diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,22575 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3219,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0009319664492078285,
+      "grad_norm": 61.721725621787705,
+      "learning_rate": 1.5527950310559006e-07,
+      "loss": 14.0072,
+      "step": 1
+    },
+    {
+      "epoch": 0.001863932898415657,
+      "grad_norm": 62.384238066990214,
+      "learning_rate": 3.1055900621118013e-07,
+      "loss": 15.6069,
+      "step": 2
+    },
+    {
+      "epoch": 0.0027958993476234857,
+      "grad_norm": 59.80926664454978,
+      "learning_rate": 4.6583850931677024e-07,
+      "loss": 14.0348,
+      "step": 3
+    },
+    {
+      "epoch": 0.003727865796831314,
+      "grad_norm": 60.945171555598776,
+      "learning_rate": 6.211180124223603e-07,
+      "loss": 13.8137,
+      "step": 4
+    },
+    {
+      "epoch": 0.004659832246039142,
+      "grad_norm": 60.49195838756151,
+      "learning_rate": 7.763975155279503e-07,
+      "loss": 13.6986,
+      "step": 5
+    },
+    {
+      "epoch": 0.005591798695246971,
+      "grad_norm": 58.601784276318604,
+      "learning_rate": 9.316770186335405e-07,
+      "loss": 14.1002,
+      "step": 6
+    },
+    {
+      "epoch": 0.0065237651444548,
+      "grad_norm": 61.27166866120001,
+      "learning_rate": 1.0869565217391306e-06,
+      "loss": 14.1581,
+      "step": 7
+    },
+    {
+      "epoch": 0.007455731593662628,
+      "grad_norm": 62.37191838036504,
+      "learning_rate": 1.2422360248447205e-06,
+      "loss": 14.0852,
+      "step": 8
+    },
+    {
+      "epoch": 0.008387698042870456,
+      "grad_norm": 61.72006159940576,
+      "learning_rate": 1.3975155279503107e-06,
+      "loss": 13.7344,
+      "step": 9
+    },
+    {
+      "epoch": 0.009319664492078284,
+      "grad_norm": 61.69178105423322,
+      "learning_rate": 1.5527950310559006e-06,
+      "loss": 14.5083,
+      "step": 10
+    },
+    {
+      "epoch": 0.010251630941286114,
+      "grad_norm": 63.67303644133823,
+      "learning_rate": 1.7080745341614908e-06,
+      "loss": 13.4483,
+      "step": 11
+    },
+    {
+      "epoch": 0.011183597390493943,
+      "grad_norm": 82.36856212896062,
+      "learning_rate": 1.863354037267081e-06,
+      "loss": 12.7749,
+      "step": 12
+    },
+    {
+      "epoch": 0.012115563839701771,
+      "grad_norm": 86.58418517974022,
+      "learning_rate": 2.018633540372671e-06,
+      "loss": 12.6339,
+      "step": 13
+    },
+    {
+      "epoch": 0.0130475302889096,
+      "grad_norm": 93.81213444993256,
+      "learning_rate": 2.173913043478261e-06,
+      "loss": 12.5021,
+      "step": 14
+    },
+    {
+      "epoch": 0.013979496738117428,
+      "grad_norm": 99.28265683864647,
+      "learning_rate": 2.329192546583851e-06,
+      "loss": 11.6273,
+      "step": 15
+    },
+    {
+      "epoch": 0.014911463187325256,
+      "grad_norm": 81.72240717811445,
+      "learning_rate": 2.484472049689441e-06,
+      "loss": 7.4106,
+      "step": 16
+    },
+    {
+      "epoch": 0.015843429636533086,
+      "grad_norm": 68.7632169189392,
+      "learning_rate": 2.639751552795031e-06,
+      "loss": 6.9629,
+      "step": 17
+    },
+    {
+      "epoch": 0.016775396085740912,
+      "grad_norm": 60.275819178473995,
+      "learning_rate": 2.7950310559006214e-06,
+      "loss": 5.9715,
+      "step": 18
+    },
+    {
+      "epoch": 0.017707362534948742,
+      "grad_norm": 47.54814695459644,
+      "learning_rate": 2.9503105590062115e-06,
+      "loss": 6.0634,
+      "step": 19
+    },
+    {
+      "epoch": 0.01863932898415657,
+      "grad_norm": 38.1649452906253,
+      "learning_rate": 3.1055900621118013e-06,
+      "loss": 5.1814,
+      "step": 20
+    },
+    {
+      "epoch": 0.0195712954333644,
+      "grad_norm": 19.062650369664226,
+      "learning_rate": 3.2608695652173914e-06,
+      "loss": 4.4849,
+      "step": 21
+    },
+    {
+      "epoch": 0.02050326188257223,
+      "grad_norm": 8.56749014373029,
+      "learning_rate": 3.4161490683229816e-06,
+      "loss": 3.7968,
+      "step": 22
+    },
+    {
+      "epoch": 0.021435228331780055,
+      "grad_norm": 7.9704056954849625,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 3.9527,
+      "step": 23
+    },
+    {
+      "epoch": 0.022367194780987885,
+      "grad_norm": 6.544989393886097,
+      "learning_rate": 3.726708074534162e-06,
+      "loss": 3.7714,
+      "step": 24
+    },
+    {
+      "epoch": 0.023299161230195712,
+      "grad_norm": 6.28072941991866,
+      "learning_rate": 3.881987577639752e-06,
+      "loss": 4.1354,
+      "step": 25
+    },
+    {
+      "epoch": 0.024231127679403542,
+      "grad_norm": 6.260233735974329,
+      "learning_rate": 4.037267080745342e-06,
+      "loss": 4.0841,
+      "step": 26
+    },
+    {
+      "epoch": 0.02516309412861137,
+      "grad_norm": 6.6832257514348905,
+      "learning_rate": 4.192546583850932e-06,
+      "loss": 3.6622,
+      "step": 27
+    },
+    {
+      "epoch": 0.0260950605778192,
+      "grad_norm": 5.605380898113528,
+      "learning_rate": 4.347826086956522e-06,
+      "loss": 3.8678,
+      "step": 28
+    },
+    {
+      "epoch": 0.02702702702702703,
+      "grad_norm": 4.621754314747825,
+      "learning_rate": 4.503105590062112e-06,
+      "loss": 3.3442,
+      "step": 29
+    },
+    {
+      "epoch": 0.027958993476234855,
+      "grad_norm": 4.063991065496303,
+      "learning_rate": 4.658385093167702e-06,
+      "loss": 3.3079,
+      "step": 30
+    },
+    {
+      "epoch": 0.028890959925442685,
+      "grad_norm": 4.656681769730814,
+      "learning_rate": 4.813664596273292e-06,
+      "loss": 3.6787,
+      "step": 31
+    },
+    {
+      "epoch": 0.02982292637465051,
+      "grad_norm": 16.226487384453293,
+      "learning_rate": 4.968944099378882e-06,
+      "loss": 3.2509,
+      "step": 32
+    },
+    {
+      "epoch": 0.03075489282385834,
+      "grad_norm": 4.415933887388792,
+      "learning_rate": 5.124223602484472e-06,
+      "loss": 3.3509,
+      "step": 33
+    },
+    {
+      "epoch": 0.03168685927306617,
+      "grad_norm": 3.4623504169414843,
+      "learning_rate": 5.279503105590062e-06,
+      "loss": 3.0353,
+      "step": 34
+    },
+    {
+      "epoch": 0.032618825722273995,
+      "grad_norm": 3.47703268264277,
+      "learning_rate": 5.4347826086956525e-06,
+      "loss": 3.3995,
+      "step": 35
+    },
+    {
+      "epoch": 0.033550792171481825,
+      "grad_norm": 3.3045640793667688,
+      "learning_rate": 5.590062111801243e-06,
+      "loss": 2.6022,
+      "step": 36
+    },
+    {
+      "epoch": 0.034482758620689655,
+      "grad_norm": 3.3348350258870094,
+      "learning_rate": 5.745341614906832e-06,
+      "loss": 2.8759,
+      "step": 37
+    },
+    {
+      "epoch": 0.035414725069897485,
+      "grad_norm": 3.3905539588052376,
+      "learning_rate": 5.900621118012423e-06,
+      "loss": 2.7106,
+      "step": 38
+    },
+    {
+      "epoch": 0.036346691519105315,
+      "grad_norm": 3.374485258668037,
+      "learning_rate": 6.055900621118013e-06,
+      "loss": 2.8883,
+      "step": 39
+    },
+    {
+      "epoch": 0.03727865796831314,
+      "grad_norm": 3.4643319173868066,
+      "learning_rate": 6.2111801242236025e-06,
+      "loss": 2.5034,
+      "step": 40
+    },
+    {
+      "epoch": 0.03821062441752097,
+      "grad_norm": 3.291382459818416,
+      "learning_rate": 6.366459627329193e-06,
+      "loss": 2.562,
+      "step": 41
+    },
+    {
+      "epoch": 0.0391425908667288,
+      "grad_norm": 4.105935686013169,
+      "learning_rate": 6.521739130434783e-06,
+      "loss": 3.4783,
+      "step": 42
+    },
+    {
+      "epoch": 0.04007455731593663,
+      "grad_norm": 3.6889192854945607,
+      "learning_rate": 6.677018633540373e-06,
+      "loss": 3.2231,
+      "step": 43
+    },
+    {
+      "epoch": 0.04100652376514446,
+      "grad_norm": 2.9093690391275397,
+      "learning_rate": 6.832298136645963e-06,
+      "loss": 2.8329,
+      "step": 44
+    },
+    {
+      "epoch": 0.04193849021435228,
+      "grad_norm": 26.982703306959177,
+      "learning_rate": 6.9875776397515525e-06,
+      "loss": 2.7035,
+      "step": 45
+    },
+    {
+      "epoch": 0.04287045666356011,
+      "grad_norm": 3.3892779731204943,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 2.8287,
+      "step": 46
+    },
+    {
+      "epoch": 0.04380242311276794,
+      "grad_norm": 3.096607987974277,
+      "learning_rate": 7.298136645962733e-06,
+      "loss": 2.5728,
+      "step": 47
+    },
+    {
+      "epoch": 0.04473438956197577,
+      "grad_norm": 2.896393292821508,
+      "learning_rate": 7.453416149068324e-06,
+      "loss": 2.7195,
+      "step": 48
+    },
+    {
+      "epoch": 0.045666356011183594,
+      "grad_norm": 6.037605922550089,
+      "learning_rate": 7.608695652173914e-06,
+      "loss": 2.7581,
+      "step": 49
+    },
+    {
+      "epoch": 0.046598322460391424,
+      "grad_norm": 2.7806171065514667,
+      "learning_rate": 7.763975155279503e-06,
+      "loss": 2.4379,
+      "step": 50
+    },
+    {
+      "epoch": 0.047530288909599254,
+      "grad_norm": 2.9659879039840935,
+      "learning_rate": 7.919254658385093e-06,
+      "loss": 2.5124,
+      "step": 51
+    },
+    {
+      "epoch": 0.048462255358807084,
+      "grad_norm": 2.5612786020393097,
+      "learning_rate": 8.074534161490684e-06,
+      "loss": 2.2447,
+      "step": 52
+    },
+    {
+      "epoch": 0.049394221808014914,
+      "grad_norm": 2.628797759217477,
+      "learning_rate": 8.229813664596275e-06,
+      "loss": 2.4054,
+      "step": 53
+    },
+    {
+      "epoch": 0.05032618825722274,
+      "grad_norm": 2.7959306781007687,
+      "learning_rate": 8.385093167701864e-06,
+      "loss": 2.4986,
+      "step": 54
+    },
+    {
+      "epoch": 0.05125815470643057,
+      "grad_norm": 2.9750968950260988,
+      "learning_rate": 8.540372670807453e-06,
+      "loss": 2.5794,
+      "step": 55
+    },
+    {
+      "epoch": 0.0521901211556384,
+      "grad_norm": 2.9760764323095157,
+      "learning_rate": 8.695652173913044e-06,
+      "loss": 2.5993,
+      "step": 56
+    },
+    {
+      "epoch": 0.05312208760484623,
+      "grad_norm": 3.1655510837767675,
+      "learning_rate": 8.850931677018634e-06,
+      "loss": 2.0999,
+      "step": 57
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 2.665330388975331,
+      "learning_rate": 9.006211180124225e-06,
+      "loss": 2.4901,
+      "step": 58
+    },
+    {
+      "epoch": 0.05498602050326188,
+      "grad_norm": 4.077398423292217,
+      "learning_rate": 9.161490683229814e-06,
+      "loss": 2.4228,
+      "step": 59
+    },
+    {
+      "epoch": 0.05591798695246971,
+      "grad_norm": 2.7618780243382632,
+      "learning_rate": 9.316770186335403e-06,
+      "loss": 2.4636,
+      "step": 60
+    },
+    {
+      "epoch": 0.05684995340167754,
+      "grad_norm": 3.0263347905429336,
+      "learning_rate": 9.472049689440994e-06,
+      "loss": 2.258,
+      "step": 61
+    },
+    {
+      "epoch": 0.05778191985088537,
+      "grad_norm": 3.064501612556207,
+      "learning_rate": 9.627329192546584e-06,
+      "loss": 2.282,
+      "step": 62
+    },
+    {
+      "epoch": 0.05871388630009319,
+      "grad_norm": 2.15159931001617,
+      "learning_rate": 9.782608695652175e-06,
+      "loss": 2.0721,
+      "step": 63
+    },
+    {
+      "epoch": 0.05964585274930102,
+      "grad_norm": 2.7240439028893944,
+      "learning_rate": 9.937888198757764e-06,
+      "loss": 2.0928,
+      "step": 64
+    },
+    {
+      "epoch": 0.06057781919850885,
+      "grad_norm": 3.3221390037142817,
+      "learning_rate": 1.0093167701863353e-05,
+      "loss": 2.7245,
+      "step": 65
+    },
+    {
+      "epoch": 0.06150978564771668,
+      "grad_norm": 2.4624267508015287,
+      "learning_rate": 1.0248447204968944e-05,
+      "loss": 2.2824,
+      "step": 66
+    },
+    {
+      "epoch": 0.06244175209692451,
+      "grad_norm": 2.8958074489580996,
+      "learning_rate": 1.0403726708074535e-05,
+      "loss": 2.6111,
+      "step": 67
+    },
+    {
+      "epoch": 0.06337371854613234,
+      "grad_norm": 2.998411433567849,
+      "learning_rate": 1.0559006211180125e-05,
+      "loss": 2.3469,
+      "step": 68
+    },
+    {
+      "epoch": 0.06430568499534017,
+      "grad_norm": 2.5934574918054403,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 2.3979,
+      "step": 69
+    },
+    {
+      "epoch": 0.06523765144454799,
+      "grad_norm": 2.3529309988282496,
+      "learning_rate": 1.0869565217391305e-05,
+      "loss": 2.0489,
+      "step": 70
+    },
+    {
+      "epoch": 0.06616961789375582,
+      "grad_norm": 2.182564182337809,
+      "learning_rate": 1.1024844720496894e-05,
+      "loss": 2.2999,
+      "step": 71
+    },
+    {
+      "epoch": 0.06710158434296365,
+      "grad_norm": 2.512231511054208,
+      "learning_rate": 1.1180124223602485e-05,
+      "loss": 2.2745,
+      "step": 72
+    },
+    {
+      "epoch": 0.06803355079217148,
+      "grad_norm": 2.6859430554096857,
+      "learning_rate": 1.1335403726708076e-05,
+      "loss": 2.4296,
+      "step": 73
+    },
+    {
+      "epoch": 0.06896551724137931,
+      "grad_norm": 2.1485871848097924,
+      "learning_rate": 1.1490683229813664e-05,
+      "loss": 2.1396,
+      "step": 74
+    },
+    {
+      "epoch": 0.06989748369058714,
+      "grad_norm": 4.651923282360382,
+      "learning_rate": 1.1645962732919255e-05,
+      "loss": 2.7152,
+      "step": 75
+    },
+    {
+      "epoch": 0.07082945013979497,
+      "grad_norm": 3.472541579939051,
+      "learning_rate": 1.1801242236024846e-05,
+      "loss": 2.3504,
+      "step": 76
+    },
+    {
+      "epoch": 0.0717614165890028,
+      "grad_norm": 2.4887841700442945,
+      "learning_rate": 1.1956521739130435e-05,
+      "loss": 2.0742,
+      "step": 77
+    },
+    {
+      "epoch": 0.07269338303821063,
+      "grad_norm": 2.5498437903552964,
+      "learning_rate": 1.2111801242236026e-05,
+      "loss": 2.2942,
+      "step": 78
+    },
+    {
+      "epoch": 0.07362534948741846,
+      "grad_norm": 2.7346464586490904,
+      "learning_rate": 1.2267080745341616e-05,
+      "loss": 2.5123,
+      "step": 79
+    },
+    {
+      "epoch": 0.07455731593662628,
+      "grad_norm": 2.8721933825497725,
+      "learning_rate": 1.2422360248447205e-05,
+      "loss": 2.3694,
+      "step": 80
+    },
+    {
+      "epoch": 0.0754892823858341,
+      "grad_norm": 3.0473875939544124,
+      "learning_rate": 1.2577639751552794e-05,
+      "loss": 2.638,
+      "step": 81
+    },
+    {
+      "epoch": 0.07642124883504194,
+      "grad_norm": 2.3459431649289613,
+      "learning_rate": 1.2732919254658385e-05,
+      "loss": 2.0385,
+      "step": 82
+    },
+    {
+      "epoch": 0.07735321528424977,
+      "grad_norm": 2.6397232208015176,
+      "learning_rate": 1.2888198757763975e-05,
+      "loss": 2.358,
+      "step": 83
+    },
+    {
+      "epoch": 0.0782851817334576,
+      "grad_norm": 2.8343070288623298,
+      "learning_rate": 1.3043478260869566e-05,
+      "loss": 2.1533,
+      "step": 84
+    },
+    {
+      "epoch": 0.07921714818266543,
+      "grad_norm": 2.9726284337597644,
+      "learning_rate": 1.3198757763975155e-05,
+      "loss": 2.2943,
+      "step": 85
+    },
+    {
+      "epoch": 0.08014911463187326,
+      "grad_norm": 2.4391210885611367,
+      "learning_rate": 1.3354037267080746e-05,
+      "loss": 2.4044,
+      "step": 86
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 2.4694806114283714,
+      "learning_rate": 1.3509316770186337e-05,
+      "loss": 2.3851,
+      "step": 87
+    },
+    {
+      "epoch": 0.08201304753028892,
+      "grad_norm": 2.8173448393387712,
+      "learning_rate": 1.3664596273291926e-05,
+      "loss": 2.2549,
+      "step": 88
+    },
+    {
+      "epoch": 0.08294501397949673,
+      "grad_norm": 2.6402739694260076,
+      "learning_rate": 1.3819875776397517e-05,
+      "loss": 2.0099,
+      "step": 89
+    },
+    {
+      "epoch": 0.08387698042870456,
+      "grad_norm": 2.559690470331959,
+      "learning_rate": 1.3975155279503105e-05,
+      "loss": 2.2524,
+      "step": 90
+    },
+    {
+      "epoch": 0.08480894687791239,
+      "grad_norm": 2.3781173683940118,
+      "learning_rate": 1.4130434782608694e-05,
+      "loss": 2.2069,
+      "step": 91
+    },
+    {
+      "epoch": 0.08574091332712022,
+      "grad_norm": 2.606480001360248,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 2.0796,
+      "step": 92
+    },
+    {
+      "epoch": 0.08667287977632805,
+      "grad_norm": 2.572676262208739,
+      "learning_rate": 1.4440993788819876e-05,
+      "loss": 2.1165,
+      "step": 93
+    },
+    {
+      "epoch": 0.08760484622553588,
+      "grad_norm": 2.4352775485071336,
+      "learning_rate": 1.4596273291925466e-05,
+      "loss": 2.0478,
+      "step": 94
+    },
+    {
+      "epoch": 0.08853681267474371,
+      "grad_norm": 2.1596138972839567,
+      "learning_rate": 1.4751552795031057e-05,
+      "loss": 1.7324,
+      "step": 95
+    },
+    {
+      "epoch": 0.08946877912395154,
+      "grad_norm": 3.003118652208737,
+      "learning_rate": 1.4906832298136648e-05,
+      "loss": 2.2822,
+      "step": 96
+    },
+    {
+      "epoch": 0.09040074557315937,
+      "grad_norm": 2.593022558430485,
+      "learning_rate": 1.5062111801242237e-05,
+      "loss": 1.9756,
+      "step": 97
+    },
+    {
+      "epoch": 0.09133271202236719,
+      "grad_norm": 2.594904115515215,
+      "learning_rate": 1.5217391304347828e-05,
+      "loss": 2.339,
+      "step": 98
+    },
+    {
+      "epoch": 0.09226467847157502,
+      "grad_norm": 2.3498773520260787,
+      "learning_rate": 1.537267080745342e-05,
+      "loss": 1.9699,
+      "step": 99
+    },
+    {
+      "epoch": 0.09319664492078285,
+      "grad_norm": 3.035076684779126,
+      "learning_rate": 1.5527950310559007e-05,
+      "loss": 2.2438,
+      "step": 100
+    },
+    {
+      "epoch": 0.09412861136999068,
+      "grad_norm": 2.5412672136378434,
+      "learning_rate": 1.5683229813664594e-05,
+      "loss": 2.0582,
+      "step": 101
+    },
+    {
+      "epoch": 0.09506057781919851,
+      "grad_norm": 3.3733341207695657,
+      "learning_rate": 1.5838509316770185e-05,
+      "loss": 2.379,
+      "step": 102
+    },
+    {
+      "epoch": 0.09599254426840634,
+      "grad_norm": 2.3680049081912533,
+      "learning_rate": 1.5993788819875776e-05,
+      "loss": 1.8185,
+      "step": 103
+    },
+    {
+      "epoch": 0.09692451071761417,
+      "grad_norm": 2.6271437705770557,
+      "learning_rate": 1.6149068322981367e-05,
+      "loss": 2.066,
+      "step": 104
+    },
+    {
+      "epoch": 0.097856477166822,
+      "grad_norm": 2.8182872792882487,
+      "learning_rate": 1.630434782608696e-05,
+      "loss": 2.2031,
+      "step": 105
+    },
+    {
+      "epoch": 0.09878844361602983,
+      "grad_norm": 2.767546876759962,
+      "learning_rate": 1.645962732919255e-05,
+      "loss": 2.0362,
+      "step": 106
+    },
+    {
+      "epoch": 0.09972041006523766,
+      "grad_norm": 2.8720814708704303,
+      "learning_rate": 1.6614906832298137e-05,
+      "loss": 2.1619,
+      "step": 107
+    },
+    {
+      "epoch": 0.10065237651444547,
+      "grad_norm": 2.6357902117111673,
+      "learning_rate": 1.6770186335403728e-05,
+      "loss": 2.21,
+      "step": 108
+    },
+    {
+      "epoch": 0.1015843429636533,
+      "grad_norm": 3.2869336103657143,
+      "learning_rate": 1.6925465838509316e-05,
+      "loss": 2.4743,
+      "step": 109
+    },
+    {
+      "epoch": 0.10251630941286113,
+      "grad_norm": 2.12994816033784,
+      "learning_rate": 1.7080745341614907e-05,
+      "loss": 2.0892,
+      "step": 110
+    },
+    {
+      "epoch": 0.10344827586206896,
+      "grad_norm": 2.3931260443047084,
+      "learning_rate": 1.7236024844720498e-05,
+      "loss": 2.1917,
+      "step": 111
+    },
+    {
+      "epoch": 0.1043802423112768,
+      "grad_norm": 2.9349110249273824,
+      "learning_rate": 1.739130434782609e-05,
+      "loss": 2.4501,
+      "step": 112
+    },
+    {
+      "epoch": 0.10531220876048462,
+      "grad_norm": 3.0356661514611902,
+      "learning_rate": 1.7546583850931676e-05,
+      "loss": 2.1248,
+      "step": 113
+    },
+    {
+      "epoch": 0.10624417520969245,
+      "grad_norm": 2.6914840795173167,
+      "learning_rate": 1.7701863354037267e-05,
+      "loss": 2.061,
+      "step": 114
+    },
+    {
+      "epoch": 0.10717614165890028,
+      "grad_norm": 2.6997595610812537,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 1.9218,
+      "step": 115
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 2.239991232506967,
+      "learning_rate": 1.801242236024845e-05,
+      "loss": 1.9662,
+      "step": 116
+    },
+    {
+      "epoch": 0.10904007455731593,
+      "grad_norm": 2.852205414263623,
+      "learning_rate": 1.816770186335404e-05,
+      "loss": 2.1973,
+      "step": 117
+    },
+    {
+      "epoch": 0.10997204100652376,
+      "grad_norm": 2.157293184684896,
+      "learning_rate": 1.8322981366459628e-05,
+      "loss": 1.9781,
+      "step": 118
+    },
+    {
+      "epoch": 0.11090400745573159,
+      "grad_norm": 2.62101263911897,
+      "learning_rate": 1.8478260869565216e-05,
+      "loss": 1.979,
+      "step": 119
+    },
+    {
+      "epoch": 0.11183597390493942,
+      "grad_norm": 1.9445728143591678,
+      "learning_rate": 1.8633540372670807e-05,
+      "loss": 1.8359,
+      "step": 120
+    },
+    {
+      "epoch": 0.11276794035414725,
+      "grad_norm": 2.2116085786872848,
+      "learning_rate": 1.8788819875776398e-05,
+      "loss": 2.0713,
+      "step": 121
+    },
+    {
+      "epoch": 0.11369990680335508,
+      "grad_norm": 2.4376595708980004,
+      "learning_rate": 1.894409937888199e-05,
+      "loss": 2.0508,
+      "step": 122
+    },
+    {
+      "epoch": 0.11463187325256291,
+      "grad_norm": 2.9128237889329505,
+      "learning_rate": 1.909937888198758e-05,
+      "loss": 2.3079,
+      "step": 123
+    },
+    {
+      "epoch": 0.11556383970177074,
+      "grad_norm": 2.805039449739058,
+      "learning_rate": 1.9254658385093167e-05,
+      "loss": 2.1642,
+      "step": 124
+    },
+    {
+      "epoch": 0.11649580615097857,
+      "grad_norm": 2.8244936060351504,
+      "learning_rate": 1.940993788819876e-05,
+      "loss": 2.098,
+      "step": 125
+    },
+    {
+      "epoch": 0.11742777260018639,
+      "grad_norm": 2.4980897522764947,
+      "learning_rate": 1.956521739130435e-05,
+      "loss": 1.9346,
+      "step": 126
+    },
+    {
+      "epoch": 0.11835973904939422,
+      "grad_norm": 2.4441576181200437,
+      "learning_rate": 1.972049689440994e-05,
+      "loss": 2.4698,
+      "step": 127
+    },
+    {
+      "epoch": 0.11929170549860205,
+      "grad_norm": 2.900097739653629,
+      "learning_rate": 1.9875776397515528e-05,
+      "loss": 2.0206,
+      "step": 128
+    },
+    {
+      "epoch": 0.12022367194780988,
+      "grad_norm": 2.076427241514277,
+      "learning_rate": 2.003105590062112e-05,
+      "loss": 1.7837,
+      "step": 129
+    },
+    {
+      "epoch": 0.1211556383970177,
+      "grad_norm": 2.5906477450586602,
+      "learning_rate": 2.0186335403726707e-05,
+      "loss": 2.1037,
+      "step": 130
+    },
+    {
+      "epoch": 0.12208760484622554,
+      "grad_norm": 2.7244098380262445,
+      "learning_rate": 2.0341614906832298e-05,
+      "loss": 2.2477,
+      "step": 131
+    },
+    {
+      "epoch": 0.12301957129543337,
+      "grad_norm": 2.5152601892455078,
+      "learning_rate": 2.049689440993789e-05,
+      "loss": 2.2193,
+      "step": 132
+    },
+    {
+      "epoch": 0.1239515377446412,
+      "grad_norm": 2.299218065417288,
+      "learning_rate": 2.065217391304348e-05,
+      "loss": 2.1009,
+      "step": 133
+    },
+    {
+      "epoch": 0.12488350419384903,
+      "grad_norm": 2.377961927494741,
+      "learning_rate": 2.080745341614907e-05,
+      "loss": 2.0521,
+      "step": 134
+    },
+    {
+      "epoch": 0.12581547064305684,
+      "grad_norm": 3.066231195866783,
+      "learning_rate": 2.096273291925466e-05,
+      "loss": 2.0245,
+      "step": 135
+    },
+    {
+      "epoch": 0.1267474370922647,
+      "grad_norm": 2.3508081800385883,
+      "learning_rate": 2.111801242236025e-05,
+      "loss": 1.9608,
+      "step": 136
+    },
+    {
+      "epoch": 0.1276794035414725,
+      "grad_norm": 2.9759483024088027,
+      "learning_rate": 2.127329192546584e-05,
+      "loss": 2.2336,
+      "step": 137
+    },
+    {
+      "epoch": 0.12861136999068035,
+      "grad_norm": 2.521857964665751,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 2.2065,
+      "step": 138
+    },
+    {
+      "epoch": 0.12954333643988816,
+      "grad_norm": 2.220087446184091,
+      "learning_rate": 2.158385093167702e-05,
+      "loss": 1.8011,
+      "step": 139
+    },
+    {
+      "epoch": 0.13047530288909598,
+      "grad_norm": 2.5658763997201324,
+      "learning_rate": 2.173913043478261e-05,
+      "loss": 2.2217,
+      "step": 140
+    },
+    {
+      "epoch": 0.13140726933830382,
+      "grad_norm": 3.010983102241095,
+      "learning_rate": 2.1894409937888198e-05,
+      "loss": 2.3117,
+      "step": 141
+    },
+    {
+      "epoch": 0.13233923578751164,
+      "grad_norm": 3.154998737005876,
+      "learning_rate": 2.204968944099379e-05,
+      "loss": 2.6896,
+      "step": 142
+    },
+    {
+      "epoch": 0.13327120223671948,
+      "grad_norm": 2.6610756068293333,
+      "learning_rate": 2.220496894409938e-05,
+      "loss": 2.2412,
+      "step": 143
+    },
+    {
+      "epoch": 0.1342031686859273,
+      "grad_norm": 2.5483418602107712,
+      "learning_rate": 2.236024844720497e-05,
+      "loss": 1.7735,
+      "step": 144
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 2.48572794723197,
+      "learning_rate": 2.2515527950310562e-05,
+      "loss": 1.9725,
+      "step": 145
+    },
+    {
+      "epoch": 0.13606710158434296,
+      "grad_norm": 2.0474353777413965,
+      "learning_rate": 2.2670807453416153e-05,
+      "loss": 1.7776,
+      "step": 146
+    },
+    {
+      "epoch": 0.1369990680335508,
+      "grad_norm": 2.810501560863117,
+      "learning_rate": 2.282608695652174e-05,
+      "loss": 2.1104,
+      "step": 147
+    },
+    {
+      "epoch": 0.13793103448275862,
+      "grad_norm": 2.8524102705310135,
+      "learning_rate": 2.2981366459627328e-05,
+      "loss": 1.9505,
+      "step": 148
+    },
+    {
+      "epoch": 0.13886300093196646,
+      "grad_norm": 2.604331618125836,
+      "learning_rate": 2.313664596273292e-05,
+      "loss": 2.0968,
+      "step": 149
+    },
+    {
+      "epoch": 0.13979496738117428,
+      "grad_norm": 2.8193654656083873,
+      "learning_rate": 2.329192546583851e-05,
+      "loss": 2.2923,
+      "step": 150
+    },
+    {
+      "epoch": 0.1407269338303821,
+      "grad_norm": 2.9676683733487703,
+      "learning_rate": 2.34472049689441e-05,
+      "loss": 2.1493,
+      "step": 151
+    },
+    {
+      "epoch": 0.14165890027958994,
+      "grad_norm": 2.3467723024414355,
+      "learning_rate": 2.3602484472049692e-05,
+      "loss": 1.8444,
+      "step": 152
+    },
+    {
+      "epoch": 0.14259086672879775,
+      "grad_norm": 2.704857609741678,
+      "learning_rate": 2.375776397515528e-05,
+      "loss": 2.1072,
+      "step": 153
+    },
+    {
+      "epoch": 0.1435228331780056,
+      "grad_norm": 2.936991594649293,
+      "learning_rate": 2.391304347826087e-05,
+      "loss": 2.1277,
+      "step": 154
+    },
+    {
+      "epoch": 0.14445479962721341,
+      "grad_norm": 2.168427639519257,
+      "learning_rate": 2.4068322981366462e-05,
+      "loss": 1.7643,
+      "step": 155
+    },
+    {
+      "epoch": 0.14538676607642126,
+      "grad_norm": 2.183387039727307,
+      "learning_rate": 2.4223602484472053e-05,
+      "loss": 1.9756,
+      "step": 156
+    },
+    {
+      "epoch": 0.14631873252562907,
+      "grad_norm": 3.1618142483695153,
+      "learning_rate": 2.437888198757764e-05,
+      "loss": 2.1595,
+      "step": 157
+    },
+    {
+      "epoch": 0.14725069897483692,
+      "grad_norm": 2.6798312427416695,
+      "learning_rate": 2.453416149068323e-05,
+      "loss": 2.2759,
+      "step": 158
+    },
+    {
+      "epoch": 0.14818266542404473,
+      "grad_norm": 2.8631016480881444,
+      "learning_rate": 2.468944099378882e-05,
+      "loss": 2.0358,
+      "step": 159
+    },
+    {
+      "epoch": 0.14911463187325255,
+      "grad_norm": 2.5893015628222456,
+      "learning_rate": 2.484472049689441e-05,
+      "loss": 1.9141,
+      "step": 160
+    },
+    {
+      "epoch": 0.1500465983224604,
+      "grad_norm": 2.6219440228411366,
+      "learning_rate": 2.5e-05,
+      "loss": 1.7873,
+      "step": 161
+    },
+    {
+      "epoch": 0.1509785647716682,
+      "grad_norm": 2.1457462041141153,
+      "learning_rate": 2.515527950310559e-05,
+      "loss": 1.8142,
+      "step": 162
+    },
+    {
+      "epoch": 0.15191053122087605,
+      "grad_norm": 2.645406011008569,
+      "learning_rate": 2.5310559006211183e-05,
+      "loss": 1.9853,
+      "step": 163
+    },
+    {
+      "epoch": 0.15284249767008387,
+      "grad_norm": 2.8945426841217663,
+      "learning_rate": 2.546583850931677e-05,
+      "loss": 2.2036,
+      "step": 164
+    },
+    {
+      "epoch": 0.15377446411929171,
+      "grad_norm": 2.3775691336642604,
+      "learning_rate": 2.5621118012422362e-05,
+      "loss": 1.6476,
+      "step": 165
+    },
+    {
+      "epoch": 0.15470643056849953,
+      "grad_norm": 2.6409185939899333,
+      "learning_rate": 2.577639751552795e-05,
+      "loss": 2.0291,
+      "step": 166
+    },
+    {
+      "epoch": 0.15563839701770738,
+      "grad_norm": 2.536773193175286,
+      "learning_rate": 2.5931677018633544e-05,
+      "loss": 2.0713,
+      "step": 167
+    },
+    {
+      "epoch": 0.1565703634669152,
+      "grad_norm": 4.223533737767007,
+      "learning_rate": 2.608695652173913e-05,
+      "loss": 1.7925,
+      "step": 168
+    },
+    {
+      "epoch": 0.157502329916123,
+      "grad_norm": 2.4574779491973886,
+      "learning_rate": 2.6242236024844723e-05,
+      "loss": 1.8933,
+      "step": 169
+    },
+    {
+      "epoch": 0.15843429636533085,
+      "grad_norm": 2.5320409609390695,
+      "learning_rate": 2.639751552795031e-05,
+      "loss": 1.9855,
+      "step": 170
+    },
+    {
+      "epoch": 0.15936626281453867,
+      "grad_norm": 2.2708046404630386,
+      "learning_rate": 2.6552795031055898e-05,
+      "loss": 1.677,
+      "step": 171
+    },
+    {
+      "epoch": 0.1602982292637465,
+      "grad_norm": 2.3606001534530776,
+      "learning_rate": 2.6708074534161492e-05,
+      "loss": 1.7578,
+      "step": 172
+    },
+    {
+      "epoch": 0.16123019571295433,
+      "grad_norm": 3.2193125405215266,
+      "learning_rate": 2.686335403726708e-05,
+      "loss": 2.1556,
+      "step": 173
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 2.741105969678947,
+      "learning_rate": 2.7018633540372674e-05,
+      "loss": 2.0784,
+      "step": 174
+    },
+    {
+      "epoch": 0.16309412861137,
+      "grad_norm": 3.211795531220288,
+      "learning_rate": 2.7173913043478262e-05,
+      "loss": 2.1257,
+      "step": 175
+    },
+    {
+      "epoch": 0.16402609506057783,
+      "grad_norm": 2.180084802121087,
+      "learning_rate": 2.7329192546583853e-05,
+      "loss": 1.748,
+      "step": 176
+    },
+    {
+      "epoch": 0.16495806150978565,
+      "grad_norm": 3.155297234326803,
+      "learning_rate": 2.748447204968944e-05,
+      "loss": 2.1488,
+      "step": 177
+    },
+    {
+      "epoch": 0.16589002795899346,
+      "grad_norm": 2.6065896146930223,
+      "learning_rate": 2.7639751552795035e-05,
+      "loss": 1.8315,
+      "step": 178
+    },
+    {
+      "epoch": 0.1668219944082013,
+      "grad_norm": 2.278269628814607,
+      "learning_rate": 2.7795031055900623e-05,
+      "loss": 1.9055,
+      "step": 179
+    },
+    {
+      "epoch": 0.16775396085740912,
+      "grad_norm": 2.7252046027184154,
+      "learning_rate": 2.795031055900621e-05,
+      "loss": 2.0025,
+      "step": 180
+    },
+    {
+      "epoch": 0.16868592730661697,
+      "grad_norm": 2.4311320320962255,
+      "learning_rate": 2.8105590062111805e-05,
+      "loss": 1.5615,
+      "step": 181
+    },
+    {
+      "epoch": 0.16961789375582478,
+      "grad_norm": 2.7273615107837217,
+      "learning_rate": 2.826086956521739e-05,
+      "loss": 1.8117,
+      "step": 182
+    },
+    {
+      "epoch": 0.17054986020503263,
+      "grad_norm": 2.7103703396253773,
+      "learning_rate": 2.8416149068322983e-05,
+      "loss": 1.7033,
+      "step": 183
+    },
+    {
+      "epoch": 0.17148182665424044,
+      "grad_norm": 2.0301768043715396,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 1.7298,
+      "step": 184
+    },
+    {
+      "epoch": 0.1724137931034483,
+      "grad_norm": 2.354568424007958,
+      "learning_rate": 2.8726708074534165e-05,
+      "loss": 1.8992,
+      "step": 185
+    },
+    {
+      "epoch": 0.1733457595526561,
+      "grad_norm": 2.313675005542693,
+      "learning_rate": 2.8881987577639753e-05,
+      "loss": 1.7996,
+      "step": 186
+    },
+    {
+      "epoch": 0.17427772600186392,
+      "grad_norm": 2.287860583732967,
+      "learning_rate": 2.9037267080745344e-05,
+      "loss": 1.7267,
+      "step": 187
+    },
+    {
+      "epoch": 0.17520969245107176,
+      "grad_norm": 2.5197866518007648,
+      "learning_rate": 2.919254658385093e-05,
+      "loss": 2.0716,
+      "step": 188
+    },
+    {
+      "epoch": 0.17614165890027958,
+      "grad_norm": 2.2253864363831104,
+      "learning_rate": 2.9347826086956526e-05,
+      "loss": 1.6939,
+      "step": 189
+    },
+    {
+      "epoch": 0.17707362534948742,
+      "grad_norm": 2.4809952435976483,
+      "learning_rate": 2.9503105590062114e-05,
+      "loss": 2.2532,
+      "step": 190
+    },
+    {
+      "epoch": 0.17800559179869524,
+      "grad_norm": 2.552320820469185,
+      "learning_rate": 2.96583850931677e-05,
+      "loss": 1.9526,
+      "step": 191
+    },
+    {
+      "epoch": 0.17893755824790308,
+      "grad_norm": 3.4084144510026273,
+      "learning_rate": 2.9813664596273296e-05,
+      "loss": 2.3914,
+      "step": 192
+    },
+    {
+      "epoch": 0.1798695246971109,
+      "grad_norm": 2.7674182703949377,
+      "learning_rate": 2.9968944099378883e-05,
+      "loss": 2.2194,
+      "step": 193
+    },
+    {
+      "epoch": 0.18080149114631874,
+      "grad_norm": 2.3790128945771865,
+      "learning_rate": 3.0124223602484474e-05,
+      "loss": 1.7589,
+      "step": 194
+    },
+    {
+      "epoch": 0.18173345759552656,
+      "grad_norm": 2.688053042196188,
+      "learning_rate": 3.0279503105590062e-05,
+      "loss": 1.9788,
+      "step": 195
+    },
+    {
+      "epoch": 0.18266542404473438,
+      "grad_norm": 2.9545705314187956,
+      "learning_rate": 3.0434782608695656e-05,
+      "loss": 2.1315,
+      "step": 196
+    },
+    {
+      "epoch": 0.18359739049394222,
+      "grad_norm": 2.7457722794892674,
+      "learning_rate": 3.059006211180124e-05,
+      "loss": 1.8244,
+      "step": 197
+    },
+    {
+      "epoch": 0.18452935694315004,
+      "grad_norm": 2.833117005482504,
+      "learning_rate": 3.074534161490684e-05,
+      "loss": 1.9776,
+      "step": 198
+    },
+    {
+      "epoch": 0.18546132339235788,
+      "grad_norm": 1.703451711114753,
+      "learning_rate": 3.090062111801242e-05,
+      "loss": 1.5718,
+      "step": 199
+    },
+    {
+      "epoch": 0.1863932898415657,
+      "grad_norm": 2.490071923167969,
+      "learning_rate": 3.1055900621118014e-05,
+      "loss": 1.9003,
+      "step": 200
+    },
+    {
+      "epoch": 0.18732525629077354,
+      "grad_norm": 2.339140769079801,
+      "learning_rate": 3.1211180124223605e-05,
+      "loss": 1.8685,
+      "step": 201
+    },
+    {
+      "epoch": 0.18825722273998136,
+      "grad_norm": 2.142885545738482,
+      "learning_rate": 3.136645962732919e-05,
+      "loss": 1.9613,
+      "step": 202
+    },
+    {
+      "epoch": 0.1891891891891892,
+      "grad_norm": 2.6166704645896948,
+      "learning_rate": 3.152173913043479e-05,
+      "loss": 2.1163,
+      "step": 203
+    },
+    {
+      "epoch": 0.19012115563839702,
+      "grad_norm": 2.2613020474340577,
+      "learning_rate": 3.167701863354037e-05,
+      "loss": 1.8604,
+      "step": 204
+    },
+    {
+      "epoch": 0.19105312208760486,
+      "grad_norm": 2.3742984523515123,
+      "learning_rate": 3.183229813664597e-05,
+      "loss": 1.9108,
+      "step": 205
+    },
+    {
+      "epoch": 0.19198508853681268,
+      "grad_norm": 2.868737566371466,
+      "learning_rate": 3.198757763975155e-05,
+      "loss": 2.1021,
+      "step": 206
+    },
+    {
+      "epoch": 0.1929170549860205,
+      "grad_norm": 2.393828314560624,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 1.6608,
+      "step": 207
+    },
+    {
+      "epoch": 0.19384902143522834,
+      "grad_norm": 2.9289872912318264,
+      "learning_rate": 3.2298136645962735e-05,
+      "loss": 2.083,
+      "step": 208
+    },
+    {
+      "epoch": 0.19478098788443615,
+      "grad_norm": 2.1108055291841787,
+      "learning_rate": 3.245341614906832e-05,
+      "loss": 1.7121,
+      "step": 209
+    },
+    {
+      "epoch": 0.195712954333644,
+      "grad_norm": 2.362105645803653,
+      "learning_rate": 3.260869565217392e-05,
+      "loss": 1.6845,
+      "step": 210
+    },
+    {
+      "epoch": 0.1966449207828518,
+      "grad_norm": 2.3959321431893272,
+      "learning_rate": 3.27639751552795e-05,
+      "loss": 1.7631,
+      "step": 211
+    },
+    {
+      "epoch": 0.19757688723205966,
+      "grad_norm": 2.4746408860607882,
+      "learning_rate": 3.29192546583851e-05,
+      "loss": 1.6922,
+      "step": 212
+    },
+    {
+      "epoch": 0.19850885368126747,
+      "grad_norm": 2.597429617856143,
+      "learning_rate": 3.307453416149068e-05,
+      "loss": 1.9128,
+      "step": 213
+    },
+    {
+      "epoch": 0.19944082013047532,
+      "grad_norm": 3.0564613037601913,
+      "learning_rate": 3.3229813664596274e-05,
+      "loss": 2.0276,
+      "step": 214
+    },
+    {
+      "epoch": 0.20037278657968313,
+      "grad_norm": 2.4496151041836294,
+      "learning_rate": 3.3385093167701865e-05,
+      "loss": 1.7769,
+      "step": 215
+    },
+    {
+      "epoch": 0.20130475302889095,
+      "grad_norm": 1.8585471876163304,
+      "learning_rate": 3.3540372670807456e-05,
+      "loss": 1.6142,
+      "step": 216
+    },
+    {
+      "epoch": 0.2022367194780988,
+      "grad_norm": 2.9790991119132193,
+      "learning_rate": 3.369565217391305e-05,
+      "loss": 2.1531,
+      "step": 217
+    },
+    {
+      "epoch": 0.2031686859273066,
+      "grad_norm": 2.4358120772151235,
+      "learning_rate": 3.385093167701863e-05,
+      "loss": 1.7797,
+      "step": 218
+    },
+    {
+      "epoch": 0.20410065237651445,
+      "grad_norm": 2.5738415787810967,
+      "learning_rate": 3.400621118012422e-05,
+      "loss": 1.9322,
+      "step": 219
+    },
+    {
+      "epoch": 0.20503261882572227,
+      "grad_norm": 3.219906695103498,
+      "learning_rate": 3.4161490683229814e-05,
+      "loss": 2.0247,
+      "step": 220
+    },
+    {
+      "epoch": 0.2059645852749301,
+      "grad_norm": 2.373711598604349,
+      "learning_rate": 3.4316770186335405e-05,
+      "loss": 1.5522,
+      "step": 221
+    },
+    {
+      "epoch": 0.20689655172413793,
+      "grad_norm": 2.1321745289693754,
+      "learning_rate": 3.4472049689440996e-05,
+      "loss": 1.8802,
+      "step": 222
+    },
+    {
+      "epoch": 0.20782851817334577,
+      "grad_norm": 1.9834783937977842,
+      "learning_rate": 3.462732919254659e-05,
+      "loss": 1.7343,
+      "step": 223
+    },
+    {
+      "epoch": 0.2087604846225536,
+      "grad_norm": 2.2372817872270483,
+      "learning_rate": 3.478260869565218e-05,
+      "loss": 1.7117,
+      "step": 224
+    },
+    {
+      "epoch": 0.2096924510717614,
+      "grad_norm": 2.9136725378993775,
+      "learning_rate": 3.493788819875777e-05,
+      "loss": 1.7654,
+      "step": 225
+    },
+    {
+      "epoch": 0.21062441752096925,
+      "grad_norm": 2.45309559843133,
+      "learning_rate": 3.509316770186335e-05,
+      "loss": 1.7888,
+      "step": 226
+    },
+    {
+      "epoch": 0.21155638397017706,
+      "grad_norm": 2.511302370240779,
+      "learning_rate": 3.524844720496895e-05,
+      "loss": 1.9258,
+      "step": 227
+    },
+    {
+      "epoch": 0.2124883504193849,
+      "grad_norm": 2.255651452675934,
+      "learning_rate": 3.5403726708074535e-05,
+      "loss": 1.6131,
+      "step": 228
+    },
+    {
+      "epoch": 0.21342031686859272,
+      "grad_norm": 2.854970508408505,
+      "learning_rate": 3.5559006211180126e-05,
+      "loss": 1.8918,
+      "step": 229
+    },
+    {
+      "epoch": 0.21435228331780057,
+      "grad_norm": 2.7033153406288197,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 1.9638,
+      "step": 230
+    },
+    {
+      "epoch": 0.21528424976700838,
+      "grad_norm": 2.467841175418176,
+      "learning_rate": 3.58695652173913e-05,
+      "loss": 1.8903,
+      "step": 231
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 2.21207856082642,
+      "learning_rate": 3.60248447204969e-05,
+      "loss": 1.724,
+      "step": 232
+    },
+    {
+      "epoch": 0.21714818266542404,
+      "grad_norm": 2.448527160238523,
+      "learning_rate": 3.618012422360248e-05,
+      "loss": 1.9403,
+      "step": 233
+    },
+    {
+      "epoch": 0.21808014911463186,
+      "grad_norm": 2.2101085986897564,
+      "learning_rate": 3.633540372670808e-05,
+      "loss": 1.6642,
+      "step": 234
+    },
+    {
+      "epoch": 0.2190121155638397,
+      "grad_norm": 2.1835487754107894,
+      "learning_rate": 3.6490683229813665e-05,
+      "loss": 1.7725,
+      "step": 235
+    },
+    {
+      "epoch": 0.21994408201304752,
+      "grad_norm": 3.139473265432975,
+      "learning_rate": 3.6645962732919256e-05,
+      "loss": 1.9177,
+      "step": 236
+    },
+    {
+      "epoch": 0.22087604846225536,
+      "grad_norm": 2.9996949291029473,
+      "learning_rate": 3.680124223602485e-05,
+      "loss": 2.292,
+      "step": 237
+    },
+    {
+      "epoch": 0.22180801491146318,
+      "grad_norm": 2.5905773029080006,
+      "learning_rate": 3.695652173913043e-05,
+      "loss": 1.885,
+      "step": 238
+    },
+    {
+      "epoch": 0.22273998136067102,
+      "grad_norm": 2.3794899873341415,
+      "learning_rate": 3.711180124223603e-05,
+      "loss": 1.6383,
+      "step": 239
+    },
+    {
+      "epoch": 0.22367194780987884,
+      "grad_norm": 2.592014980266268,
+      "learning_rate": 3.7267080745341614e-05,
+      "loss": 2.0515,
+      "step": 240
+    },
+    {
+      "epoch": 0.22460391425908668,
+      "grad_norm": 2.7570234089365506,
+      "learning_rate": 3.742236024844721e-05,
+      "loss": 2.0306,
+      "step": 241
+    },
+    {
+      "epoch": 0.2255358807082945,
+      "grad_norm": 2.759277281855068,
+      "learning_rate": 3.7577639751552796e-05,
+      "loss": 1.8731,
+      "step": 242
+    },
+    {
+      "epoch": 0.22646784715750232,
+      "grad_norm": 2.3540701218901576,
+      "learning_rate": 3.773291925465839e-05,
+      "loss": 1.7182,
+      "step": 243
+    },
+    {
+      "epoch": 0.22739981360671016,
+      "grad_norm": 2.5962567882032013,
+      "learning_rate": 3.788819875776398e-05,
+      "loss": 1.5832,
+      "step": 244
+    },
+    {
+      "epoch": 0.22833178005591798,
+      "grad_norm": 2.2728386265232157,
+      "learning_rate": 3.804347826086957e-05,
+      "loss": 1.7719,
+      "step": 245
+    },
+    {
+      "epoch": 0.22926374650512582,
+      "grad_norm": 2.3460955353657202,
+      "learning_rate": 3.819875776397516e-05,
+      "loss": 1.6519,
+      "step": 246
+    },
+    {
+      "epoch": 0.23019571295433364,
+      "grad_norm": 2.0797642408850097,
+      "learning_rate": 3.8354037267080744e-05,
+      "loss": 1.6078,
+      "step": 247
+    },
+    {
+      "epoch": 0.23112767940354148,
+      "grad_norm": 2.475905370409815,
+      "learning_rate": 3.8509316770186335e-05,
+      "loss": 1.795,
+      "step": 248
+    },
+    {
+      "epoch": 0.2320596458527493,
+      "grad_norm": 2.3704099034759105,
+      "learning_rate": 3.8664596273291926e-05,
+      "loss": 1.8821,
+      "step": 249
+    },
+    {
+      "epoch": 0.23299161230195714,
+      "grad_norm": 2.197058324956388,
+      "learning_rate": 3.881987577639752e-05,
+      "loss": 1.5014,
+      "step": 250
+    },
+    {
+      "epoch": 0.23392357875116496,
+      "grad_norm": 3.1323366792509266,
+      "learning_rate": 3.897515527950311e-05,
+      "loss": 2.1915,
+      "step": 251
+    },
+    {
+      "epoch": 0.23485554520037277,
+      "grad_norm": 2.5349189992633656,
+      "learning_rate": 3.91304347826087e-05,
+      "loss": 1.7888,
+      "step": 252
+    },
+    {
+      "epoch": 0.23578751164958062,
+      "grad_norm": 2.802730538833947,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 1.7743,
+      "step": 253
+    },
+    {
+      "epoch": 0.23671947809878843,
+      "grad_norm": 2.577034328633359,
+      "learning_rate": 3.944099378881988e-05,
+      "loss": 2.0294,
+      "step": 254
+    },
+    {
+      "epoch": 0.23765144454799628,
+      "grad_norm": 2.2119460551720187,
+      "learning_rate": 3.9596273291925465e-05,
+      "loss": 1.8061,
+      "step": 255
+    },
+    {
+      "epoch": 0.2385834109972041,
+      "grad_norm": 3.4383417637553118,
+      "learning_rate": 3.9751552795031056e-05,
+      "loss": 1.746,
+      "step": 256
+    },
+    {
+      "epoch": 0.23951537744641194,
+      "grad_norm": 2.582955720750917,
+      "learning_rate": 3.990683229813665e-05,
+      "loss": 2.1232,
+      "step": 257
+    },
+    {
+      "epoch": 0.24044734389561975,
+      "grad_norm": 2.3411837578320087,
+      "learning_rate": 4.006211180124224e-05,
+      "loss": 1.9428,
+      "step": 258
+    },
+    {
+      "epoch": 0.2413793103448276,
+      "grad_norm": 2.413742551931853,
+      "learning_rate": 4.021739130434783e-05,
+      "loss": 1.6664,
+      "step": 259
+    },
+    {
+      "epoch": 0.2423112767940354,
+      "grad_norm": 2.6553539204409895,
+      "learning_rate": 4.0372670807453414e-05,
+      "loss": 1.9972,
+      "step": 260
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 2.411670536378713,
+      "learning_rate": 4.052795031055901e-05,
+      "loss": 1.7565,
+      "step": 261
+    },
+    {
+      "epoch": 0.24417520969245107,
+      "grad_norm": 2.638793932386771,
+      "learning_rate": 4.0683229813664596e-05,
+      "loss": 1.6998,
+      "step": 262
+    },
+    {
+      "epoch": 0.2451071761416589,
+      "grad_norm": 2.466826616717204,
+      "learning_rate": 4.0838509316770193e-05,
+      "loss": 1.9753,
+      "step": 263
+    },
+    {
+      "epoch": 0.24603914259086673,
+      "grad_norm": 2.0418967612477803,
+      "learning_rate": 4.099378881987578e-05,
+      "loss": 1.5446,
+      "step": 264
+    },
+    {
+      "epoch": 0.24697110904007455,
+      "grad_norm": 2.392936576165752,
+      "learning_rate": 4.114906832298137e-05,
+      "loss": 1.7848,
+      "step": 265
+    },
+    {
+      "epoch": 0.2479030754892824,
+      "grad_norm": 2.7286905312625174,
+      "learning_rate": 4.130434782608696e-05,
+      "loss": 1.9773,
+      "step": 266
+    },
+    {
+      "epoch": 0.2488350419384902,
+      "grad_norm": 2.1925841143158573,
+      "learning_rate": 4.1459627329192544e-05,
+      "loss": 1.8416,
+      "step": 267
+    },
+    {
+      "epoch": 0.24976700838769805,
+      "grad_norm": 2.6136090925418447,
+      "learning_rate": 4.161490683229814e-05,
+      "loss": 1.8997,
+      "step": 268
+    },
+    {
+      "epoch": 0.2506989748369059,
+      "grad_norm": 2.7327312060942357,
+      "learning_rate": 4.1770186335403726e-05,
+      "loss": 2.0315,
+      "step": 269
+    },
+    {
+      "epoch": 0.2516309412861137,
+      "grad_norm": 2.0929171558933013,
+      "learning_rate": 4.192546583850932e-05,
+      "loss": 1.434,
+      "step": 270
+    },
+    {
+      "epoch": 0.25256290773532153,
+      "grad_norm": 2.396566549793493,
+      "learning_rate": 4.208074534161491e-05,
+      "loss": 1.7252,
+      "step": 271
+    },
+    {
+      "epoch": 0.2534948741845294,
+      "grad_norm": 2.3878509171632434,
+      "learning_rate": 4.22360248447205e-05,
+      "loss": 1.7377,
+      "step": 272
+    },
+    {
+      "epoch": 0.25442684063373716,
+      "grad_norm": 2.422576137712497,
+      "learning_rate": 4.239130434782609e-05,
+      "loss": 1.7881,
+      "step": 273
+    },
+    {
+      "epoch": 0.255358807082945,
+      "grad_norm": 2.236260493388797,
+      "learning_rate": 4.254658385093168e-05,
+      "loss": 1.5817,
+      "step": 274
+    },
+    {
+      "epoch": 0.25629077353215285,
+      "grad_norm": 2.3600947341629066,
+      "learning_rate": 4.270186335403727e-05,
+      "loss": 1.7472,
+      "step": 275
+    },
+    {
+      "epoch": 0.2572227399813607,
+      "grad_norm": 2.431458954851922,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 1.9279,
+      "step": 276
+    },
+    {
+      "epoch": 0.2581547064305685,
+      "grad_norm": 2.4575171738992623,
+      "learning_rate": 4.301242236024845e-05,
+      "loss": 1.5367,
+      "step": 277
+    },
+    {
+      "epoch": 0.2590866728797763,
+      "grad_norm": 2.730502878936673,
+      "learning_rate": 4.316770186335404e-05,
+      "loss": 1.8912,
+      "step": 278
+    },
+    {
+      "epoch": 0.26001863932898417,
+      "grad_norm": 2.7556361507549467,
+      "learning_rate": 4.332298136645963e-05,
+      "loss": 2.0461,
+      "step": 279
+    },
+    {
+      "epoch": 0.26095060577819196,
+      "grad_norm": 2.7121612928827172,
+      "learning_rate": 4.347826086956522e-05,
+      "loss": 1.7671,
+      "step": 280
+    },
+    {
+      "epoch": 0.2618825722273998,
+      "grad_norm": 2.0308695816402795,
+      "learning_rate": 4.363354037267081e-05,
+      "loss": 1.4448,
+      "step": 281
+    },
+    {
+      "epoch": 0.26281453867660765,
+      "grad_norm": 2.337508598486232,
+      "learning_rate": 4.3788819875776396e-05,
+      "loss": 1.7392,
+      "step": 282
+    },
+    {
+      "epoch": 0.2637465051258155,
+      "grad_norm": 2.23336940076628,
+      "learning_rate": 4.3944099378881993e-05,
+      "loss": 1.7005,
+      "step": 283
+    },
+    {
+      "epoch": 0.2646784715750233,
+      "grad_norm": 2.5640335079878684,
+      "learning_rate": 4.409937888198758e-05,
+      "loss": 1.7851,
+      "step": 284
+    },
+    {
+      "epoch": 0.2656104380242311,
+      "grad_norm": 1.9705675051918328,
+      "learning_rate": 4.425465838509317e-05,
+      "loss": 1.5371,
+      "step": 285
+    },
+    {
+      "epoch": 0.26654240447343897,
+      "grad_norm": 2.3234163546534976,
+      "learning_rate": 4.440993788819876e-05,
+      "loss": 1.9927,
+      "step": 286
+    },
+    {
+      "epoch": 0.2674743709226468,
+      "grad_norm": 1.9502587181835287,
+      "learning_rate": 4.456521739130435e-05,
+      "loss": 1.7609,
+      "step": 287
+    },
+    {
+      "epoch": 0.2684063373718546,
+      "grad_norm": 2.0307444046480843,
+      "learning_rate": 4.472049689440994e-05,
+      "loss": 1.7249,
+      "step": 288
+    },
+    {
+      "epoch": 0.26933830382106244,
+      "grad_norm": 2.509671278493603,
+      "learning_rate": 4.4875776397515526e-05,
+      "loss": 1.7396,
+      "step": 289
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 2.211146042267718,
+      "learning_rate": 4.5031055900621124e-05,
+      "loss": 1.5224,
+      "step": 290
+    },
+    {
+      "epoch": 0.2712022367194781,
+      "grad_norm": 2.4385742346969335,
+      "learning_rate": 4.518633540372671e-05,
+      "loss": 1.7423,
+      "step": 291
+    },
+    {
+      "epoch": 0.2721342031686859,
+      "grad_norm": 2.469159380839025,
+      "learning_rate": 4.5341614906832306e-05,
+      "loss": 1.7954,
+      "step": 292
+    },
+    {
+      "epoch": 0.27306616961789376,
+      "grad_norm": 2.5587994819861444,
+      "learning_rate": 4.549689440993789e-05,
+      "loss": 2.0163,
+      "step": 293
+    },
+    {
+      "epoch": 0.2739981360671016,
+      "grad_norm": 2.548684323320384,
+      "learning_rate": 4.565217391304348e-05,
+      "loss": 1.7903,
+      "step": 294
+    },
+    {
+      "epoch": 0.2749301025163094,
+      "grad_norm": 2.2307082617613507,
+      "learning_rate": 4.580745341614907e-05,
+      "loss": 1.5123,
+      "step": 295
+    },
+    {
+      "epoch": 0.27586206896551724,
+      "grad_norm": 2.725200176158627,
+      "learning_rate": 4.5962732919254656e-05,
+      "loss": 1.9998,
+      "step": 296
+    },
+    {
+      "epoch": 0.2767940354147251,
+      "grad_norm": 2.492441504693259,
+      "learning_rate": 4.6118012422360254e-05,
+      "loss": 1.9188,
+      "step": 297
+    },
+    {
+      "epoch": 0.2777260018639329,
+      "grad_norm": 2.410930483526035,
+      "learning_rate": 4.627329192546584e-05,
+      "loss": 1.6573,
+      "step": 298
+    },
+    {
+      "epoch": 0.2786579683131407,
+      "grad_norm": 2.9228571011471445,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 1.8353,
+      "step": 299
+    },
+    {
+      "epoch": 0.27958993476234856,
+      "grad_norm": 2.7232086417920116,
+      "learning_rate": 4.658385093167702e-05,
+      "loss": 1.8166,
+      "step": 300
+    },
+    {
+      "epoch": 0.2805219012115564,
+      "grad_norm": 2.7736215131371136,
+      "learning_rate": 4.673913043478261e-05,
+      "loss": 2.1412,
+      "step": 301
+    },
+    {
+      "epoch": 0.2814538676607642,
+      "grad_norm": 2.4848976962294484,
+      "learning_rate": 4.68944099378882e-05,
+      "loss": 1.691,
+      "step": 302
+    },
+    {
+      "epoch": 0.28238583410997203,
+      "grad_norm": 2.5976500776490585,
+      "learning_rate": 4.7049689440993793e-05,
+      "loss": 1.8756,
+      "step": 303
+    },
+    {
+      "epoch": 0.2833178005591799,
+      "grad_norm": 2.457831122992162,
+      "learning_rate": 4.7204968944099384e-05,
+      "loss": 1.9808,
+      "step": 304
+    },
+    {
+      "epoch": 0.2842497670083877,
+      "grad_norm": 2.4250975329031386,
+      "learning_rate": 4.736024844720497e-05,
+      "loss": 1.7863,
+      "step": 305
+    },
+    {
+      "epoch": 0.2851817334575955,
+      "grad_norm": 2.8976005705091588,
+      "learning_rate": 4.751552795031056e-05,
+      "loss": 1.6011,
+      "step": 306
+    },
+    {
+      "epoch": 0.28611369990680335,
+      "grad_norm": 2.3736509880868133,
+      "learning_rate": 4.767080745341615e-05,
+      "loss": 1.6424,
+      "step": 307
+    },
+    {
+      "epoch": 0.2870456663560112,
+      "grad_norm": 2.6993936881229272,
+      "learning_rate": 4.782608695652174e-05,
+      "loss": 1.8895,
+      "step": 308
+    },
+    {
+      "epoch": 0.287977632805219,
+      "grad_norm": 3.173897704546431,
+      "learning_rate": 4.798136645962733e-05,
+      "loss": 2.0394,
+      "step": 309
+    },
+    {
+      "epoch": 0.28890959925442683,
+      "grad_norm": 2.6041033703851078,
+      "learning_rate": 4.8136645962732924e-05,
+      "loss": 2.2275,
+      "step": 310
+    },
+    {
+      "epoch": 0.2898415657036347,
+      "grad_norm": 2.244820203964956,
+      "learning_rate": 4.829192546583851e-05,
+      "loss": 1.6167,
+      "step": 311
+    },
+    {
+      "epoch": 0.2907735321528425,
+      "grad_norm": 2.4883189094113844,
+      "learning_rate": 4.8447204968944106e-05,
+      "loss": 1.8408,
+      "step": 312
+    },
+    {
+      "epoch": 0.2917054986020503,
+      "grad_norm": 2.074975082307135,
+      "learning_rate": 4.860248447204969e-05,
+      "loss": 1.4865,
+      "step": 313
+    },
+    {
+      "epoch": 0.29263746505125815,
+      "grad_norm": 2.298863682537741,
+      "learning_rate": 4.875776397515528e-05,
+      "loss": 1.7034,
+      "step": 314
+    },
+    {
+      "epoch": 0.293569431500466,
+      "grad_norm": 2.4311984251544096,
+      "learning_rate": 4.891304347826087e-05,
+      "loss": 2.0033,
+      "step": 315
+    },
+    {
+      "epoch": 0.29450139794967384,
+      "grad_norm": 2.039876292740557,
+      "learning_rate": 4.906832298136646e-05,
+      "loss": 1.7639,
+      "step": 316
+    },
+    {
+      "epoch": 0.2954333643988816,
+      "grad_norm": 2.728785661665197,
+      "learning_rate": 4.9223602484472054e-05,
+      "loss": 1.8643,
+      "step": 317
+    },
+    {
+      "epoch": 0.29636533084808947,
+      "grad_norm": 2.164477376748269,
+      "learning_rate": 4.937888198757764e-05,
+      "loss": 1.904,
+      "step": 318
+    },
+    {
+      "epoch": 0.2972972972972973,
+      "grad_norm": 2.8682847974071666,
+      "learning_rate": 4.9534161490683236e-05,
+      "loss": 1.8029,
+      "step": 319
+    },
+    {
+      "epoch": 0.2982292637465051,
+      "grad_norm": 2.806153346320384,
+      "learning_rate": 4.968944099378882e-05,
+      "loss": 1.9412,
+      "step": 320
+    },
+    {
+      "epoch": 0.29916123019571295,
+      "grad_norm": 2.3489240683925705,
+      "learning_rate": 4.984472049689442e-05,
+      "loss": 1.6176,
+      "step": 321
+    },
+    {
+      "epoch": 0.3000931966449208,
+      "grad_norm": 2.116620057146439,
+      "learning_rate": 5e-05,
+      "loss": 1.7347,
+      "step": 322
+    },
+    {
+      "epoch": 0.30102516309412863,
+      "grad_norm": 2.577519166216306,
+      "learning_rate": 4.998274076630998e-05,
+      "loss": 1.8586,
+      "step": 323
+    },
+    {
+      "epoch": 0.3019571295433364,
+      "grad_norm": 2.2640534927864278,
+      "learning_rate": 4.996548153261996e-05,
+      "loss": 1.7437,
+      "step": 324
+    },
+    {
+      "epoch": 0.30288909599254427,
+      "grad_norm": 2.090208366793187,
+      "learning_rate": 4.994822229892993e-05,
+      "loss": 1.8101,
+      "step": 325
+    },
+    {
+      "epoch": 0.3038210624417521,
+      "grad_norm": 2.3322525102696754,
+      "learning_rate": 4.9930963065239906e-05,
+      "loss": 1.5742,
+      "step": 326
+    },
+    {
+      "epoch": 0.3047530288909599,
+      "grad_norm": 2.5443043412954105,
+      "learning_rate": 4.991370383154988e-05,
+      "loss": 1.7505,
+      "step": 327
+    },
+    {
+      "epoch": 0.30568499534016774,
+      "grad_norm": 2.410178901638368,
+      "learning_rate": 4.989644459785986e-05,
+      "loss": 1.7955,
+      "step": 328
+    },
+    {
+      "epoch": 0.3066169617893756,
+      "grad_norm": 2.555429344281188,
+      "learning_rate": 4.9879185364169836e-05,
+      "loss": 1.7335,
+      "step": 329
+    },
+    {
+      "epoch": 0.30754892823858343,
+      "grad_norm": 2.330056728482757,
+      "learning_rate": 4.986192613047981e-05,
+      "loss": 1.6738,
+      "step": 330
+    },
+    {
+      "epoch": 0.3084808946877912,
+      "grad_norm": 2.8965164114723767,
+      "learning_rate": 4.9844666896789785e-05,
+      "loss": 2.2223,
+      "step": 331
+    },
+    {
+      "epoch": 0.30941286113699906,
+      "grad_norm": 2.393327679964644,
+      "learning_rate": 4.982740766309976e-05,
+      "loss": 1.7896,
+      "step": 332
+    },
+    {
+      "epoch": 0.3103448275862069,
+      "grad_norm": 2.1004306391100274,
+      "learning_rate": 4.981014842940973e-05,
+      "loss": 1.6748,
+      "step": 333
+    },
+    {
+      "epoch": 0.31127679403541475,
+      "grad_norm": 2.218417134489124,
+      "learning_rate": 4.979288919571971e-05,
+      "loss": 1.6451,
+      "step": 334
+    },
+    {
+      "epoch": 0.31220876048462254,
+      "grad_norm": 2.0514742742422314,
+      "learning_rate": 4.977562996202969e-05,
+      "loss": 1.6261,
+      "step": 335
+    },
+    {
+      "epoch": 0.3131407269338304,
+      "grad_norm": 1.9900111482447533,
+      "learning_rate": 4.975837072833966e-05,
+      "loss": 1.4542,
+      "step": 336
+    },
+    {
+      "epoch": 0.3140726933830382,
+      "grad_norm": 2.0157859331884134,
+      "learning_rate": 4.974111149464964e-05,
+      "loss": 1.556,
+      "step": 337
+    },
+    {
+      "epoch": 0.315004659832246,
+      "grad_norm": 1.9135003057006486,
+      "learning_rate": 4.972385226095961e-05,
+      "loss": 1.7163,
+      "step": 338
+    },
+    {
+      "epoch": 0.31593662628145386,
+      "grad_norm": 2.2570057825363334,
+      "learning_rate": 4.970659302726959e-05,
+      "loss": 1.9754,
+      "step": 339
+    },
+    {
+      "epoch": 0.3168685927306617,
+      "grad_norm": 2.6067448496857297,
+      "learning_rate": 4.968933379357957e-05,
+      "loss": 1.86,
+      "step": 340
+    },
+    {
+      "epoch": 0.31780055917986955,
+      "grad_norm": 3.0630751103066824,
+      "learning_rate": 4.967207455988954e-05,
+      "loss": 1.6862,
+      "step": 341
+    },
+    {
+      "epoch": 0.31873252562907733,
+      "grad_norm": 4.030804380417776,
+      "learning_rate": 4.9654815326199515e-05,
+      "loss": 1.8608,
+      "step": 342
+    },
+    {
+      "epoch": 0.3196644920782852,
+      "grad_norm": 2.234618493377559,
+      "learning_rate": 4.9637556092509496e-05,
+      "loss": 1.9488,
+      "step": 343
+    },
+    {
+      "epoch": 0.320596458527493,
+      "grad_norm": 2.4963039989583264,
+      "learning_rate": 4.962029685881947e-05,
+      "loss": 1.9044,
+      "step": 344
+    },
+    {
+      "epoch": 0.32152842497670087,
+      "grad_norm": 2.29254369995935,
+      "learning_rate": 4.9603037625129445e-05,
+      "loss": 1.4577,
+      "step": 345
+    },
+    {
+      "epoch": 0.32246039142590865,
+      "grad_norm": 2.4531422903951734,
+      "learning_rate": 4.9585778391439426e-05,
+      "loss": 1.5813,
+      "step": 346
+    },
+    {
+      "epoch": 0.3233923578751165,
+      "grad_norm": 2.0254711915931445,
+      "learning_rate": 4.95685191577494e-05,
+      "loss": 1.6147,
+      "step": 347
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 2.1626034488689356,
+      "learning_rate": 4.9551259924059375e-05,
+      "loss": 1.6767,
+      "step": 348
+    },
+    {
+      "epoch": 0.32525629077353213,
+      "grad_norm": 1.980250864625762,
+      "learning_rate": 4.953400069036935e-05,
+      "loss": 1.5093,
+      "step": 349
+    },
+    {
+      "epoch": 0.32618825722274,
+      "grad_norm": 2.441344856125904,
+      "learning_rate": 4.951674145667933e-05,
+      "loss": 1.965,
+      "step": 350
+    },
+    {
+      "epoch": 0.3271202236719478,
+      "grad_norm": 2.290922765507851,
+      "learning_rate": 4.9499482222989304e-05,
+      "loss": 1.5906,
+      "step": 351
+    },
+    {
+      "epoch": 0.32805219012115566,
+      "grad_norm": 2.242552358846413,
+      "learning_rate": 4.948222298929928e-05,
+      "loss": 1.4928,
+      "step": 352
+    },
+    {
+      "epoch": 0.32898415657036345,
+      "grad_norm": 2.3049347047747353,
+      "learning_rate": 4.946496375560925e-05,
+      "loss": 1.7921,
+      "step": 353
+    },
+    {
+      "epoch": 0.3299161230195713,
+      "grad_norm": 3.7223064896267424,
+      "learning_rate": 4.944770452191923e-05,
+      "loss": 2.1261,
+      "step": 354
+    },
+    {
+      "epoch": 0.33084808946877914,
+      "grad_norm": 2.7846777605368414,
+      "learning_rate": 4.94304452882292e-05,
+      "loss": 1.8979,
+      "step": 355
+    },
+    {
+      "epoch": 0.3317800559179869,
+      "grad_norm": 2.6348233741644207,
+      "learning_rate": 4.9413186054539176e-05,
+      "loss": 1.9805,
+      "step": 356
+    },
+    {
+      "epoch": 0.33271202236719477,
+      "grad_norm": 2.3850668137099227,
+      "learning_rate": 4.939592682084916e-05,
+      "loss": 1.7627,
+      "step": 357
+    },
+    {
+      "epoch": 0.3336439888164026,
+      "grad_norm": 2.668982618714907,
+      "learning_rate": 4.937866758715913e-05,
+      "loss": 1.7615,
+      "step": 358
+    },
+    {
+      "epoch": 0.33457595526561046,
+      "grad_norm": 2.261335075997684,
+      "learning_rate": 4.9361408353469105e-05,
+      "loss": 1.7078,
+      "step": 359
+    },
+    {
+      "epoch": 0.33550792171481825,
+      "grad_norm": 2.3155777730724005,
+      "learning_rate": 4.934414911977908e-05,
+      "loss": 1.6028,
+      "step": 360
+    },
+    {
+      "epoch": 0.3364398881640261,
+      "grad_norm": 2.6538520688212874,
+      "learning_rate": 4.932688988608906e-05,
+      "loss": 1.7358,
+      "step": 361
+    },
+    {
+      "epoch": 0.33737185461323393,
+      "grad_norm": 2.7738145436638204,
+      "learning_rate": 4.9309630652399035e-05,
+      "loss": 1.4935,
+      "step": 362
+    },
+    {
+      "epoch": 0.3383038210624418,
+      "grad_norm": 2.614497848589433,
+      "learning_rate": 4.929237141870901e-05,
+      "loss": 1.7664,
+      "step": 363
+    },
+    {
+      "epoch": 0.33923578751164957,
+      "grad_norm": 2.451757273192885,
+      "learning_rate": 4.927511218501899e-05,
+      "loss": 1.8783,
+      "step": 364
+    },
+    {
+      "epoch": 0.3401677539608574,
+      "grad_norm": 2.4387215113417158,
+      "learning_rate": 4.9257852951328965e-05,
+      "loss": 1.7498,
+      "step": 365
+    },
+    {
+      "epoch": 0.34109972041006525,
+      "grad_norm": 2.7913734263963637,
+      "learning_rate": 4.924059371763894e-05,
+      "loss": 2.1034,
+      "step": 366
+    },
+    {
+      "epoch": 0.34203168685927304,
+      "grad_norm": 2.280202469388489,
+      "learning_rate": 4.922333448394891e-05,
+      "loss": 1.7523,
+      "step": 367
+    },
+    {
+      "epoch": 0.3429636533084809,
+      "grad_norm": 1.64597529118763,
+      "learning_rate": 4.9206075250258894e-05,
+      "loss": 1.4617,
+      "step": 368
+    },
+    {
+      "epoch": 0.34389561975768873,
+      "grad_norm": 2.4980145549794375,
+      "learning_rate": 4.918881601656887e-05,
+      "loss": 1.5815,
+      "step": 369
+    },
+    {
+      "epoch": 0.3448275862068966,
+      "grad_norm": 2.122940010221761,
+      "learning_rate": 4.917155678287884e-05,
+      "loss": 1.5202,
+      "step": 370
+    },
+    {
+      "epoch": 0.34575955265610436,
+      "grad_norm": 2.148795022386655,
+      "learning_rate": 4.915429754918882e-05,
+      "loss": 1.5989,
+      "step": 371
+    },
+    {
+      "epoch": 0.3466915191053122,
+      "grad_norm": 2.6182576395045687,
+      "learning_rate": 4.91370383154988e-05,
+      "loss": 1.7428,
+      "step": 372
+    },
+    {
+      "epoch": 0.34762348555452005,
+      "grad_norm": 2.498145152337295,
+      "learning_rate": 4.911977908180877e-05,
+      "loss": 1.8439,
+      "step": 373
+    },
+    {
+      "epoch": 0.34855545200372784,
+      "grad_norm": 2.853297876871486,
+      "learning_rate": 4.910251984811875e-05,
+      "loss": 2.0834,
+      "step": 374
+    },
+    {
+      "epoch": 0.3494874184529357,
+      "grad_norm": 2.041888050950634,
+      "learning_rate": 4.908526061442872e-05,
+      "loss": 1.6088,
+      "step": 375
+    },
+    {
+      "epoch": 0.3504193849021435,
+      "grad_norm": 2.2903338017050756,
+      "learning_rate": 4.9068001380738695e-05,
+      "loss": 1.8955,
+      "step": 376
+    },
+    {
+      "epoch": 0.35135135135135137,
+      "grad_norm": 2.200630718186369,
+      "learning_rate": 4.905074214704867e-05,
+      "loss": 1.5895,
+      "step": 377
+    },
+    {
+      "epoch": 0.35228331780055916,
+      "grad_norm": 1.9199910678254593,
+      "learning_rate": 4.9033482913358644e-05,
+      "loss": 1.772,
+      "step": 378
+    },
+    {
+      "epoch": 0.353215284249767,
+      "grad_norm": 2.3018041668110336,
+      "learning_rate": 4.9016223679668625e-05,
+      "loss": 1.5763,
+      "step": 379
+    },
+    {
+      "epoch": 0.35414725069897485,
+      "grad_norm": 2.1004780527024716,
+      "learning_rate": 4.89989644459786e-05,
+      "loss": 1.6315,
+      "step": 380
+    },
+    {
+      "epoch": 0.3550792171481827,
+      "grad_norm": 2.2283222491291332,
+      "learning_rate": 4.8981705212288574e-05,
+      "loss": 1.8745,
+      "step": 381
+    },
+    {
+      "epoch": 0.3560111835973905,
+      "grad_norm": 2.270558890598022,
+      "learning_rate": 4.896444597859855e-05,
+      "loss": 1.8489,
+      "step": 382
+    },
+    {
+      "epoch": 0.3569431500465983,
+      "grad_norm": 2.178140379060242,
+      "learning_rate": 4.894718674490853e-05,
+      "loss": 1.5983,
+      "step": 383
+    },
+    {
+      "epoch": 0.35787511649580617,
+      "grad_norm": 2.284359748413507,
+      "learning_rate": 4.89299275112185e-05,
+      "loss": 1.7317,
+      "step": 384
+    },
+    {
+      "epoch": 0.35880708294501396,
+      "grad_norm": 1.9916879538066206,
+      "learning_rate": 4.891266827752848e-05,
+      "loss": 1.8797,
+      "step": 385
+    },
+    {
+      "epoch": 0.3597390493942218,
+      "grad_norm": 1.6365215219778348,
+      "learning_rate": 4.889540904383846e-05,
+      "loss": 1.2384,
+      "step": 386
+    },
+    {
+      "epoch": 0.36067101584342964,
+      "grad_norm": 2.145298329517641,
+      "learning_rate": 4.887814981014843e-05,
+      "loss": 1.8015,
+      "step": 387
+    },
+    {
+      "epoch": 0.3616029822926375,
+      "grad_norm": 2.5121252836105596,
+      "learning_rate": 4.886089057645841e-05,
+      "loss": 2.1432,
+      "step": 388
+    },
+    {
+      "epoch": 0.3625349487418453,
+      "grad_norm": 2.2162638301125175,
+      "learning_rate": 4.884363134276838e-05,
+      "loss": 1.7253,
+      "step": 389
+    },
+    {
+      "epoch": 0.3634669151910531,
+      "grad_norm": 2.23616691052989,
+      "learning_rate": 4.882637210907836e-05,
+      "loss": 1.6069,
+      "step": 390
+    },
+    {
+      "epoch": 0.36439888164026096,
+      "grad_norm": 1.945982449988874,
+      "learning_rate": 4.880911287538834e-05,
+      "loss": 1.5307,
+      "step": 391
+    },
+    {
+      "epoch": 0.36533084808946875,
+      "grad_norm": 1.7113825442339592,
+      "learning_rate": 4.879185364169831e-05,
+      "loss": 1.4996,
+      "step": 392
+    },
+    {
+      "epoch": 0.3662628145386766,
+      "grad_norm": 2.98728256820357,
+      "learning_rate": 4.877459440800829e-05,
+      "loss": 2.1,
+      "step": 393
+    },
+    {
+      "epoch": 0.36719478098788444,
+      "grad_norm": 2.1315558068965896,
+      "learning_rate": 4.8757335174318266e-05,
+      "loss": 1.4943,
+      "step": 394
+    },
+    {
+      "epoch": 0.3681267474370923,
+      "grad_norm": 2.1443606253290834,
+      "learning_rate": 4.874007594062824e-05,
+      "loss": 1.6384,
+      "step": 395
+    },
+    {
+      "epoch": 0.36905871388630007,
+      "grad_norm": 2.498897073180855,
+      "learning_rate": 4.8722816706938215e-05,
+      "loss": 1.8995,
+      "step": 396
+    },
+    {
+      "epoch": 0.3699906803355079,
+      "grad_norm": 2.06943119158177,
+      "learning_rate": 4.870555747324819e-05,
+      "loss": 1.4368,
+      "step": 397
+    },
+    {
+      "epoch": 0.37092264678471576,
+      "grad_norm": 2.597960129568427,
+      "learning_rate": 4.8688298239558164e-05,
+      "loss": 2.0382,
+      "step": 398
+    },
+    {
+      "epoch": 0.3718546132339236,
+      "grad_norm": 2.050536951910299,
+      "learning_rate": 4.867103900586814e-05,
+      "loss": 1.4564,
+      "step": 399
+    },
+    {
+      "epoch": 0.3727865796831314,
+      "grad_norm": 2.1302706061049506,
+      "learning_rate": 4.865377977217811e-05,
+      "loss": 1.9221,
+      "step": 400
+    },
+    {
+      "epoch": 0.37371854613233924,
+      "grad_norm": 2.8828601757891805,
+      "learning_rate": 4.863652053848809e-05,
+      "loss": 2.0846,
+      "step": 401
+    },
+    {
+      "epoch": 0.3746505125815471,
+      "grad_norm": 2.557660695822736,
+      "learning_rate": 4.861926130479807e-05,
+      "loss": 1.6723,
+      "step": 402
+    },
+    {
+      "epoch": 0.37558247903075487,
+      "grad_norm": 2.4021341595178782,
+      "learning_rate": 4.860200207110804e-05,
+      "loss": 1.7845,
+      "step": 403
+    },
+    {
+      "epoch": 0.3765144454799627,
+      "grad_norm": 2.5827088622195884,
+      "learning_rate": 4.858474283741802e-05,
