812834,
+      "grad_norm": 2.0420629608721126,
+      "learning_rate": 1.9062057182351768e-06,
+      "loss": 0.2374,
+      "step": 1593
+    },
+    {
+      "epoch": 2.2259939548942107,
+      "grad_norm": 1.8896161037166734,
+      "learning_rate": 1.899826011281865e-06,
+      "loss": 0.2551,
+      "step": 1594
+    },
+    {
+      "epoch": 2.227388979307138,
+      "grad_norm": 1.906560235493038,
+      "learning_rate": 1.893454493161468e-06,
+      "loss": 0.2208,
+      "step": 1595
+    },
+    {
+      "epoch": 2.2287840037200652,
+      "grad_norm": 1.8248556594089396,
+      "learning_rate": 1.8870911807037856e-06,
+      "loss": 0.2283,
+      "step": 1596
+    },
+    {
+      "epoch": 2.2301790281329925,
+      "grad_norm": 2.1102132026488745,
+      "learning_rate": 1.8807360907169326e-06,
+      "loss": 0.2577,
+      "step": 1597
+    },
+    {
+      "epoch": 2.23157405254592,
+      "grad_norm": 1.9791563851389666,
+      "learning_rate": 1.8743892399873154e-06,
+      "loss": 0.2252,
+      "step": 1598
+    },
+    {
+      "epoch": 2.2329690769588466,
+      "grad_norm": 2.049090710717768,
+      "learning_rate": 1.868050645279576e-06,
+      "loss": 0.2368,
+      "step": 1599
+    },
+    {
+      "epoch": 2.234364101371774,
+      "grad_norm": 1.9669615346401133,
+      "learning_rate": 1.8617203233365427e-06,
+      "loss": 0.2306,
+      "step": 1600
+    },
+    {
+      "epoch": 2.235759125784701,
+      "grad_norm": 1.9525208909026137,
+      "learning_rate": 1.8553982908792e-06,
+      "loss": 0.2372,
+      "step": 1601
+    },
+    {
+      "epoch": 2.2371541501976284,
+      "grad_norm": 2.179736854299681,
+      "learning_rate": 1.8490845646066303e-06,
+      "loss": 0.2783,
+      "step": 1602
+    },
+    {
+      "epoch": 2.2385491746105557,
+      "grad_norm": 2.295070889351553,
+      "learning_rate": 1.8427791611959762e-06,
+      "loss": 0.2567,
+      "step": 1603
+    },
+    {
+      "epoch": 2.239944199023483,
+      "grad_norm": 1.8673932533442659,
+      "learning_rate": 1.8364820973024e-06,
+      "loss": 0.2154,
+      "step": 1604
+    },
+    {
+      "epoch": 2.2413392234364102,
+      "grad_norm": 2.1778592782550317,
+      "learning_rate": 1.8301933895590362e-06,
+      "loss": 0.2522,
+      "step": 1605
+    },
+    {
+      "epoch": 2.2427342478493375,
+      "grad_norm": 1.953126746209516,
+      "learning_rate": 1.8239130545769408e-06,
+      "loss": 0.2189,
+      "step": 1606
+    },
+    {
+      "epoch": 2.2441292722622648,
+      "grad_norm": 1.9964448323232684,
+      "learning_rate": 1.8176411089450618e-06,
+      "loss": 0.2524,
+      "step": 1607
+    },
+    {
+      "epoch": 2.2455242966751916,
+      "grad_norm": 1.9567449377882402,
+      "learning_rate": 1.8113775692301822e-06,
+      "loss": 0.2342,
+      "step": 1608
+    },
+    {
+      "epoch": 2.246919321088119,
+      "grad_norm": 2.1009476547919776,
+      "learning_rate": 1.8051224519768817e-06,
+      "loss": 0.2391,
+      "step": 1609
+    },
+    {
+      "epoch": 2.248314345501046,
+      "grad_norm": 1.9587113710355928,
+      "learning_rate": 1.7988757737074959e-06,
+      "loss": 0.2179,
+      "step": 1610
+    },
+    {
+      "epoch": 2.2497093699139734,
+      "grad_norm": 2.0040293159213354,
+      "learning_rate": 1.7926375509220695e-06,
+      "loss": 0.2222,
+      "step": 1611
+    },
+    {
+      "epoch": 2.2511043943269007,
+      "grad_norm": 2.0888632269335354,
+      "learning_rate": 1.7864078000983076e-06,
+      "loss": 0.2109,
+      "step": 1612
+    },
+    {
+      "epoch": 2.252499418739828,
+      "grad_norm": 2.039750099246385,
+      "learning_rate": 1.7801865376915451e-06,
+      "loss": 0.2215,
+      "step": 1613
+    },
+    {
+      "epoch": 2.2538944431527552,
+      "grad_norm": 2.0748993098994952,
+      "learning_rate": 1.7739737801346895e-06,
+      "loss": 0.2469,
+      "step": 1614
+    },
+    {
+      "epoch": 2.2552894675656825,
+      "grad_norm": 1.8716892318272644,
+      "learning_rate": 1.7677695438381831e-06,
+      "loss": 0.2573,
+      "step": 1615
+    },
+    {
+      "epoch": 2.2566844919786098,
+      "grad_norm": 2.09042965134009,
+      "learning_rate": 1.761573845189965e-06,
+      "loss": 0.2367,
+      "step": 1616
+    },
+    {
+      "epoch": 2.258079516391537,
+      "grad_norm": 1.7936694510013214,
+      "learning_rate": 1.7553867005554215e-06,
+      "loss": 0.2068,
+      "step": 1617
+    },
+    {
+      "epoch": 2.2594745408044643,
+      "grad_norm": 1.9479746253603065,
+      "learning_rate": 1.7492081262773397e-06,
+      "loss": 0.2304,
+      "step": 1618
+    },
+    {
+      "epoch": 2.260869565217391,
+      "grad_norm": 2.134206025310751,
+      "learning_rate": 1.7430381386758748e-06,
+      "loss": 0.2447,
+      "step": 1619
+    },
+    {
+      "epoch": 2.2622645896303184,
+      "grad_norm": 1.9220137353156832,
+      "learning_rate": 1.7368767540484965e-06,
+      "loss": 0.2228,
+      "step": 1620
+    },
+    {
+      "epoch": 2.2636596140432457,
+      "grad_norm": 1.9039789497758535,
+      "learning_rate": 1.7307239886699546e-06,
+      "loss": 0.2541,
+      "step": 1621
+    },
+    {
+      "epoch": 2.265054638456173,
+      "grad_norm": 2.2671396189025774,
+      "learning_rate": 1.7245798587922263e-06,
+      "loss": 0.2414,
+      "step": 1622
+    },
+    {
+      "epoch": 2.2664496628691,
+      "grad_norm": 1.8312831188481675,
+      "learning_rate": 1.7184443806444851e-06,
+      "loss": 0.2221,
+      "step": 1623
+    },
+    {
+      "epoch": 2.2678446872820275,
+      "grad_norm": 1.7794976279581878,
+      "learning_rate": 1.7123175704330514e-06,
+      "loss": 0.2194,
+      "step": 1624
+    },
+    {
+      "epoch": 2.2692397116949548,
+      "grad_norm": 1.910542117265865,
+      "learning_rate": 1.706199444341341e-06,
+      "loss": 0.2393,
+      "step": 1625
+    },
+    {
+      "epoch": 2.270634736107882,
+      "grad_norm": 2.0701674846292213,
+      "learning_rate": 1.7000900185298418e-06,
+      "loss": 0.2729,
+      "step": 1626
+    },
+    {
+      "epoch": 2.2720297605208093,
+      "grad_norm": 2.167518255074453,
+      "learning_rate": 1.6939893091360577e-06,
+      "loss": 0.2609,
+      "step": 1627
+    },
+    {
+      "epoch": 2.273424784933736,
+      "grad_norm": 2.0346558592689963,
+      "learning_rate": 1.6878973322744658e-06,
+      "loss": 0.2266,
+      "step": 1628
+    },
+    {
+      "epoch": 2.2748198093466634,
+      "grad_norm": 1.9628822161613053,
+      "learning_rate": 1.6818141040364816e-06,
+      "loss": 0.2307,
+      "step": 1629
+    },
+    {
+      "epoch": 2.2762148337595907,
+      "grad_norm": 1.9799083641231046,
+      "learning_rate": 1.6757396404904087e-06,
+      "loss": 0.2243,
+      "step": 1630
+    },
+    {
+      "epoch": 2.277609858172518,
+      "grad_norm": 1.9097000725178894,
+      "learning_rate": 1.6696739576813981e-06,
+      "loss": 0.2401,
+      "step": 1631
+    },
+    {
+      "epoch": 2.279004882585445,
+      "grad_norm": 1.9065958749954741,
+      "learning_rate": 1.6636170716314114e-06,
+      "loss": 0.2659,
+      "step": 1632
+    },
+    {
+      "epoch": 2.2803999069983725,
+      "grad_norm": 2.185791921800934,
+      "learning_rate": 1.657568998339175e-06,
+      "loss": 0.2615,
+      "step": 1633
+    },
+    {
+      "epoch": 2.2817949314112997,
+      "grad_norm": 2.2995275927534196,
+      "learning_rate": 1.6515297537801305e-06,
+      "loss": 0.2491,
+      "step": 1634
+    },
+    {
+      "epoch": 2.283189955824227,
+      "grad_norm": 1.9425924150562717,
+      "learning_rate": 1.6454993539064075e-06,
+      "loss": 0.2178,
+      "step": 1635
+    },
+    {
+      "epoch": 2.2845849802371543,
+      "grad_norm": 1.8938205464951827,
+      "learning_rate": 1.6394778146467672e-06,
+      "loss": 0.2665,
+      "step": 1636
+    },
+    {
+      "epoch": 2.2859800046500816,
+      "grad_norm": 2.1852392869275716,
+      "learning_rate": 1.6334651519065658e-06,
+      "loss": 0.2576,
+      "step": 1637
+    },
+    {
+      "epoch": 2.287375029063009,
+      "grad_norm": 2.09160013513525,
+      "learning_rate": 1.6274613815677176e-06,
+      "loss": 0.2468,
+      "step": 1638
+    },
+    {
+      "epoch": 2.2887700534759357,
+      "grad_norm": 2.2221912807762,
+      "learning_rate": 1.6214665194886474e-06,
+      "loss": 0.2001,
+      "step": 1639
+    },
+    {
+      "epoch": 2.290165077888863,
+      "grad_norm": 1.8322716677994173,
+      "learning_rate": 1.6154805815042457e-06,
+      "loss": 0.2098,
+      "step": 1640
+    },
+    {
+      "epoch": 2.29156010230179,
+      "grad_norm": 2.1005552341097244,
+      "learning_rate": 1.6095035834258365e-06,
+      "loss": 0.2459,
+      "step": 1641
+    },
+    {
+      "epoch": 2.2929551267147175,
+      "grad_norm": 1.9045088391363771,
+      "learning_rate": 1.6035355410411252e-06,
+      "loss": 0.2003,
+      "step": 1642
+    },
+    {
+      "epoch": 2.2943501511276447,
+      "grad_norm": 2.0139872656400786,
+      "learning_rate": 1.5975764701141611e-06,
+      "loss": 0.2358,
+      "step": 1643
+    },
+    {
+      "epoch": 2.295745175540572,
+      "grad_norm": 1.9262075355870194,
+      "learning_rate": 1.5916263863853e-06,
+      "loss": 0.2404,
+      "step": 1644
+    },
+    {
+      "epoch": 2.2971401999534993,
+      "grad_norm": 2.201421447614601,
+      "learning_rate": 1.585685305571159e-06,
+      "loss": 0.2669,
+      "step": 1645
+    },
+    {
+      "epoch": 2.2985352243664265,
+      "grad_norm": 2.1567312680551227,
+      "learning_rate": 1.5797532433645696e-06,
+      "loss": 0.2259,
+      "step": 1646
+    },
+    {
+      "epoch": 2.299930248779354,
+      "grad_norm": 2.078334907511813,
+      "learning_rate": 1.5738302154345475e-06,
+      "loss": 0.2349,
+      "step": 1647
+    },
+    {
+      "epoch": 2.3013252731922806,
+      "grad_norm": 2.079069425158578,
+      "learning_rate": 1.5679162374262414e-06,
+      "loss": 0.2349,
+      "step": 1648
+    },
+    {
+      "epoch": 2.302720297605208,
+      "grad_norm": 1.9252113061545926,
+      "learning_rate": 1.5620113249608943e-06,
+      "loss": 0.253,
+      "step": 1649
+    },
+    {
+      "epoch": 2.304115322018135,
+      "grad_norm": 2.0259959768639586,
+      "learning_rate": 1.5561154936358069e-06,
+      "loss": 0.2402,
+      "step": 1650
+    },
+    {
+      "epoch": 2.3055103464310625,
+      "grad_norm": 2.090859034549075,
+      "learning_rate": 1.5502287590242942e-06,
+      "loss": 0.2286,
+      "step": 1651
+    },
+    {
+      "epoch": 2.3069053708439897,
+      "grad_norm": 1.8312991160092411,
+      "learning_rate": 1.5443511366756375e-06,
+      "loss": 0.2153,
+      "step": 1652
+    },
+    {
+      "epoch": 2.308300395256917,
+      "grad_norm": 1.954872792366119,
+      "learning_rate": 1.53848264211505e-06,
+      "loss": 0.2437,
+      "step": 1653
+    },
+    {
+      "epoch": 2.3096954196698443,
+      "grad_norm": 1.7300626387812998,
+      "learning_rate": 1.5326232908436405e-06,
+      "loss": 0.2146,
+      "step": 1654
+    },
+    {
+      "epoch": 2.3110904440827715,
+      "grad_norm": 2.0097750787041124,
+      "learning_rate": 1.526773098338359e-06,
+      "loss": 0.2328,
+      "step": 1655
+    },
+    {
+      "epoch": 2.312485468495699,
+      "grad_norm": 1.9802225368429347,
+      "learning_rate": 1.5209320800519683e-06,
+      "loss": 0.2315,
+      "step": 1656
+    },
+    {
+      "epoch": 2.313880492908626,
+      "grad_norm": 2.081213044620488,
+      "learning_rate": 1.515100251412998e-06,
+      "loss": 0.2503,
+      "step": 1657
+    },
+    {
+      "epoch": 2.3152755173215533,
+      "grad_norm": 1.9855688760114059,
+      "learning_rate": 1.5092776278257027e-06,
+      "loss": 0.227,
+      "step": 1658
+    },
+    {
+      "epoch": 2.31667054173448,
+      "grad_norm": 1.924537538881194,
+      "learning_rate": 1.5034642246700203e-06,
+      "loss": 0.2401,
+      "step": 1659
+    },
+    {
+      "epoch": 2.3180655661474074,
+      "grad_norm": 2.1845890347247714,
+      "learning_rate": 1.4976600573015398e-06,
+      "loss": 0.2516,
+      "step": 1660
+    },
+    {
+      "epoch": 2.3194605905603347,
+      "grad_norm": 2.0681273481820384,
+      "learning_rate": 1.4918651410514479e-06,
+      "loss": 0.2487,
+      "step": 1661
+    },
+    {
+      "epoch": 2.320855614973262,
+      "grad_norm": 1.9894399584697364,
+      "learning_rate": 1.486079491226501e-06,
