diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,11193 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9996235648409562,
+  "eval_steps": 500,
+  "global_step": 7968,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012547838634795157,
+      "grad_norm": 25.442365646362305,
+      "learning_rate": 4.1666666666666667e-07,
+      "loss": 0.1369,
+      "step": 5
+    },
+    {
+      "epoch": 0.0025095677269590315,
+      "grad_norm": 2707.152587890625,
+      "learning_rate": 8.333333333333333e-07,
+      "loss": 0.1494,
+      "step": 10
+    },
+    {
+      "epoch": 0.003764351590438547,
+      "grad_norm": 1.4790934324264526,
+      "learning_rate": 1.25e-06,
+      "loss": 0.1436,
+      "step": 15
+    },
+    {
+      "epoch": 0.005019135453918063,
+      "grad_norm": 6611.14990234375,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.1239,
+      "step": 20
+    },
+    {
+      "epoch": 0.006273919317397578,
+      "grad_norm": 3.4831368923187256,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 0.1313,
+      "step": 25
+    },
+    {
+      "epoch": 0.007528703180877094,
+      "grad_norm": 0.8000121712684631,
+      "learning_rate": 2.5e-06,
+      "loss": 0.1203,
+      "step": 30
+    },
+    {
+      "epoch": 0.00878348704435661,
+      "grad_norm": 0.5967940092086792,
+      "learning_rate": 2.916666666666667e-06,
+      "loss": 0.1185,
+      "step": 35
+    },
+    {
+      "epoch": 0.010038270907836126,
+      "grad_norm": 3509.61083984375,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 0.1243,
+      "step": 40
+    },
+    {
+      "epoch": 0.01129305477131564,
+      "grad_norm": 4.068423271179199,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.123,
+      "step": 45
+    },
+    {
+      "epoch": 0.012547838634795156,
+      "grad_norm": 981.9570922851562,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 0.1358,
+      "step": 50
+    },
+    {
+      "epoch": 0.013802622498274672,
+      "grad_norm": 565.902099609375,
+      "learning_rate": 4.583333333333333e-06,
+      "loss": 0.1181,
+      "step": 55
+    },
+    {
+      "epoch": 0.015057406361754188,
+      "grad_norm": 1.1762653589248657,
+      "learning_rate": 5e-06,
+      "loss": 0.1138,
+      "step": 60
+    },
+    {
+      "epoch": 0.016312190225233704,
+      "grad_norm": 13.798249244689941,
+      "learning_rate": 5.416666666666667e-06,
+      "loss": 0.1303,
+      "step": 65
+    },
+    {
+      "epoch": 0.01756697408871322,
+      "grad_norm": 1.9088438749313354,
+      "learning_rate": 5.833333333333334e-06,
+      "loss": 0.128,
+      "step": 70
+    },
+    {
+      "epoch": 0.018821757952192736,
+      "grad_norm": 379.896484375,
+      "learning_rate": 6.25e-06,
+      "loss": 0.1115,
+      "step": 75
+    },
+    {
+      "epoch": 0.020076541815672252,
+      "grad_norm": 53.034820556640625,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 0.1106,
+      "step": 80
+    },
+    {
+      "epoch": 0.021331325679151768,
+      "grad_norm": 0.5621482729911804,
+      "learning_rate": 7.083333333333335e-06,
+      "loss": 0.1339,
+      "step": 85
+    },
+    {
+      "epoch": 0.02258610954263128,
+      "grad_norm": 1503.7694091796875,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.1091,
+      "step": 90
+    },
+    {
+      "epoch": 0.023840893406110796,
+      "grad_norm": 21.09101676940918,
+      "learning_rate": 7.916666666666667e-06,
+      "loss": 0.1027,
+      "step": 95
+    },
+    {
+      "epoch": 0.025095677269590312,
+      "grad_norm": 153.1422882080078,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 0.1158,
+      "step": 100
+    },
+    {
+      "epoch": 0.026350461133069828,
+      "grad_norm": 686.55615234375,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.1239,
+      "step": 105
+    },
+    {
+      "epoch": 0.027605244996549344,
+      "grad_norm": 456.04541015625,
+      "learning_rate": 9.166666666666666e-06,
+      "loss": 0.1176,
+      "step": 110
+    },
+    {
+      "epoch": 0.02886002886002886,
+      "grad_norm": 244.83534240722656,
+      "learning_rate": 9.583333333333335e-06,
+      "loss": 0.0952,
+      "step": 115
+    },
+    {
+      "epoch": 0.030114812723508376,
+      "grad_norm": 720.8134155273438,
+      "learning_rate": 1e-05,
+      "loss": 0.1181,
+      "step": 120
+    },
+    {
+      "epoch": 0.03136959658698789,
+      "grad_norm": 1.439946174621582,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 0.1147,
+      "step": 125
+    },
+    {
+      "epoch": 0.03262438045046741,
+      "grad_norm": 24.099985122680664,
+      "learning_rate": 1.0833333333333334e-05,
+      "loss": 0.1133,
+      "step": 130
+    },
+    {
+      "epoch": 0.03387916431394692,
+      "grad_norm": 10.10527515411377,
+      "learning_rate": 1.125e-05,
+      "loss": 0.1237,
+      "step": 135
+    },
+    {
+      "epoch": 0.03513394817742644,
+      "grad_norm": 171.32110595703125,
+      "learning_rate": 1.1666666666666668e-05,
+      "loss": 0.3536,
+      "step": 140
+    },
+    {
+      "epoch": 0.03638873204090595,
+      "grad_norm": 129.24395751953125,
+      "learning_rate": 1.2083333333333333e-05,
+      "loss": 0.1079,
+      "step": 145
+    },
+    {
+      "epoch": 0.03764351590438547,
+      "grad_norm": 3.056920289993286,
+      "learning_rate": 1.25e-05,
+      "loss": 0.1223,
+      "step": 150
+    },
+    {
+      "epoch": 0.038898299767864984,
+      "grad_norm": 58.41600036621094,
+      "learning_rate": 1.2916666666666668e-05,
+      "loss": 0.094,
+      "step": 155
+    },
+    {
+      "epoch": 0.040153083631344504,
+      "grad_norm": 67.04778289794922,
+      "learning_rate": 1.3333333333333333e-05,
+      "loss": 0.1181,
+      "step": 160
+    },
+    {
+      "epoch": 0.041407867494824016,
+      "grad_norm": 152.23841857910156,
+      "learning_rate": 1.375e-05,
+      "loss": 0.1122,
+      "step": 165
+    },
+    {
+      "epoch": 0.042662651358303535,
+      "grad_norm": 35.90134048461914,
+      "learning_rate": 1.416666666666667e-05,
+      "loss": 0.1152,
+      "step": 170
+    },
+    {
+      "epoch": 0.04391743522178305,
+      "grad_norm": 1.0190317630767822,
+      "learning_rate": 1.4583333333333333e-05,
+      "loss": 0.1138,
+      "step": 175
+    },
+    {
+      "epoch": 0.04517221908526256,
+      "grad_norm": 0.9965628385543823,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.1127,
+      "step": 180
+    },
+    {
+      "epoch": 0.04642700294874208,
+      "grad_norm": 1.6608027219772339,
+      "learning_rate": 1.5416666666666668e-05,
+      "loss": 0.1114,
+      "step": 185
+    },
+    {
+      "epoch": 0.04768178681222159,
+      "grad_norm": 1.2562459707260132,
+      "learning_rate": 1.5833333333333333e-05,
+      "loss": 0.1051,
+      "step": 190
+    },
+    {
+      "epoch": 0.04893657067570111,
+      "grad_norm": 0.13770118355751038,
+      "learning_rate": 1.6250000000000002e-05,
+      "loss": 0.1049,
+      "step": 195
+    },
+    {
+      "epoch": 0.050191354539180624,
+      "grad_norm": 0.3370998799800873,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 0.1044,
+      "step": 200
+    },
+    {
+      "epoch": 0.051446138402660144,
+      "grad_norm": 0.5696057081222534,
+      "learning_rate": 1.7083333333333333e-05,
+      "loss": 0.1049,
+      "step": 205
+    },
+    {
+      "epoch": 0.052700922266139656,
+      "grad_norm": 0.24888598918914795,
+      "learning_rate": 1.7500000000000002e-05,
+      "loss": 0.1065,
+      "step": 210
+    },
+    {
+      "epoch": 0.053955706129619176,
+      "grad_norm": 0.5320826172828674,
+      "learning_rate": 1.7916666666666667e-05,
+      "loss": 0.1044,
+      "step": 215
+    },
+    {
+      "epoch": 0.05521048999309869,
+      "grad_norm": 0.28355732560157776,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 0.1189,
+      "step": 220
+    },
+    {
+      "epoch": 0.05646527385657821,
+      "grad_norm": 0.20372584462165833,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 0.1038,
+      "step": 225
+    },
+    {
+      "epoch": 0.05772005772005772,
+      "grad_norm": 0.5983103513717651,
+      "learning_rate": 1.916666666666667e-05,
+      "loss": 0.0955,
+      "step": 230
+    },
+    {
+      "epoch": 0.05897484158353723,
+      "grad_norm": 0.36013373732566833,
+      "learning_rate": 1.9583333333333333e-05,
+      "loss": 0.1012,
+      "step": 235
+    },
+    {
+      "epoch": 0.06022962544701675,
+      "grad_norm": 0.20898281037807465,
+      "learning_rate": 2e-05,
+      "loss": 0.0993,
+      "step": 240
+    },
+    {
+      "epoch": 0.061484409310496264,
+      "grad_norm": 0.29076099395751953,
+      "learning_rate": 1.999997934261318e-05,
+      "loss": 0.106,
+      "step": 245
+    },
+    {
+      "epoch": 0.06273919317397578,
+      "grad_norm": 0.8422909379005432,
+      "learning_rate": 1.999991737053805e-05,
+      "loss": 0.0944,
+      "step": 250
+    },
+    {
+      "epoch": 0.0639939770374553,
+      "grad_norm": 0.165451779961586,
+      "learning_rate": 1.999981408403066e-05,
+      "loss": 0.1004,
+      "step": 255
+    },
+    {
+      "epoch": 0.06524876090093482,
+      "grad_norm": 0.14131979644298553,
+      "learning_rate": 1.9999669483517726e-05,
+      "loss": 0.0877,
+      "step": 260
+    },
+    {
+      "epoch": 0.06650354476441434,
+      "grad_norm": 0.2859463095664978,
+      "learning_rate": 1.9999483569596664e-05,
+      "loss": 0.0939,
+      "step": 265
+    },
+    {
+      "epoch": 0.06775832862789384,
+      "grad_norm": 0.26985302567481995,
+      "learning_rate": 1.9999256343035577e-05,
+      "loss": 0.1005,
+      "step": 270
+    },
+    {
+      "epoch": 0.06901311249137336,
+      "grad_norm": 0.16001251339912415,
+      "learning_rate": 1.9998987804773244e-05,
+      "loss": 0.0991,
+      "step": 275
+    },
+    {
+      "epoch": 0.07026789635485288,
+      "grad_norm": 0.28724098205566406,
+      "learning_rate": 1.9998677955919127e-05,
+      "loss": 0.1039,
+      "step": 280
+    },
+    {
+      "epoch": 0.0715226802183324,
+      "grad_norm": 0.10330358892679214,
+      "learning_rate": 1.9998326797753352e-05,
+      "loss": 0.0992,
+      "step": 285
+    },
+    {
+      "epoch": 0.0727774640818119,
+      "grad_norm": 0.630402147769928,
+      "learning_rate": 1.999793433172673e-05,
+      "loss": 0.1055,
+      "step": 290
+    },
+    {
+      "epoch": 0.07403224794529142,
+      "grad_norm": 0.3052836060523987,
+      "learning_rate": 1.9997500559460718e-05,
+      "loss": 0.0966,
+      "step": 295
+    },
+    {
+      "epoch": 0.07528703180877094,
+      "grad_norm": 0.3691750466823578,
+      "learning_rate": 1.999702548274744e-05,
+      "loss": 0.1047,
+      "step": 300
+    },
+    {
+      "epoch": 0.07654181567225045,
+      "grad_norm": 0.47557133436203003,
+      "learning_rate": 1.999650910354967e-05,
+      "loss": 0.1022,
+      "step": 305
+    },
+    {
+      "epoch": 0.07779659953572997,
+      "grad_norm": 0.21415193378925323,
+      "learning_rate": 1.999595142400081e-05,
+      "loss": 0.1075,
+      "step": 310
+    },
+    {
+      "epoch": 0.07905138339920949,
+      "grad_norm": 0.3545961081981659,
+      "learning_rate": 1.99953524464049e-05,
+      "loss": 0.096,
+      "step": 315
+    },
+    {
+      "epoch": 0.08030616726268901,
+      "grad_norm": 0.3949846029281616,
+      "learning_rate": 1.9994712173236604e-05,
+      "loss": 0.0974,
+      "step": 320
+    },
+    {
+      "epoch": 0.08156095112616851,
+      "grad_norm": 1.0949060916900635,
+      "learning_rate": 1.9994030607141196e-05,
+      "loss": 0.0973,
+      "step": 325
+    },
+    {
+      "epoch": 0.08281573498964803,
+      "grad_norm": 0.35101133584976196,
+      "learning_rate": 1.9993307750934555e-05,
+      "loss": 0.0965,
+      "step": 330
+    },
+    {
+      "epoch": 0.08407051885312755,
+      "grad_norm": 0.29526200890541077,
+      "learning_rate": 1.999254360760314e-05,
+      "loss": 0.0898,
+      "step": 335
+    },
+    {
+      "epoch": 0.08532530271660707,
+      "grad_norm": 0.09260907024145126,
+      "learning_rate": 1.999173818030399e-05,
+      "loss": 0.0948,
+      "step": 340
+    },
+    {
+      "epoch": 0.08658008658008658,
+      "grad_norm": 0.2705346345901489,
+      "learning_rate": 1.999089147236472e-05,
+      "loss": 0.0946,
+      "step": 345
+    },
+    {
+      "epoch": 0.0878348704435661,
+      "grad_norm": 0.3629266917705536,
+      "learning_rate": 1.999000348728347e-05,
+      "loss": 0.0999,
+      "step": 350
+    },
+    {
+      "epoch": 0.08908965430704562,
+      "grad_norm": 0.11858992278575897,
+      "learning_rate": 1.9989074228728942e-05,
+      "loss": 0.0931,
+      "step": 355
+    },
+    {
+      "epoch": 0.09034443817052512,
+      "grad_norm": 0.41660797595977783,
+      "learning_rate": 1.9988103700540345e-05,
+      "loss": 0.1047,
+      "step": 360
+    },
+    {
+      "epoch": 0.09159922203400464,
+      "grad_norm": 0.7686080932617188,
+      "learning_rate": 1.9987091906727387e-05,
+      "loss": 0.0968,
+      "step": 365
+    },
+    {
+      "epoch": 0.09285400589748416,
+      "grad_norm": 0.3915267884731293,
+      "learning_rate": 1.998603885147028e-05,
+      "loss": 0.0947,
+      "step": 370
+    },
+    {
+      "epoch": 0.09410878976096368,
+      "grad_norm": 0.23744699358940125,
+      "learning_rate": 1.998494453911969e-05,
+      "loss": 0.0915,
+      "step": 375
+    },
+    {
+      "epoch": 0.09536357362444318,
+      "grad_norm": 0.11716315150260925,
+      "learning_rate": 1.9983808974196752e-05,
+      "loss": 0.1005,
+      "step": 380
+    },
+    {
+      "epoch": 0.0966183574879227,
+      "grad_norm": 0.355939656496048,
+      "learning_rate": 1.9982632161393022e-05,
+      "loss": 0.0992,
+      "step": 385
+    },
+    {
+      "epoch": 0.09787314135140222,
+      "grad_norm": 0.23542310297489166,
+      "learning_rate": 1.9981414105570473e-05,
+      "loss": 0.0864,
+      "step": 390
+    },
+    {
+      "epoch": 0.09912792521488174,
+      "grad_norm": 0.5069738626480103,
+      "learning_rate": 1.9980154811761482e-05,
+      "loss": 0.0929,
+      "step": 395
+    },
+    {
+      "epoch": 0.10038270907836125,
+      "grad_norm": 0.5965291857719421,
+      "learning_rate": 1.9978854285168784e-05,
+      "loss": 0.0979,
+      "step": 400
+    },
+    {
+      "epoch": 0.10163749294184077,
+      "grad_norm": 0.557249128818512,
+      "learning_rate": 1.9977512531165484e-05,
+      "loss": 0.0994,
+      "step": 405
+    },
+    {
+      "epoch": 0.10289227680532029,
+      "grad_norm": 0.3753920793533325,
+      "learning_rate": 1.9976129555295003e-05,
+      "loss": 0.1009,
+      "step": 410
+    },
+    {
+      "epoch": 0.10414706066879979,
+      "grad_norm": 0.18703804910182953,
+      "learning_rate": 1.9974705363271076e-05,
+      "loss": 0.0938,
+      "step": 415
+    },
+    {
+      "epoch": 0.10540184453227931,
+      "grad_norm": 0.49501076340675354,
+      "learning_rate": 1.997323996097772e-05,
+      "loss": 0.083,
+      "step": 420
+    },
+    {
+      "epoch": 0.10665662839575883,
+      "grad_norm": 0.13180793821811676,
+      "learning_rate": 1.9971733354469215e-05,
+      "loss": 0.1087,
+      "step": 425
+    },
+    {
+      "epoch": 0.10791141225923835,
+      "grad_norm": 0.4710563123226166,
+      "learning_rate": 1.9970185549970066e-05,
+      "loss": 0.0977,
+      "step": 430
+    },
+    {
+      "epoch": 0.10916619612271786,
+      "grad_norm": 0.5056783556938171,
+      "learning_rate": 1.9968596553874993e-05,
+      "loss": 0.0947,
+      "step": 435
+    },
+    {
+      "epoch": 0.11042097998619738,
+      "grad_norm": 0.2544513940811157,
+      "learning_rate": 1.99669663727489e-05,
+      "loss": 0.0925,
+      "step": 440
+    },
+    {
+      "epoch": 0.1116757638496769,
+      "grad_norm": 0.512052059173584,
+      "learning_rate": 1.9965295013326843e-05,
+      "loss": 0.0984,
+      "step": 445
+    },
+    {
+      "epoch": 0.11293054771315642,
+      "grad_norm": 0.31233909726142883,
+      "learning_rate": 1.9963582482514003e-05,
+      "loss": 0.0903,
+      "step": 450
+    },
+    {
+      "epoch": 0.11418533157663592,
+      "grad_norm": 0.3753993511199951,
+      "learning_rate": 1.9961828787385662e-05,
+      "loss": 0.0985,
+      "step": 455
+    },
+    {
+      "epoch": 0.11544011544011544,
+      "grad_norm": 0.14997100830078125,
+      "learning_rate": 1.996003393518718e-05,
+      "loss": 0.0991,
+      "step": 460
+    },
+    {
+      "epoch": 0.11669489930359496,
+      "grad_norm": 0.1637381911277771,
+      "learning_rate": 1.995819793333394e-05,
+      "loss": 0.0833,
+      "step": 465
+    },
+    {
+      "epoch": 0.11794968316707446,
+      "grad_norm": 0.35544639825820923,
+      "learning_rate": 1.9956320789411338e-05,
+      "loss": 0.0901,
+      "step": 470
+    },
+    {
+      "epoch": 0.11920446703055398,
+      "grad_norm": 0.29124122858047485,
+      "learning_rate": 1.9954402511174763e-05,
+      "loss": 0.1043,
+      "step": 475
+    },
+    {
+      "epoch": 0.1204592508940335,
+      "grad_norm": 0.1517423689365387,
+      "learning_rate": 1.9952443106549535e-05,
+      "loss": 0.0937,
+      "step": 480
+    },
+    {
+      "epoch": 0.12171403475751302,
+      "grad_norm": 0.22879864275455475,
+      "learning_rate": 1.9950442583630884e-05,
+      "loss": 0.0983,
+      "step": 485
+    },
+    {
+      "epoch": 0.12296881862099253,
+      "grad_norm": 0.12301073968410492,
+      "learning_rate": 1.9948400950683932e-05,
+      "loss": 0.0937,
+      "step": 490
+    },
+    {
+      "epoch": 0.12422360248447205,
+      "grad_norm": 0.46629127860069275,
+      "learning_rate": 1.9946318216143633e-05,
+      "loss": 0.109,
+      "step": 495
+    },
+    {
+      "epoch": 0.12547838634795155,
+      "grad_norm": 0.34613123536109924,
+      "learning_rate": 1.9944194388614764e-05,
+      "loss": 0.0979,
+      "step": 500
+    },
+    {
+      "epoch": 0.12673317021143107,
+      "grad_norm": 0.3203139901161194,
+      "learning_rate": 1.9942029476871868e-05,
+      "loss": 0.105,
+      "step": 505
+    },
+    {
+      "epoch": 0.1279879540749106,
+      "grad_norm": 0.23387163877487183,
+      "learning_rate": 1.9939823489859226e-05,
+      "loss": 0.0918,
+      "step": 510
+    },
+    {
+      "epoch": 0.1292427379383901,
+      "grad_norm": 0.5645913481712341,
+      "learning_rate": 1.9937576436690822e-05,
+      "loss": 0.1008,
+      "step": 515
+    },
+    {
+      "epoch": 0.13049752180186963,
+      "grad_norm": 0.758552074432373,
+      "learning_rate": 1.9935288326650314e-05,
+      "loss": 0.1041,
+      "step": 520
+    },
+    {
+      "epoch": 0.13175230566534915,
+      "grad_norm": 0.40192508697509766,
+      "learning_rate": 1.993295916919097e-05,
+      "loss": 0.0961,
+      "step": 525
+    },
+    {
+      "epoch": 0.13300708952882867,
+      "grad_norm": 0.3283374309539795,
+      "learning_rate": 1.9930588973935653e-05,
+      "loss": 0.1099,
+      "step": 530
+    },
+    {
+      "epoch": 0.13426187339230816,
+      "grad_norm": 0.13031624257564545,
+      "learning_rate": 1.992817775067677e-05,
+      "loss": 0.0964,
+      "step": 535
+    },
+    {
+      "epoch": 0.13551665725578768,
+      "grad_norm": 0.40052691102027893,
+      "learning_rate": 1.9925725509376236e-05,
+      "loss": 0.1003,
+      "step": 540
+    },
+    {
+      "epoch": 0.1367714411192672,
+      "grad_norm": 0.20770736038684845,
+      "learning_rate": 1.992323226016543e-05,
+      "loss": 0.0977,
+      "step": 545
+    },
+    {
+      "epoch": 0.13802622498274672,
+      "grad_norm": 0.5854383111000061,
+      "learning_rate": 1.9920698013345162e-05,
+      "loss": 0.096,
+      "step": 550
+    },
+    {
+      "epoch": 0.13928100884622624,
+      "grad_norm": 0.33884522318840027,
+      "learning_rate": 1.99181227793856e-05,
+      "loss": 0.0981,
+      "step": 555
+    },
+    {
+      "epoch": 0.14053579270970576,
+      "grad_norm": 0.9304419159889221,
+      "learning_rate": 1.9915506568926283e-05,
+      "loss": 0.1069,
+      "step": 560
+    },
+    {
+      "epoch": 0.14179057657318528,
+      "grad_norm": 0.5028262138366699,
+      "learning_rate": 1.991284939277601e-05,
+      "loss": 0.1064,
+      "step": 565
+    },
+    {
+      "epoch": 0.1430453604366648,
+      "grad_norm": 0.30051878094673157,
+      "learning_rate": 1.991015126191285e-05,
+      "loss": 0.0977,
+      "step": 570
+    },
+    {
+      "epoch": 0.1443001443001443,
+      "grad_norm": 0.5658770799636841,
+      "learning_rate": 1.990741218748407e-05,
+      "loss": 0.1132,
+      "step": 575
+    },
+    {
+      "epoch": 0.1455549281636238,
+      "grad_norm": 0.1313030868768692,
+      "learning_rate": 1.9904632180806094e-05,
+      "loss": 0.0976,
+      "step": 580
+    },
+    {
+      "epoch": 0.14680971202710333,
+      "grad_norm": 0.3436416685581207,
+      "learning_rate": 1.9901811253364458e-05,
+      "loss": 0.103,
+      "step": 585
+    },
+    {
+      "epoch": 0.14806449589058285,
+      "grad_norm": 0.7141314744949341,
+      "learning_rate": 1.9898949416813757e-05,
+      "loss": 0.0994,
+      "step": 590
+    },
+    {
+      "epoch": 0.14931927975406237,
+      "grad_norm": 0.35225412249565125,
+      "learning_rate": 1.9896046682977603e-05,
+      "loss": 0.0989,
+      "step": 595
+    },
+    {
+      "epoch": 0.1505740636175419,
+      "grad_norm": 0.19844387471675873,
+      "learning_rate": 1.989310306384858e-05,
+      "loss": 0.1032,
+      "step": 600
+    },
+    {
+      "epoch": 0.1518288474810214,
+      "grad_norm": 0.21333719789981842,
+      "learning_rate": 1.989011857158818e-05,
+      "loss": 0.1,
+      "step": 605
+    },
+    {
+      "epoch": 0.1530836313445009,
+      "grad_norm": 0.3265141546726227,
+      "learning_rate": 1.9887093218526768e-05,
+      "loss": 0.0935,
+      "step": 610
+    },
+    {
+      "epoch": 0.15433841520798042,
+      "grad_norm": 0.24291647970676422,
+      "learning_rate": 1.9884027017163515e-05,
+      "loss": 0.0923,
+      "step": 615
+    },
+    {
+      "epoch": 0.15559319907145994,
+      "grad_norm": 0.4325827658176422,
+      "learning_rate": 1.9880919980166374e-05,
+      "loss": 0.089,
+      "step": 620
+    },
+    {
+      "epoch": 0.15684798293493946,
+      "grad_norm": 0.1882610321044922,
+      "learning_rate": 1.9877772120371986e-05,
+      "loss": 0.0893,
+      "step": 625
+    },
+    {
+      "epoch": 0.15810276679841898,
+      "grad_norm": 0.4826973080635071,
+      "learning_rate": 1.987458345078567e-05,
+      "loss": 0.0983,
+      "step": 630
+    },
+    {
+      "epoch": 0.1593575506618985,
+      "grad_norm": 0.5353069305419922,
+      "learning_rate": 1.9871353984581342e-05,
+      "loss": 0.0937,
+      "step": 635
+    },
+    {
+      "epoch": 0.16061233452537801,
+      "grad_norm": 0.12057194858789444,
+      "learning_rate": 1.9868083735101464e-05,
+      "loss": 0.0999,
+      "step": 640
+    },
+    {
+      "epoch": 0.1618671183888575,
+      "grad_norm": 0.4315410256385803,
+      "learning_rate": 1.9864772715857e-05,
+      "loss": 0.0982,
+      "step": 645
+    },
+    {
+      "epoch": 0.16312190225233703,
+      "grad_norm": 0.1834162324666977,
+      "learning_rate": 1.9861420940527357e-05,
+      "loss": 0.0876,
+      "step": 650
+    },
+    {
+      "epoch": 0.16437668611581654,
+      "grad_norm": 0.1956656128168106,
+      "learning_rate": 1.985802842296031e-05,
+      "loss": 0.0833,
+      "step": 655
+    },
+    {
+      "epoch": 0.16563146997929606,
+      "grad_norm": 0.08431841433048248,
+      "learning_rate": 1.9854595177171968e-05,
+      "loss": 0.0933,
+      "step": 660
+    },
+    {
+      "epoch": 0.16688625384277558,
+      "grad_norm": 0.40520185232162476,
+      "learning_rate": 1.9851121217346717e-05,
+      "loss": 0.0887,
+      "step": 665
+    },
+    {
+      "epoch": 0.1681410377062551,
+      "grad_norm": 0.661037266254425,
+      "learning_rate": 1.9847606557837138e-05,
+      "loss": 0.0971,
+      "step": 670
+    },
+    {
+      "epoch": 0.16939582156973462,
+      "grad_norm": 0.35876908898353577,
+      "learning_rate": 1.9844051213163967e-05,
+      "loss": 0.0891,
+      "step": 675
+    },
+    {
+      "epoch": 0.17065060543321414,
+      "grad_norm": 0.73444002866745,
+      "learning_rate": 1.9840455198016033e-05,
+      "loss": 0.1015,
+      "step": 680
+    },
+    {
+      "epoch": 0.17190538929669363,
+      "grad_norm": 0.5403853058815002,
+      "learning_rate": 1.9836818527250185e-05,
+      "loss": 0.094,
+      "step": 685
+    },
+    {
+      "epoch": 0.17316017316017315,
+      "grad_norm": 0.26550817489624023,
+      "learning_rate": 1.9833141215891253e-05,
+      "loss": 0.0916,
+      "step": 690
+    },
+    {
+      "epoch": 0.17441495702365267,
+      "grad_norm": 0.20519235730171204,
+      "learning_rate": 1.9829423279131962e-05,
+      "loss": 0.0938,
+      "step": 695
+    },
+    {
+      "epoch": 0.1756697408871322,
+      "grad_norm": 0.2511141896247864,
+      "learning_rate": 1.9825664732332886e-05,
+      "loss": 0.095,
+      "step": 700
+    },
+    {
+      "epoch": 0.1769245247506117,
+      "grad_norm": 0.42727699875831604,
+      "learning_rate": 1.982186559102237e-05,
+      "loss": 0.1039,
+      "step": 705
+    },
+    {
+      "epoch": 0.17817930861409123,
+      "grad_norm": 0.361933171749115,
+      "learning_rate": 1.9818025870896485e-05,
+      "loss": 0.0933,
+      "step": 710
+    },
+    {
+      "epoch": 0.17943409247757075,
+      "grad_norm": 0.081818588078022,
+      "learning_rate": 1.981414558781895e-05,
+      "loss": 0.0972,
+      "step": 715
+    },
+    {
+      "epoch": 0.18068887634105024,
+      "grad_norm": 0.24198131263256073,
+      "learning_rate": 1.9810224757821063e-05,
+      "loss": 0.08,
+      "step": 720
+    },
+    {
+      "epoch": 0.18194366020452976,
+      "grad_norm": 0.22726424038410187,
+      "learning_rate": 1.9806263397101645e-05,
+      "loss": 0.0889,
+      "step": 725
+    },
+    {
+      "epoch": 0.18319844406800928,
+      "grad_norm": 0.26447054743766785,
+      "learning_rate": 1.980226152202697e-05,
+      "loss": 0.0988,
+      "step": 730
+    },
+    {
+      "epoch": 0.1844532279314888,
+      "grad_norm": 0.20784974098205566,
+      "learning_rate": 1.9798219149130692e-05,
+      "loss": 0.1134,
+      "step": 735
+    },
+    {
+      "epoch": 0.18570801179496832,
+      "grad_norm": 0.6903693079948425,
+      "learning_rate": 1.9794136295113783e-05,
+      "loss": 0.1002,
+      "step": 740
+    },
+    {
+      "epoch": 0.18696279565844784,
+      "grad_norm": 0.9974376559257507,
+      "learning_rate": 1.9790012976844465e-05,
+      "loss": 0.0922,
+      "step": 745
+    },
+    {
+      "epoch": 0.18821757952192736,
+      "grad_norm": 0.3469702899456024,
+      "learning_rate": 1.9785849211358133e-05,
+      "loss": 0.0843,
+      "step": 750
+    },
+    {
+      "epoch": 0.18947236338540685,
+      "grad_norm": 0.25853797793388367,
+      "learning_rate": 1.9781645015857287e-05,
+      "loss": 0.0832,
+      "step": 755
+    },
+    {
+      "epoch": 0.19072714724888637,
+      "grad_norm": 0.4150042235851288,
+      "learning_rate": 1.9777400407711467e-05,
+      "loss": 0.0951,
+      "step": 760
+    },
+    {
+      "epoch": 0.1919819311123659,
+      "grad_norm": 0.8367013335227966,
+      "learning_rate": 1.9773115404457175e-05,
+      "loss": 0.1043,
+      "step": 765
+    },
+    {
+      "epoch": 0.1932367149758454,
+      "grad_norm": 0.09506336599588394,
+      "learning_rate": 1.976879002379781e-05,
+      "loss": 0.1006,
+      "step": 770
+    },
+    {
+      "epoch": 0.19449149883932493,
+      "grad_norm": 0.47608646750450134,
+      "learning_rate": 1.9764424283603577e-05,
+      "loss": 0.0969,
+      "step": 775
+    },
+    {
+      "epoch": 0.19574628270280445,
+      "grad_norm": 0.3089286684989929,
+      "learning_rate": 1.976001820191143e-05,
+      "loss": 0.081,
+      "step": 780
+    },
+    {
+      "epoch": 0.19700106656628397,
+      "grad_norm": 0.7951834201812744,
+      "learning_rate": 1.9755571796925014e-05,
+      "loss": 0.0987,
+      "step": 785
+    },
+    {
+      "epoch": 0.19825585042976349,
+      "grad_norm": 0.29604893922805786,
+      "learning_rate": 1.9751085087014533e-05,
+      "loss": 0.0905,
+      "step": 790
+    },
+    {
+      "epoch": 0.19951063429324298,
+      "grad_norm": 0.6843680143356323,
+      "learning_rate": 1.974655809071673e-05,
+      "loss": 0.1044,
+      "step": 795
+    },
+    {
+      "epoch": 0.2007654181567225,
+      "grad_norm": 0.3378108739852905,
+      "learning_rate": 1.9741990826734793e-05,
+      "loss": 0.0791,
+      "step": 800
+    },
+    {
+      "epoch": 0.20202020202020202,
+      "grad_norm": 0.2935211658477783,
+      "learning_rate": 1.9737383313938266e-05,
+      "loss": 0.0996,
+      "step": 805
+    },
+    {
+      "epoch": 0.20327498588368154,
+      "grad_norm": 0.2529759705066681,
+      "learning_rate": 1.9732735571362985e-05,
+      "loss": 0.0992,
+      "step": 810
+    },
+    {
+      "epoch": 0.20452976974716106,
+      "grad_norm": 0.2465377002954483,
+      "learning_rate": 1.9728047618210995e-05,
+      "loss": 0.0976,
+      "step": 815
+    },
+    {
+      "epoch": 0.20578455361064057,
+      "grad_norm": 0.4152264893054962,
+      "learning_rate": 1.9723319473850465e-05,
+      "loss": 0.1063,
+      "step": 820
+    },
+    {
+      "epoch": 0.2070393374741201,
+      "grad_norm": 0.3043062686920166,
+      "learning_rate": 1.971855115781562e-05,
+      "loss": 0.1005,
+      "step": 825
+    },
+    {
+      "epoch": 0.20829412133759959,
+      "grad_norm": 0.15693755447864532,
+      "learning_rate": 1.9713742689806646e-05,
+      "loss": 0.0905,
+      "step": 830
+    },
+    {
+      "epoch": 0.2095489052010791,
+      "grad_norm": 0.21518734097480774,
+      "learning_rate": 1.9708894089689622e-05,
+      "loss": 0.0952,
+      "step": 835
+    },
+    {
+      "epoch": 0.21080368906455862,
+      "grad_norm": 0.5088281631469727,
+      "learning_rate": 1.9704005377496428e-05,
+      "loss": 0.0844,
+      "step": 840
+    },
+    {
+      "epoch": 0.21205847292803814,
+      "grad_norm": 0.19282662868499756,
+      "learning_rate": 1.969907657342467e-05,
+      "loss": 0.0892,
+      "step": 845
+    },
+    {
+      "epoch": 0.21331325679151766,
+      "grad_norm": 0.3823853135108948,
+      "learning_rate": 1.969410769783759e-05,
+      "loss": 0.083,
+      "step": 850
+    },
+    {
+      "epoch": 0.21456804065499718,
+      "grad_norm": 0.31330108642578125,
+      "learning_rate": 1.9689098771263982e-05,
+      "loss": 0.0958,
+      "step": 855
+    },
+    {
+      "epoch": 0.2158228245184767,
+      "grad_norm": 0.2921331822872162,
+      "learning_rate": 1.968404981439812e-05,
+      "loss": 0.0915,
+      "step": 860
+    },
+    {
+      "epoch": 0.21707760838195622,
+      "grad_norm": 0.23992085456848145,
+      "learning_rate": 1.9678960848099646e-05,
+      "loss": 0.0945,
+      "step": 865
+    },
+    {
+      "epoch": 0.2183323922454357,
+      "grad_norm": 0.22444355487823486,
+      "learning_rate": 1.967383189339352e-05,
+      "loss": 0.0816,
+      "step": 870
+    },
+    {
+      "epoch": 0.21958717610891523,
+      "grad_norm": 0.27808722853660583,
+      "learning_rate": 1.9668662971469886e-05,
+      "loss": 0.0872,
+      "step": 875
+    },
+    {
+      "epoch": 0.22084195997239475,
+      "grad_norm": 0.4084593653678894,
+      "learning_rate": 1.9663454103684043e-05,
+      "loss": 0.1007,
+      "step": 880
+    },
+    {
+      "epoch": 0.22209674383587427,
+      "grad_norm": 0.2565450966358185,
+      "learning_rate": 1.9658205311556304e-05,
+      "loss": 0.0973,
+      "step": 885
+    },
+    {
+      "epoch": 0.2233515276993538,
+      "grad_norm": 0.19569608569145203,
+      "learning_rate": 1.9652916616771933e-05,
+      "loss": 0.0884,
+      "step": 890
+    },
+    {
+      "epoch": 0.2246063115628333,
+      "grad_norm": 0.17645685374736786,
+      "learning_rate": 1.9647588041181057e-05,
+      "loss": 0.0918,
+      "step": 895
+    },
+    {
+      "epoch": 0.22586109542631283,
+      "grad_norm": 0.5224778056144714,
+      "learning_rate": 1.9642219606798566e-05,
+      "loss": 0.0899,
+      "step": 900
+    },
+    {
+      "epoch": 0.22711587928979232,
+      "grad_norm": 0.29252490401268005,
+      "learning_rate": 1.963681133580402e-05,
+      "loss": 0.0978,
+      "step": 905
+    },
+    {
+      "epoch": 0.22837066315327184,
+      "grad_norm": 0.2631376385688782,
+      "learning_rate": 1.9631363250541577e-05,
+      "loss": 0.0961,
+      "step": 910
+    },
+    {
+      "epoch": 0.22962544701675136,
+      "grad_norm": 0.4589993953704834,
+      "learning_rate": 1.9625875373519866e-05,
+      "loss": 0.0942,
+      "step": 915
+    },
+    {
+      "epoch": 0.23088023088023088,
+      "grad_norm": 0.4137325882911682,
+      "learning_rate": 1.9620347727411933e-05,
+      "loss": 0.0917,
+      "step": 920
+    },
+    {
+      "epoch": 0.2321350147437104,
+      "grad_norm": 0.21914274990558624,
+      "learning_rate": 1.9614780335055127e-05,
+      "loss": 0.0851,
+      "step": 925
+    },
+    {
+      "epoch": 0.23338979860718992,
+      "grad_norm": 0.3876037001609802,
+      "learning_rate": 1.9609173219450998e-05,
+      "loss": 0.0878,
+      "step": 930
+    },
+    {
+      "epoch": 0.23464458247066944,
+      "grad_norm": 0.376610666513443,
+      "learning_rate": 1.9603526403765218e-05,
+      "loss": 0.0949,
+      "step": 935
+    },
+    {
+      "epoch": 0.23589936633414893,
+      "grad_norm": 0.2858617603778839,
+      "learning_rate": 1.9597839911327475e-05,
+      "loss": 0.091,
+      "step": 940
+    },
+    {
+      "epoch": 0.23715415019762845,
+      "grad_norm": 0.6040699481964111,