+      "loss": 1.554,
+      "step": 404
+    },
+    {
+      "epoch": 0.37744641192917056,
+      "grad_norm": 2.5241502602820356,
+      "learning_rate": 4.8567483603728e-05,
+      "loss": 1.7935,
+      "step": 405
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 2.1655646425490263,
+      "learning_rate": 4.855022437003797e-05,
+      "loss": 1.691,
+      "step": 406
+    },
+    {
+      "epoch": 0.3793103448275862,
+      "grad_norm": 2.1958334711738425,
+      "learning_rate": 4.8532965136347946e-05,
+      "loss": 1.5138,
+      "step": 407
+    },
+    {
+      "epoch": 0.38024231127679403,
+      "grad_norm": 2.4607434188214596,
+      "learning_rate": 4.851570590265793e-05,
+      "loss": 1.6133,
+      "step": 408
+    },
+    {
+      "epoch": 0.3811742777260019,
+      "grad_norm": 2.375502017111872,
+      "learning_rate": 4.84984466689679e-05,
+      "loss": 1.6224,
+      "step": 409
+    },
+    {
+      "epoch": 0.3821062441752097,
+      "grad_norm": 1.846830335517278,
+      "learning_rate": 4.8481187435277875e-05,
+      "loss": 1.5932,
+      "step": 410
+    },
+    {
+      "epoch": 0.3830382106244175,
+      "grad_norm": 1.914078621303638,
+      "learning_rate": 4.846392820158785e-05,
+      "loss": 1.4322,
+      "step": 411
+    },
+    {
+      "epoch": 0.38397017707362535,
+      "grad_norm": 2.5365411984122344,
+      "learning_rate": 4.844666896789783e-05,
+      "loss": 1.7428,
+      "step": 412
+    },
+    {
+      "epoch": 0.3849021435228332,
+      "grad_norm": 2.8169500561551564,
+      "learning_rate": 4.8429409734207805e-05,
+      "loss": 1.6714,
+      "step": 413
+    },
+    {
+      "epoch": 0.385834109972041,
+      "grad_norm": 2.6503629928229673,
+      "learning_rate": 4.841215050051778e-05,
+      "loss": 1.7922,
+      "step": 414
+    },
+    {
+      "epoch": 0.38676607642124883,
+      "grad_norm": 2.832006543169936,
+      "learning_rate": 4.839489126682776e-05,
+      "loss": 1.9968,
+      "step": 415
+    },
+    {
+      "epoch": 0.38769804287045667,
+      "grad_norm": 1.8543978327074264,
+      "learning_rate": 4.8377632033137735e-05,
+      "loss": 1.5264,
+      "step": 416
+    },
+    {
+      "epoch": 0.3886300093196645,
+      "grad_norm": 2.3775491994762437,
+      "learning_rate": 4.836037279944771e-05,
+      "loss": 1.9969,
+      "step": 417
+    },
+    {
+      "epoch": 0.3895619757688723,
+      "grad_norm": 1.982943143995858,
+      "learning_rate": 4.834311356575768e-05,
+      "loss": 1.4726,
+      "step": 418
+    },
+    {
+      "epoch": 0.39049394221808015,
+      "grad_norm": 1.8385000427392217,
+      "learning_rate": 4.832585433206766e-05,
+      "loss": 1.5906,
+      "step": 419
+    },
+    {
+      "epoch": 0.391425908667288,
+      "grad_norm": 2.120946234156906,
+      "learning_rate": 4.830859509837763e-05,
+      "loss": 1.3405,
+      "step": 420
+    },
+    {
+      "epoch": 0.3923578751164958,
+      "grad_norm": 2.1031538019144462,
+      "learning_rate": 4.8291335864687606e-05,
+      "loss": 1.6157,
+      "step": 421
+    },
+    {
+      "epoch": 0.3932898415657036,
+      "grad_norm": 1.8953876098842153,
+      "learning_rate": 4.827407663099758e-05,
+      "loss": 1.5679,
+      "step": 422
+    },
+    {
+      "epoch": 0.39422180801491147,
+      "grad_norm": 1.8541260748456552,
+      "learning_rate": 4.825681739730756e-05,
+      "loss": 1.5211,
+      "step": 423
+    },
+    {
+      "epoch": 0.3951537744641193,
+      "grad_norm": 2.3123530125309713,
+      "learning_rate": 4.8239558163617536e-05,
+      "loss": 1.7355,
+      "step": 424
+    },
+    {
+      "epoch": 0.3960857409133271,
+      "grad_norm": 2.584548183329704,
+      "learning_rate": 4.822229892992751e-05,
+      "loss": 1.8732,
+      "step": 425
+    },
+    {
+      "epoch": 0.39701770736253494,
+      "grad_norm": 1.879243121790465,
+      "learning_rate": 4.820503969623749e-05,
+      "loss": 1.6891,
+      "step": 426
+    },
+    {
+      "epoch": 0.3979496738117428,
+      "grad_norm": 2.1868516453271445,
+      "learning_rate": 4.8187780462547465e-05,
+      "loss": 1.4688,
+      "step": 427
+    },
+    {
+      "epoch": 0.39888164026095063,
+      "grad_norm": 2.606899480929107,
+      "learning_rate": 4.817052122885744e-05,
+      "loss": 1.7294,
+      "step": 428
+    },
+    {
+      "epoch": 0.3998136067101584,
+      "grad_norm": 2.2716693298612602,
+      "learning_rate": 4.8153261995167414e-05,
+      "loss": 1.4426,
+      "step": 429
+    },
+    {
+      "epoch": 0.40074557315936626,
+      "grad_norm": 2.201516340609497,
+      "learning_rate": 4.8136002761477395e-05,
+      "loss": 1.5739,
+      "step": 430
+    },
+    {
+      "epoch": 0.4016775396085741,
+      "grad_norm": 1.8553701814656958,
+      "learning_rate": 4.811874352778737e-05,
+      "loss": 1.5574,
+      "step": 431
+    },
+    {
+      "epoch": 0.4026095060577819,
+      "grad_norm": 2.1757393432914007,
+      "learning_rate": 4.8101484294097344e-05,
+      "loss": 1.7567,
+      "step": 432
+    },
+    {
+      "epoch": 0.40354147250698974,
+      "grad_norm": 2.3440360567455967,
+      "learning_rate": 4.808422506040732e-05,
+      "loss": 1.5784,
+      "step": 433
+    },
+    {
+      "epoch": 0.4044734389561976,
+      "grad_norm": 2.5179137854821163,
+      "learning_rate": 4.80669658267173e-05,
+      "loss": 1.5836,
+      "step": 434
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 2.2608614407466865,
+      "learning_rate": 4.804970659302727e-05,
+      "loss": 1.6252,
+      "step": 435
+    },
+    {
+      "epoch": 0.4063373718546132,
+      "grad_norm": 2.073624239652641,
+      "learning_rate": 4.803244735933725e-05,
+      "loss": 1.7098,
+      "step": 436
+    },
+    {
+      "epoch": 0.40726933830382106,
+      "grad_norm": 2.0505606287949196,
+      "learning_rate": 4.801518812564723e-05,
+      "loss": 1.6316,
+      "step": 437
+    },
+    {
+      "epoch": 0.4082013047530289,
+      "grad_norm": 2.2400900951047595,
+      "learning_rate": 4.79979288919572e-05,
+      "loss": 1.555,
+      "step": 438
+    },
+    {
+      "epoch": 0.4091332712022367,
+      "grad_norm": 1.7431591395488863,
+      "learning_rate": 4.798066965826718e-05,
+      "loss": 1.4488,
+      "step": 439
+    },
+    {
+      "epoch": 0.41006523765144454,
+      "grad_norm": 2.1020371414942,
+      "learning_rate": 4.796341042457715e-05,
+      "loss": 1.5458,
+      "step": 440
+    },
+    {
+      "epoch": 0.4109972041006524,
+      "grad_norm": 2.45063120843792,
+      "learning_rate": 4.7946151190887126e-05,
+      "loss": 1.6751,
+      "step": 441
+    },
+    {
+      "epoch": 0.4119291705498602,
+      "grad_norm": 2.8130922198082953,
+      "learning_rate": 4.79288919571971e-05,
+      "loss": 1.5425,
+      "step": 442
+    },
+    {
+      "epoch": 0.412861136999068,
+      "grad_norm": 2.3188014819975664,
+      "learning_rate": 4.7911632723507074e-05,
+      "loss": 1.7772,
+      "step": 443
+    },
+    {
+      "epoch": 0.41379310344827586,
+      "grad_norm": 2.488182091344662,
+      "learning_rate": 4.7894373489817056e-05,
+      "loss": 1.5266,
+      "step": 444
+    },
+    {
+      "epoch": 0.4147250698974837,
+      "grad_norm": 1.8674185984969744,
+      "learning_rate": 4.787711425612703e-05,
+      "loss": 1.3721,
+      "step": 445
+    },
+    {
+      "epoch": 0.41565703634669154,
+      "grad_norm": 2.103989704660401,
+      "learning_rate": 4.7859855022437004e-05,
+      "loss": 1.8408,
+      "step": 446
+    },
+    {
+      "epoch": 0.41658900279589933,
+      "grad_norm": 1.7817391660999073,
+      "learning_rate": 4.784259578874698e-05,
+      "loss": 1.6646,
+      "step": 447
+    },
+    {
+      "epoch": 0.4175209692451072,
+      "grad_norm": 2.1564076058293513,
+      "learning_rate": 4.782533655505696e-05,
+      "loss": 1.6697,
+      "step": 448
+    },
+    {
+      "epoch": 0.418452935694315,
+      "grad_norm": 2.6727658514780153,
+      "learning_rate": 4.7808077321366934e-05,
+      "loss": 1.8301,
+      "step": 449
+    },
+    {
+      "epoch": 0.4193849021435228,
+      "grad_norm": 2.21012674316267,
+      "learning_rate": 4.779081808767691e-05,
+      "loss": 1.5792,
+      "step": 450
+    },
+    {
+      "epoch": 0.42031686859273065,
+      "grad_norm": 2.1570363581765113,
+      "learning_rate": 4.777355885398688e-05,
+      "loss": 1.3867,
+      "step": 451
+    },
+    {
+      "epoch": 0.4212488350419385,
+      "grad_norm": 2.0608151172586084,
+      "learning_rate": 4.7756299620296863e-05,
+      "loss": 1.7013,
+      "step": 452
+    },
+    {
+      "epoch": 0.42218080149114634,
+      "grad_norm": 2.047618762805517,
+      "learning_rate": 4.773904038660684e-05,
+      "loss": 1.46,
+      "step": 453
+    },
+    {
+      "epoch": 0.42311276794035413,
+      "grad_norm": 2.0706682147959397,
+      "learning_rate": 4.772178115291681e-05,
+      "loss": 1.5664,
+      "step": 454
+    },
+    {
+      "epoch": 0.424044734389562,
+      "grad_norm": 1.703813213998405,
+      "learning_rate": 4.770452191922679e-05,
+      "loss": 1.349,
+      "step": 455
+    },
+    {
+      "epoch": 0.4249767008387698,
+      "grad_norm": 2.3178832450352918,
+      "learning_rate": 4.768726268553677e-05,
+      "loss": 1.7057,
+      "step": 456
+    },
+    {
+      "epoch": 0.42590866728797766,
+      "grad_norm": 2.0866938660211063,
+      "learning_rate": 4.767000345184674e-05,
+      "loss": 1.3272,
+      "step": 457
+    },
+    {
+      "epoch": 0.42684063373718545,
+      "grad_norm": 2.1860381336503063,
+      "learning_rate": 4.7652744218156716e-05,
+      "loss": 1.5189,
+      "step": 458
+    },
+    {
+      "epoch": 0.4277726001863933,
+      "grad_norm": 2.0831756364782597,
+      "learning_rate": 4.76354849844667e-05,
+      "loss": 1.4172,
+      "step": 459
+    },
+    {
+      "epoch": 0.42870456663560114,
+      "grad_norm": 2.8041496158490316,
+      "learning_rate": 4.761822575077667e-05,
+      "loss": 1.7621,
+      "step": 460
+    },
+    {
+      "epoch": 0.4296365330848089,
+      "grad_norm": 2.405500403221895,
+      "learning_rate": 4.7600966517086646e-05,
+      "loss": 1.8754,
+      "step": 461
+    },
+    {
+      "epoch": 0.43056849953401677,
+      "grad_norm": 2.5642193433593823,
+      "learning_rate": 4.758370728339662e-05,
+      "loss": 1.9845,
+      "step": 462
+    },
+    {
+      "epoch": 0.4315004659832246,
+      "grad_norm": 2.042469407019576,
+      "learning_rate": 4.7566448049706594e-05,
+      "loss": 1.7624,
+      "step": 463
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 2.1519515286834707,
+      "learning_rate": 4.754918881601657e-05,
+      "loss": 1.6899,
+      "step": 464
+    },
+    {
+      "epoch": 0.43336439888164024,
+      "grad_norm": 2.0436079914648007,
+      "learning_rate": 4.753192958232654e-05,
+      "loss": 1.6541,
+      "step": 465
+    },
+    {
+      "epoch": 0.4342963653308481,
+      "grad_norm": 2.9745504927151467,
+      "learning_rate": 4.7514670348636524e-05,
+      "loss": 1.6727,
+      "step": 466
+    },
+    {
+      "epoch": 0.43522833178005593,
+      "grad_norm": 2.5772107694272037,
+      "learning_rate": 4.74974111149465e-05,
+      "loss": 1.7472,
+      "step": 467
+    },
+    {
+      "epoch": 0.4361602982292637,
+      "grad_norm": 2.0859895671576747,
+      "learning_rate": 4.748015188125647e-05,
+      "loss": 1.5336,
+      "step": 468
+    },
+    {
+      "epoch": 0.43709226467847156,
+      "grad_norm": 2.456626686757949,
+      "learning_rate": 4.746289264756645e-05,
+      "loss": 1.7939,
+      "step": 469
+    },
+    {
+      "epoch": 0.4380242311276794,
+      "grad_norm": 2.126978753101913,
+      "learning_rate": 4.744563341387643e-05,
+      "loss": 1.6253,
+      "step": 470
+    },
+    {
+      "epoch": 0.43895619757688725,
+      "grad_norm": 2.3107007888541857,
+      "learning_rate": 4.74283741801864e-05,
+      "loss": 1.8334,
+      "step": 471
+    },
+    {
+      "epoch": 0.43988816402609504,
+      "grad_norm": 1.8329283018005555,
+      "learning_rate": 4.7411114946496376e-05,
+      "loss": 1.4513,
+      "step": 472
+    },
+    {
+      "epoch": 0.4408201304753029,
+      "grad_norm": 2.6608911361043774,
+      "learning_rate": 4.739385571280635e-05,
+      "loss": 1.963,
+      "step": 473
+    },
+    {
+      "epoch": 0.44175209692451073,
+      "grad_norm": 2.2247893706340602,
+      "learning_rate": 4.737659647911633e-05,
+      "loss": 1.7821,
+      "step": 474
+    },
+    {
+      "epoch": 0.4426840633737186,
+      "grad_norm": 2.176788988872512,
+      "learning_rate": 4.7359337245426306e-05,
+      "loss": 1.5456,
+      "step": 475
+    },
+    {
+      "epoch": 0.44361602982292636,
+      "grad_norm": 2.3650281777864706,
+      "learning_rate": 4.734207801173628e-05,
+      "loss": 2.0436,
+      "step": 476
+    },
+    {
+      "epoch": 0.4445479962721342,
+      "grad_norm": 2.288813427098428,
+      "learning_rate": 4.732481877804626e-05,
+      "loss": 1.6951,
+      "step": 477
+    },
+    {
+      "epoch": 0.44547996272134205,
+      "grad_norm": 2.1499637548972093,
+      "learning_rate": 4.7307559544356236e-05,
+      "loss": 1.5602,
+      "step": 478
+    },
+    {
+      "epoch": 0.44641192917054984,
+      "grad_norm": 2.227647930702237,
+      "learning_rate": 4.729030031066621e-05,
+      "loss": 1.7576,
+      "step": 479
+    },
+    {
+      "epoch": 0.4473438956197577,
+      "grad_norm": 2.399193690838884,
+      "learning_rate": 4.7273041076976184e-05,
+      "loss": 1.78,
+      "step": 480
+    },
+    {
+      "epoch": 0.4482758620689655,
+      "grad_norm": 1.6902784723078,
+      "learning_rate": 4.7255781843286165e-05,
+      "loss": 1.5116,
+      "step": 481
+    },
+    {
+      "epoch": 0.44920782851817337,
+      "grad_norm": 2.0255022295109453,
+      "learning_rate": 4.723852260959614e-05,
+      "loss": 1.4941,
+      "step": 482
+    },
+    {
+      "epoch": 0.45013979496738116,
+      "grad_norm": 2.7228201769861142,
+      "learning_rate": 4.7221263375906114e-05,
+      "loss": 1.6989,
+      "step": 483
+    },
+    {
+      "epoch": 0.451071761416589,
+      "grad_norm": 2.1539721213538385,
+      "learning_rate": 4.720400414221609e-05,
+      "loss": 1.4276,
+      "step": 484
+    },
+    {
+      "epoch": 0.45200372786579684,
+      "grad_norm": 1.685852493389208,
+      "learning_rate": 4.718674490852606e-05,
+      "loss": 1.5403,
+      "step": 485
+    },
+    {
+      "epoch": 0.45293569431500463,
+      "grad_norm": 1.940014178027911,
+      "learning_rate": 4.716948567483604e-05,
+      "loss": 1.5512,
+      "step": 486
+    },
+    {
+      "epoch": 0.4538676607642125,
+      "grad_norm": 2.252489482098647,
+      "learning_rate": 4.715222644114601e-05,
+      "loss": 1.5864,
+      "step": 487
+    },
+    {
+      "epoch": 0.4547996272134203,
+      "grad_norm": 2.7965217629890593,
+      "learning_rate": 4.713496720745599e-05,
+      "loss": 1.9796,
+      "step": 488
+    },
+    {
+      "epoch": 0.45573159366262816,
+      "grad_norm": 2.435749566796755,
+      "learning_rate": 4.7117707973765966e-05,
+      "loss": 2.0731,
+      "step": 489
+    },
+    {
+      "epoch": 0.45666356011183595,
+      "grad_norm": 1.7135401056445434,
+      "learning_rate": 4.710044874007594e-05,
+      "loss": 1.3982,
+      "step": 490
+    },
+    {
+      "epoch": 0.4575955265610438,
+      "grad_norm": 2.5426512811194564,
+      "learning_rate": 4.7083189506385915e-05,
+      "loss": 1.7515,
+      "step": 491
+    },
+    {
+      "epoch": 0.45852749301025164,
+      "grad_norm": 2.4066052604875994,
+      "learning_rate": 4.7065930272695896e-05,
+      "loss": 1.6365,
+      "step": 492
+    },
+    {
+      "epoch": 0.4594594594594595,
+      "grad_norm": 2.339412362619582,
+      "learning_rate": 4.704867103900587e-05,
+      "loss": 1.8434,
+      "step": 493
+    },
+    {
+      "epoch": 0.4603914259086673,
+      "grad_norm": 1.7676662446343308,
+      "learning_rate": 4.7031411805315845e-05,
+      "loss": 1.4994,
+      "step": 494
+    },
+    {
+      "epoch": 0.4613233923578751,
+      "grad_norm": 2.060161387752443,
+      "learning_rate": 4.7014152571625826e-05,
+      "loss": 1.5133,
+      "step": 495
+    },
+    {
+      "epoch": 0.46225535880708296,
+      "grad_norm": 2.1218439623159777,
+      "learning_rate": 4.69968933379358e-05,
+      "loss": 1.4787,
+      "step": 496
+    },
+    {
+      "epoch": 0.46318732525629075,
+      "grad_norm": 2.046029297384043,
+      "learning_rate": 4.6979634104245774e-05,
+      "loss": 1.4449,
+      "step": 497
+    },
+    {
+      "epoch": 0.4641192917054986,
+      "grad_norm": 2.202812263089534,
+      "learning_rate": 4.696237487055575e-05,
+      "loss": 1.4808,
+      "step": 498
+    },
+    {
+      "epoch": 0.46505125815470644,
+      "grad_norm": 2.213090070386427,
+      "learning_rate": 4.694511563686573e-05,
+      "loss": 1.5563,
+      "step": 499
+    },
+    {
+      "epoch": 0.4659832246039143,
+      "grad_norm": 2.2642902521906922,
+      "learning_rate": 4.6927856403175704e-05,
+      "loss": 1.6896,
+      "step": 500
+    },
+    {
+      "epoch": 0.46691519105312207,
+      "grad_norm": 2.363627572935832,
+      "learning_rate": 4.691059716948568e-05,
+      "loss": 1.6147,
+      "step": 501
+    },
+    {
+      "epoch": 0.4678471575023299,
+      "grad_norm": 2.594004353178262,
+      "learning_rate": 4.689333793579565e-05,
+      "loss": 1.8595,
+      "step": 502
+    },
+    {
+      "epoch": 0.46877912395153776,
+      "grad_norm": 2.480120329993103,
+      "learning_rate": 4.6876078702105634e-05,
+      "loss": 1.2406,
+      "step": 503
+    },
+    {
+      "epoch": 0.46971109040074555,
+      "grad_norm": 2.308530343085864,
+      "learning_rate": 4.685881946841561e-05,
+      "loss": 1.7078,
+      "step": 504
+    },
+    {
+      "epoch": 0.4706430568499534,
+      "grad_norm": 2.572193097908897,
+      "learning_rate": 4.684156023472558e-05,
+      "loss": 1.6665,
+      "step": 505
+    },
+    {
+      "epoch": 0.47157502329916123,
+      "grad_norm": 2.203640328851444,
+      "learning_rate": 4.6824301001035556e-05,
+      "loss": 1.5246,
+      "step": 506
+    },
+    {
+      "epoch": 0.4725069897483691,
+      "grad_norm": 2.069581087209235,
+      "learning_rate": 4.680704176734553e-05,
+      "loss": 1.5596,
+      "step": 507
+    },
+    {
+      "epoch": 0.47343895619757687,
+      "grad_norm": 1.9325602250636038,
+      "learning_rate": 4.6789782533655505e-05,
+      "loss": 1.5719,
+      "step": 508
+    },
+    {
+      "epoch": 0.4743709226467847,
+      "grad_norm": 2.0045206135055733,
+      "learning_rate": 4.677252329996548e-05,
+      "loss": 1.5289,
+      "step": 509
+    },
+    {
+      "epoch": 0.47530288909599255,
+      "grad_norm": 2.090816088376809,
+      "learning_rate": 4.675526406627546e-05,
+      "loss": 1.4785,
+      "step": 510
+    },
+    {
+      "epoch": 0.4762348555452004,
+      "grad_norm": 4.243535422032179,
+      "learning_rate": 4.6738004832585435e-05,
+      "loss": 1.7547,
+      "step": 511
+    },
+    {
+      "epoch": 0.4771668219944082,
+      "grad_norm": 2.6279180489842395,
+      "learning_rate": 4.672074559889541e-05,
+      "loss": 1.5638,
+      "step": 512
+    },
+    {
+      "epoch": 0.47809878844361603,
+      "grad_norm": 2.0259135108299087,
+      "learning_rate": 4.670348636520538e-05,
+      "loss": 1.4976,
+      "step": 513
+    },
+    {
+      "epoch": 0.4790307548928239,
+      "grad_norm": 1.8157613057786348,
+      "learning_rate": 4.6686227131515364e-05,
+      "loss": 1.4325,
+      "step": 514
+    },
+    {
+      "epoch": 0.47996272134203166,
+      "grad_norm": 1.8038558366555435,
+      "learning_rate": 4.666896789782534e-05,
+      "loss": 1.397,
+      "step": 515
+    },
+    {
+      "epoch": 0.4808946877912395,
+      "grad_norm": 1.788224779125571,
+      "learning_rate": 4.665170866413531e-05,
+      "loss": 1.4675,
+      "step": 516
+    },
+    {
+      "epoch": 0.48182665424044735,
+      "grad_norm": 1.7632704011196598,
+      "learning_rate": 4.6634449430445294e-05,
+      "loss": 1.4134,
+      "step": 517
+    },
+    {
+      "epoch": 0.4827586206896552,
+      "grad_norm": 1.8383862021651716,
+      "learning_rate": 4.661719019675527e-05,
+      "loss": 1.5264,
+      "step": 518
+    },
+    {
+      "epoch": 0.483690587138863,
+      "grad_norm": 2.3700987104034734,
+      "learning_rate": 4.659993096306524e-05,
+      "loss": 1.5728,
+      "step": 519
+    },
+    {
+      "epoch": 0.4846225535880708,
+      "grad_norm": 1.7911004416543497,
+      "learning_rate": 4.658267172937522e-05,
+      "loss": 1.4909,
+      "step": 520
+    },
+    {
+      "epoch": 0.48555452003727867,
+      "grad_norm": 1.900552746161005,
+      "learning_rate": 4.65654124956852e-05,
+      "loss": 1.3838,
+      "step": 521
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 2.5157301921906567,
+      "learning_rate": 4.654815326199517e-05,
+      "loss": 1.9063,
+      "step": 522
+    },
+    {
+      "epoch": 0.4874184529356943,
+      "grad_norm": 1.9737243407379443,
+      "learning_rate": 4.6530894028305146e-05,
+      "loss": 1.6145,
+      "step": 523
+    },
+    {
+      "epoch": 0.48835041938490215,
+      "grad_norm": 1.8881693616901634,
+      "learning_rate": 4.651363479461512e-05,
+      "loss": 1.576,
+      "step": 524
+    },
+    {
+      "epoch": 0.48928238583411,
+      "grad_norm": 2.443436294025698,
+      "learning_rate": 4.64963755609251e-05,
+      "loss": 1.5882,
+      "step": 525
+    },
+    {
+      "epoch": 0.4902143522833178,
+      "grad_norm": 2.493326953846239,
+      "learning_rate": 4.6479116327235076e-05,
+      "loss": 1.4876,
+      "step": 526
+    },
+    {
+      "epoch": 0.4911463187325256,
+      "grad_norm": 2.6237623065614453,
+      "learning_rate": 4.646185709354505e-05,
+      "loss": 1.4975,
+      "step": 527
+    },
+    {
+      "epoch": 0.49207828518173347,
+      "grad_norm": 2.030649300423701,
+      "learning_rate": 4.6444597859855025e-05,
+      "loss": 1.5145,
+      "step": 528
+    },
+    {
+      "epoch": 0.4930102516309413,
+      "grad_norm": 2.186702205039668,
+      "learning_rate": 4.6427338626165e-05,
+      "loss": 1.4249,
+      "step": 529
+    },
+    {
+      "epoch": 0.4939422180801491,
+      "grad_norm": 2.1786179030616077,
+      "learning_rate": 4.641007939247497e-05,
+      "loss": 1.7072,
+      "step": 530
+    },
+    {
+      "epoch": 0.49487418452935694,
+      "grad_norm": 2.3893347221006422,
+      "learning_rate": 4.639282015878495e-05,
+      "loss": 1.5719,
+      "step": 531
+    },
+    {
+      "epoch": 0.4958061509785648,
+      "grad_norm": 2.463158910456155,
+      "learning_rate": 4.637556092509493e-05,
+      "loss": 1.4317,
+      "step": 532
+    },
+    {
+      "epoch": 0.4967381174277726,
+      "grad_norm": 2.0428004791206615,
+      "learning_rate": 4.63583016914049e-05,
+      "loss": 1.5757,
+      "step": 533
+    },
+    {
+      "epoch": 0.4976700838769804,
+      "grad_norm": 1.7308445541184365,
+      "learning_rate": 4.634104245771488e-05,
+      "loss": 1.4441,
+      "step": 534
+    },
+    {
+      "epoch": 0.49860205032618826,
+      "grad_norm": 2.008124659456434,
+      "learning_rate": 4.632378322402486e-05,
+      "loss": 1.5902,
+      "step": 535
+    },
+    {
+      "epoch": 0.4995340167753961,
+      "grad_norm": 1.9244931854611158,
+      "learning_rate": 4.630652399033483e-05,
+      "loss": 1.5666,
+      "step": 536
+    },
+    {
+      "epoch": 0.5004659832246039,
+      "grad_norm": 2.00904767439663,
+      "learning_rate": 4.628926475664481e-05,
+      "loss": 1.7267,
+      "step": 537
+    },
+    {
+      "epoch": 0.5013979496738118,
+      "grad_norm": 1.893362747595434,
+      "learning_rate": 4.627200552295478e-05,
+      "loss": 1.4848,
+      "step": 538
+    },
+    {
+      "epoch": 0.5023299161230196,
+      "grad_norm": 1.6572557226706044,
+      "learning_rate": 4.625474628926476e-05,
+      "loss": 1.4003,
+      "step": 539
+    },
+    {
+      "epoch": 0.5032618825722274,
+      "grad_norm": 2.1294521575258227,
+      "learning_rate": 4.6237487055574736e-05,
+      "loss": 1.62,
+      "step": 540
+    },
+    {
+      "epoch": 0.5041938490214353,
+      "grad_norm": 2.1053494957939574,
+      "learning_rate": 4.622022782188471e-05,
+      "loss": 1.5438,
+      "step": 541
+    },
+    {
+      "epoch": 0.5051258154706431,
+      "grad_norm": 2.0635300379107306,
+      "learning_rate": 4.6202968588194685e-05,
+      "loss": 1.3931,
+      "step": 542
+    },
+    {
+      "epoch": 0.5060577819198508,
+      "grad_norm": 2.2376886369203253,
+      "learning_rate": 4.6185709354504666e-05,
+      "loss": 1.5011,
+      "step": 543
+    },
+    {
+      "epoch": 0.5069897483690587,
+      "grad_norm": 2.588422566201751,
+      "learning_rate": 4.616845012081464e-05,
+      "loss": 1.8323,
+      "step": 544
+    },
+    {
+      "epoch": 0.5079217148182665,
+      "grad_norm": 1.908431908508238,
+      "learning_rate": 4.6151190887124615e-05,
+      "loss": 1.3687,
+      "step": 545
+    },
+    {
+      "epoch": 0.5088536812674743,
+      "grad_norm": 2.0991775553453547,
+      "learning_rate": 4.613393165343459e-05,
+      "loss": 1.6897,
+      "step": 546
+    },
+    {
+      "epoch": 0.5097856477166822,
+      "grad_norm": 2.0358750755624695,
+      "learning_rate": 4.611667241974456e-05,
+      "loss": 1.4534,
+      "step": 547
+    },
+    {
+      "epoch": 0.51071761416589,
+      "grad_norm": 2.1769958012684927,
+      "learning_rate": 4.609941318605454e-05,
+      "loss": 1.3591,
+      "step": 548
+    },
+    {
+      "epoch": 0.5116495806150979,
+      "grad_norm": 2.0428662772781125,
+      "learning_rate": 4.608215395236451e-05,
+      "loss": 1.4024,
+      "step": 549
+    },
+    {
+      "epoch": 0.5125815470643057,
+      "grad_norm": 2.0952462839089847,
+      "learning_rate": 4.606489471867449e-05,
+      "loss": 1.601,
+      "step": 550
+    },
+    {
+      "epoch": 0.5135135135135135,
+      "grad_norm": 2.5201107128060514,
+      "learning_rate": 4.604763548498447e-05,
+      "loss": 1.7692,
+      "step": 551
+    },
+    {
+      "epoch": 0.5144454799627214,
+      "grad_norm": 1.886191876525905,
+      "learning_rate": 4.603037625129444e-05,
+      "loss": 1.4185,
+      "step": 552
+    },
+    {
+      "epoch": 0.5153774464119292,
+      "grad_norm": 2.4885193632757816,
+      "learning_rate": 4.6013117017604416e-05,
+      "loss": 1.7483,
+      "step": 553
+    },
+    {
+      "epoch": 0.516309412861137,
+      "grad_norm": 2.1675288260809555,
+      "learning_rate": 4.59958577839144e-05,
+      "loss": 1.6109,
+      "step": 554
+    },
+    {
+      "epoch": 0.5172413793103449,
+      "grad_norm": 2.0135093023580866,
+      "learning_rate": 4.597859855022437e-05,
+      "loss": 1.489,
+      "step": 555
+    },
+    {
+      "epoch": 0.5181733457595527,
+      "grad_norm": 2.0633338339463587,
+      "learning_rate": 4.5961339316534345e-05,
+      "loss": 1.6268,
+      "step": 556
+    },
+    {
+      "epoch": 0.5191053122087604,
+      "grad_norm": 1.7972401619859044,
+      "learning_rate": 4.5944080082844326e-05,
+      "loss": 1.4562,
+      "step": 557
+    },
+    {
+      "epoch": 0.5200372786579683,
+      "grad_norm": 2.3110501187445793,
+      "learning_rate": 4.59268208491543e-05,
+      "loss": 1.6862,
+      "step": 558
+    },
+    {
+      "epoch": 0.5209692451071761,
+      "grad_norm": 2.300441828563265,
+      "learning_rate": 4.5909561615464275e-05,
+      "loss": 1.77,
+      "step": 559
+    },
+    {
+      "epoch": 0.5219012115563839,
+      "grad_norm": 2.721099890008729,
+      "learning_rate": 4.589230238177425e-05,
+      "loss": 1.5825,
+      "step": 560
+    },
+    {
+      "epoch": 0.5228331780055918,
+      "grad_norm": 2.1982120984825357,
+      "learning_rate": 4.587504314808423e-05,
+      "loss": 1.4928,
+      "step": 561
+    },
+    {
+      "epoch": 0.5237651444547996,
+      "grad_norm": 2.1961946799736576,
+      "learning_rate": 4.5857783914394205e-05,
+      "loss": 1.6534,
+      "step": 562
+    },
+    {
+      "epoch": 0.5246971109040075,
+      "grad_norm": 1.9298443120380138,
+      "learning_rate": 4.584052468070418e-05,
+      "loss": 1.2161,
+      "step": 563
+    },
+    {
+      "epoch": 0.5256290773532153,
+      "grad_norm": 2.28810981409144,
+      "learning_rate": 4.582326544701415e-05,
+      "loss": 1.8269,
+      "step": 564
+    },
+    {
+      "epoch": 0.5265610438024231,
+      "grad_norm": 2.6280382710009644,
+      "learning_rate": 4.5806006213324134e-05,
+      "loss": 1.588,
+      "step": 565
+    },
+    {
+      "epoch": 0.527493010251631,
+      "grad_norm": 1.9481716964057716,
+      "learning_rate": 4.578874697963411e-05,
+      "loss": 1.483,
+      "step": 566
+    },
+    {
+      "epoch": 0.5284249767008388,
+      "grad_norm": 2.29300855416401,
+      "learning_rate": 4.577148774594408e-05,
+      "loss": 1.6411,
+      "step": 567
+    },
+    {
+      "epoch": 0.5293569431500466,
+      "grad_norm": 1.899519828235566,
+      "learning_rate": 4.575422851225406e-05,
+      "loss": 1.4391,
+      "step": 568
+    },
+    {
+      "epoch": 0.5302889095992545,
+      "grad_norm": 2.2707582741399754,
+      "learning_rate": 4.573696927856403e-05,
+      "loss": 1.7035,
+      "step": 569
+    },
+    {
+      "epoch": 0.5312208760484622,
+      "grad_norm": 2.3796578958878696,
+      "learning_rate": 4.5719710044874006e-05,
+      "loss": 1.4181,
+      "step": 570
+    },
+    {
+      "epoch": 0.53215284249767,
+      "grad_norm": 2.3783380918282866,
+      "learning_rate": 4.570245081118398e-05,
+      "loss": 1.5349,
+      "step": 571
+    },
+    {
+      "epoch": 0.5330848089468779,
+      "grad_norm": 2.2719123775694947,
+      "learning_rate": 4.568519157749396e-05,
+      "loss": 1.1881,
+      "step": 572
+    },
+    {
+      "epoch": 0.5340167753960857,
+      "grad_norm": 2.4670703290899545,
+      "learning_rate": 4.5667932343803935e-05,
+      "loss": 1.6008,
+      "step": 573
+    },
+    {
+      "epoch": 0.5349487418452936,
+      "grad_norm": 1.6054006705840036,
+      "learning_rate": 4.565067311011391e-05,
+      "loss": 1.4972,
+      "step": 574
+    },
+    {
+      "epoch": 0.5358807082945014,
+      "grad_norm": 1.6779853440982424,
+      "learning_rate": 4.563341387642389e-05,
+      "loss": 1.4041,
+      "step": 575
+    },
+    {
+      "epoch": 0.5368126747437092,
+      "grad_norm": 1.8027601379025011,
+      "learning_rate": 4.5616154642733865e-05,
+      "loss": 1.3077,
+      "step": 576
+    },
+    {
+      "epoch": 0.5377446411929171,
+      "grad_norm": 2.977914314103489,
+      "learning_rate": 4.559889540904384e-05,
+      "loss": 1.7215,
+      "step": 577
+    },
+    {
+      "epoch": 0.5386766076421249,
+      "grad_norm": 2.0671102588688046,
+      "learning_rate": 4.5581636175353814e-05,
+      "loss": 1.4861,
+      "step": 578
+    },
+    {
+      "epoch": 0.5396085740913327,
+      "grad_norm": 2.0124636279190873,
+      "learning_rate": 4.5564376941663795e-05,
+      "loss": 1.4658,
+      "step": 579
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 2.214728305228285,
+      "learning_rate": 4.554711770797377e-05,
+      "loss": 1.6103,
+      "step": 580
+    },
+    {
+      "epoch": 0.5414725069897484,
+      "grad_norm": 2.144792497296664,
+      "learning_rate": 4.552985847428374e-05,
+      "loss": 1.4633,
+      "step": 581
+    },
+    {
+      "epoch": 0.5424044734389561,
+      "grad_norm": 1.941616208042572,
+      "learning_rate": 4.551259924059372e-05,
+      "loss": 1.5104,
+      "step": 582
+    },
+    {
+      "epoch": 0.543336439888164,
+      "grad_norm": 2.5351732073916486,
+      "learning_rate": 4.54953400069037e-05,
+      "loss": 1.7182,
+      "step": 583
+    },
+    {
+      "epoch": 0.5442684063373718,
+      "grad_norm": 1.978747430572336,
+      "learning_rate": 4.547808077321367e-05,
+      "loss": 1.4487,
+      "step": 584
+    },
+    {
+      "epoch": 0.5452003727865797,
+      "grad_norm": 1.951396184706673,
+      "learning_rate": 4.546082153952365e-05,
+      "loss": 1.2185,
+      "step": 585
+    },
+    {
+      "epoch": 0.5461323392357875,
+      "grad_norm": 1.799851446871657,
+      "learning_rate": 4.544356230583363e-05,
+      "loss": 1.2074,
+      "step": 586
+    },
+    {
+      "epoch": 0.5470643056849953,
+      "grad_norm": 2.6568143361368697,
+      "learning_rate": 4.54263030721436e-05,
+      "loss": 1.6497,
+      "step": 587
+    },
+    {
+      "epoch": 0.5479962721342032,
+      "grad_norm": 2.5117642135881733,
+      "learning_rate": 4.540904383845358e-05,
+      "loss": 1.6714,
+      "step": 588
+    },
+    {
+      "epoch": 0.548928238583411,
+      "grad_norm": 1.68738159411162,
+      "learning_rate": 4.539178460476355e-05,
+      "loss": 1.3291,
+      "step": 589
+    },
+    {
+      "epoch": 0.5498602050326188,
+      "grad_norm": 1.7694583766766416,
+      "learning_rate": 4.5374525371073526e-05,
+      "loss": 1.2529,
+      "step": 590
+    },
+    {
+      "epoch": 0.5507921714818267,
+      "grad_norm": 2.4583683209980243,
+      "learning_rate": 4.53572661373835e-05,
+      "loss": 1.6661,
+      "step": 591
+    },
+    {
+      "epoch": 0.5517241379310345,
+      "grad_norm": 2.3659501594521326,
+      "learning_rate": 4.5340006903693474e-05,
+      "loss": 1.7649,
+      "step": 592
+    },
+    {
+      "epoch": 0.5526561043802423,
+      "grad_norm": 1.9126405540763456,
+      "learning_rate": 4.532274767000345e-05,
+      "loss": 1.4651,
+      "step": 593
+    },
+    {
+      "epoch": 0.5535880708294502,
+      "grad_norm": 1.9886585647878114,
+      "learning_rate": 4.530548843631343e-05,
+      "loss": 1.3025,
+      "step": 594
+    },
+    {
+      "epoch": 0.554520037278658,
+      "grad_norm": 2.0928483225367422,
+      "learning_rate": 4.5288229202623404e-05,
+      "loss": 1.4536,
+      "step": 595
+    },
+    {
+      "epoch": 0.5554520037278659,
+      "grad_norm": 2.413540546333916,
+      "learning_rate": 4.527096996893338e-05,
+      "loss": 1.5634,
+      "step": 596
+    },
+    {
+      "epoch": 0.5563839701770736,
+      "grad_norm": 2.087084201171065,
+      "learning_rate": 4.525371073524336e-05,
+      "loss": 1.5778,
+      "step": 597
+    },
+    {
+      "epoch": 0.5573159366262814,
+      "grad_norm": 2.3397773090223883,
+      "learning_rate": 4.523645150155333e-05,
+      "loss": 1.7813,
+      "step": 598
+    },
+    {
+      "epoch": 0.5582479030754893,
+      "grad_norm": 2.4547573881528604,
+      "learning_rate": 4.521919226786331e-05,
+      "loss": 1.6399,
+      "step": 599
+    },
+    {
+      "epoch": 0.5591798695246971,
+      "grad_norm": 1.854982124761483,
+      "learning_rate": 4.520193303417328e-05,
+      "loss": 1.4316,
+      "step": 600
+    },
+    {
+      "epoch": 0.5601118359739049,
+      "grad_norm": 2.001104732186021,
+      "learning_rate": 4.518467380048326e-05,
+      "loss": 1.5854,
+      "step": 601
+    },
+    {
+      "epoch": 0.5610438024231128,
+      "grad_norm": 1.7560929299757573,
+      "learning_rate": 4.516741456679324e-05,
+      "loss": 1.3943,
+      "step": 602
+    },
+    {
+      "epoch": 0.5619757688723206,
+      "grad_norm": 2.4056534036712085,
+      "learning_rate": 4.515015533310321e-05,
+      "loss": 1.7953,
+      "step": 603
+    },
+    {
+      "epoch": 0.5629077353215284,
+      "grad_norm": 1.9602174918280861,
+      "learning_rate": 4.5132896099413186e-05,
+      "loss": 1.5476,
+      "step": 604
+    },
+    {
+      "epoch": 0.5638397017707363,
+      "grad_norm": 1.985544061437955,
+      "learning_rate": 4.511563686572317e-05,
+      "loss": 1.5605,
+      "step": 605
+    },
+    {
+      "epoch": 0.5647716682199441,
+      "grad_norm": 2.470699994318334,
+      "learning_rate": 4.509837763203314e-05,
+      "loss": 1.7184,
+      "step": 606
+    },
+    {
+      "epoch": 0.5657036346691519,
+      "grad_norm": 2.2441123851055544,
+      "learning_rate": 4.5081118398343116e-05,
+      "loss": 1.4424,
+      "step": 607
+    },
+    {
+      "epoch": 0.5666356011183598,
+      "grad_norm": 2.085960360729081,
+      "learning_rate": 4.5063859164653097e-05,
+      "loss": 1.6885,
+      "step": 608
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 2.3195026603636837,
+      "learning_rate": 4.504659993096307e-05,
+      "loss": 1.6056,
+      "step": 609
+    },
+    {
+      "epoch": 0.5684995340167754,
+      "grad_norm": 1.7650612684880733,
+      "learning_rate": 4.5029340697273045e-05,
+      "loss": 1.2646,
+      "step": 610
+    },
+    {
+      "epoch": 0.5694315004659832,
+      "grad_norm": 1.8588577595621563,
+      "learning_rate": 4.501208146358302e-05,
+      "loss": 1.5862,
+      "step": 611
+    },
+    {
+      "epoch": 0.570363466915191,
+      "grad_norm": 2.5820180112729183,
+      "learning_rate": 4.4994822229892994e-05,
+      "loss": 1.7587,
+      "step": 612
+    },
+    {
+      "epoch": 0.5712954333643989,
+      "grad_norm": 2.3348056648682327,
+      "learning_rate": 4.497756299620297e-05,
+      "loss": 1.6532,
+      "step": 613
+    },
+    {
+      "epoch": 0.5722273998136067,
+      "grad_norm": 1.991459384733236,
+      "learning_rate": 4.496030376251294e-05,
+      "loss": 1.5479,
+      "step": 614
+    },
+    {
+      "epoch": 0.5731593662628145,
+      "grad_norm": 2.09197995385507,
+      "learning_rate": 4.4943044528822923e-05,
+      "loss": 1.8409,
+      "step": 615
+    },
+    {
+      "epoch": 0.5740913327120224,
+      "grad_norm": 2.034214742834668,
+      "learning_rate": 4.49257852951329e-05,
+      "loss": 1.531,
+      "step": 616
+    },
+    {
+      "epoch": 0.5750232991612302,
+      "grad_norm": 2.4329874508576412,
+      "learning_rate": 4.490852606144287e-05,
+      "loss": 1.5023,
+      "step": 617
+    },
+    {
+      "epoch": 0.575955265610438,
+      "grad_norm": 2.3400308544701316,
+      "learning_rate": 4.4891266827752846e-05,
+      "loss": 1.5878,
+      "step": 618
+    },
+    {
+      "epoch": 0.5768872320596459,
+      "grad_norm": 1.87200810834496,
+      "learning_rate": 4.487400759406283e-05,
+      "loss": 1.4448,
+      "step": 619
+    },
+    {
+      "epoch": 0.5778191985088537,
+      "grad_norm": 1.9159508691055998,
+      "learning_rate": 4.48567483603728e-05,
+      "loss": 1.3665,
+      "step": 620
+    },
+    {
+      "epoch": 0.5787511649580616,
+      "grad_norm": 1.954410176779966,
+      "learning_rate": 4.4839489126682776e-05,
+      "loss": 1.5404,
+      "step": 621
+    },
+    {
+      "epoch": 0.5796831314072693,
+      "grad_norm": 3.023195638337082,
+      "learning_rate": 4.482222989299275e-05,
+      "loss": 1.7185,
+      "step": 622
+    },
+    {
+      "epoch": 0.5806150978564771,
+      "grad_norm": 2.296024820155765,
+      "learning_rate": 4.480497065930273e-05,
+      "loss": 1.6162,
+      "step": 623
+    },
+    {
+      "epoch": 0.581547064305685,
+      "grad_norm": 1.9675511919734292,
+      "learning_rate": 4.4787711425612706e-05,
+      "loss": 1.3908,
+      "step": 624
+    },
+    {
+      "epoch": 0.5824790307548928,
+      "grad_norm": 2.056616193197246,
+      "learning_rate": 4.477045219192268e-05,
+      "loss": 1.5248,
+      "step": 625
+    },
+    {
+      "epoch": 0.5834109972041006,
+      "grad_norm": 1.537865409243945,
+      "learning_rate": 4.475319295823266e-05,
+      "loss": 1.3415,
+      "step": 626
+    },
+    {
+      "epoch": 0.5843429636533085,
+      "grad_norm": 2.1398283502243416,
+      "learning_rate": 4.4735933724542635e-05,
+      "loss": 1.6989,
+      "step": 627
+    },
+    {
+      "epoch": 0.5852749301025163,
+      "grad_norm": 1.890605699915211,
+      "learning_rate": 4.471867449085261e-05,
+      "loss": 1.456,
+      "step": 628
+    },
+    {
+      "epoch": 0.5862068965517241,
+      "grad_norm": 2.1828666945071036,
+      "learning_rate": 4.4701415257162584e-05,
+      "loss": 1.4976,
+      "step": 629
+    },
+    {
+      "epoch": 0.587138863000932,
+      "grad_norm": 2.169386577707007,
+      "learning_rate": 4.4684156023472565e-05,
+      "loss": 1.5314,
+      "step": 630
+    },
+    {
+      "epoch": 0.5880708294501398,
+      "grad_norm": 1.9426318132286957,
+      "learning_rate": 4.466689678978254e-05,
+      "loss": 1.4573,
+      "step": 631
+    },
+    {
+      "epoch": 0.5890027958993477,
+      "grad_norm": 1.6532573142489262,
+      "learning_rate": 4.4649637556092513e-05,
+      "loss": 1.2743,
+      "step": 632
+    },
+    {
+      "epoch": 0.5899347623485555,
+      "grad_norm": 1.600661978729288,
+      "learning_rate": 4.463237832240249e-05,
+      "loss": 1.1262,
+      "step": 633
+    },
+    {
+      "epoch": 0.5908667287977633,
+      "grad_norm": 2.020485077049736,
+      "learning_rate": 4.461511908871246e-05,
+      "loss": 1.3705,
+      "step": 634
+    },
+    {
+      "epoch": 0.5917986952469712,
+      "grad_norm": 1.5112208599014145,
+      "learning_rate": 4.4597859855022436e-05,
+      "loss": 1.1962,
+      "step": 635
+    },
+    {
+      "epoch": 0.5927306616961789,
+      "grad_norm": 2.502375433054268,
+      "learning_rate": 4.458060062133241e-05,
+      "loss": 1.6832,
+      "step": 636
+    },
+    {
+      "epoch": 0.5936626281453867,
+      "grad_norm": 2.2243476076283777,
+      "learning_rate": 4.456334138764239e-05,
+      "loss": 1.2587,
+      "step": 637
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 1.6818092752096212,
+      "learning_rate": 4.4546082153952366e-05,
+      "loss": 1.2492,
+      "step": 638
+    },
+    {
+      "epoch": 0.5955265610438024,
+      "grad_norm": 1.8840448101706606,
+      "learning_rate": 4.452882292026234e-05,
+      "loss": 1.2427,
+      "step": 639
+    },
+    {
+      "epoch": 0.5964585274930102,
+      "grad_norm": 3.1208691992721467,
+      "learning_rate": 4.4511563686572315e-05,
+      "loss": 1.5553,
+      "step": 640
+    },
+    {
+      "epoch": 0.5973904939422181,
+      "grad_norm": 1.8144569488818487,
+      "learning_rate": 4.4494304452882296e-05,
+      "loss": 1.2617,
+      "step": 641
+    },
+    {
+      "epoch": 0.5983224603914259,
+      "grad_norm": 1.966514046361118,
+      "learning_rate": 4.447704521919227e-05,
+      "loss": 1.4531,
+      "step": 642
+    },
+    {
+      "epoch": 0.5992544268406338,
+      "grad_norm": 1.8981843443887565,
+      "learning_rate": 4.4459785985502244e-05,
+      "loss": 1.3537,
+      "step": 643
+    },
+    {
+      "epoch": 0.6001863932898416,
+      "grad_norm": 1.9544411578209862,
+      "learning_rate": 4.444252675181222e-05,
+      "loss": 1.4942,
+      "step": 644
+    },
+    {
+      "epoch": 0.6011183597390494,
+      "grad_norm": 2.3918985303112885,
+      "learning_rate": 4.44252675181222e-05,
+      "loss": 1.4954,
+      "step": 645
+    },
+    {
+      "epoch": 0.6020503261882573,
+      "grad_norm": 2.2633203231314294,
+      "learning_rate": 4.4408008284432174e-05,
+      "loss": 1.5546,
+      "step": 646
+    },
+    {
+      "epoch": 0.6029822926374651,
+      "grad_norm": 2.5467931506305015,
+      "learning_rate": 4.439074905074215e-05,
+      "loss": 1.3353,
+      "step": 647
+    },
+    {
+      "epoch": 0.6039142590866728,
+      "grad_norm": 2.231452505736299,
+      "learning_rate": 4.437348981705213e-05,
+      "loss": 1.6501,
+      "step": 648
+    },
+    {
+      "epoch": 0.6048462255358807,
+      "grad_norm": 2.1555067656059257,
+      "learning_rate": 4.4356230583362103e-05,
+      "loss": 1.5269,
+      "step": 649
+    },
+    {
+      "epoch": 0.6057781919850885,
+      "grad_norm": 1.7223222905969742,
+      "learning_rate": 4.433897134967208e-05,
+      "loss": 1.4544,
+      "step": 650
+    },
+    {
+      "epoch": 0.6067101584342963,
+      "grad_norm": 2.114398963864951,
+      "learning_rate": 4.432171211598205e-05,
+      "loss": 1.3824,
+      "step": 651
+    },
+    {
+      "epoch": 0.6076421248835042,
+      "grad_norm": 1.5481126980559332,
+      "learning_rate": 4.430445288229203e-05,
+      "loss": 1.1607,
+      "step": 652
+    },
+    {
+      "epoch": 0.608574091332712,
+      "grad_norm": 1.7288997809114726,
+      "learning_rate": 4.428719364860201e-05,
+      "loss": 1.2306,
+      "step": 653
+    },
+    {
+      "epoch": 0.6095060577819198,
+      "grad_norm": 2.299125828718039,
+      "learning_rate": 4.426993441491198e-05,
+      "loss": 1.6753,
+      "step": 654
+    },
+    {
+      "epoch": 0.6104380242311277,
+      "grad_norm": 2.154429155007647,
+      "learning_rate": 4.4252675181221956e-05,
+      "loss": 1.557,
+      "step": 655
+    },
+    {
+      "epoch": 0.6113699906803355,
+      "grad_norm": 1.9923367916032084,
+      "learning_rate": 4.423541594753193e-05,
+      "loss": 1.2982,
+      "step": 656
+    },
+    {
+      "epoch": 0.6123019571295434,
+      "grad_norm": 1.9377944869611121,
+      "learning_rate": 4.4218156713841905e-05,
+      "loss": 1.2228,
+      "step": 657
+    },
+    {
+      "epoch": 0.6132339235787512,
+      "grad_norm": 2.0098027979343307,
+      "learning_rate": 4.420089748015188e-05,
+      "loss": 1.4284,
+      "step": 658
+    },
+    {
+      "epoch": 0.614165890027959,
+      "grad_norm": 2.0476560006611346,
+      "learning_rate": 4.418363824646186e-05,
+      "loss": 1.2779,
+      "step": 659
+    },
+    {
+      "epoch": 0.6150978564771669,
+      "grad_norm": 1.7930833758466425,
+      "learning_rate": 4.4166379012771834e-05,
+      "loss": 1.1943,
+      "step": 660
+    },
+    {
+      "epoch": 0.6160298229263746,
+      "grad_norm": 2.173823190651099,
+      "learning_rate": 4.414911977908181e-05,
+      "loss": 1.5932,
+      "step": 661
+    },
+    {
+      "epoch": 0.6169617893755824,
+      "grad_norm": 1.9740851599045526,
+      "learning_rate": 4.413186054539178e-05,
+      "loss": 1.2683,
+      "step": 662
+    },
+    {
+      "epoch": 0.6178937558247903,
+      "grad_norm": 2.037991078320121,
+      "learning_rate": 4.4114601311701764e-05,
+      "loss": 1.8075,
+      "step": 663
+    },
+    {
+      "epoch": 0.6188257222739981,
+      "grad_norm": 1.4917526470867923,
+      "learning_rate": 4.409734207801174e-05,
+      "loss": 1.0681,
+      "step": 664
+    },
+    {
+      "epoch": 0.6197576887232059,
+      "grad_norm": 4.071910897631147,
+      "learning_rate": 4.408008284432171e-05,
+      "loss": 1.6845,
+      "step": 665
+    },
+    {
+      "epoch": 0.6206896551724138,
+      "grad_norm": 2.391030076079946,
+      "learning_rate": 4.4062823610631694e-05,
+      "loss": 1.4076,
+      "step": 666
+    },
+    {
+      "epoch": 0.6216216216216216,
+      "grad_norm": 2.1388205964579443,
+      "learning_rate": 4.404556437694167e-05,
+      "loss": 1.6279,
+      "step": 667
+    },
+    {
+      "epoch": 0.6225535880708295,
+      "grad_norm": 1.8578247616086454,
+      "learning_rate": 4.402830514325164e-05,
+      "loss": 1.4541,
+      "step": 668
+    },
+    {
+      "epoch": 0.6234855545200373,
+      "grad_norm": 2.394376597240572,
+      "learning_rate": 4.4011045909561616e-05,
+      "loss": 1.4691,
+      "step": 669
+    },
+    {
+      "epoch": 0.6244175209692451,
+      "grad_norm": 2.2432393136805318,
+      "learning_rate": 4.39937866758716e-05,
+      "loss": 1.7647,
+      "step": 670
+    },
+    {
+      "epoch": 0.625349487418453,
+      "grad_norm": 2.8862716613770707,
+      "learning_rate": 4.397652744218157e-05,
+      "loss": 1.6645,
+      "step": 671
+    },
+    {
+      "epoch": 0.6262814538676608,
+      "grad_norm": 1.8085386206073675,
+      "learning_rate": 4.3959268208491546e-05,
+      "loss": 1.3769,
+      "step": 672
+    },
+    {
+      "epoch": 0.6272134203168686,
+      "grad_norm": 1.8872692456692937,
+      "learning_rate": 4.394200897480152e-05,
+      "loss": 1.4825,
+      "step": 673
+    },
+    {
+      "epoch": 0.6281453867660765,
+      "grad_norm": 2.347954041631238,
+      "learning_rate": 4.39247497411115e-05,
+      "loss": 1.7025,
+      "step": 674
+    },
+    {
+      "epoch": 0.6290773532152842,
+      "grad_norm": 2.2214395418308195,
+      "learning_rate": 4.3907490507421476e-05,
+      "loss": 1.3095,
+      "step": 675
+    },
+    {
+      "epoch": 0.630009319664492,
+      "grad_norm": 1.785031083760799,
+      "learning_rate": 4.389023127373145e-05,
+      "loss": 1.1761,
+      "step": 676
+    },
+    {
+      "epoch": 0.6309412861136999,
+      "grad_norm": 2.2407492784355134,
+      "learning_rate": 4.3872972040041424e-05,
+      "loss": 1.6408,
+      "step": 677
+    },
+    {
+      "epoch": 0.6318732525629077,
+      "grad_norm": 1.949307498919388,
+      "learning_rate": 4.38557128063514e-05,
+      "loss": 1.2396,
+      "step": 678
+    },
+    {
+      "epoch": 0.6328052190121156,
+      "grad_norm": 1.8740202132358161,
+      "learning_rate": 4.383845357266137e-05,
+      "loss": 1.4448,
+      "step": 679
+    },
+    {
+      "epoch": 0.6337371854613234,
+      "grad_norm": 2.0709993851680784,
+      "learning_rate": 4.382119433897135e-05,
+      "loss": 1.5902,
+      "step": 680
+    },
+    {
+      "epoch": 0.6346691519105312,
+      "grad_norm": 1.784696643205085,
+      "learning_rate": 4.380393510528133e-05,
+      "loss": 1.2055,
+      "step": 681
+    },
+    {
+      "epoch": 0.6356011183597391,
+      "grad_norm": 2.0272352417257617,
+      "learning_rate": 4.37866758715913e-05,
+      "loss": 1.2813,
+      "step": 682
+    },
+    {
+      "epoch": 0.6365330848089469,
+      "grad_norm": 2.3793190709127248,
+      "learning_rate": 4.376941663790128e-05,
+      "loss": 1.4366,
+      "step": 683
+    },
+    {
+      "epoch": 0.6374650512581547,
+      "grad_norm": 2.943162618700922,
+      "learning_rate": 4.375215740421125e-05,
+      "loss": 1.45,
+      "step": 684
+    },
+    {
+      "epoch": 0.6383970177073626,
+      "grad_norm": 2.066132180425921,
+      "learning_rate": 4.373489817052123e-05,
+      "loss": 1.3265,
+      "step": 685
+    },
+    {
+      "epoch": 0.6393289841565704,
+      "grad_norm": 2.831698006761152,
+      "learning_rate": 4.3717638936831206e-05,
+      "loss": 1.6661,
+      "step": 686
+    },
+    {
+      "epoch": 0.6402609506057781,
+      "grad_norm": 2.234386972952958,
+      "learning_rate": 4.370037970314118e-05,
+      "loss": 1.443,
+      "step": 687
+    },
+    {
+      "epoch": 0.641192917054986,
+      "grad_norm": 2.010901225448851,
+      "learning_rate": 4.368312046945116e-05,
+      "loss": 1.2145,
+      "step": 688
+    },
+    {
+      "epoch": 0.6421248835041938,
+      "grad_norm": 2.061949302402805,
+      "learning_rate": 4.3665861235761136e-05,
+      "loss": 1.3985,
+      "step": 689
+    },
+    {
+      "epoch": 0.6430568499534017,
+      "grad_norm": 2.0603849160359045,
+      "learning_rate": 4.364860200207111e-05,
+      "loss": 1.3857,
+      "step": 690
+    },
+    {
+      "epoch": 0.6439888164026095,
+      "grad_norm": 2.4271170679945953,
+      "learning_rate": 4.3631342768381085e-05,
+      "loss": 1.6645,
+      "step": 691
+    },
+    {
+      "epoch": 0.6449207828518173,
+      "grad_norm": 2.171453585549501,
+      "learning_rate": 4.3614083534691066e-05,
+      "loss": 1.5432,
+      "step": 692
+    },
+    {
+      "epoch": 0.6458527493010252,
+      "grad_norm": 1.8783940099237892,
+      "learning_rate": 4.359682430100104e-05,
+      "loss": 1.2328,
+      "step": 693
+    },
+    {
+      "epoch": 0.646784715750233,
+      "grad_norm": 2.4431153517789985,
+      "learning_rate": 4.3579565067311014e-05,
+      "loss": 1.3814,
+      "step": 694
+    },
+    {
+      "epoch": 0.6477166821994408,
+      "grad_norm": 1.9510088548444617,
+      "learning_rate": 4.356230583362099e-05,
+      "loss": 1.3547,
+      "step": 695
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 1.9933714622747107,
+      "learning_rate": 4.354504659993097e-05,
+      "loss": 1.3509,
+      "step": 696
+    },
+    {
+      "epoch": 0.6495806150978565,
+      "grad_norm": 2.0483590116808643,
+      "learning_rate": 4.3527787366240944e-05,
+      "loss": 1.2666,
+      "step": 697
+    },
+    {
+      "epoch": 0.6505125815470643,
+      "grad_norm": 1.853123450658554,
+      "learning_rate": 4.351052813255092e-05,
+      "loss": 1.3375,
+      "step": 698
+    },
+    {
+      "epoch": 0.6514445479962722,
+      "grad_norm": 2.3046556941988183,
+      "learning_rate": 4.349326889886089e-05,
+      "loss": 1.6646,
+      "step": 699
+    },
+    {
+      "epoch": 0.65237651444548,
+      "grad_norm": 1.804127218202496,
+      "learning_rate": 4.347600966517087e-05,
+      "loss": 1.3208,
+      "step": 700
+    },
+    {
+      "epoch": 0.6533084808946877,
+      "grad_norm": 1.8563787675645853,
+      "learning_rate": 4.345875043148084e-05,
+      "loss": 1.2022,
+      "step": 701
+    },
+    {
+      "epoch": 0.6542404473438956,
+      "grad_norm": 2.3197918764720415,
+      "learning_rate": 4.3441491197790815e-05,
+      "loss": 1.4416,
+      "step": 702
+    },
+    {
+      "epoch": 0.6551724137931034,
+      "grad_norm": 1.7865611065592508,
+      "learning_rate": 4.3424231964100796e-05,
+      "loss": 1.4572,
+      "step": 703
+    },
+    {
+      "epoch": 0.6561043802423113,
+      "grad_norm": 2.1207105702098223,
+      "learning_rate": 4.340697273041077e-05,
+      "loss": 1.4923,
+      "step": 704
+    },
+    {
+      "epoch": 0.6570363466915191,
+      "grad_norm": 1.8241505411868817,
+      "learning_rate": 4.3389713496720745e-05,
+      "loss": 1.3226,
+      "step": 705
+    },
+    {
+      "epoch": 0.6579683131407269,
+      "grad_norm": 1.7080990755629029,
+      "learning_rate": 4.3372454263030726e-05,
+      "loss": 1.2283,
+      "step": 706
+    },
+    {
+      "epoch": 0.6589002795899348,
+      "grad_norm": 2.033632968376693,
+      "learning_rate": 4.33551950293407e-05,
+      "loss": 1.4349,
+      "step": 707
+    },
+    {
+      "epoch": 0.6598322460391426,
+      "grad_norm": 2.0857120490128374,
+      "learning_rate": 4.3337935795650675e-05,
+      "loss": 1.3085,
+      "step": 708
+    },
+    {
+      "epoch": 0.6607642124883504,
+      "grad_norm": 2.0407798246043045,
+      "learning_rate": 4.332067656196065e-05,
+      "loss": 1.2574,
+      "step": 709
+    },
+    {
+      "epoch": 0.6616961789375583,
+      "grad_norm": 2.068839116664596,
+      "learning_rate": 4.330341732827063e-05,
+      "loss": 1.4416,
+      "step": 710
+    },
+    {
+      "epoch": 0.6626281453867661,
+      "grad_norm": 2.260574491875068,
+      "learning_rate": 4.3286158094580604e-05,
+      "loss": 1.3321,
+      "step": 711
+    },
+    {
+      "epoch": 0.6635601118359739,
+      "grad_norm": 2.1169301070834936,
+      "learning_rate": 4.326889886089058e-05,
+      "loss": 1.3624,
+      "step": 712
+    },
+    {
+      "epoch": 0.6644920782851818,
+      "grad_norm": 1.8193843456032963,
+      "learning_rate": 4.325163962720055e-05,
+      "loss": 1.4692,
+      "step": 713
+    },
+    {
+      "epoch": 0.6654240447343895,
+      "grad_norm": 1.6817905674293894,
+      "learning_rate": 4.3234380393510534e-05,
+      "loss": 1.0599,
+      "step": 714
+    },
+    {
+      "epoch": 0.6663560111835974,
+      "grad_norm": 1.805931496156492,
+      "learning_rate": 4.321712115982051e-05,
+      "loss": 1.569,
+      "step": 715
+    },
+    {
+      "epoch": 0.6672879776328052,
+      "grad_norm": 1.8018407869142257,
+      "learning_rate": 4.319986192613048e-05,
+      "loss": 1.4754,
+      "step": 716
+    },
+    {
+      "epoch": 0.668219944082013,
+      "grad_norm": 1.995494775005292,
+      "learning_rate": 4.3182602692440464e-05,
+      "loss": 1.3981,
+      "step": 717
+    },
+    {
+      "epoch": 0.6691519105312209,
+      "grad_norm": 2.6113966725249975,
+      "learning_rate": 4.316534345875044e-05,
+      "loss": 1.7296,
+      "step": 718
+    },
+    {
+      "epoch": 0.6700838769804287,
+      "grad_norm": 2.0923211158549644,
+      "learning_rate": 4.314808422506041e-05,
+      "loss": 1.4699,
+      "step": 719
+    },
+    {
+      "epoch": 0.6710158434296365,
+      "grad_norm": 2.215334894046195,
+      "learning_rate": 4.3130824991370387e-05,
+      "loss": 1.3603,
+      "step": 720
+    },
+    {
+      "epoch": 0.6719478098788444,
+      "grad_norm": 1.9188884894546603,
+      "learning_rate": 4.311356575768036e-05,
+      "loss": 1.6086,
+      "step": 721
+    },
+    {
+      "epoch": 0.6728797763280522,
+      "grad_norm": 2.0354708689542456,
+      "learning_rate": 4.3096306523990335e-05,
+      "loss": 1.2386,
+      "step": 722
+    },
+    {
+      "epoch": 0.67381174277726,
+      "grad_norm": 1.8583141385215627,
+      "learning_rate": 4.307904729030031e-05,
+      "loss": 1.0889,
+      "step": 723
+    },
+    {
+      "epoch": 0.6747437092264679,
+      "grad_norm": 2.0201413626752958,
+      "learning_rate": 4.3061788056610284e-05,
+      "loss": 1.5631,
+      "step": 724
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 1.784538737119986,
+      "learning_rate": 4.3044528822920265e-05,
+      "loss": 1.5267,
+      "step": 725
+    },
+    {
+      "epoch": 0.6766076421248836,
+      "grad_norm": 1.8004434488092709,
+      "learning_rate": 4.302726958923024e-05,
+      "loss": 1.3363,
+      "step": 726
+    },
+    {
+      "epoch": 0.6775396085740913,
+      "grad_norm": 1.8512769312437596,
+      "learning_rate": 4.301001035554021e-05,
+      "loss": 1.7407,
+      "step": 727
+    },
+    {
+      "epoch": 0.6784715750232991,
+      "grad_norm": 1.9048935374670353,
+      "learning_rate": 4.2992751121850194e-05,
+      "loss": 1.2801,
+      "step": 728
+    },
+    {
+      "epoch": 0.679403541472507,
+      "grad_norm": 1.7111044973481957,
+      "learning_rate": 4.297549188816017e-05,
+      "loss": 1.5748,
+      "step": 729
+    },
+    {
+      "epoch": 0.6803355079217148,
+      "grad_norm": 1.8369479940429119,
+      "learning_rate": 4.295823265447014e-05,
+      "loss": 1.3689,
+      "step": 730
+    },
+    {
+      "epoch": 0.6812674743709226,
+      "grad_norm": 2.039652703796649,
+      "learning_rate": 4.294097342078012e-05,
+      "loss": 1.5484,
+      "step": 731
+    },
+    {
+      "epoch": 0.6821994408201305,
+      "grad_norm": 1.8052333411452912,
+      "learning_rate": 4.29237141870901e-05,
+      "loss": 1.4378,
+      "step": 732
+    },
+    {
+      "epoch": 0.6831314072693383,
+      "grad_norm": 2.2083594619797005,
+      "learning_rate": 4.290645495340007e-05,
+      "loss": 1.3338,
+      "step": 733
+    },
+    {
+      "epoch": 0.6840633737185461,
+      "grad_norm": 2.068564420818082,
+      "learning_rate": 4.288919571971005e-05,
+      "loss": 1.4246,
+      "step": 734
+    },
+    {
+      "epoch": 0.684995340167754,
+      "grad_norm": 1.839746469314068,
+      "learning_rate": 4.287193648602002e-05,
+      "loss": 1.4275,
+      "step": 735
+    },
+    {
+      "epoch": 0.6859273066169618,
+      "grad_norm": 2.080353217539164,
+      "learning_rate": 4.285467725233e-05,
+      "loss": 1.4841,
+      "step": 736
+    },
+    {
+      "epoch": 0.6868592730661697,
+      "grad_norm": 2.132048131653378,
+      "learning_rate": 4.2837418018639977e-05,
+      "loss": 1.4302,
+      "step": 737
+    },
+    {
+      "epoch": 0.6877912395153775,
+      "grad_norm": 2.4429629450599926,
+      "learning_rate": 4.282015878494995e-05,
+      "loss": 1.5878,
+      "step": 738
+    },
+    {
+      "epoch": 0.6887232059645852,
+      "grad_norm": 2.0104565269910806,
+      "learning_rate": 4.2802899551259925e-05,
+      "loss": 1.4688,
+      "step": 739
+    },
+    {
+      "epoch": 0.6896551724137931,
+      "grad_norm": 1.9995526665202663,
+      "learning_rate": 4.27856403175699e-05,
+      "loss": 1.3934,
+      "step": 740
+    },
+    {
+      "epoch": 0.6905871388630009,
+      "grad_norm": 1.9560558763233773,
+      "learning_rate": 4.2768381083879874e-05,
+      "loss": 1.2263,
+      "step": 741
+    },
+    {
+      "epoch": 0.6915191053122087,
+      "grad_norm": 1.9054657448037862,
+      "learning_rate": 4.2751121850189855e-05,
+      "loss": 1.3038,
+      "step": 742
+    },
+    {
+      "epoch": 0.6924510717614166,
+      "grad_norm": 1.960450681886479,
+      "learning_rate": 4.273386261649983e-05,
+      "loss": 1.4726,
+      "step": 743
+    },
+    {
+      "epoch": 0.6933830382106244,
+      "grad_norm": 1.813966161179375,
+      "learning_rate": 4.27166033828098e-05,
+      "loss": 1.2202,
+      "step": 744
+    },
+    {
+      "epoch": 0.6943150046598322,
+      "grad_norm": 2.357755098996246,
+      "learning_rate": 4.269934414911978e-05,
+      "loss": 1.3541,
+      "step": 745
+    },
+    {
+      "epoch": 0.6952469711090401,
+      "grad_norm": 1.801295911630654,
+      "learning_rate": 4.268208491542976e-05,
+      "loss": 1.1799,
+      "step": 746
+    },
+    {
+      "epoch": 0.6961789375582479,
+      "grad_norm": 1.9166541901179965,
+      "learning_rate": 4.266482568173973e-05,
+      "loss": 1.2533,
+      "step": 747
+    },
+    {
+      "epoch": 0.6971109040074557,
+      "grad_norm": 1.9612877041495227,
+      "learning_rate": 4.264756644804971e-05,
+      "loss": 1.3527,
+      "step": 748
+    },
+    {
+      "epoch": 0.6980428704566636,
+      "grad_norm": 2.039639313233368,
+      "learning_rate": 4.263030721435968e-05,
+      "loss": 1.5973,
+      "step": 749
+    },
+    {
+      "epoch": 0.6989748369058714,
+      "grad_norm": 1.9638839081401054,
+      "learning_rate": 4.261304798066966e-05,
+      "loss": 1.404,
+      "step": 750
+    },
+    {
+      "epoch": 0.6999068033550793,
+      "grad_norm": 1.9989726012761442,
+      "learning_rate": 4.259578874697964e-05,
+      "loss": 1.3617,
+      "step": 751
+    },
+    {
+      "epoch": 0.700838769804287,
+      "grad_norm": 1.7303370186592049,
+      "learning_rate": 4.257852951328961e-05,
+      "loss": 1.1912,
+      "step": 752
+    },
+    {
+      "epoch": 0.7017707362534948,
+      "grad_norm": 1.7824327339455543,
+      "learning_rate": 4.2561270279599586e-05,
+      "loss": 1.3445,
+      "step": 753
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 1.898063850984462,
+      "learning_rate": 4.2544011045909567e-05,
+      "loss": 1.4838,
+      "step": 754
+    },
+    {
+      "epoch": 0.7036346691519105,
+      "grad_norm": 2.448783061819776,
+      "learning_rate": 4.252675181221954e-05,
+      "loss": 1.4404,
+      "step": 755
+    },
+    {
+      "epoch": 0.7045666356011183,
+      "grad_norm": 1.6681545639736448,
+      "learning_rate": 4.2509492578529515e-05,
+      "loss": 1.194,
+      "step": 756
+    },
+    {
+      "epoch": 0.7054986020503262,
+      "grad_norm": 1.643472788283256,
+      "learning_rate": 4.2492233344839496e-05,
+      "loss": 1.2979,
+      "step": 757
+    },
+    {
+      "epoch": 0.706430568499534,
+      "grad_norm": 2.1464618280800183,
+      "learning_rate": 4.247497411114947e-05,
+      "loss": 1.4137,
+      "step": 758
+    },
+    {
+      "epoch": 0.7073625349487418,
+      "grad_norm": 2.2549024898561263,
+      "learning_rate": 4.2457714877459445e-05,
+      "loss": 1.4627,
+      "step": 759
+    },
+    {
+      "epoch": 0.7082945013979497,
+      "grad_norm": 2.036954523123664,
+      "learning_rate": 4.244045564376942e-05,
+      "loss": 1.3857,
+      "step": 760
+    },
+    {
+      "epoch": 0.7092264678471575,
+      "grad_norm": 1.707606553488717,
+      "learning_rate": 4.242319641007939e-05,
+      "loss": 1.4075,
+      "step": 761
+    },
+    {
+      "epoch": 0.7101584342963654,
+      "grad_norm": 2.275791293220143,
+      "learning_rate": 4.240593717638937e-05,
+      "loss": 1.8326,
+      "step": 762
+    },
+    {
+      "epoch": 0.7110904007455732,
+      "grad_norm": 2.143537814670928,
+      "learning_rate": 4.238867794269934e-05,
+      "loss": 1.4656,
+      "step": 763
+    },
+    {
+      "epoch": 0.712022367194781,
+      "grad_norm": 2.0900351760341627,
+      "learning_rate": 4.2371418709009316e-05,
+      "loss": 1.2234,
+      "step": 764
+    },
+    {
+      "epoch": 0.7129543336439889,
+      "grad_norm": 1.74679964788102,
+      "learning_rate": 4.23541594753193e-05,
+      "loss": 1.2543,
+      "step": 765
+    },
+    {
+      "epoch": 0.7138863000931966,
+      "grad_norm": 2.259967191861325,
+      "learning_rate": 4.233690024162927e-05,
+      "loss": 1.3167,
+      "step": 766
+    },
+    {
+      "epoch": 0.7148182665424044,
+      "grad_norm": 1.9078826526391914,
+      "learning_rate": 4.2319641007939246e-05,
+      "loss": 1.0234,
+      "step": 767
+    },
+    {
+      "epoch": 0.7157502329916123,
+      "grad_norm": 1.6792087971731047,
+      "learning_rate": 4.230238177424923e-05,
+      "loss": 1.2066,
+      "step": 768
+    },
+    {
+      "epoch": 0.7166821994408201,
+      "grad_norm": 2.2839457188901373,
+      "learning_rate": 4.22851225405592e-05,
+      "loss": 1.6409,
+      "step": 769
+    },
+    {
+      "epoch": 0.7176141658900279,
+      "grad_norm": 2.3506870527025447,
+      "learning_rate": 4.2267863306869176e-05,
+      "loss": 1.4189,
+      "step": 770
+    },
+    {
+      "epoch": 0.7185461323392358,
+      "grad_norm": 1.904354725520936,
+      "learning_rate": 4.225060407317915e-05,
+      "loss": 1.1936,
+      "step": 771
+    },
+    {
+      "epoch": 0.7194780987884436,
+      "grad_norm": 1.8411913186294977,
+      "learning_rate": 4.223334483948913e-05,
+      "loss": 1.1153,
+      "step": 772
+    },
+    {
+      "epoch": 0.7204100652376515,
+      "grad_norm": 1.8844451427508484,
+      "learning_rate": 4.2216085605799105e-05,
+      "loss": 1.5136,
+      "step": 773
+    },
+    {
+      "epoch": 0.7213420316868593,
+      "grad_norm": 2.152178237184675,
+      "learning_rate": 4.219882637210908e-05,
+      "loss": 1.3257,
+      "step": 774
+    },
+    {
+      "epoch": 0.7222739981360671,
+      "grad_norm": 2.163832655916822,
+      "learning_rate": 4.2181567138419054e-05,
+      "loss": 1.4058,
+      "step": 775
+    },
+    {
+      "epoch": 0.723205964585275,
+      "grad_norm": 1.8460081974482896,
+      "learning_rate": 4.2164307904729035e-05,
+      "loss": 1.326,
+      "step": 776
+    },
+    {
+      "epoch": 0.7241379310344828,
+      "grad_norm": 2.4547352437068435,
+      "learning_rate": 4.214704867103901e-05,
+      "loss": 1.5332,
+      "step": 777
+    },
+    {
+      "epoch": 0.7250698974836906,
+      "grad_norm": 3.1781687089771635,
+      "learning_rate": 4.2129789437348983e-05,
+      "loss": 1.8521,
+      "step": 778
+    },
+    {
+      "epoch": 0.7260018639328985,
+      "grad_norm": 1.5771906706032106,
+      "learning_rate": 4.2112530203658964e-05,
+      "loss": 1.0025,
+      "step": 779
+    },
+    {
+      "epoch": 0.7269338303821062,
+      "grad_norm": 2.1156274926507503,
+      "learning_rate": 4.209527096996894e-05,
+      "loss": 1.4813,
+      "step": 780
+    },
+    {
+      "epoch": 0.727865796831314,
+      "grad_norm": 1.857741344979411,
+      "learning_rate": 4.207801173627891e-05,
+      "loss": 1.1814,
+      "step": 781
+    },
+    {
+      "epoch": 0.7287977632805219,
+      "grad_norm": 2.5538313098437317,
+      "learning_rate": 4.206075250258889e-05,
+      "loss": 1.5076,
+      "step": 782
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 2.0995761789298633,
+      "learning_rate": 4.204349326889886e-05,
+      "loss": 1.2312,
+      "step": 783
+    },
+    {
+      "epoch": 0.7306616961789375,
+      "grad_norm": 1.872984552913146,
+      "learning_rate": 4.2026234035208836e-05,
+      "loss": 1.3945,
+      "step": 784
+    },
+    {
+      "epoch": 0.7315936626281454,
+      "grad_norm": 1.4668222296614135,
+      "learning_rate": 4.200897480151881e-05,
+      "loss": 0.9713,
+      "step": 785
+    },
+    {
+      "epoch": 0.7325256290773532,
+      "grad_norm": 2.021824031584892,
+      "learning_rate": 4.1991715567828785e-05,
+      "loss": 1.4965,
+      "step": 786
+    },
+    {
+      "epoch": 0.7334575955265611,
+      "grad_norm": 2.094023751582729,
+      "learning_rate": 4.1974456334138766e-05,
+      "loss": 1.5319,
+      "step": 787
+    },
+    {
+      "epoch": 0.7343895619757689,
+      "grad_norm": 2.3430265390004137,
+      "learning_rate": 4.195719710044874e-05,
+      "loss": 1.4636,
+      "step": 788
+    },
+    {
+      "epoch": 0.7353215284249767,
+      "grad_norm": 2.5246992874395926,
+      "learning_rate": 4.1939937866758714e-05,
+      "loss": 1.7064,
+      "step": 789
+    },
+    {
+      "epoch": 0.7362534948741846,
+      "grad_norm": 1.6062859386687087,
+      "learning_rate": 4.1922678633068695e-05,
+      "loss": 1.2336,
+      "step": 790
+    },
+    {
+      "epoch": 0.7371854613233924,
+      "grad_norm": 2.2130188315021235,
+      "learning_rate": 4.190541939937867e-05,
+      "loss": 1.4448,
+      "step": 791
+    },
+    {
+      "epoch": 0.7381174277726001,
+      "grad_norm": 2.0156057902625237,
+      "learning_rate": 4.1888160165688644e-05,
+      "loss": 1.3595,
+      "step": 792
+    },
+    {
+      "epoch": 0.739049394221808,
+      "grad_norm": 1.5084027225081174,
+      "learning_rate": 4.187090093199862e-05,
+      "loss": 1.0535,
+      "step": 793
+    },
+    {
+      "epoch": 0.7399813606710158,
+      "grad_norm": 2.0420836472049184,
+      "learning_rate": 4.18536416983086e-05,
+      "loss": 1.5454,
+      "step": 794
+    },
+    {
+      "epoch": 0.7409133271202236,
+      "grad_norm": 2.1316810274358042,
+      "learning_rate": 4.1836382464618573e-05,
+      "loss": 1.4086,
+      "step": 795
+    },
+    {
+      "epoch": 0.7418452935694315,
+      "grad_norm": 1.7784492112707815,
+      "learning_rate": 4.181912323092855e-05,
+      "loss": 1.2053,
+      "step": 796
+    },
+    {
+      "epoch": 0.7427772600186393,
+      "grad_norm": 1.9875258405143827,
+      "learning_rate": 4.180186399723853e-05,
+      "loss": 1.3411,
+      "step": 797
+    },
+    {
+      "epoch": 0.7437092264678472,
+      "grad_norm": 1.9642751256388349,
+      "learning_rate": 4.17846047635485e-05,
+      "loss": 1.2397,
+      "step": 798
+    },
+    {
+      "epoch": 0.744641192917055,
+      "grad_norm": 1.786380925658126,
+      "learning_rate": 4.176734552985848e-05,
+      "loss": 1.1482,
+      "step": 799
+    },
+    {
+      "epoch": 0.7455731593662628,
+      "grad_norm": 2.0998358579921605,
+      "learning_rate": 4.175008629616845e-05,
+      "loss": 1.3326,
+      "step": 800
+    },
+    {
+      "epoch": 0.7465051258154707,
+      "grad_norm": 1.9868316414285263,
+      "learning_rate": 4.173282706247843e-05,
+      "loss": 1.3788,
+      "step": 801
+    },
+    {
+      "epoch": 0.7474370922646785,
+      "grad_norm": 1.9705824333526967,
+      "learning_rate": 4.171556782878841e-05,
+      "loss": 1.1955,
+      "step": 802
+    },
+    {
+      "epoch": 0.7483690587138863,
+      "grad_norm": 2.1457603992054697,
+      "learning_rate": 4.169830859509838e-05,
+      "loss": 1.3472,
+      "step": 803
+    },
+    {
+      "epoch": 0.7493010251630942,
+      "grad_norm": 1.998322432574611,
+      "learning_rate": 4.1681049361408356e-05,
+      "loss": 1.3968,
+      "step": 804
+    },
+    {
+      "epoch": 0.750232991612302,
+      "grad_norm": 2.1137182094204494,
+      "learning_rate": 4.166379012771833e-05,
+      "loss": 1.4327,
+      "step": 805
+    },
+    {
+      "epoch": 0.7511649580615097,
+      "grad_norm": 1.7042345434704442,
+      "learning_rate": 4.1646530894028304e-05,
+      "loss": 1.1611,
+      "step": 806
+    },
+    {
+      "epoch": 0.7520969245107176,
+      "grad_norm": 1.9415772849771413,
+      "learning_rate": 4.162927166033828e-05,
+      "loss": 1.3665,
+      "step": 807
+    },
+    {
+      "epoch": 0.7530288909599254,
+      "grad_norm": 1.9323533603052085,
+      "learning_rate": 4.161201242664826e-05,
+      "loss": 1.3674,
+      "step": 808
+    },
+    {
+      "epoch": 0.7539608574091333,
+      "grad_norm": 2.1363351137889746,
+      "learning_rate": 4.1594753192958234e-05,
+      "loss": 1.2984,