+      "loss": 0.2185,
+      "step": 1662
+    },
+    {
+      "epoch": 2.3222506393861893,
+      "grad_norm": 1.9836278430699281,
+      "learning_rate": 1.4803031231089782e-06,
+      "loss": 0.2273,
+      "step": 1663
+    },
+    {
+      "epoch": 2.3236456637991165,
+      "grad_norm": 1.856207145115784,
+      "learning_rate": 1.4745360519566382e-06,
+      "loss": 0.2257,
+      "step": 1664
+    },
+    {
+      "epoch": 2.325040688212044,
+      "grad_norm": 2.059552287079175,
+      "learning_rate": 1.4687782930026833e-06,
+      "loss": 0.2519,
+      "step": 1665
+    },
+    {
+      "epoch": 2.326435712624971,
+      "grad_norm": 1.9484607160245673,
+      "learning_rate": 1.4630298614557236e-06,
+      "loss": 0.2305,
+      "step": 1666
+    },
+    {
+      "epoch": 2.3278307370378983,
+      "grad_norm": 2.057365261689767,
+      "learning_rate": 1.4572907724997249e-06,
+      "loss": 0.2738,
+      "step": 1667
+    },
+    {
+      "epoch": 2.329225761450825,
+      "grad_norm": 1.9604906487353115,
+      "learning_rate": 1.4515610412939791e-06,
+      "loss": 0.222,
+      "step": 1668
+    },
+    {
+      "epoch": 2.3306207858637524,
+      "grad_norm": 1.8794624116326488,
+      "learning_rate": 1.445840682973062e-06,
+      "loss": 0.2429,
+      "step": 1669
+    },
+    {
+      "epoch": 2.3320158102766797,
+      "grad_norm": 2.003776327480136,
+      "learning_rate": 1.4401297126467884e-06,
+      "loss": 0.2113,
+      "step": 1670
+    },
+    {
+      "epoch": 2.333410834689607,
+      "grad_norm": 1.9226038150982783,
+      "learning_rate": 1.4344281454001751e-06,
+      "loss": 0.2057,
+      "step": 1671
+    },
+    {
+      "epoch": 2.3348058591025342,
+      "grad_norm": 2.2602193419865517,
+      "learning_rate": 1.4287359962934055e-06,
+      "loss": 0.2485,
+      "step": 1672
+    },
+    {
+      "epoch": 2.3362008835154615,
+      "grad_norm": 2.1461782426022595,
+      "learning_rate": 1.4230532803617814e-06,
+      "loss": 0.2534,
+      "step": 1673
+    },
+    {
+      "epoch": 2.337595907928389,
+      "grad_norm": 1.9316351213030138,
+      "learning_rate": 1.4173800126156916e-06,
+      "loss": 0.2418,
+      "step": 1674
+    },
+    {
+      "epoch": 2.338990932341316,
+      "grad_norm": 1.856527405239218,
+      "learning_rate": 1.411716208040566e-06,
+      "loss": 0.2237,
+      "step": 1675
+    },
+    {
+      "epoch": 2.3403859567542433,
+      "grad_norm": 2.073164314302376,
+      "learning_rate": 1.4060618815968375e-06,
+      "loss": 0.2487,
+      "step": 1676
+    },
+    {
+      "epoch": 2.3417809811671706,
+      "grad_norm": 2.025415555206487,
+      "learning_rate": 1.4004170482199054e-06,
+      "loss": 0.264,
+      "step": 1677
+    },
+    {
+      "epoch": 2.343176005580098,
+      "grad_norm": 1.8947228597498715,
+      "learning_rate": 1.3947817228200956e-06,
+      "loss": 0.209,
+      "step": 1678
+    },
+    {
+      "epoch": 2.3445710299930247,
+      "grad_norm": 2.006321036445722,
+      "learning_rate": 1.3891559202826133e-06,
+      "loss": 0.2622,
+      "step": 1679
+    },
+    {
+      "epoch": 2.345966054405952,
+      "grad_norm": 2.303900612419585,
+      "learning_rate": 1.3835396554675179e-06,
+      "loss": 0.2378,
+      "step": 1680
+    },
+    {
+      "epoch": 2.3473610788188792,
+      "grad_norm": 2.037346779671685,
+      "learning_rate": 1.37793294320967e-06,
+      "loss": 0.2608,
+      "step": 1681
+    },
+    {
+      "epoch": 2.3487561032318065,
+      "grad_norm": 1.9305248706544658,
+      "learning_rate": 1.3723357983186974e-06,
+      "loss": 0.2163,
+      "step": 1682
+    },
+    {
+      "epoch": 2.3501511276447338,
+      "grad_norm": 2.0140564946990276,
+      "learning_rate": 1.3667482355789607e-06,
+      "loss": 0.2248,
+      "step": 1683
+    },
+    {
+      "epoch": 2.351546152057661,
+      "grad_norm": 2.084958562843635,
+      "learning_rate": 1.3611702697495088e-06,
+      "loss": 0.2131,
+      "step": 1684
+    },
+    {
+      "epoch": 2.3529411764705883,
+      "grad_norm": 1.9332055631395932,
+      "learning_rate": 1.3556019155640416e-06,
+      "loss": 0.2329,
+      "step": 1685
+    },
+    {
+      "epoch": 2.3543362008835156,
+      "grad_norm": 1.8488896616988466,
+      "learning_rate": 1.350043187730868e-06,
+      "loss": 0.2395,
+      "step": 1686
+    },
+    {
+      "epoch": 2.355731225296443,
+      "grad_norm": 2.430410848237495,
+      "learning_rate": 1.34449410093287e-06,
+      "loss": 0.2343,
+      "step": 1687
+    },
+    {
+      "epoch": 2.3571262497093697,
+      "grad_norm": 1.8914353443219187,
+      "learning_rate": 1.3389546698274686e-06,
+      "loss": 0.2171,
+      "step": 1688
+    },
+    {
+      "epoch": 2.358521274122297,
+      "grad_norm": 2.1017255630100493,
+      "learning_rate": 1.333424909046574e-06,
+      "loss": 0.2495,
+      "step": 1689
+    },
+    {
+      "epoch": 2.359916298535224,
+      "grad_norm": 2.161429978690411,
+      "learning_rate": 1.327904833196556e-06,
+      "loss": 0.2586,
+      "step": 1690
+    },
+    {
+      "epoch": 2.3613113229481515,
+      "grad_norm": 1.9388415195895992,
+      "learning_rate": 1.3223944568582047e-06,
+      "loss": 0.215,
+      "step": 1691
+    },
+    {
+      "epoch": 2.3627063473610788,
+      "grad_norm": 1.9652199591809878,
+      "learning_rate": 1.3168937945866861e-06,
+      "loss": 0.2276,
+      "step": 1692
+    },
+    {
+      "epoch": 2.364101371774006,
+      "grad_norm": 2.0711223470034597,
+      "learning_rate": 1.311402860911507e-06,
+      "loss": 0.241,
+      "step": 1693
+    },
+    {
+      "epoch": 2.3654963961869333,
+      "grad_norm": 1.9743629822432749,
+      "learning_rate": 1.3059216703364814e-06,
+      "loss": 0.2167,
+      "step": 1694
+    },
+    {
+      "epoch": 2.3668914205998606,
+      "grad_norm": 1.9924830951212522,
+      "learning_rate": 1.3004502373396821e-06,
+      "loss": 0.2358,
+      "step": 1695
+    },
+    {
+      "epoch": 2.368286445012788,
+      "grad_norm": 1.9788689015808711,
+      "learning_rate": 1.2949885763734127e-06,
+      "loss": 0.2423,
+      "step": 1696
+    },
+    {
+      "epoch": 2.369681469425715,
+      "grad_norm": 1.965926856035357,
+      "learning_rate": 1.2895367018641658e-06,
+      "loss": 0.2244,
+      "step": 1697
+    },
+    {
+      "epoch": 2.3710764938386424,
+      "grad_norm": 1.932908429022025,
+      "learning_rate": 1.284094628212576e-06,
+      "loss": 0.2582,
+      "step": 1698
+    },
+    {
+      "epoch": 2.372471518251569,
+      "grad_norm": 1.9431876485395636,
+      "learning_rate": 1.278662369793398e-06,
+      "loss": 0.2202,
+      "step": 1699
+    },
+    {
+      "epoch": 2.3738665426644965,
+      "grad_norm": 2.078921253978772,
+      "learning_rate": 1.273239940955459e-06,
+      "loss": 0.2572,
+      "step": 1700
+    },
+    {
+      "epoch": 2.3752615670774238,
+      "grad_norm": 2.035585260712323,
+      "learning_rate": 1.267827356021618e-06,
+      "loss": 0.2617,
+      "step": 1701
+    },
+    {
+      "epoch": 2.376656591490351,
+      "grad_norm": 2.180569630743452,
+      "learning_rate": 1.2624246292887377e-06,
+      "loss": 0.2349,
+      "step": 1702
+    },
+    {
+      "epoch": 2.3780516159032783,
+      "grad_norm": 2.03793194210423,
+      "learning_rate": 1.2570317750276374e-06,
+      "loss": 0.258,
+      "step": 1703
+    },
+    {
+      "epoch": 2.3794466403162056,
+      "grad_norm": 2.082005999471457,
+      "learning_rate": 1.2516488074830586e-06,
+      "loss": 0.2477,
+      "step": 1704
+    },
+    {
+      "epoch": 2.380841664729133,
+      "grad_norm": 1.8400987966532008,
+      "learning_rate": 1.246275740873631e-06,
+      "loss": 0.2331,
+      "step": 1705
+    },
+    {
+      "epoch": 2.38223668914206,
+      "grad_norm": 2.2066646811850865,
+      "learning_rate": 1.2409125893918329e-06,
+      "loss": 0.2542,
+      "step": 1706
+    },
+    {
+      "epoch": 2.3836317135549874,
+      "grad_norm": 2.0611768857820496,
+      "learning_rate": 1.2355593672039462e-06,
+      "loss": 0.202,
+      "step": 1707
+    },
+    {
+      "epoch": 2.385026737967914,
+      "grad_norm": 1.914378896690255,
+      "learning_rate": 1.2302160884500337e-06,
+      "loss": 0.2609,
+      "step": 1708
+    },
+    {
+      "epoch": 2.3864217623808415,
+      "grad_norm": 2.054680958740492,
+      "learning_rate": 1.2248827672438868e-06,
+      "loss": 0.2035,
+      "step": 1709
+    },
+    {
+      "epoch": 2.3878167867937687,
+      "grad_norm": 2.0441738070012105,
+      "learning_rate": 1.2195594176729963e-06,
+      "loss": 0.2388,
+      "step": 1710
+    },
+    {
+      "epoch": 2.389211811206696,
+      "grad_norm": 2.1122807295618844,
+      "learning_rate": 1.2142460537985168e-06,
+      "loss": 0.2303,
+      "step": 1711
+    },
+    {
+      "epoch": 2.3906068356196233,
+      "grad_norm": 2.0401675272163056,
+      "learning_rate": 1.2089426896552265e-06,
+      "loss": 0.2269,
+      "step": 1712
+    },
+    {
+      "epoch": 2.3920018600325506,
+      "grad_norm": 2.029589904828905,
+      "learning_rate": 1.2036493392514847e-06,
+      "loss": 0.2637,
+      "step": 1713
+    },
+    {
+      "epoch": 2.393396884445478,
+      "grad_norm": 2.0658070957162638,
+      "learning_rate": 1.1983660165692078e-06,
+      "loss": 0.2096,
+      "step": 1714
+    },
+    {
+      "epoch": 2.394791908858405,
+      "grad_norm": 2.283039546249197,
+      "learning_rate": 1.1930927355638189e-06,
+      "loss": 0.2436,
+      "step": 1715
+    },
+    {
+      "epoch": 2.3961869332713324,
+      "grad_norm": 1.8422258840639667,
+      "learning_rate": 1.1878295101642185e-06,
+      "loss": 0.2319,
+      "step": 1716
+    },
+    {
+      "epoch": 2.3975819576842596,
+      "grad_norm": 2.0568494378113296,
+      "learning_rate": 1.182576354272748e-06,
+      "loss": 0.2276,
+      "step": 1717
+    },
+    {
+      "epoch": 2.398976982097187,
+      "grad_norm": 1.952539086426312,
+      "learning_rate": 1.1773332817651512e-06,
+      "loss": 0.2295,
+      "step": 1718
+    },
+    {
+      "epoch": 2.400372006510114,
+      "grad_norm": 1.9158228630392138,
+      "learning_rate": 1.1721003064905329e-06,
+      "loss": 0.2133,
+      "step": 1719
+    },
+    {
+      "epoch": 2.401767030923041,
+      "grad_norm": 1.8532051493631778,
+      "learning_rate": 1.1668774422713336e-06,
+      "loss": 0.2396,
+      "step": 1720
+    },
+    {
+      "epoch": 2.4031620553359683,
+      "grad_norm": 1.9414496664869467,
+      "learning_rate": 1.1616647029032818e-06,
+      "loss": 0.2383,
+      "step": 1721
+    },
+    {
+      "epoch": 2.4045570797488955,
+      "grad_norm": 2.056538780483839,
+      "learning_rate": 1.1564621021553617e-06,
+      "loss": 0.2294,
+      "step": 1722
+    },
+    {
+      "epoch": 2.405952104161823,
+      "grad_norm": 2.0066219434319716,
+      "learning_rate": 1.1512696537697804e-06,
+      "loss": 0.2628,
+      "step": 1723
+    },
+    {
+      "epoch": 2.40734712857475,
+      "grad_norm": 2.285413378087384,
+      "learning_rate": 1.1460873714619275e-06,
+      "loss": 0.2684,
+      "step": 1724
+    },
+    {
+      "epoch": 2.4087421529876774,
+      "grad_norm": 2.074862240076793,
+      "learning_rate": 1.140915268920339e-06,
+      "loss": 0.2423,
+      "step": 1725
+    },
+    {
+      "epoch": 2.4101371774006046,
+      "grad_norm": 1.9534626474259522,
+      "learning_rate": 1.13575335980666e-06,
+      "loss": 0.2245,
+      "step": 1726
+    },
+    {
+      "epoch": 2.411532201813532,
+      "grad_norm": 2.11850049560302,
+      "learning_rate": 1.130601657755616e-06,
+      "loss": 0.2221,
+      "step": 1727
+    },
+    {
+      "epoch": 2.4129272262264587,
+      "grad_norm": 2.0705807422285614,
+      "learning_rate": 1.125460176374965e-06,
+      "loss": 0.227,
+      "step": 1728
+    },
+    {
+      "epoch": 2.414322250639386,
+      "grad_norm": 1.843249546710679,
+      "learning_rate": 1.1203289292454728e-06,
+      "loss": 0.2523,
+      "step": 1729
+    },
+    {
+      "epoch": 2.4157172750523133,
+      "grad_norm": 2.136376082447783,
+      "learning_rate": 1.1152079299208724e-06,
+      "loss": 0.2486,
+      "step": 1730
+    },
+    {
+      "epoch": 2.4171122994652405,
+      "grad_norm": 2.1189074920710484,
+      "learning_rate": 1.1100971919278247e-06,
+      "loss": 0.2286,
+      "step": 1731
+    },
+    {
+      "epoch": 2.418507323878168,