+      "learning_rate": 1.959211376563139e-05,
+      "loss": 0.0866,
+      "step": 945
+    },
+    {
+      "epoch": 0.23840893406110797,
+      "grad_norm": 0.15516065061092377,
+      "learning_rate": 1.9586347990334406e-05,
+      "loss": 0.093,
+      "step": 950
+    },
+    {
+      "epoch": 0.2396637179245875,
+      "grad_norm": 0.38089755177497864,
+      "learning_rate": 1.958054260925768e-05,
+      "loss": 0.091,
+      "step": 955
+    },
+    {
+      "epoch": 0.240918501788067,
+      "grad_norm": 0.15390656888484955,
+      "learning_rate": 1.9574697646386027e-05,
+      "loss": 0.0883,
+      "step": 960
+    },
+    {
+      "epoch": 0.24217328565154653,
+      "grad_norm": 0.5063360929489136,
+      "learning_rate": 1.956881312586777e-05,
+      "loss": 0.094,
+      "step": 965
+    },
+    {
+      "epoch": 0.24342806951502605,
+      "grad_norm": 0.15385521948337555,
+      "learning_rate": 1.9562889072014682e-05,
+      "loss": 0.0942,
+      "step": 970
+    },
+    {
+      "epoch": 0.24468285337850557,
+      "grad_norm": 0.4311351180076599,
+      "learning_rate": 1.9556925509301844e-05,
+      "loss": 0.0896,
+      "step": 975
+    },
+    {
+      "epoch": 0.24593763724198506,
+      "grad_norm": 0.1626339703798294,
+      "learning_rate": 1.955092246236759e-05,
+      "loss": 0.0901,
+      "step": 980
+    },
+    {
+      "epoch": 0.24719242110546458,
+      "grad_norm": 0.1978272795677185,
+      "learning_rate": 1.954487995601337e-05,
+      "loss": 0.086,
+      "step": 985
+    },
+    {
+      "epoch": 0.2484472049689441,
+      "grad_norm": 0.5115323066711426,
+      "learning_rate": 1.953879801520366e-05,
+      "loss": 0.0966,
+      "step": 990
+    },
+    {
+      "epoch": 0.24970198883242362,
+      "grad_norm": 0.556075930595398,
+      "learning_rate": 1.9532676665065863e-05,
+      "loss": 0.0947,
+      "step": 995
+    },
+    {
+      "epoch": 0.2509567726959031,
+      "grad_norm": 0.4283283054828644,
+      "learning_rate": 1.9526515930890203e-05,
+      "loss": 0.0995,
+      "step": 1000
+    },
+    {
+      "epoch": 0.25221155655938265,
+      "grad_norm": 0.20551514625549316,
+      "learning_rate": 1.9520315838129602e-05,
+      "loss": 0.0822,
+      "step": 1005
+    },
+    {
+      "epoch": 0.25346634042286215,
+      "grad_norm": 0.2339772880077362,
+      "learning_rate": 1.9514076412399615e-05,
+      "loss": 0.0916,
+      "step": 1010
+    },
+    {
+      "epoch": 0.2547211242863417,
+      "grad_norm": 0.33765801787376404,
+      "learning_rate": 1.9507797679478282e-05,
+      "loss": 0.0834,
+      "step": 1015
+    },
+    {
+      "epoch": 0.2559759081498212,
+      "grad_norm": 0.15917867422103882,
+      "learning_rate": 1.9501479665306046e-05,
+      "loss": 0.1036,
+      "step": 1020
+    },
+    {
+      "epoch": 0.25723069201330073,
+      "grad_norm": 0.15824545919895172,
+      "learning_rate": 1.9495122395985642e-05,
+      "loss": 0.0926,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2584854758767802,
+      "grad_norm": 0.22190357744693756,
+      "learning_rate": 1.948872589778198e-05,
+      "loss": 0.0915,
+      "step": 1030
+    },
+    {
+      "epoch": 0.2597402597402597,
+      "grad_norm": 0.39479732513427734,
+      "learning_rate": 1.9482290197122054e-05,
+      "loss": 0.0954,
+      "step": 1035
+    },
+    {
+      "epoch": 0.26099504360373926,
+      "grad_norm": 0.44342002272605896,
+      "learning_rate": 1.947581532059481e-05,
+      "loss": 0.0968,
+      "step": 1040
+    },
+    {
+      "epoch": 0.26224982746721875,
+      "grad_norm": 0.18394285440444946,
+      "learning_rate": 1.946930129495106e-05,
+      "loss": 0.0968,
+      "step": 1045
+    },
+    {
+      "epoch": 0.2635046113306983,
+      "grad_norm": 0.40263161063194275,
+      "learning_rate": 1.9462748147103342e-05,
+      "loss": 0.0954,
+      "step": 1050
+    },
+    {
+      "epoch": 0.2647593951941778,
+      "grad_norm": 0.3291696608066559,
+      "learning_rate": 1.9456155904125853e-05,
+      "loss": 0.0974,
+      "step": 1055
+    },
+    {
+      "epoch": 0.26601417905765734,
+      "grad_norm": 0.17444024980068207,
+      "learning_rate": 1.9449524593254283e-05,
+      "loss": 0.0892,
+      "step": 1060
+    },
+    {
+      "epoch": 0.26726896292113683,
+      "grad_norm": 0.5202435851097107,
+      "learning_rate": 1.944285424188575e-05,
+      "loss": 0.1026,
+      "step": 1065
+    },
+    {
+      "epoch": 0.2685237467846163,
+      "grad_norm": 0.6516605615615845,
+      "learning_rate": 1.943614487757866e-05,
+      "loss": 0.0899,
+      "step": 1070
+    },
+    {
+      "epoch": 0.26977853064809587,
+      "grad_norm": 0.7146198153495789,
+      "learning_rate": 1.9429396528052594e-05,
+      "loss": 0.0834,
+      "step": 1075
+    },
+    {
+      "epoch": 0.27103331451157536,
+      "grad_norm": 0.46206018328666687,
+      "learning_rate": 1.9422609221188208e-05,
+      "loss": 0.0961,
+      "step": 1080
+    },
+    {
+      "epoch": 0.2722880983750549,
+      "grad_norm": 0.258039265871048,
+      "learning_rate": 1.9415782985027105e-05,
+      "loss": 0.0861,
+      "step": 1085
+    },
+    {
+      "epoch": 0.2735428822385344,
+      "grad_norm": 0.2528323829174042,
+      "learning_rate": 1.9408917847771732e-05,
+      "loss": 0.0903,
+      "step": 1090
+    },
+    {
+      "epoch": 0.27479766610201395,
+      "grad_norm": 0.35763663053512573,
+      "learning_rate": 1.9402013837785242e-05,
+      "loss": 0.0961,
+      "step": 1095
+    },
+    {
+      "epoch": 0.27605244996549344,
+      "grad_norm": 0.33493563532829285,
+      "learning_rate": 1.93950709835914e-05,
+      "loss": 0.086,
+      "step": 1100
+    },
+    {
+      "epoch": 0.27730723382897293,
+      "grad_norm": 0.6536179780960083,
+      "learning_rate": 1.9388089313874447e-05,
+      "loss": 0.083,
+      "step": 1105
+    },
+    {
+      "epoch": 0.2785620176924525,
+      "grad_norm": 0.6487680077552795,
+      "learning_rate": 1.9381068857478994e-05,
+      "loss": 0.0968,
+      "step": 1110
+    },
+    {
+      "epoch": 0.27981680155593197,
+      "grad_norm": 0.25090888142585754,
+      "learning_rate": 1.9374009643409895e-05,
+      "loss": 0.0785,
+      "step": 1115
+    },
+    {
+      "epoch": 0.2810715854194115,
+      "grad_norm": 0.1529439091682434,
+      "learning_rate": 1.9366911700832146e-05,
+      "loss": 0.0946,
+      "step": 1120
+    },
+    {
+      "epoch": 0.282326369282891,
+      "grad_norm": 0.2744344472885132,
+      "learning_rate": 1.935977505907072e-05,
+      "loss": 0.0827,
+      "step": 1125
+    },
+    {
+      "epoch": 0.28358115314637056,
+      "grad_norm": 0.4758091866970062,
+      "learning_rate": 1.93525997476105e-05,
+      "loss": 0.0895,
+      "step": 1130
+    },
+    {
+      "epoch": 0.28483593700985005,
+      "grad_norm": 0.5062987804412842,
+      "learning_rate": 1.9345385796096118e-05,
+      "loss": 0.1007,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2860907208733296,
+      "grad_norm": 0.44954121112823486,
+      "learning_rate": 1.933813323433186e-05,
+      "loss": 0.0915,
+      "step": 1140
+    },
+    {
+      "epoch": 0.2873455047368091,
+      "grad_norm": 0.2048788219690323,
+      "learning_rate": 1.9330842092281508e-05,
+      "loss": 0.0866,
+      "step": 1145
+    },
+    {
+      "epoch": 0.2886002886002886,
+      "grad_norm": 0.3660197854042053,
+      "learning_rate": 1.9323512400068262e-05,
+      "loss": 0.0916,
+      "step": 1150
+    },
+    {
+      "epoch": 0.2898550724637681,
+      "grad_norm": 0.5215709209442139,
+      "learning_rate": 1.931614418797457e-05,
+      "loss": 0.096,
+      "step": 1155
+    },
+    {
+      "epoch": 0.2911098563272476,
+      "grad_norm": 0.46644458174705505,
+      "learning_rate": 1.9308737486442045e-05,
+      "loss": 0.1107,
+      "step": 1160
+    },
+    {
+      "epoch": 0.29236464019072717,
+      "grad_norm": 0.24164815247058868,
+      "learning_rate": 1.9301292326071295e-05,
+      "loss": 0.0923,
+      "step": 1165
+    },
+    {
+      "epoch": 0.29361942405420666,
+      "grad_norm": 0.42168012261390686,
+      "learning_rate": 1.9293808737621837e-05,
+      "loss": 0.0998,
+      "step": 1170
+    },
+    {
+      "epoch": 0.2948742079176862,
+      "grad_norm": 0.37700730562210083,
+      "learning_rate": 1.9286286752011948e-05,
+      "loss": 0.1036,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2961289917811657,
+      "grad_norm": 0.6957244277000427,
+      "learning_rate": 1.927872640031854e-05,
+      "loss": 0.0903,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2973837756446452,
+      "grad_norm": 0.30158841609954834,
+      "learning_rate": 1.9271127713777033e-05,
+      "loss": 0.0861,
+      "step": 1185
+    },
+    {
+      "epoch": 0.29863855950812473,
+      "grad_norm": 0.41035765409469604,
+      "learning_rate": 1.9263490723781233e-05,
+      "loss": 0.0909,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2998933433716042,
+      "grad_norm": 0.1835126429796219,
+      "learning_rate": 1.9255815461883184e-05,
+      "loss": 0.0968,
+      "step": 1195
+    },
+    {
+      "epoch": 0.3011481272350838,
+      "grad_norm": 0.18504665791988373,
+      "learning_rate": 1.9248101959793066e-05,
+      "loss": 0.0995,
+      "step": 1200
+    },
+    {
+      "epoch": 0.30240291109856327,
+      "grad_norm": 0.32556024193763733,
+      "learning_rate": 1.9240350249379035e-05,
+      "loss": 0.0972,
+      "step": 1205
+    },
+    {
+      "epoch": 0.3036576949620428,
+      "grad_norm": 0.2088983952999115,
+      "learning_rate": 1.92325603626671e-05,
+      "loss": 0.0935,
+      "step": 1210
+    },
+    {
+      "epoch": 0.3049124788255223,
+      "grad_norm": 0.5765480995178223,
+      "learning_rate": 1.922473233184101e-05,
+      "loss": 0.0933,
+      "step": 1215
+    },
+    {
+      "epoch": 0.3061672626890018,
+      "grad_norm": 0.7290382981300354,
+      "learning_rate": 1.9216866189242095e-05,
+      "loss": 0.0954,
+      "step": 1220
+    },
+    {
+      "epoch": 0.30742204655248134,
+      "grad_norm": 0.2641292214393616,
+      "learning_rate": 1.9208961967369148e-05,
+      "loss": 0.0796,
+      "step": 1225
+    },
+    {
+      "epoch": 0.30867683041596083,
+      "grad_norm": 0.6574556827545166,
+      "learning_rate": 1.9201019698878272e-05,
+      "loss": 0.1038,
+      "step": 1230
+    },
+    {
+      "epoch": 0.3099316142794404,
+      "grad_norm": 0.19239722192287445,
+      "learning_rate": 1.9193039416582785e-05,
+      "loss": 0.094,
+      "step": 1235
+    },
+    {
+      "epoch": 0.3111863981429199,
+      "grad_norm": 0.14215205609798431,
+      "learning_rate": 1.918502115345303e-05,
+      "loss": 0.0965,
+      "step": 1240
+    },
+    {
+      "epoch": 0.3124411820063994,
+      "grad_norm": 0.3955984115600586,
+      "learning_rate": 1.9176964942616286e-05,
+      "loss": 0.0961,
+      "step": 1245
+    },
+    {
+      "epoch": 0.3136959658698789,
+      "grad_norm": 0.5442742109298706,
+      "learning_rate": 1.9168870817356602e-05,
+      "loss": 0.0844,
+      "step": 1250
+    },
+    {
+      "epoch": 0.3149507497333584,
+      "grad_norm": 0.2363368719816208,
+      "learning_rate": 1.916073881111468e-05,
+      "loss": 0.0882,
+      "step": 1255
+    },
+    {
+      "epoch": 0.31620553359683795,
+      "grad_norm": 0.4277198016643524,
+      "learning_rate": 1.915256895748771e-05,
+      "loss": 0.0813,
+      "step": 1260
+    },
+    {
+      "epoch": 0.31746031746031744,
+      "grad_norm": 0.15375563502311707,
+      "learning_rate": 1.9144361290229266e-05,
+      "loss": 0.1007,
+      "step": 1265
+    },
+    {
+      "epoch": 0.318715101323797,
+      "grad_norm": 0.2776433527469635,
+      "learning_rate": 1.913611584324913e-05,
+      "loss": 0.0898,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3199698851872765,
+      "grad_norm": 0.20093151926994324,
+      "learning_rate": 1.912783265061319e-05,
+      "loss": 0.0897,
+      "step": 1275
+    },
+    {
+      "epoch": 0.32122466905075603,
+      "grad_norm": 0.13896812498569489,
+      "learning_rate": 1.9119511746543265e-05,
+      "loss": 0.0945,
+      "step": 1280
+    },
+    {
+      "epoch": 0.3224794529142355,
+      "grad_norm": 0.1117207333445549,
+      "learning_rate": 1.911115316541698e-05,
+      "loss": 0.0887,
+      "step": 1285
+    },
+    {
+      "epoch": 0.323734236777715,
+      "grad_norm": 0.19571149349212646,
+      "learning_rate": 1.9102756941767625e-05,
+      "loss": 0.086,
+      "step": 1290
+    },
+    {
+      "epoch": 0.32498902064119456,
+      "grad_norm": 0.12047363817691803,
+      "learning_rate": 1.9094323110284006e-05,
+      "loss": 0.1008,
+      "step": 1295
+    },
+    {
+      "epoch": 0.32624380450467405,
+      "grad_norm": 0.2128770500421524,
+      "learning_rate": 1.9085851705810307e-05,
+      "loss": 0.1054,
+      "step": 1300
+    },
+    {
+      "epoch": 0.3274985883681536,
+      "grad_norm": 0.18241125345230103,
+      "learning_rate": 1.907734276334595e-05,
+      "loss": 0.0981,
+      "step": 1305
+    },
+    {
+      "epoch": 0.3287533722316331,
+      "grad_norm": 0.32163918018341064,
+      "learning_rate": 1.9068796318045434e-05,
+      "loss": 0.0871,
+      "step": 1310
+    },
+    {
+      "epoch": 0.33000815609511264,
+      "grad_norm": 0.6417638063430786,
+      "learning_rate": 1.90602124052182e-05,
+      "loss": 0.0959,
+      "step": 1315
+    },
+    {
+      "epoch": 0.33126293995859213,
+      "grad_norm": 0.2659170627593994,
+      "learning_rate": 1.9051591060328496e-05,
+      "loss": 0.0938,
+      "step": 1320
+    },
+    {
+      "epoch": 0.3325177238220717,
+      "grad_norm": 0.34006327390670776,
+      "learning_rate": 1.904293231899521e-05,
+      "loss": 0.0835,
+      "step": 1325
+    },
+    {
+      "epoch": 0.33377250768555117,
+      "grad_norm": 0.44254758954048157,
+      "learning_rate": 1.9034236216991738e-05,
+      "loss": 0.0894,
+      "step": 1330
+    },
+    {
+      "epoch": 0.33502729154903066,
+      "grad_norm": 0.2562454342842102,
+      "learning_rate": 1.9025502790245824e-05,
+      "loss": 0.0953,
+      "step": 1335
+    },
+    {
+      "epoch": 0.3362820754125102,
+      "grad_norm": 0.33543840050697327,
+      "learning_rate": 1.901673207483943e-05,
+      "loss": 0.0911,
+      "step": 1340
+    },
+    {
+      "epoch": 0.3375368592759897,
+      "grad_norm": 0.3265765905380249,
+      "learning_rate": 1.9007924107008563e-05,
+      "loss": 0.0885,
+      "step": 1345
+    },
+    {
+      "epoch": 0.33879164313946925,
+      "grad_norm": 0.08985516428947449,
+      "learning_rate": 1.8999078923143142e-05,
+      "loss": 0.085,
+      "step": 1350
+    },
+    {
+      "epoch": 0.34004642700294874,
+      "grad_norm": 0.5862855315208435,
+      "learning_rate": 1.899019655978685e-05,
+      "loss": 0.0956,
+      "step": 1355
+    },
+    {
+      "epoch": 0.3413012108664283,
+      "grad_norm": 0.32543689012527466,
+      "learning_rate": 1.8981277053636963e-05,
+      "loss": 0.096,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3425559947299078,
+      "grad_norm": 0.6047563552856445,
+      "learning_rate": 1.8972320441544224e-05,
+      "loss": 0.0843,
+      "step": 1365
+    },
+    {
+      "epoch": 0.34381077859338727,
+      "grad_norm": 0.28414762020111084,
+      "learning_rate": 1.8963326760512668e-05,
+      "loss": 0.0892,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3450655624568668,
+      "grad_norm": 0.337194561958313,
+      "learning_rate": 1.895429604769949e-05,
+      "loss": 0.099,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3463203463203463,
+      "grad_norm": 0.4529132843017578,
+      "learning_rate": 1.894522834041487e-05,
+      "loss": 0.0941,
+      "step": 1380
+    },
+    {
+      "epoch": 0.34757513018382585,
+      "grad_norm": 0.12156742066144943,
+      "learning_rate": 1.8936123676121844e-05,
+      "loss": 0.0896,
+      "step": 1385
+    },
+    {
+      "epoch": 0.34882991404730535,
+      "grad_norm": 0.25660645961761475,
+      "learning_rate": 1.8926982092436117e-05,
+      "loss": 0.1036,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3500846979107849,
+      "grad_norm": 0.31460127234458923,
+      "learning_rate": 1.891780362712594e-05,
+      "loss": 0.1055,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3513394817742644,
+      "grad_norm": 0.31471750140190125,
+      "learning_rate": 1.8908588318111932e-05,
+      "loss": 0.0958,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3525942656377439,
+      "grad_norm": 0.13055211305618286,
+      "learning_rate": 1.889933620346694e-05,
+      "loss": 0.1019,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3538490495012234,
+      "grad_norm": 0.13505928218364716,
+      "learning_rate": 1.8890047321415856e-05,
+      "loss": 0.0816,
+      "step": 1410
+    },
+    {
+      "epoch": 0.3551038333647029,
+      "grad_norm": 0.2555672526359558,
+      "learning_rate": 1.8880721710335495e-05,
+      "loss": 0.0998,
+      "step": 1415
+    },
+    {
+      "epoch": 0.35635861722818246,
+      "grad_norm": 0.3331303298473358,
+      "learning_rate": 1.8871359408754405e-05,
+      "loss": 0.094,
+      "step": 1420
+    },
+    {
+      "epoch": 0.35761340109166195,
+      "grad_norm": 0.17009656131267548,
+      "learning_rate": 1.8861960455352723e-05,
+      "loss": 0.0946,
+      "step": 1425
+    },
+    {
+      "epoch": 0.3588681849551415,
+      "grad_norm": 0.2334425002336502,
+      "learning_rate": 1.885252488896201e-05,
+      "loss": 0.0906,
+      "step": 1430
+    },
+    {
+      "epoch": 0.360122968818621,
+      "grad_norm": 0.21691180765628815,
+      "learning_rate": 1.8843052748565097e-05,
+      "loss": 0.091,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3613777526821005,
+      "grad_norm": 0.40637996792793274,
+      "learning_rate": 1.8833544073295918e-05,
+      "loss": 0.0856,
+      "step": 1440
+    },
+    {
+      "epoch": 0.36263253654558003,
+      "grad_norm": 0.4045301079750061,
+      "learning_rate": 1.882399890243935e-05,
+      "loss": 0.0821,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3638873204090595,
+      "grad_norm": 0.25760528445243835,
+      "learning_rate": 1.8814417275431046e-05,
+      "loss": 0.0895,
+      "step": 1450
+    },
+    {
+      "epoch": 0.36514210427253907,
+      "grad_norm": 0.3337756097316742,
+      "learning_rate": 1.8804799231857292e-05,
+      "loss": 0.0939,
+      "step": 1455
+    },
+    {
+      "epoch": 0.36639688813601856,
+      "grad_norm": 0.325998991727829,
+      "learning_rate": 1.8795144811454805e-05,
+      "loss": 0.0978,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3676516719994981,
+      "grad_norm": 0.12589257955551147,
+      "learning_rate": 1.878545405411061e-05,
+      "loss": 0.0826,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3689064558629776,
+      "grad_norm": 0.10235779732465744,
+      "learning_rate": 1.877572699986185e-05,
+      "loss": 0.0841,
+      "step": 1470
+    },
+    {
+      "epoch": 0.3701612397264571,
+      "grad_norm": 0.20532777905464172,
+      "learning_rate": 1.876596368889563e-05,
+      "loss": 0.0897,
+      "step": 1475
+    },
+    {
+      "epoch": 0.37141602358993664,
+      "grad_norm": 0.30277219414711,
+      "learning_rate": 1.8756164161548848e-05,
+      "loss": 0.0975,
+      "step": 1480
+    },
+    {
+      "epoch": 0.37267080745341613,
+      "grad_norm": 0.15634752810001373,
+      "learning_rate": 1.8746328458308034e-05,
+      "loss": 0.0879,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3739255913168957,
+      "grad_norm": 0.2814009487628937,
+      "learning_rate": 1.873645661980917e-05,
+      "loss": 0.0885,
+      "step": 1490
+    },
+    {
+      "epoch": 0.37518037518037517,
+      "grad_norm": 0.7077462673187256,
+      "learning_rate": 1.872654868683753e-05,
+      "loss": 0.1143,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3764351590438547,
+      "grad_norm": 0.14586646854877472,
+      "learning_rate": 1.8716604700327516e-05,
+      "loss": 0.1006,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3776899429073342,
+      "grad_norm": 0.23802891373634338,
+      "learning_rate": 1.8706624701362485e-05,
+      "loss": 0.0926,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3789447267708137,
+      "grad_norm": 0.41399145126342773,
+      "learning_rate": 1.8696608731174576e-05,
+      "loss": 0.0897,
+      "step": 1510
+    },
+    {
+      "epoch": 0.38019951063429325,
+      "grad_norm": 0.18691197037696838,
+      "learning_rate": 1.8686556831144545e-05,
+      "loss": 0.0843,
+      "step": 1515
+    },
+    {
+      "epoch": 0.38145429449777274,
+      "grad_norm": 0.3470015823841095,
+      "learning_rate": 1.867646904280159e-05,
+      "loss": 0.1038,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3827090783612523,
+      "grad_norm": 0.3475337624549866,
+      "learning_rate": 1.8666345407823177e-05,
+      "loss": 0.0992,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3839638622247318,
+      "grad_norm": 0.21168570220470428,
+      "learning_rate": 1.865618596803487e-05,
+      "loss": 0.0877,
+      "step": 1530
+    },
+    {
+      "epoch": 0.3852186460882113,
+      "grad_norm": 0.2576999366283417,
+      "learning_rate": 1.864599076541018e-05,
+      "loss": 0.099,
+      "step": 1535
+    },
+    {
+      "epoch": 0.3864734299516908,
+      "grad_norm": 0.42329198122024536,
+      "learning_rate": 1.8635759842070344e-05,
+      "loss": 0.0886,
+      "step": 1540
+    },
+    {
+      "epoch": 0.38772821381517036,
+      "grad_norm": 0.2144610732793808,
+      "learning_rate": 1.862549324028419e-05,
+      "loss": 0.0976,
+      "step": 1545
+    },
+    {
+      "epoch": 0.38898299767864986,
+      "grad_norm": 0.36767107248306274,
+      "learning_rate": 1.8615191002467955e-05,
+      "loss": 0.1067,
+      "step": 1550
+    },
+    {
+      "epoch": 0.39023778154212935,
+      "grad_norm": 0.32758021354675293,
+      "learning_rate": 1.8604853171185098e-05,
+      "loss": 0.0925,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3914925654056089,
+      "grad_norm": 0.11192046850919724,
+      "learning_rate": 1.859447978914614e-05,
+      "loss": 0.0815,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3927473492690884,
+      "grad_norm": 0.17995843291282654,
+      "learning_rate": 1.8584070899208468e-05,
+      "loss": 0.0823,
+      "step": 1565
+    },
+    {
+      "epoch": 0.39400213313256793,
+      "grad_norm": 0.3672018349170685,
+      "learning_rate": 1.857362654437618e-05,
+      "loss": 0.0955,
+      "step": 1570
+    },
+    {
+      "epoch": 0.3952569169960474,
+      "grad_norm": 0.11518882215023041,
+      "learning_rate": 1.8563146767799884e-05,
+      "loss": 0.0955,
+      "step": 1575
+    },
+    {
+      "epoch": 0.39651170085952697,
+      "grad_norm": 0.30487769842147827,
+      "learning_rate": 1.8552631612776554e-05,
+      "loss": 0.0842,
+      "step": 1580
+    },
+    {
+      "epoch": 0.39776648472300646,
+      "grad_norm": 0.6982203722000122,
+      "learning_rate": 1.85420811227493e-05,
+      "loss": 0.0801,
+      "step": 1585
+    },
+    {
+      "epoch": 0.39902126858648596,
+      "grad_norm": 0.22043141722679138,
+      "learning_rate": 1.853149534130724e-05,
+      "loss": 0.1039,
+      "step": 1590
+    },
+    {
+      "epoch": 0.4002760524499655,
+      "grad_norm": 0.3814282715320587,
+      "learning_rate": 1.8520874312185292e-05,
+      "loss": 0.0861,
+      "step": 1595
+    },
+    {
+      "epoch": 0.401530836313445,
+      "grad_norm": 0.14649051427841187,
+      "learning_rate": 1.8510218079263995e-05,
+      "loss": 0.09,
+      "step": 1600
+    },
+    {
+      "epoch": 0.40278562017692454,
+      "grad_norm": 0.16551180183887482,
+      "learning_rate": 1.849952668656933e-05,
+      "loss": 0.079,
+      "step": 1605
+    },
+    {
+      "epoch": 0.40404040404040403,
+      "grad_norm": 0.17042899131774902,
+      "learning_rate": 1.8488800178272553e-05,
+      "loss": 0.1073,
+      "step": 1610
+    },
+    {
+      "epoch": 0.4052951879038836,
+      "grad_norm": 0.4069715440273285,
+      "learning_rate": 1.847803859868998e-05,
+      "loss": 0.0948,
+      "step": 1615
+    },
+    {
+      "epoch": 0.40654997176736307,
+      "grad_norm": 0.18239188194274902,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 0.0934,
+      "step": 1620
+    },
+    {
+      "epoch": 0.40780475563084256,
+      "grad_norm": 0.4659058749675751,
+      "learning_rate": 1.845641040365707e-05,
+      "loss": 0.0953,
+      "step": 1625
+    },
+    {
+      "epoch": 0.4090595394943221,
+      "grad_norm": 0.1839870810508728,
+      "learning_rate": 1.844554387756313e-05,
+      "loss": 0.0874,
+      "step": 1630
+    },
+    {
+      "epoch": 0.4103143233578016,
+      "grad_norm": 0.8208329081535339,
+      "learning_rate": 1.8434642458895823e-05,
+      "loss": 0.0993,
+      "step": 1635
+    },
+    {
+      "epoch": 0.41156910722128115,
+      "grad_norm": 0.3736049234867096,
+      "learning_rate": 1.8423706192694118e-05,
+      "loss": 0.0892,
+      "step": 1640
+    },
+    {
+      "epoch": 0.41282389108476064,
+      "grad_norm": 0.1566988229751587,
+      "learning_rate": 1.841273512414095e-05,
+      "loss": 0.0893,
+      "step": 1645
+    },
+    {
+      "epoch": 0.4140786749482402,
+      "grad_norm": 0.5364568829536438,
+      "learning_rate": 1.840172929856304e-05,
+      "loss": 0.1002,
+      "step": 1650
+    },
+    {
+      "epoch": 0.4153334588117197,
+      "grad_norm": 0.41213005781173706,
+      "learning_rate": 1.8390688761430707e-05,
+      "loss": 0.0942,
+      "step": 1655
+    },
+    {
+      "epoch": 0.41658824267519917,
+      "grad_norm": 0.417540580034256,
+      "learning_rate": 1.8379613558357686e-05,
+      "loss": 0.09,
+      "step": 1660
+    },
+    {
+      "epoch": 0.4178430265386787,
+      "grad_norm": 0.30479973554611206,
+      "learning_rate": 1.836850373510092e-05,
+      "loss": 0.0819,
+      "step": 1665
+    },
+    {
+      "epoch": 0.4190978104021582,
+      "grad_norm": 0.5476796627044678,
+      "learning_rate": 1.8357359337560393e-05,
+      "loss": 0.0835,
+      "step": 1670
+    },
+    {
+      "epoch": 0.42035259426563776,
+      "grad_norm": 0.14834348857402802,
+      "learning_rate": 1.8346180411778934e-05,
+      "loss": 0.081,
+      "step": 1675
+    },
+    {
+      "epoch": 0.42160737812911725,
+      "grad_norm": 0.49090176820755005,
+      "learning_rate": 1.833496700394202e-05,
+      "loss": 0.0864,
+      "step": 1680
+    },
+    {
+      "epoch": 0.4228621619925968,
+      "grad_norm": 0.17847497761249542,
+      "learning_rate": 1.83237191603776e-05,
+      "loss": 0.0705,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4241169458560763,
+      "grad_norm": 0.22627241909503937,
+      "learning_rate": 1.831243692755587e-05,
+      "loss": 0.0757,
+      "step": 1690
+    },
+    {
+      "epoch": 0.4253717297195558,
+      "grad_norm": 0.6790867447853088,
+      "learning_rate": 1.830112035208913e-05,
+      "loss": 0.086,
+      "step": 1695
+    },
+    {
+      "epoch": 0.4266265135830353,
+      "grad_norm": 0.33606070280075073,
+      "learning_rate": 1.828976948073155e-05,
+      "loss": 0.0954,
+      "step": 1700
+    },
+    {
+      "epoch": 0.4278812974465148,
+      "grad_norm": 0.15440745651721954,
+      "learning_rate": 1.8278384360379008e-05,
+      "loss": 0.0877,
+      "step": 1705
+    },
+    {
+      "epoch": 0.42913608130999437,
+      "grad_norm": 0.3315647542476654,
+      "learning_rate": 1.8266965038068856e-05,
+      "loss": 0.0856,
+      "step": 1710
+    },
+    {
+      "epoch": 0.43039086517347386,
+      "grad_norm": 0.2767457365989685,
+      "learning_rate": 1.8255511560979782e-05,
+      "loss": 0.0946,
+      "step": 1715
+    },
+    {
+      "epoch": 0.4316456490369534,
+      "grad_norm": 0.24647273123264313,
+      "learning_rate": 1.824402397643155e-05,
+      "loss": 0.1019,
+      "step": 1720
+    },
+    {
+      "epoch": 0.4329004329004329,
+      "grad_norm": 0.5085861086845398,
+      "learning_rate": 1.823250233188487e-05,
+      "loss": 0.0915,
+      "step": 1725
+    },
+    {
+      "epoch": 0.43415521676391244,
+      "grad_norm": 0.2605689465999603,
+      "learning_rate": 1.822094667494115e-05,
+      "loss": 0.0787,
+      "step": 1730
+    },
+    {
+      "epoch": 0.43541000062739194,
+      "grad_norm": 0.2040780782699585,
+      "learning_rate": 1.8209357053342325e-05,
+      "loss": 0.0897,
+      "step": 1735
+    },
+    {
+      "epoch": 0.4366647844908714,
+      "grad_norm": 0.12423911690711975,
+      "learning_rate": 1.8197733514970655e-05,
+      "loss": 0.1005,
+      "step": 1740
+    },
+    {
+      "epoch": 0.437919568354351,
+      "grad_norm": 0.08479491621255875,
+      "learning_rate": 1.8186076107848524e-05,
+      "loss": 0.0871,
+      "step": 1745
+    },
+    {
+      "epoch": 0.43917435221783047,
+      "grad_norm": 0.2682114541530609,
+      "learning_rate": 1.8174384880138247e-05,
+      "loss": 0.0907,
+      "step": 1750
+    },
+    {
+      "epoch": 0.44042913608131,
+      "grad_norm": 0.32582300901412964,
+      "learning_rate": 1.8162659880141865e-05,
+      "loss": 0.0893,
+      "step": 1755
+    },
+    {
+      "epoch": 0.4416839199447895,
+      "grad_norm": 0.4764627516269684,
+      "learning_rate": 1.8150901156300956e-05,
+      "loss": 0.0892,
+      "step": 1760
+    },
+    {
+      "epoch": 0.44293870380826905,
+      "grad_norm": 0.2317512184381485,
+      "learning_rate": 1.8139108757196412e-05,
+      "loss": 0.0972,
+      "step": 1765
+    },
+    {
+      "epoch": 0.44419348767174854,
+      "grad_norm": 0.1146375834941864,
+      "learning_rate": 1.812728273154827e-05,
+      "loss": 0.1024,
+      "step": 1770
+    },
+    {
+      "epoch": 0.44544827153522804,
+      "grad_norm": 0.07953114807605743,
+      "learning_rate": 1.8115423128215485e-05,
+      "loss": 0.0892,
+      "step": 1775
+    },
+    {
+      "epoch": 0.4467030553987076,
+      "grad_norm": 0.5142281651496887,
+      "learning_rate": 1.810352999619574e-05,
+      "loss": 0.1007,
+      "step": 1780
+    },
+    {
+      "epoch": 0.4479578392621871,
+      "grad_norm": 0.35662662982940674,
+      "learning_rate": 1.8091603384625243e-05,
+      "loss": 0.0937,
+      "step": 1785
+    },
+    {
+      "epoch": 0.4492126231256666,
+      "grad_norm": 0.8468720316886902,
+      "learning_rate": 1.8079643342778516e-05,
+      "loss": 0.0987,
+      "step": 1790
+    },
+    {
+      "epoch": 0.4504674069891461,
+      "grad_norm": 0.34367749094963074,
+      "learning_rate": 1.80676499200682e-05,
+      "loss": 0.087,
+      "step": 1795
+    },
+    {
+      "epoch": 0.45172219085262566,
+      "grad_norm": 0.21747180819511414,
+      "learning_rate": 1.8055623166044855e-05,
+      "loss": 0.0883,
+      "step": 1800
+    },
+    {
+      "epoch": 0.45297697471610515,
+      "grad_norm": 0.4869808554649353,
+      "learning_rate": 1.8043563130396738e-05,
+      "loss": 0.0989,
+      "step": 1805
+    },
+    {
+      "epoch": 0.45423175857958464,
+      "grad_norm": 0.2503896653652191,
+      "learning_rate": 1.8031469862949618e-05,
+      "loss": 0.083,
+      "step": 1810
+    },
+    {
+      "epoch": 0.4554865424430642,
+      "grad_norm": 0.34971126914024353,
+      "learning_rate": 1.801934341366655e-05,
+      "loss": 0.0845,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4567413263065437,
+      "grad_norm": 0.6193472743034363,
+      "learning_rate": 1.800718383264769e-05,
+      "loss": 0.0966,
+      "step": 1820
+    },
+    {
+      "epoch": 0.45799611017002323,
+      "grad_norm": 0.5469531416893005,
+      "learning_rate": 1.799499117013007e-05,
+      "loss": 0.0844,
+      "step": 1825
+    },
+    {
+      "epoch": 0.4592508940335027,
+      "grad_norm": 0.2877050042152405,
+      "learning_rate": 1.7982765476487398e-05,
+      "loss": 0.0929,
+      "step": 1830
+    },
+    {
+      "epoch": 0.46050567789698227,
+      "grad_norm": 0.2750946283340454,
+      "learning_rate": 1.797050680222985e-05,
+      "loss": 0.0883,
+      "step": 1835
+    },
+    {
+      "epoch": 0.46176046176046176,
+      "grad_norm": 0.4805077016353607,
+      "learning_rate": 1.7958215198003866e-05,
+      "loss": 0.0916,
+      "step": 1840
+    },
+    {
+      "epoch": 0.46301524562394125,
+      "grad_norm": 0.4399943947792053,
+      "learning_rate": 1.7945890714591926e-05,
+      "loss": 0.099,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4642700294874208,
+      "grad_norm": 0.8317438364028931,
+      "learning_rate": 1.7933533402912354e-05,
+      "loss": 0.0836,
+      "step": 1850
+    },
+    {
+      "epoch": 0.4655248133509003,
+      "grad_norm": 0.26974841952323914,
+      "learning_rate": 1.7921143314019106e-05,
+      "loss": 0.1013,
+      "step": 1855
+    },
+    {
+      "epoch": 0.46677959721437984,
+      "grad_norm": 0.2595142722129822,
+      "learning_rate": 1.7908720499101552e-05,
+      "loss": 0.0881,
+      "step": 1860
+    },
+    {
+      "epoch": 0.46803438107785933,
+      "grad_norm": 0.18567359447479248,
+      "learning_rate": 1.789626500948427e-05,
+      "loss": 0.0929,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4692891649413389,
+      "grad_norm": 0.11302470415830612,
+      "learning_rate": 1.7883776896626836e-05,
+      "loss": 0.0882,
+      "step": 1870
+    },
+    {
+      "epoch": 0.47054394880481837,
+      "grad_norm": 0.06441876292228699,
+      "learning_rate": 1.7871256212123605e-05,
+      "loss": 0.091,
+      "step": 1875
+    },
+    {
+      "epoch": 0.47179873266829786,
+      "grad_norm": 0.5252417922019958,
+      "learning_rate": 1.78587030077035e-05,
+      "loss": 0.0995,
