diff --git "a/checkpoint-36500/trainer_state.json" "b/checkpoint-36500/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-36500/trainer_state.json"
@@ -0,0 +1,25584 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9980722166773766,
+  "eval_steps": 500,
+  "global_step": 36500,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00027344444292530865,
+      "grad_norm": 1896.0,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 6.5599,
+      "step": 10
+    },
+    {
+      "epoch": 0.0005468888858506173,
+      "grad_norm": 27.0,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 3.6252,
+      "step": 20
+    },
+    {
+      "epoch": 0.000820333328775926,
+      "grad_norm": 25.875,
+      "learning_rate": 2.9e-06,
+      "loss": 3.3786,
+      "step": 30
+    },
+    {
+      "epoch": 0.0010937777717012346,
+      "grad_norm": 138.0,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 2.8689,
+      "step": 40
+    },
+    {
+      "epoch": 0.0013672222146265433,
+      "grad_norm": 8.0,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 2.112,
+      "step": 50
+    },
+    {
+      "epoch": 0.001640666657551852,
+      "grad_norm": 9.8125,
+      "learning_rate": 5.9e-06,
+      "loss": 1.4067,
+      "step": 60
+    },
+    {
+      "epoch": 0.0019141111004771605,
+      "grad_norm": 26.75,
+      "learning_rate": 6.9e-06,
+      "loss": 1.2991,
+      "step": 70
+    },
+    {
+      "epoch": 0.002187555543402469,
+      "grad_norm": 5280.0,
+      "learning_rate": 7.9e-06,
+      "loss": 2.1783,
+      "step": 80
+    },
+    {
+      "epoch": 0.002460999986327778,
+      "grad_norm": 4.34375,
+      "learning_rate": 8.900000000000001e-06,
+      "loss": 1.1253,
+      "step": 90
+    },
+    {
+      "epoch": 0.0027344444292530866,
+      "grad_norm": 3.578125,
+      "learning_rate": 9.9e-06,
+      "loss": 1.1952,
+      "step": 100
+    },
+    {
+      "epoch": 0.0030078888721783953,
+      "grad_norm": 3.09375,
+      "learning_rate": 9.997532285925805e-06,
+      "loss": 1.0657,
+      "step": 110
+    },
+    {
+      "epoch": 0.003281333315103704,
+      "grad_norm": 3.9375,
+      "learning_rate": 9.99479038139892e-06,
+      "loss": 1.0124,
+      "step": 120
+    },
+    {
+      "epoch": 0.0035547777580290122,
+      "grad_norm": 12.5625,
+      "learning_rate": 9.992048476872036e-06,
+      "loss": 0.9081,
+      "step": 130
+    },
+    {
+      "epoch": 0.003828222200954321,
+      "grad_norm": 3.359375,
+      "learning_rate": 9.989306572345152e-06,
+      "loss": 0.9172,
+      "step": 140
+    },
+    {
+      "epoch": 0.00410166664387963,
+      "grad_norm": 3.171875,
+      "learning_rate": 9.986564667818267e-06,
+      "loss": 0.8917,
+      "step": 150
+    },
+    {
+      "epoch": 0.004375111086804938,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.983822763291383e-06,
+      "loss": 0.8512,
+      "step": 160
+    },
+    {
+      "epoch": 0.004648555529730247,
+      "grad_norm": 3.0625,
+      "learning_rate": 9.9810808587645e-06,
+      "loss": 0.898,
+      "step": 170
+    },
+    {
+      "epoch": 0.004921999972655556,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.978338954237614e-06,
+      "loss": 0.8238,
+      "step": 180
+    },
+    {
+      "epoch": 0.0051954444155808644,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.97559704971073e-06,
+      "loss": 0.8266,
+      "step": 190
+    },
+    {
+      "epoch": 0.005468888858506173,
+      "grad_norm": 3.15625,
+      "learning_rate": 9.972855145183845e-06,
+      "loss": 0.8465,
+      "step": 200
+    },
+    {
+      "epoch": 0.005742333301431482,
+      "grad_norm": 3.375,
+      "learning_rate": 9.970113240656962e-06,
+      "loss": 0.924,
+      "step": 210
+    },
+    {
+      "epoch": 0.0060157777443567905,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.967371336130078e-06,
+      "loss": 0.7541,
+      "step": 220
+    },
+    {
+      "epoch": 0.006289222187282099,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.964629431603192e-06,
+      "loss": 0.8921,
+      "step": 230
+    },
+    {
+      "epoch": 0.006562666630207408,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.961887527076309e-06,
+      "loss": 0.8874,
+      "step": 240
+    },
+    {
+      "epoch": 0.006836111073132717,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.959145622549423e-06,
+      "loss": 0.792,
+      "step": 250
+    },
+    {
+      "epoch": 0.0071095555160580245,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.95640371802254e-06,
+      "loss": 0.8166,
+      "step": 260
+    },
+    {
+      "epoch": 0.007382999958983333,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.953661813495656e-06,
+      "loss": 0.8695,
+      "step": 270
+    },
+    {
+      "epoch": 0.007656444401908642,
+      "grad_norm": 2.375,
+      "learning_rate": 9.95091990896877e-06,
+      "loss": 0.7685,
+      "step": 280
+    },
+    {
+      "epoch": 0.00792988884483395,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.948178004441887e-06,
+      "loss": 0.8273,
+      "step": 290
+    },
+    {
+      "epoch": 0.00820333328775926,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.945436099915002e-06,
+      "loss": 0.8486,
+      "step": 300
+    },
+    {
+      "epoch": 0.008476777730684568,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.942694195388118e-06,
+      "loss": 0.7712,
+      "step": 310
+    },
+    {
+      "epoch": 0.008750222173609877,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.939952290861233e-06,
+      "loss": 0.8415,
+      "step": 320
+    },
+    {
+      "epoch": 0.009023666616535185,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.937210386334349e-06,
+      "loss": 0.8591,
+      "step": 330
+    },
+    {
+      "epoch": 0.009297111059460494,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.934468481807464e-06,
+      "loss": 0.8108,
+      "step": 340
+    },
+    {
+      "epoch": 0.009570555502385803,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.93172657728058e-06,
+      "loss": 0.8237,
+      "step": 350
+    },
+    {
+      "epoch": 0.009843999945311111,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.928984672753695e-06,
+      "loss": 0.7216,
+      "step": 360
+    },
+    {
+      "epoch": 0.01011744438823642,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.926242768226811e-06,
+      "loss": 0.7826,
+      "step": 370
+    },
+    {
+      "epoch": 0.010390888831161729,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.923500863699926e-06,
+      "loss": 0.8139,
+      "step": 380
+    },
+    {
+      "epoch": 0.010664333274087038,
+      "grad_norm": 2.5,
+      "learning_rate": 9.920758959173042e-06,
+      "loss": 0.8307,
+      "step": 390
+    },
+    {
+      "epoch": 0.010937777717012346,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.918017054646158e-06,
+      "loss": 0.8096,
+      "step": 400
+    },
+    {
+      "epoch": 0.011211222159937655,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.915275150119273e-06,
+      "loss": 0.7875,
+      "step": 410
+    },
+    {
+      "epoch": 0.011484666602862964,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.91253324559239e-06,
+      "loss": 0.8298,
+      "step": 420
+    },
+    {
+      "epoch": 0.011758111045788272,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.909791341065504e-06,
+      "loss": 0.8296,
+      "step": 430
+    },
+    {
+      "epoch": 0.012031555488713581,
+      "grad_norm": 2.5,
+      "learning_rate": 9.90704943653862e-06,
+      "loss": 0.7895,
+      "step": 440
+    },
+    {
+      "epoch": 0.01230499993163889,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.904307532011736e-06,
+      "loss": 0.724,
+      "step": 450
+    },
+    {
+      "epoch": 0.012578444374564198,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.901565627484851e-06,
+      "loss": 0.7024,
+      "step": 460
+    },
+    {
+      "epoch": 0.012851888817489507,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.898823722957967e-06,
+      "loss": 0.7299,
+      "step": 470
+    },
+    {
+      "epoch": 0.013125333260414816,
+      "grad_norm": 2.625,
+      "learning_rate": 9.896081818431082e-06,
+      "loss": 0.8051,
+      "step": 480
+    },
+    {
+      "epoch": 0.013398777703340125,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.893339913904198e-06,
+      "loss": 0.7479,
+      "step": 490
+    },
+    {
+      "epoch": 0.013672222146265433,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.890598009377315e-06,
+      "loss": 0.7858,
+      "step": 500
+    },
+    {
+      "epoch": 0.013945666589190742,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.88785610485043e-06,
+      "loss": 0.7261,
+      "step": 510
+    },
+    {
+      "epoch": 0.014219111032116049,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.885114200323546e-06,
+      "loss": 0.8463,
+      "step": 520
+    },
+    {
+      "epoch": 0.014492555475041358,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.882372295796662e-06,
+      "loss": 0.7567,
+      "step": 530
+    },
+    {
+      "epoch": 0.014765999917966666,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.879630391269777e-06,
+      "loss": 0.7418,
+      "step": 540
+    },
+    {
+      "epoch": 0.015039444360891975,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.876888486742893e-06,
+      "loss": 0.7041,
+      "step": 550
+    },
+    {
+      "epoch": 0.015312888803817284,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.874146582216008e-06,
+      "loss": 0.7952,
+      "step": 560
+    },
+    {
+      "epoch": 0.015586333246742592,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.871404677689124e-06,
+      "loss": 0.6998,
+      "step": 570
+    },
+    {
+      "epoch": 0.0158597776896679,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.86866277316224e-06,
+      "loss": 0.8426,
+      "step": 580
+    },
+    {
+      "epoch": 0.01613322213259321,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.865920868635355e-06,
+      "loss": 0.7063,
+      "step": 590
+    },
+    {
+      "epoch": 0.01640666657551852,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.863178964108471e-06,
+      "loss": 0.8102,
+      "step": 600
+    },
+    {
+      "epoch": 0.016680111018443827,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.860437059581586e-06,
+      "loss": 0.8037,
+      "step": 610
+    },
+    {
+      "epoch": 0.016953555461369136,
+      "grad_norm": 2.5,
+      "learning_rate": 9.857695155054702e-06,
+      "loss": 0.7553,
+      "step": 620
+    },
+    {
+      "epoch": 0.017226999904294445,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.854953250527819e-06,
+      "loss": 0.8293,
+      "step": 630
+    },
+    {
+      "epoch": 0.017500444347219753,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.852211346000933e-06,
+      "loss": 0.7451,
+      "step": 640
+    },
+    {
+      "epoch": 0.017773888790145062,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.84946944147405e-06,
+      "loss": 0.7486,
+      "step": 650
+    },
+    {
+      "epoch": 0.01804733323307037,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.846727536947164e-06,
+      "loss": 0.6981,
+      "step": 660
+    },
+    {
+      "epoch": 0.01832077767599568,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.84398563242028e-06,
+      "loss": 0.7658,
+      "step": 670
+    },
+    {
+      "epoch": 0.018594222118920988,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.841243727893395e-06,
+      "loss": 0.7807,
+      "step": 680
+    },
+    {
+      "epoch": 0.018867666561846297,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.838501823366511e-06,
+      "loss": 0.7659,
+      "step": 690
+    },
+    {
+      "epoch": 0.019141111004771606,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.835759918839626e-06,
+      "loss": 0.7199,
+      "step": 700
+    },
+    {
+      "epoch": 0.019414555447696914,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.833018014312742e-06,
+      "loss": 0.7504,
+      "step": 710
+    },
+    {
+      "epoch": 0.019687999890622223,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.830276109785859e-06,
+      "loss": 0.7001,
+      "step": 720
+    },
+    {
+      "epoch": 0.01996144433354753,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.827534205258973e-06,
+      "loss": 0.7776,
+      "step": 730
+    },
+    {
+      "epoch": 0.02023488877647284,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.82479230073209e-06,
+      "loss": 0.6961,
+      "step": 740
+    },
+    {
+      "epoch": 0.02050833321939815,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.822050396205204e-06,
+      "loss": 0.7371,
+      "step": 750
+    },
+    {
+      "epoch": 0.020781777662323458,
+      "grad_norm": 2.15625,
+      "learning_rate": 9.81930849167832e-06,
+      "loss": 0.7172,
+      "step": 760
+    },
+    {
+      "epoch": 0.021055222105248766,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.816566587151435e-06,
+      "loss": 0.7591,
+      "step": 770
+    },
+    {
+      "epoch": 0.021328666548174075,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.813824682624552e-06,
+      "loss": 0.7655,
+      "step": 780
+    },
+    {
+      "epoch": 0.021602110991099384,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.811082778097666e-06,
+      "loss": 0.771,
+      "step": 790
+    },
+    {
+      "epoch": 0.021875555434024693,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.808340873570783e-06,
+      "loss": 0.7777,
+      "step": 800
+    },
+    {
+      "epoch": 0.02214899987695,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.805598969043899e-06,
+      "loss": 0.7443,
+      "step": 810
+    },
+    {
+      "epoch": 0.02242244431987531,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.802857064517014e-06,
+      "loss": 0.7241,
+      "step": 820
+    },
+    {
+      "epoch": 0.02269588876280062,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.80011515999013e-06,
+      "loss": 0.7555,
+      "step": 830
+    },
+    {
+      "epoch": 0.022969333205725927,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.797373255463245e-06,
+      "loss": 0.7858,
+      "step": 840
+    },
+    {
+      "epoch": 0.023242777648651236,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.79463135093636e-06,
+      "loss": 0.7149,
+      "step": 850
+    },
+    {
+      "epoch": 0.023516222091576545,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.791889446409477e-06,
+      "loss": 0.7042,
+      "step": 860
+    },
+    {
+      "epoch": 0.023789666534501853,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.789147541882592e-06,
+      "loss": 0.8091,
+      "step": 870
+    },
+    {
+      "epoch": 0.024063110977427162,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.786405637355708e-06,
+      "loss": 0.7842,
+      "step": 880
+    },
+    {
+      "epoch": 0.02433655542035247,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.783663732828824e-06,
+      "loss": 0.7569,
+      "step": 890
+    },
+    {
+      "epoch": 0.02460999986327778,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.780921828301939e-06,
+      "loss": 0.7228,
+      "step": 900
+    },
+    {
+      "epoch": 0.02488344430620309,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.778179923775055e-06,
+      "loss": 0.8507,
+      "step": 910
+    },
+    {
+      "epoch": 0.025156888749128397,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.77543801924817e-06,
+      "loss": 0.814,
+      "step": 920
+    },
+    {
+      "epoch": 0.025430333192053706,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.772696114721286e-06,
+      "loss": 0.7462,
+      "step": 930
+    },
+    {
+      "epoch": 0.025703777634979014,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.769954210194403e-06,
+      "loss": 0.7244,
+      "step": 940
+    },
+    {
+      "epoch": 0.025977222077904323,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.767212305667517e-06,
+      "loss": 0.711,
+      "step": 950
+    },
+    {
+      "epoch": 0.026250666520829632,
+      "grad_norm": 2.5,
+      "learning_rate": 9.764470401140634e-06,
+      "loss": 0.8858,
+      "step": 960
+    },
+    {
+      "epoch": 0.02652411096375494,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.761728496613748e-06,
+      "loss": 0.7479,
+      "step": 970
+    },
+    {
+      "epoch": 0.02679755540668025,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.758986592086865e-06,
+      "loss": 0.7941,
+      "step": 980
+    },
+    {
+      "epoch": 0.027070999849605558,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.756244687559981e-06,
+      "loss": 0.8549,
+      "step": 990
+    },
+    {
+      "epoch": 0.027344444292530867,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.753502783033096e-06,
+      "loss": 0.7651,
+      "step": 1000
+    },
+    {
+      "epoch": 0.027617888735456175,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.750760878506212e-06,
+      "loss": 0.8144,
+      "step": 1010
+    },
+    {
+      "epoch": 0.027891333178381484,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.748018973979327e-06,
+      "loss": 0.7076,
+      "step": 1020
+    },
+    {
+      "epoch": 0.028164777621306793,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.745277069452443e-06,
+      "loss": 0.6917,
+      "step": 1030
+    },
+    {
+      "epoch": 0.028438222064232098,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.74253516492556e-06,
+      "loss": 0.7972,
+      "step": 1040
+    },
+    {
+      "epoch": 0.028711666507157407,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.739793260398674e-06,
+      "loss": 0.7161,
+      "step": 1050
+    },
+    {
+      "epoch": 0.028985110950082715,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.73705135587179e-06,
+      "loss": 0.752,
+      "step": 1060
+    },
+    {
+      "epoch": 0.029258555393008024,
+      "grad_norm": 2.5,
+      "learning_rate": 9.734309451344905e-06,
+      "loss": 0.8185,
+      "step": 1070
+    },
+    {
+      "epoch": 0.029531999835933333,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.731567546818021e-06,
+      "loss": 0.7669,
+      "step": 1080
+    },
+    {
+      "epoch": 0.02980544427885864,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.728825642291136e-06,
+      "loss": 0.7679,
+      "step": 1090
+    },
+    {
+      "epoch": 0.03007888872178395,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.726083737764252e-06,
+      "loss": 0.7009,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03035233316470926,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.723341833237367e-06,
+      "loss": 0.7807,
+      "step": 1110
+    },
+    {
+      "epoch": 0.030625777607634568,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.720599928710483e-06,
+      "loss": 0.7801,
+      "step": 1120
+    },
+    {
+      "epoch": 0.030899222050559876,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.717858024183598e-06,
+      "loss": 0.7226,
+      "step": 1130
+    },
+    {
+      "epoch": 0.031172666493485185,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.715116119656714e-06,
+      "loss": 0.7538,
+      "step": 1140
+    },
+    {
+      "epoch": 0.0314461109364105,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.712374215129829e-06,
+      "loss": 0.8462,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0317195553793358,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.709632310602945e-06,
+      "loss": 0.6615,
+      "step": 1160
+    },
+    {
+      "epoch": 0.031992999822261114,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.706890406076061e-06,
+      "loss": 0.8396,
+      "step": 1170
+    },
+    {
+      "epoch": 0.03226644426518642,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.704148501549176e-06,
+      "loss": 0.7388,
+      "step": 1180
+    },
+    {
+      "epoch": 0.03253988870811173,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.701406597022292e-06,
+      "loss": 0.802,
+      "step": 1190
+    },
+    {
+      "epoch": 0.03281333315103704,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.698664692495409e-06,
+      "loss": 0.6973,
+      "step": 1200
+    },
+    {
+      "epoch": 0.03308677759396235,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.695922787968523e-06,
+      "loss": 0.8428,
+      "step": 1210
+    },
+    {
+      "epoch": 0.033360222036887655,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.69318088344164e-06,
+      "loss": 0.7894,
+      "step": 1220
+    },
+    {
+      "epoch": 0.03363366647981297,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.690438978914754e-06,
+      "loss": 0.7225,
+      "step": 1230
+    },
+    {
+      "epoch": 0.03390711092273827,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.68769707438787e-06,
+      "loss": 0.8022,
+      "step": 1240
+    },
+    {
+      "epoch": 0.034180555365663584,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.684955169860987e-06,
+      "loss": 0.7958,
+      "step": 1250
+    },
+    {
+      "epoch": 0.03445399980858889,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.682213265334102e-06,
+      "loss": 0.7053,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0347274442515142,
+      "grad_norm": 2.875,
+      "learning_rate": 9.679471360807218e-06,
+      "loss": 0.8154,
+      "step": 1270
+    },
+    {
+      "epoch": 0.03500088869443951,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.676729456280332e-06,
+      "loss": 0.7323,
+      "step": 1280
+    },
+    {
+      "epoch": 0.03527433313736482,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.673987551753449e-06,
+      "loss": 0.7683,
+      "step": 1290
+    },
+    {
+      "epoch": 0.035547777580290124,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.671245647226565e-06,
+      "loss": 0.7345,
+      "step": 1300
+    },
+    {
+      "epoch": 0.035821222023215436,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.66850374269968e-06,
+      "loss": 0.7208,
+      "step": 1310
+    },
+    {
+      "epoch": 0.03609466646614074,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.665761838172796e-06,
+      "loss": 0.827,
+      "step": 1320
+    },
+    {
+      "epoch": 0.036368110909066054,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.66301993364591e-06,
+      "loss": 0.8135,
+      "step": 1330
+    },
+    {
+      "epoch": 0.03664155535199136,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.660278029119027e-06,
+      "loss": 0.6895,
+      "step": 1340
+    },
+    {
+      "epoch": 0.03691499979491667,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.657536124592143e-06,
+      "loss": 0.7645,
+      "step": 1350
+    },
+    {
+      "epoch": 0.037188444237841976,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.654794220065258e-06,
+      "loss": 0.7629,
+      "step": 1360
+    },
+    {
+      "epoch": 0.03746188868076729,
+      "grad_norm": 2.171875,
+      "learning_rate": 9.652052315538374e-06,
+      "loss": 0.7334,
+      "step": 1370
+    },
+    {
+      "epoch": 0.037735333123692594,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.64931041101149e-06,
+      "loss": 0.769,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0380087775666179,
+      "grad_norm": 2.375,
+      "learning_rate": 9.646568506484605e-06,
+      "loss": 0.7722,
+      "step": 1390
+    },
+    {
+      "epoch": 0.03828222200954321,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.643826601957722e-06,
+      "loss": 0.7187,
+      "step": 1400
+    },
+    {
+      "epoch": 0.038555666452468516,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.641084697430836e-06,
+      "loss": 0.7417,
+      "step": 1410
+    },
+    {
+      "epoch": 0.03882911089539383,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.638342792903953e-06,
+      "loss": 0.7457,
+      "step": 1420
+    },
+    {
+      "epoch": 0.039102555338319134,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.635600888377067e-06,
+      "loss": 0.7441,
+      "step": 1430
+    },
+    {
+      "epoch": 0.039375999781244446,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.632858983850184e-06,
+      "loss": 0.6988,
+      "step": 1440
+    },
+    {
+      "epoch": 0.03964944422416975,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.630117079323298e-06,
+      "loss": 0.7586,
+      "step": 1450
+    },
+    {
+      "epoch": 0.03992288866709506,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.627375174796415e-06,
+      "loss": 0.7791,
+      "step": 1460
+    },
+    {
+      "epoch": 0.04019633311002037,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.62463327026953e-06,
+      "loss": 0.784,
+      "step": 1470
+    },
+    {
+      "epoch": 0.04046977755294568,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.621891365742646e-06,
+      "loss": 0.6897,
+      "step": 1480
+    },
+    {
+      "epoch": 0.040743221995870986,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.61914946121576e-06,
+      "loss": 0.7845,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0410166664387963,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.616407556688876e-06,
+      "loss": 0.7592,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0412901108817216,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.613665652161993e-06,
+      "loss": 0.6765,
+      "step": 1510
+    },
+    {
+      "epoch": 0.041563555324646916,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.610923747635107e-06,
+      "loss": 0.7223,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04183699976757222,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.608181843108224e-06,
+      "loss": 0.6825,
+      "step": 1530
+    },
+    {
+      "epoch": 0.04211044421049753,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.605439938581338e-06,
+      "loss": 0.7194,
+      "step": 1540
+    },
+    {
+      "epoch": 0.04238388865342284,
+      "grad_norm": 2.03125,
+      "learning_rate": 9.602698034054455e-06,
+      "loss": 0.6613,
+      "step": 1550
+    },
+    {
+      "epoch": 0.04265733309634815,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.599956129527571e-06,
+      "loss": 0.8118,
+      "step": 1560
+    },
+    {
+      "epoch": 0.042930777539273456,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.597214225000686e-06,
+      "loss": 0.7024,
+      "step": 1570
+    },
+    {
+      "epoch": 0.04320422198219877,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.594472320473802e-06,
+      "loss": 0.7343,
+      "step": 1580
+    },
+    {
+      "epoch": 0.04347766642512407,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.591730415946917e-06,
+      "loss": 0.6862,
+      "step": 1590
+    },
+    {
+      "epoch": 0.043751110868049385,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.588988511420033e-06,
+      "loss": 0.6886,
+      "step": 1600
+    },
+    {
+      "epoch": 0.04402455531097469,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.58624660689315e-06,
+      "loss": 0.7629,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0442979997539,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.583504702366264e-06,
+      "loss": 0.7699,
+      "step": 1620
+    },
+    {
+      "epoch": 0.04457144419682531,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.58076279783938e-06,
+      "loss": 0.7388,
+      "step": 1630
+    },
+    {
+      "epoch": 0.04484488863975062,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.578020893312495e-06,
+      "loss": 0.7558,
+      "step": 1640
+    },
+    {
+      "epoch": 0.045118333082675925,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.575278988785611e-06,
+      "loss": 0.8286,
+      "step": 1650
+    },
+    {
+      "epoch": 0.04539177752560124,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.572537084258728e-06,
+      "loss": 0.7431,
+      "step": 1660
+    },
+    {
+      "epoch": 0.04566522196852654,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.569795179731842e-06,
+      "loss": 0.7432,
+      "step": 1670
+    },
+    {
+      "epoch": 0.045938666411451855,
+      "grad_norm": 2.375,
+      "learning_rate": 9.567053275204959e-06,
+      "loss": 0.7249,
+      "step": 1680
+    },
+    {
+      "epoch": 0.04621211085437716,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.564311370678073e-06,
+      "loss": 0.7563,
+      "step": 1690
+    },
+    {
+      "epoch": 0.04648555529730247,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.56156946615119e-06,
+      "loss": 0.7134,
+      "step": 1700
+    },
+    {
+      "epoch": 0.04675899974022778,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.558827561624306e-06,
+      "loss": 0.7157,
+      "step": 1710
+    },
+    {
+      "epoch": 0.04703244418315309,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.55608565709742e-06,
+      "loss": 0.711,
+      "step": 1720
+    },
+    {
+      "epoch": 0.047305888626078395,
+      "grad_norm": 2.625,
+      "learning_rate": 9.553343752570537e-06,
+      "loss": 0.7702,
+      "step": 1730
+    },
+    {
+      "epoch": 0.04757933306900371,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.550601848043653e-06,
+      "loss": 0.786,
+      "step": 1740
+    },
+    {
+      "epoch": 0.04785277751192901,
+      "grad_norm": 2.625,
+      "learning_rate": 9.547859943516768e-06,
+      "loss": 0.6589,
+      "step": 1750
+    },
+    {
+      "epoch": 0.048126221954854324,
+      "grad_norm": 2.265625,
+      "learning_rate": 9.545118038989884e-06,
+      "loss": 0.6113,
+      "step": 1760
+    },
+    {
+      "epoch": 0.04839966639777963,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.542376134462999e-06,
+      "loss": 0.7724,
+      "step": 1770
+    },
+    {
+      "epoch": 0.04867311084070494,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.539634229936115e-06,
+      "loss": 0.7697,
+      "step": 1780
+    },
+    {
+      "epoch": 0.04894655528363025,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.53689232540923e-06,
+      "loss": 0.7129,
+      "step": 1790
+    },
+    {
+      "epoch": 0.04921999972655556,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.534150420882346e-06,
+      "loss": 0.7778,
+      "step": 1800
+    },
+    {
+      "epoch": 0.049493444169480864,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.53140851635546e-06,
+      "loss": 0.6428,
+      "step": 1810
+    },
+    {
+      "epoch": 0.04976688861240618,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.528666611828577e-06,
+      "loss": 0.7865,
+      "step": 1820
+    },
+    {
+      "epoch": 0.05004033305533148,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.525924707301693e-06,
+      "loss": 0.8194,
+      "step": 1830
+    },
+    {
+      "epoch": 0.050313777498256794,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.523182802774808e-06,
+      "loss": 0.7181,
+      "step": 1840
+    },
+    {
+      "epoch": 0.0505872219411821,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.520440898247924e-06,
+      "loss": 0.7142,
+      "step": 1850
+    },
+    {
+      "epoch": 0.05086066638410741,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.517698993721039e-06,
+      "loss": 0.7482,
+      "step": 1860
+    },
+    {
+      "epoch": 0.05113411082703272,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.514957089194155e-06,
+      "loss": 0.7046,
+      "step": 1870
+    },
+    {
+      "epoch": 0.05140755526995803,
+      "grad_norm": 3.0,
+      "learning_rate": 9.51221518466727e-06,
+      "loss": 0.7585,
+      "step": 1880
+    },
+    {
+      "epoch": 0.051680999712883334,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.509473280140386e-06,
+      "loss": 0.7295,
+      "step": 1890
+    },
+    {
+      "epoch": 0.051954444155808646,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.5067313756135e-06,
+      "loss": 0.7939,
+      "step": 1900
+    },
+    {
+      "epoch": 0.05222788859873395,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.503989471086617e-06,
+      "loss": 0.7067,
+      "step": 1910
+    },
+    {
+      "epoch": 0.052501333041659264,
+      "grad_norm": 2.5,
+      "learning_rate": 9.501247566559733e-06,
+      "loss": 0.7374,
+      "step": 1920
+    },
+    {
+      "epoch": 0.05277477748458457,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.498505662032848e-06,
+      "loss": 0.798,
+      "step": 1930
+    },
+    {
+      "epoch": 0.05304822192750988,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.495763757505964e-06,
+      "loss": 0.7785,
+      "step": 1940
+    },
+    {
+      "epoch": 0.053321666370435186,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.493021852979079e-06,
+      "loss": 0.7982,
+      "step": 1950
+    },
+    {
+      "epoch": 0.0535951108133605,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.490279948452195e-06,
+      "loss": 0.7191,
+      "step": 1960
+    },
+    {
+      "epoch": 0.053868555256285804,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.487538043925312e-06,
+      "loss": 0.6877,
+      "step": 1970
+    },
+    {
+      "epoch": 0.054141999699211116,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.484796139398426e-06,
+      "loss": 0.7858,
+      "step": 1980
+    },
+    {
+      "epoch": 0.05441544414213642,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.482054234871543e-06,
+      "loss": 0.7308,
+      "step": 1990
+    },
+    {
+      "epoch": 0.05468888858506173,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.479312330344657e-06,
+      "loss": 0.6782,
+      "step": 2000
+    },
+    {
+      "epoch": 0.05496233302798704,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.476570425817774e-06,
+      "loss": 0.7048,
+      "step": 2010
+    },
+    {
+      "epoch": 0.05523577747091235,
+      "grad_norm": 2.375,
+      "learning_rate": 9.47382852129089e-06,
+      "loss": 0.7187,
+      "step": 2020
+    },
+    {
+      "epoch": 0.055509221913837656,
+      "grad_norm": 2.234375,
+      "learning_rate": 9.471086616764005e-06,
+      "loss": 0.7757,
+      "step": 2030
+    },
+    {
+      "epoch": 0.05578266635676297,
+      "grad_norm": 2.75,
+      "learning_rate": 9.468344712237121e-06,
+      "loss": 0.6249,
+      "step": 2040
+    },
+    {
+      "epoch": 0.05605611079968827,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.465602807710236e-06,
+      "loss": 0.6795,
+      "step": 2050
+    },
+    {
+      "epoch": 0.056329555242613585,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.462860903183352e-06,
+      "loss": 0.7831,
+      "step": 2060
+    },
+    {
+      "epoch": 0.05660299968553889,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.460118998656468e-06,
+      "loss": 0.7694,
+      "step": 2070
+    },
+    {
+      "epoch": 0.056876444128464196,
+      "grad_norm": 2.5,
+      "learning_rate": 9.457377094129583e-06,
+      "loss": 0.7548,
+      "step": 2080
+    },
+    {
+      "epoch": 0.05714988857138951,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.4546351896027e-06,
+      "loss": 0.7172,
+      "step": 2090
+    },
+    {
+      "epoch": 0.05742333301431481,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.451893285075816e-06,
+      "loss": 0.6696,
+      "step": 2100
+    },
+    {
+      "epoch": 0.057696777457240125,
+      "grad_norm": 2.375,
+      "learning_rate": 9.44915138054893e-06,
+      "loss": 0.6847,
+      "step": 2110
+    },
+    {
+      "epoch": 0.05797022190016543,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.446409476022047e-06,
+      "loss": 0.6531,
+      "step": 2120
+    },
+    {
+      "epoch": 0.05824366634309074,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.443667571495161e-06,
+      "loss": 0.7827,
+      "step": 2130
+    },
+    {
+      "epoch": 0.05851711078601605,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.440925666968277e-06,
+      "loss": 0.7604,
+      "step": 2140
+    },
+    {
+      "epoch": 0.05879055522894136,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.438183762441394e-06,
+      "loss": 0.7345,
+      "step": 2150
+    },
+    {
+      "epoch": 0.059063999671866665,
+      "grad_norm": 2.640625,
+      "learning_rate": 9.435441857914508e-06,
+      "loss": 0.7002,
+      "step": 2160
+    },
+    {
+      "epoch": 0.05933744411479198,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.432699953387625e-06,
+      "loss": 0.7382,
+      "step": 2170
+    },
+    {
+      "epoch": 0.05961088855771728,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.42995804886074e-06,
+      "loss": 0.7264,
+      "step": 2180
+    },
+    {
+      "epoch": 0.059884333000642595,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.427216144333856e-06,
+      "loss": 0.7551,
+      "step": 2190
+    },
+    {
+      "epoch": 0.0601577774435679,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.42447423980697e-06,
+      "loss": 0.6975,
+      "step": 2200
+    },
+    {
+      "epoch": 0.06043122188649321,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.421732335280087e-06,
+      "loss": 0.6908,
+      "step": 2210
+    },
+    {
+      "epoch": 0.06070466632941852,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.418990430753201e-06,
+      "loss": 0.7643,
+      "step": 2220
+    },
+    {
+      "epoch": 0.06097811077234383,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.416248526226318e-06,
+      "loss": 0.7596,
+      "step": 2230
+    },
+    {
+      "epoch": 0.061251555215269135,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.413506621699432e-06,
+      "loss": 0.758,
+      "step": 2240
+    },
+    {
+      "epoch": 0.06152499965819445,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.410764717172549e-06,
+      "loss": 0.6851,
+      "step": 2250
+    },
+    {
+      "epoch": 0.06179844410111975,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.408022812645663e-06,
+      "loss": 0.7515,
+      "step": 2260
+    },
+    {
+      "epoch": 0.062071888544045065,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.40528090811878e-06,
+      "loss": 0.7356,
+      "step": 2270
+    },
+    {
+      "epoch": 0.06234533298697037,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.402539003591896e-06,
+      "loss": 0.6987,
+      "step": 2280
+    },
+    {
+      "epoch": 0.06261877742989568,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.39979709906501e-06,
+      "loss": 0.7255,
+      "step": 2290
+    },
+    {
+      "epoch": 0.062892221872821,
+      "grad_norm": 1.8828125,
+      "learning_rate": 9.397055194538127e-06,
+      "loss": 0.6714,
+      "step": 2300
+    },
+    {
+      "epoch": 0.0631656663157463,
+      "grad_norm": 2.375,
+      "learning_rate": 9.394313290011242e-06,
+      "loss": 0.7164,
+      "step": 2310
+    },
+    {
+      "epoch": 0.0634391107586716,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.391571385484358e-06,
+      "loss": 0.7498,
+      "step": 2320
+    },
+    {
+      "epoch": 0.06371255520159691,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.388829480957474e-06,
+      "loss": 0.6796,
+      "step": 2330
+    },
+    {
+      "epoch": 0.06398599964452223,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.386087576430589e-06,
+      "loss": 0.7478,
+      "step": 2340
+    },
+    {
+      "epoch": 0.06425944408744753,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.383345671903705e-06,
+      "loss": 0.7363,
+      "step": 2350
+    },
+    {
+      "epoch": 0.06453288853037284,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.38060376737682e-06,
+      "loss": 0.6932,
+      "step": 2360
+    },
+    {
+      "epoch": 0.06480633297329814,
+      "grad_norm": 2.75,
+      "learning_rate": 9.377861862849936e-06,
+      "loss": 0.7147,
+      "step": 2370
+    },
+    {
+      "epoch": 0.06507977741622346,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.375119958323052e-06,
+      "loss": 0.7282,
+      "step": 2380
+    },
+    {
+      "epoch": 0.06535322185914877,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.372378053796167e-06,
+      "loss": 0.6753,
+      "step": 2390
+    },
+    {
+      "epoch": 0.06562666630207407,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.369636149269283e-06,
+      "loss": 0.6923,
+      "step": 2400
+    },
+    {
+      "epoch": 0.06590011074499938,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.3668942447424e-06,
+      "loss": 0.6947,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0661735551879247,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.364152340215514e-06,
+      "loss": 0.6729,
+      "step": 2420
+    },
+    {
+      "epoch": 0.06644699963085,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.36141043568863e-06,
+      "loss": 0.7164,
+      "step": 2430
+    },
+    {
+      "epoch": 0.06672044407377531,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.358668531161745e-06,
+      "loss": 0.7092,
+      "step": 2440
+    },
+    {
+      "epoch": 0.06699388851670061,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.355926626634862e-06,
+      "loss": 0.791,
+      "step": 2450
+    },
+    {
+      "epoch": 0.06726733295962593,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.353184722107978e-06,
+      "loss": 0.7566,
+      "step": 2460
+    },
+    {
+      "epoch": 0.06754077740255124,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.350442817581093e-06,
+      "loss": 0.6943,
+      "step": 2470
+    },
+    {
+      "epoch": 0.06781422184547654,
+      "grad_norm": 2.375,
+      "learning_rate": 9.347700913054209e-06,
+      "loss": 0.7462,
+      "step": 2480
+    },
+    {
+      "epoch": 0.06808766628840185,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.344959008527324e-06,
+      "loss": 0.6632,
+      "step": 2490
+    },
+    {
+      "epoch": 0.06836111073132717,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.34221710400044e-06,
+      "loss": 0.7105,
+      "step": 2500
+    },
+    {
+      "epoch": 0.06863455517425247,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.339475199473556e-06,
+      "loss": 0.7333,
+      "step": 2510
+    },
+    {
+      "epoch": 0.06890799961717778,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.336733294946671e-06,
+      "loss": 0.7125,
+      "step": 2520
+    },
+    {
+      "epoch": 0.06918144406010308,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.333991390419787e-06,
+      "loss": 0.7751,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0694548885030284,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.331249485892902e-06,
+      "loss": 0.7913,
+      "step": 2540
+    },
+    {
+      "epoch": 0.06972833294595371,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.328507581366018e-06,
+      "loss": 0.7526,
+      "step": 2550
+    },
+    {
+      "epoch": 0.07000177738887901,
+      "grad_norm": 1.7734375,
+      "learning_rate": 9.325765676839133e-06,
+      "loss": 0.6748,
+      "step": 2560
+    },
+    {
+      "epoch": 0.07027522183180432,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.323023772312249e-06,
+      "loss": 0.7181,
+      "step": 2570
+    },
+    {
+      "epoch": 0.07054866627472964,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.320281867785364e-06,
+      "loss": 0.7296,
+      "step": 2580
+    },
+    {
+      "epoch": 0.07082211071765494,
+      "grad_norm": 1.984375,
+      "learning_rate": 9.31753996325848e-06,
+      "loss": 0.6338,
+      "step": 2590
+    },
+    {
+      "epoch": 0.07109555516058025,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.314798058731596e-06,
+      "loss": 0.7439,
+      "step": 2600
+    },
+    {
+      "epoch": 0.07136899960350555,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.312056154204711e-06,
+      "loss": 0.7339,
+      "step": 2610
+    },
+    {
+      "epoch": 0.07164244404643087,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.309314249677827e-06,
+      "loss": 0.7345,
+      "step": 2620
+    },
+    {
+      "epoch": 0.07191588848935618,
+      "grad_norm": 2.875,
+      "learning_rate": 9.306572345150942e-06,
+      "loss": 0.7684,
+      "step": 2630
+    },
+    {
+      "epoch": 0.07218933293228148,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.303830440624058e-06,
+      "loss": 0.6764,
+      "step": 2640
+    },
+    {
+      "epoch": 0.07246277737520679,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.301088536097173e-06,
+      "loss": 0.698,
+      "step": 2650
+    },
+    {
+      "epoch": 0.07273622181813211,
+      "grad_norm": 2.21875,
+      "learning_rate": 9.29834663157029e-06,
+      "loss": 0.7214,
+      "step": 2660
+    },
+    {
+      "epoch": 0.07300966626105741,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.295604727043404e-06,
+      "loss": 0.7003,
+      "step": 2670
+    },
+    {
+      "epoch": 0.07328311070398272,
+      "grad_norm": 2.34375,
+      "learning_rate": 9.29286282251652e-06,
+      "loss": 0.7278,
+      "step": 2680
+    },
+    {
+      "epoch": 0.07355655514690802,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.290120917989637e-06,
+      "loss": 0.7587,
+      "step": 2690
+    },
+    {
+      "epoch": 0.07382999958983334,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.287379013462751e-06,
+      "loss": 0.6787,
+      "step": 2700
+    },
+    {
+      "epoch": 0.07410344403275865,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.284637108935868e-06,
+      "loss": 0.7689,
+      "step": 2710
+    },
+    {
+      "epoch": 0.07437688847568395,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.281895204408982e-06,
+      "loss": 0.7211,
+      "step": 2720
+    },
+    {
+      "epoch": 0.07465033291860926,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.279153299882099e-06,
+      "loss": 0.7231,
+      "step": 2730
+    },
+    {
+      "epoch": 0.07492377736153458,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.276411395355215e-06,
+      "loss": 0.7228,
+      "step": 2740
+    },
+    {
+      "epoch": 0.07519722180445988,
+      "grad_norm": 2.5,
+      "learning_rate": 9.27366949082833e-06,
+      "loss": 0.6982,
+      "step": 2750
+    },
+    {
+      "epoch": 0.07547066624738519,
+      "grad_norm": 2.5,
+      "learning_rate": 9.270927586301446e-06,
+      "loss": 0.7375,
+      "step": 2760
+    },
+    {
+      "epoch": 0.07574411069031049,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.268185681774562e-06,
+      "loss": 0.7524,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0760175551332358,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.265443777247677e-06,
+      "loss": 0.7594,
+      "step": 2780
+    },
+    {
+      "epoch": 0.07629099957616112,
+      "grad_norm": 2.875,
+      "learning_rate": 9.262701872720793e-06,
+      "loss": 0.7392,
+      "step": 2790
+    },
+    {
+      "epoch": 0.07656444401908642,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.259959968193908e-06,
+      "loss": 0.6191,
+      "step": 2800
+    },
+    {
+      "epoch": 0.07683788846201173,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.257218063667024e-06,
+      "loss": 0.694,
+      "step": 2810
+    },
+    {
+      "epoch": 0.07711133290493703,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.25447615914014e-06,
+      "loss": 0.7963,
+      "step": 2820
+    },
+    {
+      "epoch": 0.07738477734786235,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.251734254613255e-06,
+      "loss": 0.7302,
+      "step": 2830
+    },
+    {
+      "epoch": 0.07765822179078766,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.248992350086371e-06,
+      "loss": 0.7355,
+      "step": 2840
+    },
+    {
+      "epoch": 0.07793166623371296,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.246250445559486e-06,
+      "loss": 0.702,
+      "step": 2850
+    },
+    {
+      "epoch": 0.07820511067663827,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.243508541032602e-06,
+      "loss": 0.6293,
+      "step": 2860
+    },
+    {
+      "epoch": 0.07847855511956359,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.240766636505719e-06,
+      "loss": 0.733,
+      "step": 2870
+    },
+    {
+      "epoch": 0.07875199956248889,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.238024731978833e-06,
+      "loss": 0.7017,
+      "step": 2880
+    },
+    {
+      "epoch": 0.0790254440054142,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.23528282745195e-06,
+      "loss": 0.7087,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0792988884483395,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.232540922925064e-06,
+      "loss": 0.7148,
+      "step": 2900
+    },
+    {
+      "epoch": 0.07957233289126482,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.22979901839818e-06,
+      "loss": 0.6705,
+      "step": 2910
+    },
+    {
+      "epoch": 0.07984577733419013,
+      "grad_norm": 2.75,
+      "learning_rate": 9.227057113871297e-06,
+      "loss": 0.8337,
+      "step": 2920
+    },
+    {
+      "epoch": 0.08011922177711543,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.224315209344412e-06,
+      "loss": 0.7417,
+      "step": 2930
+    },
+    {
+      "epoch": 0.08039266622004074,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.221573304817528e-06,
+      "loss": 0.6853,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08066611066296606,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.218831400290643e-06,
+      "loss": 0.7064,
+      "step": 2950
+    },
+    {
+      "epoch": 0.08093955510589136,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.216089495763759e-06,
+      "loss": 0.7303,
+      "step": 2960
+    },
+    {
+      "epoch": 0.08121299954881667,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.213347591236874e-06,
+      "loss": 0.7357,
+      "step": 2970
+    },
+    {
+      "epoch": 0.08148644399174197,
+      "grad_norm": 2.25,
+      "learning_rate": 9.21060568670999e-06,
+      "loss": 0.7279,
+      "step": 2980
+    },
+    {
+      "epoch": 0.08175988843466729,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.207863782183104e-06,
+      "loss": 0.7148,
+      "step": 2990
+    },
+    {
+      "epoch": 0.0820333328775926,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.20512187765622e-06,
+      "loss": 0.7621,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0823067773205179,
+      "grad_norm": 2.296875,
+      "learning_rate": 9.202379973129335e-06,
+      "loss": 0.7009,
+      "step": 3010
+    },
+    {
+      "epoch": 0.0825802217634432,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.199638068602452e-06,
+      "loss": 0.6917,
+      "step": 3020
+    },
+    {
+      "epoch": 0.08285366620636853,
+      "grad_norm": 2.515625,
+      "learning_rate": 9.196896164075566e-06,
+      "loss": 0.6573,
+      "step": 3030
+    },
+    {
+      "epoch": 0.08312711064929383,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.194154259548683e-06,
+      "loss": 0.7536,
+      "step": 3040
+    },
+    {
+      "epoch": 0.08340055509221914,
+      "grad_norm": 2.609375,
+      "learning_rate": 9.191412355021799e-06,
+      "loss": 0.7113,
+      "step": 3050
+    },
+    {
+      "epoch": 0.08367399953514444,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.188670450494914e-06,
+      "loss": 0.6865,
+      "step": 3060
+    },
+    {
+      "epoch": 0.08394744397806976,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.18592854596803e-06,
+      "loss": 0.6777,
+      "step": 3070
+    },
+    {
+      "epoch": 0.08422088842099507,
+      "grad_norm": 2.453125,
+      "learning_rate": 9.183186641441145e-06,
+      "loss": 0.7431,
+      "step": 3080
+    },
+    {
+      "epoch": 0.08449433286392037,
+      "grad_norm": 2.859375,
+      "learning_rate": 9.180444736914261e-06,
+      "loss": 0.8253,
+      "step": 3090
+    },
+    {
+      "epoch": 0.08476777730684568,
+      "grad_norm": 2.96875,
+      "learning_rate": 9.177702832387377e-06,
+      "loss": 0.6511,
+      "step": 3100
+    },
+    {
+      "epoch": 0.085041221749771,
+      "grad_norm": 2.625,
+      "learning_rate": 9.174960927860492e-06,
+      "loss": 0.6804,
+      "step": 3110
+    },
+    {
+      "epoch": 0.0853146661926963,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.172219023333608e-06,
+      "loss": 0.7164,
+      "step": 3120
+    },
+    {
+      "epoch": 0.0855881106356216,
+      "grad_norm": 2.0,
+      "learning_rate": 9.169477118806725e-06,
+      "loss": 0.6317,
+      "step": 3130
+    },
+    {
+      "epoch": 0.08586155507854691,
+      "grad_norm": 2.375,
+      "learning_rate": 9.16673521427984e-06,
+      "loss": 0.6372,
+      "step": 3140
+    },
+    {
+      "epoch": 0.08613499952147223,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.163993309752956e-06,
+      "loss": 0.8034,
+      "step": 3150
+    },
+    {
+      "epoch": 0.08640844396439754,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.16125140522607e-06,
+      "loss": 0.7131,
+      "step": 3160
+    },
+    {
+      "epoch": 0.08668188840732284,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.158509500699187e-06,
+      "loss": 0.7006,
+      "step": 3170
+    },
+    {
+      "epoch": 0.08695533285024815,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.155767596172303e-06,
+      "loss": 0.6829,
+      "step": 3180
+    },
+    {
+      "epoch": 0.08722877729317347,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.153025691645417e-06,
+      "loss": 0.7358,
+      "step": 3190
+    },
+    {
+      "epoch": 0.08750222173609877,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.150283787118534e-06,
+      "loss": 0.7103,
+      "step": 3200
+    },
+    {
+      "epoch": 0.08777566617902408,
+      "grad_norm": 3.015625,
+      "learning_rate": 9.147541882591648e-06,
+      "loss": 0.7994,
+      "step": 3210
+    },
+    {
+      "epoch": 0.08804911062194938,
+      "grad_norm": 2.65625,
+      "learning_rate": 9.144799978064765e-06,
+      "loss": 0.73,
+      "step": 3220
+    },
+    {
+      "epoch": 0.0883225550648747,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.142058073537881e-06,
+      "loss": 0.7445,
+      "step": 3230
+    },
+    {
+      "epoch": 0.0885959995078,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.139316169010996e-06,
+      "loss": 0.7588,
+      "step": 3240
+    },
+    {
+      "epoch": 0.08886944395072531,
+      "grad_norm": 2.765625,
+      "learning_rate": 9.136574264484112e-06,
+      "loss": 0.7597,
+      "step": 3250
+    },
+    {
+      "epoch": 0.08914288839365062,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.133832359957227e-06,
+      "loss": 0.6559,
+      "step": 3260
+    },
+    {
+      "epoch": 0.08941633283657593,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.131090455430343e-06,
+      "loss": 0.7017,
+      "step": 3270
+    },
+    {
+      "epoch": 0.08968977727950124,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.12834855090346e-06,
+      "loss": 0.7672,
+      "step": 3280
+    },
+    {
+      "epoch": 0.08996322172242655,
+      "grad_norm": 2.75,
+      "learning_rate": 9.125606646376574e-06,
+      "loss": 0.7921,
+      "step": 3290
+    },
+    {
+      "epoch": 0.09023666616535185,
+      "grad_norm": 2.125,
+      "learning_rate": 9.12286474184969e-06,
+      "loss": 0.7452,
+      "step": 3300
+    },
+    {
+      "epoch": 0.09051011060827717,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.120122837322805e-06,
+      "loss": 0.6165,
+      "step": 3310
+    },
+    {
+      "epoch": 0.09078355505120247,
+      "grad_norm": 2.078125,
+      "learning_rate": 9.117380932795921e-06,
+      "loss": 0.7387,
+      "step": 3320
+    },
+    {
+      "epoch": 0.09105699949412778,
+      "grad_norm": 2.375,
+      "learning_rate": 9.114639028269036e-06,
+      "loss": 0.7043,
+      "step": 3330
+    },
+    {
+      "epoch": 0.09133044393705309,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.111897123742152e-06,
+      "loss": 0.755,
+      "step": 3340
+    },
+    {
+      "epoch": 0.0916038883799784,
+      "grad_norm": 2.3125,
+      "learning_rate": 9.109155219215267e-06,
+      "loss": 0.6737,
+      "step": 3350
+    },
+    {
+      "epoch": 0.09187733282290371,
+      "grad_norm": 2.546875,
+      "learning_rate": 9.106413314688383e-06,
+      "loss": 0.7364,
+      "step": 3360
+    },
+    {
+      "epoch": 0.09215077726582901,
+      "grad_norm": 2.140625,
+      "learning_rate": 9.103671410161498e-06,
+      "loss": 0.6957,
+      "step": 3370
+    },
+    {
+      "epoch": 0.09242422170875432,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.100929505634614e-06,
+      "loss": 0.747,
+      "step": 3380
+    },
+    {
+      "epoch": 0.09269766615167964,
+      "grad_norm": 2.6875,
+      "learning_rate": 9.09818760110773e-06,
+      "loss": 0.7225,
+      "step": 3390
+    },
+    {
+      "epoch": 0.09297111059460494,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.095445696580845e-06,
+      "loss": 0.6379,
+      "step": 3400
+    },
+    {
+      "epoch": 0.09324455503753025,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.092703792053961e-06,
+      "loss": 0.7161,
+      "step": 3410
+    },
+    {
+      "epoch": 0.09351799948045555,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.089961887527076e-06,
+      "loss": 0.7085,
+      "step": 3420
+    },
+    {
+      "epoch": 0.09379144392338087,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.087219983000192e-06,
+      "loss": 0.7941,
+      "step": 3430
+    },
+    {
+      "epoch": 0.09406488836630618,
+      "grad_norm": 2.28125,
+      "learning_rate": 9.084478078473307e-06,
+      "loss": 0.6901,
+      "step": 3440
+    },
+    {
+      "epoch": 0.09433833280923148,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.081736173946423e-06,
+      "loss": 0.7395,
+      "step": 3450
+    },
+    {
+      "epoch": 0.09461177725215679,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.07899426941954e-06,
+      "loss": 0.6675,
+      "step": 3460
+    },
+    {
+      "epoch": 0.0948852216950821,
+      "grad_norm": 2.203125,
+      "learning_rate": 9.076252364892654e-06,
+      "loss": 0.6616,
+      "step": 3470
+    },
+    {
+      "epoch": 0.09515866613800741,
+      "grad_norm": 3.21875,
+      "learning_rate": 9.07351046036577e-06,
+      "loss": 0.7121,
+      "step": 3480
+    },
+    {
+      "epoch": 0.09543211058093272,
+      "grad_norm": 2.90625,
+      "learning_rate": 9.070768555838887e-06,
+      "loss": 0.6962,
+      "step": 3490
+    },
+    {
+      "epoch": 0.09570555502385802,
+      "grad_norm": 2.4375,
+      "learning_rate": 9.068026651312002e-06,
+      "loss": 0.6692,
+      "step": 3500
+    },
+    {
+      "epoch": 0.09597899946678333,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.065284746785118e-06,
+      "loss": 0.7412,
+      "step": 3510
+    },
+    {
+      "epoch": 0.09625244390970865,
+      "grad_norm": 2.828125,
+      "learning_rate": 9.062542842258233e-06,
+      "loss": 0.7908,
+      "step": 3520
+    },
+    {
+      "epoch": 0.09652588835263395,
+      "grad_norm": 2.5,
+      "learning_rate": 9.059800937731349e-06,
+      "loss": 0.7482,
+      "step": 3530
+    },
+    {
+      "epoch": 0.09679933279555926,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.057059033204465e-06,
+      "loss": 0.7422,
+      "step": 3540
+    },
+    {
+      "epoch": 0.09707277723848456,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.05431712867758e-06,
+      "loss": 0.7405,
+      "step": 3550
+    },
+    {
+      "epoch": 0.09734622168140988,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.051575224150696e-06,
+      "loss": 0.7859,
+      "step": 3560
+    },
+    {
+      "epoch": 0.09761966612433519,
+      "grad_norm": 2.46875,
+      "learning_rate": 9.048833319623811e-06,
+      "loss": 0.684,
+      "step": 3570
+    },
+    {
+      "epoch": 0.0978931105672605,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.046091415096927e-06,
+      "loss": 0.7065,
+      "step": 3580
+    },
+    {
+      "epoch": 0.0981665550101858,
+      "grad_norm": 2.015625,
+      "learning_rate": 9.043349510570044e-06,
+      "loss": 0.7461,
+      "step": 3590
+    },
+    {
+      "epoch": 0.09843999945311112,
+      "grad_norm": 2.921875,
+      "learning_rate": 9.040607606043158e-06,
+      "loss": 0.7034,
+      "step": 3600
+    },
+    {
+      "epoch": 0.09871344389603642,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.037865701516275e-06,
+      "loss": 0.7345,
+      "step": 3610
+    },
+    {
+      "epoch": 0.09898688833896173,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.035123796989389e-06,
+      "loss": 0.6923,
+      "step": 3620
+    },
+    {
+      "epoch": 0.09926033278188703,
+      "grad_norm": 2.390625,
+      "learning_rate": 9.032381892462505e-06,
+      "loss": 0.6892,
+      "step": 3630
+    },
+    {
+      "epoch": 0.09953377722481235,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.029639987935622e-06,
+      "loss": 0.6434,
+      "step": 3640
+    },
+    {
+      "epoch": 0.09980722166773766,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.026898083408736e-06,
+      "loss": 0.6552,
+      "step": 3650
+    },
+    {
+      "epoch": 0.10008066611066296,
+      "grad_norm": 2.359375,
+      "learning_rate": 9.024156178881853e-06,
+      "loss": 0.7417,
+      "step": 3660
+    },
+    {
+      "epoch": 0.10035411055358827,
+      "grad_norm": 3.265625,
+      "learning_rate": 9.021414274354967e-06,
+      "loss": 0.6971,
+      "step": 3670
+    },
+    {
+      "epoch": 0.10062755499651359,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.018672369828084e-06,
+      "loss": 0.772,
+      "step": 3680
+    },
+    {
+      "epoch": 0.1009009994394389,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.015930465301198e-06,
+      "loss": 0.6931,
+      "step": 3690
+    },
+    {
+      "epoch": 0.1011744438823642,
+      "grad_norm": 2.328125,
+      "learning_rate": 9.013188560774315e-06,
+      "loss": 0.7003,
+      "step": 3700
+    },
+    {
+      "epoch": 0.1014478883252895,
+      "grad_norm": 2.84375,
+      "learning_rate": 9.010446656247431e-06,
+      "loss": 0.7358,
+      "step": 3710
+    },
+    {
+      "epoch": 0.10172133276821482,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.007704751720546e-06,
+      "loss": 0.6683,
+      "step": 3720
+    },
+    {
+      "epoch": 0.10199477721114013,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.004962847193662e-06,
+      "loss": 0.7268,
+      "step": 3730
+    },
+    {
+      "epoch": 0.10226822165406543,
+      "grad_norm": 2.53125,
+      "learning_rate": 9.002220942666777e-06,
+      "loss": 0.7729,
+      "step": 3740
+    },
+    {
+      "epoch": 0.10254166609699074,
+      "grad_norm": 2.375,
+      "learning_rate": 8.999479038139893e-06,
+      "loss": 0.6931,
+      "step": 3750
+    },
+    {
+      "epoch": 0.10281511053991606,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.996737133613008e-06,
+      "loss": 0.6683,
+      "step": 3760
+    },
+    {
+      "epoch": 0.10308855498284136,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.993995229086124e-06,
+      "loss": 0.7224,
+      "step": 3770
+    },
+    {
+      "epoch": 0.10336199942576667,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.991253324559239e-06,
+      "loss": 0.6687,
+      "step": 3780
+    },
+    {
+      "epoch": 0.10363544386869197,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.988511420032355e-06,
+      "loss": 0.7198,
+      "step": 3790
+    },
+    {
+      "epoch": 0.10390888831161729,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.985769515505471e-06,
+      "loss": 0.7144,
+      "step": 3800
+    },
+    {
+      "epoch": 0.1041823327545426,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.983027610978586e-06,
+      "loss": 0.687,
+      "step": 3810
+    },
+    {
+      "epoch": 0.1044557771974679,
+      "grad_norm": 2.875,
+      "learning_rate": 8.980285706451702e-06,
+      "loss": 0.6806,
+      "step": 3820
+    },
+    {
+      "epoch": 0.10472922164039321,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.977543801924817e-06,
+      "loss": 0.7434,
+      "step": 3830
+    },
+    {
+      "epoch": 0.10500266608331853,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.974801897397933e-06,
+      "loss": 0.6502,
+      "step": 3840
+    },
+    {
+      "epoch": 0.10527611052624383,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.97205999287105e-06,
+      "loss": 0.7147,
+      "step": 3850
+    },
+    {
+      "epoch": 0.10554955496916914,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.969318088344164e-06,
+      "loss": 0.7403,
+      "step": 3860
+    },
+    {
+      "epoch": 0.10582299941209444,
+      "grad_norm": 3.953125,
+      "learning_rate": 8.96657618381728e-06,
+      "loss": 0.6041,
+      "step": 3870
+    },
+    {
+      "epoch": 0.10609644385501976,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.963834279290395e-06,
+      "loss": 0.7531,
+      "step": 3880
+    },
+    {
+      "epoch": 0.10636988829794507,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.961092374763511e-06,
+      "loss": 0.6987,
+      "step": 3890
+    },
+    {
+      "epoch": 0.10664333274087037,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.958350470236628e-06,
+      "loss": 0.7257,
+      "step": 3900
+    },
+    {
+      "epoch": 0.10691677718379568,
+      "grad_norm": 2.625,
+      "learning_rate": 8.955608565709742e-06,
+      "loss": 0.7666,
+      "step": 3910
+    },
+    {
+      "epoch": 0.107190221626721,
+      "grad_norm": 2.75,
+      "learning_rate": 8.952866661182859e-06,
+      "loss": 0.7229,
+      "step": 3920
+    },
+    {
+      "epoch": 0.1074636660696463,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.950124756655973e-06,
+      "loss": 0.6536,
+      "step": 3930
+    },
+    {
+      "epoch": 0.10773711051257161,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.94738285212909e-06,
+      "loss": 0.7457,
+      "step": 3940
+    },
+    {
+      "epoch": 0.10801055495549691,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.944640947602206e-06,
+      "loss": 0.6894,
+      "step": 3950
+    },
+    {
+      "epoch": 0.10828399939842223,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.94189904307532e-06,
+      "loss": 0.6934,
+      "step": 3960
+    },
+    {
+      "epoch": 0.10855744384134754,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.939157138548437e-06,
+      "loss": 0.7197,
+      "step": 3970
+    },
+    {
+      "epoch": 0.10883088828427284,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.936415234021553e-06,
+      "loss": 0.6729,
+      "step": 3980
+    },
+    {
+      "epoch": 0.10910433272719815,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.933673329494668e-06,
+      "loss": 0.7135,
+      "step": 3990
+    },
+    {
+      "epoch": 0.10937777717012347,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.930931424967784e-06,
+      "loss": 0.7233,
+      "step": 4000
+    },
+    {
+      "epoch": 0.10965122161304877,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.928189520440899e-06,
+      "loss": 0.7577,
+      "step": 4010
+    },
+    {
+      "epoch": 0.10992466605597408,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.925447615914015e-06,
+      "loss": 0.7103,
+      "step": 4020
+    },
+    {
+      "epoch": 0.11019811049889938,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.922705711387132e-06,
+      "loss": 0.7714,
+      "step": 4030
+    },
+    {
+      "epoch": 0.1104715549418247,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.919963806860246e-06,
+      "loss": 0.7234,
+      "step": 4040
+    },
+    {
+      "epoch": 0.11074499938475,
+      "grad_norm": 2.25,
+      "learning_rate": 8.917221902333362e-06,
+      "loss": 0.6417,
+      "step": 4050
+    },
+    {
+      "epoch": 0.11101844382767531,
+      "grad_norm": 2.375,
+      "learning_rate": 8.914479997806477e-06,
+      "loss": 0.7264,
+      "step": 4060
+    },
+    {
+      "epoch": 0.11129188827060062,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.911738093279593e-06,
+      "loss": 0.6653,
+      "step": 4070
+    },
+    {
+      "epoch": 0.11156533271352594,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.908996188752708e-06,
+      "loss": 0.6545,
+      "step": 4080
+    },
+    {
+      "epoch": 0.11183877715645124,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.906254284225824e-06,
+      "loss": 0.7,
+      "step": 4090
+    },
+    {
+      "epoch": 0.11211222159937655,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.903512379698939e-06,
+      "loss": 0.638,
+      "step": 4100
+    },
+    {
+      "epoch": 0.11238566604230185,
+      "grad_norm": 2.5,
+      "learning_rate": 8.900770475172055e-06,
+      "loss": 0.7191,
+      "step": 4110
+    },
+    {
+      "epoch": 0.11265911048522717,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.89802857064517e-06,
+      "loss": 0.7591,
+      "step": 4120
+    },
+    {
+      "epoch": 0.11293255492815248,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.895286666118286e-06,
+      "loss": 0.7115,
+      "step": 4130
+    },
+    {
+      "epoch": 0.11320599937107778,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.892544761591401e-06,
+      "loss": 0.7683,
+      "step": 4140
+    },
+    {
+      "epoch": 0.11347944381400309,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.889802857064517e-06,
+      "loss": 0.6935,
+      "step": 4150
+    },
+    {
+      "epoch": 0.11375288825692839,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.887060952537634e-06,
+      "loss": 0.7217,
+      "step": 4160
+    },
+    {
+      "epoch": 0.11402633269985371,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.884319048010748e-06,
+      "loss": 0.7102,
+      "step": 4170
+    },
+    {
+      "epoch": 0.11429977714277902,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.881577143483865e-06,
+      "loss": 0.7157,
+      "step": 4180
+    },
+    {
+      "epoch": 0.11457322158570432,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.87883523895698e-06,
+      "loss": 0.8331,
+      "step": 4190
+    },
+    {
+      "epoch": 0.11484666602862963,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.876093334430096e-06,
+      "loss": 0.6918,
+      "step": 4200
+    },
+    {
+      "epoch": 0.11512011047155495,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.873351429903212e-06,
+      "loss": 0.7676,
+      "step": 4210
+    },
+    {
+      "epoch": 0.11539355491448025,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.870609525376327e-06,
+      "loss": 0.7235,
+      "step": 4220
+    },
+    {
+      "epoch": 0.11566699935740556,
+      "grad_norm": 2.5,
+      "learning_rate": 8.867867620849443e-06,
+      "loss": 0.6853,
+      "step": 4230
+    },
+    {
+      "epoch": 0.11594044380033086,
+      "grad_norm": 2.625,
+      "learning_rate": 8.865125716322558e-06,
+      "loss": 0.6729,
+      "step": 4240
+    },
+    {
+      "epoch": 0.11621388824325618,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.862383811795674e-06,
+      "loss": 0.7693,
+      "step": 4250
+    },
+    {
+      "epoch": 0.11648733268618149,
+      "grad_norm": 2.109375,
+      "learning_rate": 8.85964190726879e-06,
+      "loss": 0.7707,
+      "step": 4260
+    },
+    {
+      "epoch": 0.11676077712910679,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.856900002741905e-06,
+      "loss": 0.7274,
+      "step": 4270
+    },
+    {
+      "epoch": 0.1170342215720321,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.854158098215021e-06,
+      "loss": 0.6655,
+      "step": 4280
+    },
+    {
+      "epoch": 0.11730766601495742,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.851416193688136e-06,
+      "loss": 0.7099,
+      "step": 4290
+    },
+    {
+      "epoch": 0.11758111045788272,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.848674289161252e-06,
+      "loss": 0.7477,
+      "step": 4300
+    },
+    {
+      "epoch": 0.11785455490080803,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.845932384634368e-06,
+      "loss": 0.6887,
+      "step": 4310
+    },
+    {
+      "epoch": 0.11812799934373333,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.843190480107483e-06,
+      "loss": 0.6919,
+      "step": 4320
+    },
+    {
+      "epoch": 0.11840144378665865,
+      "grad_norm": 2.25,
+      "learning_rate": 8.8404485755806e-06,
+      "loss": 0.6508,
+      "step": 4330
+    },
+    {
+      "epoch": 0.11867488822958396,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.837706671053716e-06,
+      "loss": 0.6674,
+      "step": 4340
+    },
+    {
+      "epoch": 0.11894833267250926,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.83496476652683e-06,
+      "loss": 0.7467,
+      "step": 4350
+    },
+    {
+      "epoch": 0.11922177711543457,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.832222861999947e-06,
+      "loss": 0.7059,
+      "step": 4360
+    },
+    {
+      "epoch": 0.11949522155835988,
+      "grad_norm": 2.5,
+      "learning_rate": 8.829480957473061e-06,
+      "loss": 0.6914,
+      "step": 4370
+    },
+    {
+      "epoch": 0.11976866600128519,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.826739052946178e-06,
+      "loss": 0.7489,
+      "step": 4380
+    },
+    {
+      "epoch": 0.1200421104442105,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.823997148419294e-06,
+      "loss": 0.6841,
+      "step": 4390
+    },
+    {
+      "epoch": 0.1203155548871358,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.821255243892409e-06,
+      "loss": 0.6628,
+      "step": 4400
+    },
+    {
+      "epoch": 0.12058899933006112,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.818513339365525e-06,
+      "loss": 0.7411,
+      "step": 4410
+    },
+    {
+      "epoch": 0.12086244377298642,
+      "grad_norm": 2.09375,
+      "learning_rate": 8.81577143483864e-06,
+      "loss": 0.6633,
+      "step": 4420
+    },
+    {
+      "epoch": 0.12113588821591173,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.813029530311756e-06,
+      "loss": 0.7844,
+      "step": 4430
+    },
+    {
+      "epoch": 0.12140933265883704,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.81028762578487e-06,
+      "loss": 0.7023,
+      "step": 4440
+    },
+    {
+      "epoch": 0.12168277710176235,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.807545721257987e-06,
+      "loss": 0.6457,
+      "step": 4450
+    },
+    {
+      "epoch": 0.12195622154468766,
+      "grad_norm": 2.625,
+      "learning_rate": 8.804803816731101e-06,
+      "loss": 0.725,
+      "step": 4460
+    },
+    {
+      "epoch": 0.12222966598761296,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.802061912204218e-06,
+      "loss": 0.7752,
+      "step": 4470
+    },
+    {
+      "epoch": 0.12250311043053827,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.799320007677332e-06,
+      "loss": 0.7277,
+      "step": 4480
+    },
+    {
+      "epoch": 0.12277655487346359,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.796578103150449e-06,
+      "loss": 0.6553,
+      "step": 4490
+    },
+    {
+      "epoch": 0.1230499993163889,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.793836198623565e-06,
+      "loss": 0.7371,
+      "step": 4500
+    },
+    {
+      "epoch": 0.1233234437593142,
+      "grad_norm": 2.5,
+      "learning_rate": 8.79109429409668e-06,
+      "loss": 0.6542,
+      "step": 4510
+    },
+    {
+      "epoch": 0.1235968882022395,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.788352389569796e-06,
+      "loss": 0.7508,
+      "step": 4520
+    },
+    {
+      "epoch": 0.12387033264516482,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.78561048504291e-06,
+      "loss": 0.7154,
+      "step": 4530
+    },
+    {
+      "epoch": 0.12414377708809013,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.782868580516027e-06,
+      "loss": 0.6304,
+      "step": 4540
+    },
+    {
+      "epoch": 0.12441722153101543,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.780126675989142e-06,
+      "loss": 0.7055,
+      "step": 4550
+    },
+    {
+      "epoch": 0.12469066597394074,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.777384771462258e-06,
+      "loss": 0.7053,
+      "step": 4560
+    },
+    {
+      "epoch": 0.12496411041686606,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.774642866935374e-06,