+      "grad_norm": 2.014466795940267,
+      "learning_rate": 1.104996728765887e-06,
+      "loss": 0.2471,
+      "step": 1732
+    },
+    {
+      "epoch": 2.419902348291095,
+      "grad_norm": 2.123142992349427,
+      "learning_rate": 1.0999065539074793e-06,
+      "loss": 0.2395,
+      "step": 1733
+    },
+    {
+      "epoch": 2.4212973727040223,
+      "grad_norm": 2.092780801188859,
+      "learning_rate": 1.094826680797843e-06,
+      "loss": 0.2323,
+      "step": 1734
+    },
+    {
+      "epoch": 2.4226923971169496,
+      "grad_norm": 1.9743466639593539,
+      "learning_rate": 1.0897571228550097e-06,
+      "loss": 0.2472,
+      "step": 1735
+    },
+    {
+      "epoch": 2.424087421529877,
+      "grad_norm": 2.0603635239914055,
+      "learning_rate": 1.0846978934697666e-06,
+      "loss": 0.2422,
+      "step": 1736
+    },
+    {
+      "epoch": 2.425482445942804,
+      "grad_norm": 1.878454958518566,
+      "learning_rate": 1.0796490060056142e-06,
+      "loss": 0.2313,
+      "step": 1737
+    },
+    {
+      "epoch": 2.4268774703557314,
+      "grad_norm": 2.0047687132435628,
+      "learning_rate": 1.074610473798738e-06,
+      "loss": 0.2287,
+      "step": 1738
+    },
+    {
+      "epoch": 2.4282724947686587,
+      "grad_norm": 2.130778745894436,
+      "learning_rate": 1.0695823101579728e-06,
+      "loss": 0.2227,
+      "step": 1739
+    },
+    {
+      "epoch": 2.4296675191815855,
+      "grad_norm": 1.9959834221817423,
+      "learning_rate": 1.0645645283647616e-06,
+      "loss": 0.2395,
+      "step": 1740
+    },
+    {
+      "epoch": 2.431062543594513,
+      "grad_norm": 2.076441367338032,
+      "learning_rate": 1.0595571416731293e-06,
+      "loss": 0.228,
+      "step": 1741
+    },
+    {
+      "epoch": 2.43245756800744,
+      "grad_norm": 1.9727319013347187,
+      "learning_rate": 1.0545601633096414e-06,
+      "loss": 0.2213,
+      "step": 1742
+    },
+    {
+      "epoch": 2.4338525924203673,
+      "grad_norm": 2.0619132387341015,
+      "learning_rate": 1.049573606473369e-06,
+      "loss": 0.245,
+      "step": 1743
+    },
+    {
+      "epoch": 2.4352476168332946,
+      "grad_norm": 1.9808499296055042,
+      "learning_rate": 1.0445974843358563e-06,
+      "loss": 0.2079,
+      "step": 1744
+    },
+    {
+      "epoch": 2.436642641246222,
+      "grad_norm": 1.9924612913344897,
+      "learning_rate": 1.0396318100410868e-06,
+      "loss": 0.2241,
+      "step": 1745
+    },
+    {
+      "epoch": 2.438037665659149,
+      "grad_norm": 2.0195101754838714,
+      "learning_rate": 1.0346765967054472e-06,
+      "loss": 0.2574,
+      "step": 1746
+    },
+    {
+      "epoch": 2.4394326900720764,
+      "grad_norm": 2.210865355019898,
+      "learning_rate": 1.029731857417689e-06,
+      "loss": 0.2589,
+      "step": 1747
+    },
+    {
+      "epoch": 2.4408277144850032,
+      "grad_norm": 1.9888864102360722,
+      "learning_rate": 1.0247976052389018e-06,
+      "loss": 0.2377,
+      "step": 1748
+    },
+    {
+      "epoch": 2.4422227388979305,
+      "grad_norm": 1.9963526780225826,
+      "learning_rate": 1.0198738532024715e-06,
+      "loss": 0.2201,
+      "step": 1749
+    },
+    {
+      "epoch": 2.443617763310858,
+      "grad_norm": 1.9375954149029564,
+      "learning_rate": 1.0149606143140484e-06,
+      "loss": 0.2464,
+      "step": 1750
+    },
+    {
+      "epoch": 2.445012787723785,
+      "grad_norm": 1.9490012959029157,
+      "learning_rate": 1.0100579015515156e-06,
+      "loss": 0.2387,
+      "step": 1751
+    },
+    {
+      "epoch": 2.4464078121367123,
+      "grad_norm": 2.1421767553643263,
+      "learning_rate": 1.005165727864953e-06,
+      "loss": 0.2382,
+      "step": 1752
+    },
+    {
+      "epoch": 2.4478028365496396,
+      "grad_norm": 2.0852319018269254,
+      "learning_rate": 1.0002841061765989e-06,
+      "loss": 0.2303,
+      "step": 1753
+    },
+    {
+      "epoch": 2.449197860962567,
+      "grad_norm": 1.898437822666265,
+      "learning_rate": 9.954130493808201e-07,
+      "loss": 0.2343,
+      "step": 1754
+    },
+    {
+      "epoch": 2.450592885375494,
+      "grad_norm": 2.2328708419723142,
+      "learning_rate": 9.905525703440815e-07,
+      "loss": 0.2324,
+      "step": 1755
+    },
+    {
+      "epoch": 2.4519879097884214,
+      "grad_norm": 2.098948338494277,
+      "learning_rate": 9.85702681904902e-07,
+      "loss": 0.2128,
+      "step": 1756
+    },
+    {
+      "epoch": 2.4533829342013487,
+      "grad_norm": 2.0504505351704916,
+      "learning_rate": 9.808633968738297e-07,
+      "loss": 0.2389,
+      "step": 1757
+    },
+    {
+      "epoch": 2.454777958614276,
+      "grad_norm": 2.006806960911157,
+      "learning_rate": 9.760347280334064e-07,
+      "loss": 0.2539,
+      "step": 1758
+    },
+    {
+      "epoch": 2.456172983027203,
+      "grad_norm": 1.9356692454237607,
+      "learning_rate": 9.712166881381279e-07,
+      "loss": 0.2416,
+      "step": 1759
+    },
+    {
+      "epoch": 2.45756800744013,
+      "grad_norm": 2.0609982252557013,
+      "learning_rate": 9.664092899144156e-07,
+      "loss": 0.2459,
+      "step": 1760
+    },
+    {
+      "epoch": 2.4589630318530573,
+      "grad_norm": 1.8784646017465558,
+      "learning_rate": 9.616125460605857e-07,
+      "loss": 0.219,
+      "step": 1761
+    },
+    {
+      "epoch": 2.4603580562659846,
+      "grad_norm": 1.9827145437937714,
+      "learning_rate": 9.56826469246806e-07,
+      "loss": 0.2164,
+      "step": 1762
+    },
+    {
+      "epoch": 2.461753080678912,
+      "grad_norm": 1.9828280925435495,
+      "learning_rate": 9.520510721150722e-07,
+      "loss": 0.2205,
+      "step": 1763
+    },
+    {
+      "epoch": 2.463148105091839,
+      "grad_norm": 2.1573204674909725,
+      "learning_rate": 9.472863672791721e-07,
+      "loss": 0.2797,
+      "step": 1764
+    },
+    {
+      "epoch": 2.4645431295047664,
+      "grad_norm": 2.114781710817254,
+      "learning_rate": 9.425323673246461e-07,
+      "loss": 0.2311,
+      "step": 1765
+    },
+    {
+      "epoch": 2.4659381539176937,
+      "grad_norm": 2.073720068988035,
+      "learning_rate": 9.377890848087595e-07,
+      "loss": 0.2422,
+      "step": 1766
+    },
+    {
+      "epoch": 2.467333178330621,
+      "grad_norm": 2.1924520868805963,
+      "learning_rate": 9.330565322604729e-07,
+      "loss": 0.2557,
+      "step": 1767
+    },
+    {
+      "epoch": 2.468728202743548,
+      "grad_norm": 2.1314923038601443,
+      "learning_rate": 9.283347221803985e-07,
+      "loss": 0.2365,
+      "step": 1768
+    },
+    {
+      "epoch": 2.470123227156475,
+      "grad_norm": 2.1016166593533656,
+      "learning_rate": 9.236236670407772e-07,
+      "loss": 0.2611,
+      "step": 1769
+    },
+    {
+      "epoch": 2.4715182515694023,
+      "grad_norm": 2.1958806745804416,
+      "learning_rate": 9.189233792854424e-07,
+      "loss": 0.2471,
+      "step": 1770
+    },
+    {
+      "epoch": 2.4729132759823296,
+      "grad_norm": 2.0613588674481877,
+      "learning_rate": 9.142338713297838e-07,
+      "loss": 0.2462,
+      "step": 1771
+    },
+    {
+      "epoch": 2.474308300395257,
+      "grad_norm": 1.889535779389397,
+      "learning_rate": 9.095551555607169e-07,
+      "loss": 0.2399,
+      "step": 1772
+    },
+    {
+      "epoch": 2.475703324808184,
+      "grad_norm": 2.076673087167382,
+      "learning_rate": 9.048872443366529e-07,
+      "loss": 0.2505,
+      "step": 1773
+    },
+    {
+      "epoch": 2.4770983492211114,
+      "grad_norm": 1.980949495347227,
+      "learning_rate": 9.002301499874622e-07,
+      "loss": 0.2365,
+      "step": 1774
+    },
+    {
+      "epoch": 2.4784933736340387,
+      "grad_norm": 1.972433910616226,
+      "learning_rate": 8.955838848144449e-07,
+      "loss": 0.2229,
+      "step": 1775
+    },
+    {
+      "epoch": 2.479888398046966,
+      "grad_norm": 2.205742031503069,
+      "learning_rate": 8.909484610902958e-07,
+      "loss": 0.2521,
+      "step": 1776
+    },
+    {
+      "epoch": 2.481283422459893,
+      "grad_norm": 1.9908248881502195,
+      "learning_rate": 8.863238910590704e-07,
+      "loss": 0.2096,
+      "step": 1777
+    },
+    {
+      "epoch": 2.4826784468728205,
+      "grad_norm": 1.9781717394676646,
+      "learning_rate": 8.817101869361599e-07,
+      "loss": 0.2482,
+      "step": 1778
+    },
+    {
+      "epoch": 2.4840734712857477,
+      "grad_norm": 2.1777301806732723,
+      "learning_rate": 8.77107360908253e-07,
+      "loss": 0.2652,
+      "step": 1779
+    },
+    {
+      "epoch": 2.4854684956986746,
+      "grad_norm": 2.068171178796176,
+      "learning_rate": 8.725154251333012e-07,
+      "loss": 0.2339,
+      "step": 1780
+    },
+    {
+      "epoch": 2.486863520111602,
+      "grad_norm": 2.2553151451018096,
+      "learning_rate": 8.679343917404959e-07,
+      "loss": 0.2962,
+      "step": 1781
+    },
+    {
+      "epoch": 2.488258544524529,
+      "grad_norm": 1.859927320195907,
+      "learning_rate": 8.633642728302266e-07,
+      "loss": 0.2409,
+      "step": 1782
+    },
+    {
+      "epoch": 2.4896535689374564,
+      "grad_norm": 2.097136463059436,
+      "learning_rate": 8.588050804740527e-07,
+      "loss": 0.226,
+      "step": 1783
+    },
+    {
+      "epoch": 2.4910485933503836,
+      "grad_norm": 2.031361226173633,
+      "learning_rate": 8.542568267146761e-07,
+      "loss": 0.2163,
+      "step": 1784
+    },
+    {
+      "epoch": 2.492443617763311,
+      "grad_norm": 1.8371067572490631,
+      "learning_rate": 8.49719523565904e-07,
+      "loss": 0.2184,
+      "step": 1785
+    },
+    {
+      "epoch": 2.493838642176238,
+      "grad_norm": 2.1523570658199564,
+      "learning_rate": 8.451931830126148e-07,
+      "loss": 0.2385,
+      "step": 1786
+    },
+    {
+      "epoch": 2.4952336665891655,
+      "grad_norm": 1.951878992014691,
+      "learning_rate": 8.40677817010736e-07,
+      "loss": 0.2222,
+      "step": 1787
+    },
+    {
+      "epoch": 2.4966286910020927,
+      "grad_norm": 2.1398256136886715,
+      "learning_rate": 8.361734374872032e-07,
+      "loss": 0.2424,
+      "step": 1788
+    },
+    {
+      "epoch": 2.4980237154150196,
+      "grad_norm": 1.8387718429463848,
+      "learning_rate": 8.316800563399307e-07,
+      "loss": 0.2252,
+      "step": 1789
+    },
+    {
+      "epoch": 2.499418739827947,
+      "grad_norm": 2.0758591904095205,
+      "learning_rate": 8.271976854377861e-07,
+      "loss": 0.2155,
+      "step": 1790
+    },
+    {
+      "epoch": 2.500813764240874,
+      "grad_norm": 1.9301849276761391,
+      "learning_rate": 8.227263366205523e-07,
+      "loss": 0.264,
+      "step": 1791
+    },
+    {
+      "epoch": 2.5022087886538014,
+      "grad_norm": 2.1047743582073157,
+      "learning_rate": 8.182660216988964e-07,
+      "loss": 0.3259,
+      "step": 1792
+    },
+    {
+      "epoch": 2.5036038130667286,
+      "grad_norm": 2.4983792769873623,
+      "learning_rate": 8.138167524543445e-07,
+      "loss": 0.2514,
+      "step": 1793
+    },
+    {
+      "epoch": 2.504998837479656,
+      "grad_norm": 2.2120478894270224,
+      "learning_rate": 8.09378540639243e-07,
+      "loss": 0.2515,
+      "step": 1794
+    },
+    {
+      "epoch": 2.506393861892583,
+      "grad_norm": 2.0867139416681195,
+      "learning_rate": 8.049513979767304e-07,
+      "loss": 0.2479,
+      "step": 1795
+    },
+    {
+      "epoch": 2.5077888863055104,
+      "grad_norm": 1.9359798627052334,
+      "learning_rate": 8.00535336160711e-07,
+      "loss": 0.2003,
+      "step": 1796
+    },
+    {
+      "epoch": 2.5091839107184377,
+      "grad_norm": 2.085079199029155,
+      "learning_rate": 7.96130366855819e-07,
+      "loss": 0.2429,
+      "step": 1797
+    },
+    {
+      "epoch": 2.510578935131365,
+      "grad_norm": 2.175659897075208,
+      "learning_rate": 7.917365016973866e-07,
+      "loss": 0.2206,
+      "step": 1798
+    },
+    {
+      "epoch": 2.5119739595442923,
+      "grad_norm": 2.1209361970147884,
+      "learning_rate": 7.873537522914155e-07,
+      "loss": 0.2506,
+      "step": 1799
+    },
+    {
+      "epoch": 2.5133689839572195,
+      "grad_norm": 2.1173385860772993,
+      "learning_rate": 7.829821302145485e-07,
+      "loss": 0.2345,
+      "step": 1800
+    },
+    {
+      "epoch": 2.5147640083701464,
+      "grad_norm": 1.9789184114073841,
+      "learning_rate": 7.786216470140334e-07,
+      "loss": 0.2365,
+      "step": 1801
+    },
+    {