+      "step": 809
+    },
+    {
+      "epoch": 0.7548928238583411,
+      "grad_norm": 2.1094337931657186,
+      "learning_rate": 4.157749395926821e-05,
+      "loss": 1.414,
+      "step": 810
+    },
+    {
+      "epoch": 0.7558247903075489,
+      "grad_norm": 2.1336364380351442,
+      "learning_rate": 4.156023472557818e-05,
+      "loss": 1.3668,
+      "step": 811
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 2.422460692594104,
+      "learning_rate": 4.1542975491888163e-05,
+      "loss": 1.4787,
+      "step": 812
+    },
+    {
+      "epoch": 0.7576887232059646,
+      "grad_norm": 1.9343999034722352,
+      "learning_rate": 4.152571625819814e-05,
+      "loss": 1.2479,
+      "step": 813
+    },
+    {
+      "epoch": 0.7586206896551724,
+      "grad_norm": 2.137581507042891,
+      "learning_rate": 4.150845702450811e-05,
+      "loss": 1.5358,
+      "step": 814
+    },
+    {
+      "epoch": 0.7595526561043803,
+      "grad_norm": 1.71055761925614,
+      "learning_rate": 4.1491197790818086e-05,
+      "loss": 1.1954,
+      "step": 815
+    },
+    {
+      "epoch": 0.7604846225535881,
+      "grad_norm": 2.7298276959775225,
+      "learning_rate": 4.147393855712807e-05,
+      "loss": 1.418,
+      "step": 816
+    },
+    {
+      "epoch": 0.7614165890027959,
+      "grad_norm": 1.9692873150083845,
+      "learning_rate": 4.145667932343804e-05,
+      "loss": 1.4227,
+      "step": 817
+    },
+    {
+      "epoch": 0.7623485554520038,
+      "grad_norm": 1.9923431792343884,
+      "learning_rate": 4.1439420089748016e-05,
+      "loss": 1.5921,
+      "step": 818
+    },
+    {
+      "epoch": 0.7632805219012115,
+      "grad_norm": 1.842620732185841,
+      "learning_rate": 4.1422160856058e-05,
+      "loss": 1.3445,
+      "step": 819
+    },
+    {
+      "epoch": 0.7642124883504194,
+      "grad_norm": 1.481089789525297,
+      "learning_rate": 4.140490162236797e-05,
+      "loss": 1.1006,
+      "step": 820
+    },
+    {
+      "epoch": 0.7651444547996272,
+      "grad_norm": 1.6452301114097931,
+      "learning_rate": 4.1387642388677946e-05,
+      "loss": 1.2526,
+      "step": 821
+    },
+    {
+      "epoch": 0.766076421248835,
+      "grad_norm": 2.020566238694564,
+      "learning_rate": 4.137038315498792e-05,
+      "loss": 1.3055,
+      "step": 822
+    },
+    {
+      "epoch": 0.7670083876980429,
+      "grad_norm": 1.6831511115766118,
+      "learning_rate": 4.13531239212979e-05,
+      "loss": 1.1611,
+      "step": 823
+    },
+    {
+      "epoch": 0.7679403541472507,
+      "grad_norm": 1.426846329804794,
+      "learning_rate": 4.1335864687607875e-05,
+      "loss": 1.0319,
+      "step": 824
+    },
+    {
+      "epoch": 0.7688723205964585,
+      "grad_norm": 1.895490520214705,
+      "learning_rate": 4.131860545391785e-05,
+      "loss": 1.3094,
+      "step": 825
+    },
+    {
+      "epoch": 0.7698042870456664,
+      "grad_norm": 1.7215498623007965,
+      "learning_rate": 4.1301346220227824e-05,
+      "loss": 1.2584,
+      "step": 826
+    },
+    {
+      "epoch": 0.7707362534948742,
+      "grad_norm": 1.7723558887070212,
+      "learning_rate": 4.12840869865378e-05,
+      "loss": 1.3375,
+      "step": 827
+    },
+    {
+      "epoch": 0.771668219944082,
+      "grad_norm": 1.7301957170965272,
+      "learning_rate": 4.126682775284777e-05,
+      "loss": 1.32,
+      "step": 828
+    },
+    {
+      "epoch": 0.7726001863932899,
+      "grad_norm": 1.689106302066971,
+      "learning_rate": 4.124956851915775e-05,
+      "loss": 1.0552,
+      "step": 829
+    },
+    {
+      "epoch": 0.7735321528424977,
+      "grad_norm": 1.9699612828952748,
+      "learning_rate": 4.123230928546773e-05,
+      "loss": 1.2044,
+      "step": 830
+    },
+    {
+      "epoch": 0.7744641192917054,
+      "grad_norm": 1.6998403263730744,
+      "learning_rate": 4.12150500517777e-05,
+      "loss": 1.2234,
+      "step": 831
+    },
+    {
+      "epoch": 0.7753960857409133,
+      "grad_norm": 2.2900120384883547,
+      "learning_rate": 4.1197790818087676e-05,
+      "loss": 1.2399,
+      "step": 832
+    },
+    {
+      "epoch": 0.7763280521901211,
+      "grad_norm": 2.159485964542713,
+      "learning_rate": 4.118053158439765e-05,
+      "loss": 1.2511,
+      "step": 833
+    },
+    {
+      "epoch": 0.777260018639329,
+      "grad_norm": 1.7091235436138634,
+      "learning_rate": 4.116327235070763e-05,
+      "loss": 1.2612,
+      "step": 834
+    },
+    {
+      "epoch": 0.7781919850885368,
+      "grad_norm": 1.5182654085723892,
+      "learning_rate": 4.1146013117017606e-05,
+      "loss": 1.0785,
+      "step": 835
+    },
+    {
+      "epoch": 0.7791239515377446,
+      "grad_norm": 2.1679282828742132,
+      "learning_rate": 4.112875388332758e-05,
+      "loss": 1.2214,
+      "step": 836
+    },
+    {
+      "epoch": 0.7800559179869525,
+      "grad_norm": 2.0545478392319185,
+      "learning_rate": 4.111149464963756e-05,
+      "loss": 1.3771,
+      "step": 837
+    },
+    {
+      "epoch": 0.7809878844361603,
+      "grad_norm": 2.001442035409045,
+      "learning_rate": 4.1094235415947536e-05,
+      "loss": 1.7857,
+      "step": 838
+    },
+    {
+      "epoch": 0.7819198508853681,
+      "grad_norm": 1.856445806735656,
+      "learning_rate": 4.107697618225751e-05,
+      "loss": 1.1836,
+      "step": 839
+    },
+    {
+      "epoch": 0.782851817334576,
+      "grad_norm": 2.2562098959702204,
+      "learning_rate": 4.1059716948567484e-05,
+      "loss": 1.29,
+      "step": 840
+    },
+    {
+      "epoch": 0.7837837837837838,
+      "grad_norm": 2.062451033482535,
+      "learning_rate": 4.1042457714877465e-05,
+      "loss": 1.4149,
+      "step": 841
+    },
+    {
+      "epoch": 0.7847157502329916,
+      "grad_norm": 2.2933028301082707,
+      "learning_rate": 4.102519848118744e-05,
+      "loss": 1.3148,
+      "step": 842
+    },
+    {
+      "epoch": 0.7856477166821995,
+      "grad_norm": 2.007261883429039,
+      "learning_rate": 4.1007939247497414e-05,
+      "loss": 1.4345,
+      "step": 843
+    },
+    {
+      "epoch": 0.7865796831314072,
+      "grad_norm": 2.509753874215836,
+      "learning_rate": 4.099068001380739e-05,
+      "loss": 1.3703,
+      "step": 844
+    },
+    {
+      "epoch": 0.7875116495806151,
+      "grad_norm": 1.6702236977205052,
+      "learning_rate": 4.097342078011737e-05,
+      "loss": 1.2786,
+      "step": 845
+    },
+    {
+      "epoch": 0.7884436160298229,
+      "grad_norm": 1.7552769896471143,
+      "learning_rate": 4.0956161546427344e-05,
+      "loss": 1.0768,
+      "step": 846
+    },
+    {
+      "epoch": 0.7893755824790307,
+      "grad_norm": 1.824737295339658,
+      "learning_rate": 4.093890231273732e-05,
+      "loss": 1.2695,
+      "step": 847
+    },
+    {
+      "epoch": 0.7903075489282386,
+      "grad_norm": 1.941913563066448,
+      "learning_rate": 4.092164307904729e-05,
+      "loss": 1.0388,
+      "step": 848
+    },
+    {
+      "epoch": 0.7912395153774464,
+      "grad_norm": 2.529374499949455,
+      "learning_rate": 4.0904383845357266e-05,
+      "loss": 1.3506,
+      "step": 849
+    },
+    {
+      "epoch": 0.7921714818266542,
+      "grad_norm": 1.6324730057952308,
+      "learning_rate": 4.088712461166724e-05,
+      "loss": 1.1516,
+      "step": 850
+    },
+    {
+      "epoch": 0.7931034482758621,
+      "grad_norm": 2.2040177948994266,
+      "learning_rate": 4.0869865377977215e-05,
+      "loss": 1.2941,
+      "step": 851
+    },
+    {
+      "epoch": 0.7940354147250699,
+      "grad_norm": 1.9234805625116973,
+      "learning_rate": 4.0852606144287196e-05,
+      "loss": 1.3124,
+      "step": 852
+    },
+    {
+      "epoch": 0.7949673811742777,
+      "grad_norm": 2.000901748833031,
+      "learning_rate": 4.083534691059717e-05,
+      "loss": 1.1968,
+      "step": 853
+    },
+    {
+      "epoch": 0.7958993476234856,
+      "grad_norm": 2.3159733508189757,
+      "learning_rate": 4.0818087676907145e-05,
+      "loss": 1.2288,
+      "step": 854
+    },
+    {
+      "epoch": 0.7968313140726934,
+      "grad_norm": 2.6692105464808913,
+      "learning_rate": 4.080082844321712e-05,
+      "loss": 1.4358,
+      "step": 855
+    },
+    {
+      "epoch": 0.7977632805219013,
+      "grad_norm": 1.804844490862743,
+      "learning_rate": 4.07835692095271e-05,
+      "loss": 1.1892,
+      "step": 856
+    },
+    {
+      "epoch": 0.798695246971109,
+      "grad_norm": 2.446274033262904,
+      "learning_rate": 4.0766309975837074e-05,
+      "loss": 1.3485,
+      "step": 857
+    },
+    {
+      "epoch": 0.7996272134203168,
+      "grad_norm": 2.1117499308954715,
+      "learning_rate": 4.074905074214705e-05,
+      "loss": 1.1937,
+      "step": 858
+    },
+    {
+      "epoch": 0.8005591798695247,
+      "grad_norm": 2.2191020286757692,
+      "learning_rate": 4.073179150845703e-05,
+      "loss": 1.2558,
+      "step": 859
+    },
+    {
+      "epoch": 0.8014911463187325,
+      "grad_norm": 2.1368908451809863,
+      "learning_rate": 4.0714532274767004e-05,
+      "loss": 1.3898,
+      "step": 860
+    },
+    {
+      "epoch": 0.8024231127679403,
+      "grad_norm": 2.0097238668695017,
+      "learning_rate": 4.069727304107698e-05,
+      "loss": 1.0717,
+      "step": 861
+    },
+    {
+      "epoch": 0.8033550792171482,
+      "grad_norm": 1.6938823350328949,
+      "learning_rate": 4.068001380738695e-05,
+      "loss": 1.1417,
+      "step": 862
+    },
+    {
+      "epoch": 0.804287045666356,
+      "grad_norm": 1.650642822579472,
+      "learning_rate": 4.0662754573696934e-05,
+      "loss": 0.8511,
+      "step": 863
+    },
+    {
+      "epoch": 0.8052190121155638,
+      "grad_norm": 1.5952752888137047,
+      "learning_rate": 4.064549534000691e-05,
+      "loss": 1.0349,
+      "step": 864
+    },
+    {
+      "epoch": 0.8061509785647717,
+      "grad_norm": 1.7937713667959465,
+      "learning_rate": 4.062823610631688e-05,
+      "loss": 1.2318,
+      "step": 865
+    },
+    {
+      "epoch": 0.8070829450139795,
+      "grad_norm": 1.8564857053202517,
+      "learning_rate": 4.0610976872626856e-05,
+      "loss": 1.2679,
+      "step": 866
+    },
+    {
+      "epoch": 0.8080149114631874,
+      "grad_norm": 2.273936576615098,
+      "learning_rate": 4.059371763893684e-05,
+      "loss": 1.5451,
+      "step": 867
+    },
+    {
+      "epoch": 0.8089468779123952,
+      "grad_norm": 1.9260024272008691,
+      "learning_rate": 4.057645840524681e-05,
+      "loss": 1.3364,
+      "step": 868
+    },
+    {
+      "epoch": 0.809878844361603,
+      "grad_norm": 2.1003388329173838,
+      "learning_rate": 4.0559199171556786e-05,
+      "loss": 1.3889,
+      "step": 869
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 2.1391401971501782,
+      "learning_rate": 4.054193993786676e-05,
+      "loss": 1.2977,
+      "step": 870
+    },
+    {
+      "epoch": 0.8117427772600186,
+      "grad_norm": 2.2414441700185317,
+      "learning_rate": 4.0524680704176735e-05,
+      "loss": 1.7499,
+      "step": 871
+    },
+    {
+      "epoch": 0.8126747437092264,
+      "grad_norm": 2.1325018463275134,
+      "learning_rate": 4.050742147048671e-05,
+      "loss": 1.5597,
+      "step": 872
+    },
+    {
+      "epoch": 0.8136067101584343,
+      "grad_norm": 2.836751154470441,
+      "learning_rate": 4.049016223679668e-05,
+      "loss": 1.4329,
+      "step": 873
+    },
+    {
+      "epoch": 0.8145386766076421,
+      "grad_norm": 1.8870831065862983,
+      "learning_rate": 4.0472903003106664e-05,
+      "loss": 1.3293,
+      "step": 874
+    },
+    {
+      "epoch": 0.8154706430568499,
+      "grad_norm": 2.163642855689422,
+      "learning_rate": 4.045564376941664e-05,
+      "loss": 1.3216,
+      "step": 875
+    },
+    {
+      "epoch": 0.8164026095060578,
+      "grad_norm": 2.0058550215101087,
+      "learning_rate": 4.043838453572661e-05,
+      "loss": 1.3999,
+      "step": 876
+    },
+    {
+      "epoch": 0.8173345759552656,
+      "grad_norm": 2.0304998247719364,
+      "learning_rate": 4.042112530203659e-05,
+      "loss": 1.4916,
+      "step": 877
+    },
+    {
+      "epoch": 0.8182665424044734,
+      "grad_norm": 1.8819692434670448,
+      "learning_rate": 4.040386606834657e-05,
+      "loss": 1.3905,
+      "step": 878
+    },
+    {
+      "epoch": 0.8191985088536813,
+      "grad_norm": 2.2917133729712598,
+      "learning_rate": 4.038660683465654e-05,
+      "loss": 1.4333,
+      "step": 879
+    },
+    {
+      "epoch": 0.8201304753028891,
+      "grad_norm": 1.934741303445692,
+      "learning_rate": 4.036934760096652e-05,
+      "loss": 0.9711,
+      "step": 880
+    },
+    {
+      "epoch": 0.821062441752097,
+      "grad_norm": 1.5342838793191045,
+      "learning_rate": 4.03520883672765e-05,
+      "loss": 1.3011,
+      "step": 881
+    },
+    {
+      "epoch": 0.8219944082013048,
+      "grad_norm": 1.6966617934920785,
+      "learning_rate": 4.033482913358647e-05,
+      "loss": 1.1449,
+      "step": 882
+    },
+    {
+      "epoch": 0.8229263746505125,
+      "grad_norm": 2.3849590992652168,
+      "learning_rate": 4.0317569899896447e-05,
+      "loss": 1.4187,
+      "step": 883
+    },
+    {
+      "epoch": 0.8238583410997204,
+      "grad_norm": 2.2656581406813228,
+      "learning_rate": 4.030031066620642e-05,
+      "loss": 1.3341,
+      "step": 884
+    },
+    {
+      "epoch": 0.8247903075489282,
+      "grad_norm": 1.6607117587545748,
+      "learning_rate": 4.02830514325164e-05,
+      "loss": 1.0305,
+      "step": 885
+    },
+    {
+      "epoch": 0.825722273998136,
+      "grad_norm": 1.8611972467850384,
+      "learning_rate": 4.0265792198826376e-05,
+      "loss": 1.1792,
+      "step": 886
+    },
+    {
+      "epoch": 0.8266542404473439,
+      "grad_norm": 2.1332496109255787,
+      "learning_rate": 4.024853296513635e-05,
+      "loss": 1.2068,
+      "step": 887
+    },
+    {
+      "epoch": 0.8275862068965517,
+      "grad_norm": 1.6628965926771502,
+      "learning_rate": 4.023127373144633e-05,
+      "loss": 1.1684,
+      "step": 888
+    },
+    {
+      "epoch": 0.8285181733457595,
+      "grad_norm": 1.4349256536201414,
+      "learning_rate": 4.0214014497756306e-05,
+      "loss": 0.9308,
+      "step": 889
+    },
+    {
+      "epoch": 0.8294501397949674,
+      "grad_norm": 2.138406792893459,
+      "learning_rate": 4.019675526406628e-05,
+      "loss": 1.3524,
+      "step": 890
+    },
+    {
+      "epoch": 0.8303821062441752,
+      "grad_norm": 2.418985113281917,
+      "learning_rate": 4.0179496030376254e-05,
+      "loss": 1.5346,
+      "step": 891
+    },
+    {
+      "epoch": 0.8313140726933831,
+      "grad_norm": 2.216705948623736,
+      "learning_rate": 4.016223679668623e-05,
+      "loss": 1.2678,
+      "step": 892
+    },
+    {
+      "epoch": 0.8322460391425909,
+      "grad_norm": 1.8272370788563528,
+      "learning_rate": 4.01449775629962e-05,
+      "loss": 1.2358,
+      "step": 893
+    },
+    {
+      "epoch": 0.8331780055917987,
+      "grad_norm": 1.9851865100922867,
+      "learning_rate": 4.012771832930618e-05,
+      "loss": 1.1082,
+      "step": 894
+    },
+    {
+      "epoch": 0.8341099720410066,
+      "grad_norm": 1.8460472285866196,
+      "learning_rate": 4.011045909561615e-05,
+      "loss": 1.4093,
+      "step": 895
+    },
+    {
+      "epoch": 0.8350419384902144,
+      "grad_norm": 1.6871618319167854,
+      "learning_rate": 4.009319986192613e-05,
+      "loss": 1.1534,
+      "step": 896
+    },
+    {
+      "epoch": 0.8359739049394221,
+      "grad_norm": 2.1969672893412175,
+      "learning_rate": 4.007594062823611e-05,
+      "loss": 1.1532,
+      "step": 897
+    },
+    {
+      "epoch": 0.83690587138863,
+      "grad_norm": 1.9429034432280947,
+      "learning_rate": 4.005868139454608e-05,
+      "loss": 1.213,
+      "step": 898
+    },
+    {
+      "epoch": 0.8378378378378378,
+      "grad_norm": 1.7330299218024476,
+      "learning_rate": 4.004142216085606e-05,
+      "loss": 1.1064,
+      "step": 899
+    },
+    {
+      "epoch": 0.8387698042870456,
+      "grad_norm": 2.481595656950581,
+      "learning_rate": 4.0024162927166037e-05,
+      "loss": 1.5327,
+      "step": 900
+    },
+    {
+      "epoch": 0.8397017707362535,
+      "grad_norm": 1.5500917200735969,
+      "learning_rate": 4.000690369347601e-05,
+      "loss": 1.1615,
+      "step": 901
+    },
+    {
+      "epoch": 0.8406337371854613,
+      "grad_norm": 1.9807755950064603,
+      "learning_rate": 3.9989644459785985e-05,
+      "loss": 1.2653,
+      "step": 902
+    },
+    {
+      "epoch": 0.8415657036346692,
+      "grad_norm": 1.735037955242122,
+      "learning_rate": 3.9972385226095966e-05,
+      "loss": 1.2624,
+      "step": 903
+    },
+    {
+      "epoch": 0.842497670083877,
+      "grad_norm": 1.9914753573022204,
+      "learning_rate": 3.995512599240594e-05,
+      "loss": 1.3748,
+      "step": 904
+    },
+    {
+      "epoch": 0.8434296365330848,
+      "grad_norm": 2.087494217522081,
+      "learning_rate": 3.9937866758715915e-05,
+      "loss": 1.2541,
+      "step": 905
+    },
+    {
+      "epoch": 0.8443616029822927,
+      "grad_norm": 1.891042354067379,
+      "learning_rate": 3.992060752502589e-05,
+      "loss": 1.441,
+      "step": 906
+    },
+    {
+      "epoch": 0.8452935694315005,
+      "grad_norm": 1.7129464358691793,
+      "learning_rate": 3.990334829133587e-05,
+      "loss": 1.1997,
+      "step": 907
+    },
+    {
+      "epoch": 0.8462255358807083,
+      "grad_norm": 2.1932722624089527,
+      "learning_rate": 3.9886089057645844e-05,
+      "loss": 1.13,
+      "step": 908
+    },
+    {
+      "epoch": 0.8471575023299162,
+      "grad_norm": 1.79980130696546,
+      "learning_rate": 3.986882982395582e-05,
+      "loss": 0.9927,
+      "step": 909
+    },
+    {
+      "epoch": 0.848089468779124,
+      "grad_norm": 2.133337603804457,
+      "learning_rate": 3.98515705902658e-05,
+      "loss": 1.1913,
+      "step": 910
+    },
+    {
+      "epoch": 0.8490214352283317,
+      "grad_norm": 2.064178049074544,
+      "learning_rate": 3.9834311356575774e-05,
+      "loss": 1.3737,
+      "step": 911
+    },
+    {
+      "epoch": 0.8499534016775396,
+      "grad_norm": 1.5311002238604936,
+      "learning_rate": 3.981705212288575e-05,
+      "loss": 1.1752,
+      "step": 912
+    },
+    {
+      "epoch": 0.8508853681267474,
+      "grad_norm": 1.771091847631717,
+      "learning_rate": 3.979979288919572e-05,
+      "loss": 1.0232,
+      "step": 913
+    },
+    {
+      "epoch": 0.8518173345759553,
+      "grad_norm": 2.215586422453935,
+      "learning_rate": 3.97825336555057e-05,
+      "loss": 1.4256,
+      "step": 914
+    },
+    {
+      "epoch": 0.8527493010251631,
+      "grad_norm": 1.9863848383821157,
+      "learning_rate": 3.976527442181567e-05,
+      "loss": 1.3008,
+      "step": 915
+    },
+    {
+      "epoch": 0.8536812674743709,
+      "grad_norm": 2.610380239050004,
+      "learning_rate": 3.9748015188125646e-05,
+      "loss": 1.3674,
+      "step": 916
+    },
+    {
+      "epoch": 0.8546132339235788,
+      "grad_norm": 1.5586368657982033,
+      "learning_rate": 3.973075595443562e-05,
+      "loss": 1.0763,
+      "step": 917
+    },
+    {
+      "epoch": 0.8555452003727866,
+      "grad_norm": 1.706456081272584,
+      "learning_rate": 3.97134967207456e-05,
+      "loss": 1.2932,
+      "step": 918
+    },
+    {
+      "epoch": 0.8564771668219944,
+      "grad_norm": 2.038221436995571,
+      "learning_rate": 3.9696237487055575e-05,
+      "loss": 1.2647,
+      "step": 919
+    },
+    {
+      "epoch": 0.8574091332712023,
+      "grad_norm": 2.442676125655239,
+      "learning_rate": 3.967897825336555e-05,
+      "loss": 1.3495,
+      "step": 920
+    },
+    {
+      "epoch": 0.8583410997204101,
+      "grad_norm": 2.0857532123192715,
+      "learning_rate": 3.966171901967553e-05,
+      "loss": 1.3104,
+      "step": 921
+    },
+    {
+      "epoch": 0.8592730661696178,
+      "grad_norm": 1.9578359144130342,
+      "learning_rate": 3.9644459785985505e-05,
+      "loss": 1.2101,
+      "step": 922
+    },
+    {
+      "epoch": 0.8602050326188257,
+      "grad_norm": 1.582014862254946,
+      "learning_rate": 3.962720055229548e-05,
+      "loss": 0.9202,
+      "step": 923
+    },
+    {
+      "epoch": 0.8611369990680335,
+      "grad_norm": 1.8725311649318939,
+      "learning_rate": 3.9609941318605453e-05,
+      "loss": 1.0633,
+      "step": 924
+    },
+    {
+      "epoch": 0.8620689655172413,
+      "grad_norm": 1.6868775656932953,
+      "learning_rate": 3.9592682084915434e-05,
+      "loss": 1.1282,
+      "step": 925
+    },
+    {
+      "epoch": 0.8630009319664492,
+      "grad_norm": 2.1593314472060072,
+      "learning_rate": 3.957542285122541e-05,
+      "loss": 1.1312,
+      "step": 926
+    },
+    {
+      "epoch": 0.863932898415657,
+      "grad_norm": 1.9665561344235662,
+      "learning_rate": 3.955816361753538e-05,
+      "loss": 1.1146,
+      "step": 927
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 1.8695778174936635,
+      "learning_rate": 3.9540904383845364e-05,
+      "loss": 1.1148,
+      "step": 928
+    },
+    {
+      "epoch": 0.8657968313140727,
+      "grad_norm": 1.7305471962376053,
+      "learning_rate": 3.952364515015534e-05,
+      "loss": 1.0966,
+      "step": 929
+    },
+    {
+      "epoch": 0.8667287977632805,
+      "grad_norm": 1.7610191179414965,
+      "learning_rate": 3.950638591646531e-05,
+      "loss": 0.9925,
+      "step": 930
+    },
+    {
+      "epoch": 0.8676607642124884,
+      "grad_norm": 2.108938431364582,
+      "learning_rate": 3.948912668277529e-05,
+      "loss": 1.1888,
+      "step": 931
+    },
+    {
+      "epoch": 0.8685927306616962,
+      "grad_norm": 1.5668957409982347,
+      "learning_rate": 3.947186744908527e-05,
+      "loss": 1.0915,
+      "step": 932
+    },
+    {
+      "epoch": 0.869524697110904,
+      "grad_norm": 1.811555202080944,
+      "learning_rate": 3.945460821539524e-05,
+      "loss": 1.1474,
+      "step": 933
+    },
+    {
+      "epoch": 0.8704566635601119,
+      "grad_norm": 1.5945599908322852,
+      "learning_rate": 3.943734898170522e-05,
+      "loss": 1.1682,
+      "step": 934
+    },
+    {
+      "epoch": 0.8713886300093197,
+      "grad_norm": 1.6195625043112556,
+      "learning_rate": 3.942008974801519e-05,
+      "loss": 1.1013,
+      "step": 935
+    },
+    {
+      "epoch": 0.8723205964585274,
+      "grad_norm": 1.919791013400488,
+      "learning_rate": 3.9402830514325165e-05,
+      "loss": 1.3704,
+      "step": 936
+    },
+    {
+      "epoch": 0.8732525629077353,
+      "grad_norm": 1.776057714316385,
+      "learning_rate": 3.938557128063514e-05,
+      "loss": 1.2316,
+      "step": 937
+    },
+    {
+      "epoch": 0.8741845293569431,
+      "grad_norm": 1.7638179905175073,
+      "learning_rate": 3.9368312046945114e-05,
+      "loss": 1.099,
+      "step": 938
+    },
+    {
+      "epoch": 0.875116495806151,
+      "grad_norm": 2.4127488484654647,
+      "learning_rate": 3.9351052813255095e-05,
+      "loss": 1.2786,
+      "step": 939
+    },
+    {
+      "epoch": 0.8760484622553588,
+      "grad_norm": 2.2615946487392726,
+      "learning_rate": 3.933379357956507e-05,
+      "loss": 1.1274,
+      "step": 940
+    },
+    {
+      "epoch": 0.8769804287045666,
+      "grad_norm": 1.8903153158361183,
+      "learning_rate": 3.9316534345875043e-05,
+      "loss": 1.3563,
+      "step": 941
+    },
+    {
+      "epoch": 0.8779123951537745,
+      "grad_norm": 2.3363110527228366,
+      "learning_rate": 3.929927511218502e-05,
+      "loss": 1.2991,
+      "step": 942
+    },
+    {
+      "epoch": 0.8788443616029823,
+      "grad_norm": 1.4398665766497756,
+      "learning_rate": 3.9282015878495e-05,
+      "loss": 1.0172,
+      "step": 943
+    },
+    {
+      "epoch": 0.8797763280521901,
+      "grad_norm": 1.8343377133188703,
+      "learning_rate": 3.926475664480497e-05,
+      "loss": 1.3713,
+      "step": 944
+    },
+    {
+      "epoch": 0.880708294501398,
+      "grad_norm": 1.878527196059273,
+      "learning_rate": 3.924749741111495e-05,
+      "loss": 1.2599,
+      "step": 945
+    },
+    {
+      "epoch": 0.8816402609506058,
+      "grad_norm": 2.0373353402978744,
+      "learning_rate": 3.923023817742492e-05,
+      "loss": 1.336,
+      "step": 946
+    },
+    {
+      "epoch": 0.8825722273998136,
+      "grad_norm": 1.701608101206123,
+      "learning_rate": 3.92129789437349e-05,
+      "loss": 1.1464,
+      "step": 947
+    },
+    {
+      "epoch": 0.8835041938490215,
+      "grad_norm": 1.8614761125028878,
+      "learning_rate": 3.919571971004488e-05,
+      "loss": 1.2749,
+      "step": 948
+    },
+    {
+      "epoch": 0.8844361602982292,
+      "grad_norm": 1.5604970923343298,
+      "learning_rate": 3.917846047635485e-05,
+      "loss": 1.0379,
+      "step": 949
+    },
+    {
+      "epoch": 0.8853681267474371,
+      "grad_norm": 1.5293886868866886,
+      "learning_rate": 3.916120124266483e-05,
+      "loss": 1.1665,
+      "step": 950
+    },
+    {
+      "epoch": 0.8863000931966449,
+      "grad_norm": 1.6589536440888992,
+      "learning_rate": 3.914394200897481e-05,
+      "loss": 1.2596,
+      "step": 951
+    },
+    {
+      "epoch": 0.8872320596458527,
+      "grad_norm": 1.757147258465629,
+      "learning_rate": 3.912668277528478e-05,
+      "loss": 1.1548,
+      "step": 952
+    },
+    {
+      "epoch": 0.8881640260950606,
+      "grad_norm": 1.6655738134159026,
+      "learning_rate": 3.9109423541594755e-05,
+      "loss": 1.1174,
+      "step": 953
+    },
+    {
+      "epoch": 0.8890959925442684,
+      "grad_norm": 2.0563727213954177,
+      "learning_rate": 3.909216430790473e-05,
+      "loss": 1.355,
+      "step": 954
+    },
+    {
+      "epoch": 0.8900279589934762,
+      "grad_norm": 1.8564706658003545,
+      "learning_rate": 3.9074905074214704e-05,
+      "loss": 1.2652,
+      "step": 955
+    },
+    {
+      "epoch": 0.8909599254426841,
+      "grad_norm": 1.9393694809527278,
+      "learning_rate": 3.905764584052468e-05,
+      "loss": 1.468,
+      "step": 956
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 1.6127691398307509,
+      "learning_rate": 3.904038660683465e-05,
+      "loss": 1.1037,
+      "step": 957
+    },
+    {
+      "epoch": 0.8928238583410997,
+      "grad_norm": 1.8340221689510137,
+      "learning_rate": 3.9023127373144633e-05,
+      "loss": 1.1379,
+      "step": 958
+    },
+    {
+      "epoch": 0.8937558247903076,
+      "grad_norm": 2.0677255200037097,
+      "learning_rate": 3.900586813945461e-05,
+      "loss": 1.395,
+      "step": 959
+    },
+    {
+      "epoch": 0.8946877912395154,
+      "grad_norm": 1.924031928875714,
+      "learning_rate": 3.898860890576458e-05,
+      "loss": 1.0351,
+      "step": 960
+    },
+    {
+      "epoch": 0.8956197576887233,
+      "grad_norm": 1.5979202537100732,
+      "learning_rate": 3.897134967207456e-05,
+      "loss": 1.251,
+      "step": 961
+    },
+    {
+      "epoch": 0.896551724137931,
+      "grad_norm": 1.6585790871491353,
+      "learning_rate": 3.895409043838454e-05,
+      "loss": 1.095,
+      "step": 962
+    },
+    {
+      "epoch": 0.8974836905871388,
+      "grad_norm": 2.12879821915401,
+      "learning_rate": 3.893683120469451e-05,
+      "loss": 1.1605,
+      "step": 963
+    },
+    {
+      "epoch": 0.8984156570363467,
+      "grad_norm": 2.0630635382465763,
+      "learning_rate": 3.8919571971004486e-05,
+      "loss": 1.1858,
+      "step": 964
+    },
+    {
+      "epoch": 0.8993476234855545,
+      "grad_norm": 1.7760984660356112,
+      "learning_rate": 3.890231273731447e-05,
+      "loss": 1.1775,
+      "step": 965
+    },
+    {
+      "epoch": 0.9002795899347623,
+      "grad_norm": 2.3203991397053936,
+      "learning_rate": 3.888505350362444e-05,
+      "loss": 1.4171,
+      "step": 966
+    },
+    {
+      "epoch": 0.9012115563839702,
+      "grad_norm": 1.7806821589823014,
+      "learning_rate": 3.8867794269934416e-05,
+      "loss": 1.2928,
+      "step": 967
+    },
+    {
+      "epoch": 0.902143522833178,
+      "grad_norm": 2.047997839198512,
+      "learning_rate": 3.885053503624439e-05,
+      "loss": 1.1769,
+      "step": 968
+    },
+    {
+      "epoch": 0.9030754892823858,
+      "grad_norm": 1.7629770734904493,
+      "learning_rate": 3.883327580255437e-05,
+      "loss": 1.1568,
+      "step": 969
+    },
+    {
+      "epoch": 0.9040074557315937,
+      "grad_norm": 1.4889559049917973,
+      "learning_rate": 3.8816016568864345e-05,
+      "loss": 0.8944,
+      "step": 970
+    },
+    {
+      "epoch": 0.9049394221808015,
+      "grad_norm": 2.0372245837294236,
+      "learning_rate": 3.879875733517432e-05,
+      "loss": 1.2634,
+      "step": 971
+    },
+    {
+      "epoch": 0.9058713886300093,
+      "grad_norm": 2.1339374701473566,
+      "learning_rate": 3.87814981014843e-05,
+      "loss": 1.4298,
+      "step": 972
+    },
+    {
+      "epoch": 0.9068033550792172,
+      "grad_norm": 1.7342123499766244,
+      "learning_rate": 3.8764238867794275e-05,
+      "loss": 1.1976,
+      "step": 973
+    },
+    {
+      "epoch": 0.907735321528425,
+      "grad_norm": 1.7582538464338293,
+      "learning_rate": 3.874697963410425e-05,
+      "loss": 1.221,
+      "step": 974
+    },
+    {
+      "epoch": 0.9086672879776329,
+      "grad_norm": 1.7479560872081994,
+      "learning_rate": 3.8729720400414224e-05,
+      "loss": 1.2795,
+      "step": 975
+    },
+    {
+      "epoch": 0.9095992544268406,
+      "grad_norm": 2.6023968504986374,
+      "learning_rate": 3.87124611667242e-05,
+      "loss": 1.2428,
+      "step": 976
+    },
+    {
+      "epoch": 0.9105312208760484,
+      "grad_norm": 1.7068857576621073,
+      "learning_rate": 3.869520193303417e-05,
+      "loss": 1.138,
+      "step": 977
+    },
+    {
+      "epoch": 0.9114631873252563,
+      "grad_norm": 2.0325027714642823,
+      "learning_rate": 3.8677942699344146e-05,
+      "loss": 1.2338,
+      "step": 978
+    },
+    {
+      "epoch": 0.9123951537744641,
+      "grad_norm": 1.97425683804585,
+      "learning_rate": 3.866068346565413e-05,
+      "loss": 1.355,
+      "step": 979
+    },
+    {
+      "epoch": 0.9133271202236719,
+      "grad_norm": 2.196834234870384,
+      "learning_rate": 3.86434242319641e-05,
+      "loss": 1.4425,
+      "step": 980
+    },
+    {
+      "epoch": 0.9142590866728798,
+      "grad_norm": 1.862943149843021,
+      "learning_rate": 3.8626164998274076e-05,
+      "loss": 0.9396,
+      "step": 981
+    },
+    {
+      "epoch": 0.9151910531220876,
+      "grad_norm": 1.6590761251894663,
+      "learning_rate": 3.860890576458405e-05,
+      "loss": 1.1387,
+      "step": 982
+    },
+    {
+      "epoch": 0.9161230195712954,
+      "grad_norm": 1.8555595200800876,
+      "learning_rate": 3.859164653089403e-05,
+      "loss": 1.2757,
+      "step": 983
+    },
+    {
+      "epoch": 0.9170549860205033,
+      "grad_norm": 1.6702018671035337,
+      "learning_rate": 3.8574387297204006e-05,
+      "loss": 1.083,
+      "step": 984
+    },
+    {
+      "epoch": 0.9179869524697111,
+      "grad_norm": 1.9916899803720076,
+      "learning_rate": 3.855712806351398e-05,
+      "loss": 1.1529,
+      "step": 985
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 1.8394453061978857,
+      "learning_rate": 3.8539868829823954e-05,
+      "loss": 1.1559,
+      "step": 986
+    },
+    {
+      "epoch": 0.9198508853681268,
+      "grad_norm": 1.8363195235735141,
+      "learning_rate": 3.8522609596133935e-05,
+      "loss": 1.3613,
+      "step": 987
+    },
+    {
+      "epoch": 0.9207828518173345,
+      "grad_norm": 1.7514646956870676,
+      "learning_rate": 3.850535036244391e-05,
+      "loss": 1.1544,
+      "step": 988
+    },
+    {
+      "epoch": 0.9217148182665424,
+      "grad_norm": 1.88411266184549,
+      "learning_rate": 3.8488091128753884e-05,
+      "loss": 1.3973,
+      "step": 989
+    },
+    {
+      "epoch": 0.9226467847157502,
+      "grad_norm": 1.656959862662826,
+      "learning_rate": 3.8470831895063865e-05,
+      "loss": 1.0007,
+      "step": 990
+    },
+    {
+      "epoch": 0.923578751164958,
+      "grad_norm": 1.8182106145158219,
+      "learning_rate": 3.845357266137384e-05,
+      "loss": 1.0853,
+      "step": 991
+    },
+    {
+      "epoch": 0.9245107176141659,
+      "grad_norm": 1.8771101028897115,
+      "learning_rate": 3.8436313427683814e-05,
+      "loss": 1.2561,
+      "step": 992
+    },
+    {
+      "epoch": 0.9254426840633737,
+      "grad_norm": 2.2395150236878956,
+      "learning_rate": 3.841905419399379e-05,
+      "loss": 1.2113,
+      "step": 993
+    },
+    {
+      "epoch": 0.9263746505125815,
+      "grad_norm": 1.8687216365900812,
+      "learning_rate": 3.840179496030377e-05,
+      "loss": 1.4368,
+      "step": 994
+    },
+    {
+      "epoch": 0.9273066169617894,
+      "grad_norm": 1.7658169712973384,
+      "learning_rate": 3.838453572661374e-05,
+      "loss": 1.1768,
+      "step": 995
+    },
+    {
+      "epoch": 0.9282385834109972,
+      "grad_norm": 2.0384793080415067,
+      "learning_rate": 3.836727649292372e-05,
+      "loss": 1.0056,
+      "step": 996
+    },
+    {
+      "epoch": 0.9291705498602051,
+      "grad_norm": 1.7479971181008158,
+      "learning_rate": 3.835001725923369e-05,
+      "loss": 1.0316,
+      "step": 997
+    },
+    {
+      "epoch": 0.9301025163094129,
+      "grad_norm": 1.6878499043843467,
+      "learning_rate": 3.8332758025543666e-05,
+      "loss": 1.1645,
+      "step": 998
+    },
+    {
+      "epoch": 0.9310344827586207,
+      "grad_norm": 1.5825897597315555,
+      "learning_rate": 3.831549879185364e-05,
+      "loss": 1.2877,
+      "step": 999
+    },
+    {
+      "epoch": 0.9319664492078286,
+      "grad_norm": 1.8382209423065623,
+      "learning_rate": 3.8298239558163615e-05,
+      "loss": 1.4222,
+      "step": 1000
+    },
+    {
+      "epoch": 0.9328984156570364,
+      "grad_norm": 2.0977305200714977,
+      "learning_rate": 3.8280980324473596e-05,
+      "loss": 1.2429,
+      "step": 1001
+    },
+    {
+      "epoch": 0.9338303821062441,
+      "grad_norm": 1.685769576512523,
+      "learning_rate": 3.826372109078357e-05,
+      "loss": 1.2019,
+      "step": 1002
+    },
+    {
+      "epoch": 0.934762348555452,
+      "grad_norm": 1.8144937439426314,
+      "learning_rate": 3.8246461857093544e-05,
+      "loss": 1.1162,
+      "step": 1003
+    },
+    {
+      "epoch": 0.9356943150046598,
+      "grad_norm": 1.8347330234332795,
+      "learning_rate": 3.822920262340352e-05,
+      "loss": 1.1449,
+      "step": 1004
+    },
+    {
+      "epoch": 0.9366262814538676,
+      "grad_norm": 1.6090006329275974,
+      "learning_rate": 3.82119433897135e-05,
+      "loss": 0.9298,
+      "step": 1005
+    },
+    {
+      "epoch": 0.9375582479030755,
+      "grad_norm": 1.78409391584463,
+      "learning_rate": 3.8194684156023474e-05,
+      "loss": 1.2539,
+      "step": 1006
+    },
+    {
+      "epoch": 0.9384902143522833,
+      "grad_norm": 1.886879728130471,
+      "learning_rate": 3.817742492233345e-05,
+      "loss": 1.1382,
+      "step": 1007
+    },
+    {
+      "epoch": 0.9394221808014911,
+      "grad_norm": 1.9588067920896384,
+      "learning_rate": 3.816016568864342e-05,
+      "loss": 1.1631,
+      "step": 1008
+    },
+    {
+      "epoch": 0.940354147250699,
+      "grad_norm": 2.114510221424188,
+      "learning_rate": 3.8142906454953404e-05,
+      "loss": 1.1838,
+      "step": 1009
+    },
+    {
+      "epoch": 0.9412861136999068,
+      "grad_norm": 1.7450266543759236,
+      "learning_rate": 3.812564722126338e-05,
+      "loss": 1.2967,
+      "step": 1010
+    },
+    {
+      "epoch": 0.9422180801491147,
+      "grad_norm": 1.6504258384754358,
+      "learning_rate": 3.810838798757335e-05,
+      "loss": 1.0017,
+      "step": 1011
+    },
+    {
+      "epoch": 0.9431500465983225,
+      "grad_norm": 1.600844959985762,
+      "learning_rate": 3.809112875388333e-05,
+      "loss": 0.9369,
+      "step": 1012
+    },
+    {
+      "epoch": 0.9440820130475303,
+      "grad_norm": 1.4547888076247315,
+      "learning_rate": 3.807386952019331e-05,
+      "loss": 1.067,
+      "step": 1013
+    },
+    {
+      "epoch": 0.9450139794967382,
+      "grad_norm": 2.206047258162346,
+      "learning_rate": 3.805661028650328e-05,
+      "loss": 1.4317,
+      "step": 1014
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 1.57671767988248,
+      "learning_rate": 3.8039351052813256e-05,
+      "loss": 1.0047,
+      "step": 1015
+    },
+    {
+      "epoch": 0.9468779123951537,
+      "grad_norm": 1.427284247400093,
+      "learning_rate": 3.802209181912324e-05,
+      "loss": 0.9677,
+      "step": 1016
+    },
+    {
+      "epoch": 0.9478098788443616,
+      "grad_norm": 1.4984737514653648,
+      "learning_rate": 3.800483258543321e-05,
+      "loss": 1.0217,
+      "step": 1017
+    },
+    {
+      "epoch": 0.9487418452935694,
+      "grad_norm": 1.9326372777278753,
+      "learning_rate": 3.7987573351743186e-05,
+      "loss": 0.9454,
+      "step": 1018
+    },
+    {
+      "epoch": 0.9496738117427772,
+      "grad_norm": 1.702686736979134,
+      "learning_rate": 3.797031411805316e-05,
+      "loss": 1.1,
+      "step": 1019
+    },
+    {
+      "epoch": 0.9506057781919851,
+      "grad_norm": 2.10895952155338,
+      "learning_rate": 3.7953054884363134e-05,
+      "loss": 1.1297,
+      "step": 1020
+    },
+    {
+      "epoch": 0.9515377446411929,
+      "grad_norm": 2.301710234898796,
+      "learning_rate": 3.793579565067311e-05,
+      "loss": 1.2564,
+      "step": 1021
+    },
+    {
+      "epoch": 0.9524697110904008,
+      "grad_norm": 1.7250364622907965,
+      "learning_rate": 3.791853641698308e-05,
+      "loss": 1.1774,
+      "step": 1022
+    },
+    {
+      "epoch": 0.9534016775396086,
+      "grad_norm": 2.5244686523758557,
+      "learning_rate": 3.7901277183293064e-05,
+      "loss": 1.5729,
+      "step": 1023
+    },
+    {
+      "epoch": 0.9543336439888164,
+      "grad_norm": 1.9278291234266374,
+      "learning_rate": 3.788401794960304e-05,
+      "loss": 1.1668,
+      "step": 1024
+    },
+    {
+      "epoch": 0.9552656104380243,
+      "grad_norm": 1.9162449603774727,
+      "learning_rate": 3.786675871591301e-05,
+      "loss": 1.26,
+      "step": 1025
+    },
+    {
+      "epoch": 0.9561975768872321,
+      "grad_norm": 1.7441417707972606,
+      "learning_rate": 3.784949948222299e-05,
+      "loss": 1.0245,
+      "step": 1026
+    },
+    {
+      "epoch": 0.9571295433364398,
+      "grad_norm": 1.685466332586829,
+      "learning_rate": 3.783224024853297e-05,
+      "loss": 1.0286,
+      "step": 1027
+    },
+    {
+      "epoch": 0.9580615097856477,
+      "grad_norm": 1.689013023424628,
+      "learning_rate": 3.781498101484294e-05,
+      "loss": 1.2209,
+      "step": 1028
+    },
+    {
+      "epoch": 0.9589934762348555,
+      "grad_norm": 1.4221216809620976,
+      "learning_rate": 3.7797721781152916e-05,
+      "loss": 0.8597,
+      "step": 1029
+    },
+    {
+      "epoch": 0.9599254426840633,
+      "grad_norm": 1.8252181551879918,
+      "learning_rate": 3.77804625474629e-05,
+      "loss": 0.9017,
+      "step": 1030
+    },
+    {
+      "epoch": 0.9608574091332712,
+      "grad_norm": 2.1372010254879323,
+      "learning_rate": 3.776320331377287e-05,
+      "loss": 1.1521,
+      "step": 1031
+    },
+    {
+      "epoch": 0.961789375582479,
+      "grad_norm": 1.5415888774704762,
+      "learning_rate": 3.7745944080082846e-05,
+      "loss": 1.0921,
+      "step": 1032
+    },
+    {
+      "epoch": 0.9627213420316869,
+      "grad_norm": 1.960632969914242,
+      "learning_rate": 3.772868484639282e-05,
+      "loss": 1.2761,
+      "step": 1033
+    },
+    {
+      "epoch": 0.9636533084808947,
+      "grad_norm": 1.936057924801292,
+      "learning_rate": 3.77114256127028e-05,
+      "loss": 1.1782,
+      "step": 1034
+    },
+    {
+      "epoch": 0.9645852749301025,
+      "grad_norm": 1.8487437651829857,
+      "learning_rate": 3.7694166379012776e-05,
+      "loss": 1.1489,
+      "step": 1035
+    },
+    {
+      "epoch": 0.9655172413793104,
+      "grad_norm": 1.9537365363979553,
+      "learning_rate": 3.767690714532275e-05,
+      "loss": 1.0902,
+      "step": 1036
+    },
+    {
+      "epoch": 0.9664492078285182,
+      "grad_norm": 1.6583978514383562,
+      "learning_rate": 3.7659647911632724e-05,
+      "loss": 1.0222,
+      "step": 1037
+    },
+    {
+      "epoch": 0.967381174277726,
+      "grad_norm": 1.8985353459889547,
+      "learning_rate": 3.7642388677942705e-05,
+      "loss": 1.2951,
+      "step": 1038
+    },
+    {
+      "epoch": 0.9683131407269339,
+      "grad_norm": 1.9168885555562654,
+      "learning_rate": 3.762512944425268e-05,
+      "loss": 1.0539,
+      "step": 1039
+    },
+    {
+      "epoch": 0.9692451071761417,
+      "grad_norm": 1.9967019898033327,
+      "learning_rate": 3.7607870210562654e-05,
+      "loss": 1.4437,
+      "step": 1040
+    },
+    {
+      "epoch": 0.9701770736253494,
+      "grad_norm": 1.726008233495271,
+      "learning_rate": 3.759061097687263e-05,
+      "loss": 1.1906,
+      "step": 1041
+    },
+    {
+      "epoch": 0.9711090400745573,
+      "grad_norm": 2.020487307672184,
+      "learning_rate": 3.75733517431826e-05,
+      "loss": 1.1436,
+      "step": 1042
+    },
+    {
+      "epoch": 0.9720410065237651,
+      "grad_norm": 2.180094009768937,
+      "learning_rate": 3.755609250949258e-05,
+      "loss": 1.0945,
+      "step": 1043
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 1.6286458237157577,
+      "learning_rate": 3.753883327580255e-05,
+      "loss": 1.0392,
+      "step": 1044
+    },
+    {
+      "epoch": 0.9739049394221808,
+      "grad_norm": 1.9636035369081377,
+      "learning_rate": 3.752157404211253e-05,
+      "loss": 1.2157,
+      "step": 1045
+    },
+    {
+      "epoch": 0.9748369058713886,
+      "grad_norm": 2.1860581565325,
+      "learning_rate": 3.7504314808422507e-05,
+      "loss": 1.3255,
+      "step": 1046
+    },
+    {
+      "epoch": 0.9757688723205965,
+      "grad_norm": 2.105790814037647,
+      "learning_rate": 3.748705557473248e-05,
+      "loss": 1.3569,
+      "step": 1047
+    },
+    {
+      "epoch": 0.9767008387698043,
+      "grad_norm": 1.6928546501358226,
+      "learning_rate": 3.7469796341042455e-05,
+      "loss": 1.0614,
+      "step": 1048
+    },
+    {
+      "epoch": 0.9776328052190121,
+      "grad_norm": 1.477851954240227,
+      "learning_rate": 3.7452537107352436e-05,
+      "loss": 0.8097,
+      "step": 1049
+    },
+    {
+      "epoch": 0.97856477166822,
+      "grad_norm": 2.033720836339729,
+      "learning_rate": 3.743527787366241e-05,
+      "loss": 1.2702,
+      "step": 1050
+    },
+    {
+      "epoch": 0.9794967381174278,
+      "grad_norm": 1.8830111093432438,
+      "learning_rate": 3.7418018639972385e-05,
+      "loss": 1.0035,
+      "step": 1051
+    },
+    {
+      "epoch": 0.9804287045666356,
+      "grad_norm": 1.832623372531838,
+      "learning_rate": 3.7400759406282366e-05,
+      "loss": 1.1366,
+      "step": 1052
+    },
+    {
+      "epoch": 0.9813606710158435,
+      "grad_norm": 2.42589799290065,
+      "learning_rate": 3.738350017259234e-05,
+      "loss": 1.2666,
+      "step": 1053
+    },
+    {
+      "epoch": 0.9822926374650512,
+      "grad_norm": 2.033151690500238,
+      "learning_rate": 3.7366240938902314e-05,
+      "loss": 1.2804,
+      "step": 1054
+    },
+    {
+      "epoch": 0.983224603914259,
+      "grad_norm": 1.8227413791151923,
+      "learning_rate": 3.734898170521229e-05,
+      "loss": 1.1931,
+      "step": 1055
+    },
+    {
+      "epoch": 0.9841565703634669,
+      "grad_norm": 2.018328559498595,
+      "learning_rate": 3.733172247152227e-05,
+      "loss": 1.0375,
+      "step": 1056
+    },
+    {
+      "epoch": 0.9850885368126747,
+      "grad_norm": 1.5623885886606201,
+      "learning_rate": 3.7314463237832244e-05,
+      "loss": 1.1019,
+      "step": 1057
+    },
+    {
+      "epoch": 0.9860205032618826,
+      "grad_norm": 1.8895045010094922,
+      "learning_rate": 3.729720400414222e-05,
+      "loss": 1.0188,
+      "step": 1058
+    },
+    {
+      "epoch": 0.9869524697110904,
+      "grad_norm": 1.5327718492334028,
+      "learning_rate": 3.72799447704522e-05,
+      "loss": 1.1045,
+      "step": 1059
+    },
+    {
+      "epoch": 0.9878844361602982,
+      "grad_norm": 1.9179845126778212,
+      "learning_rate": 3.7262685536762174e-05,
+      "loss": 1.1705,
+      "step": 1060
+    },
+    {
+      "epoch": 0.9888164026095061,
+      "grad_norm": 1.810809732155172,
+      "learning_rate": 3.724542630307215e-05,
+      "loss": 0.9913,
+      "step": 1061
+    },
+    {
+      "epoch": 0.9897483690587139,
+      "grad_norm": 2.1590141034184898,
+      "learning_rate": 3.722816706938212e-05,
+      "loss": 1.4165,
+      "step": 1062
+    },
+    {
+      "epoch": 0.9906803355079217,
+      "grad_norm": 1.6415441710676344,
+      "learning_rate": 3.7210907835692097e-05,
+      "loss": 1.0459,
+      "step": 1063
+    },
+    {
+      "epoch": 0.9916123019571296,
+      "grad_norm": 1.7420813114031501,
+      "learning_rate": 3.719364860200207e-05,
+      "loss": 1.1731,
+      "step": 1064
+    },
+    {
+      "epoch": 0.9925442684063374,
+      "grad_norm": 1.7703932646185374,
+      "learning_rate": 3.7176389368312045e-05,
+      "loss": 1.1108,
+      "step": 1065
+    },
+    {
+      "epoch": 0.9934762348555451,
+      "grad_norm": 1.606724699623895,
+      "learning_rate": 3.715913013462202e-05,
+      "loss": 1.2084,
+      "step": 1066
+    },
+    {
+      "epoch": 0.994408201304753,
+      "grad_norm": 1.7905535443339085,
+      "learning_rate": 3.7141870900932e-05,
+      "loss": 1.2063,
+      "step": 1067
+    },
+    {
+      "epoch": 0.9953401677539608,
+      "grad_norm": 1.493095742590078,
+      "learning_rate": 3.7124611667241975e-05,
+      "loss": 1.03,
+      "step": 1068
+    },
+    {
+      "epoch": 0.9962721342031687,
+      "grad_norm": 1.899806041130243,
+      "learning_rate": 3.710735243355195e-05,
+      "loss": 0.9744,
+      "step": 1069
+    },
+    {
+      "epoch": 0.9972041006523765,
+      "grad_norm": 1.687929326535403,
+      "learning_rate": 3.709009319986193e-05,
+      "loss": 1.1419,
+      "step": 1070
+    },
+    {
+      "epoch": 0.9981360671015843,
+      "grad_norm": 1.8103817858680404,
+      "learning_rate": 3.7072833966171904e-05,
+      "loss": 1.0608,
+      "step": 1071
+    },
+    {
+      "epoch": 0.9990680335507922,
+      "grad_norm": 1.6234190363460799,
+      "learning_rate": 3.705557473248188e-05,
+      "loss": 1.0254,
+      "step": 1072
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.719782056515951,
+      "learning_rate": 3.703831549879185e-05,
+      "loss": 0.9162,
+      "step": 1073
+    },
+    {
+      "epoch": 1.0009319664492078,
+      "grad_norm": 1.3649061714051154,
+      "learning_rate": 3.7021056265101834e-05,
+      "loss": 0.7294,
+      "step": 1074
+    },
+    {
+      "epoch": 1.0018639328984156,
+      "grad_norm": 1.7239287693065681,
+      "learning_rate": 3.700379703141181e-05,
+      "loss": 0.7831,
+      "step": 1075
+    },
+    {
+      "epoch": 1.0027958993476236,
+      "grad_norm": 1.183999306103034,
+      "learning_rate": 3.698653779772178e-05,
+      "loss": 0.6325,
+      "step": 1076
+    },
+    {
+      "epoch": 1.0037278657968314,
+      "grad_norm": 1.5506101473714162,
+      "learning_rate": 3.696927856403176e-05,
+      "loss": 0.8487,
+      "step": 1077
+    },
+    {
+      "epoch": 1.0046598322460392,
+      "grad_norm": 1.1407256229007832,
+      "learning_rate": 3.695201933034174e-05,
+      "loss": 0.7652,
+      "step": 1078
+    },
+    {
+      "epoch": 1.005591798695247,
+      "grad_norm": 1.0174557708733998,
+      "learning_rate": 3.693476009665171e-05,
+      "loss": 0.6077,
+      "step": 1079
+    },
+    {
+      "epoch": 1.0065237651444547,
+      "grad_norm": 1.4347835760684338,
+      "learning_rate": 3.6917500862961687e-05,
+      "loss": 0.737,
+      "step": 1080
+    },
+    {
+      "epoch": 1.0074557315936625,
+      "grad_norm": 1.5159358402286476,
+      "learning_rate": 3.690024162927167e-05,
+      "loss": 0.7767,
+      "step": 1081
+    },
+    {
+      "epoch": 1.0083876980428705,
+      "grad_norm": 1.5580344961809853,
+      "learning_rate": 3.688298239558164e-05,
+      "loss": 0.8459,
+      "step": 1082
+    },
+    {
+      "epoch": 1.0093196644920783,
+      "grad_norm": 1.5875688730972903,
+      "learning_rate": 3.6865723161891616e-05,
+      "loss": 0.8597,
+      "step": 1083
+    },
+    {
+      "epoch": 1.0102516309412861,
+      "grad_norm": 1.5986681688352933,
+      "learning_rate": 3.684846392820159e-05,
+      "loss": 0.7752,
+      "step": 1084
+    },
+    {
+      "epoch": 1.011183597390494,
+      "grad_norm": 1.5398778396554993,
+      "learning_rate": 3.6831204694511565e-05,
+      "loss": 0.8371,
+      "step": 1085
+    },
+    {
+      "epoch": 1.0121155638397017,
+      "grad_norm": 1.023018702911238,
+      "learning_rate": 3.681394546082154e-05,
+      "loss": 0.6048,
+      "step": 1086
+    },
+    {
+      "epoch": 1.0130475302889097,
+      "grad_norm": 1.2361953693553913,
+      "learning_rate": 3.6796686227131513e-05,
+      "loss": 0.6787,
+      "step": 1087
+    },
+    {
+      "epoch": 1.0139794967381175,
+      "grad_norm": 1.3282282555266092,
+      "learning_rate": 3.677942699344149e-05,
+      "loss": 0.8068,
+      "step": 1088
+    },
+    {
+      "epoch": 1.0149114631873253,
+      "grad_norm": 1.2305226091755068,
+      "learning_rate": 3.676216775975147e-05,
+      "loss": 0.7153,
+      "step": 1089
+    },
+    {
+      "epoch": 1.015843429636533,
+      "grad_norm": 1.6159412829355664,
+      "learning_rate": 3.674490852606144e-05,
+      "loss": 0.9165,
+      "step": 1090
+    },
+    {
+      "epoch": 1.0167753960857409,
+      "grad_norm": 1.6477025363818256,
+      "learning_rate": 3.672764929237142e-05,
+      "loss": 0.7276,
+      "step": 1091
+    },
+    {
+      "epoch": 1.0177073625349486,
+      "grad_norm": 1.3129121554565353,
+      "learning_rate": 3.67103900586814e-05,
+      "loss": 0.6889,
+      "step": 1092
+    },
+    {
+      "epoch": 1.0186393289841567,
+      "grad_norm": 1.2551388003144526,
+      "learning_rate": 3.669313082499137e-05,
+      "loss": 0.6933,
+      "step": 1093
+    },
+    {
+      "epoch": 1.0195712954333644,
+      "grad_norm": 1.7144552548478829,
+      "learning_rate": 3.667587159130135e-05,
+      "loss": 0.8115,
+      "step": 1094
+    },
+    {
+      "epoch": 1.0205032618825722,
+      "grad_norm": 1.536820606819437,
+      "learning_rate": 3.665861235761132e-05,
+      "loss": 0.7578,
+      "step": 1095
+    },
+    {
+      "epoch": 1.02143522833178,
+      "grad_norm": 1.2897613706976931,
+      "learning_rate": 3.66413531239213e-05,
+      "loss": 0.729,
+      "step": 1096
+    },
+    {
+      "epoch": 1.0223671947809878,
+      "grad_norm": 1.1853789130203745,
+      "learning_rate": 3.662409389023128e-05,
+      "loss": 0.7259,
+      "step": 1097
+    },
+    {
+      "epoch": 1.0232991612301958,
+      "grad_norm": 1.078529892147532,
+      "learning_rate": 3.660683465654125e-05,
+      "loss": 0.7253,
+      "step": 1098
+    },
+    {
+      "epoch": 1.0242311276794036,
+      "grad_norm": 1.5390571436192029,
+      "learning_rate": 3.6589575422851225e-05,
+      "loss": 0.7323,
+      "step": 1099
+    },
+    {
+      "epoch": 1.0251630941286114,
+      "grad_norm": 1.0756204149326098,
+      "learning_rate": 3.6572316189161206e-05,
+      "loss": 0.5986,
+      "step": 1100
+    },
+    {
+      "epoch": 1.0260950605778192,
+      "grad_norm": 1.3019089080829598,
+      "learning_rate": 3.655505695547118e-05,
+      "loss": 0.7095,
+      "step": 1101
+    },
+    {
+      "epoch": 1.027027027027027,
+      "grad_norm": 1.7254062433280677,
+      "learning_rate": 3.6537797721781155e-05,
+      "loss": 0.8866,
+      "step": 1102
+    },
+    {
+      "epoch": 1.0279589934762348,
+      "grad_norm": 1.2465709681907675,
+      "learning_rate": 3.6520538488091136e-05,
+      "loss": 0.7078,
+      "step": 1103
+    },
+    {
+      "epoch": 1.0288909599254428,
+      "grad_norm": 1.1588895852661276,
+      "learning_rate": 3.650327925440111e-05,
+      "loss": 0.6415,
+      "step": 1104
+    },
+    {
+      "epoch": 1.0298229263746506,
+      "grad_norm": 1.5317222718629904,
+      "learning_rate": 3.6486020020711085e-05,
+      "loss": 0.7848,
+      "step": 1105
+    },
+    {
+      "epoch": 1.0307548928238583,
+      "grad_norm": 1.315466175544975,
+      "learning_rate": 3.646876078702106e-05,
+      "loss": 0.8264,
+      "step": 1106
+    },
+    {
+      "epoch": 1.0316868592730661,
+      "grad_norm": 1.6480252144954404,
+      "learning_rate": 3.645150155333103e-05,
+      "loss": 0.8644,
+      "step": 1107
+    },
+    {
+      "epoch": 1.032618825722274,
+      "grad_norm": 1.6147400362104876,
+      "learning_rate": 3.643424231964101e-05,
+      "loss": 0.8293,
+      "step": 1108
+    },
+    {
+      "epoch": 1.0335507921714817,
+      "grad_norm": 1.865412952486786,
+      "learning_rate": 3.641698308595098e-05,
+      "loss": 0.7669,
+      "step": 1109
+    },
+    {
+      "epoch": 1.0344827586206897,
+      "grad_norm": 1.3441955854777554,
+      "learning_rate": 3.639972385226096e-05,
+      "loss": 0.7125,
+      "step": 1110
+    },
+    {
+      "epoch": 1.0354147250698975,
+      "grad_norm": 1.450279229546568,
+      "learning_rate": 3.638246461857094e-05,
+      "loss": 0.8024,
+      "step": 1111
+    },
+    {
+      "epoch": 1.0363466915191053,
+      "grad_norm": 1.3024003029692752,
+      "learning_rate": 3.636520538488091e-05,
+      "loss": 0.6845,
+      "step": 1112
+    },
+    {
+      "epoch": 1.037278657968313,
+      "grad_norm": 1.5036259258373472,
+      "learning_rate": 3.6347946151190886e-05,
+      "loss": 0.6984,
+      "step": 1113
+    },
+    {
+      "epoch": 1.0382106244175209,
+      "grad_norm": 1.5461701275542759,
+      "learning_rate": 3.633068691750087e-05,
+      "loss": 0.7051,
+      "step": 1114
+    },
+    {
+      "epoch": 1.0391425908667289,
+      "grad_norm": 1.8196020592790683,
+      "learning_rate": 3.631342768381084e-05,
+      "loss": 0.9771,
+      "step": 1115
+    },
+    {
+      "epoch": 1.0400745573159367,
+      "grad_norm": 1.3445541098807887,
+      "learning_rate": 3.6296168450120815e-05,
+      "loss": 0.631,
+      "step": 1116
+    },
+    {
+      "epoch": 1.0410065237651445,
+      "grad_norm": 1.403250153737369,
+      "learning_rate": 3.627890921643079e-05,
+      "loss": 0.7208,
+      "step": 1117
+    },
+    {
+      "epoch": 1.0419384902143523,
+      "grad_norm": 1.6369513993465379,
+      "learning_rate": 3.626164998274077e-05,
+      "loss": 0.7475,
+      "step": 1118
+    },
+    {
+      "epoch": 1.04287045666356,
+      "grad_norm": 1.2110772387486413,
+      "learning_rate": 3.6244390749050745e-05,
+      "loss": 0.6397,
+      "step": 1119
+    },
+    {
+      "epoch": 1.0438024231127678,
+      "grad_norm": 1.472689965938432,
+      "learning_rate": 3.622713151536072e-05,
+      "loss": 0.7798,
+      "step": 1120
+    },
+    {
+      "epoch": 1.0447343895619758,
+      "grad_norm": 1.4185316882859869,
+      "learning_rate": 3.62098722816707e-05,
+      "loss": 0.6951,
+      "step": 1121
+    },
+    {
+      "epoch": 1.0456663560111836,
+      "grad_norm": 1.33321574982954,
+      "learning_rate": 3.6192613047980675e-05,
+      "loss": 0.7337,
+      "step": 1122
+    },
+    {
+      "epoch": 1.0465983224603914,
+      "grad_norm": 1.4906532942889998,
+      "learning_rate": 3.617535381429065e-05,
+      "loss": 0.683,
+      "step": 1123
+    },
+    {
+      "epoch": 1.0475302889095992,
+      "grad_norm": 1.632062303219311,
+      "learning_rate": 3.615809458060062e-05,
+      "loss": 0.808,
+      "step": 1124
+    },
+    {
+      "epoch": 1.048462255358807,
+      "grad_norm": 1.5837861463219254,
+      "learning_rate": 3.6140835346910604e-05,
+      "loss": 0.7883,
+      "step": 1125
+    },
+    {
+      "epoch": 1.049394221808015,
+      "grad_norm": 1.4288475628456776,
+      "learning_rate": 3.612357611322058e-05,
+      "loss": 0.755,
+      "step": 1126
+    },
+    {
+      "epoch": 1.0503261882572228,
+      "grad_norm": 1.4502594856157338,
+      "learning_rate": 3.610631687953055e-05,
+      "loss": 0.7142,
+      "step": 1127
+    },
+    {
+      "epoch": 1.0512581547064306,
+      "grad_norm": 1.7759633483940507,
+      "learning_rate": 3.608905764584053e-05,
+      "loss": 0.7757,
+      "step": 1128
+    },
+    {
+      "epoch": 1.0521901211556384,
+      "grad_norm": 1.4893882535737215,
+      "learning_rate": 3.60717984121505e-05,
+      "loss": 0.722,
+      "step": 1129
+    },
+    {
+      "epoch": 1.0531220876048462,
+      "grad_norm": 1.0583914643553611,
+      "learning_rate": 3.6054539178460476e-05,
+      "loss": 0.6478,
+      "step": 1130
+    },
+    {
+      "epoch": 1.054054054054054,
+      "grad_norm": 1.4275920065216865,
+      "learning_rate": 3.603727994477045e-05,
+      "loss": 0.7151,
+      "step": 1131
+    },
+    {
+      "epoch": 1.054986020503262,
+      "grad_norm": 1.1482226469866856,
+      "learning_rate": 3.602002071108043e-05,
+      "loss": 0.6781,
+      "step": 1132
+    },
+    {
+      "epoch": 1.0559179869524697,
+      "grad_norm": 1.6461837673098925,
+      "learning_rate": 3.6002761477390405e-05,
+      "loss": 0.7536,
+      "step": 1133
+    },
+    {
+      "epoch": 1.0568499534016775,
+      "grad_norm": 1.1735324790427848,
+      "learning_rate": 3.598550224370038e-05,
+      "loss": 0.7512,
+      "step": 1134
+    },
+    {
+      "epoch": 1.0577819198508853,
+      "grad_norm": 1.490857840189612,
+      "learning_rate": 3.5968243010010354e-05,
+      "loss": 0.8499,
+      "step": 1135
+    },
+    {
+      "epoch": 1.058713886300093,
+      "grad_norm": 1.422188118028489,
+      "learning_rate": 3.5950983776320335e-05,
+      "loss": 0.8199,
+      "step": 1136
+    },
+    {
+      "epoch": 1.0596458527493011,
+      "grad_norm": 1.4612498620775867,
+      "learning_rate": 3.593372454263031e-05,
+      "loss": 0.7265,
+      "step": 1137
+    },
+    {
+      "epoch": 1.060577819198509,
+      "grad_norm": 1.0378680066247126,
+      "learning_rate": 3.5916465308940284e-05,
+      "loss": 0.6357,
+      "step": 1138
+    },
+    {
+      "epoch": 1.0615097856477167,
+      "grad_norm": 1.4080206116397302,
+      "learning_rate": 3.589920607525026e-05,
+      "loss": 0.7434,
+      "step": 1139
+    },
+    {
+      "epoch": 1.0624417520969245,
+      "grad_norm": 1.588490178563853,
+      "learning_rate": 3.588194684156024e-05,
+      "loss": 0.7828,
+      "step": 1140
+    },
+    {
+      "epoch": 1.0633737185461323,
+      "grad_norm": 0.9745769703075206,
+      "learning_rate": 3.586468760787021e-05,
+      "loss": 0.5924,
+      "step": 1141
+    },
+    {
+      "epoch": 1.06430568499534,
+      "grad_norm": 1.7659603284678564,
+      "learning_rate": 3.584742837418019e-05,
+      "loss": 0.7447,
+      "step": 1142
+    },
+    {
+      "epoch": 1.065237651444548,
+      "grad_norm": 1.1389213565086822,
+      "learning_rate": 3.583016914049017e-05,
+      "loss": 0.6848,
+      "step": 1143
+    },
+    {
+      "epoch": 1.0661696178937559,
+      "grad_norm": 1.1750795772662783,
+      "learning_rate": 3.581290990680014e-05,
+      "loss": 0.6331,
+      "step": 1144
+    },
+    {
+      "epoch": 1.0671015843429636,
+      "grad_norm": 1.257244877094481,
+      "learning_rate": 3.579565067311012e-05,
+      "loss": 0.7433,
+      "step": 1145
+    },
+    {
+      "epoch": 1.0680335507921714,
+      "grad_norm": 1.748441534984443,
+      "learning_rate": 3.577839143942009e-05,
+      "loss": 0.715,
+      "step": 1146
+    },
+    {
+      "epoch": 1.0689655172413792,
+      "grad_norm": 2.3172064997886705,
+      "learning_rate": 3.5761132205730066e-05,
+      "loss": 0.8705,
+      "step": 1147
+    },
+    {
+      "epoch": 1.0698974836905872,
+      "grad_norm": 1.0594903408914296,
+      "learning_rate": 3.574387297204005e-05,
+      "loss": 0.5939,
+      "step": 1148
+    },
+    {
+      "epoch": 1.070829450139795,
+      "grad_norm": 1.0372451032061418,
+      "learning_rate": 3.572661373835002e-05,
+      "loss": 0.6791,
+      "step": 1149
+    },
+    {
+      "epoch": 1.0717614165890028,
+      "grad_norm": 1.0350826173692014,
+      "learning_rate": 3.5709354504659995e-05,
+      "loss": 0.6207,
+      "step": 1150
+    },
+    {
+      "epoch": 1.0726933830382106,
+      "grad_norm": 1.8519783817756204,
+      "learning_rate": 3.569209527096997e-05,
+      "loss": 0.9722,
+      "step": 1151
+    },
+    {
+      "epoch": 1.0736253494874184,
+      "grad_norm": 1.281931882471073,
+      "learning_rate": 3.5674836037279944e-05,
+      "loss": 0.7511,
+      "step": 1152
+    },
+    {
+      "epoch": 1.0745573159366262,
+      "grad_norm": 1.3578421291802993,
+      "learning_rate": 3.565757680358992e-05,
+      "loss": 0.7461,
+      "step": 1153
+    },
+    {
+      "epoch": 1.0754892823858342,
+      "grad_norm": 1.5713246080893546,
+      "learning_rate": 3.56403175698999e-05,
+      "loss": 0.7372,
+      "step": 1154
+    },
+    {
+      "epoch": 1.076421248835042,
+      "grad_norm": 1.5146332513964957,
+      "learning_rate": 3.5623058336209874e-05,
+      "loss": 0.8491,
+      "step": 1155
+    },
+    {
+      "epoch": 1.0773532152842498,
+      "grad_norm": 1.5991119384117254,
+      "learning_rate": 3.560579910251985e-05,
+      "loss": 0.7106,
+      "step": 1156
+    },
+    {
+      "epoch": 1.0782851817334576,
+      "grad_norm": 1.4388930868823224,
+      "learning_rate": 3.558853986882982e-05,
+      "loss": 0.7468,
+      "step": 1157
+    },
+    {
+      "epoch": 1.0792171481826653,
+      "grad_norm": 1.6128240272725005,
+      "learning_rate": 3.55712806351398e-05,
+      "loss": 0.7397,
+      "step": 1158
+    },
+    {
+      "epoch": 1.0801491146318734,
+      "grad_norm": 1.254692950036553,
+      "learning_rate": 3.555402140144978e-05,
+      "loss": 0.6782,
+      "step": 1159
+    },
+    {
+      "epoch": 1.0810810810810811,
+      "grad_norm": 1.9447029946931342,
+      "learning_rate": 3.553676216775975e-05,
+      "loss": 0.8325,
+      "step": 1160
+    },
+    {
+      "epoch": 1.082013047530289,
+      "grad_norm": 1.4390692277406967,
+      "learning_rate": 3.551950293406973e-05,
+      "loss": 0.7152,
+      "step": 1161
+    },
+    {
+      "epoch": 1.0829450139794967,
+      "grad_norm": 1.4551233169168243,
+      "learning_rate": 3.550224370037971e-05,
+      "loss": 0.6998,
+      "step": 1162
+    },
+    {
+      "epoch": 1.0838769804287045,
+      "grad_norm": 1.9851651018792267,
+      "learning_rate": 3.548498446668968e-05,
+      "loss": 0.9881,
+      "step": 1163
+    },
+    {
+      "epoch": 1.0848089468779123,
+      "grad_norm": 1.4564675075482842,
+      "learning_rate": 3.5467725232999656e-05,
+      "loss": 0.7775,
+      "step": 1164
+    },
+    {
+      "epoch": 1.0857409133271203,
+      "grad_norm": 1.8433830485212255,
+      "learning_rate": 3.545046599930964e-05,
+      "loss": 0.8179,
+      "step": 1165
+    },
+    {
+      "epoch": 1.086672879776328,
+      "grad_norm": 1.1643374057117528,
+      "learning_rate": 3.543320676561961e-05,
+      "loss": 0.7712,
+      "step": 1166
+    },
+    {
+      "epoch": 1.0876048462255359,
+      "grad_norm": 1.153280875121964,
+      "learning_rate": 3.5415947531929585e-05,
+      "loss": 0.724,
+      "step": 1167
+    },
+    {
+      "epoch": 1.0885368126747437,
+      "grad_norm": 1.6764135718923063,
+      "learning_rate": 3.539868829823956e-05,
+      "loss": 0.8715,
+      "step": 1168
+    },
+    {
+      "epoch": 1.0894687791239515,
+      "grad_norm": 1.5567307783661648,
+      "learning_rate": 3.5381429064549534e-05,
+      "loss": 0.8653,
+      "step": 1169
+    },
+    {
+      "epoch": 1.0904007455731595,
+      "grad_norm": 1.2169329455276772,
+      "learning_rate": 3.536416983085951e-05,
+      "loss": 0.7602,
+      "step": 1170
+    },
+    {
+      "epoch": 1.0913327120223673,
+      "grad_norm": 1.7704996285444032,
+      "learning_rate": 3.534691059716948e-05,
+      "loss": 0.8641,
+      "step": 1171
+    },
+    {
+      "epoch": 1.092264678471575,
+      "grad_norm": 1.2265437203677807,
+      "learning_rate": 3.5329651363479464e-05,
+      "loss": 0.6821,
+      "step": 1172
+    },
+    {
+      "epoch": 1.0931966449207828,
+      "grad_norm": 1.3620187773065022,
+      "learning_rate": 3.531239212978944e-05,
+      "loss": 0.7253,
+      "step": 1173
+    },
+    {
+      "epoch": 1.0941286113699906,
+      "grad_norm": 1.262692331845872,
+      "learning_rate": 3.529513289609941e-05,
+      "loss": 0.7021,
+      "step": 1174
+    },
+    {
+      "epoch": 1.0950605778191984,
+      "grad_norm": 1.4444478828571023,
+      "learning_rate": 3.5277873662409386e-05,
+      "loss": 0.6919,
+      "step": 1175
+    },
+    {
+      "epoch": 1.0959925442684064,
+      "grad_norm": 1.2866955519636065,
+      "learning_rate": 3.526061442871937e-05,
+      "loss": 0.7446,
+      "step": 1176
+    },
+    {
+      "epoch": 1.0969245107176142,
+      "grad_norm": 1.6512237777170136,
+      "learning_rate": 3.524335519502934e-05,
+      "loss": 0.8216,
+      "step": 1177
+    },
+    {
+      "epoch": 1.097856477166822,
+      "grad_norm": 1.3239013510046813,
+      "learning_rate": 3.5226095961339316e-05,
+      "loss": 0.7652,
+      "step": 1178
+    },
+    {
+      "epoch": 1.0987884436160298,
+      "grad_norm": 1.4700112426073026,
+      "learning_rate": 3.520883672764929e-05,
+      "loss": 0.7407,
+      "step": 1179
+    },
+    {
+      "epoch": 1.0997204100652376,
+      "grad_norm": 1.4654083741098811,
+      "learning_rate": 3.519157749395927e-05,
+      "loss": 0.7702,
+      "step": 1180
+    },
+    {
+      "epoch": 1.1006523765144456,
+      "grad_norm": 1.339445808436511,
+      "learning_rate": 3.5174318260269246e-05,
+      "loss": 0.6688,
+      "step": 1181
+    },
+    {
+      "epoch": 1.1015843429636534,
+      "grad_norm": 2.986244987715836,
+      "learning_rate": 3.515705902657922e-05,
+      "loss": 0.8583,
+      "step": 1182
+    },
+    {
+      "epoch": 1.1025163094128612,
+      "grad_norm": 1.359259145192605,
+      "learning_rate": 3.51397997928892e-05,
+      "loss": 0.6748,
+      "step": 1183
+    },
+    {
+      "epoch": 1.103448275862069,
+      "grad_norm": 1.5900795954236393,
+      "learning_rate": 3.5122540559199175e-05,
+      "loss": 0.6896,
+      "step": 1184
+    },
+    {
+      "epoch": 1.1043802423112767,
+      "grad_norm": 1.6760105563294918,
+      "learning_rate": 3.510528132550915e-05,
+      "loss": 0.8926,
+      "step": 1185
+    },
+    {
+      "epoch": 1.1053122087604845,
+      "grad_norm": 1.4657661034485816,
+      "learning_rate": 3.5088022091819124e-05,
+      "loss": 0.8155,
+      "step": 1186
+    },
+    {
+      "epoch": 1.1062441752096925,
+      "grad_norm": 1.296335261500405,
+      "learning_rate": 3.5070762858129105e-05,
+      "loss": 0.7232,
+      "step": 1187
+    },
+    {
+      "epoch": 1.1071761416589003,
+      "grad_norm": 1.521408963112937,
+      "learning_rate": 3.505350362443908e-05,
+      "loss": 0.7086,
+      "step": 1188
+    },
+    {
+      "epoch": 1.1081081081081081,
+      "grad_norm": 1.4746247237709627,
+      "learning_rate": 3.5036244390749054e-05,
+      "loss": 0.765,
+      "step": 1189
+    },
+    {
+      "epoch": 1.109040074557316,
+      "grad_norm": 1.4061430956161045,
+      "learning_rate": 3.501898515705903e-05,
+      "loss": 0.7106,
+      "step": 1190
+    },
+    {
+      "epoch": 1.1099720410065237,
+      "grad_norm": 1.6874180931927614,
+      "learning_rate": 3.5001725923369e-05,
+      "loss": 0.729,
+      "step": 1191
+    },
+    {
+      "epoch": 1.1109040074557317,
+      "grad_norm": 1.452523554504755,
+      "learning_rate": 3.4984466689678977e-05,
+      "loss": 0.8508,
+      "step": 1192
+    },
+    {
+      "epoch": 1.1118359739049395,
+      "grad_norm": 2.3028927344722137,
+      "learning_rate": 3.496720745598895e-05,
+      "loss": 0.797,
+      "step": 1193
+    },
+    {
+      "epoch": 1.1127679403541473,
+      "grad_norm": 2.0101613230287283,
+      "learning_rate": 3.494994822229893e-05,
+      "loss": 0.9202,
+      "step": 1194
+    },
+    {
+      "epoch": 1.113699906803355,
+      "grad_norm": 1.2132476833045358,
+      "learning_rate": 3.4932688988608906e-05,
+      "loss": 0.7292,
+      "step": 1195
+    },
+    {
+      "epoch": 1.1146318732525629,
+      "grad_norm": 1.2963091179216977,
+      "learning_rate": 3.491542975491888e-05,
+      "loss": 0.7736,
+      "step": 1196
+    },
+    {
+      "epoch": 1.1155638397017706,
+      "grad_norm": 1.3682634341456557,
+      "learning_rate": 3.4898170521228855e-05,
+      "loss": 0.7165,
+      "step": 1197
+    },
+    {
+      "epoch": 1.1164958061509787,
+      "grad_norm": 1.4185198070432576,
+      "learning_rate": 3.4880911287538836e-05,
+      "loss": 0.7161,
+      "step": 1198
+    },
+    {
+      "epoch": 1.1174277726001864,
+      "grad_norm": 1.5795956507672817,
+      "learning_rate": 3.486365205384881e-05,
+      "loss": 0.8605,
+      "step": 1199
+    },
+    {
+      "epoch": 1.1183597390493942,
+      "grad_norm": 1.6208295472825978,
+      "learning_rate": 3.4846392820158784e-05,
+      "loss": 0.7593,
+      "step": 1200
+    },
+    {
+      "epoch": 1.119291705498602,
+      "grad_norm": 1.453058009695687,
+      "learning_rate": 3.4829133586468765e-05,
+      "loss": 0.67,
+      "step": 1201
+    },
+    {
+      "epoch": 1.1202236719478098,
+      "grad_norm": 1.3750343832689738,
+      "learning_rate": 3.481187435277874e-05,
+      "loss": 0.6646,
+      "step": 1202
+    },
+    {
+      "epoch": 1.1211556383970178,
+      "grad_norm": 1.462771294963424,
+      "learning_rate": 3.4794615119088714e-05,
+      "loss": 0.6449,
+      "step": 1203
+    },
+    {
+      "epoch": 1.1220876048462256,
+      "grad_norm": 1.180801377864553,
+      "learning_rate": 3.477735588539869e-05,
+      "loss": 0.6366,
+      "step": 1204
+    },
+    {
+      "epoch": 1.1230195712954334,
+      "grad_norm": 1.5889497429325734,
+      "learning_rate": 3.476009665170867e-05,
+      "loss": 0.8069,
+      "step": 1205
+    },
+    {
+      "epoch": 1.1239515377446412,
+      "grad_norm": 1.4057523633674185,
+      "learning_rate": 3.4742837418018644e-05,
+      "loss": 0.8894,
+      "step": 1206
+    },
+    {
+      "epoch": 1.124883504193849,
+      "grad_norm": 2.136342495317272,
+      "learning_rate": 3.472557818432862e-05,
+      "loss": 0.9474,
+      "step": 1207
+    },
+    {
+      "epoch": 1.1258154706430568,
+      "grad_norm": 1.5121500987077814,
+      "learning_rate": 3.470831895063859e-05,
+      "loss": 0.711,
+      "step": 1208
+    },
+    {
+      "epoch": 1.1267474370922648,
+      "grad_norm": 1.3560411189927002,
+      "learning_rate": 3.469105971694857e-05,
+      "loss": 0.6868,
+      "step": 1209
+    },
+    {
+      "epoch": 1.1276794035414726,
+      "grad_norm": 1.537784094842146,
+      "learning_rate": 3.467380048325855e-05,
+      "loss": 0.8373,
+      "step": 1210
+    },
+    {
+      "epoch": 1.1286113699906803,
+      "grad_norm": 1.3427875436774859,
+      "learning_rate": 3.465654124956852e-05,
+      "loss": 0.7215,
+      "step": 1211
+    },
+    {
+      "epoch": 1.1295433364398881,
+      "grad_norm": 1.2902431011871058,
+      "learning_rate": 3.4639282015878496e-05,
+      "loss": 0.706,
+      "step": 1212
+    },
+    {
+      "epoch": 1.130475302889096,
+      "grad_norm": 1.5321738889724492,
+      "learning_rate": 3.462202278218847e-05,
+      "loss": 0.714,