+      "loss": 0.6757,
+      "step": 4570
+    },
+    {
+      "epoch": 0.12523755485979135,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.771900962408489e-06,
+      "loss": 0.659,
+      "step": 4580
+    },
+    {
+      "epoch": 0.12551099930271667,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.769159057881605e-06,
+      "loss": 0.6657,
+      "step": 4590
+    },
+    {
+      "epoch": 0.125784443745642,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.76641715335472e-06,
+      "loss": 0.6331,
+      "step": 4600
+    },
+    {
+      "epoch": 0.12605788818856728,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.763675248827836e-06,
+      "loss": 0.7044,
+      "step": 4610
+    },
+    {
+      "epoch": 0.1263313326314926,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.760933344300953e-06,
+      "loss": 0.6441,
+      "step": 4620
+    },
+    {
+      "epoch": 0.12660477707441792,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.758191439774067e-06,
+      "loss": 0.6773,
+      "step": 4630
+    },
+    {
+      "epoch": 0.1268782215173432,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.755449535247184e-06,
+      "loss": 0.7616,
+      "step": 4640
+    },
+    {
+      "epoch": 0.12715166596026853,
+      "grad_norm": 3.125,
+      "learning_rate": 8.752707630720298e-06,
+      "loss": 0.7642,
+      "step": 4650
+    },
+    {
+      "epoch": 0.12742511040319382,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.749965726193415e-06,
+      "loss": 0.7509,
+      "step": 4660
+    },
+    {
+      "epoch": 0.12769855484611914,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.747223821666531e-06,
+      "loss": 0.6931,
+      "step": 4670
+    },
+    {
+      "epoch": 0.12797199928904446,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.744481917139645e-06,
+      "loss": 0.7423,
+      "step": 4680
+    },
+    {
+      "epoch": 0.12824544373196975,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.741740012612762e-06,
+      "loss": 0.7211,
+      "step": 4690
+    },
+    {
+      "epoch": 0.12851888817489507,
+      "grad_norm": 2.625,
+      "learning_rate": 8.738998108085878e-06,
+      "loss": 0.6715,
+      "step": 4700
+    },
+    {
+      "epoch": 0.1287923326178204,
+      "grad_norm": 2.625,
+      "learning_rate": 8.736256203558993e-06,
+      "loss": 0.6229,
+      "step": 4710
+    },
+    {
+      "epoch": 0.12906577706074568,
+      "grad_norm": 2.625,
+      "learning_rate": 8.733514299032109e-06,
+      "loss": 0.6676,
+      "step": 4720
+    },
+    {
+      "epoch": 0.129339221503671,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.730772394505224e-06,
+      "loss": 0.7498,
+      "step": 4730
+    },
+    {
+      "epoch": 0.1296126659465963,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.72803048997834e-06,
+      "loss": 0.7396,
+      "step": 4740
+    },
+    {
+      "epoch": 0.1298861103895216,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.725288585451456e-06,
+      "loss": 0.7093,
+      "step": 4750
+    },
+    {
+      "epoch": 0.13015955483244693,
+      "grad_norm": 3.0,
+      "learning_rate": 8.722546680924571e-06,
+      "loss": 0.8104,
+      "step": 4760
+    },
+    {
+      "epoch": 0.13043299927537222,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.719804776397687e-06,
+      "loss": 0.6919,
+      "step": 4770
+    },
+    {
+      "epoch": 0.13070644371829754,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.717062871870802e-06,
+      "loss": 0.7879,
+      "step": 4780
+    },
+    {
+      "epoch": 0.13097988816122286,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.714320967343918e-06,
+      "loss": 0.7091,
+      "step": 4790
+    },
+    {
+      "epoch": 0.13125333260414815,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.711579062817033e-06,
+      "loss": 0.7491,
+      "step": 4800
+    },
+    {
+      "epoch": 0.13152677704707347,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.70883715829015e-06,
+      "loss": 0.6757,
+      "step": 4810
+    },
+    {
+      "epoch": 0.13180022148999876,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.706095253763266e-06,
+      "loss": 0.7679,
+      "step": 4820
+    },
+    {
+      "epoch": 0.13207366593292408,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.70335334923638e-06,
+      "loss": 0.7233,
+      "step": 4830
+    },
+    {
+      "epoch": 0.1323471103758494,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.700611444709497e-06,
+      "loss": 0.6943,
+      "step": 4840
+    },
+    {
+      "epoch": 0.1326205548187747,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.697869540182611e-06,
+      "loss": 0.6535,
+      "step": 4850
+    },
+    {
+      "epoch": 0.1328939992617,
+      "grad_norm": 2.875,
+      "learning_rate": 8.695127635655728e-06,
+      "loss": 0.7324,
+      "step": 4860
+    },
+    {
+      "epoch": 0.1331674437046253,
+      "grad_norm": 2.875,
+      "learning_rate": 8.692385731128842e-06,
+      "loss": 0.7232,
+      "step": 4870
+    },
+    {
+      "epoch": 0.13344088814755062,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.689643826601959e-06,
+      "loss": 0.6952,
+      "step": 4880
+    },
+    {
+      "epoch": 0.13371433259047594,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.686901922075073e-06,
+      "loss": 0.7213,
+      "step": 4890
+    },
+    {
+      "epoch": 0.13398777703340123,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.68416001754819e-06,
+      "loss": 0.7087,
+      "step": 4900
+    },
+    {
+      "epoch": 0.13426122147632655,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.681418113021304e-06,
+      "loss": 0.6853,
+      "step": 4910
+    },
+    {
+      "epoch": 0.13453466591925187,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.67867620849442e-06,
+      "loss": 0.6944,
+      "step": 4920
+    },
+    {
+      "epoch": 0.13480811036217716,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.675934303967537e-06,
+      "loss": 0.719,
+      "step": 4930
+    },
+    {
+      "epoch": 0.13508155480510248,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.673192399440651e-06,
+      "loss": 0.7218,
+      "step": 4940
+    },
+    {
+      "epoch": 0.13535499924802777,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.670450494913768e-06,
+      "loss": 0.7447,
+      "step": 4950
+    },
+    {
+      "epoch": 0.1356284436909531,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.667708590386882e-06,
+      "loss": 0.7047,
+      "step": 4960
+    },
+    {
+      "epoch": 0.1359018881338784,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.664966685859999e-06,
+      "loss": 0.6646,
+      "step": 4970
+    },
+    {
+      "epoch": 0.1361753325768037,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.662224781333115e-06,
+      "loss": 0.7471,
+      "step": 4980
+    },
+    {
+      "epoch": 0.13644877701972902,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.65948287680623e-06,
+      "loss": 0.7759,
+      "step": 4990
+    },
+    {
+      "epoch": 0.13672222146265434,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.656740972279346e-06,
+      "loss": 0.7191,
+      "step": 5000
+    },
+    {
+      "epoch": 0.13699566590557963,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.65399906775246e-06,
+      "loss": 0.7392,
+      "step": 5010
+    },
+    {
+      "epoch": 0.13726911034850495,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.651257163225577e-06,
+      "loss": 0.7191,
+      "step": 5020
+    },
+    {
+      "epoch": 0.13754255479143024,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.648515258698693e-06,
+      "loss": 0.7622,
+      "step": 5030
+    },
+    {
+      "epoch": 0.13781599923435556,
+      "grad_norm": 2.75,
+      "learning_rate": 8.645773354171808e-06,
+      "loss": 0.7585,
+      "step": 5040
+    },
+    {
+      "epoch": 0.13808944367728088,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.643031449644924e-06,
+      "loss": 0.6846,
+      "step": 5050
+    },
+    {
+      "epoch": 0.13836288812020617,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.64028954511804e-06,
+      "loss": 0.7397,
+      "step": 5060
+    },
+    {
+      "epoch": 0.1386363325631315,
+      "grad_norm": 2.875,
+      "learning_rate": 8.637547640591155e-06,
+      "loss": 0.6789,
+      "step": 5070
+    },
+    {
+      "epoch": 0.1389097770060568,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.634805736064272e-06,
+      "loss": 0.7214,
+      "step": 5080
+    },
+    {
+      "epoch": 0.1391832214489821,
+      "grad_norm": 2.625,
+      "learning_rate": 8.632063831537386e-06,
+      "loss": 0.7123,
+      "step": 5090
+    },
+    {
+      "epoch": 0.13945666589190742,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.629321927010503e-06,
+      "loss": 0.6605,
+      "step": 5100
+    },
+    {
+      "epoch": 0.1397301103348327,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.626580022483619e-06,
+      "loss": 0.7382,
+      "step": 5110
+    },
+    {
+      "epoch": 0.14000355477775803,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.623838117956733e-06,
+      "loss": 0.7103,
+      "step": 5120
+    },
+    {
+      "epoch": 0.14027699922068335,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.62109621342985e-06,
+      "loss": 0.6646,
+      "step": 5130
+    },
+    {
+      "epoch": 0.14055044366360864,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.618354308902964e-06,
+      "loss": 0.6872,
+      "step": 5140
+    },
+    {
+      "epoch": 0.14082388810653396,
+      "grad_norm": 3.078125,
+      "learning_rate": 8.61561240437608e-06,
+      "loss": 0.708,
+      "step": 5150
+    },
+    {
+      "epoch": 0.14109733254945928,
+      "grad_norm": 2.75,
+      "learning_rate": 8.612870499849197e-06,
+      "loss": 0.6966,
+      "step": 5160
+    },
+    {
+      "epoch": 0.14137077699238457,
+      "grad_norm": 3.25,
+      "learning_rate": 8.610128595322312e-06,
+      "loss": 0.7286,
+      "step": 5170
+    },
+    {
+      "epoch": 0.14164422143530989,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.607386690795428e-06,
+      "loss": 0.6637,
+      "step": 5180
+    },
+    {
+      "epoch": 0.14191766587823518,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.604644786268543e-06,
+      "loss": 0.7338,
+      "step": 5190
+    },
+    {
+      "epoch": 0.1421911103211605,
+      "grad_norm": 3.34375,
+      "learning_rate": 8.601902881741659e-06,
+      "loss": 0.7104,
+      "step": 5200
+    },
+    {
+      "epoch": 0.14246455476408582,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.599160977214774e-06,
+      "loss": 0.7038,
+      "step": 5210
+    },
+    {
+      "epoch": 0.1427379992070111,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.59641907268789e-06,
+      "loss": 0.7704,
+      "step": 5220
+    },
+    {
+      "epoch": 0.14301144364993643,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.593677168161005e-06,
+      "loss": 0.7014,
+      "step": 5230
+    },
+    {
+      "epoch": 0.14328488809286175,
+      "grad_norm": 2.5,
+      "learning_rate": 8.590935263634121e-06,
+      "loss": 0.7237,
+      "step": 5240
+    },
+    {
+      "epoch": 0.14355833253578704,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.588193359107236e-06,
+      "loss": 0.7022,
+      "step": 5250
+    },
+    {
+      "epoch": 0.14383177697871236,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.585451454580352e-06,
+      "loss": 0.694,
+      "step": 5260
+    },
+    {
+      "epoch": 0.14410522142163765,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.582709550053468e-06,
+      "loss": 0.7499,
+      "step": 5270
+    },
+    {
+      "epoch": 0.14437866586456297,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.579967645526583e-06,
+      "loss": 0.6762,
+      "step": 5280
+    },
+    {
+      "epoch": 0.14465211030748829,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.5772257409997e-06,
+      "loss": 0.7721,
+      "step": 5290
+    },
+    {
+      "epoch": 0.14492555475041358,
+      "grad_norm": 2.625,
+      "learning_rate": 8.574483836472814e-06,
+      "loss": 0.7793,
+      "step": 5300
+    },
+    {
+      "epoch": 0.1451989991933389,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.57174193194593e-06,
+      "loss": 0.6799,
+      "step": 5310
+    },
+    {
+      "epoch": 0.14547244363626421,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.569000027419045e-06,
+      "loss": 0.6888,
+      "step": 5320
+    },
+    {
+      "epoch": 0.1457458880791895,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.566258122892161e-06,
+      "loss": 0.7632,
+      "step": 5330
+    },
+    {
+      "epoch": 0.14601933252211483,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.563516218365277e-06,
+      "loss": 0.675,
+      "step": 5340
+    },
+    {
+      "epoch": 0.14629277696504012,
+      "grad_norm": 2.046875,
+      "learning_rate": 8.560774313838392e-06,
+      "loss": 0.6194,
+      "step": 5350
+    },
+    {
+      "epoch": 0.14656622140796544,
+      "grad_norm": 2.75,
+      "learning_rate": 8.558032409311508e-06,
+      "loss": 0.7368,
+      "step": 5360
+    },
+    {
+      "epoch": 0.14683966585089075,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.555290504784625e-06,
+      "loss": 0.619,
+      "step": 5370
+    },
+    {
+      "epoch": 0.14711311029381605,
+      "grad_norm": 2.625,
+      "learning_rate": 8.55254860025774e-06,
+      "loss": 0.6897,
+      "step": 5380
+    },
+    {
+      "epoch": 0.14738655473674137,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.549806695730856e-06,
+      "loss": 0.6628,
+      "step": 5390
+    },
+    {
+      "epoch": 0.14765999917966668,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.54706479120397e-06,
+      "loss": 0.66,
+      "step": 5400
+    },
+    {
+      "epoch": 0.14793344362259198,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.544322886677087e-06,
+      "loss": 0.7292,
+      "step": 5410
+    },
+    {
+      "epoch": 0.1482068880655173,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.541580982150203e-06,
+      "loss": 0.6785,
+      "step": 5420
+    },
+    {
+      "epoch": 0.1484803325084426,
+      "grad_norm": 2.375,
+      "learning_rate": 8.538839077623318e-06,
+      "loss": 0.7086,
+      "step": 5430
+    },
+    {
+      "epoch": 0.1487537769513679,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.536097173096434e-06,
+      "loss": 0.6942,
+      "step": 5440
+    },
+    {
+      "epoch": 0.14902722139429322,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.533355268569549e-06,
+      "loss": 0.6452,
+      "step": 5450
+    },
+    {
+      "epoch": 0.14930066583721852,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.530613364042665e-06,
+      "loss": 0.6932,
+      "step": 5460
+    },
+    {
+      "epoch": 0.14957411028014383,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.527871459515781e-06,
+      "loss": 0.6652,
+      "step": 5470
+    },
+    {
+      "epoch": 0.14984755472306915,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.525129554988896e-06,
+      "loss": 0.7413,
+      "step": 5480
+    },
+    {
+      "epoch": 0.15012099916599445,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.522387650462012e-06,
+      "loss": 0.7156,
+      "step": 5490
+    },
+    {
+      "epoch": 0.15039444360891976,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.519645745935127e-06,
+      "loss": 0.694,
+      "step": 5500
+    },
+    {
+      "epoch": 0.15066788805184506,
+      "grad_norm": 3.125,
+      "learning_rate": 8.516903841408243e-06,
+      "loss": 0.7845,
+      "step": 5510
+    },
+    {
+      "epoch": 0.15094133249477037,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.51416193688136e-06,
+      "loss": 0.6495,
+      "step": 5520
+    },
+    {
+      "epoch": 0.1512147769376957,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.511420032354474e-06,
+      "loss": 0.7663,
+      "step": 5530
+    },
+    {
+      "epoch": 0.15148822138062099,
+      "grad_norm": 2.625,
+      "learning_rate": 8.50867812782759e-06,
+      "loss": 0.6831,
+      "step": 5540
+    },
+    {
+      "epoch": 0.1517616658235463,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.505936223300705e-06,
+      "loss": 0.6493,
+      "step": 5550
+    },
+    {
+      "epoch": 0.1520351102664716,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.503194318773821e-06,
+      "loss": 0.6528,
+      "step": 5560
+    },
+    {
+      "epoch": 0.15230855470939691,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.500452414246936e-06,
+      "loss": 0.7426,
+      "step": 5570
+    },
+    {
+      "epoch": 0.15258199915232223,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.497710509720052e-06,
+      "loss": 0.715,
+      "step": 5580
+    },
+    {
+      "epoch": 0.15285544359524753,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.494968605193169e-06,
+      "loss": 0.712,
+      "step": 5590
+    },
+    {
+      "epoch": 0.15312888803817284,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.492226700666283e-06,
+      "loss": 0.6856,
+      "step": 5600
+    },
+    {
+      "epoch": 0.15340233248109816,
+      "grad_norm": 2.453125,
+      "learning_rate": 8.4894847961394e-06,
+      "loss": 0.6684,
+      "step": 5610
+    },
+    {
+      "epoch": 0.15367577692402346,
+      "grad_norm": 2.328125,
+      "learning_rate": 8.486742891612514e-06,
+      "loss": 0.7006,
+      "step": 5620
+    },
+    {
+      "epoch": 0.15394922136694877,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.48400098708563e-06,
+      "loss": 0.6484,
+      "step": 5630
+    },
+    {
+      "epoch": 0.15422266580987407,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.481259082558745e-06,
+      "loss": 0.6751,
+      "step": 5640
+    },
+    {
+      "epoch": 0.15449611025279938,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.478517178031862e-06,
+      "loss": 0.6373,
+      "step": 5650
+    },
+    {
+      "epoch": 0.1547695546957247,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.475775273504976e-06,
+      "loss": 0.7138,
+      "step": 5660
+    },
+    {
+      "epoch": 0.15504299913865,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.473033368978093e-06,
+      "loss": 0.7317,
+      "step": 5670
+    },
+    {
+      "epoch": 0.15531644358157531,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.470291464451207e-06,
+      "loss": 0.7045,
+      "step": 5680
+    },
+    {
+      "epoch": 0.15558988802450063,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.467549559924324e-06,
+      "loss": 0.7913,
+      "step": 5690
+    },
+    {
+      "epoch": 0.15586333246742592,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.46480765539744e-06,
+      "loss": 0.726,
+      "step": 5700
+    },
+    {
+      "epoch": 0.15613677691035124,
+      "grad_norm": 2.203125,
+      "learning_rate": 8.462065750870555e-06,
+      "loss": 0.7121,
+      "step": 5710
+    },
+    {
+      "epoch": 0.15641022135327654,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.459323846343671e-06,
+      "loss": 0.7186,
+      "step": 5720
+    },
+    {
+      "epoch": 0.15668366579620185,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.456581941816787e-06,
+      "loss": 0.6472,
+      "step": 5730
+    },
+    {
+      "epoch": 0.15695711023912717,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.453840037289902e-06,
+      "loss": 0.7427,
+      "step": 5740
+    },
+    {
+      "epoch": 0.15723055468205246,
+      "grad_norm": 2.84375,
+      "learning_rate": 8.451098132763018e-06,
+      "loss": 0.6463,
+      "step": 5750
+    },
+    {
+      "epoch": 0.15750399912497778,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.448356228236133e-06,
+      "loss": 0.7166,
+      "step": 5760
+    },
+    {
+      "epoch": 0.1577774435679031,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.445614323709249e-06,
+      "loss": 0.7253,
+      "step": 5770
+    },
+    {
+      "epoch": 0.1580508880108284,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.442872419182365e-06,
+      "loss": 0.7311,
+      "step": 5780
+    },
+    {
+      "epoch": 0.1583243324537537,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.44013051465548e-06,
+      "loss": 0.7085,
+      "step": 5790
+    },
+    {
+      "epoch": 0.158597776896679,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.437388610128596e-06,
+      "loss": 0.6869,
+      "step": 5800
+    },
+    {
+      "epoch": 0.15887122133960432,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.434646705601711e-06,
+      "loss": 0.7748,
+      "step": 5810
+    },
+    {
+      "epoch": 0.15914466578252964,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.431904801074827e-06,
+      "loss": 0.741,
+      "step": 5820
+    },
+    {
+      "epoch": 0.15941811022545493,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.429162896547944e-06,
+      "loss": 0.6964,
+      "step": 5830
+    },
+    {
+      "epoch": 0.15969155466838025,
+      "grad_norm": 3.0,
+      "learning_rate": 8.426420992021058e-06,
+      "loss": 0.7194,
+      "step": 5840
+    },
+    {
+      "epoch": 0.15996499911130557,
+      "grad_norm": 2.875,
+      "learning_rate": 8.423679087494175e-06,
+      "loss": 0.7095,
+      "step": 5850
+    },
+    {
+      "epoch": 0.16023844355423086,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.42093718296729e-06,
+      "loss": 0.7354,
+      "step": 5860
+    },
+    {
+      "epoch": 0.16051188799715618,
+      "grad_norm": 2.34375,
+      "learning_rate": 8.418195278440406e-06,
+      "loss": 0.6978,
+      "step": 5870
+    },
+    {
+      "epoch": 0.16078533244008147,
+      "grad_norm": 2.625,
+      "learning_rate": 8.415453373913522e-06,
+      "loss": 0.6818,
+      "step": 5880
+    },
+    {
+      "epoch": 0.1610587768830068,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.412711469386637e-06,
+      "loss": 0.6714,
+      "step": 5890
+    },
+    {
+      "epoch": 0.1613322213259321,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.409969564859753e-06,
+      "loss": 0.6876,
+      "step": 5900
+    },
+    {
+      "epoch": 0.1616056657688574,
+      "grad_norm": 2.875,
+      "learning_rate": 8.40722766033287e-06,
+      "loss": 0.7288,
+      "step": 5910
+    },
+    {
+      "epoch": 0.16187911021178272,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.404485755805984e-06,
+      "loss": 0.752,
+      "step": 5920
+    },
+    {
+      "epoch": 0.16215255465470804,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.4017438512791e-06,
+      "loss": 0.6885,
+      "step": 5930
+    },
+    {
+      "epoch": 0.16242599909763333,
+      "grad_norm": 2.21875,
+      "learning_rate": 8.399001946752215e-06,
+      "loss": 0.6508,
+      "step": 5940
+    },
+    {
+      "epoch": 0.16269944354055865,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.396260042225331e-06,
+      "loss": 0.7862,
+      "step": 5950
+    },
+    {
+      "epoch": 0.16297288798348394,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.393518137698446e-06,
+      "loss": 0.7009,
+      "step": 5960
+    },
+    {
+      "epoch": 0.16324633242640926,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.390776233171562e-06,
+      "loss": 0.7271,
+      "step": 5970
+    },
+    {
+      "epoch": 0.16351977686933458,
+      "grad_norm": 2.625,
+      "learning_rate": 8.388034328644677e-06,
+      "loss": 0.6894,
+      "step": 5980
+    },
+    {
+      "epoch": 0.16379322131225987,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.385292424117793e-06,
+      "loss": 0.7306,
+      "step": 5990
+    },
+    {
+      "epoch": 0.1640666657551852,
+      "grad_norm": 2.875,
+      "learning_rate": 8.382550519590908e-06,
+      "loss": 0.8003,
+      "step": 6000
+    },
+    {
+      "epoch": 0.1643401101981105,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.379808615064024e-06,
+      "loss": 0.7013,
+      "step": 6010
+    },
+    {
+      "epoch": 0.1646135546410358,
+      "grad_norm": 2.5,
+      "learning_rate": 8.377066710537139e-06,
+      "loss": 0.6948,
+      "step": 6020
+    },
+    {
+      "epoch": 0.16488699908396112,
+      "grad_norm": 2.1875,
+      "learning_rate": 8.374324806010255e-06,
+      "loss": 0.6641,
+      "step": 6030
+    },
+    {
+      "epoch": 0.1651604435268864,
+      "grad_norm": 2.875,
+      "learning_rate": 8.37158290148337e-06,
+      "loss": 0.73,
+      "step": 6040
+    },
+    {
+      "epoch": 0.16543388796981173,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.368840996956486e-06,
+      "loss": 0.6809,
+      "step": 6050
+    },
+    {
+      "epoch": 0.16570733241273705,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.366099092429602e-06,
+      "loss": 0.6309,
+      "step": 6060
+    },
+    {
+      "epoch": 0.16598077685566234,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.363357187902717e-06,
+      "loss": 0.6947,
+      "step": 6070
+    },
+    {
+      "epoch": 0.16625422129858766,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.360615283375833e-06,
+      "loss": 0.648,
+      "step": 6080
+    },
+    {
+      "epoch": 0.16652766574151298,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.35787337884895e-06,
+      "loss": 0.7386,
+      "step": 6090
+    },
+    {
+      "epoch": 0.16680111018443827,
+      "grad_norm": 2.375,
+      "learning_rate": 8.355131474322064e-06,
+      "loss": 0.7219,
+      "step": 6100
+    },
+    {
+      "epoch": 0.1670745546273636,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.35238956979518e-06,
+      "loss": 0.7803,
+      "step": 6110
+    },
+    {
+      "epoch": 0.16734799907028888,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.349647665268295e-06,
+      "loss": 0.7205,
+      "step": 6120
+    },
+    {
+      "epoch": 0.1676214435132142,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.346905760741412e-06,
+      "loss": 0.6415,
+      "step": 6130
+    },
+    {
+      "epoch": 0.16789488795613952,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.344163856214528e-06,
+      "loss": 0.692,
+      "step": 6140
+    },
+    {
+      "epoch": 0.1681683323990648,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.341421951687643e-06,
+      "loss": 0.6347,
+      "step": 6150
+    },
+    {
+      "epoch": 0.16844177684199013,
+      "grad_norm": 2.625,
+      "learning_rate": 8.338680047160759e-06,
+      "loss": 0.7345,
+      "step": 6160
+    },
+    {
+      "epoch": 0.16871522128491545,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.335938142633873e-06,
+      "loss": 0.6842,
+      "step": 6170
+    },
+    {
+      "epoch": 0.16898866572784074,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.33319623810699e-06,
+      "loss": 0.66,
+      "step": 6180
+    },
+    {
+      "epoch": 0.16926211017076606,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.330454333580106e-06,
+      "loss": 0.717,
+      "step": 6190
+    },
+    {
+      "epoch": 0.16953555461369135,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.32771242905322e-06,
+      "loss": 0.7247,
+      "step": 6200
+    },
+    {
+      "epoch": 0.16980899905661667,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.324970524526337e-06,
+      "loss": 0.7192,
+      "step": 6210
+    },
+    {
+      "epoch": 0.170082443499542,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.322228619999452e-06,
+      "loss": 0.6518,
+      "step": 6220
+    },
+    {
+      "epoch": 0.17035588794246728,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.319486715472568e-06,
+      "loss": 0.7406,
+      "step": 6230
+    },
+    {
+      "epoch": 0.1706293323853926,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.316744810945684e-06,
+      "loss": 0.7224,
+      "step": 6240
+    },
+    {
+      "epoch": 0.1709027768283179,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.314002906418799e-06,
+      "loss": 0.8149,
+      "step": 6250
+    },
+    {
+      "epoch": 0.1711762212712432,
+      "grad_norm": 2.515625,
+      "learning_rate": 8.311261001891915e-06,
+      "loss": 0.7296,
+      "step": 6260
+    },
+    {
+      "epoch": 0.17144966571416853,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.308519097365032e-06,
+      "loss": 0.7115,
+      "step": 6270
+    },
+    {
+      "epoch": 0.17172311015709382,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.305777192838146e-06,
+      "loss": 0.7331,
+      "step": 6280
+    },
+    {
+      "epoch": 0.17199655460001914,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.303035288311263e-06,
+      "loss": 0.6872,
+      "step": 6290
+    },
+    {
+      "epoch": 0.17226999904294446,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.300293383784377e-06,
+      "loss": 0.6936,
+      "step": 6300
+    },
+    {
+      "epoch": 0.17254344348586975,
+      "grad_norm": 2.296875,
+      "learning_rate": 8.297551479257494e-06,
+      "loss": 0.6532,
+      "step": 6310
+    },
+    {
+      "epoch": 0.17281688792879507,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.294809574730608e-06,
+      "loss": 0.764,
+      "step": 6320
+    },
+    {
+      "epoch": 0.17309033237172036,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.292067670203725e-06,
+      "loss": 0.6219,
+      "step": 6330
+    },
+    {
+      "epoch": 0.17336377681464568,
+      "grad_norm": 2.75,
+      "learning_rate": 8.28932576567684e-06,
+      "loss": 0.7028,
+      "step": 6340
+    },
+    {
+      "epoch": 0.173637221257571,
+      "grad_norm": 2.40625,
+      "learning_rate": 8.286583861149956e-06,
+      "loss": 0.6943,
+      "step": 6350
+    },
+    {
+      "epoch": 0.1739106657004963,
+      "grad_norm": 2.625,
+      "learning_rate": 8.28384195662307e-06,
+      "loss": 0.6714,
+      "step": 6360
+    },
+    {
+      "epoch": 0.1741841101434216,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.281100052096187e-06,
+      "loss": 0.7102,
+      "step": 6370
+    },
+    {
+      "epoch": 0.17445755458634693,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.278358147569303e-06,
+      "loss": 0.6785,
+      "step": 6380
+    },
+    {
+      "epoch": 0.17473099902927222,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.275616243042417e-06,
+      "loss": 0.7144,
+      "step": 6390
+    },
+    {
+      "epoch": 0.17500444347219754,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.272874338515534e-06,
+      "loss": 0.7297,
+      "step": 6400
+    },
+    {
+      "epoch": 0.17527788791512283,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.270132433988648e-06,
+      "loss": 0.6153,
+      "step": 6410
+    },
+    {
+      "epoch": 0.17555133235804815,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.267390529461765e-06,
+      "loss": 0.7173,
+      "step": 6420
+    },
+    {
+      "epoch": 0.17582477680097347,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.26464862493488e-06,
+      "loss": 0.6497,
+      "step": 6430
+    },
+    {
+      "epoch": 0.17609822124389876,
+      "grad_norm": 2.75,
+      "learning_rate": 8.261906720407996e-06,
+      "loss": 0.7468,
+      "step": 6440
+    },
+    {
+      "epoch": 0.17637166568682408,
+      "grad_norm": 3.0,
+      "learning_rate": 8.259164815881112e-06,
+      "loss": 0.7303,
+      "step": 6450
+    },
+    {
+      "epoch": 0.1766451101297494,
+      "grad_norm": 3.515625,
+      "learning_rate": 8.256422911354227e-06,
+      "loss": 0.674,
+      "step": 6460
+    },
+    {
+      "epoch": 0.1769185545726747,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.253681006827343e-06,
+      "loss": 0.7902,
+      "step": 6470
+    },
+    {
+      "epoch": 0.1771919990156,
+      "grad_norm": 2.75,
+      "learning_rate": 8.250939102300458e-06,
+      "loss": 0.7388,
+      "step": 6480
+    },
+    {
+      "epoch": 0.1774654434585253,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.248197197773574e-06,
+      "loss": 0.7135,
+      "step": 6490
+    },
+    {
+      "epoch": 0.17773888790145062,
+      "grad_norm": 2.390625,
+      "learning_rate": 8.24545529324669e-06,
+      "loss": 0.678,
+      "step": 6500
+    },
+    {
+      "epoch": 0.17801233234437594,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.242713388719805e-06,
+      "loss": 0.6821,
+      "step": 6510
+    },
+    {
+      "epoch": 0.17828577678730123,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.239971484192921e-06,
+      "loss": 0.618,
+      "step": 6520
+    },
+    {
+      "epoch": 0.17855922123022655,
+      "grad_norm": 2.125,
+      "learning_rate": 8.237229579666036e-06,
+      "loss": 0.6572,
+      "step": 6530
+    },
+    {
+      "epoch": 0.17883266567315187,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.234487675139152e-06,
+      "loss": 0.7001,
+      "step": 6540
+    },
+    {
+      "epoch": 0.17910611011607716,
+      "grad_norm": 2.875,
+      "learning_rate": 8.231745770612269e-06,
+      "loss": 0.6963,
+      "step": 6550
+    },
+    {
+      "epoch": 0.17937955455900248,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.229003866085383e-06,
+      "loss": 0.6611,
+      "step": 6560
+    },
+    {
+      "epoch": 0.17965299900192777,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.2262619615585e-06,
+      "loss": 0.7419,
+      "step": 6570
+    },
+    {
+      "epoch": 0.1799264434448531,
+      "grad_norm": 3.0,
+      "learning_rate": 8.223520057031616e-06,
+      "loss": 0.7737,
+      "step": 6580
+    },
+    {
+      "epoch": 0.1801998878877784,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.22077815250473e-06,
+      "loss": 0.6804,
+      "step": 6590
+    },
+    {
+      "epoch": 0.1804733323307037,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.218036247977847e-06,
+      "loss": 0.6785,
+      "step": 6600
+    },
+    {
+      "epoch": 0.18074677677362902,
+      "grad_norm": 2.625,
+      "learning_rate": 8.215294343450961e-06,
+      "loss": 0.717,
+      "step": 6610
+    },
+    {
+      "epoch": 0.18102022121655434,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.212552438924078e-06,
+      "loss": 0.7222,
+      "step": 6620
+    },
+    {
+      "epoch": 0.18129366565947963,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.209810534397194e-06,
+      "loss": 0.6882,
+      "step": 6630
+    },
+    {
+      "epoch": 0.18156711010240495,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.207068629870309e-06,
+      "loss": 0.6836,
+      "step": 6640
+    },
+    {
+      "epoch": 0.18184055454533024,
+      "grad_norm": 2.71875,
+      "learning_rate": 8.204326725343425e-06,
+      "loss": 0.7369,
+      "step": 6650
+    },
+    {
+      "epoch": 0.18211399898825556,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.20158482081654e-06,
+      "loss": 0.7083,
+      "step": 6660
+    },
+    {
+      "epoch": 0.18238744343118088,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.198842916289656e-06,
+      "loss": 0.75,
+      "step": 6670
+    },
+    {
+      "epoch": 0.18266088787410617,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.19610101176277e-06,
+      "loss": 0.7765,
+      "step": 6680
+    },
+    {
+      "epoch": 0.1829343323170315,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.193359107235887e-06,
+      "loss": 0.7226,
+      "step": 6690
+    },
+    {
+      "epoch": 0.1832077767599568,
+      "grad_norm": 3.640625,
+      "learning_rate": 8.190617202709003e-06,
+      "loss": 0.6855,
+      "step": 6700
+    },
+    {
+      "epoch": 0.1834812212028821,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.187875298182118e-06,
+      "loss": 0.663,
+      "step": 6710
+    },
+    {
+      "epoch": 0.18375466564580742,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.185133393655234e-06,
+      "loss": 0.7216,
+      "step": 6720
+    },
+    {
+      "epoch": 0.1840281100887327,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.182391489128349e-06,
+      "loss": 0.6865,
+      "step": 6730
+    },
+    {
+      "epoch": 0.18430155453165803,
+      "grad_norm": 2.625,
+      "learning_rate": 8.179649584601465e-06,
+      "loss": 0.7234,
+      "step": 6740
+    },
+    {
+      "epoch": 0.18457499897458335,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.17690768007458e-06,
+      "loss": 0.6552,
+      "step": 6750
+    },
+    {
+      "epoch": 0.18484844341750864,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.174165775547696e-06,
+      "loss": 0.6976,
+      "step": 6760
+    },
+    {
+      "epoch": 0.18512188786043396,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.171423871020811e-06,
+      "loss": 0.6946,
+      "step": 6770
+    },
+    {
+      "epoch": 0.18539533230335928,
+      "grad_norm": 2.75,
+      "learning_rate": 8.168681966493927e-06,
+      "loss": 0.7791,
+      "step": 6780
+    },
+    {
+      "epoch": 0.18566877674628457,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.165940061967042e-06,
+      "loss": 0.707,
+      "step": 6790
+    },
+    {
+      "epoch": 0.1859422211892099,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.163198157440158e-06,
+      "loss": 0.776,
+      "step": 6800
+    },
+    {
+      "epoch": 0.18621566563213518,
+      "grad_norm": 3.09375,
+      "learning_rate": 8.160456252913274e-06,
+      "loss": 0.6908,
+      "step": 6810
+    },
+    {
+      "epoch": 0.1864891100750605,
+      "grad_norm": 2.3125,
+      "learning_rate": 8.157714348386389e-06,
+      "loss": 0.6502,
+      "step": 6820
+    },
+    {
+      "epoch": 0.18676255451798582,
+      "grad_norm": 2.359375,
+      "learning_rate": 8.154972443859505e-06,
+      "loss": 0.6662,
+      "step": 6830
+    },
+    {
+      "epoch": 0.1870359989609111,
+      "grad_norm": 2.25,
+      "learning_rate": 8.15223053933262e-06,
+      "loss": 0.6613,
+      "step": 6840
+    },
+    {
+      "epoch": 0.18730944340383643,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.149488634805736e-06,
+      "loss": 0.679,
+      "step": 6850
+    },
+    {
+      "epoch": 0.18758288784676175,
+      "grad_norm": 3.359375,
+      "learning_rate": 8.146746730278853e-06,
+      "loss": 0.7829,
+      "step": 6860
+    },
+    {
+      "epoch": 0.18785633228968704,
+      "grad_norm": 2.859375,
+      "learning_rate": 8.144004825751967e-06,
+      "loss": 0.7227,
+      "step": 6870
+    },
+    {
+      "epoch": 0.18812977673261236,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.141262921225084e-06,
+      "loss": 0.7565,
+      "step": 6880
+    },
+    {
+      "epoch": 0.18840322117553765,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.138521016698198e-06,
+      "loss": 0.6529,
+      "step": 6890
+    },
+    {
+      "epoch": 0.18867666561846297,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.135779112171315e-06,
+      "loss": 0.7092,
+      "step": 6900
+    },
+    {
+      "epoch": 0.1889501100613883,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.133037207644431e-06,
+      "loss": 0.6895,
+      "step": 6910
+    },
+    {
+      "epoch": 0.18922355450431358,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.130295303117546e-06,
+      "loss": 0.7286,
+      "step": 6920
+    },
+    {
+      "epoch": 0.1894969989472389,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.127553398590662e-06,
+      "loss": 0.6628,
+      "step": 6930
+    },
+    {
+      "epoch": 0.1897704433901642,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.124811494063778e-06,
+      "loss": 0.6949,
+      "step": 6940
+    },
+    {
+      "epoch": 0.1900438878330895,
+      "grad_norm": 2.25,
+      "learning_rate": 8.122069589536893e-06,
+      "loss": 0.6493,
+      "step": 6950
+    },
+    {
+      "epoch": 0.19031733227601483,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.11932768501001e-06,
+      "loss": 0.7083,
+      "step": 6960
+    },
+    {
+      "epoch": 0.19059077671894012,
+      "grad_norm": 2.75,
+      "learning_rate": 8.116585780483124e-06,
+      "loss": 0.6602,
+      "step": 6970
+    },
+    {
+      "epoch": 0.19086422116186544,
+      "grad_norm": 2.234375,
+      "learning_rate": 8.11384387595624e-06,
+      "loss": 0.68,
+      "step": 6980
+    },
+    {
+      "epoch": 0.19113766560479076,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.111101971429357e-06,
+      "loss": 0.8084,
+      "step": 6990
+    },
+    {
+      "epoch": 0.19141111004771605,
+      "grad_norm": 2.875,
+      "learning_rate": 8.108360066902471e-06,
+      "loss": 0.7732,
+      "step": 7000
+    },
+    {
+      "epoch": 0.19168455449064137,
+      "grad_norm": 3.015625,
+      "learning_rate": 8.105618162375588e-06,
+      "loss": 0.6623,
+      "step": 7010
+    },
+    {
+      "epoch": 0.19195799893356666,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.102876257848702e-06,
+      "loss": 0.7723,
+      "step": 7020
+    },
+    {
+      "epoch": 0.19223144337649198,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.100134353321818e-06,
+      "loss": 0.6913,
+      "step": 7030
+    },
+    {
+      "epoch": 0.1925048878194173,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.097392448794935e-06,
+      "loss": 0.7974,
+      "step": 7040
+    },
+    {
+      "epoch": 0.1927783322623426,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.09465054426805e-06,
+      "loss": 0.7618,
+      "step": 7050
+    },
+    {
+      "epoch": 0.1930517767052679,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.091908639741166e-06,
+      "loss": 0.7155,
+      "step": 7060
+    },
+    {
+      "epoch": 0.19332522114819323,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.08916673521428e-06,
+      "loss": 0.7114,
+      "step": 7070
+    },
+    {
+      "epoch": 0.19359866559111852,
+      "grad_norm": 2.640625,
+      "learning_rate": 8.086424830687397e-06,
+      "loss": 0.7111,
+      "step": 7080
+    },
+    {
+      "epoch": 0.19387211003404384,
+      "grad_norm": 3.125,
+      "learning_rate": 8.083682926160511e-06,
+      "loss": 0.6835,
+      "step": 7090
+    },
+    {
+      "epoch": 0.19414555447696913,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.080941021633628e-06,
+      "loss": 0.7392,
+      "step": 7100
+    },
+    {
+      "epoch": 0.19441899891989445,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.078199117106742e-06,
+      "loss": 0.7701,
+      "step": 7110
+    },
+    {
+      "epoch": 0.19469244336281977,
+      "grad_norm": 2.609375,
+      "learning_rate": 8.075457212579859e-06,
+      "loss": 0.6767,
+      "step": 7120
+    },
+    {
+      "epoch": 0.19496588780574506,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.072715308052973e-06,
+      "loss": 0.7154,
+      "step": 7130
+    },
+    {
+      "epoch": 0.19523933224867038,
+      "grad_norm": 2.875,
+      "learning_rate": 8.06997340352609e-06,
+      "loss": 0.6923,
+      "step": 7140
+    },
+    {
+      "epoch": 0.1955127766915957,
+      "grad_norm": 2.4375,
+      "learning_rate": 8.067231498999204e-06,
+      "loss": 0.7875,
+      "step": 7150
+    },
+    {
+      "epoch": 0.195786221134521,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.06448959447232e-06,
+      "loss": 0.7205,
+      "step": 7160
+    },
+    {
+      "epoch": 0.1960596655774463,
+      "grad_norm": 2.734375,
+      "learning_rate": 8.061747689945437e-06,
+      "loss": 0.6838,
+      "step": 7170
+    },
+    {
+      "epoch": 0.1963331100203716,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.059005785418552e-06,
+      "loss": 0.7421,
+      "step": 7180
+    },
+    {
+      "epoch": 0.19660655446329692,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.056263880891668e-06,
+      "loss": 0.7385,
+      "step": 7190
+    },
+    {
+      "epoch": 0.19687999890622224,
+      "grad_norm": 2.96875,
+      "learning_rate": 8.053521976364783e-06,
+      "loss": 0.8019,
+      "step": 7200
+    },
+    {
+      "epoch": 0.19715344334914753,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.050780071837899e-06,
+      "loss": 0.6608,
+      "step": 7210
+    },
+    {
+      "epoch": 0.19742688779207285,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.048038167311015e-06,
+      "loss": 0.6893,
+      "step": 7220
+    },
+    {
+      "epoch": 0.19770033223499817,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.04529626278413e-06,
+      "loss": 0.6946,
+      "step": 7230
+    },
+    {
+      "epoch": 0.19797377667792346,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.042554358257246e-06,
+      "loss": 0.6856,
+      "step": 7240
+    },
+    {
+      "epoch": 0.19824722112084878,
+      "grad_norm": 2.953125,
+      "learning_rate": 8.03981245373036e-06,
+      "loss": 0.7387,
+      "step": 7250
+    },
+    {
+      "epoch": 0.19852066556377407,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.037070549203477e-06,
+      "loss": 0.7115,
+      "step": 7260
+    },
+    {
+      "epoch": 0.1987941100066994,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.034328644676593e-06,
+      "loss": 0.7336,
+      "step": 7270
+    },
+    {
+      "epoch": 0.1990675544496247,
+      "grad_norm": 2.5625,
+      "learning_rate": 8.031586740149708e-06,
+      "loss": 0.6799,
+      "step": 7280
+    },
+    {
+      "epoch": 0.19934099889255,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.028844835622824e-06,
+      "loss": 0.6311,
+      "step": 7290
+    },
+    {
+      "epoch": 0.19961444333547532,
+      "grad_norm": 2.65625,
+      "learning_rate": 8.02610293109594e-06,
+      "loss": 0.7292,
+      "step": 7300
+    },
+    {
+      "epoch": 0.19988788777840064,
+      "grad_norm": 3.046875,
+      "learning_rate": 8.023361026569055e-06,
+      "loss": 0.677,
+      "step": 7310
+    },
+    {
+      "epoch": 0.20016133222132593,
+      "grad_norm": 2.5,
+      "learning_rate": 8.020619122042172e-06,
+      "loss": 0.7558,
+      "step": 7320
+    },
+    {
+      "epoch": 0.20043477666425125,
+      "grad_norm": 2.6875,
+      "learning_rate": 8.017877217515286e-06,
+      "loss": 0.6776,
+      "step": 7330
+    },
+    {
+      "epoch": 0.20070822110717654,
+      "grad_norm": 2.8125,
+      "learning_rate": 8.015135312988403e-06,
+      "loss": 0.7339,
+      "step": 7340
+    },
+    {
+      "epoch": 0.20098166555010186,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.012393408461519e-06,
+      "loss": 0.7231,
+      "step": 7350
+    },
+    {
+      "epoch": 0.20125510999302718,
+      "grad_norm": 2.484375,
+      "learning_rate": 8.009651503934634e-06,
+      "loss": 0.7257,
+      "step": 7360
+    },
+    {
+      "epoch": 0.20152855443595247,
+      "grad_norm": 2.421875,
+      "learning_rate": 8.00690959940775e-06,
+      "loss": 0.6538,
+      "step": 7370
+    },
+    {
+      "epoch": 0.2018019988788778,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.004167694880865e-06,
+      "loss": 0.6581,
+      "step": 7380
+    },
+    {
+      "epoch": 0.2020754433218031,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.001425790353981e-06,
+      "loss": 0.7096,
+      "step": 7390
+    },
+    {
+      "epoch": 0.2023488877647284,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.998683885827097e-06,
+      "loss": 0.6994,
+      "step": 7400
+    },
+    {
+      "epoch": 0.20262233220765372,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.995941981300212e-06,
+      "loss": 0.7282,
+      "step": 7410
+    },
+    {
+      "epoch": 0.202895776650579,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.993200076773328e-06,
+      "loss": 0.767,
+      "step": 7420
+    },
+    {
+      "epoch": 0.20316922109350433,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.990458172246443e-06,
+      "loss": 0.6625,
+      "step": 7430
+    },
+    {
+      "epoch": 0.20344266553642965,
+      "grad_norm": 2.75,
+      "learning_rate": 7.98771626771956e-06,
+      "loss": 0.6955,
+      "step": 7440
+    },
+    {
+      "epoch": 0.20371610997935494,
+      "grad_norm": 2.75,
+      "learning_rate": 7.984974363192674e-06,
+      "loss": 0.6762,
+      "step": 7450
+    },
+    {
+      "epoch": 0.20398955442228026,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.98223245866579e-06,
+      "loss": 0.6661,
+      "step": 7460
+    },
+    {
+      "epoch": 0.20426299886520558,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.979490554138905e-06,
+      "loss": 0.7025,
+      "step": 7470
+    },
+    {
+      "epoch": 0.20453644330813087,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.976748649612021e-06,
+      "loss": 0.6563,
+      "step": 7480
+    },
+    {
+      "epoch": 0.20480988775105619,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.974006745085137e-06,
+      "loss": 0.8056,
+      "step": 7490
+    },
+    {
+      "epoch": 0.20508333219398148,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.971264840558252e-06,
+      "loss": 0.7226,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2053567766369068,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.968522936031368e-06,
+      "loss": 0.6216,
+      "step": 7510
+    },
+    {
+      "epoch": 0.20563022107983212,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.965781031504483e-06,
+      "loss": 0.7541,
+      "step": 7520
+    },
+    {
+      "epoch": 0.2059036655227574,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.9630391269776e-06,
+      "loss": 0.7987,
+      "step": 7530
+    },
+    {
+      "epoch": 0.20617710996568273,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.960297222450714e-06,
+      "loss": 0.7113,
+      "step": 7540
+    },
+    {
+      "epoch": 0.20645055440860804,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.95755531792383e-06,
+      "loss": 0.6564,
+      "step": 7550
+    },
+    {
+      "epoch": 0.20672399885153334,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.954813413396945e-06,
+      "loss": 0.6892,
+      "step": 7560
+    },
+    {
+      "epoch": 0.20699744329445866,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.952071508870061e-06,
+      "loss": 0.7113,
+      "step": 7570
+    },
+    {
+      "epoch": 0.20727088773738395,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.949329604343178e-06,
+      "loss": 0.6635,
+      "step": 7580
+    },
+    {
+      "epoch": 0.20754433218030927,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.946587699816292e-06,
+      "loss": 0.7383,
+      "step": 7590
+    },
+    {
+      "epoch": 0.20781777662323458,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.943845795289409e-06,
+      "loss": 0.6727,
+      "step": 7600
+    },
+    {
+      "epoch": 0.20809122106615988,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.941103890762523e-06,
+      "loss": 0.7019,
+      "step": 7610
+    },
+    {
+      "epoch": 0.2083646655090852,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.93836198623564e-06,
+      "loss": 0.6564,
+      "step": 7620
+    },
+    {
+      "epoch": 0.2086381099520105,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.935620081708756e-06,
+      "loss": 0.7474,
+      "step": 7630
+    },
+    {
+      "epoch": 0.2089115543949358,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.93287817718187e-06,
+      "loss": 0.6763,
+      "step": 7640
+    },
+    {
+      "epoch": 0.20918499883786112,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.930136272654987e-06,
+      "loss": 0.7594,
+      "step": 7650
+    },
+    {
+      "epoch": 0.20945844328078642,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.927394368128103e-06,
+      "loss": 0.7537,
+      "step": 7660
+    },
+    {
+      "epoch": 0.20973188772371174,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.924652463601218e-06,
+      "loss": 0.7038,
+      "step": 7670
+    },
+    {
+      "epoch": 0.21000533216663705,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.921910559074334e-06,
+      "loss": 0.6148,
+      "step": 7680
+    },
+    {
+      "epoch": 0.21027877660956235,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.919168654547449e-06,
+      "loss": 0.7059,
+      "step": 7690
+    },
+    {
+      "epoch": 0.21055222105248766,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.916426750020565e-06,
+      "loss": 0.8172,
+      "step": 7700
+    },
+    {
+      "epoch": 0.21082566549541296,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.913684845493681e-06,
+      "loss": 0.6022,
+      "step": 7710
+    },
+    {
+      "epoch": 0.21109910993833828,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.910942940966796e-06,
+      "loss": 0.7819,
+      "step": 7720
+    },
+    {
+      "epoch": 0.2113725543812636,
+      "grad_norm": 2.5,
+      "learning_rate": 7.908201036439912e-06,
+      "loss": 0.6228,
+      "step": 7730
+    },
+    {
+      "epoch": 0.21164599882418889,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.905459131913027e-06,
+      "loss": 0.7004,
+      "step": 7740
+    },
+    {
+      "epoch": 0.2119194432671142,
+      "grad_norm": 2.375,
+      "learning_rate": 7.902717227386143e-06,
+      "loss": 0.7204,
+      "step": 7750
+    },
+    {
+      "epoch": 0.21219288771003952,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.89997532285926e-06,
+      "loss": 0.7008,
+      "step": 7760
+    },
+    {
+      "epoch": 0.21246633215296482,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.897233418332374e-06,
+      "loss": 0.7058,
+      "step": 7770
+    },
+    {
+      "epoch": 0.21273977659589013,
+      "grad_norm": 2.5,
+      "learning_rate": 7.89449151380549e-06,
+      "loss": 0.6728,
+      "step": 7780
+    },
+    {
+      "epoch": 0.21301322103881543,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.891749609278605e-06,
+      "loss": 0.6555,
+      "step": 7790
+    },
+    {
+      "epoch": 0.21328666548174074,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.889007704751722e-06,
+      "loss": 0.6773,
+      "step": 7800
+    },
+    {
+      "epoch": 0.21356010992466606,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.886265800224838e-06,
+      "loss": 0.7249,
+      "step": 7810
+    },
+    {
+      "epoch": 0.21383355436759136,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.883523895697953e-06,
+      "loss": 0.683,
+      "step": 7820
+    },
+    {
+      "epoch": 0.21410699881051667,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.880781991171069e-06,
+      "loss": 0.7468,
+      "step": 7830
+    },
+    {
+      "epoch": 0.214380443253442,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.878040086644184e-06,
+      "loss": 0.671,
+      "step": 7840
+    },
+    {
+      "epoch": 0.21465388769636728,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.8752981821173e-06,
+      "loss": 0.6613,
+      "step": 7850
+    },
+    {
+      "epoch": 0.2149273321392926,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.872556277590414e-06,
+      "loss": 0.7463,
+      "step": 7860
+    },
+    {
+      "epoch": 0.2152007765822179,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.86981437306353e-06,
+      "loss": 0.6532,
+      "step": 7870
+    },
+    {
+      "epoch": 0.21547422102514321,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.867072468536645e-06,
+      "loss": 0.6985,
+      "step": 7880
+    },
+    {
+      "epoch": 0.21574766546806853,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.864330564009762e-06,
+      "loss": 0.6782,
+      "step": 7890
+    },
+    {
+      "epoch": 0.21602110991099382,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.861588659482876e-06,
+      "loss": 0.7212,
+      "step": 7900
+    },
+    {
+      "epoch": 0.21629455435391914,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.858846754955993e-06,
+      "loss": 0.6803,
+      "step": 7910
+    },
+    {
+      "epoch": 0.21656799879684446,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.856104850429107e-06,
+      "loss": 0.7446,
+      "step": 7920
+    },
+    {
+      "epoch": 0.21684144323976975,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.853362945902224e-06,
+      "loss": 0.6921,
+      "step": 7930
+    },
+    {
+      "epoch": 0.21711488768269507,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.85062104137534e-06,
+      "loss": 0.7634,
+      "step": 7940
+    },
+    {
+      "epoch": 0.21738833212562036,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.847879136848455e-06,
+      "loss": 0.7258,
+      "step": 7950
+    },
+    {
+      "epoch": 0.21766177656854568,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.845137232321571e-06,
+      "loss": 0.7002,
+      "step": 7960
+    },
+    {
+      "epoch": 0.217935221011471,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.842395327794687e-06,
+      "loss": 0.7136,
+      "step": 7970
+    },
+    {
+      "epoch": 0.2182086654543963,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.839653423267802e-06,
+      "loss": 0.6587,
+      "step": 7980
+    },
+    {
+      "epoch": 0.2184821098973216,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.836911518740918e-06,
+      "loss": 0.7183,
+      "step": 7990
+    },
+    {
+      "epoch": 0.21875555434024693,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.834169614214033e-06,
+      "loss": 0.7181,
+      "step": 8000
+    },
+    {
+      "epoch": 0.21902899878317222,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.83142770968715e-06,
+      "loss": 0.7664,
+      "step": 8010
+    },
+    {
+      "epoch": 0.21930244322609754,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.828685805160266e-06,
+      "loss": 0.6806,
+      "step": 8020
+    },
+    {
+      "epoch": 0.21957588766902283,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.82594390063338e-06,
+      "loss": 0.7141,
+      "step": 8030
+    },
+    {
+      "epoch": 0.21984933211194815,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.823201996106497e-06,
+      "loss": 0.7348,
+      "step": 8040
+    },
+    {
+      "epoch": 0.22012277655487347,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.820460091579611e-06,
+      "loss": 0.6858,
+      "step": 8050
+    },
+    {
+      "epoch": 0.22039622099779876,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.817718187052728e-06,
+      "loss": 0.707,
+      "step": 8060
+    },
+    {
+      "epoch": 0.22066966544072408,
+      "grad_norm": 3.0,
+      "learning_rate": 7.814976282525844e-06,
+      "loss": 0.7124,
+      "step": 8070
+    },
+    {
+      "epoch": 0.2209431098836494,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.812234377998958e-06,
+      "loss": 0.6937,
+      "step": 8080
+    },
+    {
+      "epoch": 0.2212165543265747,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.809492473472075e-06,
+      "loss": 0.6625,
+      "step": 8090
+    },
+    {
+      "epoch": 0.2214899987695,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.80675056894519e-06,
+      "loss": 0.6625,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2217634432124253,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.804008664418306e-06,
+      "loss": 0.7438,
+      "step": 8110
+    },
+    {
+      "epoch": 0.22203688765535062,
+      "grad_norm": 2.875,
+      "learning_rate": 7.801266759891422e-06,
+      "loss": 0.7229,
+      "step": 8120
+    },
+    {
+      "epoch": 0.22231033209827594,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.798524855364537e-06,
+      "loss": 0.7284,
+      "step": 8130
+    },
+    {
+      "epoch": 0.22258377654120123,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.795782950837653e-06,
+      "loss": 0.7385,
+      "step": 8140
+    },
+    {
+      "epoch": 0.22285722098412655,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.79304104631077e-06,
+      "loss": 0.6546,
+      "step": 8150
+    },
+    {
+      "epoch": 0.22313066542705187,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.790299141783884e-06,
+      "loss": 0.7147,
+      "step": 8160
+    },
+    {
+      "epoch": 0.22340410986997716,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.787557237257e-06,
+      "loss": 0.678,
+      "step": 8170
+    },
+    {
+      "epoch": 0.22367755431290248,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.784815332730115e-06,
+      "loss": 0.7315,
+      "step": 8180
+    },
+    {
+      "epoch": 0.22395099875582777,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.782073428203231e-06,
+      "loss": 0.7517,
+      "step": 8190
+    },
+    {
+      "epoch": 0.2242244431987531,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.779331523676346e-06,
+      "loss": 0.678,
+      "step": 8200
+    },
+    {
+      "epoch": 0.2244978876416784,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.776589619149462e-06,
+      "loss": 0.69,
+      "step": 8210
+    },
+    {
+      "epoch": 0.2247713320846037,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.773847714622577e-06,
+      "loss": 0.6988,
+      "step": 8220
+    },
+    {
+      "epoch": 0.22504477652752902,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.771105810095693e-06,
+      "loss": 0.7531,
+      "step": 8230
+    },
+    {
+      "epoch": 0.22531822097045434,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.768363905568808e-06,
+      "loss": 0.6979,
+      "step": 8240
+    },
+    {
+      "epoch": 0.22559166541337963,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.765622001041924e-06,
+      "loss": 0.6515,
+      "step": 8250
+    },
+    {
+      "epoch": 0.22586510985630495,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.76288009651504e-06,
+      "loss": 0.5952,
+      "step": 8260
+    },
+    {
+      "epoch": 0.22613855429923024,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.760138191988155e-06,
+      "loss": 0.7456,
+      "step": 8270
+    },
+    {
+      "epoch": 0.22641199874215556,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.757396287461272e-06,
+      "loss": 0.6701,
+      "step": 8280
+    },
+    {
+      "epoch": 0.22668544318508088,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.754654382934386e-06,
+      "loss": 0.7564,
+      "step": 8290
+    },
+    {
+      "epoch": 0.22695888762800617,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.751912478407502e-06,
+      "loss": 0.7033,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2272323320709315,
+      "grad_norm": 3.21875,
+      "learning_rate": 7.749170573880617e-06,
+      "loss": 0.7193,
+      "step": 8310
+    },
+    {
+      "epoch": 0.22750577651385678,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.746428669353733e-06,
+      "loss": 0.7145,
+      "step": 8320
+    },
+    {
+      "epoch": 0.2277792209567821,
+      "grad_norm": 2.625,
+      "learning_rate": 7.74368676482685e-06,
+      "loss": 0.638,
+      "step": 8330
+    },
+    {
+      "epoch": 0.22805266539970742,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.740944860299964e-06,
+      "loss": 0.7182,
+      "step": 8340
+    },
+    {
+      "epoch": 0.2283261098426327,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.73820295577308e-06,
+      "loss": 0.641,
+      "step": 8350
+    },
+    {
+      "epoch": 0.22859955428555803,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.735461051246195e-06,
+      "loss": 0.6775,
+      "step": 8360
+    },
+    {
+      "epoch": 0.22887299872848335,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.732719146719312e-06,
+      "loss": 0.6925,
+      "step": 8370
+    },
+    {
+      "epoch": 0.22914644317140864,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.729977242192428e-06,
+      "loss": 0.6075,
+      "step": 8380
+    },
+    {
+      "epoch": 0.22941988761433396,
+      "grad_norm": 2.75,
+      "learning_rate": 7.727235337665543e-06,
+      "loss": 0.7305,
+      "step": 8390
+    },
+    {
+      "epoch": 0.22969333205725925,
+      "grad_norm": 2.5,
+      "learning_rate": 7.724493433138659e-06,
+      "loss": 0.675,
+      "step": 8400
+    },
+    {
+      "epoch": 0.22996677650018457,
+      "grad_norm": 5.75,
+      "learning_rate": 7.721751528611774e-06,
+      "loss": 0.6105,
+      "step": 8410
+    },
+    {
+      "epoch": 0.2302402209431099,
+      "grad_norm": 3.640625,
+      "learning_rate": 7.71900962408489e-06,
+      "loss": 0.7066,
+      "step": 8420
+    },
+    {
+      "epoch": 0.23051366538603518,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.716267719558006e-06,
+      "loss": 0.6597,
+      "step": 8430
+    },
+    {
+      "epoch": 0.2307871098289605,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.713525815031121e-06,
+      "loss": 0.6767,
+      "step": 8440
+    },
+    {
+      "epoch": 0.23106055427188582,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.710783910504237e-06,
+      "loss": 0.6576,
+      "step": 8450
+    },
+    {
+      "epoch": 0.2313339987148111,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.708042005977352e-06,
+      "loss": 0.7109,
+      "step": 8460
+    },
+    {
+      "epoch": 0.23160744315773643,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.705300101450468e-06,
+      "loss": 0.7459,
+      "step": 8470
+    },
+    {
+      "epoch": 0.23188088760066172,
+      "grad_norm": 3.46875,
+      "learning_rate": 7.702558196923585e-06,
+      "loss": 0.6968,
+      "step": 8480
+    },
+    {
+      "epoch": 0.23215433204358704,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.6998162923967e-06,
+      "loss": 0.6928,
+      "step": 8490
+    },
+    {
+      "epoch": 0.23242777648651236,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.697074387869816e-06,
+      "loss": 0.695,
+      "step": 8500
+    },
+    {
+      "epoch": 0.23270122092943765,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.694332483342932e-06,
+      "loss": 0.7941,
+      "step": 8510
+    },
+    {
+      "epoch": 0.23297466537236297,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.691590578816046e-06,
+      "loss": 0.6616,
+      "step": 8520
+    },
+    {
+      "epoch": 0.2332481098152883,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.688848674289163e-06,
+      "loss": 0.6886,
+      "step": 8530
+    },
+    {
+      "epoch": 0.23352155425821358,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.686106769762277e-06,
+      "loss": 0.7339,
+      "step": 8540
+    },
+    {
+      "epoch": 0.2337949987011389,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.683364865235394e-06,
+      "loss": 0.7151,
+      "step": 8550
+    },
+    {
+      "epoch": 0.2340684431440642,
+      "grad_norm": 2.421875,
+      "learning_rate": 7.680622960708508e-06,
+      "loss": 0.65,
+      "step": 8560
+    },
+    {
+      "epoch": 0.2343418875869895,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.677881056181625e-06,
+      "loss": 0.722,
+      "step": 8570
+    },
+    {
+      "epoch": 0.23461533202991483,
+      "grad_norm": 2.0625,
+      "learning_rate": 7.675139151654741e-06,
+      "loss": 0.7052,
+      "step": 8580
+    },
+    {
+      "epoch": 0.23488877647284012,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.672397247127856e-06,
+      "loss": 0.6782,
+      "step": 8590
+    },
+    {
+      "epoch": 0.23516222091576544,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.669655342600972e-06,
+      "loss": 0.7057,
+      "step": 8600
+    },
+    {
+      "epoch": 0.23543566535869076,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.666913438074087e-06,
+      "loss": 0.7619,
+      "step": 8610
+    },
+    {
+      "epoch": 0.23570910980161605,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.664171533547203e-06,
+      "loss": 0.6647,
+      "step": 8620
+    },
+    {
+      "epoch": 0.23598255424454137,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.661429629020318e-06,
+      "loss": 0.6891,
+      "step": 8630
+    },
+    {
+      "epoch": 0.23625599868746666,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.658687724493434e-06,
+      "loss": 0.7388,
+      "step": 8640
+    },
+    {
+      "epoch": 0.23652944313039198,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.655945819966549e-06,
+      "loss": 0.7183,
+      "step": 8650
+    },
+    {
+      "epoch": 0.2368028875733173,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.653203915439665e-06,
+      "loss": 0.6682,
+      "step": 8660
+    },
+    {
+      "epoch": 0.2370763320162426,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.65046201091278e-06,
+      "loss": 0.7204,
+      "step": 8670
+    },
+    {
+      "epoch": 0.2373497764591679,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.647720106385896e-06,
+      "loss": 0.7241,
+      "step": 8680
+    },
+    {
+      "epoch": 0.23762322090209323,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.644978201859012e-06,
+      "loss": 0.7432,
+      "step": 8690
+    },
+    {
+      "epoch": 0.23789666534501852,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.642236297332127e-06,
+      "loss": 0.8173,
+      "step": 8700
+    },
+    {
+      "epoch": 0.23817010978794384,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.639494392805243e-06,
+      "loss": 0.6875,
+      "step": 8710
+    },
+    {
+      "epoch": 0.23844355423086913,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.636752488278358e-06,
+      "loss": 0.69,
+      "step": 8720
+    },
+    {
+      "epoch": 0.23871699867379445,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.634010583751474e-06,
+      "loss": 0.6373,
+      "step": 8730
+    },
+    {
+      "epoch": 0.23899044311671977,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.63126867922459e-06,
+      "loss": 0.7567,
+      "step": 8740
+    },
+    {
+      "epoch": 0.23926388755964506,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.628526774697705e-06,
+      "loss": 0.6443,
+      "step": 8750
+    },
+    {
+      "epoch": 0.23953733200257038,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.625784870170821e-06,
+      "loss": 0.6936,
+      "step": 8760
+    },
+    {
+      "epoch": 0.2398107764454957,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.623042965643936e-06,
+      "loss": 0.7132,
+      "step": 8770
+    },
+    {
+      "epoch": 0.240084220888421,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.620301061117052e-06,
+      "loss": 0.6631,
+      "step": 8780
+    },
+    {
+      "epoch": 0.2403576653313463,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.617559156590169e-06,
+      "loss": 0.6444,
+      "step": 8790
+    },
+    {
+      "epoch": 0.2406311097742716,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.614817252063283e-06,
+      "loss": 0.7201,
+      "step": 8800
+    },
+    {
+      "epoch": 0.24090455421719692,
+      "grad_norm": 2.75,
+      "learning_rate": 7.6120753475364e-06,
+      "loss": 0.6874,
+      "step": 8810
+    },
+    {
+      "epoch": 0.24117799866012224,
+      "grad_norm": 2.75,
+      "learning_rate": 7.609333443009514e-06,
+      "loss": 0.6698,
+      "step": 8820
+    },
+    {
+      "epoch": 0.24145144310304753,
+      "grad_norm": 2.234375,
+      "learning_rate": 7.606591538482631e-06,
+      "loss": 0.6676,
+      "step": 8830
+    },
+    {
+      "epoch": 0.24172488754597285,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.603849633955747e-06,
+      "loss": 0.76,
+      "step": 8840
+    },
+    {
+      "epoch": 0.24199833198889817,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.601107729428862e-06,
+      "loss": 0.6583,
+      "step": 8850
+    },
+    {
+      "epoch": 0.24227177643182346,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.598365824901978e-06,
+      "loss": 0.6999,
+      "step": 8860
+    },
+    {
+      "epoch": 0.24254522087474878,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.595623920375093e-06,
+      "loss": 0.717,
+      "step": 8870
+    },
+    {
+      "epoch": 0.24281866531767407,
+      "grad_norm": 3.234375,
+      "learning_rate": 7.592882015848209e-06,
+      "loss": 0.75,
+      "step": 8880
+    },
+    {
+      "epoch": 0.2430921097605994,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.590140111321324e-06,
+      "loss": 0.6548,
+      "step": 8890
+    },
+    {
+      "epoch": 0.2433655542035247,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.58739820679444e-06,
+      "loss": 0.6962,
+      "step": 8900
+    },
+    {
+      "epoch": 0.24363899864645,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.584656302267555e-06,
+      "loss": 0.6997,
+      "step": 8910
+    },
+    {
+      "epoch": 0.24391244308937532,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.581914397740672e-06,
+      "loss": 0.7096,
+      "step": 8920
+    },
+    {
+      "epoch": 0.24418588753230064,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.579172493213786e-06,
+      "loss": 0.6557,
+      "step": 8930
+    },
+    {
+      "epoch": 0.24445933197522593,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.576430588686903e-06,
+      "loss": 0.6859,
+      "step": 8940
+    },
+    {
+      "epoch": 0.24473277641815125,
+      "grad_norm": 2.625,
+      "learning_rate": 7.573688684160017e-06,
+      "loss": 0.6274,
+      "step": 8950
+    },
+    {
+      "epoch": 0.24500622086107654,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.570946779633134e-06,
+      "loss": 0.6992,
+      "step": 8960
+    },
+    {
+      "epoch": 0.24527966530400186,
+      "grad_norm": 2.75,
+      "learning_rate": 7.56820487510625e-06,
+      "loss": 0.7478,
+      "step": 8970
+    },
+    {
+      "epoch": 0.24555310974692718,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.5654629705793646e-06,
+      "loss": 0.7116,
+      "step": 8980
+    },
+    {
+      "epoch": 0.24582655418985247,
+      "grad_norm": 2.5,
+      "learning_rate": 7.562721066052481e-06,
+      "loss": 0.6652,
+      "step": 8990
+    },
+    {
+      "epoch": 0.2460999986327778,
+      "grad_norm": 3.0,
+      "learning_rate": 7.5599791615255955e-06,
+      "loss": 0.7693,
+      "step": 9000
+    },
+    {
+      "epoch": 0.24637344307570308,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.557237256998712e-06,
+      "loss": 0.6308,
+      "step": 9010
+    },
+    {
+      "epoch": 0.2466468875186284,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.554495352471828e-06,
+      "loss": 0.6975,
+      "step": 9020
+    },
+    {
+      "epoch": 0.24692033196155372,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.551753447944943e-06,
+      "loss": 0.6442,
+      "step": 9030
+    },
+    {
+      "epoch": 0.247193776404479,
+      "grad_norm": 2.1875,
+      "learning_rate": 7.549011543418059e-06,
+      "loss": 0.6081,
+      "step": 9040
+    },
+    {
+      "epoch": 0.24746722084740433,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.546269638891175e-06,
+      "loss": 0.7126,
+      "step": 9050
+    },
+    {
+      "epoch": 0.24774066529032965,
+      "grad_norm": 2.625,
+      "learning_rate": 7.54352773436429e-06,
+      "loss": 0.6811,
+      "step": 9060
+    },
+    {
+      "epoch": 0.24801410973325494,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.540785829837406e-06,
+      "loss": 0.7447,
+      "step": 9070
+    },
+    {
+      "epoch": 0.24828755417618026,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.538043925310521e-06,
+      "loss": 0.7173,
+      "step": 9080
+    },
+    {
+      "epoch": 0.24856099861910555,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.5353020207836366e-06,
+      "loss": 0.7225,
+      "step": 9090
+    },
+    {
+      "epoch": 0.24883444306203087,
+      "grad_norm": 2.359375,
+      "learning_rate": 7.532560116256753e-06,
+      "loss": 0.6711,
+      "step": 9100
+    },
+    {
+      "epoch": 0.2491078875049562,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.5298182117298675e-06,
+      "loss": 0.6839,
+      "step": 9110
+    },
+    {
+      "epoch": 0.24938133194788148,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.527076307202984e-06,
+      "loss": 0.6785,
+      "step": 9120
+    },
+    {
+      "epoch": 0.2496547763908068,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.524334402676099e-06,
+      "loss": 0.7844,
+      "step": 9130
+    },
+    {
+      "epoch": 0.24992822083373212,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.521592498149215e-06,
+      "loss": 0.6824,
+      "step": 9140
+    },
+    {
+      "epoch": 0.25020166527665744,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.518850593622331e-06,
+      "loss": 0.7132,
+      "step": 9150
+    },
+    {
+      "epoch": 0.2504751097195827,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.516108689095446e-06,