+      "epoch": 2.5161590327830736,
+      "grad_norm": 2.0832279626065007,
+      "learning_rate": 7.742723142076991e-07,
+      "loss": 0.233,
+      "step": 1802
+    },
+    {
+      "epoch": 2.517554057196001,
+      "grad_norm": 1.9853952576197482,
+      "learning_rate": 7.699341432839203e-07,
+      "loss": 0.2465,
+      "step": 1803
+    },
+    {
+      "epoch": 2.518949081608928,
+      "grad_norm": 1.9655292208069006,
+      "learning_rate": 7.656071457015879e-07,
+      "loss": 0.2094,
+      "step": 1804
+    },
+    {
+      "epoch": 2.5203441060218554,
+      "grad_norm": 2.072957528236977,
+      "learning_rate": 7.612913328900784e-07,
+      "loss": 0.2308,
+      "step": 1805
+    },
+    {
+      "epoch": 2.5217391304347827,
+      "grad_norm": 2.050593842636747,
+      "learning_rate": 7.569867162492283e-07,
+      "loss": 0.2009,
+      "step": 1806
+    },
+    {
+      "epoch": 2.52313415484771,
+      "grad_norm": 1.9505121319227763,
+      "learning_rate": 7.526933071492959e-07,
+      "loss": 0.2628,
+      "step": 1807
+    },
+    {
+      "epoch": 2.524529179260637,
+      "grad_norm": 2.233499694853933,
+      "learning_rate": 7.484111169309399e-07,
+      "loss": 0.2267,
+      "step": 1808
+    },
+    {
+      "epoch": 2.525924203673564,
+      "grad_norm": 2.0427774995742456,
+      "learning_rate": 7.441401569051848e-07,
+      "loss": 0.2307,
+      "step": 1809
+    },
+    {
+      "epoch": 2.5273192280864913,
+      "grad_norm": 2.1523563504281604,
+      "learning_rate": 7.398804383533886e-07,
+      "loss": 0.2172,
+      "step": 1810
+    },
+    {
+      "epoch": 2.5287142524994186,
+      "grad_norm": 2.107184691359107,
+      "learning_rate": 7.356319725272165e-07,
+      "loss": 0.2354,
+      "step": 1811
+    },
+    {
+      "epoch": 2.530109276912346,
+      "grad_norm": 1.9472623249394085,
+      "learning_rate": 7.313947706486136e-07,
+      "loss": 0.2439,
+      "step": 1812
+    },
+    {
+      "epoch": 2.531504301325273,
+      "grad_norm": 2.2547362177468027,
+      "learning_rate": 7.271688439097713e-07,
+      "loss": 0.2611,
+      "step": 1813
+    },
+    {
+      "epoch": 2.5328993257382004,
+      "grad_norm": 1.9790179073787875,
+      "learning_rate": 7.229542034730952e-07,
+      "loss": 0.1908,
+      "step": 1814
+    },
+    {
+      "epoch": 2.5342943501511277,
+      "grad_norm": 2.1718993566008304,
+      "learning_rate": 7.187508604711851e-07,
+      "loss": 0.239,
+      "step": 1815
+    },
+    {
+      "epoch": 2.535689374564055,
+      "grad_norm": 2.006119781462586,
+      "learning_rate": 7.145588260067943e-07,
+      "loss": 0.2547,
+      "step": 1816
+    },
+    {
+      "epoch": 2.5370843989769822,
+      "grad_norm": 1.9707372022899972,
+      "learning_rate": 7.103781111528074e-07,
+      "loss": 0.2386,
+      "step": 1817
+    },
+    {
+      "epoch": 2.5384794233899095,
+      "grad_norm": 2.1078105247147616,
+      "learning_rate": 7.062087269522105e-07,
+      "loss": 0.2083,
+      "step": 1818
+    },
+    {
+      "epoch": 2.5398744478028368,
+      "grad_norm": 1.8096422690505354,
+      "learning_rate": 7.020506844180608e-07,
+      "loss": 0.24,
+      "step": 1819
+    },
+    {
+      "epoch": 2.541269472215764,
+      "grad_norm": 2.2907856322308047,
+      "learning_rate": 6.979039945334543e-07,
+      "loss": 0.2422,
+      "step": 1820
+    },
+    {
+      "epoch": 2.542664496628691,
+      "grad_norm": 2.0400580939554462,
+      "learning_rate": 6.937686682515044e-07,
+      "loss": 0.2353,
+      "step": 1821
+    },
+    {
+      "epoch": 2.544059521041618,
+      "grad_norm": 2.20245634299469,
+      "learning_rate": 6.896447164953057e-07,
+      "loss": 0.2535,
+      "step": 1822
+    },
+    {
+      "epoch": 2.5454545454545454,
+      "grad_norm": 2.048512795877345,
+      "learning_rate": 6.855321501579077e-07,
+      "loss": 0.2345,
+      "step": 1823
+    },
+    {
+      "epoch": 2.5468495698674727,
+      "grad_norm": 2.107797822320567,
+      "learning_rate": 6.814309801022873e-07,
+      "loss": 0.24,
+      "step": 1824
+    },
+    {
+      "epoch": 2.5482445942804,
+      "grad_norm": 2.0167190988164805,
+      "learning_rate": 6.77341217161322e-07,
+      "loss": 0.2644,
+      "step": 1825
+    },
+    {
+      "epoch": 2.5496396186933272,
+      "grad_norm": 2.0434168772720875,
+      "learning_rate": 6.732628721377533e-07,
+      "loss": 0.2187,
+      "step": 1826
+    },
+    {
+      "epoch": 2.5510346431062545,
+      "grad_norm": 2.0283320717025077,
+      "learning_rate": 6.69195955804165e-07,
+      "loss": 0.2894,
+      "step": 1827
+    },
+    {
+      "epoch": 2.5524296675191813,
+      "grad_norm": 2.184425853980423,
+      "learning_rate": 6.651404789029553e-07,
+      "loss": 0.2351,
+      "step": 1828
+    },
+    {
+      "epoch": 2.5538246919321086,
+      "grad_norm": 1.769961391111206,
+      "learning_rate": 6.610964521463032e-07,
+      "loss": 0.2367,
+      "step": 1829
+    },
+    {
+      "epoch": 2.555219716345036,
+      "grad_norm": 2.057938269523395,
+      "learning_rate": 6.570638862161449e-07,
+      "loss": 0.2136,
+      "step": 1830
+    },
+    {
+      "epoch": 2.556614740757963,
+      "grad_norm": 1.9601470016366807,
+      "learning_rate": 6.530427917641447e-07,
+      "loss": 0.2161,
+      "step": 1831
+    },
+    {
+      "epoch": 2.5580097651708904,
+      "grad_norm": 2.0539405710497296,
+      "learning_rate": 6.490331794116633e-07,
+      "loss": 0.2535,
+      "step": 1832
+    },
+    {
+      "epoch": 2.5594047895838177,
+      "grad_norm": 1.7665824386687003,
+      "learning_rate": 6.450350597497335e-07,
+      "loss": 0.2094,
+      "step": 1833
+    },
+    {
+      "epoch": 2.560799813996745,
+      "grad_norm": 1.9943622708189017,
+      "learning_rate": 6.410484433390335e-07,
+      "loss": 0.2517,
+      "step": 1834
+    },
+    {
+      "epoch": 2.562194838409672,
+      "grad_norm": 1.9325462598814633,
+      "learning_rate": 6.370733407098517e-07,
+      "loss": 0.2051,
+      "step": 1835
+    },
+    {
+      "epoch": 2.5635898628225995,
+      "grad_norm": 2.034392642432849,
+      "learning_rate": 6.331097623620697e-07,
+      "loss": 0.2411,
+      "step": 1836
+    },
+    {
+      "epoch": 2.5649848872355268,
+      "grad_norm": 1.906608056494423,
+      "learning_rate": 6.291577187651255e-07,
+      "loss": 0.2229,
+      "step": 1837
+    },
+    {
+      "epoch": 2.566379911648454,
+      "grad_norm": 2.0666019144763244,
+      "learning_rate": 6.252172203579892e-07,
+      "loss": 0.2405,
+      "step": 1838
+    },
+    {
+      "epoch": 2.5677749360613813,
+      "grad_norm": 2.0323492259191775,
+      "learning_rate": 6.212882775491352e-07,
+      "loss": 0.2727,
+      "step": 1839
+    },
+    {
+      "epoch": 2.5691699604743086,
+      "grad_norm": 1.9711644816953129,
+      "learning_rate": 6.173709007165158e-07,
+      "loss": 0.2451,
+      "step": 1840
+    },
+    {
+      "epoch": 2.5705649848872354,
+      "grad_norm": 2.1149991971809627,
+      "learning_rate": 6.134651002075315e-07,
+      "loss": 0.2332,
+      "step": 1841
+    },
+    {
+      "epoch": 2.5719600093001627,
+      "grad_norm": 2.1440606333696532,
+      "learning_rate": 6.095708863390065e-07,
+      "loss": 0.2443,
+      "step": 1842
+    },
+    {
+      "epoch": 2.57335503371309,
+      "grad_norm": 1.898581256781591,
+      "learning_rate": 6.056882693971605e-07,
+      "loss": 0.2183,
+      "step": 1843
+    },
+    {
+      "epoch": 2.574750058126017,
+      "grad_norm": 2.0720450905291448,
+      "learning_rate": 6.018172596375776e-07,
+      "loss": 0.2642,
+      "step": 1844
+    },
+    {
+      "epoch": 2.5761450825389445,
+      "grad_norm": 1.9560353823505137,
+      "learning_rate": 5.979578672851843e-07,
+      "loss": 0.2464,
+      "step": 1845
+    },
+    {
+      "epoch": 2.5775401069518717,
+      "grad_norm": 2.0786260231432077,
+      "learning_rate": 5.941101025342239e-07,
+      "loss": 0.222,
+      "step": 1846
+    },
+    {
+      "epoch": 2.578935131364799,
+      "grad_norm": 2.2179763750347634,
+      "learning_rate": 5.902739755482201e-07,
+      "loss": 0.2893,
+      "step": 1847
+    },
+    {
+      "epoch": 2.580330155777726,
+      "grad_norm": 1.98667473908548,
+      "learning_rate": 5.864494964599615e-07,
+      "loss": 0.2155,
+      "step": 1848
+    },
+    {
+      "epoch": 2.581725180190653,
+      "grad_norm": 2.0169000385789437,
+      "learning_rate": 5.826366753714707e-07,
+      "loss": 0.2356,
+      "step": 1849
+    },
+    {
+      "epoch": 2.5831202046035804,
+      "grad_norm": 1.8149669738422527,
+      "learning_rate": 5.788355223539698e-07,
+      "loss": 0.2252,
+      "step": 1850
+    },
+    {
+      "epoch": 2.5845152290165077,
+      "grad_norm": 2.1280566049946517,
+      "learning_rate": 5.750460474478675e-07,
+      "loss": 0.221,
+      "step": 1851
+    },
+    {
+      "epoch": 2.585910253429435,
+      "grad_norm": 1.7836139149162966,
+      "learning_rate": 5.712682606627251e-07,
+      "loss": 0.2277,
+      "step": 1852
+    },
+    {
+      "epoch": 2.587305277842362,
+      "grad_norm": 2.1508036868898928,
+      "learning_rate": 5.675021719772262e-07,
+      "loss": 0.2428,
+      "step": 1853
+    },
+    {
+      "epoch": 2.5887003022552895,
+      "grad_norm": 2.012263006386255,
+      "learning_rate": 5.637477913391604e-07,
+      "loss": 0.2221,
+      "step": 1854
+    },
+    {
+      "epoch": 2.5900953266682167,
+      "grad_norm": 1.9813993494667828,
+      "learning_rate": 5.600051286653884e-07,
+      "loss": 0.258,
+      "step": 1855
+    },
+    {
+      "epoch": 2.591490351081144,
+      "grad_norm": 2.1602425598412798,
+      "learning_rate": 5.562741938418187e-07,
+      "loss": 0.2299,
+      "step": 1856
+    },
+    {
+      "epoch": 2.5928853754940713,
+      "grad_norm": 2.0750208195688846,
+      "learning_rate": 5.525549967233829e-07,
+      "loss": 0.2349,
+      "step": 1857
+    },
+    {
+      "epoch": 2.5942803999069985,
+      "grad_norm": 2.0263311031326383,
+      "learning_rate": 5.488475471340099e-07,
+      "loss": 0.2325,
+      "step": 1858
+    },
+    {
+      "epoch": 2.595675424319926,
+      "grad_norm": 2.078040762901248,
+      "learning_rate": 5.451518548665946e-07,
+      "loss": 0.2428,
+      "step": 1859
+    },
+    {
+      "epoch": 2.597070448732853,
+      "grad_norm": 1.8412450414831687,
+      "learning_rate": 5.414679296829806e-07,
+      "loss": 0.2332,
+      "step": 1860
+    },
+    {
+      "epoch": 2.59846547314578,
+      "grad_norm": 1.8369491681995171,
+      "learning_rate": 5.377957813139262e-07,
+      "loss": 0.199,
+      "step": 1861
+    },
+    {
+      "epoch": 2.599860497558707,
+      "grad_norm": 2.0738049963713356,
+      "learning_rate": 5.341354194590831e-07,
+      "loss": 0.3003,
+      "step": 1862
+    },
+    {
+      "epoch": 2.6012555219716345,
+      "grad_norm": 2.2015502071142685,
+      "learning_rate": 5.304868537869706e-07,
+      "loss": 0.2218,
+      "step": 1863
+    },
+    {
+      "epoch": 2.6026505463845617,
+      "grad_norm": 1.9042994634815842,
+      "learning_rate": 5.268500939349514e-07,
+      "loss": 0.228,
+      "step": 1864
+    },
+    {
+      "epoch": 2.604045570797489,
+      "grad_norm": 1.938132636974177,
+      "learning_rate": 5.232251495091989e-07,
+      "loss": 0.2143,
+      "step": 1865
+    },
+    {
+      "epoch": 2.6054405952104163,
+      "grad_norm": 1.9382048893357533,
+      "learning_rate": 5.196120300846835e-07,
+      "loss": 0.2221,
+      "step": 1866
+    },
+    {
+      "epoch": 2.6068356196233435,
+      "grad_norm": 1.9324006507947422,
+      "learning_rate": 5.160107452051361e-07,
+      "loss": 0.2491,
+      "step": 1867
+    },
+    {
+      "epoch": 2.6082306440362704,
+      "grad_norm": 2.0427761815474397,
+      "learning_rate": 5.124213043830278e-07,
+      "loss": 0.2021,
+      "step": 1868
+    },
+    {
+      "epoch": 2.6096256684491976,
+      "grad_norm": 2.0082833497287225,
+      "learning_rate": 5.088437170995481e-07,
+      "loss": 0.2141,
+      "step": 1869
+    },
+    {
+      "epoch": 2.611020692862125,
+      "grad_norm": 1.866320494961171,
+      "learning_rate": 5.052779928045737e-07,
+      "loss": 0.2257,
+      "step": 1870
+    },
+    {
+      "epoch": 2.612415717275052,
+      "grad_norm": 2.0142624931091833,
+      "learning_rate": 5.01724140916649e-07,
+      "loss": 0.2341,
+      "step": 1871