+      "step": 1880
+    },
+    {
+      "epoch": 0.4730535165317774,
+      "grad_norm": 0.2045115977525711,
+      "learning_rate": 1.7846117335229808e-05,
+      "loss": 0.0889,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4743083003952569,
+      "grad_norm": 0.22728148102760315,
+      "learning_rate": 1.783349924669994e-05,
+      "loss": 0.1018,
+      "step": 1890
+    },
+    {
+      "epoch": 0.47556308425873645,
+      "grad_norm": 0.17804867029190063,
+      "learning_rate": 1.7820848794245243e-05,
+      "loss": 0.1001,
+      "step": 1895
+    },
+    {
+      "epoch": 0.47681786812221594,
+      "grad_norm": 0.5709582567214966,
+      "learning_rate": 1.7808166030130782e-05,
+      "loss": 0.087,
+      "step": 1900
+    },
+    {
+      "epoch": 0.4780726519856955,
+      "grad_norm": 0.19447216391563416,
+      "learning_rate": 1.779545100675511e-05,
+      "loss": 0.0881,
+      "step": 1905
+    },
+    {
+      "epoch": 0.479327435849175,
+      "grad_norm": 0.5337616801261902,
+      "learning_rate": 1.778270377665005e-05,
+      "loss": 0.0954,
+      "step": 1910
+    },
+    {
+      "epoch": 0.48058221971265447,
+      "grad_norm": 0.8371096253395081,
+      "learning_rate": 1.77699243924805e-05,
+      "loss": 0.0886,
+      "step": 1915
+    },
+    {
+      "epoch": 0.481837003576134,
+      "grad_norm": 0.21723900735378265,
+      "learning_rate": 1.77571129070442e-05,
+      "loss": 0.0899,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4830917874396135,
+      "grad_norm": 0.18002496659755707,
+      "learning_rate": 1.7744269373271507e-05,
+      "loss": 0.0876,
+      "step": 1925
+    },
+    {
+      "epoch": 0.48434657130309305,
+      "grad_norm": 0.15647904574871063,
+      "learning_rate": 1.7731393844225187e-05,
+      "loss": 0.0842,
+      "step": 1930
+    },
+    {
+      "epoch": 0.48560135516657255,
+      "grad_norm": 0.36643457412719727,
+      "learning_rate": 1.7718486373100207e-05,
+      "loss": 0.0848,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4868561390300521,
+      "grad_norm": 0.37278589606285095,
+      "learning_rate": 1.7705547013223486e-05,
+      "loss": 0.1,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4881109228935316,
+      "grad_norm": 0.18367764353752136,
+      "learning_rate": 1.7692575818053696e-05,
+      "loss": 0.0969,
+      "step": 1945
+    },
+    {
+      "epoch": 0.48936570675701113,
+      "grad_norm": 0.4618573486804962,
+      "learning_rate": 1.7679572841181033e-05,
+      "loss": 0.0854,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4906204906204906,
+      "grad_norm": 0.2800655961036682,
+      "learning_rate": 1.7666538136327007e-05,
+      "loss": 0.0942,
+      "step": 1955
+    },
+    {
+      "epoch": 0.4918752744839701,
+      "grad_norm": 0.2937714457511902,
+      "learning_rate": 1.7653471757344203e-05,
+      "loss": 0.0977,
+      "step": 1960
+    },
+    {
+      "epoch": 0.49313005834744966,
+      "grad_norm": 0.7581244707107544,
+      "learning_rate": 1.7640373758216075e-05,
+      "loss": 0.0939,
+      "step": 1965
+    },
+    {
+      "epoch": 0.49438484221092915,
+      "grad_norm": 0.21476201713085175,
+      "learning_rate": 1.7627244193056705e-05,
+      "loss": 0.0936,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4956396260744087,
+      "grad_norm": 0.1473897099494934,
+      "learning_rate": 1.7614083116110597e-05,
+      "loss": 0.0859,
+      "step": 1975
+    },
+    {
+      "epoch": 0.4968944099378882,
+      "grad_norm": 0.4635510742664337,
+      "learning_rate": 1.7600890581752435e-05,
+      "loss": 0.0984,
+      "step": 1980
+    },
+    {
+      "epoch": 0.49814919380136774,
+      "grad_norm": 0.15860065817832947,
+      "learning_rate": 1.758766664448689e-05,
+      "loss": 0.0934,
+      "step": 1985
+    },
+    {
+      "epoch": 0.49940397766484723,
+      "grad_norm": 0.10272369533777237,
+      "learning_rate": 1.7574411358948347e-05,
+      "loss": 0.0749,
+      "step": 1990
+    },
+    {
+      "epoch": 0.5006587615283268,
+      "grad_norm": 0.3367800712585449,
+      "learning_rate": 1.7561124779900723e-05,
+      "loss": 0.0809,
+      "step": 1995
+    },
+    {
+      "epoch": 0.5019135453918062,
+      "grad_norm": 0.23143918812274933,
+      "learning_rate": 1.7547806962237222e-05,
+      "loss": 0.1092,
+      "step": 2000
+    },
+    {
+      "epoch": 0.5031683292552858,
+      "grad_norm": 0.5789488554000854,
+      "learning_rate": 1.7534457960980097e-05,
+      "loss": 0.1088,
+      "step": 2005
+    },
+    {
+      "epoch": 0.5044231131187653,
+      "grad_norm": 0.2879166007041931,
+      "learning_rate": 1.7521077831280453e-05,
+      "loss": 0.0873,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5056778969822449,
+      "grad_norm": 0.6664144396781921,
+      "learning_rate": 1.750766662841799e-05,
+      "loss": 0.0984,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5069326808457243,
+      "grad_norm": 0.13004139065742493,
+      "learning_rate": 1.7494224407800792e-05,
+      "loss": 0.0879,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5081874647092038,
+      "grad_norm": 0.30764704942703247,
+      "learning_rate": 1.7480751224965083e-05,
+      "loss": 0.093,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5094422485726834,
+      "grad_norm": 0.15937355160713196,
+      "learning_rate": 1.7467247135575016e-05,
+      "loss": 0.088,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5106970324361628,
+      "grad_norm": 0.12112176418304443,
+      "learning_rate": 1.7453712195422432e-05,
+      "loss": 0.0967,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5119518162996424,
+      "grad_norm": 0.35342133045196533,
+      "learning_rate": 1.744014646042663e-05,
+      "loss": 0.0834,
+      "step": 2040
+    },
+    {
+      "epoch": 0.5132066001631219,
+      "grad_norm": 0.39632728695869446,
+      "learning_rate": 1.7426549986634135e-05,
+      "loss": 0.0866,
+      "step": 2045
+    },
+    {
+      "epoch": 0.5144613840266015,
+      "grad_norm": 0.4647534489631653,
+      "learning_rate": 1.741292283021847e-05,
+      "loss": 0.0834,
+      "step": 2050
+    },
+    {
+      "epoch": 0.5157161678900809,
+      "grad_norm": 0.3376861810684204,
+      "learning_rate": 1.7399265047479926e-05,
+      "loss": 0.0771,
+      "step": 2055
+    },
+    {
+      "epoch": 0.5169709517535604,
+      "grad_norm": 0.3715597689151764,
+      "learning_rate": 1.7385576694845324e-05,
+      "loss": 0.0886,
+      "step": 2060
+    },
+    {
+      "epoch": 0.51822573561704,
+      "grad_norm": 0.4062286615371704,
+      "learning_rate": 1.7371857828867778e-05,
+      "loss": 0.0886,
+      "step": 2065
+    },
+    {
+      "epoch": 0.5194805194805194,
+      "grad_norm": 0.2665550708770752,
+      "learning_rate": 1.7358108506226477e-05,
+      "loss": 0.0918,
+      "step": 2070
+    },
+    {
+      "epoch": 0.520735303343999,
+      "grad_norm": 0.4846581220626831,
+      "learning_rate": 1.7344328783726436e-05,
+      "loss": 0.0834,
+      "step": 2075
+    },
+    {
+      "epoch": 0.5219900872074785,
+      "grad_norm": 0.2385398894548416,
+      "learning_rate": 1.7330518718298263e-05,
+      "loss": 0.0953,
+      "step": 2080
+    },
+    {
+      "epoch": 0.5232448710709581,
+      "grad_norm": 0.5059955716133118,
+      "learning_rate": 1.7316678366997935e-05,
+      "loss": 0.0796,
+      "step": 2085
+    },
+    {
+      "epoch": 0.5244996549344375,
+      "grad_norm": 0.5229114294052124,
+      "learning_rate": 1.7302807787006547e-05,
+      "loss": 0.0953,
+      "step": 2090
+    },
+    {
+      "epoch": 0.5257544387979171,
+      "grad_norm": 0.3612153232097626,
+      "learning_rate": 1.728890703563009e-05,
+      "loss": 0.0874,
+      "step": 2095
+    },
+    {
+      "epoch": 0.5270092226613966,
+      "grad_norm": 0.2730209529399872,
+      "learning_rate": 1.7274976170299197e-05,
+      "loss": 0.0977,
+      "step": 2100
+    },
+    {
+      "epoch": 0.528264006524876,
+      "grad_norm": 0.28223976492881775,
+      "learning_rate": 1.726101524856893e-05,
+      "loss": 0.0901,
+      "step": 2105
+    },
+    {
+      "epoch": 0.5295187903883556,
+      "grad_norm": 0.24942927062511444,
+      "learning_rate": 1.724702432811852e-05,
+      "loss": 0.0978,
+      "step": 2110
+    },
+    {
+      "epoch": 0.5307735742518351,
+      "grad_norm": 0.21909397840499878,
+      "learning_rate": 1.7233003466751133e-05,
+      "loss": 0.0978,
+      "step": 2115
+    },
+    {
+      "epoch": 0.5320283581153147,
+      "grad_norm": 0.28112301230430603,
+      "learning_rate": 1.7218952722393646e-05,
+      "loss": 0.0899,
+      "step": 2120
+    },
+    {
+      "epoch": 0.5332831419787941,
+      "grad_norm": 0.23570087552070618,
+      "learning_rate": 1.7204872153096386e-05,
+      "loss": 0.0839,
+      "step": 2125
+    },
+    {
+      "epoch": 0.5345379258422737,
+      "grad_norm": 0.5193778872489929,
+      "learning_rate": 1.719076181703291e-05,
+      "loss": 0.1028,
+      "step": 2130
+    },
+    {
+      "epoch": 0.5357927097057532,
+      "grad_norm": 0.2834993600845337,
+      "learning_rate": 1.7176621772499752e-05,
+      "loss": 0.0984,
+      "step": 2135
+    },
+    {
+      "epoch": 0.5370474935692326,
+      "grad_norm": 0.41437774896621704,
+      "learning_rate": 1.716245207791618e-05,
+      "loss": 0.092,
+      "step": 2140
+    },
+    {
+      "epoch": 0.5383022774327122,
+      "grad_norm": 0.4809644818305969,
+      "learning_rate": 1.714825279182398e-05,
+      "loss": 0.0814,
+      "step": 2145
+    },
+    {
+      "epoch": 0.5395570612961917,
+      "grad_norm": 0.10345987975597382,
+      "learning_rate": 1.7134023972887164e-05,
+      "loss": 0.0736,
+      "step": 2150
+    },
+    {
+      "epoch": 0.5408118451596713,
+      "grad_norm": 0.06666211783885956,
+      "learning_rate": 1.7119765679891794e-05,
+      "loss": 0.078,
+      "step": 2155
+    },
+    {
+      "epoch": 0.5420666290231507,
+      "grad_norm": 0.4196692407131195,
+      "learning_rate": 1.7105477971745668e-05,
+      "loss": 0.0963,
+      "step": 2160
+    },
+    {
+      "epoch": 0.5433214128866303,
+      "grad_norm": 0.1562952995300293,
+      "learning_rate": 1.7091160907478137e-05,
+      "loss": 0.0777,
+      "step": 2165
+    },
+    {
+      "epoch": 0.5445761967501098,
+      "grad_norm": 0.2740727663040161,
+      "learning_rate": 1.7076814546239825e-05,
+      "loss": 0.0953,
+      "step": 2170
+    },
+    {
+      "epoch": 0.5458309806135893,
+      "grad_norm": 0.2551318407058716,
+      "learning_rate": 1.7062438947302405e-05,
+      "loss": 0.0775,
+      "step": 2175
+    },
+    {
+      "epoch": 0.5470857644770688,
+      "grad_norm": 0.9234172105789185,
+      "learning_rate": 1.704803417005833e-05,
+      "loss": 0.089,
+      "step": 2180
+    },
+    {
+      "epoch": 0.5483405483405484,
+      "grad_norm": 0.29249778389930725,
+      "learning_rate": 1.7033600274020616e-05,
+      "loss": 0.0925,
+      "step": 2185
+    },
+    {
+      "epoch": 0.5495953322040279,
+      "grad_norm": 0.1841478943824768,
+      "learning_rate": 1.7019137318822577e-05,
+      "loss": 0.0937,
+      "step": 2190
+    },
+    {
+      "epoch": 0.5508501160675073,
+      "grad_norm": 0.15575248003005981,
+      "learning_rate": 1.7004645364217584e-05,
+      "loss": 0.0894,
+      "step": 2195
+    },
+    {
+      "epoch": 0.5521048999309869,
+      "grad_norm": 0.37570688128471375,
+      "learning_rate": 1.699012447007882e-05,
+      "loss": 0.0853,
+      "step": 2200
+    },
+    {
+      "epoch": 0.5533596837944664,
+      "grad_norm": 0.49081242084503174,
+      "learning_rate": 1.6975574696399033e-05,
+      "loss": 0.0938,
+      "step": 2205
+    },
+    {
+      "epoch": 0.5546144676579459,
+      "grad_norm": 0.39014649391174316,
+      "learning_rate": 1.6960996103290282e-05,
+      "loss": 0.0876,
+      "step": 2210
+    },
+    {
+      "epoch": 0.5558692515214254,
+      "grad_norm": 0.48171404004096985,
+      "learning_rate": 1.694638875098369e-05,
+      "loss": 0.0865,
+      "step": 2215
+    },
+    {
+      "epoch": 0.557124035384905,
+      "grad_norm": 0.18437805771827698,
+      "learning_rate": 1.693175269982921e-05,
+      "loss": 0.0895,
+      "step": 2220
+    },
+    {
+      "epoch": 0.5583788192483845,
+      "grad_norm": 0.35959693789482117,
+      "learning_rate": 1.691708801029535e-05,
+      "loss": 0.0898,
+      "step": 2225
+    },
+    {
+      "epoch": 0.5596336031118639,
+      "grad_norm": 0.5558937191963196,
+      "learning_rate": 1.6902394742968945e-05,
+      "loss": 0.0974,
+      "step": 2230
+    },
+    {
+      "epoch": 0.5608883869753435,
+      "grad_norm": 0.2495732605457306,
+      "learning_rate": 1.68876729585549e-05,
+      "loss": 0.0879,
+      "step": 2235
+    },
+    {
+      "epoch": 0.562143170838823,
+      "grad_norm": 0.1926206350326538,
+      "learning_rate": 1.6872922717875923e-05,
+      "loss": 0.0889,
+      "step": 2240
+    },
+    {
+      "epoch": 0.5633979547023026,
+      "grad_norm": 0.17689497768878937,
+      "learning_rate": 1.6858144081872315e-05,
+      "loss": 0.0924,
+      "step": 2245
+    },
+    {
+      "epoch": 0.564652738565782,
+      "grad_norm": 0.10556932538747787,
+      "learning_rate": 1.6843337111601663e-05,
+      "loss": 0.086,
+      "step": 2250
+    },
+    {
+      "epoch": 0.5659075224292616,
+      "grad_norm": 0.19952704012393951,
+      "learning_rate": 1.6828501868238637e-05,
+      "loss": 0.0801,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5671623062927411,
+      "grad_norm": 0.2911912202835083,
+      "learning_rate": 1.6813638413074707e-05,
+      "loss": 0.0925,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5684170901562206,
+      "grad_norm": 0.23794835805892944,
+      "learning_rate": 1.67987468075179e-05,
+      "loss": 0.0913,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5696718740197001,
+      "grad_norm": 0.1402294784784317,
+      "learning_rate": 1.6783827113092547e-05,
+      "loss": 0.0882,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5709266578831796,
+      "grad_norm": 0.15708528459072113,
+      "learning_rate": 1.6768879391439035e-05,
+      "loss": 0.0941,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5721814417466592,
+      "grad_norm": 0.37110403180122375,
+      "learning_rate": 1.6753903704313527e-05,
+      "loss": 0.0969,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5734362256101386,
+      "grad_norm": 0.2311350554227829,
+      "learning_rate": 1.6738900113587745e-05,
+      "loss": 0.0849,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5746910094736182,
+      "grad_norm": 0.22549764811992645,
+      "learning_rate": 1.6723868681248677e-05,
+      "loss": 0.0923,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5759457933370977,
+      "grad_norm": 0.12788964807987213,
+      "learning_rate": 1.6708809469398347e-05,
+      "loss": 0.0798,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5772005772005772,
+      "grad_norm": 0.4113115668296814,
+      "learning_rate": 1.6693722540253554e-05,
+      "loss": 0.0794,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5784553610640567,
+      "grad_norm": 0.28315308690071106,
+      "learning_rate": 1.6678607956145596e-05,
+      "loss": 0.0855,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5797101449275363,
+      "grad_norm": 0.18012386560440063,
+      "learning_rate": 1.6663465779520042e-05,
+      "loss": 0.0906,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5809649287910158,
+      "grad_norm": 0.0916949212551117,
+      "learning_rate": 1.6648296072936445e-05,
+      "loss": 0.0807,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5822197126544952,
+      "grad_norm": 0.3671141266822815,
+      "learning_rate": 1.6633098899068112e-05,
+      "loss": 0.093,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5834744965179748,
+      "grad_norm": 0.16488447785377502,
+      "learning_rate": 1.6617874320701813e-05,
+      "loss": 0.0883,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5847292803814543,
+      "grad_norm": 0.291689395904541,
+      "learning_rate": 1.660262240073756e-05,
+      "loss": 0.0908,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5859840642449338,
+      "grad_norm": 0.25021910667419434,
+      "learning_rate": 1.658734320218831e-05,
+      "loss": 0.0799,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5872388481084133,
+      "grad_norm": 0.11796696484088898,
+      "learning_rate": 1.6572036788179728e-05,
+      "loss": 0.0904,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5884936319718929,
+      "grad_norm": 0.14167477190494537,
+      "learning_rate": 1.6556703221949912e-05,
+      "loss": 0.0912,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5897484158353724,
+      "grad_norm": 0.12993638217449188,
+      "learning_rate": 1.6541342566849145e-05,
+      "loss": 0.0851,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5910031996988518,
+      "grad_norm": 0.23770704865455627,
+      "learning_rate": 1.652595488633963e-05,
+      "loss": 0.0898,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5922579835623314,
+      "grad_norm": 0.5402049422264099,
+      "learning_rate": 1.6510540243995216e-05,
+      "loss": 0.0882,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5935127674258109,
+      "grad_norm": 0.16188447177410126,
+      "learning_rate": 1.6495098703501153e-05,
+      "loss": 0.0781,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5947675512892904,
+      "grad_norm": 0.33231815695762634,
+      "learning_rate": 1.6479630328653814e-05,
+      "loss": 0.0879,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5960223351527699,
+      "grad_norm": 0.3051691949367523,
+      "learning_rate": 1.6464135183360444e-05,
+      "loss": 0.1065,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5972771190162495,
+      "grad_norm": 0.09856946021318436,
+      "learning_rate": 1.6448613331638877e-05,
+      "loss": 0.0931,
+      "step": 2380
+    },
+    {
+      "epoch": 0.598531902879729,
+      "grad_norm": 0.3320614993572235,
+      "learning_rate": 1.6433064837617294e-05,
+      "loss": 0.0806,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5997866867432085,
+      "grad_norm": 0.6440458297729492,
+      "learning_rate": 1.641748976553395e-05,
+      "loss": 0.0981,
+      "step": 2390
+    },
+    {
+      "epoch": 0.601041470606688,
+      "grad_norm": 0.24246153235435486,
+      "learning_rate": 1.64018881797369e-05,
+      "loss": 0.088,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6022962544701675,
+      "grad_norm": 0.21367953717708588,
+      "learning_rate": 1.6386260144683744e-05,
+      "loss": 0.087,
+      "step": 2400
+    },
+    {
+      "epoch": 0.603551038333647,
+      "grad_norm": 0.2851586639881134,
+      "learning_rate": 1.6370605724941356e-05,
+      "loss": 0.0877,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6048058221971265,
+      "grad_norm": 0.19967404007911682,
+      "learning_rate": 1.6354924985185614e-05,
+      "loss": 0.0888,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 0.22060827910900116,
+      "learning_rate": 1.633921799020114e-05,
+      "loss": 0.0845,
+      "step": 2415
+    },
+    {
+      "epoch": 0.6073153899240856,
+      "grad_norm": 0.4157061278820038,
+      "learning_rate": 1.632348480488103e-05,
+      "loss": 0.0899,
+      "step": 2420
+    },
+    {
+      "epoch": 0.6085701737875651,
+      "grad_norm": 0.27656257152557373,
+      "learning_rate": 1.6307725494226586e-05,
+      "loss": 0.09,
+      "step": 2425
+    },
+    {
+      "epoch": 0.6098249576510446,
+      "grad_norm": 0.3946853578090668,
+      "learning_rate": 1.6291940123347033e-05,
+      "loss": 0.0863,
+      "step": 2430
+    },
+    {
+      "epoch": 0.6110797415145242,
+      "grad_norm": 0.2414676994085312,
+      "learning_rate": 1.6276128757459282e-05,
+      "loss": 0.0905,
+      "step": 2435
+    },
+    {
+      "epoch": 0.6123345253780036,
+      "grad_norm": 0.23438376188278198,
+      "learning_rate": 1.6260291461887628e-05,
+      "loss": 0.0726,
+      "step": 2440
+    },
+    {
+      "epoch": 0.6135893092414831,
+      "grad_norm": 0.20305827260017395,
+      "learning_rate": 1.6244428302063506e-05,
+      "loss": 0.0891,
+      "step": 2445
+    },
+    {
+      "epoch": 0.6148440931049627,
+      "grad_norm": 0.3930656909942627,
+      "learning_rate": 1.62285393435252e-05,
+      "loss": 0.0884,
+      "step": 2450
+    },
+    {
+      "epoch": 0.6160988769684422,
+      "grad_norm": 0.23331406712532043,
+      "learning_rate": 1.6212624651917573e-05,
+      "loss": 0.0904,
+      "step": 2455
+    },
+    {
+      "epoch": 0.6173536608319217,
+      "grad_norm": 0.34075310826301575,
+      "learning_rate": 1.6196684292991827e-05,
+      "loss": 0.0941,
+      "step": 2460
+    },
+    {
+      "epoch": 0.6186084446954012,
+      "grad_norm": 0.33201703429222107,
+      "learning_rate": 1.6180718332605185e-05,
+      "loss": 0.0838,
+      "step": 2465
+    },
+    {
+      "epoch": 0.6198632285588808,
+      "grad_norm": 0.10392648726701736,
+      "learning_rate": 1.6164726836720656e-05,
+      "loss": 0.0974,
+      "step": 2470
+    },
+    {
+      "epoch": 0.6211180124223602,
+      "grad_norm": 0.21417252719402313,
+      "learning_rate": 1.614870987140674e-05,
+      "loss": 0.0891,
+      "step": 2475
+    },
+    {
+      "epoch": 0.6223727962858397,
+      "grad_norm": 0.24638135731220245,
+      "learning_rate": 1.6132667502837164e-05,
+      "loss": 0.0827,
+      "step": 2480
+    },
+    {
+      "epoch": 0.6236275801493193,
+      "grad_norm": 0.183788001537323,
+      "learning_rate": 1.611659979729062e-05,
+      "loss": 0.0917,
+      "step": 2485
+    },
+    {
+      "epoch": 0.6248823640127988,
+      "grad_norm": 0.25125667452812195,
+      "learning_rate": 1.6100506821150455e-05,
+      "loss": 0.0888,
+      "step": 2490
+    },
+    {
+      "epoch": 0.6261371478762783,
+      "grad_norm": 0.24602219462394714,
+      "learning_rate": 1.6084388640904452e-05,
+      "loss": 0.0773,
+      "step": 2495
+    },
+    {
+      "epoch": 0.6273919317397578,
+      "grad_norm": 0.6547009944915771,
+      "learning_rate": 1.60682453231445e-05,
+      "loss": 0.1033,
+      "step": 2500
+    },
+    {
+      "epoch": 0.6286467156032374,
+      "grad_norm": 0.11527785658836365,
+      "learning_rate": 1.605207693456635e-05,
+      "loss": 0.0947,
+      "step": 2505
+    },
+    {
+      "epoch": 0.6299014994667168,
+      "grad_norm": 0.3719431459903717,
+      "learning_rate": 1.6035883541969336e-05,
+      "loss": 0.0873,
+      "step": 2510
+    },
+    {
+      "epoch": 0.6311562833301964,
+      "grad_norm": 0.2360219806432724,
+      "learning_rate": 1.601966521225609e-05,
+      "loss": 0.0826,
+      "step": 2515
+    },
+    {
+      "epoch": 0.6324110671936759,
+      "grad_norm": 0.27890655398368835,
+      "learning_rate": 1.6003422012432275e-05,
+      "loss": 0.0955,
+      "step": 2520
+    },
+    {
+      "epoch": 0.6336658510571554,
+      "grad_norm": 0.14903779327869415,
+      "learning_rate": 1.5987154009606308e-05,
+      "loss": 0.0951,
+      "step": 2525
+    },
+    {
+      "epoch": 0.6349206349206349,
+      "grad_norm": 0.20811066031455994,
+      "learning_rate": 1.5970861270989065e-05,
+      "loss": 0.0874,
+      "step": 2530
+    },
+    {
+      "epoch": 0.6361754187841144,
+      "grad_norm": 0.39913272857666016,
+      "learning_rate": 1.5954543863893638e-05,
+      "loss": 0.0846,
+      "step": 2535
+    },
+    {
+      "epoch": 0.637430202647594,
+      "grad_norm": 0.3802116811275482,
+      "learning_rate": 1.5938201855735017e-05,
+      "loss": 0.0754,
+      "step": 2540
+    },
+    {
+      "epoch": 0.6386849865110734,
+      "grad_norm": 0.4720747172832489,
+      "learning_rate": 1.592183531402984e-05,
+      "loss": 0.0872,
+      "step": 2545
+    },
+    {
+      "epoch": 0.639939770374553,
+      "grad_norm": 0.46679922938346863,
+      "learning_rate": 1.590544430639611e-05,
+      "loss": 0.0867,
+      "step": 2550
+    },
+    {
+      "epoch": 0.6411945542380325,
+      "grad_norm": 0.29669615626335144,
+      "learning_rate": 1.5889028900552897e-05,
+      "loss": 0.0988,
+      "step": 2555
+    },
+    {
+      "epoch": 0.6424493381015121,
+      "grad_norm": 0.20683659613132477,
+      "learning_rate": 1.587258916432008e-05,
+      "loss": 0.0935,
+      "step": 2560
+    },
+    {
+      "epoch": 0.6437041219649915,
+      "grad_norm": 0.3284091055393219,
+      "learning_rate": 1.5856125165618056e-05,
+      "loss": 0.0779,
+      "step": 2565
+    },
+    {
+      "epoch": 0.644958905828471,
+      "grad_norm": 0.24657922983169556,
+      "learning_rate": 1.5839636972467466e-05,
+      "loss": 0.0827,
+      "step": 2570
+    },
+    {
+      "epoch": 0.6462136896919506,
+      "grad_norm": 0.32977283000946045,
+      "learning_rate": 1.5823124652988907e-05,
+      "loss": 0.0955,
+      "step": 2575
+    },
+    {
+      "epoch": 0.64746847355543,
+      "grad_norm": 0.28097760677337646,
+      "learning_rate": 1.580658827540265e-05,
+      "loss": 0.0837,
+      "step": 2580
+    },
+    {
+      "epoch": 0.6487232574189096,
+      "grad_norm": 0.12431260943412781,
+      "learning_rate": 1.5790027908028366e-05,
+      "loss": 0.0901,
+      "step": 2585
+    },
+    {
+      "epoch": 0.6499780412823891,
+      "grad_norm": 0.22479486465454102,
+      "learning_rate": 1.5773443619284844e-05,
+      "loss": 0.084,
+      "step": 2590
+    },
+    {
+      "epoch": 0.6512328251458687,
+      "grad_norm": 0.36471477150917053,
+      "learning_rate": 1.5756835477689683e-05,
+      "loss": 0.0811,
+      "step": 2595
+    },
+    {
+      "epoch": 0.6524876090093481,
+      "grad_norm": 0.11465982347726822,
+      "learning_rate": 1.574020355185906e-05,
+      "loss": 0.0775,
+      "step": 2600
+    },
+    {
+      "epoch": 0.6537423928728276,
+      "grad_norm": 0.22774779796600342,
+      "learning_rate": 1.5723547910507392e-05,
+      "loss": 0.0893,
+      "step": 2605
+    },
+    {
+      "epoch": 0.6549971767363072,
+      "grad_norm": 0.17315097153186798,
+      "learning_rate": 1.5706868622447084e-05,
+      "loss": 0.0924,
+      "step": 2610
+    },
+    {
+      "epoch": 0.6562519605997866,
+      "grad_norm": 0.21010887622833252,
+      "learning_rate": 1.5690165756588235e-05,
+      "loss": 0.085,
+      "step": 2615
+    },
+    {
+      "epoch": 0.6575067444632662,
+      "grad_norm": 0.20467358827590942,
+      "learning_rate": 1.5673439381938365e-05,
+      "loss": 0.0857,
+      "step": 2620
+    },
+    {
+      "epoch": 0.6587615283267457,
+      "grad_norm": 0.0706283450126648,
+      "learning_rate": 1.565668956760211e-05,
+      "loss": 0.0921,
+      "step": 2625
+    },
+    {
+      "epoch": 0.6600163121902253,
+      "grad_norm": 0.10635245591402054,
+      "learning_rate": 1.563991638278094e-05,
+      "loss": 0.0901,
+      "step": 2630
+    },
+    {
+      "epoch": 0.6612710960537047,
+      "grad_norm": 0.2414940446615219,
+      "learning_rate": 1.56231198967729e-05,
+      "loss": 0.0979,
+      "step": 2635
+    },
+    {
+      "epoch": 0.6625258799171843,
+      "grad_norm": 0.3018660843372345,
+      "learning_rate": 1.560630017897229e-05,
+      "loss": 0.0774,
+      "step": 2640
+    },
+    {
+      "epoch": 0.6637806637806638,
+      "grad_norm": 0.08618688583374023,
+      "learning_rate": 1.558945729886938e-05,
+      "loss": 0.0917,
+      "step": 2645
+    },
+    {
+      "epoch": 0.6650354476441434,
+      "grad_norm": 0.3376272916793823,
+      "learning_rate": 1.5572591326050167e-05,
+      "loss": 0.0997,
+      "step": 2650
+    },
+    {
+      "epoch": 0.6662902315076228,
+      "grad_norm": 0.47232964634895325,
+      "learning_rate": 1.5555702330196024e-05,
+      "loss": 0.0984,
+      "step": 2655
+    },
+    {
+      "epoch": 0.6675450153711023,
+      "grad_norm": 0.3733605742454529,
+      "learning_rate": 1.5538790381083457e-05,
+      "loss": 0.095,
+      "step": 2660
+    },
+    {
+      "epoch": 0.6687997992345819,
+      "grad_norm": 0.197517529129982,
+      "learning_rate": 1.5521855548583807e-05,
+      "loss": 0.0882,
+      "step": 2665
+    },
+    {
+      "epoch": 0.6700545830980613,
+      "grad_norm": 0.5959796905517578,
+      "learning_rate": 1.550489790266294e-05,
+      "loss": 0.0978,
+      "step": 2670
+    },
+    {
+      "epoch": 0.6713093669615409,
+      "grad_norm": 0.3124695122241974,
+      "learning_rate": 1.5487917513381e-05,
+      "loss": 0.0877,
+      "step": 2675
+    },
+    {
+      "epoch": 0.6725641508250204,
+      "grad_norm": 0.5856664776802063,
+      "learning_rate": 1.5470914450892066e-05,
+      "loss": 0.0871,
+      "step": 2680
+    },
+    {
+      "epoch": 0.6738189346885,
+      "grad_norm": 0.11247047036886215,
+      "learning_rate": 1.5453888785443916e-05,
+      "loss": 0.0787,
+      "step": 2685
+    },
+    {
+      "epoch": 0.6750737185519794,
+      "grad_norm": 0.21094247698783875,
+      "learning_rate": 1.54368405873777e-05,
+      "loss": 0.0905,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6763285024154589,
+      "grad_norm": 0.32155776023864746,
+      "learning_rate": 1.5419769927127664e-05,
+      "loss": 0.0896,
+      "step": 2695
+    },
+    {
+      "epoch": 0.6775832862789385,
+      "grad_norm": 0.1487855464220047,
+      "learning_rate": 1.5402676875220847e-05,
+      "loss": 0.0924,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6788380701424179,
+      "grad_norm": 0.3109644651412964,
+      "learning_rate": 1.5385561502276813e-05,
+      "loss": 0.085,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6800928540058975,
+      "grad_norm": 0.13988961279392242,
+      "learning_rate": 1.536842387900733e-05,
+      "loss": 0.0806,
+      "step": 2710
+    },
+    {
+      "epoch": 0.681347637869377,
+      "grad_norm": 0.4081459045410156,
+      "learning_rate": 1.5351264076216114e-05,
+      "loss": 0.0912,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6826024217328566,
+      "grad_norm": 0.23864784836769104,
+      "learning_rate": 1.533408216479849e-05,
+      "loss": 0.088,
+      "step": 2720
+    },
+    {
+      "epoch": 0.683857205596336,
+      "grad_norm": 0.24918711185455322,
+      "learning_rate": 1.531687821574114e-05,
+      "loss": 0.0904,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6851119894598156,
+      "grad_norm": 0.23600496351718903,
+      "learning_rate": 1.5299652300121792e-05,
+      "loss": 0.0944,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6863667733232951,
+      "grad_norm": 0.1220758855342865,
+      "learning_rate": 1.5282404489108925e-05,
+      "loss": 0.0811,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6876215571867745,
+      "grad_norm": 0.1529039442539215,
+      "learning_rate": 1.5265134853961477e-05,
+      "loss": 0.0985,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6888763410502541,
+      "grad_norm": 0.3052196502685547,
+      "learning_rate": 1.524784346602856e-05,
+      "loss": 0.0917,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6901311249137336,
+      "grad_norm": 0.3731965720653534,
+      "learning_rate": 1.5230530396749148e-05,
+      "loss": 0.0906,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6913859087772132,
+      "grad_norm": 0.09618979692459106,
+      "learning_rate": 1.5213195717651793e-05,
+      "loss": 0.0914,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6926406926406926,
+      "grad_norm": 0.3307301104068756,
+      "learning_rate": 1.5195839500354337e-05,
+      "loss": 0.0859,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6938954765041722,
+      "grad_norm": 0.12252155691385269,
+      "learning_rate": 1.5178461816563594e-05,
+      "loss": 0.0881,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6951502603676517,
+      "grad_norm": 0.2949330806732178,
+      "learning_rate": 1.5161062738075068e-05,
+      "loss": 0.0782,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6964050442311311,
+      "grad_norm": 0.3150964379310608,
+      "learning_rate": 1.5143642336772663e-05,
+      "loss": 0.0866,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6976598280946107,
+      "grad_norm": 0.44747477769851685,
+      "learning_rate": 1.5126200684628372e-05,
+      "loss": 0.0943,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6989146119580902,
+      "grad_norm": 0.1062508150935173,
+      "learning_rate": 1.5108737853701981e-05,
+      "loss": 0.1027,
+      "step": 2785
+    },
+    {
+      "epoch": 0.7001693958215698,
+      "grad_norm": 0.16572189331054688,
+      "learning_rate": 1.5091253916140789e-05,
+      "loss": 0.0892,
+      "step": 2790
+    },
+    {
+      "epoch": 0.7014241796850492,
+      "grad_norm": 0.1096203476190567,
+      "learning_rate": 1.5073748944179282e-05,
+      "loss": 0.0889,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7026789635485288,
+      "grad_norm": 0.3676256835460663,
+      "learning_rate": 1.5056223010138857e-05,
+      "loss": 0.0998,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7039337474120083,
+      "grad_norm": 0.2482236921787262,
+      "learning_rate": 1.5038676186427515e-05,
+      "loss": 0.089,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7051885312754878,
+      "grad_norm": 0.13800540566444397,
+      "learning_rate": 1.5021108545539562e-05,
+      "loss": 0.0923,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7064433151389673,
+      "grad_norm": 0.33005279302597046,
+      "learning_rate": 1.5003520160055303e-05,
+      "loss": 0.0894,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7076980990024468,
+      "grad_norm": 0.34894976019859314,
+      "learning_rate": 1.4985911102640762e-05,