+      "loss": 0.7171,
+      "step": 9160
+    },
+    {
+      "epoch": 0.250748554162508,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.513366784568562e-06,
+      "loss": 0.7262,
+      "step": 9170
+    },
+    {
+      "epoch": 0.25102199860543334,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.5106248800416784e-06,
+      "loss": 0.647,
+      "step": 9180
+    },
+    {
+      "epoch": 0.25129544304835866,
+      "grad_norm": 2.5,
+      "learning_rate": 7.507882975514793e-06,
+      "loss": 0.693,
+      "step": 9190
+    },
+    {
+      "epoch": 0.251568887491284,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.505141070987909e-06,
+      "loss": 0.6096,
+      "step": 9200
+    },
+    {
+      "epoch": 0.25184233193420924,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.502399166461024e-06,
+      "loss": 0.7002,
+      "step": 9210
+    },
+    {
+      "epoch": 0.25211577637713456,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.49965726193414e-06,
+      "loss": 0.6868,
+      "step": 9220
+    },
+    {
+      "epoch": 0.2523892208200599,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.496915357407256e-06,
+      "loss": 0.7755,
+      "step": 9230
+    },
+    {
+      "epoch": 0.2526626652629852,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.494173452880371e-06,
+      "loss": 0.68,
+      "step": 9240
+    },
+    {
+      "epoch": 0.2529361097059105,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.491431548353487e-06,
+      "loss": 0.6615,
+      "step": 9250
+    },
+    {
+      "epoch": 0.25320955414883584,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.488689643826602e-06,
+      "loss": 0.7068,
+      "step": 9260
+    },
+    {
+      "epoch": 0.2534829985917611,
+      "grad_norm": 2.265625,
+      "learning_rate": 7.485947739299718e-06,
+      "loss": 0.6525,
+      "step": 9270
+    },
+    {
+      "epoch": 0.2537564430346864,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.483205834772834e-06,
+      "loss": 0.7144,
+      "step": 9280
+    },
+    {
+      "epoch": 0.25402988747761174,
+      "grad_norm": 3.140625,
+      "learning_rate": 7.48046393024595e-06,
+      "loss": 0.7296,
+      "step": 9290
+    },
+    {
+      "epoch": 0.25430333192053706,
+      "grad_norm": 2.625,
+      "learning_rate": 7.477722025719065e-06,
+      "loss": 0.7118,
+      "step": 9300
+    },
+    {
+      "epoch": 0.2545767763634624,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.4749801211921806e-06,
+      "loss": 0.6641,
+      "step": 9310
+    },
+    {
+      "epoch": 0.25485022080638764,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.472238216665296e-06,
+      "loss": 0.6263,
+      "step": 9320
+    },
+    {
+      "epoch": 0.25512366524931296,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.469496312138412e-06,
+      "loss": 0.677,
+      "step": 9330
+    },
+    {
+      "epoch": 0.2553971096922383,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.466754407611527e-06,
+      "loss": 0.6567,
+      "step": 9340
+    },
+    {
+      "epoch": 0.2556705541351636,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.464012503084643e-06,
+      "loss": 0.7111,
+      "step": 9350
+    },
+    {
+      "epoch": 0.2559439985780889,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.46127059855776e-06,
+      "loss": 0.6513,
+      "step": 9360
+    },
+    {
+      "epoch": 0.2562174430210142,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.458528694030874e-06,
+      "loss": 0.7197,
+      "step": 9370
+    },
+    {
+      "epoch": 0.2564908874639395,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.455786789503991e-06,
+      "loss": 0.7078,
+      "step": 9380
+    },
+    {
+      "epoch": 0.2567643319068648,
+      "grad_norm": 2.75,
+      "learning_rate": 7.453044884977105e-06,
+      "loss": 0.7895,
+      "step": 9390
+    },
+    {
+      "epoch": 0.25703777634979014,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.450302980450222e-06,
+      "loss": 0.5916,
+      "step": 9400
+    },
+    {
+      "epoch": 0.25731122079271546,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.447561075923337e-06,
+      "loss": 0.7061,
+      "step": 9410
+    },
+    {
+      "epoch": 0.2575846652356408,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.4448191713964525e-06,
+      "loss": 0.7298,
+      "step": 9420
+    },
+    {
+      "epoch": 0.25785810967856604,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.442077266869568e-06,
+      "loss": 0.629,
+      "step": 9430
+    },
+    {
+      "epoch": 0.25813155412149136,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.4393353623426835e-06,
+      "loss": 0.5903,
+      "step": 9440
+    },
+    {
+      "epoch": 0.2584049985644167,
+      "grad_norm": 2.75,
+      "learning_rate": 7.4365934578158e-06,
+      "loss": 0.7129,
+      "step": 9450
+    },
+    {
+      "epoch": 0.258678443007342,
+      "grad_norm": 3.453125,
+      "learning_rate": 7.433851553288915e-06,
+      "loss": 0.693,
+      "step": 9460
+    },
+    {
+      "epoch": 0.2589518874502673,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.431109648762031e-06,
+      "loss": 0.687,
+      "step": 9470
+    },
+    {
+      "epoch": 0.2592253318931926,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.428367744235146e-06,
+      "loss": 0.6796,
+      "step": 9480
+    },
+    {
+      "epoch": 0.2594987763361179,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.425625839708262e-06,
+      "loss": 0.6469,
+      "step": 9490
+    },
+    {
+      "epoch": 0.2597722207790432,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.422883935181377e-06,
+      "loss": 0.6474,
+      "step": 9500
+    },
+    {
+      "epoch": 0.26004566522196854,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.420142030654494e-06,
+      "loss": 0.7923,
+      "step": 9510
+    },
+    {
+      "epoch": 0.26031910966489386,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.417400126127608e-06,
+      "loss": 0.6564,
+      "step": 9520
+    },
+    {
+      "epoch": 0.2605925541078191,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.4146582216007245e-06,
+      "loss": 0.7091,
+      "step": 9530
+    },
+    {
+      "epoch": 0.26086599855074444,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.411916317073841e-06,
+      "loss": 0.6955,
+      "step": 9540
+    },
+    {
+      "epoch": 0.26113944299366976,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.4091744125469555e-06,
+      "loss": 0.7045,
+      "step": 9550
+    },
+    {
+      "epoch": 0.2614128874365951,
+      "grad_norm": 2.15625,
+      "learning_rate": 7.406432508020072e-06,
+      "loss": 0.6304,
+      "step": 9560
+    },
+    {
+      "epoch": 0.2616863318795204,
+      "grad_norm": 2.75,
+      "learning_rate": 7.4036906034931865e-06,
+      "loss": 0.6762,
+      "step": 9570
+    },
+    {
+      "epoch": 0.2619597763224457,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.400948698966303e-06,
+      "loss": 0.6756,
+      "step": 9580
+    },
+    {
+      "epoch": 0.262233220765371,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.398206794439418e-06,
+      "loss": 0.7616,
+      "step": 9590
+    },
+    {
+      "epoch": 0.2625066652082963,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.395464889912534e-06,
+      "loss": 0.7114,
+      "step": 9600
+    },
+    {
+      "epoch": 0.2627801096512216,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.39272298538565e-06,
+      "loss": 0.7287,
+      "step": 9610
+    },
+    {
+      "epoch": 0.26305355409414694,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.389981080858765e-06,
+      "loss": 0.7074,
+      "step": 9620
+    },
+    {
+      "epoch": 0.26332699853707225,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.387239176331881e-06,
+      "loss": 0.6533,
+      "step": 9630
+    },
+    {
+      "epoch": 0.2636004429799975,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.3844972718049965e-06,
+      "loss": 0.7169,
+      "step": 9640
+    },
+    {
+      "epoch": 0.26387388742292284,
+      "grad_norm": 2.84375,
+      "learning_rate": 7.381755367278112e-06,
+      "loss": 0.6494,
+      "step": 9650
+    },
+    {
+      "epoch": 0.26414733186584816,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.3790134627512275e-06,
+      "loss": 0.6044,
+      "step": 9660
+    },
+    {
+      "epoch": 0.2644207763087735,
+      "grad_norm": 2.75,
+      "learning_rate": 7.376271558224343e-06,
+      "loss": 0.7373,
+      "step": 9670
+    },
+    {
+      "epoch": 0.2646942207516988,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.3735296536974585e-06,
+      "loss": 0.7098,
+      "step": 9680
+    },
+    {
+      "epoch": 0.26496766519462406,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.370787749170575e-06,
+      "loss": 0.7358,
+      "step": 9690
+    },
+    {
+      "epoch": 0.2652411096375494,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.3680458446436894e-06,
+      "loss": 0.6778,
+      "step": 9700
+    },
+    {
+      "epoch": 0.2655145540804747,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.365303940116806e-06,
+      "loss": 0.7236,
+      "step": 9710
+    },
+    {
+      "epoch": 0.2657879985234,
+      "grad_norm": 2.75,
+      "learning_rate": 7.362562035589922e-06,
+      "loss": 0.7346,
+      "step": 9720
+    },
+    {
+      "epoch": 0.26606144296632533,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.359820131063037e-06,
+      "loss": 0.6637,
+      "step": 9730
+    },
+    {
+      "epoch": 0.2663348874092506,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.357078226536153e-06,
+      "loss": 0.7364,
+      "step": 9740
+    },
+    {
+      "epoch": 0.2666083318521759,
+      "grad_norm": 3.59375,
+      "learning_rate": 7.354336322009268e-06,
+      "loss": 0.7688,
+      "step": 9750
+    },
+    {
+      "epoch": 0.26688177629510124,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.351594417482384e-06,
+      "loss": 0.7163,
+      "step": 9760
+    },
+    {
+      "epoch": 0.26715522073802656,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.3488525129555e-06,
+      "loss": 0.7108,
+      "step": 9770
+    },
+    {
+      "epoch": 0.2674286651809519,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.346110608428615e-06,
+      "loss": 0.7109,
+      "step": 9780
+    },
+    {
+      "epoch": 0.2677021096238772,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.343368703901731e-06,
+      "loss": 0.7241,
+      "step": 9790
+    },
+    {
+      "epoch": 0.26797555406680246,
+      "grad_norm": 2.890625,
+      "learning_rate": 7.340626799374846e-06,
+      "loss": 0.7336,
+      "step": 9800
+    },
+    {
+      "epoch": 0.2682489985097278,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.337884894847962e-06,
+      "loss": 0.7293,
+      "step": 9810
+    },
+    {
+      "epoch": 0.2685224429526531,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.335142990321078e-06,
+      "loss": 0.7154,
+      "step": 9820
+    },
+    {
+      "epoch": 0.2687958873955784,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.332401085794193e-06,
+      "loss": 0.6306,
+      "step": 9830
+    },
+    {
+      "epoch": 0.26906933183850373,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.329659181267309e-06,
+      "loss": 0.6722,
+      "step": 9840
+    },
+    {
+      "epoch": 0.269342776281429,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.326917276740424e-06,
+      "loss": 0.7869,
+      "step": 9850
+    },
+    {
+      "epoch": 0.2696162207243543,
+      "grad_norm": 3.1875,
+      "learning_rate": 7.32417537221354e-06,
+      "loss": 0.8045,
+      "step": 9860
+    },
+    {
+      "epoch": 0.26988966516727964,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.321433467686656e-06,
+      "loss": 0.8061,
+      "step": 9870
+    },
+    {
+      "epoch": 0.27016310961020495,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.318691563159771e-06,
+      "loss": 0.6681,
+      "step": 9880
+    },
+    {
+      "epoch": 0.2704365540531303,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.315949658632887e-06,
+      "loss": 0.6793,
+      "step": 9890
+    },
+    {
+      "epoch": 0.27070999849605554,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.313207754106003e-06,
+      "loss": 0.7424,
+      "step": 9900
+    },
+    {
+      "epoch": 0.27098344293898086,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.310465849579118e-06,
+      "loss": 0.7068,
+      "step": 9910
+    },
+    {
+      "epoch": 0.2712568873819062,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.307723945052234e-06,
+      "loss": 0.7346,
+      "step": 9920
+    },
+    {
+      "epoch": 0.2715303318248315,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.304982040525349e-06,
+      "loss": 0.711,
+      "step": 9930
+    },
+    {
+      "epoch": 0.2718037762677568,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.302240135998465e-06,
+      "loss": 0.702,
+      "step": 9940
+    },
+    {
+      "epoch": 0.27207722071068213,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.2994982314715816e-06,
+      "loss": 0.6556,
+      "step": 9950
+    },
+    {
+      "epoch": 0.2723506651536074,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.296756326944696e-06,
+      "loss": 0.7083,
+      "step": 9960
+    },
+    {
+      "epoch": 0.2726241095965327,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.2940144224178125e-06,
+      "loss": 0.7038,
+      "step": 9970
+    },
+    {
+      "epoch": 0.27289755403945803,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.291272517890927e-06,
+      "loss": 0.5966,
+      "step": 9980
+    },
+    {
+      "epoch": 0.27317099848238335,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.2885306133640435e-06,
+      "loss": 0.73,
+      "step": 9990
+    },
+    {
+      "epoch": 0.2734444429253087,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.285788708837159e-06,
+      "loss": 0.6524,
+      "step": 10000
+    },
+    {
+      "epoch": 0.27371788736823394,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.2830468043102745e-06,
+      "loss": 0.6547,
+      "step": 10010
+    },
+    {
+      "epoch": 0.27399133181115926,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.28030489978339e-06,
+      "loss": 0.7095,
+      "step": 10020
+    },
+    {
+      "epoch": 0.2742647762540846,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.277562995256505e-06,
+      "loss": 0.7374,
+      "step": 10030
+    },
+    {
+      "epoch": 0.2745382206970099,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.274821090729621e-06,
+      "loss": 0.7398,
+      "step": 10040
+    },
+    {
+      "epoch": 0.2748116651399352,
+      "grad_norm": 2.09375,
+      "learning_rate": 7.272079186202737e-06,
+      "loss": 0.6579,
+      "step": 10050
+    },
+    {
+      "epoch": 0.2750851095828605,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.269337281675853e-06,
+      "loss": 0.7178,
+      "step": 10060
+    },
+    {
+      "epoch": 0.2753585540257858,
+      "grad_norm": 3.0,
+      "learning_rate": 7.266595377148968e-06,
+      "loss": 0.6846,
+      "step": 10070
+    },
+    {
+      "epoch": 0.2756319984687111,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.2638534726220845e-06,
+      "loss": 0.712,
+      "step": 10080
+    },
+    {
+      "epoch": 0.27590544291163643,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.261111568095199e-06,
+      "loss": 0.6423,
+      "step": 10090
+    },
+    {
+      "epoch": 0.27617888735456175,
+      "grad_norm": 2.625,
+      "learning_rate": 7.2583696635683155e-06,
+      "loss": 0.7389,
+      "step": 10100
+    },
+    {
+      "epoch": 0.27645233179748707,
+      "grad_norm": 3.109375,
+      "learning_rate": 7.25562775904143e-06,
+      "loss": 0.6908,
+      "step": 10110
+    },
+    {
+      "epoch": 0.27672577624041234,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.2528858545145464e-06,
+      "loss": 0.6775,
+      "step": 10120
+    },
+    {
+      "epoch": 0.27699922068333765,
+      "grad_norm": 2.90625,
+      "learning_rate": 7.250143949987663e-06,
+      "loss": 0.7069,
+      "step": 10130
+    },
+    {
+      "epoch": 0.277272665126263,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.247402045460777e-06,
+      "loss": 0.765,
+      "step": 10140
+    },
+    {
+      "epoch": 0.2775461095691883,
+      "grad_norm": 2.765625,
+      "learning_rate": 7.244660140933894e-06,
+      "loss": 0.6856,
+      "step": 10150
+    },
+    {
+      "epoch": 0.2778195540121136,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.241918236407008e-06,
+      "loss": 0.7105,
+      "step": 10160
+    },
+    {
+      "epoch": 0.2780929984550389,
+      "grad_norm": 2.203125,
+      "learning_rate": 7.239176331880125e-06,
+      "loss": 0.638,
+      "step": 10170
+    },
+    {
+      "epoch": 0.2783664428979642,
+      "grad_norm": 2.34375,
+      "learning_rate": 7.23643442735324e-06,
+      "loss": 0.7125,
+      "step": 10180
+    },
+    {
+      "epoch": 0.2786398873408895,
+      "grad_norm": 3.375,
+      "learning_rate": 7.233692522826356e-06,
+      "loss": 0.6691,
+      "step": 10190
+    },
+    {
+      "epoch": 0.27891333178381483,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.230950618299471e-06,
+      "loss": 0.6366,
+      "step": 10200
+    },
+    {
+      "epoch": 0.27918677622674015,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.228208713772587e-06,
+      "loss": 0.6189,
+      "step": 10210
+    },
+    {
+      "epoch": 0.2794602206696654,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.225466809245703e-06,
+      "loss": 0.7729,
+      "step": 10220
+    },
+    {
+      "epoch": 0.27973366511259073,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.2227249047188184e-06,
+      "loss": 0.7949,
+      "step": 10230
+    },
+    {
+      "epoch": 0.28000710955551605,
+      "grad_norm": 2.625,
+      "learning_rate": 7.219983000191934e-06,
+      "loss": 0.6995,
+      "step": 10240
+    },
+    {
+      "epoch": 0.2802805539984414,
+      "grad_norm": 2.625,
+      "learning_rate": 7.217241095665049e-06,
+      "loss": 0.7009,
+      "step": 10250
+    },
+    {
+      "epoch": 0.2805539984413667,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.214499191138166e-06,
+      "loss": 0.6849,
+      "step": 10260
+    },
+    {
+      "epoch": 0.280827442884292,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.21175728661128e-06,
+      "loss": 0.7117,
+      "step": 10270
+    },
+    {
+      "epoch": 0.2811008873272173,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.209015382084397e-06,
+      "loss": 0.7217,
+      "step": 10280
+    },
+    {
+      "epoch": 0.2813743317701426,
+      "grad_norm": 3.546875,
+      "learning_rate": 7.206273477557511e-06,
+      "loss": 0.687,
+      "step": 10290
+    },
+    {
+      "epoch": 0.2816477762130679,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.203531573030628e-06,
+      "loss": 0.7166,
+      "step": 10300
+    },
+    {
+      "epoch": 0.28192122065599323,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.200789668503744e-06,
+      "loss": 0.6192,
+      "step": 10310
+    },
+    {
+      "epoch": 0.28219466509891855,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.198047763976859e-06,
+      "loss": 0.683,
+      "step": 10320
+    },
+    {
+      "epoch": 0.2824681095418438,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.195305859449975e-06,
+      "loss": 0.7471,
+      "step": 10330
+    },
+    {
+      "epoch": 0.28274155398476913,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.19256395492309e-06,
+      "loss": 0.7108,
+      "step": 10340
+    },
+    {
+      "epoch": 0.28301499842769445,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.189822050396206e-06,
+      "loss": 0.6842,
+      "step": 10350
+    },
+    {
+      "epoch": 0.28328844287061977,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.187080145869321e-06,
+      "loss": 0.6432,
+      "step": 10360
+    },
+    {
+      "epoch": 0.2835618873135451,
+      "grad_norm": 2.75,
+      "learning_rate": 7.184338241342437e-06,
+      "loss": 0.7452,
+      "step": 10370
+    },
+    {
+      "epoch": 0.28383533175647035,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.181596336815553e-06,
+      "loss": 0.6307,
+      "step": 10380
+    },
+    {
+      "epoch": 0.2841087761993957,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.178854432288669e-06,
+      "loss": 0.6685,
+      "step": 10390
+    },
+    {
+      "epoch": 0.284382220642321,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.176112527761784e-06,
+      "loss": 0.6847,
+      "step": 10400
+    },
+    {
+      "epoch": 0.2846556650852463,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.1733706232349e-06,
+      "loss": 0.7554,
+      "step": 10410
+    },
+    {
+      "epoch": 0.28492910952817163,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.170628718708015e-06,
+      "loss": 0.7918,
+      "step": 10420
+    },
+    {
+      "epoch": 0.2852025539710969,
+      "grad_norm": 2.3125,
+      "learning_rate": 7.167886814181131e-06,
+      "loss": 0.6851,
+      "step": 10430
+    },
+    {
+      "epoch": 0.2854759984140222,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.165144909654247e-06,
+      "loss": 0.7382,
+      "step": 10440
+    },
+    {
+      "epoch": 0.28574944285694753,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.162403005127362e-06,
+      "loss": 0.7446,
+      "step": 10450
+    },
+    {
+      "epoch": 0.28602288729987285,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.159661100600478e-06,
+      "loss": 0.6616,
+      "step": 10460
+    },
+    {
+      "epoch": 0.28629633174279817,
+      "grad_norm": 2.296875,
+      "learning_rate": 7.1569191960735926e-06,
+      "loss": 0.6103,
+      "step": 10470
+    },
+    {
+      "epoch": 0.2865697761857235,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.154177291546709e-06,
+      "loss": 0.6605,
+      "step": 10480
+    },
+    {
+      "epoch": 0.28684322062864875,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.151435387019825e-06,
+      "loss": 0.6494,
+      "step": 10490
+    },
+    {
+      "epoch": 0.2871166650715741,
+      "grad_norm": 2.328125,
+      "learning_rate": 7.14869348249294e-06,
+      "loss": 0.7084,
+      "step": 10500
+    },
+    {
+      "epoch": 0.2873901095144994,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.145951577966056e-06,
+      "loss": 0.6977,
+      "step": 10510
+    },
+    {
+      "epoch": 0.2876635539574247,
+      "grad_norm": 2.75,
+      "learning_rate": 7.143209673439171e-06,
+      "loss": 0.6618,
+      "step": 10520
+    },
+    {
+      "epoch": 0.28793699840035003,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.140467768912287e-06,
+      "loss": 0.7375,
+      "step": 10530
+    },
+    {
+      "epoch": 0.2882104428432753,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.1377258643854035e-06,
+      "loss": 0.5987,
+      "step": 10540
+    },
+    {
+      "epoch": 0.2884838872862006,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.134983959858518e-06,
+      "loss": 0.6533,
+      "step": 10550
+    },
+    {
+      "epoch": 0.28875733172912593,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.1322420553316344e-06,
+      "loss": 0.6927,
+      "step": 10560
+    },
+    {
+      "epoch": 0.28903077617205125,
+      "grad_norm": 3.25,
+      "learning_rate": 7.12950015080475e-06,
+      "loss": 0.7638,
+      "step": 10570
+    },
+    {
+      "epoch": 0.28930422061497657,
+      "grad_norm": 2.484375,
+      "learning_rate": 7.126758246277865e-06,
+      "loss": 0.6918,
+      "step": 10580
+    },
+    {
+      "epoch": 0.28957766505790183,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.124016341750981e-06,
+      "loss": 0.7578,
+      "step": 10590
+    },
+    {
+      "epoch": 0.28985110950082715,
+      "grad_norm": 2.28125,
+      "learning_rate": 7.121274437224096e-06,
+      "loss": 0.6612,
+      "step": 10600
+    },
+    {
+      "epoch": 0.2901245539437525,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.118532532697212e-06,
+      "loss": 0.6589,
+      "step": 10610
+    },
+    {
+      "epoch": 0.2903979983866778,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.115790628170328e-06,
+      "loss": 0.691,
+      "step": 10620
+    },
+    {
+      "epoch": 0.2906714428296031,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.113048723643443e-06,
+      "loss": 0.7327,
+      "step": 10630
+    },
+    {
+      "epoch": 0.29094488727252843,
+      "grad_norm": 3.0,
+      "learning_rate": 7.110306819116559e-06,
+      "loss": 0.7485,
+      "step": 10640
+    },
+    {
+      "epoch": 0.2912183317154537,
+      "grad_norm": 2.75,
+      "learning_rate": 7.107564914589674e-06,
+      "loss": 0.676,
+      "step": 10650
+    },
+    {
+      "epoch": 0.291491776158379,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.10482301006279e-06,
+      "loss": 0.7206,
+      "step": 10660
+    },
+    {
+      "epoch": 0.29176522060130433,
+      "grad_norm": 2.625,
+      "learning_rate": 7.1020811055359064e-06,
+      "loss": 0.6657,
+      "step": 10670
+    },
+    {
+      "epoch": 0.29203866504422965,
+      "grad_norm": 2.609375,
+      "learning_rate": 7.099339201009021e-06,
+      "loss": 0.6122,
+      "step": 10680
+    },
+    {
+      "epoch": 0.29231210948715497,
+      "grad_norm": 2.46875,
+      "learning_rate": 7.096597296482137e-06,
+      "loss": 0.6871,
+      "step": 10690
+    },
+    {
+      "epoch": 0.29258555393008023,
+      "grad_norm": 2.640625,
+      "learning_rate": 7.093855391955252e-06,
+      "loss": 0.664,
+      "step": 10700
+    },
+    {
+      "epoch": 0.29285899837300555,
+      "grad_norm": 4.125,
+      "learning_rate": 7.091113487428368e-06,
+      "loss": 0.746,
+      "step": 10710
+    },
+    {
+      "epoch": 0.29313244281593087,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.088371582901485e-06,
+      "loss": 0.7636,
+      "step": 10720
+    },
+    {
+      "epoch": 0.2934058872588562,
+      "grad_norm": 2.21875,
+      "learning_rate": 7.085629678374599e-06,
+      "loss": 0.6021,
+      "step": 10730
+    },
+    {
+      "epoch": 0.2936793317017815,
+      "grad_norm": 3.75,
+      "learning_rate": 7.082887773847716e-06,
+      "loss": 0.6665,
+      "step": 10740
+    },
+    {
+      "epoch": 0.2939527761447068,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.080145869320831e-06,
+      "loss": 0.6754,
+      "step": 10750
+    },
+    {
+      "epoch": 0.2942262205876321,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.077403964793947e-06,
+      "loss": 0.761,
+      "step": 10760
+    },
+    {
+      "epoch": 0.2944996650305574,
+      "grad_norm": 2.875,
+      "learning_rate": 7.074662060267062e-06,
+      "loss": 0.7016,
+      "step": 10770
+    },
+    {
+      "epoch": 0.29477310947348273,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.0719201557401776e-06,
+      "loss": 0.7447,
+      "step": 10780
+    },
+    {
+      "epoch": 0.29504655391640805,
+      "grad_norm": 2.375,
+      "learning_rate": 7.069178251213293e-06,
+      "loss": 0.6428,
+      "step": 10790
+    },
+    {
+      "epoch": 0.29531999835933337,
+      "grad_norm": 2.515625,
+      "learning_rate": 7.066436346686409e-06,
+      "loss": 0.6487,
+      "step": 10800
+    },
+    {
+      "epoch": 0.29559344280225863,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.063694442159524e-06,
+      "loss": 0.7809,
+      "step": 10810
+    },
+    {
+      "epoch": 0.29586688724518395,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.06095253763264e-06,
+      "loss": 0.7382,
+      "step": 10820
+    },
+    {
+      "epoch": 0.29614033168810927,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.058210633105755e-06,
+      "loss": 0.71,
+      "step": 10830
+    },
+    {
+      "epoch": 0.2964137761310346,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.055468728578871e-06,
+      "loss": 0.6666,
+      "step": 10840
+    },
+    {
+      "epoch": 0.2966872205739599,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.052726824051988e-06,
+      "loss": 0.728,
+      "step": 10850
+    },
+    {
+      "epoch": 0.2969606650168852,
+      "grad_norm": 3.34375,
+      "learning_rate": 7.049984919525102e-06,
+      "loss": 0.695,
+      "step": 10860
+    },
+    {
+      "epoch": 0.2972341094598105,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.047243014998219e-06,
+      "loss": 0.6523,
+      "step": 10870
+    },
+    {
+      "epoch": 0.2975075539027358,
+      "grad_norm": 2.9375,
+      "learning_rate": 7.044501110471333e-06,
+      "loss": 0.7291,
+      "step": 10880
+    },
+    {
+      "epoch": 0.29778099834566113,
+      "grad_norm": 2.578125,
+      "learning_rate": 7.0417592059444496e-06,
+      "loss": 0.6904,
+      "step": 10890
+    },
+    {
+      "epoch": 0.29805444278858645,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.039017301417566e-06,
+      "loss": 0.7389,
+      "step": 10900
+    },
+    {
+      "epoch": 0.2983278872315117,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.0362753968906805e-06,
+      "loss": 0.6369,
+      "step": 10910
+    },
+    {
+      "epoch": 0.29860133167443703,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.033533492363797e-06,
+      "loss": 0.6776,
+      "step": 10920
+    },
+    {
+      "epoch": 0.29887477611736235,
+      "grad_norm": 2.953125,
+      "learning_rate": 7.030791587836912e-06,
+      "loss": 0.6531,
+      "step": 10930
+    },
+    {
+      "epoch": 0.29914822056028767,
+      "grad_norm": 2.8125,
+      "learning_rate": 7.028049683310028e-06,
+      "loss": 0.7285,
+      "step": 10940
+    },
+    {
+      "epoch": 0.299421665003213,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.025307778783143e-06,
+      "loss": 0.6325,
+      "step": 10950
+    },
+    {
+      "epoch": 0.2996951094461383,
+      "grad_norm": 2.546875,
+      "learning_rate": 7.022565874256259e-06,
+      "loss": 0.6957,
+      "step": 10960
+    },
+    {
+      "epoch": 0.29996855388906357,
+      "grad_norm": 3.171875,
+      "learning_rate": 7.019823969729374e-06,
+      "loss": 0.7209,
+      "step": 10970
+    },
+    {
+      "epoch": 0.3002419983319889,
+      "grad_norm": 2.4375,
+      "learning_rate": 7.017082065202491e-06,
+      "loss": 0.7059,
+      "step": 10980
+    },
+    {
+      "epoch": 0.3005154427749142,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.014340160675605e-06,
+      "loss": 0.7217,
+      "step": 10990
+    },
+    {
+      "epoch": 0.30078888721783953,
+      "grad_norm": 2.53125,
+      "learning_rate": 7.0115982561487216e-06,
+      "loss": 0.7045,
+      "step": 11000
+    },
+    {
+      "epoch": 0.30106233166076485,
+      "grad_norm": 3.09375,
+      "learning_rate": 7.008856351621837e-06,
+      "loss": 0.6908,
+      "step": 11010
+    },
+    {
+      "epoch": 0.3013357761036901,
+      "grad_norm": 2.390625,
+      "learning_rate": 7.0061144470949525e-06,
+      "loss": 0.7213,
+      "step": 11020
+    },
+    {
+      "epoch": 0.30160922054661543,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.003372542568069e-06,
+      "loss": 0.7418,
+      "step": 11030
+    },
+    {
+      "epoch": 0.30188266498954075,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.0006306380411835e-06,
+      "loss": 0.6833,
+      "step": 11040
+    },
+    {
+      "epoch": 0.30215610943246607,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.9978887335143e-06,
+      "loss": 0.7172,
+      "step": 11050
+    },
+    {
+      "epoch": 0.3024295538753914,
+      "grad_norm": 2.875,
+      "learning_rate": 6.9951468289874145e-06,
+      "loss": 0.7148,
+      "step": 11060
+    },
+    {
+      "epoch": 0.30270299831831665,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.992404924460531e-06,
+      "loss": 0.7217,
+      "step": 11070
+    },
+    {
+      "epoch": 0.30297644276124197,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.989663019933647e-06,
+      "loss": 0.6593,
+      "step": 11080
+    },
+    {
+      "epoch": 0.3032498872041673,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.986921115406762e-06,
+      "loss": 0.7034,
+      "step": 11090
+    },
+    {
+      "epoch": 0.3035233316470926,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.984179210879878e-06,
+      "loss": 0.7046,
+      "step": 11100
+    },
+    {
+      "epoch": 0.30379677609001793,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.9814373063529936e-06,
+      "loss": 0.6705,
+      "step": 11110
+    },
+    {
+      "epoch": 0.3040702205329432,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.978695401826109e-06,
+      "loss": 0.7415,
+      "step": 11120
+    },
+    {
+      "epoch": 0.3043436649758685,
+      "grad_norm": 2.875,
+      "learning_rate": 6.9759534972992245e-06,
+      "loss": 0.6866,
+      "step": 11130
+    },
+    {
+      "epoch": 0.30461710941879383,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.97321159277234e-06,
+      "loss": 0.729,
+      "step": 11140
+    },
+    {
+      "epoch": 0.30489055386171915,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.9704696882454555e-06,
+      "loss": 0.681,
+      "step": 11150
+    },
+    {
+      "epoch": 0.30516399830464447,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.967727783718572e-06,
+      "loss": 0.6561,
+      "step": 11160
+    },
+    {
+      "epoch": 0.3054374427475698,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.964985879191687e-06,
+      "loss": 0.6162,
+      "step": 11170
+    },
+    {
+      "epoch": 0.30571088719049505,
+      "grad_norm": 2.625,
+      "learning_rate": 6.962243974664803e-06,
+      "loss": 0.6543,
+      "step": 11180
+    },
+    {
+      "epoch": 0.30598433163342037,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.959502070137918e-06,
+      "loss": 0.7339,
+      "step": 11190
+    },
+    {
+      "epoch": 0.3062577760763457,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.956760165611034e-06,
+      "loss": 0.6975,
+      "step": 11200
+    },
+    {
+      "epoch": 0.306531220519271,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.95401826108415e-06,
+      "loss": 0.6994,
+      "step": 11210
+    },
+    {
+      "epoch": 0.3068046649621963,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.951276356557265e-06,
+      "loss": 0.6846,
+      "step": 11220
+    },
+    {
+      "epoch": 0.3070781094051216,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.948534452030381e-06,
+      "loss": 0.6327,
+      "step": 11230
+    },
+    {
+      "epoch": 0.3073515538480469,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.945792547503496e-06,
+      "loss": 0.6395,
+      "step": 11240
+    },
+    {
+      "epoch": 0.30762499829097223,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.943050642976612e-06,
+      "loss": 0.6184,
+      "step": 11250
+    },
+    {
+      "epoch": 0.30789844273389755,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.940308738449728e-06,
+      "loss": 0.7189,
+      "step": 11260
+    },
+    {
+      "epoch": 0.30817188717682287,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.937566833922843e-06,
+      "loss": 0.6948,
+      "step": 11270
+    },
+    {
+      "epoch": 0.30844533161974813,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.934824929395959e-06,
+      "loss": 0.7341,
+      "step": 11280
+    },
+    {
+      "epoch": 0.30871877606267345,
+      "grad_norm": 2.875,
+      "learning_rate": 6.932083024869075e-06,
+      "loss": 0.7089,
+      "step": 11290
+    },
+    {
+      "epoch": 0.30899222050559877,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.92934112034219e-06,
+      "loss": 0.7374,
+      "step": 11300
+    },
+    {
+      "epoch": 0.3092656649485241,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.926599215815306e-06,
+      "loss": 0.7596,
+      "step": 11310
+    },
+    {
+      "epoch": 0.3095391093914494,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.923857311288421e-06,
+      "loss": 0.6692,
+      "step": 11320
+    },
+    {
+      "epoch": 0.3098125538343747,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.9211154067615376e-06,
+      "loss": 0.7374,
+      "step": 11330
+    },
+    {
+      "epoch": 0.3100859982773,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.918373502234653e-06,
+      "loss": 0.6903,
+      "step": 11340
+    },
+    {
+      "epoch": 0.3103594427202253,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.9156315977077685e-06,
+      "loss": 0.7144,
+      "step": 11350
+    },
+    {
+      "epoch": 0.31063288716315063,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.912889693180884e-06,
+      "loss": 0.6638,
+      "step": 11360
+    },
+    {
+      "epoch": 0.31090633160607595,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.9101477886539995e-06,
+      "loss": 0.7098,
+      "step": 11370
+    },
+    {
+      "epoch": 0.31117977604900127,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.907405884127115e-06,
+      "loss": 0.6232,
+      "step": 11380
+    },
+    {
+      "epoch": 0.31145322049192653,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.904663979600231e-06,
+      "loss": 0.7413,
+      "step": 11390
+    },
+    {
+      "epoch": 0.31172666493485185,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.901922075073346e-06,
+      "loss": 0.7506,
+      "step": 11400
+    },
+    {
+      "epoch": 0.31200010937777717,
+      "grad_norm": 3.234375,
+      "learning_rate": 6.899180170546462e-06,
+      "loss": 0.6956,
+      "step": 11410
+    },
+    {
+      "epoch": 0.3122735538207025,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.896438266019577e-06,
+      "loss": 0.8072,
+      "step": 11420
+    },
+    {
+      "epoch": 0.3125469982636278,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.893696361492693e-06,
+      "loss": 0.755,
+      "step": 11430
+    },
+    {
+      "epoch": 0.31282044270655307,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.8909544569658095e-06,
+      "loss": 0.7009,
+      "step": 11440
+    },
+    {
+      "epoch": 0.3130938871494784,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.888212552438924e-06,
+      "loss": 0.644,
+      "step": 11450
+    },
+    {
+      "epoch": 0.3133673315924037,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.8854706479120405e-06,
+      "loss": 0.6101,
+      "step": 11460
+    },
+    {
+      "epoch": 0.313640776035329,
+      "grad_norm": 2.75,
+      "learning_rate": 6.882728743385156e-06,
+      "loss": 0.7123,
+      "step": 11470
+    },
+    {
+      "epoch": 0.31391422047825435,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.8799868388582715e-06,
+      "loss": 0.5996,
+      "step": 11480
+    },
+    {
+      "epoch": 0.31418766492117967,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.877244934331388e-06,
+      "loss": 0.7435,
+      "step": 11490
+    },
+    {
+      "epoch": 0.31446110936410493,
+      "grad_norm": 3.484375,
+      "learning_rate": 6.8745030298045024e-06,
+      "loss": 0.7611,
+      "step": 11500
+    },
+    {
+      "epoch": 0.31473455380703025,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.871761125277619e-06,
+      "loss": 0.6308,
+      "step": 11510
+    },
+    {
+      "epoch": 0.31500799824995557,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.869019220750734e-06,
+      "loss": 0.655,
+      "step": 11520
+    },
+    {
+      "epoch": 0.3152814426928809,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.86627731622385e-06,
+      "loss": 0.7073,
+      "step": 11530
+    },
+    {
+      "epoch": 0.3155548871358062,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.863535411696965e-06,
+      "loss": 0.6477,
+      "step": 11540
+    },
+    {
+      "epoch": 0.31582833157873147,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.860793507170081e-06,
+      "loss": 0.7341,
+      "step": 11550
+    },
+    {
+      "epoch": 0.3161017760216568,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.858051602643196e-06,
+      "loss": 0.7324,
+      "step": 11560
+    },
+    {
+      "epoch": 0.3163752204645821,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.8553096981163125e-06,
+      "loss": 0.6826,
+      "step": 11570
+    },
+    {
+      "epoch": 0.3166486649075074,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.852567793589427e-06,
+      "loss": 0.687,
+      "step": 11580
+    },
+    {
+      "epoch": 0.31692210935043275,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.8498258890625435e-06,
+      "loss": 0.7307,
+      "step": 11590
+    },
+    {
+      "epoch": 0.317195553793358,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.847083984535658e-06,
+      "loss": 0.6419,
+      "step": 11600
+    },
+    {
+      "epoch": 0.31746899823628333,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.8443420800087744e-06,
+      "loss": 0.6693,
+      "step": 11610
+    },
+    {
+      "epoch": 0.31774244267920865,
+      "grad_norm": 2.5,
+      "learning_rate": 6.841600175481891e-06,
+      "loss": 0.6882,
+      "step": 11620
+    },
+    {
+      "epoch": 0.31801588712213397,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.838858270955005e-06,
+      "loss": 0.6351,
+      "step": 11630
+    },
+    {
+      "epoch": 0.3182893315650593,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.836116366428122e-06,
+      "loss": 0.767,
+      "step": 11640
+    },
+    {
+      "epoch": 0.3185627760079846,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.833374461901238e-06,
+      "loss": 0.72,
+      "step": 11650
+    },
+    {
+      "epoch": 0.31883622045090987,
+      "grad_norm": 2.75,
+      "learning_rate": 6.830632557374353e-06,
+      "loss": 0.653,
+      "step": 11660
+    },
+    {
+      "epoch": 0.3191096648938352,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.827890652847469e-06,
+      "loss": 0.6554,
+      "step": 11670
+    },
+    {
+      "epoch": 0.3193831093367605,
+      "grad_norm": 2.375,
+      "learning_rate": 6.825148748320584e-06,
+      "loss": 0.5828,
+      "step": 11680
+    },
+    {
+      "epoch": 0.3196565537796858,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.8224068437937e-06,
+      "loss": 0.7171,
+      "step": 11690
+    },
+    {
+      "epoch": 0.31992999822261114,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.8196649392668155e-06,
+      "loss": 0.6954,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3202034426655364,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.816923034739931e-06,
+      "loss": 0.7713,
+      "step": 11710
+    },
+    {
+      "epoch": 0.3204768871084617,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.8141811302130464e-06,
+      "loss": 0.728,
+      "step": 11720
+    },
+    {
+      "epoch": 0.32075033155138705,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.811439225686162e-06,
+      "loss": 0.7415,
+      "step": 11730
+    },
+    {
+      "epoch": 0.32102377599431237,
+      "grad_norm": 2.5,
+      "learning_rate": 6.808697321159277e-06,
+      "loss": 0.6014,
+      "step": 11740
+    },
+    {
+      "epoch": 0.3212972204372377,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.805955416632394e-06,
+      "loss": 0.6663,
+      "step": 11750
+    },
+    {
+      "epoch": 0.32157066488016295,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.803213512105508e-06,
+      "loss": 0.7049,
+      "step": 11760
+    },
+    {
+      "epoch": 0.32184410932308827,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.800471607578625e-06,
+      "loss": 0.6885,
+      "step": 11770
+    },
+    {
+      "epoch": 0.3221175537660136,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.797729703051741e-06,
+      "loss": 0.6875,
+      "step": 11780
+    },
+    {
+      "epoch": 0.3223909982089389,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.794987798524856e-06,
+      "loss": 0.7229,
+      "step": 11790
+    },
+    {
+      "epoch": 0.3226644426518642,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.792245893997972e-06,
+      "loss": 0.7529,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3229378870947895,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.789503989471087e-06,
+      "loss": 0.7187,
+      "step": 11810
+    },
+    {
+      "epoch": 0.3232113315377148,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.786762084944203e-06,
+      "loss": 0.7406,
+      "step": 11820
+    },
+    {
+      "epoch": 0.3234847759806401,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.784020180417319e-06,
+      "loss": 0.7044,
+      "step": 11830
+    },
+    {
+      "epoch": 0.32375822042356545,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.781278275890434e-06,
+      "loss": 0.6815,
+      "step": 11840
+    },
+    {
+      "epoch": 0.32403166486649077,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.77853637136355e-06,
+      "loss": 0.6475,
+      "step": 11850
+    },
+    {
+      "epoch": 0.3243051093094161,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.775794466836665e-06,
+      "loss": 0.6687,
+      "step": 11860
+    },
+    {
+      "epoch": 0.32457855375234135,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.773052562309781e-06,
+      "loss": 0.6364,
+      "step": 11870
+    },
+    {
+      "epoch": 0.32485199819526667,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.770310657782897e-06,
+      "loss": 0.6746,
+      "step": 11880
+    },
+    {
+      "epoch": 0.325125442638192,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.767568753256012e-06,
+      "loss": 0.6738,
+      "step": 11890
+    },
+    {
+      "epoch": 0.3253988870811173,
+      "grad_norm": 2.875,
+      "learning_rate": 6.764826848729128e-06,
+      "loss": 0.5849,
+      "step": 11900
+    },
+    {
+      "epoch": 0.3256723315240426,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.762084944202243e-06,
+      "loss": 0.7135,
+      "step": 11910
+    },
+    {
+      "epoch": 0.3259457759669679,
+      "grad_norm": 2.625,
+      "learning_rate": 6.759343039675359e-06,
+      "loss": 0.7883,
+      "step": 11920
+    },
+    {
+      "epoch": 0.3262192204098932,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.756601135148475e-06,
+      "loss": 0.6686,
+      "step": 11930
+    },
+    {
+      "epoch": 0.3264926648528185,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.75385923062159e-06,
+      "loss": 0.7054,
+      "step": 11940
+    },
+    {
+      "epoch": 0.32676610929574385,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.751117326094706e-06,
+      "loss": 0.6563,
+      "step": 11950
+    },
+    {
+      "epoch": 0.32703955373866916,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.748375421567822e-06,
+      "loss": 0.7064,
+      "step": 11960
+    },
+    {
+      "epoch": 0.32731299818159443,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.745633517040937e-06,
+      "loss": 0.6405,
+      "step": 11970
+    },
+    {
+      "epoch": 0.32758644262451975,
+      "grad_norm": 2.015625,
+      "learning_rate": 6.742891612514053e-06,
+      "loss": 0.6231,
+      "step": 11980
+    },
+    {
+      "epoch": 0.32785988706744507,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.740149707987168e-06,
+      "loss": 0.6941,
+      "step": 11990
+    },
+    {
+      "epoch": 0.3281333315103704,
+      "grad_norm": 3.0,
+      "learning_rate": 6.737407803460284e-06,
+      "loss": 0.6867,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3284067759532957,
+      "grad_norm": 2.078125,
+      "learning_rate": 6.7346658989334005e-06,
+      "loss": 0.6154,
+      "step": 12010
+    },
+    {
+      "epoch": 0.328680220396221,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.731923994406515e-06,
+      "loss": 0.6894,
+      "step": 12020
+    },
+    {
+      "epoch": 0.3289536648391463,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.7291820898796315e-06,
+      "loss": 0.663,
+      "step": 12030
+    },
+    {
+      "epoch": 0.3292271092820716,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.726440185352746e-06,
+      "loss": 0.7265,
+      "step": 12040
+    },
+    {
+      "epoch": 0.3295005537249969,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.723698280825862e-06,
+      "loss": 0.7118,
+      "step": 12050
+    },
+    {
+      "epoch": 0.32977399816792224,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.720956376298978e-06,
+      "loss": 0.7145,
+      "step": 12060
+    },
+    {
+      "epoch": 0.33004744261084756,
+      "grad_norm": 3.21875,
+      "learning_rate": 6.718214471772093e-06,
+      "loss": 0.7637,
+      "step": 12070
+    },
+    {
+      "epoch": 0.3303208870537728,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.715472567245209e-06,
+      "loss": 0.6981,
+      "step": 12080
+    },
+    {
+      "epoch": 0.33059433149669815,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.712730662718324e-06,
+      "loss": 0.7494,
+      "step": 12090
+    },
+    {
+      "epoch": 0.33086777593962347,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.70998875819144e-06,
+      "loss": 0.6865,
+      "step": 12100
+    },
+    {
+      "epoch": 0.3311412203825488,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.707246853664556e-06,
+      "loss": 0.7279,
+      "step": 12110
+    },
+    {
+      "epoch": 0.3314146648254741,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.704504949137672e-06,
+      "loss": 0.7386,
+      "step": 12120
+    },
+    {
+      "epoch": 0.33168810926839937,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.701763044610787e-06,
+      "loss": 0.6342,
+      "step": 12130
+    },
+    {
+      "epoch": 0.3319615537113247,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.6990211400839035e-06,
+      "loss": 0.6769,
+      "step": 12140
+    },
+    {
+      "epoch": 0.33223499815425,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.696279235557018e-06,
+      "loss": 0.6702,
+      "step": 12150
+    },
+    {
+      "epoch": 0.3325084425971753,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.693537331030134e-06,
+      "loss": 0.7738,
+      "step": 12160
+    },
+    {
+      "epoch": 0.33278188704010064,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.690795426503249e-06,
+      "loss": 0.7227,
+      "step": 12170
+    },
+    {
+      "epoch": 0.33305533148302596,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.688053521976365e-06,
+      "loss": 0.7018,
+      "step": 12180
+    },
+    {
+      "epoch": 0.3333287759259512,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.685311617449482e-06,
+      "loss": 0.6944,
+      "step": 12190
+    },
+    {
+      "epoch": 0.33360222036887655,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.682569712922596e-06,
+      "loss": 0.6985,
+      "step": 12200
+    },
+    {
+      "epoch": 0.33387566481180186,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.679827808395713e-06,
+      "loss": 0.685,
+      "step": 12210
+    },
+    {
+      "epoch": 0.3341491092547272,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.677085903868827e-06,
+      "loss": 0.7549,
+      "step": 12220
+    },
+    {
+      "epoch": 0.3344225536976525,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.674343999341944e-06,
+      "loss": 0.705,
+      "step": 12230
+    },
+    {
+      "epoch": 0.33469599814057777,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.671602094815059e-06,
+      "loss": 0.825,
+      "step": 12240
+    },
+    {
+      "epoch": 0.3349694425835031,
+      "grad_norm": 2.375,
+      "learning_rate": 6.668860190288175e-06,
+      "loss": 0.7136,
+      "step": 12250
+    },
+    {
+      "epoch": 0.3352428870264284,
+      "grad_norm": 2.625,
+      "learning_rate": 6.66611828576129e-06,
+      "loss": 0.6553,
+      "step": 12260
+    },
+    {
+      "epoch": 0.3355163314693537,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.6633763812344056e-06,
+      "loss": 0.6817,
+      "step": 12270
+    },
+    {
+      "epoch": 0.33578977591227904,
+      "grad_norm": 3.0,
+      "learning_rate": 6.660634476707522e-06,
+      "loss": 0.7475,
+      "step": 12280
+    },
+    {
+      "epoch": 0.3360632203552043,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.657892572180637e-06,
+      "loss": 0.7105,
+      "step": 12290
+    },
+    {
+      "epoch": 0.3363366647981296,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.655150667653753e-06,
+      "loss": 0.6866,
+      "step": 12300
+    },
+    {
+      "epoch": 0.33661010924105494,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.652408763126868e-06,
+      "loss": 0.7252,
+      "step": 12310
+    },
+    {
+      "epoch": 0.33688355368398026,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.649666858599985e-06,
+      "loss": 0.683,
+      "step": 12320
+    },
+    {
+      "epoch": 0.3371569981269056,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.646924954073099e-06,
+      "loss": 0.6286,
+      "step": 12330
+    },
+    {
+      "epoch": 0.3374304425698309,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.644183049546216e-06,
+      "loss": 0.7069,
+      "step": 12340
+    },
+    {
+      "epoch": 0.33770388701275617,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.64144114501933e-06,
+      "loss": 0.725,
+      "step": 12350
+    },
+    {
+      "epoch": 0.3379773314556815,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.638699240492447e-06,
+      "loss": 0.7511,
+      "step": 12360
+    },
+    {
+      "epoch": 0.3382507758986068,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.635957335965563e-06,
+      "loss": 0.6582,
+      "step": 12370
+    },
+    {
+      "epoch": 0.3385242203415321,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.6332154314386776e-06,
+      "loss": 0.6827,
+      "step": 12380
+    },
+    {
+      "epoch": 0.33879766478445744,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.630473526911794e-06,
+      "loss": 0.6445,
+      "step": 12390
+    },
+    {
+      "epoch": 0.3390711092273827,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.6277316223849085e-06,
+      "loss": 0.6738,
+      "step": 12400
+    },
+    {
+      "epoch": 0.339344553670308,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.624989717858025e-06,
+      "loss": 0.6651,
+      "step": 12410
+    },
+    {
+      "epoch": 0.33961799811323334,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.62224781333114e-06,
+      "loss": 0.6363,
+      "step": 12420
+    },
+    {
+      "epoch": 0.33989144255615866,
+      "grad_norm": 2.5,
+      "learning_rate": 6.619505908804256e-06,
+      "loss": 0.7364,
+      "step": 12430
+    },
+    {
+      "epoch": 0.340164886999084,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.616764004277372e-06,
+      "loss": 0.6807,
+      "step": 12440
+    },
+    {
+      "epoch": 0.34043833144200925,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.614022099750487e-06,
+      "loss": 0.7298,
+      "step": 12450
+    },
+    {
+      "epoch": 0.34071177588493456,
+      "grad_norm": 2.703125,
+      "learning_rate": 6.611280195223603e-06,
+      "loss": 0.7365,
+      "step": 12460
+    },
+    {
+      "epoch": 0.3409852203278599,
+      "grad_norm": 3.125,
+      "learning_rate": 6.608538290696719e-06,
+      "loss": 0.7325,
+      "step": 12470
+    },
+    {
+      "epoch": 0.3412586647707852,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.605796386169834e-06,
+      "loss": 0.6979,
+      "step": 12480
+    },
+    {
+      "epoch": 0.3415321092137105,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.6030544816429496e-06,
+      "loss": 0.6499,
+      "step": 12490
+    },
+    {
+      "epoch": 0.3418055536566358,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.600312577116066e-06,
+      "loss": 0.6311,
+      "step": 12500
+    },
+    {
+      "epoch": 0.3420789980995611,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.5975706725891805e-06,
+      "loss": 0.7192,
+      "step": 12510
+    },
+    {
+      "epoch": 0.3423524425424864,
+      "grad_norm": 2.3125,
+      "learning_rate": 6.594828768062297e-06,
+      "loss": 0.7258,
+      "step": 12520
+    },
+    {
+      "epoch": 0.34262588698541174,
+      "grad_norm": 3.34375,
+      "learning_rate": 6.5920868635354115e-06,
+      "loss": 0.6524,
+      "step": 12530
+    },
+    {
+      "epoch": 0.34289933142833706,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.589344959008528e-06,
+      "loss": 0.6693,
+      "step": 12540
+    },
+    {
+      "epoch": 0.3431727758712624,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.586603054481644e-06,
+      "loss": 0.7093,
+      "step": 12550
+    },
+    {
+      "epoch": 0.34344622031418764,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.583861149954759e-06,
+      "loss": 0.7526,
+      "step": 12560
+    },
+    {
+      "epoch": 0.34371966475711296,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.581119245427875e-06,
+      "loss": 0.6705,
+      "step": 12570
+    },
+    {
+      "epoch": 0.3439931092000383,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.57837734090099e-06,
+      "loss": 0.6943,
+      "step": 12580
+    },
+    {
+      "epoch": 0.3442665536429636,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.575635436374106e-06,
+      "loss": 0.7304,
+      "step": 12590
+    },
+    {
+      "epoch": 0.3445399980858889,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.572893531847222e-06,
+      "loss": 0.7203,
+      "step": 12600
+    },
+    {
+      "epoch": 0.3448134425288142,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.570151627320337e-06,
+      "loss": 0.6275,
+      "step": 12610
+    },
+    {
+      "epoch": 0.3450868869717395,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.567409722793453e-06,
+      "loss": 0.6899,
+      "step": 12620
+    },
+    {
+      "epoch": 0.3453603314146648,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.564667818266568e-06,
+      "loss": 0.7792,
+      "step": 12630
+    },
+    {
+      "epoch": 0.34563377585759014,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.561925913739684e-06,
+      "loss": 0.6737,
+      "step": 12640
+    },
+    {
+      "epoch": 0.34590722030051546,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.5591840092128e-06,
+      "loss": 0.7049,
+      "step": 12650
+    },
+    {
+      "epoch": 0.3461806647434407,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.556442104685915e-06,
+      "loss": 0.763,
+      "step": 12660
+    },
+    {
+      "epoch": 0.34645410918636604,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.553700200159031e-06,
+      "loss": 0.6805,
+      "step": 12670
+    },
+    {
+      "epoch": 0.34672755362929136,
+      "grad_norm": 3.53125,
+      "learning_rate": 6.550958295632147e-06,
+      "loss": 0.7152,
+      "step": 12680
+    },
+    {
+      "epoch": 0.3470009980722167,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.548216391105262e-06,
+      "loss": 0.7106,
+      "step": 12690
+    },
+    {
+      "epoch": 0.347274442515142,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.545474486578378e-06,
+      "loss": 0.6086,
+      "step": 12700
+    },
+    {
+      "epoch": 0.3475478869580673,
+      "grad_norm": 3.0,
+      "learning_rate": 6.542732582051493e-06,
+      "loss": 0.6425,
+      "step": 12710
+    },
+    {
+      "epoch": 0.3478213314009926,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.539990677524609e-06,
+      "loss": 0.7287,
+      "step": 12720
+    },
+    {
+      "epoch": 0.3480947758439179,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.537248772997725e-06,
+      "loss": 0.7655,
+      "step": 12730
+    },
+    {
+      "epoch": 0.3483682202868432,
+      "grad_norm": 3.28125,
+      "learning_rate": 6.53450686847084e-06,
+      "loss": 0.7742,
+      "step": 12740
+    },
+    {
+      "epoch": 0.34864166472976854,
+      "grad_norm": 2.625,
+      "learning_rate": 6.531764963943956e-06,
+      "loss": 0.678,
+      "step": 12750
+    },
+    {
+      "epoch": 0.34891510917269386,
+      "grad_norm": 2.75,
+      "learning_rate": 6.529023059417071e-06,
+      "loss": 0.7807,
+      "step": 12760
+    },
+    {
+      "epoch": 0.3491885536156191,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.526281154890187e-06,
+      "loss": 0.6971,
+      "step": 12770
+    },
+    {
+      "epoch": 0.34946199805854444,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.523539250363304e-06,
+      "loss": 0.6924,
+      "step": 12780
+    },
+    {
+      "epoch": 0.34973544250146976,
+      "grad_norm": 2.875,
+      "learning_rate": 6.520797345836418e-06,
+      "loss": 0.7426,
+      "step": 12790
+    },
+    {
+      "epoch": 0.3500088869443951,
+      "grad_norm": 2.625,
+      "learning_rate": 6.518055441309535e-06,
+      "loss": 0.754,
+      "step": 12800
+    },
+    {
+      "epoch": 0.3502823313873204,
+      "grad_norm": 2.75,
+      "learning_rate": 6.515313536782649e-06,
+      "loss": 0.7247,
+      "step": 12810
+    },
+    {
+      "epoch": 0.35055577583024566,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.5125716322557655e-06,
+      "loss": 0.6583,
+      "step": 12820
+    },
+    {
+      "epoch": 0.350829220273171,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.509829727728881e-06,
+      "loss": 0.6691,
+      "step": 12830
+    },
+    {
+      "epoch": 0.3511026647160963,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.5070878232019965e-06,
+      "loss": 0.6095,
+      "step": 12840
+    },
+    {
+      "epoch": 0.3513761091590216,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.504345918675112e-06,
+      "loss": 0.7547,
+      "step": 12850
+    },
+    {
+      "epoch": 0.35164955360194694,
+      "grad_norm": 3.0,
+      "learning_rate": 6.501604014148228e-06,
+      "loss": 0.7327,
+      "step": 12860
+    },
+    {
+      "epoch": 0.35192299804487226,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.498862109621343e-06,
+      "loss": 0.6493,
+      "step": 12870
+    },
+    {
+      "epoch": 0.3521964424877975,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.496120205094459e-06,
+      "loss": 0.6474,
+      "step": 12880
+    },
+    {
+      "epoch": 0.35246988693072284,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.493378300567575e-06,
+      "loss": 0.66,
+      "step": 12890
+    },
+    {
+      "epoch": 0.35274333137364816,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.49063639604069e-06,
+      "loss": 0.6339,
+      "step": 12900
+    },
+    {
+      "epoch": 0.3530167758165735,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.4878944915138066e-06,
+      "loss": 0.6468,
+      "step": 12910
+    },
+    {
+      "epoch": 0.3532902202594988,
+      "grad_norm": 2.75,
+      "learning_rate": 6.485152586986921e-06,
+      "loss": 0.6921,
+      "step": 12920
+    },
+    {
+      "epoch": 0.35356366470242406,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.4824106824600375e-06,
+      "loss": 0.7402,
+      "step": 12930
+    },
+    {
+      "epoch": 0.3538371091453494,
+      "grad_norm": 3.015625,
+      "learning_rate": 6.479668777933152e-06,
+      "loss": 0.7112,
+      "step": 12940
+    },
+    {
+      "epoch": 0.3541105535882747,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.4769268734062685e-06,
+      "loss": 0.6417,
+      "step": 12950
+    },
+    {
+      "epoch": 0.3543839980312,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.474184968879385e-06,
+      "loss": 0.6665,
+      "step": 12960
+    },
+    {
+      "epoch": 0.35465744247412534,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.4714430643524995e-06,
+      "loss": 0.6761,
+      "step": 12970
+    },
+    {
+      "epoch": 0.3549308869170506,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.468701159825616e-06,
+      "loss": 0.8327,
+      "step": 12980
+    },
+    {
+      "epoch": 0.3552043313599759,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.4659592552987304e-06,
+      "loss": 0.695,
+      "step": 12990
+    },
+    {
+      "epoch": 0.35547777580290124,
+      "grad_norm": 2.375,
+      "learning_rate": 6.463217350771847e-06,
+      "loss": 0.6741,
+      "step": 13000
+    },
+    {
+      "epoch": 0.35575122024582656,
+      "grad_norm": 3.375,
+      "learning_rate": 6.460475446244962e-06,
+      "loss": 0.7361,
+      "step": 13010
+    },
+    {
+      "epoch": 0.3560246646887519,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.457733541718078e-06,
+      "loss": 0.7786,
+      "step": 13020
+    },
+    {
+      "epoch": 0.3562981091316772,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.454991637191193e-06,
+      "loss": 0.7425,
+      "step": 13030
+    },
+    {
+      "epoch": 0.35657155357460246,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.4522497326643095e-06,
+      "loss": 0.6,
+      "step": 13040
+    },
+    {
+      "epoch": 0.3568449980175278,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.449507828137425e-06,
+      "loss": 0.643,
+      "step": 13050
+    },
+    {
+      "epoch": 0.3571184424604531,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.4467659236105405e-06,
+      "loss": 0.6522,
+      "step": 13060
+    },
+    {
+      "epoch": 0.3573918869033784,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.444024019083656e-06,
+      "loss": 0.6744,
+      "step": 13070
+    },
+    {
+      "epoch": 0.35766533134630374,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.4412821145567715e-06,
+      "loss": 0.6567,
+      "step": 13080
+    },
+    {
+      "epoch": 0.357938775789229,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.438540210029888e-06,
+      "loss": 0.6331,
+      "step": 13090
+    },
+    {
+      "epoch": 0.3582122202321543,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.4357983055030024e-06,
+      "loss": 0.7281,
+      "step": 13100
+    },
+    {
+      "epoch": 0.35848566467507964,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.433056400976119e-06,
+      "loss": 0.6825,
+      "step": 13110
+    },
+    {
+      "epoch": 0.35875910911800496,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.430314496449233e-06,
+      "loss": 0.6924,
+      "step": 13120
+    },
+    {
+      "epoch": 0.3590325535609303,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.42757259192235e-06,
+      "loss": 0.6499,
+      "step": 13130
+    },
+    {
+      "epoch": 0.35930599800385554,
+      "grad_norm": 2.625,
+      "learning_rate": 6.424830687395466e-06,
+      "loss": 0.6913,
+      "step": 13140
+    },
+    {
+      "epoch": 0.35957944244678086,
+      "grad_norm": 2.1875,
+      "learning_rate": 6.422088782868581e-06,
+      "loss": 0.708,
+      "step": 13150
+    },
+    {
+      "epoch": 0.3598528868897062,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.419346878341697e-06,
+      "loss": 0.7335,
+      "step": 13160
+    },
+    {
+      "epoch": 0.3601263313326315,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.4166049738148125e-06,
+      "loss": 0.7103,
+      "step": 13170
+    },
+    {
+      "epoch": 0.3603997757755568,
+      "grad_norm": 2.09375,
+      "learning_rate": 6.413863069287928e-06,
+      "loss": 0.6485,
+      "step": 13180
+    },
+    {
+      "epoch": 0.3606732202184821,
+      "grad_norm": 3.0,
+      "learning_rate": 6.4111211647610435e-06,
+      "loss": 0.6755,
+      "step": 13190
+    },
+    {
+      "epoch": 0.3609466646614074,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.408379260234159e-06,
+      "loss": 0.6169,
+      "step": 13200
+    },
+    {
+      "epoch": 0.3612201091043327,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.405637355707275e-06,
+      "loss": 0.7475,
+      "step": 13210
+    },
+    {
+      "epoch": 0.36149355354725804,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.402895451180391e-06,
+      "loss": 0.7178,
+      "step": 13220
+    },
+    {
+      "epoch": 0.36176699799018336,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.400153546653506e-06,
+      "loss": 0.7109,
+      "step": 13230
+    },
+    {
+      "epoch": 0.3620404424331087,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.397411642126622e-06,
+      "loss": 0.7612,
+      "step": 13240
+    },
+    {
+      "epoch": 0.36231388687603394,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.394669737599737e-06,
+      "loss": 0.6037,
+      "step": 13250
+    },
+    {
+      "epoch": 0.36258733131895926,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.391927833072853e-06,
+      "loss": 0.7557,
+      "step": 13260
+    },
+    {
+      "epoch": 0.3628607757618846,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.389185928545969e-06,
+      "loss": 0.679,
+      "step": 13270
+    },
+    {
+      "epoch": 0.3631342202048099,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.386444024019084e-06,
+      "loss": 0.718,
+      "step": 13280
+    },
+    {
+      "epoch": 0.3634076646477352,
+      "grad_norm": 2.828125,
+      "learning_rate": 6.3837021194922e-06,
+      "loss": 0.6763,
+      "step": 13290
+    },
+    {
+      "epoch": 0.3636811090906605,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.380960214965315e-06,
+      "loss": 0.6543,
+      "step": 13300
+    },
+    {
+      "epoch": 0.3639545535335858,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.378218310438431e-06,
+      "loss": 0.6607,
+      "step": 13310
+    },
+    {
+      "epoch": 0.3642279979765111,
+      "grad_norm": 2.46875,
+      "learning_rate": 6.375476405911547e-06,
+      "loss": 0.7173,
+      "step": 13320
+    },
+    {
+      "epoch": 0.36450144241943644,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.372734501384662e-06,
+      "loss": 0.6754,
+      "step": 13330
+    },
+    {
+      "epoch": 0.36477488686236176,
+      "grad_norm": 2.171875,
+      "learning_rate": 6.369992596857778e-06,
+      "loss": 0.6021,
+      "step": 13340
+    },
+    {
+      "epoch": 0.365048331305287,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.367250692330894e-06,
+      "loss": 0.7117,
+      "step": 13350
+    },
+    {
+      "epoch": 0.36532177574821234,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.364508787804009e-06,
+      "loss": 0.6762,
+      "step": 13360
+    },
+    {
+      "epoch": 0.36559522019113766,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.3617668832771255e-06,
+      "loss": 0.788,
+      "step": 13370
+    },
+    {
+      "epoch": 0.365868664634063,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.35902497875024e-06,
+      "loss": 0.7206,
+      "step": 13380
+    },
+    {
+      "epoch": 0.3661421090769883,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.3562830742233565e-06,
+      "loss": 0.6886,
+      "step": 13390
+    },
+    {
+      "epoch": 0.3664155535199136,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.353541169696472e-06,
+      "loss": 0.664,
+      "step": 13400
+    },
+    {
+      "epoch": 0.3666889979628389,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.3507992651695874e-06,
+      "loss": 0.7328,
+      "step": 13410
+    },
+    {
+      "epoch": 0.3669624424057642,
+      "grad_norm": 2.328125,
+      "learning_rate": 6.348057360642703e-06,
+      "loss": 0.6768,
+      "step": 13420
+    },
+    {
+      "epoch": 0.3672358868486895,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.345315456115818e-06,
+      "loss": 0.71,
+      "step": 13430
+    },
+    {
+      "epoch": 0.36750933129161484,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.342573551588934e-06,
+      "loss": 0.6394,
+      "step": 13440
+    },
+    {
+      "epoch": 0.36778277573454016,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.33983164706205e-06,
+      "loss": 0.684,
+      "step": 13450
+    },
+    {
+      "epoch": 0.3680562201774654,
+      "grad_norm": 2.03125,
+      "learning_rate": 6.337089742535165e-06,
+      "loss": 0.6506,
+      "step": 13460
+    },
+    {
+      "epoch": 0.36832966462039074,
+      "grad_norm": 2.5,
+      "learning_rate": 6.334347838008281e-06,
+      "loss": 0.7561,
+      "step": 13470
+    },
+    {
+      "epoch": 0.36860310906331606,
+      "grad_norm": 2.75,
+      "learning_rate": 6.331605933481396e-06,
+      "loss": 0.6451,
+      "step": 13480
+    },
+    {
+      "epoch": 0.3688765535062414,
+      "grad_norm": 2.875,
+      "learning_rate": 6.328864028954512e-06,
+      "loss": 0.7024,
+      "step": 13490
+    },
+    {
+      "epoch": 0.3691499979491667,
+      "grad_norm": 3.0,
+      "learning_rate": 6.3261221244276285e-06,
+      "loss": 0.7214,
+      "step": 13500
+    },
+    {
+      "epoch": 0.36942344239209196,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.323380219900743e-06,
+      "loss": 0.6148,
+      "step": 13510
+    },
+    {
+      "epoch": 0.3696968868350173,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.3206383153738594e-06,
+      "loss": 0.7339,
+      "step": 13520
+    },
+    {
+      "epoch": 0.3699703312779426,
+      "grad_norm": 3.0,
+      "learning_rate": 6.317896410846976e-06,
+      "loss": 0.7109,
+      "step": 13530
+    },
+    {
+      "epoch": 0.3702437757208679,
+      "grad_norm": 2.984375,
+      "learning_rate": 6.31515450632009e-06,
+      "loss": 0.7465,
+      "step": 13540
+    },
+    {
+      "epoch": 0.37051722016379324,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.312412601793207e-06,
+      "loss": 0.7331,
+      "step": 13550
+    },
+    {
+      "epoch": 0.37079066460671856,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.309670697266321e-06,
+      "loss": 0.6578,
+      "step": 13560
+    },
+    {
+      "epoch": 0.3710641090496438,
+      "grad_norm": 2.875,
+      "learning_rate": 6.306928792739438e-06,
+      "loss": 0.6959,
+      "step": 13570
+    },
+    {
+      "epoch": 0.37133755349256914,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.304186888212553e-06,
+      "loss": 0.7239,
+      "step": 13580
+    },
+    {
+      "epoch": 0.37161099793549446,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.301444983685669e-06,
+      "loss": 0.7013,
+      "step": 13590
+    },
+    {
+      "epoch": 0.3718844423784198,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.298703079158784e-06,
+      "loss": 0.7123,
+      "step": 13600
+    },
+    {
+      "epoch": 0.3721578868213451,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.2959611746319e-06,
+      "loss": 0.6951,
+      "step": 13610
+    },
+    {
+      "epoch": 0.37243133126427036,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.293219270105015e-06,
+      "loss": 0.7117,
+      "step": 13620
+    },
+    {
+      "epoch": 0.3727047757071957,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.2904773655781314e-06,
+      "loss": 0.7431,
+      "step": 13630
+    },
+    {
+      "epoch": 0.372978220150121,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.287735461051246e-06,
+      "loss": 0.8444,
+      "step": 13640
+    },
+    {
+      "epoch": 0.3732516645930463,
+      "grad_norm": 2.578125,
+      "learning_rate": 6.284993556524362e-06,
+      "loss": 0.7474,
+      "step": 13650
+    },
+    {
+      "epoch": 0.37352510903597164,
+      "grad_norm": 2.5,
+      "learning_rate": 6.282251651997477e-06,
+      "loss": 0.7127,
+      "step": 13660
+    },
+    {
+      "epoch": 0.3737985534788969,
+      "grad_norm": 3.109375,
+      "learning_rate": 6.279509747470593e-06,
+      "loss": 0.7059,
+      "step": 13670
+    },
+    {
+      "epoch": 0.3740719979218222,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.27676784294371e-06,
+      "loss": 0.6749,
+      "step": 13680
+    },
+    {
+      "epoch": 0.37434544236474754,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.274025938416824e-06,
+      "loss": 0.6664,
+      "step": 13690
+    },
+    {
+      "epoch": 0.37461888680767286,
+      "grad_norm": 2.875,
+      "learning_rate": 6.271284033889941e-06,
+      "loss": 0.7008,
+      "step": 13700
+    },
+    {
+      "epoch": 0.3748923312505982,
+      "grad_norm": 3.40625,
+      "learning_rate": 6.268542129363057e-06,
+      "loss": 0.6752,
+      "step": 13710
+    },
+    {
+      "epoch": 0.3751657756935235,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.265800224836172e-06,
+      "loss": 0.7562,
+      "step": 13720
+    },
+    {