+      "step": 1213
+    },
+    {
+      "epoch": 1.131407269338304,
+      "grad_norm": 1.3542330787043868,
+      "learning_rate": 3.4604763548498445e-05,
+      "loss": 0.7385,
+      "step": 1214
+    },
+    {
+      "epoch": 1.1323392357875117,
+      "grad_norm": 1.4379777066067934,
+      "learning_rate": 3.458750431480842e-05,
+      "loss": 0.7733,
+      "step": 1215
+    },
+    {
+      "epoch": 1.1332712022367195,
+      "grad_norm": 1.01214024067799,
+      "learning_rate": 3.45702450811184e-05,
+      "loss": 0.6375,
+      "step": 1216
+    },
+    {
+      "epoch": 1.1342031686859273,
+      "grad_norm": 1.5406765766372088,
+      "learning_rate": 3.4552985847428374e-05,
+      "loss": 0.7784,
+      "step": 1217
+    },
+    {
+      "epoch": 1.135135135135135,
+      "grad_norm": 1.393892744193624,
+      "learning_rate": 3.453572661373835e-05,
+      "loss": 0.8303,
+      "step": 1218
+    },
+    {
+      "epoch": 1.1360671015843429,
+      "grad_norm": 1.26115146355603,
+      "learning_rate": 3.451846738004832e-05,
+      "loss": 0.6876,
+      "step": 1219
+    },
+    {
+      "epoch": 1.1369990680335509,
+      "grad_norm": 2.094520573160802,
+      "learning_rate": 3.4501208146358304e-05,
+      "loss": 0.9334,
+      "step": 1220
+    },
+    {
+      "epoch": 1.1379310344827587,
+      "grad_norm": 1.4351623104147457,
+      "learning_rate": 3.448394891266828e-05,
+      "loss": 0.763,
+      "step": 1221
+    },
+    {
+      "epoch": 1.1388630009319665,
+      "grad_norm": 1.133047111732924,
+      "learning_rate": 3.446668967897825e-05,
+      "loss": 0.6213,
+      "step": 1222
+    },
+    {
+      "epoch": 1.1397949673811743,
+      "grad_norm": 1.3637895241974394,
+      "learning_rate": 3.4449430445288234e-05,
+      "loss": 0.7969,
+      "step": 1223
+    },
+    {
+      "epoch": 1.140726933830382,
+      "grad_norm": 1.3666425370277386,
+      "learning_rate": 3.443217121159821e-05,
+      "loss": 0.765,
+      "step": 1224
+    },
+    {
+      "epoch": 1.14165890027959,
+      "grad_norm": 1.6668434545584325,
+      "learning_rate": 3.441491197790818e-05,
+      "loss": 0.7946,
+      "step": 1225
+    },
+    {
+      "epoch": 1.1425908667287978,
+      "grad_norm": 1.2340857217957313,
+      "learning_rate": 3.4397652744218157e-05,
+      "loss": 0.7121,
+      "step": 1226
+    },
+    {
+      "epoch": 1.1435228331780056,
+      "grad_norm": 1.3098558947267365,
+      "learning_rate": 3.438039351052814e-05,
+      "loss": 0.7325,
+      "step": 1227
+    },
+    {
+      "epoch": 1.1444547996272134,
+      "grad_norm": 1.3802340525160324,
+      "learning_rate": 3.436313427683811e-05,
+      "loss": 0.7011,
+      "step": 1228
+    },
+    {
+      "epoch": 1.1453867660764212,
+      "grad_norm": 1.3348720521589563,
+      "learning_rate": 3.4345875043148086e-05,
+      "loss": 0.7302,
+      "step": 1229
+    },
+    {
+      "epoch": 1.146318732525629,
+      "grad_norm": 1.5262091858833498,
+      "learning_rate": 3.432861580945806e-05,
+      "loss": 0.7731,
+      "step": 1230
+    },
+    {
+      "epoch": 1.147250698974837,
+      "grad_norm": 1.5381927796808261,
+      "learning_rate": 3.431135657576804e-05,
+      "loss": 0.8106,
+      "step": 1231
+    },
+    {
+      "epoch": 1.1481826654240448,
+      "grad_norm": 1.30347270853192,
+      "learning_rate": 3.4294097342078016e-05,
+      "loss": 0.634,
+      "step": 1232
+    },
+    {
+      "epoch": 1.1491146318732526,
+      "grad_norm": 1.2155423714326736,
+      "learning_rate": 3.427683810838799e-05,
+      "loss": 0.7018,
+      "step": 1233
+    },
+    {
+      "epoch": 1.1500465983224604,
+      "grad_norm": 1.3685471567953102,
+      "learning_rate": 3.4259578874697964e-05,
+      "loss": 0.8134,
+      "step": 1234
+    },
+    {
+      "epoch": 1.1509785647716682,
+      "grad_norm": 1.2626951754885825,
+      "learning_rate": 3.424231964100794e-05,
+      "loss": 0.6848,
+      "step": 1235
+    },
+    {
+      "epoch": 1.1519105312208762,
+      "grad_norm": 1.0273421003947283,
+      "learning_rate": 3.422506040731791e-05,
+      "loss": 0.6366,
+      "step": 1236
+    },
+    {
+      "epoch": 1.152842497670084,
+      "grad_norm": 1.4214607431595907,
+      "learning_rate": 3.420780117362789e-05,
+      "loss": 0.6507,
+      "step": 1237
+    },
+    {
+      "epoch": 1.1537744641192917,
+      "grad_norm": 1.3692228464219929,
+      "learning_rate": 3.419054193993787e-05,
+      "loss": 0.7155,
+      "step": 1238
+    },
+    {
+      "epoch": 1.1547064305684995,
+      "grad_norm": 1.2444240483172668,
+      "learning_rate": 3.417328270624784e-05,
+      "loss": 0.6111,
+      "step": 1239
+    },
+    {
+      "epoch": 1.1556383970177073,
+      "grad_norm": 1.508014742045498,
+      "learning_rate": 3.415602347255782e-05,
+      "loss": 0.8421,
+      "step": 1240
+    },
+    {
+      "epoch": 1.156570363466915,
+      "grad_norm": 1.5298555665673337,
+      "learning_rate": 3.41387642388678e-05,
+      "loss": 0.8193,
+      "step": 1241
+    },
+    {
+      "epoch": 1.157502329916123,
+      "grad_norm": 1.3241589238633034,
+      "learning_rate": 3.412150500517777e-05,
+      "loss": 0.7172,
+      "step": 1242
+    },
+    {
+      "epoch": 1.158434296365331,
+      "grad_norm": 1.325024849865998,
+      "learning_rate": 3.410424577148775e-05,
+      "loss": 0.7217,
+      "step": 1243
+    },
+    {
+      "epoch": 1.1593662628145387,
+      "grad_norm": 1.891737863058202,
+      "learning_rate": 3.408698653779772e-05,
+      "loss": 0.8857,
+      "step": 1244
+    },
+    {
+      "epoch": 1.1602982292637465,
+      "grad_norm": 1.5845881092218366,
+      "learning_rate": 3.40697273041077e-05,
+      "loss": 0.7522,
+      "step": 1245
+    },
+    {
+      "epoch": 1.1612301957129543,
+      "grad_norm": 1.6718145664112418,
+      "learning_rate": 3.4052468070417676e-05,
+      "loss": 0.7265,
+      "step": 1246
+    },
+    {
+      "epoch": 1.1621621621621623,
+      "grad_norm": 1.4959781894825384,
+      "learning_rate": 3.403520883672765e-05,
+      "loss": 0.7902,
+      "step": 1247
+    },
+    {
+      "epoch": 1.16309412861137,
+      "grad_norm": 1.1105281050645424,
+      "learning_rate": 3.4017949603037625e-05,
+      "loss": 0.6054,
+      "step": 1248
+    },
+    {
+      "epoch": 1.1640260950605779,
+      "grad_norm": 1.4474794499013945,
+      "learning_rate": 3.4000690369347606e-05,
+      "loss": 0.7194,
+      "step": 1249
+    },
+    {
+      "epoch": 1.1649580615097856,
+      "grad_norm": 1.341240098891626,
+      "learning_rate": 3.398343113565758e-05,
+      "loss": 0.7067,
+      "step": 1250
+    },
+    {
+      "epoch": 1.1658900279589934,
+      "grad_norm": 1.778980150614401,
+      "learning_rate": 3.3966171901967554e-05,
+      "loss": 0.7627,
+      "step": 1251
+    },
+    {
+      "epoch": 1.1668219944082012,
+      "grad_norm": 1.2703966197391652,
+      "learning_rate": 3.3948912668277536e-05,
+      "loss": 0.64,
+      "step": 1252
+    },
+    {
+      "epoch": 1.167753960857409,
+      "grad_norm": 1.32364497557943,
+      "learning_rate": 3.393165343458751e-05,
+      "loss": 0.6831,
+      "step": 1253
+    },
+    {
+      "epoch": 1.168685927306617,
+      "grad_norm": 1.5158363211402257,
+      "learning_rate": 3.3914394200897484e-05,
+      "loss": 0.763,
+      "step": 1254
+    },
+    {
+      "epoch": 1.1696178937558248,
+      "grad_norm": 1.243813495628069,
+      "learning_rate": 3.389713496720746e-05,
+      "loss": 0.7165,
+      "step": 1255
+    },
+    {
+      "epoch": 1.1705498602050326,
+      "grad_norm": 1.4048847252295722,
+      "learning_rate": 3.387987573351743e-05,
+      "loss": 0.7691,
+      "step": 1256
+    },
+    {
+      "epoch": 1.1714818266542404,
+      "grad_norm": 1.6454763698206067,
+      "learning_rate": 3.386261649982741e-05,
+      "loss": 0.8321,
+      "step": 1257
+    },
+    {
+      "epoch": 1.1724137931034484,
+      "grad_norm": 1.6022755854892032,
+      "learning_rate": 3.384535726613738e-05,
+      "loss": 0.6785,
+      "step": 1258
+    },
+    {
+      "epoch": 1.1733457595526562,
+      "grad_norm": 1.6817601247260072,
+      "learning_rate": 3.3828098032447356e-05,
+      "loss": 0.7596,
+      "step": 1259
+    },
+    {
+      "epoch": 1.174277726001864,
+      "grad_norm": 1.5148919449456182,
+      "learning_rate": 3.381083879875734e-05,
+      "loss": 0.7504,
+      "step": 1260
+    },
+    {
+      "epoch": 1.1752096924510718,
+      "grad_norm": 1.2403796010532109,
+      "learning_rate": 3.379357956506731e-05,
+      "loss": 0.677,
+      "step": 1261
+    },
+    {
+      "epoch": 1.1761416589002796,
+      "grad_norm": 1.6013035943414615,
+      "learning_rate": 3.3776320331377285e-05,
+      "loss": 0.764,
+      "step": 1262
+    },
+    {
+      "epoch": 1.1770736253494873,
+      "grad_norm": 1.2672829541392943,
+      "learning_rate": 3.3759061097687266e-05,
+      "loss": 0.6603,
+      "step": 1263
+    },
+    {
+      "epoch": 1.1780055917986951,
+      "grad_norm": 1.3814181935476173,
+      "learning_rate": 3.374180186399724e-05,
+      "loss": 0.6628,
+      "step": 1264
+    },
+    {
+      "epoch": 1.1789375582479031,
+      "grad_norm": 2.0640397943190183,
+      "learning_rate": 3.3724542630307215e-05,
+      "loss": 1.0156,
+      "step": 1265
+    },
+    {
+      "epoch": 1.179869524697111,
+      "grad_norm": 1.137631594522119,
+      "learning_rate": 3.370728339661719e-05,
+      "loss": 0.6939,
+      "step": 1266
+    },
+    {
+      "epoch": 1.1808014911463187,
+      "grad_norm": 1.1639433021188843,
+      "learning_rate": 3.369002416292717e-05,
+      "loss": 0.693,
+      "step": 1267
+    },
+    {
+      "epoch": 1.1817334575955265,
+      "grad_norm": 1.463716112517946,
+      "learning_rate": 3.3672764929237145e-05,
+      "loss": 0.7383,
+      "step": 1268
+    },
+    {
+      "epoch": 1.1826654240447343,
+      "grad_norm": 3.332954545425625,
+      "learning_rate": 3.365550569554712e-05,
+      "loss": 1.0486,
+      "step": 1269
+    },
+    {
+      "epoch": 1.1835973904939423,
+      "grad_norm": 1.4931589556792706,
+      "learning_rate": 3.363824646185709e-05,
+      "loss": 0.6973,
+      "step": 1270
+    },
+    {
+      "epoch": 1.18452935694315,
+      "grad_norm": 1.391964702354487,
+      "learning_rate": 3.3620987228167074e-05,
+      "loss": 0.8136,
+      "step": 1271
+    },
+    {
+      "epoch": 1.1854613233923579,
+      "grad_norm": 1.3799787078810164,
+      "learning_rate": 3.360372799447705e-05,
+      "loss": 0.6889,
+      "step": 1272
+    },
+    {
+      "epoch": 1.1863932898415657,
+      "grad_norm": 1.5094332630475207,
+      "learning_rate": 3.358646876078702e-05,
+      "loss": 0.7112,
+      "step": 1273
+    },
+    {
+      "epoch": 1.1873252562907735,
+      "grad_norm": 1.4158505521535538,
+      "learning_rate": 3.3569209527097004e-05,
+      "loss": 0.7239,
+      "step": 1274
+    },
+    {
+      "epoch": 1.1882572227399812,
+      "grad_norm": 1.712231069037708,
+      "learning_rate": 3.355195029340698e-05,
+      "loss": 0.8908,
+      "step": 1275
+    },
+    {
+      "epoch": 1.1891891891891893,
+      "grad_norm": 1.6692247829412796,
+      "learning_rate": 3.353469105971695e-05,
+      "loss": 0.8136,
+      "step": 1276
+    },
+    {
+      "epoch": 1.190121155638397,
+      "grad_norm": 1.4525298457746871,
+      "learning_rate": 3.351743182602693e-05,
+      "loss": 0.7977,
+      "step": 1277
+    },
+    {
+      "epoch": 1.1910531220876048,
+      "grad_norm": 1.2312441209133833,
+      "learning_rate": 3.35001725923369e-05,
+      "loss": 0.6929,
+      "step": 1278
+    },
+    {
+      "epoch": 1.1919850885368126,
+      "grad_norm": 1.4340754786597731,
+      "learning_rate": 3.3482913358646875e-05,
+      "loss": 0.8067,
+      "step": 1279
+    },
+    {
+      "epoch": 1.1929170549860204,
+      "grad_norm": 1.3909037881047654,
+      "learning_rate": 3.346565412495685e-05,
+      "loss": 0.7509,
+      "step": 1280
+    },
+    {
+      "epoch": 1.1938490214352284,
+      "grad_norm": 1.2013848942720886,
+      "learning_rate": 3.344839489126683e-05,
+      "loss": 0.6971,
+      "step": 1281
+    },
+    {
+      "epoch": 1.1947809878844362,
+      "grad_norm": 1.3759631590939345,
+      "learning_rate": 3.3431135657576805e-05,
+      "loss": 0.8664,
+      "step": 1282
+    },
+    {
+      "epoch": 1.195712954333644,
+      "grad_norm": 1.2058135809976276,
+      "learning_rate": 3.341387642388678e-05,
+      "loss": 0.7231,
+      "step": 1283
+    },
+    {
+      "epoch": 1.1966449207828518,
+      "grad_norm": 1.3975425212360018,
+      "learning_rate": 3.3396617190196753e-05,
+      "loss": 0.7862,
+      "step": 1284
+    },
+    {
+      "epoch": 1.1975768872320596,
+      "grad_norm": 1.4547981650881183,
+      "learning_rate": 3.3379357956506735e-05,
+      "loss": 0.8848,
+      "step": 1285
+    },
+    {
+      "epoch": 1.1985088536812674,
+      "grad_norm": 1.313203172364521,
+      "learning_rate": 3.336209872281671e-05,
+      "loss": 0.7364,
+      "step": 1286
+    },
+    {
+      "epoch": 1.1994408201304754,
+      "grad_norm": 1.727724799978353,
+      "learning_rate": 3.334483948912668e-05,
+      "loss": 0.829,
+      "step": 1287
+    },
+    {
+      "epoch": 1.2003727865796832,
+      "grad_norm": 1.1523256587549546,
+      "learning_rate": 3.332758025543666e-05,
+      "loss": 0.7201,
+      "step": 1288
+    },
+    {
+      "epoch": 1.201304753028891,
+      "grad_norm": 1.1000111696990804,
+      "learning_rate": 3.331032102174664e-05,
+      "loss": 0.611,
+      "step": 1289
+    },
+    {
+      "epoch": 1.2022367194780987,
+      "grad_norm": 1.2078596675098292,
+      "learning_rate": 3.329306178805661e-05,
+      "loss": 0.6758,
+      "step": 1290
+    },
+    {
+      "epoch": 1.2031686859273065,
+      "grad_norm": 1.364835137803524,
+      "learning_rate": 3.327580255436659e-05,
+      "loss": 0.6657,
+      "step": 1291
+    },
+    {
+      "epoch": 1.2041006523765145,
+      "grad_norm": 1.7003539151649592,
+      "learning_rate": 3.325854332067657e-05,
+      "loss": 0.7117,
+      "step": 1292
+    },
+    {
+      "epoch": 1.2050326188257223,
+      "grad_norm": 1.2575115717198357,
+      "learning_rate": 3.324128408698654e-05,
+      "loss": 0.7161,
+      "step": 1293
+    },
+    {
+      "epoch": 1.2059645852749301,
+      "grad_norm": 1.4499481132076104,
+      "learning_rate": 3.322402485329652e-05,
+      "loss": 0.845,
+      "step": 1294
+    },
+    {
+      "epoch": 1.206896551724138,
+      "grad_norm": 1.380671367893695,
+      "learning_rate": 3.320676561960649e-05,
+      "loss": 0.6404,
+      "step": 1295
+    },
+    {
+      "epoch": 1.2078285181733457,
+      "grad_norm": 1.5558462900945862,
+      "learning_rate": 3.318950638591647e-05,
+      "loss": 0.7812,
+      "step": 1296
+    },
+    {
+      "epoch": 1.2087604846225535,
+      "grad_norm": 1.3402612632751973,
+      "learning_rate": 3.3172247152226446e-05,
+      "loss": 0.6908,
+      "step": 1297
+    },
+    {
+      "epoch": 1.2096924510717615,
+      "grad_norm": 1.5283430281410657,
+      "learning_rate": 3.315498791853642e-05,
+      "loss": 0.8126,
+      "step": 1298
+    },
+    {
+      "epoch": 1.2106244175209693,
+      "grad_norm": 1.5204837870324759,
+      "learning_rate": 3.3137728684846395e-05,
+      "loss": 0.8256,
+      "step": 1299
+    },
+    {
+      "epoch": 1.211556383970177,
+      "grad_norm": 1.6336424635170927,
+      "learning_rate": 3.312046945115637e-05,
+      "loss": 0.8615,
+      "step": 1300
+    },
+    {
+      "epoch": 1.2124883504193849,
+      "grad_norm": 1.2505479129894008,
+      "learning_rate": 3.3103210217466344e-05,
+      "loss": 0.6853,
+      "step": 1301
+    },
+    {
+      "epoch": 1.2134203168685926,
+      "grad_norm": 1.3086887059537167,
+      "learning_rate": 3.308595098377632e-05,
+      "loss": 0.855,
+      "step": 1302
+    },
+    {
+      "epoch": 1.2143522833178007,
+      "grad_norm": 1.8764180696086892,
+      "learning_rate": 3.30686917500863e-05,
+      "loss": 0.8537,
+      "step": 1303
+    },
+    {
+      "epoch": 1.2152842497670084,
+      "grad_norm": 2.791848227395523,
+      "learning_rate": 3.305143251639627e-05,
+      "loss": 0.764,
+      "step": 1304
+    },
+    {
+      "epoch": 1.2162162162162162,
+      "grad_norm": 1.4073250297561573,
+      "learning_rate": 3.303417328270625e-05,
+      "loss": 0.6354,
+      "step": 1305
+    },
+    {
+      "epoch": 1.217148182665424,
+      "grad_norm": 1.669963731091551,
+      "learning_rate": 3.301691404901622e-05,
+      "loss": 0.7308,
+      "step": 1306
+    },
+    {
+      "epoch": 1.2180801491146318,
+      "grad_norm": 1.2575440487537823,
+      "learning_rate": 3.29996548153262e-05,
+      "loss": 0.6856,
+      "step": 1307
+    },
+    {
+      "epoch": 1.2190121155638396,
+      "grad_norm": 1.3785360912388753,
+      "learning_rate": 3.298239558163618e-05,
+      "loss": 0.6777,
+      "step": 1308
+    },
+    {
+      "epoch": 1.2199440820130476,
+      "grad_norm": 1.1557129116445568,
+      "learning_rate": 3.296513634794615e-05,
+      "loss": 0.6588,
+      "step": 1309
+    },
+    {
+      "epoch": 1.2208760484622554,
+      "grad_norm": 1.3471972580244775,
+      "learning_rate": 3.2947877114256126e-05,
+      "loss": 0.6824,
+      "step": 1310
+    },
+    {
+      "epoch": 1.2218080149114632,
+      "grad_norm": 2.4034868712593718,
+      "learning_rate": 3.293061788056611e-05,
+      "loss": 1.0678,
+      "step": 1311
+    },
+    {
+      "epoch": 1.222739981360671,
+      "grad_norm": 1.2086844077773633,
+      "learning_rate": 3.291335864687608e-05,
+      "loss": 0.6745,
+      "step": 1312
+    },
+    {
+      "epoch": 1.2236719478098788,
+      "grad_norm": 1.5147381652634342,
+      "learning_rate": 3.2896099413186055e-05,
+      "loss": 0.7135,
+      "step": 1313
+    },
+    {
+      "epoch": 1.2246039142590868,
+      "grad_norm": 1.2776263416915479,
+      "learning_rate": 3.2878840179496036e-05,
+      "loss": 0.6925,
+      "step": 1314
+    },
+    {
+      "epoch": 1.2255358807082946,
+      "grad_norm": 1.305135519802972,
+      "learning_rate": 3.286158094580601e-05,
+      "loss": 0.6948,
+      "step": 1315
+    },
+    {
+      "epoch": 1.2264678471575023,
+      "grad_norm": 1.3743011824276925,
+      "learning_rate": 3.2844321712115985e-05,
+      "loss": 0.7222,
+      "step": 1316
+    },
+    {
+      "epoch": 1.2273998136067101,
+      "grad_norm": 1.3932919294250117,
+      "learning_rate": 3.282706247842596e-05,
+      "loss": 0.6985,
+      "step": 1317
+    },
+    {
+      "epoch": 1.228331780055918,
+      "grad_norm": 1.3280039061715228,
+      "learning_rate": 3.280980324473594e-05,
+      "loss": 0.662,
+      "step": 1318
+    },
+    {
+      "epoch": 1.2292637465051257,
+      "grad_norm": 1.3782065408402946,
+      "learning_rate": 3.2792544011045915e-05,
+      "loss": 0.8422,
+      "step": 1319
+    },
+    {
+      "epoch": 1.2301957129543337,
+      "grad_norm": 1.5301132984480748,
+      "learning_rate": 3.277528477735589e-05,
+      "loss": 0.7373,
+      "step": 1320
+    },
+    {
+      "epoch": 1.2311276794035415,
+      "grad_norm": 0.97566106662221,
+      "learning_rate": 3.275802554366586e-05,
+      "loss": 0.6491,
+      "step": 1321
+    },
+    {
+      "epoch": 1.2320596458527493,
+      "grad_norm": 1.2319440860579822,
+      "learning_rate": 3.274076630997584e-05,
+      "loss": 0.7067,
+      "step": 1322
+    },
+    {
+      "epoch": 1.232991612301957,
+      "grad_norm": 1.1269313723565566,
+      "learning_rate": 3.272350707628581e-05,
+      "loss": 0.6509,
+      "step": 1323
+    },
+    {
+      "epoch": 1.2339235787511649,
+      "grad_norm": 1.278063597037127,
+      "learning_rate": 3.2706247842595786e-05,
+      "loss": 0.7476,
+      "step": 1324
+    },
+    {
+      "epoch": 1.2348555452003729,
+      "grad_norm": 0.9943540806711418,
+      "learning_rate": 3.268898860890577e-05,
+      "loss": 0.6157,
+      "step": 1325
+    },
+    {
+      "epoch": 1.2357875116495807,
+      "grad_norm": 1.0339951604708595,
+      "learning_rate": 3.267172937521574e-05,
+      "loss": 0.6374,
+      "step": 1326
+    },
+    {
+      "epoch": 1.2367194780987885,
+      "grad_norm": 1.562915110293172,
+      "learning_rate": 3.2654470141525716e-05,
+      "loss": 0.7909,
+      "step": 1327
+    },
+    {
+      "epoch": 1.2376514445479962,
+      "grad_norm": 1.5439063587720254,
+      "learning_rate": 3.263721090783569e-05,
+      "loss": 0.9021,
+      "step": 1328
+    },
+    {
+      "epoch": 1.238583410997204,
+      "grad_norm": 1.1444741939148921,
+      "learning_rate": 3.261995167414567e-05,
+      "loss": 0.6594,
+      "step": 1329
+    },
+    {
+      "epoch": 1.2395153774464118,
+      "grad_norm": 1.4597913122938997,
+      "learning_rate": 3.2602692440455645e-05,
+      "loss": 0.667,
+      "step": 1330
+    },
+    {
+      "epoch": 1.2404473438956198,
+      "grad_norm": 1.4471786702951388,
+      "learning_rate": 3.258543320676562e-05,
+      "loss": 0.7916,
+      "step": 1331
+    },
+    {
+      "epoch": 1.2413793103448276,
+      "grad_norm": 1.0539304743368196,
+      "learning_rate": 3.25681739730756e-05,
+      "loss": 0.6564,
+      "step": 1332
+    },
+    {
+      "epoch": 1.2423112767940354,
+      "grad_norm": 1.0037750862963772,
+      "learning_rate": 3.2550914739385575e-05,
+      "loss": 0.5818,
+      "step": 1333
+    },
+    {
+      "epoch": 1.2432432432432432,
+      "grad_norm": 1.373627524042673,
+      "learning_rate": 3.253365550569555e-05,
+      "loss": 0.7756,
+      "step": 1334
+    },
+    {
+      "epoch": 1.244175209692451,
+      "grad_norm": 1.4208343907605052,
+      "learning_rate": 3.2516396272005524e-05,
+      "loss": 0.7271,
+      "step": 1335
+    },
+    {
+      "epoch": 1.245107176141659,
+      "grad_norm": 1.375302856331067,
+      "learning_rate": 3.2499137038315505e-05,
+      "loss": 0.7568,
+      "step": 1336
+    },
+    {
+      "epoch": 1.2460391425908668,
+      "grad_norm": 1.3863383267271319,
+      "learning_rate": 3.248187780462548e-05,
+      "loss": 0.7866,
+      "step": 1337
+    },
+    {
+      "epoch": 1.2469711090400746,
+      "grad_norm": 1.6094679697908614,
+      "learning_rate": 3.246461857093545e-05,
+      "loss": 0.6813,
+      "step": 1338
+    },
+    {
+      "epoch": 1.2479030754892824,
+      "grad_norm": 1.4933625968902746,
+      "learning_rate": 3.244735933724543e-05,
+      "loss": 0.8044,
+      "step": 1339
+    },
+    {
+      "epoch": 1.2488350419384902,
+      "grad_norm": 1.4009308156711724,
+      "learning_rate": 3.243010010355541e-05,
+      "loss": 0.6558,
+      "step": 1340
+    },
+    {
+      "epoch": 1.249767008387698,
+      "grad_norm": 1.5721212652423324,
+      "learning_rate": 3.241284086986538e-05,
+      "loss": 0.7524,
+      "step": 1341
+    },
+    {
+      "epoch": 1.250698974836906,
+      "grad_norm": 1.6523096890397586,
+      "learning_rate": 3.239558163617536e-05,
+      "loss": 0.7207,
+      "step": 1342
+    },
+    {
+      "epoch": 1.2516309412861137,
+      "grad_norm": 1.5359297030264936,
+      "learning_rate": 3.237832240248533e-05,
+      "loss": 0.8222,
+      "step": 1343
+    },
+    {
+      "epoch": 1.2525629077353215,
+      "grad_norm": 1.5224045993244149,
+      "learning_rate": 3.2361063168795306e-05,
+      "loss": 0.6752,
+      "step": 1344
+    },
+    {
+      "epoch": 1.2534948741845293,
+      "grad_norm": 1.2615001991805024,
+      "learning_rate": 3.234380393510528e-05,
+      "loss": 0.6901,
+      "step": 1345
+    },
+    {
+      "epoch": 1.254426840633737,
+      "grad_norm": 1.475470854381166,
+      "learning_rate": 3.2326544701415254e-05,
+      "loss": 0.7042,
+      "step": 1346
+    },
+    {
+      "epoch": 1.2553588070829451,
+      "grad_norm": 1.4581184770318525,
+      "learning_rate": 3.2309285467725235e-05,
+      "loss": 0.765,
+      "step": 1347
+    },
+    {
+      "epoch": 1.256290773532153,
+      "grad_norm": 1.2483092646926612,
+      "learning_rate": 3.229202623403521e-05,
+      "loss": 0.8107,
+      "step": 1348
+    },
+    {
+      "epoch": 1.2572227399813607,
+      "grad_norm": 1.7690403723766353,
+      "learning_rate": 3.2274767000345184e-05,
+      "loss": 0.8365,
+      "step": 1349
+    },
+    {
+      "epoch": 1.2581547064305685,
+      "grad_norm": 1.2473354728534776,
+      "learning_rate": 3.225750776665516e-05,
+      "loss": 0.6762,
+      "step": 1350
+    },
+    {
+      "epoch": 1.2590866728797763,
+      "grad_norm": 1.3958802236184433,
+      "learning_rate": 3.224024853296514e-05,
+      "loss": 0.7516,
+      "step": 1351
+    },
+    {
+      "epoch": 1.260018639328984,
+      "grad_norm": 1.5597976732126446,
+      "learning_rate": 3.2222989299275114e-05,
+      "loss": 0.7922,
+      "step": 1352
+    },
+    {
+      "epoch": 1.2609506057781918,
+      "grad_norm": 1.525302396148938,
+      "learning_rate": 3.220573006558509e-05,
+      "loss": 0.6952,
+      "step": 1353
+    },
+    {
+      "epoch": 1.2618825722273999,
+      "grad_norm": 1.041903575366251,
+      "learning_rate": 3.218847083189507e-05,
+      "loss": 0.6318,
+      "step": 1354
+    },
+    {
+      "epoch": 1.2628145386766076,
+      "grad_norm": 1.5649627887492368,
+      "learning_rate": 3.217121159820504e-05,
+      "loss": 0.7875,
+      "step": 1355
+    },
+    {
+      "epoch": 1.2637465051258154,
+      "grad_norm": 1.1890127490158091,
+      "learning_rate": 3.215395236451502e-05,
+      "loss": 0.7424,
+      "step": 1356
+    },
+    {
+      "epoch": 1.2646784715750232,
+      "grad_norm": 1.263993585880285,
+      "learning_rate": 3.213669313082499e-05,
+      "loss": 0.6251,
+      "step": 1357
+    },
+    {
+      "epoch": 1.2656104380242312,
+      "grad_norm": 1.8496162330496884,
+      "learning_rate": 3.211943389713497e-05,
+      "loss": 0.7589,
+      "step": 1358
+    },
+    {
+      "epoch": 1.266542404473439,
+      "grad_norm": 1.2365255172781422,
+      "learning_rate": 3.210217466344495e-05,
+      "loss": 0.6174,
+      "step": 1359
+    },
+    {
+      "epoch": 1.2674743709226468,
+      "grad_norm": 1.8107410315656784,
+      "learning_rate": 3.208491542975492e-05,
+      "loss": 0.8286,
+      "step": 1360
+    },
+    {
+      "epoch": 1.2684063373718546,
+      "grad_norm": 1.4158023991085815,
+      "learning_rate": 3.2067656196064896e-05,
+      "loss": 0.7465,
+      "step": 1361
+    },
+    {
+      "epoch": 1.2693383038210624,
+      "grad_norm": 1.0844632533352248,
+      "learning_rate": 3.205039696237487e-05,
+      "loss": 0.6136,
+      "step": 1362
+    },
+    {
+      "epoch": 1.2702702702702702,
+      "grad_norm": 1.473451764450536,
+      "learning_rate": 3.2033137728684844e-05,
+      "loss": 0.6616,
+      "step": 1363
+    },
+    {
+      "epoch": 1.271202236719478,
+      "grad_norm": 1.239950567765819,
+      "learning_rate": 3.201587849499482e-05,
+      "loss": 0.6429,
+      "step": 1364
+    },
+    {
+      "epoch": 1.272134203168686,
+      "grad_norm": 1.6556697237673317,
+      "learning_rate": 3.19986192613048e-05,
+      "loss": 0.7233,
+      "step": 1365
+    },
+    {
+      "epoch": 1.2730661696178938,
+      "grad_norm": 1.3207382478008765,
+      "learning_rate": 3.1981360027614774e-05,
+      "loss": 0.7185,
+      "step": 1366
+    },
+    {
+      "epoch": 1.2739981360671015,
+      "grad_norm": 1.3207448722245285,
+      "learning_rate": 3.196410079392475e-05,
+      "loss": 0.7158,
+      "step": 1367
+    },
+    {
+      "epoch": 1.2749301025163093,
+      "grad_norm": 1.2842845243280097,
+      "learning_rate": 3.194684156023472e-05,
+      "loss": 0.7554,
+      "step": 1368
+    },
+    {
+      "epoch": 1.2758620689655173,
+      "grad_norm": 1.2207423836090032,
+      "learning_rate": 3.1929582326544704e-05,
+      "loss": 0.6563,
+      "step": 1369
+    },
+    {
+      "epoch": 1.2767940354147251,
+      "grad_norm": 1.9235161522994055,
+      "learning_rate": 3.191232309285468e-05,
+      "loss": 0.875,
+      "step": 1370
+    },
+    {
+      "epoch": 1.277726001863933,
+      "grad_norm": 1.5915466591184841,
+      "learning_rate": 3.189506385916465e-05,
+      "loss": 0.9335,
+      "step": 1371
+    },
+    {
+      "epoch": 1.2786579683131407,
+      "grad_norm": 1.1944473231283887,
+      "learning_rate": 3.187780462547463e-05,
+      "loss": 0.6642,
+      "step": 1372
+    },
+    {
+      "epoch": 1.2795899347623485,
+      "grad_norm": 1.024355277454074,
+      "learning_rate": 3.186054539178461e-05,
+      "loss": 0.6022,
+      "step": 1373
+    },
+    {
+      "epoch": 1.2805219012115563,
+      "grad_norm": 1.3756689951158794,
+      "learning_rate": 3.184328615809458e-05,
+      "loss": 0.7764,
+      "step": 1374
+    },
+    {
+      "epoch": 1.281453867660764,
+      "grad_norm": 1.2938057483946184,
+      "learning_rate": 3.1826026924404556e-05,
+      "loss": 0.6184,
+      "step": 1375
+    },
+    {
+      "epoch": 1.282385834109972,
+      "grad_norm": 1.164778660584826,
+      "learning_rate": 3.180876769071454e-05,
+      "loss": 0.6765,
+      "step": 1376
+    },
+    {
+      "epoch": 1.2833178005591799,
+      "grad_norm": 1.0736576616833164,
+      "learning_rate": 3.179150845702451e-05,
+      "loss": 0.663,
+      "step": 1377
+    },
+    {
+      "epoch": 1.2842497670083877,
+      "grad_norm": 1.2725649828088756,
+      "learning_rate": 3.1774249223334486e-05,
+      "loss": 0.7794,
+      "step": 1378
+    },
+    {
+      "epoch": 1.2851817334575955,
+      "grad_norm": 1.463271425470718,
+      "learning_rate": 3.175698998964446e-05,
+      "loss": 0.7278,
+      "step": 1379
+    },
+    {
+      "epoch": 1.2861136999068035,
+      "grad_norm": 1.4054346251351435,
+      "learning_rate": 3.173973075595444e-05,
+      "loss": 0.653,
+      "step": 1380
+    },
+    {
+      "epoch": 1.2870456663560113,
+      "grad_norm": 1.260821563635683,
+      "learning_rate": 3.1722471522264415e-05,
+      "loss": 0.5885,
+      "step": 1381
+    },
+    {
+      "epoch": 1.287977632805219,
+      "grad_norm": 1.1038897646998014,
+      "learning_rate": 3.170521228857439e-05,
+      "loss": 0.5783,
+      "step": 1382
+    },
+    {
+      "epoch": 1.2889095992544268,
+      "grad_norm": 1.242682749553793,
+      "learning_rate": 3.1687953054884364e-05,
+      "loss": 0.6986,
+      "step": 1383
+    },
+    {
+      "epoch": 1.2898415657036346,
+      "grad_norm": 1.2079198445419437,
+      "learning_rate": 3.167069382119434e-05,
+      "loss": 0.7422,
+      "step": 1384
+    },
+    {
+      "epoch": 1.2907735321528424,
+      "grad_norm": 1.2043388360534122,
+      "learning_rate": 3.165343458750431e-05,
+      "loss": 0.6504,
+      "step": 1385
+    },
+    {
+      "epoch": 1.2917054986020502,
+      "grad_norm": 1.3048600656115052,
+      "learning_rate": 3.163617535381429e-05,
+      "loss": 0.6546,
+      "step": 1386
+    },
+    {
+      "epoch": 1.2926374650512582,
+      "grad_norm": 1.5668823596321173,
+      "learning_rate": 3.161891612012427e-05,
+      "loss": 0.7678,
+      "step": 1387
+    },
+    {
+      "epoch": 1.293569431500466,
+      "grad_norm": 1.2106289662707606,
+      "learning_rate": 3.160165688643424e-05,
+      "loss": 0.7132,
+      "step": 1388
+    },
+    {
+      "epoch": 1.2945013979496738,
+      "grad_norm": 1.2216423581756501,
+      "learning_rate": 3.1584397652744217e-05,
+      "loss": 0.6597,
+      "step": 1389
+    },
+    {
+      "epoch": 1.2954333643988816,
+      "grad_norm": 1.4168098381604592,
+      "learning_rate": 3.156713841905419e-05,
+      "loss": 0.6796,
+      "step": 1390
+    },
+    {
+      "epoch": 1.2963653308480896,
+      "grad_norm": 1.2647847852819936,
+      "learning_rate": 3.154987918536417e-05,
+      "loss": 0.6268,
+      "step": 1391
+    },
+    {
+      "epoch": 1.2972972972972974,
+      "grad_norm": 1.383375437381754,
+      "learning_rate": 3.1532619951674146e-05,
+      "loss": 0.7212,
+      "step": 1392
+    },
+    {
+      "epoch": 1.2982292637465052,
+      "grad_norm": 1.2499128926696708,
+      "learning_rate": 3.151536071798412e-05,
+      "loss": 0.7532,
+      "step": 1393
+    },
+    {
+      "epoch": 1.299161230195713,
+      "grad_norm": 1.139996877678223,
+      "learning_rate": 3.14981014842941e-05,
+      "loss": 0.6872,
+      "step": 1394
+    },
+    {
+      "epoch": 1.3000931966449207,
+      "grad_norm": 1.4132001170705772,
+      "learning_rate": 3.1480842250604076e-05,
+      "loss": 0.7058,
+      "step": 1395
+    },
+    {
+      "epoch": 1.3010251630941285,
+      "grad_norm": 1.7059791730388112,
+      "learning_rate": 3.146358301691405e-05,
+      "loss": 0.7887,
+      "step": 1396
+    },
+    {
+      "epoch": 1.3019571295433363,
+      "grad_norm": 1.3934338079103297,
+      "learning_rate": 3.1446323783224024e-05,
+      "loss": 0.6382,
+      "step": 1397
+    },
+    {
+      "epoch": 1.3028890959925443,
+      "grad_norm": 1.0347770468342754,
+      "learning_rate": 3.1429064549534006e-05,
+      "loss": 0.6117,
+      "step": 1398
+    },
+    {
+      "epoch": 1.303821062441752,
+      "grad_norm": 1.6605552674398847,
+      "learning_rate": 3.141180531584398e-05,
+      "loss": 0.6009,
+      "step": 1399
+    },
+    {
+      "epoch": 1.30475302889096,
+      "grad_norm": 1.1209001190121353,
+      "learning_rate": 3.1394546082153954e-05,
+      "loss": 0.6587,
+      "step": 1400
+    },
+    {
+      "epoch": 1.3056849953401677,
+      "grad_norm": 1.2703854633057727,
+      "learning_rate": 3.137728684846393e-05,
+      "loss": 0.7087,
+      "step": 1401
+    },
+    {
+      "epoch": 1.3066169617893757,
+      "grad_norm": 1.562357129894783,
+      "learning_rate": 3.136002761477391e-05,
+      "loss": 0.7734,
+      "step": 1402
+    },
+    {
+      "epoch": 1.3075489282385835,
+      "grad_norm": 1.4833748769752175,
+      "learning_rate": 3.1342768381083884e-05,
+      "loss": 0.6998,
+      "step": 1403
+    },
+    {
+      "epoch": 1.3084808946877913,
+      "grad_norm": 1.619131904479025,
+      "learning_rate": 3.132550914739386e-05,
+      "loss": 0.735,
+      "step": 1404
+    },
+    {
+      "epoch": 1.309412861136999,
+      "grad_norm": 1.7606254828378527,
+      "learning_rate": 3.130824991370383e-05,
+      "loss": 0.7982,
+      "step": 1405
+    },
+    {
+      "epoch": 1.3103448275862069,
+      "grad_norm": 1.26255164592409,
+      "learning_rate": 3.129099068001381e-05,
+      "loss": 0.6815,
+      "step": 1406
+    },
+    {
+      "epoch": 1.3112767940354146,
+      "grad_norm": 1.4411589778664657,
+      "learning_rate": 3.127373144632378e-05,
+      "loss": 0.7135,
+      "step": 1407
+    },
+    {
+      "epoch": 1.3122087604846224,
+      "grad_norm": 1.4045882030880255,
+      "learning_rate": 3.1256472212633755e-05,
+      "loss": 0.7154,
+      "step": 1408
+    },
+    {
+      "epoch": 1.3131407269338304,
+      "grad_norm": 1.3400633097909807,
+      "learning_rate": 3.1239212978943736e-05,
+      "loss": 0.7148,
+      "step": 1409
+    },
+    {
+      "epoch": 1.3140726933830382,
+      "grad_norm": 0.9242001257183926,
+      "learning_rate": 3.122195374525371e-05,
+      "loss": 0.6008,
+      "step": 1410
+    },
+    {
+      "epoch": 1.315004659832246,
+      "grad_norm": 1.3829772827949818,
+      "learning_rate": 3.1204694511563685e-05,
+      "loss": 0.7145,
+      "step": 1411
+    },
+    {
+      "epoch": 1.3159366262814538,
+      "grad_norm": 1.4704979621818994,
+      "learning_rate": 3.118743527787366e-05,
+      "loss": 0.6771,
+      "step": 1412
+    },
+    {
+      "epoch": 1.3168685927306618,
+      "grad_norm": 1.4530786115721288,
+      "learning_rate": 3.117017604418364e-05,
+      "loss": 0.6566,
+      "step": 1413
+    },
+    {
+      "epoch": 1.3178005591798696,
+      "grad_norm": 1.2336880658913025,
+      "learning_rate": 3.1152916810493615e-05,
+      "loss": 0.7689,
+      "step": 1414
+    },
+    {
+      "epoch": 1.3187325256290774,
+      "grad_norm": 1.2108300826816694,
+      "learning_rate": 3.113565757680359e-05,
+      "loss": 0.7469,
+      "step": 1415
+    },
+    {
+      "epoch": 1.3196644920782852,
+      "grad_norm": 1.2926355618829761,
+      "learning_rate": 3.111839834311357e-05,
+      "loss": 0.6144,
+      "step": 1416
+    },
+    {
+      "epoch": 1.320596458527493,
+      "grad_norm": 1.4249241802907169,
+      "learning_rate": 3.1101139109423544e-05,
+      "loss": 0.6392,
+      "step": 1417
+    },
+    {
+      "epoch": 1.3215284249767008,
+      "grad_norm": 1.6740549907203979,
+      "learning_rate": 3.108387987573352e-05,
+      "loss": 0.8382,
+      "step": 1418
+    },
+    {
+      "epoch": 1.3224603914259085,
+      "grad_norm": 1.2738545955480534,
+      "learning_rate": 3.106662064204349e-05,
+      "loss": 0.6991,
+      "step": 1419
+    },
+    {
+      "epoch": 1.3233923578751166,
+      "grad_norm": 1.3921230563838707,
+      "learning_rate": 3.1049361408353474e-05,
+      "loss": 0.6689,
+      "step": 1420
+    },
+    {
+      "epoch": 1.3243243243243243,
+      "grad_norm": 1.2074005356311064,
+      "learning_rate": 3.103210217466345e-05,
+      "loss": 0.6615,
+      "step": 1421
+    },
+    {
+      "epoch": 1.3252562907735321,
+      "grad_norm": 1.5313727399433208,
+      "learning_rate": 3.101484294097342e-05,
+      "loss": 0.8302,
+      "step": 1422
+    },
+    {
+      "epoch": 1.32618825722274,
+      "grad_norm": 2.011708516623336,
+      "learning_rate": 3.0997583707283403e-05,
+      "loss": 0.859,
+      "step": 1423
+    },
+    {
+      "epoch": 1.327120223671948,
+      "grad_norm": 1.7007932685047795,
+      "learning_rate": 3.098032447359338e-05,
+      "loss": 0.7695,
+      "step": 1424
+    },
+    {
+      "epoch": 1.3280521901211557,
+      "grad_norm": 1.4336211828037448,
+      "learning_rate": 3.096306523990335e-05,
+      "loss": 0.7427,
+      "step": 1425
+    },
+    {
+      "epoch": 1.3289841565703635,
+      "grad_norm": 1.5097236879439422,
+      "learning_rate": 3.0945806006213326e-05,
+      "loss": 0.7847,
+      "step": 1426
+    },
+    {
+      "epoch": 1.3299161230195713,
+      "grad_norm": 1.3418383266564842,
+      "learning_rate": 3.09285467725233e-05,
+      "loss": 0.688,
+      "step": 1427
+    },
+    {
+      "epoch": 1.330848089468779,
+      "grad_norm": 1.123126121685922,
+      "learning_rate": 3.0911287538833275e-05,
+      "loss": 0.6325,
+      "step": 1428
+    },
+    {
+      "epoch": 1.3317800559179869,
+      "grad_norm": 1.2383019714691208,
+      "learning_rate": 3.089402830514325e-05,
+      "loss": 0.7294,
+      "step": 1429
+    },
+    {
+      "epoch": 1.3327120223671947,
+      "grad_norm": 1.6956577905080201,
+      "learning_rate": 3.0876769071453223e-05,
+      "loss": 0.8168,
+      "step": 1430
+    },
+    {
+      "epoch": 1.3336439888164027,
+      "grad_norm": 1.2258914280406548,
+      "learning_rate": 3.0859509837763205e-05,
+      "loss": 0.6893,
+      "step": 1431
+    },
+    {
+      "epoch": 1.3345759552656105,
+      "grad_norm": 1.3593892131543595,
+      "learning_rate": 3.084225060407318e-05,
+      "loss": 0.6794,
+      "step": 1432
+    },
+    {
+      "epoch": 1.3355079217148182,
+      "grad_norm": 1.5811268339335902,
+      "learning_rate": 3.082499137038315e-05,
+      "loss": 0.8741,
+      "step": 1433
+    },
+    {
+      "epoch": 1.336439888164026,
+      "grad_norm": 1.2527925062734924,
+      "learning_rate": 3.0807732136693134e-05,
+      "loss": 0.6471,
+      "step": 1434
+    },
+    {
+      "epoch": 1.337371854613234,
+      "grad_norm": 0.9917046308675229,
+      "learning_rate": 3.079047290300311e-05,
+      "loss": 0.6889,
+      "step": 1435
+    },
+    {
+      "epoch": 1.3383038210624418,
+      "grad_norm": 1.133886819608572,
+      "learning_rate": 3.077321366931308e-05,
+      "loss": 0.633,
+      "step": 1436
+    },
+    {
+      "epoch": 1.3392357875116496,
+      "grad_norm": 1.2652204316905835,
+      "learning_rate": 3.075595443562306e-05,
+      "loss": 0.633,
+      "step": 1437
+    },
+    {
+      "epoch": 1.3401677539608574,
+      "grad_norm": 1.2009913042620608,
+      "learning_rate": 3.073869520193304e-05,
+      "loss": 0.7911,
+      "step": 1438
+    },
+    {
+      "epoch": 1.3410997204100652,
+      "grad_norm": 1.0090314642488032,
+      "learning_rate": 3.072143596824301e-05,
+      "loss": 0.6152,
+      "step": 1439
+    },
+    {
+      "epoch": 1.342031686859273,
+      "grad_norm": 1.1028414508150655,
+      "learning_rate": 3.070417673455299e-05,
+      "loss": 0.6238,
+      "step": 1440
+    },
+    {
+      "epoch": 1.3429636533084808,
+      "grad_norm": 1.2217490846265446,
+      "learning_rate": 3.068691750086296e-05,
+      "loss": 0.7309,
+      "step": 1441
+    },
+    {
+      "epoch": 1.3438956197576888,
+      "grad_norm": 1.5986878227931975,
+      "learning_rate": 3.066965826717294e-05,
+      "loss": 0.7549,
+      "step": 1442
+    },
+    {
+      "epoch": 1.3448275862068966,
+      "grad_norm": 1.2733051572820622,
+      "learning_rate": 3.0652399033482916e-05,
+      "loss": 0.6401,
+      "step": 1443
+    },
+    {
+      "epoch": 1.3457595526561044,
+      "grad_norm": 1.2711765630482712,
+      "learning_rate": 3.063513979979289e-05,
+      "loss": 0.7334,
+      "step": 1444
+    },
+    {
+      "epoch": 1.3466915191053122,
+      "grad_norm": 1.6247563954376198,
+      "learning_rate": 3.061788056610287e-05,
+      "loss": 0.7913,
+      "step": 1445
+    },
+    {
+      "epoch": 1.3476234855545202,
+      "grad_norm": 1.4426420979261014,
+      "learning_rate": 3.0600621332412846e-05,
+      "loss": 0.7732,
+      "step": 1446
+    },
+    {
+      "epoch": 1.348555452003728,
+      "grad_norm": 1.465107549861714,
+      "learning_rate": 3.058336209872282e-05,
+      "loss": 0.778,
+      "step": 1447
+    },
+    {
+      "epoch": 1.3494874184529357,
+      "grad_norm": 1.4679234889576245,
+      "learning_rate": 3.0566102865032795e-05,
+      "loss": 0.6758,
+      "step": 1448
+    },
+    {
+      "epoch": 1.3504193849021435,
+      "grad_norm": 1.2280217329325505,
+      "learning_rate": 3.054884363134277e-05,
+      "loss": 0.6792,
+      "step": 1449
+    },
+    {
+      "epoch": 1.3513513513513513,
+      "grad_norm": 1.586980583518615,
+      "learning_rate": 3.053158439765274e-05,
+      "loss": 0.7336,
+      "step": 1450
+    },
+    {
+      "epoch": 1.352283317800559,
+      "grad_norm": 0.9715670367624714,
+      "learning_rate": 3.051432516396272e-05,
+      "loss": 0.6211,
+      "step": 1451
+    },
+    {
+      "epoch": 1.353215284249767,
+      "grad_norm": 1.4335007145342678,
+      "learning_rate": 3.0497065930272695e-05,
+      "loss": 0.7056,
+      "step": 1452
+    },
+    {
+      "epoch": 1.354147250698975,
+      "grad_norm": 1.0616891095257752,
+      "learning_rate": 3.0479806696582676e-05,
+      "loss": 0.6004,
+      "step": 1453
+    },
+    {
+      "epoch": 1.3550792171481827,
+      "grad_norm": 1.1991244271485644,
+      "learning_rate": 3.046254746289265e-05,
+      "loss": 0.6128,
+      "step": 1454
+    },
+    {
+      "epoch": 1.3560111835973905,
+      "grad_norm": 1.0840430320822245,
+      "learning_rate": 3.0445288229202625e-05,
+      "loss": 0.6291,
+      "step": 1455
+    },
+    {
+      "epoch": 1.3569431500465983,
+      "grad_norm": 1.6455572501469424,
+      "learning_rate": 3.0428028995512602e-05,
+      "loss": 0.7259,
+      "step": 1456
+    },
+    {
+      "epoch": 1.3578751164958063,
+      "grad_norm": 1.654776234985187,
+      "learning_rate": 3.0410769761822577e-05,
+      "loss": 0.7687,
+      "step": 1457
+    },
+    {
+      "epoch": 1.358807082945014,
+      "grad_norm": 1.4206843075363922,
+      "learning_rate": 3.039351052813255e-05,
+      "loss": 0.7231,
+      "step": 1458
+    },
+    {
+      "epoch": 1.3597390493942219,
+      "grad_norm": 1.162624430879057,
+      "learning_rate": 3.0376251294442525e-05,
+      "loss": 0.6717,
+      "step": 1459
+    },
+    {
+      "epoch": 1.3606710158434296,
+      "grad_norm": 1.3773294396505447,
+      "learning_rate": 3.0358992060752506e-05,
+      "loss": 0.6623,
+      "step": 1460
+    },
+    {
+      "epoch": 1.3616029822926374,
+      "grad_norm": 1.050736445571792,
+      "learning_rate": 3.034173282706248e-05,
+      "loss": 0.615,
+      "step": 1461
+    },
+    {
+      "epoch": 1.3625349487418452,
+      "grad_norm": 1.417611631934681,
+      "learning_rate": 3.0324473593372455e-05,
+      "loss": 0.7239,
+      "step": 1462
+    },
+    {
+      "epoch": 1.363466915191053,
+      "grad_norm": 1.4088205158572462,
+      "learning_rate": 3.0307214359682433e-05,
+      "loss": 0.7078,
+      "step": 1463
+    },
+    {
+      "epoch": 1.364398881640261,
+      "grad_norm": 1.2103238990628493,
+      "learning_rate": 3.0289955125992407e-05,
+      "loss": 0.6749,
+      "step": 1464
+    },
+    {
+      "epoch": 1.3653308480894688,
+      "grad_norm": 1.517598593806043,
+      "learning_rate": 3.027269589230238e-05,
+      "loss": 0.7681,
+      "step": 1465
+    },
+    {
+      "epoch": 1.3662628145386766,
+      "grad_norm": 1.3311376040310963,
+      "learning_rate": 3.0255436658612356e-05,
+      "loss": 0.6523,
+      "step": 1466
+    },
+    {
+      "epoch": 1.3671947809878844,
+      "grad_norm": 1.063009235600076,
+      "learning_rate": 3.0238177424922337e-05,
+      "loss": 0.6064,
+      "step": 1467
+    },
+    {
+      "epoch": 1.3681267474370924,
+      "grad_norm": 1.2694388060894486,
+      "learning_rate": 3.022091819123231e-05,
+      "loss": 0.6943,
+      "step": 1468
+    },
+    {
+      "epoch": 1.3690587138863002,
+      "grad_norm": 1.539997227611018,
+      "learning_rate": 3.0203658957542285e-05,
+      "loss": 0.7163,
+      "step": 1469
+    },
+    {
+      "epoch": 1.369990680335508,
+      "grad_norm": 1.041187445225499,
+      "learning_rate": 3.018639972385226e-05,
+      "loss": 0.6803,
+      "step": 1470
+    },
+    {
+      "epoch": 1.3709226467847158,
+      "grad_norm": 1.050801350042604,
+      "learning_rate": 3.016914049016224e-05,
+      "loss": 0.6281,
+      "step": 1471
+    },
+    {
+      "epoch": 1.3718546132339235,
+      "grad_norm": 1.3831525894663346,
+      "learning_rate": 3.0151881256472215e-05,
+      "loss": 0.7196,
+      "step": 1472
+    },
+    {
+      "epoch": 1.3727865796831313,
+      "grad_norm": 1.3299321263214208,
+      "learning_rate": 3.013462202278219e-05,
+      "loss": 0.7127,
+      "step": 1473
+    },
+    {
+      "epoch": 1.3737185461323391,
+      "grad_norm": 1.176993664008512,
+      "learning_rate": 3.0117362789092167e-05,
+      "loss": 0.6384,
+      "step": 1474
+    },
+    {
+      "epoch": 1.3746505125815471,
+      "grad_norm": 1.3079184771137549,
+      "learning_rate": 3.010010355540214e-05,
+      "loss": 0.6993,
+      "step": 1475
+    },
+    {
+      "epoch": 1.375582479030755,
+      "grad_norm": 1.447089167724556,
+      "learning_rate": 3.0082844321712115e-05,
+      "loss": 0.6434,
+      "step": 1476
+    },
+    {
+      "epoch": 1.3765144454799627,
+      "grad_norm": 1.15221109718126,
+      "learning_rate": 3.006558508802209e-05,
+      "loss": 0.6037,
+      "step": 1477
+    },
+    {
+      "epoch": 1.3774464119291705,
+      "grad_norm": 1.2839257042012429,
+      "learning_rate": 3.004832585433207e-05,
+      "loss": 0.6835,
+      "step": 1478
+    },
+    {
+      "epoch": 1.3783783783783785,
+      "grad_norm": 1.5965656724758037,
+      "learning_rate": 3.0031066620642045e-05,
+      "loss": 0.7169,
+      "step": 1479
+    },
+    {
+      "epoch": 1.3793103448275863,
+      "grad_norm": 2.0622628358324215,
+      "learning_rate": 3.001380738695202e-05,
+      "loss": 0.8467,
+      "step": 1480
+    },
+    {
+      "epoch": 1.380242311276794,
+      "grad_norm": 1.575061711015695,
+      "learning_rate": 2.9996548153261994e-05,
+      "loss": 0.7091,
+      "step": 1481
+    },
+    {
+      "epoch": 1.3811742777260019,
+      "grad_norm": 1.466404192303706,
+      "learning_rate": 2.9979288919571975e-05,
+      "loss": 0.7023,
+      "step": 1482
+    },
+    {
+      "epoch": 1.3821062441752097,
+      "grad_norm": 1.3057021672508815,
+      "learning_rate": 2.996202968588195e-05,
+      "loss": 0.6925,
+      "step": 1483
+    },
+    {
+      "epoch": 1.3830382106244175,
+      "grad_norm": 1.3507478479805866,
+      "learning_rate": 2.9944770452191923e-05,
+      "loss": 0.7589,
+      "step": 1484
+    },
+    {
+      "epoch": 1.3839701770736252,
+      "grad_norm": 1.2884321491856237,
+      "learning_rate": 2.99275112185019e-05,
+      "loss": 0.6667,
+      "step": 1485
+    },
+    {
+      "epoch": 1.3849021435228333,
+      "grad_norm": 1.4128459825790871,
+      "learning_rate": 2.9910251984811875e-05,
+      "loss": 0.6953,
+      "step": 1486
+    },
+    {
+      "epoch": 1.385834109972041,
+      "grad_norm": 1.1848704182298018,
+      "learning_rate": 2.989299275112185e-05,
+      "loss": 0.6873,
+      "step": 1487
+    },
+    {
+      "epoch": 1.3867660764212488,
+      "grad_norm": 1.6429794497999894,
+      "learning_rate": 2.9875733517431824e-05,
+      "loss": 0.7304,
+      "step": 1488
+    },
+    {
+      "epoch": 1.3876980428704566,
+      "grad_norm": 1.4442234375607448,
+      "learning_rate": 2.9858474283741805e-05,
+      "loss": 0.7308,
+      "step": 1489
+    },
+    {
+      "epoch": 1.3886300093196646,
+      "grad_norm": 1.1793880473467218,
+      "learning_rate": 2.984121505005178e-05,
+      "loss": 0.7012,
+      "step": 1490
+    },
+    {
+      "epoch": 1.3895619757688724,
+      "grad_norm": 1.5551485689796642,
+      "learning_rate": 2.9823955816361753e-05,
+      "loss": 0.7487,
+      "step": 1491
+    },
+    {
+      "epoch": 1.3904939422180802,
+      "grad_norm": 1.2216678832299062,
+      "learning_rate": 2.9806696582671728e-05,
+      "loss": 0.7194,
+      "step": 1492
+    },
+    {
+      "epoch": 1.391425908667288,
+      "grad_norm": 1.3820805112863366,
+      "learning_rate": 2.978943734898171e-05,
+      "loss": 0.6497,
+      "step": 1493
+    },
+    {
+      "epoch": 1.3923578751164958,
+      "grad_norm": 1.5559493041817283,
+      "learning_rate": 2.9772178115291683e-05,
+      "loss": 0.7119,
+      "step": 1494
+    },
+    {
+      "epoch": 1.3932898415657036,
+      "grad_norm": 1.5566013250321595,
+      "learning_rate": 2.9754918881601657e-05,
+      "loss": 0.7847,
+      "step": 1495
+    },
+    {
+      "epoch": 1.3942218080149114,
+      "grad_norm": 1.0901296667948104,
+      "learning_rate": 2.9737659647911635e-05,
+      "loss": 0.7086,
+      "step": 1496
+    },
+    {
+      "epoch": 1.3951537744641194,
+      "grad_norm": 1.6177310717489113,
+      "learning_rate": 2.972040041422161e-05,
+      "loss": 0.9221,
+      "step": 1497
+    },
+    {
+      "epoch": 1.3960857409133272,
+      "grad_norm": 1.3510018063362947,
+      "learning_rate": 2.9703141180531584e-05,
+      "loss": 0.6294,
+      "step": 1498
+    },
+    {
+      "epoch": 1.397017707362535,
+      "grad_norm": 1.2750913581769905,
+      "learning_rate": 2.9685881946841558e-05,
+      "loss": 0.6733,
+      "step": 1499
+    },
+    {
+      "epoch": 1.3979496738117427,
+      "grad_norm": 1.5280552758072488,
+      "learning_rate": 2.966862271315154e-05,
+      "loss": 0.9504,
+      "step": 1500
+    },
+    {
+      "epoch": 1.3988816402609507,
+      "grad_norm": 1.173954536915472,
+      "learning_rate": 2.9651363479461513e-05,
+      "loss": 0.744,
+      "step": 1501
+    },
+    {
+      "epoch": 1.3998136067101585,
+      "grad_norm": 1.3492511325290324,
+      "learning_rate": 2.9634104245771488e-05,
+      "loss": 0.6394,
+      "step": 1502
+    },
+    {
+      "epoch": 1.4007455731593663,
+      "grad_norm": 1.3509537273938315,
+      "learning_rate": 2.9616845012081462e-05,
+      "loss": 0.6794,
+      "step": 1503
+    },
+    {
+      "epoch": 1.401677539608574,
+      "grad_norm": 1.0962134589403443,
+      "learning_rate": 2.9599585778391443e-05,
+      "loss": 0.6486,
+      "step": 1504
+    },
+    {
+      "epoch": 1.402609506057782,
+      "grad_norm": 1.648792231247164,
+      "learning_rate": 2.9582326544701417e-05,
+      "loss": 0.8347,
+      "step": 1505
+    },
+    {
+      "epoch": 1.4035414725069897,
+      "grad_norm": 1.001538163151916,
+      "learning_rate": 2.956506731101139e-05,
+      "loss": 0.5706,
+      "step": 1506
+    },
+    {
+      "epoch": 1.4044734389561975,
+      "grad_norm": 1.4285913760662683,
+      "learning_rate": 2.954780807732137e-05,
+      "loss": 0.6691,
+      "step": 1507
+    },
+    {
+      "epoch": 1.4054054054054055,
+      "grad_norm": 0.9380207791669793,
+      "learning_rate": 2.9530548843631343e-05,
+      "loss": 0.6092,
+      "step": 1508
+    },
+    {
+      "epoch": 1.4063373718546133,
+      "grad_norm": 1.4125349436690586,
+      "learning_rate": 2.9513289609941318e-05,
+      "loss": 0.8663,
+      "step": 1509
+    },
+    {
+      "epoch": 1.407269338303821,
+      "grad_norm": 1.0263481643386272,
+      "learning_rate": 2.9496030376251292e-05,
+      "loss": 0.5899,
+      "step": 1510
+    },
+    {
+      "epoch": 1.4082013047530288,
+      "grad_norm": 1.6509000788411372,
+      "learning_rate": 2.9478771142561273e-05,
+      "loss": 0.8923,
+      "step": 1511
+    },
+    {
+      "epoch": 1.4091332712022366,
+      "grad_norm": 1.3649055847102667,
+      "learning_rate": 2.9461511908871247e-05,
+      "loss": 0.6936,
+      "step": 1512
+    },
+    {
+      "epoch": 1.4100652376514446,
+      "grad_norm": 1.1662229971049451,
+      "learning_rate": 2.944425267518122e-05,
+      "loss": 0.6918,
+      "step": 1513
+    },
+    {
+      "epoch": 1.4109972041006524,
+      "grad_norm": 1.2429052308283854,
+      "learning_rate": 2.9426993441491203e-05,
+      "loss": 0.6292,
+      "step": 1514
+    },
+    {
+      "epoch": 1.4119291705498602,
+      "grad_norm": 1.31956090386342,
+      "learning_rate": 2.9409734207801177e-05,
+      "loss": 0.7525,
+      "step": 1515
+    },
+    {
+      "epoch": 1.412861136999068,
+      "grad_norm": 1.2065017351695229,
+      "learning_rate": 2.939247497411115e-05,
+      "loss": 0.6549,
+      "step": 1516
+    },
+    {
+      "epoch": 1.4137931034482758,
+      "grad_norm": 0.9933996776878173,
+      "learning_rate": 2.9375215740421126e-05,
+      "loss": 0.7119,
+      "step": 1517
+    },
+    {
+      "epoch": 1.4147250698974836,
+      "grad_norm": 1.077896283741522,
+      "learning_rate": 2.9357956506731103e-05,
+      "loss": 0.594,
+      "step": 1518
+    },
+    {
+      "epoch": 1.4156570363466916,
+      "grad_norm": 1.4785250645343382,
+      "learning_rate": 2.9340697273041078e-05,
+      "loss": 0.6756,
+      "step": 1519
+    },
+    {
+      "epoch": 1.4165890027958994,
+      "grad_norm": 1.6665057149486426,
+      "learning_rate": 2.9323438039351052e-05,
+      "loss": 0.812,
+      "step": 1520
+    },
+    {
+      "epoch": 1.4175209692451072,
+      "grad_norm": 1.7436562794751123,
+      "learning_rate": 2.9306178805661026e-05,
+      "loss": 0.7301,
+      "step": 1521
+    },
+    {
+      "epoch": 1.418452935694315,
+      "grad_norm": 1.725837953157957,
+      "learning_rate": 2.9288919571971007e-05,
+      "loss": 0.7915,
+      "step": 1522
+    },
+    {
+      "epoch": 1.4193849021435228,
+      "grad_norm": 1.092082033575554,
+      "learning_rate": 2.927166033828098e-05,
+      "loss": 0.5868,
+      "step": 1523
+    },
+    {
+      "epoch": 1.4203168685927308,
+      "grad_norm": 1.2968359108845546,
+      "learning_rate": 2.9254401104590956e-05,
+      "loss": 0.6698,
+      "step": 1524
+    },
+    {
+      "epoch": 1.4212488350419386,
+      "grad_norm": 1.411518552852907,
+      "learning_rate": 2.9237141870900937e-05,
+      "loss": 0.6897,
+      "step": 1525
+    },
+    {
+      "epoch": 1.4221808014911463,
+      "grad_norm": 1.2159345237094008,
+      "learning_rate": 2.921988263721091e-05,
+      "loss": 0.6791,
+      "step": 1526
+    },
+    {
+      "epoch": 1.4231127679403541,
+      "grad_norm": 1.090834088769425,
+      "learning_rate": 2.9202623403520885e-05,
+      "loss": 0.6417,
+      "step": 1527
+    },
+    {
+      "epoch": 1.424044734389562,
+      "grad_norm": 1.2133242817191137,
+      "learning_rate": 2.918536416983086e-05,
+      "loss": 0.6403,
+      "step": 1528
+    },
+    {
+      "epoch": 1.4249767008387697,
+      "grad_norm": 1.389177516394,
+      "learning_rate": 2.9168104936140837e-05,
+      "loss": 0.7351,
+      "step": 1529
+    },
+    {
+      "epoch": 1.4259086672879777,
+      "grad_norm": 1.379825578896807,
+      "learning_rate": 2.9150845702450812e-05,
+      "loss": 0.6623,
+      "step": 1530
+    },
+    {
+      "epoch": 1.4268406337371855,
+      "grad_norm": 1.1938995675837658,
+      "learning_rate": 2.9133586468760786e-05,
+      "loss": 0.6757,
+      "step": 1531
+    },
+    {
+      "epoch": 1.4277726001863933,
+      "grad_norm": 1.0292795365917813,
+      "learning_rate": 2.911632723507076e-05,
+      "loss": 0.6047,
+      "step": 1532
+    },
+    {
+      "epoch": 1.428704566635601,
+      "grad_norm": 1.2033466308333367,
+      "learning_rate": 2.909906800138074e-05,
+      "loss": 0.6484,
+      "step": 1533
+    },
+    {
+      "epoch": 1.4296365330848089,
+      "grad_norm": 1.764554129430913,
+      "learning_rate": 2.9081808767690716e-05,
+      "loss": 0.6988,
+      "step": 1534
+    },
+    {
+      "epoch": 1.4305684995340169,
+      "grad_norm": 1.2409729305274877,
+      "learning_rate": 2.906454953400069e-05,
+      "loss": 0.7387,
+      "step": 1535
+    },
+    {
+      "epoch": 1.4315004659832247,
+      "grad_norm": 1.1439366900140613,
+      "learning_rate": 2.904729030031067e-05,
+      "loss": 0.7389,
+      "step": 1536
+    },
+    {
+      "epoch": 1.4324324324324325,
+      "grad_norm": 1.314470261509201,
+      "learning_rate": 2.9030031066620645e-05,
+      "loss": 0.74,
+      "step": 1537
+    },
+    {
+      "epoch": 1.4333643988816402,
+      "grad_norm": 1.061406568278389,
+      "learning_rate": 2.901277183293062e-05,
+      "loss": 0.726,
+      "step": 1538
+    },
+    {
+      "epoch": 1.434296365330848,
+      "grad_norm": 1.4526517994059174,
+      "learning_rate": 2.8995512599240594e-05,
+      "loss": 0.6782,
+      "step": 1539
+    },
+    {
+      "epoch": 1.4352283317800558,
+      "grad_norm": 1.0913279467576507,
+      "learning_rate": 2.897825336555057e-05,
+      "loss": 0.5746,
+      "step": 1540
+    },
+    {
+      "epoch": 1.4361602982292636,
+      "grad_norm": 1.3168650597393097,
+      "learning_rate": 2.8960994131860546e-05,
+      "loss": 0.6597,
+      "step": 1541
+    },
+    {
+      "epoch": 1.4370922646784716,
+      "grad_norm": 1.4353706974900369,
+      "learning_rate": 2.894373489817052e-05,
+      "loss": 0.8263,
+      "step": 1542
+    },
+    {
+      "epoch": 1.4380242311276794,
+      "grad_norm": 1.4380190160255972,
+      "learning_rate": 2.8926475664480494e-05,
+      "loss": 0.6645,
+      "step": 1543
+    },
+    {
+      "epoch": 1.4389561975768872,
+      "grad_norm": 1.1988692660198903,
+      "learning_rate": 2.8909216430790476e-05,
+      "loss": 0.6235,
+      "step": 1544
+    },
+    {
+      "epoch": 1.439888164026095,
+      "grad_norm": 1.2912605476747534,
+      "learning_rate": 2.889195719710045e-05,
+      "loss": 0.679,
+      "step": 1545
+    },
+    {
+      "epoch": 1.440820130475303,
+      "grad_norm": 1.206541782611264,
+      "learning_rate": 2.8874697963410424e-05,
+      "loss": 0.6193,
+      "step": 1546
+    },
+    {
+      "epoch": 1.4417520969245108,
+      "grad_norm": 1.268381376913714,
+      "learning_rate": 2.8857438729720405e-05,
+      "loss": 0.6601,
+      "step": 1547
+    },
+    {
+      "epoch": 1.4426840633737186,
+      "grad_norm": 1.6295893918119015,
+      "learning_rate": 2.884017949603038e-05,
+      "loss": 0.6843,
+      "step": 1548
+    },
+    {
+      "epoch": 1.4436160298229264,
+      "grad_norm": 1.137766327664887,
+      "learning_rate": 2.8822920262340354e-05,
+      "loss": 0.6967,
+      "step": 1549
+    },
+    {
+      "epoch": 1.4445479962721341,
+      "grad_norm": 1.2300928164697897,
+      "learning_rate": 2.8805661028650328e-05,
+      "loss": 0.7014,
+      "step": 1550
+    },
+    {
+      "epoch": 1.445479962721342,
+      "grad_norm": 1.4727513149606155,
+      "learning_rate": 2.8788401794960306e-05,
+      "loss": 0.7241,
+      "step": 1551
+    },
+    {
+      "epoch": 1.4464119291705497,
+      "grad_norm": 1.3817728193375332,
+      "learning_rate": 2.877114256127028e-05,
+      "loss": 0.6399,
+      "step": 1552
+    },
+    {
+      "epoch": 1.4473438956197577,
+      "grad_norm": 1.4232379048886006,
+      "learning_rate": 2.8753883327580254e-05,
+      "loss": 0.7211,
+      "step": 1553
+    },
+    {
+      "epoch": 1.4482758620689655,
+      "grad_norm": 1.6166131053183028,
+      "learning_rate": 2.8736624093890235e-05,
+      "loss": 0.7205,
+      "step": 1554
+    },
+    {
+      "epoch": 1.4492078285181733,
+      "grad_norm": 1.3857320291950994,
+      "learning_rate": 2.871936486020021e-05,
+      "loss": 0.7239,
+      "step": 1555
+    },
+    {
+      "epoch": 1.450139794967381,
+      "grad_norm": 1.1035784872393433,
+      "learning_rate": 2.8702105626510184e-05,
+      "loss": 0.6717,
+      "step": 1556
+    },
+    {
+      "epoch": 1.4510717614165891,
+      "grad_norm": 1.0533275285790846,
+      "learning_rate": 2.8684846392820158e-05,
+      "loss": 0.5844,
+      "step": 1557
+    },
+    {
+      "epoch": 1.452003727865797,
+      "grad_norm": 1.0487445181693136,
+      "learning_rate": 2.866758715913014e-05,
+      "loss": 0.5489,
+      "step": 1558
+    },
+    {
+      "epoch": 1.4529356943150047,
+      "grad_norm": 1.219592840801566,
+      "learning_rate": 2.8650327925440114e-05,
+      "loss": 0.6315,
+      "step": 1559
+    },
+    {
+      "epoch": 1.4538676607642125,
+      "grad_norm": 1.475334903689729,
+      "learning_rate": 2.8633068691750088e-05,
+      "loss": 0.6669,
+      "step": 1560
+    },
+    {
+      "epoch": 1.4547996272134203,
+      "grad_norm": 1.38949531798538,
+      "learning_rate": 2.8615809458060062e-05,
+      "loss": 0.6708,
+      "step": 1561
+    },
+    {
+      "epoch": 1.455731593662628,
+      "grad_norm": 1.4133172377337586,
+      "learning_rate": 2.859855022437004e-05,
+      "loss": 0.6653,
+      "step": 1562
+    },
+    {
+      "epoch": 1.4566635601118358,
+      "grad_norm": 1.3024638823576031,
+      "learning_rate": 2.8581290990680014e-05,
+      "loss": 0.6954,
+      "step": 1563
+    },
+    {
+      "epoch": 1.4575955265610439,
+      "grad_norm": 1.2787935857092214,
+      "learning_rate": 2.856403175698999e-05,
+      "loss": 0.6357,
+      "step": 1564
+    },
+    {
+      "epoch": 1.4585274930102516,
+      "grad_norm": 1.8854183509792923,
+      "learning_rate": 2.854677252329997e-05,
+      "loss": 0.8694,
+      "step": 1565
+    },
+    {
+      "epoch": 1.4594594594594594,
+      "grad_norm": 1.1094775350034873,
+      "learning_rate": 2.8529513289609944e-05,
+      "loss": 0.6786,
+      "step": 1566
+    },
+    {
+      "epoch": 1.4603914259086672,
+      "grad_norm": 2.616278250330937,
+      "learning_rate": 2.8512254055919918e-05,
+      "loss": 0.8902,
+      "step": 1567
+    },
+    {
+      "epoch": 1.4613233923578752,
+      "grad_norm": 1.6766611761677763,
+      "learning_rate": 2.8494994822229892e-05,
+      "loss": 0.8732,
+      "step": 1568
+    },
+    {
+      "epoch": 1.462255358807083,
+      "grad_norm": 1.2123340868758168,
+      "learning_rate": 2.8477735588539873e-05,
+      "loss": 0.6235,
+      "step": 1569
+    },
+    {
+      "epoch": 1.4631873252562908,
+      "grad_norm": 1.0900533778210026,
+      "learning_rate": 2.8460476354849848e-05,
+      "loss": 0.6348,
+      "step": 1570
+    },
+    {
+      "epoch": 1.4641192917054986,
+      "grad_norm": 1.27984101026434,
+      "learning_rate": 2.8443217121159822e-05,
+      "loss": 0.6542,
+      "step": 1571
+    },
+    {
+      "epoch": 1.4650512581547064,
+      "grad_norm": 1.8232854184763883,
+      "learning_rate": 2.8425957887469796e-05,
+      "loss": 0.7895,
+      "step": 1572
+    },
+    {
+      "epoch": 1.4659832246039142,
+      "grad_norm": 1.3247562278802847,
+      "learning_rate": 2.8408698653779774e-05,
+      "loss": 0.7357,
+      "step": 1573
+    },
+    {
+      "epoch": 1.466915191053122,
+      "grad_norm": 1.2242564455002753,
+      "learning_rate": 2.8391439420089748e-05,
+      "loss": 0.639,
+      "step": 1574
+    },
+    {
+      "epoch": 1.46784715750233,
+      "grad_norm": 1.2514087546779797,
+      "learning_rate": 2.8374180186399723e-05,
+      "loss": 0.6656,
+      "step": 1575
+    },
+    {
+      "epoch": 1.4687791239515378,
+      "grad_norm": 1.1735151334447145,
+      "learning_rate": 2.8356920952709704e-05,
+      "loss": 0.6123,
+      "step": 1576
+    },
+    {
+      "epoch": 1.4697110904007455,
+      "grad_norm": 1.1480774685101183,
+      "learning_rate": 2.8339661719019678e-05,
+      "loss": 0.6923,
+      "step": 1577
+    },
+    {
+      "epoch": 1.4706430568499533,
+      "grad_norm": 1.2974513086371868,
+      "learning_rate": 2.8322402485329652e-05,
+      "loss": 0.6535,
+      "step": 1578
+    },
+    {
+      "epoch": 1.4715750232991613,
+      "grad_norm": 1.6008911690561058,
+      "learning_rate": 2.8305143251639626e-05,
+      "loss": 0.7819,
+      "step": 1579
+    },
+    {
+      "epoch": 1.4725069897483691,
+      "grad_norm": 1.027293876619349,
+      "learning_rate": 2.8287884017949608e-05,
+      "loss": 0.6049,
+      "step": 1580
+    },
+    {
+      "epoch": 1.473438956197577,
+      "grad_norm": 1.0871944235554436,
+      "learning_rate": 2.8270624784259582e-05,
+      "loss": 0.6131,
+      "step": 1581
+    },
+    {
+      "epoch": 1.4743709226467847,
+      "grad_norm": 1.234028614362018,
+      "learning_rate": 2.8253365550569556e-05,
+      "loss": 0.6117,
+      "step": 1582
+    },
+    {
+      "epoch": 1.4753028890959925,
+      "grad_norm": 0.9922985859268607,
+      "learning_rate": 2.823610631687953e-05,
+      "loss": 0.5821,
+      "step": 1583
+    },
+    {
+      "epoch": 1.4762348555452003,
+      "grad_norm": 1.0438009608018461,
+      "learning_rate": 2.8218847083189508e-05,
+      "loss": 0.5861,
+      "step": 1584
+    },
+    {
+      "epoch": 1.477166821994408,
+      "grad_norm": 1.1123710329026872,
+      "learning_rate": 2.8201587849499482e-05,
+      "loss": 0.6266,
+      "step": 1585
+    },
+    {
+      "epoch": 1.478098788443616,
+      "grad_norm": 1.000177307954422,
+      "learning_rate": 2.8184328615809457e-05,
+      "loss": 0.6636,
+      "step": 1586
+    },
+    {
+      "epoch": 1.4790307548928239,
+      "grad_norm": 1.2669751924585102,
+      "learning_rate": 2.8167069382119438e-05,
+      "loss": 0.6662,
+      "step": 1587
+    },
+    {
+      "epoch": 1.4799627213420317,
+      "grad_norm": 1.4108052318582616,
+      "learning_rate": 2.8149810148429412e-05,
+      "loss": 0.6421,
+      "step": 1588
+    },
+    {
+      "epoch": 1.4808946877912395,
+      "grad_norm": 1.381495694855027,
+      "learning_rate": 2.8132550914739386e-05,
+      "loss": 0.7405,
+      "step": 1589
+    },
+    {
+      "epoch": 1.4818266542404475,
+      "grad_norm": 1.083626679411281,
+      "learning_rate": 2.811529168104936e-05,
+      "loss": 0.7067,
+      "step": 1590
+    },
+    {
+      "epoch": 1.4827586206896552,
+      "grad_norm": 1.2247758053543085,
+      "learning_rate": 2.809803244735934e-05,
+      "loss": 0.6862,
+      "step": 1591
+    },
+    {
+      "epoch": 1.483690587138863,
+      "grad_norm": 1.5364587720112584,
+      "learning_rate": 2.8080773213669316e-05,
+      "loss": 0.6894,
+      "step": 1592
+    },
+    {
+      "epoch": 1.4846225535880708,
+      "grad_norm": 1.241780389423068,
+      "learning_rate": 2.806351397997929e-05,
+      "loss": 0.6782,
+      "step": 1593
+    },
+    {
+      "epoch": 1.4855545200372786,
+      "grad_norm": 1.1441567325245137,
+      "learning_rate": 2.8046254746289268e-05,
+      "loss": 0.6889,
+      "step": 1594
+    },
+    {
+      "epoch": 1.4864864864864864,
+      "grad_norm": 1.4865173056939582,
+      "learning_rate": 2.8028995512599242e-05,
+      "loss": 0.882,
+      "step": 1595
+    },
+    {
+      "epoch": 1.4874184529356942,
+      "grad_norm": 1.5975626455694452,
+      "learning_rate": 2.8011736278909217e-05,
+      "loss": 0.7019,
+      "step": 1596
+    },
+    {
+      "epoch": 1.4883504193849022,
+      "grad_norm": 1.3971572276949673,
+      "learning_rate": 2.799447704521919e-05,
+      "loss": 0.7237,
+      "step": 1597
+    },
+    {
+      "epoch": 1.48928238583411,
+      "grad_norm": 1.0960793615216782,
+      "learning_rate": 2.7977217811529172e-05,
+      "loss": 0.6351,
+      "step": 1598
+    },
+    {
+      "epoch": 1.4902143522833178,
+      "grad_norm": 1.0043675593925032,
+      "learning_rate": 2.7959958577839146e-05,
+      "loss": 0.6233,
+      "step": 1599
+    },
+    {
+      "epoch": 1.4911463187325256,
+      "grad_norm": 1.4339801234387062,
+      "learning_rate": 2.794269934414912e-05,
+      "loss": 0.6536,
+      "step": 1600
+    },
+    {
+      "epoch": 1.4920782851817336,
+      "grad_norm": 1.3345300263517301,
+      "learning_rate": 2.7925440110459095e-05,
+      "loss": 0.7921,
+      "step": 1601
+    },
+    {
+      "epoch": 1.4930102516309414,
+      "grad_norm": 1.2786940437926917,
+      "learning_rate": 2.7908180876769076e-05,
+      "loss": 0.6916,
+      "step": 1602
+    },
+    {
+      "epoch": 1.4939422180801492,
+      "grad_norm": 1.0833445207153354,
+      "learning_rate": 2.789092164307905e-05,
+      "loss": 0.6112,
+      "step": 1603
+    },
+    {
+      "epoch": 1.494874184529357,
+      "grad_norm": 1.3229650084899984,
+      "learning_rate": 2.7873662409389024e-05,
+      "loss": 0.695,
+      "step": 1604
+    },
+    {
+      "epoch": 1.4958061509785647,
+      "grad_norm": 1.3335140414731783,
+      "learning_rate": 2.7856403175699002e-05,
+      "loss": 0.7089,
+      "step": 1605
+    },
+    {
+      "epoch": 1.4967381174277725,
+      "grad_norm": 1.3803774661581936,
+      "learning_rate": 2.7839143942008976e-05,
+      "loss": 0.7243,
+      "step": 1606
+    },
+    {
+      "epoch": 1.4976700838769803,
+      "grad_norm": 1.2496910522096227,
+      "learning_rate": 2.782188470831895e-05,
+      "loss": 0.6888,
+      "step": 1607
+    },
+    {
+      "epoch": 1.4986020503261883,
+      "grad_norm": 1.2428110448496856,
+      "learning_rate": 2.7804625474628925e-05,
+      "loss": 0.6978,
+      "step": 1608
+    },
+    {
+      "epoch": 1.499534016775396,
+      "grad_norm": 0.8890682346548967,
+      "learning_rate": 2.7787366240938906e-05,
+      "loss": 0.5593,
+      "step": 1609
+    },
+    {
+      "epoch": 1.500465983224604,
+      "grad_norm": 1.2432485814266476,
+      "learning_rate": 2.777010700724888e-05,
+      "loss": 0.6336,
+      "step": 1610
+    },
+    {
+      "epoch": 1.501397949673812,
+      "grad_norm": 1.2541572281912619,
+      "learning_rate": 2.7752847773558855e-05,
+      "loss": 0.5927,
+      "step": 1611
+    },
+    {
+      "epoch": 1.5023299161230197,
+      "grad_norm": 1.546104089554132,
+      "learning_rate": 2.773558853986883e-05,
+      "loss": 0.6768,
+      "step": 1612
+    },
+    {
+      "epoch": 1.5032618825722275,
+      "grad_norm": 1.3596007272764212,
+      "learning_rate": 2.771832930617881e-05,
+      "loss": 0.6555,
+      "step": 1613
+    },
+    {
+      "epoch": 1.5041938490214353,
+      "grad_norm": 1.219687613838576,
+      "learning_rate": 2.7701070072488784e-05,
+      "loss": 0.6265,
+      "step": 1614
+    },
+    {
+      "epoch": 1.505125815470643,
+      "grad_norm": 1.3965883460799928,