+      "loss": 0.0975,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7089528828659264,
+      "grad_norm": 0.2921280860900879,
+      "learning_rate": 1.4968281446047357e-05,
+      "loss": 0.089,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7102076667294058,
+      "grad_norm": 0.3317979872226715,
+      "learning_rate": 1.4950631263111615e-05,
+      "loss": 0.0889,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7114624505928854,
+      "grad_norm": 0.16637681424617767,
+      "learning_rate": 1.4932960626754867e-05,
+      "loss": 0.0788,
+      "step": 2835
+    },
+    {
+      "epoch": 0.7127172344563649,
+      "grad_norm": 0.15462718904018402,
+      "learning_rate": 1.491526960998295e-05,
+      "loss": 0.0887,
+      "step": 2840
+    },
+    {
+      "epoch": 0.7139720183198444,
+      "grad_norm": 0.23913145065307617,
+      "learning_rate": 1.4897558285885896e-05,
+      "loss": 0.0849,
+      "step": 2845
+    },
+    {
+      "epoch": 0.7152268021833239,
+      "grad_norm": 0.44012489914894104,
+      "learning_rate": 1.487982672763764e-05,
+      "loss": 0.0802,
+      "step": 2850
+    },
+    {
+      "epoch": 0.7164815860468035,
+      "grad_norm": 0.16587427258491516,
+      "learning_rate": 1.4862075008495718e-05,
+      "loss": 0.0913,
+      "step": 2855
+    },
+    {
+      "epoch": 0.717736369910283,
+      "grad_norm": 0.4225013256072998,
+      "learning_rate": 1.4844303201800949e-05,
+      "loss": 0.0828,
+      "step": 2860
+    },
+    {
+      "epoch": 0.7189911537737624,
+      "grad_norm": 0.3589664399623871,
+      "learning_rate": 1.4826511380977155e-05,
+      "loss": 0.0861,
+      "step": 2865
+    },
+    {
+      "epoch": 0.720245937637242,
+      "grad_norm": 0.13328975439071655,
+      "learning_rate": 1.4808699619530841e-05,
+      "loss": 0.0897,
+      "step": 2870
+    },
+    {
+      "epoch": 0.7215007215007215,
+      "grad_norm": 0.5425406098365784,
+      "learning_rate": 1.479086799105089e-05,
+      "loss": 0.0789,
+      "step": 2875
+    },
+    {
+      "epoch": 0.722755505364201,
+      "grad_norm": 0.22545233368873596,
+      "learning_rate": 1.4773016569208283e-05,
+      "loss": 0.0814,
+      "step": 2880
+    },
+    {
+      "epoch": 0.7240102892276805,
+      "grad_norm": 0.2654625177383423,
+      "learning_rate": 1.4755145427755755e-05,
+      "loss": 0.0846,
+      "step": 2885
+    },
+    {
+      "epoch": 0.7252650730911601,
+      "grad_norm": 0.4927530288696289,
+      "learning_rate": 1.4737254640527525e-05,
+      "loss": 0.0892,
+      "step": 2890
+    },
+    {
+      "epoch": 0.7265198569546396,
+      "grad_norm": 0.317210853099823,
+      "learning_rate": 1.4719344281438977e-05,
+      "loss": 0.0858,
+      "step": 2895
+    },
+    {
+      "epoch": 0.727774640818119,
+      "grad_norm": 0.3418554961681366,
+      "learning_rate": 1.4701414424486353e-05,
+      "loss": 0.0859,
+      "step": 2900
+    },
+    {
+      "epoch": 0.7290294246815986,
+      "grad_norm": 0.21974579989910126,
+      "learning_rate": 1.4683465143746452e-05,
+      "loss": 0.0988,
+      "step": 2905
+    },
+    {
+      "epoch": 0.7302842085450781,
+      "grad_norm": 0.35090965032577515,
+      "learning_rate": 1.466549651337632e-05,
+      "loss": 0.0909,
+      "step": 2910
+    },
+    {
+      "epoch": 0.7315389924085576,
+      "grad_norm": 0.27880313992500305,
+      "learning_rate": 1.4647508607612952e-05,
+      "loss": 0.0907,
+      "step": 2915
+    },
+    {
+      "epoch": 0.7327937762720371,
+      "grad_norm": 0.12414207309484482,
+      "learning_rate": 1.4629501500772962e-05,
+      "loss": 0.0912,
+      "step": 2920
+    },
+    {
+      "epoch": 0.7340485601355167,
+      "grad_norm": 0.3848329782485962,
+      "learning_rate": 1.4611475267252318e-05,
+      "loss": 0.0813,
+      "step": 2925
+    },
+    {
+      "epoch": 0.7353033439989962,
+      "grad_norm": 0.07426943629980087,
+      "learning_rate": 1.4593429981525985e-05,
+      "loss": 0.0862,
+      "step": 2930
+    },
+    {
+      "epoch": 0.7365581278624757,
+      "grad_norm": 0.1772748976945877,
+      "learning_rate": 1.4575365718147655e-05,
+      "loss": 0.111,
+      "step": 2935
+    },
+    {
+      "epoch": 0.7378129117259552,
+      "grad_norm": 0.23796626925468445,
+      "learning_rate": 1.4557282551749428e-05,
+      "loss": 0.0852,
+      "step": 2940
+    },
+    {
+      "epoch": 0.7390676955894347,
+      "grad_norm": 0.22509251534938812,
+      "learning_rate": 1.4539180557041494e-05,
+      "loss": 0.0804,
+      "step": 2945
+    },
+    {
+      "epoch": 0.7403224794529142,
+      "grad_norm": 0.29270699620246887,
+      "learning_rate": 1.452105980881183e-05,
+      "loss": 0.0925,
+      "step": 2950
+    },
+    {
+      "epoch": 0.7415772633163937,
+      "grad_norm": 0.27449578046798706,
+      "learning_rate": 1.4502920381925905e-05,
+      "loss": 0.084,
+      "step": 2955
+    },
+    {
+      "epoch": 0.7428320471798733,
+      "grad_norm": 0.23798063397407532,
+      "learning_rate": 1.4484762351326344e-05,
+      "loss": 0.0894,
+      "step": 2960
+    },
+    {
+      "epoch": 0.7440868310433528,
+      "grad_norm": 0.1779652088880539,
+      "learning_rate": 1.4466585792032644e-05,
+      "loss": 0.0829,
+      "step": 2965
+    },
+    {
+      "epoch": 0.7453416149068323,
+      "grad_norm": 0.10709716379642487,
+      "learning_rate": 1.4448390779140844e-05,
+      "loss": 0.0889,
+      "step": 2970
+    },
+    {
+      "epoch": 0.7465963987703118,
+      "grad_norm": 0.3699777126312256,
+      "learning_rate": 1.4430177387823232e-05,
+      "loss": 0.0925,
+      "step": 2975
+    },
+    {
+      "epoch": 0.7478511826337914,
+      "grad_norm": 0.23383845388889313,
+      "learning_rate": 1.4411945693328017e-05,
+      "loss": 0.0802,
+      "step": 2980
+    },
+    {
+      "epoch": 0.7491059664972708,
+      "grad_norm": 0.17309200763702393,
+      "learning_rate": 1.4393695770979038e-05,
+      "loss": 0.092,
+      "step": 2985
+    },
+    {
+      "epoch": 0.7503607503607503,
+      "grad_norm": 0.08418245613574982,
+      "learning_rate": 1.4375427696175434e-05,
+      "loss": 0.088,
+      "step": 2990
+    },
+    {
+      "epoch": 0.7516155342242299,
+      "grad_norm": 0.4208645820617676,
+      "learning_rate": 1.4357141544391342e-05,
+      "loss": 0.0946,
+      "step": 2995
+    },
+    {
+      "epoch": 0.7528703180877094,
+      "grad_norm": 0.10623873025178909,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.0876,
+      "step": 3000
+    },
+    {
+      "epoch": 0.7541251019511889,
+      "grad_norm": 0.13417641818523407,
+      "learning_rate": 1.4320515312151352e-05,
+      "loss": 0.0853,
+      "step": 3005
+    },
+    {
+      "epoch": 0.7553798858146684,
+      "grad_norm": 0.11833745986223221,
+      "learning_rate": 1.4302175383015907e-05,
+      "loss": 0.0923,
+      "step": 3010
+    },
+    {
+      "epoch": 0.756634669678148,
+      "grad_norm": 0.13997751474380493,
+      "learning_rate": 1.4283817679540246e-05,
+      "loss": 0.0842,
+      "step": 3015
+    },
+    {
+      "epoch": 0.7578894535416274,
+      "grad_norm": 0.4091770648956299,
+      "learning_rate": 1.4265442277568808e-05,
+      "loss": 0.0869,
+      "step": 3020
+    },
+    {
+      "epoch": 0.759144237405107,
+      "grad_norm": 0.38319870829582214,
+      "learning_rate": 1.4247049253019148e-05,
+      "loss": 0.0816,
+      "step": 3025
+    },
+    {
+      "epoch": 0.7603990212685865,
+      "grad_norm": 0.44087597727775574,
+      "learning_rate": 1.4228638681881633e-05,
+      "loss": 0.0925,
+      "step": 3030
+    },
+    {
+      "epoch": 0.761653805132066,
+      "grad_norm": 0.10818012803792953,
+      "learning_rate": 1.4210210640219117e-05,
+      "loss": 0.0873,
+      "step": 3035
+    },
+    {
+      "epoch": 0.7629085889955455,
+      "grad_norm": 0.3094254434108734,
+      "learning_rate": 1.4191765204166643e-05,
+      "loss": 0.0881,
+      "step": 3040
+    },
+    {
+      "epoch": 0.764163372859025,
+      "grad_norm": 0.14086200296878815,
+      "learning_rate": 1.4173302449931107e-05,
+      "loss": 0.0874,
+      "step": 3045
+    },
+    {
+      "epoch": 0.7654181567225046,
+      "grad_norm": 0.10829006880521774,
+      "learning_rate": 1.4154822453790963e-05,
+      "loss": 0.0813,
+      "step": 3050
+    },
+    {
+      "epoch": 0.7666729405859841,
+      "grad_norm": 0.21264980733394623,
+      "learning_rate": 1.4136325292095899e-05,
+      "loss": 0.0823,
+      "step": 3055
+    },
+    {
+      "epoch": 0.7679277244494636,
+      "grad_norm": 0.07161971926689148,
+      "learning_rate": 1.4117811041266518e-05,
+      "loss": 0.0878,
+      "step": 3060
+    },
+    {
+      "epoch": 0.7691825083129431,
+      "grad_norm": 0.24035820364952087,
+      "learning_rate": 1.4099279777794026e-05,
+      "loss": 0.0811,
+      "step": 3065
+    },
+    {
+      "epoch": 0.7704372921764227,
+      "grad_norm": 0.17685869336128235,
+      "learning_rate": 1.4080731578239917e-05,
+      "loss": 0.0854,
+      "step": 3070
+    },
+    {
+      "epoch": 0.7716920760399021,
+      "grad_norm": 0.33490845561027527,
+      "learning_rate": 1.4062166519235665e-05,
+      "loss": 0.0875,
+      "step": 3075
+    },
+    {
+      "epoch": 0.7729468599033816,
+      "grad_norm": 0.24075230956077576,
+      "learning_rate": 1.4043584677482383e-05,
+      "loss": 0.0942,
+      "step": 3080
+    },
+    {
+      "epoch": 0.7742016437668612,
+      "grad_norm": 0.16541673243045807,
+      "learning_rate": 1.4024986129750535e-05,
+      "loss": 0.0924,
+      "step": 3085
+    },
+    {
+      "epoch": 0.7754564276303407,
+      "grad_norm": 0.36374107003211975,
+      "learning_rate": 1.40063709528796e-05,
+      "loss": 0.0959,
+      "step": 3090
+    },
+    {
+      "epoch": 0.7767112114938202,
+      "grad_norm": 0.19677504897117615,
+      "learning_rate": 1.3987739223777756e-05,
+      "loss": 0.0841,
+      "step": 3095
+    },
+    {
+      "epoch": 0.7779659953572997,
+      "grad_norm": 0.24688193202018738,
+      "learning_rate": 1.3969091019421573e-05,
+      "loss": 0.0795,
+      "step": 3100
+    },
+    {
+      "epoch": 0.7792207792207793,
+      "grad_norm": 0.4386122524738312,
+      "learning_rate": 1.3950426416855685e-05,
+      "loss": 0.091,
+      "step": 3105
+    },
+    {
+      "epoch": 0.7804755630842587,
+      "grad_norm": 0.281536340713501,
+      "learning_rate": 1.3931745493192473e-05,
+      "loss": 0.0809,
+      "step": 3110
+    },
+    {
+      "epoch": 0.7817303469477382,
+      "grad_norm": 0.3444419801235199,
+      "learning_rate": 1.391304832561175e-05,
+      "loss": 0.0819,
+      "step": 3115
+    },
+    {
+      "epoch": 0.7829851308112178,
+      "grad_norm": 0.3660227060317993,
+      "learning_rate": 1.3894334991360448e-05,
+      "loss": 0.0908,
+      "step": 3120
+    },
+    {
+      "epoch": 0.7842399146746973,
+      "grad_norm": 0.15960213541984558,
+      "learning_rate": 1.3875605567752275e-05,
+      "loss": 0.0817,
+      "step": 3125
+    },
+    {
+      "epoch": 0.7854946985381768,
+      "grad_norm": 0.10506850481033325,
+      "learning_rate": 1.3856860132167423e-05,
+      "loss": 0.0867,
+      "step": 3130
+    },
+    {
+      "epoch": 0.7867494824016563,
+      "grad_norm": 0.28331154584884644,
+      "learning_rate": 1.3838098762052237e-05,
+      "loss": 0.0899,
+      "step": 3135
+    },
+    {
+      "epoch": 0.7880042662651359,
+      "grad_norm": 0.07872291654348373,
+      "learning_rate": 1.381932153491889e-05,
+      "loss": 0.0864,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7892590501286153,
+      "grad_norm": 0.26927658915519714,
+      "learning_rate": 1.3800528528345074e-05,
+      "loss": 0.0928,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7905138339920948,
+      "grad_norm": 0.4041297733783722,
+      "learning_rate": 1.378171981997367e-05,
+      "loss": 0.0891,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7917686178555744,
+      "grad_norm": 0.29741108417510986,
+      "learning_rate": 1.3762895487512426e-05,
+      "loss": 0.0827,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7930234017190539,
+      "grad_norm": 0.11423831433057785,
+      "learning_rate": 1.3744055608733654e-05,
+      "loss": 0.0853,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7942781855825334,
+      "grad_norm": 0.07605528086423874,
+      "learning_rate": 1.3725200261473879e-05,
+      "loss": 0.0958,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7955329694460129,
+      "grad_norm": 0.27038949728012085,
+      "learning_rate": 1.3706329523633546e-05,
+      "loss": 0.0866,
+      "step": 3170
+    },
+    {
+      "epoch": 0.7967877533094925,
+      "grad_norm": 0.17523351311683655,
+      "learning_rate": 1.3687443473176678e-05,
+      "loss": 0.092,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7980425371729719,
+      "grad_norm": 0.14444954693317413,
+      "learning_rate": 1.3668542188130567e-05,
+      "loss": 0.0937,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7992973210364515,
+      "grad_norm": 0.09110607951879501,
+      "learning_rate": 1.3649625746585442e-05,
+      "loss": 0.0894,
+      "step": 3185
+    },
+    {
+      "epoch": 0.800552104899931,
+      "grad_norm": 0.24283112585544586,
+      "learning_rate": 1.3630694226694159e-05,
+      "loss": 0.0879,
+      "step": 3190
+    },
+    {
+      "epoch": 0.8018068887634106,
+      "grad_norm": 0.2000175565481186,
+      "learning_rate": 1.3611747706671859e-05,
+      "loss": 0.0809,
+      "step": 3195
+    },
+    {
+      "epoch": 0.80306167262689,
+      "grad_norm": 0.13179725408554077,
+      "learning_rate": 1.3592786264795659e-05,
+      "loss": 0.0792,
+      "step": 3200
+    },
+    {
+      "epoch": 0.8043164564903695,
+      "grad_norm": 0.17673130333423615,
+      "learning_rate": 1.357380997940433e-05,
+      "loss": 0.0819,
+      "step": 3205
+    },
+    {
+      "epoch": 0.8055712403538491,
+      "grad_norm": 0.1775507628917694,
+      "learning_rate": 1.3554818928897965e-05,
+      "loss": 0.094,
+      "step": 3210
+    },
+    {
+      "epoch": 0.8068260242173285,
+      "grad_norm": 0.10443057119846344,
+      "learning_rate": 1.3535813191737663e-05,
+      "loss": 0.0921,
+      "step": 3215
+    },
+    {
+      "epoch": 0.8080808080808081,
+      "grad_norm": 0.45298513770103455,
+      "learning_rate": 1.351679284644519e-05,
+      "loss": 0.092,
+      "step": 3220
+    },
+    {
+      "epoch": 0.8093355919442876,
+      "grad_norm": 0.1088099479675293,
+      "learning_rate": 1.3497757971602677e-05,
+      "loss": 0.0945,
+      "step": 3225
+    },
+    {
+      "epoch": 0.8105903758077672,
+      "grad_norm": 0.11513658612966537,
+      "learning_rate": 1.3478708645852272e-05,
+      "loss": 0.0906,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8118451596712466,
+      "grad_norm": 0.15573541820049286,
+      "learning_rate": 1.3459644947895844e-05,
+      "loss": 0.0895,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8130999435347261,
+      "grad_norm": 0.548417866230011,
+      "learning_rate": 1.344056695649462e-05,
+      "loss": 0.0957,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8143547273982057,
+      "grad_norm": 0.24571377038955688,
+      "learning_rate": 1.3421474750468893e-05,
+      "loss": 0.0852,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8156095112616851,
+      "grad_norm": 0.1693846732378006,
+      "learning_rate": 1.3402368408697681e-05,
+      "loss": 0.0962,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8168642951251647,
+      "grad_norm": 0.12856455147266388,
+      "learning_rate": 1.3383248010118404e-05,
+      "loss": 0.0926,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8181190789886442,
+      "grad_norm": 0.3115483820438385,
+      "learning_rate": 1.336411363372655e-05,
+      "loss": 0.0943,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8193738628521238,
+      "grad_norm": 0.44211283326148987,
+      "learning_rate": 1.3344965358575368e-05,
+      "loss": 0.102,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8206286467156032,
+      "grad_norm": 0.1086539626121521,
+      "learning_rate": 1.3325803263775521e-05,
+      "loss": 0.0968,
+      "step": 3270
+    },
+    {
+      "epoch": 0.8218834305790828,
+      "grad_norm": 0.24857747554779053,
+      "learning_rate": 1.3306627428494769e-05,
+      "loss": 0.097,
+      "step": 3275
+    },
+    {
+      "epoch": 0.8231382144425623,
+      "grad_norm": 0.5873998403549194,
+      "learning_rate": 1.3287437931957642e-05,
+      "loss": 0.0848,
+      "step": 3280
+    },
+    {
+      "epoch": 0.8243929983060417,
+      "grad_norm": 0.2507260739803314,
+      "learning_rate": 1.3268234853445113e-05,
+      "loss": 0.0917,
+      "step": 3285
+    },
+    {
+      "epoch": 0.8256477821695213,
+      "grad_norm": 0.2204979509115219,
+      "learning_rate": 1.3249018272294261e-05,
+      "loss": 0.083,
+      "step": 3290
+    },
+    {
+      "epoch": 0.8269025660330008,
+      "grad_norm": 0.19563211500644684,
+      "learning_rate": 1.3229788267897958e-05,
+      "loss": 0.0803,
+      "step": 3295
+    },
+    {
+      "epoch": 0.8281573498964804,
+      "grad_norm": 0.24326251447200775,
+      "learning_rate": 1.3210544919704539e-05,
+      "loss": 0.078,
+      "step": 3300
+    },
+    {
+      "epoch": 0.8294121337599598,
+      "grad_norm": 0.08204273879528046,
+      "learning_rate": 1.319128830721745e-05,
+      "loss": 0.0866,
+      "step": 3305
+    },
+    {
+      "epoch": 0.8306669176234394,
+      "grad_norm": 0.3777156174182892,
+      "learning_rate": 1.317201850999496e-05,
+      "loss": 0.0892,
+      "step": 3310
+    },
+    {
+      "epoch": 0.8319217014869189,
+      "grad_norm": 0.15763606131076813,
+      "learning_rate": 1.315273560764979e-05,
+      "loss": 0.0947,
+      "step": 3315
+    },
+    {
+      "epoch": 0.8331764853503983,
+      "grad_norm": 0.4323311448097229,
+      "learning_rate": 1.3133439679848824e-05,
+      "loss": 0.091,
+      "step": 3320
+    },
+    {
+      "epoch": 0.8344312692138779,
+      "grad_norm": 0.51838618516922,
+      "learning_rate": 1.3114130806312744e-05,
+      "loss": 0.0783,
+      "step": 3325
+    },
+    {
+      "epoch": 0.8356860530773574,
+      "grad_norm": 0.2101561725139618,
+      "learning_rate": 1.3094809066815731e-05,
+      "loss": 0.0797,
+      "step": 3330
+    },
+    {
+      "epoch": 0.836940836940837,
+      "grad_norm": 0.35820868611335754,
+      "learning_rate": 1.3075474541185104e-05,
+      "loss": 0.0904,
+      "step": 3335
+    },
+    {
+      "epoch": 0.8381956208043164,
+      "grad_norm": 0.17395086586475372,
+      "learning_rate": 1.3056127309301027e-05,
+      "loss": 0.0773,
+      "step": 3340
+    },
+    {
+      "epoch": 0.839450404667796,
+      "grad_norm": 0.22628264129161835,
+      "learning_rate": 1.3036767451096148e-05,
+      "loss": 0.0833,
+      "step": 3345
+    },
+    {
+      "epoch": 0.8407051885312755,
+      "grad_norm": 0.14669981598854065,
+      "learning_rate": 1.3017395046555284e-05,
+      "loss": 0.0866,
+      "step": 3350
+    },
+    {
+      "epoch": 0.841959972394755,
+      "grad_norm": 0.45152294635772705,
+      "learning_rate": 1.2998010175715081e-05,
+      "loss": 0.0955,
+      "step": 3355
+    },
+    {
+      "epoch": 0.8432147562582345,
+      "grad_norm": 0.14131848514080048,
+      "learning_rate": 1.2978612918663702e-05,
+      "loss": 0.0805,
+      "step": 3360
+    },
+    {
+      "epoch": 0.844469540121714,
+      "grad_norm": 0.4170994460582733,
+      "learning_rate": 1.2959203355540466e-05,
+      "loss": 0.0885,
+      "step": 3365
+    },
+    {
+      "epoch": 0.8457243239851936,
+      "grad_norm": 0.2687641382217407,
+      "learning_rate": 1.2939781566535551e-05,
+      "loss": 0.0859,
+      "step": 3370
+    },
+    {
+      "epoch": 0.846979107848673,
+      "grad_norm": 0.15929897129535675,
+      "learning_rate": 1.2920347631889637e-05,
+      "loss": 0.0876,
+      "step": 3375
+    },
+    {
+      "epoch": 0.8482338917121526,
+      "grad_norm": 0.1964198797941208,
+      "learning_rate": 1.2900901631893585e-05,
+      "loss": 0.0966,
+      "step": 3380
+    },
+    {
+      "epoch": 0.8494886755756321,
+      "grad_norm": 0.21609993278980255,
+      "learning_rate": 1.28814436468881e-05,
+      "loss": 0.0811,
+      "step": 3385
+    },
+    {
+      "epoch": 0.8507434594391116,
+      "grad_norm": 0.24047359824180603,
+      "learning_rate": 1.2861973757263416e-05,
+      "loss": 0.0858,
+      "step": 3390
+    },
+    {
+      "epoch": 0.8519982433025911,
+      "grad_norm": 0.25712600350379944,
+      "learning_rate": 1.2842492043458929e-05,
+      "loss": 0.0793,
+      "step": 3395
+    },
+    {
+      "epoch": 0.8532530271660707,
+      "grad_norm": 0.1645137518644333,
+      "learning_rate": 1.2822998585962909e-05,
+      "loss": 0.0801,
+      "step": 3400
+    },
+    {
+      "epoch": 0.8545078110295502,
+      "grad_norm": 0.601332426071167,
+      "learning_rate": 1.280349346531213e-05,
+      "loss": 0.0891,
+      "step": 3405
+    },
+    {
+      "epoch": 0.8557625948930296,
+      "grad_norm": 0.3988315761089325,
+      "learning_rate": 1.2783976762091554e-05,
+      "loss": 0.0885,
+      "step": 3410
+    },
+    {
+      "epoch": 0.8570173787565092,
+      "grad_norm": 0.37066715955734253,
+      "learning_rate": 1.2764448556934001e-05,
+      "loss": 0.0935,
+      "step": 3415
+    },
+    {
+      "epoch": 0.8582721626199887,
+      "grad_norm": 0.1261286735534668,
+      "learning_rate": 1.274490893051981e-05,
+      "loss": 0.0824,
+      "step": 3420
+    },
+    {
+      "epoch": 0.8595269464834682,
+      "grad_norm": 0.11686058342456818,
+      "learning_rate": 1.2725357963576506e-05,
+      "loss": 0.0963,
+      "step": 3425
+    },
+    {
+      "epoch": 0.8607817303469477,
+      "grad_norm": 0.1499367356300354,
+      "learning_rate": 1.2705795736878461e-05,
+      "loss": 0.0899,
+      "step": 3430
+    },
+    {
+      "epoch": 0.8620365142104273,
+      "grad_norm": 0.25637149810791016,
+      "learning_rate": 1.268622233124658e-05,
+      "loss": 0.09,
+      "step": 3435
+    },
+    {
+      "epoch": 0.8632912980739068,
+      "grad_norm": 0.2606950104236603,
+      "learning_rate": 1.2666637827547935e-05,
+      "loss": 0.0928,
+      "step": 3440
+    },
+    {
+      "epoch": 0.8645460819373862,
+      "grad_norm": 0.17637059092521667,
+      "learning_rate": 1.264704230669547e-05,
+      "loss": 0.0999,
+      "step": 3445
+    },
+    {
+      "epoch": 0.8658008658008658,
+      "grad_norm": 0.21137331426143646,
+      "learning_rate": 1.2627435849647629e-05,
+      "loss": 0.0871,
+      "step": 3450
+    },
+    {
+      "epoch": 0.8670556496643453,
+      "grad_norm": 0.40598878264427185,
+      "learning_rate": 1.2607818537408047e-05,
+      "loss": 0.0821,
+      "step": 3455
+    },
+    {
+      "epoch": 0.8683104335278249,
+      "grad_norm": 0.11776316165924072,
+      "learning_rate": 1.2588190451025209e-05,
+      "loss": 0.0987,
+      "step": 3460
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 0.11056658625602722,
+      "learning_rate": 1.2568551671592106e-05,
+      "loss": 0.0863,
+      "step": 3465
+    },
+    {
+      "epoch": 0.8708200012547839,
+      "grad_norm": 0.29835495352745056,
+      "learning_rate": 1.2548902280245909e-05,
+      "loss": 0.0876,
+      "step": 3470
+    },
+    {
+      "epoch": 0.8720747851182634,
+      "grad_norm": 0.3364979922771454,
+      "learning_rate": 1.252924235816764e-05,
+      "loss": 0.0788,
+      "step": 3475
+    },
+    {
+      "epoch": 0.8733295689817429,
+      "grad_norm": 0.1180335134267807,
+      "learning_rate": 1.2509571986581814e-05,
+      "loss": 0.0799,
+      "step": 3480
+    },
+    {
+      "epoch": 0.8745843528452224,
+      "grad_norm": 0.2995622754096985,
+      "learning_rate": 1.2489891246756131e-05,
+      "loss": 0.0964,
+      "step": 3485
+    },
+    {
+      "epoch": 0.875839136708702,
+      "grad_norm": 0.23864442110061646,
+      "learning_rate": 1.2470200220001122e-05,
+      "loss": 0.0982,
+      "step": 3490
+    },
+    {
+      "epoch": 0.8770939205721815,
+      "grad_norm": 0.6277647018432617,
+      "learning_rate": 1.245049898766982e-05,
+      "loss": 0.0722,
+      "step": 3495
+    },
+    {
+      "epoch": 0.8783487044356609,
+      "grad_norm": 0.4860200881958008,
+      "learning_rate": 1.2430787631157414e-05,
+      "loss": 0.0952,
+      "step": 3500
+    },
+    {
+      "epoch": 0.8796034882991405,
+      "grad_norm": 0.30421602725982666,
+      "learning_rate": 1.2411066231900935e-05,
+      "loss": 0.0901,
+      "step": 3505
+    },
+    {
+      "epoch": 0.88085827216262,
+      "grad_norm": 0.20121634006500244,
+      "learning_rate": 1.239133487137889e-05,
+      "loss": 0.0833,
+      "step": 3510
+    },
+    {
+      "epoch": 0.8821130560260995,
+      "grad_norm": 0.3852861225605011,
+      "learning_rate": 1.2371593631110953e-05,
+      "loss": 0.0862,
+      "step": 3515
+    },
+    {
+      "epoch": 0.883367839889579,
+      "grad_norm": 0.45018061995506287,
+      "learning_rate": 1.2351842592657612e-05,
+      "loss": 0.098,
+      "step": 3520
+    },
+    {
+      "epoch": 0.8846226237530586,
+      "grad_norm": 0.40172022581100464,
+      "learning_rate": 1.2332081837619836e-05,
+      "loss": 0.0899,
+      "step": 3525
+    },
+    {
+      "epoch": 0.8858774076165381,
+      "grad_norm": 0.2906516492366791,
+      "learning_rate": 1.2312311447638731e-05,
+      "loss": 0.0852,
+      "step": 3530
+    },
+    {
+      "epoch": 0.8871321914800175,
+      "grad_norm": 0.39876627922058105,
+      "learning_rate": 1.2292531504395223e-05,
+      "loss": 0.0938,
+      "step": 3535
+    },
+    {
+      "epoch": 0.8883869753434971,
+      "grad_norm": 0.1180957704782486,
+      "learning_rate": 1.2272742089609694e-05,
+      "loss": 0.0862,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8896417592069766,
+      "grad_norm": 0.1698416918516159,
+      "learning_rate": 1.2252943285041662e-05,
+      "loss": 0.0873,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8908965430704561,
+      "grad_norm": 0.3837547302246094,
+      "learning_rate": 1.2233135172489453e-05,
+      "loss": 0.0888,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8921513269339356,
+      "grad_norm": 0.09981489181518555,
+      "learning_rate": 1.221331783378982e-05,
+      "loss": 0.0802,
+      "step": 3555
+    },
+    {
+      "epoch": 0.8934061107974152,
+      "grad_norm": 0.21812370419502258,
+      "learning_rate": 1.2193491350817657e-05,
+      "loss": 0.0749,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8946608946608947,
+      "grad_norm": 0.6983737349510193,
+      "learning_rate": 1.2173655805485627e-05,
+      "loss": 0.0881,
+      "step": 3565
+    },
+    {
+      "epoch": 0.8959156785243741,
+      "grad_norm": 0.3287610411643982,
+      "learning_rate": 1.2153811279743841e-05,
+      "loss": 0.0859,
+      "step": 3570
+    },
+    {
+      "epoch": 0.8971704623878537,
+      "grad_norm": 0.3805353045463562,
+      "learning_rate": 1.2133957855579501e-05,
+      "loss": 0.0874,
+      "step": 3575
+    },
+    {
+      "epoch": 0.8984252462513332,
+      "grad_norm": 0.3398018181324005,
+      "learning_rate": 1.2114095615016585e-05,
+      "loss": 0.0892,
+      "step": 3580
+    },
+    {
+      "epoch": 0.8996800301148127,
+      "grad_norm": 0.2726369798183441,
+      "learning_rate": 1.2094224640115488e-05,
+      "loss": 0.0846,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9009348139782922,
+      "grad_norm": 0.2575783133506775,
+      "learning_rate": 1.2074345012972694e-05,
+      "loss": 0.0863,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9021895978417718,
+      "grad_norm": 0.18294771015644073,
+      "learning_rate": 1.2054456815720432e-05,
+      "loss": 0.0792,
+      "step": 3595
+    },
+    {
+      "epoch": 0.9034443817052513,
+      "grad_norm": 0.2737415134906769,
+      "learning_rate": 1.2034560130526341e-05,
+      "loss": 0.0811,
+      "step": 3600
+    },
+    {
+      "epoch": 0.9046991655687308,
+      "grad_norm": 0.22107063233852386,
+      "learning_rate": 1.2014655039593119e-05,
+      "loss": 0.0937,
+      "step": 3605
+    },
+    {
+      "epoch": 0.9059539494322103,
+      "grad_norm": 0.37395817041397095,
+      "learning_rate": 1.1994741625158206e-05,
+      "loss": 0.0813,
+      "step": 3610
+    },
+    {
+      "epoch": 0.9072087332956899,
+      "grad_norm": 0.3313354253768921,
+      "learning_rate": 1.1974819969493421e-05,
+      "loss": 0.0789,
+      "step": 3615
+    },
+    {
+      "epoch": 0.9084635171591693,
+      "grad_norm": 0.3608105182647705,
+      "learning_rate": 1.195489015490463e-05,
+      "loss": 0.0859,
+      "step": 3620
+    },
+    {
+      "epoch": 0.9097183010226488,
+      "grad_norm": 0.37519270181655884,
+      "learning_rate": 1.1934952263731411e-05,
+      "loss": 0.0919,
+      "step": 3625
+    },
+    {
+      "epoch": 0.9109730848861284,
+      "grad_norm": 0.5066865086555481,
+      "learning_rate": 1.1915006378346719e-05,
+      "loss": 0.1,
+      "step": 3630
+    },
+    {
+      "epoch": 0.9122278687496079,
+      "grad_norm": 0.22127996385097504,
+      "learning_rate": 1.1895052581156516e-05,
+      "loss": 0.0854,
+      "step": 3635
+    },
+    {
+      "epoch": 0.9134826526130874,
+      "grad_norm": 0.18663233518600464,
+      "learning_rate": 1.1875090954599472e-05,
+      "loss": 0.0962,
+      "step": 3640
+    },
+    {
+      "epoch": 0.9147374364765669,
+      "grad_norm": 0.5338375568389893,
+      "learning_rate": 1.1855121581146591e-05,
+      "loss": 0.086,
+      "step": 3645
+    },
+    {
+      "epoch": 0.9159922203400465,
+      "grad_norm": 0.10081567615270615,
+      "learning_rate": 1.183514454330089e-05,
+      "loss": 0.0863,
+      "step": 3650
+    },
+    {
+      "epoch": 0.9172470042035259,
+      "grad_norm": 0.09096982330083847,
+      "learning_rate": 1.1815159923597044e-05,
+      "loss": 0.084,
+      "step": 3655
+    },
+    {
+      "epoch": 0.9185017880670054,
+      "grad_norm": 0.2180848866701126,
+      "learning_rate": 1.1795167804601062e-05,
+      "loss": 0.0916,
+      "step": 3660
+    },
+    {
+      "epoch": 0.919756571930485,
+      "grad_norm": 0.2790067791938782,
+      "learning_rate": 1.177516826890993e-05,
+      "loss": 0.0914,
+      "step": 3665
+    },
+    {
+      "epoch": 0.9210113557939645,
+      "grad_norm": 0.26915568113327026,
+      "learning_rate": 1.1755161399151277e-05,
+      "loss": 0.0887,
+      "step": 3670
+    },
+    {
+      "epoch": 0.922266139657444,
+      "grad_norm": 0.050155360251665115,
+      "learning_rate": 1.1735147277983027e-05,
+      "loss": 0.0957,
+      "step": 3675
+    },
+    {
+      "epoch": 0.9235209235209235,
+      "grad_norm": 0.07628196477890015,
+      "learning_rate": 1.1715125988093075e-05,
+      "loss": 0.0895,
+      "step": 3680
+    },
+    {
+      "epoch": 0.9247757073844031,
+      "grad_norm": 0.23782391846179962,
+      "learning_rate": 1.1695097612198929e-05,
+      "loss": 0.0878,
+      "step": 3685
+    },
+    {
+      "epoch": 0.9260304912478825,
+      "grad_norm": 0.28425681591033936,
+      "learning_rate": 1.1675062233047365e-05,
+      "loss": 0.0928,
+      "step": 3690
+    },
+    {
+      "epoch": 0.927285275111362,
+      "grad_norm": 0.36293289065361023,
+      "learning_rate": 1.16550199334141e-05,
+      "loss": 0.0976,
+      "step": 3695
+    },
+    {
+      "epoch": 0.9285400589748416,
+      "grad_norm": 0.5268048048019409,
+      "learning_rate": 1.1634970796103442e-05,
+      "loss": 0.0995,
+      "step": 3700
+    },
+    {
+      "epoch": 0.9297948428383211,
+      "grad_norm": 0.18220224976539612,
+      "learning_rate": 1.1614914903947952e-05,
+      "loss": 0.0895,
+      "step": 3705
+    },
+    {
+      "epoch": 0.9310496267018006,
+      "grad_norm": 0.18643346428871155,
+      "learning_rate": 1.1594852339808082e-05,
+      "loss": 0.0977,
+      "step": 3710
+    },
+    {
+      "epoch": 0.9323044105652801,
+      "grad_norm": 0.1516513228416443,
+      "learning_rate": 1.1574783186571876e-05,
+      "loss": 0.0947,
+      "step": 3715
+    },
+    {
+      "epoch": 0.9335591944287597,
+      "grad_norm": 0.27426013350486755,
+      "learning_rate": 1.155470752715458e-05,
+      "loss": 0.0885,
+      "step": 3720
+    },
+    {
+      "epoch": 0.9348139782922391,
+      "grad_norm": 0.2292599081993103,
+      "learning_rate": 1.1534625444498325e-05,
+      "loss": 0.0774,
+      "step": 3725
+    },
+    {
+      "epoch": 0.9360687621557187,
+      "grad_norm": 0.2502307593822479,
+      "learning_rate": 1.1514537021571784e-05,
+      "loss": 0.0816,
+      "step": 3730
+    },
+    {
+      "epoch": 0.9373235460191982,
+      "grad_norm": 0.1261693835258484,
+      "learning_rate": 1.1494442341369819e-05,
+      "loss": 0.0859,
+      "step": 3735
+    },
+    {
+      "epoch": 0.9385783298826778,
+      "grad_norm": 0.18473610281944275,
+      "learning_rate": 1.1474341486913146e-05,
+      "loss": 0.0804,
+      "step": 3740
+    },
+    {
+      "epoch": 0.9398331137461572,
+      "grad_norm": 0.1409272700548172,
+      "learning_rate": 1.1454234541247995e-05,
+      "loss": 0.0985,
+      "step": 3745
+    },
+    {
+      "epoch": 0.9410878976096367,
+      "grad_norm": 0.32273778319358826,
+      "learning_rate": 1.1434121587445752e-05,
+      "loss": 0.0901,
+      "step": 3750
+    },
+    {
+      "epoch": 0.9423426814731163,
+      "grad_norm": 0.12227199226617813,
+      "learning_rate": 1.1414002708602632e-05,
+      "loss": 0.0928,
+      "step": 3755
+    },
+    {
+      "epoch": 0.9435974653365957,
+      "grad_norm": 0.2481284737586975,