+    },
+    {
+      "epoch": 2.6138107416879794,
+      "grad_norm": 2.0232328561691997,
+      "learning_rate": 4.981821708229545e-07,
+      "loss": 0.2512,
+      "step": 1872
+    },
+    {
+      "epoch": 2.6152057661009067,
+      "grad_norm": 2.1058063752183744,
+      "learning_rate": 4.946520918792886e-07,
+      "loss": 0.2182,
+      "step": 1873
+    },
+    {
+      "epoch": 2.616600790513834,
+      "grad_norm": 2.1036496155790463,
+      "learning_rate": 4.911339134100401e-07,
+      "loss": 0.2533,
+      "step": 1874
+    },
+    {
+      "epoch": 2.6179958149267613,
+      "grad_norm": 2.1879285031785325,
+      "learning_rate": 4.87627644708163e-07,
+      "loss": 0.239,
+      "step": 1875
+    },
+    {
+      "epoch": 2.6193908393396885,
+      "grad_norm": 2.006704466585366,
+      "learning_rate": 4.841332950351535e-07,
+      "loss": 0.2441,
+      "step": 1876
+    },
+    {
+      "epoch": 2.620785863752616,
+      "grad_norm": 2.007471298063676,
+      "learning_rate": 4.806508736210253e-07,
+      "loss": 0.2365,
+      "step": 1877
+    },
+    {
+      "epoch": 2.622180888165543,
+      "grad_norm": 1.789418537428787,
+      "learning_rate": 4.771803896642812e-07,
+      "loss": 0.2165,
+      "step": 1878
+    },
+    {
+      "epoch": 2.6235759125784703,
+      "grad_norm": 1.9228615618742761,
+      "learning_rate": 4.737218523318965e-07,
+      "loss": 0.2542,
+      "step": 1879
+    },
+    {
+      "epoch": 2.6249709369913976,
+      "grad_norm": 1.970234498952072,
+      "learning_rate": 4.7027527075929e-07,
+      "loss": 0.2212,
+      "step": 1880
+    },
+    {
+      "epoch": 2.6263659614043244,
+      "grad_norm": 2.071241627988333,
+      "learning_rate": 4.6684065405029677e-07,
+      "loss": 0.2363,
+      "step": 1881
+    },
+    {
+      "epoch": 2.6277609858172517,
+      "grad_norm": 1.9258080559377377,
+      "learning_rate": 4.6341801127715303e-07,
+      "loss": 0.2257,
+      "step": 1882
+    },
+    {
+      "epoch": 2.629156010230179,
+      "grad_norm": 1.9922325956159916,
+      "learning_rate": 4.6000735148046316e-07,
+      "loss": 0.2529,
+      "step": 1883
+    },
+    {
+      "epoch": 2.6305510346431062,
+      "grad_norm": 2.0813559532867,
+      "learning_rate": 4.566086836691791e-07,
+      "loss": 0.2275,
+      "step": 1884
+    },
+    {
+      "epoch": 2.6319460590560335,
+      "grad_norm": 1.9067412469761382,
+      "learning_rate": 4.532220168205798e-07,
+      "loss": 0.24,
+      "step": 1885
+    },
+    {
+      "epoch": 2.633341083468961,
+      "grad_norm": 2.0801424081538906,
+      "learning_rate": 4.498473598802444e-07,
+      "loss": 0.2085,
+      "step": 1886
+    },
+    {
+      "epoch": 2.634736107881888,
+      "grad_norm": 2.0163287177348277,
+      "learning_rate": 4.464847217620266e-07,
+      "loss": 0.2322,
+      "step": 1887
+    },
+    {
+      "epoch": 2.636131132294815,
+      "grad_norm": 2.0311176630581267,
+      "learning_rate": 4.4313411134803584e-07,
+      "loss": 0.2464,
+      "step": 1888
+    },
+    {
+      "epoch": 2.637526156707742,
+      "grad_norm": 2.0628314868379554,
+      "learning_rate": 4.397955374886104e-07,
+      "loss": 0.2192,
+      "step": 1889
+    },
+    {
+      "epoch": 2.6389211811206694,
+      "grad_norm": 2.1050032920387385,
+      "learning_rate": 4.364690090022938e-07,
+      "loss": 0.2369,
+      "step": 1890
+    },
+    {
+      "epoch": 2.6403162055335967,
+      "grad_norm": 1.9286471955876885,
+      "learning_rate": 4.331545346758159e-07,
+      "loss": 0.2322,
+      "step": 1891
+    },
+    {
+      "epoch": 2.641711229946524,
+      "grad_norm": 2.0048011858542334,
+      "learning_rate": 4.2985212326406456e-07,
+      "loss": 0.2121,
+      "step": 1892
+    },
+    {
+      "epoch": 2.6431062543594512,
+      "grad_norm": 1.9806271015465544,
+      "learning_rate": 4.265617834900637e-07,
+      "loss": 0.2134,
+      "step": 1893
+    },
+    {
+      "epoch": 2.6445012787723785,
+      "grad_norm": 1.9317814362926222,
+      "learning_rate": 4.2328352404495346e-07,
+      "loss": 0.2035,
+      "step": 1894
+    },
+    {
+      "epoch": 2.6458963031853058,
+      "grad_norm": 2.0239142062051663,
+      "learning_rate": 4.2001735358796316e-07,
+      "loss": 0.2007,
+      "step": 1895
+    },
+    {
+      "epoch": 2.647291327598233,
+      "grad_norm": 1.7416542864642734,
+      "learning_rate": 4.167632807463895e-07,
+      "loss": 0.2022,
+      "step": 1896
+    },
+    {
+      "epoch": 2.6486863520111603,
+      "grad_norm": 1.9831774018825685,
+      "learning_rate": 4.135213141155769e-07,
+      "loss": 0.2296,
+      "step": 1897
+    },
+    {
+      "epoch": 2.6500813764240876,
+      "grad_norm": 2.116068772685149,
+      "learning_rate": 4.1029146225889103e-07,
+      "loss": 0.2356,
+      "step": 1898
+    },
+    {
+      "epoch": 2.651476400837015,
+      "grad_norm": 1.9642948045482904,
+      "learning_rate": 4.0707373370769634e-07,
+      "loss": 0.2321,
+      "step": 1899
+    },
+    {
+      "epoch": 2.652871425249942,
+      "grad_norm": 2.0222515744267042,
+      "learning_rate": 4.0386813696133564e-07,
+      "loss": 0.2363,
+      "step": 1900
+    },
+    {
+      "epoch": 2.654266449662869,
+      "grad_norm": 1.9783354095753165,
+      "learning_rate": 4.0067468048710756e-07,
+      "loss": 0.2194,
+      "step": 1901
+    },
+    {
+      "epoch": 2.6556614740757962,
+      "grad_norm": 1.9211770592630109,
+      "learning_rate": 3.974933727202412e-07,
+      "loss": 0.229,
+      "step": 1902
+    },
+    {
+      "epoch": 2.6570564984887235,
+      "grad_norm": 1.8639767908818732,
+      "learning_rate": 3.943242220638777e-07,
+      "loss": 0.2152,
+      "step": 1903
+    },
+    {
+      "epoch": 2.6584515229016508,
+      "grad_norm": 2.0239531405139997,
+      "learning_rate": 3.911672368890462e-07,
+      "loss": 0.2426,
+      "step": 1904
+    },
+    {
+      "epoch": 2.659846547314578,
+      "grad_norm": 2.1025627641545572,
+      "learning_rate": 3.8802242553464096e-07,
+      "loss": 0.2313,
+      "step": 1905
+    },
+    {
+      "epoch": 2.6612415717275053,
+      "grad_norm": 2.0146577001145123,
+      "learning_rate": 3.8488979630739996e-07,
+      "loss": 0.2298,
+      "step": 1906
+    },
+    {
+      "epoch": 2.6626365961404326,
+      "grad_norm": 2.0364327114333003,
+      "learning_rate": 3.8176935748188425e-07,
+      "loss": 0.1858,
+      "step": 1907
+    },
+    {
+      "epoch": 2.66403162055336,
+      "grad_norm": 1.7558229059450832,
+      "learning_rate": 3.78661117300454e-07,
+      "loss": 0.207,
+      "step": 1908
+    },
+    {
+      "epoch": 2.6654266449662867,
+      "grad_norm": 1.9582911012714252,
+      "learning_rate": 3.755650839732489e-07,
+      "loss": 0.2123,
+      "step": 1909
+    },
+    {
+      "epoch": 2.666821669379214,
+      "grad_norm": 1.8840012884467325,
+      "learning_rate": 3.7248126567816454e-07,
+      "loss": 0.2562,
+      "step": 1910
+    },
+    {
+      "epoch": 2.668216693792141,
+      "grad_norm": 2.061501597369465,
+      "learning_rate": 3.694096705608319e-07,
+      "loss": 0.2368,
+      "step": 1911
+    },
+    {
+      "epoch": 2.6696117182050685,
+      "grad_norm": 2.1513760140631075,
+      "learning_rate": 3.6635030673459413e-07,
+      "loss": 0.2218,
+      "step": 1912
+    },
+    {
+      "epoch": 2.6710067426179958,
+      "grad_norm": 2.065018071288163,
+      "learning_rate": 3.6330318228049e-07,
+      "loss": 0.2468,
+      "step": 1913
+    },
+    {
+      "epoch": 2.672401767030923,
+      "grad_norm": 1.9139784162080573,
+      "learning_rate": 3.6026830524722443e-07,
+      "loss": 0.2407,
+      "step": 1914
+    },
+    {
+      "epoch": 2.6737967914438503,
+      "grad_norm": 1.9445091905357286,
+      "learning_rate": 3.572456836511551e-07,
+      "loss": 0.2458,
+      "step": 1915
+    },
+    {
+      "epoch": 2.6751918158567776,
+      "grad_norm": 1.94322248049399,
+      "learning_rate": 3.5423532547626816e-07,
+      "loss": 0.2207,
+      "step": 1916
+    },
+    {
+      "epoch": 2.676586840269705,
+      "grad_norm": 1.9210010764056549,
+      "learning_rate": 3.5123723867415527e-07,
+      "loss": 0.1981,
+      "step": 1917
+    },
+    {
+      "epoch": 2.677981864682632,
+      "grad_norm": 1.9185403011532343,
+      "learning_rate": 3.4825143116399454e-07,
+      "loss": 0.2194,
+      "step": 1918
+    },
+    {
+      "epoch": 2.6793768890955594,
+      "grad_norm": 2.256457880823731,
+      "learning_rate": 3.452779108325316e-07,
+      "loss": 0.2209,
+      "step": 1919
+    },
+    {
+      "epoch": 2.6807719135084866,
+      "grad_norm": 1.9724953467337907,
+      "learning_rate": 3.4231668553405316e-07,
+      "loss": 0.222,
+      "step": 1920
+    },
+    {
+      "epoch": 2.6821669379214135,
+      "grad_norm": 2.0819922224865044,
+      "learning_rate": 3.39367763090373e-07,
+      "loss": 0.237,
+      "step": 1921
+    },
+    {
+      "epoch": 2.6835619623343407,
+      "grad_norm": 2.0287374188952945,
+      "learning_rate": 3.3643115129080695e-07,
+      "loss": 0.238,
+      "step": 1922
+    },
+    {
+      "epoch": 2.684956986747268,
+      "grad_norm": 2.012873321631083,
+      "learning_rate": 3.3350685789215133e-07,
+      "loss": 0.2264,
+      "step": 1923
+    },
+    {
+      "epoch": 2.6863520111601953,
+      "grad_norm": 2.100733946305469,
+      "learning_rate": 3.3059489061866625e-07,
+      "loss": 0.2447,
+      "step": 1924
+    },
+    {
+      "epoch": 2.6877470355731226,
+      "grad_norm": 2.1754655463163273,
+      "learning_rate": 3.276952571620556e-07,
+      "loss": 0.2521,
+      "step": 1925
+    },
+    {
+      "epoch": 2.68914205998605,
+      "grad_norm": 2.030678162445352,
+      "learning_rate": 3.248079651814395e-07,
+      "loss": 0.2302,
+      "step": 1926
+    },
+    {
+      "epoch": 2.690537084398977,
+      "grad_norm": 1.827439538967311,
+      "learning_rate": 3.2193302230334455e-07,
+      "loss": 0.1955,
+      "step": 1927
+    },
+    {
+      "epoch": 2.6919321088119044,
+      "grad_norm": 1.826638985874322,
+      "learning_rate": 3.190704361216751e-07,
+      "loss": 0.2122,
+      "step": 1928
+    },
+    {
+      "epoch": 2.693327133224831,
+      "grad_norm": 2.0614544727358597,
+      "learning_rate": 3.162202141976956e-07,
+      "loss": 0.266,
+      "step": 1929
+    },
+    {
+      "epoch": 2.6947221576377585,
+      "grad_norm": 2.20309115743746,
+      "learning_rate": 3.133823640600137e-07,
+      "loss": 0.2578,
+      "step": 1930
+    },
+    {
+      "epoch": 2.6961171820506857,
+      "grad_norm": 1.9253571598225367,
+      "learning_rate": 3.105568932045577e-07,
+      "loss": 0.2494,
+      "step": 1931
+    },
+    {
+      "epoch": 2.697512206463613,
+      "grad_norm": 2.082806360246401,
+      "learning_rate": 3.077438090945573e-07,
+      "loss": 0.2233,
+      "step": 1932
+    },
+    {
+      "epoch": 2.6989072308765403,
+      "grad_norm": 2.1368559158587876,
+      "learning_rate": 3.0494311916052234e-07,
+      "loss": 0.2472,
+      "step": 1933
+    },
+    {
+      "epoch": 2.7003022552894675,
+      "grad_norm": 2.1399120491029393,
+      "learning_rate": 3.021548308002248e-07,
+      "loss": 0.2454,
+      "step": 1934
+    },
+    {
+      "epoch": 2.701697279702395,
+      "grad_norm": 2.0184949022912746,
+      "learning_rate": 2.9937895137868046e-07,
+      "loss": 0.2334,
+      "step": 1935
+    },
+    {
+      "epoch": 2.703092304115322,
+      "grad_norm": 2.12813933018869,
+      "learning_rate": 2.9661548822812636e-07,
+      "loss": 0.2145,
+      "step": 1936
+    },
+    {
+      "epoch": 2.7044873285282494,
+      "grad_norm": 1.8207230076654493,
+      "learning_rate": 2.9386444864800355e-07,
+      "loss": 0.2533,
+      "step": 1937
+    },
+    {
+      "epoch": 2.7058823529411766,
+      "grad_norm": 1.9816052407797478,
+      "learning_rate": 2.911258399049394e-07,
+      "loss": 0.2681,
+      "step": 1938
+    },
+    {
+      "epoch": 2.707277377354104,
+      "grad_norm": 2.1675186182836783,
+      "learning_rate": 2.8839966923272286e-07,
+      "loss": 0.2283,
+      "step": 1939
+    },
+    {
+      "epoch": 2.708672401767031,
+      "grad_norm": 2.026735541307535,
+      "learning_rate": 2.8568594383229067e-07,
+      "loss": 0.2516,
+      "step": 1940
+    },
+    {
+      "epoch": 2.710067426179958,
+      "grad_norm": 2.1369075066967844,
+      "learning_rate": 2.8298467087170655e-07,