+      "learning_rate": 2.768381083879876e-05,
+      "loss": 0.7022,
+      "step": 1615
+    },
+    {
+      "epoch": 1.5060577819198508,
+      "grad_norm": 1.0024675306656592,
+      "learning_rate": 2.7666551605108736e-05,
+      "loss": 0.5864,
+      "step": 1616
+    },
+    {
+      "epoch": 1.5069897483690586,
+      "grad_norm": 1.2371922664085906,
+      "learning_rate": 2.764929237141871e-05,
+      "loss": 0.5841,
+      "step": 1617
+    },
+    {
+      "epoch": 1.5079217148182664,
+      "grad_norm": 1.3152903490661658,
+      "learning_rate": 2.7632033137728685e-05,
+      "loss": 0.7049,
+      "step": 1618
+    },
+    {
+      "epoch": 1.5088536812674742,
+      "grad_norm": 1.2379278069557027,
+      "learning_rate": 2.761477390403866e-05,
+      "loss": 0.7016,
+      "step": 1619
+    },
+    {
+      "epoch": 1.5097856477166822,
+      "grad_norm": 1.5517370811800237,
+      "learning_rate": 2.759751467034864e-05,
+      "loss": 0.7016,
+      "step": 1620
+    },
+    {
+      "epoch": 1.51071761416589,
+      "grad_norm": 1.4404552369945078,
+      "learning_rate": 2.7580255436658614e-05,
+      "loss": 0.6579,
+      "step": 1621
+    },
+    {
+      "epoch": 1.511649580615098,
+      "grad_norm": 1.4656848539440317,
+      "learning_rate": 2.756299620296859e-05,
+      "loss": 0.7065,
+      "step": 1622
+    },
+    {
+      "epoch": 1.5125815470643058,
+      "grad_norm": 1.483616773620366,
+      "learning_rate": 2.7545736969278563e-05,
+      "loss": 0.7129,
+      "step": 1623
+    },
+    {
+      "epoch": 1.5135135135135136,
+      "grad_norm": 1.2431371619000349,
+      "learning_rate": 2.7528477735588544e-05,
+      "loss": 0.7804,
+      "step": 1624
+    },
+    {
+      "epoch": 1.5144454799627214,
+      "grad_norm": 1.1514062019423479,
+      "learning_rate": 2.751121850189852e-05,
+      "loss": 0.6248,
+      "step": 1625
+    },
+    {
+      "epoch": 1.5153774464119292,
+      "grad_norm": 1.2099434051522568,
+      "learning_rate": 2.7493959268208493e-05,
+      "loss": 0.5901,
+      "step": 1626
+    },
+    {
+      "epoch": 1.516309412861137,
+      "grad_norm": 1.2879245623271236,
+      "learning_rate": 2.747670003451847e-05,
+      "loss": 0.7356,
+      "step": 1627
+    },
+    {
+      "epoch": 1.5172413793103448,
+      "grad_norm": 0.9265175427968619,
+      "learning_rate": 2.7459440800828445e-05,
+      "loss": 0.6052,
+      "step": 1628
+    },
+    {
+      "epoch": 1.5181733457595525,
+      "grad_norm": 1.4279267276794818,
+      "learning_rate": 2.744218156713842e-05,
+      "loss": 0.683,
+      "step": 1629
+    },
+    {
+      "epoch": 1.5191053122087603,
+      "grad_norm": 1.2894113149497017,
+      "learning_rate": 2.7424922333448393e-05,
+      "loss": 0.7333,
+      "step": 1630
+    },
+    {
+      "epoch": 1.5200372786579683,
+      "grad_norm": 1.9064300988744205,
+      "learning_rate": 2.7407663099758374e-05,
+      "loss": 0.7088,
+      "step": 1631
+    },
+    {
+      "epoch": 1.5209692451071761,
+      "grad_norm": 1.0711415917208083,
+      "learning_rate": 2.739040386606835e-05,
+      "loss": 0.6204,
+      "step": 1632
+    },
+    {
+      "epoch": 1.521901211556384,
+      "grad_norm": 1.180738951954075,
+      "learning_rate": 2.7373144632378323e-05,
+      "loss": 0.6952,
+      "step": 1633
+    },
+    {
+      "epoch": 1.522833178005592,
+      "grad_norm": 1.1114980786956732,
+      "learning_rate": 2.7355885398688297e-05,
+      "loss": 0.6358,
+      "step": 1634
+    },
+    {
+      "epoch": 1.5237651444547997,
+      "grad_norm": 1.793319786893991,
+      "learning_rate": 2.7338626164998278e-05,
+      "loss": 0.8592,
+      "step": 1635
+    },
+    {
+      "epoch": 1.5246971109040075,
+      "grad_norm": 1.1344578073765976,
+      "learning_rate": 2.7321366931308253e-05,
+      "loss": 0.6456,
+      "step": 1636
+    },
+    {
+      "epoch": 1.5256290773532153,
+      "grad_norm": 1.0631593436997226,
+      "learning_rate": 2.7304107697618227e-05,
+      "loss": 0.6212,
+      "step": 1637
+    },
+    {
+      "epoch": 1.526561043802423,
+      "grad_norm": 1.0103127488237815,
+      "learning_rate": 2.7286848463928204e-05,
+      "loss": 0.6361,
+      "step": 1638
+    },
+    {
+      "epoch": 1.5274930102516309,
+      "grad_norm": 1.4904661273599438,
+      "learning_rate": 2.726958923023818e-05,
+      "loss": 0.6541,
+      "step": 1639
+    },
+    {
+      "epoch": 1.5284249767008387,
+      "grad_norm": 1.3870180644731687,
+      "learning_rate": 2.7252329996548153e-05,
+      "loss": 0.791,
+      "step": 1640
+    },
+    {
+      "epoch": 1.5293569431500464,
+      "grad_norm": 0.9871106415844279,
+      "learning_rate": 2.7235070762858127e-05,
+      "loss": 0.6329,
+      "step": 1641
+    },
+    {
+      "epoch": 1.5302889095992545,
+      "grad_norm": 1.5002122179874264,
+      "learning_rate": 2.721781152916811e-05,
+      "loss": 0.6769,
+      "step": 1642
+    },
+    {
+      "epoch": 1.5312208760484622,
+      "grad_norm": 1.144933949431654,
+      "learning_rate": 2.7200552295478083e-05,
+      "loss": 0.6421,
+      "step": 1643
+    },
+    {
+      "epoch": 1.53215284249767,
+      "grad_norm": 1.3719918897343923,
+      "learning_rate": 2.7183293061788057e-05,
+      "loss": 0.6924,
+      "step": 1644
+    },
+    {
+      "epoch": 1.533084808946878,
+      "grad_norm": 1.2217537010671708,
+      "learning_rate": 2.7166033828098038e-05,
+      "loss": 0.6358,
+      "step": 1645
+    },
+    {
+      "epoch": 1.5340167753960858,
+      "grad_norm": 0.9995933986097731,
+      "learning_rate": 2.7148774594408012e-05,
+      "loss": 0.5932,
+      "step": 1646
+    },
+    {
+      "epoch": 1.5349487418452936,
+      "grad_norm": 1.134005811704675,
+      "learning_rate": 2.7131515360717987e-05,
+      "loss": 0.6219,
+      "step": 1647
+    },
+    {
+      "epoch": 1.5358807082945014,
+      "grad_norm": 1.140655499096173,
+      "learning_rate": 2.711425612702796e-05,
+      "loss": 0.6009,
+      "step": 1648
+    },
+    {
+      "epoch": 1.5368126747437092,
+      "grad_norm": 1.0526214965710194,
+      "learning_rate": 2.709699689333794e-05,
+      "loss": 0.6094,
+      "step": 1649
+    },
+    {
+      "epoch": 1.537744641192917,
+      "grad_norm": 1.2415215858594573,
+      "learning_rate": 2.7079737659647913e-05,
+      "loss": 0.6171,
+      "step": 1650
+    },
+    {
+      "epoch": 1.5386766076421248,
+      "grad_norm": 1.1379036142423589,
+      "learning_rate": 2.7062478425957887e-05,
+      "loss": 0.6435,
+      "step": 1651
+    },
+    {
+      "epoch": 1.5396085740913326,
+      "grad_norm": 1.3947898426349725,
+      "learning_rate": 2.704521919226786e-05,
+      "loss": 0.7231,
+      "step": 1652
+    },
+    {
+      "epoch": 1.5405405405405406,
+      "grad_norm": 1.3681386271196188,
+      "learning_rate": 2.7027959958577843e-05,
+      "loss": 0.6678,
+      "step": 1653
+    },
+    {
+      "epoch": 1.5414725069897484,
+      "grad_norm": 1.1923524740771254,
+      "learning_rate": 2.7010700724887817e-05,
+      "loss": 0.6078,
+      "step": 1654
+    },
+    {
+      "epoch": 1.5424044734389561,
+      "grad_norm": 0.978791936593239,
+      "learning_rate": 2.699344149119779e-05,
+      "loss": 0.6061,
+      "step": 1655
+    },
+    {
+      "epoch": 1.5433364398881642,
+      "grad_norm": 0.9273410085427988,
+      "learning_rate": 2.6976182257507772e-05,
+      "loss": 0.5721,
+      "step": 1656
+    },
+    {
+      "epoch": 1.544268406337372,
+      "grad_norm": 1.3625277897098695,
+      "learning_rate": 2.6958923023817746e-05,
+      "loss": 0.7103,
+      "step": 1657
+    },
+    {
+      "epoch": 1.5452003727865797,
+      "grad_norm": 1.1775965319385153,
+      "learning_rate": 2.694166379012772e-05,
+      "loss": 0.6798,
+      "step": 1658
+    },
+    {
+      "epoch": 1.5461323392357875,
+      "grad_norm": 1.3871526143276205,
+      "learning_rate": 2.6924404556437695e-05,
+      "loss": 0.6347,
+      "step": 1659
+    },
+    {
+      "epoch": 1.5470643056849953,
+      "grad_norm": 1.3155663931905026,
+      "learning_rate": 2.6907145322747673e-05,
+      "loss": 0.6538,
+      "step": 1660
+    },
+    {
+      "epoch": 1.547996272134203,
+      "grad_norm": 1.3277978663790355,
+      "learning_rate": 2.6889886089057647e-05,
+      "loss": 0.6827,
+      "step": 1661
+    },
+    {
+      "epoch": 1.5489282385834109,
+      "grad_norm": 1.2433948182366332,
+      "learning_rate": 2.687262685536762e-05,
+      "loss": 0.6683,
+      "step": 1662
+    },
+    {
+      "epoch": 1.5498602050326187,
+      "grad_norm": 1.2360089706531139,
+      "learning_rate": 2.6855367621677596e-05,
+      "loss": 0.6293,
+      "step": 1663
+    },
+    {
+      "epoch": 1.5507921714818267,
+      "grad_norm": 1.3152871709091323,
+      "learning_rate": 2.6838108387987577e-05,
+      "loss": 0.6686,
+      "step": 1664
+    },
+    {
+      "epoch": 1.5517241379310345,
+      "grad_norm": 1.0501251898372068,
+      "learning_rate": 2.682084915429755e-05,
+      "loss": 0.5537,
+      "step": 1665
+    },
+    {
+      "epoch": 1.5526561043802423,
+      "grad_norm": 1.240023877136927,
+      "learning_rate": 2.6803589920607525e-05,
+      "loss": 0.7192,
+      "step": 1666
+    },
+    {
+      "epoch": 1.5535880708294503,
+      "grad_norm": 1.5703796155258554,
+      "learning_rate": 2.6786330686917503e-05,
+      "loss": 0.8272,
+      "step": 1667
+    },
+    {
+      "epoch": 1.554520037278658,
+      "grad_norm": 1.0119247899075052,
+      "learning_rate": 2.6769071453227477e-05,
+      "loss": 0.5858,
+      "step": 1668
+    },
+    {
+      "epoch": 1.5554520037278659,
+      "grad_norm": 1.2071802569159358,
+      "learning_rate": 2.675181221953745e-05,
+      "loss": 0.6511,
+      "step": 1669
+    },
+    {
+      "epoch": 1.5563839701770736,
+      "grad_norm": 1.4656094817757574,
+      "learning_rate": 2.673455298584743e-05,
+      "loss": 0.7041,
+      "step": 1670
+    },
+    {
+      "epoch": 1.5573159366262814,
+      "grad_norm": 1.3556283622945007,
+      "learning_rate": 2.6717293752157407e-05,
+      "loss": 0.7203,
+      "step": 1671
+    },
+    {
+      "epoch": 1.5582479030754892,
+      "grad_norm": 1.0469996773140187,
+      "learning_rate": 2.670003451846738e-05,
+      "loss": 0.6002,
+      "step": 1672
+    },
+    {
+      "epoch": 1.559179869524697,
+      "grad_norm": 1.360468980726922,
+      "learning_rate": 2.6682775284777355e-05,
+      "loss": 0.653,
+      "step": 1673
+    },
+    {
+      "epoch": 1.5601118359739048,
+      "grad_norm": 1.104847541941314,
+      "learning_rate": 2.666551605108733e-05,
+      "loss": 0.5993,
+      "step": 1674
+    },
+    {
+      "epoch": 1.5610438024231128,
+      "grad_norm": 2.085628798237078,
+      "learning_rate": 2.664825681739731e-05,
+      "loss": 0.7451,
+      "step": 1675
+    },
+    {
+      "epoch": 1.5619757688723206,
+      "grad_norm": 1.84491117948682,
+      "learning_rate": 2.6630997583707285e-05,
+      "loss": 0.8095,
+      "step": 1676
+    },
+    {
+      "epoch": 1.5629077353215284,
+      "grad_norm": 1.520274109617452,
+      "learning_rate": 2.661373835001726e-05,
+      "loss": 0.7029,
+      "step": 1677
+    },
+    {
+      "epoch": 1.5638397017707364,
+      "grad_norm": 1.0652080195719709,
+      "learning_rate": 2.6596479116327237e-05,
+      "loss": 0.6536,
+      "step": 1678
+    },
+    {
+      "epoch": 1.5647716682199442,
+      "grad_norm": 0.9664552217054523,
+      "learning_rate": 2.657921988263721e-05,
+      "loss": 0.5465,
+      "step": 1679
+    },
+    {
+      "epoch": 1.565703634669152,
+      "grad_norm": 1.5291820410848262,
+      "learning_rate": 2.6561960648947186e-05,
+      "loss": 0.8001,
+      "step": 1680
+    },
+    {
+      "epoch": 1.5666356011183598,
+      "grad_norm": 0.869845405355291,
+      "learning_rate": 2.654470141525716e-05,
+      "loss": 0.6063,
+      "step": 1681
+    },
+    {
+      "epoch": 1.5675675675675675,
+      "grad_norm": 1.2330376246203556,
+      "learning_rate": 2.652744218156714e-05,
+      "loss": 0.6621,
+      "step": 1682
+    },
+    {
+      "epoch": 1.5684995340167753,
+      "grad_norm": 0.9723224137623186,
+      "learning_rate": 2.6510182947877115e-05,
+      "loss": 0.621,
+      "step": 1683
+    },
+    {
+      "epoch": 1.5694315004659831,
+      "grad_norm": 1.1920943259904626,
+      "learning_rate": 2.649292371418709e-05,
+      "loss": 0.6316,
+      "step": 1684
+    },
+    {
+      "epoch": 1.570363466915191,
+      "grad_norm": 1.1443142725719502,
+      "learning_rate": 2.647566448049707e-05,
+      "loss": 0.5781,
+      "step": 1685
+    },
+    {
+      "epoch": 1.571295433364399,
+      "grad_norm": 1.483005890009273,
+      "learning_rate": 2.6458405246807045e-05,
+      "loss": 0.6015,
+      "step": 1686
+    },
+    {
+      "epoch": 1.5722273998136067,
+      "grad_norm": 1.387039748007008,
+      "learning_rate": 2.644114601311702e-05,
+      "loss": 0.7395,
+      "step": 1687
+    },
+    {
+      "epoch": 1.5731593662628145,
+      "grad_norm": 1.366028551121252,
+      "learning_rate": 2.6423886779426994e-05,
+      "loss": 0.7303,
+      "step": 1688
+    },
+    {
+      "epoch": 1.5740913327120225,
+      "grad_norm": 1.3076925144459381,
+      "learning_rate": 2.640662754573697e-05,
+      "loss": 0.7576,
+      "step": 1689
+    },
+    {
+      "epoch": 1.5750232991612303,
+      "grad_norm": 1.3205526615640268,
+      "learning_rate": 2.6389368312046945e-05,
+      "loss": 0.6596,
+      "step": 1690
+    },
+    {
+      "epoch": 1.575955265610438,
+      "grad_norm": 1.7227806133259924,
+      "learning_rate": 2.637210907835692e-05,
+      "loss": 0.7403,
+      "step": 1691
+    },
+    {
+      "epoch": 1.5768872320596459,
+      "grad_norm": 1.4072815357781874,
+      "learning_rate": 2.6354849844666894e-05,
+      "loss": 0.6935,
+      "step": 1692
+    },
+    {
+      "epoch": 1.5778191985088537,
+      "grad_norm": 1.291922183454139,
+      "learning_rate": 2.6337590610976875e-05,
+      "loss": 0.6904,
+      "step": 1693
+    },
+    {
+      "epoch": 1.5787511649580614,
+      "grad_norm": 1.0454985084548625,
+      "learning_rate": 2.632033137728685e-05,
+      "loss": 0.5904,
+      "step": 1694
+    },
+    {
+      "epoch": 1.5796831314072692,
+      "grad_norm": 1.0786381976504482,
+      "learning_rate": 2.6303072143596824e-05,
+      "loss": 0.6273,
+      "step": 1695
+    },
+    {
+      "epoch": 1.580615097856477,
+      "grad_norm": 1.5801822563071939,
+      "learning_rate": 2.6285812909906805e-05,
+      "loss": 0.8003,
+      "step": 1696
+    },
+    {
+      "epoch": 1.581547064305685,
+      "grad_norm": 1.4291884916192497,
+      "learning_rate": 2.626855367621678e-05,
+      "loss": 0.7287,
+      "step": 1697
+    },
+    {
+      "epoch": 1.5824790307548928,
+      "grad_norm": 1.0834615839625221,
+      "learning_rate": 2.6251294442526753e-05,
+      "loss": 0.6604,
+      "step": 1698
+    },
+    {
+      "epoch": 1.5834109972041006,
+      "grad_norm": 1.391415273499712,
+      "learning_rate": 2.6234035208836728e-05,
+      "loss": 0.7414,
+      "step": 1699
+    },
+    {
+      "epoch": 1.5843429636533086,
+      "grad_norm": 1.082062198664995,
+      "learning_rate": 2.6216775975146705e-05,
+      "loss": 0.6371,
+      "step": 1700
+    },
+    {
+      "epoch": 1.5852749301025164,
+      "grad_norm": 0.90523842826735,
+      "learning_rate": 2.619951674145668e-05,
+      "loss": 0.5741,
+      "step": 1701
+    },
+    {
+      "epoch": 1.5862068965517242,
+      "grad_norm": 1.1309136237099402,
+      "learning_rate": 2.6182257507766654e-05,
+      "loss": 0.6846,
+      "step": 1702
+    },
+    {
+      "epoch": 1.587138863000932,
+      "grad_norm": 1.3786738581442326,
+      "learning_rate": 2.6164998274076628e-05,
+      "loss": 0.6144,
+      "step": 1703
+    },
+    {
+      "epoch": 1.5880708294501398,
+      "grad_norm": 1.2579152404631226,
+      "learning_rate": 2.614773904038661e-05,
+      "loss": 0.6612,
+      "step": 1704
+    },
+    {
+      "epoch": 1.5890027958993476,
+      "grad_norm": 1.3560818666456103,
+      "learning_rate": 2.6130479806696584e-05,
+      "loss": 0.708,
+      "step": 1705
+    },
+    {
+      "epoch": 1.5899347623485554,
+      "grad_norm": 1.368084341436182,
+      "learning_rate": 2.6113220573006558e-05,
+      "loss": 0.652,
+      "step": 1706
+    },
+    {
+      "epoch": 1.5908667287977631,
+      "grad_norm": 1.4602617588633766,
+      "learning_rate": 2.609596133931654e-05,
+      "loss": 0.674,
+      "step": 1707
+    },
+    {
+      "epoch": 1.5917986952469712,
+      "grad_norm": 1.1072727937207356,
+      "learning_rate": 2.6078702105626513e-05,
+      "loss": 0.7037,
+      "step": 1708
+    },
+    {
+      "epoch": 1.592730661696179,
+      "grad_norm": 1.8889504061801077,
+      "learning_rate": 2.6061442871936487e-05,
+      "loss": 0.7016,
+      "step": 1709
+    },
+    {
+      "epoch": 1.5936626281453867,
+      "grad_norm": 1.398276527537707,
+      "learning_rate": 2.6044183638246462e-05,
+      "loss": 0.7419,
+      "step": 1710
+    },
+    {
+      "epoch": 1.5945945945945947,
+      "grad_norm": 1.5372571293957447,
+      "learning_rate": 2.602692440455644e-05,
+      "loss": 0.6484,
+      "step": 1711
+    },
+    {
+      "epoch": 1.5955265610438025,
+      "grad_norm": 1.1278836802489343,
+      "learning_rate": 2.6009665170866414e-05,
+      "loss": 0.7012,
+      "step": 1712
+    },
+    {
+      "epoch": 1.5964585274930103,
+      "grad_norm": 1.238313260628469,
+      "learning_rate": 2.5992405937176388e-05,
+      "loss": 0.6815,
+      "step": 1713
+    },
+    {
+      "epoch": 1.597390493942218,
+      "grad_norm": 1.1561783184075605,
+      "learning_rate": 2.5975146703486362e-05,
+      "loss": 0.6253,
+      "step": 1714
+    },
+    {
+      "epoch": 1.598322460391426,
+      "grad_norm": 1.0313042009228033,
+      "learning_rate": 2.5957887469796343e-05,
+      "loss": 0.5686,
+      "step": 1715
+    },
+    {
+      "epoch": 1.5992544268406337,
+      "grad_norm": 1.0386813593747728,
+      "learning_rate": 2.5940628236106318e-05,
+      "loss": 0.6178,
+      "step": 1716
+    },
+    {
+      "epoch": 1.6001863932898415,
+      "grad_norm": 1.737254512686763,
+      "learning_rate": 2.5923369002416292e-05,
+      "loss": 0.7115,
+      "step": 1717
+    },
+    {
+      "epoch": 1.6011183597390493,
+      "grad_norm": 1.6452449885689562,
+      "learning_rate": 2.5906109768726273e-05,
+      "loss": 0.7024,
+      "step": 1718
+    },
+    {
+      "epoch": 1.6020503261882573,
+      "grad_norm": 1.6269230777690065,
+      "learning_rate": 2.5888850535036247e-05,
+      "loss": 0.6783,
+      "step": 1719
+    },
+    {
+      "epoch": 1.602982292637465,
+      "grad_norm": 1.362279175639183,
+      "learning_rate": 2.587159130134622e-05,
+      "loss": 0.6417,
+      "step": 1720
+    },
+    {
+      "epoch": 1.6039142590866728,
+      "grad_norm": 1.150673966461896,
+      "learning_rate": 2.5854332067656196e-05,
+      "loss": 0.6506,
+      "step": 1721
+    },
+    {
+      "epoch": 1.6048462255358809,
+      "grad_norm": 1.04546320961687,
+      "learning_rate": 2.5837072833966174e-05,
+      "loss": 0.6554,
+      "step": 1722
+    },
+    {
+      "epoch": 1.6057781919850886,
+      "grad_norm": 1.2023266958710301,
+      "learning_rate": 2.5819813600276148e-05,
+      "loss": 0.6248,
+      "step": 1723
+    },
+    {
+      "epoch": 1.6067101584342964,
+      "grad_norm": 1.6802021200987811,
+      "learning_rate": 2.5802554366586122e-05,
+      "loss": 0.7301,
+      "step": 1724
+    },
+    {
+      "epoch": 1.6076421248835042,
+      "grad_norm": 1.1554810026778783,
+      "learning_rate": 2.5785295132896096e-05,
+      "loss": 0.6227,
+      "step": 1725
+    },
+    {
+      "epoch": 1.608574091332712,
+      "grad_norm": 1.1344174854718498,
+      "learning_rate": 2.5768035899206078e-05,
+      "loss": 0.6357,
+      "step": 1726
+    },
+    {
+      "epoch": 1.6095060577819198,
+      "grad_norm": 1.0062887947554384,
+      "learning_rate": 2.5750776665516052e-05,
+      "loss": 0.6134,
+      "step": 1727
+    },
+    {
+      "epoch": 1.6104380242311276,
+      "grad_norm": 1.2675293596250348,
+      "learning_rate": 2.5733517431826026e-05,
+      "loss": 0.7213,
+      "step": 1728
+    },
+    {
+      "epoch": 1.6113699906803354,
+      "grad_norm": 1.0067736951421091,
+      "learning_rate": 2.5716258198136007e-05,
+      "loss": 0.5779,
+      "step": 1729
+    },
+    {
+      "epoch": 1.6123019571295434,
+      "grad_norm": 1.1005636534813592,
+      "learning_rate": 2.569899896444598e-05,
+      "loss": 0.6037,
+      "step": 1730
+    },
+    {
+      "epoch": 1.6132339235787512,
+      "grad_norm": 1.3403981716520503,
+      "learning_rate": 2.5681739730755956e-05,
+      "loss": 0.6731,
+      "step": 1731
+    },
+    {
+      "epoch": 1.614165890027959,
+      "grad_norm": 1.4442470205686595,
+      "learning_rate": 2.566448049706593e-05,
+      "loss": 0.7078,
+      "step": 1732
+    },
+    {
+      "epoch": 1.615097856477167,
+      "grad_norm": 1.059970571278762,
+      "learning_rate": 2.5647221263375908e-05,
+      "loss": 0.6046,
+      "step": 1733
+    },
+    {
+      "epoch": 1.6160298229263748,
+      "grad_norm": 0.9926081825534805,
+      "learning_rate": 2.5629962029685882e-05,
+      "loss": 0.5644,
+      "step": 1734
+    },
+    {
+      "epoch": 1.6169617893755825,
+      "grad_norm": 1.5522192634392629,
+      "learning_rate": 2.5612702795995856e-05,
+      "loss": 0.6899,
+      "step": 1735
+    },
+    {
+      "epoch": 1.6178937558247903,
+      "grad_norm": 1.476917341901594,
+      "learning_rate": 2.5595443562305837e-05,
+      "loss": 0.7058,
+      "step": 1736
+    },
+    {
+      "epoch": 1.6188257222739981,
+      "grad_norm": 0.8884651317021021,
+      "learning_rate": 2.557818432861581e-05,
+      "loss": 0.5477,
+      "step": 1737
+    },
+    {
+      "epoch": 1.619757688723206,
+      "grad_norm": 1.1163083901918809,
+      "learning_rate": 2.5560925094925786e-05,
+      "loss": 0.6037,
+      "step": 1738
+    },
+    {
+      "epoch": 1.6206896551724137,
+      "grad_norm": 1.4808847347782188,
+      "learning_rate": 2.554366586123576e-05,
+      "loss": 0.7984,
+      "step": 1739
+    },
+    {
+      "epoch": 1.6216216216216215,
+      "grad_norm": 1.209189089626313,
+      "learning_rate": 2.552640662754574e-05,
+      "loss": 0.644,
+      "step": 1740
+    },
+    {
+      "epoch": 1.6225535880708295,
+      "grad_norm": 1.1761983899449917,
+      "learning_rate": 2.5509147393855716e-05,
+      "loss": 0.6983,
+      "step": 1741
+    },
+    {
+      "epoch": 1.6234855545200373,
+      "grad_norm": 1.2083659609803374,
+      "learning_rate": 2.549188816016569e-05,
+      "loss": 0.6957,
+      "step": 1742
+    },
+    {
+      "epoch": 1.624417520969245,
+      "grad_norm": 1.3024684508398487,
+      "learning_rate": 2.5474628926475664e-05,
+      "loss": 0.6613,
+      "step": 1743
+    },
+    {
+      "epoch": 1.625349487418453,
+      "grad_norm": 1.2877921944659902,
+      "learning_rate": 2.5457369692785642e-05,
+      "loss": 0.6781,
+      "step": 1744
+    },
+    {
+      "epoch": 1.6262814538676609,
+      "grad_norm": 1.9005953875433472,
+      "learning_rate": 2.5440110459095616e-05,
+      "loss": 1.0214,
+      "step": 1745
+    },
+    {
+      "epoch": 1.6272134203168687,
+      "grad_norm": 1.2751078056513536,
+      "learning_rate": 2.542285122540559e-05,
+      "loss": 0.661,
+      "step": 1746
+    },
+    {
+      "epoch": 1.6281453867660765,
+      "grad_norm": 1.0359871306403223,
+      "learning_rate": 2.540559199171557e-05,
+      "loss": 0.6388,
+      "step": 1747
+    },
+    {
+      "epoch": 1.6290773532152842,
+      "grad_norm": 0.9697954009021853,
+      "learning_rate": 2.5388332758025546e-05,
+      "loss": 0.6167,
+      "step": 1748
+    },
+    {
+      "epoch": 1.630009319664492,
+      "grad_norm": 1.274601828635102,
+      "learning_rate": 2.537107352433552e-05,
+      "loss": 0.6177,
+      "step": 1749
+    },
+    {
+      "epoch": 1.6309412861136998,
+      "grad_norm": 1.678253211345998,
+      "learning_rate": 2.5353814290645494e-05,
+      "loss": 0.675,
+      "step": 1750
+    },
+    {
+      "epoch": 1.6318732525629076,
+      "grad_norm": 0.9938251273065561,
+      "learning_rate": 2.5336555056955475e-05,
+      "loss": 0.5739,
+      "step": 1751
+    },
+    {
+      "epoch": 1.6328052190121156,
+      "grad_norm": 0.9819624583231132,
+      "learning_rate": 2.531929582326545e-05,
+      "loss": 0.5825,
+      "step": 1752
+    },
+    {
+      "epoch": 1.6337371854613234,
+      "grad_norm": 1.1844414755628616,
+      "learning_rate": 2.5302036589575424e-05,
+      "loss": 0.647,
+      "step": 1753
+    },
+    {
+      "epoch": 1.6346691519105312,
+      "grad_norm": 1.1053838240052911,
+      "learning_rate": 2.5284777355885398e-05,
+      "loss": 0.5517,
+      "step": 1754
+    },
+    {
+      "epoch": 1.6356011183597392,
+      "grad_norm": 1.1099339058032094,
+      "learning_rate": 2.5267518122195376e-05,
+      "loss": 0.6703,
+      "step": 1755
+    },
+    {
+      "epoch": 1.636533084808947,
+      "grad_norm": 1.2520885795688683,
+      "learning_rate": 2.525025888850535e-05,
+      "loss": 0.6061,
+      "step": 1756
+    },
+    {
+      "epoch": 1.6374650512581548,
+      "grad_norm": 1.1949889917091705,
+      "learning_rate": 2.5232999654815325e-05,
+      "loss": 0.6153,
+      "step": 1757
+    },
+    {
+      "epoch": 1.6383970177073626,
+      "grad_norm": 3.5400724177321607,
+      "learning_rate": 2.5215740421125306e-05,
+      "loss": 0.6084,
+      "step": 1758
+    },
+    {
+      "epoch": 1.6393289841565704,
+      "grad_norm": 1.387109236525189,
+      "learning_rate": 2.519848118743528e-05,
+      "loss": 0.6727,
+      "step": 1759
+    },
+    {
+      "epoch": 1.6402609506057781,
+      "grad_norm": 1.266293964913095,
+      "learning_rate": 2.5181221953745254e-05,
+      "loss": 0.6434,
+      "step": 1760
+    },
+    {
+      "epoch": 1.641192917054986,
+      "grad_norm": 1.1337445373130723,
+      "learning_rate": 2.516396272005523e-05,
+      "loss": 0.6372,
+      "step": 1761
+    },
+    {
+      "epoch": 1.6421248835041937,
+      "grad_norm": 0.9670728498902161,
+      "learning_rate": 2.514670348636521e-05,
+      "loss": 0.5735,
+      "step": 1762
+    },
+    {
+      "epoch": 1.6430568499534017,
+      "grad_norm": 0.7816421383406547,
+      "learning_rate": 2.5129444252675184e-05,
+      "loss": 0.5691,
+      "step": 1763
+    },
+    {
+      "epoch": 1.6439888164026095,
+      "grad_norm": 1.1160900309582416,
+      "learning_rate": 2.5112185018985158e-05,
+      "loss": 0.6562,
+      "step": 1764
+    },
+    {
+      "epoch": 1.6449207828518173,
+      "grad_norm": 1.3147029987410832,
+      "learning_rate": 2.5094925785295132e-05,
+      "loss": 0.6616,
+      "step": 1765
+    },
+    {
+      "epoch": 1.6458527493010253,
+      "grad_norm": 1.0727681820171777,
+      "learning_rate": 2.507766655160511e-05,
+      "loss": 0.6106,
+      "step": 1766
+    },
+    {
+      "epoch": 1.646784715750233,
+      "grad_norm": 1.0412029783002497,
+      "learning_rate": 2.5060407317915084e-05,
+      "loss": 0.6254,
+      "step": 1767
+    },
+    {
+      "epoch": 1.647716682199441,
+      "grad_norm": 1.002038077607325,
+      "learning_rate": 2.504314808422506e-05,
+      "loss": 0.535,
+      "step": 1768
+    },
+    {
+      "epoch": 1.6486486486486487,
+      "grad_norm": 1.3062420842273914,
+      "learning_rate": 2.502588885053504e-05,
+      "loss": 0.6816,
+      "step": 1769
+    },
+    {
+      "epoch": 1.6495806150978565,
+      "grad_norm": 1.5686893464767089,
+      "learning_rate": 2.5008629616845014e-05,
+      "loss": 0.6673,
+      "step": 1770
+    },
+    {
+      "epoch": 1.6505125815470643,
+      "grad_norm": 1.0833815263988689,
+      "learning_rate": 2.499137038315499e-05,
+      "loss": 0.5888,
+      "step": 1771
+    },
+    {
+      "epoch": 1.651444547996272,
+      "grad_norm": 1.1667373599816016,
+      "learning_rate": 2.4974111149464966e-05,
+      "loss": 0.7315,
+      "step": 1772
+    },
+    {
+      "epoch": 1.6523765144454798,
+      "grad_norm": 1.1081514953116238,
+      "learning_rate": 2.495685191577494e-05,
+      "loss": 0.6649,
+      "step": 1773
+    },
+    {
+      "epoch": 1.6533084808946876,
+      "grad_norm": 1.1446630656878367,
+      "learning_rate": 2.4939592682084918e-05,
+      "loss": 0.6003,
+      "step": 1774
+    },
+    {
+      "epoch": 1.6542404473438956,
+      "grad_norm": 1.0621993863561598,
+      "learning_rate": 2.4922333448394892e-05,
+      "loss": 0.5983,
+      "step": 1775
+    },
+    {
+      "epoch": 1.6551724137931034,
+      "grad_norm": 1.4098320279216427,
+      "learning_rate": 2.4905074214704867e-05,
+      "loss": 0.704,
+      "step": 1776
+    },
+    {
+      "epoch": 1.6561043802423114,
+      "grad_norm": 1.6304656682212002,
+      "learning_rate": 2.4887814981014844e-05,
+      "loss": 0.7354,
+      "step": 1777
+    },
+    {
+      "epoch": 1.6570363466915192,
+      "grad_norm": 1.7699194345605915,
+      "learning_rate": 2.487055574732482e-05,
+      "loss": 0.7127,
+      "step": 1778
+    },
+    {
+      "epoch": 1.657968313140727,
+      "grad_norm": 1.103545852136451,
+      "learning_rate": 2.4853296513634796e-05,
+      "loss": 0.6306,
+      "step": 1779
+    },
+    {
+      "epoch": 1.6589002795899348,
+      "grad_norm": 1.5798720250664617,
+      "learning_rate": 2.483603727994477e-05,
+      "loss": 0.626,
+      "step": 1780
+    },
+    {
+      "epoch": 1.6598322460391426,
+      "grad_norm": 1.160798029612555,
+      "learning_rate": 2.4818778046254748e-05,
+      "loss": 0.6099,
+      "step": 1781
+    },
+    {
+      "epoch": 1.6607642124883504,
+      "grad_norm": 1.0722010633489931,
+      "learning_rate": 2.4801518812564722e-05,
+      "loss": 0.6424,
+      "step": 1782
+    },
+    {
+      "epoch": 1.6616961789375582,
+      "grad_norm": 1.1188402309543923,
+      "learning_rate": 2.47842595788747e-05,
+      "loss": 0.5805,
+      "step": 1783
+    },
+    {
+      "epoch": 1.662628145386766,
+      "grad_norm": 1.7800033354513818,
+      "learning_rate": 2.4767000345184674e-05,
+      "loss": 0.7993,
+      "step": 1784
+    },
+    {
+      "epoch": 1.6635601118359737,
+      "grad_norm": 1.4336166595150355,
+      "learning_rate": 2.4749741111494652e-05,
+      "loss": 0.6839,
+      "step": 1785
+    },
+    {
+      "epoch": 1.6644920782851818,
+      "grad_norm": 1.0878529980246274,
+      "learning_rate": 2.4732481877804626e-05,
+      "loss": 0.5998,
+      "step": 1786
+    },
+    {
+      "epoch": 1.6654240447343895,
+      "grad_norm": 1.0247845812072947,
+      "learning_rate": 2.47152226441146e-05,
+      "loss": 0.5713,
+      "step": 1787
+    },
+    {
+      "epoch": 1.6663560111835976,
+      "grad_norm": 1.1024094639790762,
+      "learning_rate": 2.469796341042458e-05,
+      "loss": 0.629,
+      "step": 1788
+    },
+    {
+      "epoch": 1.6672879776328053,
+      "grad_norm": 1.123256656895968,
+      "learning_rate": 2.4680704176734553e-05,
+      "loss": 0.6181,
+      "step": 1789
+    },
+    {
+      "epoch": 1.6682199440820131,
+      "grad_norm": 1.244917387734725,
+      "learning_rate": 2.466344494304453e-05,
+      "loss": 0.6123,
+      "step": 1790
+    },
+    {
+      "epoch": 1.669151910531221,
+      "grad_norm": 0.9391514013344326,
+      "learning_rate": 2.4646185709354505e-05,
+      "loss": 0.5897,
+      "step": 1791
+    },
+    {
+      "epoch": 1.6700838769804287,
+      "grad_norm": 1.0618725074899185,
+      "learning_rate": 2.4628926475664482e-05,
+      "loss": 0.6531,
+      "step": 1792
+    },
+    {
+      "epoch": 1.6710158434296365,
+      "grad_norm": 1.075418631020261,
+      "learning_rate": 2.4611667241974457e-05,
+      "loss": 0.6221,
+      "step": 1793
+    },
+    {
+      "epoch": 1.6719478098788443,
+      "grad_norm": 1.3623611895724654,
+      "learning_rate": 2.4594408008284434e-05,
+      "loss": 0.6273,
+      "step": 1794
+    },
+    {
+      "epoch": 1.672879776328052,
+      "grad_norm": 1.222025289172515,
+      "learning_rate": 2.457714877459441e-05,
+      "loss": 0.6195,
+      "step": 1795
+    },
+    {
+      "epoch": 1.6738117427772599,
+      "grad_norm": 1.3186295685134308,
+      "learning_rate": 2.4559889540904386e-05,
+      "loss": 0.673,
+      "step": 1796
+    },
+    {
+      "epoch": 1.6747437092264679,
+      "grad_norm": 0.9221555595630763,
+      "learning_rate": 2.454263030721436e-05,
+      "loss": 0.5118,
+      "step": 1797
+    },
+    {
+      "epoch": 1.6756756756756757,
+      "grad_norm": 1.1401527016611062,
+      "learning_rate": 2.4525371073524335e-05,
+      "loss": 0.6801,
+      "step": 1798
+    },
+    {
+      "epoch": 1.6766076421248837,
+      "grad_norm": 0.8995209050859603,
+      "learning_rate": 2.4508111839834313e-05,
+      "loss": 0.4944,
+      "step": 1799
+    },
+    {
+      "epoch": 1.6775396085740915,
+      "grad_norm": 0.8693711507649573,
+      "learning_rate": 2.4490852606144287e-05,
+      "loss": 0.5555,
+      "step": 1800
+    },
+    {
+      "epoch": 1.6784715750232992,
+      "grad_norm": 1.1190176553757805,
+      "learning_rate": 2.4473593372454264e-05,
+      "loss": 0.6383,
+      "step": 1801
+    },
+    {
+      "epoch": 1.679403541472507,
+      "grad_norm": 0.9132651778307891,
+      "learning_rate": 2.445633413876424e-05,
+      "loss": 0.5561,
+      "step": 1802
+    },
+    {
+      "epoch": 1.6803355079217148,
+      "grad_norm": 0.8834840893094342,
+      "learning_rate": 2.4439074905074216e-05,
+      "loss": 0.5865,
+      "step": 1803
+    },
+    {
+      "epoch": 1.6812674743709226,
+      "grad_norm": 1.327345340030878,
+      "learning_rate": 2.442181567138419e-05,
+      "loss": 0.5907,
+      "step": 1804
+    },
+    {
+      "epoch": 1.6821994408201304,
+      "grad_norm": 1.1911063363439731,
+      "learning_rate": 2.440455643769417e-05,
+      "loss": 0.6908,
+      "step": 1805
+    },
+    {
+      "epoch": 1.6831314072693382,
+      "grad_norm": 1.2022527603013453,
+      "learning_rate": 2.4387297204004146e-05,
+      "loss": 0.6321,
+      "step": 1806
+    },
+    {
+      "epoch": 1.684063373718546,
+      "grad_norm": 1.2387015498784535,
+      "learning_rate": 2.437003797031412e-05,
+      "loss": 0.6108,
+      "step": 1807
+    },
+    {
+      "epoch": 1.684995340167754,
+      "grad_norm": 1.0857868999017488,
+      "learning_rate": 2.4352778736624095e-05,
+      "loss": 0.6185,
+      "step": 1808
+    },
+    {
+      "epoch": 1.6859273066169618,
+      "grad_norm": 1.1216897542981945,
+      "learning_rate": 2.433551950293407e-05,
+      "loss": 0.6323,
+      "step": 1809
+    },
+    {
+      "epoch": 1.6868592730661698,
+      "grad_norm": 1.2804275404281993,
+      "learning_rate": 2.4318260269244047e-05,
+      "loss": 0.6428,
+      "step": 1810
+    },
+    {
+      "epoch": 1.6877912395153776,
+      "grad_norm": 1.4648764401161554,
+      "learning_rate": 2.430100103555402e-05,
+      "loss": 0.7402,
+      "step": 1811
+    },
+    {
+      "epoch": 1.6887232059645854,
+      "grad_norm": 1.3989009537331611,
+      "learning_rate": 2.4283741801864e-05,
+      "loss": 0.5587,
+      "step": 1812
+    },
+    {
+      "epoch": 1.6896551724137931,
+      "grad_norm": 1.129357182270499,
+      "learning_rate": 2.4266482568173973e-05,
+      "loss": 0.6367,
+      "step": 1813
+    },
+    {
+      "epoch": 1.690587138863001,
+      "grad_norm": 1.0672300975791298,
+      "learning_rate": 2.424922333448395e-05,
+      "loss": 0.5773,
+      "step": 1814
+    },
+    {
+      "epoch": 1.6915191053122087,
+      "grad_norm": 1.4720092665886397,
+      "learning_rate": 2.4231964100793925e-05,
+      "loss": 0.7826,
+      "step": 1815
+    },
+    {
+      "epoch": 1.6924510717614165,
+      "grad_norm": 1.146824907293526,
+      "learning_rate": 2.4214704867103903e-05,
+      "loss": 0.5965,
+      "step": 1816
+    },
+    {
+      "epoch": 1.6933830382106243,
+      "grad_norm": 1.0928433647521956,
+      "learning_rate": 2.419744563341388e-05,
+      "loss": 0.6367,
+      "step": 1817
+    },
+    {
+      "epoch": 1.694315004659832,
+      "grad_norm": 1.133766168278627,
+      "learning_rate": 2.4180186399723855e-05,
+      "loss": 0.6317,
+      "step": 1818
+    },
+    {
+      "epoch": 1.69524697110904,
+      "grad_norm": 1.0531296617112893,
+      "learning_rate": 2.416292716603383e-05,
+      "loss": 0.5734,
+      "step": 1819
+    },
+    {
+      "epoch": 1.696178937558248,
+      "grad_norm": 1.0945362116515505,
+      "learning_rate": 2.4145667932343803e-05,
+      "loss": 0.654,
+      "step": 1820
+    },
+    {
+      "epoch": 1.6971109040074557,
+      "grad_norm": 1.1211404712686097,
+      "learning_rate": 2.412840869865378e-05,
+      "loss": 0.6083,
+      "step": 1821
+    },
+    {
+      "epoch": 1.6980428704566637,
+      "grad_norm": 0.9481956615874035,
+      "learning_rate": 2.4111149464963755e-05,
+      "loss": 0.5765,
+      "step": 1822
+    },
+    {
+      "epoch": 1.6989748369058715,
+      "grad_norm": 1.4363848122267164,
+      "learning_rate": 2.4093890231273733e-05,
+      "loss": 0.6934,
+      "step": 1823
+    },
+    {
+      "epoch": 1.6999068033550793,
+      "grad_norm": 1.1916636464943198,
+      "learning_rate": 2.4076630997583707e-05,
+      "loss": 0.7256,
+      "step": 1824
+    },
+    {
+      "epoch": 1.700838769804287,
+      "grad_norm": 1.1176393216846996,
+      "learning_rate": 2.4059371763893685e-05,
+      "loss": 0.6126,
+      "step": 1825
+    },
+    {
+      "epoch": 1.7017707362534948,
+      "grad_norm": 1.1609479078969058,
+      "learning_rate": 2.404211253020366e-05,
+      "loss": 0.5633,
+      "step": 1826
+    },
+    {
+      "epoch": 1.7027027027027026,
+      "grad_norm": 1.4119359384343377,
+      "learning_rate": 2.4024853296513637e-05,
+      "loss": 0.7359,
+      "step": 1827
+    },
+    {
+      "epoch": 1.7036346691519104,
+      "grad_norm": 0.8642954447085843,
+      "learning_rate": 2.4007594062823614e-05,
+      "loss": 0.5671,
+      "step": 1828
+    },
+    {
+      "epoch": 1.7045666356011182,
+      "grad_norm": 1.3074113866605,
+      "learning_rate": 2.399033482913359e-05,
+      "loss": 0.6598,
+      "step": 1829
+    },
+    {
+      "epoch": 1.7054986020503262,
+      "grad_norm": 0.9718752701737383,
+      "learning_rate": 2.3973075595443563e-05,
+      "loss": 0.6311,
+      "step": 1830
+    },
+    {
+      "epoch": 1.706430568499534,
+      "grad_norm": 1.1303982484304747,
+      "learning_rate": 2.3955816361753537e-05,
+      "loss": 0.644,
+      "step": 1831
+    },
+    {
+      "epoch": 1.7073625349487418,
+      "grad_norm": 1.002524555412059,
+      "learning_rate": 2.3938557128063515e-05,
+      "loss": 0.5504,
+      "step": 1832
+    },
+    {
+      "epoch": 1.7082945013979498,
+      "grad_norm": 1.2918662212874363,
+      "learning_rate": 2.392129789437349e-05,
+      "loss": 0.6473,
+      "step": 1833
+    },
+    {
+      "epoch": 1.7092264678471576,
+      "grad_norm": 1.203364162854513,
+      "learning_rate": 2.3904038660683467e-05,
+      "loss": 0.6043,
+      "step": 1834
+    },
+    {
+      "epoch": 1.7101584342963654,
+      "grad_norm": 1.6187652411449944,
+      "learning_rate": 2.388677942699344e-05,
+      "loss": 0.66,
+      "step": 1835
+    },
+    {
+      "epoch": 1.7110904007455732,
+      "grad_norm": 1.5667276046526304,
+      "learning_rate": 2.386952019330342e-05,
+      "loss": 0.6942,
+      "step": 1836
+    },
+    {
+      "epoch": 1.712022367194781,
+      "grad_norm": 1.2256309374547794,
+      "learning_rate": 2.3852260959613397e-05,
+      "loss": 0.64,
+      "step": 1837
+    },
+    {
+      "epoch": 1.7129543336439887,
+      "grad_norm": 1.0553870222950965,
+      "learning_rate": 2.383500172592337e-05,
+      "loss": 0.5814,
+      "step": 1838
+    },
+    {
+      "epoch": 1.7138863000931965,
+      "grad_norm": 1.3299311749879814,
+      "learning_rate": 2.381774249223335e-05,
+      "loss": 0.6776,
+      "step": 1839
+    },
+    {
+      "epoch": 1.7148182665424043,
+      "grad_norm": 1.1737777769142494,
+      "learning_rate": 2.3800483258543323e-05,
+      "loss": 0.6642,
+      "step": 1840
+    },
+    {
+      "epoch": 1.7157502329916123,
+      "grad_norm": 1.1327051691769159,
+      "learning_rate": 2.3783224024853297e-05,
+      "loss": 0.6476,
+      "step": 1841
+    },
+    {
+      "epoch": 1.7166821994408201,
+      "grad_norm": 1.763850064326684,
+      "learning_rate": 2.376596479116327e-05,
+      "loss": 0.7692,
+      "step": 1842
+    },
+    {
+      "epoch": 1.717614165890028,
+      "grad_norm": 1.150202004708297,
+      "learning_rate": 2.374870555747325e-05,
+      "loss": 0.6687,
+      "step": 1843
+    },
+    {
+      "epoch": 1.718546132339236,
+      "grad_norm": 1.3153597340698444,
+      "learning_rate": 2.3731446323783223e-05,
+      "loss": 0.6127,
+      "step": 1844
+    },
+    {
+      "epoch": 1.7194780987884437,
+      "grad_norm": 1.4270325276515936,
+      "learning_rate": 2.37141870900932e-05,
+      "loss": 0.6084,
+      "step": 1845
+    },
+    {
+      "epoch": 1.7204100652376515,
+      "grad_norm": 1.2798680104128126,
+      "learning_rate": 2.3696927856403175e-05,
+      "loss": 0.7178,
+      "step": 1846
+    },
+    {
+      "epoch": 1.7213420316868593,
+      "grad_norm": 1.1518894666430022,
+      "learning_rate": 2.3679668622713153e-05,
+      "loss": 0.5984,
+      "step": 1847
+    },
+    {
+      "epoch": 1.722273998136067,
+      "grad_norm": 0.9934134187314226,
+      "learning_rate": 2.366240938902313e-05,
+      "loss": 0.5866,
+      "step": 1848
+    },
+    {
+      "epoch": 1.7232059645852749,
+      "grad_norm": 1.0878319163560595,
+      "learning_rate": 2.3645150155333105e-05,
+      "loss": 0.6878,
+      "step": 1849
+    },
+    {
+      "epoch": 1.7241379310344827,
+      "grad_norm": 1.1897453861616598,
+      "learning_rate": 2.3627890921643083e-05,
+      "loss": 0.6331,
+      "step": 1850
+    },
+    {
+      "epoch": 1.7250698974836904,
+      "grad_norm": 1.2904951412526642,
+      "learning_rate": 2.3610631687953057e-05,
+      "loss": 0.6468,
+      "step": 1851
+    },
+    {
+      "epoch": 1.7260018639328985,
+      "grad_norm": 0.9638644761348705,
+      "learning_rate": 2.359337245426303e-05,
+      "loss": 0.5923,
+      "step": 1852
+    },
+    {
+      "epoch": 1.7269338303821062,
+      "grad_norm": 1.3663415073630136,
+      "learning_rate": 2.3576113220573005e-05,
+      "loss": 0.63,
+      "step": 1853
+    },
+    {
+      "epoch": 1.727865796831314,
+      "grad_norm": 1.1034754506724276,
+      "learning_rate": 2.3558853986882983e-05,
+      "loss": 0.5827,
+      "step": 1854
+    },
+    {
+      "epoch": 1.728797763280522,
+      "grad_norm": 1.2401623877148598,
+      "learning_rate": 2.3541594753192957e-05,
+      "loss": 0.5994,
+      "step": 1855
+    },
+    {
+      "epoch": 1.7297297297297298,
+      "grad_norm": 0.924436635136358,
+      "learning_rate": 2.3524335519502935e-05,
+      "loss": 0.5322,
+      "step": 1856
+    },
+    {
+      "epoch": 1.7306616961789376,
+      "grad_norm": 1.1917845380071341,
+      "learning_rate": 2.3507076285812913e-05,
+      "loss": 0.6997,
+      "step": 1857
+    },
+    {
+      "epoch": 1.7315936626281454,
+      "grad_norm": 0.8958323494320585,
+      "learning_rate": 2.3489817052122887e-05,
+      "loss": 0.541,
+      "step": 1858
+    },
+    {
+      "epoch": 1.7325256290773532,
+      "grad_norm": 1.1404520294305953,
+      "learning_rate": 2.3472557818432865e-05,
+      "loss": 0.6647,
+      "step": 1859
+    },
+    {
+      "epoch": 1.733457595526561,
+      "grad_norm": 1.6475751380698487,
+      "learning_rate": 2.345529858474284e-05,
+      "loss": 0.6542,
+      "step": 1860
+    },
+    {
+      "epoch": 1.7343895619757688,
+      "grad_norm": 1.4164832019329043,
+      "learning_rate": 2.3438039351052817e-05,
+      "loss": 0.6418,
+      "step": 1861
+    },
+    {
+      "epoch": 1.7353215284249766,
+      "grad_norm": 1.1341473252687682,
+      "learning_rate": 2.342078011736279e-05,
+      "loss": 0.5903,
+      "step": 1862
+    },
+    {
+      "epoch": 1.7362534948741846,
+      "grad_norm": 1.1111012937163045,
+      "learning_rate": 2.3403520883672765e-05,
+      "loss": 0.6165,
+      "step": 1863
+    },
+    {
+      "epoch": 1.7371854613233924,
+      "grad_norm": 0.9442670318037693,
+      "learning_rate": 2.338626164998274e-05,
+      "loss": 0.5883,
+      "step": 1864
+    },
+    {
+      "epoch": 1.7381174277726001,
+      "grad_norm": 0.8834664973716261,
+      "learning_rate": 2.3369002416292717e-05,
+      "loss": 0.5829,
+      "step": 1865
+    },
+    {
+      "epoch": 1.7390493942218082,
+      "grad_norm": 1.5150241711745227,
+      "learning_rate": 2.335174318260269e-05,
+      "loss": 0.6805,
+      "step": 1866
+    },
+    {
+      "epoch": 1.739981360671016,
+      "grad_norm": 1.1023475468253516,
+      "learning_rate": 2.333448394891267e-05,
+      "loss": 0.569,
+      "step": 1867
+    },
+    {
+      "epoch": 1.7409133271202237,
+      "grad_norm": 1.4731633863024602,
+      "learning_rate": 2.3317224715222647e-05,
+      "loss": 0.6083,
+      "step": 1868
+    },
+    {
+      "epoch": 1.7418452935694315,
+      "grad_norm": 1.1528995092797338,
+      "learning_rate": 2.329996548153262e-05,
+      "loss": 0.5853,
+      "step": 1869
+    },
+    {
+      "epoch": 1.7427772600186393,
+      "grad_norm": 1.0469566744199388,
+      "learning_rate": 2.32827062478426e-05,
+      "loss": 0.6132,
+      "step": 1870
+    },
+    {
+      "epoch": 1.743709226467847,
+      "grad_norm": 0.8593575427547471,
+      "learning_rate": 2.3265447014152573e-05,
+      "loss": 0.5534,
+      "step": 1871
+    },
+    {
+      "epoch": 1.7446411929170549,
+      "grad_norm": 0.97039687388224,
+      "learning_rate": 2.324818778046255e-05,
+      "loss": 0.5732,
+      "step": 1872
+    },
+    {
+      "epoch": 1.7455731593662627,
+      "grad_norm": 0.9751490518820631,
+      "learning_rate": 2.3230928546772525e-05,
+      "loss": 0.5889,
+      "step": 1873
+    },
+    {
+      "epoch": 1.7465051258154707,
+      "grad_norm": 1.8297201880091312,
+      "learning_rate": 2.32136693130825e-05,
+      "loss": 0.698,
+      "step": 1874
+    },
+    {
+      "epoch": 1.7474370922646785,
+      "grad_norm": 0.9776619740118907,
+      "learning_rate": 2.3196410079392474e-05,
+      "loss": 0.5581,
+      "step": 1875
+    },
+    {
+      "epoch": 1.7483690587138863,
+      "grad_norm": 1.215652541771515,
+      "learning_rate": 2.317915084570245e-05,
+      "loss": 0.7031,
+      "step": 1876
+    },
+    {
+      "epoch": 1.7493010251630943,
+      "grad_norm": 0.9228872153323642,
+      "learning_rate": 2.316189161201243e-05,
+      "loss": 0.5665,
+      "step": 1877
+    },
+    {
+      "epoch": 1.750232991612302,
+      "grad_norm": 1.0883874802851552,
+      "learning_rate": 2.3144632378322403e-05,
+      "loss": 0.5569,
+      "step": 1878
+    },
+    {
+      "epoch": 1.7511649580615098,
+      "grad_norm": 0.9254378739043877,
+      "learning_rate": 2.312737314463238e-05,
+      "loss": 0.5664,
+      "step": 1879
+    },
+    {
+      "epoch": 1.7520969245107176,
+      "grad_norm": 1.6684956877277997,
+      "learning_rate": 2.3110113910942355e-05,
+      "loss": 0.6817,
+      "step": 1880
+    },
+    {
+      "epoch": 1.7530288909599254,
+      "grad_norm": 1.4334984063461196,
+      "learning_rate": 2.3092854677252333e-05,
+      "loss": 0.7343,
+      "step": 1881
+    },
+    {
+      "epoch": 1.7539608574091332,
+      "grad_norm": 1.2454845685451412,
+      "learning_rate": 2.3075595443562307e-05,
+      "loss": 0.6631,
+      "step": 1882
+    },
+    {
+      "epoch": 1.754892823858341,
+      "grad_norm": 1.061145232027552,
+      "learning_rate": 2.305833620987228e-05,
+      "loss": 0.6015,
+      "step": 1883
+    },
+    {
+      "epoch": 1.7558247903075488,
+      "grad_norm": 1.1355812281976112,
+      "learning_rate": 2.3041076976182256e-05,
+      "loss": 0.5676,
+      "step": 1884
+    },
+    {
+      "epoch": 1.7567567567567568,
+      "grad_norm": 0.8496769682249257,
+      "learning_rate": 2.3023817742492234e-05,
+      "loss": 0.5529,
+      "step": 1885
+    },
+    {
+      "epoch": 1.7576887232059646,
+      "grad_norm": 1.3638865275979815,
+      "learning_rate": 2.3006558508802208e-05,
+      "loss": 0.6851,
+      "step": 1886
+    },
+    {
+      "epoch": 1.7586206896551724,
+      "grad_norm": 0.9206323464900638,
+      "learning_rate": 2.2989299275112186e-05,
+      "loss": 0.5679,
+      "step": 1887
+    },
+    {
+      "epoch": 1.7595526561043804,
+      "grad_norm": 1.518191250610075,
+      "learning_rate": 2.2972040041422163e-05,
+      "loss": 0.7253,
+      "step": 1888
+    },
+    {
+      "epoch": 1.7604846225535882,
+      "grad_norm": 1.065541520257435,
+      "learning_rate": 2.2954780807732138e-05,
+      "loss": 0.5404,
+      "step": 1889
+    },
+    {
+      "epoch": 1.761416589002796,
+      "grad_norm": 1.1811145171031898,
+      "learning_rate": 2.2937521574042115e-05,
+      "loss": 0.7338,
+      "step": 1890
+    },
+    {
+      "epoch": 1.7623485554520038,
+      "grad_norm": 1.4335317063482358,
+      "learning_rate": 2.292026234035209e-05,
+      "loss": 0.6424,
+      "step": 1891
+    },
+    {
+      "epoch": 1.7632805219012115,
+      "grad_norm": 0.8309651894220671,
+      "learning_rate": 2.2903003106662067e-05,
+      "loss": 0.5412,
+      "step": 1892
+    },
+    {
+      "epoch": 1.7642124883504193,
+      "grad_norm": 2.1767125918665298,
+      "learning_rate": 2.288574387297204e-05,
+      "loss": 0.6477,
+      "step": 1893
+    },
+    {
+      "epoch": 1.7651444547996271,
+      "grad_norm": 0.9137020767199686,
+      "learning_rate": 2.2868484639282016e-05,
+      "loss": 0.5612,
+      "step": 1894
+    },
+    {
+      "epoch": 1.766076421248835,
+      "grad_norm": 1.357184691999994,
+      "learning_rate": 2.285122540559199e-05,
+      "loss": 0.7192,
+      "step": 1895
+    },
+    {
+      "epoch": 1.767008387698043,
+      "grad_norm": 1.093226244406422,
+      "learning_rate": 2.2833966171901968e-05,
+      "loss": 0.6154,
+      "step": 1896
+    },
+    {
+      "epoch": 1.7679403541472507,
+      "grad_norm": 0.9399340105948784,
+      "learning_rate": 2.2816706938211945e-05,
+      "loss": 0.5672,
+      "step": 1897
+    },
+    {
+      "epoch": 1.7688723205964585,
+      "grad_norm": 1.0419431882688304,
+      "learning_rate": 2.279944770452192e-05,
+      "loss": 0.6046,
+      "step": 1898
+    },
+    {
+      "epoch": 1.7698042870456665,
+      "grad_norm": 1.462170216895989,
+      "learning_rate": 2.2782188470831897e-05,
+      "loss": 0.6748,
+      "step": 1899
+    },
+    {
+      "epoch": 1.7707362534948743,
+      "grad_norm": 1.0665472073620583,
+      "learning_rate": 2.276492923714187e-05,
+      "loss": 0.5864,
+      "step": 1900
+    },
+    {
+      "epoch": 1.771668219944082,
+      "grad_norm": 1.1455748597502171,
+      "learning_rate": 2.274767000345185e-05,
+      "loss": 0.5627,
+      "step": 1901
+    },
+    {
+      "epoch": 1.7726001863932899,
+      "grad_norm": 1.2750369361826668,
+      "learning_rate": 2.2730410769761824e-05,
+      "loss": 0.6621,
+      "step": 1902
+    },
+    {
+      "epoch": 1.7735321528424977,
+      "grad_norm": 1.0990684169804714,
+      "learning_rate": 2.27131515360718e-05,
+      "loss": 0.6571,
+      "step": 1903
+    },
+    {
+      "epoch": 1.7744641192917054,
+      "grad_norm": 1.153915327711749,
+      "learning_rate": 2.2695892302381776e-05,
+      "loss": 0.695,
+      "step": 1904
+    },
+    {
+      "epoch": 1.7753960857409132,
+      "grad_norm": 0.9991044967885865,
+      "learning_rate": 2.267863306869175e-05,
+      "loss": 0.6426,
+      "step": 1905
+    },
+    {
+      "epoch": 1.776328052190121,
+      "grad_norm": 1.337081564376702,
+      "learning_rate": 2.2661373835001724e-05,
+      "loss": 0.6602,
+      "step": 1906
+    },
+    {
+      "epoch": 1.777260018639329,
+      "grad_norm": 0.8582645348704847,
+      "learning_rate": 2.2644114601311702e-05,
+      "loss": 0.5331,
+      "step": 1907
+    },
+    {
+      "epoch": 1.7781919850885368,
+      "grad_norm": 0.8568362612500653,
+      "learning_rate": 2.262685536762168e-05,
+      "loss": 0.5629,
+      "step": 1908
+    },
+    {
+      "epoch": 1.7791239515377446,
+      "grad_norm": 1.1074040671612873,
+      "learning_rate": 2.2609596133931654e-05,
+      "loss": 0.6479,
+      "step": 1909
+    },
+    {
+      "epoch": 1.7800559179869526,
+      "grad_norm": 0.9263226386493979,
+      "learning_rate": 2.259233690024163e-05,
+      "loss": 0.5869,
+      "step": 1910
+    },
+    {
+      "epoch": 1.7809878844361604,
+      "grad_norm": 1.58261608381876,
+      "learning_rate": 2.2575077666551606e-05,
+      "loss": 0.6624,
+      "step": 1911
+    },
+    {
+      "epoch": 1.7819198508853682,
+      "grad_norm": 1.539267791937469,
+      "learning_rate": 2.2557818432861583e-05,
+      "loss": 0.5793,
+      "step": 1912
+    },
+    {
+      "epoch": 1.782851817334576,
+      "grad_norm": 1.198503452090489,
+      "learning_rate": 2.2540559199171558e-05,
+      "loss": 0.6198,
+      "step": 1913
+    },
+    {
+      "epoch": 1.7837837837837838,
+      "grad_norm": 1.093989251896788,
+      "learning_rate": 2.2523299965481535e-05,
+      "loss": 0.5668,
+      "step": 1914
+    },
+    {
+      "epoch": 1.7847157502329916,
+      "grad_norm": 0.8487886476605565,
+      "learning_rate": 2.250604073179151e-05,
+      "loss": 0.5274,
+      "step": 1915
+    },
+    {
+      "epoch": 1.7856477166821993,
+      "grad_norm": 1.133360505427394,
+      "learning_rate": 2.2488781498101484e-05,
+      "loss": 0.6114,
+      "step": 1916
+    },
+    {
+      "epoch": 1.7865796831314071,
+      "grad_norm": 1.0981259919958533,
+      "learning_rate": 2.2471522264411462e-05,
+      "loss": 0.6735,
+      "step": 1917
+    },
+    {
+      "epoch": 1.7875116495806151,
+      "grad_norm": 1.57566431426833,
+      "learning_rate": 2.2454263030721436e-05,
+      "loss": 0.6675,
+      "step": 1918
+    },
+    {
+      "epoch": 1.788443616029823,
+      "grad_norm": 1.178741010942802,
+      "learning_rate": 2.2437003797031414e-05,
+      "loss": 0.7827,
+      "step": 1919
+    },
+    {
+      "epoch": 1.7893755824790307,
+      "grad_norm": 1.4481018270742043,
+      "learning_rate": 2.2419744563341388e-05,
+      "loss": 0.7855,
+      "step": 1920
+    },
+    {
+      "epoch": 1.7903075489282387,
+      "grad_norm": 1.1777181060187965,
+      "learning_rate": 2.2402485329651366e-05,
+      "loss": 0.6621,
+      "step": 1921
+    },
+    {
+      "epoch": 1.7912395153774465,
+      "grad_norm": 1.227071871424058,
+      "learning_rate": 2.238522609596134e-05,
+      "loss": 0.6401,
+      "step": 1922
+    },
+    {
+      "epoch": 1.7921714818266543,
+      "grad_norm": 0.9033700629673592,
+      "learning_rate": 2.2367966862271318e-05,
+      "loss": 0.5312,
+      "step": 1923
+    },
+    {
+      "epoch": 1.793103448275862,
+      "grad_norm": 1.2996986860233097,
+      "learning_rate": 2.2350707628581292e-05,
+      "loss": 0.5816,
+      "step": 1924
+    },
+    {
+      "epoch": 1.7940354147250699,
+      "grad_norm": 1.9680760896607858,
+      "learning_rate": 2.233344839489127e-05,
+      "loss": 0.797,
+      "step": 1925
+    },
+    {
+      "epoch": 1.7949673811742777,
+      "grad_norm": 1.2472668251612056,
+      "learning_rate": 2.2316189161201244e-05,
+      "loss": 0.6666,
+      "step": 1926
+    },
+    {
+      "epoch": 1.7958993476234855,
+      "grad_norm": 1.1215539488273578,
+      "learning_rate": 2.2298929927511218e-05,
+      "loss": 0.6625,
+      "step": 1927
+    },
+    {
+      "epoch": 1.7968313140726933,
+      "grad_norm": 0.9326493940139129,
+      "learning_rate": 2.2281670693821196e-05,
+      "loss": 0.5936,
+      "step": 1928
+    },
+    {
+      "epoch": 1.7977632805219013,
+      "grad_norm": 1.1807180745971873,
+      "learning_rate": 2.226441146013117e-05,
+      "loss": 0.6228,
+      "step": 1929
+    },
+    {
+      "epoch": 1.798695246971109,
+      "grad_norm": 1.2046393560499216,
+      "learning_rate": 2.2247152226441148e-05,
+      "loss": 0.6933,
+      "step": 1930
+    },
+    {
+      "epoch": 1.7996272134203168,
+      "grad_norm": 1.2262016196627537,
+      "learning_rate": 2.2229892992751122e-05,
+      "loss": 0.6704,
+      "step": 1931
+    },
+    {
+      "epoch": 1.8005591798695249,
+      "grad_norm": 1.238384131441919,
+      "learning_rate": 2.22126337590611e-05,
+      "loss": 0.618,
+      "step": 1932
+    },
+    {
+      "epoch": 1.8014911463187326,
+      "grad_norm": 0.8822476448714115,
+      "learning_rate": 2.2195374525371074e-05,
+      "loss": 0.5705,
+      "step": 1933
+    },
+    {
+      "epoch": 1.8024231127679404,
+      "grad_norm": 0.7593321485271978,
+      "learning_rate": 2.2178115291681052e-05,
+      "loss": 0.5248,
+      "step": 1934
+    },
+    {
+      "epoch": 1.8033550792171482,
+      "grad_norm": 1.1474370409487828,
+      "learning_rate": 2.2160856057991026e-05,
+      "loss": 0.5685,
+      "step": 1935
+    },
+    {
+      "epoch": 1.804287045666356,
+      "grad_norm": 0.7515755400300772,
+      "learning_rate": 2.2143596824301004e-05,
+      "loss": 0.5304,
+      "step": 1936
+    },
+    {
+      "epoch": 1.8052190121155638,
+      "grad_norm": 0.9209374021619059,
+      "learning_rate": 2.2126337590610978e-05,
+      "loss": 0.5392,
+      "step": 1937
+    },
+    {
+      "epoch": 1.8061509785647716,
+      "grad_norm": 0.8810223556116109,
+      "learning_rate": 2.2109078356920952e-05,
+      "loss": 0.5462,
+      "step": 1938
+    },
+    {
+      "epoch": 1.8070829450139794,
+      "grad_norm": 0.8496269737442576,
+      "learning_rate": 2.209181912323093e-05,
+      "loss": 0.5612,
+      "step": 1939
+    },
+    {
+      "epoch": 1.8080149114631874,
+      "grad_norm": 1.1798186838024534,
+      "learning_rate": 2.2074559889540904e-05,
+      "loss": 0.6174,
+      "step": 1940
+    },
+    {
+      "epoch": 1.8089468779123952,
+      "grad_norm": 1.5535500778466846,
+      "learning_rate": 2.2057300655850882e-05,
+      "loss": 0.7269,
+      "step": 1941
+    },
+    {
+      "epoch": 1.809878844361603,
+      "grad_norm": 0.9923120768450053,
+      "learning_rate": 2.2040041422160856e-05,
+      "loss": 0.6175,
+      "step": 1942
+    },
+    {
+      "epoch": 1.810810810810811,
+      "grad_norm": 1.1337941948499748,
+      "learning_rate": 2.2022782188470834e-05,
+      "loss": 0.6043,
+      "step": 1943
+    },
+    {
+      "epoch": 1.8117427772600188,
+      "grad_norm": 1.0029335977778526,
+      "learning_rate": 2.2005522954780808e-05,
+      "loss": 0.5998,
+      "step": 1944
+    },
+    {
+      "epoch": 1.8126747437092265,
+      "grad_norm": 1.0770479728096305,
+      "learning_rate": 2.1988263721090786e-05,
+      "loss": 0.5802,
+      "step": 1945
+    },
+    {
+      "epoch": 1.8136067101584343,
+      "grad_norm": 0.7956173628178483,
+      "learning_rate": 2.197100448740076e-05,
+      "loss": 0.5009,
+      "step": 1946
+    },
+    {
+      "epoch": 1.8145386766076421,
+      "grad_norm": 1.1202069627273175,
+      "learning_rate": 2.1953745253710738e-05,
+      "loss": 0.5479,
+      "step": 1947
+    },
+    {
+      "epoch": 1.81547064305685,
+      "grad_norm": 1.0884099365401205,
+      "learning_rate": 2.1936486020020712e-05,
+      "loss": 0.6093,
+      "step": 1948
+    },
+    {
+      "epoch": 1.8164026095060577,
+      "grad_norm": 1.3779817020091394,
+      "learning_rate": 2.1919226786330686e-05,
+      "loss": 0.6458,
+      "step": 1949
+    },
+    {
+      "epoch": 1.8173345759552655,
+      "grad_norm": 1.3915038296158557,
+      "learning_rate": 2.1901967552640664e-05,
+      "loss": 0.6038,
+      "step": 1950
+    },
+    {
+      "epoch": 1.8182665424044733,
+      "grad_norm": 1.0748606011943511,
+      "learning_rate": 2.188470831895064e-05,
+      "loss": 0.5975,
+      "step": 1951
+    },
+    {
+      "epoch": 1.8191985088536813,
+      "grad_norm": 1.2075249373146624,
+      "learning_rate": 2.1867449085260616e-05,
+      "loss": 0.5888,
+      "step": 1952
+    },
+    {
+      "epoch": 1.820130475302889,
+      "grad_norm": 0.8650546269368669,
+      "learning_rate": 2.185018985157059e-05,
+      "loss": 0.5287,
+      "step": 1953
+    },
+    {
+      "epoch": 1.821062441752097,
+      "grad_norm": 0.9857913924047544,
+      "learning_rate": 2.1832930617880568e-05,
+      "loss": 0.551,
+      "step": 1954
+    },
+    {
+      "epoch": 1.8219944082013049,
+      "grad_norm": 0.8326000675532326,
+      "learning_rate": 2.1815671384190542e-05,
+      "loss": 0.5252,
+      "step": 1955
+    },
+    {
+      "epoch": 1.8229263746505127,
+      "grad_norm": 0.7614729067852178,
+      "learning_rate": 2.179841215050052e-05,
+      "loss": 0.5009,
+      "step": 1956
+    },
+    {
+      "epoch": 1.8238583410997204,
+      "grad_norm": 1.0363783728850249,
+      "learning_rate": 2.1781152916810494e-05,
+      "loss": 0.6053,
+      "step": 1957
+    },
+    {
+      "epoch": 1.8247903075489282,
+      "grad_norm": 1.044405952143845,
+      "learning_rate": 2.1763893683120472e-05,
+      "loss": 0.5742,
+      "step": 1958
+    },
+    {
+      "epoch": 1.825722273998136,
+      "grad_norm": 1.273749831402287,
+      "learning_rate": 2.1746634449430446e-05,
+      "loss": 0.6093,
+      "step": 1959
+    },
+    {
+      "epoch": 1.8266542404473438,
+      "grad_norm": 1.3121226394301886,
+      "learning_rate": 2.172937521574042e-05,
+      "loss": 0.6694,
+      "step": 1960
+    },
+    {
+      "epoch": 1.8275862068965516,
+      "grad_norm": 0.9349057715890917,
+      "learning_rate": 2.1712115982050398e-05,
+      "loss": 0.5999,
+      "step": 1961
+    },
+    {
+      "epoch": 1.8285181733457594,
+      "grad_norm": 1.2536529357188306,
+      "learning_rate": 2.1694856748360373e-05,
+      "loss": 0.6528,
+      "step": 1962
+    },
+    {
+      "epoch": 1.8294501397949674,
+      "grad_norm": 1.0113248196952564,
+      "learning_rate": 2.167759751467035e-05,
+      "loss": 0.5362,
+      "step": 1963
+    },
+    {
+      "epoch": 1.8303821062441752,
+      "grad_norm": 1.1953130891943429,
+      "learning_rate": 2.1660338280980324e-05,
+      "loss": 0.6181,
+      "step": 1964
+    },
+    {
+      "epoch": 1.8313140726933832,
+      "grad_norm": 1.2302616845137173,
+      "learning_rate": 2.1643079047290302e-05,
+      "loss": 0.6601,
+      "step": 1965
+    },
+    {
+      "epoch": 1.832246039142591,
+      "grad_norm": 1.3568358846428987,
+      "learning_rate": 2.1625819813600276e-05,
+      "loss": 0.7284,
+      "step": 1966
+    },
+    {
+      "epoch": 1.8331780055917988,
+      "grad_norm": 1.242384742318171,
+      "learning_rate": 2.1608560579910254e-05,
+      "loss": 0.6528,
+      "step": 1967
+    },
+    {
+      "epoch": 1.8341099720410066,
+      "grad_norm": 0.9734313621054808,
+      "learning_rate": 2.1591301346220232e-05,
+      "loss": 0.5517,
+      "step": 1968
+    },
+    {
+      "epoch": 1.8350419384902144,
+      "grad_norm": 1.0545925282781727,
+      "learning_rate": 2.1574042112530206e-05,
+      "loss": 0.6333,
+      "step": 1969
+    },
+    {
+      "epoch": 1.8359739049394221,
+      "grad_norm": 0.8945364601415817,
+      "learning_rate": 2.155678287884018e-05,
+      "loss": 0.5444,
+      "step": 1970
+    },
+    {
+      "epoch": 1.83690587138863,
+      "grad_norm": 1.3436037357414483,
+      "learning_rate": 2.1539523645150155e-05,
+      "loss": 0.6296,
+      "step": 1971
+    },
+    {
+      "epoch": 1.8378378378378377,
+      "grad_norm": 0.9269471897809335,
+      "learning_rate": 2.1522264411460132e-05,
+      "loss": 0.5447,
+      "step": 1972
+    },
+    {
+      "epoch": 1.8387698042870455,
+      "grad_norm": 1.0067903021472,
+      "learning_rate": 2.1505005177770107e-05,
+      "loss": 0.5971,
+      "step": 1973
+    },
+    {
+      "epoch": 1.8397017707362535,
+      "grad_norm": 0.9304226990189358,
+      "learning_rate": 2.1487745944080084e-05,
+      "loss": 0.5903,
+      "step": 1974
+    },
+    {
+      "epoch": 1.8406337371854613,
+      "grad_norm": 1.1425166057867586,
+      "learning_rate": 2.147048671039006e-05,
+      "loss": 0.5774,
+      "step": 1975
+    },
+    {
+      "epoch": 1.8415657036346693,
+      "grad_norm": 1.2920408895917637,
+      "learning_rate": 2.1453227476700036e-05,
+      "loss": 0.6365,
+      "step": 1976
+    },
+    {
+      "epoch": 1.842497670083877,
+      "grad_norm": 0.9994298720944137,
+      "learning_rate": 2.143596824301001e-05,
+      "loss": 0.5765,
+      "step": 1977
+    },
+    {
+      "epoch": 1.843429636533085,
+      "grad_norm": 0.8908078595888925,
+      "learning_rate": 2.1418709009319988e-05,
+      "loss": 0.5227,
+      "step": 1978
+    },
+    {
+      "epoch": 1.8443616029822927,
+      "grad_norm": 1.08399598393069,
+      "learning_rate": 2.1401449775629963e-05,
+      "loss": 0.6224,
+      "step": 1979
+    },
+    {
+      "epoch": 1.8452935694315005,
+      "grad_norm": 0.897800017256895,
+      "learning_rate": 2.1384190541939937e-05,
+      "loss": 0.54,
+      "step": 1980
+    },
+    {
+      "epoch": 1.8462255358807083,
+      "grad_norm": 1.3048045843204872,
+      "learning_rate": 2.1366931308249915e-05,
+      "loss": 0.5583,
+      "step": 1981
+    },
+    {
+      "epoch": 1.847157502329916,
+      "grad_norm": 1.2710748726442536,
+      "learning_rate": 2.134967207455989e-05,
+      "loss": 0.5775,
+      "step": 1982
+    },
+    {
+      "epoch": 1.8480894687791238,
+      "grad_norm": 1.1039279177194934,
+      "learning_rate": 2.1332412840869866e-05,
+      "loss": 0.6022,
+      "step": 1983
+    },
+    {
+      "epoch": 1.8490214352283316,
+      "grad_norm": 1.0770016648359935,
+      "learning_rate": 2.131515360717984e-05,
+      "loss": 0.6078,
+      "step": 1984
+    },
+    {
+      "epoch": 1.8499534016775396,
+      "grad_norm": 0.8219149218005288,
+      "learning_rate": 2.129789437348982e-05,
+      "loss": 0.5968,
+      "step": 1985
+    },
+    {
+      "epoch": 1.8508853681267474,
+      "grad_norm": 1.012200142306708,
+      "learning_rate": 2.1280635139799793e-05,
+      "loss": 0.6281,
+      "step": 1986
+    },
+    {
+      "epoch": 1.8518173345759554,
+      "grad_norm": 1.2409131222287944,
+      "learning_rate": 2.126337590610977e-05,
+      "loss": 0.6011,
+      "step": 1987
+    },
+    {
+      "epoch": 1.8527493010251632,
+      "grad_norm": 1.0176894050946066,
+      "learning_rate": 2.1246116672419748e-05,
+      "loss": 0.6067,
+      "step": 1988
+    },
+    {
+      "epoch": 1.853681267474371,
+      "grad_norm": 1.3356149677977376,
+      "learning_rate": 2.1228857438729722e-05,
+      "loss": 0.6112,
+      "step": 1989
+    },
+    {
+      "epoch": 1.8546132339235788,
+      "grad_norm": 1.0424369909129465,
+      "learning_rate": 2.1211598205039697e-05,
+      "loss": 0.6472,
+      "step": 1990
+    },
+    {
+      "epoch": 1.8555452003727866,
+      "grad_norm": 0.8661443489325285,
+      "learning_rate": 2.119433897134967e-05,
+      "loss": 0.5579,
+      "step": 1991
+    },
+    {
+      "epoch": 1.8564771668219944,
+      "grad_norm": 1.047966122031923,
+      "learning_rate": 2.117707973765965e-05,
+      "loss": 0.5806,
+      "step": 1992
+    },
+    {
+      "epoch": 1.8574091332712022,
+      "grad_norm": 1.6229990490111457,
+      "learning_rate": 2.1159820503969623e-05,
+      "loss": 0.6406,
+      "step": 1993
+    },
+    {
+      "epoch": 1.85834109972041,
+      "grad_norm": 1.144021716165569,
+      "learning_rate": 2.11425612702796e-05,
+      "loss": 0.5937,
+      "step": 1994
+    },
+    {
+      "epoch": 1.8592730661696177,
+      "grad_norm": 0.9889567325303206,
+      "learning_rate": 2.1125302036589575e-05,
+      "loss": 0.6189,
+      "step": 1995
+    },
+    {
+      "epoch": 1.8602050326188257,
+      "grad_norm": 1.025174166165424,
+      "learning_rate": 2.1108042802899553e-05,
+      "loss": 0.517,
+      "step": 1996
+    },
+    {
+      "epoch": 1.8611369990680335,
+      "grad_norm": 1.467415028637867,
+      "learning_rate": 2.1090783569209527e-05,
+      "loss": 0.5904,
+      "step": 1997
+    },
+    {
+      "epoch": 1.8620689655172413,
+      "grad_norm": 1.1675726778485223,
+      "learning_rate": 2.1073524335519505e-05,
+      "loss": 0.5365,
+      "step": 1998
+    },
+    {
+      "epoch": 1.8630009319664493,
+      "grad_norm": 1.2641330953106678,
+      "learning_rate": 2.1056265101829482e-05,
+      "loss": 0.7179,
+      "step": 1999
+    },
+    {
+      "epoch": 1.8639328984156571,
+      "grad_norm": 0.9860106304957966,
+      "learning_rate": 2.1039005868139457e-05,
+      "loss": 0.543,
+      "step": 2000
+    },
+    {
+      "epoch": 1.864864864864865,
+      "grad_norm": 1.0747776168538723,
+      "learning_rate": 2.102174663444943e-05,
+      "loss": 0.5708,
+      "step": 2001
+    },
+    {
+      "epoch": 1.8657968313140727,
+      "grad_norm": 1.1526143571923282,
+      "learning_rate": 2.1004487400759405e-05,
+      "loss": 0.5785,
+      "step": 2002
+    },
+    {
+      "epoch": 1.8667287977632805,
+      "grad_norm": 1.1357846065867523,
+      "learning_rate": 2.0987228167069383e-05,
+      "loss": 0.6154,
+      "step": 2003
+    },
+    {
+      "epoch": 1.8676607642124883,
+      "grad_norm": 0.8225601703913166,
+      "learning_rate": 2.0969968933379357e-05,
+      "loss": 0.5778,
+      "step": 2004
+    },
+    {
+      "epoch": 1.868592730661696,
+      "grad_norm": 1.1868606813707556,
+      "learning_rate": 2.0952709699689335e-05,
+      "loss": 0.6249,
+      "step": 2005
+    },
+    {
+      "epoch": 1.8695246971109039,
+      "grad_norm": 0.7208231055829799,
+      "learning_rate": 2.093545046599931e-05,
+      "loss": 0.53,
+      "step": 2006
+    },
+    {
+      "epoch": 1.8704566635601119,
+      "grad_norm": 1.2948864974540513,
+      "learning_rate": 2.0918191232309287e-05,
+      "loss": 0.683,
+      "step": 2007
+    },
+    {
+      "epoch": 1.8713886300093197,
+      "grad_norm": 1.213022322945868,
+      "learning_rate": 2.0900931998619264e-05,
+      "loss": 0.6242,
+      "step": 2008
+    },
+    {
+      "epoch": 1.8723205964585274,
+      "grad_norm": 1.3825292207722202,
+      "learning_rate": 2.088367276492924e-05,
+      "loss": 0.5899,
+      "step": 2009
+    },
+    {
+      "epoch": 1.8732525629077355,
+      "grad_norm": 1.0707203962616725,
+      "learning_rate": 2.0866413531239216e-05,
+      "loss": 0.5518,
+      "step": 2010
+    },
+    {
+      "epoch": 1.8741845293569432,
+      "grad_norm": 0.9583850682293293,
+      "learning_rate": 2.084915429754919e-05,
+      "loss": 0.5776,
+      "step": 2011
+    },
+    {
+      "epoch": 1.875116495806151,
+      "grad_norm": 0.8310641724698572,
+      "learning_rate": 2.0831895063859165e-05,
+      "loss": 0.4888,
+      "step": 2012
+    },
+    {
+      "epoch": 1.8760484622553588,
+      "grad_norm": 0.6998269929848457,
+      "learning_rate": 2.081463583016914e-05,
+      "loss": 0.5277,
+      "step": 2013
+    },
+    {
+      "epoch": 1.8769804287045666,
+      "grad_norm": 1.068308489503118,
+      "learning_rate": 2.0797376596479117e-05,
+      "loss": 0.5559,
+      "step": 2014
+    },
+    {
+      "epoch": 1.8779123951537744,
+      "grad_norm": 0.8260013805265983,
+      "learning_rate": 2.078011736278909e-05,
+      "loss": 0.5222,
+      "step": 2015
+    },
+    {
+      "epoch": 1.8788443616029822,
+      "grad_norm": 1.1754466087720743,
+      "learning_rate": 2.076285812909907e-05,