+      "epoch": 0.37543922013644876,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.263058320309288e-06,
+      "loss": 0.6944,
+      "step": 13730
+    },
+    {
+      "epoch": 0.3757126645793741,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.260316415782403e-06,
+      "loss": 0.7294,
+      "step": 13740
+    },
+    {
+      "epoch": 0.3759861090222994,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.257574511255519e-06,
+      "loss": 0.7171,
+      "step": 13750
+    },
+    {
+      "epoch": 0.3762595534652247,
+      "grad_norm": 2.765625,
+      "learning_rate": 6.254832606728634e-06,
+      "loss": 0.6723,
+      "step": 13760
+    },
+    {
+      "epoch": 0.37653299790815004,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.25209070220175e-06,
+      "loss": 0.6431,
+      "step": 13770
+    },
+    {
+      "epoch": 0.3768064423510753,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.249348797674865e-06,
+      "loss": 0.6719,
+      "step": 13780
+    },
+    {
+      "epoch": 0.3770798867940006,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.246606893147981e-06,
+      "loss": 0.6918,
+      "step": 13790
+    },
+    {
+      "epoch": 0.37735333123692594,
+      "grad_norm": 2.625,
+      "learning_rate": 6.243864988621096e-06,
+      "loss": 0.6617,
+      "step": 13800
+    },
+    {
+      "epoch": 0.37762677567985126,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.241123084094213e-06,
+      "loss": 0.7124,
+      "step": 13810
+    },
+    {
+      "epoch": 0.3779002201227766,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.238381179567327e-06,
+      "loss": 0.5449,
+      "step": 13820
+    },
+    {
+      "epoch": 0.37817366456570184,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.235639275040444e-06,
+      "loss": 0.7055,
+      "step": 13830
+    },
+    {
+      "epoch": 0.37844710900862716,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.232897370513559e-06,
+      "loss": 0.6791,
+      "step": 13840
+    },
+    {
+      "epoch": 0.3787205534515525,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.230155465986675e-06,
+      "loss": 0.7459,
+      "step": 13850
+    },
+    {
+      "epoch": 0.3789939978944778,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.227413561459791e-06,
+      "loss": 0.7265,
+      "step": 13860
+    },
+    {
+      "epoch": 0.3792674423374031,
+      "grad_norm": 2.59375,
+      "learning_rate": 6.2246716569329055e-06,
+      "loss": 0.6804,
+      "step": 13870
+    },
+    {
+      "epoch": 0.3795408867803284,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.221929752406022e-06,
+      "loss": 0.6714,
+      "step": 13880
+    },
+    {
+      "epoch": 0.3798143312232537,
+      "grad_norm": 3.328125,
+      "learning_rate": 6.219187847879138e-06,
+      "loss": 0.7088,
+      "step": 13890
+    },
+    {
+      "epoch": 0.380087775666179,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.216445943352253e-06,
+      "loss": 0.6568,
+      "step": 13900
+    },
+    {
+      "epoch": 0.38036122010910434,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.213704038825369e-06,
+      "loss": 0.7247,
+      "step": 13910
+    },
+    {
+      "epoch": 0.38063466455202966,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.210962134298484e-06,
+      "loss": 0.6368,
+      "step": 13920
+    },
+    {
+      "epoch": 0.380908108994955,
+      "grad_norm": 2.421875,
+      "learning_rate": 6.2082202297716e-06,
+      "loss": 0.6921,
+      "step": 13930
+    },
+    {
+      "epoch": 0.38118155343788024,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.205478325244716e-06,
+      "loss": 0.6863,
+      "step": 13940
+    },
+    {
+      "epoch": 0.38145499788080556,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.202736420717831e-06,
+      "loss": 0.6825,
+      "step": 13950
+    },
+    {
+      "epoch": 0.3817284423237309,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.199994516190947e-06,
+      "loss": 0.7176,
+      "step": 13960
+    },
+    {
+      "epoch": 0.3820018867666562,
+      "grad_norm": 3.09375,
+      "learning_rate": 6.197252611664062e-06,
+      "loss": 0.6778,
+      "step": 13970
+    },
+    {
+      "epoch": 0.3822753312095815,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.1945107071371775e-06,
+      "loss": 0.7316,
+      "step": 13980
+    },
+    {
+      "epoch": 0.3825487756525068,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.191768802610294e-06,
+      "loss": 0.7122,
+      "step": 13990
+    },
+    {
+      "epoch": 0.3828222200954321,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.189026898083409e-06,
+      "loss": 0.6813,
+      "step": 14000
+    },
+    {
+      "epoch": 0.3830956645383574,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.186284993556525e-06,
+      "loss": 0.7347,
+      "step": 14010
+    },
+    {
+      "epoch": 0.38336910898128274,
+      "grad_norm": 2.921875,
+      "learning_rate": 6.18354308902964e-06,
+      "loss": 0.7135,
+      "step": 14020
+    },
+    {
+      "epoch": 0.38364255342420805,
+      "grad_norm": 3.046875,
+      "learning_rate": 6.180801184502756e-06,
+      "loss": 0.7529,
+      "step": 14030
+    },
+    {
+      "epoch": 0.3839159978671333,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.178059279975872e-06,
+      "loss": 0.7285,
+      "step": 14040
+    },
+    {
+      "epoch": 0.38418944231005864,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.175317375448987e-06,
+      "loss": 0.7314,
+      "step": 14050
+    },
+    {
+      "epoch": 0.38446288675298396,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.172575470922103e-06,
+      "loss": 0.7224,
+      "step": 14060
+    },
+    {
+      "epoch": 0.3847363311959093,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.169833566395219e-06,
+      "loss": 0.6825,
+      "step": 14070
+    },
+    {
+      "epoch": 0.3850097756388346,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.167091661868334e-06,
+      "loss": 0.7531,
+      "step": 14080
+    },
+    {
+      "epoch": 0.3852832200817599,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.16434975734145e-06,
+      "loss": 0.6188,
+      "step": 14090
+    },
+    {
+      "epoch": 0.3855566645246852,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.161607852814565e-06,
+      "loss": 0.7209,
+      "step": 14100
+    },
+    {
+      "epoch": 0.3858301089676105,
+      "grad_norm": 3.0625,
+      "learning_rate": 6.158865948287681e-06,
+      "loss": 0.7137,
+      "step": 14110
+    },
+    {
+      "epoch": 0.3861035534105358,
+      "grad_norm": 3.265625,
+      "learning_rate": 6.156124043760797e-06,
+      "loss": 0.7105,
+      "step": 14120
+    },
+    {
+      "epoch": 0.38637699785346113,
+      "grad_norm": 2.25,
+      "learning_rate": 6.153382139233912e-06,
+      "loss": 0.6481,
+      "step": 14130
+    },
+    {
+      "epoch": 0.38665044229638645,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.150640234707028e-06,
+      "loss": 0.647,
+      "step": 14140
+    },
+    {
+      "epoch": 0.3869238867393117,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.147898330180143e-06,
+      "loss": 0.7331,
+      "step": 14150
+    },
+    {
+      "epoch": 0.38719733118223704,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.14515642565326e-06,
+      "loss": 0.6945,
+      "step": 14160
+    },
+    {
+      "epoch": 0.38747077562516236,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.142414521126375e-06,
+      "loss": 0.7456,
+      "step": 14170
+    },
+    {
+      "epoch": 0.3877442200680877,
+      "grad_norm": 2.515625,
+      "learning_rate": 6.1396726165994906e-06,
+      "loss": 0.6746,
+      "step": 14180
+    },
+    {
+      "epoch": 0.388017664511013,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.136930712072606e-06,
+      "loss": 0.6863,
+      "step": 14190
+    },
+    {
+      "epoch": 0.38829110895393826,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.1341888075457215e-06,
+      "loss": 0.779,
+      "step": 14200
+    },
+    {
+      "epoch": 0.3885645533968636,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.131446903018837e-06,
+      "loss": 0.6763,
+      "step": 14210
+    },
+    {
+      "epoch": 0.3888379978397889,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.128704998491953e-06,
+      "loss": 0.6615,
+      "step": 14220
+    },
+    {
+      "epoch": 0.3891114422827142,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.125963093965068e-06,
+      "loss": 0.6847,
+      "step": 14230
+    },
+    {
+      "epoch": 0.38938488672563953,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.123221189438184e-06,
+      "loss": 0.6729,
+      "step": 14240
+    },
+    {
+      "epoch": 0.38965833116856485,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.120479284911301e-06,
+      "loss": 0.6582,
+      "step": 14250
+    },
+    {
+      "epoch": 0.3899317756114901,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.117737380384415e-06,
+      "loss": 0.7595,
+      "step": 14260
+    },
+    {
+      "epoch": 0.39020522005441544,
+      "grad_norm": 2.625,
+      "learning_rate": 6.114995475857532e-06,
+      "loss": 0.6754,
+      "step": 14270
+    },
+    {
+      "epoch": 0.39047866449734076,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.112253571330646e-06,
+      "loss": 0.7241,
+      "step": 14280
+    },
+    {
+      "epoch": 0.3907521089402661,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.1095116668037626e-06,
+      "loss": 0.6959,
+      "step": 14290
+    },
+    {
+      "epoch": 0.3910255533831914,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.106769762276878e-06,
+      "loss": 0.7203,
+      "step": 14300
+    },
+    {
+      "epoch": 0.39129899782611666,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.1040278577499935e-06,
+      "loss": 0.725,
+      "step": 14310
+    },
+    {
+      "epoch": 0.391572442269042,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.10128595322311e-06,
+      "loss": 0.7359,
+      "step": 14320
+    },
+    {
+      "epoch": 0.3918458867119673,
+      "grad_norm": 2.265625,
+      "learning_rate": 6.0985440486962245e-06,
+      "loss": 0.6918,
+      "step": 14330
+    },
+    {
+      "epoch": 0.3921193311548926,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.095802144169341e-06,
+      "loss": 0.6649,
+      "step": 14340
+    },
+    {
+      "epoch": 0.39239277559781793,
+      "grad_norm": 3.640625,
+      "learning_rate": 6.093060239642456e-06,
+      "loss": 0.7633,
+      "step": 14350
+    },
+    {
+      "epoch": 0.3926662200407432,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.090318335115572e-06,
+      "loss": 0.6541,
+      "step": 14360
+    },
+    {
+      "epoch": 0.3929396644836685,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.087576430588687e-06,
+      "loss": 0.6348,
+      "step": 14370
+    },
+    {
+      "epoch": 0.39321310892659384,
+      "grad_norm": 2.625,
+      "learning_rate": 6.084834526061803e-06,
+      "loss": 0.6468,
+      "step": 14380
+    },
+    {
+      "epoch": 0.39348655336951915,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.082092621534918e-06,
+      "loss": 0.7333,
+      "step": 14390
+    },
+    {
+      "epoch": 0.3937599978124445,
+      "grad_norm": 2.953125,
+      "learning_rate": 6.0793507170080346e-06,
+      "loss": 0.7096,
+      "step": 14400
+    },
+    {
+      "epoch": 0.3940334422553698,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.076608812481149e-06,
+      "loss": 0.7081,
+      "step": 14410
+    },
+    {
+      "epoch": 0.39430688669829506,
+      "grad_norm": 3.15625,
+      "learning_rate": 6.0738669079542655e-06,
+      "loss": 0.7049,
+      "step": 14420
+    },
+    {
+      "epoch": 0.3945803311412204,
+      "grad_norm": 2.9375,
+      "learning_rate": 6.071125003427382e-06,
+      "loss": 0.6019,
+      "step": 14430
+    },
+    {
+      "epoch": 0.3948537755841457,
+      "grad_norm": 2.875,
+      "learning_rate": 6.0683830989004965e-06,
+      "loss": 0.6275,
+      "step": 14440
+    },
+    {
+      "epoch": 0.395127220027071,
+      "grad_norm": 3.171875,
+      "learning_rate": 6.065641194373613e-06,
+      "loss": 0.7434,
+      "step": 14450
+    },
+    {
+      "epoch": 0.39540066446999633,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.0628992898467275e-06,
+      "loss": 0.6777,
+      "step": 14460
+    },
+    {
+      "epoch": 0.3956741089129216,
+      "grad_norm": 2.75,
+      "learning_rate": 6.060157385319844e-06,
+      "loss": 0.6286,
+      "step": 14470
+    },
+    {
+      "epoch": 0.3959475533558469,
+      "grad_norm": 2.40625,
+      "learning_rate": 6.05741548079296e-06,
+      "loss": 0.6864,
+      "step": 14480
+    },
+    {
+      "epoch": 0.39622099779877223,
+      "grad_norm": 2.75,
+      "learning_rate": 6.054673576266075e-06,
+      "loss": 0.7356,
+      "step": 14490
+    },
+    {
+      "epoch": 0.39649444224169755,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.051931671739191e-06,
+      "loss": 0.7372,
+      "step": 14500
+    },
+    {
+      "epoch": 0.3967678866846229,
+      "grad_norm": 3.03125,
+      "learning_rate": 6.049189767212306e-06,
+      "loss": 0.6744,
+      "step": 14510
+    },
+    {
+      "epoch": 0.39704133112754814,
+      "grad_norm": 3.140625,
+      "learning_rate": 6.046447862685422e-06,
+      "loss": 0.6878,
+      "step": 14520
+    },
+    {
+      "epoch": 0.39731477557047346,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.0437059581585375e-06,
+      "loss": 0.6424,
+      "step": 14530
+    },
+    {
+      "epoch": 0.3975882200133988,
+      "grad_norm": 2.4375,
+      "learning_rate": 6.040964053631653e-06,
+      "loss": 0.6246,
+      "step": 14540
+    },
+    {
+      "epoch": 0.3978616644563241,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.0382221491047685e-06,
+      "loss": 0.6473,
+      "step": 14550
+    },
+    {
+      "epoch": 0.3981351088992494,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.035480244577885e-06,
+      "loss": 0.7126,
+      "step": 14560
+    },
+    {
+      "epoch": 0.3984085533421747,
+      "grad_norm": 2.34375,
+      "learning_rate": 6.0327383400509995e-06,
+      "loss": 0.6815,
+      "step": 14570
+    },
+    {
+      "epoch": 0.3986819977851,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.029996435524116e-06,
+      "loss": 0.727,
+      "step": 14580
+    },
+    {
+      "epoch": 0.3989554422280253,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.02725453099723e-06,
+      "loss": 0.6589,
+      "step": 14590
+    },
+    {
+      "epoch": 0.39922888667095063,
+      "grad_norm": 3.390625,
+      "learning_rate": 6.024512626470347e-06,
+      "loss": 0.7331,
+      "step": 14600
+    },
+    {
+      "epoch": 0.39950233111387595,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.021770721943463e-06,
+      "loss": 0.639,
+      "step": 14610
+    },
+    {
+      "epoch": 0.39977577555680127,
+      "grad_norm": 2.140625,
+      "learning_rate": 6.019028817416578e-06,
+      "loss": 0.7135,
+      "step": 14620
+    },
+    {
+      "epoch": 0.40004921999972654,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.016286912889694e-06,
+      "loss": 0.6038,
+      "step": 14630
+    },
+    {
+      "epoch": 0.40032266444265185,
+      "grad_norm": 2.734375,
+      "learning_rate": 6.013545008362809e-06,
+      "loss": 0.6952,
+      "step": 14640
+    },
+    {
+      "epoch": 0.4005961088855772,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.010803103835925e-06,
+      "loss": 0.6579,
+      "step": 14650
+    },
+    {
+      "epoch": 0.4008695533285025,
+      "grad_norm": 2.6875,
+      "learning_rate": 6.008061199309041e-06,
+      "loss": 0.7139,
+      "step": 14660
+    },
+    {
+      "epoch": 0.4011429977714278,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.005319294782156e-06,
+      "loss": 0.636,
+      "step": 14670
+    },
+    {
+      "epoch": 0.4014164422143531,
+      "grad_norm": 2.453125,
+      "learning_rate": 6.002577390255272e-06,
+      "loss": 0.595,
+      "step": 14680
+    },
+    {
+      "epoch": 0.4016898866572784,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.999835485728387e-06,
+      "loss": 0.7053,
+      "step": 14690
+    },
+    {
+      "epoch": 0.4019633311002037,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.997093581201503e-06,
+      "loss": 0.7116,
+      "step": 14700
+    },
+    {
+      "epoch": 0.40223677554312903,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.994351676674619e-06,
+      "loss": 0.6998,
+      "step": 14710
+    },
+    {
+      "epoch": 0.40251021998605435,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.991609772147734e-06,
+      "loss": 0.6792,
+      "step": 14720
+    },
+    {
+      "epoch": 0.4027836644289796,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.98886786762085e-06,
+      "loss": 0.7268,
+      "step": 14730
+    },
+    {
+      "epoch": 0.40305710887190493,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.986125963093966e-06,
+      "loss": 0.7736,
+      "step": 14740
+    },
+    {
+      "epoch": 0.40333055331483025,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.983384058567081e-06,
+      "loss": 0.6195,
+      "step": 14750
+    },
+    {
+      "epoch": 0.4036039977577556,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.980642154040197e-06,
+      "loss": 0.6881,
+      "step": 14760
+    },
+    {
+      "epoch": 0.4038774422006809,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.977900249513312e-06,
+      "loss": 0.7166,
+      "step": 14770
+    },
+    {
+      "epoch": 0.4041508866436062,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.975158344986428e-06,
+      "loss": 0.7411,
+      "step": 14780
+    },
+    {
+      "epoch": 0.4044243310865315,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.972416440459544e-06,
+      "loss": 0.6692,
+      "step": 14790
+    },
+    {
+      "epoch": 0.4046977755294568,
+      "grad_norm": 2.5,
+      "learning_rate": 5.969674535932659e-06,
+      "loss": 0.6725,
+      "step": 14800
+    },
+    {
+      "epoch": 0.4049712199723821,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.966932631405775e-06,
+      "loss": 0.6516,
+      "step": 14810
+    },
+    {
+      "epoch": 0.40524466441530743,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.96419072687889e-06,
+      "loss": 0.6855,
+      "step": 14820
+    },
+    {
+      "epoch": 0.40551810885823275,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.961448822352006e-06,
+      "loss": 0.6546,
+      "step": 14830
+    },
+    {
+      "epoch": 0.405791553301158,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.9587069178251225e-06,
+      "loss": 0.7487,
+      "step": 14840
+    },
+    {
+      "epoch": 0.40606499774408333,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.955965013298237e-06,
+      "loss": 0.7097,
+      "step": 14850
+    },
+    {
+      "epoch": 0.40633844218700865,
+      "grad_norm": 3.5,
+      "learning_rate": 5.9532231087713535e-06,
+      "loss": 0.7235,
+      "step": 14860
+    },
+    {
+      "epoch": 0.40661188662993397,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.950481204244468e-06,
+      "loss": 0.6973,
+      "step": 14870
+    },
+    {
+      "epoch": 0.4068853310728593,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.9477392997175845e-06,
+      "loss": 0.768,
+      "step": 14880
+    },
+    {
+      "epoch": 0.40715877551578455,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.9449973951907e-06,
+      "loss": 0.6414,
+      "step": 14890
+    },
+    {
+      "epoch": 0.4074322199587099,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.9422554906638154e-06,
+      "loss": 0.6941,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4077056644016352,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.939513586136931e-06,
+      "loss": 0.825,
+      "step": 14910
+    },
+    {
+      "epoch": 0.4079791088445605,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.936771681610047e-06,
+      "loss": 0.6626,
+      "step": 14920
+    },
+    {
+      "epoch": 0.40825255328748583,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.934029777083162e-06,
+      "loss": 0.7031,
+      "step": 14930
+    },
+    {
+      "epoch": 0.40852599773041115,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.931287872556278e-06,
+      "loss": 0.7209,
+      "step": 14940
+    },
+    {
+      "epoch": 0.4087994421733364,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.928545968029394e-06,
+      "loss": 0.5933,
+      "step": 14950
+    },
+    {
+      "epoch": 0.40907288661626173,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.925804063502509e-06,
+      "loss": 0.7659,
+      "step": 14960
+    },
+    {
+      "epoch": 0.40934633105918705,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.9230621589756255e-06,
+      "loss": 0.5858,
+      "step": 14970
+    },
+    {
+      "epoch": 0.40961977550211237,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.92032025444874e-06,
+      "loss": 0.7104,
+      "step": 14980
+    },
+    {
+      "epoch": 0.4098932199450377,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.9175783499218565e-06,
+      "loss": 0.7616,
+      "step": 14990
+    },
+    {
+      "epoch": 0.41016666438796295,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.914836445394971e-06,
+      "loss": 0.6702,
+      "step": 15000
+    },
+    {
+      "epoch": 0.4104401088308883,
+      "grad_norm": 2.625,
+      "learning_rate": 5.9120945408680874e-06,
+      "loss": 0.6992,
+      "step": 15010
+    },
+    {
+      "epoch": 0.4107135532738136,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.909352636341204e-06,
+      "loss": 0.7395,
+      "step": 15020
+    },
+    {
+      "epoch": 0.4109869977167389,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.906610731814318e-06,
+      "loss": 0.7298,
+      "step": 15030
+    },
+    {
+      "epoch": 0.41126044215966423,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.903868827287435e-06,
+      "loss": 0.7603,
+      "step": 15040
+    },
+    {
+      "epoch": 0.4115338866025895,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.901126922760549e-06,
+      "loss": 0.7144,
+      "step": 15050
+    },
+    {
+      "epoch": 0.4118073310455148,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.898385018233666e-06,
+      "loss": 0.7735,
+      "step": 15060
+    },
+    {
+      "epoch": 0.41208077548844013,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.895643113706781e-06,
+      "loss": 0.6999,
+      "step": 15070
+    },
+    {
+      "epoch": 0.41235421993136545,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.892901209179897e-06,
+      "loss": 0.726,
+      "step": 15080
+    },
+    {
+      "epoch": 0.41262766437429077,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.890159304653012e-06,
+      "loss": 0.5768,
+      "step": 15090
+    },
+    {
+      "epoch": 0.4129011088172161,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.8874174001261285e-06,
+      "loss": 0.7366,
+      "step": 15100
+    },
+    {
+      "epoch": 0.41317455326014135,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.884675495599244e-06,
+      "loss": 0.6978,
+      "step": 15110
+    },
+    {
+      "epoch": 0.41344799770306667,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.8819335910723594e-06,
+      "loss": 0.6102,
+      "step": 15120
+    },
+    {
+      "epoch": 0.413721442145992,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.879191686545475e-06,
+      "loss": 0.7044,
+      "step": 15130
+    },
+    {
+      "epoch": 0.4139948865889173,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.87644978201859e-06,
+      "loss": 0.6246,
+      "step": 15140
+    },
+    {
+      "epoch": 0.41426833103184263,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.873707877491707e-06,
+      "loss": 0.6282,
+      "step": 15150
+    },
+    {
+      "epoch": 0.4145417754747679,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.870965972964821e-06,
+      "loss": 0.7281,
+      "step": 15160
+    },
+    {
+      "epoch": 0.4148152199176932,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.868224068437938e-06,
+      "loss": 0.6464,
+      "step": 15170
+    },
+    {
+      "epoch": 0.41508866436061853,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.865482163911052e-06,
+      "loss": 0.7392,
+      "step": 15180
+    },
+    {
+      "epoch": 0.41536210880354385,
+      "grad_norm": 3.25,
+      "learning_rate": 5.862740259384169e-06,
+      "loss": 0.724,
+      "step": 15190
+    },
+    {
+      "epoch": 0.41563555324646917,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.859998354857285e-06,
+      "loss": 0.66,
+      "step": 15200
+    },
+    {
+      "epoch": 0.41590899768939443,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.8572564503304e-06,
+      "loss": 0.7156,
+      "step": 15210
+    },
+    {
+      "epoch": 0.41618244213231975,
+      "grad_norm": 3.0,
+      "learning_rate": 5.854514545803516e-06,
+      "loss": 0.6355,
+      "step": 15220
+    },
+    {
+      "epoch": 0.41645588657524507,
+      "grad_norm": 2.21875,
+      "learning_rate": 5.851772641276631e-06,
+      "loss": 0.7075,
+      "step": 15230
+    },
+    {
+      "epoch": 0.4167293310181704,
+      "grad_norm": 2.25,
+      "learning_rate": 5.849030736749747e-06,
+      "loss": 0.6611,
+      "step": 15240
+    },
+    {
+      "epoch": 0.4170027754610957,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.846288832222862e-06,
+      "loss": 0.7198,
+      "step": 15250
+    },
+    {
+      "epoch": 0.417276219904021,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.843546927695978e-06,
+      "loss": 0.7012,
+      "step": 15260
+    },
+    {
+      "epoch": 0.4175496643469463,
+      "grad_norm": 2.625,
+      "learning_rate": 5.840805023169094e-06,
+      "loss": 0.6907,
+      "step": 15270
+    },
+    {
+      "epoch": 0.4178231087898716,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.83806311864221e-06,
+      "loss": 0.6457,
+      "step": 15280
+    },
+    {
+      "epoch": 0.41809655323279693,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.835321214115325e-06,
+      "loss": 0.7494,
+      "step": 15290
+    },
+    {
+      "epoch": 0.41836999767572225,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.832579309588441e-06,
+      "loss": 0.7189,
+      "step": 15300
+    },
+    {
+      "epoch": 0.41864344211864757,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.829837405061556e-06,
+      "loss": 0.6462,
+      "step": 15310
+    },
+    {
+      "epoch": 0.41891688656157283,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.827095500534672e-06,
+      "loss": 0.6437,
+      "step": 15320
+    },
+    {
+      "epoch": 0.41919033100449815,
+      "grad_norm": 3.1875,
+      "learning_rate": 5.824353596007788e-06,
+      "loss": 0.7495,
+      "step": 15330
+    },
+    {
+      "epoch": 0.41946377544742347,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.8216116914809026e-06,
+      "loss": 0.7473,
+      "step": 15340
+    },
+    {
+      "epoch": 0.4197372198903488,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.818869786954019e-06,
+      "loss": 0.6484,
+      "step": 15350
+    },
+    {
+      "epoch": 0.4200106643332741,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.8161278824271335e-06,
+      "loss": 0.7195,
+      "step": 15360
+    },
+    {
+      "epoch": 0.42028410877619937,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.81338597790025e-06,
+      "loss": 0.7102,
+      "step": 15370
+    },
+    {
+      "epoch": 0.4205575532191247,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.810644073373366e-06,
+      "loss": 0.6974,
+      "step": 15380
+    },
+    {
+      "epoch": 0.42083099766205,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.807902168846481e-06,
+      "loss": 0.7082,
+      "step": 15390
+    },
+    {
+      "epoch": 0.42110444210497533,
+      "grad_norm": 2.875,
+      "learning_rate": 5.805160264319597e-06,
+      "loss": 0.744,
+      "step": 15400
+    },
+    {
+      "epoch": 0.42137788654790065,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.802418359792712e-06,
+      "loss": 0.7514,
+      "step": 15410
+    },
+    {
+      "epoch": 0.4216513309908259,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.799676455265828e-06,
+      "loss": 0.6245,
+      "step": 15420
+    },
+    {
+      "epoch": 0.42192477543375123,
+      "grad_norm": 2.875,
+      "learning_rate": 5.7969345507389445e-06,
+      "loss": 0.6849,
+      "step": 15430
+    },
+    {
+      "epoch": 0.42219821987667655,
+      "grad_norm": 2.75,
+      "learning_rate": 5.794192646212059e-06,
+      "loss": 0.6904,
+      "step": 15440
+    },
+    {
+      "epoch": 0.42247166431960187,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.791450741685175e-06,
+      "loss": 0.6454,
+      "step": 15450
+    },
+    {
+      "epoch": 0.4227451087625272,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.788708837158291e-06,
+      "loss": 0.6852,
+      "step": 15460
+    },
+    {
+      "epoch": 0.4230185532054525,
+      "grad_norm": 2.75,
+      "learning_rate": 5.785966932631406e-06,
+      "loss": 0.6395,
+      "step": 15470
+    },
+    {
+      "epoch": 0.42329199764837777,
+      "grad_norm": 3.34375,
+      "learning_rate": 5.783225028104522e-06,
+      "loss": 0.6912,
+      "step": 15480
+    },
+    {
+      "epoch": 0.4235654420913031,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.780483123577637e-06,
+      "loss": 0.6453,
+      "step": 15490
+    },
+    {
+      "epoch": 0.4238388865342284,
+      "grad_norm": 2.5,
+      "learning_rate": 5.777741219050753e-06,
+      "loss": 0.6702,
+      "step": 15500
+    },
+    {
+      "epoch": 0.42411233097715373,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.774999314523869e-06,
+      "loss": 0.7166,
+      "step": 15510
+    },
+    {
+      "epoch": 0.42438577542007905,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.772257409996984e-06,
+      "loss": 0.7389,
+      "step": 15520
+    },
+    {
+      "epoch": 0.4246592198630043,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.7695155054701e-06,
+      "loss": 0.65,
+      "step": 15530
+    },
+    {
+      "epoch": 0.42493266430592963,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.766773600943215e-06,
+      "loss": 0.7133,
+      "step": 15540
+    },
+    {
+      "epoch": 0.42520610874885495,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.764031696416331e-06,
+      "loss": 0.6706,
+      "step": 15550
+    },
+    {
+      "epoch": 0.42547955319178027,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.761289791889447e-06,
+      "loss": 0.7512,
+      "step": 15560
+    },
+    {
+      "epoch": 0.4257529976347056,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.758547887362562e-06,
+      "loss": 0.6969,
+      "step": 15570
+    },
+    {
+      "epoch": 0.42602644207763085,
+      "grad_norm": 3.359375,
+      "learning_rate": 5.755805982835678e-06,
+      "loss": 0.7272,
+      "step": 15580
+    },
+    {
+      "epoch": 0.42629988652055617,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.753064078308793e-06,
+      "loss": 0.703,
+      "step": 15590
+    },
+    {
+      "epoch": 0.4265733309634815,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.750322173781909e-06,
+      "loss": 0.6982,
+      "step": 15600
+    },
+    {
+      "epoch": 0.4268467754064068,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.747580269255026e-06,
+      "loss": 0.7823,
+      "step": 15610
+    },
+    {
+      "epoch": 0.42712021984933213,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.74483836472814e-06,
+      "loss": 0.7053,
+      "step": 15620
+    },
+    {
+      "epoch": 0.42739366429225745,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.742096460201257e-06,
+      "loss": 0.6289,
+      "step": 15630
+    },
+    {
+      "epoch": 0.4276671087351827,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.739354555674372e-06,
+      "loss": 0.7626,
+      "step": 15640
+    },
+    {
+      "epoch": 0.42794055317810803,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.736612651147488e-06,
+      "loss": 0.6943,
+      "step": 15650
+    },
+    {
+      "epoch": 0.42821399762103335,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.733870746620603e-06,
+      "loss": 0.6606,
+      "step": 15660
+    },
+    {
+      "epoch": 0.42848744206395867,
+      "grad_norm": 3.21875,
+      "learning_rate": 5.7311288420937186e-06,
+      "loss": 0.7333,
+      "step": 15670
+    },
+    {
+      "epoch": 0.428760886506884,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.728386937566834e-06,
+      "loss": 0.7562,
+      "step": 15680
+    },
+    {
+      "epoch": 0.42903433094980925,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.72564503303995e-06,
+      "loss": 0.6959,
+      "step": 15690
+    },
+    {
+      "epoch": 0.42930777539273457,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.722903128513065e-06,
+      "loss": 0.64,
+      "step": 15700
+    },
+    {
+      "epoch": 0.4295812198356599,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.720161223986181e-06,
+      "loss": 0.7431,
+      "step": 15710
+    },
+    {
+      "epoch": 0.4298546642785852,
+      "grad_norm": 2.171875,
+      "learning_rate": 5.717419319459297e-06,
+      "loss": 0.659,
+      "step": 15720
+    },
+    {
+      "epoch": 0.4301281087215105,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.714677414932412e-06,
+      "loss": 0.6002,
+      "step": 15730
+    },
+    {
+      "epoch": 0.4304015531644358,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.711935510405529e-06,
+      "loss": 0.6399,
+      "step": 15740
+    },
+    {
+      "epoch": 0.4306749976073611,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.709193605878643e-06,
+      "loss": 0.6559,
+      "step": 15750
+    },
+    {
+      "epoch": 0.43094844205028643,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.70645170135176e-06,
+      "loss": 0.7101,
+      "step": 15760
+    },
+    {
+      "epoch": 0.43122188649321175,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.703709796824876e-06,
+      "loss": 0.7159,
+      "step": 15770
+    },
+    {
+      "epoch": 0.43149533093613707,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.7009678922979906e-06,
+      "loss": 0.6853,
+      "step": 15780
+    },
+    {
+      "epoch": 0.4317687753790624,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.698225987771107e-06,
+      "loss": 0.7091,
+      "step": 15790
+    },
+    {
+      "epoch": 0.43204221982198765,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.6954840832442215e-06,
+      "loss": 0.7035,
+      "step": 15800
+    },
+    {
+      "epoch": 0.43231566426491297,
+      "grad_norm": 2.625,
+      "learning_rate": 5.692742178717338e-06,
+      "loss": 0.6755,
+      "step": 15810
+    },
+    {
+      "epoch": 0.4325891087078383,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.690000274190453e-06,
+      "loss": 0.6707,
+      "step": 15820
+    },
+    {
+      "epoch": 0.4328625531507636,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.687258369663569e-06,
+      "loss": 0.7217,
+      "step": 15830
+    },
+    {
+      "epoch": 0.4331359975936889,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.684516465136684e-06,
+      "loss": 0.695,
+      "step": 15840
+    },
+    {
+      "epoch": 0.4334094420366142,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.6817745606098e-06,
+      "loss": 0.7308,
+      "step": 15850
+    },
+    {
+      "epoch": 0.4336828864795395,
+      "grad_norm": 2.625,
+      "learning_rate": 5.679032656082915e-06,
+      "loss": 0.6607,
+      "step": 15860
+    },
+    {
+      "epoch": 0.43395633092246483,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.676290751556032e-06,
+      "loss": 0.7026,
+      "step": 15870
+    },
+    {
+      "epoch": 0.43422977536539015,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.673548847029147e-06,
+      "loss": 0.6638,
+      "step": 15880
+    },
+    {
+      "epoch": 0.43450321980831547,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.6708069425022626e-06,
+      "loss": 0.634,
+      "step": 15890
+    },
+    {
+      "epoch": 0.43477666425124073,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.668065037975378e-06,
+      "loss": 0.8034,
+      "step": 15900
+    },
+    {
+      "epoch": 0.43505010869416605,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.6653231334484935e-06,
+      "loss": 0.6354,
+      "step": 15910
+    },
+    {
+      "epoch": 0.43532355313709137,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.66258122892161e-06,
+      "loss": 0.6598,
+      "step": 15920
+    },
+    {
+      "epoch": 0.4355969975800167,
+      "grad_norm": 2.75,
+      "learning_rate": 5.6598393243947245e-06,
+      "loss": 0.6071,
+      "step": 15930
+    },
+    {
+      "epoch": 0.435870442022942,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.657097419867841e-06,
+      "loss": 0.6799,
+      "step": 15940
+    },
+    {
+      "epoch": 0.43614388646586727,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.654355515340957e-06,
+      "loss": 0.7165,
+      "step": 15950
+    },
+    {
+      "epoch": 0.4364173309087926,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.651613610814072e-06,
+      "loss": 0.704,
+      "step": 15960
+    },
+    {
+      "epoch": 0.4366907753517179,
+      "grad_norm": 2.875,
+      "learning_rate": 5.648871706287188e-06,
+      "loss": 0.7388,
+      "step": 15970
+    },
+    {
+      "epoch": 0.4369642197946432,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.646129801760303e-06,
+      "loss": 0.6585,
+      "step": 15980
+    },
+    {
+      "epoch": 0.43723766423756855,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.643387897233419e-06,
+      "loss": 0.5972,
+      "step": 15990
+    },
+    {
+      "epoch": 0.43751110868049387,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.6406459927065345e-06,
+      "loss": 0.6451,
+      "step": 16000
+    },
+    {
+      "epoch": 0.43778455312341913,
+      "grad_norm": 2.875,
+      "learning_rate": 5.63790408817965e-06,
+      "loss": 0.6892,
+      "step": 16010
+    },
+    {
+      "epoch": 0.43805799756634445,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.6351621836527655e-06,
+      "loss": 0.7897,
+      "step": 16020
+    },
+    {
+      "epoch": 0.43833144200926977,
+      "grad_norm": 2.75,
+      "learning_rate": 5.632420279125881e-06,
+      "loss": 0.624,
+      "step": 16030
+    },
+    {
+      "epoch": 0.4386048864521951,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.629678374598997e-06,
+      "loss": 0.7036,
+      "step": 16040
+    },
+    {
+      "epoch": 0.4388783308951204,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.626936470072113e-06,
+      "loss": 0.6196,
+      "step": 16050
+    },
+    {
+      "epoch": 0.43915177533804567,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.624194565545228e-06,
+      "loss": 0.6539,
+      "step": 16060
+    },
+    {
+      "epoch": 0.439425219780971,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.621452661018344e-06,
+      "loss": 0.7136,
+      "step": 16070
+    },
+    {
+      "epoch": 0.4396986642238963,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.618710756491459e-06,
+      "loss": 0.7133,
+      "step": 16080
+    },
+    {
+      "epoch": 0.4399721086668216,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.615968851964575e-06,
+      "loss": 0.6938,
+      "step": 16090
+    },
+    {
+      "epoch": 0.44024555310974695,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.613226947437691e-06,
+      "loss": 0.6777,
+      "step": 16100
+    },
+    {
+      "epoch": 0.4405189975526722,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.610485042910806e-06,
+      "loss": 0.6498,
+      "step": 16110
+    },
+    {
+      "epoch": 0.44079244199559753,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.607743138383922e-06,
+      "loss": 0.6162,
+      "step": 16120
+    },
+    {
+      "epoch": 0.44106588643852285,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.605001233857038e-06,
+      "loss": 0.6843,
+      "step": 16130
+    },
+    {
+      "epoch": 0.44133933088144817,
+      "grad_norm": 2.375,
+      "learning_rate": 5.602259329330153e-06,
+      "loss": 0.6719,
+      "step": 16140
+    },
+    {
+      "epoch": 0.4416127753243735,
+      "grad_norm": 2.875,
+      "learning_rate": 5.599517424803269e-06,
+      "loss": 0.6539,
+      "step": 16150
+    },
+    {
+      "epoch": 0.4418862197672988,
+      "grad_norm": 3.125,
+      "learning_rate": 5.596775520276384e-06,
+      "loss": 0.6956,
+      "step": 16160
+    },
+    {
+      "epoch": 0.44215966421022407,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.5940336157495e-06,
+      "loss": 0.6771,
+      "step": 16170
+    },
+    {
+      "epoch": 0.4424331086531494,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.591291711222616e-06,
+      "loss": 0.6583,
+      "step": 16180
+    },
+    {
+      "epoch": 0.4427065530960747,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.588549806695731e-06,
+      "loss": 0.7375,
+      "step": 16190
+    },
+    {
+      "epoch": 0.442979997539,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.5858079021688476e-06,
+      "loss": 0.7138,
+      "step": 16200
+    },
+    {
+      "epoch": 0.44325344198192534,
+      "grad_norm": 2.265625,
+      "learning_rate": 5.583065997641962e-06,
+      "loss": 0.6597,
+      "step": 16210
+    },
+    {
+      "epoch": 0.4435268864248506,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.5803240931150785e-06,
+      "loss": 0.6809,
+      "step": 16220
+    },
+    {
+      "epoch": 0.4438003308677759,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.577582188588194e-06,
+      "loss": 0.694,
+      "step": 16230
+    },
+    {
+      "epoch": 0.44407377531070125,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.5748402840613095e-06,
+      "loss": 0.6712,
+      "step": 16240
+    },
+    {
+      "epoch": 0.44434721975362657,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.572098379534425e-06,
+      "loss": 0.7146,
+      "step": 16250
+    },
+    {
+      "epoch": 0.4446206641965519,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.5693564750075405e-06,
+      "loss": 0.767,
+      "step": 16260
+    },
+    {
+      "epoch": 0.44489410863947715,
+      "grad_norm": 3.125,
+      "learning_rate": 5.566614570480656e-06,
+      "loss": 0.6735,
+      "step": 16270
+    },
+    {
+      "epoch": 0.44516755308240247,
+      "grad_norm": 3.125,
+      "learning_rate": 5.563872665953772e-06,
+      "loss": 0.6784,
+      "step": 16280
+    },
+    {
+      "epoch": 0.4454409975253278,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.561130761426887e-06,
+      "loss": 0.7688,
+      "step": 16290
+    },
+    {
+      "epoch": 0.4457144419682531,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.558388856900003e-06,
+      "loss": 0.7248,
+      "step": 16300
+    },
+    {
+      "epoch": 0.4459878864111784,
+      "grad_norm": 2.546875,
+      "learning_rate": 5.5556469523731196e-06,
+      "loss": 0.7285,
+      "step": 16310
+    },
+    {
+      "epoch": 0.44626133085410374,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.552905047846234e-06,
+      "loss": 0.7176,
+      "step": 16320
+    },
+    {
+      "epoch": 0.446534775297029,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.5501631433193505e-06,
+      "loss": 0.7016,
+      "step": 16330
+    },
+    {
+      "epoch": 0.4468082197399543,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.547421238792465e-06,
+      "loss": 0.6685,
+      "step": 16340
+    },
+    {
+      "epoch": 0.44708166418287965,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.5446793342655815e-06,
+      "loss": 0.7495,
+      "step": 16350
+    },
+    {
+      "epoch": 0.44735510862580496,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.541937429738698e-06,
+      "loss": 0.6678,
+      "step": 16360
+    },
+    {
+      "epoch": 0.4476285530687303,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.5391955252118125e-06,
+      "loss": 0.7367,
+      "step": 16370
+    },
+    {
+      "epoch": 0.44790199751165555,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.536453620684929e-06,
+      "loss": 0.6578,
+      "step": 16380
+    },
+    {
+      "epoch": 0.44817544195458087,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.5337117161580434e-06,
+      "loss": 0.6856,
+      "step": 16390
+    },
+    {
+      "epoch": 0.4484488863975062,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.53096981163116e-06,
+      "loss": 0.7409,
+      "step": 16400
+    },
+    {
+      "epoch": 0.4487223308404315,
+      "grad_norm": 3.125,
+      "learning_rate": 5.528227907104275e-06,
+      "loss": 0.6849,
+      "step": 16410
+    },
+    {
+      "epoch": 0.4489957752833568,
+      "grad_norm": 3.375,
+      "learning_rate": 5.525486002577391e-06,
+      "loss": 0.8204,
+      "step": 16420
+    },
+    {
+      "epoch": 0.4492692197262821,
+      "grad_norm": 2.359375,
+      "learning_rate": 5.522744098050506e-06,
+      "loss": 0.6096,
+      "step": 16430
+    },
+    {
+      "epoch": 0.4495426641692074,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.520002193523622e-06,
+      "loss": 0.7223,
+      "step": 16440
+    },
+    {
+      "epoch": 0.4498161086121327,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.517260288996737e-06,
+      "loss": 0.7118,
+      "step": 16450
+    },
+    {
+      "epoch": 0.45008955305505804,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.5145183844698535e-06,
+      "loss": 0.685,
+      "step": 16460
+    },
+    {
+      "epoch": 0.45036299749798336,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.511776479942968e-06,
+      "loss": 0.757,
+      "step": 16470
+    },
+    {
+      "epoch": 0.4506364419409087,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.5090345754160845e-06,
+      "loss": 0.6923,
+      "step": 16480
+    },
+    {
+      "epoch": 0.45090988638383395,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.506292670889201e-06,
+      "loss": 0.7471,
+      "step": 16490
+    },
+    {
+      "epoch": 0.45118333082675927,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.503550766362315e-06,
+      "loss": 0.6723,
+      "step": 16500
+    },
+    {
+      "epoch": 0.4514567752696846,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.500808861835432e-06,
+      "loss": 0.6824,
+      "step": 16510
+    },
+    {
+      "epoch": 0.4517302197126099,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.498066957308546e-06,
+      "loss": 0.6889,
+      "step": 16520
+    },
+    {
+      "epoch": 0.4520036641555352,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.495325052781663e-06,
+      "loss": 0.6781,
+      "step": 16530
+    },
+    {
+      "epoch": 0.4522771085984605,
+      "grad_norm": 2.625,
+      "learning_rate": 5.492583148254779e-06,
+      "loss": 0.7044,
+      "step": 16540
+    },
+    {
+      "epoch": 0.4525505530413858,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.489841243727894e-06,
+      "loss": 0.6045,
+      "step": 16550
+    },
+    {
+      "epoch": 0.4528239974843111,
+      "grad_norm": 3.125,
+      "learning_rate": 5.48709933920101e-06,
+      "loss": 0.7622,
+      "step": 16560
+    },
+    {
+      "epoch": 0.45309744192723644,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.484357434674125e-06,
+      "loss": 0.705,
+      "step": 16570
+    },
+    {
+      "epoch": 0.45337088637016176,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.481615530147241e-06,
+      "loss": 0.7218,
+      "step": 16580
+    },
+    {
+      "epoch": 0.453644330813087,
+      "grad_norm": 2.234375,
+      "learning_rate": 5.4788736256203565e-06,
+      "loss": 0.6365,
+      "step": 16590
+    },
+    {
+      "epoch": 0.45391777525601235,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.476131721093472e-06,
+      "loss": 0.6723,
+      "step": 16600
+    },
+    {
+      "epoch": 0.45419121969893766,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.473389816566587e-06,
+      "loss": 0.7121,
+      "step": 16610
+    },
+    {
+      "epoch": 0.454464664141863,
+      "grad_norm": 2.296875,
+      "learning_rate": 5.470647912039703e-06,
+      "loss": 0.6517,
+      "step": 16620
+    },
+    {
+      "epoch": 0.4547381085847883,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.467906007512818e-06,
+      "loss": 0.6533,
+      "step": 16630
+    },
+    {
+      "epoch": 0.45501155302771357,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.465164102985935e-06,
+      "loss": 0.7379,
+      "step": 16640
+    },
+    {
+      "epoch": 0.4552849974706389,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.462422198459049e-06,
+      "loss": 0.7042,
+      "step": 16650
+    },
+    {
+      "epoch": 0.4555584419135642,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.459680293932166e-06,
+      "loss": 0.7023,
+      "step": 16660
+    },
+    {
+      "epoch": 0.4558318863564895,
+      "grad_norm": 2.25,
+      "learning_rate": 5.456938389405282e-06,
+      "loss": 0.6324,
+      "step": 16670
+    },
+    {
+      "epoch": 0.45610533079941484,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.454196484878397e-06,
+      "loss": 0.6947,
+      "step": 16680
+    },
+    {
+      "epoch": 0.45637877524234016,
+      "grad_norm": 3.65625,
+      "learning_rate": 5.451454580351513e-06,
+      "loss": 0.7733,
+      "step": 16690
+    },
+    {
+      "epoch": 0.4566522196852654,
+      "grad_norm": 2.75,
+      "learning_rate": 5.448712675824628e-06,
+      "loss": 0.7789,
+      "step": 16700
+    },
+    {
+      "epoch": 0.45692566412819074,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.445970771297744e-06,
+      "loss": 0.6233,
+      "step": 16710
+    },
+    {
+      "epoch": 0.45719910857111606,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.44322886677086e-06,
+      "loss": 0.6149,
+      "step": 16720
+    },
+    {
+      "epoch": 0.4574725530140414,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.440486962243975e-06,
+      "loss": 0.635,
+      "step": 16730
+    },
+    {
+      "epoch": 0.4577459974569667,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.437745057717091e-06,
+      "loss": 0.7475,
+      "step": 16740
+    },
+    {
+      "epoch": 0.45801944189989197,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.435003153190206e-06,
+      "loss": 0.5786,
+      "step": 16750
+    },
+    {
+      "epoch": 0.4582928863428173,
+      "grad_norm": 2.625,
+      "learning_rate": 5.432261248663322e-06,
+      "loss": 0.6695,
+      "step": 16760
+    },
+    {
+      "epoch": 0.4585663307857426,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.429519344136438e-06,
+      "loss": 0.7091,
+      "step": 16770
+    },
+    {
+      "epoch": 0.4588397752286679,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.426777439609553e-06,
+      "loss": 0.7884,
+      "step": 16780
+    },
+    {
+      "epoch": 0.45911321967159324,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.424035535082669e-06,
+      "loss": 0.7404,
+      "step": 16790
+    },
+    {
+      "epoch": 0.4593866641145185,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.421293630555784e-06,
+      "loss": 0.6936,
+      "step": 16800
+    },
+    {
+      "epoch": 0.4596601085574438,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.4185517260289e-06,
+      "loss": 0.7395,
+      "step": 16810
+    },
+    {
+      "epoch": 0.45993355300036914,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.415809821502016e-06,
+      "loss": 0.6455,
+      "step": 16820
+    },
+    {
+      "epoch": 0.46020699744329446,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.413067916975131e-06,
+      "loss": 0.7302,
+      "step": 16830
+    },
+    {
+      "epoch": 0.4604804418862198,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.410326012448247e-06,
+      "loss": 0.777,
+      "step": 16840
+    },
+    {
+      "epoch": 0.4607538863291451,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.407584107921363e-06,
+      "loss": 0.7175,
+      "step": 16850
+    },
+    {
+      "epoch": 0.46102733077207037,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.404842203394478e-06,
+      "loss": 0.6448,
+      "step": 16860
+    },
+    {
+      "epoch": 0.4613007752149957,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.402100298867594e-06,
+      "loss": 0.6362,
+      "step": 16870
+    },
+    {
+      "epoch": 0.461574219657921,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.399358394340709e-06,
+      "loss": 0.6346,
+      "step": 16880
+    },
+    {
+      "epoch": 0.4618476641008463,
+      "grad_norm": 3.296875,
+      "learning_rate": 5.396616489813825e-06,
+      "loss": 0.6802,
+      "step": 16890
+    },
+    {
+      "epoch": 0.46212110854377164,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.3938745852869415e-06,
+      "loss": 0.633,
+      "step": 16900
+    },
+    {
+      "epoch": 0.4623945529866969,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.391132680760056e-06,
+      "loss": 0.6539,
+      "step": 16910
+    },
+    {
+      "epoch": 0.4626679974296222,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.3883907762331724e-06,
+      "loss": 0.675,
+      "step": 16920
+    },
+    {
+      "epoch": 0.46294144187254754,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.385648871706287e-06,
+      "loss": 0.7065,
+      "step": 16930
+    },
+    {
+      "epoch": 0.46321488631547286,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.382906967179403e-06,
+      "loss": 0.7009,
+      "step": 16940
+    },
+    {
+      "epoch": 0.4634883307583982,
+      "grad_norm": 2.625,
+      "learning_rate": 5.380165062652519e-06,
+      "loss": 0.7301,
+      "step": 16950
+    },
+    {
+      "epoch": 0.46376177520132345,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.377423158125634e-06,
+      "loss": 0.683,
+      "step": 16960
+    },
+    {
+      "epoch": 0.46403521964424876,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.37468125359875e-06,
+      "loss": 0.7845,
+      "step": 16970
+    },
+    {
+      "epoch": 0.4643086640871741,
+      "grad_norm": 3.234375,
+      "learning_rate": 5.371939349071865e-06,
+      "loss": 0.8183,
+      "step": 16980
+    },
+    {
+      "epoch": 0.4645821085300994,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.369197444544982e-06,
+      "loss": 0.6598,
+      "step": 16990
+    },
+    {
+      "epoch": 0.4648555529730247,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.366455540018097e-06,
+      "loss": 0.673,
+      "step": 17000
+    },
+    {
+      "epoch": 0.46512899741595004,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.363713635491213e-06,
+      "loss": 0.7532,
+      "step": 17010
+    },
+    {
+      "epoch": 0.4654024418588753,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.360971730964328e-06,
+      "loss": 0.6747,
+      "step": 17020
+    },
+    {
+      "epoch": 0.4656758863018006,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.3582298264374444e-06,
+      "loss": 0.6817,
+      "step": 17030
+    },
+    {
+      "epoch": 0.46594933074472594,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.355487921910559e-06,
+      "loss": 0.6773,
+      "step": 17040
+    },
+    {
+      "epoch": 0.46622277518765126,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.352746017383675e-06,
+      "loss": 0.7744,
+      "step": 17050
+    },
+    {
+      "epoch": 0.4664962196305766,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.35000411285679e-06,
+      "loss": 0.6683,
+      "step": 17060
+    },
+    {
+      "epoch": 0.46676966407350184,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.347262208329906e-06,
+      "loss": 0.7371,
+      "step": 17070
+    },
+    {
+      "epoch": 0.46704310851642716,
+      "grad_norm": 2.875,
+      "learning_rate": 5.344520303803023e-06,
+      "loss": 0.7589,
+      "step": 17080
+    },
+    {
+      "epoch": 0.4673165529593525,
+      "grad_norm": 3.203125,
+      "learning_rate": 5.341778399276137e-06,
+      "loss": 0.7275,
+      "step": 17090
+    },
+    {
+      "epoch": 0.4675899974022778,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.339036494749254e-06,
+      "loss": 0.7315,
+      "step": 17100
+    },
+    {
+      "epoch": 0.4678634418452031,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.336294590222368e-06,
+      "loss": 0.6282,
+      "step": 17110
+    },
+    {
+      "epoch": 0.4681368862881284,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.333552685695485e-06,
+      "loss": 0.6921,
+      "step": 17120
+    },
+    {
+      "epoch": 0.4684103307310537,
+      "grad_norm": 3.46875,
+      "learning_rate": 5.3308107811686e-06,
+      "loss": 0.7187,
+      "step": 17130
+    },
+    {
+      "epoch": 0.468683775173979,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.328068876641716e-06,
+      "loss": 0.7035,
+      "step": 17140
+    },
+    {
+      "epoch": 0.46895721961690434,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.325326972114832e-06,
+      "loss": 0.7026,
+      "step": 17150
+    },
+    {
+      "epoch": 0.46923066405982966,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.322585067587947e-06,
+      "loss": 0.735,
+      "step": 17160
+    },
+    {
+      "epoch": 0.469504108502755,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.319843163061063e-06,
+      "loss": 0.7784,
+      "step": 17170
+    },
+    {
+      "epoch": 0.46977755294568024,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.317101258534178e-06,
+      "loss": 0.7082,
+      "step": 17180
+    },
+    {
+      "epoch": 0.47005099738860556,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.314359354007294e-06,
+      "loss": 0.7477,
+      "step": 17190
+    },
+    {
+      "epoch": 0.4703244418315309,
+      "grad_norm": 3.0,
+      "learning_rate": 5.311617449480409e-06,
+      "loss": 0.7412,
+      "step": 17200
+    },
+    {
+      "epoch": 0.4705978862744562,
+      "grad_norm": 2.453125,
+      "learning_rate": 5.308875544953526e-06,
+      "loss": 0.7294,
+      "step": 17210
+    },
+    {
+      "epoch": 0.4708713307173815,
+      "grad_norm": 2.875,
+      "learning_rate": 5.30613364042664e-06,
+      "loss": 0.6907,
+      "step": 17220
+    },
+    {
+      "epoch": 0.4711447751603068,
+      "grad_norm": 2.75,
+      "learning_rate": 5.303391735899757e-06,
+      "loss": 0.6658,
+      "step": 17230
+    },
+    {
+      "epoch": 0.4714182196032321,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.300649831372871e-06,
+      "loss": 0.6401,
+      "step": 17240
+    },
+    {
+      "epoch": 0.4716916640461574,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.297907926845988e-06,
+      "loss": 0.695,
+      "step": 17250
+    },
+    {
+      "epoch": 0.47196510848908274,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.295166022319104e-06,
+      "loss": 0.6576,
+      "step": 17260
+    },
+    {
+      "epoch": 0.47223855293200806,
+      "grad_norm": 2.75,
+      "learning_rate": 5.2924241177922185e-06,
+      "loss": 0.7626,
+      "step": 17270
+    },
+    {
+      "epoch": 0.4725119973749333,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.289682213265335e-06,
+      "loss": 0.706,
+      "step": 17280
+    },
+    {
+      "epoch": 0.47278544181785864,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.2869403087384495e-06,
+      "loss": 0.7233,
+      "step": 17290
+    },
+    {
+      "epoch": 0.47305888626078396,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.284198404211566e-06,
+      "loss": 0.6719,
+      "step": 17300
+    },
+    {
+      "epoch": 0.4733323307037093,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.281456499684682e-06,
+      "loss": 0.7049,
+      "step": 17310
+    },
+    {
+      "epoch": 0.4736057751466346,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.278714595157797e-06,
+      "loss": 0.691,
+      "step": 17320
+    },
+    {
+      "epoch": 0.47387921958955986,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.275972690630913e-06,
+      "loss": 0.7139,
+      "step": 17330
+    },
+    {
+      "epoch": 0.4741526640324852,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.273230786104029e-06,
+      "loss": 0.6278,
+      "step": 17340
+    },
+    {
+      "epoch": 0.4744261084754105,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.270488881577144e-06,
+      "loss": 0.5927,
+      "step": 17350
+    },
+    {
+      "epoch": 0.4746995529183358,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.26774697705026e-06,
+      "loss": 0.6849,
+      "step": 17360
+    },
+    {
+      "epoch": 0.47497299736126114,
+      "grad_norm": 2.28125,
+      "learning_rate": 5.265005072523375e-06,
+      "loss": 0.7456,
+      "step": 17370
+    },
+    {
+      "epoch": 0.47524644180418646,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.2622631679964905e-06,
+      "loss": 0.6039,
+      "step": 17380
+    },
+    {
+      "epoch": 0.4755198862471117,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.259521263469607e-06,
+      "loss": 0.7238,
+      "step": 17390
+    },
+    {
+      "epoch": 0.47579333069003704,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.2567793589427215e-06,
+      "loss": 0.7005,
+      "step": 17400
+    },
+    {
+      "epoch": 0.47606677513296236,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.254037454415838e-06,
+      "loss": 0.7948,
+      "step": 17410
+    },
+    {
+      "epoch": 0.4763402195758877,
+      "grad_norm": 3.078125,
+      "learning_rate": 5.2512955498889525e-06,
+      "loss": 0.6585,
+      "step": 17420
+    },
+    {
+      "epoch": 0.476613664018813,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.248553645362069e-06,
+      "loss": 0.7341,
+      "step": 17430
+    },
+    {
+      "epoch": 0.47688710846173826,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.245811740835185e-06,
+      "loss": 0.6731,
+      "step": 17440
+    },
+    {
+      "epoch": 0.4771605529046636,
+      "grad_norm": 3.0,
+      "learning_rate": 5.2430698363083e-06,
+      "loss": 0.7003,
+      "step": 17450
+    },
+    {
+      "epoch": 0.4774339973475889,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.240327931781416e-06,
+      "loss": 0.6762,
+      "step": 17460
+    },
+    {
+      "epoch": 0.4777074417905142,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.237586027254531e-06,
+      "loss": 0.678,
+      "step": 17470
+    },
+    {
+      "epoch": 0.47798088623343954,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.234844122727647e-06,
+      "loss": 0.7466,
+      "step": 17480
+    },
+    {
+      "epoch": 0.4782543306763648,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.232102218200763e-06,
+      "loss": 0.6788,
+      "step": 17490
+    },
+    {
+      "epoch": 0.4785277751192901,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.229360313673878e-06,
+      "loss": 0.6385,
+      "step": 17500
+    },
+    {
+      "epoch": 0.47880121956221544,
+      "grad_norm": 3.578125,
+      "learning_rate": 5.226618409146994e-06,
+      "loss": 0.6834,
+      "step": 17510
+    },
+    {
+      "epoch": 0.47907466400514076,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.22387650462011e-06,
+      "loss": 0.6528,
+      "step": 17520
+    },
+    {
+      "epoch": 0.4793481084480661,
+      "grad_norm": 2.03125,
+      "learning_rate": 5.221134600093225e-06,
+      "loss": 0.6356,
+      "step": 17530
+    },
+    {
+      "epoch": 0.4796215528909914,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.218392695566341e-06,
+      "loss": 0.6317,
+      "step": 17540
+    },
+    {
+      "epoch": 0.47989499733391666,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.215650791039456e-06,
+      "loss": 0.6527,
+      "step": 17550
+    },
+    {
+      "epoch": 0.480168441776842,
+      "grad_norm": 2.328125,
+      "learning_rate": 5.212908886512572e-06,
+      "loss": 0.6464,
+      "step": 17560
+    },
+    {
+      "epoch": 0.4804418862197673,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.210166981985688e-06,
+      "loss": 0.641,
+      "step": 17570
+    },
+    {
+      "epoch": 0.4807153306626926,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.207425077458803e-06,
+      "loss": 0.7125,
+      "step": 17580
+    },
+    {
+      "epoch": 0.48098877510561794,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.204683172931919e-06,
+      "loss": 0.6778,
+      "step": 17590
+    },
+    {
+      "epoch": 0.4812622195485432,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.201941268405034e-06,
+      "loss": 0.7131,
+      "step": 17600
+    },
+    {
+      "epoch": 0.4815356639914685,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.19919936387815e-06,
+      "loss": 0.7327,
+      "step": 17610
+    },
+    {
+      "epoch": 0.48180910843439384,
+      "grad_norm": 3.125,
+      "learning_rate": 5.196457459351266e-06,
+      "loss": 0.6985,
+      "step": 17620
+    },
+    {
+      "epoch": 0.48208255287731916,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.193715554824381e-06,
+      "loss": 0.7478,
+      "step": 17630
+    },
+    {
+      "epoch": 0.4823559973202445,
+      "grad_norm": 2.59375,
+      "learning_rate": 5.190973650297497e-06,
+      "loss": 0.7122,
+      "step": 17640
+    },
+    {
+      "epoch": 0.48262944176316974,
+      "grad_norm": 2.390625,
+      "learning_rate": 5.188231745770612e-06,
+      "loss": 0.684,
+      "step": 17650
+    },
+    {
+      "epoch": 0.48290288620609506,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.185489841243728e-06,
+      "loss": 0.711,
+      "step": 17660
+    },
+    {
+      "epoch": 0.4831763306490204,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.182747936716845e-06,
+      "loss": 0.6777,
+      "step": 17670
+    },
+    {
+      "epoch": 0.4834497750919457,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.180006032189959e-06,
+      "loss": 0.7224,
+      "step": 17680
+    },
+    {
+      "epoch": 0.483723219534871,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.1772641276630756e-06,
+      "loss": 0.7227,
+      "step": 17690
+    },
+    {
+      "epoch": 0.48399666397779634,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.174522223136191e-06,
+      "loss": 0.7225,
+      "step": 17700
+    },
+    {
+      "epoch": 0.4842701084207216,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.1717803186093065e-06,
+      "loss": 0.6859,
+      "step": 17710
+    },
+    {
+      "epoch": 0.4845435528636469,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.169038414082422e-06,
+      "loss": 0.6958,
+      "step": 17720
+    },
+    {
+      "epoch": 0.48481699730657224,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.1662965095555375e-06,
+      "loss": 0.7592,
+      "step": 17730
+    },
+    {
+      "epoch": 0.48509044174949756,
+      "grad_norm": 2.625,
+      "learning_rate": 5.163554605028653e-06,
+      "loss": 0.63,
+      "step": 17740
+    },
+    {
+      "epoch": 0.4853638861924229,
+      "grad_norm": 3.140625,
+      "learning_rate": 5.160812700501769e-06,
+      "loss": 0.7445,
+      "step": 17750
+    },
+    {
+      "epoch": 0.48563733063534814,
+      "grad_norm": 3.40625,
+      "learning_rate": 5.158070795974884e-06,
+      "loss": 0.6839,
+      "step": 17760
+    },
+    {
+      "epoch": 0.48591077507827346,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.155328891448e-06,
+      "loss": 0.6781,
+      "step": 17770
+    },
+    {
+      "epoch": 0.4861842195211988,
+      "grad_norm": 3.484375,
+      "learning_rate": 5.152586986921116e-06,
+      "loss": 0.753,
+      "step": 17780
+    },
+    {
+      "epoch": 0.4864576639641241,
+      "grad_norm": 2.40625,
+      "learning_rate": 5.149845082394231e-06,
+      "loss": 0.7041,
+      "step": 17790
+    },
+    {
+      "epoch": 0.4867311084070494,
+      "grad_norm": 3.265625,
+      "learning_rate": 5.1471031778673476e-06,
+      "loss": 0.6776,
+      "step": 17800
+    },
+    {
+      "epoch": 0.4870045528499747,
+      "grad_norm": 3.03125,
+      "learning_rate": 5.144361273340462e-06,
+      "loss": 0.7598,
+      "step": 17810
+    },
+    {
+      "epoch": 0.4872779972929,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.1416193688135785e-06,
+      "loss": 0.7051,
+      "step": 17820
+    },
+    {
+      "epoch": 0.4875514417358253,
+      "grad_norm": 2.4375,
+      "learning_rate": 5.138877464286693e-06,
+      "loss": 0.615,
+      "step": 17830
+    },
+    {
+      "epoch": 0.48782488617875064,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.1361355597598095e-06,
+      "loss": 0.6564,
+      "step": 17840
+    },
+    {
+      "epoch": 0.48809833062167596,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.133393655232926e-06,
+      "loss": 0.7284,
+      "step": 17850
+    },
+    {
+      "epoch": 0.4883717750646013,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.1306517507060405e-06,
+      "loss": 0.6361,
+      "step": 17860
+    },
+    {
+      "epoch": 0.48864521950752654,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.127909846179157e-06,
+      "loss": 0.6992,
+      "step": 17870
+    },
+    {
+      "epoch": 0.48891866395045186,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.125167941652272e-06,
+      "loss": 0.6776,
+      "step": 17880
+    },
+    {
+      "epoch": 0.4891921083933772,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.122426037125388e-06,
+      "loss": 0.5967,
+      "step": 17890
+    },
+    {
+      "epoch": 0.4894655528363025,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.119684132598503e-06,
+      "loss": 0.6571,
+      "step": 17900
+    },
+    {
+      "epoch": 0.4897389972792278,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.116942228071619e-06,
+      "loss": 0.7763,
+      "step": 17910
+    },
+    {
+      "epoch": 0.4900124417221531,
+      "grad_norm": 2.53125,
+      "learning_rate": 5.114200323544734e-06,
+      "loss": 0.6491,
+      "step": 17920
+    },
+    {
+      "epoch": 0.4902858861650784,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.1114584190178505e-06,
+      "loss": 0.6642,
+      "step": 17930
+    },
+    {
+      "epoch": 0.4905593306080037,
+      "grad_norm": 2.90625,
+      "learning_rate": 5.108716514490966e-06,
+      "loss": 0.7604,
+      "step": 17940
+    },
+    {
+      "epoch": 0.49083277505092904,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.1059746099640815e-06,
+      "loss": 0.6204,
+      "step": 17950
+    },
+    {
+      "epoch": 0.49110621949385436,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.103232705437197e-06,
+      "loss": 0.7275,
+      "step": 17960
+    },
+    {
+      "epoch": 0.4913796639367796,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.1004908009103124e-06,
+      "loss": 0.7532,
+      "step": 17970
+    },
+    {
+      "epoch": 0.49165310837970494,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.097748896383429e-06,
+      "loss": 0.6374,
+      "step": 17980
+    },
+    {
+      "epoch": 0.49192655282263026,
+      "grad_norm": 2.3125,
+      "learning_rate": 5.095006991856543e-06,
+      "loss": 0.7289,
+      "step": 17990
+    },
+    {
+      "epoch": 0.4921999972655556,
+      "grad_norm": 3.015625,
+      "learning_rate": 5.09226508732966e-06,
+      "loss": 0.7089,
+      "step": 18000
+    },
+    {
+      "epoch": 0.4924734417084809,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.089523182802774e-06,
+      "loss": 0.6699,
+      "step": 18010
+    },
+    {
+      "epoch": 0.49274688615140616,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.086781278275891e-06,
+      "loss": 0.6772,
+      "step": 18020
+    },
+    {
+      "epoch": 0.4930203305943315,
+      "grad_norm": 2.984375,
+      "learning_rate": 5.084039373749007e-06,
+      "loss": 0.6919,
+      "step": 18030
+    },
+    {
+      "epoch": 0.4932937750372568,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.081297469222122e-06,
+      "loss": 0.6288,
+      "step": 18040
+    },
+    {
+      "epoch": 0.4935672194801821,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.078555564695238e-06,
+      "loss": 0.7086,
+      "step": 18050
+    },
+    {
+      "epoch": 0.49384066392310744,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.0758136601683535e-06,
+      "loss": 0.6064,
+      "step": 18060
+    },
+    {
+      "epoch": 0.49411410836603276,
+      "grad_norm": 2.96875,
+      "learning_rate": 5.073071755641469e-06,
+      "loss": 0.6964,
+      "step": 18070
+    },
+    {
+      "epoch": 0.494387552808958,
+      "grad_norm": 2.515625,
+      "learning_rate": 5.070329851114585e-06,
+      "loss": 0.6649,
+      "step": 18080
+    },
+    {
+      "epoch": 0.49466099725188334,
+      "grad_norm": 3.171875,
+      "learning_rate": 5.0675879465877e-06,
+      "loss": 0.6752,
+      "step": 18090
+    },
+    {
+      "epoch": 0.49493444169480866,
+      "grad_norm": 3.390625,
+      "learning_rate": 5.064846042060816e-06,
+      "loss": 0.7428,
+      "step": 18100
+    },
+    {
+      "epoch": 0.495207886137734,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.062104137533932e-06,
+      "loss": 0.6359,
+      "step": 18110
+    },
+    {
+      "epoch": 0.4954813305806593,
+      "grad_norm": 2.703125,
+      "learning_rate": 5.059362233007047e-06,
+      "loss": 0.6202,
+      "step": 18120
+    },
+    {
+      "epoch": 0.49575477502358456,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.056620328480163e-06,
+      "loss": 0.7494,
+      "step": 18130
+    },
+    {
+      "epoch": 0.4960282194665099,
+      "grad_norm": 2.375,
+      "learning_rate": 5.053878423953278e-06,
+      "loss": 0.7305,
+      "step": 18140
+    },
+    {
+      "epoch": 0.4963016639094352,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.051136519426394e-06,
+      "loss": 0.7331,
+      "step": 18150
+    },
+    {
+      "epoch": 0.4965751083523605,
+      "grad_norm": 3.28125,
+      "learning_rate": 5.04839461489951e-06,
+      "loss": 0.7122,
+      "step": 18160
+    },
+    {
+      "epoch": 0.49684855279528584,
+      "grad_norm": 2.859375,
+      "learning_rate": 5.045652710372625e-06,
+      "loss": 0.6936,
+      "step": 18170
+    },
+    {
+      "epoch": 0.4971219972382111,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.042910805845741e-06,
+      "loss": 0.6649,
+      "step": 18180
+    },
+    {
+      "epoch": 0.4973954416811364,
+      "grad_norm": 2.421875,
+      "learning_rate": 5.040168901318856e-06,
+      "loss": 0.6629,
+      "step": 18190
+    },
+    {
+      "epoch": 0.49766888612406174,
+      "grad_norm": 2.796875,
+      "learning_rate": 5.037426996791972e-06,
+      "loss": 0.6654,
+      "step": 18200
+    },
+    {
+      "epoch": 0.49794233056698706,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.034685092265088e-06,
+      "loss": 0.6939,
+      "step": 18210
+    },
+    {
+      "epoch": 0.4982157750099124,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.031943187738203e-06,
+      "loss": 0.6656,
+      "step": 18220
+    },
+    {
+      "epoch": 0.4984892194528377,
+      "grad_norm": 2.5625,
+      "learning_rate": 5.029201283211319e-06,
+      "loss": 0.7188,
+      "step": 18230
+    },
+    {
+      "epoch": 0.49876266389576296,
+      "grad_norm": 2.46875,
+      "learning_rate": 5.0264593786844355e-06,
+      "loss": 0.6231,
+      "step": 18240
+    },
+    {
+      "epoch": 0.4990361083386883,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.02371747415755e-06,
+      "loss": 0.6916,
+      "step": 18250
+    },
+    {
+      "epoch": 0.4993095527816136,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.0209755696306665e-06,
+      "loss": 0.6731,
+      "step": 18260
+    },
+    {
+      "epoch": 0.4995829972245389,
+      "grad_norm": 2.640625,
+      "learning_rate": 5.018233665103781e-06,
+      "loss": 0.645,
+      "step": 18270
+    },
+    {
+      "epoch": 0.49985644166746424,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.0154917605768975e-06,