+      "learning_rate": 1.1393877987839329e-05,
+      "loss": 0.0853,
+      "step": 3760
+    },
+    {
+      "epoch": 0.9448522492000753,
+      "grad_norm": 0.25302889943122864,
+      "learning_rate": 1.1373747508300668e-05,
+      "loss": 0.0729,
+      "step": 3765
+    },
+    {
+      "epoch": 0.9461070330635548,
+      "grad_norm": 0.28753232955932617,
+      "learning_rate": 1.1353611353155272e-05,
+      "loss": 0.0916,
+      "step": 3770
+    },
+    {
+      "epoch": 0.9473618169270344,
+      "grad_norm": 0.308005154132843,
+      "learning_rate": 1.133346960559521e-05,
+      "loss": 0.0896,
+      "step": 3775
+    },
+    {
+      "epoch": 0.9486166007905138,
+      "grad_norm": 0.22970429062843323,
+      "learning_rate": 1.1313322348835658e-05,
+      "loss": 0.0871,
+      "step": 3780
+    },
+    {
+      "epoch": 0.9498713846539933,
+      "grad_norm": 0.3094078600406647,
+      "learning_rate": 1.1293169666114546e-05,
+      "loss": 0.0929,
+      "step": 3785
+    },
+    {
+      "epoch": 0.9511261685174729,
+      "grad_norm": 0.17183607816696167,
+      "learning_rate": 1.127301164069223e-05,
+      "loss": 0.0901,
+      "step": 3790
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.3873693346977234,
+      "learning_rate": 1.1252848355851136e-05,
+      "loss": 0.0829,
+      "step": 3795
+    },
+    {
+      "epoch": 0.9536357362444319,
+      "grad_norm": 0.42757073044776917,
+      "learning_rate": 1.1232679894895417e-05,
+      "loss": 0.0855,
+      "step": 3800
+    },
+    {
+      "epoch": 0.9548905201079114,
+      "grad_norm": 0.49850013852119446,
+      "learning_rate": 1.1212506341150615e-05,
+      "loss": 0.0834,
+      "step": 3805
+    },
+    {
+      "epoch": 0.956145303971391,
+      "grad_norm": 0.14661534130573273,
+      "learning_rate": 1.1192327777963313e-05,
+      "loss": 0.0798,
+      "step": 3810
+    },
+    {
+      "epoch": 0.9574000878348704,
+      "grad_norm": 0.29823562502861023,
+      "learning_rate": 1.117214428870078e-05,
+      "loss": 0.0786,
+      "step": 3815
+    },
+    {
+      "epoch": 0.95865487169835,
+      "grad_norm": 0.12722723186016083,
+      "learning_rate": 1.1151955956750652e-05,
+      "loss": 0.0807,
+      "step": 3820
+    },
+    {
+      "epoch": 0.9599096555618295,
+      "grad_norm": 0.10096994042396545,
+      "learning_rate": 1.1131762865520566e-05,
+      "loss": 0.0847,
+      "step": 3825
+    },
+    {
+      "epoch": 0.9611644394253089,
+      "grad_norm": 0.1767021268606186,
+      "learning_rate": 1.1111565098437815e-05,
+      "loss": 0.1014,
+      "step": 3830
+    },
+    {
+      "epoch": 0.9624192232887885,
+      "grad_norm": 0.2014637440443039,
+      "learning_rate": 1.1091362738949024e-05,
+      "loss": 0.091,
+      "step": 3835
+    },
+    {
+      "epoch": 0.963674007152268,
+      "grad_norm": 0.2271222472190857,
+      "learning_rate": 1.1071155870519777e-05,
+      "loss": 0.0887,
+      "step": 3840
+    },
+    {
+      "epoch": 0.9649287910157476,
+      "grad_norm": 0.15430879592895508,
+      "learning_rate": 1.1050944576634298e-05,
+      "loss": 0.0842,
+      "step": 3845
+    },
+    {
+      "epoch": 0.966183574879227,
+      "grad_norm": 0.14442376792430878,
+      "learning_rate": 1.1030728940795087e-05,
+      "loss": 0.0935,
+      "step": 3850
+    },
+    {
+      "epoch": 0.9674383587427066,
+      "grad_norm": 0.1097768247127533,
+      "learning_rate": 1.101050904652259e-05,
+      "loss": 0.0842,
+      "step": 3855
+    },
+    {
+      "epoch": 0.9686931426061861,
+      "grad_norm": 0.14131766557693481,
+      "learning_rate": 1.0990284977354841e-05,
+      "loss": 0.0848,
+      "step": 3860
+    },
+    {
+      "epoch": 0.9699479264696657,
+      "grad_norm": 0.2724064588546753,
+      "learning_rate": 1.097005681684712e-05,
+      "loss": 0.0792,
+      "step": 3865
+    },
+    {
+      "epoch": 0.9712027103331451,
+      "grad_norm": 0.15965338051319122,
+      "learning_rate": 1.094982464857162e-05,
+      "loss": 0.0699,
+      "step": 3870
+    },
+    {
+      "epoch": 0.9724574941966246,
+      "grad_norm": 0.25648635625839233,
+      "learning_rate": 1.0929588556117086e-05,
+      "loss": 0.0869,
+      "step": 3875
+    },
+    {
+      "epoch": 0.9737122780601042,
+      "grad_norm": 0.1974741518497467,
+      "learning_rate": 1.0909348623088472e-05,
+      "loss": 0.0814,
+      "step": 3880
+    },
+    {
+      "epoch": 0.9749670619235836,
+      "grad_norm": 0.1373465359210968,
+      "learning_rate": 1.0889104933106604e-05,
+      "loss": 0.0868,
+      "step": 3885
+    },
+    {
+      "epoch": 0.9762218457870632,
+      "grad_norm": 0.2142958790063858,
+      "learning_rate": 1.0868857569807831e-05,
+      "loss": 0.0888,
+      "step": 3890
+    },
+    {
+      "epoch": 0.9774766296505427,
+      "grad_norm": 0.48555663228034973,
+      "learning_rate": 1.0848606616843673e-05,
+      "loss": 0.0839,
+      "step": 3895
+    },
+    {
+      "epoch": 0.9787314135140223,
+      "grad_norm": 0.2687907814979553,
+      "learning_rate": 1.0828352157880489e-05,
+      "loss": 0.0737,
+      "step": 3900
+    },
+    {
+      "epoch": 0.9799861973775017,
+      "grad_norm": 0.18711687624454498,
+      "learning_rate": 1.0808094276599113e-05,
+      "loss": 0.0958,
+      "step": 3905
+    },
+    {
+      "epoch": 0.9812409812409812,
+      "grad_norm": 0.09380491077899933,
+      "learning_rate": 1.0787833056694526e-05,
+      "loss": 0.0839,
+      "step": 3910
+    },
+    {
+      "epoch": 0.9824957651044608,
+      "grad_norm": 0.2556704878807068,
+      "learning_rate": 1.0767568581875494e-05,
+      "loss": 0.0801,
+      "step": 3915
+    },
+    {
+      "epoch": 0.9837505489679402,
+      "grad_norm": 0.17845787107944489,
+      "learning_rate": 1.0747300935864245e-05,
+      "loss": 0.0906,
+      "step": 3920
+    },
+    {
+      "epoch": 0.9850053328314198,
+      "grad_norm": 0.20369689166545868,
+      "learning_rate": 1.0727030202396091e-05,
+      "loss": 0.0813,
+      "step": 3925
+    },
+    {
+      "epoch": 0.9862601166948993,
+      "grad_norm": 0.3633408546447754,
+      "learning_rate": 1.0706756465219114e-05,
+      "loss": 0.091,
+      "step": 3930
+    },
+    {
+      "epoch": 0.9875149005583789,
+      "grad_norm": 0.14127115905284882,
+      "learning_rate": 1.0686479808093798e-05,
+      "loss": 0.1013,
+      "step": 3935
+    },
+    {
+      "epoch": 0.9887696844218583,
+      "grad_norm": 0.43319714069366455,
+      "learning_rate": 1.0666200314792695e-05,
+      "loss": 0.0854,
+      "step": 3940
+    },
+    {
+      "epoch": 0.9900244682853379,
+      "grad_norm": 0.5234352946281433,
+      "learning_rate": 1.064591806910007e-05,
+      "loss": 0.0799,
+      "step": 3945
+    },
+    {
+      "epoch": 0.9912792521488174,
+      "grad_norm": 0.10436911880970001,
+      "learning_rate": 1.062563315481156e-05,
+      "loss": 0.0774,
+      "step": 3950
+    },
+    {
+      "epoch": 0.9925340360122968,
+      "grad_norm": 0.18167732656002045,
+      "learning_rate": 1.0605345655733839e-05,
+      "loss": 0.0922,
+      "step": 3955
+    },
+    {
+      "epoch": 0.9937888198757764,
+      "grad_norm": 0.08805394172668457,
+      "learning_rate": 1.058505565568424e-05,
+      "loss": 0.0908,
+      "step": 3960
+    },
+    {
+      "epoch": 0.9950436037392559,
+      "grad_norm": 0.186781644821167,
+      "learning_rate": 1.056476323849044e-05,
+      "loss": 0.1014,
+      "step": 3965
+    },
+    {
+      "epoch": 0.9962983876027355,
+      "grad_norm": 0.3317610025405884,
+      "learning_rate": 1.0544468487990105e-05,
+      "loss": 0.0959,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9975531714662149,
+      "grad_norm": 0.159254252910614,
+      "learning_rate": 1.0524171488030537e-05,
+      "loss": 0.0961,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9988079553296945,
+      "grad_norm": 0.5758376717567444,
+      "learning_rate": 1.0503872322468331e-05,
+      "loss": 0.0859,
+      "step": 3980
+    },
+    {
+      "epoch": 1.000062739193174,
+      "grad_norm": 0.1497868299484253,
+      "learning_rate": 1.048357107516903e-05,
+      "loss": 0.0865,
+      "step": 3985
+    },
+    {
+      "epoch": 1.0013175230566536,
+      "grad_norm": 0.3601112365722656,
+      "learning_rate": 1.0463267830006779e-05,
+      "loss": 0.0877,
+      "step": 3990
+    },
+    {
+      "epoch": 1.002572306920133,
+      "grad_norm": 0.06848477572202682,
+      "learning_rate": 1.0442962670863971e-05,
+      "loss": 0.0809,
+      "step": 3995
+    },
+    {
+      "epoch": 1.0038270907836124,
+      "grad_norm": 0.40660738945007324,
+      "learning_rate": 1.0422655681630917e-05,
+      "loss": 0.0925,
+      "step": 4000
+    },
+    {
+      "epoch": 1.005081874647092,
+      "grad_norm": 0.12791000306606293,
+      "learning_rate": 1.040234694620548e-05,
+      "loss": 0.0877,
+      "step": 4005
+    },
+    {
+      "epoch": 1.0063366585105715,
+      "grad_norm": 0.445340096950531,
+      "learning_rate": 1.0382036548492743e-05,
+      "loss": 0.0683,
+      "step": 4010
+    },
+    {
+      "epoch": 1.007591442374051,
+      "grad_norm": 0.29247570037841797,
+      "learning_rate": 1.0361724572404654e-05,
+      "loss": 0.0869,
+      "step": 4015
+    },
+    {
+      "epoch": 1.0088462262375306,
+      "grad_norm": 0.10048985481262207,
+      "learning_rate": 1.034141110185968e-05,
+      "loss": 0.072,
+      "step": 4020
+    },
+    {
+      "epoch": 1.0101010101010102,
+      "grad_norm": 0.19421495497226715,
+      "learning_rate": 1.0321096220782469e-05,
+      "loss": 0.0859,
+      "step": 4025
+    },
+    {
+      "epoch": 1.0113557939644897,
+      "grad_norm": 0.34872177243232727,
+      "learning_rate": 1.0300780013103488e-05,
+      "loss": 0.09,
+      "step": 4030
+    },
+    {
+      "epoch": 1.012610577827969,
+      "grad_norm": 0.3325657844543457,
+      "learning_rate": 1.028046256275869e-05,
+      "loss": 0.097,
+      "step": 4035
+    },
+    {
+      "epoch": 1.0138653616914486,
+      "grad_norm": 0.3407081365585327,
+      "learning_rate": 1.0260143953689165e-05,
+      "loss": 0.0872,
+      "step": 4040
+    },
+    {
+      "epoch": 1.0151201455549281,
+      "grad_norm": 0.18198151886463165,
+      "learning_rate": 1.0239824269840784e-05,
+      "loss": 0.0884,
+      "step": 4045
+    },
+    {
+      "epoch": 1.0163749294184077,
+      "grad_norm": 0.08795715123414993,
+      "learning_rate": 1.0219503595163857e-05,
+      "loss": 0.09,
+      "step": 4050
+    },
+    {
+      "epoch": 1.0176297132818872,
+      "grad_norm": 0.07186929136514664,
+      "learning_rate": 1.0199182013612797e-05,
+      "loss": 0.0783,
+      "step": 4055
+    },
+    {
+      "epoch": 1.0188844971453668,
+      "grad_norm": 0.36318737268447876,
+      "learning_rate": 1.017885960914576e-05,
+      "loss": 0.0852,
+      "step": 4060
+    },
+    {
+      "epoch": 1.0201392810088463,
+      "grad_norm": 0.350274920463562,
+      "learning_rate": 1.0158536465724291e-05,
+      "loss": 0.0723,
+      "step": 4065
+    },
+    {
+      "epoch": 1.0213940648723256,
+      "grad_norm": 0.3536963164806366,
+      "learning_rate": 1.0138212667313003e-05,
+      "loss": 0.0898,
+      "step": 4070
+    },
+    {
+      "epoch": 1.0226488487358052,
+      "grad_norm": 0.37890106439590454,
+      "learning_rate": 1.011788829787921e-05,
+      "loss": 0.0871,
+      "step": 4075
+    },
+    {
+      "epoch": 1.0239036325992847,
+      "grad_norm": 0.2589329183101654,
+      "learning_rate": 1.0097563441392582e-05,
+      "loss": 0.0964,
+      "step": 4080
+    },
+    {
+      "epoch": 1.0251584164627643,
+      "grad_norm": 0.3065876066684723,
+      "learning_rate": 1.0077238181824804e-05,
+      "loss": 0.0933,
+      "step": 4085
+    },
+    {
+      "epoch": 1.0264132003262438,
+      "grad_norm": 0.11984225362539291,
+      "learning_rate": 1.0056912603149229e-05,
+      "loss": 0.0845,
+      "step": 4090
+    },
+    {
+      "epoch": 1.0276679841897234,
+      "grad_norm": 0.19489920139312744,
+      "learning_rate": 1.0036586789340518e-05,
+      "loss": 0.071,
+      "step": 4095
+    },
+    {
+      "epoch": 1.028922768053203,
+      "grad_norm": 0.4552467167377472,
+      "learning_rate": 1.001626082437432e-05,
+      "loss": 0.0869,
+      "step": 4100
+    },
+    {
+      "epoch": 1.0301775519166823,
+      "grad_norm": 0.2607716917991638,
+      "learning_rate": 9.995934792226892e-06,
+      "loss": 0.0903,
+      "step": 4105
+    },
+    {
+      "epoch": 1.0314323357801618,
+      "grad_norm": 0.2067851722240448,
+      "learning_rate": 9.975608776874775e-06,
+      "loss": 0.0818,
+      "step": 4110
+    },
+    {
+      "epoch": 1.0326871196436413,
+      "grad_norm": 0.1449105590581894,
+      "learning_rate": 9.955282862294447e-06,
+      "loss": 0.0903,
+      "step": 4115
+    },
+    {
+      "epoch": 1.033941903507121,
+      "grad_norm": 0.3527059853076935,
+      "learning_rate": 9.93495713246196e-06,
+      "loss": 0.0982,
+      "step": 4120
+    },
+    {
+      "epoch": 1.0351966873706004,
+      "grad_norm": 0.38260769844055176,
+      "learning_rate": 9.91463167135261e-06,
+      "loss": 0.0798,
+      "step": 4125
+    },
+    {
+      "epoch": 1.03645147123408,
+      "grad_norm": 0.19611142575740814,
+      "learning_rate": 9.894306562940576e-06,
+      "loss": 0.0771,
+      "step": 4130
+    },
+    {
+      "epoch": 1.0377062550975595,
+      "grad_norm": 0.09190942347049713,
+      "learning_rate": 9.873981891198585e-06,
+      "loss": 0.0887,
+      "step": 4135
+    },
+    {
+      "epoch": 1.0389610389610389,
+      "grad_norm": 0.3059125244617462,
+      "learning_rate": 9.853657740097558e-06,
+      "loss": 0.0804,
+      "step": 4140
+    },
+    {
+      "epoch": 1.0402158228245184,
+      "grad_norm": 0.2917691171169281,
+      "learning_rate": 9.833334193606266e-06,
+      "loss": 0.0876,
+      "step": 4145
+    },
+    {
+      "epoch": 1.041470606687998,
+      "grad_norm": 0.10409116744995117,
+      "learning_rate": 9.81301133569098e-06,
+      "loss": 0.0951,
+      "step": 4150
+    },
+    {
+      "epoch": 1.0427253905514775,
+      "grad_norm": 0.23820455372333527,
+      "learning_rate": 9.792689250315126e-06,
+      "loss": 0.0873,
+      "step": 4155
+    },
+    {
+      "epoch": 1.043980174414957,
+      "grad_norm": 0.3840187191963196,
+      "learning_rate": 9.772368021438943e-06,
+      "loss": 0.0859,
+      "step": 4160
+    },
+    {
+      "epoch": 1.0452349582784366,
+      "grad_norm": 0.2977546751499176,
+      "learning_rate": 9.752047733019132e-06,
+      "loss": 0.0833,
+      "step": 4165
+    },
+    {
+      "epoch": 1.0464897421419161,
+      "grad_norm": 0.3379094898700714,
+      "learning_rate": 9.731728469008493e-06,
+      "loss": 0.0799,
+      "step": 4170
+    },
+    {
+      "epoch": 1.0477445260053955,
+      "grad_norm": 0.18591825664043427,
+      "learning_rate": 9.711410313355614e-06,
+      "loss": 0.0772,
+      "step": 4175
+    },
+    {
+      "epoch": 1.048999309868875,
+      "grad_norm": 0.1972389817237854,
+      "learning_rate": 9.691093350004492e-06,
+      "loss": 0.0912,
+      "step": 4180
+    },
+    {
+      "epoch": 1.0502540937323546,
+      "grad_norm": 0.2270268350839615,
+      "learning_rate": 9.670777662894205e-06,
+      "loss": 0.0733,
+      "step": 4185
+    },
+    {
+      "epoch": 1.0515088775958341,
+      "grad_norm": 0.3008296489715576,
+      "learning_rate": 9.650463335958551e-06,
+      "loss": 0.0821,
+      "step": 4190
+    },
+    {
+      "epoch": 1.0527636614593137,
+      "grad_norm": 0.11666283011436462,
+      "learning_rate": 9.630150453125711e-06,
+      "loss": 0.0782,
+      "step": 4195
+    },
+    {
+      "epoch": 1.0540184453227932,
+      "grad_norm": 0.1857229471206665,
+      "learning_rate": 9.609839098317902e-06,
+      "loss": 0.0811,
+      "step": 4200
+    },
+    {
+      "epoch": 1.0552732291862728,
+      "grad_norm": 0.24505260586738586,
+      "learning_rate": 9.589529355451028e-06,
+      "loss": 0.0747,
+      "step": 4205
+    },
+    {
+      "epoch": 1.056528013049752,
+      "grad_norm": 0.12762409448623657,
+      "learning_rate": 9.569221308434336e-06,
+      "loss": 0.0813,
+      "step": 4210
+    },
+    {
+      "epoch": 1.0577827969132316,
+      "grad_norm": 0.14525793492794037,
+      "learning_rate": 9.548915041170049e-06,
+      "loss": 0.0814,
+      "step": 4215
+    },
+    {
+      "epoch": 1.0590375807767112,
+      "grad_norm": 0.4550996422767639,
+      "learning_rate": 9.528610637553063e-06,
+      "loss": 0.087,
+      "step": 4220
+    },
+    {
+      "epoch": 1.0602923646401907,
+      "grad_norm": 0.11074529588222504,
+      "learning_rate": 9.508308181470556e-06,
+      "loss": 0.0885,
+      "step": 4225
+    },
+    {
+      "epoch": 1.0615471485036703,
+      "grad_norm": 0.2155328094959259,
+      "learning_rate": 9.488007756801672e-06,
+      "loss": 0.0873,
+      "step": 4230
+    },
+    {
+      "epoch": 1.0628019323671498,
+      "grad_norm": 0.0720556378364563,
+      "learning_rate": 9.467709447417149e-06,
+      "loss": 0.0787,
+      "step": 4235
+    },
+    {
+      "epoch": 1.0640567162306294,
+      "grad_norm": 0.5692139267921448,
+      "learning_rate": 9.447413337178994e-06,
+      "loss": 0.0746,
+      "step": 4240
+    },
+    {
+      "epoch": 1.0653115000941087,
+      "grad_norm": 0.42540258169174194,
+      "learning_rate": 9.42711950994013e-06,
+      "loss": 0.0977,
+      "step": 4245
+    },
+    {
+      "epoch": 1.0665662839575882,
+      "grad_norm": 0.13704167306423187,
+      "learning_rate": 9.406828049544046e-06,
+      "loss": 0.0815,
+      "step": 4250
+    },
+    {
+      "epoch": 1.0678210678210678,
+      "grad_norm": 0.11324842274188995,
+      "learning_rate": 9.386539039824446e-06,
+      "loss": 0.0853,
+      "step": 4255
+    },
+    {
+      "epoch": 1.0690758516845473,
+      "grad_norm": 0.09248417615890503,
+      "learning_rate": 9.366252564604914e-06,
+      "loss": 0.0751,
+      "step": 4260
+    },
+    {
+      "epoch": 1.0703306355480269,
+      "grad_norm": 0.11014354974031448,
+      "learning_rate": 9.34596870769857e-06,
+      "loss": 0.0888,
+      "step": 4265
+    },
+    {
+      "epoch": 1.0715854194115064,
+      "grad_norm": 0.2357379049062729,
+      "learning_rate": 9.325687552907708e-06,
+      "loss": 0.0806,
+      "step": 4270
+    },
+    {
+      "epoch": 1.072840203274986,
+      "grad_norm": 0.23039399087429047,
+      "learning_rate": 9.305409184023455e-06,
+      "loss": 0.0899,
+      "step": 4275
+    },
+    {
+      "epoch": 1.0740949871384653,
+      "grad_norm": 0.3211176097393036,
+      "learning_rate": 9.285133684825435e-06,
+      "loss": 0.0921,
+      "step": 4280
+    },
+    {
+      "epoch": 1.0753497710019448,
+      "grad_norm": 0.08523392677307129,
+      "learning_rate": 9.264861139081417e-06,
+      "loss": 0.0742,
+      "step": 4285
+    },
+    {
+      "epoch": 1.0766045548654244,
+      "grad_norm": 0.446898490190506,
+      "learning_rate": 9.244591630546964e-06,
+      "loss": 0.0888,
+      "step": 4290
+    },
+    {
+      "epoch": 1.077859338728904,
+      "grad_norm": 0.30059829354286194,
+      "learning_rate": 9.224325242965088e-06,
+      "loss": 0.0796,
+      "step": 4295
+    },
+    {
+      "epoch": 1.0791141225923835,
+      "grad_norm": 0.1692143827676773,
+      "learning_rate": 9.204062060065915e-06,
+      "loss": 0.0988,
+      "step": 4300
+    },
+    {
+      "epoch": 1.080368906455863,
+      "grad_norm": 0.5262351632118225,
+      "learning_rate": 9.18380216556632e-06,
+      "loss": 0.0817,
+      "step": 4305
+    },
+    {
+      "epoch": 1.0816236903193426,
+      "grad_norm": 0.2138439118862152,
+      "learning_rate": 9.163545643169607e-06,
+      "loss": 0.0921,
+      "step": 4310
+    },
+    {
+      "epoch": 1.082878474182822,
+      "grad_norm": 0.2504398822784424,
+      "learning_rate": 9.143292576565142e-06,
+      "loss": 0.0747,
+      "step": 4315
+    },
+    {
+      "epoch": 1.0841332580463015,
+      "grad_norm": 0.2210868000984192,
+      "learning_rate": 9.123043049427996e-06,
+      "loss": 0.0843,
+      "step": 4320
+    },
+    {
+      "epoch": 1.085388041909781,
+      "grad_norm": 0.20180197060108185,
+      "learning_rate": 9.102797145418644e-06,
+      "loss": 0.0854,
+      "step": 4325
+    },
+    {
+      "epoch": 1.0866428257732605,
+      "grad_norm": 0.3638245761394501,
+      "learning_rate": 9.082554948182577e-06,
+      "loss": 0.0809,
+      "step": 4330
+    },
+    {
+      "epoch": 1.08789760963674,
+      "grad_norm": 0.28940561413764954,
+      "learning_rate": 9.062316541349978e-06,
+      "loss": 0.0806,
+      "step": 4335
+    },
+    {
+      "epoch": 1.0891523935002196,
+      "grad_norm": 0.15019264817237854,
+      "learning_rate": 9.042082008535361e-06,
+      "loss": 0.081,
+      "step": 4340
+    },
+    {
+      "epoch": 1.0904071773636992,
+      "grad_norm": 0.4596641957759857,
+      "learning_rate": 9.021851433337243e-06,
+      "loss": 0.0886,
+      "step": 4345
+    },
+    {
+      "epoch": 1.0916619612271785,
+      "grad_norm": 0.3484109342098236,
+      "learning_rate": 9.001624899337785e-06,
+      "loss": 0.0997,
+      "step": 4350
+    },
+    {
+      "epoch": 1.092916745090658,
+      "grad_norm": 0.4524548053741455,
+      "learning_rate": 8.981402490102464e-06,
+      "loss": 0.0719,
+      "step": 4355
+    },
+    {
+      "epoch": 1.0941715289541376,
+      "grad_norm": 0.18216001987457275,
+      "learning_rate": 8.961184289179695e-06,
+      "loss": 0.0866,
+      "step": 4360
+    },
+    {
+      "epoch": 1.0954263128176172,
+      "grad_norm": 0.3416767120361328,
+      "learning_rate": 8.94097038010052e-06,
+      "loss": 0.0759,
+      "step": 4365
+    },
+    {
+      "epoch": 1.0966810966810967,
+      "grad_norm": 0.549856960773468,
+      "learning_rate": 8.920760846378248e-06,
+      "loss": 0.0875,
+      "step": 4370
+    },
+    {
+      "epoch": 1.0979358805445762,
+      "grad_norm": 0.12520906329154968,
+      "learning_rate": 8.900555771508114e-06,
+      "loss": 0.086,
+      "step": 4375
+    },
+    {
+      "epoch": 1.0991906644080558,
+      "grad_norm": 0.17632810771465302,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.084,
+      "step": 4380
+    },
+    {
+      "epoch": 1.1004454482715351,
+      "grad_norm": 0.28502127528190613,
+      "learning_rate": 8.860159332212719e-06,
+      "loss": 0.0813,
+      "step": 4385
+    },
+    {
+      "epoch": 1.1017002321350147,
+      "grad_norm": 0.43483448028564453,
+      "learning_rate": 8.83996813468443e-06,
+      "loss": 0.0871,
+      "step": 4390
+    },
+    {
+      "epoch": 1.1029550159984942,
+      "grad_norm": 0.28160524368286133,
+      "learning_rate": 8.81978172980154e-06,
+      "loss": 0.0886,
+      "step": 4395
+    },
+    {
+      "epoch": 1.1042097998619738,
+      "grad_norm": 0.4156465232372284,
+      "learning_rate": 8.799600200963716e-06,
+      "loss": 0.083,
+      "step": 4400
+    },
+    {
+      "epoch": 1.1054645837254533,
+      "grad_norm": 0.24867035448551178,
+      "learning_rate": 8.77942363155049e-06,
+      "loss": 0.0785,
+      "step": 4405
+    },
+    {
+      "epoch": 1.1067193675889329,
+      "grad_norm": 0.3078720271587372,
+      "learning_rate": 8.7592521049209e-06,
+      "loss": 0.0817,
+      "step": 4410
+    },
+    {
+      "epoch": 1.1079741514524124,
+      "grad_norm": 0.12757770717144012,
+      "learning_rate": 8.739085704413161e-06,
+      "loss": 0.0773,
+      "step": 4415
+    },
+    {
+      "epoch": 1.1092289353158917,
+      "grad_norm": 0.2089099884033203,
+      "learning_rate": 8.718924513344288e-06,
+      "loss": 0.0826,
+      "step": 4420
+    },
+    {
+      "epoch": 1.1104837191793713,
+      "grad_norm": 0.28598687052726746,
+      "learning_rate": 8.698768615009789e-06,
+      "loss": 0.0898,
+      "step": 4425
+    },
+    {
+      "epoch": 1.1117385030428508,
+      "grad_norm": 0.16409343481063843,
+      "learning_rate": 8.678618092683307e-06,
+      "loss": 0.0904,
+      "step": 4430
+    },
+    {
+      "epoch": 1.1129932869063304,
+      "grad_norm": 0.17756353318691254,
+      "learning_rate": 8.658473029616264e-06,
+      "loss": 0.077,
+      "step": 4435
+    },
+    {
+      "epoch": 1.11424807076981,
+      "grad_norm": 0.27898117899894714,
+      "learning_rate": 8.638333509037537e-06,
+      "loss": 0.0785,
+      "step": 4440
+    },
+    {
+      "epoch": 1.1155028546332895,
+      "grad_norm": 0.12959200143814087,
+      "learning_rate": 8.61819961415309e-06,
+      "loss": 0.0752,
+      "step": 4445
+    },
+    {
+      "epoch": 1.116757638496769,
+      "grad_norm": 0.0896739810705185,
+      "learning_rate": 8.598071428145663e-06,
+      "loss": 0.0962,
+      "step": 4450
+    },
+    {
+      "epoch": 1.1180124223602483,
+      "grad_norm": 0.37898027896881104,
+      "learning_rate": 8.577949034174395e-06,
+      "loss": 0.0864,
+      "step": 4455
+    },
+    {
+      "epoch": 1.1192672062237279,
+      "grad_norm": 0.19253282248973846,
+      "learning_rate": 8.55783251537451e-06,
+      "loss": 0.0883,
+      "step": 4460
+    },
+    {
+      "epoch": 1.1205219900872074,
+      "grad_norm": 0.35812661051750183,
+      "learning_rate": 8.537721954856942e-06,
+      "loss": 0.0839,
+      "step": 4465
+    },
+    {
+      "epoch": 1.121776773950687,
+      "grad_norm": 0.3062339127063751,
+      "learning_rate": 8.517617435708011e-06,
+      "loss": 0.0808,
+      "step": 4470
+    },
+    {
+      "epoch": 1.1230315578141665,
+      "grad_norm": 0.13663077354431152,
+      "learning_rate": 8.497519040989096e-06,
+      "loss": 0.0961,
+      "step": 4475
+    },
+    {
+      "epoch": 1.124286341677646,
+      "grad_norm": 0.2609081566333771,
+      "learning_rate": 8.477426853736257e-06,
+      "loss": 0.0929,
+      "step": 4480
+    },
+    {
+      "epoch": 1.1255411255411256,
+      "grad_norm": 0.12208747863769531,
+      "learning_rate": 8.457340956959905e-06,
+      "loss": 0.0911,
+      "step": 4485
+    },
+    {
+      "epoch": 1.126795909404605,
+      "grad_norm": 0.19168135523796082,
+      "learning_rate": 8.437261433644472e-06,
+      "loss": 0.0863,
+      "step": 4490
+    },
+    {
+      "epoch": 1.1280506932680845,
+      "grad_norm": 0.20487892627716064,
+      "learning_rate": 8.417188366748051e-06,
+      "loss": 0.0883,
+      "step": 4495
+    },
+    {
+      "epoch": 1.129305477131564,
+      "grad_norm": 0.1829683482646942,
+      "learning_rate": 8.397121839202069e-06,
+      "loss": 0.0905,
+      "step": 4500
+    },
+    {
+      "epoch": 1.1305602609950436,
+      "grad_norm": 0.13941001892089844,
+      "learning_rate": 8.377061933910924e-06,
+      "loss": 0.078,
+      "step": 4505
+    },
+    {
+      "epoch": 1.1318150448585231,
+      "grad_norm": 0.1624182164669037,
+      "learning_rate": 8.357008733751664e-06,
+      "loss": 0.0877,
+      "step": 4510
+    },
+    {
+      "epoch": 1.1330698287220027,
+      "grad_norm": 0.44617176055908203,
+      "learning_rate": 8.33696232157363e-06,
+      "loss": 0.0875,
+      "step": 4515
+    },
+    {
+      "epoch": 1.1343246125854822,
+      "grad_norm": 0.1490097939968109,
+      "learning_rate": 8.316922780198126e-06,
+      "loss": 0.0755,
+      "step": 4520
+    },
+    {
+      "epoch": 1.1355793964489616,
+      "grad_norm": 0.2881050109863281,
+      "learning_rate": 8.296890192418052e-06,
+      "loss": 0.0865,
+      "step": 4525
+    },
+    {
+      "epoch": 1.136834180312441,
+      "grad_norm": 0.19701127707958221,
+      "learning_rate": 8.276864640997602e-06,
+      "loss": 0.0837,
+      "step": 4530
+    },
+    {
+      "epoch": 1.1380889641759206,
+      "grad_norm": 0.4736967086791992,
+      "learning_rate": 8.256846208671882e-06,
+      "loss": 0.0843,
+      "step": 4535
+    },
+    {
+      "epoch": 1.1393437480394002,
+      "grad_norm": 0.22454383969306946,
+      "learning_rate": 8.236834978146597e-06,
+      "loss": 0.0805,
+      "step": 4540
+    },
+    {
+      "epoch": 1.1405985319028797,
+      "grad_norm": 0.1956356167793274,
+      "learning_rate": 8.216831032097689e-06,
+      "loss": 0.0877,
+      "step": 4545
+    },
+    {
+      "epoch": 1.1418533157663593,
+      "grad_norm": 0.3192068338394165,
+      "learning_rate": 8.196834453171008e-06,
+      "loss": 0.0773,
+      "step": 4550
+    },
+    {
+      "epoch": 1.1431080996298388,
+      "grad_norm": 0.21237006783485413,
+      "learning_rate": 8.17684532398197e-06,
+      "loss": 0.0799,
+      "step": 4555
+    },
+    {
+      "epoch": 1.1443628834933182,
+      "grad_norm": 0.3464091718196869,
+      "learning_rate": 8.15686372711521e-06,
+      "loss": 0.0946,
+      "step": 4560
+    },
+    {
+      "epoch": 1.1456176673567977,
+      "grad_norm": 0.44841843843460083,
+      "learning_rate": 8.136889745124241e-06,
+      "loss": 0.0937,
+      "step": 4565
+    },
+    {
+      "epoch": 1.1468724512202773,
+      "grad_norm": 0.12974920868873596,
+      "learning_rate": 8.116923460531117e-06,
+      "loss": 0.0866,
+      "step": 4570
+    },
+    {
+      "epoch": 1.1481272350837568,
+      "grad_norm": 0.2840186655521393,
+      "learning_rate": 8.09696495582609e-06,
+      "loss": 0.0759,
+      "step": 4575
+    },
+    {
+      "epoch": 1.1493820189472364,
+      "grad_norm": 0.2896987497806549,
+      "learning_rate": 8.077014313467274e-06,
+      "loss": 0.0905,
+      "step": 4580
+    },
+    {
+      "epoch": 1.150636802810716,
+      "grad_norm": 0.08831676840782166,
+      "learning_rate": 8.057071615880297e-06,
+      "loss": 0.0855,
+      "step": 4585
+    },
+    {
+      "epoch": 1.1518915866741954,
+      "grad_norm": 0.310893714427948,
+      "learning_rate": 8.037136945457959e-06,
+      "loss": 0.0868,
+      "step": 4590
+    },
+    {
+      "epoch": 1.1531463705376748,
+      "grad_norm": 0.2327936589717865,
+      "learning_rate": 8.017210384559901e-06,
+      "loss": 0.0661,
+      "step": 4595
+    },
+    {
+      "epoch": 1.1544011544011543,
+      "grad_norm": 0.23894554376602173,
+      "learning_rate": 7.997292015512257e-06,
+      "loss": 0.0816,
+      "step": 4600
+    },
+    {
+      "epoch": 1.1556559382646339,
+      "grad_norm": 0.2750934064388275,
+      "learning_rate": 7.977381920607324e-06,
+      "loss": 0.0724,
+      "step": 4605
+    },
+    {
+      "epoch": 1.1569107221281134,
+      "grad_norm": 0.2656705677509308,
+      "learning_rate": 7.957480182103198e-06,
+      "loss": 0.0839,
+      "step": 4610
+    },
+    {
+      "epoch": 1.158165505991593,
+      "grad_norm": 0.08596844226121902,
+      "learning_rate": 7.93758688222347e-06,
+      "loss": 0.0873,
+      "step": 4615
+    },
+    {
+      "epoch": 1.1594202898550725,
+      "grad_norm": 0.2484734207391739,
+      "learning_rate": 7.91770210315685e-06,
+      "loss": 0.084,
+      "step": 4620
+    },
+    {
+      "epoch": 1.160675073718552,
+      "grad_norm": 0.3806692361831665,
+      "learning_rate": 7.897825927056865e-06,
+      "loss": 0.0856,
+      "step": 4625
+    },
+    {
+      "epoch": 1.1619298575820314,
+      "grad_norm": 0.26702991127967834,
+      "learning_rate": 7.877958436041475e-06,
+      "loss": 0.0804,
+      "step": 4630
+    },
+    {
+      "epoch": 1.163184641445511,
+      "grad_norm": 0.19116578996181488,
+      "learning_rate": 7.858099712192774e-06,
+      "loss": 0.0882,
+      "step": 4635
+    },
+    {
+      "epoch": 1.1644394253089905,
+      "grad_norm": 0.08863260596990585,
+      "learning_rate": 7.83824983755663e-06,
+      "loss": 0.0761,
+      "step": 4640
+    },
+    {
+      "epoch": 1.16569420917247,
+      "grad_norm": 0.22460100054740906,
+      "learning_rate": 7.818408894142351e-06,
+      "loss": 0.0905,
+      "step": 4645
+    },
+    {
+      "epoch": 1.1669489930359496,
+      "grad_norm": 0.2022630125284195,
+      "learning_rate": 7.798576963922347e-06,
+      "loss": 0.086,
+      "step": 4650
+    },
+    {
+      "epoch": 1.1682037768994291,
+      "grad_norm": 0.17606490850448608,
+      "learning_rate": 7.778754128831782e-06,
+      "loss": 0.0742,
+      "step": 4655
+    },
+    {
+      "epoch": 1.1694585607629087,
+      "grad_norm": 0.1387161910533905,
+      "learning_rate": 7.75894047076826e-06,
+      "loss": 0.0817,
+      "step": 4660
+    },
+    {
+      "epoch": 1.170713344626388,
+      "grad_norm": 0.09419666230678558,
+      "learning_rate": 7.739136071591455e-06,
+      "loss": 0.0918,
+      "step": 4665
+    },
+    {
+      "epoch": 1.1719681284898675,
+      "grad_norm": 0.30808964371681213,
+      "learning_rate": 7.719341013122795e-06,
+      "loss": 0.0724,
+      "step": 4670
+    },
+    {
+      "epoch": 1.173222912353347,
+      "grad_norm": 0.43276599049568176,
+      "learning_rate": 7.699555377145113e-06,
+      "loss": 0.09,
+      "step": 4675
+    },
+    {
+      "epoch": 1.1744776962168266,
+      "grad_norm": 0.1630372256040573,
+      "learning_rate": 7.679779245402321e-06,
+      "loss": 0.0795,
+      "step": 4680
+    },
+    {
+      "epoch": 1.1757324800803062,
+      "grad_norm": 0.1593324989080429,
+      "learning_rate": 7.660012699599062e-06,
+      "loss": 0.0891,
+      "step": 4685
+    },
+    {
+      "epoch": 1.1769872639437857,
+      "grad_norm": 0.19340068101882935,
+      "learning_rate": 7.640255821400364e-06,
+      "loss": 0.098,
+      "step": 4690
+    },
+    {
+      "epoch": 1.1782420478072653,
+      "grad_norm": 0.10026690363883972,
+      "learning_rate": 7.620508692431327e-06,
+      "loss": 0.0809,
+      "step": 4695
+    },
+    {
+      "epoch": 1.1794968316707446,
+      "grad_norm": 0.24709242582321167,
+      "learning_rate": 7.600771394276767e-06,
+      "loss": 0.09,
+      "step": 4700