+      "loss": 0.5829,
+      "step": 2016
+    },
+    {
+      "epoch": 1.87977632805219,
+      "grad_norm": 0.98589029705321,
+      "learning_rate": 2.0745598895409043e-05,
+      "loss": 0.5467,
+      "step": 2017
+    },
+    {
+      "epoch": 1.880708294501398,
+      "grad_norm": 1.111334784191955,
+      "learning_rate": 2.072833966171902e-05,
+      "loss": 0.5641,
+      "step": 2018
+    },
+    {
+      "epoch": 1.8816402609506058,
+      "grad_norm": 1.5022233428989058,
+      "learning_rate": 2.0711080428029e-05,
+      "loss": 0.6385,
+      "step": 2019
+    },
+    {
+      "epoch": 1.8825722273998136,
+      "grad_norm": 0.8799199240559048,
+      "learning_rate": 2.0693821194338973e-05,
+      "loss": 0.5542,
+      "step": 2020
+    },
+    {
+      "epoch": 1.8835041938490216,
+      "grad_norm": 1.0407435371255258,
+      "learning_rate": 2.067656196064895e-05,
+      "loss": 0.6519,
+      "step": 2021
+    },
+    {
+      "epoch": 1.8844361602982294,
+      "grad_norm": 0.8972765183184099,
+      "learning_rate": 2.0659302726958925e-05,
+      "loss": 0.5156,
+      "step": 2022
+    },
+    {
+      "epoch": 1.8853681267474371,
+      "grad_norm": 1.0855693280397318,
+      "learning_rate": 2.06420434932689e-05,
+      "loss": 0.5366,
+      "step": 2023
+    },
+    {
+      "epoch": 1.886300093196645,
+      "grad_norm": 1.0059732580141991,
+      "learning_rate": 2.0624784259578873e-05,
+      "loss": 0.5489,
+      "step": 2024
+    },
+    {
+      "epoch": 1.8872320596458527,
+      "grad_norm": 1.0361067900344854,
+      "learning_rate": 2.060752502588885e-05,
+      "loss": 0.5849,
+      "step": 2025
+    },
+    {
+      "epoch": 1.8881640260950605,
+      "grad_norm": 1.0722767827514017,
+      "learning_rate": 2.0590265792198825e-05,
+      "loss": 0.6166,
+      "step": 2026
+    },
+    {
+      "epoch": 1.8890959925442683,
+      "grad_norm": 0.9793613187875295,
+      "learning_rate": 2.0573006558508803e-05,
+      "loss": 0.5414,
+      "step": 2027
+    },
+    {
+      "epoch": 1.890027958993476,
+      "grad_norm": 1.080453490071538,
+      "learning_rate": 2.055574732481878e-05,
+      "loss": 0.6513,
+      "step": 2028
+    },
+    {
+      "epoch": 1.890959925442684,
+      "grad_norm": 0.9902910444459541,
+      "learning_rate": 2.0538488091128755e-05,
+      "loss": 0.5601,
+      "step": 2029
+    },
+    {
+      "epoch": 1.8918918918918919,
+      "grad_norm": 1.0162537606657187,
+      "learning_rate": 2.0521228857438733e-05,
+      "loss": 0.5615,
+      "step": 2030
+    },
+    {
+      "epoch": 1.8928238583410997,
+      "grad_norm": 1.1372023167969398,
+      "learning_rate": 2.0503969623748707e-05,
+      "loss": 0.568,
+      "step": 2031
+    },
+    {
+      "epoch": 1.8937558247903077,
+      "grad_norm": 1.3998056334211773,
+      "learning_rate": 2.0486710390058685e-05,
+      "loss": 0.7791,
+      "step": 2032
+    },
+    {
+      "epoch": 1.8946877912395155,
+      "grad_norm": 1.0649182633355276,
+      "learning_rate": 2.046945115636866e-05,
+      "loss": 0.6519,
+      "step": 2033
+    },
+    {
+      "epoch": 1.8956197576887233,
+      "grad_norm": 0.9090363784041056,
+      "learning_rate": 2.0452191922678633e-05,
+      "loss": 0.5549,
+      "step": 2034
+    },
+    {
+      "epoch": 1.896551724137931,
+      "grad_norm": 1.104750145744069,
+      "learning_rate": 2.0434932688988608e-05,
+      "loss": 0.5856,
+      "step": 2035
+    },
+    {
+      "epoch": 1.8974836905871388,
+      "grad_norm": 0.9024100269432067,
+      "learning_rate": 2.0417673455298585e-05,
+      "loss": 0.5788,
+      "step": 2036
+    },
+    {
+      "epoch": 1.8984156570363466,
+      "grad_norm": 0.9311690886441268,
+      "learning_rate": 2.040041422160856e-05,
+      "loss": 0.5132,
+      "step": 2037
+    },
+    {
+      "epoch": 1.8993476234855544,
+      "grad_norm": 0.877327439509272,
+      "learning_rate": 2.0383154987918537e-05,
+      "loss": 0.5494,
+      "step": 2038
+    },
+    {
+      "epoch": 1.9002795899347622,
+      "grad_norm": 0.8506003071956204,
+      "learning_rate": 2.0365895754228515e-05,
+      "loss": 0.5589,
+      "step": 2039
+    },
+    {
+      "epoch": 1.9012115563839702,
+      "grad_norm": 0.7833481160242857,
+      "learning_rate": 2.034863652053849e-05,
+      "loss": 0.5725,
+      "step": 2040
+    },
+    {
+      "epoch": 1.902143522833178,
+      "grad_norm": 0.9969523849194736,
+      "learning_rate": 2.0331377286848467e-05,
+      "loss": 0.5607,
+      "step": 2041
+    },
+    {
+      "epoch": 1.9030754892823858,
+      "grad_norm": 0.8330436521001651,
+      "learning_rate": 2.031411805315844e-05,
+      "loss": 0.5656,
+      "step": 2042
+    },
+    {
+      "epoch": 1.9040074557315938,
+      "grad_norm": 0.8388515841886794,
+      "learning_rate": 2.029685881946842e-05,
+      "loss": 0.5706,
+      "step": 2043
+    },
+    {
+      "epoch": 1.9049394221808016,
+      "grad_norm": 1.0069010229112443,
+      "learning_rate": 2.0279599585778393e-05,
+      "loss": 0.5519,
+      "step": 2044
+    },
+    {
+      "epoch": 1.9058713886300094,
+      "grad_norm": 1.211466970695071,
+      "learning_rate": 2.0262340352088367e-05,
+      "loss": 0.6312,
+      "step": 2045
+    },
+    {
+      "epoch": 1.9068033550792172,
+      "grad_norm": 0.944595779153726,
+      "learning_rate": 2.024508111839834e-05,
+      "loss": 0.5829,
+      "step": 2046
+    },
+    {
+      "epoch": 1.907735321528425,
+      "grad_norm": 0.8748280890169862,
+      "learning_rate": 2.022782188470832e-05,
+      "loss": 0.6061,
+      "step": 2047
+    },
+    {
+      "epoch": 1.9086672879776327,
+      "grad_norm": 1.0003142442308326,
+      "learning_rate": 2.0210562651018294e-05,
+      "loss": 0.5792,
+      "step": 2048
+    },
+    {
+      "epoch": 1.9095992544268405,
+      "grad_norm": 1.3242180394202876,
+      "learning_rate": 2.019330341732827e-05,
+      "loss": 0.6578,
+      "step": 2049
+    },
+    {
+      "epoch": 1.9105312208760483,
+      "grad_norm": 0.9248488735032984,
+      "learning_rate": 2.017604418363825e-05,
+      "loss": 0.5764,
+      "step": 2050
+    },
+    {
+      "epoch": 1.9114631873252563,
+      "grad_norm": 1.0963385167462363,
+      "learning_rate": 2.0158784949948223e-05,
+      "loss": 0.5695,
+      "step": 2051
+    },
+    {
+      "epoch": 1.9123951537744641,
+      "grad_norm": 0.8858877530346347,
+      "learning_rate": 2.01415257162582e-05,
+      "loss": 0.5318,
+      "step": 2052
+    },
+    {
+      "epoch": 1.913327120223672,
+      "grad_norm": 1.1954804346290728,
+      "learning_rate": 2.0124266482568175e-05,
+      "loss": 0.5596,
+      "step": 2053
+    },
+    {
+      "epoch": 1.91425908667288,
+      "grad_norm": 1.0816007820067384,
+      "learning_rate": 2.0107007248878153e-05,
+      "loss": 0.5617,
+      "step": 2054
+    },
+    {
+      "epoch": 1.9151910531220877,
+      "grad_norm": 0.7799856978672806,
+      "learning_rate": 2.0089748015188127e-05,
+      "loss": 0.5197,
+      "step": 2055
+    },
+    {
+      "epoch": 1.9161230195712955,
+      "grad_norm": 0.9795343280258604,
+      "learning_rate": 2.00724887814981e-05,
+      "loss": 0.526,
+      "step": 2056
+    },
+    {
+      "epoch": 1.9170549860205033,
+      "grad_norm": 0.9916938210918606,
+      "learning_rate": 2.0055229547808076e-05,
+      "loss": 0.6017,
+      "step": 2057
+    },
+    {
+      "epoch": 1.917986952469711,
+      "grad_norm": 1.3913137235146924,
+      "learning_rate": 2.0037970314118053e-05,
+      "loss": 0.6896,
+      "step": 2058
+    },
+    {
+      "epoch": 1.9189189189189189,
+      "grad_norm": 1.0655327338410783,
+      "learning_rate": 2.002071108042803e-05,
+      "loss": 0.5976,
+      "step": 2059
+    },
+    {
+      "epoch": 1.9198508853681266,
+      "grad_norm": 0.6312729993083249,
+      "learning_rate": 2.0003451846738005e-05,
+      "loss": 0.5003,
+      "step": 2060
+    },
+    {
+      "epoch": 1.9207828518173344,
+      "grad_norm": 1.0377782252543253,
+      "learning_rate": 1.9986192613047983e-05,
+      "loss": 0.5762,
+      "step": 2061
+    },
+    {
+      "epoch": 1.9217148182665424,
+      "grad_norm": 0.9069063403142814,
+      "learning_rate": 1.9968933379357957e-05,
+      "loss": 0.5424,
+      "step": 2062
+    },
+    {
+      "epoch": 1.9226467847157502,
+      "grad_norm": 1.163119618051086,
+      "learning_rate": 1.9951674145667935e-05,
+      "loss": 0.5487,
+      "step": 2063
+    },
+    {
+      "epoch": 1.923578751164958,
+      "grad_norm": 0.9100753558691642,
+      "learning_rate": 1.993441491197791e-05,
+      "loss": 0.5388,
+      "step": 2064
+    },
+    {
+      "epoch": 1.924510717614166,
+      "grad_norm": 1.066601217419813,
+      "learning_rate": 1.9917155678287887e-05,
+      "loss": 0.6044,
+      "step": 2065
+    },
+    {
+      "epoch": 1.9254426840633738,
+      "grad_norm": 0.9238240571043157,
+      "learning_rate": 1.989989644459786e-05,
+      "loss": 0.5618,
+      "step": 2066
+    },
+    {
+      "epoch": 1.9263746505125816,
+      "grad_norm": 0.839008188610973,
+      "learning_rate": 1.9882637210907836e-05,
+      "loss": 0.5096,
+      "step": 2067
+    },
+    {
+      "epoch": 1.9273066169617894,
+      "grad_norm": 0.9998429534961613,
+      "learning_rate": 1.986537797721781e-05,
+      "loss": 0.6459,
+      "step": 2068
+    },
+    {
+      "epoch": 1.9282385834109972,
+      "grad_norm": 1.0517554537122884,
+      "learning_rate": 1.9848118743527788e-05,
+      "loss": 0.5466,
+      "step": 2069
+    },
+    {
+      "epoch": 1.929170549860205,
+      "grad_norm": 0.8263719307372474,
+      "learning_rate": 1.9830859509837765e-05,
+      "loss": 0.5502,
+      "step": 2070
+    },
+    {
+      "epoch": 1.9301025163094128,
+      "grad_norm": 0.9862819218502602,
+      "learning_rate": 1.981360027614774e-05,
+      "loss": 0.6436,
+      "step": 2071
+    },
+    {
+      "epoch": 1.9310344827586206,
+      "grad_norm": 0.8857024743694311,
+      "learning_rate": 1.9796341042457717e-05,
+      "loss": 0.5692,
+      "step": 2072
+    },
+    {
+      "epoch": 1.9319664492078286,
+      "grad_norm": 1.2303841937020252,
+      "learning_rate": 1.977908180876769e-05,
+      "loss": 0.5618,
+      "step": 2073
+    },
+    {
+      "epoch": 1.9328984156570364,
+      "grad_norm": 1.2309133177947897,
+      "learning_rate": 1.976182257507767e-05,
+      "loss": 0.5605,
+      "step": 2074
+    },
+    {
+      "epoch": 1.9338303821062441,
+      "grad_norm": 1.2880367384236016,
+      "learning_rate": 1.9744563341387643e-05,
+      "loss": 0.5931,
+      "step": 2075
+    },
+    {
+      "epoch": 1.9347623485554521,
+      "grad_norm": 1.2428494808116008,
+      "learning_rate": 1.972730410769762e-05,
+      "loss": 0.6533,
+      "step": 2076
+    },
+    {
+      "epoch": 1.93569431500466,
+      "grad_norm": 1.4573802813592105,
+      "learning_rate": 1.9710044874007595e-05,
+      "loss": 0.654,
+      "step": 2077
+    },
+    {
+      "epoch": 1.9366262814538677,
+      "grad_norm": 0.9653610039154155,
+      "learning_rate": 1.969278564031757e-05,
+      "loss": 0.5597,
+      "step": 2078
+    },
+    {
+      "epoch": 1.9375582479030755,
+      "grad_norm": 0.8370241788505649,
+      "learning_rate": 1.9675526406627547e-05,
+      "loss": 0.5394,
+      "step": 2079
+    },
+    {
+      "epoch": 1.9384902143522833,
+      "grad_norm": 1.1258210012426917,
+      "learning_rate": 1.9658267172937522e-05,
+      "loss": 0.5608,
+      "step": 2080
+    },
+    {
+      "epoch": 1.939422180801491,
+      "grad_norm": 1.4496464803711915,
+      "learning_rate": 1.96410079392475e-05,
+      "loss": 0.6165,
+      "step": 2081
+    },
+    {
+      "epoch": 1.9403541472506989,
+      "grad_norm": 1.0786407591564404,
+      "learning_rate": 1.9623748705557474e-05,
+      "loss": 0.5797,
+      "step": 2082
+    },
+    {
+      "epoch": 1.9412861136999067,
+      "grad_norm": 1.0493352777140044,
+      "learning_rate": 1.960648947186745e-05,
+      "loss": 0.6183,
+      "step": 2083
+    },
+    {
+      "epoch": 1.9422180801491147,
+      "grad_norm": 0.9617192193331552,
+      "learning_rate": 1.9589230238177426e-05,
+      "loss": 0.5786,
+      "step": 2084
+    },
+    {
+      "epoch": 1.9431500465983225,
+      "grad_norm": 1.2626541412619758,
+      "learning_rate": 1.9571971004487403e-05,
+      "loss": 0.6119,
+      "step": 2085
+    },
+    {
+      "epoch": 1.9440820130475303,
+      "grad_norm": 1.0546329422105452,
+      "learning_rate": 1.9554711770797378e-05,
+      "loss": 0.6109,
+      "step": 2086
+    },
+    {
+      "epoch": 1.9450139794967383,
+      "grad_norm": 1.1157264171267396,
+      "learning_rate": 1.9537452537107352e-05,
+      "loss": 0.6015,
+      "step": 2087
+    },
+    {
+      "epoch": 1.945945945945946,
+      "grad_norm": 0.9469355927327956,
+      "learning_rate": 1.9520193303417326e-05,
+      "loss": 0.5515,
+      "step": 2088
+    },
+    {
+      "epoch": 1.9468779123951538,
+      "grad_norm": 1.074240153490205,
+      "learning_rate": 1.9502934069727304e-05,
+      "loss": 0.5943,
+      "step": 2089
+    },
+    {
+      "epoch": 1.9478098788443616,
+      "grad_norm": 1.1780032259705728,
+      "learning_rate": 1.948567483603728e-05,
+      "loss": 0.599,
+      "step": 2090
+    },
+    {
+      "epoch": 1.9487418452935694,
+      "grad_norm": 1.0661853072837046,
+      "learning_rate": 1.9468415602347256e-05,
+      "loss": 0.6655,
+      "step": 2091
+    },
+    {
+      "epoch": 1.9496738117427772,
+      "grad_norm": 0.9704404764637414,
+      "learning_rate": 1.9451156368657234e-05,
+      "loss": 0.5746,
+      "step": 2092
+    },
+    {
+      "epoch": 1.950605778191985,
+      "grad_norm": 0.8555382877231249,
+      "learning_rate": 1.9433897134967208e-05,
+      "loss": 0.5468,
+      "step": 2093
+    },
+    {
+      "epoch": 1.9515377446411928,
+      "grad_norm": 0.9864101457731763,
+      "learning_rate": 1.9416637901277185e-05,
+      "loss": 0.5373,
+      "step": 2094
+    },
+    {
+      "epoch": 1.9524697110904008,
+      "grad_norm": 0.7104258911767494,
+      "learning_rate": 1.939937866758716e-05,
+      "loss": 0.5199,
+      "step": 2095
+    },
+    {
+      "epoch": 1.9534016775396086,
+      "grad_norm": 1.0671456682103353,
+      "learning_rate": 1.9382119433897137e-05,
+      "loss": 0.5551,
+      "step": 2096
+    },
+    {
+      "epoch": 1.9543336439888164,
+      "grad_norm": 0.9660497873694887,
+      "learning_rate": 1.9364860200207112e-05,
+      "loss": 0.5821,
+      "step": 2097
+    },
+    {
+      "epoch": 1.9552656104380244,
+      "grad_norm": 0.8513739869572055,
+      "learning_rate": 1.9347600966517086e-05,
+      "loss": 0.5479,
+      "step": 2098
+    },
+    {
+      "epoch": 1.9561975768872322,
+      "grad_norm": 1.1598295882732985,
+      "learning_rate": 1.9330341732827064e-05,
+      "loss": 0.6186,
+      "step": 2099
+    },
+    {
+      "epoch": 1.95712954333644,
+      "grad_norm": 1.4451489573441703,
+      "learning_rate": 1.9313082499137038e-05,
+      "loss": 0.6814,
+      "step": 2100
+    },
+    {
+      "epoch": 1.9580615097856477,
+      "grad_norm": 0.8254373368037576,
+      "learning_rate": 1.9295823265447016e-05,
+      "loss": 0.4983,
+      "step": 2101
+    },
+    {
+      "epoch": 1.9589934762348555,
+      "grad_norm": 0.8104698937724144,
+      "learning_rate": 1.927856403175699e-05,
+      "loss": 0.4894,
+      "step": 2102
+    },
+    {
+      "epoch": 1.9599254426840633,
+      "grad_norm": 1.2678566370587647,
+      "learning_rate": 1.9261304798066968e-05,
+      "loss": 0.5903,
+      "step": 2103
+    },
+    {
+      "epoch": 1.9608574091332711,
+      "grad_norm": 1.0186712139934984,
+      "learning_rate": 1.9244045564376942e-05,
+      "loss": 0.576,
+      "step": 2104
+    },
+    {
+      "epoch": 1.961789375582479,
+      "grad_norm": 1.046595286387772,
+      "learning_rate": 1.922678633068692e-05,
+      "loss": 0.5824,
+      "step": 2105
+    },
+    {
+      "epoch": 1.962721342031687,
+      "grad_norm": 1.459544426496336,
+      "learning_rate": 1.9209527096996894e-05,
+      "loss": 0.5755,
+      "step": 2106
+    },
+    {
+      "epoch": 1.9636533084808947,
+      "grad_norm": 1.2398606928979499,
+      "learning_rate": 1.919226786330687e-05,
+      "loss": 0.5511,
+      "step": 2107
+    },
+    {
+      "epoch": 1.9645852749301025,
+      "grad_norm": 0.7805635354899032,
+      "learning_rate": 1.9175008629616846e-05,
+      "loss": 0.6029,
+      "step": 2108
+    },
+    {
+      "epoch": 1.9655172413793105,
+      "grad_norm": 1.350832787476808,
+      "learning_rate": 1.915774939592682e-05,
+      "loss": 0.6548,
+      "step": 2109
+    },
+    {
+      "epoch": 1.9664492078285183,
+      "grad_norm": 1.074978383548324,
+      "learning_rate": 1.9140490162236798e-05,
+      "loss": 0.5355,
+      "step": 2110
+    },
+    {
+      "epoch": 1.967381174277726,
+      "grad_norm": 0.9740676458715976,
+      "learning_rate": 1.9123230928546772e-05,
+      "loss": 0.5704,
+      "step": 2111
+    },
+    {
+      "epoch": 1.9683131407269339,
+      "grad_norm": 0.8084312904229535,
+      "learning_rate": 1.910597169485675e-05,
+      "loss": 0.5472,
+      "step": 2112
+    },
+    {
+      "epoch": 1.9692451071761417,
+      "grad_norm": 1.1081136695138898,
+      "learning_rate": 1.9088712461166724e-05,
+      "loss": 0.6042,
+      "step": 2113
+    },
+    {
+      "epoch": 1.9701770736253494,
+      "grad_norm": 0.8269152005745667,
+      "learning_rate": 1.9071453227476702e-05,
+      "loss": 0.5186,
+      "step": 2114
+    },
+    {
+      "epoch": 1.9711090400745572,
+      "grad_norm": 1.1909996546415142,
+      "learning_rate": 1.9054193993786676e-05,
+      "loss": 0.6097,
+      "step": 2115
+    },
+    {
+      "epoch": 1.972041006523765,
+      "grad_norm": 1.0917568932224977,
+      "learning_rate": 1.9036934760096654e-05,
+      "loss": 0.5485,
+      "step": 2116
+    },
+    {
+      "epoch": 1.972972972972973,
+      "grad_norm": 1.1078259833136614,
+      "learning_rate": 1.9019675526406628e-05,
+      "loss": 0.7332,
+      "step": 2117
+    },
+    {
+      "epoch": 1.9739049394221808,
+      "grad_norm": 0.9758931246415697,
+      "learning_rate": 1.9002416292716606e-05,
+      "loss": 0.5501,
+      "step": 2118
+    },
+    {
+      "epoch": 1.9748369058713886,
+      "grad_norm": 0.6572694653576517,
+      "learning_rate": 1.898515705902658e-05,
+      "loss": 0.4903,
+      "step": 2119
+    },
+    {
+      "epoch": 1.9757688723205966,
+      "grad_norm": 1.3391613284300623,
+      "learning_rate": 1.8967897825336554e-05,
+      "loss": 0.5351,
+      "step": 2120
+    },
+    {
+      "epoch": 1.9767008387698044,
+      "grad_norm": 0.8970846991309143,
+      "learning_rate": 1.8950638591646532e-05,
+      "loss": 0.5889,
+      "step": 2121
+    },
+    {
+      "epoch": 1.9776328052190122,
+      "grad_norm": 0.9970316551450316,
+      "learning_rate": 1.8933379357956506e-05,
+      "loss": 0.6178,
+      "step": 2122
+    },
+    {
+      "epoch": 1.97856477166822,
+      "grad_norm": 1.0039500320999704,
+      "learning_rate": 1.8916120124266484e-05,
+      "loss": 0.5563,
+      "step": 2123
+    },
+    {
+      "epoch": 1.9794967381174278,
+      "grad_norm": 0.6875082306540664,
+      "learning_rate": 1.8898860890576458e-05,
+      "loss": 0.5101,
+      "step": 2124
+    },
+    {
+      "epoch": 1.9804287045666356,
+      "grad_norm": 0.8537439159548043,
+      "learning_rate": 1.8881601656886436e-05,
+      "loss": 0.523,
+      "step": 2125
+    },
+    {
+      "epoch": 1.9813606710158433,
+      "grad_norm": 0.9757022910018104,
+      "learning_rate": 1.886434242319641e-05,
+      "loss": 0.6009,
+      "step": 2126
+    },
+    {
+      "epoch": 1.9822926374650511,
+      "grad_norm": 1.194395283897404,
+      "learning_rate": 1.8847083189506388e-05,
+      "loss": 0.6589,
+      "step": 2127
+    },
+    {
+      "epoch": 1.983224603914259,
+      "grad_norm": 0.9262163524590606,
+      "learning_rate": 1.8829823955816362e-05,
+      "loss": 0.5942,
+      "step": 2128
+    },
+    {
+      "epoch": 1.984156570363467,
+      "grad_norm": 1.0931926216966186,
+      "learning_rate": 1.881256472212634e-05,
+      "loss": 0.6068,
+      "step": 2129
+    },
+    {
+      "epoch": 1.9850885368126747,
+      "grad_norm": 0.8568075716577603,
+      "learning_rate": 1.8795305488436314e-05,
+      "loss": 0.5705,
+      "step": 2130
+    },
+    {
+      "epoch": 1.9860205032618827,
+      "grad_norm": 0.89547209322041,
+      "learning_rate": 1.877804625474629e-05,
+      "loss": 0.5062,
+      "step": 2131
+    },
+    {
+      "epoch": 1.9869524697110905,
+      "grad_norm": 1.0553710274039452,
+      "learning_rate": 1.8760787021056266e-05,
+      "loss": 0.5603,
+      "step": 2132
+    },
+    {
+      "epoch": 1.9878844361602983,
+      "grad_norm": 0.8864337439236525,
+      "learning_rate": 1.874352778736624e-05,
+      "loss": 0.5396,
+      "step": 2133
+    },
+    {
+      "epoch": 1.988816402609506,
+      "grad_norm": 0.9451275786423879,
+      "learning_rate": 1.8726268553676218e-05,
+      "loss": 0.523,
+      "step": 2134
+    },
+    {
+      "epoch": 1.9897483690587139,
+      "grad_norm": 0.9842237101940886,
+      "learning_rate": 1.8709009319986192e-05,
+      "loss": 0.551,
+      "step": 2135
+    },
+    {
+      "epoch": 1.9906803355079217,
+      "grad_norm": 1.184204336613584,
+      "learning_rate": 1.869175008629617e-05,
+      "loss": 0.6928,
+      "step": 2136
+    },
+    {
+      "epoch": 1.9916123019571295,
+      "grad_norm": 0.856419659081117,
+      "learning_rate": 1.8674490852606144e-05,
+      "loss": 0.567,
+      "step": 2137
+    },
+    {
+      "epoch": 1.9925442684063372,
+      "grad_norm": 1.1896144586519641,
+      "learning_rate": 1.8657231618916122e-05,
+      "loss": 0.5339,
+      "step": 2138
+    },
+    {
+      "epoch": 1.993476234855545,
+      "grad_norm": 1.3093505999756982,
+      "learning_rate": 1.86399723852261e-05,
+      "loss": 0.5884,
+      "step": 2139
+    },
+    {
+      "epoch": 1.994408201304753,
+      "grad_norm": 0.9704591592554302,
+      "learning_rate": 1.8622713151536074e-05,
+      "loss": 0.583,
+      "step": 2140
+    },
+    {
+      "epoch": 1.9953401677539608,
+      "grad_norm": 0.9010590120103972,
+      "learning_rate": 1.8605453917846048e-05,
+      "loss": 0.5124,
+      "step": 2141
+    },
+    {
+      "epoch": 1.9962721342031688,
+      "grad_norm": 0.9433443299412205,
+      "learning_rate": 1.8588194684156023e-05,
+      "loss": 0.5943,
+      "step": 2142
+    },
+    {
+      "epoch": 1.9972041006523766,
+      "grad_norm": 0.8958061105215889,
+      "learning_rate": 1.8570935450466e-05,
+      "loss": 0.5202,
+      "step": 2143
+    },
+    {
+      "epoch": 1.9981360671015844,
+      "grad_norm": 1.0645650951599746,
+      "learning_rate": 1.8553676216775975e-05,
+      "loss": 0.6102,
+      "step": 2144
+    },
+    {
+      "epoch": 1.9990680335507922,
+      "grad_norm": 1.1396271702288097,
+      "learning_rate": 1.8536416983085952e-05,
+      "loss": 0.5655,
+      "step": 2145
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9110444158687907,
+      "learning_rate": 1.8519157749395927e-05,
+      "loss": 0.5236,
+      "step": 2146
+    },
+    {
+      "epoch": 2.000931966449208,
+      "grad_norm": 0.566928705999753,
+      "learning_rate": 1.8501898515705904e-05,
+      "loss": 0.4785,
+      "step": 2147
+    },
+    {
+      "epoch": 2.0018639328984156,
+      "grad_norm": 0.6129387618329587,
+      "learning_rate": 1.848463928201588e-05,
+      "loss": 0.4593,
+      "step": 2148
+    },
+    {
+      "epoch": 2.0027958993476234,
+      "grad_norm": 0.6474300574508146,
+      "learning_rate": 1.8467380048325856e-05,
+      "loss": 0.4707,
+      "step": 2149
+    },
+    {
+      "epoch": 2.003727865796831,
+      "grad_norm": 0.7321502310689297,
+      "learning_rate": 1.8450120814635834e-05,
+      "loss": 0.4612,
+      "step": 2150
+    },
+    {
+      "epoch": 2.004659832246039,
+      "grad_norm": 0.6012496057620604,
+      "learning_rate": 1.8432861580945808e-05,
+      "loss": 0.4908,
+      "step": 2151
+    },
+    {
+      "epoch": 2.005591798695247,
+      "grad_norm": 0.6802449254150075,
+      "learning_rate": 1.8415602347255782e-05,
+      "loss": 0.4964,
+      "step": 2152
+    },
+    {
+      "epoch": 2.006523765144455,
+      "grad_norm": 0.5170621203668989,
+      "learning_rate": 1.8398343113565757e-05,
+      "loss": 0.4901,
+      "step": 2153
+    },
+    {
+      "epoch": 2.0074557315936628,
+      "grad_norm": 0.5457548163180715,
+      "learning_rate": 1.8381083879875734e-05,
+      "loss": 0.4953,
+      "step": 2154
+    },
+    {
+      "epoch": 2.0083876980428705,
+      "grad_norm": 0.7291009125990316,
+      "learning_rate": 1.836382464618571e-05,
+      "loss": 0.4738,
+      "step": 2155
+    },
+    {
+      "epoch": 2.0093196644920783,
+      "grad_norm": 0.655862564948919,
+      "learning_rate": 1.8346565412495686e-05,
+      "loss": 0.4877,
+      "step": 2156
+    },
+    {
+      "epoch": 2.010251630941286,
+      "grad_norm": 0.5282698996797063,
+      "learning_rate": 1.832930617880566e-05,
+      "loss": 0.4738,
+      "step": 2157
+    },
+    {
+      "epoch": 2.011183597390494,
+      "grad_norm": 0.8259304450769719,
+      "learning_rate": 1.831204694511564e-05,
+      "loss": 0.4511,
+      "step": 2158
+    },
+    {
+      "epoch": 2.0121155638397017,
+      "grad_norm": 0.5022028311886687,
+      "learning_rate": 1.8294787711425613e-05,
+      "loss": 0.4595,
+      "step": 2159
+    },
+    {
+      "epoch": 2.0130475302889095,
+      "grad_norm": 0.8851220382977533,
+      "learning_rate": 1.827752847773559e-05,
+      "loss": 0.5042,
+      "step": 2160
+    },
+    {
+      "epoch": 2.0139794967381173,
+      "grad_norm": 0.7735614358788059,
+      "learning_rate": 1.8260269244045568e-05,
+      "loss": 0.5184,
+      "step": 2161
+    },
+    {
+      "epoch": 2.014911463187325,
+      "grad_norm": 0.5179144605505243,
+      "learning_rate": 1.8243010010355542e-05,
+      "loss": 0.4523,
+      "step": 2162
+    },
+    {
+      "epoch": 2.0158434296365333,
+      "grad_norm": 0.7925617896482355,
+      "learning_rate": 1.8225750776665517e-05,
+      "loss": 0.4864,
+      "step": 2163
+    },
+    {
+      "epoch": 2.016775396085741,
+      "grad_norm": 0.5859131227708972,
+      "learning_rate": 1.820849154297549e-05,
+      "loss": 0.4525,
+      "step": 2164
+    },
+    {
+      "epoch": 2.017707362534949,
+      "grad_norm": 0.5792998920740682,
+      "learning_rate": 1.819123230928547e-05,
+      "loss": 0.4887,
+      "step": 2165
+    },
+    {
+      "epoch": 2.0186393289841567,
+      "grad_norm": 0.5252773792517837,
+      "learning_rate": 1.8173973075595443e-05,
+      "loss": 0.4538,
+      "step": 2166
+    },
+    {
+      "epoch": 2.0195712954333644,
+      "grad_norm": 0.5946128100882362,
+      "learning_rate": 1.815671384190542e-05,
+      "loss": 0.4972,
+      "step": 2167
+    },
+    {
+      "epoch": 2.0205032618825722,
+      "grad_norm": 0.8545136393866403,
+      "learning_rate": 1.8139454608215395e-05,
+      "loss": 0.5303,
+      "step": 2168
+    },
+    {
+      "epoch": 2.02143522833178,
+      "grad_norm": 0.6515188025987148,
+      "learning_rate": 1.8122195374525372e-05,
+      "loss": 0.4878,
+      "step": 2169
+    },
+    {
+      "epoch": 2.022367194780988,
+      "grad_norm": 0.6741958936727683,
+      "learning_rate": 1.810493614083535e-05,
+      "loss": 0.4719,
+      "step": 2170
+    },
+    {
+      "epoch": 2.0232991612301956,
+      "grad_norm": 0.5427049965059899,
+      "learning_rate": 1.8087676907145324e-05,
+      "loss": 0.4734,
+      "step": 2171
+    },
+    {
+      "epoch": 2.0242311276794034,
+      "grad_norm": 1.2335648322566286,
+      "learning_rate": 1.8070417673455302e-05,
+      "loss": 0.5267,
+      "step": 2172
+    },
+    {
+      "epoch": 2.025163094128611,
+      "grad_norm": 0.8858702948709529,
+      "learning_rate": 1.8053158439765276e-05,
+      "loss": 0.516,
+      "step": 2173
+    },
+    {
+      "epoch": 2.0260950605778194,
+      "grad_norm": 0.8832572437499593,
+      "learning_rate": 1.803589920607525e-05,
+      "loss": 0.483,
+      "step": 2174
+    },
+    {
+      "epoch": 2.027027027027027,
+      "grad_norm": 0.7766376606672916,
+      "learning_rate": 1.8018639972385225e-05,
+      "loss": 0.4427,
+      "step": 2175
+    },
+    {
+      "epoch": 2.027958993476235,
+      "grad_norm": 0.7585317657293112,
+      "learning_rate": 1.8001380738695203e-05,
+      "loss": 0.459,
+      "step": 2176
+    },
+    {
+      "epoch": 2.0288909599254428,
+      "grad_norm": 0.744802140481006,
+      "learning_rate": 1.7984121505005177e-05,
+      "loss": 0.501,
+      "step": 2177
+    },
+    {
+      "epoch": 2.0298229263746506,
+      "grad_norm": 0.5133019456237099,
+      "learning_rate": 1.7966862271315155e-05,
+      "loss": 0.4796,
+      "step": 2178
+    },
+    {
+      "epoch": 2.0307548928238583,
+      "grad_norm": 0.600621823000531,
+      "learning_rate": 1.794960303762513e-05,
+      "loss": 0.4896,
+      "step": 2179
+    },
+    {
+      "epoch": 2.031686859273066,
+      "grad_norm": 0.7353623124629926,
+      "learning_rate": 1.7932343803935107e-05,
+      "loss": 0.4993,
+      "step": 2180
+    },
+    {
+      "epoch": 2.032618825722274,
+      "grad_norm": 0.4894721984828612,
+      "learning_rate": 1.7915084570245084e-05,
+      "loss": 0.4771,
+      "step": 2181
+    },
+    {
+      "epoch": 2.0335507921714817,
+      "grad_norm": 0.6414967632596235,
+      "learning_rate": 1.789782533655506e-05,
+      "loss": 0.4957,
+      "step": 2182
+    },
+    {
+      "epoch": 2.0344827586206895,
+      "grad_norm": 0.6500597570935097,
+      "learning_rate": 1.7880566102865033e-05,
+      "loss": 0.4859,
+      "step": 2183
+    },
+    {
+      "epoch": 2.0354147250698973,
+      "grad_norm": 0.4241973659903767,
+      "learning_rate": 1.786330686917501e-05,
+      "loss": 0.4304,
+      "step": 2184
+    },
+    {
+      "epoch": 2.0363466915191055,
+      "grad_norm": 0.5865214619692485,
+      "learning_rate": 1.7846047635484985e-05,
+      "loss": 0.4379,
+      "step": 2185
+    },
+    {
+      "epoch": 2.0372786579683133,
+      "grad_norm": 0.4854623650441569,
+      "learning_rate": 1.782878840179496e-05,
+      "loss": 0.45,
+      "step": 2186
+    },
+    {
+      "epoch": 2.038210624417521,
+      "grad_norm": 0.8949817650481162,
+      "learning_rate": 1.7811529168104937e-05,
+      "loss": 0.513,
+      "step": 2187
+    },
+    {
+      "epoch": 2.039142590866729,
+      "grad_norm": 1.0544213653505983,
+      "learning_rate": 1.779426993441491e-05,
+      "loss": 0.4795,
+      "step": 2188
+    },
+    {
+      "epoch": 2.0400745573159367,
+      "grad_norm": 0.946471244430725,
+      "learning_rate": 1.777701070072489e-05,
+      "loss": 0.4687,
+      "step": 2189
+    },
+    {
+      "epoch": 2.0410065237651445,
+      "grad_norm": 0.8137692158123455,
+      "learning_rate": 1.7759751467034866e-05,
+      "loss": 0.4936,
+      "step": 2190
+    },
+    {
+      "epoch": 2.0419384902143523,
+      "grad_norm": 0.6192572428377031,
+      "learning_rate": 1.774249223334484e-05,
+      "loss": 0.5154,
+      "step": 2191
+    },
+    {
+      "epoch": 2.04287045666356,
+      "grad_norm": 0.6717623678812484,
+      "learning_rate": 1.772523299965482e-05,
+      "loss": 0.4875,
+      "step": 2192
+    },
+    {
+      "epoch": 2.043802423112768,
+      "grad_norm": 0.6548132204602367,
+      "learning_rate": 1.7707973765964793e-05,
+      "loss": 0.4583,
+      "step": 2193
+    },
+    {
+      "epoch": 2.0447343895619756,
+      "grad_norm": 0.745452045424192,
+      "learning_rate": 1.7690714532274767e-05,
+      "loss": 0.4936,
+      "step": 2194
+    },
+    {
+      "epoch": 2.0456663560111834,
+      "grad_norm": 0.5550838913455136,
+      "learning_rate": 1.767345529858474e-05,
+      "loss": 0.5082,
+      "step": 2195
+    },
+    {
+      "epoch": 2.0465983224603916,
+      "grad_norm": 0.6871792225793496,
+      "learning_rate": 1.765619606489472e-05,
+      "loss": 0.4988,
+      "step": 2196
+    },
+    {
+      "epoch": 2.0475302889095994,
+      "grad_norm": 0.6822505370775842,
+      "learning_rate": 1.7638936831204693e-05,
+      "loss": 0.4598,
+      "step": 2197
+    },
+    {
+      "epoch": 2.048462255358807,
+      "grad_norm": 0.5214010603024606,
+      "learning_rate": 1.762167759751467e-05,
+      "loss": 0.491,
+      "step": 2198
+    },
+    {
+      "epoch": 2.049394221808015,
+      "grad_norm": 0.8538780551987547,
+      "learning_rate": 1.7604418363824645e-05,
+      "loss": 0.4911,
+      "step": 2199
+    },
+    {
+      "epoch": 2.050326188257223,
+      "grad_norm": 0.6570411213426649,
+      "learning_rate": 1.7587159130134623e-05,
+      "loss": 0.4963,
+      "step": 2200
+    },
+    {
+      "epoch": 2.0512581547064306,
+      "grad_norm": 0.6448282516197589,
+      "learning_rate": 1.75698998964446e-05,
+      "loss": 0.4758,
+      "step": 2201
+    },
+    {
+      "epoch": 2.0521901211556384,
+      "grad_norm": 0.5905260464770311,
+      "learning_rate": 1.7552640662754575e-05,
+      "loss": 0.4835,
+      "step": 2202
+    },
+    {
+      "epoch": 2.053122087604846,
+      "grad_norm": 0.4445270080977586,
+      "learning_rate": 1.7535381429064553e-05,
+      "loss": 0.4546,
+      "step": 2203
+    },
+    {
+      "epoch": 2.054054054054054,
+      "grad_norm": 0.5278006161431348,
+      "learning_rate": 1.7518122195374527e-05,
+      "loss": 0.4409,
+      "step": 2204
+    },
+    {
+      "epoch": 2.0549860205032617,
+      "grad_norm": 0.5317658465500724,
+      "learning_rate": 1.75008629616845e-05,
+      "loss": 0.472,
+      "step": 2205
+    },
+    {
+      "epoch": 2.0559179869524695,
+      "grad_norm": 1.0268444794630966,
+      "learning_rate": 1.7483603727994475e-05,
+      "loss": 0.5266,
+      "step": 2206
+    },
+    {
+      "epoch": 2.0568499534016778,
+      "grad_norm": 0.6100587845488897,
+      "learning_rate": 1.7466344494304453e-05,
+      "loss": 0.4681,
+      "step": 2207
+    },
+    {
+      "epoch": 2.0577819198508855,
+      "grad_norm": 0.5059590650844958,
+      "learning_rate": 1.7449085260614427e-05,
+      "loss": 0.4461,
+      "step": 2208
+    },
+    {
+      "epoch": 2.0587138863000933,
+      "grad_norm": 0.6721508317954468,
+      "learning_rate": 1.7431826026924405e-05,
+      "loss": 0.486,
+      "step": 2209
+    },
+    {
+      "epoch": 2.059645852749301,
+      "grad_norm": 0.5336224712372963,
+      "learning_rate": 1.7414566793234383e-05,
+      "loss": 0.4535,
+      "step": 2210
+    },
+    {
+      "epoch": 2.060577819198509,
+      "grad_norm": 0.7359003834586001,
+      "learning_rate": 1.7397307559544357e-05,
+      "loss": 0.4765,
+      "step": 2211
+    },
+    {
+      "epoch": 2.0615097856477167,
+      "grad_norm": 0.42584829321650863,
+      "learning_rate": 1.7380048325854335e-05,
+      "loss": 0.4305,
+      "step": 2212
+    },
+    {
+      "epoch": 2.0624417520969245,
+      "grad_norm": 0.7328226891511267,
+      "learning_rate": 1.736278909216431e-05,
+      "loss": 0.492,
+      "step": 2213
+    },
+    {
+      "epoch": 2.0633737185461323,
+      "grad_norm": 0.5673881105895888,
+      "learning_rate": 1.7345529858474287e-05,
+      "loss": 0.4609,
+      "step": 2214
+    },
+    {
+      "epoch": 2.06430568499534,
+      "grad_norm": 0.5460259910462523,
+      "learning_rate": 1.732827062478426e-05,
+      "loss": 0.4814,
+      "step": 2215
+    },
+    {
+      "epoch": 2.065237651444548,
+      "grad_norm": 0.5142601013688192,
+      "learning_rate": 1.7311011391094235e-05,
+      "loss": 0.4262,
+      "step": 2216
+    },
+    {
+      "epoch": 2.0661696178937556,
+      "grad_norm": 0.6916833035279921,
+      "learning_rate": 1.729375215740421e-05,
+      "loss": 0.4959,
+      "step": 2217
+    },
+    {
+      "epoch": 2.0671015843429634,
+      "grad_norm": 0.6313947181556666,
+      "learning_rate": 1.7276492923714187e-05,
+      "loss": 0.4505,
+      "step": 2218
+    },
+    {
+      "epoch": 2.0680335507921717,
+      "grad_norm": 0.7673643761138421,
+      "learning_rate": 1.725923369002416e-05,
+      "loss": 0.4972,
+      "step": 2219
+    },
+    {
+      "epoch": 2.0689655172413794,
+      "grad_norm": 0.6284367221529983,
+      "learning_rate": 1.724197445633414e-05,
+      "loss": 0.4721,
+      "step": 2220
+    },
+    {
+      "epoch": 2.0698974836905872,
+      "grad_norm": 0.671783206633783,
+      "learning_rate": 1.7224715222644117e-05,
+      "loss": 0.4698,
+      "step": 2221
+    },
+    {
+      "epoch": 2.070829450139795,
+      "grad_norm": 0.6769021015928172,
+      "learning_rate": 1.720745598895409e-05,
+      "loss": 0.4765,
+      "step": 2222
+    },
+    {
+      "epoch": 2.071761416589003,
+      "grad_norm": 0.5829249736220897,
+      "learning_rate": 1.719019675526407e-05,
+      "loss": 0.4626,
+      "step": 2223
+    },
+    {
+      "epoch": 2.0726933830382106,
+      "grad_norm": 0.709323624684184,
+      "learning_rate": 1.7172937521574043e-05,
+      "loss": 0.4894,
+      "step": 2224
+    },
+    {
+      "epoch": 2.0736253494874184,
+      "grad_norm": 0.6782530742691311,
+      "learning_rate": 1.715567828788402e-05,
+      "loss": 0.4573,
+      "step": 2225
+    },
+    {
+      "epoch": 2.074557315936626,
+      "grad_norm": 0.44967840404274917,
+      "learning_rate": 1.7138419054193995e-05,
+      "loss": 0.4216,
+      "step": 2226
+    },
+    {
+      "epoch": 2.075489282385834,
+      "grad_norm": 0.6819023450102143,
+      "learning_rate": 1.712115982050397e-05,
+      "loss": 0.4829,
+      "step": 2227
+    },
+    {
+      "epoch": 2.0764212488350418,
+      "grad_norm": 0.7511608222652812,
+      "learning_rate": 1.7103900586813944e-05,
+      "loss": 0.4823,
+      "step": 2228
+    },
+    {
+      "epoch": 2.0773532152842495,
+      "grad_norm": 0.6827716835081838,
+      "learning_rate": 1.708664135312392e-05,
+      "loss": 0.4614,
+      "step": 2229
+    },
+    {
+      "epoch": 2.0782851817334578,
+      "grad_norm": 0.40431521926602293,
+      "learning_rate": 1.70693821194339e-05,
+      "loss": 0.4394,
+      "step": 2230
+    },
+    {
+      "epoch": 2.0792171481826656,
+      "grad_norm": 0.5460038801543602,
+      "learning_rate": 1.7052122885743873e-05,
+      "loss": 0.5006,
+      "step": 2231
+    },
+    {
+      "epoch": 2.0801491146318734,
+      "grad_norm": 0.6132841310810708,
+      "learning_rate": 1.703486365205385e-05,
+      "loss": 0.4485,
+      "step": 2232
+    },
+    {
+      "epoch": 2.081081081081081,
+      "grad_norm": 0.6924148053431719,
+      "learning_rate": 1.7017604418363825e-05,
+      "loss": 0.4774,
+      "step": 2233
+    },
+    {
+      "epoch": 2.082013047530289,
+      "grad_norm": 0.5788138053753754,
+      "learning_rate": 1.7000345184673803e-05,
+      "loss": 0.4555,
+      "step": 2234
+    },
+    {
+      "epoch": 2.0829450139794967,
+      "grad_norm": 0.6541422987828025,
+      "learning_rate": 1.6983085950983777e-05,
+      "loss": 0.4602,
+      "step": 2235
+    },
+    {
+      "epoch": 2.0838769804287045,
+      "grad_norm": 0.8146347388747902,
+      "learning_rate": 1.6965826717293755e-05,
+      "loss": 0.4983,
+      "step": 2236
+    },
+    {
+      "epoch": 2.0848089468779123,
+      "grad_norm": 0.6759951075551538,
+      "learning_rate": 1.694856748360373e-05,
+      "loss": 0.5052,
+      "step": 2237
+    },
+    {
+      "epoch": 2.08574091332712,
+      "grad_norm": 0.7278526097088425,
+      "learning_rate": 1.6931308249913704e-05,
+      "loss": 0.5228,
+      "step": 2238
+    },
+    {
+      "epoch": 2.086672879776328,
+      "grad_norm": 0.9022709680497087,
+      "learning_rate": 1.6914049016223678e-05,
+      "loss": 0.5028,
+      "step": 2239
+    },
+    {
+      "epoch": 2.0876048462255357,
+      "grad_norm": 0.6248180089778431,
+      "learning_rate": 1.6896789782533655e-05,
+      "loss": 0.4889,
+      "step": 2240
+    },
+    {
+      "epoch": 2.088536812674744,
+      "grad_norm": 0.5846484649380559,
+      "learning_rate": 1.6879530548843633e-05,
+      "loss": 0.4687,
+      "step": 2241
+    },
+    {
+      "epoch": 2.0894687791239517,
+      "grad_norm": 0.7658923446974368,
+      "learning_rate": 1.6862271315153607e-05,
+      "loss": 0.4736,
+      "step": 2242
+    },
+    {
+      "epoch": 2.0904007455731595,
+      "grad_norm": 0.8354809791252094,
+      "learning_rate": 1.6845012081463585e-05,
+      "loss": 0.5037,
+      "step": 2243
+    },
+    {
+      "epoch": 2.0913327120223673,
+      "grad_norm": 0.5575714117278624,
+      "learning_rate": 1.682775284777356e-05,
+      "loss": 0.47,
+      "step": 2244
+    },
+    {
+      "epoch": 2.092264678471575,
+      "grad_norm": 0.7181172734885467,
+      "learning_rate": 1.6810493614083537e-05,
+      "loss": 0.4763,
+      "step": 2245
+    },
+    {
+      "epoch": 2.093196644920783,
+      "grad_norm": 0.8376814383762288,
+      "learning_rate": 1.679323438039351e-05,
+      "loss": 0.4978,
+      "step": 2246
+    },
+    {
+      "epoch": 2.0941286113699906,
+      "grad_norm": 0.5821406967808621,
+      "learning_rate": 1.677597514670349e-05,
+      "loss": 0.4876,
+      "step": 2247
+    },
+    {
+      "epoch": 2.0950605778191984,
+      "grad_norm": 0.7977300372544451,
+      "learning_rate": 1.6758715913013463e-05,
+      "loss": 0.5049,
+      "step": 2248
+    },
+    {
+      "epoch": 2.095992544268406,
+      "grad_norm": 0.5760417124837485,
+      "learning_rate": 1.6741456679323438e-05,
+      "loss": 0.4601,
+      "step": 2249
+    },
+    {
+      "epoch": 2.096924510717614,
+      "grad_norm": 0.7574404147158804,
+      "learning_rate": 1.6724197445633415e-05,
+      "loss": 0.4834,
+      "step": 2250
+    },
+    {
+      "epoch": 2.0978564771668218,
+      "grad_norm": 0.7149985434483032,
+      "learning_rate": 1.670693821194339e-05,
+      "loss": 0.4957,
+      "step": 2251
+    },
+    {
+      "epoch": 2.09878844361603,
+      "grad_norm": 0.4950244991259179,
+      "learning_rate": 1.6689678978253367e-05,
+      "loss": 0.462,
+      "step": 2252
+    },
+    {
+      "epoch": 2.099720410065238,
+      "grad_norm": 0.8345321202956306,
+      "learning_rate": 1.667241974456334e-05,
+      "loss": 0.5205,
+      "step": 2253
+    },
+    {
+      "epoch": 2.1006523765144456,
+      "grad_norm": 0.7275502782665934,
+      "learning_rate": 1.665516051087332e-05,
+      "loss": 0.4676,
+      "step": 2254
+    },
+    {
+      "epoch": 2.1015843429636534,
+      "grad_norm": 0.45279550242213445,
+      "learning_rate": 1.6637901277183294e-05,
+      "loss": 0.4801,
+      "step": 2255
+    },
+    {
+      "epoch": 2.102516309412861,
+      "grad_norm": 0.7244144770381175,
+      "learning_rate": 1.662064204349327e-05,
+      "loss": 0.5137,
+      "step": 2256
+    },
+    {
+      "epoch": 2.103448275862069,
+      "grad_norm": 0.5989304605300362,
+      "learning_rate": 1.6603382809803246e-05,
+      "loss": 0.4658,
+      "step": 2257
+    },
+    {
+      "epoch": 2.1043802423112767,
+      "grad_norm": 0.5079896806552823,
+      "learning_rate": 1.6586123576113223e-05,
+      "loss": 0.4693,
+      "step": 2258
+    },
+    {
+      "epoch": 2.1053122087604845,
+      "grad_norm": 0.663794845751385,
+      "learning_rate": 1.6568864342423197e-05,
+      "loss": 0.4828,
+      "step": 2259
+    },
+    {
+      "epoch": 2.1062441752096923,
+      "grad_norm": 1.1142321995391056,
+      "learning_rate": 1.6551605108733172e-05,
+      "loss": 0.5214,
+      "step": 2260
+    },
+    {
+      "epoch": 2.1071761416589,
+      "grad_norm": 0.5480168625691394,
+      "learning_rate": 1.653434587504315e-05,
+      "loss": 0.4761,
+      "step": 2261
+    },
+    {
+      "epoch": 2.108108108108108,
+      "grad_norm": 0.7284194170208476,
+      "learning_rate": 1.6517086641353124e-05,
+      "loss": 0.4609,
+      "step": 2262
+    },
+    {
+      "epoch": 2.109040074557316,
+      "grad_norm": 0.6647932436476448,
+      "learning_rate": 1.64998274076631e-05,
+      "loss": 0.4856,
+      "step": 2263
+    },
+    {
+      "epoch": 2.109972041006524,
+      "grad_norm": 0.6611900227128719,
+      "learning_rate": 1.6482568173973076e-05,
+      "loss": 0.4706,
+      "step": 2264
+    },
+    {
+      "epoch": 2.1109040074557317,
+      "grad_norm": 0.7520872601758197,
+      "learning_rate": 1.6465308940283053e-05,
+      "loss": 0.4796,
+      "step": 2265
+    },
+    {
+      "epoch": 2.1118359739049395,
+      "grad_norm": 0.6567219442506047,
+      "learning_rate": 1.6448049706593028e-05,
+      "loss": 0.4681,
+      "step": 2266
+    },
+    {
+      "epoch": 2.1127679403541473,
+      "grad_norm": 0.5838219004734899,
+      "learning_rate": 1.6430790472903005e-05,
+      "loss": 0.4785,
+      "step": 2267
+    },
+    {
+      "epoch": 2.113699906803355,
+      "grad_norm": 0.8007480552918937,
+      "learning_rate": 1.641353123921298e-05,
+      "loss": 0.5012,
+      "step": 2268
+    },
+    {
+      "epoch": 2.114631873252563,
+      "grad_norm": 0.7200117358834833,
+      "learning_rate": 1.6396272005522957e-05,
+      "loss": 0.4943,
+      "step": 2269
+    },
+    {
+      "epoch": 2.1155638397017706,
+      "grad_norm": 0.7444915246052246,
+      "learning_rate": 1.637901277183293e-05,
+      "loss": 0.5265,
+      "step": 2270
+    },
+    {
+      "epoch": 2.1164958061509784,
+      "grad_norm": 0.6047494767195583,
+      "learning_rate": 1.6361753538142906e-05,
+      "loss": 0.504,
+      "step": 2271
+    },
+    {
+      "epoch": 2.117427772600186,
+      "grad_norm": 1.0189361576365759,
+      "learning_rate": 1.6344494304452884e-05,
+      "loss": 0.4713,
+      "step": 2272
+    },
+    {
+      "epoch": 2.118359739049394,
+      "grad_norm": 0.563856475920108,
+      "learning_rate": 1.6327235070762858e-05,
+      "loss": 0.4567,
+      "step": 2273
+    },
+    {
+      "epoch": 2.1192917054986022,
+      "grad_norm": 0.5997285151733354,
+      "learning_rate": 1.6309975837072836e-05,
+      "loss": 0.4816,
+      "step": 2274
+    },
+    {
+      "epoch": 2.12022367194781,
+      "grad_norm": 0.791422238183526,
+      "learning_rate": 1.629271660338281e-05,
+      "loss": 0.4959,
+      "step": 2275
+    },
+    {
+      "epoch": 2.121155638397018,
+      "grad_norm": 0.6761843603965912,
+      "learning_rate": 1.6275457369692788e-05,
+      "loss": 0.475,
+      "step": 2276
+    },
+    {
+      "epoch": 2.1220876048462256,
+      "grad_norm": 0.7683013668546091,
+      "learning_rate": 1.6258198136002762e-05,
+      "loss": 0.4483,
+      "step": 2277
+    },
+    {
+      "epoch": 2.1230195712954334,
+      "grad_norm": 0.7507769401429873,
+      "learning_rate": 1.624093890231274e-05,
+      "loss": 0.484,
+      "step": 2278
+    },
+    {
+      "epoch": 2.123951537744641,
+      "grad_norm": 0.5146850682587676,
+      "learning_rate": 1.6223679668622714e-05,
+      "loss": 0.4542,
+      "step": 2279
+    },
+    {
+      "epoch": 2.124883504193849,
+      "grad_norm": 0.6637668678429601,
+      "learning_rate": 1.620642043493269e-05,
+      "loss": 0.4887,
+      "step": 2280
+    },
+    {
+      "epoch": 2.1258154706430568,
+      "grad_norm": 0.6528890954249629,
+      "learning_rate": 1.6189161201242666e-05,
+      "loss": 0.4527,
+      "step": 2281
+    },
+    {
+      "epoch": 2.1267474370922645,
+      "grad_norm": 0.44283771071642414,
+      "learning_rate": 1.617190196755264e-05,
+      "loss": 0.4439,
+      "step": 2282
+    },
+    {
+      "epoch": 2.1276794035414723,
+      "grad_norm": 0.7147252146440377,
+      "learning_rate": 1.6154642733862618e-05,
+      "loss": 0.4901,
+      "step": 2283
+    },
+    {
+      "epoch": 2.12861136999068,
+      "grad_norm": 0.7435158575561752,
+      "learning_rate": 1.6137383500172592e-05,
+      "loss": 0.4682,
+      "step": 2284
+    },
+    {
+      "epoch": 2.1295433364398884,
+      "grad_norm": 0.5847749110464453,
+      "learning_rate": 1.612012426648257e-05,
+      "loss": 0.4399,
+      "step": 2285
+    },
+    {
+      "epoch": 2.130475302889096,
+      "grad_norm": 0.5708046934401089,
+      "learning_rate": 1.6102865032792544e-05,
+      "loss": 0.4848,
+      "step": 2286
+    },
+    {
+      "epoch": 2.131407269338304,
+      "grad_norm": 0.7501370489182962,
+      "learning_rate": 1.608560579910252e-05,
+      "loss": 0.5026,
+      "step": 2287
+    },
+    {
+      "epoch": 2.1323392357875117,
+      "grad_norm": 0.6768070970701205,
+      "learning_rate": 1.6068346565412496e-05,
+      "loss": 0.4658,
+      "step": 2288
+    },
+    {
+      "epoch": 2.1332712022367195,
+      "grad_norm": 0.8031098875140986,
+      "learning_rate": 1.6051087331722474e-05,
+      "loss": 0.4772,
+      "step": 2289
+    },
+    {
+      "epoch": 2.1342031686859273,
+      "grad_norm": 0.5573760250977855,
+      "learning_rate": 1.6033828098032448e-05,
+      "loss": 0.4681,
+      "step": 2290
+    },
+    {
+      "epoch": 2.135135135135135,
+      "grad_norm": 0.6031184220102512,
+      "learning_rate": 1.6016568864342422e-05,
+      "loss": 0.4557,
+      "step": 2291
+    },
+    {
+      "epoch": 2.136067101584343,
+      "grad_norm": 0.5888046700891744,
+      "learning_rate": 1.59993096306524e-05,
+      "loss": 0.5118,
+      "step": 2292
+    },
+    {
+      "epoch": 2.1369990680335507,
+      "grad_norm": 0.5861744531563393,
+      "learning_rate": 1.5982050396962374e-05,
+      "loss": 0.4299,
+      "step": 2293
+    },
+    {
+      "epoch": 2.1379310344827585,
+      "grad_norm": 0.6846205720909125,
+      "learning_rate": 1.5964791163272352e-05,
+      "loss": 0.5028,
+      "step": 2294
+    },
+    {
+      "epoch": 2.1388630009319662,
+      "grad_norm": 0.6617475007018842,
+      "learning_rate": 1.5947531929582326e-05,
+      "loss": 0.4831,
+      "step": 2295
+    },
+    {
+      "epoch": 2.1397949673811745,
+      "grad_norm": 0.6570836776711707,
+      "learning_rate": 1.5930272695892304e-05,
+      "loss": 0.499,
+      "step": 2296
+    },
+    {
+      "epoch": 2.1407269338303823,
+      "grad_norm": 0.6503430874663836,
+      "learning_rate": 1.5913013462202278e-05,
+      "loss": 0.488,
+      "step": 2297
+    },
+    {
+      "epoch": 2.14165890027959,
+      "grad_norm": 0.5961769805063115,
+      "learning_rate": 1.5895754228512256e-05,
+      "loss": 0.4883,
+      "step": 2298
+    },
+    {
+      "epoch": 2.142590866728798,
+      "grad_norm": 1.764940423818458,
+      "learning_rate": 1.587849499482223e-05,
+      "loss": 0.4639,
+      "step": 2299
+    },
+    {
+      "epoch": 2.1435228331780056,
+      "grad_norm": 1.020981564211258,
+      "learning_rate": 1.5861235761132208e-05,
+      "loss": 0.5469,
+      "step": 2300
+    },
+    {
+      "epoch": 2.1444547996272134,
+      "grad_norm": 1.053531399942582,
+      "learning_rate": 1.5843976527442182e-05,
+      "loss": 0.5164,
+      "step": 2301
+    },
+    {
+      "epoch": 2.145386766076421,
+      "grad_norm": 0.9104532424272882,
+      "learning_rate": 1.5826717293752156e-05,
+      "loss": 0.5211,
+      "step": 2302
+    },
+    {
+      "epoch": 2.146318732525629,
+      "grad_norm": 0.5955033253523766,
+      "learning_rate": 1.5809458060062134e-05,
+      "loss": 0.4658,
+      "step": 2303
+    },
+    {
+      "epoch": 2.147250698974837,
+      "grad_norm": 0.4836762903370474,
+      "learning_rate": 1.5792198826372108e-05,
+      "loss": 0.4356,
+      "step": 2304
+    },
+    {
+      "epoch": 2.1481826654240446,
+      "grad_norm": 0.6381174793569486,
+      "learning_rate": 1.5774939592682086e-05,
+      "loss": 0.4729,
+      "step": 2305
+    },
+    {
+      "epoch": 2.1491146318732524,
+      "grad_norm": 0.7388174230671309,
+      "learning_rate": 1.575768035899206e-05,
+      "loss": 0.4551,
+      "step": 2306
+    },
+    {
+      "epoch": 2.1500465983224606,
+      "grad_norm": 0.6123915024549202,
+      "learning_rate": 1.5740421125302038e-05,
+      "loss": 0.5149,
+      "step": 2307
+    },
+    {
+      "epoch": 2.1509785647716684,
+      "grad_norm": 0.7504411652382771,
+      "learning_rate": 1.5723161891612012e-05,
+      "loss": 0.4871,
+      "step": 2308
+    },
+    {
+      "epoch": 2.151910531220876,
+      "grad_norm": 0.5068535783226923,
+      "learning_rate": 1.570590265792199e-05,
+      "loss": 0.4728,
+      "step": 2309
+    },
+    {
+      "epoch": 2.152842497670084,
+      "grad_norm": 0.5932028118999028,
+      "learning_rate": 1.5688643424231964e-05,
+      "loss": 0.4388,
+      "step": 2310
+    },
+    {
+      "epoch": 2.1537744641192917,
+      "grad_norm": 0.7379574776954992,
+      "learning_rate": 1.5671384190541942e-05,
+      "loss": 0.5105,
+      "step": 2311
+    },
+    {
+      "epoch": 2.1547064305684995,
+      "grad_norm": 0.9536589915672642,
+      "learning_rate": 1.5654124956851916e-05,
+      "loss": 0.5282,
+      "step": 2312
+    },
+    {
+      "epoch": 2.1556383970177073,
+      "grad_norm": 0.6004429131798287,
+      "learning_rate": 1.563686572316189e-05,
+      "loss": 0.4841,
+      "step": 2313
+    },
+    {
+      "epoch": 2.156570363466915,
+      "grad_norm": 0.6887713823470023,
+      "learning_rate": 1.5619606489471868e-05,
+      "loss": 0.4546,
+      "step": 2314
+    },
+    {
+      "epoch": 2.157502329916123,
+      "grad_norm": 0.8599328883584993,
+      "learning_rate": 1.5602347255781842e-05,
+      "loss": 0.499,
+      "step": 2315
+    },
+    {
+      "epoch": 2.1584342963653307,
+      "grad_norm": 0.736763380572327,
+      "learning_rate": 1.558508802209182e-05,
+      "loss": 0.5155,
+      "step": 2316
+    },
+    {
+      "epoch": 2.1593662628145385,
+      "grad_norm": 0.8268015068724023,
+      "learning_rate": 1.5567828788401794e-05,
+      "loss": 0.4811,
+      "step": 2317
+    },
+    {
+      "epoch": 2.1602982292637467,
+      "grad_norm": 0.60978338665944,
+      "learning_rate": 1.5550569554711772e-05,
+      "loss": 0.4543,
+      "step": 2318
+    },
+    {
+      "epoch": 2.1612301957129545,
+      "grad_norm": 0.6150558327470734,
+      "learning_rate": 1.5533310321021746e-05,
+      "loss": 0.4467,
+      "step": 2319
+    },
+    {
+      "epoch": 2.1621621621621623,
+      "grad_norm": 0.6409648953354372,
+      "learning_rate": 1.5516051087331724e-05,
+      "loss": 0.4613,
+      "step": 2320
+    },
+    {
+      "epoch": 2.16309412861137,
+      "grad_norm": 0.684941962960559,
+      "learning_rate": 1.5498791853641702e-05,
+      "loss": 0.4745,
+      "step": 2321
+    },
+    {
+      "epoch": 2.164026095060578,
+      "grad_norm": 0.7069905078625647,
+      "learning_rate": 1.5481532619951676e-05,
+      "loss": 0.5056,
+      "step": 2322
+    },
+    {
+      "epoch": 2.1649580615097856,
+      "grad_norm": 0.5505696954236167,
+      "learning_rate": 1.546427338626165e-05,
+      "loss": 0.4448,
+      "step": 2323
+    },
+    {
+      "epoch": 2.1658900279589934,
+      "grad_norm": 0.4543732964425905,
+      "learning_rate": 1.5447014152571625e-05,
+      "loss": 0.4385,
+      "step": 2324
+    },
+    {
+      "epoch": 2.1668219944082012,
+      "grad_norm": 0.9734565241566739,
+      "learning_rate": 1.5429754918881602e-05,
+      "loss": 0.5139,
+      "step": 2325
+    },
+    {
+      "epoch": 2.167753960857409,
+      "grad_norm": 0.6790221606207381,
+      "learning_rate": 1.5412495685191577e-05,
+      "loss": 0.4818,
+      "step": 2326
+    },
+    {
+      "epoch": 2.168685927306617,
+      "grad_norm": 0.7794944978617138,
+      "learning_rate": 1.5395236451501554e-05,
+      "loss": 0.4617,
+      "step": 2327
+    },
+    {
+      "epoch": 2.1696178937558246,
+      "grad_norm": 0.7193267150568577,
+      "learning_rate": 1.537797721781153e-05,
+      "loss": 0.4866,
+      "step": 2328
+    },
+    {
+      "epoch": 2.170549860205033,
+      "grad_norm": 1.018540261467336,
+      "learning_rate": 1.5360717984121506e-05,
+      "loss": 0.4982,
+      "step": 2329
+    },
+    {
+      "epoch": 2.1714818266542406,
+      "grad_norm": 0.6328641083827854,
+      "learning_rate": 1.534345875043148e-05,
+      "loss": 0.4595,
+      "step": 2330
+    },
+    {
+      "epoch": 2.1724137931034484,
+      "grad_norm": 0.6988587628243841,
+      "learning_rate": 1.5326199516741458e-05,
+      "loss": 0.474,
+      "step": 2331
+    },
+    {
+      "epoch": 2.173345759552656,
+      "grad_norm": 0.5501375290753291,
+      "learning_rate": 1.5308940283051436e-05,
+      "loss": 0.4626,
+      "step": 2332
+    },
+    {
+      "epoch": 2.174277726001864,
+      "grad_norm": 0.7052015127489573,
+      "learning_rate": 1.529168104936141e-05,
+      "loss": 0.4761,
+      "step": 2333
+    },
+    {
+      "epoch": 2.1752096924510718,
+      "grad_norm": 0.5537450450113306,
+      "learning_rate": 1.5274421815671384e-05,
+      "loss": 0.475,
+      "step": 2334
+    },
+    {
+      "epoch": 2.1761416589002796,
+      "grad_norm": 0.6420549347361532,
+      "learning_rate": 1.525716258198136e-05,
+      "loss": 0.4543,
+      "step": 2335
+    },
+    {
+      "epoch": 2.1770736253494873,
+      "grad_norm": 1.0456298871915028,
+      "learning_rate": 1.5239903348291338e-05,
+      "loss": 0.4788,
+      "step": 2336
+    },
+    {
+      "epoch": 2.178005591798695,
+      "grad_norm": 0.7050186915898987,
+      "learning_rate": 1.5222644114601312e-05,
+      "loss": 0.4972,
+      "step": 2337
+    },
+    {
+      "epoch": 2.178937558247903,
+      "grad_norm": 0.6073394108991068,
+      "learning_rate": 1.5205384880911288e-05,
+      "loss": 0.4571,
+      "step": 2338
+    },
+    {
+      "epoch": 2.1798695246971107,
+      "grad_norm": 0.4716393331202839,
+      "learning_rate": 1.5188125647221263e-05,
+      "loss": 0.4738,
+      "step": 2339
+    },
+    {
+      "epoch": 2.180801491146319,
+      "grad_norm": 0.7512213184056765,
+      "learning_rate": 1.517086641353124e-05,
+      "loss": 0.4581,
+      "step": 2340
+    },
+    {
+      "epoch": 2.1817334575955267,
+      "grad_norm": 0.6642512141172897,
+      "learning_rate": 1.5153607179841216e-05,
+      "loss": 0.4835,
+      "step": 2341
+    },
+    {
+      "epoch": 2.1826654240447345,
+      "grad_norm": 0.6336698111692686,
+      "learning_rate": 1.513634794615119e-05,
+      "loss": 0.4946,
+      "step": 2342
+    },
+    {
+      "epoch": 2.1835973904939423,
+      "grad_norm": 0.6026433290688719,
+      "learning_rate": 1.5119088712461168e-05,
+      "loss": 0.4765,
+      "step": 2343
+    },
+    {
+      "epoch": 2.18452935694315,
+      "grad_norm": 0.6482339199026333,
+      "learning_rate": 1.5101829478771143e-05,
+      "loss": 0.4764,
+      "step": 2344
+    },
+    {
+      "epoch": 2.185461323392358,
+      "grad_norm": 0.4427627446504915,
+      "learning_rate": 1.508457024508112e-05,
+      "loss": 0.468,
+      "step": 2345
+    },
+    {
+      "epoch": 2.1863932898415657,
+      "grad_norm": 0.7186788121365741,
+      "learning_rate": 1.5067311011391095e-05,
+      "loss": 0.4803,
+      "step": 2346
+    },
+    {
+      "epoch": 2.1873252562907735,
+      "grad_norm": 0.7817911841155176,
+      "learning_rate": 1.505005177770107e-05,
+      "loss": 0.4547,
+      "step": 2347
+    },
+    {
+      "epoch": 2.1882572227399812,
+      "grad_norm": 0.7240135107253599,
+      "learning_rate": 1.5032792544011045e-05,
+      "loss": 0.4491,
+      "step": 2348
+    },
+    {
+      "epoch": 2.189189189189189,
+      "grad_norm": 0.5315025605056282,
+      "learning_rate": 1.5015533310321023e-05,
+      "loss": 0.4151,
+      "step": 2349
+    },
+    {
+      "epoch": 2.190121155638397,
+      "grad_norm": 0.6441072159235413,
+      "learning_rate": 1.4998274076630997e-05,
+      "loss": 0.476,
+      "step": 2350
+    },
+    {
+      "epoch": 2.191053122087605,
+      "grad_norm": 0.659544020181842,
+      "learning_rate": 1.4981014842940974e-05,
+      "loss": 0.4691,
+      "step": 2351
+    },
+    {
+      "epoch": 2.191985088536813,
+      "grad_norm": 0.6192534326267729,
+      "learning_rate": 1.496375560925095e-05,
+      "loss": 0.4905,
+      "step": 2352
+    },
+    {
+      "epoch": 2.1929170549860206,
+      "grad_norm": 0.7102424825496126,
+      "learning_rate": 1.4946496375560925e-05,
+      "loss": 0.4866,
+      "step": 2353
+    },
+    {
+      "epoch": 2.1938490214352284,
+      "grad_norm": 0.5552095762753042,
+      "learning_rate": 1.4929237141870902e-05,
+      "loss": 0.4876,
+      "step": 2354
+    },
+    {
+      "epoch": 2.194780987884436,
+      "grad_norm": 0.7465616256192912,
+      "learning_rate": 1.4911977908180877e-05,
+      "loss": 0.5051,
+      "step": 2355
+    },
+    {
+      "epoch": 2.195712954333644,
+      "grad_norm": 0.9104632884601167,
+      "learning_rate": 1.4894718674490854e-05,
+      "loss": 0.5311,
+      "step": 2356
+    },
+    {
+      "epoch": 2.196644920782852,
+      "grad_norm": 0.5963125216446196,
+      "learning_rate": 1.4877459440800829e-05,
+      "loss": 0.4708,
+      "step": 2357
+    },
+    {
+      "epoch": 2.1975768872320596,
+      "grad_norm": 0.7622086518455418,
+      "learning_rate": 1.4860200207110805e-05,
+      "loss": 0.4807,
+      "step": 2358
+    },
+    {
+      "epoch": 2.1985088536812674,
+      "grad_norm": 0.8093385850311505,
+      "learning_rate": 1.4842940973420779e-05,
+      "loss": 0.4672,
+      "step": 2359
+    },
+    {
+      "epoch": 2.199440820130475,
+      "grad_norm": 0.708426835438159,
+      "learning_rate": 1.4825681739730757e-05,
+      "loss": 0.4923,
+      "step": 2360
+    },
+    {
+      "epoch": 2.200372786579683,
+      "grad_norm": 0.7035484912647353,
+      "learning_rate": 1.4808422506040731e-05,
+      "loss": 0.4543,
+      "step": 2361
+    },
+    {
+      "epoch": 2.201304753028891,
+      "grad_norm": 0.8404641620863175,
+      "learning_rate": 1.4791163272350709e-05,
+      "loss": 0.4955,
+      "step": 2362
+    },
+    {
+      "epoch": 2.202236719478099,
+      "grad_norm": 0.7329327900895163,
+      "learning_rate": 1.4773904038660685e-05,
+      "loss": 0.4517,
+      "step": 2363
+    },
+    {
+      "epoch": 2.2031686859273067,
+      "grad_norm": 0.6388644220818431,
+      "learning_rate": 1.4756644804970659e-05,
+      "loss": 0.4603,
+      "step": 2364
+    },
+    {
+      "epoch": 2.2041006523765145,
+      "grad_norm": 0.7834673377559241,
+      "learning_rate": 1.4739385571280637e-05,
+      "loss": 0.5078,
+      "step": 2365
+    },
+    {
+      "epoch": 2.2050326188257223,
+      "grad_norm": 0.6444983245070262,
+      "learning_rate": 1.472212633759061e-05,
+      "loss": 0.4672,
+      "step": 2366
+    },
+    {
+      "epoch": 2.20596458527493,
+      "grad_norm": 0.5930635912883404,
+      "learning_rate": 1.4704867103900589e-05,
+      "loss": 0.46,
+      "step": 2367
+    },
+    {
+      "epoch": 2.206896551724138,
+      "grad_norm": 0.6169291922390854,
+      "learning_rate": 1.4687607870210563e-05,
+      "loss": 0.4502,
+      "step": 2368
+    },
+    {
+      "epoch": 2.2078285181733457,
+      "grad_norm": 0.6905588356949659,
+      "learning_rate": 1.4670348636520539e-05,
+      "loss": 0.4598,
+      "step": 2369
+    },
+    {
+      "epoch": 2.2087604846225535,
+      "grad_norm": 0.5497061863795298,
+      "learning_rate": 1.4653089402830513e-05,
+      "loss": 0.4594,
+      "step": 2370
+    },
+    {
+      "epoch": 2.2096924510717613,
+      "grad_norm": 0.6242867607352662,
+      "learning_rate": 1.463583016914049e-05,
+      "loss": 0.4597,
+      "step": 2371
+    },
+    {
+      "epoch": 2.210624417520969,
+      "grad_norm": 0.6175564289700358,
+      "learning_rate": 1.4618570935450468e-05,
+      "loss": 0.4646,
+      "step": 2372
+    },
+    {
+      "epoch": 2.2115563839701773,
+      "grad_norm": 0.6457000525787986,
+      "learning_rate": 1.4601311701760443e-05,
+      "loss": 0.4474,
+      "step": 2373
+    },
+    {
+      "epoch": 2.212488350419385,
+      "grad_norm": 0.6819581316276697,
+      "learning_rate": 1.4584052468070419e-05,
+      "loss": 0.4712,
+      "step": 2374
+    },
+    {
+      "epoch": 2.213420316868593,
+      "grad_norm": 1.4057697729751928,
+      "learning_rate": 1.4566793234380393e-05,
+      "loss": 0.5241,
+      "step": 2375
+    },
+    {
+      "epoch": 2.2143522833178007,
+      "grad_norm": 0.6188835987247284,
+      "learning_rate": 1.454953400069037e-05,
+      "loss": 0.4871,
+      "step": 2376
+    },
+    {
+      "epoch": 2.2152842497670084,
+      "grad_norm": 0.7318695236674685,
+      "learning_rate": 1.4532274767000345e-05,
+      "loss": 0.4798,
+      "step": 2377
+    },
+    {
+      "epoch": 2.2162162162162162,
+      "grad_norm": 0.8368647751829434,
+      "learning_rate": 1.4515015533310323e-05,
+      "loss": 0.4568,
+      "step": 2378
+    },
+    {
+      "epoch": 2.217148182665424,
+      "grad_norm": 0.6177409251535368,
+      "learning_rate": 1.4497756299620297e-05,
+      "loss": 0.4953,
+      "step": 2379
+    },
+    {
+      "epoch": 2.218080149114632,
+      "grad_norm": 0.5476970007892535,
+      "learning_rate": 1.4480497065930273e-05,
+      "loss": 0.4558,
+      "step": 2380
+    },
+    {
+      "epoch": 2.2190121155638396,
+      "grad_norm": 0.6218256449350521,
+      "learning_rate": 1.4463237832240247e-05,
+      "loss": 0.5006,
+      "step": 2381
+    },
+    {
+      "epoch": 2.2199440820130474,
+      "grad_norm": 0.6918495773763064,
+      "learning_rate": 1.4445978598550225e-05,
+      "loss": 0.4758,
+      "step": 2382
+    },
+    {
+      "epoch": 2.220876048462255,
+      "grad_norm": 0.9173706546457704,
+      "learning_rate": 1.4428719364860203e-05,
+      "loss": 0.4634,
+      "step": 2383
+    },
+    {
+      "epoch": 2.2218080149114634,
+      "grad_norm": 0.5296428234849653,
+      "learning_rate": 1.4411460131170177e-05,
+      "loss": 0.4582,
+      "step": 2384
+    },
+    {
+      "epoch": 2.222739981360671,
+      "grad_norm": 0.522698721273033,
+      "learning_rate": 1.4394200897480153e-05,
+      "loss": 0.4334,
+      "step": 2385
+    },
+    {
+      "epoch": 2.223671947809879,
+      "grad_norm": 0.5114336228495995,
+      "learning_rate": 1.4376941663790127e-05,
+      "loss": 0.4497,
+      "step": 2386
+    },
+    {
+      "epoch": 2.2246039142590868,
+      "grad_norm": 0.5947598616326761,
+      "learning_rate": 1.4359682430100105e-05,
+      "loss": 0.4827,
+      "step": 2387
+    },
+    {
+      "epoch": 2.2255358807082946,
+      "grad_norm": 0.6959517976967228,
+      "learning_rate": 1.4342423196410079e-05,
+      "loss": 0.4719,
+      "step": 2388
+    },
+    {
+      "epoch": 2.2264678471575023,
+      "grad_norm": 0.5835154909041277,
+      "learning_rate": 1.4325163962720057e-05,
+      "loss": 0.4714,
+      "step": 2389
+    },
+    {
+      "epoch": 2.22739981360671,
+      "grad_norm": 0.5526059960835007,
+      "learning_rate": 1.4307904729030031e-05,
+      "loss": 0.4473,
+      "step": 2390
+    },
+    {
+      "epoch": 2.228331780055918,
+      "grad_norm": 0.6370155914211028,
+      "learning_rate": 1.4290645495340007e-05,
+      "loss": 0.4946,
+      "step": 2391
+    },
+    {
+      "epoch": 2.2292637465051257,
+      "grad_norm": 0.8830636187136927,
+      "learning_rate": 1.4273386261649985e-05,
+      "loss": 0.513,
+      "step": 2392
+    },
+    {
+      "epoch": 2.2301957129543335,
+      "grad_norm": 0.5965218961687848,
+      "learning_rate": 1.4256127027959959e-05,
+      "loss": 0.4903,
+      "step": 2393
+    },
+    {
+      "epoch": 2.2311276794035413,
+      "grad_norm": 0.5520725958705621,
+      "learning_rate": 1.4238867794269937e-05,
+      "loss": 0.4618,
+      "step": 2394
+    },
+    {
+      "epoch": 2.2320596458527495,
+      "grad_norm": 0.6645760478238856,
+      "learning_rate": 1.4221608560579911e-05,
+      "loss": 0.4555,
+      "step": 2395
+    },
+    {
+      "epoch": 2.2329916123019573,
+      "grad_norm": 0.46920973072054645,
+      "learning_rate": 1.4204349326889887e-05,
+      "loss": 0.4472,
+      "step": 2396
+    },
+    {
+      "epoch": 2.233923578751165,
+      "grad_norm": 0.5024083063771777,
+      "learning_rate": 1.4187090093199861e-05,
+      "loss": 0.4502,
+      "step": 2397
+    },
+    {
+      "epoch": 2.234855545200373,
+      "grad_norm": 0.6146461472209771,
+      "learning_rate": 1.4169830859509839e-05,
+      "loss": 0.4873,
+      "step": 2398
+    },
+    {
+      "epoch": 2.2357875116495807,
+      "grad_norm": 0.7749108417968369,
+      "learning_rate": 1.4152571625819813e-05,
+      "loss": 0.4747,
+      "step": 2399
+    },
+    {
+      "epoch": 2.2367194780987885,
+      "grad_norm": 0.7738176873713739,
+      "learning_rate": 1.4135312392129791e-05,
+      "loss": 0.4737,
+      "step": 2400
+    },
+    {
+      "epoch": 2.2376514445479962,
+      "grad_norm": 0.5518893276639418,
+      "learning_rate": 1.4118053158439765e-05,
+      "loss": 0.4589,
+      "step": 2401
+    },
+    {
+      "epoch": 2.238583410997204,
+      "grad_norm": 0.4504056865931002,
+      "learning_rate": 1.4100793924749741e-05,
+      "loss": 0.4161,
+      "step": 2402
+    },
+    {
+      "epoch": 2.239515377446412,
+      "grad_norm": 0.633770677210652,
+      "learning_rate": 1.4083534691059719e-05,
+      "loss": 0.4385,
+      "step": 2403
+    },
+    {
+      "epoch": 2.2404473438956196,
+      "grad_norm": 0.9080889945805233,
+      "learning_rate": 1.4066275457369693e-05,
+      "loss": 0.4725,
+      "step": 2404
+    },
+    {
+      "epoch": 2.2413793103448274,
+      "grad_norm": 0.783898022151696,
+      "learning_rate": 1.404901622367967e-05,
+      "loss": 0.4717,
+      "step": 2405
+    },
+    {
+      "epoch": 2.2423112767940356,
+      "grad_norm": 0.7275603265278269,
+      "learning_rate": 1.4031756989989645e-05,
+      "loss": 0.5015,
+      "step": 2406
+    },
+    {
+      "epoch": 2.2432432432432434,
+      "grad_norm": 0.6420207893567118,
+      "learning_rate": 1.4014497756299621e-05,
+      "loss": 0.4565,
+      "step": 2407
+    },
+    {
+      "epoch": 2.244175209692451,
+      "grad_norm": 0.8487255345630685,
+      "learning_rate": 1.3997238522609595e-05,
+      "loss": 0.5218,
+      "step": 2408
+    },
+    {
+      "epoch": 2.245107176141659,
+      "grad_norm": 0.5559935713015115,
+      "learning_rate": 1.3979979288919573e-05,
+      "loss": 0.4446,
+      "step": 2409
+    },
+    {
+      "epoch": 2.246039142590867,
+      "grad_norm": 0.5577908463881104,
+      "learning_rate": 1.3962720055229547e-05,
+      "loss": 0.4764,
+      "step": 2410
+    },
+    {
+      "epoch": 2.2469711090400746,
+      "grad_norm": 0.678382347796411,
+      "learning_rate": 1.3945460821539525e-05,
+      "loss": 0.4845,
+      "step": 2411
+    },
+    {
+      "epoch": 2.2479030754892824,
+      "grad_norm": 0.4994090480625536,
+      "learning_rate": 1.3928201587849501e-05,
+      "loss": 0.435,
+      "step": 2412
+    },
+    {
+      "epoch": 2.24883504193849,
+      "grad_norm": 0.49198300904970277,
+      "learning_rate": 1.3910942354159475e-05,
+      "loss": 0.4761,
+      "step": 2413
+    },
+    {
+      "epoch": 2.249767008387698,
+      "grad_norm": 0.6212578643296621,
+      "learning_rate": 1.3893683120469453e-05,
+      "loss": 0.4751,
+      "step": 2414
+    },
+    {
+      "epoch": 2.2506989748369057,
+      "grad_norm": 0.5641801351892461,
+      "learning_rate": 1.3876423886779427e-05,
+      "loss": 0.4727,
+      "step": 2415
+    },
+    {
+      "epoch": 2.2516309412861135,
+      "grad_norm": 0.46397999203316803,
+      "learning_rate": 1.3859164653089405e-05,
+      "loss": 0.4574,
+      "step": 2416
+    },
+    {
+      "epoch": 2.2525629077353218,
+      "grad_norm": 0.5245591539733104,
+      "learning_rate": 1.384190541939938e-05,
+      "loss": 0.47,
+      "step": 2417
+    },
+    {