+      "loss": 0.7123,
+      "step": 18280
+    },
+    {
+      "epoch": 0.5001298861103896,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.012749856050013e-06,
+      "loss": 0.6978,
+      "step": 18290
+    },
+    {
+      "epoch": 0.5004033305533149,
+      "grad_norm": 2.625,
+      "learning_rate": 5.0100079515231284e-06,
+      "loss": 0.6407,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5006767749962402,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.007266046996244e-06,
+      "loss": 0.7159,
+      "step": 18310
+    },
+    {
+      "epoch": 0.5009502194391654,
+      "grad_norm": 2.34375,
+      "learning_rate": 5.004524142469359e-06,
+      "loss": 0.6525,
+      "step": 18320
+    },
+    {
+      "epoch": 0.5012236638820907,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.001782237942475e-06,
+      "loss": 0.7188,
+      "step": 18330
+    },
+    {
+      "epoch": 0.501497108325016,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.99904033341559e-06,
+      "loss": 0.6594,
+      "step": 18340
+    },
+    {
+      "epoch": 0.5017705527679414,
+      "grad_norm": 2.75,
+      "learning_rate": 4.996298428888707e-06,
+      "loss": 0.6811,
+      "step": 18350
+    },
+    {
+      "epoch": 0.5020439972108667,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.993556524361822e-06,
+      "loss": 0.7097,
+      "step": 18360
+    },
+    {
+      "epoch": 0.502317441653792,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.990814619834938e-06,
+      "loss": 0.6983,
+      "step": 18370
+    },
+    {
+      "epoch": 0.5025908860967173,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.988072715308053e-06,
+      "loss": 0.6817,
+      "step": 18380
+    },
+    {
+      "epoch": 0.5028643305396426,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.985330810781169e-06,
+      "loss": 0.7247,
+      "step": 18390
+    },
+    {
+      "epoch": 0.503137774982568,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.982588906254285e-06,
+      "loss": 0.6772,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5034112194254933,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.9798470017274004e-06,
+      "loss": 0.7281,
+      "step": 18410
+    },
+    {
+      "epoch": 0.5036846638684185,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.977105097200516e-06,
+      "loss": 0.717,
+      "step": 18420
+    },
+    {
+      "epoch": 0.5039581083113438,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.974363192673631e-06,
+      "loss": 0.6924,
+      "step": 18430
+    },
+    {
+      "epoch": 0.5042315527542691,
+      "grad_norm": 2.625,
+      "learning_rate": 4.971621288146748e-06,
+      "loss": 0.752,
+      "step": 18440
+    },
+    {
+      "epoch": 0.5045049971971944,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.968879383619863e-06,
+      "loss": 0.7681,
+      "step": 18450
+    },
+    {
+      "epoch": 0.5047784416401198,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.966137479092979e-06,
+      "loss": 0.6581,
+      "step": 18460
+    },
+    {
+      "epoch": 0.5050518860830451,
+      "grad_norm": 3.0,
+      "learning_rate": 4.963395574566094e-06,
+      "loss": 0.7711,
+      "step": 18470
+    },
+    {
+      "epoch": 0.5053253305259704,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.96065367003921e-06,
+      "loss": 0.6294,
+      "step": 18480
+    },
+    {
+      "epoch": 0.5055987749688957,
+      "grad_norm": 2.75,
+      "learning_rate": 4.957911765512325e-06,
+      "loss": 0.71,
+      "step": 18490
+    },
+    {
+      "epoch": 0.505872219411821,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.955169860985441e-06,
+      "loss": 0.6588,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5061456638547464,
+      "grad_norm": 3.75,
+      "learning_rate": 4.952427956458556e-06,
+      "loss": 0.7516,
+      "step": 18510
+    },
+    {
+      "epoch": 0.5064191082976717,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.949686051931672e-06,
+      "loss": 0.6872,
+      "step": 18520
+    },
+    {
+      "epoch": 0.5066925527405969,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.946944147404788e-06,
+      "loss": 0.6495,
+      "step": 18530
+    },
+    {
+      "epoch": 0.5069659971835222,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.944202242877903e-06,
+      "loss": 0.7672,
+      "step": 18540
+    },
+    {
+      "epoch": 0.5072394416264475,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.941460338351019e-06,
+      "loss": 0.7386,
+      "step": 18550
+    },
+    {
+      "epoch": 0.5075128860693728,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.938718433824134e-06,
+      "loss": 0.7186,
+      "step": 18560
+    },
+    {
+      "epoch": 0.5077863305122982,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.93597652929725e-06,
+      "loss": 0.7155,
+      "step": 18570
+    },
+    {
+      "epoch": 0.5080597749552235,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.933234624770366e-06,
+      "loss": 0.6686,
+      "step": 18580
+    },
+    {
+      "epoch": 0.5083332193981488,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.930492720243482e-06,
+      "loss": 0.682,
+      "step": 18590
+    },
+    {
+      "epoch": 0.5086066638410741,
+      "grad_norm": 3.0,
+      "learning_rate": 4.927750815716597e-06,
+      "loss": 0.6497,
+      "step": 18600
+    },
+    {
+      "epoch": 0.5088801082839994,
+      "grad_norm": 2.296875,
+      "learning_rate": 4.925008911189713e-06,
+      "loss": 0.6369,
+      "step": 18610
+    },
+    {
+      "epoch": 0.5091535527269248,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.922267006662829e-06,
+      "loss": 0.6718,
+      "step": 18620
+    },
+    {
+      "epoch": 0.5094269971698501,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.919525102135944e-06,
+      "loss": 0.6862,
+      "step": 18630
+    },
+    {
+      "epoch": 0.5097004416127753,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.91678319760906e-06,
+      "loss": 0.6695,
+      "step": 18640
+    },
+    {
+      "epoch": 0.5099738860557006,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.914041293082175e-06,
+      "loss": 0.7417,
+      "step": 18650
+    },
+    {
+      "epoch": 0.5102473304986259,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.911299388555291e-06,
+      "loss": 0.7005,
+      "step": 18660
+    },
+    {
+      "epoch": 0.5105207749415512,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.908557484028406e-06,
+      "loss": 0.7466,
+      "step": 18670
+    },
+    {
+      "epoch": 0.5107942193844766,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.905815579501522e-06,
+      "loss": 0.6777,
+      "step": 18680
+    },
+    {
+      "epoch": 0.5110676638274019,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.903073674974637e-06,
+      "loss": 0.6993,
+      "step": 18690
+    },
+    {
+      "epoch": 0.5113411082703272,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.900331770447753e-06,
+      "loss": 0.6274,
+      "step": 18700
+    },
+    {
+      "epoch": 0.5116145527132525,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.897589865920869e-06,
+      "loss": 0.7551,
+      "step": 18710
+    },
+    {
+      "epoch": 0.5118879971561778,
+      "grad_norm": 3.328125,
+      "learning_rate": 4.894847961393985e-06,
+      "loss": 0.6499,
+      "step": 18720
+    },
+    {
+      "epoch": 0.5121614415991032,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.8921060568671e-06,
+      "loss": 0.6998,
+      "step": 18730
+    },
+    {
+      "epoch": 0.5124348860420284,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.8893641523402156e-06,
+      "loss": 0.5908,
+      "step": 18740
+    },
+    {
+      "epoch": 0.5127083304849537,
+      "grad_norm": 3.453125,
+      "learning_rate": 4.886622247813331e-06,
+      "loss": 0.7284,
+      "step": 18750
+    },
+    {
+      "epoch": 0.512981774927879,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.883880343286447e-06,
+      "loss": 0.6817,
+      "step": 18760
+    },
+    {
+      "epoch": 0.5132552193708043,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.881138438759563e-06,
+      "loss": 0.6707,
+      "step": 18770
+    },
+    {
+      "epoch": 0.5135286638137296,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.878396534232678e-06,
+      "loss": 0.7267,
+      "step": 18780
+    },
+    {
+      "epoch": 0.513802108256655,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.875654629705794e-06,
+      "loss": 0.7265,
+      "step": 18790
+    },
+    {
+      "epoch": 0.5140755526995803,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.87291272517891e-06,
+      "loss": 0.6385,
+      "step": 18800
+    },
+    {
+      "epoch": 0.5143489971425056,
+      "grad_norm": 3.125,
+      "learning_rate": 4.870170820652026e-06,
+      "loss": 0.7145,
+      "step": 18810
+    },
+    {
+      "epoch": 0.5146224415854309,
+      "grad_norm": 3.46875,
+      "learning_rate": 4.867428916125141e-06,
+      "loss": 0.6973,
+      "step": 18820
+    },
+    {
+      "epoch": 0.5148958860283562,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.864687011598257e-06,
+      "loss": 0.7059,
+      "step": 18830
+    },
+    {
+      "epoch": 0.5151693304712816,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.861945107071372e-06,
+      "loss": 0.6898,
+      "step": 18840
+    },
+    {
+      "epoch": 0.5154427749142068,
+      "grad_norm": 3.0,
+      "learning_rate": 4.8592032025444876e-06,
+      "loss": 0.7593,
+      "step": 18850
+    },
+    {
+      "epoch": 0.5157162193571321,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.856461298017603e-06,
+      "loss": 0.7073,
+      "step": 18860
+    },
+    {
+      "epoch": 0.5159896638000574,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.853719393490719e-06,
+      "loss": 0.7068,
+      "step": 18870
+    },
+    {
+      "epoch": 0.5162631082429827,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.850977488963835e-06,
+      "loss": 0.6895,
+      "step": 18880
+    },
+    {
+      "epoch": 0.516536552685908,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.84823558443695e-06,
+      "loss": 0.6914,
+      "step": 18890
+    },
+    {
+      "epoch": 0.5168099971288334,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.845493679910066e-06,
+      "loss": 0.6822,
+      "step": 18900
+    },
+    {
+      "epoch": 0.5170834415717587,
+      "grad_norm": 3.25,
+      "learning_rate": 4.842751775383181e-06,
+      "loss": 0.654,
+      "step": 18910
+    },
+    {
+      "epoch": 0.517356886014684,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.840009870856297e-06,
+      "loss": 0.6853,
+      "step": 18920
+    },
+    {
+      "epoch": 0.5176303304576093,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.837267966329412e-06,
+      "loss": 0.6625,
+      "step": 18930
+    },
+    {
+      "epoch": 0.5179037749005346,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.834526061802529e-06,
+      "loss": 0.6396,
+      "step": 18940
+    },
+    {
+      "epoch": 0.5181772193434598,
+      "grad_norm": 3.203125,
+      "learning_rate": 4.831784157275644e-06,
+      "loss": 0.7489,
+      "step": 18950
+    },
+    {
+      "epoch": 0.5184506637863852,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.8290422527487596e-06,
+      "loss": 0.7637,
+      "step": 18960
+    },
+    {
+      "epoch": 0.5187241082293105,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.826300348221875e-06,
+      "loss": 0.7256,
+      "step": 18970
+    },
+    {
+      "epoch": 0.5189975526722358,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.823558443694991e-06,
+      "loss": 0.6722,
+      "step": 18980
+    },
+    {
+      "epoch": 0.5192709971151611,
+      "grad_norm": 2.375,
+      "learning_rate": 4.820816539168107e-06,
+      "loss": 0.7188,
+      "step": 18990
+    },
+    {
+      "epoch": 0.5195444415580864,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.818074634641222e-06,
+      "loss": 0.6522,
+      "step": 19000
+    },
+    {
+      "epoch": 0.5198178860010118,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.815332730114338e-06,
+      "loss": 0.6654,
+      "step": 19010
+    },
+    {
+      "epoch": 0.5200913304439371,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.812590825587453e-06,
+      "loss": 0.7387,
+      "step": 19020
+    },
+    {
+      "epoch": 0.5203647748868624,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.80984892106057e-06,
+      "loss": 0.6441,
+      "step": 19030
+    },
+    {
+      "epoch": 0.5206382193297877,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.807107016533685e-06,
+      "loss": 0.6674,
+      "step": 19040
+    },
+    {
+      "epoch": 0.520911663772713,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.804365112006801e-06,
+      "loss": 0.7107,
+      "step": 19050
+    },
+    {
+      "epoch": 0.5211851082156382,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.801623207479916e-06,
+      "loss": 0.7812,
+      "step": 19060
+    },
+    {
+      "epoch": 0.5214585526585636,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.7988813029530316e-06,
+      "loss": 0.6002,
+      "step": 19070
+    },
+    {
+      "epoch": 0.5217319971014889,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.796139398426147e-06,
+      "loss": 0.6217,
+      "step": 19080
+    },
+    {
+      "epoch": 0.5220054415444142,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7933974938992625e-06,
+      "loss": 0.6772,
+      "step": 19090
+    },
+    {
+      "epoch": 0.5222788859873395,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.790655589372378e-06,
+      "loss": 0.5856,
+      "step": 19100
+    },
+    {
+      "epoch": 0.5225523304302648,
+      "grad_norm": 3.125,
+      "learning_rate": 4.787913684845494e-06,
+      "loss": 0.6684,
+      "step": 19110
+    },
+    {
+      "epoch": 0.5228257748731902,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.78517178031861e-06,
+      "loss": 0.6603,
+      "step": 19120
+    },
+    {
+      "epoch": 0.5230992193161155,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.782429875791725e-06,
+      "loss": 0.6754,
+      "step": 19130
+    },
+    {
+      "epoch": 0.5233726637590408,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.779687971264841e-06,
+      "loss": 0.6245,
+      "step": 19140
+    },
+    {
+      "epoch": 0.5236461082019661,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.776946066737956e-06,
+      "loss": 0.7126,
+      "step": 19150
+    },
+    {
+      "epoch": 0.5239195526448914,
+      "grad_norm": 3.578125,
+      "learning_rate": 4.774204162211073e-06,
+      "loss": 0.668,
+      "step": 19160
+    },
+    {
+      "epoch": 0.5241929970878166,
+      "grad_norm": 3.21875,
+      "learning_rate": 4.771462257684188e-06,
+      "loss": 0.7577,
+      "step": 19170
+    },
+    {
+      "epoch": 0.524466441530742,
+      "grad_norm": 2.5,
+      "learning_rate": 4.7687203531573036e-06,
+      "loss": 0.7556,
+      "step": 19180
+    },
+    {
+      "epoch": 0.5247398859736673,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.765978448630419e-06,
+      "loss": 0.7141,
+      "step": 19190
+    },
+    {
+      "epoch": 0.5250133304165926,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.763236544103535e-06,
+      "loss": 0.6281,
+      "step": 19200
+    },
+    {
+      "epoch": 0.5252867748595179,
+      "grad_norm": 2.75,
+      "learning_rate": 4.760494639576651e-06,
+      "loss": 0.7107,
+      "step": 19210
+    },
+    {
+      "epoch": 0.5255602193024432,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.757752735049766e-06,
+      "loss": 0.726,
+      "step": 19220
+    },
+    {
+      "epoch": 0.5258336637453686,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.755010830522882e-06,
+      "loss": 0.6732,
+      "step": 19230
+    },
+    {
+      "epoch": 0.5261071081882939,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.752268925995997e-06,
+      "loss": 0.7208,
+      "step": 19240
+    },
+    {
+      "epoch": 0.5263805526312192,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.749527021469113e-06,
+      "loss": 0.6911,
+      "step": 19250
+    },
+    {
+      "epoch": 0.5266539970741445,
+      "grad_norm": 3.0,
+      "learning_rate": 4.746785116942228e-06,
+      "loss": 0.7489,
+      "step": 19260
+    },
+    {
+      "epoch": 0.5269274415170697,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.744043212415344e-06,
+      "loss": 0.6225,
+      "step": 19270
+    },
+    {
+      "epoch": 0.527200885959995,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.741301307888459e-06,
+      "loss": 0.6847,
+      "step": 19280
+    },
+    {
+      "epoch": 0.5274743304029204,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7385594033615755e-06,
+      "loss": 0.7516,
+      "step": 19290
+    },
+    {
+      "epoch": 0.5277477748458457,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.735817498834691e-06,
+      "loss": 0.7448,
+      "step": 19300
+    },
+    {
+      "epoch": 0.528021219288771,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.7330755943078065e-06,
+      "loss": 0.7301,
+      "step": 19310
+    },
+    {
+      "epoch": 0.5282946637316963,
+      "grad_norm": 3.125,
+      "learning_rate": 4.730333689780922e-06,
+      "loss": 0.6996,
+      "step": 19320
+    },
+    {
+      "epoch": 0.5285681081746216,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.7275917852540375e-06,
+      "loss": 0.6248,
+      "step": 19330
+    },
+    {
+      "epoch": 0.528841552617547,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.724849880727154e-06,
+      "loss": 0.7418,
+      "step": 19340
+    },
+    {
+      "epoch": 0.5291149970604723,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.722107976200269e-06,
+      "loss": 0.6927,
+      "step": 19350
+    },
+    {
+      "epoch": 0.5293884415033976,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.719366071673385e-06,
+      "loss": 0.7105,
+      "step": 19360
+    },
+    {
+      "epoch": 0.5296618859463229,
+      "grad_norm": 2.34375,
+      "learning_rate": 4.7166241671465e-06,
+      "loss": 0.6705,
+      "step": 19370
+    },
+    {
+      "epoch": 0.5299353303892481,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.713882262619617e-06,
+      "loss": 0.6294,
+      "step": 19380
+    },
+    {
+      "epoch": 0.5302087748321734,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.711140358092732e-06,
+      "loss": 0.6245,
+      "step": 19390
+    },
+    {
+      "epoch": 0.5304822192750988,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.7083984535658475e-06,
+      "loss": 0.7179,
+      "step": 19400
+    },
+    {
+      "epoch": 0.5307556637180241,
+      "grad_norm": 2.5,
+      "learning_rate": 4.705656549038963e-06,
+      "loss": 0.6198,
+      "step": 19410
+    },
+    {
+      "epoch": 0.5310291081609494,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.7029146445120785e-06,
+      "loss": 0.7195,
+      "step": 19420
+    },
+    {
+      "epoch": 0.5313025526038747,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.700172739985194e-06,
+      "loss": 0.6926,
+      "step": 19430
+    },
+    {
+      "epoch": 0.5315759970468,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6974308354583095e-06,
+      "loss": 0.6234,
+      "step": 19440
+    },
+    {
+      "epoch": 0.5318494414897253,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.694688930931425e-06,
+      "loss": 0.6912,
+      "step": 19450
+    },
+    {
+      "epoch": 0.5321228859326507,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.6919470264045404e-06,
+      "loss": 0.6584,
+      "step": 19460
+    },
+    {
+      "epoch": 0.532396330375576,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.689205121877657e-06,
+      "loss": 0.7239,
+      "step": 19470
+    },
+    {
+      "epoch": 0.5326697748185012,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.686463217350772e-06,
+      "loss": 0.7124,
+      "step": 19480
+    },
+    {
+      "epoch": 0.5329432192614265,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.683721312823888e-06,
+      "loss": 0.6204,
+      "step": 19490
+    },
+    {
+      "epoch": 0.5332166637043518,
+      "grad_norm": 3.390625,
+      "learning_rate": 4.680979408297003e-06,
+      "loss": 0.7051,
+      "step": 19500
+    },
+    {
+      "epoch": 0.5334901081472772,
+      "grad_norm": 3.09375,
+      "learning_rate": 4.678237503770119e-06,
+      "loss": 0.7524,
+      "step": 19510
+    },
+    {
+      "epoch": 0.5337635525902025,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.675495599243235e-06,
+      "loss": 0.679,
+      "step": 19520
+    },
+    {
+      "epoch": 0.5340369970331278,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.6727536947163505e-06,
+      "loss": 0.7024,
+      "step": 19530
+    },
+    {
+      "epoch": 0.5343104414760531,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.670011790189466e-06,
+      "loss": 0.6619,
+      "step": 19540
+    },
+    {
+      "epoch": 0.5345838859189784,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.6672698856625815e-06,
+      "loss": 0.6081,
+      "step": 19550
+    },
+    {
+      "epoch": 0.5348573303619037,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.664527981135698e-06,
+      "loss": 0.6983,
+      "step": 19560
+    },
+    {
+      "epoch": 0.5351307748048291,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.661786076608813e-06,
+      "loss": 0.6668,
+      "step": 19570
+    },
+    {
+      "epoch": 0.5354042192477544,
+      "grad_norm": 2.625,
+      "learning_rate": 4.659044172081929e-06,
+      "loss": 0.6601,
+      "step": 19580
+    },
+    {
+      "epoch": 0.5356776636906796,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.656302267555044e-06,
+      "loss": 0.7087,
+      "step": 19590
+    },
+    {
+      "epoch": 0.5359511081336049,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.65356036302816e-06,
+      "loss": 0.7753,
+      "step": 19600
+    },
+    {
+      "epoch": 0.5362245525765302,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.650818458501275e-06,
+      "loss": 0.7579,
+      "step": 19610
+    },
+    {
+      "epoch": 0.5364979970194556,
+      "grad_norm": 2.875,
+      "learning_rate": 4.648076553974391e-06,
+      "loss": 0.7217,
+      "step": 19620
+    },
+    {
+      "epoch": 0.5367714414623809,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.645334649447506e-06,
+      "loss": 0.7523,
+      "step": 19630
+    },
+    {
+      "epoch": 0.5370448859053062,
+      "grad_norm": 3.359375,
+      "learning_rate": 4.642592744920622e-06,
+      "loss": 0.6836,
+      "step": 19640
+    },
+    {
+      "epoch": 0.5373183303482315,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.639850840393738e-06,
+      "loss": 0.7166,
+      "step": 19650
+    },
+    {
+      "epoch": 0.5375917747911568,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.6371089358668535e-06,
+      "loss": 0.7123,
+      "step": 19660
+    },
+    {
+      "epoch": 0.5378652192340821,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.634367031339969e-06,
+      "loss": 0.7586,
+      "step": 19670
+    },
+    {
+      "epoch": 0.5381386636770075,
+      "grad_norm": 2.75,
+      "learning_rate": 4.6316251268130844e-06,
+      "loss": 0.6322,
+      "step": 19680
+    },
+    {
+      "epoch": 0.5384121081199328,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.6288832222862e-06,
+      "loss": 0.6921,
+      "step": 19690
+    },
+    {
+      "epoch": 0.538685552562858,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.626141317759316e-06,
+      "loss": 0.664,
+      "step": 19700
+    },
+    {
+      "epoch": 0.5389589970057833,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.623399413232432e-06,
+      "loss": 0.7034,
+      "step": 19710
+    },
+    {
+      "epoch": 0.5392324414487086,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.620657508705547e-06,
+      "loss": 0.6747,
+      "step": 19720
+    },
+    {
+      "epoch": 0.539505885891634,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.617915604178663e-06,
+      "loss": 0.6734,
+      "step": 19730
+    },
+    {
+      "epoch": 0.5397793303345593,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.615173699651779e-06,
+      "loss": 0.6742,
+      "step": 19740
+    },
+    {
+      "epoch": 0.5400527747774846,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.6124317951248945e-06,
+      "loss": 0.7365,
+      "step": 19750
+    },
+    {
+      "epoch": 0.5403262192204099,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.60968989059801e-06,
+      "loss": 0.7404,
+      "step": 19760
+    },
+    {
+      "epoch": 0.5405996636633352,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.6069479860711255e-06,
+      "loss": 0.6442,
+      "step": 19770
+    },
+    {
+      "epoch": 0.5408731081062605,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.604206081544241e-06,
+      "loss": 0.6004,
+      "step": 19780
+    },
+    {
+      "epoch": 0.5411465525491859,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.601464177017356e-06,
+      "loss": 0.8411,
+      "step": 19790
+    },
+    {
+      "epoch": 0.5414199969921111,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.598722272490472e-06,
+      "loss": 0.7472,
+      "step": 19800
+    },
+    {
+      "epoch": 0.5416934414350364,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.595980367963588e-06,
+      "loss": 0.639,
+      "step": 19810
+    },
+    {
+      "epoch": 0.5419668858779617,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.593238463436704e-06,
+      "loss": 0.7474,
+      "step": 19820
+    },
+    {
+      "epoch": 0.542240330320887,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.590496558909819e-06,
+      "loss": 0.7224,
+      "step": 19830
+    },
+    {
+      "epoch": 0.5425137747638124,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.587754654382935e-06,
+      "loss": 0.7052,
+      "step": 19840
+    },
+    {
+      "epoch": 0.5427872192067377,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.58501274985605e-06,
+      "loss": 0.7569,
+      "step": 19850
+    },
+    {
+      "epoch": 0.543060663649663,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.582270845329166e-06,
+      "loss": 0.7447,
+      "step": 19860
+    },
+    {
+      "epoch": 0.5433341080925883,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.579528940802281e-06,
+      "loss": 0.6715,
+      "step": 19870
+    },
+    {
+      "epoch": 0.5436075525355136,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.5767870362753975e-06,
+      "loss": 0.6665,
+      "step": 19880
+    },
+    {
+      "epoch": 0.543880996978439,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.574045131748513e-06,
+      "loss": 0.6988,
+      "step": 19890
+    },
+    {
+      "epoch": 0.5441544414213643,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.571303227221628e-06,
+      "loss": 0.6768,
+      "step": 19900
+    },
+    {
+      "epoch": 0.5444278858642895,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.568561322694744e-06,
+      "loss": 0.649,
+      "step": 19910
+    },
+    {
+      "epoch": 0.5447013303072148,
+      "grad_norm": 3.484375,
+      "learning_rate": 4.56581941816786e-06,
+      "loss": 0.6657,
+      "step": 19920
+    },
+    {
+      "epoch": 0.5449747747501401,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.563077513640976e-06,
+      "loss": 0.6554,
+      "step": 19930
+    },
+    {
+      "epoch": 0.5452482191930654,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.560335609114091e-06,
+      "loss": 0.6416,
+      "step": 19940
+    },
+    {
+      "epoch": 0.5455216636359908,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.557593704587207e-06,
+      "loss": 0.6985,
+      "step": 19950
+    },
+    {
+      "epoch": 0.5457951080789161,
+      "grad_norm": 3.34375,
+      "learning_rate": 4.554851800060322e-06,
+      "loss": 0.7019,
+      "step": 19960
+    },
+    {
+      "epoch": 0.5460685525218414,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.5521098955334385e-06,
+      "loss": 0.6751,
+      "step": 19970
+    },
+    {
+      "epoch": 0.5463419969647667,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.549367991006554e-06,
+      "loss": 0.65,
+      "step": 19980
+    },
+    {
+      "epoch": 0.546615441407692,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.5466260864796695e-06,
+      "loss": 0.591,
+      "step": 19990
+    },
+    {
+      "epoch": 0.5468888858506173,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.543884181952785e-06,
+      "loss": 0.6397,
+      "step": 20000
+    },
+    {
+      "epoch": 0.5471623302935427,
+      "grad_norm": 3.03125,
+      "learning_rate": 4.5411422774259e-06,
+      "loss": 0.7258,
+      "step": 20010
+    },
+    {
+      "epoch": 0.5474357747364679,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.538400372899016e-06,
+      "loss": 0.7103,
+      "step": 20020
+    },
+    {
+      "epoch": 0.5477092191793932,
+      "grad_norm": 2.609375,
+      "learning_rate": 4.535658468372131e-06,
+      "loss": 0.7195,
+      "step": 20030
+    },
+    {
+      "epoch": 0.5479826636223185,
+      "grad_norm": 2.5,
+      "learning_rate": 4.532916563845247e-06,
+      "loss": 0.6332,
+      "step": 20040
+    },
+    {
+      "epoch": 0.5482561080652438,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.530174659318362e-06,
+      "loss": 0.7484,
+      "step": 20050
+    },
+    {
+      "epoch": 0.5485295525081691,
+      "grad_norm": 2.328125,
+      "learning_rate": 4.527432754791479e-06,
+      "loss": 0.6646,
+      "step": 20060
+    },
+    {
+      "epoch": 0.5488029969510945,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.524690850264594e-06,
+      "loss": 0.6387,
+      "step": 20070
+    },
+    {
+      "epoch": 0.5490764413940198,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.52194894573771e-06,
+      "loss": 0.7343,
+      "step": 20080
+    },
+    {
+      "epoch": 0.5493498858369451,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.519207041210825e-06,
+      "loss": 0.7343,
+      "step": 20090
+    },
+    {
+      "epoch": 0.5496233302798704,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.5164651366839414e-06,
+      "loss": 0.7184,
+      "step": 20100
+    },
+    {
+      "epoch": 0.5498967747227957,
+      "grad_norm": 2.5,
+      "learning_rate": 4.513723232157057e-06,
+      "loss": 0.6625,
+      "step": 20110
+    },
+    {
+      "epoch": 0.550170219165721,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.510981327630172e-06,
+      "loss": 0.6399,
+      "step": 20120
+    },
+    {
+      "epoch": 0.5504436636086463,
+      "grad_norm": 2.625,
+      "learning_rate": 4.508239423103288e-06,
+      "loss": 0.7136,
+      "step": 20130
+    },
+    {
+      "epoch": 0.5507171080515716,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.505497518576403e-06,
+      "loss": 0.7013,
+      "step": 20140
+    },
+    {
+      "epoch": 0.5509905524944969,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.50275561404952e-06,
+      "loss": 0.655,
+      "step": 20150
+    },
+    {
+      "epoch": 0.5512639969374222,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.500013709522635e-06,
+      "loss": 0.7694,
+      "step": 20160
+    },
+    {
+      "epoch": 0.5515374413803475,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.497271804995751e-06,
+      "loss": 0.7054,
+      "step": 20170
+    },
+    {
+      "epoch": 0.5518108858232729,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.494529900468866e-06,
+      "loss": 0.6754,
+      "step": 20180
+    },
+    {
+      "epoch": 0.5520843302661982,
+      "grad_norm": 2.375,
+      "learning_rate": 4.491787995941982e-06,
+      "loss": 0.7457,
+      "step": 20190
+    },
+    {
+      "epoch": 0.5523577747091235,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.489046091415097e-06,
+      "loss": 0.6481,
+      "step": 20200
+    },
+    {
+      "epoch": 0.5526312191520488,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.486304186888213e-06,
+      "loss": 0.6407,
+      "step": 20210
+    },
+    {
+      "epoch": 0.5529046635949741,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.483562282361328e-06,
+      "loss": 0.6862,
+      "step": 20220
+    },
+    {
+      "epoch": 0.5531781080378994,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.4808203778344436e-06,
+      "loss": 0.6917,
+      "step": 20230
+    },
+    {
+      "epoch": 0.5534515524808247,
+      "grad_norm": 2.75,
+      "learning_rate": 4.47807847330756e-06,
+      "loss": 0.6999,
+      "step": 20240
+    },
+    {
+      "epoch": 0.55372499692375,
+      "grad_norm": 2.203125,
+      "learning_rate": 4.475336568780675e-06,
+      "loss": 0.6537,
+      "step": 20250
+    },
+    {
+      "epoch": 0.5539984413666753,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.472594664253791e-06,
+      "loss": 0.659,
+      "step": 20260
+    },
+    {
+      "epoch": 0.5542718858096006,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.469852759726906e-06,
+      "loss": 0.6239,
+      "step": 20270
+    },
+    {
+      "epoch": 0.554545330252526,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.467110855200023e-06,
+      "loss": 0.6467,
+      "step": 20280
+    },
+    {
+      "epoch": 0.5548187746954513,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.464368950673138e-06,
+      "loss": 0.6364,
+      "step": 20290
+    },
+    {
+      "epoch": 0.5550922191383766,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.461627046146254e-06,
+      "loss": 0.7023,
+      "step": 20300
+    },
+    {
+      "epoch": 0.5553656635813019,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.458885141619369e-06,
+      "loss": 0.7158,
+      "step": 20310
+    },
+    {
+      "epoch": 0.5556391080242272,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.456143237092485e-06,
+      "loss": 0.6542,
+      "step": 20320
+    },
+    {
+      "epoch": 0.5559125524671524,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.453401332565601e-06,
+      "loss": 0.7751,
+      "step": 20330
+    },
+    {
+      "epoch": 0.5561859969100778,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.450659428038716e-06,
+      "loss": 0.7021,
+      "step": 20340
+    },
+    {
+      "epoch": 0.5564594413530031,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.447917523511832e-06,
+      "loss": 0.7211,
+      "step": 20350
+    },
+    {
+      "epoch": 0.5567328857959284,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.445175618984947e-06,
+      "loss": 0.6943,
+      "step": 20360
+    },
+    {
+      "epoch": 0.5570063302388537,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.442433714458063e-06,
+      "loss": 0.748,
+      "step": 20370
+    },
+    {
+      "epoch": 0.557279774681779,
+      "grad_norm": 2.25,
+      "learning_rate": 4.439691809931178e-06,
+      "loss": 0.6461,
+      "step": 20380
+    },
+    {
+      "epoch": 0.5575532191247043,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.436949905404294e-06,
+      "loss": 0.7508,
+      "step": 20390
+    },
+    {
+      "epoch": 0.5578266635676297,
+      "grad_norm": 3.0,
+      "learning_rate": 4.434208000877409e-06,
+      "loss": 0.6852,
+      "step": 20400
+    },
+    {
+      "epoch": 0.558100108010555,
+      "grad_norm": 2.625,
+      "learning_rate": 4.431466096350526e-06,
+      "loss": 0.6415,
+      "step": 20410
+    },
+    {
+      "epoch": 0.5583735524534803,
+      "grad_norm": 3.609375,
+      "learning_rate": 4.428724191823641e-06,
+      "loss": 0.746,
+      "step": 20420
+    },
+    {
+      "epoch": 0.5586469968964056,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.425982287296757e-06,
+      "loss": 0.622,
+      "step": 20430
+    },
+    {
+      "epoch": 0.5589204413393308,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.423240382769872e-06,
+      "loss": 0.7444,
+      "step": 20440
+    },
+    {
+      "epoch": 0.5591938857822562,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.4204984782429875e-06,
+      "loss": 0.7601,
+      "step": 20450
+    },
+    {
+      "epoch": 0.5594673302251815,
+      "grad_norm": 2.28125,
+      "learning_rate": 4.417756573716104e-06,
+      "loss": 0.6773,
+      "step": 20460
+    },
+    {
+      "epoch": 0.5597407746681068,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.415014669189219e-06,
+      "loss": 0.6696,
+      "step": 20470
+    },
+    {
+      "epoch": 0.5600142191110321,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.412272764662335e-06,
+      "loss": 0.7401,
+      "step": 20480
+    },
+    {
+      "epoch": 0.5602876635539574,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.40953086013545e-06,
+      "loss": 0.7083,
+      "step": 20490
+    },
+    {
+      "epoch": 0.5605611079968827,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.406788955608567e-06,
+      "loss": 0.6916,
+      "step": 20500
+    },
+    {
+      "epoch": 0.5608345524398081,
+      "grad_norm": 3.71875,
+      "learning_rate": 4.404047051081682e-06,
+      "loss": 0.7645,
+      "step": 20510
+    },
+    {
+      "epoch": 0.5611079968827334,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.401305146554798e-06,
+      "loss": 0.6941,
+      "step": 20520
+    },
+    {
+      "epoch": 0.5613814413256587,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.398563242027913e-06,
+      "loss": 0.7158,
+      "step": 20530
+    },
+    {
+      "epoch": 0.561654885768584,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.395821337501029e-06,
+      "loss": 0.6711,
+      "step": 20540
+    },
+    {
+      "epoch": 0.5619283302115092,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.393079432974144e-06,
+      "loss": 0.6702,
+      "step": 20550
+    },
+    {
+      "epoch": 0.5622017746544345,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.3903375284472595e-06,
+      "loss": 0.6488,
+      "step": 20560
+    },
+    {
+      "epoch": 0.5624752190973599,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.387595623920375e-06,
+      "loss": 0.6939,
+      "step": 20570
+    },
+    {
+      "epoch": 0.5627486635402852,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.3848537193934905e-06,
+      "loss": 0.6448,
+      "step": 20580
+    },
+    {
+      "epoch": 0.5630221079832105,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.382111814866607e-06,
+      "loss": 0.6943,
+      "step": 20590
+    },
+    {
+      "epoch": 0.5632955524261358,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.379369910339722e-06,
+      "loss": 0.7163,
+      "step": 20600
+    },
+    {
+      "epoch": 0.5635689968690611,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.376628005812838e-06,
+      "loss": 0.6511,
+      "step": 20610
+    },
+    {
+      "epoch": 0.5638424413119865,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.373886101285953e-06,
+      "loss": 0.628,
+      "step": 20620
+    },
+    {
+      "epoch": 0.5641158857549118,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.371144196759069e-06,
+      "loss": 0.6502,
+      "step": 20630
+    },
+    {
+      "epoch": 0.5643893301978371,
+      "grad_norm": 3.234375,
+      "learning_rate": 4.368402292232185e-06,
+      "loss": 0.7693,
+      "step": 20640
+    },
+    {
+      "epoch": 0.5646627746407623,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.365660387705301e-06,
+      "loss": 0.6744,
+      "step": 20650
+    },
+    {
+      "epoch": 0.5649362190836876,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.362918483178416e-06,
+      "loss": 0.6402,
+      "step": 20660
+    },
+    {
+      "epoch": 0.565209663526613,
+      "grad_norm": 2.625,
+      "learning_rate": 4.3601765786515315e-06,
+      "loss": 0.686,
+      "step": 20670
+    },
+    {
+      "epoch": 0.5654831079695383,
+      "grad_norm": 3.0,
+      "learning_rate": 4.357434674124648e-06,
+      "loss": 0.6589,
+      "step": 20680
+    },
+    {
+      "epoch": 0.5657565524124636,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.354692769597763e-06,
+      "loss": 0.6699,
+      "step": 20690
+    },
+    {
+      "epoch": 0.5660299968553889,
+      "grad_norm": 3.140625,
+      "learning_rate": 4.351950865070879e-06,
+      "loss": 0.7462,
+      "step": 20700
+    },
+    {
+      "epoch": 0.5663034412983142,
+      "grad_norm": 3.0625,
+      "learning_rate": 4.349208960543994e-06,
+      "loss": 0.747,
+      "step": 20710
+    },
+    {
+      "epoch": 0.5665768857412395,
+      "grad_norm": 2.171875,
+      "learning_rate": 4.34646705601711e-06,
+      "loss": 0.6914,
+      "step": 20720
+    },
+    {
+      "epoch": 0.5668503301841649,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.343725151490225e-06,
+      "loss": 0.6306,
+      "step": 20730
+    },
+    {
+      "epoch": 0.5671237746270902,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.340983246963341e-06,
+      "loss": 0.6822,
+      "step": 20740
+    },
+    {
+      "epoch": 0.5673972190700155,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.338241342436457e-06,
+      "loss": 0.6707,
+      "step": 20750
+    },
+    {
+      "epoch": 0.5676706635129407,
+      "grad_norm": 2.5,
+      "learning_rate": 4.3354994379095726e-06,
+      "loss": 0.6745,
+      "step": 20760
+    },
+    {
+      "epoch": 0.567944107955866,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.332757533382688e-06,
+      "loss": 0.6819,
+      "step": 20770
+    },
+    {
+      "epoch": 0.5682175523987913,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.3300156288558035e-06,
+      "loss": 0.702,
+      "step": 20780
+    },
+    {
+      "epoch": 0.5684909968417167,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.327273724328919e-06,
+      "loss": 0.6993,
+      "step": 20790
+    },
+    {
+      "epoch": 0.568764441284642,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.3245318198020345e-06,
+      "loss": 0.6789,
+      "step": 20800
+    },
+    {
+      "epoch": 0.5690378857275673,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.32178991527515e-06,
+      "loss": 0.6539,
+      "step": 20810
+    },
+    {
+      "epoch": 0.5693113301704926,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.319048010748266e-06,
+      "loss": 0.7513,
+      "step": 20820
+    },
+    {
+      "epoch": 0.5695847746134179,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.316306106221382e-06,
+      "loss": 0.6603,
+      "step": 20830
+    },
+    {
+      "epoch": 0.5698582190563433,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.313564201694497e-06,
+      "loss": 0.6716,
+      "step": 20840
+    },
+    {
+      "epoch": 0.5701316634992686,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.310822297167613e-06,
+      "loss": 0.6961,
+      "step": 20850
+    },
+    {
+      "epoch": 0.5704051079421938,
+      "grad_norm": 2.421875,
+      "learning_rate": 4.308080392640729e-06,
+      "loss": 0.653,
+      "step": 20860
+    },
+    {
+      "epoch": 0.5706785523851191,
+      "grad_norm": 2.625,
+      "learning_rate": 4.3053384881138446e-06,
+      "loss": 0.7163,
+      "step": 20870
+    },
+    {
+      "epoch": 0.5709519968280444,
+      "grad_norm": 3.15625,
+      "learning_rate": 4.30259658358696e-06,
+      "loss": 0.7386,
+      "step": 20880
+    },
+    {
+      "epoch": 0.5712254412709697,
+      "grad_norm": 2.4375,
+      "learning_rate": 4.2998546790600755e-06,
+      "loss": 0.6396,
+      "step": 20890
+    },
+    {
+      "epoch": 0.5714988857138951,
+      "grad_norm": 2.140625,
+      "learning_rate": 4.297112774533191e-06,
+      "loss": 0.7141,
+      "step": 20900
+    },
+    {
+      "epoch": 0.5717723301568204,
+      "grad_norm": 2.875,
+      "learning_rate": 4.294370870006307e-06,
+      "loss": 0.6805,
+      "step": 20910
+    },
+    {
+      "epoch": 0.5720457745997457,
+      "grad_norm": 2.21875,
+      "learning_rate": 4.291628965479423e-06,
+      "loss": 0.7246,
+      "step": 20920
+    },
+    {
+      "epoch": 0.572319219042671,
+      "grad_norm": 3.0,
+      "learning_rate": 4.288887060952538e-06,
+      "loss": 0.6721,
+      "step": 20930
+    },
+    {
+      "epoch": 0.5725926634855963,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.286145156425654e-06,
+      "loss": 0.6246,
+      "step": 20940
+    },
+    {
+      "epoch": 0.5728661079285217,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.283403251898769e-06,
+      "loss": 0.6247,
+      "step": 20950
+    },
+    {
+      "epoch": 0.573139552371447,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.280661347371885e-06,
+      "loss": 0.7205,
+      "step": 20960
+    },
+    {
+      "epoch": 0.5734129968143722,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.277919442845e-06,
+      "loss": 0.6722,
+      "step": 20970
+    },
+    {
+      "epoch": 0.5736864412572975,
+      "grad_norm": 2.625,
+      "learning_rate": 4.275177538318116e-06,
+      "loss": 0.6769,
+      "step": 20980
+    },
+    {
+      "epoch": 0.5739598857002228,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.272435633791231e-06,
+      "loss": 0.8026,
+      "step": 20990
+    },
+    {
+      "epoch": 0.5742333301431481,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.2696937292643475e-06,
+      "loss": 0.6418,
+      "step": 21000
+    },
+    {
+      "epoch": 0.5745067745860735,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.266951824737463e-06,
+      "loss": 0.7091,
+      "step": 21010
+    },
+    {
+      "epoch": 0.5747802190289988,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.2642099202105785e-06,
+      "loss": 0.7538,
+      "step": 21020
+    },
+    {
+      "epoch": 0.5750536634719241,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.261468015683694e-06,
+      "loss": 0.6391,
+      "step": 21030
+    },
+    {
+      "epoch": 0.5753271079148494,
+      "grad_norm": 2.875,
+      "learning_rate": 4.25872611115681e-06,
+      "loss": 0.7346,
+      "step": 21040
+    },
+    {
+      "epoch": 0.5756005523577747,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.255984206629926e-06,
+      "loss": 0.7234,
+      "step": 21050
+    },
+    {
+      "epoch": 0.5758739968007001,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.253242302103041e-06,
+      "loss": 0.6384,
+      "step": 21060
+    },
+    {
+      "epoch": 0.5761474412436254,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.250500397576157e-06,
+      "loss": 0.7078,
+      "step": 21070
+    },
+    {
+      "epoch": 0.5764208856865506,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.247758493049272e-06,
+      "loss": 0.7438,
+      "step": 21080
+    },
+    {
+      "epoch": 0.5766943301294759,
+      "grad_norm": 2.578125,
+      "learning_rate": 4.2450165885223886e-06,
+      "loss": 0.6692,
+      "step": 21090
+    },
+    {
+      "epoch": 0.5769677745724012,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.242274683995504e-06,
+      "loss": 0.6652,
+      "step": 21100
+    },
+    {
+      "epoch": 0.5772412190153265,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.2395327794686195e-06,
+      "loss": 0.6446,
+      "step": 21110
+    },
+    {
+      "epoch": 0.5775146634582519,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.236790874941735e-06,
+      "loss": 0.6881,
+      "step": 21120
+    },
+    {
+      "epoch": 0.5777881079011772,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.2340489704148505e-06,
+      "loss": 0.6918,
+      "step": 21130
+    },
+    {
+      "epoch": 0.5780615523441025,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.231307065887966e-06,
+      "loss": 0.6761,
+      "step": 21140
+    },
+    {
+      "epoch": 0.5783349967870278,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.2285651613610815e-06,
+      "loss": 0.7023,
+      "step": 21150
+    },
+    {
+      "epoch": 0.5786084412299531,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.225823256834197e-06,
+      "loss": 0.7169,
+      "step": 21160
+    },
+    {
+      "epoch": 0.5788818856728785,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.223081352307312e-06,
+      "loss": 0.7241,
+      "step": 21170
+    },
+    {
+      "epoch": 0.5791553301158037,
+      "grad_norm": 2.84375,
+      "learning_rate": 4.220339447780429e-06,
+      "loss": 0.6779,
+      "step": 21180
+    },
+    {
+      "epoch": 0.579428774558729,
+      "grad_norm": 3.265625,
+      "learning_rate": 4.217597543253544e-06,
+      "loss": 0.6845,
+      "step": 21190
+    },
+    {
+      "epoch": 0.5797022190016543,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.21485563872666e-06,
+      "loss": 0.6401,
+      "step": 21200
+    },
+    {
+      "epoch": 0.5799756634445796,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.212113734199775e-06,
+      "loss": 0.6713,
+      "step": 21210
+    },
+    {
+      "epoch": 0.580249107887505,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.2093718296728915e-06,
+      "loss": 0.6344,
+      "step": 21220
+    },
+    {
+      "epoch": 0.5805225523304303,
+      "grad_norm": 2.5,
+      "learning_rate": 4.206629925146007e-06,
+      "loss": 0.7343,
+      "step": 21230
+    },
+    {
+      "epoch": 0.5807959967733556,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.2038880206191225e-06,
+      "loss": 0.7955,
+      "step": 21240
+    },
+    {
+      "epoch": 0.5810694412162809,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.201146116092238e-06,
+      "loss": 0.7284,
+      "step": 21250
+    },
+    {
+      "epoch": 0.5813428856592062,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.1984042115653534e-06,
+      "loss": 0.7046,
+      "step": 21260
+    },
+    {
+      "epoch": 0.5816163301021315,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.19566230703847e-06,
+      "loss": 0.782,
+      "step": 21270
+    },
+    {
+      "epoch": 0.5818897745450569,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.192920402511585e-06,
+      "loss": 0.7421,
+      "step": 21280
+    },
+    {
+      "epoch": 0.5821632189879821,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.190178497984701e-06,
+      "loss": 0.6807,
+      "step": 21290
+    },
+    {
+      "epoch": 0.5824366634309074,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.187436593457816e-06,
+      "loss": 0.7554,
+      "step": 21300
+    },
+    {
+      "epoch": 0.5827101078738327,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.184694688930932e-06,
+      "loss": 0.605,
+      "step": 21310
+    },
+    {
+      "epoch": 0.582983552316758,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.181952784404047e-06,
+      "loss": 0.7022,
+      "step": 21320
+    },
+    {
+      "epoch": 0.5832569967596833,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.179210879877163e-06,
+      "loss": 0.717,
+      "step": 21330
+    },
+    {
+      "epoch": 0.5835304412026087,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.176468975350278e-06,
+      "loss": 0.737,
+      "step": 21340
+    },
+    {
+      "epoch": 0.583803885645534,
+      "grad_norm": 3.296875,
+      "learning_rate": 4.173727070823394e-06,
+      "loss": 0.674,
+      "step": 21350
+    },
+    {
+      "epoch": 0.5840773300884593,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.17098516629651e-06,
+      "loss": 0.7124,
+      "step": 21360
+    },
+    {
+      "epoch": 0.5843507745313846,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.1682432617696254e-06,
+      "loss": 0.6624,
+      "step": 21370
+    },
+    {
+      "epoch": 0.5846242189743099,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.165501357242741e-06,
+      "loss": 0.6499,
+      "step": 21380
+    },
+    {
+      "epoch": 0.5848976634172353,
+      "grad_norm": 2.390625,
+      "learning_rate": 4.162759452715856e-06,
+      "loss": 0.6917,
+      "step": 21390
+    },
+    {
+      "epoch": 0.5851711078601605,
+      "grad_norm": 2.71875,
+      "learning_rate": 4.160017548188973e-06,
+      "loss": 0.7452,
+      "step": 21400
+    },
+    {
+      "epoch": 0.5854445523030858,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.157275643662088e-06,
+      "loss": 0.6747,
+      "step": 21410
+    },
+    {
+      "epoch": 0.5857179967460111,
+      "grad_norm": 2.5,
+      "learning_rate": 4.154533739135204e-06,
+      "loss": 0.6442,
+      "step": 21420
+    },
+    {
+      "epoch": 0.5859914411889364,
+      "grad_norm": 2.59375,
+      "learning_rate": 4.151791834608319e-06,
+      "loss": 0.7073,
+      "step": 21430
+    },
+    {
+      "epoch": 0.5862648856318617,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.149049930081435e-06,
+      "loss": 0.7086,
+      "step": 21440
+    },
+    {
+      "epoch": 0.5865383300747871,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.146308025554551e-06,
+      "loss": 0.6645,
+      "step": 21450
+    },
+    {
+      "epoch": 0.5868117745177124,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.1435661210276665e-06,
+      "loss": 0.7121,
+      "step": 21460
+    },
+    {
+      "epoch": 0.5870852189606377,
+      "grad_norm": 3.015625,
+      "learning_rate": 4.140824216500782e-06,
+      "loss": 0.7212,
+      "step": 21470
+    },
+    {
+      "epoch": 0.587358663403563,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.1380823119738974e-06,
+      "loss": 0.6601,
+      "step": 21480
+    },
+    {
+      "epoch": 0.5876321078464883,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.135340407447013e-06,
+      "loss": 0.7365,
+      "step": 21490
+    },
+    {
+      "epoch": 0.5879055522894135,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.132598502920128e-06,
+      "loss": 0.5913,
+      "step": 21500
+    },
+    {
+      "epoch": 0.5881789967323389,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.129856598393244e-06,
+      "loss": 0.6874,
+      "step": 21510
+    },
+    {
+      "epoch": 0.5884524411752642,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.127114693866359e-06,
+      "loss": 0.7438,
+      "step": 21520
+    },
+    {
+      "epoch": 0.5887258856181895,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.124372789339475e-06,
+      "loss": 0.6046,
+      "step": 21530
+    },
+    {
+      "epoch": 0.5889993300611148,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.121630884812591e-06,
+      "loss": 0.6267,
+      "step": 21540
+    },
+    {
+      "epoch": 0.5892727745040401,
+      "grad_norm": 3.171875,
+      "learning_rate": 4.118888980285707e-06,
+      "loss": 0.7884,
+      "step": 21550
+    },
+    {
+      "epoch": 0.5895462189469655,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.116147075758822e-06,
+      "loss": 0.6817,
+      "step": 21560
+    },
+    {
+      "epoch": 0.5898196633898908,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.113405171231938e-06,
+      "loss": 0.774,
+      "step": 21570
+    },
+    {
+      "epoch": 0.5900931078328161,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.110663266705054e-06,
+      "loss": 0.8388,
+      "step": 21580
+    },
+    {
+      "epoch": 0.5903665522757414,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.1079213621781694e-06,
+      "loss": 0.7202,
+      "step": 21590
+    },
+    {
+      "epoch": 0.5906399967186667,
+      "grad_norm": 3.546875,
+      "learning_rate": 4.105179457651285e-06,
+      "loss": 0.7443,
+      "step": 21600
+    },
+    {
+      "epoch": 0.590913441161592,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.1024375531244e-06,
+      "loss": 0.6917,
+      "step": 21610
+    },
+    {
+      "epoch": 0.5911868856045173,
+      "grad_norm": 2.53125,
+      "learning_rate": 4.099695648597516e-06,
+      "loss": 0.6694,
+      "step": 21620
+    },
+    {
+      "epoch": 0.5914603300474426,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.096953744070632e-06,
+      "loss": 0.735,
+      "step": 21630
+    },
+    {
+      "epoch": 0.5917337744903679,
+      "grad_norm": 2.40625,
+      "learning_rate": 4.094211839543748e-06,
+      "loss": 0.681,
+      "step": 21640
+    },
+    {
+      "epoch": 0.5920072189332932,
+      "grad_norm": 3.125,
+      "learning_rate": 4.091469935016863e-06,
+      "loss": 0.6902,
+      "step": 21650
+    },
+    {
+      "epoch": 0.5922806633762185,
+      "grad_norm": 3.1875,
+      "learning_rate": 4.088728030489979e-06,
+      "loss": 0.6081,
+      "step": 21660
+    },
+    {
+      "epoch": 0.5925541078191439,
+      "grad_norm": 2.890625,
+      "learning_rate": 4.085986125963094e-06,
+      "loss": 0.7042,
+      "step": 21670
+    },
+    {
+      "epoch": 0.5928275522620692,
+      "grad_norm": 2.515625,
+      "learning_rate": 4.08324422143621e-06,
+      "loss": 0.7268,
+      "step": 21680
+    },
+    {
+      "epoch": 0.5931009967049945,
+      "grad_norm": 2.640625,
+      "learning_rate": 4.080502316909325e-06,
+      "loss": 0.7241,
+      "step": 21690
+    },
+    {
+      "epoch": 0.5933744411479198,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.0777604123824414e-06,
+      "loss": 0.7943,
+      "step": 21700
+    },
+    {
+      "epoch": 0.593647885590845,
+      "grad_norm": 3.65625,
+      "learning_rate": 4.075018507855557e-06,
+      "loss": 0.7843,
+      "step": 21710
+    },
+    {
+      "epoch": 0.5939213300337703,
+      "grad_norm": 3.375,
+      "learning_rate": 4.072276603328672e-06,
+      "loss": 0.7699,
+      "step": 21720
+    },
+    {
+      "epoch": 0.5941947744766957,
+      "grad_norm": 2.734375,
+      "learning_rate": 4.069534698801788e-06,
+      "loss": 0.6464,
+      "step": 21730
+    },
+    {
+      "epoch": 0.594468218919621,
+      "grad_norm": 2.546875,
+      "learning_rate": 4.066792794274903e-06,
+      "loss": 0.6841,
+      "step": 21740
+    },
+    {
+      "epoch": 0.5947416633625463,
+      "grad_norm": 2.96875,
+      "learning_rate": 4.064050889748019e-06,
+      "loss": 0.7134,
+      "step": 21750
+    },
+    {
+      "epoch": 0.5950151078054716,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.061308985221135e-06,
+      "loss": 0.7008,
+      "step": 21760
+    },
+    {
+      "epoch": 0.5952885522483969,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.058567080694251e-06,
+      "loss": 0.7227,
+      "step": 21770
+    },
+    {
+      "epoch": 0.5955619966913223,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.055825176167366e-06,
+      "loss": 0.6985,
+      "step": 21780
+    },
+    {
+      "epoch": 0.5958354411342476,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.053083271640482e-06,
+      "loss": 0.7251,
+      "step": 21790
+    },
+    {
+      "epoch": 0.5961088855771729,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.050341367113598e-06,
+      "loss": 0.6982,
+      "step": 21800
+    },
+    {
+      "epoch": 0.5963823300200982,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.0475994625867134e-06,
+      "loss": 0.7228,
+      "step": 21810
+    },
+    {
+      "epoch": 0.5966557744630234,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.044857558059829e-06,
+      "loss": 0.7625,
+      "step": 21820
+    },
+    {
+      "epoch": 0.5969292189059487,
+      "grad_norm": 3.25,
+      "learning_rate": 4.042115653532944e-06,
+      "loss": 0.6867,
+      "step": 21830
+    },
+    {
+      "epoch": 0.5972026633488741,
+      "grad_norm": 2.875,
+      "learning_rate": 4.03937374900606e-06,
+      "loss": 0.7144,
+      "step": 21840
+    },
+    {
+      "epoch": 0.5974761077917994,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.036631844479175e-06,
+      "loss": 0.6852,
+      "step": 21850
+    },
+    {
+      "epoch": 0.5977495522347247,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.033889939952292e-06,
+      "loss": 0.7112,
+      "step": 21860
+    },
+    {
+      "epoch": 0.59802299667765,
+      "grad_norm": 2.90625,
+      "learning_rate": 4.031148035425407e-06,
+      "loss": 0.7088,
+      "step": 21870
+    },
+    {
+      "epoch": 0.5982964411205753,
+      "grad_norm": 3.0,
+      "learning_rate": 4.028406130898523e-06,
+      "loss": 0.7043,
+      "step": 21880
+    },
+    {
+      "epoch": 0.5985698855635007,
+      "grad_norm": 2.625,
+      "learning_rate": 4.025664226371638e-06,
+      "loss": 0.6846,
+      "step": 21890
+    },
+    {
+      "epoch": 0.598843330006426,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.022922321844754e-06,
+      "loss": 0.7059,
+      "step": 21900
+    },
+    {
+      "epoch": 0.5991167744493513,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.020180417317869e-06,
+      "loss": 0.6306,
+      "step": 21910
+    },
+    {
+      "epoch": 0.5993902188922766,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.0174385127909846e-06,
+      "loss": 0.7424,
+      "step": 21920
+    },
+    {
+      "epoch": 0.5996636633352018,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.0146966082641e-06,
+      "loss": 0.6839,
+      "step": 21930
+    },
+    {
+      "epoch": 0.5999371077781271,
+      "grad_norm": 2.671875,
+      "learning_rate": 4.011954703737216e-06,
+      "loss": 0.6605,
+      "step": 21940
+    },
+    {
+      "epoch": 0.6002105522210525,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.009212799210332e-06,
+      "loss": 0.6775,
+      "step": 21950
+    },
+    {
+      "epoch": 0.6004839966639778,
+      "grad_norm": 2.859375,
+      "learning_rate": 4.006470894683447e-06,
+      "loss": 0.7025,
+      "step": 21960
+    },
+    {
+      "epoch": 0.6007574411069031,
+      "grad_norm": 2.625,
+      "learning_rate": 4.003728990156563e-06,
+      "loss": 0.7513,
+      "step": 21970
+    },
+    {
+      "epoch": 0.6010308855498284,
+      "grad_norm": 2.796875,
+      "learning_rate": 4.000987085629679e-06,
+      "loss": 0.7388,
+      "step": 21980
+    },
+    {
+      "epoch": 0.6013043299927537,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.998245181102795e-06,
+      "loss": 0.6949,
+      "step": 21990
+    },
+    {
+      "epoch": 0.6015777744356791,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.99550327657591e-06,
+      "loss": 0.7143,
+      "step": 22000
+    },
+    {
+      "epoch": 0.6018512188786044,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.992761372049026e-06,
+      "loss": 0.6596,
+      "step": 22010
+    },
+    {
+      "epoch": 0.6021246633215297,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.990019467522141e-06,
+      "loss": 0.7035,
+      "step": 22020
+    },
+    {
+      "epoch": 0.6023981077644549,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.987277562995257e-06,
+      "loss": 0.7225,
+      "step": 22030
+    },
+    {
+      "epoch": 0.6026715522073802,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.984535658468373e-06,
+      "loss": 0.6985,
+      "step": 22040
+    },
+    {
+      "epoch": 0.6029449966503055,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.981793753941488e-06,
+      "loss": 0.6738,
+      "step": 22050
+    },
+    {
+      "epoch": 0.6032184410932309,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.979051849414604e-06,
+      "loss": 0.6349,
+      "step": 22060
+    },
+    {
+      "epoch": 0.6034918855361562,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.976309944887719e-06,
+      "loss": 0.7237,
+      "step": 22070
+    },
+    {
+      "epoch": 0.6037653299790815,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.973568040360835e-06,
+      "loss": 0.6546,
+      "step": 22080
+    },
+    {
+      "epoch": 0.6040387744220068,
+      "grad_norm": 2.625,
+      "learning_rate": 3.97082613583395e-06,
+      "loss": 0.7055,
+      "step": 22090
+    },
+    {
+      "epoch": 0.6043122188649321,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.968084231307066e-06,
+      "loss": 0.6938,
+      "step": 22100
+    },
+    {
+      "epoch": 0.6045856633078575,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.965342326780181e-06,
+      "loss": 0.7283,
+      "step": 22110
+    },
+    {
+      "epoch": 0.6048591077507828,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.962600422253298e-06,
+      "loss": 0.7663,
+      "step": 22120
+    },
+    {
+      "epoch": 0.6051325521937081,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.959858517726413e-06,
+      "loss": 0.6901,
+      "step": 22130
+    },
+    {
+      "epoch": 0.6054059966366333,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.9571166131995286e-06,
+      "loss": 0.6238,
+      "step": 22140
+    },
+    {
+      "epoch": 0.6056794410795586,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.954374708672644e-06,
+      "loss": 0.6461,
+      "step": 22150
+    },
+    {
+      "epoch": 0.6059528855224839,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.95163280414576e-06,
+      "loss": 0.626,
+      "step": 22160
+    },
+    {
+      "epoch": 0.6062263299654093,
+      "grad_norm": 2.875,
+      "learning_rate": 3.948890899618876e-06,
+      "loss": 0.7233,
+      "step": 22170
+    },
+    {
+      "epoch": 0.6064997744083346,
+      "grad_norm": 2.875,
+      "learning_rate": 3.946148995091991e-06,
+      "loss": 0.6262,
+      "step": 22180
+    },
+    {
+      "epoch": 0.6067732188512599,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.943407090565107e-06,
+      "loss": 0.7057,
+      "step": 22190
+    },
+    {
+      "epoch": 0.6070466632941852,
+      "grad_norm": 3.125,
+      "learning_rate": 3.940665186038222e-06,
+      "loss": 0.7417,
+      "step": 22200
+    },
+    {
+      "epoch": 0.6073201077371105,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.937923281511339e-06,
+      "loss": 0.6634,
+      "step": 22210
+    },
+    {
+      "epoch": 0.6075935521800359,
+      "grad_norm": 2.875,
+      "learning_rate": 3.935181376984454e-06,
+      "loss": 0.7111,
+      "step": 22220
+    },
+    {
+      "epoch": 0.6078669966229612,
+      "grad_norm": 3.125,
+      "learning_rate": 3.93243947245757e-06,
+      "loss": 0.7821,
+      "step": 22230
+    },
+    {
+      "epoch": 0.6081404410658864,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.929697567930685e-06,
+      "loss": 0.645,
+      "step": 22240
+    },
+    {
+      "epoch": 0.6084138855088117,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.9269556634038006e-06,
+      "loss": 0.6601,
+      "step": 22250
+    },
+    {
+      "epoch": 0.608687329951737,
+      "grad_norm": 3.0,
+      "learning_rate": 3.924213758876916e-06,
+      "loss": 0.6969,
+      "step": 22260
+    },
+    {
+      "epoch": 0.6089607743946623,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.9214718543500315e-06,
+      "loss": 0.7205,
+      "step": 22270
+    },
+    {
+      "epoch": 0.6092342188375877,
+      "grad_norm": 3.25,
+      "learning_rate": 3.918729949823147e-06,
+      "loss": 0.6232,
+      "step": 22280
+    },
+    {
+      "epoch": 0.609507663280513,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.9159880452962625e-06,
+      "loss": 0.7378,
+      "step": 22290
+    },
+    {
+      "epoch": 0.6097811077234383,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.913246140769379e-06,
+      "loss": 0.6416,
+      "step": 22300
+    },
+    {
+      "epoch": 0.6100545521663636,
+      "grad_norm": 2.625,
+      "learning_rate": 3.910504236242494e-06,
+      "loss": 0.6597,
+      "step": 22310
+    },
+    {
+      "epoch": 0.6103279966092889,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.90776233171561e-06,
+      "loss": 0.699,
+      "step": 22320
+    },
+    {
+      "epoch": 0.6106014410522143,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.905020427188725e-06,
+      "loss": 0.6792,
+      "step": 22330
+    },
+    {
+      "epoch": 0.6108748854951396,
+      "grad_norm": 2.75,
+      "learning_rate": 3.902278522661842e-06,
+      "loss": 0.7344,
+      "step": 22340
+    },
+    {
+      "epoch": 0.6111483299380648,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.899536618134957e-06,
+      "loss": 0.6968,
+      "step": 22350
+    },
+    {
+      "epoch": 0.6114217743809901,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.8967947136080726e-06,
+      "loss": 0.6598,
+      "step": 22360
+    },
+    {
+      "epoch": 0.6116952188239154,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.894052809081188e-06,
+      "loss": 0.6956,
+      "step": 22370
+    },
+    {
+      "epoch": 0.6119686632668407,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.8913109045543035e-06,
+      "loss": 0.6943,
+      "step": 22380
+    },
+    {
+      "epoch": 0.6122421077097661,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.88856900002742e-06,
+      "loss": 0.658,
+      "step": 22390
+    },
+    {
+      "epoch": 0.6125155521526914,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.885827095500535e-06,
+      "loss": 0.6777,
+      "step": 22400
+    },
+    {
+      "epoch": 0.6127889965956167,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.883085190973651e-06,
+      "loss": 0.6925,
+      "step": 22410
+    },
+    {
+      "epoch": 0.613062441038542,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.880343286446766e-06,
+      "loss": 0.6553,
+      "step": 22420
+    },
+    {
+      "epoch": 0.6133358854814673,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.877601381919882e-06,
+      "loss": 0.7011,
+      "step": 22430
+    },
+    {
+      "epoch": 0.6136093299243927,
+      "grad_norm": 2.625,
+      "learning_rate": 3.874859477392997e-06,
+      "loss": 0.688,
+      "step": 22440
+    },
+    {
+      "epoch": 0.613882774367318,
+      "grad_norm": 2.0625,
+      "learning_rate": 3.872117572866113e-06,
+      "loss": 0.618,
+      "step": 22450
+    },
+    {
+      "epoch": 0.6141562188102432,
+      "grad_norm": 2.875,
+      "learning_rate": 3.869375668339228e-06,
+      "loss": 0.6841,
+      "step": 22460
+    },
+    {
+      "epoch": 0.6144296632531685,
+      "grad_norm": 2.375,
+      "learning_rate": 3.866633763812344e-06,
+      "loss": 0.6258,
+      "step": 22470
+    },
+    {
+      "epoch": 0.6147031076960938,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.86389185928546e-06,
+      "loss": 0.6363,
+      "step": 22480
+    },
+    {
+      "epoch": 0.6149765521390191,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.8611499547585755e-06,
+      "loss": 0.7634,
+      "step": 22490
+    },
+    {
+      "epoch": 0.6152499965819445,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.858408050231691e-06,
+      "loss": 0.7057,
+      "step": 22500
+    },
+    {
+      "epoch": 0.6155234410248698,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.8556661457048065e-06,
+      "loss": 0.6706,
+      "step": 22510
+    },
+    {
+      "epoch": 0.6157968854677951,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.852924241177923e-06,
+      "loss": 0.6494,
+      "step": 22520
+    },
+    {
+      "epoch": 0.6160703299107204,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.850182336651038e-06,
+      "loss": 0.7279,
+      "step": 22530
+    },
+    {
+      "epoch": 0.6163437743536457,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.847440432124154e-06,
+      "loss": 0.669,
+      "step": 22540
+    },
+    {
+      "epoch": 0.616617218796571,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.844698527597269e-06,
+      "loss": 0.6135,
+      "step": 22550
+    },
+    {
+      "epoch": 0.6168906632394963,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.841956623070385e-06,
+      "loss": 0.719,
+      "step": 22560
+    },
+    {
+      "epoch": 0.6171641076824216,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.839214718543501e-06,
+      "loss": 0.6439,
+      "step": 22570
+    },
+    {
+      "epoch": 0.6174375521253469,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.8364728140166165e-06,
+      "loss": 0.6999,
+      "step": 22580
+    },
+    {
+      "epoch": 0.6177109965682722,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.833730909489732e-06,
+      "loss": 0.6648,
+      "step": 22590
+    },
+    {
+      "epoch": 0.6179844410111975,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.8309890049628475e-06,
+      "loss": 0.6728,
+      "step": 22600
+    },
+    {
+      "epoch": 0.6182578854541229,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.828247100435963e-06,
+      "loss": 0.6993,
+      "step": 22610
+    },
+    {
+      "epoch": 0.6185313298970482,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.8255051959090785e-06,
+      "loss": 0.6778,
+      "step": 22620
+    },
+    {
+      "epoch": 0.6188047743399735,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.822763291382194e-06,
+      "loss": 0.6924,
+      "step": 22630
+    },
+    {
+      "epoch": 0.6190782187828988,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.82002138685531e-06,
+      "loss": 0.7111,
+      "step": 22640
+    },
+    {
+      "epoch": 0.6193516632258241,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.817279482328426e-06,
+      "loss": 0.6996,
+      "step": 22650
+    },
+    {
+      "epoch": 0.6196251076687495,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.8145375778015413e-06,
+      "loss": 0.6589,
+      "step": 22660
+    },
+    {
+      "epoch": 0.6198985521116747,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.8117956732746567e-06,
+      "loss": 0.6479,
+      "step": 22670
+    },
+    {
+      "epoch": 0.6201719965546,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.8090537687477726e-06,
+      "loss": 0.7018,
+      "step": 22680
+    },
+    {
+      "epoch": 0.6204454409975253,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.806311864220888e-06,
+      "loss": 0.6675,
+      "step": 22690
+    },
+    {
+      "epoch": 0.6207188854404506,
+      "grad_norm": 2.625,
+      "learning_rate": 3.803569959694004e-06,
+      "loss": 0.7075,
+      "step": 22700
+    },
+    {
+      "epoch": 0.6209923298833759,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.8008280551671195e-06,
+      "loss": 0.6355,
+      "step": 22710
+    },
+    {
+      "epoch": 0.6212657743263013,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.798086150640235e-06,
+      "loss": 0.7422,
+      "step": 22720
+    },
+    {
+      "epoch": 0.6215392187692266,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.7953442461133505e-06,
+      "loss": 0.6899,
+      "step": 22730
+    },
+    {
+      "epoch": 0.6218126632121519,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.792602341586466e-06,
+      "loss": 0.6843,
+      "step": 22740
+    },
+    {
+      "epoch": 0.6220861076550772,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.789860437059582e-06,
+      "loss": 0.7052,
+      "step": 22750
+    },
+    {
+      "epoch": 0.6223595520980025,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.7871185325326978e-06,
+      "loss": 0.6484,
+      "step": 22760
+    },
+    {
+      "epoch": 0.6226329965409279,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.7843766280058132e-06,
+      "loss": 0.6369,
+      "step": 22770
+    },
+    {
+      "epoch": 0.6229064409838531,
+      "grad_norm": 2.28125,
+      "learning_rate": 3.7816347234789287e-06,
+      "loss": 0.6478,
+      "step": 22780
+    },
+    {
+      "epoch": 0.6231798854267784,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.7788928189520446e-06,
+      "loss": 0.6345,
+      "step": 22790
+    },
+    {
+      "epoch": 0.6234533298697037,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.77615091442516e-06,
+      "loss": 0.6116,
+      "step": 22800
+    },
+    {
+      "epoch": 0.623726774312629,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.7734090098982756e-06,
+      "loss": 0.7744,
+      "step": 22810
+    },
+    {
+      "epoch": 0.6240002187555543,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.770667105371391e-06,
+      "loss": 0.6581,
+      "step": 22820
+    },
+    {
+      "epoch": 0.6242736631984797,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.7679252008445066e-06,
+      "loss": 0.6244,
+      "step": 22830
+    },
+    {
+      "epoch": 0.624547107641405,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.765183296317623e-06,
+      "loss": 0.6519,
+      "step": 22840
+    },
+    {
+      "epoch": 0.6248205520843303,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.7624413917907384e-06,
+      "loss": 0.6796,
+      "step": 22850
+    },
+    {