+    },
+    {
+      "epoch": 1.1807516155342241,
+      "grad_norm": 0.10687977075576782,
+      "learning_rate": 7.5810440084808855e-06,
+      "loss": 0.0948,
+      "step": 4705
+    },
+    {
+      "epoch": 1.1820063993977037,
+      "grad_norm": 0.5234541893005371,
+      "learning_rate": 7.561326616546932e-06,
+      "loss": 0.0857,
+      "step": 4710
+    },
+    {
+      "epoch": 1.1832611832611832,
+      "grad_norm": 0.5169919729232788,
+      "learning_rate": 7.541619299936859e-06,
+      "loss": 0.0716,
+      "step": 4715
+    },
+    {
+      "epoch": 1.1845159671246628,
+      "grad_norm": 0.13368743658065796,
+      "learning_rate": 7.521922140071003e-06,
+      "loss": 0.082,
+      "step": 4720
+    },
+    {
+      "epoch": 1.1857707509881423,
+      "grad_norm": 0.32512664794921875,
+      "learning_rate": 7.50223521832773e-06,
+      "loss": 0.0923,
+      "step": 4725
+    },
+    {
+      "epoch": 1.1870255348516219,
+      "grad_norm": 0.15283085405826569,
+      "learning_rate": 7.482558616043123e-06,
+      "loss": 0.0913,
+      "step": 4730
+    },
+    {
+      "epoch": 1.1882803187151012,
+      "grad_norm": 0.1149832084774971,
+      "learning_rate": 7.462892414510605e-06,
+      "loss": 0.0795,
+      "step": 4735
+    },
+    {
+      "epoch": 1.1895351025785807,
+      "grad_norm": 0.1783827543258667,
+      "learning_rate": 7.443236694980649e-06,
+      "loss": 0.0922,
+      "step": 4740
+    },
+    {
+      "epoch": 1.1907898864420603,
+      "grad_norm": 0.3042134940624237,
+      "learning_rate": 7.423591538660416e-06,
+      "loss": 0.1,
+      "step": 4745
+    },
+    {
+      "epoch": 1.1920446703055398,
+      "grad_norm": 0.32708263397216797,
+      "learning_rate": 7.4039570267134266e-06,
+      "loss": 0.0874,
+      "step": 4750
+    },
+    {
+      "epoch": 1.1932994541690194,
+      "grad_norm": 0.3249336779117584,
+      "learning_rate": 7.384333240259216e-06,
+      "loss": 0.0855,
+      "step": 4755
+    },
+    {
+      "epoch": 1.194554238032499,
+      "grad_norm": 0.3625956177711487,
+      "learning_rate": 7.364720260373017e-06,
+      "loss": 0.0819,
+      "step": 4760
+    },
+    {
+      "epoch": 1.1958090218959785,
+      "grad_norm": 0.13294678926467896,
+      "learning_rate": 7.345118168085412e-06,
+      "loss": 0.0896,
+      "step": 4765
+    },
+    {
+      "epoch": 1.1970638057594578,
+      "grad_norm": 0.29470571875572205,
+      "learning_rate": 7.325527044382004e-06,
+      "loss": 0.0828,
+      "step": 4770
+    },
+    {
+      "epoch": 1.1983185896229374,
+      "grad_norm": 0.23991379141807556,
+      "learning_rate": 7.3059469702030725e-06,
+      "loss": 0.0769,
+      "step": 4775
+    },
+    {
+      "epoch": 1.199573373486417,
+      "grad_norm": 0.30648887157440186,
+      "learning_rate": 7.286378026443252e-06,
+      "loss": 0.0908,
+      "step": 4780
+    },
+    {
+      "epoch": 1.2008281573498965,
+      "grad_norm": 0.10455156862735748,
+      "learning_rate": 7.2668202939511946e-06,
+      "loss": 0.0842,
+      "step": 4785
+    },
+    {
+      "epoch": 1.202082941213376,
+      "grad_norm": 0.22708240151405334,
+      "learning_rate": 7.2472738535292295e-06,
+      "loss": 0.0885,
+      "step": 4790
+    },
+    {
+      "epoch": 1.2033377250768555,
+      "grad_norm": 0.22052530944347382,
+      "learning_rate": 7.227738785933025e-06,
+      "loss": 0.0749,
+      "step": 4795
+    },
+    {
+      "epoch": 1.204592508940335,
+      "grad_norm": 0.2625311315059662,
+      "learning_rate": 7.208215171871277e-06,
+      "loss": 0.0835,
+      "step": 4800
+    },
+    {
+      "epoch": 1.2058472928038144,
+      "grad_norm": 0.20223468542099,
+      "learning_rate": 7.188703092005353e-06,
+      "loss": 0.0855,
+      "step": 4805
+    },
+    {
+      "epoch": 1.207102076667294,
+      "grad_norm": 0.17523938417434692,
+      "learning_rate": 7.169202626948973e-06,
+      "loss": 0.0833,
+      "step": 4810
+    },
+    {
+      "epoch": 1.2083568605307735,
+      "grad_norm": 0.11394723504781723,
+      "learning_rate": 7.149713857267862e-06,
+      "loss": 0.0769,
+      "step": 4815
+    },
+    {
+      "epoch": 1.209611644394253,
+      "grad_norm": 0.28187689185142517,
+      "learning_rate": 7.130236863479434e-06,
+      "loss": 0.0908,
+      "step": 4820
+    },
+    {
+      "epoch": 1.2108664282577326,
+      "grad_norm": 0.22345809638500214,
+      "learning_rate": 7.110771726052446e-06,
+      "loss": 0.087,
+      "step": 4825
+    },
+    {
+      "epoch": 1.2121212121212122,
+      "grad_norm": 0.19860218465328217,
+      "learning_rate": 7.091318525406671e-06,
+      "loss": 0.078,
+      "step": 4830
+    },
+    {
+      "epoch": 1.2133759959846917,
+      "grad_norm": 0.3908574879169464,
+      "learning_rate": 7.071877341912576e-06,
+      "loss": 0.0926,
+      "step": 4835
+    },
+    {
+      "epoch": 1.2146307798481712,
+      "grad_norm": 0.18266786634922028,
+      "learning_rate": 7.052448255890958e-06,
+      "loss": 0.0858,
+      "step": 4840
+    },
+    {
+      "epoch": 1.2158855637116506,
+      "grad_norm": 0.2580263316631317,
+      "learning_rate": 7.033031347612655e-06,
+      "loss": 0.0777,
+      "step": 4845
+    },
+    {
+      "epoch": 1.2171403475751301,
+      "grad_norm": 0.22712433338165283,
+      "learning_rate": 7.013626697298182e-06,
+      "loss": 0.0771,
+      "step": 4850
+    },
+    {
+      "epoch": 1.2183951314386097,
+      "grad_norm": 0.5155333876609802,
+      "learning_rate": 6.994234385117414e-06,
+      "loss": 0.0997,
+      "step": 4855
+    },
+    {
+      "epoch": 1.2196499153020892,
+      "grad_norm": 0.16782841086387634,
+      "learning_rate": 6.974854491189243e-06,
+      "loss": 0.0891,
+      "step": 4860
+    },
+    {
+      "epoch": 1.2209046991655688,
+      "grad_norm": 0.2045108526945114,
+      "learning_rate": 6.95548709558127e-06,
+      "loss": 0.0846,
+      "step": 4865
+    },
+    {
+      "epoch": 1.2221594830290483,
+      "grad_norm": 0.46996012330055237,
+      "learning_rate": 6.9361322783094465e-06,
+      "loss": 0.0845,
+      "step": 4870
+    },
+    {
+      "epoch": 1.2234142668925279,
+      "grad_norm": 0.3139326870441437,
+      "learning_rate": 6.916790119337766e-06,
+      "loss": 0.0842,
+      "step": 4875
+    },
+    {
+      "epoch": 1.2246690507560072,
+      "grad_norm": 0.27236974239349365,
+      "learning_rate": 6.897460698577918e-06,
+      "loss": 0.0929,
+      "step": 4880
+    },
+    {
+      "epoch": 1.2259238346194867,
+      "grad_norm": 0.3481730818748474,
+      "learning_rate": 6.878144095888964e-06,
+      "loss": 0.0832,
+      "step": 4885
+    },
+    {
+      "epoch": 1.2271786184829663,
+      "grad_norm": 0.11182388663291931,
+      "learning_rate": 6.858840391077017e-06,
+      "loss": 0.0888,
+      "step": 4890
+    },
+    {
+      "epoch": 1.2284334023464458,
+      "grad_norm": 0.5058387517929077,
+      "learning_rate": 6.839549663894897e-06,
+      "loss": 0.0784,
+      "step": 4895
+    },
+    {
+      "epoch": 1.2296881862099254,
+      "grad_norm": 0.3258366584777832,
+      "learning_rate": 6.820271994041796e-06,
+      "loss": 0.0834,
+      "step": 4900
+    },
+    {
+      "epoch": 1.230942970073405,
+      "grad_norm": 0.11910035461187363,
+      "learning_rate": 6.8010074611629815e-06,
+      "loss": 0.0806,
+      "step": 4905
+    },
+    {
+      "epoch": 1.2321977539368845,
+      "grad_norm": 0.10795161128044128,
+      "learning_rate": 6.781756144849431e-06,
+      "loss": 0.0877,
+      "step": 4910
+    },
+    {
+      "epoch": 1.2334525378003638,
+      "grad_norm": 0.07859393954277039,
+      "learning_rate": 6.762518124637525e-06,
+      "loss": 0.0789,
+      "step": 4915
+    },
+    {
+      "epoch": 1.2347073216638433,
+      "grad_norm": 0.36102941632270813,
+      "learning_rate": 6.743293480008703e-06,
+      "loss": 0.0973,
+      "step": 4920
+    },
+    {
+      "epoch": 1.2359621055273229,
+      "grad_norm": 0.07983565330505371,
+      "learning_rate": 6.724082290389151e-06,
+      "loss": 0.0922,
+      "step": 4925
+    },
+    {
+      "epoch": 1.2372168893908024,
+      "grad_norm": 0.18033474683761597,
+      "learning_rate": 6.704884635149467e-06,
+      "loss": 0.083,
+      "step": 4930
+    },
+    {
+      "epoch": 1.238471673254282,
+      "grad_norm": 0.33377745747566223,
+      "learning_rate": 6.685700593604329e-06,
+      "loss": 0.0769,
+      "step": 4935
+    },
+    {
+      "epoch": 1.2397264571177615,
+      "grad_norm": 0.09846628457307816,
+      "learning_rate": 6.666530245012168e-06,
+      "loss": 0.0848,
+      "step": 4940
+    },
+    {
+      "epoch": 1.240981240981241,
+      "grad_norm": 0.28946545720100403,
+      "learning_rate": 6.647373668574841e-06,
+      "loss": 0.0846,
+      "step": 4945
+    },
+    {
+      "epoch": 1.2422360248447206,
+      "grad_norm": 0.2697726786136627,
+      "learning_rate": 6.628230943437319e-06,
+      "loss": 0.0741,
+      "step": 4950
+    },
+    {
+      "epoch": 1.2434908087082,
+      "grad_norm": 0.24468262493610382,
+      "learning_rate": 6.609102148687333e-06,
+      "loss": 0.084,
+      "step": 4955
+    },
+    {
+      "epoch": 1.2447455925716795,
+      "grad_norm": 0.7213346362113953,
+      "learning_rate": 6.589987363355068e-06,
+      "loss": 0.0909,
+      "step": 4960
+    },
+    {
+      "epoch": 1.246000376435159,
+      "grad_norm": 0.1924159675836563,
+      "learning_rate": 6.570886666412823e-06,
+      "loss": 0.0967,
+      "step": 4965
+    },
+    {
+      "epoch": 1.2472551602986386,
+      "grad_norm": 0.2051101177930832,
+      "learning_rate": 6.551800136774697e-06,
+      "loss": 0.0744,
+      "step": 4970
+    },
+    {
+      "epoch": 1.2485099441621181,
+      "grad_norm": 0.2706952393054962,
+      "learning_rate": 6.532727853296257e-06,
+      "loss": 0.0763,
+      "step": 4975
+    },
+    {
+      "epoch": 1.2497647280255977,
+      "grad_norm": 0.21088339388370514,
+      "learning_rate": 6.513669894774209e-06,
+      "loss": 0.0879,
+      "step": 4980
+    },
+    {
+      "epoch": 1.2510195118890772,
+      "grad_norm": 0.339167982339859,
+      "learning_rate": 6.494626339946075e-06,
+      "loss": 0.0867,
+      "step": 4985
+    },
+    {
+      "epoch": 1.2522742957525566,
+      "grad_norm": 0.2325342744588852,
+      "learning_rate": 6.47559726748987e-06,
+      "loss": 0.088,
+      "step": 4990
+    },
+    {
+      "epoch": 1.253529079616036,
+      "grad_norm": 0.24122120440006256,
+      "learning_rate": 6.456582756023781e-06,
+      "loss": 0.082,
+      "step": 4995
+    },
+    {
+      "epoch": 1.2547838634795156,
+      "grad_norm": 0.4205479025840759,
+      "learning_rate": 6.437582884105835e-06,
+      "loss": 0.0825,
+      "step": 5000
+    },
+    {
+      "epoch": 1.2560386473429952,
+      "grad_norm": 0.4158347547054291,
+      "learning_rate": 6.41859773023356e-06,
+      "loss": 0.0845,
+      "step": 5005
+    },
+    {
+      "epoch": 1.2572934312064747,
+      "grad_norm": 0.14699125289916992,
+      "learning_rate": 6.399627372843699e-06,
+      "loss": 0.0871,
+      "step": 5010
+    },
+    {
+      "epoch": 1.258548215069954,
+      "grad_norm": 0.15239457786083221,
+      "learning_rate": 6.380671890311852e-06,
+      "loss": 0.0886,
+      "step": 5015
+    },
+    {
+      "epoch": 1.2598029989334338,
+      "grad_norm": 0.35437726974487305,
+      "learning_rate": 6.361731360952169e-06,
+      "loss": 0.0941,
+      "step": 5020
+    },
+    {
+      "epoch": 1.2610577827969132,
+      "grad_norm": 0.19353091716766357,
+      "learning_rate": 6.342805863017012e-06,
+      "loss": 0.0911,
+      "step": 5025
+    },
+    {
+      "epoch": 1.2623125666603927,
+      "grad_norm": 0.19099637866020203,
+      "learning_rate": 6.323895474696651e-06,
+      "loss": 0.0929,
+      "step": 5030
+    },
+    {
+      "epoch": 1.2635673505238723,
+      "grad_norm": 0.25522175431251526,
+      "learning_rate": 6.305000274118926e-06,
+      "loss": 0.0896,
+      "step": 5035
+    },
+    {
+      "epoch": 1.2648221343873518,
+      "grad_norm": 0.35186290740966797,
+      "learning_rate": 6.286120339348935e-06,
+      "loss": 0.0717,
+      "step": 5040
+    },
+    {
+      "epoch": 1.2660769182508314,
+      "grad_norm": 0.2007788121700287,
+      "learning_rate": 6.267255748388697e-06,
+      "loss": 0.0861,
+      "step": 5045
+    },
+    {
+      "epoch": 1.2673317021143107,
+      "grad_norm": 0.07917312532663345,
+      "learning_rate": 6.248406579176838e-06,
+      "loss": 0.0829,
+      "step": 5050
+    },
+    {
+      "epoch": 1.2685864859777904,
+      "grad_norm": 0.2649421989917755,
+      "learning_rate": 6.229572909588282e-06,
+      "loss": 0.0802,
+      "step": 5055
+    },
+    {
+      "epoch": 1.2698412698412698,
+      "grad_norm": 0.13473555445671082,
+      "learning_rate": 6.2107548174339085e-06,
+      "loss": 0.0731,
+      "step": 5060
+    },
+    {
+      "epoch": 1.2710960537047493,
+      "grad_norm": 0.3079237639904022,
+      "learning_rate": 6.1919523804602335e-06,
+      "loss": 0.0874,
+      "step": 5065
+    },
+    {
+      "epoch": 1.2723508375682289,
+      "grad_norm": 0.1939849853515625,
+      "learning_rate": 6.173165676349103e-06,
+      "loss": 0.0782,
+      "step": 5070
+    },
+    {
+      "epoch": 1.2736056214317084,
+      "grad_norm": 0.3249664902687073,
+      "learning_rate": 6.15439478271736e-06,
+      "loss": 0.078,
+      "step": 5075
+    },
+    {
+      "epoch": 1.274860405295188,
+      "grad_norm": 0.21080322563648224,
+      "learning_rate": 6.135639777116526e-06,
+      "loss": 0.0777,
+      "step": 5080
+    },
+    {
+      "epoch": 1.2761151891586673,
+      "grad_norm": 0.24793751537799835,
+      "learning_rate": 6.116900737032484e-06,
+      "loss": 0.0737,
+      "step": 5085
+    },
+    {
+      "epoch": 1.277369973022147,
+      "grad_norm": 0.2202165126800537,
+      "learning_rate": 6.0981777398851504e-06,
+      "loss": 0.0823,
+      "step": 5090
+    },
+    {
+      "epoch": 1.2786247568856264,
+      "grad_norm": 0.1270621120929718,
+      "learning_rate": 6.079470863028164e-06,
+      "loss": 0.0787,
+      "step": 5095
+    },
+    {
+      "epoch": 1.279879540749106,
+      "grad_norm": 0.1102348044514656,
+      "learning_rate": 6.0607801837485665e-06,
+      "loss": 0.0749,
+      "step": 5100
+    },
+    {
+      "epoch": 1.2811343246125855,
+      "grad_norm": 0.21010182797908783,
+      "learning_rate": 6.042105779266479e-06,
+      "loss": 0.0831,
+      "step": 5105
+    },
+    {
+      "epoch": 1.282389108476065,
+      "grad_norm": 0.6771642565727234,
+      "learning_rate": 6.023447726734771e-06,
+      "loss": 0.0867,
+      "step": 5110
+    },
+    {
+      "epoch": 1.2836438923395446,
+      "grad_norm": 0.22212505340576172,
+      "learning_rate": 6.004806103238771e-06,
+      "loss": 0.0935,
+      "step": 5115
+    },
+    {
+      "epoch": 1.284898676203024,
+      "grad_norm": 0.18700364232063293,
+      "learning_rate": 5.986180985795927e-06,
+      "loss": 0.0867,
+      "step": 5120
+    },
+    {
+      "epoch": 1.2861534600665037,
+      "grad_norm": 0.13250069320201874,
+      "learning_rate": 5.967572451355486e-06,
+      "loss": 0.0897,
+      "step": 5125
+    },
+    {
+      "epoch": 1.287408243929983,
+      "grad_norm": 0.4712745249271393,
+      "learning_rate": 5.9489805767981845e-06,
+      "loss": 0.085,
+      "step": 5130
+    },
+    {
+      "epoch": 1.2886630277934625,
+      "grad_norm": 0.1509675532579422,
+      "learning_rate": 5.9304054389359354e-06,
+      "loss": 0.082,
+      "step": 5135
+    },
+    {
+      "epoch": 1.289917811656942,
+      "grad_norm": 0.37453317642211914,
+      "learning_rate": 5.911847114511497e-06,
+      "loss": 0.0786,
+      "step": 5140
+    },
+    {
+      "epoch": 1.2911725955204216,
+      "grad_norm": 0.18769773840904236,
+      "learning_rate": 5.893305680198175e-06,
+      "loss": 0.0829,
+      "step": 5145
+    },
+    {
+      "epoch": 1.2924273793839012,
+      "grad_norm": 0.4405484199523926,
+      "learning_rate": 5.874781212599475e-06,
+      "loss": 0.0804,
+      "step": 5150
+    },
+    {
+      "epoch": 1.2936821632473805,
+      "grad_norm": 0.20101556181907654,
+      "learning_rate": 5.856273788248819e-06,
+      "loss": 0.0827,
+      "step": 5155
+    },
+    {
+      "epoch": 1.2949369471108603,
+      "grad_norm": 0.2524980902671814,
+      "learning_rate": 5.837783483609214e-06,
+      "loss": 0.0766,
+      "step": 5160
+    },
+    {
+      "epoch": 1.2961917309743396,
+      "grad_norm": 0.2875005304813385,
+      "learning_rate": 5.819310375072935e-06,
+      "loss": 0.0839,
+      "step": 5165
+    },
+    {
+      "epoch": 1.2974465148378191,
+      "grad_norm": 0.3446800410747528,
+      "learning_rate": 5.800854538961213e-06,
+      "loss": 0.0906,
+      "step": 5170
+    },
+    {
+      "epoch": 1.2987012987012987,
+      "grad_norm": 0.15456603467464447,
+      "learning_rate": 5.782416051523909e-06,
+      "loss": 0.0811,
+      "step": 5175
+    },
+    {
+      "epoch": 1.2999560825647782,
+      "grad_norm": 0.3075577914714813,
+      "learning_rate": 5.763994988939223e-06,
+      "loss": 0.0808,
+      "step": 5180
+    },
+    {
+      "epoch": 1.3012108664282578,
+      "grad_norm": 0.4799324572086334,
+      "learning_rate": 5.745591427313365e-06,
+      "loss": 0.0871,
+      "step": 5185
+    },
+    {
+      "epoch": 1.3024656502917373,
+      "grad_norm": 0.1860457956790924,
+      "learning_rate": 5.727205442680218e-06,
+      "loss": 0.0799,
+      "step": 5190
+    },
+    {
+      "epoch": 1.3037204341552169,
+      "grad_norm": 0.2134033590555191,
+      "learning_rate": 5.708837111001069e-06,
+      "loss": 0.0885,
+      "step": 5195
+    },
+    {
+      "epoch": 1.3049752180186962,
+      "grad_norm": 0.40847694873809814,
+      "learning_rate": 5.690486508164268e-06,
+      "loss": 0.084,
+      "step": 5200
+    },
+    {
+      "epoch": 1.3062300018821758,
+      "grad_norm": 0.21664460003376007,
+      "learning_rate": 5.672153709984909e-06,
+      "loss": 0.085,
+      "step": 5205
+    },
+    {
+      "epoch": 1.3074847857456553,
+      "grad_norm": 0.22716949880123138,
+      "learning_rate": 5.653838792204538e-06,
+      "loss": 0.0807,
+      "step": 5210
+    },
+    {
+      "epoch": 1.3087395696091348,
+      "grad_norm": 0.2670564353466034,
+      "learning_rate": 5.6355418304908226e-06,
+      "loss": 0.086,
+      "step": 5215
+    },
+    {
+      "epoch": 1.3099943534726144,
+      "grad_norm": 0.18262840807437897,
+      "learning_rate": 5.617262900437239e-06,
+      "loss": 0.0784,
+      "step": 5220
+    },
+    {
+      "epoch": 1.311249137336094,
+      "grad_norm": 0.41220247745513916,
+      "learning_rate": 5.599002077562779e-06,
+      "loss": 0.0788,
+      "step": 5225
+    },
+    {
+      "epoch": 1.3125039211995735,
+      "grad_norm": 0.41886112093925476,
+      "learning_rate": 5.580759437311624e-06,
+      "loss": 0.0912,
+      "step": 5230
+    },
+    {
+      "epoch": 1.3137587050630528,
+      "grad_norm": 0.30091163516044617,
+      "learning_rate": 5.562535055052818e-06,
+      "loss": 0.084,
+      "step": 5235
+    },
+    {
+      "epoch": 1.3150134889265324,
+      "grad_norm": 0.38127401471138,
+      "learning_rate": 5.544329006079987e-06,
+      "loss": 0.0769,
+      "step": 5240
+    },
+    {
+      "epoch": 1.316268272790012,
+      "grad_norm": 0.13492530584335327,
+      "learning_rate": 5.526141365611018e-06,
+      "loss": 0.0795,
+      "step": 5245
+    },
+    {
+      "epoch": 1.3175230566534915,
+      "grad_norm": 0.27479758858680725,
+      "learning_rate": 5.507972208787728e-06,
+      "loss": 0.0838,
+      "step": 5250
+    },
+    {
+      "epoch": 1.318777840516971,
+      "grad_norm": 0.1254061758518219,
+      "learning_rate": 5.489821610675579e-06,
+      "loss": 0.0773,
+      "step": 5255
+    },
+    {
+      "epoch": 1.3200326243804505,
+      "grad_norm": 0.15454204380512238,
+      "learning_rate": 5.471689646263358e-06,
+      "loss": 0.0793,
+      "step": 5260
+    },
+    {
+      "epoch": 1.32128740824393,
+      "grad_norm": 0.2265045940876007,
+      "learning_rate": 5.453576390462861e-06,
+      "loss": 0.0881,
+      "step": 5265
+    },
+    {
+      "epoch": 1.3225421921074094,
+      "grad_norm": 0.14053259789943695,
+      "learning_rate": 5.435481918108603e-06,
+      "loss": 0.0836,
+      "step": 5270
+    },
+    {
+      "epoch": 1.323796975970889,
+      "grad_norm": 0.15693189203739166,
+      "learning_rate": 5.41740630395748e-06,
+      "loss": 0.0816,
+      "step": 5275
+    },
+    {
+      "epoch": 1.3250517598343685,
+      "grad_norm": 0.15856729447841644,
+      "learning_rate": 5.399349622688479e-06,
+      "loss": 0.0702,
+      "step": 5280
+    },
+    {
+      "epoch": 1.326306543697848,
+      "grad_norm": 0.23238258063793182,
+      "learning_rate": 5.3813119489023766e-06,
+      "loss": 0.0839,
+      "step": 5285
+    },
+    {
+      "epoch": 1.3275613275613276,
+      "grad_norm": 0.08567748963832855,
+      "learning_rate": 5.363293357121422e-06,
+      "loss": 0.0899,
+      "step": 5290
+    },
+    {
+      "epoch": 1.3288161114248072,
+      "grad_norm": 0.0990014597773552,
+      "learning_rate": 5.345293921789e-06,
+      "loss": 0.0812,
+      "step": 5295
+    },
+    {
+      "epoch": 1.3300708952882867,
+      "grad_norm": 0.2629724144935608,
+      "learning_rate": 5.32731371726938e-06,
+      "loss": 0.08,
+      "step": 5300
+    },
+    {
+      "epoch": 1.331325679151766,
+      "grad_norm": 0.22977909445762634,
+      "learning_rate": 5.309352817847374e-06,
+      "loss": 0.0822,
+      "step": 5305
+    },
+    {
+      "epoch": 1.3325804630152456,
+      "grad_norm": 0.1563161015510559,
+      "learning_rate": 5.291411297728027e-06,
+      "loss": 0.0814,
+      "step": 5310
+    },
+    {
+      "epoch": 1.3338352468787251,
+      "grad_norm": 0.0872747004032135,
+      "learning_rate": 5.273489231036321e-06,
+      "loss": 0.0872,
+      "step": 5315
+    },
+    {
+      "epoch": 1.3350900307422047,
+      "grad_norm": 0.24948804080486298,
+      "learning_rate": 5.255586691816874e-06,
+      "loss": 0.0708,
+      "step": 5320
+    },
+    {
+      "epoch": 1.3363448146056842,
+      "grad_norm": 0.17048531770706177,
+      "learning_rate": 5.237703754033616e-06,
+      "loss": 0.0789,
+      "step": 5325
+    },
+    {
+      "epoch": 1.3375995984691638,
+      "grad_norm": 0.33402279019355774,
+      "learning_rate": 5.219840491569503e-06,
+      "loss": 0.0754,
+      "step": 5330
+    },
+    {
+      "epoch": 1.3388543823326433,
+      "grad_norm": 0.25611111521720886,
+      "learning_rate": 5.2019969782262046e-06,
+      "loss": 0.0792,
+      "step": 5335
+    },
+    {
+      "epoch": 1.3401091661961226,
+      "grad_norm": 0.392952561378479,
+      "learning_rate": 5.184173287723782e-06,
+      "loss": 0.0684,
+      "step": 5340
+    },
+    {
+      "epoch": 1.3413639500596022,
+      "grad_norm": 0.11270631104707718,
+      "learning_rate": 5.166369493700412e-06,
+      "loss": 0.0853,
+      "step": 5345
+    },
+    {
+      "epoch": 1.3426187339230817,
+      "grad_norm": 0.3116929531097412,
+      "learning_rate": 5.148585669712074e-06,
+      "loss": 0.0821,
+      "step": 5350
+    },
+    {
+      "epoch": 1.3438735177865613,
+      "grad_norm": 0.0726916715502739,
+      "learning_rate": 5.130821889232228e-06,
+      "loss": 0.092,
+      "step": 5355
+    },
+    {
+      "epoch": 1.3451283016500408,
+      "grad_norm": 0.06830952316522598,
+      "learning_rate": 5.113078225651529e-06,
+      "loss": 0.085,
+      "step": 5360
+    },
+    {
+      "epoch": 1.3463830855135204,
+      "grad_norm": 0.18756920099258423,
+      "learning_rate": 5.095354752277526e-06,
+      "loss": 0.0887,
+      "step": 5365
+    },
+    {
+      "epoch": 1.347637869377,
+      "grad_norm": 0.29784929752349854,
+      "learning_rate": 5.0776515423343445e-06,
+      "loss": 0.0919,
+      "step": 5370
+    },
+    {
+      "epoch": 1.3488926532404792,
+      "grad_norm": 0.24415934085845947,
+      "learning_rate": 5.059968668962401e-06,
+      "loss": 0.0904,
+      "step": 5375
+    },
+    {
+      "epoch": 1.3501474371039588,
+      "grad_norm": 0.10524599254131317,
+      "learning_rate": 5.042306205218082e-06,
+      "loss": 0.0899,
+      "step": 5380
+    },
+    {
+      "epoch": 1.3514022209674383,
+      "grad_norm": 0.13240844011306763,
+      "learning_rate": 5.024664224073454e-06,
+      "loss": 0.0838,
+      "step": 5385
+    },
+    {
+      "epoch": 1.3526570048309179,
+      "grad_norm": 0.17067807912826538,
+      "learning_rate": 5.007042798415969e-06,
+      "loss": 0.0824,
+      "step": 5390
+    },
+    {
+      "epoch": 1.3539117886943974,
+      "grad_norm": 0.1325388103723526,
+      "learning_rate": 4.989442001048151e-06,
+      "loss": 0.0839,
+      "step": 5395
+    },
+    {
+      "epoch": 1.355166572557877,
+      "grad_norm": 0.4756743907928467,
+      "learning_rate": 4.971861904687283e-06,
+      "loss": 0.0775,
+      "step": 5400
+    },
+    {
+      "epoch": 1.3564213564213565,
+      "grad_norm": 0.2643389105796814,
+      "learning_rate": 4.954302581965143e-06,
+      "loss": 0.089,
+      "step": 5405
+    },
+    {
+      "epoch": 1.3576761402848359,
+      "grad_norm": 0.33239737153053284,
+      "learning_rate": 4.93676410542768e-06,
+      "loss": 0.0945,
+      "step": 5410
+    },
+    {
+      "epoch": 1.3589309241483154,
+      "grad_norm": 0.315899133682251,
+      "learning_rate": 4.919246547534709e-06,
+      "loss": 0.0858,
+      "step": 5415
+    },
+    {
+      "epoch": 1.360185708011795,
+      "grad_norm": 0.7093632817268372,
+      "learning_rate": 4.901749980659617e-06,
+      "loss": 0.0778,
+      "step": 5420
+    },
+    {
+      "epoch": 1.3614404918752745,
+      "grad_norm": 0.17570814490318298,
+      "learning_rate": 4.884274477089085e-06,
+      "loss": 0.0856,
+      "step": 5425
+    },
+    {
+      "epoch": 1.362695275738754,
+      "grad_norm": 0.1181860864162445,
+      "learning_rate": 4.866820109022752e-06,
+      "loss": 0.0844,
+      "step": 5430
+    },
+    {
+      "epoch": 1.3639500596022336,
+      "grad_norm": 0.18417999148368835,
+      "learning_rate": 4.84938694857295e-06,
+      "loss": 0.0687,
+      "step": 5435
+    },
+    {
+      "epoch": 1.3652048434657131,
+      "grad_norm": 0.10456906259059906,
+      "learning_rate": 4.831975067764387e-06,
+      "loss": 0.0765,
+      "step": 5440
+    },
+    {
+      "epoch": 1.3664596273291925,
+      "grad_norm": 0.16597416996955872,
+      "learning_rate": 4.814584538533848e-06,
+      "loss": 0.0748,
+      "step": 5445
+    },
+    {
+      "epoch": 1.367714411192672,
+      "grad_norm": 0.08436016738414764,
+      "learning_rate": 4.797215432729913e-06,
+      "loss": 0.0748,
+      "step": 5450
+    },
+    {
+      "epoch": 1.3689691950561516,
+      "grad_norm": 0.2234387844800949,
+      "learning_rate": 4.779867822112658e-06,
+      "loss": 0.0795,
+      "step": 5455
+    },
+    {
+      "epoch": 1.370223978919631,
+      "grad_norm": 0.2501266300678253,
+      "learning_rate": 4.762541778353337e-06,
+      "loss": 0.0785,
+      "step": 5460
+    },
+    {
+      "epoch": 1.3714787627831106,
+      "grad_norm": 0.14667633175849915,
+      "learning_rate": 4.745237373034103e-06,
+      "loss": 0.0836,
+      "step": 5465
+    },
+    {
+      "epoch": 1.3727335466465902,
+      "grad_norm": 0.12082650512456894,
+      "learning_rate": 4.727954677647724e-06,
+      "loss": 0.0827,
+      "step": 5470
+    },
+    {
+      "epoch": 1.3739883305100697,
+      "grad_norm": 0.25682708621025085,
+      "learning_rate": 4.7106937635972565e-06,
+      "loss": 0.0806,
+      "step": 5475
+    },
+    {
+      "epoch": 1.375243114373549,
+      "grad_norm": 0.11401407420635223,
+      "learning_rate": 4.693454702195784e-06,
+      "loss": 0.085,
+      "step": 5480
+    },
+    {
+      "epoch": 1.3764978982370286,
+      "grad_norm": 0.13626615703105927,
+      "learning_rate": 4.676237564666095e-06,
+      "loss": 0.072,
+      "step": 5485
+    },
+    {
+      "epoch": 1.3777526821005082,
+      "grad_norm": 0.274503231048584,
+      "learning_rate": 4.659042422140399e-06,
+      "loss": 0.0869,
+      "step": 5490
+    },
+    {
+      "epoch": 1.3790074659639877,
+      "grad_norm": 0.3412606716156006,
+      "learning_rate": 4.6418693456600424e-06,
+      "loss": 0.0894,
+      "step": 5495
+    },
+    {
+      "epoch": 1.3802622498274673,
+      "grad_norm": 0.17439858615398407,
+      "learning_rate": 4.62471840617521e-06,
+      "loss": 0.0855,
+      "step": 5500
+    },
+    {
+      "epoch": 1.3815170336909468,
+      "grad_norm": 0.2983112037181854,
+      "learning_rate": 4.607589674544603e-06,
+      "loss": 0.0829,
+      "step": 5505
+    },
+    {
+      "epoch": 1.3827718175544264,
+      "grad_norm": 0.32758232951164246,
+      "learning_rate": 4.590483221535198e-06,
+      "loss": 0.0889,
+      "step": 5510
+    },
+    {
+      "epoch": 1.3840266014179057,
+      "grad_norm": 0.34168171882629395,
+      "learning_rate": 4.573399117821922e-06,
+      "loss": 0.0872,
+      "step": 5515
+    },
+    {
+      "epoch": 1.3852813852813852,
+      "grad_norm": 0.21264196932315826,
+      "learning_rate": 4.556337433987359e-06,
+      "loss": 0.0772,
+      "step": 5520
+    },
+    {
+      "epoch": 1.3865361691448648,
+      "grad_norm": 0.11897042393684387,
+      "learning_rate": 4.539298240521463e-06,
+      "loss": 0.0796,
+      "step": 5525
+    },
+    {
+      "epoch": 1.3877909530083443,
+      "grad_norm": 0.20385603606700897,
+      "learning_rate": 4.522281607821288e-06,
+      "loss": 0.0698,
+      "step": 5530
+    },
+    {
+      "epoch": 1.3890457368718239,
+      "grad_norm": 0.3090057671070099,
+      "learning_rate": 4.505287606190658e-06,
+      "loss": 0.0735,
+      "step": 5535
+    },
+    {
+      "epoch": 1.3903005207353034,
+      "grad_norm": 0.21723704040050507,
+      "learning_rate": 4.488316305839911e-06,
+      "loss": 0.0877,
+      "step": 5540
+    },
+    {
+      "epoch": 1.391555304598783,
+      "grad_norm": 0.2535438537597656,
+      "learning_rate": 4.471367776885589e-06,
+      "loss": 0.0802,
+      "step": 5545
+    },
+    {
+      "epoch": 1.3928100884622623,
+      "grad_norm": 0.2798052728176117,
+      "learning_rate": 4.454442089350151e-06,
+      "loss": 0.0828,
+      "step": 5550
+    },
+    {
+      "epoch": 1.3940648723257418,
+      "grad_norm": 0.1968078762292862,
+      "learning_rate": 4.437539313161697e-06,
+      "loss": 0.0878,
+      "step": 5555
+    },
+    {
+      "epoch": 1.3953196561892214,
+      "grad_norm": 0.3160542845726013,
+      "learning_rate": 4.420659518153667e-06,
+      "loss": 0.0843,
+      "step": 5560
+    },
+    {
+      "epoch": 1.396574440052701,
+      "grad_norm": 0.10868985950946808,
+      "learning_rate": 4.403802774064548e-06,
+      "loss": 0.0898,
+      "step": 5565
+    },
+    {
+      "epoch": 1.3978292239161805,
+      "grad_norm": 0.18841342628002167,
+      "learning_rate": 4.386969150537593e-06,
+      "loss": 0.0839,
+      "step": 5570
+    },
+    {
+      "epoch": 1.39908400777966,
+      "grad_norm": 0.20673134922981262,
+      "learning_rate": 4.370158717120544e-06,
+      "loss": 0.0807,
+      "step": 5575
+    },
+    {
+      "epoch": 1.4003387916431396,
+      "grad_norm": 0.2445622682571411,
+      "learning_rate": 4.35337154326532e-06,
+      "loss": 0.0825,
+      "step": 5580
+    },
+    {
+      "epoch": 1.401593575506619,
+      "grad_norm": 0.2141215056180954,
+      "learning_rate": 4.336607698327755e-06,
+      "loss": 0.0852,
+      "step": 5585
+    },
+    {
+      "epoch": 1.4028483593700984,
+      "grad_norm": 0.2787396311759949,
+      "learning_rate": 4.3198672515672925e-06,
+      "loss": 0.0829,
+      "step": 5590
+    },
+    {
+      "epoch": 1.404103143233578,
+      "grad_norm": 0.24689462780952454,
+      "learning_rate": 4.303150272146706e-06,
+      "loss": 0.0864,
+      "step": 5595
+    },
+    {
+      "epoch": 1.4053579270970575,
+      "grad_norm": 0.17422537505626678,
+      "learning_rate": 4.286456829131821e-06,
+      "loss": 0.086,
+      "step": 5600
+    },
+    {
+      "epoch": 1.406612710960537,
+      "grad_norm": 0.20011769235134125,
+      "learning_rate": 4.269786991491222e-06,
+      "loss": 0.0812,
+      "step": 5605
+    },
+    {
+      "epoch": 1.4078674948240166,
+      "grad_norm": 0.2414385825395584,
+      "learning_rate": 4.253140828095964e-06,
+      "loss": 0.0834,
+      "step": 5610
+    },
+    {
+      "epoch": 1.4091222786874962,
+      "grad_norm": 0.2885581851005554,
+      "learning_rate": 4.236518407719289e-06,
+      "loss": 0.0765,
+      "step": 5615
+    },
+    {
+      "epoch": 1.4103770625509755,
+      "grad_norm": 0.20998211205005646,
+      "learning_rate": 4.219919799036359e-06,
+      "loss": 0.0748,
+      "step": 5620
+    },
+    {
+      "epoch": 1.411631846414455,
+      "grad_norm": 0.26928192377090454,
+      "learning_rate": 4.203345070623947e-06,
+      "loss": 0.0923,
+      "step": 5625
+    },
+    {
+      "epoch": 1.4128866302779346,
+      "grad_norm": 0.28317445516586304,
+      "learning_rate": 4.186794290960162e-06,
+      "loss": 0.08,
+      "step": 5630
+    },
+    {
+      "epoch": 1.4141414141414141,
+      "grad_norm": 0.08973658084869385,
+      "learning_rate": 4.170267528424185e-06,
+      "loss": 0.0775,
+      "step": 5635
+    },
+    {
+      "epoch": 1.4153961980048937,
+      "grad_norm": 0.4094654321670532,
+      "learning_rate": 4.153764851295954e-06,
+      "loss": 0.0969,
+      "step": 5640
+    },
+    {
+      "epoch": 1.4166509818683732,