+      "epoch": 2.2534948741845295,
+      "grad_norm": 0.6561859719957489,
+      "learning_rate": 1.3824646185709355e-05,
+      "loss": 0.4966,
+      "step": 2418
+    },
+    {
+      "epoch": 2.2544268406337373,
+      "grad_norm": 0.6188798838627673,
+      "learning_rate": 1.380738695201933e-05,
+      "loss": 0.4589,
+      "step": 2419
+    },
+    {
+      "epoch": 2.255358807082945,
+      "grad_norm": 0.5934255647964382,
+      "learning_rate": 1.3790127718329307e-05,
+      "loss": 0.4508,
+      "step": 2420
+    },
+    {
+      "epoch": 2.256290773532153,
+      "grad_norm": 0.5548464045045394,
+      "learning_rate": 1.3772868484639282e-05,
+      "loss": 0.4613,
+      "step": 2421
+    },
+    {
+      "epoch": 2.2572227399813607,
+      "grad_norm": 0.6105800847200287,
+      "learning_rate": 1.375560925094926e-05,
+      "loss": 0.468,
+      "step": 2422
+    },
+    {
+      "epoch": 2.2581547064305685,
+      "grad_norm": 0.46211661314704816,
+      "learning_rate": 1.3738350017259235e-05,
+      "loss": 0.4446,
+      "step": 2423
+    },
+    {
+      "epoch": 2.2590866728797763,
+      "grad_norm": 0.6590839834264538,
+      "learning_rate": 1.372109078356921e-05,
+      "loss": 0.48,
+      "step": 2424
+    },
+    {
+      "epoch": 2.260018639328984,
+      "grad_norm": 0.4950317599144523,
+      "learning_rate": 1.3703831549879187e-05,
+      "loss": 0.4645,
+      "step": 2425
+    },
+    {
+      "epoch": 2.260950605778192,
+      "grad_norm": 0.7509654755138561,
+      "learning_rate": 1.3686572316189161e-05,
+      "loss": 0.4819,
+      "step": 2426
+    },
+    {
+      "epoch": 2.2618825722273996,
+      "grad_norm": 0.6229722077964742,
+      "learning_rate": 1.3669313082499139e-05,
+      "loss": 0.4558,
+      "step": 2427
+    },
+    {
+      "epoch": 2.262814538676608,
+      "grad_norm": 0.5876475745456535,
+      "learning_rate": 1.3652053848809113e-05,
+      "loss": 0.4649,
+      "step": 2428
+    },
+    {
+      "epoch": 2.2637465051258157,
+      "grad_norm": 0.7638779489291221,
+      "learning_rate": 1.363479461511909e-05,
+      "loss": 0.4756,
+      "step": 2429
+    },
+    {
+      "epoch": 2.2646784715750234,
+      "grad_norm": 0.5745305424377443,
+      "learning_rate": 1.3617535381429064e-05,
+      "loss": 0.4689,
+      "step": 2430
+    },
+    {
+      "epoch": 2.2656104380242312,
+      "grad_norm": 0.7527075554170519,
+      "learning_rate": 1.3600276147739041e-05,
+      "loss": 0.449,
+      "step": 2431
+    },
+    {
+      "epoch": 2.266542404473439,
+      "grad_norm": 0.8562787449051629,
+      "learning_rate": 1.3583016914049019e-05,
+      "loss": 0.4963,
+      "step": 2432
+    },
+    {
+      "epoch": 2.267474370922647,
+      "grad_norm": 0.7788940620636965,
+      "learning_rate": 1.3565757680358993e-05,
+      "loss": 0.5055,
+      "step": 2433
+    },
+    {
+      "epoch": 2.2684063373718546,
+      "grad_norm": 0.774872712719315,
+      "learning_rate": 1.354849844666897e-05,
+      "loss": 0.4845,
+      "step": 2434
+    },
+    {
+      "epoch": 2.2693383038210624,
+      "grad_norm": 0.6128490898798151,
+      "learning_rate": 1.3531239212978944e-05,
+      "loss": 0.4918,
+      "step": 2435
+    },
+    {
+      "epoch": 2.27027027027027,
+      "grad_norm": 0.8011235054555816,
+      "learning_rate": 1.3513979979288921e-05,
+      "loss": 0.4732,
+      "step": 2436
+    },
+    {
+      "epoch": 2.271202236719478,
+      "grad_norm": 0.627293986597918,
+      "learning_rate": 1.3496720745598896e-05,
+      "loss": 0.4765,
+      "step": 2437
+    },
+    {
+      "epoch": 2.2721342031686858,
+      "grad_norm": 0.489210343407143,
+      "learning_rate": 1.3479461511908873e-05,
+      "loss": 0.4755,
+      "step": 2438
+    },
+    {
+      "epoch": 2.273066169617894,
+      "grad_norm": 0.7720362961483633,
+      "learning_rate": 1.3462202278218848e-05,
+      "loss": 0.4528,
+      "step": 2439
+    },
+    {
+      "epoch": 2.2739981360671018,
+      "grad_norm": 0.45519727515350195,
+      "learning_rate": 1.3444943044528824e-05,
+      "loss": 0.4655,
+      "step": 2440
+    },
+    {
+      "epoch": 2.2749301025163096,
+      "grad_norm": 0.5378439228582352,
+      "learning_rate": 1.3427683810838798e-05,
+      "loss": 0.4733,
+      "step": 2441
+    },
+    {
+      "epoch": 2.2758620689655173,
+      "grad_norm": 0.6075671637672431,
+      "learning_rate": 1.3410424577148775e-05,
+      "loss": 0.4848,
+      "step": 2442
+    },
+    {
+      "epoch": 2.276794035414725,
+      "grad_norm": 0.8991921915071338,
+      "learning_rate": 1.3393165343458751e-05,
+      "loss": 0.489,
+      "step": 2443
+    },
+    {
+      "epoch": 2.277726001863933,
+      "grad_norm": 0.47670824307926263,
+      "learning_rate": 1.3375906109768726e-05,
+      "loss": 0.4726,
+      "step": 2444
+    },
+    {
+      "epoch": 2.2786579683131407,
+      "grad_norm": 0.6436615886278689,
+      "learning_rate": 1.3358646876078703e-05,
+      "loss": 0.4518,
+      "step": 2445
+    },
+    {
+      "epoch": 2.2795899347623485,
+      "grad_norm": 0.7649615772118803,
+      "learning_rate": 1.3341387642388678e-05,
+      "loss": 0.4818,
+      "step": 2446
+    },
+    {
+      "epoch": 2.2805219012115563,
+      "grad_norm": 0.6677650414460492,
+      "learning_rate": 1.3324128408698655e-05,
+      "loss": 0.4815,
+      "step": 2447
+    },
+    {
+      "epoch": 2.281453867660764,
+      "grad_norm": 0.6654266971817278,
+      "learning_rate": 1.330686917500863e-05,
+      "loss": 0.4623,
+      "step": 2448
+    },
+    {
+      "epoch": 2.282385834109972,
+      "grad_norm": 0.7650262460803089,
+      "learning_rate": 1.3289609941318606e-05,
+      "loss": 0.5026,
+      "step": 2449
+    },
+    {
+      "epoch": 2.28331780055918,
+      "grad_norm": 0.5030390810638123,
+      "learning_rate": 1.327235070762858e-05,
+      "loss": 0.4657,
+      "step": 2450
+    },
+    {
+      "epoch": 2.284249767008388,
+      "grad_norm": 0.6472476529528671,
+      "learning_rate": 1.3255091473938558e-05,
+      "loss": 0.4912,
+      "step": 2451
+    },
+    {
+      "epoch": 2.2851817334575957,
+      "grad_norm": 0.6674388570564048,
+      "learning_rate": 1.3237832240248535e-05,
+      "loss": 0.4915,
+      "step": 2452
+    },
+    {
+      "epoch": 2.2861136999068035,
+      "grad_norm": 0.6185809472420627,
+      "learning_rate": 1.322057300655851e-05,
+      "loss": 0.4545,
+      "step": 2453
+    },
+    {
+      "epoch": 2.2870456663560113,
+      "grad_norm": 0.5580534352841428,
+      "learning_rate": 1.3203313772868486e-05,
+      "loss": 0.4614,
+      "step": 2454
+    },
+    {
+      "epoch": 2.287977632805219,
+      "grad_norm": 0.6835545284977311,
+      "learning_rate": 1.318605453917846e-05,
+      "loss": 0.4852,
+      "step": 2455
+    },
+    {
+      "epoch": 2.288909599254427,
+      "grad_norm": 0.6791645306550724,
+      "learning_rate": 1.3168795305488438e-05,
+      "loss": 0.4808,
+      "step": 2456
+    },
+    {
+      "epoch": 2.2898415657036346,
+      "grad_norm": 0.5448811128258174,
+      "learning_rate": 1.3151536071798412e-05,
+      "loss": 0.4554,
+      "step": 2457
+    },
+    {
+      "epoch": 2.2907735321528424,
+      "grad_norm": 0.46296341442676975,
+      "learning_rate": 1.313427683810839e-05,
+      "loss": 0.4526,
+      "step": 2458
+    },
+    {
+      "epoch": 2.29170549860205,
+      "grad_norm": 0.6777756631701423,
+      "learning_rate": 1.3117017604418364e-05,
+      "loss": 0.4809,
+      "step": 2459
+    },
+    {
+      "epoch": 2.292637465051258,
+      "grad_norm": 0.5002413180342221,
+      "learning_rate": 1.309975837072834e-05,
+      "loss": 0.4702,
+      "step": 2460
+    },
+    {
+      "epoch": 2.293569431500466,
+      "grad_norm": 0.6860480557390517,
+      "learning_rate": 1.3082499137038314e-05,
+      "loss": 0.4962,
+      "step": 2461
+    },
+    {
+      "epoch": 2.294501397949674,
+      "grad_norm": 0.4766972086027654,
+      "learning_rate": 1.3065239903348292e-05,
+      "loss": 0.4476,
+      "step": 2462
+    },
+    {
+      "epoch": 2.295433364398882,
+      "grad_norm": 0.7804236198244641,
+      "learning_rate": 1.304798066965827e-05,
+      "loss": 0.4852,
+      "step": 2463
+    },
+    {
+      "epoch": 2.2963653308480896,
+      "grad_norm": 0.5715404616859929,
+      "learning_rate": 1.3030721435968244e-05,
+      "loss": 0.4603,
+      "step": 2464
+    },
+    {
+      "epoch": 2.2972972972972974,
+      "grad_norm": 0.7329336123958825,
+      "learning_rate": 1.301346220227822e-05,
+      "loss": 0.4691,
+      "step": 2465
+    },
+    {
+      "epoch": 2.298229263746505,
+      "grad_norm": 0.7047421994830567,
+      "learning_rate": 1.2996202968588194e-05,
+      "loss": 0.4778,
+      "step": 2466
+    },
+    {
+      "epoch": 2.299161230195713,
+      "grad_norm": 0.7217966382577247,
+      "learning_rate": 1.2978943734898172e-05,
+      "loss": 0.4747,
+      "step": 2467
+    },
+    {
+      "epoch": 2.3000931966449207,
+      "grad_norm": 0.6595456351431601,
+      "learning_rate": 1.2961684501208146e-05,
+      "loss": 0.4781,
+      "step": 2468
+    },
+    {
+      "epoch": 2.3010251630941285,
+      "grad_norm": 0.9272218990423602,
+      "learning_rate": 1.2944425267518124e-05,
+      "loss": 0.4652,
+      "step": 2469
+    },
+    {
+      "epoch": 2.3019571295433363,
+      "grad_norm": 0.6722568031838895,
+      "learning_rate": 1.2927166033828098e-05,
+      "loss": 0.4957,
+      "step": 2470
+    },
+    {
+      "epoch": 2.302889095992544,
+      "grad_norm": 0.6220519866227968,
+      "learning_rate": 1.2909906800138074e-05,
+      "loss": 0.4935,
+      "step": 2471
+    },
+    {
+      "epoch": 2.3038210624417523,
+      "grad_norm": 0.5672119236458582,
+      "learning_rate": 1.2892647566448048e-05,
+      "loss": 0.4635,
+      "step": 2472
+    },
+    {
+      "epoch": 2.3047530288909597,
+      "grad_norm": 0.7215339559349118,
+      "learning_rate": 1.2875388332758026e-05,
+      "loss": 0.5023,
+      "step": 2473
+    },
+    {
+      "epoch": 2.305684995340168,
+      "grad_norm": 0.588126557060427,
+      "learning_rate": 1.2858129099068004e-05,
+      "loss": 0.4476,
+      "step": 2474
+    },
+    {
+      "epoch": 2.3066169617893757,
+      "grad_norm": 0.6373130029083878,
+      "learning_rate": 1.2840869865377978e-05,
+      "loss": 0.4687,
+      "step": 2475
+    },
+    {
+      "epoch": 2.3075489282385835,
+      "grad_norm": 0.7976578822110334,
+      "learning_rate": 1.2823610631687954e-05,
+      "loss": 0.4724,
+      "step": 2476
+    },
+    {
+      "epoch": 2.3084808946877913,
+      "grad_norm": 1.0089966789922347,
+      "learning_rate": 1.2806351397997928e-05,
+      "loss": 0.5078,
+      "step": 2477
+    },
+    {
+      "epoch": 2.309412861136999,
+      "grad_norm": 0.4519387573937339,
+      "learning_rate": 1.2789092164307906e-05,
+      "loss": 0.4556,
+      "step": 2478
+    },
+    {
+      "epoch": 2.310344827586207,
+      "grad_norm": 0.9145055739063183,
+      "learning_rate": 1.277183293061788e-05,
+      "loss": 0.4637,
+      "step": 2479
+    },
+    {
+      "epoch": 2.3112767940354146,
+      "grad_norm": 0.8868270218186437,
+      "learning_rate": 1.2754573696927858e-05,
+      "loss": 0.481,
+      "step": 2480
+    },
+    {
+      "epoch": 2.3122087604846224,
+      "grad_norm": 0.7076279483542248,
+      "learning_rate": 1.2737314463237832e-05,
+      "loss": 0.461,
+      "step": 2481
+    },
+    {
+      "epoch": 2.31314072693383,
+      "grad_norm": 0.6438017717829422,
+      "learning_rate": 1.2720055229547808e-05,
+      "loss": 0.4907,
+      "step": 2482
+    },
+    {
+      "epoch": 2.3140726933830384,
+      "grad_norm": 0.4456065029938304,
+      "learning_rate": 1.2702795995857786e-05,
+      "loss": 0.4509,
+      "step": 2483
+    },
+    {
+      "epoch": 2.315004659832246,
+      "grad_norm": 0.6106245869792355,
+      "learning_rate": 1.268553676216776e-05,
+      "loss": 0.4568,
+      "step": 2484
+    },
+    {
+      "epoch": 2.315936626281454,
+      "grad_norm": 0.7153330302938837,
+      "learning_rate": 1.2668277528477738e-05,
+      "loss": 0.4812,
+      "step": 2485
+    },
+    {
+      "epoch": 2.316868592730662,
+      "grad_norm": 0.6970749610601421,
+      "learning_rate": 1.2651018294787712e-05,
+      "loss": 0.4683,
+      "step": 2486
+    },
+    {
+      "epoch": 2.3178005591798696,
+      "grad_norm": 0.6354347905090649,
+      "learning_rate": 1.2633759061097688e-05,
+      "loss": 0.4789,
+      "step": 2487
+    },
+    {
+      "epoch": 2.3187325256290774,
+      "grad_norm": 0.48519592385488275,
+      "learning_rate": 1.2616499827407662e-05,
+      "loss": 0.4472,
+      "step": 2488
+    },
+    {
+      "epoch": 2.319664492078285,
+      "grad_norm": 0.4744682108197207,
+      "learning_rate": 1.259924059371764e-05,
+      "loss": 0.4552,
+      "step": 2489
+    },
+    {
+      "epoch": 2.320596458527493,
+      "grad_norm": 0.4963637380275077,
+      "learning_rate": 1.2581981360027614e-05,
+      "loss": 0.4379,
+      "step": 2490
+    },
+    {
+      "epoch": 2.3215284249767008,
+      "grad_norm": 0.5179224756467232,
+      "learning_rate": 1.2564722126337592e-05,
+      "loss": 0.4741,
+      "step": 2491
+    },
+    {
+      "epoch": 2.3224603914259085,
+      "grad_norm": 0.5587145966410871,
+      "learning_rate": 1.2547462892647566e-05,
+      "loss": 0.4821,
+      "step": 2492
+    },
+    {
+      "epoch": 2.3233923578751163,
+      "grad_norm": 0.6219701915958086,
+      "learning_rate": 1.2530203658957542e-05,
+      "loss": 0.4958,
+      "step": 2493
+    },
+    {
+      "epoch": 2.3243243243243246,
+      "grad_norm": 0.5966447001735513,
+      "learning_rate": 1.251294442526752e-05,
+      "loss": 0.4679,
+      "step": 2494
+    },
+    {
+      "epoch": 2.325256290773532,
+      "grad_norm": 0.500673796582164,
+      "learning_rate": 1.2495685191577494e-05,
+      "loss": 0.4792,
+      "step": 2495
+    },
+    {
+      "epoch": 2.32618825722274,
+      "grad_norm": 0.5108884214638594,
+      "learning_rate": 1.247842595788747e-05,
+      "loss": 0.4585,
+      "step": 2496
+    },
+    {
+      "epoch": 2.327120223671948,
+      "grad_norm": 0.753778844375792,
+      "learning_rate": 1.2461166724197446e-05,
+      "loss": 0.4793,
+      "step": 2497
+    },
+    {
+      "epoch": 2.3280521901211557,
+      "grad_norm": 0.6005229651966238,
+      "learning_rate": 1.2443907490507422e-05,
+      "loss": 0.4861,
+      "step": 2498
+    },
+    {
+      "epoch": 2.3289841565703635,
+      "grad_norm": 0.6792011238250208,
+      "learning_rate": 1.2426648256817398e-05,
+      "loss": 0.4639,
+      "step": 2499
+    },
+    {
+      "epoch": 2.3299161230195713,
+      "grad_norm": 0.8124175519495141,
+      "learning_rate": 1.2409389023127374e-05,
+      "loss": 0.459,
+      "step": 2500
+    },
+    {
+      "epoch": 2.330848089468779,
+      "grad_norm": 0.5310092710360167,
+      "learning_rate": 1.239212978943735e-05,
+      "loss": 0.4706,
+      "step": 2501
+    },
+    {
+      "epoch": 2.331780055917987,
+      "grad_norm": 0.5187350194894175,
+      "learning_rate": 1.2374870555747326e-05,
+      "loss": 0.4278,
+      "step": 2502
+    },
+    {
+      "epoch": 2.3327120223671947,
+      "grad_norm": 0.6802925441493386,
+      "learning_rate": 1.23576113220573e-05,
+      "loss": 0.4666,
+      "step": 2503
+    },
+    {
+      "epoch": 2.3336439888164024,
+      "grad_norm": 0.5075234741674223,
+      "learning_rate": 1.2340352088367276e-05,
+      "loss": 0.4386,
+      "step": 2504
+    },
+    {
+      "epoch": 2.3345759552656107,
+      "grad_norm": 0.7415535092715253,
+      "learning_rate": 1.2323092854677252e-05,
+      "loss": 0.494,
+      "step": 2505
+    },
+    {
+      "epoch": 2.335507921714818,
+      "grad_norm": 0.4602730819564903,
+      "learning_rate": 1.2305833620987228e-05,
+      "loss": 0.4376,
+      "step": 2506
+    },
+    {
+      "epoch": 2.3364398881640263,
+      "grad_norm": 0.4541235349723257,
+      "learning_rate": 1.2288574387297204e-05,
+      "loss": 0.458,
+      "step": 2507
+    },
+    {
+      "epoch": 2.337371854613234,
+      "grad_norm": 0.6392908494166601,
+      "learning_rate": 1.227131515360718e-05,
+      "loss": 0.4961,
+      "step": 2508
+    },
+    {
+      "epoch": 2.338303821062442,
+      "grad_norm": 0.4558858439839222,
+      "learning_rate": 1.2254055919917156e-05,
+      "loss": 0.4862,
+      "step": 2509
+    },
+    {
+      "epoch": 2.3392357875116496,
+      "grad_norm": 0.5365253244449198,
+      "learning_rate": 1.2236796686227132e-05,
+      "loss": 0.48,
+      "step": 2510
+    },
+    {
+      "epoch": 2.3401677539608574,
+      "grad_norm": 0.8370555945340878,
+      "learning_rate": 1.2219537452537108e-05,
+      "loss": 0.4868,
+      "step": 2511
+    },
+    {
+      "epoch": 2.341099720410065,
+      "grad_norm": 0.5690862842824498,
+      "learning_rate": 1.2202278218847084e-05,
+      "loss": 0.4579,
+      "step": 2512
+    },
+    {
+      "epoch": 2.342031686859273,
+      "grad_norm": 0.7018644167240901,
+      "learning_rate": 1.218501898515706e-05,
+      "loss": 0.4976,
+      "step": 2513
+    },
+    {
+      "epoch": 2.3429636533084808,
+      "grad_norm": 0.8597665514134802,
+      "learning_rate": 1.2167759751467034e-05,
+      "loss": 0.4862,
+      "step": 2514
+    },
+    {
+      "epoch": 2.3438956197576886,
+      "grad_norm": 0.7078845804881593,
+      "learning_rate": 1.215050051777701e-05,
+      "loss": 0.506,
+      "step": 2515
+    },
+    {
+      "epoch": 2.344827586206897,
+      "grad_norm": 0.9375764887374336,
+      "learning_rate": 1.2133241284086986e-05,
+      "loss": 0.5009,
+      "step": 2516
+    },
+    {
+      "epoch": 2.345759552656104,
+      "grad_norm": 0.5290935952067052,
+      "learning_rate": 1.2115982050396962e-05,
+      "loss": 0.4629,
+      "step": 2517
+    },
+    {
+      "epoch": 2.3466915191053124,
+      "grad_norm": 0.6400046397706056,
+      "learning_rate": 1.209872281670694e-05,
+      "loss": 0.4676,
+      "step": 2518
+    },
+    {
+      "epoch": 2.34762348555452,
+      "grad_norm": 0.61701377769622,
+      "learning_rate": 1.2081463583016914e-05,
+      "loss": 0.4806,
+      "step": 2519
+    },
+    {
+      "epoch": 2.348555452003728,
+      "grad_norm": 0.5824265186246463,
+      "learning_rate": 1.206420434932689e-05,
+      "loss": 0.4692,
+      "step": 2520
+    },
+    {
+      "epoch": 2.3494874184529357,
+      "grad_norm": 0.47933860169303943,
+      "learning_rate": 1.2046945115636866e-05,
+      "loss": 0.4519,
+      "step": 2521
+    },
+    {
+      "epoch": 2.3504193849021435,
+      "grad_norm": 0.5855840124349118,
+      "learning_rate": 1.2029685881946842e-05,
+      "loss": 0.4637,
+      "step": 2522
+    },
+    {
+      "epoch": 2.3513513513513513,
+      "grad_norm": 0.7498524475812263,
+      "learning_rate": 1.2012426648256818e-05,
+      "loss": 0.4817,
+      "step": 2523
+    },
+    {
+      "epoch": 2.352283317800559,
+      "grad_norm": 0.6363694326394274,
+      "learning_rate": 1.1995167414566794e-05,
+      "loss": 0.4781,
+      "step": 2524
+    },
+    {
+      "epoch": 2.353215284249767,
+      "grad_norm": 0.47266568394116804,
+      "learning_rate": 1.1977908180876769e-05,
+      "loss": 0.458,
+      "step": 2525
+    },
+    {
+      "epoch": 2.3541472506989747,
+      "grad_norm": 0.8191301556136495,
+      "learning_rate": 1.1960648947186745e-05,
+      "loss": 0.4931,
+      "step": 2526
+    },
+    {
+      "epoch": 2.355079217148183,
+      "grad_norm": 0.5343335920727327,
+      "learning_rate": 1.194338971349672e-05,
+      "loss": 0.4653,
+      "step": 2527
+    },
+    {
+      "epoch": 2.3560111835973903,
+      "grad_norm": 0.6676404175809636,
+      "learning_rate": 1.1926130479806698e-05,
+      "loss": 0.4626,
+      "step": 2528
+    },
+    {
+      "epoch": 2.3569431500465985,
+      "grad_norm": 0.579031032241606,
+      "learning_rate": 1.1908871246116674e-05,
+      "loss": 0.4596,
+      "step": 2529
+    },
+    {
+      "epoch": 2.3578751164958063,
+      "grad_norm": 0.8533389021226002,
+      "learning_rate": 1.1891612012426649e-05,
+      "loss": 0.4828,
+      "step": 2530
+    },
+    {
+      "epoch": 2.358807082945014,
+      "grad_norm": 0.49968063211107333,
+      "learning_rate": 1.1874352778736625e-05,
+      "loss": 0.4695,
+      "step": 2531
+    },
+    {
+      "epoch": 2.359739049394222,
+      "grad_norm": 0.7329894717646723,
+      "learning_rate": 1.18570935450466e-05,
+      "loss": 0.4803,
+      "step": 2532
+    },
+    {
+      "epoch": 2.3606710158434296,
+      "grad_norm": 0.6409622088971437,
+      "learning_rate": 1.1839834311356576e-05,
+      "loss": 0.4623,
+      "step": 2533
+    },
+    {
+      "epoch": 2.3616029822926374,
+      "grad_norm": 0.6933513616231758,
+      "learning_rate": 1.1822575077666552e-05,
+      "loss": 0.4807,
+      "step": 2534
+    },
+    {
+      "epoch": 2.362534948741845,
+      "grad_norm": 0.9239434623114717,
+      "learning_rate": 1.1805315843976528e-05,
+      "loss": 0.4676,
+      "step": 2535
+    },
+    {
+      "epoch": 2.363466915191053,
+      "grad_norm": 0.5870786729352476,
+      "learning_rate": 1.1788056610286503e-05,
+      "loss": 0.4437,
+      "step": 2536
+    },
+    {
+      "epoch": 2.364398881640261,
+      "grad_norm": 1.1008962963315685,
+      "learning_rate": 1.1770797376596479e-05,
+      "loss": 0.4959,
+      "step": 2537
+    },
+    {
+      "epoch": 2.3653308480894686,
+      "grad_norm": 0.5527722735813548,
+      "learning_rate": 1.1753538142906456e-05,
+      "loss": 0.468,
+      "step": 2538
+    },
+    {
+      "epoch": 2.3662628145386764,
+      "grad_norm": 0.8268961957670478,
+      "learning_rate": 1.1736278909216432e-05,
+      "loss": 0.4504,
+      "step": 2539
+    },
+    {
+      "epoch": 2.3671947809878846,
+      "grad_norm": 0.6482546654510687,
+      "learning_rate": 1.1719019675526408e-05,
+      "loss": 0.4715,
+      "step": 2540
+    },
+    {
+      "epoch": 2.3681267474370924,
+      "grad_norm": 0.5796496252667613,
+      "learning_rate": 1.1701760441836383e-05,
+      "loss": 0.4424,
+      "step": 2541
+    },
+    {
+      "epoch": 2.3690587138863,
+      "grad_norm": 0.5096061786164054,
+      "learning_rate": 1.1684501208146359e-05,
+      "loss": 0.471,
+      "step": 2542
+    },
+    {
+      "epoch": 2.369990680335508,
+      "grad_norm": 0.5420974277721192,
+      "learning_rate": 1.1667241974456335e-05,
+      "loss": 0.5118,
+      "step": 2543
+    },
+    {
+      "epoch": 2.3709226467847158,
+      "grad_norm": 0.7562457252599775,
+      "learning_rate": 1.164998274076631e-05,
+      "loss": 0.4945,
+      "step": 2544
+    },
+    {
+      "epoch": 2.3718546132339235,
+      "grad_norm": 0.6585446866185437,
+      "learning_rate": 1.1632723507076287e-05,
+      "loss": 0.4662,
+      "step": 2545
+    },
+    {
+      "epoch": 2.3727865796831313,
+      "grad_norm": 0.5652058824816795,
+      "learning_rate": 1.1615464273386263e-05,
+      "loss": 0.4725,
+      "step": 2546
+    },
+    {
+      "epoch": 2.373718546132339,
+      "grad_norm": 0.8807093672710197,
+      "learning_rate": 1.1598205039696237e-05,
+      "loss": 0.5068,
+      "step": 2547
+    },
+    {
+      "epoch": 2.374650512581547,
+      "grad_norm": 0.5129988315554527,
+      "learning_rate": 1.1580945806006215e-05,
+      "loss": 0.4413,
+      "step": 2548
+    },
+    {
+      "epoch": 2.3755824790307547,
+      "grad_norm": 0.5474442533781154,
+      "learning_rate": 1.156368657231619e-05,
+      "loss": 0.4492,
+      "step": 2549
+    },
+    {
+      "epoch": 2.3765144454799625,
+      "grad_norm": 0.6561225093313061,
+      "learning_rate": 1.1546427338626167e-05,
+      "loss": 0.4697,
+      "step": 2550
+    },
+    {
+      "epoch": 2.3774464119291707,
+      "grad_norm": 0.9442656207338072,
+      "learning_rate": 1.152916810493614e-05,
+      "loss": 0.5076,
+      "step": 2551
+    },
+    {
+      "epoch": 2.3783783783783785,
+      "grad_norm": 0.5433615170587079,
+      "learning_rate": 1.1511908871246117e-05,
+      "loss": 0.4423,
+      "step": 2552
+    },
+    {
+      "epoch": 2.3793103448275863,
+      "grad_norm": 0.6325461401171075,
+      "learning_rate": 1.1494649637556093e-05,
+      "loss": 0.4808,
+      "step": 2553
+    },
+    {
+      "epoch": 2.380242311276794,
+      "grad_norm": 0.7029888280734687,
+      "learning_rate": 1.1477390403866069e-05,
+      "loss": 0.4583,
+      "step": 2554
+    },
+    {
+      "epoch": 2.381174277726002,
+      "grad_norm": 0.5866246563965455,
+      "learning_rate": 1.1460131170176045e-05,
+      "loss": 0.4541,
+      "step": 2555
+    },
+    {
+      "epoch": 2.3821062441752097,
+      "grad_norm": 0.9254833912570127,
+      "learning_rate": 1.144287193648602e-05,
+      "loss": 0.5103,
+      "step": 2556
+    },
+    {
+      "epoch": 2.3830382106244175,
+      "grad_norm": 0.5717913040940031,
+      "learning_rate": 1.1425612702795995e-05,
+      "loss": 0.4567,
+      "step": 2557
+    },
+    {
+      "epoch": 2.3839701770736252,
+      "grad_norm": 0.6656745422632571,
+      "learning_rate": 1.1408353469105973e-05,
+      "loss": 0.4688,
+      "step": 2558
+    },
+    {
+      "epoch": 2.384902143522833,
+      "grad_norm": 0.8274057474194152,
+      "learning_rate": 1.1391094235415949e-05,
+      "loss": 0.4916,
+      "step": 2559
+    },
+    {
+      "epoch": 2.385834109972041,
+      "grad_norm": 0.4657843517850824,
+      "learning_rate": 1.1373835001725925e-05,
+      "loss": 0.4335,
+      "step": 2560
+    },
+    {
+      "epoch": 2.3867660764212486,
+      "grad_norm": 0.676323713078396,
+      "learning_rate": 1.13565757680359e-05,
+      "loss": 0.5171,
+      "step": 2561
+    },
+    {
+      "epoch": 2.387698042870457,
+      "grad_norm": 0.5997498068111198,
+      "learning_rate": 1.1339316534345875e-05,
+      "loss": 0.4589,
+      "step": 2562
+    },
+    {
+      "epoch": 2.3886300093196646,
+      "grad_norm": 0.7571211859921984,
+      "learning_rate": 1.1322057300655851e-05,
+      "loss": 0.5,
+      "step": 2563
+    },
+    {
+      "epoch": 2.3895619757688724,
+      "grad_norm": 0.5600081162924884,
+      "learning_rate": 1.1304798066965827e-05,
+      "loss": 0.4635,
+      "step": 2564
+    },
+    {
+      "epoch": 2.39049394221808,
+      "grad_norm": 0.6249901428619805,
+      "learning_rate": 1.1287538833275803e-05,
+      "loss": 0.4872,
+      "step": 2565
+    },
+    {
+      "epoch": 2.391425908667288,
+      "grad_norm": 0.6864003857178489,
+      "learning_rate": 1.1270279599585779e-05,
+      "loss": 0.4621,
+      "step": 2566
+    },
+    {
+      "epoch": 2.392357875116496,
+      "grad_norm": 0.5296490060204666,
+      "learning_rate": 1.1253020365895755e-05,
+      "loss": 0.4726,
+      "step": 2567
+    },
+    {
+      "epoch": 2.3932898415657036,
+      "grad_norm": 0.45854474262857103,
+      "learning_rate": 1.1235761132205731e-05,
+      "loss": 0.4372,
+      "step": 2568
+    },
+    {
+      "epoch": 2.3942218080149114,
+      "grad_norm": 0.6013564289402252,
+      "learning_rate": 1.1218501898515707e-05,
+      "loss": 0.4936,
+      "step": 2569
+    },
+    {
+      "epoch": 2.395153774464119,
+      "grad_norm": 0.4645440077748411,
+      "learning_rate": 1.1201242664825683e-05,
+      "loss": 0.4616,
+      "step": 2570
+    },
+    {
+      "epoch": 2.396085740913327,
+      "grad_norm": 0.6123643239079283,
+      "learning_rate": 1.1183983431135659e-05,
+      "loss": 0.4592,
+      "step": 2571
+    },
+    {
+      "epoch": 2.3970177073625347,
+      "grad_norm": 0.6145466582687095,
+      "learning_rate": 1.1166724197445635e-05,
+      "loss": 0.4788,
+      "step": 2572
+    },
+    {
+      "epoch": 2.397949673811743,
+      "grad_norm": 0.6140286596843099,
+      "learning_rate": 1.1149464963755609e-05,
+      "loss": 0.473,
+      "step": 2573
+    },
+    {
+      "epoch": 2.3988816402609507,
+      "grad_norm": 0.6287288089295646,
+      "learning_rate": 1.1132205730065585e-05,
+      "loss": 0.4648,
+      "step": 2574
+    },
+    {
+      "epoch": 2.3998136067101585,
+      "grad_norm": 0.8412654912119008,
+      "learning_rate": 1.1114946496375561e-05,
+      "loss": 0.4462,
+      "step": 2575
+    },
+    {
+      "epoch": 2.4007455731593663,
+      "grad_norm": 0.6322828970497746,
+      "learning_rate": 1.1097687262685537e-05,
+      "loss": 0.4728,
+      "step": 2576
+    },
+    {
+      "epoch": 2.401677539608574,
+      "grad_norm": 0.4869478216479367,
+      "learning_rate": 1.1080428028995513e-05,
+      "loss": 0.4489,
+      "step": 2577
+    },
+    {
+      "epoch": 2.402609506057782,
+      "grad_norm": 0.5010605315196897,
+      "learning_rate": 1.1063168795305489e-05,
+      "loss": 0.4534,
+      "step": 2578
+    },
+    {
+      "epoch": 2.4035414725069897,
+      "grad_norm": 0.6871612701484738,
+      "learning_rate": 1.1045909561615465e-05,
+      "loss": 0.4746,
+      "step": 2579
+    },
+    {
+      "epoch": 2.4044734389561975,
+      "grad_norm": 0.6796813560163076,
+      "learning_rate": 1.1028650327925441e-05,
+      "loss": 0.4429,
+      "step": 2580
+    },
+    {
+      "epoch": 2.4054054054054053,
+      "grad_norm": 0.6156226134134076,
+      "learning_rate": 1.1011391094235417e-05,
+      "loss": 0.4934,
+      "step": 2581
+    },
+    {
+      "epoch": 2.406337371854613,
+      "grad_norm": 0.7764457345241804,
+      "learning_rate": 1.0994131860545393e-05,
+      "loss": 0.4946,
+      "step": 2582
+    },
+    {
+      "epoch": 2.407269338303821,
+      "grad_norm": 0.7501042813810537,
+      "learning_rate": 1.0976872626855369e-05,
+      "loss": 0.493,
+      "step": 2583
+    },
+    {
+      "epoch": 2.408201304753029,
+      "grad_norm": 0.48477011964330313,
+      "learning_rate": 1.0959613393165343e-05,
+      "loss": 0.497,
+      "step": 2584
+    },
+    {
+      "epoch": 2.409133271202237,
+      "grad_norm": 0.5955596878813851,
+      "learning_rate": 1.094235415947532e-05,
+      "loss": 0.4716,
+      "step": 2585
+    },
+    {
+      "epoch": 2.4100652376514446,
+      "grad_norm": 0.6306059618408234,
+      "learning_rate": 1.0925094925785295e-05,
+      "loss": 0.4586,
+      "step": 2586
+    },
+    {
+      "epoch": 2.4109972041006524,
+      "grad_norm": 0.6907483499716874,
+      "learning_rate": 1.0907835692095271e-05,
+      "loss": 0.4963,
+      "step": 2587
+    },
+    {
+      "epoch": 2.4119291705498602,
+      "grad_norm": 0.39705152334600863,
+      "learning_rate": 1.0890576458405247e-05,
+      "loss": 0.4214,
+      "step": 2588
+    },
+    {
+      "epoch": 2.412861136999068,
+      "grad_norm": 0.5661326554180909,
+      "learning_rate": 1.0873317224715223e-05,
+      "loss": 0.4672,
+      "step": 2589
+    },
+    {
+      "epoch": 2.413793103448276,
+      "grad_norm": 0.6322057494834953,
+      "learning_rate": 1.0856057991025199e-05,
+      "loss": 0.456,
+      "step": 2590
+    },
+    {
+      "epoch": 2.4147250698974836,
+      "grad_norm": 0.46619159579132796,
+      "learning_rate": 1.0838798757335175e-05,
+      "loss": 0.4356,
+      "step": 2591
+    },
+    {
+      "epoch": 2.4156570363466914,
+      "grad_norm": 0.7944215092832493,
+      "learning_rate": 1.0821539523645151e-05,
+      "loss": 0.47,
+      "step": 2592
+    },
+    {
+      "epoch": 2.416589002795899,
+      "grad_norm": 0.6030687143705372,
+      "learning_rate": 1.0804280289955127e-05,
+      "loss": 0.4791,
+      "step": 2593
+    },
+    {
+      "epoch": 2.417520969245107,
+      "grad_norm": 0.4943648313422949,
+      "learning_rate": 1.0787021056265103e-05,
+      "loss": 0.4728,
+      "step": 2594
+    },
+    {
+      "epoch": 2.418452935694315,
+      "grad_norm": 0.5215932367824578,
+      "learning_rate": 1.0769761822575077e-05,
+      "loss": 0.4554,
+      "step": 2595
+    },
+    {
+      "epoch": 2.419384902143523,
+      "grad_norm": 0.7600357700959313,
+      "learning_rate": 1.0752502588885053e-05,
+      "loss": 0.5034,
+      "step": 2596
+    },
+    {
+      "epoch": 2.4203168685927308,
+      "grad_norm": 0.6464581052399756,
+      "learning_rate": 1.073524335519503e-05,
+      "loss": 0.4852,
+      "step": 2597
+    },
+    {
+      "epoch": 2.4212488350419386,
+      "grad_norm": 0.5926886344537118,
+      "learning_rate": 1.0717984121505005e-05,
+      "loss": 0.4612,
+      "step": 2598
+    },
+    {
+      "epoch": 2.4221808014911463,
+      "grad_norm": 0.5126405016605213,
+      "learning_rate": 1.0700724887814981e-05,
+      "loss": 0.434,
+      "step": 2599
+    },
+    {
+      "epoch": 2.423112767940354,
+      "grad_norm": 0.45003098150803206,
+      "learning_rate": 1.0683465654124957e-05,
+      "loss": 0.4413,
+      "step": 2600
+    },
+    {
+      "epoch": 2.424044734389562,
+      "grad_norm": 0.8060213424413334,
+      "learning_rate": 1.0666206420434933e-05,
+      "loss": 0.4678,
+      "step": 2601
+    },
+    {
+      "epoch": 2.4249767008387697,
+      "grad_norm": 0.6280145917004484,
+      "learning_rate": 1.064894718674491e-05,
+      "loss": 0.4591,
+      "step": 2602
+    },
+    {
+      "epoch": 2.4259086672879775,
+      "grad_norm": 0.7275708040015814,
+      "learning_rate": 1.0631687953054885e-05,
+      "loss": 0.4913,
+      "step": 2603
+    },
+    {
+      "epoch": 2.4268406337371853,
+      "grad_norm": 0.792570918475759,
+      "learning_rate": 1.0614428719364861e-05,
+      "loss": 0.4973,
+      "step": 2604
+    },
+    {
+      "epoch": 2.427772600186393,
+      "grad_norm": 0.5222102607944934,
+      "learning_rate": 1.0597169485674835e-05,
+      "loss": 0.4724,
+      "step": 2605
+    },
+    {
+      "epoch": 2.4287045666356013,
+      "grad_norm": 0.6538104062675449,
+      "learning_rate": 1.0579910251984811e-05,
+      "loss": 0.4813,
+      "step": 2606
+    },
+    {
+      "epoch": 2.429636533084809,
+      "grad_norm": 0.5267102905825599,
+      "learning_rate": 1.0562651018294787e-05,
+      "loss": 0.4465,
+      "step": 2607
+    },
+    {
+      "epoch": 2.430568499534017,
+      "grad_norm": 0.7082468655051846,
+      "learning_rate": 1.0545391784604763e-05,
+      "loss": 0.4698,
+      "step": 2608
+    },
+    {
+      "epoch": 2.4315004659832247,
+      "grad_norm": 0.4620036157872774,
+      "learning_rate": 1.0528132550914741e-05,
+      "loss": 0.4304,
+      "step": 2609
+    },
+    {
+      "epoch": 2.4324324324324325,
+      "grad_norm": 0.4870715178722162,
+      "learning_rate": 1.0510873317224715e-05,
+      "loss": 0.4483,
+      "step": 2610
+    },
+    {
+      "epoch": 2.4333643988816402,
+      "grad_norm": 0.651765477901892,
+      "learning_rate": 1.0493614083534691e-05,
+      "loss": 0.4954,
+      "step": 2611
+    },
+    {
+      "epoch": 2.434296365330848,
+      "grad_norm": 0.5138786772385858,
+      "learning_rate": 1.0476354849844667e-05,
+      "loss": 0.4699,
+      "step": 2612
+    },
+    {
+      "epoch": 2.435228331780056,
+      "grad_norm": 0.6019658616923824,
+      "learning_rate": 1.0459095616154643e-05,
+      "loss": 0.4369,
+      "step": 2613
+    },
+    {
+      "epoch": 2.4361602982292636,
+      "grad_norm": 0.5750895956731356,
+      "learning_rate": 1.044183638246462e-05,
+      "loss": 0.447,
+      "step": 2614
+    },
+    {
+      "epoch": 2.4370922646784714,
+      "grad_norm": 0.4117704102745728,
+      "learning_rate": 1.0424577148774595e-05,
+      "loss": 0.4452,
+      "step": 2615
+    },
+    {
+      "epoch": 2.438024231127679,
+      "grad_norm": 0.675937912968999,
+      "learning_rate": 1.040731791508457e-05,
+      "loss": 0.4603,
+      "step": 2616
+    },
+    {
+      "epoch": 2.4389561975768874,
+      "grad_norm": 0.6224067299975431,
+      "learning_rate": 1.0390058681394546e-05,
+      "loss": 0.4891,
+      "step": 2617
+    },
+    {
+      "epoch": 2.439888164026095,
+      "grad_norm": 0.5218808889726185,
+      "learning_rate": 1.0372799447704522e-05,
+      "loss": 0.442,
+      "step": 2618
+    },
+    {
+      "epoch": 2.440820130475303,
+      "grad_norm": 0.6967646726103929,
+      "learning_rate": 1.03555402140145e-05,
+      "loss": 0.4833,
+      "step": 2619
+    },
+    {
+      "epoch": 2.441752096924511,
+      "grad_norm": 0.5469250684060446,
+      "learning_rate": 1.0338280980324475e-05,
+      "loss": 0.4672,
+      "step": 2620
+    },
+    {
+      "epoch": 2.4426840633737186,
+      "grad_norm": 0.7402081403346545,
+      "learning_rate": 1.032102174663445e-05,
+      "loss": 0.5103,
+      "step": 2621
+    },
+    {
+      "epoch": 2.4436160298229264,
+      "grad_norm": 0.5901182573254528,
+      "learning_rate": 1.0303762512944426e-05,
+      "loss": 0.4632,
+      "step": 2622
+    },
+    {
+      "epoch": 2.444547996272134,
+      "grad_norm": 0.7742989412007412,
+      "learning_rate": 1.0286503279254402e-05,
+      "loss": 0.4695,
+      "step": 2623
+    },
+    {
+      "epoch": 2.445479962721342,
+      "grad_norm": 0.7645449135792415,
+      "learning_rate": 1.0269244045564377e-05,
+      "loss": 0.4625,
+      "step": 2624
+    },
+    {
+      "epoch": 2.4464119291705497,
+      "grad_norm": 0.5842988794581588,
+      "learning_rate": 1.0251984811874353e-05,
+      "loss": 0.4577,
+      "step": 2625
+    },
+    {
+      "epoch": 2.4473438956197575,
+      "grad_norm": 0.6130249010842859,
+      "learning_rate": 1.023472557818433e-05,
+      "loss": 0.4644,
+      "step": 2626
+    },
+    {
+      "epoch": 2.4482758620689653,
+      "grad_norm": 0.632634852706617,
+      "learning_rate": 1.0217466344494304e-05,
+      "loss": 0.4557,
+      "step": 2627
+    },
+    {
+      "epoch": 2.4492078285181735,
+      "grad_norm": 0.8048175667710407,
+      "learning_rate": 1.020020711080428e-05,
+      "loss": 0.4939,
+      "step": 2628
+    },
+    {
+      "epoch": 2.4501397949673813,
+      "grad_norm": 0.4980864571242558,
+      "learning_rate": 1.0182947877114257e-05,
+      "loss": 0.4663,
+      "step": 2629
+    },
+    {
+      "epoch": 2.451071761416589,
+      "grad_norm": 0.6536788290291905,
+      "learning_rate": 1.0165688643424233e-05,
+      "loss": 0.4627,
+      "step": 2630
+    },
+    {
+      "epoch": 2.452003727865797,
+      "grad_norm": 0.5410894145342322,
+      "learning_rate": 1.014842940973421e-05,
+      "loss": 0.4626,
+      "step": 2631
+    },
+    {
+      "epoch": 2.4529356943150047,
+      "grad_norm": 0.6488561951534914,
+      "learning_rate": 1.0131170176044184e-05,
+      "loss": 0.4582,
+      "step": 2632
+    },
+    {
+      "epoch": 2.4538676607642125,
+      "grad_norm": 0.6995951171682977,
+      "learning_rate": 1.011391094235416e-05,
+      "loss": 0.4842,
+      "step": 2633
+    },
+    {
+      "epoch": 2.4547996272134203,
+      "grad_norm": 0.6074451974912711,
+      "learning_rate": 1.0096651708664136e-05,
+      "loss": 0.4258,
+      "step": 2634
+    },
+    {
+      "epoch": 2.455731593662628,
+      "grad_norm": 0.5310472754736989,
+      "learning_rate": 1.0079392474974112e-05,
+      "loss": 0.4523,
+      "step": 2635
+    },
+    {
+      "epoch": 2.456663560111836,
+      "grad_norm": 0.6272202401042832,
+      "learning_rate": 1.0062133241284088e-05,
+      "loss": 0.4432,
+      "step": 2636
+    },
+    {
+      "epoch": 2.4575955265610436,
+      "grad_norm": 0.7291508024798216,
+      "learning_rate": 1.0044874007594064e-05,
+      "loss": 0.4916,
+      "step": 2637
+    },
+    {
+      "epoch": 2.4585274930102514,
+      "grad_norm": 1.1744306741557693,
+      "learning_rate": 1.0027614773904038e-05,
+      "loss": 0.4395,
+      "step": 2638
+    },
+    {
+      "epoch": 2.4594594594594597,
+      "grad_norm": 0.5471766476906619,
+      "learning_rate": 1.0010355540214016e-05,
+      "loss": 0.4574,
+      "step": 2639
+    },
+    {
+      "epoch": 2.4603914259086674,
+      "grad_norm": 0.7194110908900565,
+      "learning_rate": 9.993096306523992e-06,
+      "loss": 0.4702,
+      "step": 2640
+    },
+    {
+      "epoch": 2.4613233923578752,
+      "grad_norm": 0.6689596475542468,
+      "learning_rate": 9.975837072833968e-06,
+      "loss": 0.4736,
+      "step": 2641
+    },
+    {
+      "epoch": 2.462255358807083,
+      "grad_norm": 0.5902788735430107,
+      "learning_rate": 9.958577839143944e-06,
+      "loss": 0.4638,
+      "step": 2642
+    },
+    {
+      "epoch": 2.463187325256291,
+      "grad_norm": 0.8077022583911018,
+      "learning_rate": 9.941318605453918e-06,
+      "loss": 0.4775,
+      "step": 2643
+    },
+    {
+      "epoch": 2.4641192917054986,
+      "grad_norm": 0.5370347396862831,
+      "learning_rate": 9.924059371763894e-06,
+      "loss": 0.4564,
+      "step": 2644
+    },
+    {
+      "epoch": 2.4650512581547064,
+      "grad_norm": 0.5134991236511792,
+      "learning_rate": 9.90680013807387e-06,
+      "loss": 0.4198,
+      "step": 2645
+    },
+    {
+      "epoch": 2.465983224603914,
+      "grad_norm": 0.6038093086126517,
+      "learning_rate": 9.889540904383846e-06,
+      "loss": 0.4818,
+      "step": 2646
+    },
+    {
+      "epoch": 2.466915191053122,
+      "grad_norm": 0.6553200248734071,
+      "learning_rate": 9.872281670693822e-06,
+      "loss": 0.4718,
+      "step": 2647
+    },
+    {
+      "epoch": 2.4678471575023297,
+      "grad_norm": 0.9975369854750485,
+      "learning_rate": 9.855022437003798e-06,
+      "loss": 0.495,
+      "step": 2648
+    },
+    {
+      "epoch": 2.4687791239515375,
+      "grad_norm": 0.4211582017642352,
+      "learning_rate": 9.837763203313774e-06,
+      "loss": 0.4373,
+      "step": 2649
+    },
+    {
+      "epoch": 2.4697110904007458,
+      "grad_norm": 0.4514513881056568,
+      "learning_rate": 9.82050396962375e-06,
+      "loss": 0.4729,
+      "step": 2650
+    },
+    {
+      "epoch": 2.4706430568499536,
+      "grad_norm": 0.595911144513283,
+      "learning_rate": 9.803244735933726e-06,
+      "loss": 0.4565,
+      "step": 2651
+    },
+    {
+      "epoch": 2.4715750232991613,
+      "grad_norm": 0.4129028723459551,
+      "learning_rate": 9.785985502243702e-06,
+      "loss": 0.4619,
+      "step": 2652
+    },
+    {
+      "epoch": 2.472506989748369,
+      "grad_norm": 0.8495631026619511,
+      "learning_rate": 9.768726268553676e-06,
+      "loss": 0.4694,
+      "step": 2653
+    },
+    {
+      "epoch": 2.473438956197577,
+      "grad_norm": 0.584292005495433,
+      "learning_rate": 9.751467034863652e-06,
+      "loss": 0.4674,
+      "step": 2654
+    },
+    {
+      "epoch": 2.4743709226467847,
+      "grad_norm": 0.5550002735291573,
+      "learning_rate": 9.734207801173628e-06,
+      "loss": 0.4711,
+      "step": 2655
+    },
+    {
+      "epoch": 2.4753028890959925,
+      "grad_norm": 0.4581620057361661,
+      "learning_rate": 9.716948567483604e-06,
+      "loss": 0.4544,
+      "step": 2656
+    },
+    {
+      "epoch": 2.4762348555452003,
+      "grad_norm": 0.7533500101196682,
+      "learning_rate": 9.69968933379358e-06,
+      "loss": 0.4568,
+      "step": 2657
+    },
+    {
+      "epoch": 2.477166821994408,
+      "grad_norm": 0.5511007865030613,
+      "learning_rate": 9.682430100103556e-06,
+      "loss": 0.4512,
+      "step": 2658
+    },
+    {
+      "epoch": 2.478098788443616,
+      "grad_norm": 0.4622942226381634,
+      "learning_rate": 9.665170866413532e-06,
+      "loss": 0.4568,
+      "step": 2659
+    },
+    {
+      "epoch": 2.4790307548928237,
+      "grad_norm": 0.5333084996841727,
+      "learning_rate": 9.647911632723508e-06,
+      "loss": 0.4585,
+      "step": 2660
+    },
+    {
+      "epoch": 2.479962721342032,
+      "grad_norm": 0.5860228701419025,
+      "learning_rate": 9.630652399033484e-06,
+      "loss": 0.4663,
+      "step": 2661
+    },
+    {
+      "epoch": 2.4808946877912397,
+      "grad_norm": 0.6395771548787595,
+      "learning_rate": 9.61339316534346e-06,
+      "loss": 0.4844,
+      "step": 2662
+    },
+    {
+      "epoch": 2.4818266542404475,
+      "grad_norm": 0.37679772952718615,
+      "learning_rate": 9.596133931653436e-06,
+      "loss": 0.4666,
+      "step": 2663
+    },
+    {
+      "epoch": 2.4827586206896552,
+      "grad_norm": 0.6332297146113721,
+      "learning_rate": 9.57887469796341e-06,
+      "loss": 0.4421,
+      "step": 2664
+    },
+    {
+      "epoch": 2.483690587138863,
+      "grad_norm": 0.5712204121228958,
+      "learning_rate": 9.561615464273386e-06,
+      "loss": 0.4435,
+      "step": 2665
+    },
+    {
+      "epoch": 2.484622553588071,
+      "grad_norm": 1.0083354709224472,
+      "learning_rate": 9.544356230583362e-06,
+      "loss": 0.4714,
+      "step": 2666
+    },
+    {
+      "epoch": 2.4855545200372786,
+      "grad_norm": 0.6409340655195032,
+      "learning_rate": 9.527096996893338e-06,
+      "loss": 0.4928,
+      "step": 2667
+    },
+    {
+      "epoch": 2.4864864864864864,
+      "grad_norm": 0.6668652050552927,
+      "learning_rate": 9.509837763203314e-06,
+      "loss": 0.4436,
+      "step": 2668
+    },
+    {
+      "epoch": 2.487418452935694,
+      "grad_norm": 0.6740341246116914,
+      "learning_rate": 9.49257852951329e-06,
+      "loss": 0.4574,
+      "step": 2669
+    },
+    {
+      "epoch": 2.488350419384902,
+      "grad_norm": 0.43189086323631315,
+      "learning_rate": 9.475319295823266e-06,
+      "loss": 0.4602,
+      "step": 2670
+    },
+    {
+      "epoch": 2.4892823858341098,
+      "grad_norm": 0.7860855363391492,
+      "learning_rate": 9.458060062133242e-06,
+      "loss": 0.4956,
+      "step": 2671
+    },
+    {
+      "epoch": 2.490214352283318,
+      "grad_norm": 0.41379431658264065,
+      "learning_rate": 9.440800828443218e-06,
+      "loss": 0.4363,
+      "step": 2672
+    },
+    {
+      "epoch": 2.491146318732526,
+      "grad_norm": 0.7016849467126017,
+      "learning_rate": 9.423541594753194e-06,
+      "loss": 0.5206,
+      "step": 2673
+    },
+    {
+      "epoch": 2.4920782851817336,
+      "grad_norm": 0.6391338195072787,
+      "learning_rate": 9.40628236106317e-06,
+      "loss": 0.4595,
+      "step": 2674
+    },
+    {
+      "epoch": 2.4930102516309414,
+      "grad_norm": 0.521147901238414,
+      "learning_rate": 9.389023127373144e-06,
+      "loss": 0.4788,
+      "step": 2675
+    },
+    {
+      "epoch": 2.493942218080149,
+      "grad_norm": 0.742968859548861,
+      "learning_rate": 9.37176389368312e-06,
+      "loss": 0.4774,
+      "step": 2676
+    },
+    {
+      "epoch": 2.494874184529357,
+      "grad_norm": 0.6236159182027795,
+      "learning_rate": 9.354504659993096e-06,
+      "loss": 0.4744,
+      "step": 2677
+    },
+    {
+      "epoch": 2.4958061509785647,
+      "grad_norm": 0.5869045917466519,
+      "learning_rate": 9.337245426303072e-06,
+      "loss": 0.4597,
+      "step": 2678
+    },
+    {
+      "epoch": 2.4967381174277725,
+      "grad_norm": 0.680373165140525,
+      "learning_rate": 9.31998619261305e-06,
+      "loss": 0.477,
+      "step": 2679
+    },
+    {
+      "epoch": 2.4976700838769803,
+      "grad_norm": 0.6376179477085885,
+      "learning_rate": 9.302726958923024e-06,
+      "loss": 0.4869,
+      "step": 2680
+    },
+    {
+      "epoch": 2.498602050326188,
+      "grad_norm": 0.43133109476885007,
+      "learning_rate": 9.285467725233e-06,
+      "loss": 0.4606,
+      "step": 2681
+    },
+    {
+      "epoch": 2.499534016775396,
+      "grad_norm": 0.4486462517289343,
+      "learning_rate": 9.268208491542976e-06,
+      "loss": 0.432,
+      "step": 2682
+    },
+    {
+      "epoch": 2.500465983224604,
+      "grad_norm": 0.4945637231143179,
+      "learning_rate": 9.250949257852952e-06,
+      "loss": 0.4384,
+      "step": 2683
+    },
+    {
+      "epoch": 2.501397949673812,
+      "grad_norm": 0.5075913669803361,
+      "learning_rate": 9.233690024162928e-06,
+      "loss": 0.4398,
+      "step": 2684
+    },
+    {
+      "epoch": 2.5023299161230197,
+      "grad_norm": 0.8362133628655926,
+      "learning_rate": 9.216430790472904e-06,
+      "loss": 0.4932,
+      "step": 2685
+    },
+    {
+      "epoch": 2.5032618825722275,
+      "grad_norm": 0.7995939479688543,
+      "learning_rate": 9.199171556782878e-06,
+      "loss": 0.509,
+      "step": 2686
+    },
+    {
+      "epoch": 2.5041938490214353,
+      "grad_norm": 0.6379026645466169,
+      "learning_rate": 9.181912323092854e-06,
+      "loss": 0.4718,
+      "step": 2687
+    },
+    {
+      "epoch": 2.505125815470643,
+      "grad_norm": 0.6379052374910722,
+      "learning_rate": 9.16465308940283e-06,
+      "loss": 0.4685,
+      "step": 2688
+    },
+    {
+      "epoch": 2.506057781919851,
+      "grad_norm": 0.6940768156140386,
+      "learning_rate": 9.147393855712806e-06,
+      "loss": 0.4937,
+      "step": 2689
+    },
+    {
+      "epoch": 2.5069897483690586,
+      "grad_norm": 0.48561915761665747,
+      "learning_rate": 9.130134622022784e-06,
+      "loss": 0.4735,
+      "step": 2690
+    },
+    {
+      "epoch": 2.5079217148182664,
+      "grad_norm": 0.7887354288357781,
+      "learning_rate": 9.112875388332758e-06,
+      "loss": 0.4594,
+      "step": 2691
+    },
+    {
+      "epoch": 2.508853681267474,
+      "grad_norm": 0.5910643158118858,
+      "learning_rate": 9.095616154642734e-06,
+      "loss": 0.4508,
+      "step": 2692
+    },
+    {
+      "epoch": 2.509785647716682,
+      "grad_norm": 0.7968545124176015,
+      "learning_rate": 9.07835692095271e-06,
+      "loss": 0.4763,
+      "step": 2693
+    },
+    {
+      "epoch": 2.5107176141658902,
+      "grad_norm": 0.44883591692332986,
+      "learning_rate": 9.061097687262686e-06,
+      "loss": 0.4496,
+      "step": 2694
+    },
+    {
+      "epoch": 2.511649580615098,
+      "grad_norm": 0.5275420854594353,
+      "learning_rate": 9.043838453572662e-06,
+      "loss": 0.4521,
+      "step": 2695
+    },
+    {
+      "epoch": 2.512581547064306,
+      "grad_norm": 0.430722778095508,
+      "learning_rate": 9.026579219882638e-06,
+      "loss": 0.4541,
+      "step": 2696
+    },
+    {
+      "epoch": 2.5135135135135136,
+      "grad_norm": 0.5890044327404811,
+      "learning_rate": 9.009319986192612e-06,
+      "loss": 0.4608,
+      "step": 2697
+    },
+    {
+      "epoch": 2.5144454799627214,
+      "grad_norm": 0.6758807863568486,
+      "learning_rate": 8.992060752502588e-06,
+      "loss": 0.4758,
+      "step": 2698
+    },
+    {
+      "epoch": 2.515377446411929,
+      "grad_norm": 0.47714511908316376,
+      "learning_rate": 8.974801518812564e-06,
+      "loss": 0.4331,
+      "step": 2699
+    },
+    {
+      "epoch": 2.516309412861137,
+      "grad_norm": 0.6663715771268065,
+      "learning_rate": 8.957542285122542e-06,
+      "loss": 0.4634,
+      "step": 2700
+    },
+    {
+      "epoch": 2.5172413793103448,
+      "grad_norm": 0.565595472738953,
+      "learning_rate": 8.940283051432516e-06,
+      "loss": 0.4475,
+      "step": 2701
+    },
+    {
+      "epoch": 2.5181733457595525,
+      "grad_norm": 0.5696824458385397,
+      "learning_rate": 8.923023817742492e-06,
+      "loss": 0.4875,
+      "step": 2702
+    },
+    {
+      "epoch": 2.5191053122087603,
+      "grad_norm": 0.5068526006901345,
+      "learning_rate": 8.905764584052468e-06,
+      "loss": 0.4307,
+      "step": 2703
+    },
+    {
+      "epoch": 2.520037278657968,
+      "grad_norm": 0.8814190198201064,
+      "learning_rate": 8.888505350362444e-06,
+      "loss": 0.4641,
+      "step": 2704
+    },
+    {
+      "epoch": 2.5209692451071763,
+      "grad_norm": 0.8242091122868973,
+      "learning_rate": 8.87124611667242e-06,
+      "loss": 0.5065,
+      "step": 2705
+    },
+    {
+      "epoch": 2.5219012115563837,
+      "grad_norm": 0.6709373321773219,
+      "learning_rate": 8.853986882982396e-06,
+      "loss": 0.4734,
+      "step": 2706
+    },
+    {
+      "epoch": 2.522833178005592,
+      "grad_norm": 0.7278391432210212,
+      "learning_rate": 8.83672764929237e-06,
+      "loss": 0.51,
+      "step": 2707
+    },
+    {
+      "epoch": 2.5237651444547997,
+      "grad_norm": 0.5919073507372234,
+      "learning_rate": 8.819468415602347e-06,
+      "loss": 0.4801,
+      "step": 2708
+    },
+    {
+      "epoch": 2.5246971109040075,
+      "grad_norm": 0.4426565762983607,
+      "learning_rate": 8.802209181912323e-06,
+      "loss": 0.4364,
+      "step": 2709
+    },
+    {
+      "epoch": 2.5256290773532153,
+      "grad_norm": 0.5884219393037792,
+      "learning_rate": 8.7849499482223e-06,
+      "loss": 0.4556,
+      "step": 2710
+    },
+    {
+      "epoch": 2.526561043802423,
+      "grad_norm": 0.5374942194583858,
+      "learning_rate": 8.767690714532276e-06,
+      "loss": 0.4635,
+      "step": 2711
+    },
+    {
+      "epoch": 2.527493010251631,
+      "grad_norm": 1.286844473876209,
+      "learning_rate": 8.75043148084225e-06,
+      "loss": 0.4678,
+      "step": 2712
+    },
+    {
+      "epoch": 2.5284249767008387,
+      "grad_norm": 0.7328503068676866,
+      "learning_rate": 8.733172247152227e-06,
+      "loss": 0.4804,
+      "step": 2713
+    },
+    {
+      "epoch": 2.5293569431500464,
+      "grad_norm": 0.6748066773071,
+      "learning_rate": 8.715913013462203e-06,
+      "loss": 0.4608,
+      "step": 2714
+    },
+    {
+      "epoch": 2.5302889095992542,
+      "grad_norm": 1.344721025169338,
+      "learning_rate": 8.698653779772179e-06,
+      "loss": 0.5644,
+      "step": 2715
+    },
+    {
+      "epoch": 2.5312208760484625,
+      "grad_norm": 0.5954564090822319,
+      "learning_rate": 8.681394546082154e-06,
+      "loss": 0.4429,
+      "step": 2716
+    },
+    {
+      "epoch": 2.53215284249767,
+      "grad_norm": 0.7937089071135105,
+      "learning_rate": 8.66413531239213e-06,
+      "loss": 0.5219,
+      "step": 2717
+    },
+    {
+      "epoch": 2.533084808946878,
+      "grad_norm": 0.6011347546511806,
+      "learning_rate": 8.646876078702105e-06,
+      "loss": 0.4838,
+      "step": 2718
+    },
+    {
+      "epoch": 2.534016775396086,
+      "grad_norm": 0.434812496054289,
+      "learning_rate": 8.62961684501208e-06,
+      "loss": 0.4534,
+      "step": 2719
+    },
+    {
+      "epoch": 2.5349487418452936,
+      "grad_norm": 0.4256876254405259,
+      "learning_rate": 8.612357611322058e-06,
+      "loss": 0.4419,
+      "step": 2720
+    },
+    {
+      "epoch": 2.5358807082945014,
+      "grad_norm": 0.48745925936353834,
+      "learning_rate": 8.595098377632034e-06,
+      "loss": 0.4333,
+      "step": 2721
+    },
+    {
+      "epoch": 2.536812674743709,
+      "grad_norm": 0.42976162280147534,
+      "learning_rate": 8.57783914394201e-06,
+      "loss": 0.4591,
+      "step": 2722
+    },
+    {
+      "epoch": 2.537744641192917,
+      "grad_norm": 0.7090292348473063,
+      "learning_rate": 8.560579910251985e-06,
+      "loss": 0.4558,
+      "step": 2723
+    },
+    {
+      "epoch": 2.5386766076421248,
+      "grad_norm": 0.45329545967680035,
+      "learning_rate": 8.54332067656196e-06,
+      "loss": 0.4285,
+      "step": 2724
+    },
+    {
+      "epoch": 2.5396085740913326,
+      "grad_norm": 0.588688014968632,
+      "learning_rate": 8.526061442871937e-06,
+      "loss": 0.4542,
+      "step": 2725
+    },
+    {
+      "epoch": 2.5405405405405403,
+      "grad_norm": 0.754944600692452,
+      "learning_rate": 8.508802209181913e-06,
+      "loss": 0.4547,
+      "step": 2726
+    },
+    {
+      "epoch": 2.5414725069897486,
+      "grad_norm": 0.5433795257248808,
+      "learning_rate": 8.491542975491889e-06,
+      "loss": 0.4629,
+      "step": 2727
+    },
+    {
+      "epoch": 2.542404473438956,
+      "grad_norm": 0.535591545685451,
+      "learning_rate": 8.474283741801865e-06,
+      "loss": 0.4443,
+      "step": 2728
+    },
+    {
+      "epoch": 2.543336439888164,
+      "grad_norm": 0.4294512935349525,
+      "learning_rate": 8.457024508111839e-06,
+      "loss": 0.4507,
+      "step": 2729
+    },
+    {
+      "epoch": 2.544268406337372,
+      "grad_norm": 0.5449301019233997,
+      "learning_rate": 8.439765274421817e-06,
+      "loss": 0.448,
+      "step": 2730
+    },
+    {
+      "epoch": 2.5452003727865797,
+      "grad_norm": 0.4325937890954186,
+      "learning_rate": 8.422506040731793e-06,
+      "loss": 0.4535,
+      "step": 2731
+    },
+    {
+      "epoch": 2.5461323392357875,
+      "grad_norm": 0.7892452366694063,
+      "learning_rate": 8.405246807041769e-06,
+      "loss": 0.4592,
+      "step": 2732
+    },
+    {
+      "epoch": 2.5470643056849953,
+      "grad_norm": 0.7005064234540066,
+      "learning_rate": 8.387987573351745e-06,
+      "loss": 0.475,
+      "step": 2733
+    },
+    {
+      "epoch": 2.547996272134203,
+      "grad_norm": 0.4515349917452982,
+      "learning_rate": 8.370728339661719e-06,
+      "loss": 0.4537,
+      "step": 2734
+    },
+    {
+      "epoch": 2.548928238583411,
+      "grad_norm": 0.4794942502322273,
+      "learning_rate": 8.353469105971695e-06,
+      "loss": 0.4576,
+      "step": 2735
+    },
+    {
+      "epoch": 2.5498602050326187,
+      "grad_norm": 0.5421693890348075,
+      "learning_rate": 8.33620987228167e-06,
+      "loss": 0.4339,
+      "step": 2736
+    },
+    {
+      "epoch": 2.5507921714818265,
+      "grad_norm": 0.5169241055311681,
+      "learning_rate": 8.318950638591647e-06,
+      "loss": 0.448,
+      "step": 2737
+    },
+    {
+      "epoch": 2.5517241379310347,
+      "grad_norm": 0.5092218073466208,
+      "learning_rate": 8.301691404901623e-06,
+      "loss": 0.4397,
+      "step": 2738
+    },
+    {
+      "epoch": 2.552656104380242,
+      "grad_norm": 1.0235291170453298,
+      "learning_rate": 8.284432171211599e-06,
+      "loss": 0.4638,
+      "step": 2739
+    },
+    {
+      "epoch": 2.5535880708294503,
+      "grad_norm": 0.5030768707617875,
+      "learning_rate": 8.267172937521575e-06,
+      "loss": 0.46,
+      "step": 2740
+    },
+    {
+      "epoch": 2.554520037278658,
+      "grad_norm": 0.5393795011547167,
+      "learning_rate": 8.24991370383155e-06,
+      "loss": 0.4438,
+      "step": 2741
+    },
+    {
+      "epoch": 2.555452003727866,
+      "grad_norm": 0.5559359490937911,
+      "learning_rate": 8.232654470141527e-06,
+      "loss": 0.4351,
+      "step": 2742
+    },
+    {
+      "epoch": 2.5563839701770736,
+      "grad_norm": 0.6798293959155822,
+      "learning_rate": 8.215395236451503e-06,
+      "loss": 0.497,
+      "step": 2743
+    },
+    {
+      "epoch": 2.5573159366262814,
+      "grad_norm": 0.6156138015543042,
+      "learning_rate": 8.198136002761479e-06,
+      "loss": 0.4781,
+      "step": 2744
+    },
+    {
+      "epoch": 2.558247903075489,
+      "grad_norm": 0.459151201130525,
+      "learning_rate": 8.180876769071453e-06,
+      "loss": 0.4518,
+      "step": 2745
+    },
+    {
+      "epoch": 2.559179869524697,
+      "grad_norm": 0.7184442006467706,
+      "learning_rate": 8.163617535381429e-06,
+      "loss": 0.4722,
+      "step": 2746
+    },
+    {
+      "epoch": 2.560111835973905,
+      "grad_norm": 0.9897694857227671,
+      "learning_rate": 8.146358301691405e-06,
+      "loss": 0.4478,
+      "step": 2747
+    },
+    {
+      "epoch": 2.5610438024231126,
+      "grad_norm": 0.4468471820900582,
+      "learning_rate": 8.129099068001381e-06,
+      "loss": 0.4687,
+      "step": 2748
+    },
+    {
+      "epoch": 2.561975768872321,
+      "grad_norm": 0.7796957448791411,
+      "learning_rate": 8.111839834311357e-06,
+      "loss": 0.4726,
+      "step": 2749
+    },
+    {
+      "epoch": 2.562907735321528,
+      "grad_norm": 0.6331809156842901,
+      "learning_rate": 8.094580600621333e-06,
+      "loss": 0.4784,
+      "step": 2750
+    },
+    {
+      "epoch": 2.5638397017707364,
+      "grad_norm": 0.4566789345391017,
+      "learning_rate": 8.077321366931309e-06,
+      "loss": 0.4341,
+      "step": 2751
+    },
+    {
+      "epoch": 2.564771668219944,
+      "grad_norm": 0.5044911420505094,
+      "learning_rate": 8.060062133241285e-06,
+      "loss": 0.4496,
+      "step": 2752
+    },
+    {
+      "epoch": 2.565703634669152,
+      "grad_norm": 0.4666235225245669,
+      "learning_rate": 8.04280289955126e-06,
+      "loss": 0.4634,
+      "step": 2753
+    },
+    {
+      "epoch": 2.5666356011183598,
+      "grad_norm": 0.6471465874687904,
+      "learning_rate": 8.025543665861237e-06,
+      "loss": 0.4715,
+      "step": 2754
+    },
+    {
+      "epoch": 2.5675675675675675,
+      "grad_norm": 0.3911274564540263,
+      "learning_rate": 8.008284432171211e-06,
+      "loss": 0.4319,
+      "step": 2755
+    },
+    {
+      "epoch": 2.5684995340167753,
+      "grad_norm": 0.48143941753710173,
+      "learning_rate": 7.991025198481187e-06,
+      "loss": 0.451,
+      "step": 2756
+    },
+    {
+      "epoch": 2.569431500465983,
+      "grad_norm": 0.41424181380833797,
+      "learning_rate": 7.973765964791163e-06,
+      "loss": 0.4385,
+      "step": 2757
+    },
+    {
+      "epoch": 2.570363466915191,
+      "grad_norm": 0.4317131523114312,
+      "learning_rate": 7.956506731101139e-06,
+      "loss": 0.4352,
+      "step": 2758
+    },
+    {
+      "epoch": 2.5712954333643987,
+      "grad_norm": 0.43468755532713693,
+      "learning_rate": 7.939247497411115e-06,
+      "loss": 0.4505,
+      "step": 2759
+    },
+    {
+      "epoch": 2.572227399813607,
+      "grad_norm": 0.659645981733673,
+      "learning_rate": 7.921988263721091e-06,
+      "loss": 0.5005,
+      "step": 2760
+    },
+    {
+      "epoch": 2.5731593662628143,
+      "grad_norm": 0.5362778970467418,
+      "learning_rate": 7.904729030031067e-06,
+      "loss": 0.4717,
+      "step": 2761
+    },
+    {
+      "epoch": 2.5740913327120225,
+      "grad_norm": 0.4763693860579989,
+      "learning_rate": 7.887469796341043e-06,
+      "loss": 0.451,
+      "step": 2762
+    },
+    {
+      "epoch": 2.5750232991612303,
+      "grad_norm": 0.9186431976594777,
+      "learning_rate": 7.870210562651019e-06,
+      "loss": 0.5134,
+      "step": 2763
+    },
+    {
+      "epoch": 2.575955265610438,
+      "grad_norm": 0.4431580753813944,
+      "learning_rate": 7.852951328960995e-06,
+      "loss": 0.4479,
+      "step": 2764
+    },
+    {
+      "epoch": 2.576887232059646,
+      "grad_norm": 0.601140810003059,
+      "learning_rate": 7.835692095270971e-06,
+      "loss": 0.4681,
+      "step": 2765
+    },
+    {
+      "epoch": 2.5778191985088537,
+      "grad_norm": 0.6356230230527788,
+      "learning_rate": 7.818432861580945e-06,
+      "loss": 0.4713,
+      "step": 2766
+    },
+    {
+      "epoch": 2.5787511649580614,
+      "grad_norm": 0.6492595088365869,
+      "learning_rate": 7.801173627890921e-06,
+      "loss": 0.4375,
+      "step": 2767
+    },
+    {
+      "epoch": 2.5796831314072692,
+      "grad_norm": 0.4414682782762809,
+      "learning_rate": 7.783914394200897e-06,
+      "loss": 0.446,
+      "step": 2768
+    },
+    {
+      "epoch": 2.580615097856477,
+      "grad_norm": 0.6391864125851909,
+      "learning_rate": 7.766655160510873e-06,
+      "loss": 0.4824,
+      "step": 2769
+    },
+    {
+      "epoch": 2.581547064305685,
+      "grad_norm": 0.6085334175250952,
+      "learning_rate": 7.749395926820851e-06,
+      "loss": 0.4267,
+      "step": 2770
+    },
+    {
+      "epoch": 2.582479030754893,
+      "grad_norm": 0.7815653112649558,
+      "learning_rate": 7.732136693130825e-06,
+      "loss": 0.5029,
+      "step": 2771
+    },
+    {
+      "epoch": 2.5834109972041004,
+      "grad_norm": 0.7075010104728098,
+      "learning_rate": 7.714877459440801e-06,
+      "loss": 0.4595,
+      "step": 2772
+    },
+    {
+      "epoch": 2.5843429636533086,
+      "grad_norm": 0.5847058955326776,
+      "learning_rate": 7.697618225750777e-06,
+      "loss": 0.4769,
+      "step": 2773
+    },
+    {
+      "epoch": 2.5852749301025164,
+      "grad_norm": 0.6120968579804072,
+      "learning_rate": 7.680358992060753e-06,
+      "loss": 0.4377,
+      "step": 2774
+    },
+    {
+      "epoch": 2.586206896551724,
+      "grad_norm": 0.5144060022442597,
+      "learning_rate": 7.663099758370729e-06,
+      "loss": 0.4551,
+      "step": 2775
+    },
+    {
+      "epoch": 2.587138863000932,
+      "grad_norm": 0.595904993785559,
+      "learning_rate": 7.645840524680705e-06,
+      "loss": 0.4482,
+      "step": 2776
+    },
+    {
+      "epoch": 2.5880708294501398,
+      "grad_norm": 0.5845611910998538,
+      "learning_rate": 7.62858129099068e-06,
+      "loss": 0.4565,
+      "step": 2777
+    },
+    {
+      "epoch": 2.5890027958993476,
+      "grad_norm": 0.6235843730609826,
+      "learning_rate": 7.611322057300656e-06,
+      "loss": 0.4739,
+      "step": 2778
+    },
+    {
+      "epoch": 2.5899347623485554,
+      "grad_norm": 0.5531199023343577,
+      "learning_rate": 7.594062823610631e-06,
+      "loss": 0.4636,
+      "step": 2779
+    },
+    {
+      "epoch": 2.590866728797763,
+      "grad_norm": 0.4928865200059164,
+      "learning_rate": 7.576803589920608e-06,
+      "loss": 0.4629,
+      "step": 2780
+    },
+    {
+      "epoch": 2.591798695246971,
+      "grad_norm": 0.4737144398892941,
+      "learning_rate": 7.559544356230584e-06,
+      "loss": 0.4501,
+      "step": 2781
+    },
+    {
+      "epoch": 2.592730661696179,
+      "grad_norm": 0.7376000817148254,
+      "learning_rate": 7.54228512254056e-06,
+      "loss": 0.4882,
+      "step": 2782
+    },
+    {
+      "epoch": 2.5936626281453865,
+      "grad_norm": 0.4848287207222289,
+      "learning_rate": 7.525025888850535e-06,
+      "loss": 0.4596,
+      "step": 2783
+    },
+    {
+      "epoch": 2.5945945945945947,
+      "grad_norm": 0.5178805143987473,
+      "learning_rate": 7.507766655160511e-06,
+      "loss": 0.4366,
+      "step": 2784
+    },
+    {
+      "epoch": 2.5955265610438025,
+      "grad_norm": 0.5940042526432996,
+      "learning_rate": 7.490507421470487e-06,
+      "loss": 0.4622,
+      "step": 2785
+    },
+    {
+      "epoch": 2.5964585274930103,
+      "grad_norm": 0.4623429427447852,
+      "learning_rate": 7.473248187780462e-06,
+      "loss": 0.4806,
+      "step": 2786
+    },
+    {
+      "epoch": 2.597390493942218,
+      "grad_norm": 0.49032611438312385,
+      "learning_rate": 7.455988954090438e-06,
+      "loss": 0.4604,
+      "step": 2787
+    },
+    {
+      "epoch": 2.598322460391426,
+      "grad_norm": 0.7924903323291542,
+      "learning_rate": 7.438729720400414e-06,
+      "loss": 0.4822,
+      "step": 2788
+    },
+    {
+      "epoch": 2.5992544268406337,
+      "grad_norm": 0.6077156587962159,
+      "learning_rate": 7.4214704867103895e-06,
+      "loss": 0.4639,
+      "step": 2789
+    },
+    {
+      "epoch": 2.6001863932898415,
+      "grad_norm": 0.7115929144073315,
+      "learning_rate": 7.4042112530203655e-06,
+      "loss": 0.4619,
+      "step": 2790
+    },
+    {
+      "epoch": 2.6011183597390493,
+      "grad_norm": 0.6088156733748421,
+      "learning_rate": 7.386952019330342e-06,
+      "loss": 0.4879,
+      "step": 2791
+    },
+    {
+      "epoch": 2.602050326188257,
+      "grad_norm": 0.4974076584904852,
+      "learning_rate": 7.369692785640318e-06,
+      "loss": 0.4615,
+      "step": 2792
+    },
+    {
+      "epoch": 2.6029822926374653,
+      "grad_norm": 0.5670408333455431,
+      "learning_rate": 7.352433551950294e-06,
+      "loss": 0.4639,
+      "step": 2793
+    },
+    {
+      "epoch": 2.6039142590866726,
+      "grad_norm": 0.5802055394319428,
+      "learning_rate": 7.335174318260269e-06,
+      "loss": 0.4703,
+      "step": 2794
+    },
+    {
+      "epoch": 2.604846225535881,
+      "grad_norm": 0.5472276389668972,
+      "learning_rate": 7.317915084570245e-06,
+      "loss": 0.4615,
+      "step": 2795
+    },
+    {
+      "epoch": 2.6057781919850886,
+      "grad_norm": 0.500281628953395,
+      "learning_rate": 7.300655850880221e-06,
+      "loss": 0.4302,
+      "step": 2796
+    },
+    {
+      "epoch": 2.6067101584342964,
+      "grad_norm": 0.9169538237181648,
+      "learning_rate": 7.2833966171901965e-06,
+      "loss": 0.4978,
+      "step": 2797
+    },
+    {
+      "epoch": 2.607642124883504,
+      "grad_norm": 0.6104579633155641,
+      "learning_rate": 7.2661373835001725e-06,
+      "loss": 0.4597,
+      "step": 2798
+    },
+    {
+      "epoch": 2.608574091332712,
+      "grad_norm": 0.5684384706358154,
+      "learning_rate": 7.2488781498101485e-06,
+      "loss": 0.4699,
+      "step": 2799
+    },
+    {
+      "epoch": 2.60950605778192,
+      "grad_norm": 0.7671740917379521,
+      "learning_rate": 7.231618916120124e-06,
+      "loss": 0.4926,
+      "step": 2800
+    },
+    {
+      "epoch": 2.6104380242311276,
+      "grad_norm": 0.6448547532592918,
+      "learning_rate": 7.214359682430101e-06,
+      "loss": 0.4759,
+      "step": 2801
+    },
+    {
+      "epoch": 2.6113699906803354,
+      "grad_norm": 0.7094665907577117,
+      "learning_rate": 7.1971004487400764e-06,
+      "loss": 0.4706,
+      "step": 2802
+    },
+    {
+      "epoch": 2.612301957129543,
+      "grad_norm": 0.7473967723914514,
+      "learning_rate": 7.179841215050052e-06,
+      "loss": 0.4701,
+      "step": 2803
+    },
+    {
+      "epoch": 2.6132339235787514,
+      "grad_norm": 0.5759452103438203,
+      "learning_rate": 7.162581981360028e-06,
+      "loss": 0.452,
+      "step": 2804
+    },
+    {
+      "epoch": 2.6141658900279587,
+      "grad_norm": 0.6100071371221686,
+      "learning_rate": 7.1453227476700035e-06,
+      "loss": 0.4512,
+      "step": 2805
+    },
+    {
+      "epoch": 2.615097856477167,
+      "grad_norm": 0.5225100310142425,
+      "learning_rate": 7.1280635139799795e-06,
+      "loss": 0.4671,
+      "step": 2806
+    },
+    {
+      "epoch": 2.6160298229263748,
+      "grad_norm": 0.6068029667916598,
+      "learning_rate": 7.1108042802899555e-06,
+      "loss": 0.4554,
+      "step": 2807
+    },
+    {
+      "epoch": 2.6169617893755825,
+      "grad_norm": 0.5870305635909009,
+      "learning_rate": 7.093545046599931e-06,
+      "loss": 0.4662,
+      "step": 2808
+    },
+    {
+      "epoch": 2.6178937558247903,
+      "grad_norm": 0.41552238109879236,
+      "learning_rate": 7.076285812909907e-06,
+      "loss": 0.4616,
+      "step": 2809
+    },
+    {
+      "epoch": 2.618825722273998,
+      "grad_norm": 0.5764094295667503,
+      "learning_rate": 7.059026579219883e-06,
+      "loss": 0.4389,
+      "step": 2810
+    },
+    {
+      "epoch": 2.619757688723206,
+      "grad_norm": 0.5194341901458867,
+      "learning_rate": 7.0417673455298594e-06,
+      "loss": 0.4704,
+      "step": 2811
+    },
+    {
+      "epoch": 2.6206896551724137,
+      "grad_norm": 0.671413955941483,
+      "learning_rate": 7.024508111839835e-06,
+      "loss": 0.4625,
+      "step": 2812
+    },
+    {
+      "epoch": 2.6216216216216215,
+      "grad_norm": 0.6871630389314787,
+      "learning_rate": 7.0072488781498106e-06,
+      "loss": 0.4859,
+      "step": 2813
+    },
+    {
+      "epoch": 2.6225535880708293,
+      "grad_norm": 0.49071135389686027,
+      "learning_rate": 6.9899896444597865e-06,
+      "loss": 0.4633,
+      "step": 2814
+    },
+    {
+      "epoch": 2.6234855545200375,
+      "grad_norm": 0.7648939038445447,
+      "learning_rate": 6.9727304107697625e-06,
+      "loss": 0.4478,
+      "step": 2815
+    },
+    {
+      "epoch": 2.624417520969245,
+      "grad_norm": 0.6562139447173746,
+      "learning_rate": 6.955471177079738e-06,
+      "loss": 0.4786,
+      "step": 2816
+    },
+    {
+      "epoch": 2.625349487418453,
+      "grad_norm": 0.5896461319373711,
+      "learning_rate": 6.938211943389714e-06,
+      "loss": 0.462,
+      "step": 2817
+    },
+    {
+      "epoch": 2.626281453867661,
+      "grad_norm": 0.7275444853960985,
+      "learning_rate": 6.92095270969969e-06,
+      "loss": 0.5016,
+      "step": 2818
+    },
+    {
+      "epoch": 2.6272134203168687,
+      "grad_norm": 0.4724491494872858,
+      "learning_rate": 6.903693476009665e-06,
+      "loss": 0.4615,
+      "step": 2819