+      "loss": 0.2544,
+      "step": 1941
+    },
+    {
+      "epoch": 2.7114624505928853,
+      "grad_norm": 2.0393735418050447,
+      "learning_rate": 2.8029585748614196e-07,
+      "loss": 0.2399,
+      "step": 1942
+    },
+    {
+      "epoch": 2.7128574750058125,
+      "grad_norm": 2.0244415010257995,
+      "learning_rate": 2.7761951077785676e-07,
+      "loss": 0.242,
+      "step": 1943
+    },
+    {
+      "epoch": 2.71425249941874,
+      "grad_norm": 2.00987027151913,
+      "learning_rate": 2.749556378161833e-07,
+      "loss": 0.2226,
+      "step": 1944
+    },
+    {
+      "epoch": 2.715647523831667,
+      "grad_norm": 2.0689668056181496,
+      "learning_rate": 2.723042456375036e-07,
+      "loss": 0.2224,
+      "step": 1945
+    },
+    {
+      "epoch": 2.7170425482445943,
+      "grad_norm": 1.9797277784560094,
+      "learning_rate": 2.696653412452327e-07,
+      "loss": 0.2256,
+      "step": 1946
+    },
+    {
+      "epoch": 2.7184375726575216,
+      "grad_norm": 1.93324183395182,
+      "learning_rate": 2.6703893160980266e-07,
+      "loss": 0.2318,
+      "step": 1947
+    },
+    {
+      "epoch": 2.719832597070449,
+      "grad_norm": 1.9660338970378355,
+      "learning_rate": 2.6442502366863854e-07,
+      "loss": 0.2466,
+      "step": 1948
+    },
+    {
+      "epoch": 2.7212276214833757,
+      "grad_norm": 2.041748793231265,
+      "learning_rate": 2.618236243261452e-07,
+      "loss": 0.2324,
+      "step": 1949
+    },
+    {
+      "epoch": 2.722622645896303,
+      "grad_norm": 2.0989128634489678,
+      "learning_rate": 2.592347404536888e-07,
+      "loss": 0.2329,
+      "step": 1950
+    },
+    {
+      "epoch": 2.7240176703092303,
+      "grad_norm": 2.1530826233490403,
+      "learning_rate": 2.566583788895721e-07,
+      "loss": 0.2358,
+      "step": 1951
+    },
+    {
+      "epoch": 2.7254126947221575,
+      "grad_norm": 1.9728780674087554,
+      "learning_rate": 2.5409454643902543e-07,
+      "loss": 0.2383,
+      "step": 1952
+    },
+    {
+      "epoch": 2.726807719135085,
+      "grad_norm": 2.0580855265191556,
+      "learning_rate": 2.5154324987418434e-07,
+      "loss": 0.2428,
+      "step": 1953
+    },
+    {
+      "epoch": 2.728202743548012,
+      "grad_norm": 2.010908338058876,
+      "learning_rate": 2.4900449593406984e-07,
+      "loss": 0.2235,
+      "step": 1954
+    },
+    {
+      "epoch": 2.7295977679609393,
+      "grad_norm": 1.9669057953101938,
+      "learning_rate": 2.4647829132457446e-07,
+      "loss": 0.2263,
+      "step": 1955
+    },
+    {
+      "epoch": 2.7309927923738666,
+      "grad_norm": 1.820511916019583,
+      "learning_rate": 2.439646427184428e-07,
+      "loss": 0.2199,
+      "step": 1956
+    },
+    {
+      "epoch": 2.732387816786794,
+      "grad_norm": 2.1325860915646766,
+      "learning_rate": 2.4146355675525145e-07,
+      "loss": 0.2324,
+      "step": 1957
+    },
+    {
+      "epoch": 2.733782841199721,
+      "grad_norm": 1.96368327913333,
+      "learning_rate": 2.389750400413965e-07,
+      "loss": 0.2493,
+      "step": 1958
+    },
+    {
+      "epoch": 2.7351778656126484,
+      "grad_norm": 1.9907140449319412,
+      "learning_rate": 2.364990991500743e-07,
+      "loss": 0.2276,
+      "step": 1959
+    },
+    {
+      "epoch": 2.7365728900255757,
+      "grad_norm": 2.199391943371498,
+      "learning_rate": 2.340357406212601e-07,
+      "loss": 0.2066,
+      "step": 1960
+    },
+    {
+      "epoch": 2.7379679144385025,
+      "grad_norm": 1.8746179327096653,
+      "learning_rate": 2.315849709616963e-07,
+      "loss": 0.1987,
+      "step": 1961
+    },
+    {
+      "epoch": 2.73936293885143,
+      "grad_norm": 1.8545723296564522,
+      "learning_rate": 2.2914679664487237e-07,
+      "loss": 0.1947,
+      "step": 1962
+    },
+    {
+      "epoch": 2.740757963264357,
+      "grad_norm": 2.0292994633820993,
+      "learning_rate": 2.2672122411100727e-07,
+      "loss": 0.2357,
+      "step": 1963
+    },
+    {
+      "epoch": 2.7421529876772843,
+      "grad_norm": 1.9882351397477116,
+      "learning_rate": 2.2430825976703485e-07,
+      "loss": 0.2226,
+      "step": 1964
+    },
+    {
+      "epoch": 2.7435480120902116,
+      "grad_norm": 2.0461627757851364,
+      "learning_rate": 2.2190790998658561e-07,
+      "loss": 0.2651,
+      "step": 1965
+    },
+    {
+      "epoch": 2.744943036503139,
+      "grad_norm": 2.1018494441005884,
+      "learning_rate": 2.1952018110996843e-07,
+      "loss": 0.2186,
+      "step": 1966
+    },
+    {
+      "epoch": 2.746338060916066,
+      "grad_norm": 2.1297808725292278,
+      "learning_rate": 2.1714507944415708e-07,
+      "loss": 0.2523,
+      "step": 1967
+    },
+    {
+      "epoch": 2.7477330853289934,
+      "grad_norm": 1.9107037042938857,
+      "learning_rate": 2.1478261126276989e-07,
+      "loss": 0.2196,
+      "step": 1968
+    },
+    {
+      "epoch": 2.7491281097419202,
+      "grad_norm": 2.288966469409701,
+      "learning_rate": 2.1243278280605517e-07,
+      "loss": 0.2388,
+      "step": 1969
+    },
+    {
+      "epoch": 2.7505231341548475,
+      "grad_norm": 2.0113282340570344,
+      "learning_rate": 2.1009560028087627e-07,
+      "loss": 0.2187,
+      "step": 1970
+    },
+    {
+      "epoch": 2.7519181585677748,
+      "grad_norm": 2.0577383008463404,
+      "learning_rate": 2.0777106986069162e-07,
+      "loss": 0.2341,
+      "step": 1971
+    },
+    {
+      "epoch": 2.753313182980702,
+      "grad_norm": 1.979246023089912,
+      "learning_rate": 2.0545919768554078e-07,
+      "loss": 0.2139,
+      "step": 1972
+    },
+    {
+      "epoch": 2.7547082073936293,
+      "grad_norm": 2.1145076730283408,
+      "learning_rate": 2.0315998986202902e-07,
+      "loss": 0.2503,
+      "step": 1973
+    },
+    {
+      "epoch": 2.7561032318065566,
+      "grad_norm": 2.038214733990826,
+      "learning_rate": 2.0087345246330714e-07,
+      "loss": 0.2273,
+      "step": 1974
+    },
+    {
+      "epoch": 2.757498256219484,
+      "grad_norm": 2.085222579466815,
+      "learning_rate": 1.985995915290595e-07,
+      "loss": 0.2365,
+      "step": 1975
+    },
+    {
+      "epoch": 2.758893280632411,
+      "grad_norm": 1.961930537056812,
+      "learning_rate": 1.9633841306548717e-07,
+      "loss": 0.2116,
+      "step": 1976
+    },
+    {
+      "epoch": 2.7602883050453384,
+      "grad_norm": 1.9746414569427433,
+      "learning_rate": 1.9408992304529252e-07,
+      "loss": 0.2496,
+      "step": 1977
+    },
+    {
+      "epoch": 2.7616833294582657,
+      "grad_norm": 1.9416879339980468,
+      "learning_rate": 1.9185412740765962e-07,
+      "loss": 0.1972,
+      "step": 1978
+    },
+    {
+      "epoch": 2.763078353871193,
+      "grad_norm": 1.9526605625268896,
+      "learning_rate": 1.8963103205824397e-07,
+      "loss": 0.223,
+      "step": 1979
+    },
+    {
+      "epoch": 2.76447337828412,
+      "grad_norm": 2.034376339460357,
+      "learning_rate": 1.8742064286915329e-07,
+      "loss": 0.2364,
+      "step": 1980
+    },
+    {
+      "epoch": 2.765868402697047,
+      "grad_norm": 1.928888877253871,
+      "learning_rate": 1.8522296567893282e-07,
+      "loss": 0.249,
+      "step": 1981
+    },
+    {
+      "epoch": 2.7672634271099743,
+      "grad_norm": 2.179741053262633,
+      "learning_rate": 1.830380062925513e-07,
+      "loss": 0.2226,
+      "step": 1982
+    },
+    {
+      "epoch": 2.7686584515229016,
+      "grad_norm": 2.1545101813981384,
+      "learning_rate": 1.8086577048138432e-07,
+      "loss": 0.2366,
+      "step": 1983
+    },
+    {
+      "epoch": 2.770053475935829,
+      "grad_norm": 1.8941469814993368,
+      "learning_rate": 1.787062639831988e-07,
+      "loss": 0.2195,
+      "step": 1984
+    },
+    {
+      "epoch": 2.771448500348756,
+      "grad_norm": 2.129029126050336,
+      "learning_rate": 1.7655949250213743e-07,
+      "loss": 0.2158,
+      "step": 1985
+    },
+    {
+      "epoch": 2.7728435247616834,
+      "grad_norm": 1.9373194190810024,
+      "learning_rate": 1.7442546170870654e-07,
+      "loss": 0.2227,
+      "step": 1986
+    },
+    {
+      "epoch": 2.7742385491746107,
+      "grad_norm": 2.022970081615447,
+      "learning_rate": 1.7230417723975766e-07,
+      "loss": 0.2392,
+      "step": 1987
+    },
+    {
+      "epoch": 2.775633573587538,
+      "grad_norm": 2.0136815131628034,
+      "learning_rate": 1.7019564469847372e-07,
+      "loss": 0.2143,
+      "step": 1988
+    },
+    {
+      "epoch": 2.7770285980004648,
+      "grad_norm": 1.8758190104617118,
+      "learning_rate": 1.6809986965435675e-07,
+      "loss": 0.2052,
+      "step": 1989
+    },
+    {
+      "epoch": 2.778423622413392,
+      "grad_norm": 2.1315337163239265,
+      "learning_rate": 1.660168576432092e-07,
+      "loss": 0.2435,
+      "step": 1990
+    },
+    {
+      "epoch": 2.7798186468263193,
+      "grad_norm": 2.126578038961216,
+      "learning_rate": 1.6394661416711977e-07,
+      "loss": 0.2229,
+      "step": 1991
+    },
+    {
+      "epoch": 2.7812136712392466,
+      "grad_norm": 1.9817111978138984,
+      "learning_rate": 1.6188914469445372e-07,
+      "loss": 0.2236,
+      "step": 1992
+    },
+    {
+      "epoch": 2.782608695652174,
+      "grad_norm": 2.020978049322112,
+      "learning_rate": 1.5984445465983156e-07,
+      "loss": 0.2337,
+      "step": 1993
+    },
+    {
+      "epoch": 2.784003720065101,
+      "grad_norm": 1.9219890439740848,
+      "learning_rate": 1.5781254946412029e-07,
+      "loss": 0.2174,
+      "step": 1994
+    },
+    {
+      "epoch": 2.7853987444780284,
+      "grad_norm": 1.9668911299656417,
+      "learning_rate": 1.5579343447441663e-07,
+      "loss": 0.216,
+      "step": 1995
+    },
+    {
+      "epoch": 2.7867937688909556,
+      "grad_norm": 1.8683512982901405,
+      "learning_rate": 1.5378711502403164e-07,
+      "loss": 0.192,
+      "step": 1996
+    },
+    {
+      "epoch": 2.788188793303883,
+      "grad_norm": 1.96884738892342,
+      "learning_rate": 1.5179359641247948e-07,
+      "loss": 0.2368,
+      "step": 1997
+    },
+    {
+      "epoch": 2.78958381771681,
+      "grad_norm": 2.0694150339442574,
+      "learning_rate": 1.4981288390546188e-07,
+      "loss": 0.2447,
+      "step": 1998
+    },
+    {
+      "epoch": 2.7909788421297375,
+      "grad_norm": 2.0875718503144425,
+      "learning_rate": 1.4784498273485436e-07,
+      "loss": 0.2278,
+      "step": 1999
+    },
+    {
+      "epoch": 2.7923738665426647,
+      "grad_norm": 2.080303164509914,
+      "learning_rate": 1.458898980986917e-07,
+      "loss": 0.2392,
+      "step": 2000
+    },
+    {
+      "epoch": 2.793768890955592,
+      "grad_norm": 2.1332960074779685,
+      "learning_rate": 1.4394763516115573e-07,
+      "loss": 0.2309,
+      "step": 2001
+    },
+    {
+      "epoch": 2.795163915368519,
+      "grad_norm": 2.1368019185944966,
+      "learning_rate": 1.4201819905256043e-07,
+      "loss": 0.2307,
+      "step": 2002
+    },
+    {
+      "epoch": 2.796558939781446,
+      "grad_norm": 2.1158557952062895,
+      "learning_rate": 1.4010159486933906e-07,
+      "loss": 0.2509,
+      "step": 2003
+    },
+    {
+      "epoch": 2.7979539641943734,
+      "grad_norm": 2.0558983826210717,
+      "learning_rate": 1.3819782767403034e-07,
+      "loss": 0.2195,
+      "step": 2004
+    },
+    {
+      "epoch": 2.7993489886073006,
+      "grad_norm": 1.9261173004873333,
+      "learning_rate": 1.363069024952668e-07,
+      "loss": 0.2365,
+      "step": 2005
+    },
+    {
+      "epoch": 2.800744013020228,
+      "grad_norm": 1.9189366322448516,
+      "learning_rate": 1.344288243277575e-07,
+      "loss": 0.2437,
+      "step": 2006
+    },
+    {
+      "epoch": 2.802139037433155,
+      "grad_norm": 2.010005253000098,
+      "learning_rate": 1.3256359813227758e-07,
+      "loss": 0.2238,
+      "step": 2007
+    },
+    {
+      "epoch": 2.8035340618460824,
+      "grad_norm": 1.8560267537251662,
+      "learning_rate": 1.3071122883565657e-07,
+      "loss": 0.2472,
+      "step": 2008
+    },
+    {
+      "epoch": 2.8049290862590093,
+      "grad_norm": 2.0899891570381963,
+      "learning_rate": 1.288717213307622e-07,
+      "loss": 0.2465,
+      "step": 2009
+    },
+    {
+      "epoch": 2.8063241106719365,
+      "grad_norm": 2.0680954666113034,
+      "learning_rate": 1.2704508047649e-07,
+      "loss": 0.2487,
+      "step": 2010
+    },
+    {
+      "epoch": 2.807719135084864,