+      "grad_norm": 0.17649400234222412,
+      "learning_rate": 4.137286327755913e-06,
+      "loss": 0.0809,
+      "step": 5645
+    },
+    {
+      "epoch": 1.4179057657318528,
+      "grad_norm": 0.1127919927239418,
+      "learning_rate": 4.120832025884705e-06,
+      "loss": 0.0782,
+      "step": 5650
+    },
+    {
+      "epoch": 1.419160549595332,
+      "grad_norm": 0.5189236998558044,
+      "learning_rate": 4.104402013662901e-06,
+      "loss": 0.0867,
+      "step": 5655
+    },
+    {
+      "epoch": 1.4204153334588117,
+      "grad_norm": 0.4165945053100586,
+      "learning_rate": 4.0879963589707305e-06,
+      "loss": 0.0836,
+      "step": 5660
+    },
+    {
+      "epoch": 1.4216701173222912,
+      "grad_norm": 0.28989869356155396,
+      "learning_rate": 4.071615129587787e-06,
+      "loss": 0.0818,
+      "step": 5665
+    },
+    {
+      "epoch": 1.4229249011857708,
+      "grad_norm": 0.101304791867733,
+      "learning_rate": 4.055258393192746e-06,
+      "loss": 0.0827,
+      "step": 5670
+    },
+    {
+      "epoch": 1.4241796850492503,
+      "grad_norm": 0.2473038285970688,
+      "learning_rate": 4.038926217363089e-06,
+      "loss": 0.0805,
+      "step": 5675
+    },
+    {
+      "epoch": 1.4254344689127298,
+      "grad_norm": 0.2448158711194992,
+      "learning_rate": 4.022618669574839e-06,
+      "loss": 0.0763,
+      "step": 5680
+    },
+    {
+      "epoch": 1.4266892527762094,
+      "grad_norm": 0.5424969792366028,
+      "learning_rate": 4.006335817202256e-06,
+      "loss": 0.0836,
+      "step": 5685
+    },
+    {
+      "epoch": 1.4279440366396887,
+      "grad_norm": 0.2771781086921692,
+      "learning_rate": 3.990077727517573e-06,
+      "loss": 0.0783,
+      "step": 5690
+    },
+    {
+      "epoch": 1.4291988205031683,
+      "grad_norm": 0.05690968781709671,
+      "learning_rate": 3.973844467690727e-06,
+      "loss": 0.0729,
+      "step": 5695
+    },
+    {
+      "epoch": 1.4304536043666478,
+      "grad_norm": 0.26809534430503845,
+      "learning_rate": 3.957636104789056e-06,
+      "loss": 0.0766,
+      "step": 5700
+    },
+    {
+      "epoch": 1.4317083882301274,
+      "grad_norm": 0.31916674971580505,
+      "learning_rate": 3.94145270577705e-06,
+      "loss": 0.0988,
+      "step": 5705
+    },
+    {
+      "epoch": 1.432963172093607,
+      "grad_norm": 0.5278578996658325,
+      "learning_rate": 3.925294337516051e-06,
+      "loss": 0.0853,
+      "step": 5710
+    },
+    {
+      "epoch": 1.4342179559570865,
+      "grad_norm": 0.15941303968429565,
+      "learning_rate": 3.909161066763999e-06,
+      "loss": 0.0872,
+      "step": 5715
+    },
+    {
+      "epoch": 1.435472739820566,
+      "grad_norm": 0.46407434344291687,
+      "learning_rate": 3.893052960175128e-06,
+      "loss": 0.0746,
+      "step": 5720
+    },
+    {
+      "epoch": 1.4367275236840453,
+      "grad_norm": 0.13292035460472107,
+      "learning_rate": 3.876970084299722e-06,
+      "loss": 0.0829,
+      "step": 5725
+    },
+    {
+      "epoch": 1.4379823075475249,
+      "grad_norm": 0.18183240294456482,
+      "learning_rate": 3.860912505583819e-06,
+      "loss": 0.0797,
+      "step": 5730
+    },
+    {
+      "epoch": 1.4392370914110044,
+      "grad_norm": 0.16055604815483093,
+      "learning_rate": 3.844880290368935e-06,
+      "loss": 0.0805,
+      "step": 5735
+    },
+    {
+      "epoch": 1.440491875274484,
+      "grad_norm": 0.28802794218063354,
+      "learning_rate": 3.828873504891813e-06,
+      "loss": 0.0795,
+      "step": 5740
+    },
+    {
+      "epoch": 1.4417466591379635,
+      "grad_norm": 0.11848758161067963,
+      "learning_rate": 3.8128922152841188e-06,
+      "loss": 0.0856,
+      "step": 5745
+    },
+    {
+      "epoch": 1.443001443001443,
+      "grad_norm": 0.131498321890831,
+      "learning_rate": 3.7969364875721914e-06,
+      "loss": 0.084,
+      "step": 5750
+    },
+    {
+      "epoch": 1.4442562268649226,
+      "grad_norm": 0.1645013391971588,
+      "learning_rate": 3.78100638767676e-06,
+      "loss": 0.0635,
+      "step": 5755
+    },
+    {
+      "epoch": 1.445511010728402,
+      "grad_norm": 0.1245434433221817,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.0708,
+      "step": 5760
+    },
+    {
+      "epoch": 1.4467657945918815,
+      "grad_norm": 0.5209866762161255,
+      "learning_rate": 3.7492233344886073e-06,
+      "loss": 0.0801,
+      "step": 5765
+    },
+    {
+      "epoch": 1.448020578455361,
+      "grad_norm": 0.19441930949687958,
+      "learning_rate": 3.7333705125068576e-06,
+      "loss": 0.0711,
+      "step": 5770
+    },
+    {
+      "epoch": 1.4492753623188406,
+      "grad_norm": 0.27351874113082886,
+      "learning_rate": 3.71754358096299e-06,
+      "loss": 0.071,
+      "step": 5775
+    },
+    {
+      "epoch": 1.4505301461823201,
+      "grad_norm": 0.2211633026599884,
+      "learning_rate": 3.7017426052456086e-06,
+      "loss": 0.0878,
+      "step": 5780
+    },
+    {
+      "epoch": 1.4517849300457997,
+      "grad_norm": 0.24535594880580902,
+      "learning_rate": 3.685967650636095e-06,
+      "loss": 0.0805,
+      "step": 5785
+    },
+    {
+      "epoch": 1.4530397139092792,
+      "grad_norm": 0.24481262266635895,
+      "learning_rate": 3.6702187823083147e-06,
+      "loss": 0.0893,
+      "step": 5790
+    },
+    {
+      "epoch": 1.4542944977727585,
+      "grad_norm": 0.16357897222042084,
+      "learning_rate": 3.6544960653283544e-06,
+      "loss": 0.0862,
+      "step": 5795
+    },
+    {
+      "epoch": 1.455549281636238,
+      "grad_norm": 0.34867826104164124,
+      "learning_rate": 3.6387995646542727e-06,
+      "loss": 0.0774,
+      "step": 5800
+    },
+    {
+      "epoch": 1.4568040654997176,
+      "grad_norm": 0.1586691290140152,
+      "learning_rate": 3.6231293451357994e-06,
+      "loss": 0.0822,
+      "step": 5805
+    },
+    {
+      "epoch": 1.4580588493631972,
+      "grad_norm": 0.1658533662557602,
+      "learning_rate": 3.6074854715140983e-06,
+      "loss": 0.0921,
+      "step": 5810
+    },
+    {
+      "epoch": 1.4593136332266767,
+      "grad_norm": 0.26845037937164307,
+      "learning_rate": 3.591868008421472e-06,
+      "loss": 0.075,
+      "step": 5815
+    },
+    {
+      "epoch": 1.4605684170901563,
+      "grad_norm": 0.2408105581998825,
+      "learning_rate": 3.5762770203811225e-06,
+      "loss": 0.0927,
+      "step": 5820
+    },
+    {
+      "epoch": 1.4618232009536358,
+      "grad_norm": 0.15939578413963318,
+      "learning_rate": 3.560712571806858e-06,
+      "loss": 0.0794,
+      "step": 5825
+    },
+    {
+      "epoch": 1.4630779848171152,
+      "grad_norm": 0.07994109392166138,
+      "learning_rate": 3.5451747270028527e-06,
+      "loss": 0.0799,
+      "step": 5830
+    },
+    {
+      "epoch": 1.4643327686805947,
+      "grad_norm": 0.09967489540576935,
+      "learning_rate": 3.5296635501633558e-06,
+      "loss": 0.08,
+      "step": 5835
+    },
+    {
+      "epoch": 1.4655875525440742,
+      "grad_norm": 0.1677519679069519,
+      "learning_rate": 3.5141791053724405e-06,
+      "loss": 0.0809,
+      "step": 5840
+    },
+    {
+      "epoch": 1.4668423364075538,
+      "grad_norm": 0.21794256567955017,
+      "learning_rate": 3.4987214566037477e-06,
+      "loss": 0.0792,
+      "step": 5845
+    },
+    {
+      "epoch": 1.4680971202710333,
+      "grad_norm": 0.2627575993537903,
+      "learning_rate": 3.483290667720196e-06,
+      "loss": 0.0942,
+      "step": 5850
+    },
+    {
+      "epoch": 1.4693519041345129,
+      "grad_norm": 0.15493802726268768,
+      "learning_rate": 3.4678868024737456e-06,
+      "loss": 0.0864,
+      "step": 5855
+    },
+    {
+      "epoch": 1.4706066879979924,
+      "grad_norm": 0.44874611496925354,
+      "learning_rate": 3.452509924505113e-06,
+      "loss": 0.0773,
+      "step": 5860
+    },
+    {
+      "epoch": 1.4718614718614718,
+      "grad_norm": 0.1353214532136917,
+      "learning_rate": 3.437160097343526e-06,
+      "loss": 0.0789,
+      "step": 5865
+    },
+    {
+      "epoch": 1.4731162557249513,
+      "grad_norm": 0.14763125777244568,
+      "learning_rate": 3.4218373844064433e-06,
+      "loss": 0.0824,
+      "step": 5870
+    },
+    {
+      "epoch": 1.4743710395884309,
+      "grad_norm": 0.21625354886054993,
+      "learning_rate": 3.4065418489993118e-06,
+      "loss": 0.0736,
+      "step": 5875
+    },
+    {
+      "epoch": 1.4756258234519104,
+      "grad_norm": 0.24612949788570404,
+      "learning_rate": 3.3912735543152864e-06,
+      "loss": 0.0773,
+      "step": 5880
+    },
+    {
+      "epoch": 1.47688060731539,
+      "grad_norm": 0.1752692461013794,
+      "learning_rate": 3.376032563434979e-06,
+      "loss": 0.0826,
+      "step": 5885
+    },
+    {
+      "epoch": 1.4781353911788695,
+      "grad_norm": 0.2644670605659485,
+      "learning_rate": 3.3608189393262037e-06,
+      "loss": 0.0869,
+      "step": 5890
+    },
+    {
+      "epoch": 1.479390175042349,
+      "grad_norm": 0.20996436476707458,
+      "learning_rate": 3.345632744843702e-06,
+      "loss": 0.0902,
+      "step": 5895
+    },
+    {
+      "epoch": 1.4806449589058284,
+      "grad_norm": 0.2922036647796631,
+      "learning_rate": 3.3304740427288886e-06,
+      "loss": 0.0896,
+      "step": 5900
+    },
+    {
+      "epoch": 1.481899742769308,
+      "grad_norm": 0.37115779519081116,
+      "learning_rate": 3.3153428956096046e-06,
+      "loss": 0.0876,
+      "step": 5905
+    },
+    {
+      "epoch": 1.4831545266327875,
+      "grad_norm": 0.23147353529930115,
+      "learning_rate": 3.3002393659998357e-06,
+      "loss": 0.0778,
+      "step": 5910
+    },
+    {
+      "epoch": 1.484409310496267,
+      "grad_norm": 0.19222880899906158,
+      "learning_rate": 3.2851635162994788e-06,
+      "loss": 0.0884,
+      "step": 5915
+    },
+    {
+      "epoch": 1.4856640943597466,
+      "grad_norm": 0.1700582355260849,
+      "learning_rate": 3.27011540879406e-06,
+      "loss": 0.0831,
+      "step": 5920
+    },
+    {
+      "epoch": 1.486918878223226,
+      "grad_norm": 0.10918935388326645,
+      "learning_rate": 3.2550951056545e-06,
+      "loss": 0.0816,
+      "step": 5925
+    },
+    {
+      "epoch": 1.4881736620867057,
+      "grad_norm": 0.21468310058116913,
+      "learning_rate": 3.2401026689368363e-06,
+      "loss": 0.0807,
+      "step": 5930
+    },
+    {
+      "epoch": 1.489428445950185,
+      "grad_norm": 0.3127315044403076,
+      "learning_rate": 3.2251381605819876e-06,
+      "loss": 0.0958,
+      "step": 5935
+    },
+    {
+      "epoch": 1.4906832298136645,
+      "grad_norm": 0.21210235357284546,
+      "learning_rate": 3.210201642415477e-06,
+      "loss": 0.0748,
+      "step": 5940
+    },
+    {
+      "epoch": 1.491938013677144,
+      "grad_norm": 0.20049948990345,
+      "learning_rate": 3.1952931761471893e-06,
+      "loss": 0.0795,
+      "step": 5945
+    },
+    {
+      "epoch": 1.4931927975406236,
+      "grad_norm": 0.15716488659381866,
+      "learning_rate": 3.180412823371123e-06,
+      "loss": 0.0909,
+      "step": 5950
+    },
+    {
+      "epoch": 1.4944475814041032,
+      "grad_norm": 0.32477137446403503,
+      "learning_rate": 3.1655606455651134e-06,
+      "loss": 0.0792,
+      "step": 5955
+    },
+    {
+      "epoch": 1.4957023652675827,
+      "grad_norm": 0.17082227766513824,
+      "learning_rate": 3.1507367040905943e-06,
+      "loss": 0.0898,
+      "step": 5960
+    },
+    {
+      "epoch": 1.4969571491310623,
+      "grad_norm": 0.1650392711162567,
+      "learning_rate": 3.135941060192348e-06,
+      "loss": 0.0843,
+      "step": 5965
+    },
+    {
+      "epoch": 1.4982119329945416,
+      "grad_norm": 0.10460920631885529,
+      "learning_rate": 3.121173774998245e-06,
+      "loss": 0.0879,
+      "step": 5970
+    },
+    {
+      "epoch": 1.4994667168580211,
+      "grad_norm": 0.13847452402114868,
+      "learning_rate": 3.106434909518985e-06,
+      "loss": 0.0928,
+      "step": 5975
+    },
+    {
+      "epoch": 1.5007215007215007,
+      "grad_norm": 0.1215418130159378,
+      "learning_rate": 3.091724524647861e-06,
+      "loss": 0.0807,
+      "step": 5980
+    },
+    {
+      "epoch": 1.5019762845849802,
+      "grad_norm": 0.26588910818099976,
+      "learning_rate": 3.0770426811604946e-06,
+      "loss": 0.081,
+      "step": 5985
+    },
+    {
+      "epoch": 1.5032310684484598,
+      "grad_norm": 0.4260256290435791,
+      "learning_rate": 3.0623894397145837e-06,
+      "loss": 0.0778,
+      "step": 5990
+    },
+    {
+      "epoch": 1.5044858523119393,
+      "grad_norm": 0.20512565970420837,
+      "learning_rate": 3.0477648608496726e-06,
+      "loss": 0.0746,
+      "step": 5995
+    },
+    {
+      "epoch": 1.5057406361754189,
+      "grad_norm": 0.3236079812049866,
+      "learning_rate": 3.0331690049868733e-06,
+      "loss": 0.0818,
+      "step": 6000
+    },
+    {
+      "epoch": 1.5069954200388982,
+      "grad_norm": 0.5079814791679382,
+      "learning_rate": 3.018601932428632e-06,
+      "loss": 0.0782,
+      "step": 6005
+    },
+    {
+      "epoch": 1.508250203902378,
+      "grad_norm": 0.15258589386940002,
+      "learning_rate": 3.004063703358484e-06,
+      "loss": 0.0885,
+      "step": 6010
+    },
+    {
+      "epoch": 1.5095049877658573,
+      "grad_norm": 0.3337784707546234,
+      "learning_rate": 2.9895543778407875e-06,
+      "loss": 0.0973,
+      "step": 6015
+    },
+    {
+      "epoch": 1.5107597716293368,
+      "grad_norm": 0.1551678329706192,
+      "learning_rate": 2.9750740158205005e-06,
+      "loss": 0.0877,
+      "step": 6020
+    },
+    {
+      "epoch": 1.5120145554928164,
+      "grad_norm": 0.2518864870071411,
+      "learning_rate": 2.960622677122903e-06,
+      "loss": 0.0777,
+      "step": 6025
+    },
+    {
+      "epoch": 1.513269339356296,
+      "grad_norm": 0.22310376167297363,
+      "learning_rate": 2.9462004214533803e-06,
+      "loss": 0.0812,
+      "step": 6030
+    },
+    {
+      "epoch": 1.5145241232197755,
+      "grad_norm": 0.23948289453983307,
+      "learning_rate": 2.9318073083971486e-06,
+      "loss": 0.0879,
+      "step": 6035
+    },
+    {
+      "epoch": 1.5157789070832548,
+      "grad_norm": 0.24216221272945404,
+      "learning_rate": 2.9174433974190365e-06,
+      "loss": 0.0858,
+      "step": 6040
+    },
+    {
+      "epoch": 1.5170336909467346,
+      "grad_norm": 0.19065602123737335,
+      "learning_rate": 2.9031087478632116e-06,
+      "loss": 0.0868,
+      "step": 6045
+    },
+    {
+      "epoch": 1.518288474810214,
+      "grad_norm": 0.2692478597164154,
+      "learning_rate": 2.8888034189529524e-06,
+      "loss": 0.0737,
+      "step": 6050
+    },
+    {
+      "epoch": 1.5195432586736934,
+      "grad_norm": 0.20995523035526276,
+      "learning_rate": 2.874527469790408e-06,
+      "loss": 0.0824,
+      "step": 6055
+    },
+    {
+      "epoch": 1.520798042537173,
+      "grad_norm": 0.10530667006969452,
+      "learning_rate": 2.860280959356336e-06,
+      "loss": 0.0783,
+      "step": 6060
+    },
+    {
+      "epoch": 1.5220528264006525,
+      "grad_norm": 0.39904889464378357,
+      "learning_rate": 2.846063946509868e-06,
+      "loss": 0.0917,
+      "step": 6065
+    },
+    {
+      "epoch": 1.523307610264132,
+      "grad_norm": 0.2294001579284668,
+      "learning_rate": 2.8318764899882745e-06,
+      "loss": 0.0794,
+      "step": 6070
+    },
+    {
+      "epoch": 1.5245623941276114,
+      "grad_norm": 0.31115418672561646,
+      "learning_rate": 2.8177186484067143e-06,
+      "loss": 0.0803,
+      "step": 6075
+    },
+    {
+      "epoch": 1.5258171779910912,
+      "grad_norm": 0.21022425591945648,
+      "learning_rate": 2.803590480257985e-06,
+      "loss": 0.0809,
+      "step": 6080
+    },
+    {
+      "epoch": 1.5270719618545705,
+      "grad_norm": 0.4131692051887512,
+      "learning_rate": 2.7894920439122907e-06,
+      "loss": 0.0776,
+      "step": 6085
+    },
+    {
+      "epoch": 1.52832674571805,
+      "grad_norm": 0.08908002078533173,
+      "learning_rate": 2.77542339761701e-06,
+      "loss": 0.0797,
+      "step": 6090
+    },
+    {
+      "epoch": 1.5295815295815296,
+      "grad_norm": 0.13143645226955414,
+      "learning_rate": 2.7613845994964296e-06,
+      "loss": 0.0838,
+      "step": 6095
+    },
+    {
+      "epoch": 1.5308363134450091,
+      "grad_norm": 0.3454774022102356,
+      "learning_rate": 2.7473757075515305e-06,
+      "loss": 0.0785,
+      "step": 6100
+    },
+    {
+      "epoch": 1.5320910973084887,
+      "grad_norm": 0.2272334098815918,
+      "learning_rate": 2.7333967796597317e-06,
+      "loss": 0.0773,
+      "step": 6105
+    },
+    {
+      "epoch": 1.533345881171968,
+      "grad_norm": 0.2906242311000824,
+      "learning_rate": 2.7194478735746543e-06,
+      "loss": 0.0897,
+      "step": 6110
+    },
+    {
+      "epoch": 1.5346006650354478,
+      "grad_norm": 0.21385426819324493,
+      "learning_rate": 2.70552904692589e-06,
+      "loss": 0.0788,
+      "step": 6115
+    },
+    {
+      "epoch": 1.5358554488989271,
+      "grad_norm": 0.2976323962211609,
+      "learning_rate": 2.691640357218759e-06,
+      "loss": 0.0921,
+      "step": 6120
+    },
+    {
+      "epoch": 1.5371102327624067,
+      "grad_norm": 0.20897260308265686,
+      "learning_rate": 2.6777818618340667e-06,
+      "loss": 0.0831,
+      "step": 6125
+    },
+    {
+      "epoch": 1.5383650166258862,
+      "grad_norm": 0.22841490805149078,
+      "learning_rate": 2.663953618027869e-06,
+      "loss": 0.0829,
+      "step": 6130
+    },
+    {
+      "epoch": 1.5396198004893658,
+      "grad_norm": 0.34173956513404846,
+      "learning_rate": 2.6501556829312492e-06,
+      "loss": 0.0851,
+      "step": 6135
+    },
+    {
+      "epoch": 1.5408745843528453,
+      "grad_norm": 0.10126471519470215,
+      "learning_rate": 2.6363881135500567e-06,
+      "loss": 0.0763,
+      "step": 6140
+    },
+    {
+      "epoch": 1.5421293682163246,
+      "grad_norm": 0.1869829148054123,
+      "learning_rate": 2.6226509667646993e-06,
+      "loss": 0.0882,
+      "step": 6145
+    },
+    {
+      "epoch": 1.5433841520798044,
+      "grad_norm": 0.08933491259813309,
+      "learning_rate": 2.6089442993298854e-06,
+      "loss": 0.0928,
+      "step": 6150
+    },
+    {
+      "epoch": 1.5446389359432837,
+      "grad_norm": 0.2921507656574249,
+      "learning_rate": 2.595268167874396e-06,
+      "loss": 0.0824,
+      "step": 6155
+    },
+    {
+      "epoch": 1.5458937198067633,
+      "grad_norm": 0.27281445264816284,
+      "learning_rate": 2.581622628900868e-06,
+      "loss": 0.0822,
+      "step": 6160
+    },
+    {
+      "epoch": 1.5471485036702428,
+      "grad_norm": 0.20460334420204163,
+      "learning_rate": 2.568007738785533e-06,
+      "loss": 0.0819,
+      "step": 6165
+    },
+    {
+      "epoch": 1.5484032875337224,
+      "grad_norm": 0.5249515175819397,
+      "learning_rate": 2.5544235537779962e-06,
+      "loss": 0.0768,
+      "step": 6170
+    },
+    {
+      "epoch": 1.549658071397202,
+      "grad_norm": 0.2648284137248993,
+      "learning_rate": 2.540870130001015e-06,
+      "loss": 0.0843,
+      "step": 6175
+    },
+    {
+      "epoch": 1.5509128552606812,
+      "grad_norm": 0.47869521379470825,
+      "learning_rate": 2.5273475234502565e-06,
+      "loss": 0.089,
+      "step": 6180
+    },
+    {
+      "epoch": 1.552167639124161,
+      "grad_norm": 0.3596646785736084,
+      "learning_rate": 2.5138557899940595e-06,
+      "loss": 0.0802,
+      "step": 6185
+    },
+    {
+      "epoch": 1.5534224229876403,
+      "grad_norm": 0.13940221071243286,
+      "learning_rate": 2.5003949853732135e-06,
+      "loss": 0.0844,
+      "step": 6190
+    },
+    {
+      "epoch": 1.5546772068511199,
+      "grad_norm": 0.11128222197294235,
+      "learning_rate": 2.486965165200733e-06,
+      "loss": 0.0806,
+      "step": 6195
+    },
+    {
+      "epoch": 1.5559319907145994,
+      "grad_norm": 0.1166953295469284,
+      "learning_rate": 2.4735663849616098e-06,
+      "loss": 0.0896,
+      "step": 6200
+    },
+    {
+      "epoch": 1.557186774578079,
+      "grad_norm": 0.2891371250152588,
+      "learning_rate": 2.460198700012608e-06,
+      "loss": 0.0809,
+      "step": 6205
+    },
+    {
+      "epoch": 1.5584415584415585,
+      "grad_norm": 0.3150325417518616,
+      "learning_rate": 2.4468621655820125e-06,
+      "loss": 0.0852,
+      "step": 6210
+    },
+    {
+      "epoch": 1.5596963423050378,
+      "grad_norm": 0.4029218554496765,
+      "learning_rate": 2.433556836769411e-06,
+      "loss": 0.0751,
+      "step": 6215
+    },
+    {
+      "epoch": 1.5609511261685176,
+      "grad_norm": 0.19041700661182404,
+      "learning_rate": 2.420282768545469e-06,
+      "loss": 0.0852,
+      "step": 6220
+    },
+    {
+      "epoch": 1.562205910031997,
+      "grad_norm": 0.12898743152618408,
+      "learning_rate": 2.4070400157517036e-06,
+      "loss": 0.0909,
+      "step": 6225
+    },
+    {
+      "epoch": 1.5634606938954765,
+      "grad_norm": 0.07892940193414688,
+      "learning_rate": 2.3938286331002458e-06,
+      "loss": 0.0879,
+      "step": 6230
+    },
+    {
+      "epoch": 1.564715477758956,
+      "grad_norm": 0.21748192608356476,
+      "learning_rate": 2.380648675173619e-06,
+      "loss": 0.0877,
+      "step": 6235
+    },
+    {
+      "epoch": 1.5659702616224356,
+      "grad_norm": 0.13437186181545258,
+      "learning_rate": 2.367500196424529e-06,
+      "loss": 0.0812,
+      "step": 6240
+    },
+    {
+      "epoch": 1.5672250454859151,
+      "grad_norm": 0.21154047548770905,
+      "learning_rate": 2.3543832511756113e-06,
+      "loss": 0.0896,
+      "step": 6245
+    },
+    {
+      "epoch": 1.5684798293493944,
+      "grad_norm": 0.16429923474788666,
+      "learning_rate": 2.3412978936192343e-06,
+      "loss": 0.0875,
+      "step": 6250
+    },
+    {
+      "epoch": 1.5697346132128742,
+      "grad_norm": 0.29867643117904663,
+      "learning_rate": 2.328244177817254e-06,
+      "loss": 0.0784,
+      "step": 6255
+    },
+    {
+      "epoch": 1.5709893970763535,
+      "grad_norm": 0.0878303125500679,
+      "learning_rate": 2.315222157700797e-06,
+      "loss": 0.0809,
+      "step": 6260
+    },
+    {
+      "epoch": 1.572244180939833,
+      "grad_norm": 0.24482104182243347,
+      "learning_rate": 2.3022318870700533e-06,
+      "loss": 0.0865,
+      "step": 6265
+    },
+    {
+      "epoch": 1.5734989648033126,
+      "grad_norm": 0.3083607852458954,
+      "learning_rate": 2.289273419594027e-06,
+      "loss": 0.0895,
+      "step": 6270
+    },
+    {
+      "epoch": 1.5747537486667922,
+      "grad_norm": 0.3820037543773651,
+      "learning_rate": 2.2763468088103315e-06,
+      "loss": 0.08,
+      "step": 6275
+    },
+    {
+      "epoch": 1.5760085325302717,
+      "grad_norm": 0.09033305943012238,
+      "learning_rate": 2.263452108124968e-06,
+      "loss": 0.0795,
+      "step": 6280
+    },
+    {
+      "epoch": 1.577263316393751,
+      "grad_norm": 0.1263839453458786,
+      "learning_rate": 2.250589370812105e-06,
+      "loss": 0.0771,
+      "step": 6285
+    },
+    {
+      "epoch": 1.5785181002572308,
+      "grad_norm": 0.45043227076530457,
+      "learning_rate": 2.237758650013847e-06,
+      "loss": 0.075,
+      "step": 6290
+    },
+    {
+      "epoch": 1.5797728841207102,
+      "grad_norm": 0.10394462198019028,
+      "learning_rate": 2.2249599987400237e-06,
+      "loss": 0.0894,
+      "step": 6295
+    },
+    {
+      "epoch": 1.5810276679841897,
+      "grad_norm": 0.394104540348053,
+      "learning_rate": 2.2121934698679793e-06,
+      "loss": 0.0932,
+      "step": 6300
+    },
+    {
+      "epoch": 1.5822824518476692,
+      "grad_norm": 0.12392072379589081,
+      "learning_rate": 2.1994591161423327e-06,
+      "loss": 0.078,
+      "step": 6305
+    },
+    {
+      "epoch": 1.5835372357111488,
+      "grad_norm": 0.3277107775211334,
+      "learning_rate": 2.186756990174783e-06,
+      "loss": 0.0815,
+      "step": 6310
+    },
+    {
+      "epoch": 1.5847920195746283,
+      "grad_norm": 0.35756587982177734,
+      "learning_rate": 2.174087144443875e-06,
+      "loss": 0.0915,
+      "step": 6315
+    },
+    {
+      "epoch": 1.5860468034381077,
+      "grad_norm": 0.3044070303440094,
+      "learning_rate": 2.161449631294785e-06,
+      "loss": 0.0844,
+      "step": 6320
+    },
+    {
+      "epoch": 1.5873015873015874,
+      "grad_norm": 0.1297033727169037,
+      "learning_rate": 2.148844502939117e-06,
+      "loss": 0.0834,
+      "step": 6325
+    },
+    {
+      "epoch": 1.5885563711650668,
+      "grad_norm": 0.1339099109172821,
+      "learning_rate": 2.1362718114546777e-06,
+      "loss": 0.0886,
+      "step": 6330
+    },
+    {
+      "epoch": 1.5898111550285463,
+      "grad_norm": 0.15922382473945618,
+      "learning_rate": 2.1237316087852465e-06,
+      "loss": 0.0965,
+      "step": 6335
+    },
+    {
+      "epoch": 1.5910659388920259,
+      "grad_norm": 0.19360999763011932,
+      "learning_rate": 2.111223946740394e-06,
+      "loss": 0.0896,
+      "step": 6340
+    },
+    {
+      "epoch": 1.5923207227555054,
+      "grad_norm": 0.11068299412727356,
+      "learning_rate": 2.0987488769952436e-06,
+      "loss": 0.0759,
+      "step": 6345
+    },
+    {
+      "epoch": 1.593575506618985,
+      "grad_norm": 0.0986902266740799,
+      "learning_rate": 2.0863064510902586e-06,
+      "loss": 0.0912,
+      "step": 6350
+    },
+    {
+      "epoch": 1.5948302904824643,
+      "grad_norm": 0.23040059208869934,
+      "learning_rate": 2.0738967204310455e-06,
+      "loss": 0.0857,
+      "step": 6355
+    },
+    {
+      "epoch": 1.596085074345944,
+      "grad_norm": 0.1746888905763626,
+      "learning_rate": 2.0615197362881234e-06,
+      "loss": 0.0814,
+      "step": 6360
+    },
+    {
+      "epoch": 1.5973398582094234,
+      "grad_norm": 0.19930033385753632,
+      "learning_rate": 2.0491755497967183e-06,
+      "loss": 0.0848,
+      "step": 6365
+    },
+    {
+      "epoch": 1.598594642072903,
+      "grad_norm": 0.17192873358726501,
+      "learning_rate": 2.0368642119565617e-06,
+      "loss": 0.088,
+      "step": 6370
+    },
+    {
+      "epoch": 1.5998494259363825,
+      "grad_norm": 0.26457393169403076,
+      "learning_rate": 2.024585773631671e-06,
+      "loss": 0.0843,
+      "step": 6375
+    },
+    {
+      "epoch": 1.601104209799862,
+      "grad_norm": 0.15081310272216797,
+      "learning_rate": 2.012340285550126e-06,
+      "loss": 0.0822,
+      "step": 6380
+    },
+    {
+      "epoch": 1.6023589936633416,
+      "grad_norm": 0.1679522842168808,
+      "learning_rate": 2.0001277983038904e-06,
+      "loss": 0.0895,
+      "step": 6385
+    },
+    {
+      "epoch": 1.6036137775268209,
+      "grad_norm": 0.12190816551446915,
+      "learning_rate": 1.9879483623485786e-06,
+      "loss": 0.0764,
+      "step": 6390
+    },
+    {
+      "epoch": 1.6048685613903007,
+      "grad_norm": 0.30287179350852966,
+      "learning_rate": 1.975802028003253e-06,
+      "loss": 0.076,
+      "step": 6395
+    },
+    {
+      "epoch": 1.60612334525378,
+      "grad_norm": 0.3039589524269104,
+      "learning_rate": 1.963688845450218e-06,
+      "loss": 0.0883,
+      "step": 6400
+    },
+    {
+      "epoch": 1.6073781291172595,
+      "grad_norm": 0.10443169623613358,
+      "learning_rate": 1.9516088647348164e-06,
+      "loss": 0.0792,
+      "step": 6405
+    },
+    {
+      "epoch": 1.608632912980739,
+      "grad_norm": 0.10957269370555878,
+      "learning_rate": 1.9395621357652117e-06,
+      "loss": 0.0862,
+      "step": 6410
+    },
+    {
+      "epoch": 1.6098876968442186,
+      "grad_norm": 0.14673763513565063,
+      "learning_rate": 1.9275487083121946e-06,
+      "loss": 0.0808,
+      "step": 6415
+    },
+    {
+      "epoch": 1.6111424807076982,
+      "grad_norm": 0.32992008328437805,
+      "learning_rate": 1.9155686320089684e-06,
+      "loss": 0.0935,
+      "step": 6420
+    },
+    {
+      "epoch": 1.6123972645711775,
+      "grad_norm": 0.17369329929351807,
+      "learning_rate": 1.9036219563509439e-06,
+      "loss": 0.0838,
+      "step": 6425
+    },
+    {
+      "epoch": 1.6136520484346573,
+      "grad_norm": 0.22150751948356628,
+      "learning_rate": 1.891708730695544e-06,
+      "loss": 0.0873,
+      "step": 6430
+    },
+    {
+      "epoch": 1.6149068322981366,
+      "grad_norm": 0.18616770207881927,
+      "learning_rate": 1.8798290042619949e-06,
+      "loss": 0.0934,
+      "step": 6435
+    },
+    {
+      "epoch": 1.6161616161616161,
+      "grad_norm": 0.314390629529953,
+      "learning_rate": 1.8679828261311073e-06,
+      "loss": 0.0769,
+      "step": 6440
+    },
+    {
+      "epoch": 1.6174164000250957,
+      "grad_norm": 0.35372182726860046,
+      "learning_rate": 1.8561702452451047e-06,
+      "loss": 0.0808,
+      "step": 6445
+    },
+    {
+      "epoch": 1.6186711838885752,
+      "grad_norm": 0.16589049994945526,
+      "learning_rate": 1.8443913104073984e-06,
+      "loss": 0.0827,
+      "step": 6450
+    },
+    {
+      "epoch": 1.6199259677520548,
+      "grad_norm": 0.18396617472171783,
+      "learning_rate": 1.83264607028239e-06,
+      "loss": 0.0795,
+      "step": 6455
+    },
+    {
+      "epoch": 1.621180751615534,
+      "grad_norm": 0.0726623684167862,
+      "learning_rate": 1.82093457339527e-06,
+      "loss": 0.0816,
+      "step": 6460
+    },
+    {
+      "epoch": 1.6224355354790139,
+      "grad_norm": 0.37276962399482727,
+      "learning_rate": 1.809256868131828e-06,
+      "loss": 0.0881,
+      "step": 6465
+    },
+    {
+      "epoch": 1.6236903193424932,
+      "grad_norm": 0.2739810645580292,
+      "learning_rate": 1.7976130027382332e-06,
+      "loss": 0.0772,
+      "step": 6470
+    },
+    {
+      "epoch": 1.6249451032059727,
+      "grad_norm": 0.18522068858146667,
+      "learning_rate": 1.786003025320856e-06,
+      "loss": 0.0858,
+      "step": 6475
+    },
+    {
+      "epoch": 1.6261998870694523,
+      "grad_norm": 0.19277586042881012,
+      "learning_rate": 1.774426983846058e-06,
+      "loss": 0.082,
+      "step": 6480
+    },
+    {
+      "epoch": 1.6274546709329318,
+      "grad_norm": 0.36260509490966797,
+      "learning_rate": 1.7628849261399839e-06,
+      "loss": 0.0751,
+      "step": 6485
+    },
+    {
+      "epoch": 1.6287094547964114,
+      "grad_norm": 0.2178385704755783,
+      "learning_rate": 1.7513768998883896e-06,
+      "loss": 0.083,
+      "step": 6490
+    },
+    {
+      "epoch": 1.6299642386598907,
+      "grad_norm": 0.1693069338798523,
+      "learning_rate": 1.7399029526364254e-06,
+      "loss": 0.0827,
+      "step": 6495
+    },
+    {
+      "epoch": 1.6312190225233705,
+      "grad_norm": 0.2321835607290268,
+      "learning_rate": 1.7284631317884448e-06,
+      "loss": 0.0859,
+      "step": 6500
+    },
+    {
+      "epoch": 1.6324738063868498,
+      "grad_norm": 0.21040475368499756,
+      "learning_rate": 1.7170574846078037e-06,
+      "loss": 0.0829,
+      "step": 6505
+    },
+    {
+      "epoch": 1.6337285902503293,
+      "grad_norm": 0.1375027894973755,
+      "learning_rate": 1.7056860582166823e-06,
+      "loss": 0.0771,
+      "step": 6510
+    },
+    {
+      "epoch": 1.634983374113809,
+      "grad_norm": 0.07659853994846344,
+      "learning_rate": 1.6943488995958647e-06,
+      "loss": 0.0859,
+      "step": 6515
+    },
+    {
+      "epoch": 1.6362381579772884,
+      "grad_norm": 0.13758735358715057,
+      "learning_rate": 1.6830460555845719e-06,
+      "loss": 0.0846,
+      "step": 6520
+    },
+    {
+      "epoch": 1.637492941840768,
+      "grad_norm": 0.2629936933517456,
+      "learning_rate": 1.6717775728802432e-06,
+      "loss": 0.083,
+      "step": 6525
+    },
+    {
+      "epoch": 1.6387477257042473,
+      "grad_norm": 0.184475377202034,
+      "learning_rate": 1.6605434980383594e-06,
+      "loss": 0.0836,
+      "step": 6530
+    },
+    {
+      "epoch": 1.640002509567727,
+      "grad_norm": 0.4641217887401581,
+      "learning_rate": 1.649343877472248e-06,
+      "loss": 0.0808,
+      "step": 6535
+    },
+    {
+      "epoch": 1.6412572934312064,
+      "grad_norm": 0.23128898441791534,
+      "learning_rate": 1.638178757452894e-06,
+      "loss": 0.0745,
+      "step": 6540
+    },
+    {
+      "epoch": 1.642512077294686,
+      "grad_norm": 0.09639477729797363,
+      "learning_rate": 1.627048184108726e-06,
+      "loss": 0.0849,
+      "step": 6545
+    },
+    {
+      "epoch": 1.6437668611581655,
+      "grad_norm": 0.305663138628006,
+      "learning_rate": 1.6159522034254628e-06,
+      "loss": 0.0756,
+      "step": 6550
+    },
+    {
+      "epoch": 1.645021645021645,
+      "grad_norm": 0.14679822325706482,
+      "learning_rate": 1.604890861245898e-06,
+      "loss": 0.0698,
+      "step": 6555
+    },
+    {
+      "epoch": 1.6462764288851246,
+      "grad_norm": 0.16359837353229523,
+      "learning_rate": 1.593864203269716e-06,
+      "loss": 0.079,
+      "step": 6560
+    },
+    {
+      "epoch": 1.647531212748604,
+      "grad_norm": 0.12374599277973175,
+      "learning_rate": 1.582872275053301e-06,
+      "loss": 0.0822,
+      "step": 6565
+    },
+    {
+      "epoch": 1.6487859966120837,
+      "grad_norm": 0.2265704870223999,
+      "learning_rate": 1.5719151220095596e-06,
+      "loss": 0.0856,
+      "step": 6570
+    },
+    {
+      "epoch": 1.650040780475563,
+      "grad_norm": 0.1844184547662735,
+      "learning_rate": 1.5609927894077193e-06,
+      "loss": 0.072,
+      "step": 6575
+    },
+    {
+      "epoch": 1.6512955643390426,
+      "grad_norm": 0.0777512937784195,
+      "learning_rate": 1.5501053223731532e-06,
+      "loss": 0.0754,
+      "step": 6580
+    },
+    {
+      "epoch": 1.6525503482025221,