+    },
+    {
+      "epoch": 2.6281453867660765,
+      "grad_norm": 0.6436236975390671,
+      "learning_rate": 6.886434242319641e-06,
+      "loss": 0.4892,
+      "step": 2820
+    },
+    {
+      "epoch": 2.6290773532152842,
+      "grad_norm": 0.4942212001008256,
+      "learning_rate": 6.869175008629618e-06,
+      "loss": 0.4682,
+      "step": 2821
+    },
+    {
+      "epoch": 2.630009319664492,
+      "grad_norm": 0.7226991190294342,
+      "learning_rate": 6.8519157749395936e-06,
+      "loss": 0.5035,
+      "step": 2822
+    },
+    {
+      "epoch": 2.6309412861137,
+      "grad_norm": 0.5079189980462787,
+      "learning_rate": 6.8346565412495696e-06,
+      "loss": 0.4551,
+      "step": 2823
+    },
+    {
+      "epoch": 2.6318732525629076,
+      "grad_norm": 0.48799837267294305,
+      "learning_rate": 6.817397307559545e-06,
+      "loss": 0.4429,
+      "step": 2824
+    },
+    {
+      "epoch": 2.6328052190121154,
+      "grad_norm": 0.9650668351957619,
+      "learning_rate": 6.800138073869521e-06,
+      "loss": 0.4839,
+      "step": 2825
+    },
+    {
+      "epoch": 2.6337371854613236,
+      "grad_norm": 0.5016444799526778,
+      "learning_rate": 6.782878840179497e-06,
+      "loss": 0.445,
+      "step": 2826
+    },
+    {
+      "epoch": 2.634669151910531,
+      "grad_norm": 0.5791714594646905,
+      "learning_rate": 6.765619606489472e-06,
+      "loss": 0.4803,
+      "step": 2827
+    },
+    {
+      "epoch": 2.635601118359739,
+      "grad_norm": 0.8290959610240466,
+      "learning_rate": 6.748360372799448e-06,
+      "loss": 0.4402,
+      "step": 2828
+    },
+    {
+      "epoch": 2.636533084808947,
+      "grad_norm": 0.5216277643493261,
+      "learning_rate": 6.731101139109424e-06,
+      "loss": 0.4684,
+      "step": 2829
+    },
+    {
+      "epoch": 2.637465051258155,
+      "grad_norm": 0.6561461740101371,
+      "learning_rate": 6.713841905419399e-06,
+      "loss": 0.4662,
+      "step": 2830
+    },
+    {
+      "epoch": 2.6383970177073626,
+      "grad_norm": 0.47852078227102823,
+      "learning_rate": 6.696582671729376e-06,
+      "loss": 0.4461,
+      "step": 2831
+    },
+    {
+      "epoch": 2.6393289841565704,
+      "grad_norm": 0.544896743723724,
+      "learning_rate": 6.679323438039352e-06,
+      "loss": 0.4426,
+      "step": 2832
+    },
+    {
+      "epoch": 2.640260950605778,
+      "grad_norm": 0.6623425381073896,
+      "learning_rate": 6.662064204349328e-06,
+      "loss": 0.4693,
+      "step": 2833
+    },
+    {
+      "epoch": 2.641192917054986,
+      "grad_norm": 0.5661389973848081,
+      "learning_rate": 6.644804970659303e-06,
+      "loss": 0.4582,
+      "step": 2834
+    },
+    {
+      "epoch": 2.6421248835041937,
+      "grad_norm": 0.7493509365195714,
+      "learning_rate": 6.627545736969279e-06,
+      "loss": 0.4999,
+      "step": 2835
+    },
+    {
+      "epoch": 2.6430568499534015,
+      "grad_norm": 0.7777086833900426,
+      "learning_rate": 6.610286503279255e-06,
+      "loss": 0.464,
+      "step": 2836
+    },
+    {
+      "epoch": 2.6439888164026097,
+      "grad_norm": 0.550974388688788,
+      "learning_rate": 6.59302726958923e-06,
+      "loss": 0.4493,
+      "step": 2837
+    },
+    {
+      "epoch": 2.644920782851817,
+      "grad_norm": 0.5322626714711487,
+      "learning_rate": 6.575768035899206e-06,
+      "loss": 0.4875,
+      "step": 2838
+    },
+    {
+      "epoch": 2.6458527493010253,
+      "grad_norm": 0.7441865138322257,
+      "learning_rate": 6.558508802209182e-06,
+      "loss": 0.4603,
+      "step": 2839
+    },
+    {
+      "epoch": 2.646784715750233,
+      "grad_norm": 0.7150608044914993,
+      "learning_rate": 6.541249568519157e-06,
+      "loss": 0.4728,
+      "step": 2840
+    },
+    {
+      "epoch": 2.647716682199441,
+      "grad_norm": 0.5664238376611254,
+      "learning_rate": 6.523990334829135e-06,
+      "loss": 0.4441,
+      "step": 2841
+    },
+    {
+      "epoch": 2.6486486486486487,
+      "grad_norm": 0.5257648863778882,
+      "learning_rate": 6.50673110113911e-06,
+      "loss": 0.4603,
+      "step": 2842
+    },
+    {
+      "epoch": 2.6495806150978565,
+      "grad_norm": 0.5234722322677103,
+      "learning_rate": 6.489471867449086e-06,
+      "loss": 0.4624,
+      "step": 2843
+    },
+    {
+      "epoch": 2.6505125815470643,
+      "grad_norm": 0.5130326208978339,
+      "learning_rate": 6.472212633759062e-06,
+      "loss": 0.4595,
+      "step": 2844
+    },
+    {
+      "epoch": 2.651444547996272,
+      "grad_norm": 0.5355494833687672,
+      "learning_rate": 6.454953400069037e-06,
+      "loss": 0.4627,
+      "step": 2845
+    },
+    {
+      "epoch": 2.65237651444548,
+      "grad_norm": 0.41870194942122,
+      "learning_rate": 6.437694166379013e-06,
+      "loss": 0.4357,
+      "step": 2846
+    },
+    {
+      "epoch": 2.6533084808946876,
+      "grad_norm": 0.3835944871017278,
+      "learning_rate": 6.420434932688989e-06,
+      "loss": 0.4635,
+      "step": 2847
+    },
+    {
+      "epoch": 2.654240447343896,
+      "grad_norm": 0.8017453750969926,
+      "learning_rate": 6.403175698998964e-06,
+      "loss": 0.4889,
+      "step": 2848
+    },
+    {
+      "epoch": 2.655172413793103,
+      "grad_norm": 0.6212068924195258,
+      "learning_rate": 6.38591646530894e-06,
+      "loss": 0.4534,
+      "step": 2849
+    },
+    {
+      "epoch": 2.6561043802423114,
+      "grad_norm": 0.6886306243763836,
+      "learning_rate": 6.368657231618916e-06,
+      "loss": 0.5013,
+      "step": 2850
+    },
+    {
+      "epoch": 2.6570363466915192,
+      "grad_norm": 0.7745879698847642,
+      "learning_rate": 6.351397997928893e-06,
+      "loss": 0.468,
+      "step": 2851
+    },
+    {
+      "epoch": 2.657968313140727,
+      "grad_norm": 0.5800031025060908,
+      "learning_rate": 6.334138764238869e-06,
+      "loss": 0.4589,
+      "step": 2852
+    },
+    {
+      "epoch": 2.658900279589935,
+      "grad_norm": 0.6894369010947564,
+      "learning_rate": 6.316879530548844e-06,
+      "loss": 0.4806,
+      "step": 2853
+    },
+    {
+      "epoch": 2.6598322460391426,
+      "grad_norm": 0.47668189807647093,
+      "learning_rate": 6.29962029685882e-06,
+      "loss": 0.447,
+      "step": 2854
+    },
+    {
+      "epoch": 2.6607642124883504,
+      "grad_norm": 0.55049455977286,
+      "learning_rate": 6.282361063168796e-06,
+      "loss": 0.477,
+      "step": 2855
+    },
+    {
+      "epoch": 2.661696178937558,
+      "grad_norm": 0.4586225229634722,
+      "learning_rate": 6.265101829478771e-06,
+      "loss": 0.461,
+      "step": 2856
+    },
+    {
+      "epoch": 2.662628145386766,
+      "grad_norm": 0.4782471183757928,
+      "learning_rate": 6.247842595788747e-06,
+      "loss": 0.4743,
+      "step": 2857
+    },
+    {
+      "epoch": 2.6635601118359737,
+      "grad_norm": 0.47874685789374055,
+      "learning_rate": 6.230583362098723e-06,
+      "loss": 0.4408,
+      "step": 2858
+    },
+    {
+      "epoch": 2.664492078285182,
+      "grad_norm": 0.7988690000379604,
+      "learning_rate": 6.213324128408699e-06,
+      "loss": 0.4717,
+      "step": 2859
+    },
+    {
+      "epoch": 2.6654240447343893,
+      "grad_norm": 0.5546974622033471,
+      "learning_rate": 6.196064894718675e-06,
+      "loss": 0.4966,
+      "step": 2860
+    },
+    {
+      "epoch": 2.6663560111835976,
+      "grad_norm": 0.6150267134545601,
+      "learning_rate": 6.17880566102865e-06,
+      "loss": 0.4585,
+      "step": 2861
+    },
+    {
+      "epoch": 2.6672879776328053,
+      "grad_norm": 0.777674358272175,
+      "learning_rate": 6.161546427338626e-06,
+      "loss": 0.502,
+      "step": 2862
+    },
+    {
+      "epoch": 2.668219944082013,
+      "grad_norm": 0.569957095232017,
+      "learning_rate": 6.144287193648602e-06,
+      "loss": 0.4524,
+      "step": 2863
+    },
+    {
+      "epoch": 2.669151910531221,
+      "grad_norm": 0.5344545415323133,
+      "learning_rate": 6.127027959958578e-06,
+      "loss": 0.4513,
+      "step": 2864
+    },
+    {
+      "epoch": 2.6700838769804287,
+      "grad_norm": 0.6193465704327614,
+      "learning_rate": 6.109768726268554e-06,
+      "loss": 0.4947,
+      "step": 2865
+    },
+    {
+      "epoch": 2.6710158434296365,
+      "grad_norm": 0.6179082193568326,
+      "learning_rate": 6.09250949257853e-06,
+      "loss": 0.4694,
+      "step": 2866
+    },
+    {
+      "epoch": 2.6719478098788443,
+      "grad_norm": 0.4674066276509293,
+      "learning_rate": 6.075250258888505e-06,
+      "loss": 0.4633,
+      "step": 2867
+    },
+    {
+      "epoch": 2.672879776328052,
+      "grad_norm": 0.5876079272473999,
+      "learning_rate": 6.057991025198481e-06,
+      "loss": 0.4743,
+      "step": 2868
+    },
+    {
+      "epoch": 2.67381174277726,
+      "grad_norm": 0.47493635580817156,
+      "learning_rate": 6.040731791508457e-06,
+      "loss": 0.4334,
+      "step": 2869
+    },
+    {
+      "epoch": 2.674743709226468,
+      "grad_norm": 0.43317193708665636,
+      "learning_rate": 6.023472557818433e-06,
+      "loss": 0.4262,
+      "step": 2870
+    },
+    {
+      "epoch": 2.6756756756756754,
+      "grad_norm": 0.5387471731795906,
+      "learning_rate": 6.006213324128409e-06,
+      "loss": 0.4829,
+      "step": 2871
+    },
+    {
+      "epoch": 2.6766076421248837,
+      "grad_norm": 1.128555113892862,
+      "learning_rate": 5.988954090438384e-06,
+      "loss": 0.5098,
+      "step": 2872
+    },
+    {
+      "epoch": 2.6775396085740915,
+      "grad_norm": 0.6971500429401541,
+      "learning_rate": 5.97169485674836e-06,
+      "loss": 0.4646,
+      "step": 2873
+    },
+    {
+      "epoch": 2.6784715750232992,
+      "grad_norm": 0.33957951179736956,
+      "learning_rate": 5.954435623058337e-06,
+      "loss": 0.4535,
+      "step": 2874
+    },
+    {
+      "epoch": 2.679403541472507,
+      "grad_norm": 0.5440251971618096,
+      "learning_rate": 5.937176389368312e-06,
+      "loss": 0.5091,
+      "step": 2875
+    },
+    {
+      "epoch": 2.680335507921715,
+      "grad_norm": 0.7419548982172388,
+      "learning_rate": 5.919917155678288e-06,
+      "loss": 0.4886,
+      "step": 2876
+    },
+    {
+      "epoch": 2.6812674743709226,
+      "grad_norm": 0.509404795360994,
+      "learning_rate": 5.902657921988264e-06,
+      "loss": 0.4654,
+      "step": 2877
+    },
+    {
+      "epoch": 2.6821994408201304,
+      "grad_norm": 0.7673490267143867,
+      "learning_rate": 5.885398688298239e-06,
+      "loss": 0.4917,
+      "step": 2878
+    },
+    {
+      "epoch": 2.683131407269338,
+      "grad_norm": 0.46463767834612374,
+      "learning_rate": 5.868139454608216e-06,
+      "loss": 0.4307,
+      "step": 2879
+    },
+    {
+      "epoch": 2.684063373718546,
+      "grad_norm": 0.48474713450771834,
+      "learning_rate": 5.850880220918191e-06,
+      "loss": 0.4692,
+      "step": 2880
+    },
+    {
+      "epoch": 2.684995340167754,
+      "grad_norm": 0.5591452369716908,
+      "learning_rate": 5.833620987228167e-06,
+      "loss": 0.4307,
+      "step": 2881
+    },
+    {
+      "epoch": 2.6859273066169616,
+      "grad_norm": 0.3799909398529947,
+      "learning_rate": 5.816361753538143e-06,
+      "loss": 0.4299,
+      "step": 2882
+    },
+    {
+      "epoch": 2.68685927306617,
+      "grad_norm": 0.4904750138381915,
+      "learning_rate": 5.7991025198481184e-06,
+      "loss": 0.4471,
+      "step": 2883
+    },
+    {
+      "epoch": 2.6877912395153776,
+      "grad_norm": 0.5503602619846519,
+      "learning_rate": 5.781843286158095e-06,
+      "loss": 0.4721,
+      "step": 2884
+    },
+    {
+      "epoch": 2.6887232059645854,
+      "grad_norm": 0.6120346016515286,
+      "learning_rate": 5.76458405246807e-06,
+      "loss": 0.4808,
+      "step": 2885
+    },
+    {
+      "epoch": 2.689655172413793,
+      "grad_norm": 0.5102086195372048,
+      "learning_rate": 5.747324818778046e-06,
+      "loss": 0.4648,
+      "step": 2886
+    },
+    {
+      "epoch": 2.690587138863001,
+      "grad_norm": 0.6131362706017909,
+      "learning_rate": 5.730065585088022e-06,
+      "loss": 0.4742,
+      "step": 2887
+    },
+    {
+      "epoch": 2.6915191053122087,
+      "grad_norm": 0.5730080623664798,
+      "learning_rate": 5.7128063513979975e-06,
+      "loss": 0.4664,
+      "step": 2888
+    },
+    {
+      "epoch": 2.6924510717614165,
+      "grad_norm": 0.5539556444748182,
+      "learning_rate": 5.695547117707974e-06,
+      "loss": 0.4608,
+      "step": 2889
+    },
+    {
+      "epoch": 2.6933830382106243,
+      "grad_norm": 0.4042165003805518,
+      "learning_rate": 5.67828788401795e-06,
+      "loss": 0.4328,
+      "step": 2890
+    },
+    {
+      "epoch": 2.694315004659832,
+      "grad_norm": 0.7288746272034748,
+      "learning_rate": 5.6610286503279255e-06,
+      "loss": 0.4801,
+      "step": 2891
+    },
+    {
+      "epoch": 2.6952469711090403,
+      "grad_norm": 0.753249296390701,
+      "learning_rate": 5.6437694166379015e-06,
+      "loss": 0.4723,
+      "step": 2892
+    },
+    {
+      "epoch": 2.6961789375582477,
+      "grad_norm": 0.5329254905422398,
+      "learning_rate": 5.6265101829478774e-06,
+      "loss": 0.4645,
+      "step": 2893
+    },
+    {
+      "epoch": 2.697110904007456,
+      "grad_norm": 0.43856047497795825,
+      "learning_rate": 5.609250949257853e-06,
+      "loss": 0.4645,
+      "step": 2894
+    },
+    {
+      "epoch": 2.6980428704566637,
+      "grad_norm": 0.5386527312187392,
+      "learning_rate": 5.591991715567829e-06,
+      "loss": 0.4603,
+      "step": 2895
+    },
+    {
+      "epoch": 2.6989748369058715,
+      "grad_norm": 0.5913107238344869,
+      "learning_rate": 5.5747324818778045e-06,
+      "loss": 0.4407,
+      "step": 2896
+    },
+    {
+      "epoch": 2.6999068033550793,
+      "grad_norm": 0.5315147078648054,
+      "learning_rate": 5.5574732481877805e-06,
+      "loss": 0.4434,
+      "step": 2897
+    },
+    {
+      "epoch": 2.700838769804287,
+      "grad_norm": 0.5085313953505588,
+      "learning_rate": 5.5402140144977565e-06,
+      "loss": 0.4513,
+      "step": 2898
+    },
+    {
+      "epoch": 2.701770736253495,
+      "grad_norm": 0.40141021411641664,
+      "learning_rate": 5.5229547808077325e-06,
+      "loss": 0.4314,
+      "step": 2899
+    },
+    {
+      "epoch": 2.7027027027027026,
+      "grad_norm": 0.47738129956386555,
+      "learning_rate": 5.5056955471177085e-06,
+      "loss": 0.4516,
+      "step": 2900
+    },
+    {
+      "epoch": 2.7036346691519104,
+      "grad_norm": 0.35338138863279345,
+      "learning_rate": 5.4884363134276845e-06,
+      "loss": 0.4287,
+      "step": 2901
+    },
+    {
+      "epoch": 2.704566635601118,
+      "grad_norm": 0.47480084703496284,
+      "learning_rate": 5.47117707973766e-06,
+      "loss": 0.4384,
+      "step": 2902
+    },
+    {
+      "epoch": 2.7054986020503264,
+      "grad_norm": 0.5224188142229801,
+      "learning_rate": 5.453917846047636e-06,
+      "loss": 0.4562,
+      "step": 2903
+    },
+    {
+      "epoch": 2.706430568499534,
+      "grad_norm": 0.6178210856857508,
+      "learning_rate": 5.4366586123576116e-06,
+      "loss": 0.4677,
+      "step": 2904
+    },
+    {
+      "epoch": 2.707362534948742,
+      "grad_norm": 0.39978104258538283,
+      "learning_rate": 5.4193993786675876e-06,
+      "loss": 0.4523,
+      "step": 2905
+    },
+    {
+      "epoch": 2.70829450139795,
+      "grad_norm": 0.4772807696925502,
+      "learning_rate": 5.4021401449775635e-06,
+      "loss": 0.5016,
+      "step": 2906
+    },
+    {
+      "epoch": 2.7092264678471576,
+      "grad_norm": 0.7176369900040714,
+      "learning_rate": 5.384880911287539e-06,
+      "loss": 0.4945,
+      "step": 2907
+    },
+    {
+      "epoch": 2.7101584342963654,
+      "grad_norm": 0.5027760016807082,
+      "learning_rate": 5.367621677597515e-06,
+      "loss": 0.436,
+      "step": 2908
+    },
+    {
+      "epoch": 2.711090400745573,
+      "grad_norm": 0.4583265525486921,
+      "learning_rate": 5.350362443907491e-06,
+      "loss": 0.4455,
+      "step": 2909
+    },
+    {
+      "epoch": 2.712022367194781,
+      "grad_norm": 0.5147430687096223,
+      "learning_rate": 5.333103210217467e-06,
+      "loss": 0.4775,
+      "step": 2910
+    },
+    {
+      "epoch": 2.7129543336439887,
+      "grad_norm": 0.44460283926056515,
+      "learning_rate": 5.315843976527443e-06,
+      "loss": 0.4461,
+      "step": 2911
+    },
+    {
+      "epoch": 2.7138863000931965,
+      "grad_norm": 0.5614531510963195,
+      "learning_rate": 5.298584742837418e-06,
+      "loss": 0.4771,
+      "step": 2912
+    },
+    {
+      "epoch": 2.7148182665424043,
+      "grad_norm": 0.6346464066295525,
+      "learning_rate": 5.281325509147394e-06,
+      "loss": 0.4981,
+      "step": 2913
+    },
+    {
+      "epoch": 2.7157502329916126,
+      "grad_norm": 0.6655794236366804,
+      "learning_rate": 5.2640662754573706e-06,
+      "loss": 0.4414,
+      "step": 2914
+    },
+    {
+      "epoch": 2.71668219944082,
+      "grad_norm": 0.7900329002290922,
+      "learning_rate": 5.246807041767346e-06,
+      "loss": 0.4708,
+      "step": 2915
+    },
+    {
+      "epoch": 2.717614165890028,
+      "grad_norm": 0.5082586744552482,
+      "learning_rate": 5.229547808077322e-06,
+      "loss": 0.4533,
+      "step": 2916
+    },
+    {
+      "epoch": 2.718546132339236,
+      "grad_norm": 0.40888594376667986,
+      "learning_rate": 5.212288574387298e-06,
+      "loss": 0.4311,
+      "step": 2917
+    },
+    {
+      "epoch": 2.7194780987884437,
+      "grad_norm": 0.4604672282168814,
+      "learning_rate": 5.195029340697273e-06,
+      "loss": 0.4513,
+      "step": 2918
+    },
+    {
+      "epoch": 2.7204100652376515,
+      "grad_norm": 0.5390106273072446,
+      "learning_rate": 5.17777010700725e-06,
+      "loss": 0.4771,
+      "step": 2919
+    },
+    {
+      "epoch": 2.7213420316868593,
+      "grad_norm": 0.5357472260907231,
+      "learning_rate": 5.160510873317225e-06,
+      "loss": 0.4448,
+      "step": 2920
+    },
+    {
+      "epoch": 2.722273998136067,
+      "grad_norm": 0.7506995010291565,
+      "learning_rate": 5.143251639627201e-06,
+      "loss": 0.4986,
+      "step": 2921
+    },
+    {
+      "epoch": 2.723205964585275,
+      "grad_norm": 0.4025396059934461,
+      "learning_rate": 5.125992405937177e-06,
+      "loss": 0.4483,
+      "step": 2922
+    },
+    {
+      "epoch": 2.7241379310344827,
+      "grad_norm": 0.46324350873942055,
+      "learning_rate": 5.108733172247152e-06,
+      "loss": 0.4238,
+      "step": 2923
+    },
+    {
+      "epoch": 2.7250698974836904,
+      "grad_norm": 0.45384528382434475,
+      "learning_rate": 5.091473938557129e-06,
+      "loss": 0.4542,
+      "step": 2924
+    },
+    {
+      "epoch": 2.7260018639328987,
+      "grad_norm": 0.49038419391946697,
+      "learning_rate": 5.074214704867105e-06,
+      "loss": 0.4483,
+      "step": 2925
+    },
+    {
+      "epoch": 2.726933830382106,
+      "grad_norm": 0.6349655621536351,
+      "learning_rate": 5.05695547117708e-06,
+      "loss": 0.4786,
+      "step": 2926
+    },
+    {
+      "epoch": 2.7278657968313142,
+      "grad_norm": 0.5353874990771713,
+      "learning_rate": 5.039696237487056e-06,
+      "loss": 0.4807,
+      "step": 2927
+    },
+    {
+      "epoch": 2.728797763280522,
+      "grad_norm": 0.33713748918170566,
+      "learning_rate": 5.022437003797032e-06,
+      "loss": 0.4186,
+      "step": 2928
+    },
+    {
+      "epoch": 2.72972972972973,
+      "grad_norm": 0.6759083986578773,
+      "learning_rate": 5.005177770107008e-06,
+      "loss": 0.479,
+      "step": 2929
+    },
+    {
+      "epoch": 2.7306616961789376,
+      "grad_norm": 0.8180551725330117,
+      "learning_rate": 4.987918536416984e-06,
+      "loss": 0.4778,
+      "step": 2930
+    },
+    {
+      "epoch": 2.7315936626281454,
+      "grad_norm": 0.40302650585152533,
+      "learning_rate": 4.970659302726959e-06,
+      "loss": 0.43,
+      "step": 2931
+    },
+    {
+      "epoch": 2.732525629077353,
+      "grad_norm": 0.6509983234367867,
+      "learning_rate": 4.953400069036935e-06,
+      "loss": 0.4691,
+      "step": 2932
+    },
+    {
+      "epoch": 2.733457595526561,
+      "grad_norm": 0.5397392291217326,
+      "learning_rate": 4.936140835346911e-06,
+      "loss": 0.4563,
+      "step": 2933
+    },
+    {
+      "epoch": 2.7343895619757688,
+      "grad_norm": 0.5485464007892952,
+      "learning_rate": 4.918881601656887e-06,
+      "loss": 0.4686,
+      "step": 2934
+    },
+    {
+      "epoch": 2.7353215284249766,
+      "grad_norm": 0.5062794315939673,
+      "learning_rate": 4.901622367966863e-06,
+      "loss": 0.4354,
+      "step": 2935
+    },
+    {
+      "epoch": 2.736253494874185,
+      "grad_norm": 0.6088270317539726,
+      "learning_rate": 4.884363134276838e-06,
+      "loss": 0.4733,
+      "step": 2936
+    },
+    {
+      "epoch": 2.737185461323392,
+      "grad_norm": 0.8596031675828355,
+      "learning_rate": 4.867103900586814e-06,
+      "loss": 0.5136,
+      "step": 2937
+    },
+    {
+      "epoch": 2.7381174277726004,
+      "grad_norm": 0.5139544619913953,
+      "learning_rate": 4.84984466689679e-06,
+      "loss": 0.4586,
+      "step": 2938
+    },
+    {
+      "epoch": 2.739049394221808,
+      "grad_norm": 0.5074228107646732,
+      "learning_rate": 4.832585433206766e-06,
+      "loss": 0.4518,
+      "step": 2939
+    },
+    {
+      "epoch": 2.739981360671016,
+      "grad_norm": 0.6163699506184281,
+      "learning_rate": 4.815326199516742e-06,
+      "loss": 0.4704,
+      "step": 2940
+    },
+    {
+      "epoch": 2.7409133271202237,
+      "grad_norm": 0.4467830550812222,
+      "learning_rate": 4.798066965826718e-06,
+      "loss": 0.4473,
+      "step": 2941
+    },
+    {
+      "epoch": 2.7418452935694315,
+      "grad_norm": 0.4660374036754048,
+      "learning_rate": 4.780807732136693e-06,
+      "loss": 0.4719,
+      "step": 2942
+    },
+    {
+      "epoch": 2.7427772600186393,
+      "grad_norm": 0.3786510460493873,
+      "learning_rate": 4.763548498446669e-06,
+      "loss": 0.4271,
+      "step": 2943
+    },
+    {
+      "epoch": 2.743709226467847,
+      "grad_norm": 0.5968217629560336,
+      "learning_rate": 4.746289264756645e-06,
+      "loss": 0.4548,
+      "step": 2944
+    },
+    {
+      "epoch": 2.744641192917055,
+      "grad_norm": 0.393343045764733,
+      "learning_rate": 4.729030031066621e-06,
+      "loss": 0.4288,
+      "step": 2945
+    },
+    {
+      "epoch": 2.7455731593662627,
+      "grad_norm": 0.51507227838259,
+      "learning_rate": 4.711770797376597e-06,
+      "loss": 0.4706,
+      "step": 2946
+    },
+    {
+      "epoch": 2.746505125815471,
+      "grad_norm": 0.5737977475728446,
+      "learning_rate": 4.694511563686572e-06,
+      "loss": 0.4621,
+      "step": 2947
+    },
+    {
+      "epoch": 2.7474370922646782,
+      "grad_norm": 0.6652467132170027,
+      "learning_rate": 4.677252329996548e-06,
+      "loss": 0.4665,
+      "step": 2948
+    },
+    {
+      "epoch": 2.7483690587138865,
+      "grad_norm": 0.37630220654015956,
+      "learning_rate": 4.659993096306525e-06,
+      "loss": 0.4558,
+      "step": 2949
+    },
+    {
+      "epoch": 2.7493010251630943,
+      "grad_norm": 0.45553558172222136,
+      "learning_rate": 4.6427338626165e-06,
+      "loss": 0.4658,
+      "step": 2950
+    },
+    {
+      "epoch": 2.750232991612302,
+      "grad_norm": 0.35632130411795526,
+      "learning_rate": 4.625474628926476e-06,
+      "loss": 0.4257,
+      "step": 2951
+    },
+    {
+      "epoch": 2.75116495806151,
+      "grad_norm": 0.3141578767871785,
+      "learning_rate": 4.608215395236452e-06,
+      "loss": 0.4139,
+      "step": 2952
+    },
+    {
+      "epoch": 2.7520969245107176,
+      "grad_norm": 0.7464816256455357,
+      "learning_rate": 4.590956161546427e-06,
+      "loss": 0.4965,
+      "step": 2953
+    },
+    {
+      "epoch": 2.7530288909599254,
+      "grad_norm": 0.6956665401682156,
+      "learning_rate": 4.573696927856403e-06,
+      "loss": 0.4783,
+      "step": 2954
+    },
+    {
+      "epoch": 2.753960857409133,
+      "grad_norm": 0.5436606293966075,
+      "learning_rate": 4.556437694166379e-06,
+      "loss": 0.4391,
+      "step": 2955
+    },
+    {
+      "epoch": 2.754892823858341,
+      "grad_norm": 0.7079658350500012,
+      "learning_rate": 4.539178460476355e-06,
+      "loss": 0.4787,
+      "step": 2956
+    },
+    {
+      "epoch": 2.755824790307549,
+      "grad_norm": 0.3831913196572551,
+      "learning_rate": 4.521919226786331e-06,
+      "loss": 0.4333,
+      "step": 2957
+    },
+    {
+      "epoch": 2.756756756756757,
+      "grad_norm": 0.5629224987767827,
+      "learning_rate": 4.504659993096306e-06,
+      "loss": 0.4663,
+      "step": 2958
+    },
+    {
+      "epoch": 2.7576887232059644,
+      "grad_norm": 0.40302644817236977,
+      "learning_rate": 4.487400759406282e-06,
+      "loss": 0.4483,
+      "step": 2959
+    },
+    {
+      "epoch": 2.7586206896551726,
+      "grad_norm": 0.6969032051452534,
+      "learning_rate": 4.470141525716258e-06,
+      "loss": 0.4536,
+      "step": 2960
+    },
+    {
+      "epoch": 2.7595526561043804,
+      "grad_norm": 0.8182990059044091,
+      "learning_rate": 4.452882292026234e-06,
+      "loss": 0.5053,
+      "step": 2961
+    },
+    {
+      "epoch": 2.760484622553588,
+      "grad_norm": 0.7249961395317606,
+      "learning_rate": 4.43562305833621e-06,
+      "loss": 0.4849,
+      "step": 2962
+    },
+    {
+      "epoch": 2.761416589002796,
+      "grad_norm": 0.38939871600364756,
+      "learning_rate": 4.418363824646185e-06,
+      "loss": 0.4368,
+      "step": 2963
+    },
+    {
+      "epoch": 2.7623485554520038,
+      "grad_norm": 0.5141002527444304,
+      "learning_rate": 4.401104590956161e-06,
+      "loss": 0.4184,
+      "step": 2964
+    },
+    {
+      "epoch": 2.7632805219012115,
+      "grad_norm": 0.7749447446775217,
+      "learning_rate": 4.383845357266138e-06,
+      "loss": 0.4785,
+      "step": 2965
+    },
+    {
+      "epoch": 2.7642124883504193,
+      "grad_norm": 0.5185648744188299,
+      "learning_rate": 4.366586123576113e-06,
+      "loss": 0.4628,
+      "step": 2966
+    },
+    {
+      "epoch": 2.765144454799627,
+      "grad_norm": 0.4356542600031992,
+      "learning_rate": 4.349326889886089e-06,
+      "loss": 0.4277,
+      "step": 2967
+    },
+    {
+      "epoch": 2.766076421248835,
+      "grad_norm": 0.5080646195211846,
+      "learning_rate": 4.332067656196065e-06,
+      "loss": 0.4657,
+      "step": 2968
+    },
+    {
+      "epoch": 2.767008387698043,
+      "grad_norm": 0.5474107405743946,
+      "learning_rate": 4.31480842250604e-06,
+      "loss": 0.4804,
+      "step": 2969
+    },
+    {
+      "epoch": 2.7679403541472505,
+      "grad_norm": 0.4785139945891191,
+      "learning_rate": 4.297549188816017e-06,
+      "loss": 0.454,
+      "step": 2970
+    },
+    {
+      "epoch": 2.7688723205964587,
+      "grad_norm": 0.5387537839515654,
+      "learning_rate": 4.280289955125992e-06,
+      "loss": 0.4596,
+      "step": 2971
+    },
+    {
+      "epoch": 2.7698042870456665,
+      "grad_norm": 0.5087919811467633,
+      "learning_rate": 4.263030721435968e-06,
+      "loss": 0.4583,
+      "step": 2972
+    },
+    {
+      "epoch": 2.7707362534948743,
+      "grad_norm": 0.4535307642156473,
+      "learning_rate": 4.245771487745944e-06,
+      "loss": 0.4457,
+      "step": 2973
+    },
+    {
+      "epoch": 2.771668219944082,
+      "grad_norm": 0.49635034973274533,
+      "learning_rate": 4.2285122540559194e-06,
+      "loss": 0.4638,
+      "step": 2974
+    },
+    {
+      "epoch": 2.77260018639329,
+      "grad_norm": 0.4547019670743561,
+      "learning_rate": 4.211253020365896e-06,
+      "loss": 0.4512,
+      "step": 2975
+    },
+    {
+      "epoch": 2.7735321528424977,
+      "grad_norm": 0.5541756410936961,
+      "learning_rate": 4.193993786675872e-06,
+      "loss": 0.4682,
+      "step": 2976
+    },
+    {
+      "epoch": 2.7744641192917054,
+      "grad_norm": 0.5265586392549226,
+      "learning_rate": 4.176734552985847e-06,
+      "loss": 0.4544,
+      "step": 2977
+    },
+    {
+      "epoch": 2.7753960857409132,
+      "grad_norm": 0.6748012933743899,
+      "learning_rate": 4.159475319295823e-06,
+      "loss": 0.4726,
+      "step": 2978
+    },
+    {
+      "epoch": 2.776328052190121,
+      "grad_norm": 0.3392209757701532,
+      "learning_rate": 4.142216085605799e-06,
+      "loss": 0.4356,
+      "step": 2979
+    },
+    {
+      "epoch": 2.7772600186393293,
+      "grad_norm": 0.8135256297453759,
+      "learning_rate": 4.124956851915775e-06,
+      "loss": 0.4606,
+      "step": 2980
+    },
+    {
+      "epoch": 2.7781919850885366,
+      "grad_norm": 0.6196437396146822,
+      "learning_rate": 4.107697618225751e-06,
+      "loss": 0.4491,
+      "step": 2981
+    },
+    {
+      "epoch": 2.779123951537745,
+      "grad_norm": 0.5159308995426249,
+      "learning_rate": 4.0904383845357265e-06,
+      "loss": 0.4997,
+      "step": 2982
+    },
+    {
+      "epoch": 2.7800559179869526,
+      "grad_norm": 0.5336249397939358,
+      "learning_rate": 4.0731791508457025e-06,
+      "loss": 0.447,
+      "step": 2983
+    },
+    {
+      "epoch": 2.7809878844361604,
+      "grad_norm": 0.39322086956698055,
+      "learning_rate": 4.0559199171556784e-06,
+      "loss": 0.4561,
+      "step": 2984
+    },
+    {
+      "epoch": 2.781919850885368,
+      "grad_norm": 0.5851189829630689,
+      "learning_rate": 4.0386606834656544e-06,
+      "loss": 0.477,
+      "step": 2985
+    },
+    {
+      "epoch": 2.782851817334576,
+      "grad_norm": 0.8434465363118101,
+      "learning_rate": 4.02140144977563e-06,
+      "loss": 0.4939,
+      "step": 2986
+    },
+    {
+      "epoch": 2.7837837837837838,
+      "grad_norm": 0.5014810873275847,
+      "learning_rate": 4.0041422160856055e-06,
+      "loss": 0.466,
+      "step": 2987
+    },
+    {
+      "epoch": 2.7847157502329916,
+      "grad_norm": 0.7238366373911529,
+      "learning_rate": 3.9868829823955815e-06,
+      "loss": 0.4461,
+      "step": 2988
+    },
+    {
+      "epoch": 2.7856477166821993,
+      "grad_norm": 0.8011162508076747,
+      "learning_rate": 3.9696237487055575e-06,
+      "loss": 0.4752,
+      "step": 2989
+    },
+    {
+      "epoch": 2.786579683131407,
+      "grad_norm": 0.4776349014475215,
+      "learning_rate": 3.9523645150155335e-06,
+      "loss": 0.4465,
+      "step": 2990
+    },
+    {
+      "epoch": 2.7875116495806154,
+      "grad_norm": 0.5694009465625394,
+      "learning_rate": 3.9351052813255095e-06,
+      "loss": 0.4476,
+      "step": 2991
+    },
+    {
+      "epoch": 2.7884436160298227,
+      "grad_norm": 0.3636957817365545,
+      "learning_rate": 3.9178460476354855e-06,
+      "loss": 0.4171,
+      "step": 2992
+    },
+    {
+      "epoch": 2.789375582479031,
+      "grad_norm": 0.893127340490257,
+      "learning_rate": 3.900586813945461e-06,
+      "loss": 0.4761,
+      "step": 2993
+    },
+    {
+      "epoch": 2.7903075489282387,
+      "grad_norm": 0.374772605049121,
+      "learning_rate": 3.883327580255437e-06,
+      "loss": 0.4489,
+      "step": 2994
+    },
+    {
+      "epoch": 2.7912395153774465,
+      "grad_norm": 0.48489454916836455,
+      "learning_rate": 3.8660683465654126e-06,
+      "loss": 0.4448,
+      "step": 2995
+    },
+    {
+      "epoch": 2.7921714818266543,
+      "grad_norm": 0.4588613812928874,
+      "learning_rate": 3.8488091128753886e-06,
+      "loss": 0.4567,
+      "step": 2996
+    },
+    {
+      "epoch": 2.793103448275862,
+      "grad_norm": 0.6433198593960924,
+      "learning_rate": 3.8315498791853645e-06,
+      "loss": 0.4555,
+      "step": 2997
+    },
+    {
+      "epoch": 2.79403541472507,
+      "grad_norm": 0.881408908344404,
+      "learning_rate": 3.81429064549534e-06,
+      "loss": 0.4836,
+      "step": 2998
+    },
+    {
+      "epoch": 2.7949673811742777,
+      "grad_norm": 0.669888860939141,
+      "learning_rate": 3.7970314118053157e-06,
+      "loss": 0.4674,
+      "step": 2999
+    },
+    {
+      "epoch": 2.7958993476234855,
+      "grad_norm": 0.6973804264252439,
+      "learning_rate": 3.779772178115292e-06,
+      "loss": 0.421,
+      "step": 3000
+    },
+    {
+      "epoch": 2.7968313140726933,
+      "grad_norm": 0.4985490325982957,
+      "learning_rate": 3.7625129444252676e-06,
+      "loss": 0.4696,
+      "step": 3001
+    },
+    {
+      "epoch": 2.7977632805219015,
+      "grad_norm": 0.6699343745174512,
+      "learning_rate": 3.7452537107352436e-06,
+      "loss": 0.4808,
+      "step": 3002
+    },
+    {
+      "epoch": 2.798695246971109,
+      "grad_norm": 0.47743513329601817,
+      "learning_rate": 3.727994477045219e-06,
+      "loss": 0.4287,
+      "step": 3003
+    },
+    {
+      "epoch": 2.799627213420317,
+      "grad_norm": 0.8029101582627323,
+      "learning_rate": 3.7107352433551947e-06,
+      "loss": 0.5012,
+      "step": 3004
+    },
+    {
+      "epoch": 2.800559179869525,
+      "grad_norm": 0.6520888461016684,
+      "learning_rate": 3.693476009665171e-06,
+      "loss": 0.4813,
+      "step": 3005
+    },
+    {
+      "epoch": 2.8014911463187326,
+      "grad_norm": 0.6258830772783726,
+      "learning_rate": 3.676216775975147e-06,
+      "loss": 0.5032,
+      "step": 3006
+    },
+    {
+      "epoch": 2.8024231127679404,
+      "grad_norm": 0.9720300882535665,
+      "learning_rate": 3.6589575422851227e-06,
+      "loss": 0.5593,
+      "step": 3007
+    },
+    {
+      "epoch": 2.803355079217148,
+      "grad_norm": 0.5300781132354346,
+      "learning_rate": 3.6416983085950983e-06,
+      "loss": 0.4516,
+      "step": 3008
+    },
+    {
+      "epoch": 2.804287045666356,
+      "grad_norm": 0.5772896188980297,
+      "learning_rate": 3.6244390749050742e-06,
+      "loss": 0.4449,
+      "step": 3009
+    },
+    {
+      "epoch": 2.805219012115564,
+      "grad_norm": 0.6190085734471573,
+      "learning_rate": 3.6071798412150506e-06,
+      "loss": 0.4605,
+      "step": 3010
+    },
+    {
+      "epoch": 2.8061509785647716,
+      "grad_norm": 0.7002568806058398,
+      "learning_rate": 3.589920607525026e-06,
+      "loss": 0.4606,
+      "step": 3011
+    },
+    {
+      "epoch": 2.8070829450139794,
+      "grad_norm": 0.2687442456142916,
+      "learning_rate": 3.5726613738350018e-06,
+      "loss": 0.4135,
+      "step": 3012
+    },
+    {
+      "epoch": 2.8080149114631876,
+      "grad_norm": 0.4645540203098398,
+      "learning_rate": 3.5554021401449777e-06,
+      "loss": 0.4335,
+      "step": 3013
+    },
+    {
+      "epoch": 2.808946877912395,
+      "grad_norm": 0.7317668427757814,
+      "learning_rate": 3.5381429064549533e-06,
+      "loss": 0.4689,
+      "step": 3014
+    },
+    {
+      "epoch": 2.809878844361603,
+      "grad_norm": 0.39132565959472626,
+      "learning_rate": 3.5208836727649297e-06,
+      "loss": 0.4494,
+      "step": 3015
+    },
+    {
+      "epoch": 2.810810810810811,
+      "grad_norm": 0.5509308038196162,
+      "learning_rate": 3.5036244390749053e-06,
+      "loss": 0.465,
+      "step": 3016
+    },
+    {
+      "epoch": 2.8117427772600188,
+      "grad_norm": 0.6163367115020035,
+      "learning_rate": 3.4863652053848813e-06,
+      "loss": 0.4301,
+      "step": 3017
+    },
+    {
+      "epoch": 2.8126747437092265,
+      "grad_norm": 0.3649240566242538,
+      "learning_rate": 3.469105971694857e-06,
+      "loss": 0.4221,
+      "step": 3018
+    },
+    {
+      "epoch": 2.8136067101584343,
+      "grad_norm": 0.5388814628673582,
+      "learning_rate": 3.4518467380048324e-06,
+      "loss": 0.4575,
+      "step": 3019
+    },
+    {
+      "epoch": 2.814538676607642,
+      "grad_norm": 0.462567890506607,
+      "learning_rate": 3.434587504314809e-06,
+      "loss": 0.4527,
+      "step": 3020
+    },
+    {
+      "epoch": 2.81547064305685,
+      "grad_norm": 0.936930720336926,
+      "learning_rate": 3.4173282706247848e-06,
+      "loss": 0.5307,
+      "step": 3021
+    },
+    {
+      "epoch": 2.8164026095060577,
+      "grad_norm": 0.5839731316197087,
+      "learning_rate": 3.4000690369347603e-06,
+      "loss": 0.4711,
+      "step": 3022
+    },
+    {
+      "epoch": 2.8173345759552655,
+      "grad_norm": 0.5130330721892797,
+      "learning_rate": 3.382809803244736e-06,
+      "loss": 0.451,
+      "step": 3023
+    },
+    {
+      "epoch": 2.8182665424044733,
+      "grad_norm": 0.5003653785945739,
+      "learning_rate": 3.365550569554712e-06,
+      "loss": 0.457,
+      "step": 3024
+    },
+    {
+      "epoch": 2.819198508853681,
+      "grad_norm": 0.6054447447982518,
+      "learning_rate": 3.348291335864688e-06,
+      "loss": 0.4837,
+      "step": 3025
+    },
+    {
+      "epoch": 2.8201304753028893,
+      "grad_norm": 0.48602650421133925,
+      "learning_rate": 3.331032102174664e-06,
+      "loss": 0.4802,
+      "step": 3026
+    },
+    {
+      "epoch": 2.821062441752097,
+      "grad_norm": 0.5322358675083603,
+      "learning_rate": 3.3137728684846394e-06,
+      "loss": 0.4626,
+      "step": 3027
+    },
+    {
+      "epoch": 2.821994408201305,
+      "grad_norm": 0.6264224469417519,
+      "learning_rate": 3.296513634794615e-06,
+      "loss": 0.5004,
+      "step": 3028
+    },
+    {
+      "epoch": 2.8229263746505127,
+      "grad_norm": 0.51594764844145,
+      "learning_rate": 3.279254401104591e-06,
+      "loss": 0.4648,
+      "step": 3029
+    },
+    {
+      "epoch": 2.8238583410997204,
+      "grad_norm": 0.6255470428844947,
+      "learning_rate": 3.2619951674145674e-06,
+      "loss": 0.4551,
+      "step": 3030
+    },
+    {
+      "epoch": 2.8247903075489282,
+      "grad_norm": 0.6055014721482498,
+      "learning_rate": 3.244735933724543e-06,
+      "loss": 0.4786,
+      "step": 3031
+    },
+    {
+      "epoch": 2.825722273998136,
+      "grad_norm": 0.6382371906739547,
+      "learning_rate": 3.2274767000345185e-06,
+      "loss": 0.4653,
+      "step": 3032
+    },
+    {
+      "epoch": 2.826654240447344,
+      "grad_norm": 0.40786559795815835,
+      "learning_rate": 3.2102174663444945e-06,
+      "loss": 0.438,
+      "step": 3033
+    },
+    {
+      "epoch": 2.8275862068965516,
+      "grad_norm": 0.394534193950971,
+      "learning_rate": 3.19295823265447e-06,
+      "loss": 0.4386,
+      "step": 3034
+    },
+    {
+      "epoch": 2.8285181733457594,
+      "grad_norm": 0.5134174133195191,
+      "learning_rate": 3.1756989989644464e-06,
+      "loss": 0.479,
+      "step": 3035
+    },
+    {
+      "epoch": 2.829450139794967,
+      "grad_norm": 0.4716177969168895,
+      "learning_rate": 3.158439765274422e-06,
+      "loss": 0.4265,
+      "step": 3036
+    },
+    {
+      "epoch": 2.8303821062441754,
+      "grad_norm": 0.6617044504483117,
+      "learning_rate": 3.141180531584398e-06,
+      "loss": 0.4852,
+      "step": 3037
+    },
+    {
+      "epoch": 2.831314072693383,
+      "grad_norm": 0.6379173713752976,
+      "learning_rate": 3.1239212978943735e-06,
+      "loss": 0.4792,
+      "step": 3038
+    },
+    {
+      "epoch": 2.832246039142591,
+      "grad_norm": 0.3356725796363077,
+      "learning_rate": 3.1066620642043495e-06,
+      "loss": 0.4453,
+      "step": 3039
+    },
+    {
+      "epoch": 2.8331780055917988,
+      "grad_norm": 0.668529178267048,
+      "learning_rate": 3.089402830514325e-06,
+      "loss": 0.4906,
+      "step": 3040
+    },
+    {
+      "epoch": 2.8341099720410066,
+      "grad_norm": 0.4446723075895617,
+      "learning_rate": 3.072143596824301e-06,
+      "loss": 0.4297,
+      "step": 3041
+    },
+    {
+      "epoch": 2.8350419384902144,
+      "grad_norm": 0.4241329047791564,
+      "learning_rate": 3.054884363134277e-06,
+      "loss": 0.4378,
+      "step": 3042
+    },
+    {
+      "epoch": 2.835973904939422,
+      "grad_norm": 0.47792410060158225,
+      "learning_rate": 3.0376251294442526e-06,
+      "loss": 0.4532,
+      "step": 3043
+    },
+    {
+      "epoch": 2.83690587138863,
+      "grad_norm": 0.7930715293436557,
+      "learning_rate": 3.0203658957542286e-06,
+      "loss": 0.4838,
+      "step": 3044
+    },
+    {
+      "epoch": 2.8378378378378377,
+      "grad_norm": 0.42900236300741357,
+      "learning_rate": 3.0031066620642046e-06,
+      "loss": 0.4453,
+      "step": 3045
+    },
+    {
+      "epoch": 2.8387698042870455,
+      "grad_norm": 0.4089677536110977,
+      "learning_rate": 2.98584742837418e-06,
+      "loss": 0.4471,
+      "step": 3046
+    },
+    {
+      "epoch": 2.8397017707362533,
+      "grad_norm": 0.4732417164787061,
+      "learning_rate": 2.968588194684156e-06,
+      "loss": 0.448,
+      "step": 3047
+    },
+    {
+      "epoch": 2.8406337371854615,
+      "grad_norm": 0.3636988285052308,
+      "learning_rate": 2.951328960994132e-06,
+      "loss": 0.4232,
+      "step": 3048
+    },
+    {
+      "epoch": 2.8415657036346693,
+      "grad_norm": 0.5809749226844074,
+      "learning_rate": 2.934069727304108e-06,
+      "loss": 0.4885,
+      "step": 3049
+    },
+    {
+      "epoch": 2.842497670083877,
+      "grad_norm": 0.7492830545928535,
+      "learning_rate": 2.9168104936140837e-06,
+      "loss": 0.4662,
+      "step": 3050
+    },
+    {
+      "epoch": 2.843429636533085,
+      "grad_norm": 0.38978371334683176,
+      "learning_rate": 2.8995512599240592e-06,
+      "loss": 0.4378,
+      "step": 3051
+    },
+    {
+      "epoch": 2.8443616029822927,
+      "grad_norm": 0.44616049540890856,
+      "learning_rate": 2.882292026234035e-06,
+      "loss": 0.4713,
+      "step": 3052
+    },
+    {
+      "epoch": 2.8452935694315005,
+      "grad_norm": 0.3908183832093615,
+      "learning_rate": 2.865032792544011e-06,
+      "loss": 0.4258,
+      "step": 3053
+    },
+    {
+      "epoch": 2.8462255358807083,
+      "grad_norm": 0.6241789384893534,
+      "learning_rate": 2.847773558853987e-06,
+      "loss": 0.4786,
+      "step": 3054
+    },
+    {
+      "epoch": 2.847157502329916,
+      "grad_norm": 0.4228137653772456,
+      "learning_rate": 2.8305143251639627e-06,
+      "loss": 0.4612,
+      "step": 3055
+    },
+    {
+      "epoch": 2.848089468779124,
+      "grad_norm": 0.41944436202240687,
+      "learning_rate": 2.8132550914739387e-06,
+      "loss": 0.4464,
+      "step": 3056
+    },
+    {
+      "epoch": 2.8490214352283316,
+      "grad_norm": 0.8495616910898022,
+      "learning_rate": 2.7959958577839147e-06,
+      "loss": 0.4871,
+      "step": 3057
+    },
+    {
+      "epoch": 2.8499534016775394,
+      "grad_norm": 0.5543170833559826,
+      "learning_rate": 2.7787366240938903e-06,
+      "loss": 0.4527,
+      "step": 3058
+    },
+    {
+      "epoch": 2.8508853681267476,
+      "grad_norm": 0.49634821906725984,
+      "learning_rate": 2.7614773904038662e-06,
+      "loss": 0.464,
+      "step": 3059
+    },
+    {
+      "epoch": 2.8518173345759554,
+      "grad_norm": 0.5662827425209963,
+      "learning_rate": 2.7442181567138422e-06,
+      "loss": 0.4463,
+      "step": 3060
+    },
+    {
+      "epoch": 2.852749301025163,
+      "grad_norm": 0.4589194667601999,
+      "learning_rate": 2.726958923023818e-06,
+      "loss": 0.4549,
+      "step": 3061
+    },
+    {
+      "epoch": 2.853681267474371,
+      "grad_norm": 0.6352089458853764,
+      "learning_rate": 2.7096996893337938e-06,
+      "loss": 0.4662,
+      "step": 3062
+    },
+    {
+      "epoch": 2.854613233923579,
+      "grad_norm": 0.5007153522058602,
+      "learning_rate": 2.6924404556437693e-06,
+      "loss": 0.4557,
+      "step": 3063
+    },
+    {
+      "epoch": 2.8555452003727866,
+      "grad_norm": 0.5444260125836435,
+      "learning_rate": 2.6751812219537453e-06,
+      "loss": 0.4666,
+      "step": 3064
+    },
+    {
+      "epoch": 2.8564771668219944,
+      "grad_norm": 0.8219136927326881,
+      "learning_rate": 2.6579219882637213e-06,
+      "loss": 0.4741,
+      "step": 3065
+    },
+    {
+      "epoch": 2.857409133271202,
+      "grad_norm": 0.5107541973384392,
+      "learning_rate": 2.640662754573697e-06,
+      "loss": 0.4655,
+      "step": 3066
+    },
+    {
+      "epoch": 2.85834109972041,
+      "grad_norm": 0.42468548194978506,
+      "learning_rate": 2.623403520883673e-06,
+      "loss": 0.4351,
+      "step": 3067
+    },
+    {
+      "epoch": 2.8592730661696177,
+      "grad_norm": 0.6019667737839063,
+      "learning_rate": 2.606144287193649e-06,
+      "loss": 0.4391,
+      "step": 3068
+    },
+    {
+      "epoch": 2.8602050326188255,
+      "grad_norm": 0.42923980121553856,
+      "learning_rate": 2.588885053503625e-06,
+      "loss": 0.4513,
+      "step": 3069
+    },
+    {
+      "epoch": 2.8611369990680338,
+      "grad_norm": 0.7949502761518862,
+      "learning_rate": 2.5716258198136004e-06,
+      "loss": 0.4703,
+      "step": 3070
+    },
+    {
+      "epoch": 2.862068965517241,
+      "grad_norm": 0.4134652886881098,
+      "learning_rate": 2.554366586123576e-06,
+      "loss": 0.4364,
+      "step": 3071
+    },
+    {
+      "epoch": 2.8630009319664493,
+      "grad_norm": 0.6102080101466645,
+      "learning_rate": 2.5371073524335523e-06,
+      "loss": 0.4852,
+      "step": 3072
+    },
+    {
+      "epoch": 2.863932898415657,
+      "grad_norm": 0.40018299383362854,
+      "learning_rate": 2.519848118743528e-06,
+      "loss": 0.4464,
+      "step": 3073
+    },
+    {
+      "epoch": 2.864864864864865,
+      "grad_norm": 0.5720266702548175,
+      "learning_rate": 2.502588885053504e-06,
+      "loss": 0.4579,
+      "step": 3074
+    },
+    {
+      "epoch": 2.8657968313140727,
+      "grad_norm": 0.5567179710290511,
+      "learning_rate": 2.4853296513634795e-06,
+      "loss": 0.4534,
+      "step": 3075
+    },
+    {
+      "epoch": 2.8667287977632805,
+      "grad_norm": 0.5600159091122943,
+      "learning_rate": 2.4680704176734554e-06,
+      "loss": 0.4582,
+      "step": 3076
+    },
+    {
+      "epoch": 2.8676607642124883,
+      "grad_norm": 0.6595586585854746,
+      "learning_rate": 2.4508111839834314e-06,
+      "loss": 0.4692,
+      "step": 3077
+    },
+    {
+      "epoch": 2.868592730661696,
+      "grad_norm": 0.6003230031273922,
+      "learning_rate": 2.433551950293407e-06,
+      "loss": 0.4652,
+      "step": 3078
+    },
+    {
+      "epoch": 2.869524697110904,
+      "grad_norm": 0.4528322524196316,
+      "learning_rate": 2.416292716603383e-06,
+      "loss": 0.4695,
+      "step": 3079
+    },
+    {
+      "epoch": 2.8704566635601116,
+      "grad_norm": 0.5413324970801714,
+      "learning_rate": 2.399033482913359e-06,
+      "loss": 0.4754,
+      "step": 3080
+    },
+    {
+      "epoch": 2.87138863000932,
+      "grad_norm": 0.3387656138883765,
+      "learning_rate": 2.3817742492233345e-06,
+      "loss": 0.4403,
+      "step": 3081
+    },
+    {
+      "epoch": 2.872320596458527,
+      "grad_norm": 0.7337382736160383,
+      "learning_rate": 2.3645150155333105e-06,
+      "loss": 0.4577,
+      "step": 3082
+    },
+    {
+      "epoch": 2.8732525629077355,
+      "grad_norm": 0.6579078126458809,
+      "learning_rate": 2.347255781843286e-06,
+      "loss": 0.4543,
+      "step": 3083
+    },
+    {
+      "epoch": 2.8741845293569432,
+      "grad_norm": 0.560331412728498,
+      "learning_rate": 2.3299965481532625e-06,
+      "loss": 0.4663,
+      "step": 3084
+    },
+    {
+      "epoch": 2.875116495806151,
+      "grad_norm": 0.3921722192721106,
+      "learning_rate": 2.312737314463238e-06,
+      "loss": 0.4208,
+      "step": 3085
+    },
+    {
+      "epoch": 2.876048462255359,
+      "grad_norm": 0.44810668049249147,
+      "learning_rate": 2.2954780807732136e-06,
+      "loss": 0.4213,
+      "step": 3086
+    },
+    {
+      "epoch": 2.8769804287045666,
+      "grad_norm": 0.46526293821698594,
+      "learning_rate": 2.2782188470831896e-06,
+      "loss": 0.4634,
+      "step": 3087
+    },
+    {
+      "epoch": 2.8779123951537744,
+      "grad_norm": 0.518443410569851,
+      "learning_rate": 2.2609596133931656e-06,
+      "loss": 0.4419,
+      "step": 3088
+    },
+    {
+      "epoch": 2.878844361602982,
+      "grad_norm": 0.5378156587326649,
+      "learning_rate": 2.243700379703141e-06,
+      "loss": 0.4508,
+      "step": 3089
+    },
+    {
+      "epoch": 2.87977632805219,
+      "grad_norm": 0.44754003015756255,
+      "learning_rate": 2.226441146013117e-06,
+      "loss": 0.454,
+      "step": 3090
+    },
+    {
+      "epoch": 2.8807082945013978,
+      "grad_norm": 0.4646332956684967,
+      "learning_rate": 2.2091819123230927e-06,
+      "loss": 0.4425,
+      "step": 3091
+    },
+    {
+      "epoch": 2.881640260950606,
+      "grad_norm": 0.5488593499095793,
+      "learning_rate": 2.191922678633069e-06,
+      "loss": 0.4506,
+      "step": 3092
+    },
+    {
+      "epoch": 2.8825722273998133,
+      "grad_norm": 0.4479275135283868,
+      "learning_rate": 2.1746634449430446e-06,
+      "loss": 0.4505,
+      "step": 3093
+    },
+    {
+      "epoch": 2.8835041938490216,
+      "grad_norm": 0.4479429904361553,
+      "learning_rate": 2.15740421125302e-06,
+      "loss": 0.454,
+      "step": 3094
+    },
+    {
+      "epoch": 2.8844361602982294,
+      "grad_norm": 0.7173551470305217,
+      "learning_rate": 2.140144977562996e-06,
+      "loss": 0.4652,
+      "step": 3095
+    },
+    {
+      "epoch": 2.885368126747437,
+      "grad_norm": 0.5362678213637534,
+      "learning_rate": 2.122885743872972e-06,
+      "loss": 0.4671,
+      "step": 3096
+    },
+    {
+      "epoch": 2.886300093196645,
+      "grad_norm": 0.4778391400595037,
+      "learning_rate": 2.105626510182948e-06,
+      "loss": 0.4466,
+      "step": 3097
+    },
+    {
+      "epoch": 2.8872320596458527,
+      "grad_norm": 0.9313004122281242,
+      "learning_rate": 2.0883672764929237e-06,
+      "loss": 0.5028,
+      "step": 3098
+    },
+    {
+      "epoch": 2.8881640260950605,
+      "grad_norm": 0.6348041841819364,
+      "learning_rate": 2.0711080428028997e-06,
+      "loss": 0.4562,
+      "step": 3099
+    },
+    {
+      "epoch": 2.8890959925442683,
+      "grad_norm": 0.4970895005430311,
+      "learning_rate": 2.0538488091128757e-06,
+      "loss": 0.4558,
+      "step": 3100
+    },
+    {
+      "epoch": 2.890027958993476,
+      "grad_norm": 0.6207675110143074,
+      "learning_rate": 2.0365895754228512e-06,
+      "loss": 0.4601,
+      "step": 3101
+    },
+    {
+      "epoch": 2.890959925442684,
+      "grad_norm": 0.5312867923298549,
+      "learning_rate": 2.0193303417328272e-06,
+      "loss": 0.446,
+      "step": 3102
+    },
+    {
+      "epoch": 2.891891891891892,
+      "grad_norm": 0.42419560973556664,
+      "learning_rate": 2.0020711080428028e-06,
+      "loss": 0.4345,
+      "step": 3103
+    },
+    {
+      "epoch": 2.8928238583410995,
+      "grad_norm": 0.3922116705609888,
+      "learning_rate": 1.9848118743527788e-06,
+      "loss": 0.4266,
+      "step": 3104
+    },
+    {
+      "epoch": 2.8937558247903077,
+      "grad_norm": 0.5819569556758024,
+      "learning_rate": 1.9675526406627547e-06,
+      "loss": 0.4816,
+      "step": 3105
+    },
+    {
+      "epoch": 2.8946877912395155,
+      "grad_norm": 0.5711498533002392,
+      "learning_rate": 1.9502934069727303e-06,
+      "loss": 0.4596,
+      "step": 3106
+    },
+    {
+      "epoch": 2.8956197576887233,
+      "grad_norm": 0.6023576299292633,
+      "learning_rate": 1.9330341732827063e-06,
+      "loss": 0.4547,
+      "step": 3107
+    },
+    {
+      "epoch": 2.896551724137931,
+      "grad_norm": 0.47211867904839655,
+      "learning_rate": 1.9157749395926823e-06,
+      "loss": 0.4649,
+      "step": 3108
+    },
+    {
+      "epoch": 2.897483690587139,
+      "grad_norm": 0.6020846711800932,
+      "learning_rate": 1.8985157059026578e-06,
+      "loss": 0.4917,
+      "step": 3109
+    },
+    {
+      "epoch": 2.8984156570363466,
+      "grad_norm": 0.5554514307280167,
+      "learning_rate": 1.8812564722126338e-06,
+      "loss": 0.4584,
+      "step": 3110
+    },
+    {
+      "epoch": 2.8993476234855544,
+      "grad_norm": 0.5399288894128479,
+      "learning_rate": 1.8639972385226096e-06,
+      "loss": 0.4623,
+      "step": 3111
+    },
+    {
+      "epoch": 2.900279589934762,
+      "grad_norm": 0.534044850283184,
+      "learning_rate": 1.8467380048325856e-06,
+      "loss": 0.4466,
+      "step": 3112
+    },
+    {
+      "epoch": 2.90121155638397,
+      "grad_norm": 0.5515341803175019,
+      "learning_rate": 1.8294787711425613e-06,
+      "loss": 0.4632,
+      "step": 3113
+    },
+    {
+      "epoch": 2.9021435228331782,
+      "grad_norm": 0.5133855801855504,
+      "learning_rate": 1.8122195374525371e-06,
+      "loss": 0.4085,
+      "step": 3114
+    },
+    {
+      "epoch": 2.9030754892823856,
+      "grad_norm": 0.5788268075918206,
+      "learning_rate": 1.794960303762513e-06,
+      "loss": 0.4546,
+      "step": 3115
+    },
+    {
+      "epoch": 2.904007455731594,
+      "grad_norm": 0.5296669675194036,
+      "learning_rate": 1.7777010700724889e-06,
+      "loss": 0.4557,
+      "step": 3116
+    },
+    {
+      "epoch": 2.9049394221808016,
+      "grad_norm": 0.38109465385921876,
+      "learning_rate": 1.7604418363824649e-06,
+      "loss": 0.4448,
+      "step": 3117
+    },
+    {
+      "epoch": 2.9058713886300094,
+      "grad_norm": 0.5425604494561077,
+      "learning_rate": 1.7431826026924406e-06,
+      "loss": 0.4495,
+      "step": 3118
+    },
+    {
+      "epoch": 2.906803355079217,
+      "grad_norm": 0.6046124149929537,
+      "learning_rate": 1.7259233690024162e-06,
+      "loss": 0.4628,
+      "step": 3119
+    },
+    {
+      "epoch": 2.907735321528425,
+      "grad_norm": 0.41039330522460166,
+      "learning_rate": 1.7086641353123924e-06,
+      "loss": 0.4504,
+      "step": 3120
+    },
+    {
+      "epoch": 2.9086672879776327,
+      "grad_norm": 0.4404080567822284,
+      "learning_rate": 1.691404901622368e-06,
+      "loss": 0.4256,
+      "step": 3121
+    },
+    {
+      "epoch": 2.9095992544268405,
+      "grad_norm": 0.8281859280381998,
+      "learning_rate": 1.674145667932344e-06,
+      "loss": 0.4813,
+      "step": 3122
+    },
+    {
+      "epoch": 2.9105312208760483,
+      "grad_norm": 0.5632589625063799,
+      "learning_rate": 1.6568864342423197e-06,
+      "loss": 0.4595,
+      "step": 3123
+    },
+    {
+      "epoch": 2.911463187325256,
+      "grad_norm": 0.44366746902272675,
+      "learning_rate": 1.6396272005522955e-06,
+      "loss": 0.4312,
+      "step": 3124
+    },
+    {
+      "epoch": 2.9123951537744643,
+      "grad_norm": 0.6648066888391041,
+      "learning_rate": 1.6223679668622715e-06,
+      "loss": 0.4642,
+      "step": 3125
+    },
+    {
+      "epoch": 2.9133271202236717,
+      "grad_norm": 0.8582580763347424,
+      "learning_rate": 1.6051087331722472e-06,
+      "loss": 0.4245,
+      "step": 3126
+    },
+    {
+      "epoch": 2.91425908667288,
+      "grad_norm": 0.3401404798488896,
+      "learning_rate": 1.5878494994822232e-06,
+      "loss": 0.4205,
+      "step": 3127
+    },
+    {
+      "epoch": 2.9151910531220877,
+      "grad_norm": 0.6921578260088357,
+      "learning_rate": 1.570590265792199e-06,
+      "loss": 0.4554,
+      "step": 3128
+    },
+    {
+      "epoch": 2.9161230195712955,
+      "grad_norm": 1.0486136575918463,
+      "learning_rate": 1.5533310321021748e-06,
+      "loss": 0.5112,
+      "step": 3129
+    },
+    {
+      "epoch": 2.9170549860205033,
+      "grad_norm": 0.5169037116100644,
+      "learning_rate": 1.5360717984121505e-06,
+      "loss": 0.4796,
+      "step": 3130
+    },
+    {
+      "epoch": 2.917986952469711,
+      "grad_norm": 0.4955031924290418,
+      "learning_rate": 1.5188125647221263e-06,
+      "loss": 0.4311,
+      "step": 3131
+    },
+    {
+      "epoch": 2.918918918918919,
+      "grad_norm": 0.44220299416796455,
+      "learning_rate": 1.5015533310321023e-06,
+      "loss": 0.4729,
+      "step": 3132
+    },
+    {
+      "epoch": 2.9198508853681266,
+      "grad_norm": 0.4028448925629502,
+      "learning_rate": 1.484294097342078e-06,
+      "loss": 0.4475,
+      "step": 3133
+    },
+    {
+      "epoch": 2.9207828518173344,
+      "grad_norm": 0.4229412236240877,
+      "learning_rate": 1.467034863652054e-06,
+      "loss": 0.451,
+      "step": 3134
+    },
+    {
+      "epoch": 2.9217148182665422,
+      "grad_norm": 0.4657299802798299,
+      "learning_rate": 1.4497756299620296e-06,
+      "loss": 0.4637,
+      "step": 3135
+    },
+    {
+      "epoch": 2.9226467847157505,
+      "grad_norm": 0.42477171763418026,
+      "learning_rate": 1.4325163962720056e-06,
+      "loss": 0.4383,
+      "step": 3136
+    },
+    {
+      "epoch": 2.923578751164958,
+      "grad_norm": 0.7791322379668747,
+      "learning_rate": 1.4152571625819814e-06,
+      "loss": 0.4559,
+      "step": 3137
+    },
+    {
+      "epoch": 2.924510717614166,
+      "grad_norm": 0.7201675480962665,
+      "learning_rate": 1.3979979288919574e-06,
+      "loss": 0.4715,
+      "step": 3138
+    },
+    {
+      "epoch": 2.925442684063374,
+      "grad_norm": 0.5070129934750162,
+      "learning_rate": 1.3807386952019331e-06,
+      "loss": 0.4302,
+      "step": 3139
+    },
+    {
+      "epoch": 2.9263746505125816,
+      "grad_norm": 0.49823272700317695,
+      "learning_rate": 1.363479461511909e-06,
+      "loss": 0.4557,
+      "step": 3140
+    },
+    {
+      "epoch": 2.9273066169617894,
+      "grad_norm": 0.43033289392499974,
+      "learning_rate": 1.3462202278218847e-06,
+      "loss": 0.4293,
+      "step": 3141
+    },
+    {
+      "epoch": 2.928238583410997,
+      "grad_norm": 0.5639827979104144,
+      "learning_rate": 1.3289609941318607e-06,
+      "loss": 0.4514,
+      "step": 3142
+    },
+    {
+      "epoch": 2.929170549860205,
+      "grad_norm": 0.7798428453884083,
+      "learning_rate": 1.3117017604418364e-06,
+      "loss": 0.5204,
+      "step": 3143
+    },
+    {
+      "epoch": 2.9301025163094128,
+      "grad_norm": 0.43432597030357417,
+      "learning_rate": 1.2944425267518124e-06,
+      "loss": 0.4787,
+      "step": 3144
+    },
+    {
+      "epoch": 2.9310344827586206,
+      "grad_norm": 0.6117411184710216,
+      "learning_rate": 1.277183293061788e-06,
+      "loss": 0.4579,
+      "step": 3145
+    },
+    {
+      "epoch": 2.9319664492078283,
+      "grad_norm": 0.8135145024832577,
+      "learning_rate": 1.259924059371764e-06,
+      "loss": 0.5235,
+      "step": 3146
+    },
+    {
+      "epoch": 2.9328984156570366,
+      "grad_norm": 0.49618506752721153,
+      "learning_rate": 1.2426648256817397e-06,
+      "loss": 0.4612,
+      "step": 3147
+    },
+    {
+      "epoch": 2.933830382106244,
+      "grad_norm": 0.6074089915781699,
+      "learning_rate": 1.2254055919917157e-06,
+      "loss": 0.4471,
+      "step": 3148
+    },
+    {
+      "epoch": 2.934762348555452,
+      "grad_norm": 0.3683163084019119,
+      "learning_rate": 1.2081463583016915e-06,
+      "loss": 0.4616,
+      "step": 3149
+    },
+    {
+      "epoch": 2.93569431500466,
+      "grad_norm": 0.5211588219260656,
+      "learning_rate": 1.1908871246116673e-06,
+      "loss": 0.4447,
+      "step": 3150
+    },
+    {
+      "epoch": 2.9366262814538677,
+      "grad_norm": 0.6620594964717723,
+      "learning_rate": 1.173627890921643e-06,
+      "loss": 0.4663,
+      "step": 3151
+    },
+    {
+      "epoch": 2.9375582479030755,
+      "grad_norm": 0.4523696309158488,
+      "learning_rate": 1.156368657231619e-06,
+      "loss": 0.4602,
+      "step": 3152
+    },
+    {
+      "epoch": 2.9384902143522833,
+      "grad_norm": 0.5261570888560383,
+      "learning_rate": 1.1391094235415948e-06,
+      "loss": 0.4417,
+      "step": 3153
+    },
+    {
+      "epoch": 2.939422180801491,
+      "grad_norm": 0.6310839192427549,
+      "learning_rate": 1.1218501898515706e-06,
+      "loss": 0.477,
+      "step": 3154
+    },
+    {
+      "epoch": 2.940354147250699,
+      "grad_norm": 0.36176638815121775,
+      "learning_rate": 1.1045909561615463e-06,
+      "loss": 0.4422,
+      "step": 3155
+    },
+    {
+      "epoch": 2.9412861136999067,
+      "grad_norm": 0.7220923979996495,
+      "learning_rate": 1.0873317224715223e-06,
+      "loss": 0.467,
+      "step": 3156
+    },
+    {
+      "epoch": 2.9422180801491145,
+      "grad_norm": 0.5701829115915455,
+      "learning_rate": 1.070072488781498e-06,
+      "loss": 0.4891,
+      "step": 3157
+    },
+    {
+      "epoch": 2.9431500465983227,
+      "grad_norm": 0.47281743474164173,
+      "learning_rate": 1.052813255091474e-06,
+      "loss": 0.4503,
+      "step": 3158
+    },
+    {
+      "epoch": 2.94408201304753,
+      "grad_norm": 0.5982577545702359,
+      "learning_rate": 1.0355540214014498e-06,
+      "loss": 0.4447,
+      "step": 3159
+    },
+    {
+      "epoch": 2.9450139794967383,
+      "grad_norm": 0.49732472255684645,
+      "learning_rate": 1.0182947877114256e-06,
+      "loss": 0.4639,
+      "step": 3160
+    },
+    {
+      "epoch": 2.945945945945946,
+      "grad_norm": 0.791898860602047,
+      "learning_rate": 1.0010355540214014e-06,
+      "loss": 0.5008,
+      "step": 3161
+    },
+    {
+      "epoch": 2.946877912395154,
+      "grad_norm": 0.32617359950637176,
+      "learning_rate": 9.837763203313774e-07,
+      "loss": 0.4142,
+      "step": 3162
+    },
+    {
+      "epoch": 2.9478098788443616,
+      "grad_norm": 0.6066221636468496,
+      "learning_rate": 9.665170866413531e-07,
+      "loss": 0.4741,
+      "step": 3163
+    },
+    {
+      "epoch": 2.9487418452935694,
+      "grad_norm": 0.6293605963267502,
+      "learning_rate": 9.492578529513289e-07,
+      "loss": 0.4513,
+      "step": 3164
+    },
+    {
+      "epoch": 2.949673811742777,
+      "grad_norm": 0.4627528301170894,
+      "learning_rate": 9.319986192613048e-07,
+      "loss": 0.4355,
+      "step": 3165
+    },
+    {
+      "epoch": 2.950605778191985,
+      "grad_norm": 0.8499015958433566,
+      "learning_rate": 9.147393855712807e-07,
+      "loss": 0.4855,
+      "step": 3166
+    },
+    {
+      "epoch": 2.951537744641193,
+      "grad_norm": 0.38342110369458016,
+      "learning_rate": 8.974801518812566e-07,
+      "loss": 0.4413,
+      "step": 3167
+    },
+    {
+      "epoch": 2.9524697110904006,
+      "grad_norm": 0.47579888615412697,
+      "learning_rate": 8.802209181912324e-07,
+      "loss": 0.4753,
+      "step": 3168
+    },
+    {
+      "epoch": 2.953401677539609,
+      "grad_norm": 0.8118338418183578,
+      "learning_rate": 8.629616845012081e-07,
+      "loss": 0.4656,
+      "step": 3169
+    },
+    {
+      "epoch": 2.954333643988816,
+      "grad_norm": 0.6785862347166333,
+      "learning_rate": 8.45702450811184e-07,
+      "loss": 0.4642,
+      "step": 3170
+    },
+    {
+      "epoch": 2.9552656104380244,
+      "grad_norm": 0.6908262371370925,
+      "learning_rate": 8.284432171211599e-07,
+      "loss": 0.4843,
+      "step": 3171
+    },
+    {
+      "epoch": 2.956197576887232,
+      "grad_norm": 0.5803572425847229,
+      "learning_rate": 8.111839834311357e-07,
+      "loss": 0.4717,
+      "step": 3172
+    },
+    {
+      "epoch": 2.95712954333644,
+      "grad_norm": 0.5399629376082125,
+      "learning_rate": 7.939247497411116e-07,
+      "loss": 0.4684,
+      "step": 3173
+    },
+    {
+      "epoch": 2.9580615097856477,
+      "grad_norm": 0.631392518230111,
+      "learning_rate": 7.766655160510874e-07,
+      "loss": 0.4193,
+      "step": 3174
+    },
+    {
+      "epoch": 2.9589934762348555,
+      "grad_norm": 0.40699558107171574,
+      "learning_rate": 7.594062823610632e-07,
+      "loss": 0.4521,
+      "step": 3175
+    },
+    {
+      "epoch": 2.9599254426840633,
+      "grad_norm": 0.6120716853701015,
+      "learning_rate": 7.42147048671039e-07,
+      "loss": 0.4486,
+      "step": 3176
+    },
+    {
+      "epoch": 2.960857409133271,
+      "grad_norm": 0.4780758593142827,
+      "learning_rate": 7.248878149810148e-07,
+      "loss": 0.4729,
+      "step": 3177
+    },
+    {
+      "epoch": 2.961789375582479,
+      "grad_norm": 0.5804345433749947,
+      "learning_rate": 7.076285812909907e-07,
+      "loss": 0.4579,
+      "step": 3178
+    },
+    {
+      "epoch": 2.9627213420316867,
+      "grad_norm": 0.6086834159975631,
+      "learning_rate": 6.903693476009666e-07,
+      "loss": 0.4679,
+      "step": 3179
+    },
+    {
+      "epoch": 2.963653308480895,
+      "grad_norm": 0.4693658127708384,
+      "learning_rate": 6.731101139109423e-07,
+      "loss": 0.4608,
+      "step": 3180
+    },
+    {
+      "epoch": 2.9645852749301023,
+      "grad_norm": 0.5522406953884424,
+      "learning_rate": 6.558508802209182e-07,
+      "loss": 0.4419,
+      "step": 3181
+    },
+    {
+      "epoch": 2.9655172413793105,
+      "grad_norm": 0.548762664131316,
+      "learning_rate": 6.38591646530894e-07,
+      "loss": 0.4544,
+      "step": 3182
+    },
+    {
+      "epoch": 2.9664492078285183,
+      "grad_norm": 0.7480586647483534,
+      "learning_rate": 6.213324128408699e-07,
+      "loss": 0.4992,
+      "step": 3183
+    },
+    {
+      "epoch": 2.967381174277726,
+      "grad_norm": 0.595607643229728,
+      "learning_rate": 6.040731791508457e-07,
+      "loss": 0.4587,
+      "step": 3184
+    },
+    {
+      "epoch": 2.968313140726934,
+      "grad_norm": 0.5018484988525125,
+      "learning_rate": 5.868139454608215e-07,
+      "loss": 0.4296,
+      "step": 3185
+    },
+    {
+      "epoch": 2.9692451071761417,
+      "grad_norm": 0.550996650433628,
+      "learning_rate": 5.695547117707974e-07,
+      "loss": 0.4534,
+      "step": 3186
+    },
+    {
+      "epoch": 2.9701770736253494,
+      "grad_norm": 0.47959418582363117,
+      "learning_rate": 5.522954780807732e-07,
+      "loss": 0.4438,
+      "step": 3187
+    },
+    {
+      "epoch": 2.9711090400745572,
+      "grad_norm": 0.5435130221834207,
+      "learning_rate": 5.35036244390749e-07,
+      "loss": 0.4538,
+      "step": 3188
+    },
+    {
+      "epoch": 2.972041006523765,
+      "grad_norm": 0.46576023043065334,
+      "learning_rate": 5.177770107007249e-07,
+      "loss": 0.4582,
+      "step": 3189
+    },
+    {
+      "epoch": 2.972972972972973,
+      "grad_norm": 0.6173305946160804,
+      "learning_rate": 5.005177770107007e-07,
+      "loss": 0.4562,
+      "step": 3190
+    },
+    {
+      "epoch": 2.973904939422181,
+      "grad_norm": 1.2231789106430275,
+      "learning_rate": 4.832585433206766e-07,
+      "loss": 0.4964,
+      "step": 3191
+    },
+    {
+      "epoch": 2.9748369058713884,
+      "grad_norm": 0.37717487801725935,
+      "learning_rate": 4.659993096306524e-07,
+      "loss": 0.4396,
+      "step": 3192
+    },
+    {
+      "epoch": 2.9757688723205966,
+      "grad_norm": 0.6042372949555815,
+      "learning_rate": 4.487400759406283e-07,
+      "loss": 0.4495,
+      "step": 3193
+    },
+    {
+      "epoch": 2.9767008387698044,
+      "grad_norm": 0.6007813722000637,
+      "learning_rate": 4.3148084225060405e-07,
+      "loss": 0.4669,
+      "step": 3194
+    },
+    {
+      "epoch": 2.977632805219012,
+      "grad_norm": 0.49011560042458174,
+      "learning_rate": 4.142216085605799e-07,
+      "loss": 0.4521,
+      "step": 3195
+    },
+    {
+      "epoch": 2.97856477166822,
+      "grad_norm": 0.4527952491269564,
+      "learning_rate": 3.969623748705558e-07,
+      "loss": 0.4578,
+      "step": 3196
+    },
+    {
+      "epoch": 2.9794967381174278,
+      "grad_norm": 0.5236472598566186,
+      "learning_rate": 3.797031411805316e-07,
+      "loss": 0.4596,
+      "step": 3197
+    },
+    {
+      "epoch": 2.9804287045666356,
+      "grad_norm": 0.5310750498421181,
+      "learning_rate": 3.624439074905074e-07,
+      "loss": 0.4673,
+      "step": 3198
+    },
+    {
+      "epoch": 2.9813606710158433,
+      "grad_norm": 0.476742728583095,
+      "learning_rate": 3.451846738004833e-07,
+      "loss": 0.4717,
+      "step": 3199
+    },
+    {
+      "epoch": 2.982292637465051,
+      "grad_norm": 0.7662481626304551,
+      "learning_rate": 3.279254401104591e-07,
+      "loss": 0.4718,
+      "step": 3200
+    },
+    {
+      "epoch": 2.983224603914259,
+      "grad_norm": 0.5218164580669101,
+      "learning_rate": 3.1066620642043493e-07,
+      "loss": 0.4641,
+      "step": 3201
+    },
+    {
+      "epoch": 2.984156570363467,
+      "grad_norm": 0.48482526572563406,
+      "learning_rate": 2.9340697273041076e-07,
+      "loss": 0.4435,
+      "step": 3202
+    },
+    {
+      "epoch": 2.9850885368126745,
+      "grad_norm": 0.5737480642446772,
+      "learning_rate": 2.761477390403866e-07,
+      "loss": 0.4444,
+      "step": 3203
+    },
+    {
+      "epoch": 2.9860205032618827,
+      "grad_norm": 0.5067983097670704,
+      "learning_rate": 2.5888850535036246e-07,
+      "loss": 0.4651,
+      "step": 3204
+    },
+    {
+      "epoch": 2.9869524697110905,
+      "grad_norm": 0.45072081264679525,
+      "learning_rate": 2.416292716603383e-07,
+      "loss": 0.4401,
+      "step": 3205
+    },
+    {
+      "epoch": 2.9878844361602983,
+      "grad_norm": 0.5325490134139788,
+      "learning_rate": 2.2437003797031414e-07,
+      "loss": 0.4589,
+      "step": 3206
+    },
+    {
+      "epoch": 2.988816402609506,
+      "grad_norm": 0.4612227271218359,
+      "learning_rate": 2.0711080428028996e-07,
+      "loss": 0.4753,
+      "step": 3207
+    },
+    {
+      "epoch": 2.989748369058714,
+      "grad_norm": 2.7779661876711557,
+      "learning_rate": 1.898515705902658e-07,
+      "loss": 0.738,
+      "step": 3208
+    },
+    {
+      "epoch": 2.9906803355079217,
+      "grad_norm": 0.3636693726345704,
+      "learning_rate": 1.7259233690024164e-07,
+      "loss": 0.4487,
+      "step": 3209
+    },
+    {
+      "epoch": 2.9916123019571295,
+      "grad_norm": 0.4657839084944504,
+      "learning_rate": 1.5533310321021747e-07,
+      "loss": 0.4433,
+      "step": 3210
+    },
+    {
+      "epoch": 2.9925442684063372,
+      "grad_norm": 0.5553330987319657,
+      "learning_rate": 1.380738695201933e-07,
+      "loss": 0.4647,
+      "step": 3211
+    },
+    {
+      "epoch": 2.993476234855545,
+      "grad_norm": 0.5362358190178296,
+      "learning_rate": 1.2081463583016914e-07,
+      "loss": 0.4695,
+      "step": 3212
+    },
+    {
+      "epoch": 2.9944082013047533,
+      "grad_norm": 0.3759098920246909,
+      "learning_rate": 1.0355540214014498e-07,
+      "loss": 0.4516,
+      "step": 3213
+    },
+    {
+      "epoch": 2.9953401677539606,
+      "grad_norm": 0.4368967240351239,
+      "learning_rate": 8.629616845012082e-08,
+      "loss": 0.4599,
+      "step": 3214
+    },
+    {
+      "epoch": 2.996272134203169,
+      "grad_norm": 0.4355689108975172,
+      "learning_rate": 6.903693476009665e-08,
+      "loss": 0.4404,
+      "step": 3215
+    },
+    {
+      "epoch": 2.9972041006523766,
+      "grad_norm": 0.6032327522487283,
+      "learning_rate": 5.177770107007249e-08,
+      "loss": 0.4532,
+      "step": 3216
+    },
+    {
+      "epoch": 2.9981360671015844,
+      "grad_norm": 0.4472418693348213,
+      "learning_rate": 3.451846738004832e-08,
+      "loss": 0.4607,
+      "step": 3217
+    },
+    {
+      "epoch": 2.999068033550792,
+      "grad_norm": 0.41522093079343253,
+      "learning_rate": 1.725923369002416e-08,
+      "loss": 0.4236,
+      "step": 3218
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5896578401611983,
+      "learning_rate": 0.0,
+      "loss": 0.4816,
+      "step": 3219
+    },
+    {
+      "epoch": 3.0,
+      "step": 3219,
+      "total_flos": 0.0,
+      "train_loss": 0.982338245137577,
+      "train_runtime": 213089.2567,
+      "train_samples_per_second": 0.242,
+      "train_steps_per_second": 0.015
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3219,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}