+      "epoch": 0.6250939965272556,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.759699487263854e-06,
+      "loss": 0.6712,
+      "step": 22860
+    },
+    {
+      "epoch": 0.6253674409701809,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.7569575827369693e-06,
+      "loss": 0.7151,
+      "step": 22870
+    },
+    {
+      "epoch": 0.6256408854131061,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.7542156782100852e-06,
+      "loss": 0.7102,
+      "step": 22880
+    },
+    {
+      "epoch": 0.6259143298560315,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.7514737736832007e-06,
+      "loss": 0.6646,
+      "step": 22890
+    },
+    {
+      "epoch": 0.6261877742989568,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.748731869156316e-06,
+      "loss": 0.6981,
+      "step": 22900
+    },
+    {
+      "epoch": 0.6264612187418821,
+      "grad_norm": 3.328125,
+      "learning_rate": 3.7459899646294317e-06,
+      "loss": 0.668,
+      "step": 22910
+    },
+    {
+      "epoch": 0.6267346631848074,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.743248060102547e-06,
+      "loss": 0.7357,
+      "step": 22920
+    },
+    {
+      "epoch": 0.6270081076277327,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.7405061555756635e-06,
+      "loss": 0.6612,
+      "step": 22930
+    },
+    {
+      "epoch": 0.627281552070658,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.737764251048779e-06,
+      "loss": 0.6727,
+      "step": 22940
+    },
+    {
+      "epoch": 0.6275549965135834,
+      "grad_norm": 2.625,
+      "learning_rate": 3.7350223465218945e-06,
+      "loss": 0.7183,
+      "step": 22950
+    },
+    {
+      "epoch": 0.6278284409565087,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.73228044199501e-06,
+      "loss": 0.7308,
+      "step": 22960
+    },
+    {
+      "epoch": 0.628101885399434,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.729538537468126e-06,
+      "loss": 0.6869,
+      "step": 22970
+    },
+    {
+      "epoch": 0.6283753298423593,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.7267966329412413e-06,
+      "loss": 0.7522,
+      "step": 22980
+    },
+    {
+      "epoch": 0.6286487742852845,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.724054728414357e-06,
+      "loss": 0.6818,
+      "step": 22990
+    },
+    {
+      "epoch": 0.6289222187282099,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.7213128238874723e-06,
+      "loss": 0.6618,
+      "step": 23000
+    },
+    {
+      "epoch": 0.6291956631711352,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.7185709193605878e-06,
+      "loss": 0.7353,
+      "step": 23010
+    },
+    {
+      "epoch": 0.6294691076140605,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.715829014833704e-06,
+      "loss": 0.6523,
+      "step": 23020
+    },
+    {
+      "epoch": 0.6297425520569858,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.7130871103068196e-06,
+      "loss": 0.6072,
+      "step": 23030
+    },
+    {
+      "epoch": 0.6300159964999111,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.710345205779935e-06,
+      "loss": 0.7253,
+      "step": 23040
+    },
+    {
+      "epoch": 0.6302894409428365,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.7076033012530506e-06,
+      "loss": 0.7064,
+      "step": 23050
+    },
+    {
+      "epoch": 0.6305628853857618,
+      "grad_norm": 3.0,
+      "learning_rate": 3.7048613967261665e-06,
+      "loss": 0.7274,
+      "step": 23060
+    },
+    {
+      "epoch": 0.6308363298286871,
+      "grad_norm": 2.625,
+      "learning_rate": 3.702119492199282e-06,
+      "loss": 0.796,
+      "step": 23070
+    },
+    {
+      "epoch": 0.6311097742716124,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.6993775876723974e-06,
+      "loss": 0.7455,
+      "step": 23080
+    },
+    {
+      "epoch": 0.6313832187145376,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.696635683145513e-06,
+      "loss": 0.6813,
+      "step": 23090
+    },
+    {
+      "epoch": 0.6316566631574629,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.6938937786186292e-06,
+      "loss": 0.6736,
+      "step": 23100
+    },
+    {
+      "epoch": 0.6319301076003883,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.6911518740917447e-06,
+      "loss": 0.7878,
+      "step": 23110
+    },
+    {
+      "epoch": 0.6322035520433136,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.68840996956486e-06,
+      "loss": 0.6719,
+      "step": 23120
+    },
+    {
+      "epoch": 0.6324769964862389,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.6856680650379757e-06,
+      "loss": 0.6668,
+      "step": 23130
+    },
+    {
+      "epoch": 0.6327504409291642,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.682926160511091e-06,
+      "loss": 0.6749,
+      "step": 23140
+    },
+    {
+      "epoch": 0.6330238853720895,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.680184255984207e-06,
+      "loss": 0.6816,
+      "step": 23150
+    },
+    {
+      "epoch": 0.6332973298150149,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.6774423514573225e-06,
+      "loss": 0.6832,
+      "step": 23160
+    },
+    {
+      "epoch": 0.6335707742579402,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.674700446930438e-06,
+      "loss": 0.6274,
+      "step": 23170
+    },
+    {
+      "epoch": 0.6338442187008655,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.6719585424035535e-06,
+      "loss": 0.7433,
+      "step": 23180
+    },
+    {
+      "epoch": 0.6341176631437908,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.66921663787667e-06,
+      "loss": 0.6965,
+      "step": 23190
+    },
+    {
+      "epoch": 0.634391107586716,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.6664747333497853e-06,
+      "loss": 0.6988,
+      "step": 23200
+    },
+    {
+      "epoch": 0.6346645520296413,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.663732828822901e-06,
+      "loss": 0.7373,
+      "step": 23210
+    },
+    {
+      "epoch": 0.6349379964725667,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.6609909242960163e-06,
+      "loss": 0.7152,
+      "step": 23220
+    },
+    {
+      "epoch": 0.635211440915492,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.6582490197691318e-06,
+      "loss": 0.7013,
+      "step": 23230
+    },
+    {
+      "epoch": 0.6354848853584173,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.6555071152422477e-06,
+      "loss": 0.7327,
+      "step": 23240
+    },
+    {
+      "epoch": 0.6357583298013426,
+      "grad_norm": 3.171875,
+      "learning_rate": 3.652765210715363e-06,
+      "loss": 0.6706,
+      "step": 23250
+    },
+    {
+      "epoch": 0.6360317742442679,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.6500233061884786e-06,
+      "loss": 0.638,
+      "step": 23260
+    },
+    {
+      "epoch": 0.6363052186871933,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.647281401661594e-06,
+      "loss": 0.5764,
+      "step": 23270
+    },
+    {
+      "epoch": 0.6365786631301186,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.6445394971347105e-06,
+      "loss": 0.702,
+      "step": 23280
+    },
+    {
+      "epoch": 0.6368521075730439,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.641797592607826e-06,
+      "loss": 0.664,
+      "step": 23290
+    },
+    {
+      "epoch": 0.6371255520159692,
+      "grad_norm": 3.0,
+      "learning_rate": 3.6390556880809414e-06,
+      "loss": 0.7008,
+      "step": 23300
+    },
+    {
+      "epoch": 0.6373989964588944,
+      "grad_norm": 3.0,
+      "learning_rate": 3.636313783554057e-06,
+      "loss": 0.6567,
+      "step": 23310
+    },
+    {
+      "epoch": 0.6376724409018197,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.6335718790271724e-06,
+      "loss": 0.709,
+      "step": 23320
+    },
+    {
+      "epoch": 0.6379458853447451,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.6308299745002883e-06,
+      "loss": 0.7129,
+      "step": 23330
+    },
+    {
+      "epoch": 0.6382193297876704,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.6280880699734038e-06,
+      "loss": 0.6956,
+      "step": 23340
+    },
+    {
+      "epoch": 0.6384927742305957,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.6253461654465192e-06,
+      "loss": 0.6821,
+      "step": 23350
+    },
+    {
+      "epoch": 0.638766218673521,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.6226042609196347e-06,
+      "loss": 0.658,
+      "step": 23360
+    },
+    {
+      "epoch": 0.6390396631164463,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.619862356392751e-06,
+      "loss": 0.6904,
+      "step": 23370
+    },
+    {
+      "epoch": 0.6393131075593717,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.6171204518658665e-06,
+      "loss": 0.6809,
+      "step": 23380
+    },
+    {
+      "epoch": 0.639586552002297,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.614378547338982e-06,
+      "loss": 0.6679,
+      "step": 23390
+    },
+    {
+      "epoch": 0.6398599964452223,
+      "grad_norm": 2.015625,
+      "learning_rate": 3.6116366428120975e-06,
+      "loss": 0.58,
+      "step": 23400
+    },
+    {
+      "epoch": 0.6401334408881475,
+      "grad_norm": 2.75,
+      "learning_rate": 3.608894738285213e-06,
+      "loss": 0.6336,
+      "step": 23410
+    },
+    {
+      "epoch": 0.6404068853310728,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.606152833758329e-06,
+      "loss": 0.6392,
+      "step": 23420
+    },
+    {
+      "epoch": 0.6406803297739981,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.6034109292314444e-06,
+      "loss": 0.7383,
+      "step": 23430
+    },
+    {
+      "epoch": 0.6409537742169235,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.60066902470456e-06,
+      "loss": 0.7506,
+      "step": 23440
+    },
+    {
+      "epoch": 0.6412272186598488,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.5979271201776753e-06,
+      "loss": 0.7083,
+      "step": 23450
+    },
+    {
+      "epoch": 0.6415006631027741,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.5951852156507917e-06,
+      "loss": 0.6915,
+      "step": 23460
+    },
+    {
+      "epoch": 0.6417741075456994,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.592443311123907e-06,
+      "loss": 0.6901,
+      "step": 23470
+    },
+    {
+      "epoch": 0.6420475519886247,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.5897014065970226e-06,
+      "loss": 0.6709,
+      "step": 23480
+    },
+    {
+      "epoch": 0.64232099643155,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.586959502070138e-06,
+      "loss": 0.7622,
+      "step": 23490
+    },
+    {
+      "epoch": 0.6425944408744754,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.5842175975432536e-06,
+      "loss": 0.7031,
+      "step": 23500
+    },
+    {
+      "epoch": 0.6428678853174007,
+      "grad_norm": 3.421875,
+      "learning_rate": 3.5814756930163695e-06,
+      "loss": 0.8086,
+      "step": 23510
+    },
+    {
+      "epoch": 0.6431413297603259,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.578733788489485e-06,
+      "loss": 0.7047,
+      "step": 23520
+    },
+    {
+      "epoch": 0.6434147742032512,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.5759918839626005e-06,
+      "loss": 0.6328,
+      "step": 23530
+    },
+    {
+      "epoch": 0.6436882186461765,
+      "grad_norm": 3.28125,
+      "learning_rate": 3.5732499794357164e-06,
+      "loss": 0.6807,
+      "step": 23540
+    },
+    {
+      "epoch": 0.6439616630891019,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.5705080749088323e-06,
+      "loss": 0.7675,
+      "step": 23550
+    },
+    {
+      "epoch": 0.6442351075320272,
+      "grad_norm": 2.796875,
+      "learning_rate": 3.5677661703819478e-06,
+      "loss": 0.7683,
+      "step": 23560
+    },
+    {
+      "epoch": 0.6445085519749525,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.5650242658550632e-06,
+      "loss": 0.7118,
+      "step": 23570
+    },
+    {
+      "epoch": 0.6447819964178778,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.5622823613281787e-06,
+      "loss": 0.733,
+      "step": 23580
+    },
+    {
+      "epoch": 0.6450554408608031,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.559540456801294e-06,
+      "loss": 0.7149,
+      "step": 23590
+    },
+    {
+      "epoch": 0.6453288853037285,
+      "grad_norm": 2.375,
+      "learning_rate": 3.55679855227441e-06,
+      "loss": 0.718,
+      "step": 23600
+    },
+    {
+      "epoch": 0.6456023297466538,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.5540566477475256e-06,
+      "loss": 0.6758,
+      "step": 23610
+    },
+    {
+      "epoch": 0.645875774189579,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.5513147432206415e-06,
+      "loss": 0.745,
+      "step": 23620
+    },
+    {
+      "epoch": 0.6461492186325043,
+      "grad_norm": 3.0,
+      "learning_rate": 3.548572838693757e-06,
+      "loss": 0.6931,
+      "step": 23630
+    },
+    {
+      "epoch": 0.6464226630754296,
+      "grad_norm": 3.25,
+      "learning_rate": 3.545830934166873e-06,
+      "loss": 0.7065,
+      "step": 23640
+    },
+    {
+      "epoch": 0.6466961075183549,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.5430890296399884e-06,
+      "loss": 0.7518,
+      "step": 23650
+    },
+    {
+      "epoch": 0.6469695519612803,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.540347125113104e-06,
+      "loss": 0.7066,
+      "step": 23660
+    },
+    {
+      "epoch": 0.6472429964042056,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.5376052205862193e-06,
+      "loss": 0.6887,
+      "step": 23670
+    },
+    {
+      "epoch": 0.6475164408471309,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.534863316059335e-06,
+      "loss": 0.6952,
+      "step": 23680
+    },
+    {
+      "epoch": 0.6477898852900562,
+      "grad_norm": 2.625,
+      "learning_rate": 3.5321214115324507e-06,
+      "loss": 0.6967,
+      "step": 23690
+    },
+    {
+      "epoch": 0.6480633297329815,
+      "grad_norm": 2.625,
+      "learning_rate": 3.5293795070055666e-06,
+      "loss": 0.6667,
+      "step": 23700
+    },
+    {
+      "epoch": 0.6483367741759068,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.526637602478682e-06,
+      "loss": 0.6994,
+      "step": 23710
+    },
+    {
+      "epoch": 0.6486102186188322,
+      "grad_norm": 3.25,
+      "learning_rate": 3.5238956979517976e-06,
+      "loss": 0.7374,
+      "step": 23720
+    },
+    {
+      "epoch": 0.6488836630617574,
+      "grad_norm": 3.234375,
+      "learning_rate": 3.5211537934249135e-06,
+      "loss": 0.6126,
+      "step": 23730
+    },
+    {
+      "epoch": 0.6491571075046827,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.518411888898029e-06,
+      "loss": 0.7083,
+      "step": 23740
+    },
+    {
+      "epoch": 0.649430551947608,
+      "grad_norm": 3.125,
+      "learning_rate": 3.5156699843711445e-06,
+      "loss": 0.7015,
+      "step": 23750
+    },
+    {
+      "epoch": 0.6497039963905333,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.51292807984426e-06,
+      "loss": 0.6931,
+      "step": 23760
+    },
+    {
+      "epoch": 0.6499774408334587,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.5101861753173754e-06,
+      "loss": 0.7422,
+      "step": 23770
+    },
+    {
+      "epoch": 0.650250885276384,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.5074442707904917e-06,
+      "loss": 0.6667,
+      "step": 23780
+    },
+    {
+      "epoch": 0.6505243297193093,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.5047023662636072e-06,
+      "loss": 0.7285,
+      "step": 23790
+    },
+    {
+      "epoch": 0.6507977741622346,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.5019604617367227e-06,
+      "loss": 0.6477,
+      "step": 23800
+    },
+    {
+      "epoch": 0.6510712186051599,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.499218557209838e-06,
+      "loss": 0.7161,
+      "step": 23810
+    },
+    {
+      "epoch": 0.6513446630480852,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.496476652682954e-06,
+      "loss": 0.7416,
+      "step": 23820
+    },
+    {
+      "epoch": 0.6516181074910106,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.4937347481560696e-06,
+      "loss": 0.6478,
+      "step": 23830
+    },
+    {
+      "epoch": 0.6518915519339358,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.490992843629185e-06,
+      "loss": 0.7007,
+      "step": 23840
+    },
+    {
+      "epoch": 0.6521649963768611,
+      "grad_norm": 2.75,
+      "learning_rate": 3.4882509391023005e-06,
+      "loss": 0.7133,
+      "step": 23850
+    },
+    {
+      "epoch": 0.6524384408197864,
+      "grad_norm": 3.390625,
+      "learning_rate": 3.485509034575416e-06,
+      "loss": 0.7781,
+      "step": 23860
+    },
+    {
+      "epoch": 0.6527118852627117,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.4827671300485324e-06,
+      "loss": 0.7424,
+      "step": 23870
+    },
+    {
+      "epoch": 0.652985329705637,
+      "grad_norm": 3.3125,
+      "learning_rate": 3.480025225521648e-06,
+      "loss": 0.7535,
+      "step": 23880
+    },
+    {
+      "epoch": 0.6532587741485624,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.4772833209947633e-06,
+      "loss": 0.6907,
+      "step": 23890
+    },
+    {
+      "epoch": 0.6535322185914877,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.474541416467879e-06,
+      "loss": 0.7119,
+      "step": 23900
+    },
+    {
+      "epoch": 0.653805663034413,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.4717995119409947e-06,
+      "loss": 0.7625,
+      "step": 23910
+    },
+    {
+      "epoch": 0.6540791074773383,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.46905760741411e-06,
+      "loss": 0.7243,
+      "step": 23920
+    },
+    {
+      "epoch": 0.6543525519202636,
+      "grad_norm": 2.75,
+      "learning_rate": 3.4663157028872257e-06,
+      "loss": 0.6832,
+      "step": 23930
+    },
+    {
+      "epoch": 0.6546259963631889,
+      "grad_norm": 3.0,
+      "learning_rate": 3.463573798360341e-06,
+      "loss": 0.73,
+      "step": 23940
+    },
+    {
+      "epoch": 0.6548994408061142,
+      "grad_norm": 2.265625,
+      "learning_rate": 3.4608318938334566e-06,
+      "loss": 0.6125,
+      "step": 23950
+    },
+    {
+      "epoch": 0.6551728852490395,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.458089989306573e-06,
+      "loss": 0.7195,
+      "step": 23960
+    },
+    {
+      "epoch": 0.6554463296919648,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.4553480847796884e-06,
+      "loss": 0.6719,
+      "step": 23970
+    },
+    {
+      "epoch": 0.6557197741348901,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.452606180252804e-06,
+      "loss": 0.7712,
+      "step": 23980
+    },
+    {
+      "epoch": 0.6559932185778155,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.4498642757259194e-06,
+      "loss": 0.7246,
+      "step": 23990
+    },
+    {
+      "epoch": 0.6562666630207408,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.4471223711990353e-06,
+      "loss": 0.734,
+      "step": 24000
+    },
+    {
+      "epoch": 0.6565401074636661,
+      "grad_norm": 2.15625,
+      "learning_rate": 3.444380466672151e-06,
+      "loss": 0.6182,
+      "step": 24010
+    },
+    {
+      "epoch": 0.6568135519065914,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.4416385621452663e-06,
+      "loss": 0.6635,
+      "step": 24020
+    },
+    {
+      "epoch": 0.6570869963495167,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.4388966576183818e-06,
+      "loss": 0.7769,
+      "step": 24030
+    },
+    {
+      "epoch": 0.657360440792442,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.4361547530914972e-06,
+      "loss": 0.7696,
+      "step": 24040
+    },
+    {
+      "epoch": 0.6576338852353673,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.4334128485646136e-06,
+      "loss": 0.6662,
+      "step": 24050
+    },
+    {
+      "epoch": 0.6579073296782926,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.430670944037729e-06,
+      "loss": 0.6823,
+      "step": 24060
+    },
+    {
+      "epoch": 0.6581807741212179,
+      "grad_norm": 2.75,
+      "learning_rate": 3.4279290395108445e-06,
+      "loss": 0.7321,
+      "step": 24070
+    },
+    {
+      "epoch": 0.6584542185641432,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.42518713498396e-06,
+      "loss": 0.6752,
+      "step": 24080
+    },
+    {
+      "epoch": 0.6587276630070685,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.422445230457076e-06,
+      "loss": 0.6701,
+      "step": 24090
+    },
+    {
+      "epoch": 0.6590011074499939,
+      "grad_norm": 3.53125,
+      "learning_rate": 3.4197033259301914e-06,
+      "loss": 0.7158,
+      "step": 24100
+    },
+    {
+      "epoch": 0.6592745518929192,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.416961421403307e-06,
+      "loss": 0.728,
+      "step": 24110
+    },
+    {
+      "epoch": 0.6595479963358445,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.4142195168764224e-06,
+      "loss": 0.7289,
+      "step": 24120
+    },
+    {
+      "epoch": 0.6598214407787698,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.411477612349538e-06,
+      "loss": 0.7723,
+      "step": 24130
+    },
+    {
+      "epoch": 0.6600948852216951,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.408735707822654e-06,
+      "loss": 0.8101,
+      "step": 24140
+    },
+    {
+      "epoch": 0.6603683296646204,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.4059938032957697e-06,
+      "loss": 0.6973,
+      "step": 24150
+    },
+    {
+      "epoch": 0.6606417741075457,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.403251898768885e-06,
+      "loss": 0.6472,
+      "step": 24160
+    },
+    {
+      "epoch": 0.660915218550471,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.4005099942420006e-06,
+      "loss": 0.7256,
+      "step": 24170
+    },
+    {
+      "epoch": 0.6611886629933963,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.3977680897151165e-06,
+      "loss": 0.7473,
+      "step": 24180
+    },
+    {
+      "epoch": 0.6614621074363216,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.395026185188232e-06,
+      "loss": 0.7221,
+      "step": 24190
+    },
+    {
+      "epoch": 0.6617355518792469,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.3922842806613475e-06,
+      "loss": 0.732,
+      "step": 24200
+    },
+    {
+      "epoch": 0.6620089963221722,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.389542376134463e-06,
+      "loss": 0.7411,
+      "step": 24210
+    },
+    {
+      "epoch": 0.6622824407650976,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.3868004716075785e-06,
+      "loss": 0.639,
+      "step": 24220
+    },
+    {
+      "epoch": 0.6625558852080229,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.384058567080695e-06,
+      "loss": 0.6492,
+      "step": 24230
+    },
+    {
+      "epoch": 0.6628293296509482,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.3813166625538103e-06,
+      "loss": 0.6646,
+      "step": 24240
+    },
+    {
+      "epoch": 0.6631027740938735,
+      "grad_norm": 2.5,
+      "learning_rate": 3.3785747580269258e-06,
+      "loss": 0.6614,
+      "step": 24250
+    },
+    {
+      "epoch": 0.6633762185367987,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.3758328535000412e-06,
+      "loss": 0.6843,
+      "step": 24260
+    },
+    {
+      "epoch": 0.663649662979724,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.373090948973157e-06,
+      "loss": 0.6717,
+      "step": 24270
+    },
+    {
+      "epoch": 0.6639231074226494,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.3703490444462726e-06,
+      "loss": 0.7756,
+      "step": 24280
+    },
+    {
+      "epoch": 0.6641965518655747,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.367607139919388e-06,
+      "loss": 0.6996,
+      "step": 24290
+    },
+    {
+      "epoch": 0.6644699963085,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.3648652353925036e-06,
+      "loss": 0.7749,
+      "step": 24300
+    },
+    {
+      "epoch": 0.6647434407514253,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.362123330865619e-06,
+      "loss": 0.6611,
+      "step": 24310
+    },
+    {
+      "epoch": 0.6650168851943506,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.3593814263387354e-06,
+      "loss": 0.734,
+      "step": 24320
+    },
+    {
+      "epoch": 0.665290329637276,
+      "grad_norm": 3.03125,
+      "learning_rate": 3.356639521811851e-06,
+      "loss": 0.678,
+      "step": 24330
+    },
+    {
+      "epoch": 0.6655637740802013,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.3538976172849664e-06,
+      "loss": 0.7496,
+      "step": 24340
+    },
+    {
+      "epoch": 0.6658372185231266,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.351155712758082e-06,
+      "loss": 0.6766,
+      "step": 24350
+    },
+    {
+      "epoch": 0.6661106629660519,
+      "grad_norm": 2.296875,
+      "learning_rate": 3.3484138082311977e-06,
+      "loss": 0.6793,
+      "step": 24360
+    },
+    {
+      "epoch": 0.6663841074089771,
+      "grad_norm": 3.09375,
+      "learning_rate": 3.3456719037043132e-06,
+      "loss": 0.7391,
+      "step": 24370
+    },
+    {
+      "epoch": 0.6666575518519025,
+      "grad_norm": 9.0,
+      "learning_rate": 3.3429299991774287e-06,
+      "loss": 0.6607,
+      "step": 24380
+    },
+    {
+      "epoch": 0.6669309962948278,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.340188094650544e-06,
+      "loss": 0.7463,
+      "step": 24390
+    },
+    {
+      "epoch": 0.6672044407377531,
+      "grad_norm": 3.453125,
+      "learning_rate": 3.33744619012366e-06,
+      "loss": 0.7348,
+      "step": 24400
+    },
+    {
+      "epoch": 0.6674778851806784,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.334704285596776e-06,
+      "loss": 0.7383,
+      "step": 24410
+    },
+    {
+      "epoch": 0.6677513296236037,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.3319623810698915e-06,
+      "loss": 0.7071,
+      "step": 24420
+    },
+    {
+      "epoch": 0.668024774066529,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.329220476543007e-06,
+      "loss": 0.6869,
+      "step": 24430
+    },
+    {
+      "epoch": 0.6682982185094544,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.3264785720161225e-06,
+      "loss": 0.7147,
+      "step": 24440
+    },
+    {
+      "epoch": 0.6685716629523797,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.3237366674892384e-06,
+      "loss": 0.7004,
+      "step": 24450
+    },
+    {
+      "epoch": 0.668845107395305,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.320994762962354e-06,
+      "loss": 0.6808,
+      "step": 24460
+    },
+    {
+      "epoch": 0.6691185518382302,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.3182528584354693e-06,
+      "loss": 0.7618,
+      "step": 24470
+    },
+    {
+      "epoch": 0.6693919962811555,
+      "grad_norm": 2.75,
+      "learning_rate": 3.3155109539085852e-06,
+      "loss": 0.7601,
+      "step": 24480
+    },
+    {
+      "epoch": 0.6696654407240809,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.312769049381701e-06,
+      "loss": 0.7842,
+      "step": 24490
+    },
+    {
+      "epoch": 0.6699388851670062,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.3100271448548166e-06,
+      "loss": 0.6621,
+      "step": 24500
+    },
+    {
+      "epoch": 0.6702123296099315,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.307285240327932e-06,
+      "loss": 0.7139,
+      "step": 24510
+    },
+    {
+      "epoch": 0.6704857740528568,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.3045433358010476e-06,
+      "loss": 0.6101,
+      "step": 24520
+    },
+    {
+      "epoch": 0.6707592184957821,
+      "grad_norm": 2.375,
+      "learning_rate": 3.301801431274163e-06,
+      "loss": 0.6756,
+      "step": 24530
+    },
+    {
+      "epoch": 0.6710326629387074,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.299059526747279e-06,
+      "loss": 0.6592,
+      "step": 24540
+    },
+    {
+      "epoch": 0.6713061073816328,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.2963176222203944e-06,
+      "loss": 0.7061,
+      "step": 24550
+    },
+    {
+      "epoch": 0.6715795518245581,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.2935757176935104e-06,
+      "loss": 0.6514,
+      "step": 24560
+    },
+    {
+      "epoch": 0.6718529962674834,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.290833813166626e-06,
+      "loss": 0.7673,
+      "step": 24570
+    },
+    {
+      "epoch": 0.6721264407104086,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2880919086397417e-06,
+      "loss": 0.6626,
+      "step": 24580
+    },
+    {
+      "epoch": 0.6723998851533339,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.2853500041128572e-06,
+      "loss": 0.6736,
+      "step": 24590
+    },
+    {
+      "epoch": 0.6726733295962593,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.2826080995859727e-06,
+      "loss": 0.7229,
+      "step": 24600
+    },
+    {
+      "epoch": 0.6729467740391846,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.279866195059088e-06,
+      "loss": 0.6412,
+      "step": 24610
+    },
+    {
+      "epoch": 0.6732202184821099,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.2771242905322037e-06,
+      "loss": 0.631,
+      "step": 24620
+    },
+    {
+      "epoch": 0.6734936629250352,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.2743823860053196e-06,
+      "loss": 0.7024,
+      "step": 24630
+    },
+    {
+      "epoch": 0.6737671073679605,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.2716404814784355e-06,
+      "loss": 0.698,
+      "step": 24640
+    },
+    {
+      "epoch": 0.6740405518108858,
+      "grad_norm": 2.75,
+      "learning_rate": 3.268898576951551e-06,
+      "loss": 0.7555,
+      "step": 24650
+    },
+    {
+      "epoch": 0.6743139962538112,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.2661566724246664e-06,
+      "loss": 0.6749,
+      "step": 24660
+    },
+    {
+      "epoch": 0.6745874406967365,
+      "grad_norm": 2.75,
+      "learning_rate": 3.2634147678977823e-06,
+      "loss": 0.68,
+      "step": 24670
+    },
+    {
+      "epoch": 0.6748608851396618,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.260672863370898e-06,
+      "loss": 0.7131,
+      "step": 24680
+    },
+    {
+      "epoch": 0.675134329582587,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.2579309588440133e-06,
+      "loss": 0.7191,
+      "step": 24690
+    },
+    {
+      "epoch": 0.6754077740255123,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.255189054317129e-06,
+      "loss": 0.7465,
+      "step": 24700
+    },
+    {
+      "epoch": 0.6756812184684376,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.2524471497902443e-06,
+      "loss": 0.6861,
+      "step": 24710
+    },
+    {
+      "epoch": 0.675954662911363,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.2497052452633606e-06,
+      "loss": 0.6912,
+      "step": 24720
+    },
+    {
+      "epoch": 0.6762281073542883,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.246963340736476e-06,
+      "loss": 0.7229,
+      "step": 24730
+    },
+    {
+      "epoch": 0.6765015517972136,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.2442214362095916e-06,
+      "loss": 0.7592,
+      "step": 24740
+    },
+    {
+      "epoch": 0.6767749962401389,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.241479531682707e-06,
+      "loss": 0.712,
+      "step": 24750
+    },
+    {
+      "epoch": 0.6770484406830642,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.238737627155823e-06,
+      "loss": 0.6363,
+      "step": 24760
+    },
+    {
+      "epoch": 0.6773218851259896,
+      "grad_norm": 3.21875,
+      "learning_rate": 3.2359957226289384e-06,
+      "loss": 0.7003,
+      "step": 24770
+    },
+    {
+      "epoch": 0.6775953295689149,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.233253818102054e-06,
+      "loss": 0.7144,
+      "step": 24780
+    },
+    {
+      "epoch": 0.6778687740118401,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.2305119135751694e-06,
+      "loss": 0.7053,
+      "step": 24790
+    },
+    {
+      "epoch": 0.6781422184547654,
+      "grad_norm": 2.9375,
+      "learning_rate": 3.227770009048285e-06,
+      "loss": 0.62,
+      "step": 24800
+    },
+    {
+      "epoch": 0.6784156628976907,
+      "grad_norm": 2.34375,
+      "learning_rate": 3.2250281045214012e-06,
+      "loss": 0.6657,
+      "step": 24810
+    },
+    {
+      "epoch": 0.678689107340616,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.2222861999945167e-06,
+      "loss": 0.7047,
+      "step": 24820
+    },
+    {
+      "epoch": 0.6789625517835414,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.219544295467632e-06,
+      "loss": 0.6404,
+      "step": 24830
+    },
+    {
+      "epoch": 0.6792359962264667,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.2168023909407477e-06,
+      "loss": 0.6494,
+      "step": 24840
+    },
+    {
+      "epoch": 0.679509440669392,
+      "grad_norm": 2.875,
+      "learning_rate": 3.2140604864138636e-06,
+      "loss": 0.6778,
+      "step": 24850
+    },
+    {
+      "epoch": 0.6797828851123173,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.211318581886979e-06,
+      "loss": 0.7099,
+      "step": 24860
+    },
+    {
+      "epoch": 0.6800563295552426,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.2085766773600945e-06,
+      "loss": 0.7202,
+      "step": 24870
+    },
+    {
+      "epoch": 0.680329773998168,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.20583477283321e-06,
+      "loss": 0.7683,
+      "step": 24880
+    },
+    {
+      "epoch": 0.6806032184410933,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.2030928683063255e-06,
+      "loss": 0.6545,
+      "step": 24890
+    },
+    {
+      "epoch": 0.6808766628840185,
+      "grad_norm": 2.46875,
+      "learning_rate": 3.200350963779442e-06,
+      "loss": 0.7011,
+      "step": 24900
+    },
+    {
+      "epoch": 0.6811501073269438,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.1976090592525573e-06,
+      "loss": 0.6686,
+      "step": 24910
+    },
+    {
+      "epoch": 0.6814235517698691,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.1948671547256728e-06,
+      "loss": 0.6578,
+      "step": 24920
+    },
+    {
+      "epoch": 0.6816969962127944,
+      "grad_norm": 2.546875,
+      "learning_rate": 3.1921252501987883e-06,
+      "loss": 0.6826,
+      "step": 24930
+    },
+    {
+      "epoch": 0.6819704406557198,
+      "grad_norm": 2.875,
+      "learning_rate": 3.189383345671904e-06,
+      "loss": 0.7305,
+      "step": 24940
+    },
+    {
+      "epoch": 0.6822438850986451,
+      "grad_norm": 3.375,
+      "learning_rate": 3.1866414411450197e-06,
+      "loss": 0.781,
+      "step": 24950
+    },
+    {
+      "epoch": 0.6825173295415704,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.183899536618135e-06,
+      "loss": 0.7408,
+      "step": 24960
+    },
+    {
+      "epoch": 0.6827907739844957,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.1811576320912506e-06,
+      "loss": 0.6196,
+      "step": 24970
+    },
+    {
+      "epoch": 0.683064218427421,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.178415727564366e-06,
+      "loss": 0.7858,
+      "step": 24980
+    },
+    {
+      "epoch": 0.6833376628703464,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.1756738230374824e-06,
+      "loss": 0.7179,
+      "step": 24990
+    },
+    {
+      "epoch": 0.6836111073132716,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.172931918510598e-06,
+      "loss": 0.6825,
+      "step": 25000
+    },
+    {
+      "epoch": 0.6838845517561969,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.1701900139837134e-06,
+      "loss": 0.6931,
+      "step": 25010
+    },
+    {
+      "epoch": 0.6841579961991222,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.167448109456829e-06,
+      "loss": 0.6828,
+      "step": 25020
+    },
+    {
+      "epoch": 0.6844314406420475,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.1647062049299448e-06,
+      "loss": 0.7768,
+      "step": 25030
+    },
+    {
+      "epoch": 0.6847048850849728,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.1619643004030603e-06,
+      "loss": 0.7115,
+      "step": 25040
+    },
+    {
+      "epoch": 0.6849783295278982,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.1592223958761757e-06,
+      "loss": 0.702,
+      "step": 25050
+    },
+    {
+      "epoch": 0.6852517739708235,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.1564804913492912e-06,
+      "loss": 0.6722,
+      "step": 25060
+    },
+    {
+      "epoch": 0.6855252184137488,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.1537385868224067e-06,
+      "loss": 0.7131,
+      "step": 25070
+    },
+    {
+      "epoch": 0.6857986628566741,
+      "grad_norm": 3.125,
+      "learning_rate": 3.150996682295523e-06,
+      "loss": 0.6656,
+      "step": 25080
+    },
+    {
+      "epoch": 0.6860721072995994,
+      "grad_norm": 2.40625,
+      "learning_rate": 3.1482547777686385e-06,
+      "loss": 0.6308,
+      "step": 25090
+    },
+    {
+      "epoch": 0.6863455517425248,
+      "grad_norm": 3.15625,
+      "learning_rate": 3.145512873241754e-06,
+      "loss": 0.6882,
+      "step": 25100
+    },
+    {
+      "epoch": 0.68661899618545,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.1427709687148695e-06,
+      "loss": 0.6682,
+      "step": 25110
+    },
+    {
+      "epoch": 0.6868924406283753,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.1400290641879854e-06,
+      "loss": 0.7355,
+      "step": 25120
+    },
+    {
+      "epoch": 0.6871658850713006,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.137287159661101e-06,
+      "loss": 0.7702,
+      "step": 25130
+    },
+    {
+      "epoch": 0.6874393295142259,
+      "grad_norm": 2.421875,
+      "learning_rate": 3.1345452551342164e-06,
+      "loss": 0.7137,
+      "step": 25140
+    },
+    {
+      "epoch": 0.6877127739571512,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.131803350607332e-06,
+      "loss": 0.5851,
+      "step": 25150
+    },
+    {
+      "epoch": 0.6879862184000766,
+      "grad_norm": 2.6875,
+      "learning_rate": 3.1290614460804473e-06,
+      "loss": 0.6763,
+      "step": 25160
+    },
+    {
+      "epoch": 0.6882596628430019,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.1263195415535636e-06,
+      "loss": 0.6181,
+      "step": 25170
+    },
+    {
+      "epoch": 0.6885331072859272,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.123577637026679e-06,
+      "loss": 0.6899,
+      "step": 25180
+    },
+    {
+      "epoch": 0.6888065517288525,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.1208357324997946e-06,
+      "loss": 0.6408,
+      "step": 25190
+    },
+    {
+      "epoch": 0.6890799961717778,
+      "grad_norm": 2.75,
+      "learning_rate": 3.11809382797291e-06,
+      "loss": 0.763,
+      "step": 25200
+    },
+    {
+      "epoch": 0.6893534406147032,
+      "grad_norm": 2.75,
+      "learning_rate": 3.115351923446026e-06,
+      "loss": 0.6963,
+      "step": 25210
+    },
+    {
+      "epoch": 0.6896268850576284,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.1126100189191415e-06,
+      "loss": 0.695,
+      "step": 25220
+    },
+    {
+      "epoch": 0.6899003295005537,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.109868114392257e-06,
+      "loss": 0.7055,
+      "step": 25230
+    },
+    {
+      "epoch": 0.690173773943479,
+      "grad_norm": 3.546875,
+      "learning_rate": 3.1071262098653724e-06,
+      "loss": 0.7645,
+      "step": 25240
+    },
+    {
+      "epoch": 0.6904472183864043,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.104384305338488e-06,
+      "loss": 0.7013,
+      "step": 25250
+    },
+    {
+      "epoch": 0.6907206628293296,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.1016424008116043e-06,
+      "loss": 0.7244,
+      "step": 25260
+    },
+    {
+      "epoch": 0.690994107272255,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.0989004962847197e-06,
+      "loss": 0.6992,
+      "step": 25270
+    },
+    {
+      "epoch": 0.6912675517151803,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.0961585917578352e-06,
+      "loss": 0.7223,
+      "step": 25280
+    },
+    {
+      "epoch": 0.6915409961581056,
+      "grad_norm": 2.90625,
+      "learning_rate": 3.0934166872309507e-06,
+      "loss": 0.6981,
+      "step": 25290
+    },
+    {
+      "epoch": 0.6918144406010309,
+      "grad_norm": 2.984375,
+      "learning_rate": 3.0906747827040666e-06,
+      "loss": 0.7413,
+      "step": 25300
+    },
+    {
+      "epoch": 0.6920878850439562,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.087932878177182e-06,
+      "loss": 0.7292,
+      "step": 25310
+    },
+    {
+      "epoch": 0.6923613294868814,
+      "grad_norm": 2.890625,
+      "learning_rate": 3.0851909736502976e-06,
+      "loss": 0.7068,
+      "step": 25320
+    },
+    {
+      "epoch": 0.6926347739298068,
+      "grad_norm": 3.296875,
+      "learning_rate": 3.082449069123413e-06,
+      "loss": 0.7829,
+      "step": 25330
+    },
+    {
+      "epoch": 0.6929082183727321,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.0797071645965285e-06,
+      "loss": 0.761,
+      "step": 25340
+    },
+    {
+      "epoch": 0.6931816628156574,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.076965260069645e-06,
+      "loss": 0.7247,
+      "step": 25350
+    },
+    {
+      "epoch": 0.6934551072585827,
+      "grad_norm": 2.625,
+      "learning_rate": 3.0742233555427603e-06,
+      "loss": 0.7451,
+      "step": 25360
+    },
+    {
+      "epoch": 0.693728551701508,
+      "grad_norm": 2.765625,
+      "learning_rate": 3.071481451015876e-06,
+      "loss": 0.6505,
+      "step": 25370
+    },
+    {
+      "epoch": 0.6940019961444334,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.0687395464889913e-06,
+      "loss": 0.6476,
+      "step": 25380
+    },
+    {
+      "epoch": 0.6942754405873587,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.0659976419621072e-06,
+      "loss": 0.6482,
+      "step": 25390
+    },
+    {
+      "epoch": 0.694548885030284,
+      "grad_norm": 2.875,
+      "learning_rate": 3.0632557374352227e-06,
+      "loss": 0.8068,
+      "step": 25400
+    },
+    {
+      "epoch": 0.6948223294732093,
+      "grad_norm": 2.625,
+      "learning_rate": 3.060513832908338e-06,
+      "loss": 0.6539,
+      "step": 25410
+    },
+    {
+      "epoch": 0.6950957739161346,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.0577719283814537e-06,
+      "loss": 0.6979,
+      "step": 25420
+    },
+    {
+      "epoch": 0.6953692183590598,
+      "grad_norm": 3.109375,
+      "learning_rate": 3.0550300238545696e-06,
+      "loss": 0.6004,
+      "step": 25430
+    },
+    {
+      "epoch": 0.6956426628019852,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.0522881193276855e-06,
+      "loss": 0.7148,
+      "step": 25440
+    },
+    {
+      "epoch": 0.6959161072449105,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.049546214800801e-06,
+      "loss": 0.7252,
+      "step": 25450
+    },
+    {
+      "epoch": 0.6961895516878358,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.0468043102739164e-06,
+      "loss": 0.7433,
+      "step": 25460
+    },
+    {
+      "epoch": 0.6964629961307611,
+      "grad_norm": 2.96875,
+      "learning_rate": 3.044062405747032e-06,
+      "loss": 0.6297,
+      "step": 25470
+    },
+    {
+      "epoch": 0.6967364405736864,
+      "grad_norm": 2.609375,
+      "learning_rate": 3.041320501220148e-06,
+      "loss": 0.6938,
+      "step": 25480
+    },
+    {
+      "epoch": 0.6970098850166118,
+      "grad_norm": 3.125,
+      "learning_rate": 3.0385785966932633e-06,
+      "loss": 0.6621,
+      "step": 25490
+    },
+    {
+      "epoch": 0.6972833294595371,
+      "grad_norm": 2.921875,
+      "learning_rate": 3.0358366921663788e-06,
+      "loss": 0.6341,
+      "step": 25500
+    },
+    {
+      "epoch": 0.6975567739024624,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.0330947876394947e-06,
+      "loss": 0.6799,
+      "step": 25510
+    },
+    {
+      "epoch": 0.6978302183453877,
+      "grad_norm": 2.5,
+      "learning_rate": 3.03035288311261e-06,
+      "loss": 0.7273,
+      "step": 25520
+    },
+    {
+      "epoch": 0.698103662788313,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.027610978585726e-06,
+      "loss": 0.6838,
+      "step": 25530
+    },
+    {
+      "epoch": 0.6983771072312382,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.0248690740588416e-06,
+      "loss": 0.6762,
+      "step": 25540
+    },
+    {
+      "epoch": 0.6986505516741636,
+      "grad_norm": 2.125,
+      "learning_rate": 3.022127169531957e-06,
+      "loss": 0.6554,
+      "step": 25550
+    },
+    {
+      "epoch": 0.6989239961170889,
+      "grad_norm": 2.59375,
+      "learning_rate": 3.0193852650050725e-06,
+      "loss": 0.665,
+      "step": 25560
+    },
+    {
+      "epoch": 0.6991974405600142,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.0166433604781884e-06,
+      "loss": 0.6735,
+      "step": 25570
+    },
+    {
+      "epoch": 0.6994708850029395,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.013901455951304e-06,
+      "loss": 0.7068,
+      "step": 25580
+    },
+    {
+      "epoch": 0.6997443294458648,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.01115955142442e-06,
+      "loss": 0.6789,
+      "step": 25590
+    },
+    {
+      "epoch": 0.7000177738887902,
+      "grad_norm": 3.140625,
+      "learning_rate": 3.0084176468975353e-06,
+      "loss": 0.7074,
+      "step": 25600
+    },
+    {
+      "epoch": 0.7002912183317155,
+      "grad_norm": 3.015625,
+      "learning_rate": 3.0056757423706508e-06,
+      "loss": 0.6785,
+      "step": 25610
+    },
+    {
+      "epoch": 0.7005646627746408,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.0029338378437667e-06,
+      "loss": 0.6724,
+      "step": 25620
+    },
+    {
+      "epoch": 0.7008381072175661,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.000191933316882e-06,
+      "loss": 0.687,
+      "step": 25630
+    },
+    {
+      "epoch": 0.7011115516604913,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.9974500287899977e-06,
+      "loss": 0.7426,
+      "step": 25640
+    },
+    {
+      "epoch": 0.7013849961034166,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.994708124263113e-06,
+      "loss": 0.6782,
+      "step": 25650
+    },
+    {
+      "epoch": 0.701658440546342,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.991966219736229e-06,
+      "loss": 0.6829,
+      "step": 25660
+    },
+    {
+      "epoch": 0.7019318849892673,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.989224315209345e-06,
+      "loss": 0.6529,
+      "step": 25670
+    },
+    {
+      "epoch": 0.7022053294321926,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.9864824106824604e-06,
+      "loss": 0.6697,
+      "step": 25680
+    },
+    {
+      "epoch": 0.7024787738751179,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.983740506155576e-06,
+      "loss": 0.6845,
+      "step": 25690
+    },
+    {
+      "epoch": 0.7027522183180432,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9809986016286914e-06,
+      "loss": 0.7535,
+      "step": 25700
+    },
+    {
+      "epoch": 0.7030256627609686,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.9782566971018073e-06,
+      "loss": 0.6802,
+      "step": 25710
+    },
+    {
+      "epoch": 0.7032991072038939,
+      "grad_norm": 3.125,
+      "learning_rate": 2.9755147925749228e-06,
+      "loss": 0.7259,
+      "step": 25720
+    },
+    {
+      "epoch": 0.7035725516468192,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.9727728880480383e-06,
+      "loss": 0.6913,
+      "step": 25730
+    },
+    {
+      "epoch": 0.7038459960897445,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.9700309835211537e-06,
+      "loss": 0.6758,
+      "step": 25740
+    },
+    {
+      "epoch": 0.7041194405326697,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.96728907899427e-06,
+      "loss": 0.7054,
+      "step": 25750
+    },
+    {
+      "epoch": 0.704392884975595,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.9645471744673856e-06,
+      "loss": 0.6368,
+      "step": 25760
+    },
+    {
+      "epoch": 0.7046663294185204,
+      "grad_norm": 3.28125,
+      "learning_rate": 2.961805269940501e-06,
+      "loss": 0.6907,
+      "step": 25770
+    },
+    {
+      "epoch": 0.7049397738614457,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.9590633654136165e-06,
+      "loss": 0.6433,
+      "step": 25780
+    },
+    {
+      "epoch": 0.705213218304371,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.9563214608867324e-06,
+      "loss": 0.7706,
+      "step": 25790
+    },
+    {
+      "epoch": 0.7054866627472963,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.953579556359848e-06,
+      "loss": 0.6803,
+      "step": 25800
+    },
+    {
+      "epoch": 0.7057601071902216,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.9508376518329634e-06,
+      "loss": 0.7153,
+      "step": 25810
+    },
+    {
+      "epoch": 0.706033551633147,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.948095747306079e-06,
+      "loss": 0.6899,
+      "step": 25820
+    },
+    {
+      "epoch": 0.7063069960760723,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.9453538427791944e-06,
+      "loss": 0.7005,
+      "step": 25830
+    },
+    {
+      "epoch": 0.7065804405189976,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.9426119382523107e-06,
+      "loss": 0.7019,
+      "step": 25840
+    },
+    {
+      "epoch": 0.7068538849619228,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.939870033725426e-06,
+      "loss": 0.7224,
+      "step": 25850
+    },
+    {
+      "epoch": 0.7071273294048481,
+      "grad_norm": 2.875,
+      "learning_rate": 2.9371281291985416e-06,
+      "loss": 0.6899,
+      "step": 25860
+    },
+    {
+      "epoch": 0.7074007738477734,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.934386224671657e-06,
+      "loss": 0.6789,
+      "step": 25870
+    },
+    {
+      "epoch": 0.7076742182906988,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.931644320144773e-06,
+      "loss": 0.6677,
+      "step": 25880
+    },
+    {
+      "epoch": 0.7079476627336241,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.9289024156178885e-06,
+      "loss": 0.6555,
+      "step": 25890
+    },
+    {
+      "epoch": 0.7082211071765494,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.926160511091004e-06,
+      "loss": 0.7629,
+      "step": 25900
+    },
+    {
+      "epoch": 0.7084945516194747,
+      "grad_norm": 3.921875,
+      "learning_rate": 2.9234186065641195e-06,
+      "loss": 0.6896,
+      "step": 25910
+    },
+    {
+      "epoch": 0.7087679960624,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.920676702037235e-06,
+      "loss": 0.6545,
+      "step": 25920
+    },
+    {
+      "epoch": 0.7090414405053254,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.9179347975103513e-06,
+      "loss": 0.6048,
+      "step": 25930
+    },
+    {
+      "epoch": 0.7093148849482507,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.9151928929834668e-06,
+      "loss": 0.7316,
+      "step": 25940
+    },
+    {
+      "epoch": 0.709588329391176,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.9124509884565823e-06,
+      "loss": 0.7251,
+      "step": 25950
+    },
+    {
+      "epoch": 0.7098617738341012,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.9097090839296977e-06,
+      "loss": 0.7502,
+      "step": 25960
+    },
+    {
+      "epoch": 0.7101352182770265,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.9069671794028136e-06,
+      "loss": 0.6837,
+      "step": 25970
+    },
+    {
+      "epoch": 0.7104086627199518,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.904225274875929e-06,
+      "loss": 0.6053,
+      "step": 25980
+    },
+    {
+      "epoch": 0.7106821071628772,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.9014833703490446e-06,
+      "loss": 0.7697,
+      "step": 25990
+    },
+    {
+      "epoch": 0.7109555516058025,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.89874146582216e-06,
+      "loss": 0.6528,
+      "step": 26000
+    },
+    {
+      "epoch": 0.7112289960487278,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.8959995612952756e-06,
+      "loss": 0.7213,
+      "step": 26010
+    },
+    {
+      "epoch": 0.7115024404916531,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.893257656768392e-06,
+      "loss": 0.7147,
+      "step": 26020
+    },
+    {
+      "epoch": 0.7117758849345784,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.8905157522415074e-06,
+      "loss": 0.7372,
+      "step": 26030
+    },
+    {
+      "epoch": 0.7120493293775038,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.887773847714623e-06,
+      "loss": 0.6489,
+      "step": 26040
+    },
+    {
+      "epoch": 0.7123227738204291,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.8850319431877383e-06,
+      "loss": 0.7039,
+      "step": 26050
+    },
+    {
+      "epoch": 0.7125962182633544,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.8822900386608542e-06,
+      "loss": 0.7271,
+      "step": 26060
+    },
+    {
+      "epoch": 0.7128696627062796,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.8795481341339697e-06,
+      "loss": 0.6807,
+      "step": 26070
+    },
+    {
+      "epoch": 0.7131431071492049,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.876806229607085e-06,
+      "loss": 0.668,
+      "step": 26080
+    },
+    {
+      "epoch": 0.7134165515921302,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.8740643250802007e-06,
+      "loss": 0.7168,
+      "step": 26090
+    },
+    {
+      "epoch": 0.7136899960350556,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.871322420553316e-06,
+      "loss": 0.7389,
+      "step": 26100
+    },
+    {
+      "epoch": 0.7139634404779809,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.8685805160264325e-06,
+      "loss": 0.5996,
+      "step": 26110
+    },
+    {
+      "epoch": 0.7142368849209062,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.865838611499548e-06,
+      "loss": 0.7028,
+      "step": 26120
+    },
+    {
+      "epoch": 0.7145103293638315,
+      "grad_norm": 2.625,
+      "learning_rate": 2.8630967069726635e-06,
+      "loss": 0.6481,
+      "step": 26130
+    },
+    {
+      "epoch": 0.7147837738067568,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.860354802445779e-06,
+      "loss": 0.7945,
+      "step": 26140
+    },
+    {
+      "epoch": 0.7150572182496822,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.857612897918895e-06,
+      "loss": 0.6749,
+      "step": 26150
+    },
+    {
+      "epoch": 0.7153306626926075,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.8548709933920103e-06,
+      "loss": 0.6878,
+      "step": 26160
+    },
+    {
+      "epoch": 0.7156041071355327,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.852129088865126e-06,
+      "loss": 0.6528,
+      "step": 26170
+    },
+    {
+      "epoch": 0.715877551578458,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.8493871843382413e-06,
+      "loss": 0.7319,
+      "step": 26180
+    },
+    {
+      "epoch": 0.7161509960213833,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.8466452798113568e-06,
+      "loss": 0.6188,
+      "step": 26190
+    },
+    {
+      "epoch": 0.7164244404643086,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.843903375284473e-06,
+      "loss": 0.7615,
+      "step": 26200
+    },
+    {
+      "epoch": 0.716697884907234,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8411614707575886e-06,
+      "loss": 0.65,
+      "step": 26210
+    },
+    {
+      "epoch": 0.7169713293501593,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.838419566230704e-06,
+      "loss": 0.7356,
+      "step": 26220
+    },
+    {
+      "epoch": 0.7172447737930846,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.8356776617038196e-06,
+      "loss": 0.6764,
+      "step": 26230
+    },
+    {
+      "epoch": 0.7175182182360099,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.8329357571769355e-06,
+      "loss": 0.6949,
+      "step": 26240
+    },
+    {
+      "epoch": 0.7177916626789352,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.830193852650051e-06,
+      "loss": 0.6178,
+      "step": 26250
+    },
+    {
+      "epoch": 0.7180651071218606,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.8274519481231664e-06,
+      "loss": 0.6146,
+      "step": 26260
+    },
+    {
+      "epoch": 0.7183385515647859,
+      "grad_norm": 2.875,
+      "learning_rate": 2.824710043596282e-06,
+      "loss": 0.7565,
+      "step": 26270
+    },
+    {
+      "epoch": 0.7186119960077111,
+      "grad_norm": 2.34375,
+      "learning_rate": 2.8219681390693974e-06,
+      "loss": 0.6574,
+      "step": 26280
+    },
+    {
+      "epoch": 0.7188854404506364,
+      "grad_norm": 2.75,
+      "learning_rate": 2.8192262345425137e-06,
+      "loss": 0.6707,
+      "step": 26290
+    },
+    {
+      "epoch": 0.7191588848935617,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.816484330015629e-06,
+      "loss": 0.7539,
+      "step": 26300
+    },
+    {
+      "epoch": 0.719432329336487,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.8137424254887447e-06,
+      "loss": 0.6744,
+      "step": 26310
+    },
+    {
+      "epoch": 0.7197057737794124,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.81100052096186e-06,
+      "loss": 0.6788,
+      "step": 26320
+    },
+    {
+      "epoch": 0.7199792182223377,
+      "grad_norm": 2.625,
+      "learning_rate": 2.808258616434976e-06,
+      "loss": 0.6764,
+      "step": 26330
+    },
+    {
+      "epoch": 0.720252662665263,
+      "grad_norm": 3.0,
+      "learning_rate": 2.8055167119080916e-06,
+      "loss": 0.7155,
+      "step": 26340
+    },
+    {
+      "epoch": 0.7205261071081883,
+      "grad_norm": 2.875,
+      "learning_rate": 2.802774807381207e-06,
+      "loss": 0.733,
+      "step": 26350
+    },
+    {
+      "epoch": 0.7207995515511136,
+      "grad_norm": 2.5,
+      "learning_rate": 2.8000329028543225e-06,
+      "loss": 0.6468,
+      "step": 26360
+    },
+    {
+      "epoch": 0.721072995994039,
+      "grad_norm": 3.0,
+      "learning_rate": 2.7972909983274384e-06,
+      "loss": 0.6206,
+      "step": 26370
+    },
+    {
+      "epoch": 0.7213464404369642,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.7945490938005543e-06,
+      "loss": 0.7199,
+      "step": 26380
+    },
+    {
+      "epoch": 0.7216198848798895,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.79180718927367e-06,
+      "loss": 0.6472,
+      "step": 26390
+    },
+    {
+      "epoch": 0.7218933293228148,
+      "grad_norm": 3.296875,
+      "learning_rate": 2.7890652847467853e-06,
+      "loss": 0.7543,
+      "step": 26400
+    },
+    {
+      "epoch": 0.7221667737657401,
+      "grad_norm": 3.328125,
+      "learning_rate": 2.7863233802199008e-06,
+      "loss": 0.7146,
+      "step": 26410
+    },
+    {
+      "epoch": 0.7224402182086654,
+      "grad_norm": 2.21875,
+      "learning_rate": 2.7835814756930167e-06,
+      "loss": 0.6151,
+      "step": 26420
+    },
+    {
+      "epoch": 0.7227136626515908,
+      "grad_norm": 3.359375,
+      "learning_rate": 2.780839571166132e-06,
+      "loss": 0.7219,
+      "step": 26430
+    },
+    {
+      "epoch": 0.7229871070945161,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.7780976666392476e-06,
+      "loss": 0.75,
+      "step": 26440
+    },
+    {
+      "epoch": 0.7232605515374414,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.7753557621123635e-06,
+      "loss": 0.7328,
+      "step": 26450
+    },
+    {
+      "epoch": 0.7235339959803667,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.772613857585479e-06,
+      "loss": 0.6746,
+      "step": 26460
+    },
+    {
+      "epoch": 0.723807440423292,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.769871953058595e-06,
+      "loss": 0.6206,
+      "step": 26470
+    },
+    {
+      "epoch": 0.7240808848662174,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.7671300485317104e-06,
+      "loss": 0.7274,
+      "step": 26480
+    },
+    {
+      "epoch": 0.7243543293091426,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.764388144004826e-06,
+      "loss": 0.7153,
+      "step": 26490
+    },
+    {
+      "epoch": 0.7246277737520679,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.7616462394779414e-06,
+      "loss": 0.7339,
+      "step": 26500
+    },
+    {
+      "epoch": 0.7249012181949932,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.7589043349510573e-06,
+      "loss": 0.6517,
+      "step": 26510
+    },
+    {
+      "epoch": 0.7251746626379185,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.7561624304241728e-06,
+      "loss": 0.7495,
+      "step": 26520
+    },
+    {
+      "epoch": 0.7254481070808438,
+      "grad_norm": 3.46875,
+      "learning_rate": 2.7534205258972887e-06,
+      "loss": 0.755,
+      "step": 26530
+    },
+    {
+      "epoch": 0.7257215515237692,
+      "grad_norm": 3.9375,
+      "learning_rate": 2.750678621370404e-06,
+      "loss": 0.7211,
+      "step": 26540
+    },
+    {
+      "epoch": 0.7259949959666945,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.7479367168435196e-06,
+      "loss": 0.707,
+      "step": 26550
+    },
+    {
+      "epoch": 0.7262684404096198,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.7451948123166355e-06,
+      "loss": 0.6846,
+      "step": 26560
+    },
+    {
+      "epoch": 0.7265418848525451,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.742452907789751e-06,
+      "loss": 0.6856,
+      "step": 26570
+    },
+    {
+      "epoch": 0.7268153292954704,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7397110032628665e-06,
+      "loss": 0.6944,
+      "step": 26580
+    },
+    {
+      "epoch": 0.7270887737383958,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.736969098735982e-06,
+      "loss": 0.6306,
+      "step": 26590
+    },
+    {
+      "epoch": 0.727362218181321,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.734227194209098e-06,
+      "loss": 0.6281,
+      "step": 26600
+    },
+    {
+      "epoch": 0.7276356626242463,
+      "grad_norm": 2.5,
+      "learning_rate": 2.731485289682214e-06,
+      "loss": 0.6322,
+      "step": 26610
+    },
+    {
+      "epoch": 0.7279091070671716,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.7287433851553293e-06,
+      "loss": 0.621,
+      "step": 26620
+    },
+    {
+      "epoch": 0.7281825515100969,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.7260014806284448e-06,
+      "loss": 0.6693,
+      "step": 26630
+    },
+    {
+      "epoch": 0.7284559959530222,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.7232595761015602e-06,
+      "loss": 0.6598,
+      "step": 26640
+    },
+    {
+      "epoch": 0.7287294403959476,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.720517671574676e-06,
+      "loss": 0.6647,
+      "step": 26650
+    },
+    {
+      "epoch": 0.7290028848388729,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.7177757670477916e-06,
+      "loss": 0.6328,
+      "step": 26660
+    },
+    {
+      "epoch": 0.7292763292817982,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.715033862520907e-06,
+      "loss": 0.7776,
+      "step": 26670
+    },
+    {
+      "epoch": 0.7295497737247235,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7122919579940226e-06,
+      "loss": 0.6284,
+      "step": 26680
+    },
+    {
+      "epoch": 0.7298232181676488,
+      "grad_norm": 2.875,
+      "learning_rate": 2.709550053467139e-06,
+      "loss": 0.6576,
+      "step": 26690
+    },
+    {
+      "epoch": 0.730096662610574,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.7068081489402544e-06,
+      "loss": 0.7837,
+      "step": 26700
+    },
+    {
+      "epoch": 0.7303701070534994,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.70406624441337e-06,
+      "loss": 0.7047,
+      "step": 26710
+    },
+    {
+      "epoch": 0.7306435514964247,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.7013243398864854e-06,
+      "loss": 0.753,
+      "step": 26720
+    },
+    {
+      "epoch": 0.73091699593935,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.698582435359601e-06,
+      "loss": 0.7367,
+      "step": 26730
+    },
+    {
+      "epoch": 0.7311904403822753,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.6958405308327168e-06,
+      "loss": 0.6614,
+      "step": 26740
+    },
+    {
+      "epoch": 0.7314638848252006,
+      "grad_norm": 2.375,
+      "learning_rate": 2.6930986263058322e-06,
+      "loss": 0.7485,
+      "step": 26750
+    },
+    {
+      "epoch": 0.731737329268126,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.6903567217789477e-06,
+      "loss": 0.6772,
+      "step": 26760
+    },
+    {
+      "epoch": 0.7320107737110513,
+      "grad_norm": 2.75,
+      "learning_rate": 2.687614817252063e-06,
+      "loss": 0.6751,
+      "step": 26770
+    },
+    {
+      "epoch": 0.7322842181539766,
+      "grad_norm": 3.34375,
+      "learning_rate": 2.6848729127251795e-06,
+      "loss": 0.711,
+      "step": 26780
+    },
+    {
+      "epoch": 0.7325576625969019,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.682131008198295e-06,
+      "loss": 0.6291,
+      "step": 26790
+    },
+    {
+      "epoch": 0.7328311070398272,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.6793891036714105e-06,
+      "loss": 0.6424,
+      "step": 26800
+    },
+    {
+      "epoch": 0.7331045514827524,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.676647199144526e-06,
+      "loss": 0.6943,
+      "step": 26810
+    },
+    {
+      "epoch": 0.7333779959256778,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.6739052946176415e-06,
+      "loss": 0.7476,
+      "step": 26820
+    },
+    {
+      "epoch": 0.7336514403686031,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.6711633900907574e-06,
+      "loss": 0.6419,
+      "step": 26830
+    },
+    {
+      "epoch": 0.7339248848115284,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.668421485563873e-06,
+      "loss": 0.6848,
+      "step": 26840
+    },
+    {
+      "epoch": 0.7341983292544537,
+      "grad_norm": 2.625,
+      "learning_rate": 2.6656795810369883e-06,
+      "loss": 0.6847,
+      "step": 26850
+    },
+    {
+      "epoch": 0.734471773697379,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.662937676510104e-06,
+      "loss": 0.7222,
+      "step": 26860
+    },
+    {
+      "epoch": 0.7347452181403044,
+      "grad_norm": 3.25,
+      "learning_rate": 2.66019577198322e-06,
+      "loss": 0.6762,
+      "step": 26870
+    },
+    {
+      "epoch": 0.7350186625832297,
+      "grad_norm": 2.625,
+      "learning_rate": 2.6574538674563356e-06,
+      "loss": 0.6395,
+      "step": 26880
+    },
+    {
+      "epoch": 0.735292107026155,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.654711962929451e-06,
+      "loss": 0.6612,
+      "step": 26890
+    },
+    {
+      "epoch": 0.7355655514690803,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.6519700584025666e-06,
+      "loss": 0.6933,
+      "step": 26900
+    },
+    {
+      "epoch": 0.7358389959120056,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.649228153875682e-06,
+      "loss": 0.7057,
+      "step": 26910
+    },
+    {
+      "epoch": 0.7361124403549308,
+      "grad_norm": 2.875,
+      "learning_rate": 2.646486249348798e-06,
+      "loss": 0.6481,
+      "step": 26920
+    },
+    {
+      "epoch": 0.7363858847978562,
+      "grad_norm": 2.625,
+      "learning_rate": 2.6437443448219135e-06,
+      "loss": 0.6462,
+      "step": 26930
+    },
+    {
+      "epoch": 0.7366593292407815,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.641002440295029e-06,
+      "loss": 0.6532,
+      "step": 26940
+    },
+    {
+      "epoch": 0.7369327736837068,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.6382605357681444e-06,
+      "loss": 0.679,
+      "step": 26950
+    },
+    {
+      "epoch": 0.7372062181266321,
+      "grad_norm": 3.578125,
+      "learning_rate": 2.6355186312412608e-06,
+      "loss": 0.7106,
+      "step": 26960
+    },
+    {
+      "epoch": 0.7374796625695574,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.6327767267143762e-06,
+      "loss": 0.6571,
+      "step": 26970
+    },
+    {
+      "epoch": 0.7377531070124828,
+      "grad_norm": 2.5,
+      "learning_rate": 2.6300348221874917e-06,
+      "loss": 0.6749,
+      "step": 26980
+    },
+    {
+      "epoch": 0.7380265514554081,
+      "grad_norm": 2.625,
+      "learning_rate": 2.627292917660607e-06,
+      "loss": 0.7128,
+      "step": 26990
+    },
+    {
+      "epoch": 0.7382999958983334,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.6245510131337227e-06,
+      "loss": 0.6555,
+      "step": 27000
+    },
+    {
+      "epoch": 0.7385734403412587,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.6218091086068386e-06,
+      "loss": 0.6867,
+      "step": 27010
+    },
+    {
+      "epoch": 0.7388468847841839,
+      "grad_norm": 3.671875,
+      "learning_rate": 2.619067204079954e-06,
+      "loss": 0.7057,
+      "step": 27020
+    },
+    {
+      "epoch": 0.7391203292271092,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.6163252995530696e-06,
+      "loss": 0.6199,
+      "step": 27030
+    },
+    {
+      "epoch": 0.7393937736700346,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.613583395026185e-06,
+      "loss": 0.684,
+      "step": 27040
+    },
+    {
+      "epoch": 0.7396672181129599,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.6108414904993014e-06,
+      "loss": 0.7214,
+      "step": 27050
+    },
+    {
+      "epoch": 0.7399406625558852,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.608099585972417e-06,
+      "loss": 0.7273,
+      "step": 27060
+    },
+    {
+      "epoch": 0.7402141069988105,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.6053576814455323e-06,
+      "loss": 0.7282,
+      "step": 27070
+    },
+    {
+      "epoch": 0.7404875514417358,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.602615776918648e-06,
+      "loss": 0.7235,
+      "step": 27080
+    },
+    {
+      "epoch": 0.7407609958846612,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.5998738723917637e-06,
+      "loss": 0.673,
+      "step": 27090
+    },
+    {
+      "epoch": 0.7410344403275865,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.597131967864879e-06,
+      "loss": 0.6798,
+      "step": 27100
+    },
+    {
+      "epoch": 0.7413078847705118,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.5943900633379947e-06,
+      "loss": 0.7259,
+      "step": 27110
+    },
+    {
+      "epoch": 0.7415813292134371,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.59164815881111e-06,
+      "loss": 0.6767,
+      "step": 27120
+    },
+    {
+      "epoch": 0.7418547736563623,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.5889062542842256e-06,
+      "loss": 0.6648,
+      "step": 27130
+    },
+    {
+      "epoch": 0.7421282180992876,
+      "grad_norm": 2.4375,
+      "learning_rate": 2.586164349757342e-06,
+      "loss": 0.6945,
+      "step": 27140
+    },
+    {
+      "epoch": 0.742401662542213,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.5834224452304575e-06,
+      "loss": 0.6622,
+      "step": 27150
+    },
+    {
+      "epoch": 0.7426751069851383,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.580680540703573e-06,
+      "loss": 0.6715,
+      "step": 27160
+    },
+    {
+      "epoch": 0.7429485514280636,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.5779386361766884e-06,
+      "loss": 0.6882,
+      "step": 27170
+    },
+    {
+      "epoch": 0.7432219958709889,
+      "grad_norm": 3.21875,
+      "learning_rate": 2.5751967316498043e-06,
+      "loss": 0.6983,
+      "step": 27180
+    },
+    {
+      "epoch": 0.7434954403139142,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.57245482712292e-06,
+      "loss": 0.6911,
+      "step": 27190
+    },
+    {
+      "epoch": 0.7437688847568396,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.5697129225960353e-06,
+      "loss": 0.7332,
+      "step": 27200
+    },
+    {
+      "epoch": 0.7440423291997649,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.5669710180691508e-06,
+      "loss": 0.6356,
+      "step": 27210
+    },
+    {
+      "epoch": 0.7443157736426902,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.5642291135422662e-06,
+      "loss": 0.6714,
+      "step": 27220
+    },
+    {
+      "epoch": 0.7445892180856154,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.5614872090153826e-06,
+      "loss": 0.6495,
+      "step": 27230
+    },
+    {
+      "epoch": 0.7448626625285407,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.558745304488498e-06,
+      "loss": 0.7544,
+      "step": 27240
+    },
+    {
+      "epoch": 0.745136106971466,
+      "grad_norm": 3.59375,
+      "learning_rate": 2.5560033999616135e-06,
+      "loss": 0.7132,
+      "step": 27250
+    },
+    {
+      "epoch": 0.7454095514143914,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.553261495434729e-06,
+      "loss": 0.7109,
+      "step": 27260
+    },
+    {
+      "epoch": 0.7456829958573167,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.550519590907845e-06,
+      "loss": 0.6141,
+      "step": 27270
+    },
+    {
+      "epoch": 0.745956440300242,
+      "grad_norm": 2.5,
+      "learning_rate": 2.5477776863809604e-06,
+      "loss": 0.7022,
+      "step": 27280
+    },
+    {
+      "epoch": 0.7462298847431673,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.545035781854076e-06,
+      "loss": 0.691,
+      "step": 27290
+    },
+    {
+      "epoch": 0.7465033291860926,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.5422938773271914e-06,
+      "loss": 0.7147,
+      "step": 27300
+    },
+    {
+      "epoch": 0.746776773629018,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.5395519728003073e-06,
+      "loss": 0.6676,
+      "step": 27310
+    },
+    {
+      "epoch": 0.7470502180719433,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.536810068273423e-06,
+      "loss": 0.6751,
+      "step": 27320
+    },
+    {
+      "epoch": 0.7473236625148686,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.5340681637465387e-06,
+      "loss": 0.7644,
+      "step": 27330
+    },
+    {
+      "epoch": 0.7475971069577938,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.531326259219654e-06,
+      "loss": 0.7566,
+      "step": 27340
+    },
+    {
+      "epoch": 0.7478705514007191,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.5285843546927696e-06,
+      "loss": 0.6961,
+      "step": 27350
+    },
+    {
+      "epoch": 0.7481439958436444,
+      "grad_norm": 3.4375,
+      "learning_rate": 2.5258424501658855e-06,
+      "loss": 0.7344,
+      "step": 27360
+    },
+    {
+      "epoch": 0.7484174402865698,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.523100545639001e-06,
+      "loss": 0.6898,
+      "step": 27370
+    },
+    {
+      "epoch": 0.7486908847294951,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.5203586411121165e-06,
+      "loss": 0.6549,
+      "step": 27380
+    },
+    {
+      "epoch": 0.7489643291724204,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.5176167365852324e-06,
+      "loss": 0.6599,
+      "step": 27390
+    },
+    {
+      "epoch": 0.7492377736153457,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.514874832058348e-06,
+      "loss": 0.7372,
+      "step": 27400
+    },
+    {
+      "epoch": 0.749511218058271,
+      "grad_norm": 2.75,
+      "learning_rate": 2.512132927531464e-06,
+      "loss": 0.6892,
+      "step": 27410
+    },
+    {
+      "epoch": 0.7497846625011964,