+      "grad_norm": 2.003905225505703,
+      "learning_rate": 1.2523131109774822e-07,
+      "loss": 0.2023,
+      "step": 2011
+    },
+    {
+      "epoch": 2.809114159497791,
+      "grad_norm": 1.711566864626143,
+      "learning_rate": 1.234304179854473e-07,
+      "loss": 0.2171,
+      "step": 2012
+    },
+    {
+      "epoch": 2.8105091839107184,
+      "grad_norm": 2.1299908201771305,
+      "learning_rate": 1.2164240589648436e-07,
+      "loss": 0.2289,
+      "step": 2013
+    },
+    {
+      "epoch": 2.8119042083236456,
+      "grad_norm": 2.0155206237844103,
+      "learning_rate": 1.1986727955373588e-07,
+      "loss": 0.2267,
+      "step": 2014
+    },
+    {
+      "epoch": 2.813299232736573,
+      "grad_norm": 2.0292985224449844,
+      "learning_rate": 1.1810504364603737e-07,
+      "loss": 0.2363,
+      "step": 2015
+    },
+    {
+      "epoch": 2.8146942571495,
+      "grad_norm": 2.064222853753681,
+      "learning_rate": 1.163557028281792e-07,
+      "loss": 0.2097,
+      "step": 2016
+    },
+    {
+      "epoch": 2.8160892815624274,
+      "grad_norm": 2.0007115496451053,
+      "learning_rate": 1.146192617208891e-07,
+      "loss": 0.1964,
+      "step": 2017
+    },
+    {
+      "epoch": 2.8174843059753547,
+      "grad_norm": 1.8324329173352072,
+      "learning_rate": 1.128957249108209e-07,
+      "loss": 0.2091,
+      "step": 2018
+    },
+    {
+      "epoch": 2.818879330388282,
+      "grad_norm": 1.9879711284900963,
+      "learning_rate": 1.1118509695054236e-07,
+      "loss": 0.2097,
+      "step": 2019
+    },
+    {
+      "epoch": 2.8202743548012092,
+      "grad_norm": 1.9307293719213443,
+      "learning_rate": 1.094873823585263e-07,
+      "loss": 0.2057,
+      "step": 2020
+    },
+    {
+      "epoch": 2.8216693792141365,
+      "grad_norm": 1.8756053670010457,
+      "learning_rate": 1.0780258561913281e-07,
+      "loss": 0.2375,
+      "step": 2021
+    },
+    {
+      "epoch": 2.8230644036270633,
+      "grad_norm": 2.2006935107651433,
+      "learning_rate": 1.0613071118260321e-07,
+      "loss": 0.2365,
+      "step": 2022
+    },
+    {
+      "epoch": 2.8244594280399906,
+      "grad_norm": 2.2031758088252467,
+      "learning_rate": 1.0447176346504439e-07,
+      "loss": 0.2507,
+      "step": 2023
+    },
+    {
+      "epoch": 2.825854452452918,
+      "grad_norm": 2.0331823697534164,
+      "learning_rate": 1.0282574684841784e-07,
+      "loss": 0.2503,
+      "step": 2024
+    },
+    {
+      "epoch": 2.827249476865845,
+      "grad_norm": 2.069416903471262,
+      "learning_rate": 1.011926656805301e-07,
+      "loss": 0.2135,
+      "step": 2025
+    },
+    {
+      "epoch": 2.8286445012787724,
+      "grad_norm": 1.8308454776524232,
+      "learning_rate": 9.957252427501951e-08,
+      "loss": 0.2098,
+      "step": 2026
+    },
+    {
+      "epoch": 2.8300395256916997,
+      "grad_norm": 1.8951158726710682,
+      "learning_rate": 9.796532691134453e-08,
+      "loss": 0.2391,
+      "step": 2027
+    },
+    {
+      "epoch": 2.831434550104627,
+      "grad_norm": 1.9942746715013322,
+      "learning_rate": 9.637107783477484e-08,
+      "loss": 0.2211,
+      "step": 2028
+    },
+    {
+      "epoch": 2.832829574517554,
+      "grad_norm": 2.074390628687549,
+      "learning_rate": 9.478978125637583e-08,
+      "loss": 0.2056,
+      "step": 2029
+    },
+    {
+      "epoch": 2.834224598930481,
+      "grad_norm": 2.0197859477607314,
+      "learning_rate": 9.322144135300137e-08,
+      "loss": 0.2356,
+      "step": 2030
+    },
+    {
+      "epoch": 2.8356196233434083,
+      "grad_norm": 2.0090938538458474,
+      "learning_rate": 9.166606226728103e-08,
+      "loss": 0.2069,
+      "step": 2031
+    },
+    {
+      "epoch": 2.8370146477563356,
+      "grad_norm": 1.784604391867768,
+      "learning_rate": 9.012364810761121e-08,
+      "loss": 0.2401,
+      "step": 2032
+    },
+    {
+      "epoch": 2.838409672169263,
+      "grad_norm": 2.3621317749274486,
+      "learning_rate": 8.859420294814014e-08,
+      "loss": 0.235,
+      "step": 2033
+    },
+    {
+      "epoch": 2.83980469658219,
+      "grad_norm": 1.9238377220229415,
+      "learning_rate": 8.70777308287618e-08,
+      "loss": 0.2546,
+      "step": 2034
+    },
+    {
+      "epoch": 2.8411997209951174,
+      "grad_norm": 2.2386775669658077,
+      "learning_rate": 8.557423575510037e-08,
+      "loss": 0.2552,
+      "step": 2035
+    },
+    {
+      "epoch": 2.8425947454080447,
+      "grad_norm": 2.164637819234815,
+      "learning_rate": 8.408372169850521e-08,
+      "loss": 0.245,
+      "step": 2036
+    },
+    {
+      "epoch": 2.843989769820972,
+      "grad_norm": 2.1267060866199334,
+      "learning_rate": 8.26061925960353e-08,
+      "loss": 0.2627,
+      "step": 2037
+    },
+    {
+      "epoch": 2.8453847942338992,
+      "grad_norm": 2.175840094462751,
+      "learning_rate": 8.114165235045268e-08,
+      "loss": 0.2387,
+      "step": 2038
+    },
+    {
+      "epoch": 2.8467798186468265,
+      "grad_norm": 2.041608543433892,
+      "learning_rate": 7.969010483020845e-08,
+      "loss": 0.228,
+      "step": 2039
+    },
+    {
+      "epoch": 2.8481748430597538,
+      "grad_norm": 1.8317288740398268,
+      "learning_rate": 7.825155386943784e-08,
+      "loss": 0.2349,
+      "step": 2040
+    },
+    {
+      "epoch": 2.849569867472681,
+      "grad_norm": 2.076140243471252,
+      "learning_rate": 7.682600326794353e-08,
+      "loss": 0.2427,
+      "step": 2041
+    },
+    {
+      "epoch": 2.850964891885608,
+      "grad_norm": 1.930151853189304,
+      "learning_rate": 7.541345679118961e-08,
+      "loss": 0.2302,
+      "step": 2042
+    },
+    {
+      "epoch": 2.852359916298535,
+      "grad_norm": 2.198044174123011,
+      "learning_rate": 7.401391817029257e-08,
+      "loss": 0.227,
+      "step": 2043
+    },
+    {
+      "epoch": 2.8537549407114624,
+      "grad_norm": 2.025855994758739,
+      "learning_rate": 7.262739110200923e-08,
+      "loss": 0.2281,
+      "step": 2044
+    },
+    {
+      "epoch": 2.8551499651243897,
+      "grad_norm": 2.076594409784964,
+      "learning_rate": 7.125387924872552e-08,
+      "loss": 0.2552,
+      "step": 2045
+    },
+    {
+      "epoch": 2.856544989537317,
+      "grad_norm": 1.9583920141971498,
+      "learning_rate": 6.98933862384521e-08,
+      "loss": 0.2123,
+      "step": 2046
+    },
+    {
+      "epoch": 2.857940013950244,
+      "grad_norm": 2.1218586078077317,
+      "learning_rate": 6.854591566480884e-08,
+      "loss": 0.1864,
+      "step": 2047
+    },
+    {
+      "epoch": 2.8593350383631715,
+      "grad_norm": 1.9283990614176407,
+      "learning_rate": 6.721147108701864e-08,
+      "loss": 0.2352,
+      "step": 2048
+    },
+    {
+      "epoch": 2.8607300627760983,
+      "grad_norm": 2.0442219825500283,
+      "learning_rate": 6.589005602989862e-08,
+      "loss": 0.2299,
+      "step": 2049
+    },
+    {
+      "epoch": 2.8621250871890256,
+      "grad_norm": 2.296981149362681,
+      "learning_rate": 6.458167398384896e-08,
+      "loss": 0.2346,
+      "step": 2050
+    },
+    {
+      "epoch": 2.863520111601953,
+      "grad_norm": 2.0309143626789488,
+      "learning_rate": 6.328632840484294e-08,
+      "loss": 0.2003,
+      "step": 2051
+    },
+    {
+      "epoch": 2.86491513601488,
+      "grad_norm": 1.8135178412153783,
+      "learning_rate": 6.200402271442085e-08,
+      "loss": 0.2241,
+      "step": 2052
+    },
+    {
+      "epoch": 2.8663101604278074,
+      "grad_norm": 2.057910995863822,
+      "learning_rate": 6.073476029967884e-08,
+      "loss": 0.2304,
+      "step": 2053
+    },
+    {
+      "epoch": 2.8677051848407347,
+      "grad_norm": 2.013113048138251,
+      "learning_rate": 5.947854451326007e-08,
+      "loss": 0.2461,
+      "step": 2054
+    },
+    {
+      "epoch": 2.869100209253662,
+      "grad_norm": 2.0255907815051835,
+      "learning_rate": 5.823537867334694e-08,
+      "loss": 0.2136,
+      "step": 2055
+    },
+    {
+      "epoch": 2.870495233666589,
+      "grad_norm": 1.9199247970965672,
+      "learning_rate": 5.7005266063650534e-08,
+      "loss": 0.2243,
+      "step": 2056
+    },
+    {
+      "epoch": 2.8718902580795165,
+      "grad_norm": 2.156967447355663,
+      "learning_rate": 5.5788209933403944e-08,
+      "loss": 0.2805,
+      "step": 2057
+    },
+    {
+      "epoch": 2.8732852824924437,
+      "grad_norm": 1.9305083322744667,
+      "learning_rate": 5.4584213497351766e-08,
+      "loss": 0.2263,
+      "step": 2058
+    },
+    {
+      "epoch": 2.874680306905371,
+      "grad_norm": 1.976227997733238,
+      "learning_rate": 5.339327993574339e-08,
+      "loss": 0.2099,
+      "step": 2059
+    },
+    {
+      "epoch": 2.8760753313182983,
+      "grad_norm": 1.9714973369866737,
+      "learning_rate": 5.221541239432415e-08,
+      "loss": 0.1981,
+      "step": 2060
+    },
+    {
+      "epoch": 2.8774703557312256,
+      "grad_norm": 2.036301034568653,
+      "learning_rate": 5.1050613984324756e-08,
+      "loss": 0.1994,
+      "step": 2061
+    },
+    {
+      "epoch": 2.8788653801441524,
+      "grad_norm": 2.1703868645790667,
+      "learning_rate": 4.989888778245744e-08,
+      "loss": 0.2516,
+      "step": 2062
+    },
+    {
+      "epoch": 2.8802604045570797,
+      "grad_norm": 1.9606077688525048,
+      "learning_rate": 4.8760236830903697e-08,
+      "loss": 0.2377,
+      "step": 2063
+    },
+    {
+      "epoch": 2.881655428970007,
+      "grad_norm": 2.0531863218211885,
+      "learning_rate": 4.763466413730822e-08,
+      "loss": 0.2451,
+      "step": 2064
+    },
+    {
+      "epoch": 2.883050453382934,
+      "grad_norm": 2.0517963200516096,
+      "learning_rate": 4.65221726747711e-08,
+      "loss": 0.2393,
+      "step": 2065
+    },
+    {
+      "epoch": 2.8844454777958615,
+      "grad_norm": 2.157910418174818,
+      "learning_rate": 4.542276538183954e-08,
+      "loss": 0.2225,
+      "step": 2066
+    },
+    {
+      "epoch": 2.8858405022087887,
+      "grad_norm": 2.0408160528092667,
+      "learning_rate": 4.433644516249891e-08,
+      "loss": 0.246,
+      "step": 2067
+    },
+    {
+      "epoch": 2.887235526621716,
+      "grad_norm": 2.093617123831178,
+      "learning_rate": 4.326321488616836e-08,
+      "loss": 0.2508,
+      "step": 2068
+    },
+    {
+      "epoch": 2.888630551034643,
+      "grad_norm": 2.1884025061010144,
+      "learning_rate": 4.220307738768859e-08,
+      "loss": 0.2237,
+      "step": 2069
+    },
+    {
+      "epoch": 2.89002557544757,
+      "grad_norm": 1.9657784524140614,
+      "learning_rate": 4.11560354673185e-08,
+      "loss": 0.246,
+      "step": 2070
+    },
+    {
+      "epoch": 2.8914205998604974,
+      "grad_norm": 1.8686482185482467,
+      "learning_rate": 4.0122091890726354e-08,
+      "loss": 0.2291,
+      "step": 2071
+    },
+    {
+      "epoch": 2.8928156242734246,
+      "grad_norm": 1.9914728882718662,
+      "learning_rate": 3.9101249388981965e-08,
+      "loss": 0.2018,
+      "step": 2072
+    },
+    {
+      "epoch": 2.894210648686352,
+      "grad_norm": 1.9919379835112403,
+      "learning_rate": 3.809351065854894e-08,
+      "loss": 0.2586,
+      "step": 2073
+    },
+    {
+      "epoch": 2.895605673099279,
+      "grad_norm": 2.0949017501890173,
+      "learning_rate": 3.709887836128023e-08,
+      "loss": 0.2614,
+      "step": 2074
+    },
+    {
+      "epoch": 2.8970006975122065,
+      "grad_norm": 2.0514498538853267,
+      "learning_rate": 3.611735512440706e-08,
+      "loss": 0.2514,
+      "step": 2075
+    },
+    {
+      "epoch": 2.8983957219251337,
+      "grad_norm": 2.1514559104030906,
+      "learning_rate": 3.5148943540536105e-08,
+      "loss": 0.2265,
+      "step": 2076
+    },
+    {
+      "epoch": 2.899790746338061,
+      "grad_norm": 1.9415093712506253,
+      "learning_rate": 3.4193646167640646e-08,
+      "loss": 0.2521,
+      "step": 2077
+    },
+    {
+      "epoch": 2.9011857707509883,
+      "grad_norm": 2.2275104214143084,
+      "learning_rate": 3.325146552905223e-08,
+      "loss": 0.228,
+      "step": 2078
+    },
+    {
+      "epoch": 2.9025807951639155,
+      "grad_norm": 1.9328981753760448,
+      "learning_rate": 3.2322404113457886e-08,
+      "loss": 0.2535,
+      "step": 2079
+    },
+    {
+      "epoch": 2.903975819576843,
+      "grad_norm": 2.0722145311097027,
+      "learning_rate": 3.1406464374890144e-08,
+      "loss": 0.2293,
+      "step": 2080
+    },