+      "grad_norm": 3.125,
+      "learning_rate": 2.5093910230045793e-06,
+      "loss": 0.7421,
+      "step": 27420
+    },
+    {
+      "epoch": 0.7500581069441217,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.5066491184776948e-06,
+      "loss": 0.6903,
+      "step": 27430
+    },
+    {
+      "epoch": 0.750331551387047,
+      "grad_norm": 2.625,
+      "learning_rate": 2.5039072139508102e-06,
+      "loss": 0.657,
+      "step": 27440
+    },
+    {
+      "epoch": 0.7506049958299722,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.501165309423926e-06,
+      "loss": 0.6459,
+      "step": 27450
+    },
+    {
+      "epoch": 0.7508784402728975,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.4984234048970416e-06,
+      "loss": 0.7908,
+      "step": 27460
+    },
+    {
+      "epoch": 0.7511518847158228,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.4956815003701575e-06,
+      "loss": 0.7309,
+      "step": 27470
+    },
+    {
+      "epoch": 0.7514253291587482,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.492939595843273e-06,
+      "loss": 0.676,
+      "step": 27480
+    },
+    {
+      "epoch": 0.7516987736016735,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.4901976913163885e-06,
+      "loss": 0.6583,
+      "step": 27490
+    },
+    {
+      "epoch": 0.7519722180445988,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.487455786789504e-06,
+      "loss": 0.6944,
+      "step": 27500
+    },
+    {
+      "epoch": 0.7522456624875241,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.48471388226262e-06,
+      "loss": 0.7356,
+      "step": 27510
+    },
+    {
+      "epoch": 0.7525191069304494,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.4819719777357354e-06,
+      "loss": 0.7559,
+      "step": 27520
+    },
+    {
+      "epoch": 0.7527925513733748,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.4792300732088513e-06,
+      "loss": 0.6029,
+      "step": 27530
+    },
+    {
+      "epoch": 0.7530659958163001,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.4764881686819668e-06,
+      "loss": 0.7059,
+      "step": 27540
+    },
+    {
+      "epoch": 0.7533394402592253,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.4737462641550827e-06,
+      "loss": 0.6691,
+      "step": 27550
+    },
+    {
+      "epoch": 0.7536128847021506,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.471004359628198e-06,
+      "loss": 0.6429,
+      "step": 27560
+    },
+    {
+      "epoch": 0.7538863291450759,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.4682624551013136e-06,
+      "loss": 0.669,
+      "step": 27570
+    },
+    {
+      "epoch": 0.7541597735880012,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.465520550574429e-06,
+      "loss": 0.6678,
+      "step": 27580
+    },
+    {
+      "epoch": 0.7544332180309266,
+      "grad_norm": 3.734375,
+      "learning_rate": 2.4627786460475446e-06,
+      "loss": 0.7395,
+      "step": 27590
+    },
+    {
+      "epoch": 0.7547066624738519,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.4600367415206605e-06,
+      "loss": 0.6557,
+      "step": 27600
+    },
+    {
+      "epoch": 0.7549801069167772,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.457294836993776e-06,
+      "loss": 0.7038,
+      "step": 27610
+    },
+    {
+      "epoch": 0.7552535513597025,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.454552932466892e-06,
+      "loss": 0.7161,
+      "step": 27620
+    },
+    {
+      "epoch": 0.7555269958026278,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.4518110279400074e-06,
+      "loss": 0.7045,
+      "step": 27630
+    },
+    {
+      "epoch": 0.7558004402455532,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.4490691234131233e-06,
+      "loss": 0.7158,
+      "step": 27640
+    },
+    {
+      "epoch": 0.7560738846884785,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.4463272188862387e-06,
+      "loss": 0.6648,
+      "step": 27650
+    },
+    {
+      "epoch": 0.7563473291314037,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.4435853143593542e-06,
+      "loss": 0.7042,
+      "step": 27660
+    },
+    {
+      "epoch": 0.756620773574329,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.4408434098324697e-06,
+      "loss": 0.7044,
+      "step": 27670
+    },
+    {
+      "epoch": 0.7568942180172543,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.438101505305585e-06,
+      "loss": 0.6179,
+      "step": 27680
+    },
+    {
+      "epoch": 0.7571676624601796,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.435359600778701e-06,
+      "loss": 0.6136,
+      "step": 27690
+    },
+    {
+      "epoch": 0.757441106903105,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.4326176962518166e-06,
+      "loss": 0.7061,
+      "step": 27700
+    },
+    {
+      "epoch": 0.7577145513460303,
+      "grad_norm": 3.375,
+      "learning_rate": 2.4298757917249325e-06,
+      "loss": 0.6911,
+      "step": 27710
+    },
+    {
+      "epoch": 0.7579879957889556,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.427133887198048e-06,
+      "loss": 0.6592,
+      "step": 27720
+    },
+    {
+      "epoch": 0.7582614402318809,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.424391982671164e-06,
+      "loss": 0.6529,
+      "step": 27730
+    },
+    {
+      "epoch": 0.7585348846748062,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.4216500781442794e-06,
+      "loss": 0.7146,
+      "step": 27740
+    },
+    {
+      "epoch": 0.7588083291177316,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.418908173617395e-06,
+      "loss": 0.6487,
+      "step": 27750
+    },
+    {
+      "epoch": 0.7590817735606568,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.4161662690905103e-06,
+      "loss": 0.6496,
+      "step": 27760
+    },
+    {
+      "epoch": 0.7593552180035821,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.413424364563626e-06,
+      "loss": 0.6842,
+      "step": 27770
+    },
+    {
+      "epoch": 0.7596286624465074,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.4106824600367417e-06,
+      "loss": 0.6868,
+      "step": 27780
+    },
+    {
+      "epoch": 0.7599021068894327,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.407940555509857e-06,
+      "loss": 0.6998,
+      "step": 27790
+    },
+    {
+      "epoch": 0.760175551332358,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.405198650982973e-06,
+      "loss": 0.7066,
+      "step": 27800
+    },
+    {
+      "epoch": 0.7604489957752834,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.4024567464560886e-06,
+      "loss": 0.6898,
+      "step": 27810
+    },
+    {
+      "epoch": 0.7607224402182087,
+      "grad_norm": 2.5,
+      "learning_rate": 2.3997148419292045e-06,
+      "loss": 0.6877,
+      "step": 27820
+    },
+    {
+      "epoch": 0.760995884661134,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.39697293740232e-06,
+      "loss": 0.6975,
+      "step": 27830
+    },
+    {
+      "epoch": 0.7612693291040593,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.3942310328754354e-06,
+      "loss": 0.665,
+      "step": 27840
+    },
+    {
+      "epoch": 0.7615427735469846,
+      "grad_norm": 3.125,
+      "learning_rate": 2.391489128348551e-06,
+      "loss": 0.7079,
+      "step": 27850
+    },
+    {
+      "epoch": 0.76181621798991,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.388747223821667e-06,
+      "loss": 0.6674,
+      "step": 27860
+    },
+    {
+      "epoch": 0.7620896624328352,
+      "grad_norm": 3.234375,
+      "learning_rate": 2.3860053192947823e-06,
+      "loss": 0.6587,
+      "step": 27870
+    },
+    {
+      "epoch": 0.7623631068757605,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.383263414767898e-06,
+      "loss": 0.7108,
+      "step": 27880
+    },
+    {
+      "epoch": 0.7626365513186858,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.3805215102410137e-06,
+      "loss": 0.6927,
+      "step": 27890
+    },
+    {
+      "epoch": 0.7629099957616111,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.377779605714129e-06,
+      "loss": 0.6825,
+      "step": 27900
+    },
+    {
+      "epoch": 0.7631834402045364,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.375037701187245e-06,
+      "loss": 0.6959,
+      "step": 27910
+    },
+    {
+      "epoch": 0.7634568846474618,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.3722957966603606e-06,
+      "loss": 0.8114,
+      "step": 27920
+    },
+    {
+      "epoch": 0.7637303290903871,
+      "grad_norm": 2.625,
+      "learning_rate": 2.369553892133476e-06,
+      "loss": 0.6609,
+      "step": 27930
+    },
+    {
+      "epoch": 0.7640037735333124,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.366811987606592e-06,
+      "loss": 0.6765,
+      "step": 27940
+    },
+    {
+      "epoch": 0.7642772179762377,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.3640700830797074e-06,
+      "loss": 0.6795,
+      "step": 27950
+    },
+    {
+      "epoch": 0.764550662419163,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.361328178552823e-06,
+      "loss": 0.6434,
+      "step": 27960
+    },
+    {
+      "epoch": 0.7648241068620883,
+      "grad_norm": 2.75,
+      "learning_rate": 2.3585862740259384e-06,
+      "loss": 0.678,
+      "step": 27970
+    },
+    {
+      "epoch": 0.7650975513050136,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.3558443694990543e-06,
+      "loss": 0.6875,
+      "step": 27980
+    },
+    {
+      "epoch": 0.7653709957479389,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.35310246497217e-06,
+      "loss": 0.7308,
+      "step": 27990
+    },
+    {
+      "epoch": 0.7656444401908642,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.3503605604452857e-06,
+      "loss": 0.6993,
+      "step": 28000
+    },
+    {
+      "epoch": 0.7659178846337895,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.347618655918401e-06,
+      "loss": 0.7111,
+      "step": 28010
+    },
+    {
+      "epoch": 0.7661913290767148,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.3448767513915167e-06,
+      "loss": 0.6926,
+      "step": 28020
+    },
+    {
+      "epoch": 0.7664647735196402,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.3421348468646326e-06,
+      "loss": 0.6278,
+      "step": 28030
+    },
+    {
+      "epoch": 0.7667382179625655,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.339392942337748e-06,
+      "loss": 0.636,
+      "step": 28040
+    },
+    {
+      "epoch": 0.7670116624054908,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.3366510378108635e-06,
+      "loss": 0.6023,
+      "step": 28050
+    },
+    {
+      "epoch": 0.7672851068484161,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.333909133283979e-06,
+      "loss": 0.6754,
+      "step": 28060
+    },
+    {
+      "epoch": 0.7675585512913414,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.331167228757095e-06,
+      "loss": 0.6969,
+      "step": 28070
+    },
+    {
+      "epoch": 0.7678319957342666,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.3284253242302104e-06,
+      "loss": 0.7725,
+      "step": 28080
+    },
+    {
+      "epoch": 0.768105440177192,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.3256834197033263e-06,
+      "loss": 0.6004,
+      "step": 28090
+    },
+    {
+      "epoch": 0.7683788846201173,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.322941515176442e-06,
+      "loss": 0.7091,
+      "step": 28100
+    },
+    {
+      "epoch": 0.7686523290630426,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.3201996106495577e-06,
+      "loss": 0.6678,
+      "step": 28110
+    },
+    {
+      "epoch": 0.7689257735059679,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.317457706122673e-06,
+      "loss": 0.6659,
+      "step": 28120
+    },
+    {
+      "epoch": 0.7691992179488932,
+      "grad_norm": 3.125,
+      "learning_rate": 2.3147158015957887e-06,
+      "loss": 0.623,
+      "step": 28130
+    },
+    {
+      "epoch": 0.7694726623918186,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.311973897068904e-06,
+      "loss": 0.595,
+      "step": 28140
+    },
+    {
+      "epoch": 0.7697461068347439,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.3092319925420196e-06,
+      "loss": 0.7364,
+      "step": 28150
+    },
+    {
+      "epoch": 0.7700195512776692,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.3064900880151355e-06,
+      "loss": 0.6746,
+      "step": 28160
+    },
+    {
+      "epoch": 0.7702929957205945,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.303748183488251e-06,
+      "loss": 0.6785,
+      "step": 28170
+    },
+    {
+      "epoch": 0.7705664401635198,
+      "grad_norm": 2.328125,
+      "learning_rate": 2.301006278961367e-06,
+      "loss": 0.6948,
+      "step": 28180
+    },
+    {
+      "epoch": 0.770839884606445,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.2982643744344824e-06,
+      "loss": 0.7128,
+      "step": 28190
+    },
+    {
+      "epoch": 0.7711133290493704,
+      "grad_norm": 3.0,
+      "learning_rate": 2.2955224699075983e-06,
+      "loss": 0.7539,
+      "step": 28200
+    },
+    {
+      "epoch": 0.7713867734922957,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.2927805653807138e-06,
+      "loss": 0.6673,
+      "step": 28210
+    },
+    {
+      "epoch": 0.771660217935221,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.2900386608538293e-06,
+      "loss": 0.7276,
+      "step": 28220
+    },
+    {
+      "epoch": 0.7719336623781463,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.2872967563269448e-06,
+      "loss": 0.609,
+      "step": 28230
+    },
+    {
+      "epoch": 0.7722071068210716,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.2845548518000602e-06,
+      "loss": 0.7507,
+      "step": 28240
+    },
+    {
+      "epoch": 0.772480551263997,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.281812947273176e-06,
+      "loss": 0.6425,
+      "step": 28250
+    },
+    {
+      "epoch": 0.7727539957069223,
+      "grad_norm": 3.15625,
+      "learning_rate": 2.2790710427462916e-06,
+      "loss": 0.7171,
+      "step": 28260
+    },
+    {
+      "epoch": 0.7730274401498476,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.2763291382194075e-06,
+      "loss": 0.652,
+      "step": 28270
+    },
+    {
+      "epoch": 0.7733008845927729,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.273587233692523e-06,
+      "loss": 0.6733,
+      "step": 28280
+    },
+    {
+      "epoch": 0.7735743290356982,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.270845329165639e-06,
+      "loss": 0.6628,
+      "step": 28290
+    },
+    {
+      "epoch": 0.7738477734786234,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.2681034246387544e-06,
+      "loss": 0.7505,
+      "step": 28300
+    },
+    {
+      "epoch": 0.7741212179215488,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.26536152011187e-06,
+      "loss": 0.731,
+      "step": 28310
+    },
+    {
+      "epoch": 0.7743946623644741,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.2626196155849854e-06,
+      "loss": 0.655,
+      "step": 28320
+    },
+    {
+      "epoch": 0.7746681068073994,
+      "grad_norm": 2.875,
+      "learning_rate": 2.2598777110581013e-06,
+      "loss": 0.7311,
+      "step": 28330
+    },
+    {
+      "epoch": 0.7749415512503247,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.2571358065312167e-06,
+      "loss": 0.7112,
+      "step": 28340
+    },
+    {
+      "epoch": 0.77521499569325,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.2543939020043322e-06,
+      "loss": 0.6663,
+      "step": 28350
+    },
+    {
+      "epoch": 0.7754884401361753,
+      "grad_norm": 3.1875,
+      "learning_rate": 2.251651997477448e-06,
+      "loss": 0.6556,
+      "step": 28360
+    },
+    {
+      "epoch": 0.7757618845791007,
+      "grad_norm": 2.625,
+      "learning_rate": 2.2489100929505636e-06,
+      "loss": 0.6644,
+      "step": 28370
+    },
+    {
+      "epoch": 0.776035329022026,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.2461681884236795e-06,
+      "loss": 0.7091,
+      "step": 28380
+    },
+    {
+      "epoch": 0.7763087734649513,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.243426283896795e-06,
+      "loss": 0.6637,
+      "step": 28390
+    },
+    {
+      "epoch": 0.7765822179078765,
+      "grad_norm": 2.515625,
+      "learning_rate": 2.2406843793699105e-06,
+      "loss": 0.7008,
+      "step": 28400
+    },
+    {
+      "epoch": 0.7768556623508018,
+      "grad_norm": 3.03125,
+      "learning_rate": 2.2379424748430264e-06,
+      "loss": 0.725,
+      "step": 28410
+    },
+    {
+      "epoch": 0.7771291067937272,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.235200570316142e-06,
+      "loss": 0.7132,
+      "step": 28420
+    },
+    {
+      "epoch": 0.7774025512366525,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.2324586657892574e-06,
+      "loss": 0.7478,
+      "step": 28430
+    },
+    {
+      "epoch": 0.7776759956795778,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.229716761262373e-06,
+      "loss": 0.6303,
+      "step": 28440
+    },
+    {
+      "epoch": 0.7779494401225031,
+      "grad_norm": 3.171875,
+      "learning_rate": 2.2269748567354887e-06,
+      "loss": 0.6339,
+      "step": 28450
+    },
+    {
+      "epoch": 0.7782228845654284,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.2242329522086042e-06,
+      "loss": 0.6326,
+      "step": 28460
+    },
+    {
+      "epoch": 0.7784963290083537,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.22149104768172e-06,
+      "loss": 0.6357,
+      "step": 28470
+    },
+    {
+      "epoch": 0.7787697734512791,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.2187491431548356e-06,
+      "loss": 0.6949,
+      "step": 28480
+    },
+    {
+      "epoch": 0.7790432178942044,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.216007238627951e-06,
+      "loss": 0.6841,
+      "step": 28490
+    },
+    {
+      "epoch": 0.7793166623371297,
+      "grad_norm": 3.015625,
+      "learning_rate": 2.213265334101067e-06,
+      "loss": 0.6528,
+      "step": 28500
+    },
+    {
+      "epoch": 0.7795901067800549,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.2105234295741825e-06,
+      "loss": 0.7035,
+      "step": 28510
+    },
+    {
+      "epoch": 0.7798635512229802,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.207781525047298e-06,
+      "loss": 0.7023,
+      "step": 28520
+    },
+    {
+      "epoch": 0.7801369956659056,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.2050396205204134e-06,
+      "loss": 0.7383,
+      "step": 28530
+    },
+    {
+      "epoch": 0.7804104401088309,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.2022977159935294e-06,
+      "loss": 0.6561,
+      "step": 28540
+    },
+    {
+      "epoch": 0.7806838845517562,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.199555811466645e-06,
+      "loss": 0.6811,
+      "step": 28550
+    },
+    {
+      "epoch": 0.7809573289946815,
+      "grad_norm": 3.0625,
+      "learning_rate": 2.1968139069397607e-06,
+      "loss": 0.6894,
+      "step": 28560
+    },
+    {
+      "epoch": 0.7812307734376068,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.1940720024128762e-06,
+      "loss": 0.7475,
+      "step": 28570
+    },
+    {
+      "epoch": 0.7815042178805321,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.1913300978859917e-06,
+      "loss": 0.6895,
+      "step": 28580
+    },
+    {
+      "epoch": 0.7817776623234575,
+      "grad_norm": 2.578125,
+      "learning_rate": 2.1885881933591076e-06,
+      "loss": 0.6484,
+      "step": 28590
+    },
+    {
+      "epoch": 0.7820511067663828,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.185846288832223e-06,
+      "loss": 0.7171,
+      "step": 28600
+    },
+    {
+      "epoch": 0.782324551209308,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.1831043843053386e-06,
+      "loss": 0.686,
+      "step": 28610
+    },
+    {
+      "epoch": 0.7825979956522333,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.180362479778454e-06,
+      "loss": 0.681,
+      "step": 28620
+    },
+    {
+      "epoch": 0.7828714400951586,
+      "grad_norm": 2.203125,
+      "learning_rate": 2.17762057525157e-06,
+      "loss": 0.6183,
+      "step": 28630
+    },
+    {
+      "epoch": 0.783144884538084,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.1748786707246854e-06,
+      "loss": 0.6628,
+      "step": 28640
+    },
+    {
+      "epoch": 0.7834183289810093,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.1721367661978013e-06,
+      "loss": 0.7031,
+      "step": 28650
+    },
+    {
+      "epoch": 0.7836917734239346,
+      "grad_norm": 2.25,
+      "learning_rate": 2.169394861670917e-06,
+      "loss": 0.5876,
+      "step": 28660
+    },
+    {
+      "epoch": 0.7839652178668599,
+      "grad_norm": 3.265625,
+      "learning_rate": 2.1666529571440323e-06,
+      "loss": 0.7004,
+      "step": 28670
+    },
+    {
+      "epoch": 0.7842386623097852,
+      "grad_norm": 2.734375,
+      "learning_rate": 2.1639110526171482e-06,
+      "loss": 0.6449,
+      "step": 28680
+    },
+    {
+      "epoch": 0.7845121067527105,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.1611691480902637e-06,
+      "loss": 0.6954,
+      "step": 28690
+    },
+    {
+      "epoch": 0.7847855511956359,
+      "grad_norm": 2.75,
+      "learning_rate": 2.158427243563379e-06,
+      "loss": 0.6598,
+      "step": 28700
+    },
+    {
+      "epoch": 0.7850589956385612,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.1556853390364947e-06,
+      "loss": 0.717,
+      "step": 28710
+    },
+    {
+      "epoch": 0.7853324400814864,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.1529434345096106e-06,
+      "loss": 0.616,
+      "step": 28720
+    },
+    {
+      "epoch": 0.7856058845244117,
+      "grad_norm": 2.40625,
+      "learning_rate": 2.150201529982726e-06,
+      "loss": 0.6828,
+      "step": 28730
+    },
+    {
+      "epoch": 0.785879328967337,
+      "grad_norm": 2.375,
+      "learning_rate": 2.147459625455842e-06,
+      "loss": 0.742,
+      "step": 28740
+    },
+    {
+      "epoch": 0.7861527734102624,
+      "grad_norm": 3.203125,
+      "learning_rate": 2.1447177209289574e-06,
+      "loss": 0.63,
+      "step": 28750
+    },
+    {
+      "epoch": 0.7864262178531877,
+      "grad_norm": 2.78125,
+      "learning_rate": 2.1419758164020733e-06,
+      "loss": 0.6715,
+      "step": 28760
+    },
+    {
+      "epoch": 0.786699662296113,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.139233911875189e-06,
+      "loss": 0.6895,
+      "step": 28770
+    },
+    {
+      "epoch": 0.7869731067390383,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.1364920073483043e-06,
+      "loss": 0.6485,
+      "step": 28780
+    },
+    {
+      "epoch": 0.7872465511819636,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.1337501028214198e-06,
+      "loss": 0.7096,
+      "step": 28790
+    },
+    {
+      "epoch": 0.787519995624889,
+      "grad_norm": 2.75,
+      "learning_rate": 2.1310081982945353e-06,
+      "loss": 0.6793,
+      "step": 28800
+    },
+    {
+      "epoch": 0.7877934400678143,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.128266293767651e-06,
+      "loss": 0.676,
+      "step": 28810
+    },
+    {
+      "epoch": 0.7880668845107396,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.1255243892407667e-06,
+      "loss": 0.6814,
+      "step": 28820
+    },
+    {
+      "epoch": 0.7883403289536648,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.1227824847138826e-06,
+      "loss": 0.7144,
+      "step": 28830
+    },
+    {
+      "epoch": 0.7886137733965901,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.120040580186998e-06,
+      "loss": 0.7356,
+      "step": 28840
+    },
+    {
+      "epoch": 0.7888872178395154,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.117298675660114e-06,
+      "loss": 0.664,
+      "step": 28850
+    },
+    {
+      "epoch": 0.7891606622824408,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.1145567711332294e-06,
+      "loss": 0.706,
+      "step": 28860
+    },
+    {
+      "epoch": 0.7894341067253661,
+      "grad_norm": 2.9375,
+      "learning_rate": 2.111814866606345e-06,
+      "loss": 0.6952,
+      "step": 28870
+    },
+    {
+      "epoch": 0.7897075511682914,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.1090729620794604e-06,
+      "loss": 0.6487,
+      "step": 28880
+    },
+    {
+      "epoch": 0.7899809956112167,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.1063310575525763e-06,
+      "loss": 0.6579,
+      "step": 28890
+    },
+    {
+      "epoch": 0.790254440054142,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.1035891530256918e-06,
+      "loss": 0.7273,
+      "step": 28900
+    },
+    {
+      "epoch": 0.7905278844970673,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.1008472484988073e-06,
+      "loss": 0.5908,
+      "step": 28910
+    },
+    {
+      "epoch": 0.7908013289399927,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.098105343971923e-06,
+      "loss": 0.6259,
+      "step": 28920
+    },
+    {
+      "epoch": 0.7910747733829179,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.0953634394450387e-06,
+      "loss": 0.6401,
+      "step": 28930
+    },
+    {
+      "epoch": 0.7913482178258432,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.0926215349181546e-06,
+      "loss": 0.706,
+      "step": 28940
+    },
+    {
+      "epoch": 0.7916216622687685,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.08987963039127e-06,
+      "loss": 0.6401,
+      "step": 28950
+    },
+    {
+      "epoch": 0.7918951067116938,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.0871377258643855e-06,
+      "loss": 0.6159,
+      "step": 28960
+    },
+    {
+      "epoch": 0.7921685511546191,
+      "grad_norm": 2.828125,
+      "learning_rate": 2.0843958213375014e-06,
+      "loss": 0.6776,
+      "step": 28970
+    },
+    {
+      "epoch": 0.7924419955975445,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.081653916810617e-06,
+      "loss": 0.6629,
+      "step": 28980
+    },
+    {
+      "epoch": 0.7927154400404698,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.0789120122837324e-06,
+      "loss": 0.647,
+      "step": 28990
+    },
+    {
+      "epoch": 0.7929888844833951,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.076170107756848e-06,
+      "loss": 0.6613,
+      "step": 29000
+    },
+    {
+      "epoch": 0.7932623289263204,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.0734282032299638e-06,
+      "loss": 0.6359,
+      "step": 29010
+    },
+    {
+      "epoch": 0.7935357733692457,
+      "grad_norm": 2.90625,
+      "learning_rate": 2.0706862987030793e-06,
+      "loss": 0.7339,
+      "step": 29020
+    },
+    {
+      "epoch": 0.7938092178121711,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.067944394176195e-06,
+      "loss": 0.7012,
+      "step": 29030
+    },
+    {
+      "epoch": 0.7940826622550963,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.0652024896493106e-06,
+      "loss": 0.6424,
+      "step": 29040
+    },
+    {
+      "epoch": 0.7943561066980216,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.062460585122426e-06,
+      "loss": 0.7177,
+      "step": 29050
+    },
+    {
+      "epoch": 0.7946295511409469,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.059718680595542e-06,
+      "loss": 0.6562,
+      "step": 29060
+    },
+    {
+      "epoch": 0.7949029955838722,
+      "grad_norm": 3.484375,
+      "learning_rate": 2.0569767760686575e-06,
+      "loss": 0.7826,
+      "step": 29070
+    },
+    {
+      "epoch": 0.7951764400267975,
+      "grad_norm": 3.078125,
+      "learning_rate": 2.054234871541773e-06,
+      "loss": 0.6957,
+      "step": 29080
+    },
+    {
+      "epoch": 0.7954498844697229,
+      "grad_norm": 2.609375,
+      "learning_rate": 2.0514929670148885e-06,
+      "loss": 0.7241,
+      "step": 29090
+    },
+    {
+      "epoch": 0.7957233289126482,
+      "grad_norm": 3.0,
+      "learning_rate": 2.0487510624880044e-06,
+      "loss": 0.741,
+      "step": 29100
+    },
+    {
+      "epoch": 0.7959967733555735,
+      "grad_norm": 2.65625,
+      "learning_rate": 2.04600915796112e-06,
+      "loss": 0.7424,
+      "step": 29110
+    },
+    {
+      "epoch": 0.7962702177984988,
+      "grad_norm": 2.5,
+      "learning_rate": 2.0432672534342358e-06,
+      "loss": 0.6503,
+      "step": 29120
+    },
+    {
+      "epoch": 0.7965436622414241,
+      "grad_norm": 2.125,
+      "learning_rate": 2.0405253489073513e-06,
+      "loss": 0.6394,
+      "step": 29130
+    },
+    {
+      "epoch": 0.7968171066843494,
+      "grad_norm": 2.375,
+      "learning_rate": 2.0377834443804667e-06,
+      "loss": 0.6781,
+      "step": 29140
+    },
+    {
+      "epoch": 0.7970905511272747,
+      "grad_norm": 3.109375,
+      "learning_rate": 2.0350415398535826e-06,
+      "loss": 0.7336,
+      "step": 29150
+    },
+    {
+      "epoch": 0.7973639955702,
+      "grad_norm": 2.96875,
+      "learning_rate": 2.032299635326698e-06,
+      "loss": 0.7016,
+      "step": 29160
+    },
+    {
+      "epoch": 0.7976374400131253,
+      "grad_norm": 2.75,
+      "learning_rate": 2.0295577307998136e-06,
+      "loss": 0.6697,
+      "step": 29170
+    },
+    {
+      "epoch": 0.7979108844560506,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.026815826272929e-06,
+      "loss": 0.6629,
+      "step": 29180
+    },
+    {
+      "epoch": 0.798184328898976,
+      "grad_norm": 2.671875,
+      "learning_rate": 2.024073921746045e-06,
+      "loss": 0.6533,
+      "step": 29190
+    },
+    {
+      "epoch": 0.7984577733419013,
+      "grad_norm": 3.09375,
+      "learning_rate": 2.0213320172191605e-06,
+      "loss": 0.6641,
+      "step": 29200
+    },
+    {
+      "epoch": 0.7987312177848266,
+      "grad_norm": 2.765625,
+      "learning_rate": 2.0185901126922764e-06,
+      "loss": 0.6704,
+      "step": 29210
+    },
+    {
+      "epoch": 0.7990046622277519,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.015848208165392e-06,
+      "loss": 0.7237,
+      "step": 29220
+    },
+    {
+      "epoch": 0.7992781066706772,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.0131063036385073e-06,
+      "loss": 0.6812,
+      "step": 29230
+    },
+    {
+      "epoch": 0.7995515511136025,
+      "grad_norm": 2.875,
+      "learning_rate": 2.0103643991116233e-06,
+      "loss": 0.7029,
+      "step": 29240
+    },
+    {
+      "epoch": 0.7998249955565278,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.0076224945847387e-06,
+      "loss": 0.6595,
+      "step": 29250
+    },
+    {
+      "epoch": 0.8000984399994531,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.0048805900578542e-06,
+      "loss": 0.6282,
+      "step": 29260
+    },
+    {
+      "epoch": 0.8003718844423784,
+      "grad_norm": 2.546875,
+      "learning_rate": 2.0021386855309697e-06,
+      "loss": 0.7454,
+      "step": 29270
+    },
+    {
+      "epoch": 0.8006453288853037,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.9993967810040856e-06,
+      "loss": 0.6795,
+      "step": 29280
+    },
+    {
+      "epoch": 0.800918773328229,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.996654876477201e-06,
+      "loss": 0.6849,
+      "step": 29290
+    },
+    {
+      "epoch": 0.8011922177711543,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.993912971950317e-06,
+      "loss": 0.7439,
+      "step": 29300
+    },
+    {
+      "epoch": 0.8014656622140797,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.9911710674234325e-06,
+      "loss": 0.6776,
+      "step": 29310
+    },
+    {
+      "epoch": 0.801739106657005,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.988429162896548e-06,
+      "loss": 0.7098,
+      "step": 29320
+    },
+    {
+      "epoch": 0.8020125510999303,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.985687258369664e-06,
+      "loss": 0.6648,
+      "step": 29330
+    },
+    {
+      "epoch": 0.8022859955428556,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.9829453538427793e-06,
+      "loss": 0.7303,
+      "step": 29340
+    },
+    {
+      "epoch": 0.8025594399857809,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.980203449315895e-06,
+      "loss": 0.7445,
+      "step": 29350
+    },
+    {
+      "epoch": 0.8028328844287062,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9774615447890107e-06,
+      "loss": 0.7238,
+      "step": 29360
+    },
+    {
+      "epoch": 0.8031063288716315,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.974719640262126e-06,
+      "loss": 0.6743,
+      "step": 29370
+    },
+    {
+      "epoch": 0.8033797733145568,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.9719777357352417e-06,
+      "loss": 0.6651,
+      "step": 29380
+    },
+    {
+      "epoch": 0.8036532177574821,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.9692358312083576e-06,
+      "loss": 0.7182,
+      "step": 29390
+    },
+    {
+      "epoch": 0.8039266622004074,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.966493926681473e-06,
+      "loss": 0.7179,
+      "step": 29400
+    },
+    {
+      "epoch": 0.8042001066433327,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.9637520221545886e-06,
+      "loss": 0.6637,
+      "step": 29410
+    },
+    {
+      "epoch": 0.8044735510862581,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.9610101176277045e-06,
+      "loss": 0.6719,
+      "step": 29420
+    },
+    {
+      "epoch": 0.8047469955291834,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.95826821310082e-06,
+      "loss": 0.7492,
+      "step": 29430
+    },
+    {
+      "epoch": 0.8050204399721087,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.955526308573936e-06,
+      "loss": 0.6767,
+      "step": 29440
+    },
+    {
+      "epoch": 0.805293884415034,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.9527844040470513e-06,
+      "loss": 0.6514,
+      "step": 29450
+    },
+    {
+      "epoch": 0.8055673288579592,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.950042499520167e-06,
+      "loss": 0.6393,
+      "step": 29460
+    },
+    {
+      "epoch": 0.8058407733008845,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.9473005949932823e-06,
+      "loss": 0.8214,
+      "step": 29470
+    },
+    {
+      "epoch": 0.8061142177438099,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.944558690466398e-06,
+      "loss": 0.6726,
+      "step": 29480
+    },
+    {
+      "epoch": 0.8063876621867352,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9418167859395137e-06,
+      "loss": 0.731,
+      "step": 29490
+    },
+    {
+      "epoch": 0.8066611066296605,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.9390748814126296e-06,
+      "loss": 0.7451,
+      "step": 29500
+    },
+    {
+      "epoch": 0.8069345510725858,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.936332976885745e-06,
+      "loss": 0.6598,
+      "step": 29510
+    },
+    {
+      "epoch": 0.8072079955155111,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.9335910723588606e-06,
+      "loss": 0.6708,
+      "step": 29520
+    },
+    {
+      "epoch": 0.8074814399584365,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.9308491678319765e-06,
+      "loss": 0.6284,
+      "step": 29530
+    },
+    {
+      "epoch": 0.8077548844013618,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.928107263305092e-06,
+      "loss": 0.6692,
+      "step": 29540
+    },
+    {
+      "epoch": 0.8080283288442871,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.9253653587782074e-06,
+      "loss": 0.6918,
+      "step": 29550
+    },
+    {
+      "epoch": 0.8083017732872124,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.922623454251323e-06,
+      "loss": 0.6994,
+      "step": 29560
+    },
+    {
+      "epoch": 0.8085752177301376,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.919881549724439e-06,
+      "loss": 0.7276,
+      "step": 29570
+    },
+    {
+      "epoch": 0.808848662173063,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.9171396451975543e-06,
+      "loss": 0.6658,
+      "step": 29580
+    },
+    {
+      "epoch": 0.8091221066159883,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.91439774067067e-06,
+      "loss": 0.649,
+      "step": 29590
+    },
+    {
+      "epoch": 0.8093955510589136,
+      "grad_norm": 2.625,
+      "learning_rate": 1.9116558361437857e-06,
+      "loss": 0.6513,
+      "step": 29600
+    },
+    {
+      "epoch": 0.8096689955018389,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.908913931616901e-06,
+      "loss": 0.7674,
+      "step": 29610
+    },
+    {
+      "epoch": 0.8099424399447642,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.9061720270900169e-06,
+      "loss": 0.6799,
+      "step": 29620
+    },
+    {
+      "epoch": 0.8102158843876895,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.9034301225631323e-06,
+      "loss": 0.644,
+      "step": 29630
+    },
+    {
+      "epoch": 0.8104893288306149,
+      "grad_norm": 3.78125,
+      "learning_rate": 1.9006882180362482e-06,
+      "loss": 0.6633,
+      "step": 29640
+    },
+    {
+      "epoch": 0.8107627732735402,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.8979463135093637e-06,
+      "loss": 0.7557,
+      "step": 29650
+    },
+    {
+      "epoch": 0.8110362177164655,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.8952044089824794e-06,
+      "loss": 0.6416,
+      "step": 29660
+    },
+    {
+      "epoch": 0.8113096621593908,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.892462504455595e-06,
+      "loss": 0.7213,
+      "step": 29670
+    },
+    {
+      "epoch": 0.811583106602316,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.8897205999287108e-06,
+      "loss": 0.634,
+      "step": 29680
+    },
+    {
+      "epoch": 0.8118565510452413,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.8869786954018263e-06,
+      "loss": 0.6895,
+      "step": 29690
+    },
+    {
+      "epoch": 0.8121299954881667,
+      "grad_norm": 2.140625,
+      "learning_rate": 1.8842367908749418e-06,
+      "loss": 0.6604,
+      "step": 29700
+    },
+    {
+      "epoch": 0.812403439931092,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.8814948863480575e-06,
+      "loss": 0.684,
+      "step": 29710
+    },
+    {
+      "epoch": 0.8126768843740173,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.878752981821173e-06,
+      "loss": 0.6527,
+      "step": 29720
+    },
+    {
+      "epoch": 0.8129503288169426,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.8760110772942889e-06,
+      "loss": 0.6745,
+      "step": 29730
+    },
+    {
+      "epoch": 0.8132237732598679,
+      "grad_norm": 3.421875,
+      "learning_rate": 1.8732691727674043e-06,
+      "loss": 0.763,
+      "step": 29740
+    },
+    {
+      "epoch": 0.8134972177027933,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.87052726824052e-06,
+      "loss": 0.6093,
+      "step": 29750
+    },
+    {
+      "epoch": 0.8137706621457186,
+      "grad_norm": 2.28125,
+      "learning_rate": 1.8677853637136355e-06,
+      "loss": 0.6659,
+      "step": 29760
+    },
+    {
+      "epoch": 0.8140441065886439,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.8650434591867514e-06,
+      "loss": 0.7346,
+      "step": 29770
+    },
+    {
+      "epoch": 0.8143175510315691,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.862301554659867e-06,
+      "loss": 0.6485,
+      "step": 29780
+    },
+    {
+      "epoch": 0.8145909954744944,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.8595596501329824e-06,
+      "loss": 0.6405,
+      "step": 29790
+    },
+    {
+      "epoch": 0.8148644399174197,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.856817745606098e-06,
+      "loss": 0.6973,
+      "step": 29800
+    },
+    {
+      "epoch": 0.8151378843603451,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.8540758410792138e-06,
+      "loss": 0.6907,
+      "step": 29810
+    },
+    {
+      "epoch": 0.8154113288032704,
+      "grad_norm": 2.875,
+      "learning_rate": 1.8513339365523295e-06,
+      "loss": 0.6947,
+      "step": 29820
+    },
+    {
+      "epoch": 0.8156847732461957,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.848592032025445e-06,
+      "loss": 0.6937,
+      "step": 29830
+    },
+    {
+      "epoch": 0.815958217689121,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.8458501274985606e-06,
+      "loss": 0.6427,
+      "step": 29840
+    },
+    {
+      "epoch": 0.8162316621320463,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.8431082229716763e-06,
+      "loss": 0.676,
+      "step": 29850
+    },
+    {
+      "epoch": 0.8165051065749717,
+      "grad_norm": 2.375,
+      "learning_rate": 1.840366318444792e-06,
+      "loss": 0.7,
+      "step": 29860
+    },
+    {
+      "epoch": 0.816778551017897,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.8376244139179075e-06,
+      "loss": 0.6519,
+      "step": 29870
+    },
+    {
+      "epoch": 0.8170519954608223,
+      "grad_norm": 2.625,
+      "learning_rate": 1.834882509391023e-06,
+      "loss": 0.7279,
+      "step": 29880
+    },
+    {
+      "epoch": 0.8173254399037475,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.832140604864139e-06,
+      "loss": 0.7716,
+      "step": 29890
+    },
+    {
+      "epoch": 0.8175988843466728,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.8293987003372544e-06,
+      "loss": 0.6372,
+      "step": 29900
+    },
+    {
+      "epoch": 0.8178723287895981,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.82665679581037e-06,
+      "loss": 0.7262,
+      "step": 29910
+    },
+    {
+      "epoch": 0.8181457732325235,
+      "grad_norm": 3.9375,
+      "learning_rate": 1.8239148912834856e-06,
+      "loss": 0.6635,
+      "step": 29920
+    },
+    {
+      "epoch": 0.8184192176754488,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.8211729867566015e-06,
+      "loss": 0.6248,
+      "step": 29930
+    },
+    {
+      "epoch": 0.8186926621183741,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.818431082229717e-06,
+      "loss": 0.7272,
+      "step": 29940
+    },
+    {
+      "epoch": 0.8189661065612994,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.8156891777028326e-06,
+      "loss": 0.7776,
+      "step": 29950
+    },
+    {
+      "epoch": 0.8192395510042247,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.8129472731759481e-06,
+      "loss": 0.6713,
+      "step": 29960
+    },
+    {
+      "epoch": 0.8195129954471501,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.8102053686490636e-06,
+      "loss": 0.7753,
+      "step": 29970
+    },
+    {
+      "epoch": 0.8197864398900754,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.8074634641221795e-06,
+      "loss": 0.6909,
+      "step": 29980
+    },
+    {
+      "epoch": 0.8200598843330006,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.804721559595295e-06,
+      "loss": 0.7323,
+      "step": 29990
+    },
+    {
+      "epoch": 0.8203333287759259,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.8019796550684107e-06,
+      "loss": 0.747,
+      "step": 30000
+    },
+    {
+      "epoch": 0.8206067732188512,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.7992377505415262e-06,
+      "loss": 0.6893,
+      "step": 30010
+    },
+    {
+      "epoch": 0.8208802176617765,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.796495846014642e-06,
+      "loss": 0.6364,
+      "step": 30020
+    },
+    {
+      "epoch": 0.8211536621047019,
+      "grad_norm": 3.25,
+      "learning_rate": 1.7937539414877576e-06,
+      "loss": 0.6876,
+      "step": 30030
+    },
+    {
+      "epoch": 0.8214271065476272,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.7910120369608732e-06,
+      "loss": 0.7402,
+      "step": 30040
+    },
+    {
+      "epoch": 0.8217005509905525,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.7882701324339887e-06,
+      "loss": 0.6785,
+      "step": 30050
+    },
+    {
+      "epoch": 0.8219739954334778,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.7855282279071042e-06,
+      "loss": 0.7366,
+      "step": 30060
+    },
+    {
+      "epoch": 0.8222474398764031,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.7827863233802201e-06,
+      "loss": 0.7553,
+      "step": 30070
+    },
+    {
+      "epoch": 0.8225208843193285,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.7800444188533356e-06,
+      "loss": 0.6867,
+      "step": 30080
+    },
+    {
+      "epoch": 0.8227943287622538,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.7773025143264513e-06,
+      "loss": 0.6681,
+      "step": 30090
+    },
+    {
+      "epoch": 0.823067773205179,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.7745606097995668e-06,
+      "loss": 0.6665,
+      "step": 30100
+    },
+    {
+      "epoch": 0.8233412176481043,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.7718187052726827e-06,
+      "loss": 0.6859,
+      "step": 30110
+    },
+    {
+      "epoch": 0.8236146620910296,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.7690768007457982e-06,
+      "loss": 0.6791,
+      "step": 30120
+    },
+    {
+      "epoch": 0.823888106533955,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.7663348962189139e-06,
+      "loss": 0.6777,
+      "step": 30130
+    },
+    {
+      "epoch": 0.8241615509768803,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.7635929916920293e-06,
+      "loss": 0.6707,
+      "step": 30140
+    },
+    {
+      "epoch": 0.8244349954198056,
+      "grad_norm": 3.125,
+      "learning_rate": 1.7608510871651452e-06,
+      "loss": 0.6685,
+      "step": 30150
+    },
+    {
+      "epoch": 0.8247084398627309,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.7581091826382607e-06,
+      "loss": 0.7005,
+      "step": 30160
+    },
+    {
+      "epoch": 0.8249818843056562,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.7553672781113762e-06,
+      "loss": 0.6219,
+      "step": 30170
+    },
+    {
+      "epoch": 0.8252553287485815,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.752625373584492e-06,
+      "loss": 0.6975,
+      "step": 30180
+    },
+    {
+      "epoch": 0.8255287731915069,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.7498834690576074e-06,
+      "loss": 0.7406,
+      "step": 30190
+    },
+    {
+      "epoch": 0.8258022176344322,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.7471415645307233e-06,
+      "loss": 0.7332,
+      "step": 30200
+    },
+    {
+      "epoch": 0.8260756620773574,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.7443996600038388e-06,
+      "loss": 0.6965,
+      "step": 30210
+    },
+    {
+      "epoch": 0.8263491065202827,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.7416577554769545e-06,
+      "loss": 0.6646,
+      "step": 30220
+    },
+    {
+      "epoch": 0.826622550963208,
+      "grad_norm": 2.75,
+      "learning_rate": 1.73891585095007e-06,
+      "loss": 0.6773,
+      "step": 30230
+    },
+    {
+      "epoch": 0.8268959954061333,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.7361739464231858e-06,
+      "loss": 0.6588,
+      "step": 30240
+    },
+    {
+      "epoch": 0.8271694398490587,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.7334320418963013e-06,
+      "loss": 0.7025,
+      "step": 30250
+    },
+    {
+      "epoch": 0.827442884291984,
+      "grad_norm": 3.0,
+      "learning_rate": 1.7306901373694168e-06,
+      "loss": 0.6728,
+      "step": 30260
+    },
+    {
+      "epoch": 0.8277163287349093,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.7279482328425325e-06,
+      "loss": 0.7441,
+      "step": 30270
+    },
+    {
+      "epoch": 0.8279897731778346,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.7252063283156482e-06,
+      "loss": 0.6255,
+      "step": 30280
+    },
+    {
+      "epoch": 0.8282632176207599,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.7224644237887639e-06,
+      "loss": 0.7182,
+      "step": 30290
+    },
+    {
+      "epoch": 0.8285366620636853,
+      "grad_norm": 2.296875,
+      "learning_rate": 1.7197225192618794e-06,
+      "loss": 0.6508,
+      "step": 30300
+    },
+    {
+      "epoch": 0.8288101065066105,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.716980614734995e-06,
+      "loss": 0.7083,
+      "step": 30310
+    },
+    {
+      "epoch": 0.8290835509495358,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.7142387102081108e-06,
+      "loss": 0.6775,
+      "step": 30320
+    },
+    {
+      "epoch": 0.8293569953924611,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.7114968056812265e-06,
+      "loss": 0.6648,
+      "step": 30330
+    },
+    {
+      "epoch": 0.8296304398353864,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.708754901154342e-06,
+      "loss": 0.6443,
+      "step": 30340
+    },
+    {
+      "epoch": 0.8299038842783117,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.7060129966274574e-06,
+      "loss": 0.6466,
+      "step": 30350
+    },
+    {
+      "epoch": 0.8301773287212371,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.7032710921005733e-06,
+      "loss": 0.6562,
+      "step": 30360
+    },
+    {
+      "epoch": 0.8304507731641624,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.7005291875736888e-06,
+      "loss": 0.7088,
+      "step": 30370
+    },
+    {
+      "epoch": 0.8307242176070877,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.6977872830468045e-06,
+      "loss": 0.7132,
+      "step": 30380
+    },
+    {
+      "epoch": 0.830997662050013,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.69504537851992e-06,
+      "loss": 0.7735,
+      "step": 30390
+    },
+    {
+      "epoch": 0.8312711064929383,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.6923034739930359e-06,
+      "loss": 0.6987,
+      "step": 30400
+    },
+    {
+      "epoch": 0.8315445509358637,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.6895615694661514e-06,
+      "loss": 0.6989,
+      "step": 30410
+    },
+    {
+      "epoch": 0.8318179953787889,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.686819664939267e-06,
+      "loss": 0.6231,
+      "step": 30420
+    },
+    {
+      "epoch": 0.8320914398217142,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.6840777604123825e-06,
+      "loss": 0.7272,
+      "step": 30430
+    },
+    {
+      "epoch": 0.8323648842646395,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.681335855885498e-06,
+      "loss": 0.6673,
+      "step": 30440
+    },
+    {
+      "epoch": 0.8326383287075648,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.678593951358614e-06,
+      "loss": 0.6205,
+      "step": 30450
+    },
+    {
+      "epoch": 0.8329117731504901,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.6758520468317294e-06,
+      "loss": 0.6708,
+      "step": 30460
+    },
+    {
+      "epoch": 0.8331852175934155,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.6731101423048451e-06,
+      "loss": 0.602,
+      "step": 30470
+    },
+    {
+      "epoch": 0.8334586620363408,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.6703682377779606e-06,
+      "loss": 0.6627,
+      "step": 30480
+    },
+    {
+      "epoch": 0.8337321064792661,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.6676263332510765e-06,
+      "loss": 0.6648,
+      "step": 30490
+    },
+    {
+      "epoch": 0.8340055509221914,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.664884428724192e-06,
+      "loss": 0.6699,
+      "step": 30500
+    },
+    {
+      "epoch": 0.8342789953651167,
+      "grad_norm": 3.375,
+      "learning_rate": 1.6621425241973077e-06,
+      "loss": 0.7066,
+      "step": 30510
+    },
+    {
+      "epoch": 0.834552439808042,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.6594006196704232e-06,
+      "loss": 0.6297,
+      "step": 30520
+    },
+    {
+      "epoch": 0.8348258842509673,
+      "grad_norm": 2.75,
+      "learning_rate": 1.6566587151435386e-06,
+      "loss": 0.6894,
+      "step": 30530
+    },
+    {
+      "epoch": 0.8350993286938926,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.6539168106166545e-06,
+      "loss": 0.6887,
+      "step": 30540
+    },
+    {
+      "epoch": 0.8353727731368179,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.65117490608977e-06,
+      "loss": 0.6956,
+      "step": 30550
+    },
+    {
+      "epoch": 0.8356462175797432,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.6484330015628857e-06,
+      "loss": 0.6743,
+      "step": 30560
+    },
+    {
+      "epoch": 0.8359196620226685,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.6456910970360012e-06,
+      "loss": 0.7092,
+      "step": 30570
+    },
+    {
+      "epoch": 0.8361931064655939,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.642949192509117e-06,
+      "loss": 0.5998,
+      "step": 30580
+    },
+    {
+      "epoch": 0.8364665509085192,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.6402072879822326e-06,
+      "loss": 0.7203,
+      "step": 30590
+    },
+    {
+      "epoch": 0.8367399953514445,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.6374653834553483e-06,
+      "loss": 0.7093,
+      "step": 30600
+    },
+    {
+      "epoch": 0.8370134397943698,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.6347234789284638e-06,
+      "loss": 0.7403,
+      "step": 30610
+    },
+    {
+      "epoch": 0.8372868842372951,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.6319815744015792e-06,
+      "loss": 0.6797,
+      "step": 30620
+    },
+    {
+      "epoch": 0.8375603286802203,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.6292396698746952e-06,
+      "loss": 0.6737,
+      "step": 30630
+    },
+    {
+      "epoch": 0.8378337731231457,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.6264977653478106e-06,
+      "loss": 0.7157,
+      "step": 30640
+    },
+    {
+      "epoch": 0.838107217566071,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.6237558608209263e-06,
+      "loss": 0.7188,
+      "step": 30650
+    },
+    {
+      "epoch": 0.8383806620089963,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.6210139562940418e-06,
+      "loss": 0.7357,
+      "step": 30660
+    },
+    {
+      "epoch": 0.8386541064519216,
+      "grad_norm": 2.875,
+      "learning_rate": 1.6182720517671577e-06,
+      "loss": 0.7572,
+      "step": 30670
+    },
+    {
+      "epoch": 0.8389275508948469,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.6155301472402732e-06,
+      "loss": 0.7094,
+      "step": 30680
+    },
+    {
+      "epoch": 0.8392009953377723,
+      "grad_norm": 3.0,
+      "learning_rate": 1.6127882427133889e-06,
+      "loss": 0.7043,
+      "step": 30690
+    },
+    {
+      "epoch": 0.8394744397806976,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.6100463381865044e-06,
+      "loss": 0.6885,
+      "step": 30700
+    },
+    {
+      "epoch": 0.8397478842236229,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.60730443365962e-06,
+      "loss": 0.7758,
+      "step": 30710
+    },
+    {
+      "epoch": 0.8400213286665482,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.6045625291327358e-06,
+      "loss": 0.7035,
+      "step": 30720
+    },
+    {
+      "epoch": 0.8402947731094735,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.6018206246058512e-06,
+      "loss": 0.7599,
+      "step": 30730
+    },
+    {
+      "epoch": 0.8405682175523987,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.599078720078967e-06,
+      "loss": 0.7077,
+      "step": 30740
+    },
+    {
+      "epoch": 0.8408416619953241,
+      "grad_norm": 3.0,
+      "learning_rate": 1.5963368155520826e-06,
+      "loss": 0.6576,
+      "step": 30750
+    },
+    {
+      "epoch": 0.8411151064382494,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.5935949110251983e-06,
+      "loss": 0.7599,
+      "step": 30760
+    },
+    {
+      "epoch": 0.8413885508811747,
+      "grad_norm": 2.75,
+      "learning_rate": 1.5908530064983138e-06,
+      "loss": 0.6395,
+      "step": 30770
+    },
+    {
+      "epoch": 0.8416619953241,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.5881111019714295e-06,
+      "loss": 0.7263,
+      "step": 30780
+    },
+    {
+      "epoch": 0.8419354397670253,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.5853691974445452e-06,
+      "loss": 0.6954,
+      "step": 30790
+    },
+    {
+      "epoch": 0.8422088842099507,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.5826272929176609e-06,
+      "loss": 0.7124,
+      "step": 30800
+    },
+    {
+      "epoch": 0.842482328652876,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.5798853883907764e-06,
+      "loss": 0.6956,
+      "step": 30810
+    },
+    {
+      "epoch": 0.8427557730958013,
+      "grad_norm": 3.265625,
+      "learning_rate": 1.5771434838638918e-06,
+      "loss": 0.7108,
+      "step": 30820
+    },
+    {
+      "epoch": 0.8430292175387266,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.5744015793370078e-06,
+      "loss": 0.6347,
+      "step": 30830
+    },
+    {
+      "epoch": 0.8433026619816518,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.5716596748101232e-06,
+      "loss": 0.6504,
+      "step": 30840
+    },
+    {
+      "epoch": 0.8435761064245771,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.568917770283239e-06,
+      "loss": 0.7289,
+      "step": 30850
+    },
+    {
+      "epoch": 0.8438495508675025,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.5661758657563544e-06,
+      "loss": 0.6362,
+      "step": 30860
+    },
+    {
+      "epoch": 0.8441229953104278,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.5634339612294703e-06,
+      "loss": 0.707,
+      "step": 30870
+    },
+    {
+      "epoch": 0.8443964397533531,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.5606920567025858e-06,
+      "loss": 0.7408,
+      "step": 30880
+    },
+    {
+      "epoch": 0.8446698841962784,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.5579501521757015e-06,
+      "loss": 0.6686,
+      "step": 30890
+    },
+    {
+      "epoch": 0.8449433286392037,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.555208247648817e-06,
+      "loss": 0.7821,
+      "step": 30900
+    },
+    {
+      "epoch": 0.8452167730821291,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.5524663431219325e-06,
+      "loss": 0.6423,
+      "step": 30910
+    },
+    {
+      "epoch": 0.8454902175250544,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.5497244385950484e-06,
+      "loss": 0.7212,
+      "step": 30920
+    },
+    {
+      "epoch": 0.8457636619679797,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.5469825340681638e-06,
+      "loss": 0.7124,
+      "step": 30930
+    },
+    {
+      "epoch": 0.846037106410905,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.5442406295412795e-06,
+      "loss": 0.7046,
+      "step": 30940
+    },
+    {
+      "epoch": 0.8463105508538302,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.541498725014395e-06,
+      "loss": 0.6397,
+      "step": 30950
+    },
+    {
+      "epoch": 0.8465839952967555,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.538756820487511e-06,
+      "loss": 0.6764,
+      "step": 30960
+    },
+    {
+      "epoch": 0.8468574397396809,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.5360149159606264e-06,
+      "loss": 0.6684,
+      "step": 30970
+    },
+    {
+      "epoch": 0.8471308841826062,
+      "grad_norm": 2.875,
+      "learning_rate": 1.533273011433742e-06,
+      "loss": 0.7348,
+      "step": 30980
+    },
+    {
+      "epoch": 0.8474043286255315,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.5305311069068576e-06,
+      "loss": 0.7689,
+      "step": 30990
+    },
+    {
+      "epoch": 0.8476777730684568,
+      "grad_norm": 2.625,
+      "learning_rate": 1.527789202379973e-06,
+      "loss": 0.7193,
+      "step": 31000
+    },
+    {
+      "epoch": 0.8479512175113821,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.525047297853089e-06,
+      "loss": 0.751,
+      "step": 31010
+    },
+    {
+      "epoch": 0.8482246619543075,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.5223053933262045e-06,
+      "loss": 0.6922,
+      "step": 31020
+    },
+    {
+      "epoch": 0.8484981063972328,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.5195634887993201e-06,
+      "loss": 0.6646,
+      "step": 31030
+    },
+    {
+      "epoch": 0.8487715508401581,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.5168215842724356e-06,
+      "loss": 0.7276,
+      "step": 31040
+    },
+    {
+      "epoch": 0.8490449952830834,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.5140796797455515e-06,
+      "loss": 0.7017,
+      "step": 31050
+    },
+    {
+      "epoch": 0.8493184397260086,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.511337775218667e-06,
+      "loss": 0.6766,
+      "step": 31060
+    },
+    {
+      "epoch": 0.8495918841689339,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.5085958706917827e-06,
+      "loss": 0.6522,
+      "step": 31070
+    },
+    {
+      "epoch": 0.8498653286118593,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.5058539661648982e-06,
+      "loss": 0.5558,
+      "step": 31080
+    },
+    {
+      "epoch": 0.8501387730547846,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.5031120616380137e-06,
+      "loss": 0.7203,
+      "step": 31090
+    },
+    {
+      "epoch": 0.8504122174977099,
+      "grad_norm": 2.5,
+      "learning_rate": 1.5003701571111296e-06,
+      "loss": 0.699,
+      "step": 31100
+    },
+    {
+      "epoch": 0.8506856619406352,
+      "grad_norm": 3.234375,
+      "learning_rate": 1.497628252584245e-06,
+      "loss": 0.6766,
+      "step": 31110
+    },
+    {
+      "epoch": 0.8509591063835605,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.4948863480573608e-06,
+      "loss": 0.6129,
+      "step": 31120
+    },
+    {
+      "epoch": 0.8512325508264859,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.4921444435304762e-06,
+      "loss": 0.6463,
+      "step": 31130
+    },
+    {
+      "epoch": 0.8515059952694112,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.4894025390035921e-06,
+      "loss": 0.7253,
+      "step": 31140
+    },
+    {
+      "epoch": 0.8517794397123365,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.4866606344767076e-06,
+      "loss": 0.6665,
+      "step": 31150
+    },
+    {
+      "epoch": 0.8520528841552617,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.4839187299498233e-06,
+      "loss": 0.6528,
+      "step": 31160
+    },
+    {
+      "epoch": 0.852326328598187,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4811768254229388e-06,
+      "loss": 0.6397,
+      "step": 31170
+    },
+    {
+      "epoch": 0.8525997730411123,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.4784349208960545e-06,
+      "loss": 0.684,
+      "step": 31180
+    },
+    {
+      "epoch": 0.8528732174840377,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.4756930163691702e-06,
+      "loss": 0.6894,
+      "step": 31190
+    },
+    {
+      "epoch": 0.853146661926963,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.4729511118422857e-06,
+      "loss": 0.7683,
+      "step": 31200
+    },
+    {
+      "epoch": 0.8534201063698883,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.4702092073154014e-06,
+      "loss": 0.6482,
+      "step": 31210
+    },
+    {
+      "epoch": 0.8536935508128136,
+      "grad_norm": 2.34375,
+      "learning_rate": 1.467467302788517e-06,
+      "loss": 0.6837,
+      "step": 31220
+    },
+    {
+      "epoch": 0.8539669952557389,
+      "grad_norm": 2.359375,
+      "learning_rate": 1.4647253982616328e-06,
+      "loss": 0.6255,
+      "step": 31230
+    },
+    {
+      "epoch": 0.8542404396986643,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.4619834937347482e-06,
+      "loss": 0.6877,
+      "step": 31240
+    },
+    {
+      "epoch": 0.8545138841415896,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.459241589207864e-06,
+      "loss": 0.736,
+      "step": 31250
+    },
+    {
+      "epoch": 0.8547873285845149,
+      "grad_norm": 3.078125,
+      "learning_rate": 1.4564996846809796e-06,
+      "loss": 0.7007,
+      "step": 31260
+    },
+    {
+      "epoch": 0.8550607730274401,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.453757780154095e-06,
+      "loss": 0.7401,
+      "step": 31270
+    },
+    {
+      "epoch": 0.8553342174703654,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.4510158756272108e-06,
+      "loss": 0.7197,
+      "step": 31280
+    },
+    {
+      "epoch": 0.8556076619132907,
+      "grad_norm": 2.75,
+      "learning_rate": 1.4482739711003263e-06,
+      "loss": 0.6267,
+      "step": 31290
+    },
+    {
+      "epoch": 0.8558811063562161,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.4455320665734422e-06,
+      "loss": 0.6282,
+      "step": 31300
+    },
+    {
+      "epoch": 0.8561545507991414,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.4427901620465577e-06,
+      "loss": 0.6542,
+      "step": 31310
+    },
+    {
+      "epoch": 0.8564279952420667,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.4400482575196734e-06,
+      "loss": 0.6859,
+      "step": 31320
+    },
+    {
+      "epoch": 0.856701439684992,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.4373063529927888e-06,
+      "loss": 0.666,
+      "step": 31330
+    },
+    {
+      "epoch": 0.8569748841279173,
+      "grad_norm": 3.09375,
+      "learning_rate": 1.4345644484659047e-06,
+      "loss": 0.6532,
+      "step": 31340
+    },
+    {
+      "epoch": 0.8572483285708427,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.4318225439390202e-06,
+      "loss": 0.6707,
+      "step": 31350
+    },
+    {
+      "epoch": 0.857521773013768,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.4290806394121357e-06,
+      "loss": 0.6709,
+      "step": 31360
+    },
+    {
+      "epoch": 0.8577952174566932,
+      "grad_norm": 3.328125,
+      "learning_rate": 1.4263387348852514e-06,
+      "loss": 0.6642,
+      "step": 31370
+    },
+    {
+      "epoch": 0.8580686618996185,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.4235968303583669e-06,
+      "loss": 0.6578,
+      "step": 31380
+    },
+    {
+      "epoch": 0.8583421063425438,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.4208549258314828e-06,
+      "loss": 0.6275,
+      "step": 31390
+    },
+    {
+      "epoch": 0.8586155507854691,
+      "grad_norm": 3.15625,
+      "learning_rate": 1.4181130213045983e-06,
+      "loss": 0.6046,
+      "step": 31400
+    },
+    {
+      "epoch": 0.8588889952283945,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.415371116777714e-06,
+      "loss": 0.7618,
+      "step": 31410
+    },
+    {
+      "epoch": 0.8591624396713198,
+      "grad_norm": 2.875,
+      "learning_rate": 1.4126292122508294e-06,
+      "loss": 0.6928,
+      "step": 31420
+    },
+    {
+      "epoch": 0.8594358841142451,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.4098873077239454e-06,
+      "loss": 0.7199,
+      "step": 31430
+    },
+    {
+      "epoch": 0.8597093285571704,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.4071454031970608e-06,
+      "loss": 0.7412,
+      "step": 31440
+    },
+    {
+      "epoch": 0.8599827730000957,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.4044034986701765e-06,
+      "loss": 0.7258,
+      "step": 31450
+    },
+    {
+      "epoch": 0.860256217443021,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.401661594143292e-06,
+      "loss": 0.7338,
+      "step": 31460
+    },
+    {
+      "epoch": 0.8605296618859464,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.3989196896164075e-06,
+      "loss": 0.6748,
+      "step": 31470
+    },
+    {
+      "epoch": 0.8608031063288716,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.3961777850895234e-06,
+      "loss": 0.6383,
+      "step": 31480
+    },
+    {
+      "epoch": 0.8610765507717969,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.3934358805626389e-06,
+      "loss": 0.6804,
+      "step": 31490
+    },
+    {
+      "epoch": 0.8613499952147222,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.3906939760357546e-06,
+      "loss": 0.7167,
+      "step": 31500
+    },
+    {
+      "epoch": 0.8616234396576475,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.38795207150887e-06,
+      "loss": 0.7357,
+      "step": 31510
+    },
+    {
+      "epoch": 0.8618968841005729,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.385210166981986e-06,
+      "loss": 0.6955,
+      "step": 31520
+    },
+    {
+      "epoch": 0.8621703285434982,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.3824682624551014e-06,
+      "loss": 0.6083,
+      "step": 31530
+    },
+    {
+      "epoch": 0.8624437729864235,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.3797263579282171e-06,
+      "loss": 0.6843,
+      "step": 31540
+    },
+    {
+      "epoch": 0.8627172174293488,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.3769844534013326e-06,
+      "loss": 0.6709,
+      "step": 31550
+    },
+    {
+      "epoch": 0.8629906618722741,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.374242548874448e-06,
+      "loss": 0.6906,
+      "step": 31560
+    },
+    {
+      "epoch": 0.8632641063151995,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.371500644347564e-06,
+      "loss": 0.7971,
+      "step": 31570
+    },
+    {
+      "epoch": 0.8635375507581248,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.3687587398206795e-06,
+      "loss": 0.6573,
+      "step": 31580
+    },
+    {
+      "epoch": 0.86381099520105,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.3660168352937952e-06,
+      "loss": 0.6587,
+      "step": 31590
+    },
+    {
+      "epoch": 0.8640844396439753,
+      "grad_norm": 2.875,
+      "learning_rate": 1.3632749307669107e-06,
+      "loss": 0.7308,
+      "step": 31600
+    },
+    {
+      "epoch": 0.8643578840869006,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.3605330262400266e-06,
+      "loss": 0.6517,
+      "step": 31610
+    },
+    {
+      "epoch": 0.8646313285298259,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.357791121713142e-06,
+      "loss": 0.7239,
+      "step": 31620
+    },
+    {
+      "epoch": 0.8649047729727513,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.3550492171862577e-06,
+      "loss": 0.6933,
+      "step": 31630
+    },
+    {
+      "epoch": 0.8651782174156766,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3523073126593732e-06,
+      "loss": 0.6891,
+      "step": 31640
+    },
+    {
+      "epoch": 0.8654516618586019,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.349565408132489e-06,
+      "loss": 0.7191,
+      "step": 31650
+    },
+    {
+      "epoch": 0.8657251063015272,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.3468235036056046e-06,
+      "loss": 0.7518,
+      "step": 31660
+    },
+    {
+      "epoch": 0.8659985507444525,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.34408159907872e-06,
+      "loss": 0.6219,
+      "step": 31670
+    },
+    {
+      "epoch": 0.8662719951873779,
+      "grad_norm": 3.125,
+      "learning_rate": 1.3413396945518358e-06,
+      "loss": 0.7036,
+      "step": 31680
+    },
+    {
+      "epoch": 0.8665454396303031,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.3385977900249515e-06,
+      "loss": 0.5871,
+      "step": 31690
+    },
+    {
+      "epoch": 0.8668188840732284,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.3358558854980672e-06,
+      "loss": 0.6961,
+      "step": 31700
+    },
+    {
+      "epoch": 0.8670923285161537,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.3331139809711827e-06,
+      "loss": 0.6831,
+      "step": 31710
+    },
+    {
+      "epoch": 0.867365772959079,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.3303720764442984e-06,
+      "loss": 0.6597,
+      "step": 31720
+    },
+    {
+      "epoch": 0.8676392174020043,
+      "grad_norm": 3.296875,
+      "learning_rate": 1.327630171917414e-06,
+      "loss": 0.727,
+      "step": 31730
+    },
+    {
+      "epoch": 0.8679126618449297,
+      "grad_norm": 3.390625,
+      "learning_rate": 1.3248882673905295e-06,
+      "loss": 0.6901,
+      "step": 31740
+    },
+    {
+      "epoch": 0.868186106287855,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.3221463628636452e-06,
+      "loss": 0.6138,
+      "step": 31750
+    },
+    {
+      "epoch": 0.8684595507307803,
+      "grad_norm": 3.125,
+      "learning_rate": 1.3194044583367607e-06,
+      "loss": 0.7354,
+      "step": 31760
+    },
+    {
+      "epoch": 0.8687329951737056,
+      "grad_norm": 3.28125,
+      "learning_rate": 1.3166625538098766e-06,
+      "loss": 0.7129,
+      "step": 31770
+    },
+    {
+      "epoch": 0.8690064396166309,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.313920649282992e-06,
+      "loss": 0.6773,
+      "step": 31780
+    },
+    {
+      "epoch": 0.8692798840595563,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.3111787447561078e-06,
+      "loss": 0.6331,
+      "step": 31790
+    },
+    {
+      "epoch": 0.8695533285024815,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.3084368402292233e-06,
+      "loss": 0.6363,
+      "step": 31800
+    },
+    {
+      "epoch": 0.8698267729454068,
+      "grad_norm": 2.5,
+      "learning_rate": 1.3056949357023392e-06,
+      "loss": 0.7217,
+      "step": 31810
+    },
+    {
+      "epoch": 0.8701002173883321,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.3029530311754547e-06,
+      "loss": 0.6198,
+      "step": 31820
+    },
+    {
+      "epoch": 0.8703736618312574,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.3002111266485701e-06,
+      "loss": 0.7566,
+      "step": 31830
+    },
+    {
+      "epoch": 0.8706471062741827,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.2974692221216858e-06,
+      "loss": 0.7346,
+      "step": 31840
+    },
+    {
+      "epoch": 0.870920550717108,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.2947273175948013e-06,
+      "loss": 0.7018,
+      "step": 31850
+    },
+    {
+      "epoch": 0.8711939951600334,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.2919854130679172e-06,
+      "loss": 0.657,
+      "step": 31860
+    },
+    {
+      "epoch": 0.8714674396029587,
+      "grad_norm": 2.40625,
+      "learning_rate": 1.2892435085410327e-06,
+      "loss": 0.6274,
+      "step": 31870
+    },
+    {
+      "epoch": 0.871740884045884,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.2865016040141484e-06,
+      "loss": 0.7189,
+      "step": 31880
+    },
+    {
+      "epoch": 0.8720143284888093,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.2837596994872639e-06,
+      "loss": 0.6678,
+      "step": 31890
+    },
+    {
+      "epoch": 0.8722877729317345,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.2810177949603798e-06,
+      "loss": 0.7377,
+      "step": 31900
+    },
+    {
+      "epoch": 0.8725612173746599,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2782758904334953e-06,
+      "loss": 0.6595,
+      "step": 31910
+    },
+    {
+      "epoch": 0.8728346618175852,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.2755339859066107e-06,
+      "loss": 0.6391,
+      "step": 31920
+    },
+    {
+      "epoch": 0.8731081062605105,
+      "grad_norm": 3.375,
+      "learning_rate": 1.2727920813797264e-06,
+      "loss": 0.7499,
+      "step": 31930
+    },
+    {
+      "epoch": 0.8733815507034358,
+      "grad_norm": 3.0,
+      "learning_rate": 1.270050176852842e-06,
+      "loss": 0.7519,
+      "step": 31940
+    },
+    {
+      "epoch": 0.8736549951463611,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.2673082723259578e-06,
+      "loss": 0.6991,
+      "step": 31950
+    },
+    {
+      "epoch": 0.8739284395892865,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.2645663677990733e-06,
+      "loss": 0.7351,
+      "step": 31960
+    },
+    {
+      "epoch": 0.8742018840322118,
+      "grad_norm": 3.4375,
+      "learning_rate": 1.261824463272189e-06,
+      "loss": 0.7028,
+      "step": 31970
+    },
+    {
+      "epoch": 0.8744753284751371,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2590825587453045e-06,
+      "loss": 0.6779,
+      "step": 31980
+    },
+    {
+      "epoch": 0.8747487729180624,
+      "grad_norm": 2.6875,
+      "learning_rate": 1.2563406542184204e-06,
+      "loss": 0.6255,
+      "step": 31990
+    },
+    {
+      "epoch": 0.8750222173609877,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.2535987496915359e-06,
+      "loss": 0.6857,
+      "step": 32000
+    },
+    {
+      "epoch": 0.8752956618039129,
+      "grad_norm": 2.75,
+      "learning_rate": 1.2508568451646514e-06,
+      "loss": 0.631,
+      "step": 32010
+    },
+    {
+      "epoch": 0.8755691062468383,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.248114940637767e-06,
+      "loss": 0.6383,
+      "step": 32020
+    },
+    {
+      "epoch": 0.8758425506897636,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.2453730361108827e-06,
+      "loss": 0.7158,
+      "step": 32030
+    },
+    {
+      "epoch": 0.8761159951326889,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2426311315839984e-06,
+      "loss": 0.6703,
+      "step": 32040
+    },
+    {
+      "epoch": 0.8763894395756142,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.2398892270571141e-06,
+      "loss": 0.6235,
+      "step": 32050
+    },
+    {
+      "epoch": 0.8766628840185395,
+      "grad_norm": 2.375,
+      "learning_rate": 1.2371473225302296e-06,
+      "loss": 0.6921,
+      "step": 32060
+    },
+    {
+      "epoch": 0.8769363284614649,
+      "grad_norm": 3.46875,
+      "learning_rate": 1.234405418003345e-06,
+      "loss": 0.8106,
+      "step": 32070
+    },
+    {
+      "epoch": 0.8772097729043902,
+      "grad_norm": 3.703125,
+      "learning_rate": 1.2316635134764608e-06,
+      "loss": 0.7295,
+      "step": 32080
+    },
+    {
+      "epoch": 0.8774832173473155,
+      "grad_norm": 2.625,
+      "learning_rate": 1.2289216089495765e-06,
+      "loss": 0.6944,
+      "step": 32090
+    },
+    {
+      "epoch": 0.8777566617902408,
+      "grad_norm": 2.5625,
+      "learning_rate": 1.2261797044226922e-06,
+      "loss": 0.6818,
+      "step": 32100
+    },
+    {
+      "epoch": 0.8780301062331661,
+      "grad_norm": 2.265625,
+      "learning_rate": 1.2234377998958077e-06,
+      "loss": 0.5916,
+      "step": 32110
+    },
+    {
+      "epoch": 0.8783035506760913,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.2206958953689234e-06,
+      "loss": 0.6403,
+      "step": 32120
+    },
+    {
+      "epoch": 0.8785769951190167,
+      "grad_norm": 3.0,
+      "learning_rate": 1.217953990842039e-06,
+      "loss": 0.6867,
+      "step": 32130
+    },
+    {
+      "epoch": 0.878850439561942,
+      "grad_norm": 3.25,
+      "learning_rate": 1.2152120863151547e-06,
+      "loss": 0.8073,
+      "step": 32140
+    },
+    {
+      "epoch": 0.8791238840048673,
+      "grad_norm": 2.578125,
+      "learning_rate": 1.2124701817882702e-06,
+      "loss": 0.6828,
+      "step": 32150
+    },
+    {
+      "epoch": 0.8793973284477926,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.209728277261386e-06,
+      "loss": 0.6987,
+      "step": 32160
+    },
+    {
+      "epoch": 0.8796707728907179,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.2069863727345014e-06,
+      "loss": 0.6322,
+      "step": 32170
+    },
+    {
+      "epoch": 0.8799442173336433,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.204244468207617e-06,
+      "loss": 0.7233,
+      "step": 32180
+    },
+    {
+      "epoch": 0.8802176617765686,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.2015025636807328e-06,
+      "loss": 0.7719,
+      "step": 32190
+    },
+    {
+      "epoch": 0.8804911062194939,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.1987606591538485e-06,
+      "loss": 0.6685,
+      "step": 32200
+    },
+    {
+      "epoch": 0.8807645506624192,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.196018754626964e-06,
+      "loss": 0.7149,
+      "step": 32210
+    },
+    {
+      "epoch": 0.8810379951053444,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.1932768501000797e-06,
+      "loss": 0.6962,
+      "step": 32220
+    },
+    {
+      "epoch": 0.8813114395482697,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.1905349455731953e-06,
+      "loss": 0.6712,
+      "step": 32230
+    },
+    {
+      "epoch": 0.8815848839911951,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.1877930410463108e-06,
+      "loss": 0.7176,
+      "step": 32240
+    },
+    {
+      "epoch": 0.8818583284341204,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.1850511365194265e-06,
+      "loss": 0.6493,
+      "step": 32250
+    },
+    {
+      "epoch": 0.8821317728770457,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.182309231992542e-06,
+      "loss": 0.7206,
+      "step": 32260
+    },
+    {
+      "epoch": 0.882405217319971,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.1795673274656577e-06,
+      "loss": 0.6704,
+      "step": 32270
+    },
+    {
+      "epoch": 0.8826786617628963,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.1768254229387734e-06,
+      "loss": 0.6022,
+      "step": 32280
+    },
+    {
+      "epoch": 0.8829521062058217,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.174083518411889e-06,
+      "loss": 0.7023,
+      "step": 32290
+    },
+    {
+      "epoch": 0.883225550648747,
+      "grad_norm": 2.328125,
+      "learning_rate": 1.1713416138850046e-06,
+      "loss": 0.684,
+      "step": 32300
+    },
+    {
+      "epoch": 0.8834989950916723,
+      "grad_norm": 2.46875,
+      "learning_rate": 1.1685997093581203e-06,
+      "loss": 0.718,
+      "step": 32310
+    },
+    {
+      "epoch": 0.8837724395345976,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.165857804831236e-06,
+      "loss": 0.6877,
+      "step": 32320
+    },
+    {
+      "epoch": 0.8840458839775228,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.1631159003043516e-06,
+      "loss": 0.6762,
+      "step": 32330
+    },
+    {
+      "epoch": 0.8843193284204481,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.1603739957774671e-06,
+      "loss": 0.6931,
+      "step": 32340
+    },
+    {
+      "epoch": 0.8845927728633735,
+      "grad_norm": 3.34375,
+      "learning_rate": 1.1576320912505826e-06,
+      "loss": 0.6549,
+      "step": 32350
+    },
+    {
+      "epoch": 0.8848662173062988,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.1548901867236983e-06,
+      "loss": 0.7437,
+      "step": 32360
+    },
+    {
+      "epoch": 0.8851396617492241,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.152148282196814e-06,
+      "loss": 0.6882,
+      "step": 32370
+    },
+    {
+      "epoch": 0.8854131061921494,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.1494063776699297e-06,
+      "loss": 0.7683,
+      "step": 32380
+    },
+    {
+      "epoch": 0.8856865506350747,
+      "grad_norm": 3.109375,
+      "learning_rate": 1.1466644731430452e-06,
+      "loss": 0.6753,
+      "step": 32390
+    },
+    {
+      "epoch": 0.885959995078,
+      "grad_norm": 2.625,
+      "learning_rate": 1.1439225686161609e-06,
+      "loss": 0.5924,
+      "step": 32400
+    },
+    {
+      "epoch": 0.8862334395209254,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.1411806640892766e-06,
+      "loss": 0.7129,
+      "step": 32410
+    },
+    {
+      "epoch": 0.8865068839638507,
+      "grad_norm": 2.5,
+      "learning_rate": 1.1384387595623923e-06,
+      "loss": 0.7051,
+      "step": 32420
+    },
+    {
+      "epoch": 0.886780328406776,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.1356968550355077e-06,
+      "loss": 0.6628,
+      "step": 32430
+    },
+    {
+      "epoch": 0.8870537728497012,
+      "grad_norm": 2.484375,
+      "learning_rate": 1.1329549505086234e-06,
+      "loss": 0.6724,
+      "step": 32440
+    },
+    {
+      "epoch": 0.8873272172926265,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.130213045981739e-06,
+      "loss": 0.7243,
+      "step": 32450
+    },
+    {
+      "epoch": 0.8876006617355519,
+      "grad_norm": 2.703125,
+      "learning_rate": 1.1274711414548546e-06,
+      "loss": 0.7248,
+      "step": 32460
+    },
+    {
+      "epoch": 0.8878741061784772,
+      "grad_norm": 2.421875,
+      "learning_rate": 1.1247292369279703e-06,
+      "loss": 0.6673,
+      "step": 32470
+    },
+    {
+      "epoch": 0.8881475506214025,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.121987332401086e-06,
+      "loss": 0.7378,
+      "step": 32480
+    },
+    {
+      "epoch": 0.8884209950643278,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.1192454278742015e-06,
+      "loss": 0.667,
+      "step": 32490
+    },
+    {
+      "epoch": 0.8886944395072531,
+      "grad_norm": 3.21875,
+      "learning_rate": 1.1165035233473172e-06,
+      "loss": 0.6808,
+      "step": 32500
+    },
+    {
+      "epoch": 0.8889678839501785,
+      "grad_norm": 2.53125,
+      "learning_rate": 1.1137616188204329e-06,
+      "loss": 0.643,
+      "step": 32510
+    },
+    {
+      "epoch": 0.8892413283931038,
+      "grad_norm": 3.1875,
+      "learning_rate": 1.1110197142935483e-06,
+      "loss": 0.6712,
+      "step": 32520
+    },
+    {
+      "epoch": 0.8895147728360291,
+      "grad_norm": 2.875,
+      "learning_rate": 1.108277809766664e-06,
+      "loss": 0.7407,
+      "step": 32530
+    },
+    {
+      "epoch": 0.8897882172789543,
+      "grad_norm": 3.125,
+      "learning_rate": 1.1055359052397795e-06,
+      "loss": 0.6496,
+      "step": 32540
+    },
+    {
+      "epoch": 0.8900616617218796,
+      "grad_norm": 3.0,
+      "learning_rate": 1.1027940007128952e-06,
+      "loss": 0.7009,
+      "step": 32550
+    },
+    {
+      "epoch": 0.8903351061648049,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.100052096186011e-06,
+      "loss": 0.7813,
+      "step": 32560
+    },
+    {
+      "epoch": 0.8906085506077303,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.0973101916591266e-06,
+      "loss": 0.6824,
+      "step": 32570
+    },
+    {
+      "epoch": 0.8908819950506556,
+      "grad_norm": 3.53125,
+      "learning_rate": 1.094568287132242e-06,
+      "loss": 0.6851,
+      "step": 32580
+    },
+    {
+      "epoch": 0.8911554394935809,
+      "grad_norm": 2.5,
+      "learning_rate": 1.0918263826053578e-06,
+      "loss": 0.7132,
+      "step": 32590
+    },
+    {
+      "epoch": 0.8914288839365062,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0890844780784735e-06,
+      "loss": 0.6753,
+      "step": 32600
+    },
+    {
+      "epoch": 0.8917023283794315,
+      "grad_norm": 2.390625,
+      "learning_rate": 1.0863425735515892e-06,
+      "loss": 0.7204,
+      "step": 32610
+    },
+    {
+      "epoch": 0.8919757728223568,
+      "grad_norm": 3.015625,
+      "learning_rate": 1.0836006690247046e-06,
+      "loss": 0.677,
+      "step": 32620
+    },
+    {
+      "epoch": 0.8922492172652822,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.0808587644978201e-06,
+      "loss": 0.6179,
+      "step": 32630
+    },
+    {
+      "epoch": 0.8925226617082075,
+      "grad_norm": 2.75,
+      "learning_rate": 1.0781168599709358e-06,
+      "loss": 0.7296,
+      "step": 32640
+    },
+    {
+      "epoch": 0.8927961061511327,
+      "grad_norm": 3.40625,
+      "learning_rate": 1.0753749554440515e-06,
+      "loss": 0.7376,
+      "step": 32650
+    },
+    {
+      "epoch": 0.893069550594058,
+      "grad_norm": 2.375,
+      "learning_rate": 1.0726330509171672e-06,
+      "loss": 0.6736,
+      "step": 32660
+    },
+    {
+      "epoch": 0.8933429950369833,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.0698911463902827e-06,
+      "loss": 0.7426,
+      "step": 32670
+    },
+    {
+      "epoch": 0.8936164394799087,
+      "grad_norm": 2.609375,
+      "learning_rate": 1.0671492418633984e-06,
+      "loss": 0.6668,
+      "step": 32680
+    },
+    {
+      "epoch": 0.893889883922834,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.064407337336514e-06,
+      "loss": 0.7112,
+      "step": 32690
+    },
+    {
+      "epoch": 0.8941633283657593,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.0616654328096298e-06,
+      "loss": 0.7007,
+      "step": 32700
+    },
+    {
+      "epoch": 0.8944367728086846,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0589235282827453e-06,
+      "loss": 0.7086,
+      "step": 32710
+    },
+    {
+      "epoch": 0.8947102172516099,
+      "grad_norm": 2.75,
+      "learning_rate": 1.056181623755861e-06,
+      "loss": 0.5982,
+      "step": 32720
+    },
+    {
+      "epoch": 0.8949836616945352,
+      "grad_norm": 3.5,
+      "learning_rate": 1.0534397192289764e-06,
+      "loss": 0.6493,
+      "step": 32730
+    },
+    {
+      "epoch": 0.8952571061374606,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.0506978147020921e-06,
+      "loss": 0.7261,
+      "step": 32740
+    },
+    {
+      "epoch": 0.8955305505803858,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.0479559101752078e-06,
+      "loss": 0.6131,
+      "step": 32750
+    },
+    {
+      "epoch": 0.8958039950233111,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.0452140056483235e-06,
+      "loss": 0.6638,
+      "step": 32760
+    },
+    {
+      "epoch": 0.8960774394662364,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.042472101121439e-06,
+      "loss": 0.5816,
+      "step": 32770
+    },
+    {
+      "epoch": 0.8963508839091617,
+      "grad_norm": 2.9375,
+      "learning_rate": 1.0397301965945547e-06,
+      "loss": 0.705,
+      "step": 32780
+    },
+    {
+      "epoch": 0.896624328352087,
+      "grad_norm": 2.453125,
+      "learning_rate": 1.0369882920676704e-06,
+      "loss": 0.6765,
+      "step": 32790
+    },
+    {
+      "epoch": 0.8968977727950124,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.0342463875407859e-06,
+      "loss": 0.7313,
+      "step": 32800
+    },
+    {
+      "epoch": 0.8971712172379377,
+      "grad_norm": 2.828125,
+      "learning_rate": 1.0315044830139016e-06,
+      "loss": 0.7061,
+      "step": 32810
+    },
+    {
+      "epoch": 0.897444661680863,
+      "grad_norm": 3.171875,
+      "learning_rate": 1.028762578487017e-06,
+      "loss": 0.6994,
+      "step": 32820
+    },
+    {
+      "epoch": 0.8977181061237883,
+      "grad_norm": 2.75,
+      "learning_rate": 1.0260206739601327e-06,
+      "loss": 0.7142,
+      "step": 32830
+    },
+    {
+      "epoch": 0.8979915505667136,
+      "grad_norm": 3.0,
+      "learning_rate": 1.0232787694332484e-06,
+      "loss": 0.7611,
+      "step": 32840
+    },
+    {
+      "epoch": 0.898264995009639,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0205368649063641e-06,
+      "loss": 0.6377,
+      "step": 32850
+    },
+    {
+      "epoch": 0.8985384394525642,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.0177949603794796e-06,
+      "loss": 0.666,
+      "step": 32860
+    },
+    {
+      "epoch": 0.8988118838954895,
+      "grad_norm": 2.96875,
+      "learning_rate": 1.0150530558525953e-06,
+      "loss": 0.6963,
+      "step": 32870
+    },
+    {
+      "epoch": 0.8990853283384148,
+      "grad_norm": 3.0625,
+      "learning_rate": 1.012311151325711e-06,
+      "loss": 0.7274,
+      "step": 32880
+    },
+    {
+      "epoch": 0.8993587727813401,
+      "grad_norm": 2.84375,
+      "learning_rate": 1.0095692467988265e-06,
+      "loss": 0.712,
+      "step": 32890
+    },
+    {
+      "epoch": 0.8996322172242655,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.0068273422719422e-06,
+      "loss": 0.6751,
+      "step": 32900
+    },
+    {
+      "epoch": 0.8999056616671908,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.0040854377450579e-06,
+      "loss": 0.6699,
+      "step": 32910
+    },
+    {
+      "epoch": 0.9001791061101161,
+      "grad_norm": 3.203125,
+      "learning_rate": 1.0013435332181733e-06,
+      "loss": 0.6858,
+      "step": 32920
+    },
+    {
+      "epoch": 0.9004525505530414,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.98601628691289e-07,
+      "loss": 0.7005,
+      "step": 32930
+    },
+    {
+      "epoch": 0.9007259949959667,
+      "grad_norm": 3.046875,
+      "learning_rate": 9.958597241644047e-07,
+      "loss": 0.6379,
+      "step": 32940
+    },
+    {
+      "epoch": 0.900999439438892,
+      "grad_norm": 3.03125,
+      "learning_rate": 9.931178196375204e-07,
+      "loss": 0.6823,
+      "step": 32950
+    },
+    {
+      "epoch": 0.9012728838818174,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.90375915110636e-07,
+      "loss": 0.7349,
+      "step": 32960
+    },
+    {
+      "epoch": 0.9015463283247426,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.876340105837516e-07,
+      "loss": 0.6317,
+      "step": 32970
+    },
+    {
+      "epoch": 0.9018197727676679,
+      "grad_norm": 2.875,
+      "learning_rate": 9.848921060568673e-07,
+      "loss": 0.7727,
+      "step": 32980
+    },
+    {
+      "epoch": 0.9020932172105932,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.821502015299828e-07,
+      "loss": 0.6228,
+      "step": 32990
+    },
+    {
+      "epoch": 0.9023666616535185,
+      "grad_norm": 3.078125,
+      "learning_rate": 9.794082970030985e-07,
+      "loss": 0.656,
+      "step": 33000
+    },
+    {
+      "epoch": 0.9026401060964439,
+      "grad_norm": 2.5625,
+      "learning_rate": 9.76666392476214e-07,
+      "loss": 0.6878,
+      "step": 33010
+    },
+    {
+      "epoch": 0.9029135505393692,
+      "grad_norm": 2.734375,
+      "learning_rate": 9.739244879493296e-07,
+      "loss": 0.705,
+      "step": 33020
+    },
+    {
+      "epoch": 0.9031869949822945,
+      "grad_norm": 2.578125,
+      "learning_rate": 9.711825834224453e-07,
+      "loss": 0.7055,
+      "step": 33030
+    },
+    {
+      "epoch": 0.9034604394252198,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.68440678895561e-07,
+      "loss": 0.6558,
+      "step": 33040
+    },
+    {
+      "epoch": 0.9037338838681451,
+      "grad_norm": 2.8125,
+      "learning_rate": 9.656987743686765e-07,
+      "loss": 0.7187,
+      "step": 33050
+    },
+    {
+      "epoch": 0.9040073283110704,
+      "grad_norm": 2.9375,
+      "learning_rate": 9.629568698417922e-07,
+      "loss": 0.6986,
+      "step": 33060
+    },
+    {
+      "epoch": 0.9042807727539957,
+      "grad_norm": 2.875,
+      "learning_rate": 9.60214965314908e-07,
+      "loss": 0.6936,
+      "step": 33070
+    },
+    {
+      "epoch": 0.904554217196921,
+      "grad_norm": 3.0,
+      "learning_rate": 9.574730607880234e-07,
+      "loss": 0.6345,
+      "step": 33080
+    },
+    {
+      "epoch": 0.9048276616398463,
+      "grad_norm": 2.953125,
+      "learning_rate": 9.54731156261139e-07,
+      "loss": 0.7363,
+      "step": 33090
+    },
+    {
+      "epoch": 0.9051011060827716,
+      "grad_norm": 2.625,
+      "learning_rate": 9.519892517342547e-07,
+      "loss": 0.6912,
+      "step": 33100
+    },
+    {
+      "epoch": 0.9053745505256969,
+      "grad_norm": 2.484375,
+      "learning_rate": 9.492473472073703e-07,
+      "loss": 0.7091,
+      "step": 33110
+    },
+    {
+      "epoch": 0.9056479949686222,
+      "grad_norm": 3.0,
+      "learning_rate": 9.46505442680486e-07,
+      "loss": 0.7431,
+      "step": 33120
+    },
+    {
+      "epoch": 0.9059214394115476,
+      "grad_norm": 2.59375,
+      "learning_rate": 9.437635381536015e-07,
+      "loss": 0.686,
+      "step": 33130
+    },
+    {
+      "epoch": 0.9061948838544729,
+      "grad_norm": 7.8125,
+      "learning_rate": 9.410216336267172e-07,
+      "loss": 0.6041,
+      "step": 33140
+    },
+    {
+      "epoch": 0.9064683282973982,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.382797290998328e-07,
+      "loss": 0.6435,
+      "step": 33150
+    },
+    {
+      "epoch": 0.9067417727403235,
+      "grad_norm": 2.796875,
+      "learning_rate": 9.355378245729485e-07,
+      "loss": 0.637,
+      "step": 33160
+    },
+    {
+      "epoch": 0.9070152171832488,
+      "grad_norm": 2.40625,
+      "learning_rate": 9.32795920046064e-07,
+      "loss": 0.6794,
+      "step": 33170
+    },
+    {
+      "epoch": 0.907288661626174,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.300540155191797e-07,
+      "loss": 0.6549,
+      "step": 33180
+    },
+    {
+      "epoch": 0.9075621060690994,
+      "grad_norm": 3.328125,
+      "learning_rate": 9.273121109922953e-07,
+      "loss": 0.7215,
+      "step": 33190
+    },
+    {
+      "epoch": 0.9078355505120247,
+      "grad_norm": 3.28125,
+      "learning_rate": 9.24570206465411e-07,
+      "loss": 0.7975,
+      "step": 33200
+    },
+    {
+      "epoch": 0.90810899495495,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.218283019385266e-07,
+      "loss": 0.7061,
+      "step": 33210
+    },
+    {
+      "epoch": 0.9083824393978753,
+      "grad_norm": 3.125,
+      "learning_rate": 9.190863974116422e-07,
+      "loss": 0.7391,
+      "step": 33220
+    },
+    {
+      "epoch": 0.9086558838408006,
+      "grad_norm": 2.703125,
+      "learning_rate": 9.163444928847578e-07,
+      "loss": 0.6997,
+      "step": 33230
+    },
+    {
+      "epoch": 0.908929328283726,
+      "grad_norm": 2.984375,
+      "learning_rate": 9.136025883578735e-07,
+      "loss": 0.6592,
+      "step": 33240
+    },
+    {
+      "epoch": 0.9092027727266513,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.108606838309891e-07,
+      "loss": 0.755,
+      "step": 33250
+    },
+    {
+      "epoch": 0.9094762171695766,
+      "grad_norm": 2.671875,
+      "learning_rate": 9.081187793041046e-07,
+      "loss": 0.6718,
+      "step": 33260
+    },
+    {
+      "epoch": 0.9097496616125019,
+      "grad_norm": 2.890625,
+      "learning_rate": 9.053768747772203e-07,
+      "loss": 0.7111,
+      "step": 33270
+    },
+    {
+      "epoch": 0.9100231060554271,
+      "grad_norm": 3.109375,
+      "learning_rate": 9.026349702503359e-07,
+      "loss": 0.7134,
+      "step": 33280
+    },
+    {
+      "epoch": 0.9102965504983525,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.998930657234516e-07,
+      "loss": 0.6471,
+      "step": 33290
+    },
+    {
+      "epoch": 0.9105699949412778,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.971511611965672e-07,
+      "loss": 0.663,
+      "step": 33300
+    },
+    {
+      "epoch": 0.9108434393842031,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.944092566696829e-07,
+      "loss": 0.7223,
+      "step": 33310
+    },
+    {
+      "epoch": 0.9111168838271284,
+      "grad_norm": 3.171875,
+      "learning_rate": 8.916673521427984e-07,
+      "loss": 0.7169,
+      "step": 33320
+    },
+    {
+      "epoch": 0.9113903282700537,
+      "grad_norm": 3.15625,
+      "learning_rate": 8.889254476159141e-07,
+      "loss": 0.8168,
+      "step": 33330
+    },
+    {
+      "epoch": 0.911663772712979,
+      "grad_norm": 2.53125,
+      "learning_rate": 8.861835430890297e-07,
+      "loss": 0.7459,
+      "step": 33340
+    },
+    {
+      "epoch": 0.9119372171559044,
+      "grad_norm": 2.890625,
+      "learning_rate": 8.834416385621454e-07,
+      "loss": 0.6906,
+      "step": 33350
+    },
+    {
+      "epoch": 0.9122106615988297,
+      "grad_norm": 2.78125,
+      "learning_rate": 8.806997340352609e-07,
+      "loss": 0.6756,
+      "step": 33360
+    },
+    {
+      "epoch": 0.912484106041755,
+      "grad_norm": 2.75,
+      "learning_rate": 8.779578295083766e-07,
+      "loss": 0.7747,
+      "step": 33370
+    },
+    {
+      "epoch": 0.9127575504846803,
+      "grad_norm": 2.828125,
+      "learning_rate": 8.752159249814922e-07,
+      "loss": 0.6701,
+      "step": 33380
+    },
+    {
+      "epoch": 0.9130309949276055,
+      "grad_norm": 2.984375,
+      "learning_rate": 8.724740204546079e-07,
+      "loss": 0.6761,
+      "step": 33390
+    },
+    {
+      "epoch": 0.9133044393705309,
+      "grad_norm": 2.265625,
+      "learning_rate": 8.697321159277235e-07,
+      "loss": 0.6425,
+      "step": 33400
+    },
+    {
+      "epoch": 0.9135778838134562,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.669902114008392e-07,
+      "loss": 0.6261,
+      "step": 33410
+    },
+    {
+      "epoch": 0.9138513282563815,
+      "grad_norm": 2.578125,
+      "learning_rate": 8.642483068739547e-07,
+      "loss": 0.6579,
+      "step": 33420
+    },
+    {
+      "epoch": 0.9141247726993068,
+      "grad_norm": 2.546875,
+      "learning_rate": 8.615064023470704e-07,
+      "loss": 0.7225,
+      "step": 33430
+    },
+    {
+      "epoch": 0.9143982171422321,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.58764497820186e-07,
+      "loss": 0.636,
+      "step": 33440
+    },
+    {
+      "epoch": 0.9146716615851574,
+      "grad_norm": 3.234375,
+      "learning_rate": 8.560225932933015e-07,
+      "loss": 0.7805,
+      "step": 33450
+    },
+    {
+      "epoch": 0.9149451060280828,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.532806887664172e-07,
+      "loss": 0.7072,
+      "step": 33460
+    },
+    {
+      "epoch": 0.9152185504710081,
+      "grad_norm": 2.28125,
+      "learning_rate": 8.505387842395328e-07,
+      "loss": 0.6916,
+      "step": 33470
+    },
+    {
+      "epoch": 0.9154919949139334,
+      "grad_norm": 2.46875,
+      "learning_rate": 8.477968797126485e-07,
+      "loss": 0.6482,
+      "step": 33480
+    },
+    {
+      "epoch": 0.9157654393568587,
+      "grad_norm": 2.59375,
+      "learning_rate": 8.450549751857641e-07,
+      "loss": 0.6627,
+      "step": 33490
+    },
+    {
+      "epoch": 0.9160388837997839,
+      "grad_norm": 2.9375,
+      "learning_rate": 8.423130706588798e-07,
+      "loss": 0.6866,
+      "step": 33500
+    },
+    {
+      "epoch": 0.9163123282427093,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.395711661319954e-07,
+      "loss": 0.6808,
+      "step": 33510
+    },
+    {
+      "epoch": 0.9165857726856346,
+      "grad_norm": 2.796875,
+      "learning_rate": 8.36829261605111e-07,
+      "loss": 0.6833,
+      "step": 33520
+    },
+    {
+      "epoch": 0.9168592171285599,
+      "grad_norm": 3.28125,
+      "learning_rate": 8.340873570782266e-07,
+      "loss": 0.7627,
+      "step": 33530
+    },
+    {
+      "epoch": 0.9171326615714852,
+      "grad_norm": 3.1875,
+      "learning_rate": 8.313454525513421e-07,
+      "loss": 0.7339,
+      "step": 33540
+    },
+    {
+      "epoch": 0.9174061060144105,
+      "grad_norm": 3.671875,
+      "learning_rate": 8.286035480244578e-07,
+      "loss": 0.6587,
+      "step": 33550
+    },
+    {
+      "epoch": 0.9176795504573358,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.258616434975734e-07,
+      "loss": 0.5575,
+      "step": 33560
+    },
+    {
+      "epoch": 0.9179529949002612,
+      "grad_norm": 2.921875,
+      "learning_rate": 8.231197389706891e-07,
+      "loss": 0.7591,
+      "step": 33570
+    },
+    {
+      "epoch": 0.9182264393431865,
+      "grad_norm": 2.671875,
+      "learning_rate": 8.203778344438047e-07,
+      "loss": 0.6615,
+      "step": 33580
+    },
+    {
+      "epoch": 0.9184998837861118,
+      "grad_norm": 2.765625,
+      "learning_rate": 8.176359299169204e-07,
+      "loss": 0.697,
+      "step": 33590
+    },
+    {
+      "epoch": 0.918773328229037,
+      "grad_norm": 3.0625,
+      "learning_rate": 8.14894025390036e-07,
+      "loss": 0.6257,
+      "step": 33600
+    },
+    {
+      "epoch": 0.9190467726719623,
+      "grad_norm": 2.703125,
+      "learning_rate": 8.121521208631517e-07,
+      "loss": 0.6974,
+      "step": 33610
+    },
+    {
+      "epoch": 0.9193202171148877,
+      "grad_norm": 2.15625,
+      "learning_rate": 8.094102163362672e-07,
+      "loss": 0.6564,
+      "step": 33620
+    },
+    {
+      "epoch": 0.919593661557813,
+      "grad_norm": 3.109375,
+      "learning_rate": 8.066683118093829e-07,
+      "loss": 0.7374,
+      "step": 33630
+    },
+    {
+      "epoch": 0.9198671060007383,
+      "grad_norm": 2.375,
+      "learning_rate": 8.039264072824984e-07,
+      "loss": 0.6139,
+      "step": 33640
+    },
+    {
+      "epoch": 0.9201405504436636,
+      "grad_norm": 3.203125,
+      "learning_rate": 8.011845027556141e-07,
+      "loss": 0.7125,
+      "step": 33650
+    },
+    {
+      "epoch": 0.9204139948865889,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.984425982287297e-07,
+      "loss": 0.6622,
+      "step": 33660
+    },
+    {
+      "epoch": 0.9206874393295142,
+      "grad_norm": 3.03125,
+      "learning_rate": 7.957006937018454e-07,
+      "loss": 0.6447,
+      "step": 33670
+    },
+    {
+      "epoch": 0.9209608837724396,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.92958789174961e-07,
+      "loss": 0.6918,
+      "step": 33680
+    },
+    {
+      "epoch": 0.9212343282153649,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.902168846480767e-07,
+      "loss": 0.7583,
+      "step": 33690
+    },
+    {
+      "epoch": 0.9215077726582902,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.874749801211923e-07,
+      "loss": 0.7628,
+      "step": 33700
+    },
+    {
+      "epoch": 0.9217812171012154,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.84733075594308e-07,
+      "loss": 0.6778,
+      "step": 33710
+    },
+    {
+      "epoch": 0.9220546615441407,
+      "grad_norm": 2.703125,
+      "learning_rate": 7.819911710674235e-07,
+      "loss": 0.7393,
+      "step": 33720
+    },
+    {
+      "epoch": 0.922328105987066,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.79249266540539e-07,
+      "loss": 0.6959,
+      "step": 33730
+    },
+    {
+      "epoch": 0.9226015504299914,
+      "grad_norm": 2.984375,
+      "learning_rate": 7.765073620136547e-07,
+      "loss": 0.7236,
+      "step": 33740
+    },
+    {
+      "epoch": 0.9228749948729167,
+      "grad_norm": 2.59375,
+      "learning_rate": 7.737654574867703e-07,
+      "loss": 0.6686,
+      "step": 33750
+    },
+    {
+      "epoch": 0.923148439315842,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.71023552959886e-07,
+      "loss": 0.6145,
+      "step": 33760
+    },
+    {
+      "epoch": 0.9234218837587673,
+      "grad_norm": 2.78125,
+      "learning_rate": 7.682816484330016e-07,
+      "loss": 0.6953,
+      "step": 33770
+    },
+    {
+      "epoch": 0.9236953282016926,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.655397439061173e-07,
+      "loss": 0.7735,
+      "step": 33780
+    },
+    {
+      "epoch": 0.923968772644618,
+      "grad_norm": 2.921875,
+      "learning_rate": 7.627978393792329e-07,
+      "loss": 0.6689,
+      "step": 33790
+    },
+    {
+      "epoch": 0.9242422170875433,
+      "grad_norm": 2.828125,
+      "learning_rate": 7.600559348523486e-07,
+      "loss": 0.6683,
+      "step": 33800
+    },
+    {
+      "epoch": 0.9245156615304686,
+      "grad_norm": 2.6875,
+      "learning_rate": 7.573140303254642e-07,
+      "loss": 0.6512,
+      "step": 33810
+    },
+    {
+      "epoch": 0.9247891059733938,
+      "grad_norm": 3.0,
+      "learning_rate": 7.545721257985797e-07,
+      "loss": 0.7211,
+      "step": 33820
+    },
+    {
+      "epoch": 0.9250625504163191,
+      "grad_norm": 3.015625,
+      "learning_rate": 7.518302212716953e-07,
+      "loss": 0.7131,
+      "step": 33830
+    },
+    {
+      "epoch": 0.9253359948592444,
+      "grad_norm": 2.796875,
+      "learning_rate": 7.49088316744811e-07,
+      "loss": 0.6708,
+      "step": 33840
+    },
+    {
+      "epoch": 0.9256094393021698,
+      "grad_norm": 3.84375,
+      "learning_rate": 7.463464122179266e-07,
+      "loss": 0.7524,
+      "step": 33850
+    },
+    {
+      "epoch": 0.9258828837450951,
+      "grad_norm": 3.15625,
+      "learning_rate": 7.436045076910423e-07,
+      "loss": 0.649,
+      "step": 33860
+    },
+    {
+      "epoch": 0.9261563281880204,
+      "grad_norm": 2.40625,
+      "learning_rate": 7.408626031641579e-07,
+      "loss": 0.6876,
+      "step": 33870
+    },
+    {
+      "epoch": 0.9264297726309457,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.381206986372736e-07,
+      "loss": 0.7787,
+      "step": 33880
+    },
+    {
+      "epoch": 0.926703217073871,
+      "grad_norm": 2.671875,
+      "learning_rate": 7.353787941103892e-07,
+      "loss": 0.692,
+      "step": 33890
+    },
+    {
+      "epoch": 0.9269766615167964,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.326368895835049e-07,
+      "loss": 0.6546,
+      "step": 33900
+    },
+    {
+      "epoch": 0.9272501059597217,
+      "grad_norm": 2.71875,
+      "learning_rate": 7.298949850566204e-07,
+      "loss": 0.7151,
+      "step": 33910
+    },
+    {
+      "epoch": 0.9275235504026469,
+      "grad_norm": 2.65625,
+      "learning_rate": 7.271530805297359e-07,
+      "loss": 0.7019,
+      "step": 33920
+    },
+    {
+      "epoch": 0.9277969948455722,
+      "grad_norm": 3.0625,
+      "learning_rate": 7.244111760028516e-07,
+      "loss": 0.7618,
+      "step": 33930
+    },
+    {
+      "epoch": 0.9280704392884975,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.216692714759672e-07,
+      "loss": 0.7042,
+      "step": 33940
+    },
+    {
+      "epoch": 0.9283438837314228,
+      "grad_norm": 3.296875,
+      "learning_rate": 7.189273669490829e-07,
+      "loss": 0.7013,
+      "step": 33950
+    },
+    {
+      "epoch": 0.9286173281743482,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.161854624221985e-07,
+      "loss": 0.7193,
+      "step": 33960
+    },
+    {
+      "epoch": 0.9288907726172735,
+      "grad_norm": 2.75,
+      "learning_rate": 7.134435578953142e-07,
+      "loss": 0.7073,
+      "step": 33970
+    },
+    {
+      "epoch": 0.9291642170601988,
+      "grad_norm": 2.859375,
+      "learning_rate": 7.107016533684298e-07,
+      "loss": 0.6908,
+      "step": 33980
+    },
+    {
+      "epoch": 0.9294376615031241,
+      "grad_norm": 3.078125,
+      "learning_rate": 7.079597488415455e-07,
+      "loss": 0.7555,
+      "step": 33990
+    },
+    {
+      "epoch": 0.9297111059460494,
+      "grad_norm": 3.046875,
+      "learning_rate": 7.052178443146611e-07,
+      "loss": 0.7599,
+      "step": 34000
+    },
+    {
+      "epoch": 0.9299845503889748,
+      "grad_norm": 2.734375,
+      "learning_rate": 7.024759397877765e-07,
+      "loss": 0.7087,
+      "step": 34010
+    },
+    {
+      "epoch": 0.9302579948319001,
+      "grad_norm": 2.78125,
+      "learning_rate": 6.997340352608922e-07,
+      "loss": 0.6177,
+      "step": 34020
+    },
+    {
+      "epoch": 0.9305314392748253,
+      "grad_norm": 3.078125,
+      "learning_rate": 6.969921307340078e-07,
+      "loss": 0.6574,
+      "step": 34030
+    },
+    {
+      "epoch": 0.9308048837177506,
+      "grad_norm": 3.3125,
+      "learning_rate": 6.942502262071235e-07,
+      "loss": 0.7505,
+      "step": 34040
+    },
+    {
+      "epoch": 0.9310783281606759,
+      "grad_norm": 2.65625,
+      "learning_rate": 6.915083216802391e-07,
+      "loss": 0.6506,
+      "step": 34050
+    },
+    {
+      "epoch": 0.9313517726036012,
+      "grad_norm": 2.8125,
+      "learning_rate": 6.887664171533548e-07,
+      "loss": 0.7158,
+      "step": 34060
+    },
+    {
+      "epoch": 0.9316252170465266,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.860245126264704e-07,
+      "loss": 0.7064,
+      "step": 34070
+    },
+    {
+      "epoch": 0.9318986614894519,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.832826080995861e-07,
+      "loss": 0.6666,
+      "step": 34080
+    },
+    {
+      "epoch": 0.9321721059323772,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.805407035727017e-07,
+      "loss": 0.7401,
+      "step": 34090
+    },
+    {
+      "epoch": 0.9324455503753025,
+      "grad_norm": 3.203125,
+      "learning_rate": 6.777987990458173e-07,
+      "loss": 0.7309,
+      "step": 34100
+    },
+    {
+      "epoch": 0.9327189948182278,
+      "grad_norm": 2.640625,
+      "learning_rate": 6.750568945189328e-07,
+      "loss": 0.5635,
+      "step": 34110
+    },
+    {
+      "epoch": 0.9329924392611532,
+      "grad_norm": 2.296875,
+      "learning_rate": 6.723149899920485e-07,
+      "loss": 0.7179,
+      "step": 34120
+    },
+    {
+      "epoch": 0.9332658837040784,
+      "grad_norm": 2.234375,
+      "learning_rate": 6.695730854651641e-07,
+      "loss": 0.5347,
+      "step": 34130
+    },
+    {
+      "epoch": 0.9335393281470037,
+      "grad_norm": 2.5,
+      "learning_rate": 6.668311809382798e-07,
+      "loss": 0.6241,
+      "step": 34140
+    },
+    {
+      "epoch": 0.933812772589929,
+      "grad_norm": 2.5,
+      "learning_rate": 6.640892764113954e-07,
+      "loss": 0.6597,
+      "step": 34150
+    },
+    {
+      "epoch": 0.9340862170328543,
+      "grad_norm": 2.390625,
+      "learning_rate": 6.613473718845111e-07,
+      "loss": 0.6212,
+      "step": 34160
+    },
+    {
+      "epoch": 0.9343596614757796,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.586054673576267e-07,
+      "loss": 0.7027,
+      "step": 34170
+    },
+    {
+      "epoch": 0.934633105918705,
+      "grad_norm": 2.890625,
+      "learning_rate": 6.558635628307424e-07,
+      "loss": 0.6964,
+      "step": 34180
+    },
+    {
+      "epoch": 0.9349065503616303,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.531216583038579e-07,
+      "loss": 0.7089,
+      "step": 34190
+    },
+    {
+      "epoch": 0.9351799948045556,
+      "grad_norm": 2.359375,
+      "learning_rate": 6.503797537769735e-07,
+      "loss": 0.6494,
+      "step": 34200
+    },
+    {
+      "epoch": 0.9354534392474809,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.476378492500892e-07,
+      "loss": 0.6194,
+      "step": 34210
+    },
+    {
+      "epoch": 0.9357268836904062,
+      "grad_norm": 3.1875,
+      "learning_rate": 6.448959447232047e-07,
+      "loss": 0.6999,
+      "step": 34220
+    },
+    {
+      "epoch": 0.9360003281333316,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.421540401963204e-07,
+      "loss": 0.6851,
+      "step": 34230
+    },
+    {
+      "epoch": 0.9362737725762568,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.39412135669436e-07,
+      "loss": 0.7426,
+      "step": 34240
+    },
+    {
+      "epoch": 0.9365472170191821,
+      "grad_norm": 2.609375,
+      "learning_rate": 6.366702311425517e-07,
+      "loss": 0.628,
+      "step": 34250
+    },
+    {
+      "epoch": 0.9368206614621074,
+      "grad_norm": 2.28125,
+      "learning_rate": 6.339283266156673e-07,
+      "loss": 0.6801,
+      "step": 34260
+    },
+    {
+      "epoch": 0.9370941059050327,
+      "grad_norm": 2.75,
+      "learning_rate": 6.31186422088783e-07,
+      "loss": 0.6944,
+      "step": 34270
+    },
+    {
+      "epoch": 0.937367550347958,
+      "grad_norm": 2.546875,
+      "learning_rate": 6.284445175618986e-07,
+      "loss": 0.7342,
+      "step": 34280
+    },
+    {
+      "epoch": 0.9376409947908834,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.257026130350141e-07,
+      "loss": 0.7032,
+      "step": 34290
+    },
+    {
+      "epoch": 0.9379144392338087,
+      "grad_norm": 2.796875,
+      "learning_rate": 6.229607085081299e-07,
+      "loss": 0.7003,
+      "step": 34300
+    },
+    {
+      "epoch": 0.938187883676734,
+      "grad_norm": 2.484375,
+      "learning_rate": 6.202188039812453e-07,
+      "loss": 0.6951,
+      "step": 34310
+    },
+    {
+      "epoch": 0.9384613281196593,
+      "grad_norm": 2.96875,
+      "learning_rate": 6.17476899454361e-07,
+      "loss": 0.6513,
+      "step": 34320
+    },
+    {
+      "epoch": 0.9387347725625846,
+      "grad_norm": 2.5625,
+      "learning_rate": 6.147349949274766e-07,
+      "loss": 0.7214,
+      "step": 34330
+    },
+    {
+      "epoch": 0.93900821700551,
+      "grad_norm": 2.71875,
+      "learning_rate": 6.119930904005923e-07,
+      "loss": 0.6893,
+      "step": 34340
+    },
+    {
+      "epoch": 0.9392816614484352,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.092511858737079e-07,
+      "loss": 0.6913,
+      "step": 34350
+    },
+    {
+      "epoch": 0.9395551058913605,
+      "grad_norm": 2.859375,
+      "learning_rate": 6.065092813468235e-07,
+      "loss": 0.7542,
+      "step": 34360
+    },
+    {
+      "epoch": 0.9398285503342858,
+      "grad_norm": 3.0,
+      "learning_rate": 6.037673768199392e-07,
+      "loss": 0.7434,
+      "step": 34370
+    },
+    {
+      "epoch": 0.9401019947772111,
+      "grad_norm": 2.671875,
+      "learning_rate": 6.010254722930548e-07,
+      "loss": 0.6511,
+      "step": 34380
+    },
+    {
+      "epoch": 0.9403754392201364,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.982835677661705e-07,
+      "loss": 0.6766,
+      "step": 34390
+    },
+    {
+      "epoch": 0.9406488836630618,
+      "grad_norm": 3.109375,
+      "learning_rate": 5.955416632392861e-07,
+      "loss": 0.6432,
+      "step": 34400
+    },
+    {
+      "epoch": 0.9409223281059871,
+      "grad_norm": 2.71875,
+      "learning_rate": 5.927997587124016e-07,
+      "loss": 0.7139,
+      "step": 34410
+    },
+    {
+      "epoch": 0.9411957725489124,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.900578541855173e-07,
+      "loss": 0.7345,
+      "step": 34420
+    },
+    {
+      "epoch": 0.9414692169918377,
+      "grad_norm": 2.84375,
+      "learning_rate": 5.873159496586329e-07,
+      "loss": 0.6788,
+      "step": 34430
+    },
+    {
+      "epoch": 0.941742661434763,
+      "grad_norm": 2.609375,
+      "learning_rate": 5.845740451317486e-07,
+      "loss": 0.6549,
+      "step": 34440
+    },
+    {
+      "epoch": 0.9420161058776882,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.818321406048642e-07,
+      "loss": 0.7004,
+      "step": 34450
+    },
+    {
+      "epoch": 0.9422895503206136,
+      "grad_norm": 2.828125,
+      "learning_rate": 5.790902360779798e-07,
+      "loss": 0.6704,
+      "step": 34460
+    },
+    {
+      "epoch": 0.9425629947635389,
+      "grad_norm": 3.15625,
+      "learning_rate": 5.763483315510955e-07,
+      "loss": 0.716,
+      "step": 34470
+    },
+    {
+      "epoch": 0.9428364392064642,
+      "grad_norm": 3.25,
+      "learning_rate": 5.736064270242111e-07,
+      "loss": 0.7607,
+      "step": 34480
+    },
+    {
+      "epoch": 0.9431098836493895,
+      "grad_norm": 3.3125,
+      "learning_rate": 5.708645224973267e-07,
+      "loss": 0.7615,
+      "step": 34490
+    },
+    {
+      "epoch": 0.9433833280923148,
+      "grad_norm": 2.6875,
+      "learning_rate": 5.681226179704423e-07,
+      "loss": 0.6662,
+      "step": 34500
+    },
+    {
+      "epoch": 0.9436567725352402,
+      "grad_norm": 3.0625,
+      "learning_rate": 5.65380713443558e-07,
+      "loss": 0.7541,
+      "step": 34510
+    },
+    {
+      "epoch": 0.9439302169781655,
+      "grad_norm": 2.875,
+      "learning_rate": 5.626388089166735e-07,
+      "loss": 0.7429,
+      "step": 34520
+    },
+    {
+      "epoch": 0.9442036614210908,
+      "grad_norm": 3.046875,
+      "learning_rate": 5.598969043897892e-07,
+      "loss": 0.6663,
+      "step": 34530
+    },
+    {
+      "epoch": 0.9444771058640161,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.571549998629048e-07,
+      "loss": 0.7608,
+      "step": 34540
+    },
+    {
+      "epoch": 0.9447505503069414,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.544130953360204e-07,
+      "loss": 0.7172,
+      "step": 34550
+    },
+    {
+      "epoch": 0.9450239947498666,
+      "grad_norm": 2.9375,
+      "learning_rate": 5.516711908091361e-07,
+      "loss": 0.6975,
+      "step": 34560
+    },
+    {
+      "epoch": 0.945297439192792,
+      "grad_norm": 2.875,
+      "learning_rate": 5.489292862822517e-07,
+      "loss": 0.6997,
+      "step": 34570
+    },
+    {
+      "epoch": 0.9455708836357173,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.461873817553674e-07,
+      "loss": 0.6904,
+      "step": 34580
+    },
+    {
+      "epoch": 0.9458443280786426,
+      "grad_norm": 2.734375,
+      "learning_rate": 5.43445477228483e-07,
+      "loss": 0.6558,
+      "step": 34590
+    },
+    {
+      "epoch": 0.9461177725215679,
+      "grad_norm": 2.953125,
+      "learning_rate": 5.407035727015986e-07,
+      "loss": 0.7171,
+      "step": 34600
+    },
+    {
+      "epoch": 0.9463912169644932,
+      "grad_norm": 2.8125,
+      "learning_rate": 5.379616681747143e-07,
+      "loss": 0.7054,
+      "step": 34610
+    },
+    {
+      "epoch": 0.9466646614074186,
+      "grad_norm": 3.125,
+      "learning_rate": 5.352197636478298e-07,
+      "loss": 0.7574,
+      "step": 34620
+    },
+    {
+      "epoch": 0.9469381058503439,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.324778591209454e-07,
+      "loss": 0.6892,
+      "step": 34630
+    },
+    {
+      "epoch": 0.9472115502932692,
+      "grad_norm": 2.921875,
+      "learning_rate": 5.29735954594061e-07,
+      "loss": 0.7158,
+      "step": 34640
+    },
+    {
+      "epoch": 0.9474849947361945,
+      "grad_norm": 3.09375,
+      "learning_rate": 5.269940500671767e-07,
+      "loss": 0.6707,
+      "step": 34650
+    },
+    {
+      "epoch": 0.9477584391791197,
+      "grad_norm": 2.578125,
+      "learning_rate": 5.242521455402923e-07,
+      "loss": 0.5821,
+      "step": 34660
+    },
+    {
+      "epoch": 0.948031883622045,
+      "grad_norm": 2.484375,
+      "learning_rate": 5.21510241013408e-07,
+      "loss": 0.6984,
+      "step": 34670
+    },
+    {
+      "epoch": 0.9483053280649704,
+      "grad_norm": 3.421875,
+      "learning_rate": 5.187683364865236e-07,
+      "loss": 0.7214,
+      "step": 34680
+    },
+    {
+      "epoch": 0.9485787725078957,
+      "grad_norm": 2.5,
+      "learning_rate": 5.160264319596392e-07,
+      "loss": 0.7149,
+      "step": 34690
+    },
+    {
+      "epoch": 0.948852216950821,
+      "grad_norm": 2.671875,
+      "learning_rate": 5.132845274327549e-07,
+      "loss": 0.6753,
+      "step": 34700
+    },
+    {
+      "epoch": 0.9491256613937463,
+      "grad_norm": 2.65625,
+      "learning_rate": 5.105426229058704e-07,
+      "loss": 0.6647,
+      "step": 34710
+    },
+    {
+      "epoch": 0.9493991058366716,
+      "grad_norm": 2.890625,
+      "learning_rate": 5.078007183789861e-07,
+      "loss": 0.6846,
+      "step": 34720
+    },
+    {
+      "epoch": 0.949672550279597,
+      "grad_norm": 2.765625,
+      "learning_rate": 5.050588138521017e-07,
+      "loss": 0.747,
+      "step": 34730
+    },
+    {
+      "epoch": 0.9499459947225223,
+      "grad_norm": 4.78125,
+      "learning_rate": 5.023169093252173e-07,
+      "loss": 0.6727,
+      "step": 34740
+    },
+    {
+      "epoch": 0.9502194391654476,
+      "grad_norm": 2.453125,
+      "learning_rate": 4.99575004798333e-07,
+      "loss": 0.6553,
+      "step": 34750
+    },
+    {
+      "epoch": 0.9504928836083729,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.968331002714486e-07,
+      "loss": 0.7261,
+      "step": 34760
+    },
+    {
+      "epoch": 0.9507663280512981,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.940911957445642e-07,
+      "loss": 0.693,
+      "step": 34770
+    },
+    {
+      "epoch": 0.9510397724942234,
+      "grad_norm": 3.0,
+      "learning_rate": 4.913492912176798e-07,
+      "loss": 0.6444,
+      "step": 34780
+    },
+    {
+      "epoch": 0.9513132169371488,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.886073866907955e-07,
+      "loss": 0.733,
+      "step": 34790
+    },
+    {
+      "epoch": 0.9515866613800741,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.858654821639111e-07,
+      "loss": 0.6775,
+      "step": 34800
+    },
+    {
+      "epoch": 0.9518601058229994,
+      "grad_norm": 3.4375,
+      "learning_rate": 4.831235776370268e-07,
+      "loss": 0.6965,
+      "step": 34810
+    },
+    {
+      "epoch": 0.9521335502659247,
+      "grad_norm": 3.0,
+      "learning_rate": 4.803816731101423e-07,
+      "loss": 0.6736,
+      "step": 34820
+    },
+    {
+      "epoch": 0.95240699470885,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.776397685832579e-07,
+      "loss": 0.7333,
+      "step": 34830
+    },
+    {
+      "epoch": 0.9526804391517754,
+      "grad_norm": 2.9375,
+      "learning_rate": 4.748978640563736e-07,
+      "loss": 0.7,
+      "step": 34840
+    },
+    {
+      "epoch": 0.9529538835947007,
+      "grad_norm": 2.765625,
+      "learning_rate": 4.7215595952948926e-07,
+      "loss": 0.6343,
+      "step": 34850
+    },
+    {
+      "epoch": 0.953227328037626,
+      "grad_norm": 2.46875,
+      "learning_rate": 4.694140550026049e-07,
+      "loss": 0.6286,
+      "step": 34860
+    },
+    {
+      "epoch": 0.9535007724805513,
+      "grad_norm": 3.046875,
+      "learning_rate": 4.6667215047572044e-07,
+      "loss": 0.6234,
+      "step": 34870
+    },
+    {
+      "epoch": 0.9537742169234765,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.639302459488361e-07,
+      "loss": 0.6814,
+      "step": 34880
+    },
+    {
+      "epoch": 0.9540476613664018,
+      "grad_norm": 3.125,
+      "learning_rate": 4.611883414219517e-07,
+      "loss": 0.6411,
+      "step": 34890
+    },
+    {
+      "epoch": 0.9543211058093272,
+      "grad_norm": 3.109375,
+      "learning_rate": 4.5844643689506736e-07,
+      "loss": 0.6912,
+      "step": 34900
+    },
+    {
+      "epoch": 0.9545945502522525,
+      "grad_norm": 2.625,
+      "learning_rate": 4.5570453236818295e-07,
+      "loss": 0.7091,
+      "step": 34910
+    },
+    {
+      "epoch": 0.9548679946951778,
+      "grad_norm": 2.359375,
+      "learning_rate": 4.529626278412986e-07,
+      "loss": 0.6023,
+      "step": 34920
+    },
+    {
+      "epoch": 0.9551414391381031,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.5022072331441423e-07,
+      "loss": 0.6433,
+      "step": 34930
+    },
+    {
+      "epoch": 0.9554148835810284,
+      "grad_norm": 2.5,
+      "learning_rate": 4.4747881878752987e-07,
+      "loss": 0.6087,
+      "step": 34940
+    },
+    {
+      "epoch": 0.9556883280239538,
+      "grad_norm": 2.984375,
+      "learning_rate": 4.447369142606455e-07,
+      "loss": 0.7129,
+      "step": 34950
+    },
+    {
+      "epoch": 0.9559617724668791,
+      "grad_norm": 2.703125,
+      "learning_rate": 4.419950097337611e-07,
+      "loss": 0.668,
+      "step": 34960
+    },
+    {
+      "epoch": 0.9562352169098044,
+      "grad_norm": 3.40625,
+      "learning_rate": 4.3925310520687674e-07,
+      "loss": 0.6594,
+      "step": 34970
+    },
+    {
+      "epoch": 0.9565086613527296,
+      "grad_norm": 2.5625,
+      "learning_rate": 4.365112006799924e-07,
+      "loss": 0.7197,
+      "step": 34980
+    },
+    {
+      "epoch": 0.9567821057956549,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.33769296153108e-07,
+      "loss": 0.7347,
+      "step": 34990
+    },
+    {
+      "epoch": 0.9570555502385802,
+      "grad_norm": 3.28125,
+      "learning_rate": 4.3102739162622356e-07,
+      "loss": 0.6882,
+      "step": 35000
+    },
+    {
+      "epoch": 0.9573289946815056,
+      "grad_norm": 2.921875,
+      "learning_rate": 4.282854870993392e-07,
+      "loss": 0.7005,
+      "step": 35010
+    },
+    {
+      "epoch": 0.9576024391244309,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.2554358257245484e-07,
+      "loss": 0.7138,
+      "step": 35020
+    },
+    {
+      "epoch": 0.9578758835673562,
+      "grad_norm": 3.59375,
+      "learning_rate": 4.228016780455705e-07,
+      "loss": 0.7006,
+      "step": 35030
+    },
+    {
+      "epoch": 0.9581493280102815,
+      "grad_norm": 2.5,
+      "learning_rate": 4.200597735186861e-07,
+      "loss": 0.6811,
+      "step": 35040
+    },
+    {
+      "epoch": 0.9584227724532068,
+      "grad_norm": 2.875,
+      "learning_rate": 4.173178689918017e-07,
+      "loss": 0.7468,
+      "step": 35050
+    },
+    {
+      "epoch": 0.9586962168961322,
+      "grad_norm": 2.484375,
+      "learning_rate": 4.1457596446491735e-07,
+      "loss": 0.7405,
+      "step": 35060
+    },
+    {
+      "epoch": 0.9589696613390575,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.11834059938033e-07,
+      "loss": 0.7261,
+      "step": 35070
+    },
+    {
+      "epoch": 0.9592431057819828,
+      "grad_norm": 2.953125,
+      "learning_rate": 4.0909215541114863e-07,
+      "loss": 0.702,
+      "step": 35080
+    },
+    {
+      "epoch": 0.959516550224908,
+      "grad_norm": 2.78125,
+      "learning_rate": 4.0635025088426427e-07,
+      "loss": 0.6498,
+      "step": 35090
+    },
+    {
+      "epoch": 0.9597899946678333,
+      "grad_norm": 2.6875,
+      "learning_rate": 4.0360834635737986e-07,
+      "loss": 0.6982,
+      "step": 35100
+    },
+    {
+      "epoch": 0.9600634391107586,
+      "grad_norm": 2.828125,
+      "learning_rate": 4.008664418304955e-07,
+      "loss": 0.7046,
+      "step": 35110
+    },
+    {
+      "epoch": 0.960336883553684,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.9812453730361114e-07,
+      "loss": 0.6958,
+      "step": 35120
+    },
+    {
+      "epoch": 0.9606103279966093,
+      "grad_norm": 2.734375,
+      "learning_rate": 3.953826327767268e-07,
+      "loss": 0.7164,
+      "step": 35130
+    },
+    {
+      "epoch": 0.9608837724395346,
+      "grad_norm": 2.578125,
+      "learning_rate": 3.926407282498423e-07,
+      "loss": 0.6473,
+      "step": 35140
+    },
+    {
+      "epoch": 0.9611572168824599,
+      "grad_norm": 2.640625,
+      "learning_rate": 3.8989882372295796e-07,
+      "loss": 0.6317,
+      "step": 35150
+    },
+    {
+      "epoch": 0.9614306613253852,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.871569191960736e-07,
+      "loss": 0.7089,
+      "step": 35160
+    },
+    {
+      "epoch": 0.9617041057683106,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.8441501466918924e-07,
+      "loss": 0.6676,
+      "step": 35170
+    },
+    {
+      "epoch": 0.9619775502112359,
+      "grad_norm": 2.5,
+      "learning_rate": 3.816731101423049e-07,
+      "loss": 0.7055,
+      "step": 35180
+    },
+    {
+      "epoch": 0.9622509946541612,
+      "grad_norm": 2.5625,
+      "learning_rate": 3.7893120561542047e-07,
+      "loss": 0.7123,
+      "step": 35190
+    },
+    {
+      "epoch": 0.9625244390970864,
+      "grad_norm": 3.078125,
+      "learning_rate": 3.761893010885361e-07,
+      "loss": 0.7269,
+      "step": 35200
+    },
+    {
+      "epoch": 0.9627978835400117,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.7344739656165175e-07,
+      "loss": 0.6884,
+      "step": 35210
+    },
+    {
+      "epoch": 0.963071327982937,
+      "grad_norm": 2.8125,
+      "learning_rate": 3.707054920347674e-07,
+      "loss": 0.6583,
+      "step": 35220
+    },
+    {
+      "epoch": 0.9633447724258624,
+      "grad_norm": 3.0625,
+      "learning_rate": 3.6796358750788303e-07,
+      "loss": 0.7579,
+      "step": 35230
+    },
+    {
+      "epoch": 0.9636182168687877,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.652216829809986e-07,
+      "loss": 0.6611,
+      "step": 35240
+    },
+    {
+      "epoch": 0.963891661311713,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.6247977845411426e-07,
+      "loss": 0.6366,
+      "step": 35250
+    },
+    {
+      "epoch": 0.9641651057546383,
+      "grad_norm": 3.203125,
+      "learning_rate": 3.597378739272299e-07,
+      "loss": 0.7212,
+      "step": 35260
+    },
+    {
+      "epoch": 0.9644385501975636,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.5699596940034554e-07,
+      "loss": 0.7117,
+      "step": 35270
+    },
+    {
+      "epoch": 0.964711994640489,
+      "grad_norm": 3.046875,
+      "learning_rate": 3.5425406487346113e-07,
+      "loss": 0.6765,
+      "step": 35280
+    },
+    {
+      "epoch": 0.9649854390834143,
+      "grad_norm": 2.703125,
+      "learning_rate": 3.5151216034657677e-07,
+      "loss": 0.6882,
+      "step": 35290
+    },
+    {
+      "epoch": 0.9652588835263395,
+      "grad_norm": 2.859375,
+      "learning_rate": 3.487702558196924e-07,
+      "loss": 0.7082,
+      "step": 35300
+    },
+    {
+      "epoch": 0.9655323279692648,
+      "grad_norm": 2.625,
+      "learning_rate": 3.4602835129280805e-07,
+      "loss": 0.7668,
+      "step": 35310
+    },
+    {
+      "epoch": 0.9658057724121901,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.432864467659237e-07,
+      "loss": 0.6504,
+      "step": 35320
+    },
+    {
+      "epoch": 0.9660792168551154,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.405445422390392e-07,
+      "loss": 0.6301,
+      "step": 35330
+    },
+    {
+      "epoch": 0.9663526612980408,
+      "grad_norm": 2.359375,
+      "learning_rate": 3.3780263771215487e-07,
+      "loss": 0.6349,
+      "step": 35340
+    },
+    {
+      "epoch": 0.9666261057409661,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.350607331852705e-07,
+      "loss": 0.7404,
+      "step": 35350
+    },
+    {
+      "epoch": 0.9668995501838914,
+      "grad_norm": 2.65625,
+      "learning_rate": 3.3231882865838615e-07,
+      "loss": 0.7021,
+      "step": 35360
+    },
+    {
+      "epoch": 0.9671729946268167,
+      "grad_norm": 2.875,
+      "learning_rate": 3.295769241315018e-07,
+      "loss": 0.7051,
+      "step": 35370
+    },
+    {
+      "epoch": 0.967446439069742,
+      "grad_norm": 2.71875,
+      "learning_rate": 3.268350196046174e-07,
+      "loss": 0.6914,
+      "step": 35380
+    },
+    {
+      "epoch": 0.9677198835126674,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.24093115077733e-07,
+      "loss": 0.7076,
+      "step": 35390
+    },
+    {
+      "epoch": 0.9679933279555927,
+      "grad_norm": 2.515625,
+      "learning_rate": 3.2135121055084866e-07,
+      "loss": 0.7532,
+      "step": 35400
+    },
+    {
+      "epoch": 0.9682667723985179,
+      "grad_norm": 2.953125,
+      "learning_rate": 3.186093060239643e-07,
+      "loss": 0.7188,
+      "step": 35410
+    },
+    {
+      "epoch": 0.9685402168414432,
+      "grad_norm": 2.453125,
+      "learning_rate": 3.158674014970799e-07,
+      "loss": 0.6316,
+      "step": 35420
+    },
+    {
+      "epoch": 0.9688136612843685,
+      "grad_norm": 2.78125,
+      "learning_rate": 3.1312549697019553e-07,
+      "loss": 0.6367,
+      "step": 35430
+    },
+    {
+      "epoch": 0.9690871057272938,
+      "grad_norm": 2.53125,
+      "learning_rate": 3.1038359244331117e-07,
+      "loss": 0.7353,
+      "step": 35440
+    },
+    {
+      "epoch": 0.9693605501702192,
+      "grad_norm": 3.359375,
+      "learning_rate": 3.076416879164268e-07,
+      "loss": 0.667,
+      "step": 35450
+    },
+    {
+      "epoch": 0.9696339946131445,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.048997833895424e-07,
+      "loss": 0.7658,
+      "step": 35460
+    },
+    {
+      "epoch": 0.9699074390560698,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.0215787886265804e-07,
+      "loss": 0.7402,
+      "step": 35470
+    },
+    {
+      "epoch": 0.9701808834989951,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.994159743357736e-07,
+      "loss": 0.6522,
+      "step": 35480
+    },
+    {
+      "epoch": 0.9704543279419204,
+      "grad_norm": 2.5625,
+      "learning_rate": 2.9667406980888927e-07,
+      "loss": 0.6638,
+      "step": 35490
+    },
+    {
+      "epoch": 0.9707277723848458,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.939321652820049e-07,
+      "loss": 0.7189,
+      "step": 35500
+    },
+    {
+      "epoch": 0.971001216827771,
+      "grad_norm": 2.5,
+      "learning_rate": 2.9119026075512055e-07,
+      "loss": 0.7112,
+      "step": 35510
+    },
+    {
+      "epoch": 0.9712746612706963,
+      "grad_norm": 2.53125,
+      "learning_rate": 2.884483562282362e-07,
+      "loss": 0.7113,
+      "step": 35520
+    },
+    {
+      "epoch": 0.9715481057136216,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.857064517013518e-07,
+      "loss": 0.7018,
+      "step": 35530
+    },
+    {
+      "epoch": 0.9718215501565469,
+      "grad_norm": 2.484375,
+      "learning_rate": 2.829645471744674e-07,
+      "loss": 0.6085,
+      "step": 35540
+    },
+    {
+      "epoch": 0.9720949945994722,
+      "grad_norm": 2.75,
+      "learning_rate": 2.80222642647583e-07,
+      "loss": 0.6941,
+      "step": 35550
+    },
+    {
+      "epoch": 0.9723684390423976,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.7748073812069865e-07,
+      "loss": 0.6904,
+      "step": 35560
+    },
+    {
+      "epoch": 0.9726418834853229,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.747388335938143e-07,
+      "loss": 0.5828,
+      "step": 35570
+    },
+    {
+      "epoch": 0.9729153279282482,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.7199692906692993e-07,
+      "loss": 0.7111,
+      "step": 35580
+    },
+    {
+      "epoch": 0.9731887723711735,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.6925502454004557e-07,
+      "loss": 0.7078,
+      "step": 35590
+    },
+    {
+      "epoch": 0.9734622168140988,
+      "grad_norm": 2.921875,
+      "learning_rate": 2.6651312001316116e-07,
+      "loss": 0.6168,
+      "step": 35600
+    },
+    {
+      "epoch": 0.9737356612570242,
+      "grad_norm": 2.84375,
+      "learning_rate": 2.637712154862768e-07,
+      "loss": 0.685,
+      "step": 35610
+    },
+    {
+      "epoch": 0.9740091056999494,
+      "grad_norm": 2.640625,
+      "learning_rate": 2.610293109593924e-07,
+      "loss": 0.6182,
+      "step": 35620
+    },
+    {
+      "epoch": 0.9742825501428747,
+      "grad_norm": 2.625,
+      "learning_rate": 2.58287406432508e-07,
+      "loss": 0.7358,
+      "step": 35630
+    },
+    {
+      "epoch": 0.9745559945858,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.5554550190562367e-07,
+      "loss": 0.7443,
+      "step": 35640
+    },
+    {
+      "epoch": 0.9748294390287253,
+      "grad_norm": 2.875,
+      "learning_rate": 2.528035973787393e-07,
+      "loss": 0.6795,
+      "step": 35650
+    },
+    {
+      "epoch": 0.9751028834716506,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.5006169285185495e-07,
+      "loss": 0.7345,
+      "step": 35660
+    },
+    {
+      "epoch": 0.975376327914576,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.4731978832497054e-07,
+      "loss": 0.6703,
+      "step": 35670
+    },
+    {
+      "epoch": 0.9756497723575013,
+      "grad_norm": 2.6875,
+      "learning_rate": 2.445778837980862e-07,
+      "loss": 0.6477,
+      "step": 35680
+    },
+    {
+      "epoch": 0.9759232168004266,
+      "grad_norm": 2.625,
+      "learning_rate": 2.4183597927120176e-07,
+      "loss": 0.6718,
+      "step": 35690
+    },
+    {
+      "epoch": 0.9761966612433519,
+      "grad_norm": 2.71875,
+      "learning_rate": 2.390940747443174e-07,
+      "loss": 0.7363,
+      "step": 35700
+    },
+    {
+      "epoch": 0.9764701056862772,
+      "grad_norm": 2.421875,
+      "learning_rate": 2.3635217021743305e-07,
+      "loss": 0.7118,
+      "step": 35710
+    },
+    {
+      "epoch": 0.9767435501292026,
+      "grad_norm": 3.5,
+      "learning_rate": 2.3361026569054869e-07,
+      "loss": 0.8073,
+      "step": 35720
+    },
+    {
+      "epoch": 0.9770169945721278,
+      "grad_norm": 2.453125,
+      "learning_rate": 2.3086836116366433e-07,
+      "loss": 0.7192,
+      "step": 35730
+    },
+    {
+      "epoch": 0.9772904390150531,
+      "grad_norm": 2.859375,
+      "learning_rate": 2.2812645663677991e-07,
+      "loss": 0.7012,
+      "step": 35740
+    },
+    {
+      "epoch": 0.9775638834579784,
+      "grad_norm": 2.875,
+      "learning_rate": 2.2538455210989556e-07,
+      "loss": 0.663,
+      "step": 35750
+    },
+    {
+      "epoch": 0.9778373279009037,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.2264264758301117e-07,
+      "loss": 0.6944,
+      "step": 35760
+    },
+    {
+      "epoch": 0.978110772343829,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.199007430561268e-07,
+      "loss": 0.7547,
+      "step": 35770
+    },
+    {
+      "epoch": 0.9783842167867544,
+      "grad_norm": 3.140625,
+      "learning_rate": 2.1715883852924243e-07,
+      "loss": 0.7007,
+      "step": 35780
+    },
+    {
+      "epoch": 0.9786576612296797,
+      "grad_norm": 2.46875,
+      "learning_rate": 2.1441693400235807e-07,
+      "loss": 0.6133,
+      "step": 35790
+    },
+    {
+      "epoch": 0.978931105672605,
+      "grad_norm": 2.890625,
+      "learning_rate": 2.1167502947547365e-07,
+      "loss": 0.7038,
+      "step": 35800
+    },
+    {
+      "epoch": 0.9792045501155303,
+      "grad_norm": 2.8125,
+      "learning_rate": 2.089331249485893e-07,
+      "loss": 0.7063,
+      "step": 35810
+    },
+    {
+      "epoch": 0.9794779945584556,
+      "grad_norm": 2.984375,
+      "learning_rate": 2.0619122042170494e-07,
+      "loss": 0.705,
+      "step": 35820
+    },
+    {
+      "epoch": 0.9797514390013808,
+      "grad_norm": 3.046875,
+      "learning_rate": 2.0344931589482055e-07,
+      "loss": 0.6751,
+      "step": 35830
+    },
+    {
+      "epoch": 0.9800248834443062,
+      "grad_norm": 2.703125,
+      "learning_rate": 2.007074113679362e-07,
+      "loss": 0.7438,
+      "step": 35840
+    },
+    {
+      "epoch": 0.9802983278872315,
+      "grad_norm": 3.140625,
+      "learning_rate": 1.979655068410518e-07,
+      "loss": 0.7276,
+      "step": 35850
+    },
+    {
+      "epoch": 0.9805717723301568,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.9522360231416745e-07,
+      "loss": 0.6596,
+      "step": 35860
+    },
+    {
+      "epoch": 0.9808452167730821,
+      "grad_norm": 2.625,
+      "learning_rate": 1.9248169778728303e-07,
+      "loss": 0.7635,
+      "step": 35870
+    },
+    {
+      "epoch": 0.9811186612160074,
+      "grad_norm": 3.65625,
+      "learning_rate": 1.8973979326039867e-07,
+      "loss": 0.7105,
+      "step": 35880
+    },
+    {
+      "epoch": 0.9813921056589328,
+      "grad_norm": 2.25,
+      "learning_rate": 1.8699788873351431e-07,
+      "loss": 0.6777,
+      "step": 35890
+    },
+    {
+      "epoch": 0.9816655501018581,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.8425598420662993e-07,
+      "loss": 0.7574,
+      "step": 35900
+    },
+    {
+      "epoch": 0.9819389945447834,
+      "grad_norm": 2.78125,
+      "learning_rate": 1.8151407967974557e-07,
+      "loss": 0.7161,
+      "step": 35910
+    },
+    {
+      "epoch": 0.9822124389877087,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.7877217515286118e-07,
+      "loss": 0.7063,
+      "step": 35920
+    },
+    {
+      "epoch": 0.982485883430634,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.7603027062597683e-07,
+      "loss": 0.6843,
+      "step": 35930
+    },
+    {
+      "epoch": 0.9827593278735592,
+      "grad_norm": 2.734375,
+      "learning_rate": 1.7328836609909244e-07,
+      "loss": 0.7016,
+      "step": 35940
+    },
+    {
+      "epoch": 0.9830327723164846,
+      "grad_norm": 2.859375,
+      "learning_rate": 1.7054646157220808e-07,
+      "loss": 0.6936,
+      "step": 35950
+    },
+    {
+      "epoch": 0.9833062167594099,
+      "grad_norm": 2.796875,
+      "learning_rate": 1.6780455704532372e-07,
+      "loss": 0.7624,
+      "step": 35960
+    },
+    {
+      "epoch": 0.9835796612023352,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.650626525184393e-07,
+      "loss": 0.6882,
+      "step": 35970
+    },
+    {
+      "epoch": 0.9838531056452605,
+      "grad_norm": 2.59375,
+      "learning_rate": 1.6232074799155495e-07,
+      "loss": 0.6692,
+      "step": 35980
+    },
+    {
+      "epoch": 0.9841265500881858,
+      "grad_norm": 2.875,
+      "learning_rate": 1.5957884346467056e-07,
+      "loss": 0.7824,
+      "step": 35990
+    },
+    {
+      "epoch": 0.9843999945311112,
+      "grad_norm": 2.546875,
+      "learning_rate": 1.568369389377862e-07,
+      "loss": 0.712,
+      "step": 36000
+    },
+    {
+      "epoch": 0.9846734389740365,
+      "grad_norm": 2.515625,
+      "learning_rate": 1.5409503441090182e-07,
+      "loss": 0.7134,
+      "step": 36010
+    },
+    {
+      "epoch": 0.9849468834169618,
+      "grad_norm": 2.75,
+      "learning_rate": 1.5135312988401746e-07,
+      "loss": 0.6855,
+      "step": 36020
+    },
+    {
+      "epoch": 0.9852203278598871,
+      "grad_norm": 2.90625,
+      "learning_rate": 1.4861122535713307e-07,
+      "loss": 0.615,
+      "step": 36030
+    },
+    {
+      "epoch": 0.9854937723028123,
+      "grad_norm": 3.03125,
+      "learning_rate": 1.458693208302487e-07,
+      "loss": 0.7006,
+      "step": 36040
+    },
+    {
+      "epoch": 0.9857672167457376,
+      "grad_norm": 2.890625,
+      "learning_rate": 1.4312741630336433e-07,
+      "loss": 0.7225,
+      "step": 36050
+    },
+    {
+      "epoch": 0.986040661188663,
+      "grad_norm": 2.640625,
+      "learning_rate": 1.4038551177647997e-07,
+      "loss": 0.6622,
+      "step": 36060
+    },
+    {
+      "epoch": 0.9863141056315883,
+      "grad_norm": 2.4375,
+      "learning_rate": 1.3764360724959558e-07,
+      "loss": 0.652,
+      "step": 36070
+    },
+    {
+      "epoch": 0.9865875500745136,
+      "grad_norm": 2.8125,
+      "learning_rate": 1.349017027227112e-07,
+      "loss": 0.7124,
+      "step": 36080
+    },
+    {
+      "epoch": 0.9868609945174389,
+      "grad_norm": 3.046875,
+      "learning_rate": 1.3215979819582684e-07,
+      "loss": 0.7341,
+      "step": 36090
+    },
+    {
+      "epoch": 0.9871344389603642,
+      "grad_norm": 2.765625,
+      "learning_rate": 1.2941789366894245e-07,
+      "loss": 0.7236,
+      "step": 36100
+    },
+    {
+      "epoch": 0.9874078834032896,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.2667598914205807e-07,
+      "loss": 0.7345,
+      "step": 36110
+    },
+    {
+      "epoch": 0.9876813278462149,
+      "grad_norm": 3.3125,
+      "learning_rate": 1.239340846151737e-07,
+      "loss": 0.7393,
+      "step": 36120
+    },
+    {
+      "epoch": 0.9879547722891402,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.2119218008828935e-07,
+      "loss": 0.6382,
+      "step": 36130
+    },
+    {
+      "epoch": 0.9882282167320655,
+      "grad_norm": 2.375,
+      "learning_rate": 1.1845027556140496e-07,
+      "loss": 0.6763,
+      "step": 36140
+    },
+    {
+      "epoch": 0.9885016611749907,
+      "grad_norm": 2.71875,
+      "learning_rate": 1.1570837103452059e-07,
+      "loss": 0.6947,
+      "step": 36150
+    },
+    {
+      "epoch": 0.988775105617916,
+      "grad_norm": 2.671875,
+      "learning_rate": 1.1296646650763622e-07,
+      "loss": 0.7217,
+      "step": 36160
+    },
+    {
+      "epoch": 0.9890485500608414,
+      "grad_norm": 2.984375,
+      "learning_rate": 1.1022456198075183e-07,
+      "loss": 0.7436,
+      "step": 36170
+    },
+    {
+      "epoch": 0.9893219945037667,
+      "grad_norm": 2.75,
+      "learning_rate": 1.0748265745386746e-07,
+      "loss": 0.6377,
+      "step": 36180
+    },
+    {
+      "epoch": 0.989595438946692,
+      "grad_norm": 2.65625,
+      "learning_rate": 1.0474075292698309e-07,
+      "loss": 0.6429,
+      "step": 36190
+    },
+    {
+      "epoch": 0.9898688833896173,
+      "grad_norm": 3.484375,
+      "learning_rate": 1.019988484000987e-07,
+      "loss": 0.696,
+      "step": 36200
+    },
+    {
+      "epoch": 0.9901423278325426,
+      "grad_norm": 2.78125,
+      "learning_rate": 9.925694387321434e-08,
+      "loss": 0.7024,
+      "step": 36210
+    },
+    {
+      "epoch": 0.990415772275468,
+      "grad_norm": 3.296875,
+      "learning_rate": 9.651503934632997e-08,
+      "loss": 0.6419,
+      "step": 36220
+    },
+    {
+      "epoch": 0.9906892167183933,
+      "grad_norm": 2.421875,
+      "learning_rate": 9.37731348194456e-08,
+      "loss": 0.6579,
+      "step": 36230
+    },
+    {
+      "epoch": 0.9909626611613186,
+      "grad_norm": 2.71875,
+      "learning_rate": 9.103123029256123e-08,
+      "loss": 0.695,
+      "step": 36240
+    },
+    {
+      "epoch": 0.9912361056042439,
+      "grad_norm": 2.75,
+      "learning_rate": 8.828932576567684e-08,
+      "loss": 0.6821,
+      "step": 36250
+    },
+    {
+      "epoch": 0.9915095500471691,
+      "grad_norm": 3.03125,
+      "learning_rate": 8.554742123879247e-08,
+      "loss": 0.7694,
+      "step": 36260
+    },
+    {
+      "epoch": 0.9917829944900944,
+      "grad_norm": 2.90625,
+      "learning_rate": 8.28055167119081e-08,
+      "loss": 0.7164,
+      "step": 36270
+    },
+    {
+      "epoch": 0.9920564389330198,
+      "grad_norm": 3.375,
+      "learning_rate": 8.006361218502374e-08,
+      "loss": 0.7653,
+      "step": 36280
+    },
+    {
+      "epoch": 0.9923298833759451,
+      "grad_norm": 2.96875,
+      "learning_rate": 7.732170765813935e-08,
+      "loss": 0.6463,
+      "step": 36290
+    },
+    {
+      "epoch": 0.9926033278188704,
+      "grad_norm": 2.453125,
+      "learning_rate": 7.457980313125498e-08,
+      "loss": 0.6799,
+      "step": 36300
+    },
+    {
+      "epoch": 0.9928767722617957,
+      "grad_norm": 2.5625,
+      "learning_rate": 7.18378986043706e-08,
+      "loss": 0.7258,
+      "step": 36310
+    },
+    {
+      "epoch": 0.993150216704721,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.909599407748622e-08,
+      "loss": 0.7094,
+      "step": 36320
+    },
+    {
+      "epoch": 0.9934236611476464,
+      "grad_norm": 2.203125,
+      "learning_rate": 6.635408955060186e-08,
+      "loss": 0.654,
+      "step": 36330
+    },
+    {
+      "epoch": 0.9936971055905717,
+      "grad_norm": 2.90625,
+      "learning_rate": 6.361218502371748e-08,
+      "loss": 0.6714,
+      "step": 36340
+    },
+    {
+      "epoch": 0.993970550033497,
+      "grad_norm": 2.84375,
+      "learning_rate": 6.08702804968331e-08,
+      "loss": 0.7055,
+      "step": 36350
+    },
+    {
+      "epoch": 0.9942439944764222,
+      "grad_norm": 2.375,
+      "learning_rate": 5.812837596994873e-08,
+      "loss": 0.6541,
+      "step": 36360
+    },
+    {
+      "epoch": 0.9945174389193475,
+      "grad_norm": 2.875,
+      "learning_rate": 5.538647144306436e-08,
+      "loss": 0.6878,
+      "step": 36370
+    },
+    {
+      "epoch": 0.9947908833622728,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.2644566916179985e-08,
+      "loss": 0.6804,
+      "step": 36380
+    },
+    {
+      "epoch": 0.9950643278051982,
+      "grad_norm": 2.65625,
+      "learning_rate": 4.9902662389295606e-08,
+      "loss": 0.618,
+      "step": 36390
+    },
+    {
+      "epoch": 0.9953377722481235,
+      "grad_norm": 2.8125,
+      "learning_rate": 4.7160757862411234e-08,
+      "loss": 0.7042,
+      "step": 36400
+    },
+    {
+      "epoch": 0.9956112166910488,
+      "grad_norm": 3.125,
+      "learning_rate": 4.441885333552686e-08,
+      "loss": 0.6653,
+      "step": 36410
+    },
+    {
+      "epoch": 0.9958846611339741,
+      "grad_norm": 3.078125,
+      "learning_rate": 4.167694880864249e-08,
+      "loss": 0.6694,
+      "step": 36420
+    },
+    {
+      "epoch": 0.9961581055768994,
+      "grad_norm": 2.625,
+      "learning_rate": 3.8935044281758117e-08,
+      "loss": 0.6993,
+      "step": 36430
+    },
+    {
+      "epoch": 0.9964315500198248,
+      "grad_norm": 2.671875,
+      "learning_rate": 3.619313975487374e-08,
+      "loss": 0.6414,
+      "step": 36440
+    },
+    {
+      "epoch": 0.9967049944627501,
+      "grad_norm": 2.84375,
+      "learning_rate": 3.3451235227989365e-08,
+      "loss": 0.6967,
+      "step": 36450
+    },
+    {
+      "epoch": 0.9969784389056754,
+      "grad_norm": 2.828125,
+      "learning_rate": 3.0709330701104986e-08,
+      "loss": 0.6618,
+      "step": 36460
+    },
+    {
+      "epoch": 0.9972518833486006,
+      "grad_norm": 2.953125,
+      "learning_rate": 2.7967426174220617e-08,
+      "loss": 0.7492,
+      "step": 36470
+    },
+    {
+      "epoch": 0.9975253277915259,
+      "grad_norm": 2.796875,
+      "learning_rate": 2.522552164733624e-08,
+      "loss": 0.6612,
+      "step": 36480
+    },
+    {
+      "epoch": 0.9977987722344512,
+      "grad_norm": 2.59375,
+      "learning_rate": 2.248361712045187e-08,
+      "loss": 0.6469,
+      "step": 36490
+    },
+    {
+      "epoch": 0.9980722166773766,
+      "grad_norm": 2.921875,
+      "learning_rate": 1.9741712593567493e-08,
+      "loss": 0.7214,
+      "step": 36500
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 36571,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5.3039385772032e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}