+    {
+      "epoch": 2.90537084398977,
+      "grad_norm": 1.9769610016608057,
+      "learning_rate": 3.0503648732722046e-08,
+      "loss": 0.2477,
+      "step": 2081
+    },
+    {
+      "epoch": 2.906765868402697,
+      "grad_norm": 2.050363313420139,
+      "learning_rate": 2.9613959571660468e-08,
+      "loss": 0.2157,
+      "step": 2082
+    },
+    {
+      "epoch": 2.908160892815624,
+      "grad_norm": 1.9951482436366297,
+      "learning_rate": 2.8737399241740016e-08,
+      "loss": 0.2189,
+      "step": 2083
+    },
+    {
+      "epoch": 2.9095559172285514,
+      "grad_norm": 2.081109307380316,
+      "learning_rate": 2.7873970058316934e-08,
+      "loss": 0.2289,
+      "step": 2084
+    },
+    {
+      "epoch": 2.9109509416414787,
+      "grad_norm": 1.9864262205196008,
+      "learning_rate": 2.7023674302061875e-08,
+      "loss": 0.2243,
+      "step": 2085
+    },
+    {
+      "epoch": 2.912345966054406,
+      "grad_norm": 1.9215362456747311,
+      "learning_rate": 2.6186514218954905e-08,
+      "loss": 0.2063,
+      "step": 2086
+    },
+    {
+      "epoch": 2.9137409904673333,
+      "grad_norm": 2.0398028688026666,
+      "learning_rate": 2.5362492020280517e-08,
+      "loss": 0.2177,
+      "step": 2087
+    },
+    {
+      "epoch": 2.9151360148802605,
+      "grad_norm": 1.9687613018790115,
+      "learning_rate": 2.4551609882619288e-08,
+      "loss": 0.2492,
+      "step": 2088
+    },
+    {
+      "epoch": 2.9165310392931874,
+      "grad_norm": 2.2202058691346918,
+      "learning_rate": 2.3753869947843457e-08,
+      "loss": 0.2173,
+      "step": 2089
+    },
+    {
+      "epoch": 2.9179260637061146,
+      "grad_norm": 2.070982150627721,
+      "learning_rate": 2.296927432311358e-08,
+      "loss": 0.2468,
+      "step": 2090
+    },
+    {
+      "epoch": 2.919321088119042,
+      "grad_norm": 2.0497098444086728,
+      "learning_rate": 2.2197825080867432e-08,
+      "loss": 0.2247,
+      "step": 2091
+    },
+    {
+      "epoch": 2.920716112531969,
+      "grad_norm": 2.1761110942805066,
+      "learning_rate": 2.1439524258819456e-08,
+      "loss": 0.2491,
+      "step": 2092
+    },
+    {
+      "epoch": 2.9221111369448964,
+      "grad_norm": 1.8677894906476675,
+      "learning_rate": 2.0694373859954653e-08,
+      "loss": 0.2173,
+      "step": 2093
+    },
+    {
+      "epoch": 2.9235061613578237,
+      "grad_norm": 2.0166862523401874,
+      "learning_rate": 1.99623758525197e-08,
+      "loss": 0.2368,
+      "step": 2094
+    },
+    {
+      "epoch": 2.924901185770751,
+      "grad_norm": 1.9058734582520152,
+      "learning_rate": 1.9243532170023504e-08,
+      "loss": 0.221,
+      "step": 2095
+    },
+    {
+      "epoch": 2.9262962101836782,
+      "grad_norm": 2.008476408905422,
+      "learning_rate": 1.8537844711227215e-08,
+      "loss": 0.2049,
+      "step": 2096
+    },
+    {
+      "epoch": 2.9276912345966055,
+      "grad_norm": 1.9038713013473556,
+      "learning_rate": 1.7845315340140334e-08,
+      "loss": 0.2097,
+      "step": 2097
+    },
+    {
+      "epoch": 2.929086259009533,
+      "grad_norm": 1.7358824031830948,
+      "learning_rate": 1.7165945886018498e-08,
+      "loss": 0.1912,
+      "step": 2098
+    },
+    {
+      "epoch": 2.93048128342246,
+      "grad_norm": 2.002733171009674,
+      "learning_rate": 1.6499738143354594e-08,
+      "loss": 0.2142,
+      "step": 2099
+    },
+    {
+      "epoch": 2.9318763078353873,
+      "grad_norm": 2.0905328726237693,
+      "learning_rate": 1.584669387187765e-08,
+      "loss": 0.2267,
+      "step": 2100
+    },
+    {
+      "epoch": 2.9332713322483146,
+      "grad_norm": 1.9108917257839144,
+      "learning_rate": 1.520681479654562e-08,
+      "loss": 0.219,
+      "step": 2101
+    },
+    {
+      "epoch": 2.9346663566612414,
+      "grad_norm": 1.979846709755045,
+      "learning_rate": 1.4580102607541502e-08,
+      "loss": 0.2275,
+      "step": 2102
+    },
+    {
+      "epoch": 2.9360613810741687,
+      "grad_norm": 2.0046838206360227,
+      "learning_rate": 1.3966558960269994e-08,
+      "loss": 0.2082,
+      "step": 2103
+    },
+    {
+      "epoch": 2.937456405487096,
+      "grad_norm": 1.680760440099357,
+      "learning_rate": 1.3366185475351957e-08,
+      "loss": 0.2073,
+      "step": 2104
+    },
+    {
+      "epoch": 2.9388514299000232,
+      "grad_norm": 1.8009860986680282,
+      "learning_rate": 1.2778983738620521e-08,
+      "loss": 0.2326,
+      "step": 2105
+    },
+    {
+      "epoch": 2.9402464543129505,
+      "grad_norm": 2.0317108503401413,
+      "learning_rate": 1.2204955301116095e-08,
+      "loss": 0.2094,
+      "step": 2106
+    },
+    {
+      "epoch": 2.941641478725878,
+      "grad_norm": 1.8968224765396366,
+      "learning_rate": 1.164410167908414e-08,
+      "loss": 0.2212,
+      "step": 2107
+    },
+    {
+      "epoch": 2.943036503138805,
+      "grad_norm": 2.186384787215692,
+      "learning_rate": 1.109642435396907e-08,
+      "loss": 0.2125,
+      "step": 2108
+    },
+    {
+      "epoch": 2.9444315275517323,
+      "grad_norm": 1.9506984878795102,
+      "learning_rate": 1.0561924772412024e-08,
+      "loss": 0.2243,
+      "step": 2109
+    },
+    {
+      "epoch": 2.945826551964659,
+      "grad_norm": 2.378673020739392,
+      "learning_rate": 1.0040604346245319e-08,
+      "loss": 0.2682,
+      "step": 2110
+    },
+    {
+      "epoch": 2.9472215763775864,
+      "grad_norm": 1.724154583426404,
+      "learning_rate": 9.532464452491341e-09,
+      "loss": 0.2187,
+      "step": 2111
+    },
+    {
+      "epoch": 2.9486166007905137,
+      "grad_norm": 2.030830581414915,
+      "learning_rate": 9.037506433355325e-09,
+      "loss": 0.2265,
+      "step": 2112
+    },
+    {
+      "epoch": 2.950011625203441,
+      "grad_norm": 1.988165972381228,
+      "learning_rate": 8.555731596224803e-09,
+      "loss": 0.2115,
+      "step": 2113
+    },
+    {
+      "epoch": 2.9514066496163682,
+      "grad_norm": 2.0277716193856565,
+      "learning_rate": 8.087141213665717e-09,
+      "loss": 0.2384,
+      "step": 2114
+    },
+    {
+      "epoch": 2.9528016740292955,
+      "grad_norm": 1.9367303791899442,
+      "learning_rate": 7.631736523416867e-09,
+      "loss": 0.2296,
+      "step": 2115
+    },
+    {
+      "epoch": 2.9541966984422228,
+      "grad_norm": 2.0304044614481076,
+      "learning_rate": 7.1895187283899104e-09,
+      "loss": 0.1834,
+      "step": 2116
+    },
+    {
+      "epoch": 2.95559172285515,
+      "grad_norm": 2.1117935403333057,
+      "learning_rate": 6.760488996662706e-09,
+      "loss": 0.2321,
+      "step": 2117
+    },
+    {
+      "epoch": 2.9569867472680773,
+      "grad_norm": 1.9903261162086323,
+      "learning_rate": 6.3446484614798635e-09,
+      "loss": 0.197,
+      "step": 2118
+    },
+    {
+      "epoch": 2.9583817716810046,
+      "grad_norm": 1.9522993896671246,
+      "learning_rate": 5.941998221247192e-09,
+      "loss": 0.2359,
+      "step": 2119
+    },
+    {
+      "epoch": 2.959776796093932,
+      "grad_norm": 2.0867324978745687,
+      "learning_rate": 5.552539339528373e-09,
+      "loss": 0.2762,
+      "step": 2120
+    },
+    {
+      "epoch": 2.961171820506859,
+      "grad_norm": 1.850158984024511,
+      "learning_rate": 5.176272845045516e-09,
+      "loss": 0.2286,
+      "step": 2121
+    },
+    {
+      "epoch": 2.962566844919786,
+      "grad_norm": 1.9371278211250633,
+      "learning_rate": 4.813199731671381e-09,
+      "loss": 0.2036,
+      "step": 2122
+    },
+    {
+      "epoch": 2.963961869332713,
+      "grad_norm": 1.8464454139037667,
+      "learning_rate": 4.463320958432716e-09,
+      "loss": 0.1943,
+      "step": 2123
+    },
+    {
+      "epoch": 2.9653568937456405,
+      "grad_norm": 1.8850080704487493,
+      "learning_rate": 4.1266374495024795e-09,
+      "loss": 0.219,
+      "step": 2124
+    },
+    {
+      "epoch": 2.9667519181585678,
+      "grad_norm": 1.9760584152176828,
+      "learning_rate": 3.803150094200403e-09,
+      "loss": 0.2182,
+      "step": 2125
+    },
+    {
+      "epoch": 2.968146942571495,
+      "grad_norm": 1.9903736085094599,
+      "learning_rate": 3.4928597469885416e-09,
+      "loss": 0.2141,
+      "step": 2126
+    },
+    {
+      "epoch": 2.9695419669844223,
+      "grad_norm": 1.9826949238836484,
+      "learning_rate": 3.1957672274723907e-09,
+      "loss": 0.23,
+      "step": 2127
+    },
+    {
+      "epoch": 2.9709369913973496,
+      "grad_norm": 2.111435765657799,
+      "learning_rate": 2.9118733203942207e-09,
+      "loss": 0.2378,
+      "step": 2128
+    },
+    {
+      "epoch": 2.972332015810277,
+      "grad_norm": 2.0857714315335563,
+      "learning_rate": 2.6411787756353e-09,
+      "loss": 0.2335,
+      "step": 2129
+    },
+    {
+      "epoch": 2.9737270402232037,
+      "grad_norm": 1.9118431896872314,
+      "learning_rate": 2.3836843082108987e-09,
+      "loss": 0.2453,
+      "step": 2130
+    },
+    {
+      "epoch": 2.975122064636131,
+      "grad_norm": 1.8350015472662786,
+      "learning_rate": 2.1393905982691752e-09,
+      "loss": 0.2151,
+      "step": 2131
+    },
+    {
+      "epoch": 2.976517089049058,
+      "grad_norm": 2.0193930356279313,
+      "learning_rate": 1.9082982910911817e-09,
+      "loss": 0.2255,
+      "step": 2132
+    },
+    {
+      "epoch": 2.9779121134619855,
+      "grad_norm": 1.9778096181618947,
+      "learning_rate": 1.6904079970853083e-09,
+      "loss": 0.2194,
+      "step": 2133
+    },
+    {
+      "epoch": 2.9793071378749127,
+      "grad_norm": 1.869492540650723,
+      "learning_rate": 1.4857202917900604e-09,
+      "loss": 0.2221,
+      "step": 2134
+    },
+    {
+      "epoch": 2.98070216228784,
+      "grad_norm": 2.142769641662634,
+      "learning_rate": 1.2942357158701734e-09,
+      "loss": 0.2172,
+      "step": 2135
+    },
+    {
+      "epoch": 2.9820971867007673,
+      "grad_norm": 2.054017180212049,
+      "learning_rate": 1.1159547751143918e-09,
+      "loss": 0.2406,
+      "step": 2136
+    },
+    {
+      "epoch": 2.9834922111136946,
+      "grad_norm": 1.9105149879043084,
+      "learning_rate": 9.508779404360235e-10,
+      "loss": 0.2521,
+      "step": 2137
+    },
+    {
+      "epoch": 2.984887235526622,
+      "grad_norm": 2.112972558040102,
+      "learning_rate": 7.990056478707209e-10,
+      "loss": 0.2252,
+      "step": 2138
+    },
+    {
+      "epoch": 2.986282259939549,
+      "grad_norm": 1.898422373087025,
+      "learning_rate": 6.603382985759244e-10,
+      "loss": 0.2268,
+      "step": 2139
+    },
+    {
+      "epoch": 2.9876772843524764,
+      "grad_norm": 2.2452219566881477,
+      "learning_rate": 5.348762588286427e-10,
+      "loss": 0.2651,
+      "step": 2140
+    },
+    {
+      "epoch": 2.9890723087654036,
+      "grad_norm": 2.1778533368799264,
+      "learning_rate": 4.2261986002600783e-10,
+      "loss": 0.2521,
+      "step": 2141
+    },
+    {
+      "epoch": 2.9904673331783305,
+      "grad_norm": 2.1442535651439116,
+      "learning_rate": 3.235693986830546e-10,
+      "loss": 0.2588,
+      "step": 2142
+    },
+    {
+      "epoch": 2.9918623575912577,
+      "grad_norm": 2.119886361748237,
+      "learning_rate": 2.3772513643327555e-10,
+      "loss": 0.2408,
+      "step": 2143
+    },
+    {
+      "epoch": 2.993257382004185,
+      "grad_norm": 1.921028784949589,
+      "learning_rate": 1.650873000258457e-10,
+      "loss": 0.2254,
+      "step": 2144
+    },
+    {
+      "epoch": 2.9946524064171123,
+      "grad_norm": 2.0573904423649356,
+      "learning_rate": 1.0565608132728778e-10,
+      "loss": 0.2305,
+      "step": 2145
+    },
+    {
+      "epoch": 2.9960474308300395,
+      "grad_norm": 2.019354134088264,
+      "learning_rate": 5.943163732036183e-11,
+      "loss": 0.2517,
+      "step": 2146
+    },
+    {
+      "epoch": 2.997442455242967,
+      "grad_norm": 2.110601779062811,
+      "learning_rate": 2.6414090102400147e-11,
+      "loss": 0.2434,
+      "step": 2147
+    },
+    {
+      "epoch": 2.998837479655894,
+      "grad_norm": 2.144832780412135,
+      "learning_rate": 6.6035268864173e-12,
+      "loss": 0.2413,
+      "step": 2148
+    },
+    {
+      "epoch": 2.998837479655894,
+      "step": 2148,
+      "total_flos": 3.265949006466908e+17,
+      "train_loss": 0.4850710421606371,
+      "train_runtime": 18100.1171,
+      "train_samples_per_second": 2.851,
+      "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
+}