+      "grad_norm": 0.3023741841316223,
+      "learning_rate": 1.5392527658871813e-06,
+      "loss": 0.0788,
+      "step": 6585
+    },
+    {
+      "epoch": 1.6538051320660017,
+      "grad_norm": 0.42789706587791443,
+      "learning_rate": 1.5284351647868956e-06,
+      "loss": 0.0671,
+      "step": 6590
+    },
+    {
+      "epoch": 1.6550599159294812,
+      "grad_norm": 0.1094786524772644,
+      "learning_rate": 1.5176525637649708e-06,
+      "loss": 0.0815,
+      "step": 6595
+    },
+    {
+      "epoch": 1.6563146997929605,
+      "grad_norm": 0.2907751798629761,
+      "learning_rate": 1.5069050073694813e-06,
+      "loss": 0.0796,
+      "step": 6600
+    },
+    {
+      "epoch": 1.6575694836564403,
+      "grad_norm": 0.11227094382047653,
+      "learning_rate": 1.4961925400037102e-06,
+      "loss": 0.0883,
+      "step": 6605
+    },
+    {
+      "epoch": 1.6588242675199196,
+      "grad_norm": 0.34500670433044434,
+      "learning_rate": 1.4855152059259737e-06,
+      "loss": 0.0809,
+      "step": 6610
+    },
+    {
+      "epoch": 1.6600790513833992,
+      "grad_norm": 0.1734425127506256,
+      "learning_rate": 1.474873049249439e-06,
+      "loss": 0.0819,
+      "step": 6615
+    },
+    {
+      "epoch": 1.6613338352468787,
+      "grad_norm": 0.16895422339439392,
+      "learning_rate": 1.4642661139419302e-06,
+      "loss": 0.0874,
+      "step": 6620
+    },
+    {
+      "epoch": 1.6625886191103583,
+      "grad_norm": 0.3084584176540375,
+      "learning_rate": 1.453694443825766e-06,
+      "loss": 0.0811,
+      "step": 6625
+    },
+    {
+      "epoch": 1.6638434029738378,
+      "grad_norm": 0.09408409893512726,
+      "learning_rate": 1.4431580825775604e-06,
+      "loss": 0.0826,
+      "step": 6630
+    },
+    {
+      "epoch": 1.6650981868373171,
+      "grad_norm": 0.06825320422649384,
+      "learning_rate": 1.4326570737280488e-06,
+      "loss": 0.0747,
+      "step": 6635
+    },
+    {
+      "epoch": 1.666352970700797,
+      "grad_norm": 0.30554118752479553,
+      "learning_rate": 1.4221914606619135e-06,
+      "loss": 0.0773,
+      "step": 6640
+    },
+    {
+      "epoch": 1.6676077545642762,
+      "grad_norm": 0.3845992982387543,
+      "learning_rate": 1.4117612866176022e-06,
+      "loss": 0.0893,
+      "step": 6645
+    },
+    {
+      "epoch": 1.6688625384277558,
+      "grad_norm": 0.08592917770147324,
+      "learning_rate": 1.4013665946871347e-06,
+      "loss": 0.0817,
+      "step": 6650
+    },
+    {
+      "epoch": 1.6701173222912353,
+      "grad_norm": 0.12979847192764282,
+      "learning_rate": 1.391007427815949e-06,
+      "loss": 0.0754,
+      "step": 6655
+    },
+    {
+      "epoch": 1.6713721061547149,
+      "grad_norm": 0.2696765661239624,
+      "learning_rate": 1.3806838288027113e-06,
+      "loss": 0.0782,
+      "step": 6660
+    },
+    {
+      "epoch": 1.6726268900181944,
+      "grad_norm": 0.20548641681671143,
+      "learning_rate": 1.3703958402991345e-06,
+      "loss": 0.0815,
+      "step": 6665
+    },
+    {
+      "epoch": 1.6738816738816737,
+      "grad_norm": 0.4033859968185425,
+      "learning_rate": 1.36014350480981e-06,
+      "loss": 0.0826,
+      "step": 6670
+    },
+    {
+      "epoch": 1.6751364577451535,
+      "grad_norm": 0.29467886686325073,
+      "learning_rate": 1.3499268646920317e-06,
+      "loss": 0.0746,
+      "step": 6675
+    },
+    {
+      "epoch": 1.6763912416086328,
+      "grad_norm": 0.17417952418327332,
+      "learning_rate": 1.339745962155613e-06,
+      "loss": 0.0739,
+      "step": 6680
+    },
+    {
+      "epoch": 1.6776460254721124,
+      "grad_norm": 0.30542805790901184,
+      "learning_rate": 1.329600839262728e-06,
+      "loss": 0.0873,
+      "step": 6685
+    },
+    {
+      "epoch": 1.678900809335592,
+      "grad_norm": 0.48298323154449463,
+      "learning_rate": 1.3194915379277195e-06,
+      "loss": 0.0866,
+      "step": 6690
+    },
+    {
+      "epoch": 1.6801555931990715,
+      "grad_norm": 0.2085920125246048,
+      "learning_rate": 1.3094180999169348e-06,
+      "loss": 0.0801,
+      "step": 6695
+    },
+    {
+      "epoch": 1.681410377062551,
+      "grad_norm": 0.22730374336242676,
+      "learning_rate": 1.299380566848557e-06,
+      "loss": 0.0725,
+      "step": 6700
+    },
+    {
+      "epoch": 1.6826651609260304,
+      "grad_norm": 0.2080390900373459,
+      "learning_rate": 1.2893789801924328e-06,
+      "loss": 0.0912,
+      "step": 6705
+    },
+    {
+      "epoch": 1.6839199447895101,
+      "grad_norm": 0.2778363525867462,
+      "learning_rate": 1.2794133812698794e-06,
+      "loss": 0.0804,
+      "step": 6710
+    },
+    {
+      "epoch": 1.6851747286529895,
+      "grad_norm": 0.11341708898544312,
+      "learning_rate": 1.269483811253549e-06,
+      "loss": 0.0884,
+      "step": 6715
+    },
+    {
+      "epoch": 1.686429512516469,
+      "grad_norm": 0.11615065485239029,
+      "learning_rate": 1.259590311167238e-06,
+      "loss": 0.0817,
+      "step": 6720
+    },
+    {
+      "epoch": 1.6876842963799485,
+      "grad_norm": 0.13391433656215668,
+      "learning_rate": 1.2497329218857135e-06,
+      "loss": 0.0831,
+      "step": 6725
+    },
+    {
+      "epoch": 1.688939080243428,
+      "grad_norm": 0.36114147305488586,
+      "learning_rate": 1.2399116841345605e-06,
+      "loss": 0.0841,
+      "step": 6730
+    },
+    {
+      "epoch": 1.6901938641069076,
+      "grad_norm": 0.13782304525375366,
+      "learning_rate": 1.230126638489998e-06,
+      "loss": 0.0861,
+      "step": 6735
+    },
+    {
+      "epoch": 1.691448647970387,
+      "grad_norm": 0.7375853061676025,
+      "learning_rate": 1.2203778253787191e-06,
+      "loss": 0.0926,
+      "step": 6740
+    },
+    {
+      "epoch": 1.6927034318338667,
+      "grad_norm": 0.15916559100151062,
+      "learning_rate": 1.2106652850777257e-06,
+      "loss": 0.0725,
+      "step": 6745
+    },
+    {
+      "epoch": 1.693958215697346,
+      "grad_norm": 0.38723546266555786,
+      "learning_rate": 1.2009890577141625e-06,
+      "loss": 0.082,
+      "step": 6750
+    },
+    {
+      "epoch": 1.6952129995608256,
+      "grad_norm": 0.2894856035709381,
+      "learning_rate": 1.1913491832651359e-06,
+      "loss": 0.0908,
+      "step": 6755
+    },
+    {
+      "epoch": 1.6964677834243052,
+      "grad_norm": 0.10918894410133362,
+      "learning_rate": 1.181745701557574e-06,
+      "loss": 0.079,
+      "step": 6760
+    },
+    {
+      "epoch": 1.6977225672877847,
+      "grad_norm": 0.27660104632377625,
+      "learning_rate": 1.1721786522680445e-06,
+      "loss": 0.0791,
+      "step": 6765
+    },
+    {
+      "epoch": 1.6989773511512642,
+      "grad_norm": 0.13653279840946198,
+      "learning_rate": 1.1626480749225932e-06,
+      "loss": 0.0804,
+      "step": 6770
+    },
+    {
+      "epoch": 1.7002321350147436,
+      "grad_norm": 0.29900994896888733,
+      "learning_rate": 1.1531540088965842e-06,
+      "loss": 0.0865,
+      "step": 6775
+    },
+    {
+      "epoch": 1.7014869188782233,
+      "grad_norm": 0.10758146643638611,
+      "learning_rate": 1.143696493414539e-06,
+      "loss": 0.0898,
+      "step": 6780
+    },
+    {
+      "epoch": 1.7027417027417027,
+      "grad_norm": 0.40997231006622314,
+      "learning_rate": 1.134275567549965e-06,
+      "loss": 0.0932,
+      "step": 6785
+    },
+    {
+      "epoch": 1.7039964866051822,
+      "grad_norm": 0.1141999289393425,
+      "learning_rate": 1.124891270225208e-06,
+      "loss": 0.0903,
+      "step": 6790
+    },
+    {
+      "epoch": 1.7052512704686618,
+      "grad_norm": 0.6266759037971497,
+      "learning_rate": 1.1155436402112785e-06,
+      "loss": 0.0825,
+      "step": 6795
+    },
+    {
+      "epoch": 1.7065060543321413,
+      "grad_norm": 0.16094879806041718,
+      "learning_rate": 1.1062327161276965e-06,
+      "loss": 0.0833,
+      "step": 6800
+    },
+    {
+      "epoch": 1.7077608381956209,
+      "grad_norm": 0.2038365751504898,
+      "learning_rate": 1.0969585364423352e-06,
+      "loss": 0.091,
+      "step": 6805
+    },
+    {
+      "epoch": 1.7090156220591002,
+      "grad_norm": 0.1327473223209381,
+      "learning_rate": 1.0877211394712617e-06,
+      "loss": 0.0799,
+      "step": 6810
+    },
+    {
+      "epoch": 1.71027040592258,
+      "grad_norm": 0.3634199798107147,
+      "learning_rate": 1.0785205633785666e-06,
+      "loss": 0.0812,
+      "step": 6815
+    },
+    {
+      "epoch": 1.7115251897860593,
+      "grad_norm": 0.27833735942840576,
+      "learning_rate": 1.0693568461762238e-06,
+      "loss": 0.0825,
+      "step": 6820
+    },
+    {
+      "epoch": 1.7127799736495388,
+      "grad_norm": 0.3433382511138916,
+      "learning_rate": 1.0602300257239262e-06,
+      "loss": 0.0731,
+      "step": 6825
+    },
+    {
+      "epoch": 1.7140347575130184,
+      "grad_norm": 0.5607882738113403,
+      "learning_rate": 1.0511401397289233e-06,
+      "loss": 0.0769,
+      "step": 6830
+    },
+    {
+      "epoch": 1.715289541376498,
+      "grad_norm": 0.3127841353416443,
+      "learning_rate": 1.0420872257458725e-06,
+      "loss": 0.0901,
+      "step": 6835
+    },
+    {
+      "epoch": 1.7165443252399775,
+      "grad_norm": 0.10875175893306732,
+      "learning_rate": 1.0330713211766864e-06,
+      "loss": 0.0897,
+      "step": 6840
+    },
+    {
+      "epoch": 1.7177991091034568,
+      "grad_norm": 0.5205060839653015,
+      "learning_rate": 1.0240924632703676e-06,
+      "loss": 0.0774,
+      "step": 6845
+    },
+    {
+      "epoch": 1.7190538929669366,
+      "grad_norm": 0.15961715579032898,
+      "learning_rate": 1.0151506891228636e-06,
+      "loss": 0.0765,
+      "step": 6850
+    },
+    {
+      "epoch": 1.7203086768304159,
+      "grad_norm": 0.6552174687385559,
+      "learning_rate": 1.0062460356769189e-06,
+      "loss": 0.0854,
+      "step": 6855
+    },
+    {
+      "epoch": 1.7215634606938954,
+      "grad_norm": 0.10555896908044815,
+      "learning_rate": 9.973785397218982e-07,
+      "loss": 0.0886,
+      "step": 6860
+    },
+    {
+      "epoch": 1.722818244557375,
+      "grad_norm": 0.21054576337337494,
+      "learning_rate": 9.88548237893664e-07,
+      "loss": 0.0841,
+      "step": 6865
+    },
+    {
+      "epoch": 1.7240730284208545,
+      "grad_norm": 0.10787925869226456,
+      "learning_rate": 9.79755166674411e-07,
+      "loss": 0.076,
+      "step": 6870
+    },
+    {
+      "epoch": 1.725327812284334,
+      "grad_norm": 0.24701017141342163,
+      "learning_rate": 9.709993623925118e-07,
+      "loss": 0.0804,
+      "step": 6875
+    },
+    {
+      "epoch": 1.7265825961478134,
+      "grad_norm": 0.2187194675207138,
+      "learning_rate": 9.622808612223722e-07,
+      "loss": 0.0794,
+      "step": 6880
+    },
+    {
+      "epoch": 1.7278373800112932,
+      "grad_norm": 0.16389504075050354,
+      "learning_rate": 9.535996991842855e-07,
+      "loss": 0.09,
+      "step": 6885
+    },
+    {
+      "epoch": 1.7290921638747725,
+      "grad_norm": 0.33097338676452637,
+      "learning_rate": 9.449559121442731e-07,
+      "loss": 0.0903,
+      "step": 6890
+    },
+    {
+      "epoch": 1.730346947738252,
+      "grad_norm": 0.1447478085756302,
+      "learning_rate": 9.363495358139485e-07,
+      "loss": 0.0778,
+      "step": 6895
+    },
+    {
+      "epoch": 1.7316017316017316,
+      "grad_norm": 0.24518853425979614,
+      "learning_rate": 9.277806057503592e-07,
+      "loss": 0.0972,
+      "step": 6900
+    },
+    {
+      "epoch": 1.7328565154652111,
+      "grad_norm": 0.09704507142305374,
+      "learning_rate": 9.192491573558438e-07,
+      "loss": 0.0884,
+      "step": 6905
+    },
+    {
+      "epoch": 1.7341112993286907,
+      "grad_norm": 0.1697137951850891,
+      "learning_rate": 9.107552258778907e-07,
+      "loss": 0.0799,
+      "step": 6910
+    },
+    {
+      "epoch": 1.73536608319217,
+      "grad_norm": 0.11926918476819992,
+      "learning_rate": 9.022988464089888e-07,
+      "loss": 0.0726,
+      "step": 6915
+    },
+    {
+      "epoch": 1.7366208670556498,
+      "grad_norm": 0.16987641155719757,
+      "learning_rate": 8.9388005388647e-07,
+      "loss": 0.0835,
+      "step": 6920
+    },
+    {
+      "epoch": 1.737875650919129,
+      "grad_norm": 0.24412445724010468,
+      "learning_rate": 8.854988830923905e-07,
+      "loss": 0.0727,
+      "step": 6925
+    },
+    {
+      "epoch": 1.7391304347826086,
+      "grad_norm": 0.33428534865379333,
+      "learning_rate": 8.771553686533684e-07,
+      "loss": 0.0844,
+      "step": 6930
+    },
+    {
+      "epoch": 1.7403852186460882,
+      "grad_norm": 0.4698134958744049,
+      "learning_rate": 8.688495450404444e-07,
+      "loss": 0.0815,
+      "step": 6935
+    },
+    {
+      "epoch": 1.7416400025095677,
+      "grad_norm": 0.3183957040309906,
+      "learning_rate": 8.605814465689366e-07,
+      "loss": 0.0779,
+      "step": 6940
+    },
+    {
+      "epoch": 1.7428947863730473,
+      "grad_norm": 0.1109415665268898,
+      "learning_rate": 8.523511073983127e-07,
+      "loss": 0.0782,
+      "step": 6945
+    },
+    {
+      "epoch": 1.7441495702365266,
+      "grad_norm": 0.14376573264598846,
+      "learning_rate": 8.441585615320269e-07,
+      "loss": 0.0865,
+      "step": 6950
+    },
+    {
+      "epoch": 1.7454043541000064,
+      "grad_norm": 0.19789855182170868,
+      "learning_rate": 8.360038428174022e-07,
+      "loss": 0.0932,
+      "step": 6955
+    },
+    {
+      "epoch": 1.7466591379634857,
+      "grad_norm": 0.2379181683063507,
+      "learning_rate": 8.278869849454718e-07,
+      "loss": 0.0809,
+      "step": 6960
+    },
+    {
+      "epoch": 1.7479139218269653,
+      "grad_norm": 0.3969587981700897,
+      "learning_rate": 8.198080214508486e-07,
+      "loss": 0.0843,
+      "step": 6965
+    },
+    {
+      "epoch": 1.7491687056904448,
+      "grad_norm": 0.23574315011501312,
+      "learning_rate": 8.117669857115895e-07,
+      "loss": 0.0853,
+      "step": 6970
+    },
+    {
+      "epoch": 1.7504234895539243,
+      "grad_norm": 0.3272697925567627,
+      "learning_rate": 8.037639109490524e-07,
+      "loss": 0.0818,
+      "step": 6975
+    },
+    {
+      "epoch": 1.751678273417404,
+      "grad_norm": 0.2460939586162567,
+      "learning_rate": 7.957988302277597e-07,
+      "loss": 0.0811,
+      "step": 6980
+    },
+    {
+      "epoch": 1.7529330572808832,
+      "grad_norm": 0.3471406102180481,
+      "learning_rate": 7.87871776455259e-07,
+      "loss": 0.0651,
+      "step": 6985
+    },
+    {
+      "epoch": 1.754187841144363,
+      "grad_norm": 0.22290903329849243,
+      "learning_rate": 7.799827823819972e-07,
+      "loss": 0.0793,
+      "step": 6990
+    },
+    {
+      "epoch": 1.7554426250078423,
+      "grad_norm": 0.1208643913269043,
+      "learning_rate": 7.721318806011713e-07,
+      "loss": 0.0765,
+      "step": 6995
+    },
+    {
+      "epoch": 1.7566974088713219,
+      "grad_norm": 0.176291823387146,
+      "learning_rate": 7.643191035486086e-07,
+      "loss": 0.0892,
+      "step": 7000
+    },
+    {
+      "epoch": 1.7579521927348014,
+      "grad_norm": 0.518677294254303,
+      "learning_rate": 7.56544483502617e-07,
+      "loss": 0.0762,
+      "step": 7005
+    },
+    {
+      "epoch": 1.759206976598281,
+      "grad_norm": 0.24120575189590454,
+      "learning_rate": 7.488080525838636e-07,
+      "loss": 0.0979,
+      "step": 7010
+    },
+    {
+      "epoch": 1.7604617604617605,
+      "grad_norm": 0.16398218274116516,
+      "learning_rate": 7.411098427552377e-07,
+      "loss": 0.0815,
+      "step": 7015
+    },
+    {
+      "epoch": 1.7617165443252398,
+      "grad_norm": 0.15948547422885895,
+      "learning_rate": 7.334498858217231e-07,
+      "loss": 0.082,
+      "step": 7020
+    },
+    {
+      "epoch": 1.7629713281887196,
+      "grad_norm": 0.163728266954422,
+      "learning_rate": 7.258282134302519e-07,
+      "loss": 0.0888,
+      "step": 7025
+    },
+    {
+      "epoch": 1.764226112052199,
+      "grad_norm": 0.18428277969360352,
+      "learning_rate": 7.182448570695944e-07,
+      "loss": 0.0796,
+      "step": 7030
+    },
+    {
+      "epoch": 1.7654808959156785,
+      "grad_norm": 0.2015131711959839,
+      "learning_rate": 7.106998480702165e-07,
+      "loss": 0.0741,
+      "step": 7035
+    },
+    {
+      "epoch": 1.766735679779158,
+      "grad_norm": 0.24829888343811035,
+      "learning_rate": 7.031932176041522e-07,
+      "loss": 0.0785,
+      "step": 7040
+    },
+    {
+      "epoch": 1.7679904636426376,
+      "grad_norm": 0.15142977237701416,
+      "learning_rate": 6.957249966848711e-07,
+      "loss": 0.0837,
+      "step": 7045
+    },
+    {
+      "epoch": 1.7692452475061171,
+      "grad_norm": 0.25044482946395874,
+      "learning_rate": 6.882952161671652e-07,
+      "loss": 0.083,
+      "step": 7050
+    },
+    {
+      "epoch": 1.7705000313695964,
+      "grad_norm": 0.27821311354637146,
+      "learning_rate": 6.809039067469991e-07,
+      "loss": 0.0851,
+      "step": 7055
+    },
+    {
+      "epoch": 1.7717548152330762,
+      "grad_norm": 0.21314631402492523,
+      "learning_rate": 6.735510989614047e-07,
+      "loss": 0.0891,
+      "step": 7060
+    },
+    {
+      "epoch": 1.7730095990965555,
+      "grad_norm": 0.24855484068393707,
+      "learning_rate": 6.662368231883388e-07,
+      "loss": 0.0828,
+      "step": 7065
+    },
+    {
+      "epoch": 1.774264382960035,
+      "grad_norm": 0.09157968312501907,
+      "learning_rate": 6.589611096465642e-07,
+      "loss": 0.0827,
+      "step": 7070
+    },
+    {
+      "epoch": 1.7755191668235146,
+      "grad_norm": 0.17919617891311646,
+      "learning_rate": 6.517239883955295e-07,
+      "loss": 0.082,
+      "step": 7075
+    },
+    {
+      "epoch": 1.7767739506869942,
+      "grad_norm": 0.32990872859954834,
+      "learning_rate": 6.445254893352381e-07,
+      "loss": 0.0847,
+      "step": 7080
+    },
+    {
+      "epoch": 1.7780287345504737,
+      "grad_norm": 0.20937122404575348,
+      "learning_rate": 6.373656422061247e-07,
+      "loss": 0.0729,
+      "step": 7085
+    },
+    {
+      "epoch": 1.779283518413953,
+      "grad_norm": 0.34432610869407654,
+      "learning_rate": 6.302444765889337e-07,
+      "loss": 0.0836,
+      "step": 7090
+    },
+    {
+      "epoch": 1.7805383022774328,
+      "grad_norm": 0.23893016576766968,
+      "learning_rate": 6.23162021904603e-07,
+      "loss": 0.085,
+      "step": 7095
+    },
+    {
+      "epoch": 1.7817930861409121,
+      "grad_norm": 0.26055169105529785,
+      "learning_rate": 6.161183074141319e-07,
+      "loss": 0.0836,
+      "step": 7100
+    },
+    {
+      "epoch": 1.7830478700043917,
+      "grad_norm": 0.3493628203868866,
+      "learning_rate": 6.091133622184664e-07,
+      "loss": 0.0879,
+      "step": 7105
+    },
+    {
+      "epoch": 1.7843026538678712,
+      "grad_norm": 0.15807025134563446,
+      "learning_rate": 6.021472152583818e-07,
+      "loss": 0.0846,
+      "step": 7110
+    },
+    {
+      "epoch": 1.7855574377313508,
+      "grad_norm": 0.19779103994369507,
+      "learning_rate": 5.952198953143539e-07,
+      "loss": 0.0914,
+      "step": 7115
+    },
+    {
+      "epoch": 1.7868122215948303,
+      "grad_norm": 0.2763902246952057,
+      "learning_rate": 5.883314310064492e-07,
+      "loss": 0.0752,
+      "step": 7120
+    },
+    {
+      "epoch": 1.7880670054583097,
+      "grad_norm": 0.22704683244228363,
+      "learning_rate": 5.814818507942055e-07,
+      "loss": 0.0785,
+      "step": 7125
+    },
+    {
+      "epoch": 1.7893217893217894,
+      "grad_norm": 0.3757963478565216,
+      "learning_rate": 5.746711829765017e-07,
+      "loss": 0.0814,
+      "step": 7130
+    },
+    {
+      "epoch": 1.7905765731852687,
+      "grad_norm": 0.1545080542564392,
+      "learning_rate": 5.678994556914618e-07,
+      "loss": 0.0743,
+      "step": 7135
+    },
+    {
+      "epoch": 1.7918313570487485,
+      "grad_norm": 0.18506070971488953,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 0.0831,
+      "step": 7140
+    },
+    {
+      "epoch": 1.7930861409122278,
+      "grad_norm": 0.17606933414936066,
+      "learning_rate": 5.544729344673294e-07,
+      "loss": 0.0753,
+      "step": 7145
+    },
+    {
+      "epoch": 1.7943409247757074,
+      "grad_norm": 0.30049043893814087,
+      "learning_rate": 5.47818195999602e-07,
+      "loss": 0.0893,
+      "step": 7150
+    },
+    {
+      "epoch": 1.795595708639187,
+      "grad_norm": 0.07360345870256424,
+      "learning_rate": 5.412025090070483e-07,
+      "loss": 0.0791,
+      "step": 7155
+    },
+    {
+      "epoch": 1.7968504925026663,
+      "grad_norm": 0.21283096075057983,
+      "learning_rate": 5.346259008222243e-07,
+      "loss": 0.0813,
+      "step": 7160
+    },
+    {
+      "epoch": 1.798105276366146,
+      "grad_norm": 0.38110825419425964,
+      "learning_rate": 5.280883986162433e-07,
+      "loss": 0.0791,
+      "step": 7165
+    },
+    {
+      "epoch": 1.7993600602296254,
+      "grad_norm": 0.12914220988750458,
+      "learning_rate": 5.215900293986431e-07,
+      "loss": 0.0758,
+      "step": 7170
+    },
+    {
+      "epoch": 1.8006148440931051,
+      "grad_norm": 0.5188528299331665,
+      "learning_rate": 5.151308200172911e-07,
+      "loss": 0.0823,
+      "step": 7175
+    },
+    {
+      "epoch": 1.8018696279565845,
+      "grad_norm": 0.23793403804302216,
+      "learning_rate": 5.087107971582628e-07,
+      "loss": 0.0786,
+      "step": 7180
+    },
+    {
+      "epoch": 1.803124411820064,
+      "grad_norm": 0.18523241579532623,
+      "learning_rate": 5.02329987345741e-07,
+      "loss": 0.0864,
+      "step": 7185
+    },
+    {
+      "epoch": 1.8043791956835435,
+      "grad_norm": 0.18162751197814941,
+      "learning_rate": 4.959884169418949e-07,
+      "loss": 0.079,
+      "step": 7190
+    },
+    {
+      "epoch": 1.8056339795470229,
+      "grad_norm": 0.15588027238845825,
+      "learning_rate": 4.896861121467778e-07,
+      "loss": 0.0791,
+      "step": 7195
+    },
+    {
+      "epoch": 1.8068887634105026,
+      "grad_norm": 0.16926927864551544,
+      "learning_rate": 4.834230989982214e-07,
+      "loss": 0.0762,
+      "step": 7200
+    },
+    {
+      "epoch": 1.808143547273982,
+      "grad_norm": 0.29264163970947266,
+      "learning_rate": 4.77199403371722e-07,
+      "loss": 0.0805,
+      "step": 7205
+    },
+    {
+      "epoch": 1.8093983311374617,
+      "grad_norm": 0.15263986587524414,
+      "learning_rate": 4.7101505098033575e-07,
+      "loss": 0.0893,
+      "step": 7210
+    },
+    {
+      "epoch": 1.810653115000941,
+      "grad_norm": 0.12115020304918289,
+      "learning_rate": 4.6487006737457765e-07,
+      "loss": 0.0773,
+      "step": 7215
+    },
+    {
+      "epoch": 1.8119078988644206,
+      "grad_norm": 0.08451762050390244,
+      "learning_rate": 4.5876447794230504e-07,
+      "loss": 0.0772,
+      "step": 7220
+    },
+    {
+      "epoch": 1.8131626827279002,
+      "grad_norm": 0.2606160640716553,
+      "learning_rate": 4.5269830790862444e-07,
+      "loss": 0.0787,
+      "step": 7225
+    },
+    {
+      "epoch": 1.8144174665913795,
+      "grad_norm": 0.18426910042762756,
+      "learning_rate": 4.4667158233577925e-07,
+      "loss": 0.0788,
+      "step": 7230
+    },
+    {
+      "epoch": 1.8156722504548592,
+      "grad_norm": 0.1966276913881302,
+      "learning_rate": 4.40684326123052e-07,
+      "loss": 0.0955,
+      "step": 7235
+    },
+    {
+      "epoch": 1.8169270343183386,
+      "grad_norm": 0.19326919317245483,
+      "learning_rate": 4.3473656400665256e-07,
+      "loss": 0.0984,
+      "step": 7240
+    },
+    {
+      "epoch": 1.8181818181818183,
+      "grad_norm": 0.2253500372171402,
+      "learning_rate": 4.2882832055962885e-07,
+      "loss": 0.0747,
+      "step": 7245
+    },
+    {
+      "epoch": 1.8194366020452977,
+      "grad_norm": 0.28359681367874146,
+      "learning_rate": 4.22959620191753e-07,
+      "loss": 0.0751,
+      "step": 7250
+    },
+    {
+      "epoch": 1.8206913859087772,
+      "grad_norm": 0.12297849357128143,
+      "learning_rate": 4.171304871494264e-07,
+      "loss": 0.0708,
+      "step": 7255
+    },
+    {
+      "epoch": 1.8219461697722568,
+      "grad_norm": 0.049460649490356445,
+      "learning_rate": 4.113409455155837e-07,
+      "loss": 0.0758,
+      "step": 7260
+    },
+    {
+      "epoch": 1.823200953635736,
+      "grad_norm": 0.1316637098789215,
+      "learning_rate": 4.0559101920958243e-07,
+      "loss": 0.075,
+      "step": 7265
+    },
+    {
+      "epoch": 1.8244557374992159,
+      "grad_norm": 0.11034820973873138,
+      "learning_rate": 3.9988073198711564e-07,
+      "loss": 0.0816,
+      "step": 7270
+    },
+    {
+      "epoch": 1.8257105213626952,
+      "grad_norm": 0.5154989957809448,
+      "learning_rate": 3.942101074401028e-07,
+      "loss": 0.0787,
+      "step": 7275
+    },
+    {
+      "epoch": 1.826965305226175,
+      "grad_norm": 0.4702956974506378,
+      "learning_rate": 3.885791689966023e-07,
+      "loss": 0.0862,
+      "step": 7280
+    },
+    {
+      "epoch": 1.8282200890896543,
+      "grad_norm": 0.41101551055908203,
+      "learning_rate": 3.8298793992070814e-07,
+      "loss": 0.0888,
+      "step": 7285
+    },
+    {
+      "epoch": 1.8294748729531338,
+      "grad_norm": 0.16913031041622162,
+      "learning_rate": 3.774364433124578e-07,
+      "loss": 0.0761,
+      "step": 7290
+    },
+    {
+      "epoch": 1.8307296568166134,
+      "grad_norm": 0.1570875644683838,
+      "learning_rate": 3.7192470210773435e-07,
+      "loss": 0.0713,
+      "step": 7295
+    },
+    {
+      "epoch": 1.8319844406800927,
+      "grad_norm": 0.45383045077323914,
+      "learning_rate": 3.6645273907816805e-07,
+      "loss": 0.0774,
+      "step": 7300
+    },
+    {
+      "epoch": 1.8332392245435725,
+      "grad_norm": 0.12889161705970764,
+      "learning_rate": 3.6102057683105596e-07,
+      "loss": 0.0824,
+      "step": 7305
+    },
+    {
+      "epoch": 1.8344940084070518,
+      "grad_norm": 0.34759390354156494,
+      "learning_rate": 3.5562823780924906e-07,
+      "loss": 0.0912,
+      "step": 7310
+    },
+    {
+      "epoch": 1.8357487922705316,
+      "grad_norm": 0.35057494044303894,
+      "learning_rate": 3.5027574429107536e-07,
+      "loss": 0.0732,
+      "step": 7315
+    },
+    {
+      "epoch": 1.8370035761340109,
+      "grad_norm": 0.1586838960647583,
+      "learning_rate": 3.4496311839024133e-07,
+      "loss": 0.0896,
+      "step": 7320
+    },
+    {
+      "epoch": 1.8382583599974904,
+      "grad_norm": 0.13935531675815582,
+      "learning_rate": 3.396903820557385e-07,
+      "loss": 0.0796,
+      "step": 7325
+    },
+    {
+      "epoch": 1.83951314386097,
+      "grad_norm": 0.20520001649856567,
+      "learning_rate": 3.344575570717612e-07,
+      "loss": 0.0839,
+      "step": 7330
+    },
+    {
+      "epoch": 1.8407679277244493,
+      "grad_norm": 0.13597147166728973,
+      "learning_rate": 3.292646650576037e-07,
+      "loss": 0.0852,
+      "step": 7335
+    },
+    {
+      "epoch": 1.842022711587929,
+      "grad_norm": 0.1146201342344284,
+      "learning_rate": 3.2411172746758424e-07,
+      "loss": 0.0845,
+      "step": 7340
+    },
+    {
+      "epoch": 1.8432774954514084,
+      "grad_norm": 0.24954962730407715,
+      "learning_rate": 3.1899876559094657e-07,
+      "loss": 0.0705,
+      "step": 7345
+    },
+    {
+      "epoch": 1.8445322793148882,
+      "grad_norm": 0.43199557065963745,
+      "learning_rate": 3.1392580055177867e-07,
+      "loss": 0.0894,
+      "step": 7350
+    },
+    {
+      "epoch": 1.8457870631783675,
+      "grad_norm": 0.3313222825527191,
+      "learning_rate": 3.0889285330891973e-07,
+      "loss": 0.0888,
+      "step": 7355
+    },
+    {
+      "epoch": 1.847041847041847,
+      "grad_norm": 0.08865600824356079,
+      "learning_rate": 3.038999446558755e-07,
+      "loss": 0.086,
+      "step": 7360
+    },
+    {
+      "epoch": 1.8482966309053266,
+      "grad_norm": 0.21259349584579468,
+      "learning_rate": 2.989470952207385e-07,
+      "loss": 0.0911,
+      "step": 7365
+    },
+    {
+      "epoch": 1.849551414768806,
+      "grad_norm": 0.28370821475982666,
+      "learning_rate": 2.940343254660905e-07,
+      "loss": 0.0883,
+      "step": 7370
+    },
+    {
+      "epoch": 1.8508061986322857,
+      "grad_norm": 0.1723158210515976,
+      "learning_rate": 2.891616556889321e-07,
+      "loss": 0.0745,
+      "step": 7375
+    },
+    {
+      "epoch": 1.852060982495765,
+      "grad_norm": 0.2103600651025772,
+      "learning_rate": 2.843291060205855e-07,
+      "loss": 0.0821,
+      "step": 7380
+    },
+    {
+      "epoch": 1.8533157663592448,
+      "grad_norm": 0.19018976390361786,
+      "learning_rate": 2.7953669642662107e-07,
+      "loss": 0.0885,
+      "step": 7385
+    },
+    {
+      "epoch": 1.854570550222724,
+      "grad_norm": 0.38891276717185974,
+      "learning_rate": 2.747844467067706e-07,
+      "loss": 0.0832,
+      "step": 7390
+    },
+    {
+      "epoch": 1.8558253340862036,
+      "grad_norm": 0.17385374009609222,
+      "learning_rate": 2.7007237649484763e-07,
+      "loss": 0.0788,
+      "step": 7395
+    },
+    {
+      "epoch": 1.8570801179496832,
+      "grad_norm": 0.453713059425354,
+      "learning_rate": 2.654005052586628e-07,
+      "loss": 0.0777,
+      "step": 7400
+    },
+    {
+      "epoch": 1.8583349018131625,
+      "grad_norm": 0.20061570405960083,
+      "learning_rate": 2.607688522999441e-07,
+      "loss": 0.0742,
+      "step": 7405
+    },
+    {
+      "epoch": 1.8595896856766423,
+      "grad_norm": 0.1047121062874794,
+      "learning_rate": 2.5617743675426354e-07,
+      "loss": 0.0917,
+      "step": 7410
+    },
+    {
+      "epoch": 1.8608444695401216,
+      "grad_norm": 0.24845577776432037,
+      "learning_rate": 2.516262775909506e-07,
+      "loss": 0.0806,
+      "step": 7415
+    },
+    {
+      "epoch": 1.8620992534036014,
+      "grad_norm": 0.3548727035522461,
+      "learning_rate": 2.471153936130133e-07,
+      "loss": 0.0913,
+      "step": 7420
+    },
+    {
+      "epoch": 1.8633540372670807,
+      "grad_norm": 0.2546403706073761,
+      "learning_rate": 2.4264480345707053e-07,
+      "loss": 0.0855,
+      "step": 7425
+    },
+    {
+      "epoch": 1.8646088211305603,
+      "grad_norm": 0.2284100353717804,
+      "learning_rate": 2.3821452559326218e-07,
+      "loss": 0.0849,
+      "step": 7430
+    },
+    {
+      "epoch": 1.8658636049940398,
+      "grad_norm": 0.25266894698143005,
+      "learning_rate": 2.3382457832518134e-07,
+      "loss": 0.0778,
+      "step": 7435
+    },
+    {
+      "epoch": 1.8671183888575191,
+      "grad_norm": 0.19911006093025208,
+      "learning_rate": 2.294749797897955e-07,
+      "loss": 0.0781,
+      "step": 7440
+    },
+    {
+      "epoch": 1.868373172720999,
+      "grad_norm": 0.35023069381713867,
+      "learning_rate": 2.2516574795737323e-07,
+      "loss": 0.0787,
+      "step": 7445
+    },
+    {
+      "epoch": 1.8696279565844782,
+      "grad_norm": 0.19626636803150177,
+      "learning_rate": 2.2089690063140766e-07,
+      "loss": 0.0833,
+      "step": 7450
+    },
+    {
+      "epoch": 1.870882740447958,
+      "grad_norm": 0.1876339167356491,
+      "learning_rate": 2.1666845544854542e-07,
+      "loss": 0.084,
+      "step": 7455
+    },
+    {
+      "epoch": 1.8721375243114373,
+      "grad_norm": 0.09253226220607758,
+      "learning_rate": 2.1248042987851325e-07,
+      "loss": 0.0875,
+      "step": 7460
+    },
+    {
+      "epoch": 1.8733923081749169,
+      "grad_norm": 0.31013625860214233,
+      "learning_rate": 2.083328412240404e-07,
+      "loss": 0.083,
+      "step": 7465
+    },
+    {
+      "epoch": 1.8746470920383964,
+      "grad_norm": 0.17299607396125793,
+      "learning_rate": 2.0422570662079866e-07,
+      "loss": 0.0887,
+      "step": 7470
+    },
+    {
+      "epoch": 1.8759018759018757,
+      "grad_norm": 0.1518164724111557,
+      "learning_rate": 2.0015904303732126e-07,
+      "loss": 0.0711,
+      "step": 7475
+    },
+    {
+      "epoch": 1.8771566597653555,
+      "grad_norm": 0.16225028038024902,
+      "learning_rate": 1.961328672749352e-07,
+      "loss": 0.0743,
+      "step": 7480
+    },
+    {
+      "epoch": 1.8784114436288348,
+      "grad_norm": 0.39483585953712463,
+      "learning_rate": 1.921471959676957e-07,
+      "loss": 0.0856,
+      "step": 7485
+    },
+    {
+      "epoch": 1.8796662274923146,
+      "grad_norm": 0.43932509422302246,
+      "learning_rate": 1.8820204558231415e-07,
+      "loss": 0.0813,
+      "step": 7490
+    },
+    {
+      "epoch": 1.880921011355794,
+      "grad_norm": 0.29697397351264954,
+      "learning_rate": 1.8429743241808795e-07,
+      "loss": 0.079,
+      "step": 7495
+    },
+    {
+      "epoch": 1.8821757952192735,
+      "grad_norm": 0.1230466440320015,
+      "learning_rate": 1.804333726068408e-07,
+      "loss": 0.0918,
+      "step": 7500
+    },
+    {
+      "epoch": 1.883430579082753,
+      "grad_norm": 0.18823686242103577,
+      "learning_rate": 1.766098821128459e-07,
+      "loss": 0.0755,
+      "step": 7505
+    },
+    {
+      "epoch": 1.8846853629462323,
+      "grad_norm": 0.3333011865615845,
+      "learning_rate": 1.7282697673276837e-07,
+      "loss": 0.0908,
+      "step": 7510
+    },
+    {
+      "epoch": 1.8859401468097121,
+      "grad_norm": 0.15476582944393158,
+      "learning_rate": 1.6908467209559853e-07,
+      "loss": 0.0869,
+      "step": 7515
+    },
+    {
+      "epoch": 1.8871949306731914,
+      "grad_norm": 0.20987029373645782,
+      "learning_rate": 1.6538298366257975e-07,
+      "loss": 0.074,
+      "step": 7520
+    },
+    {
+      "epoch": 1.8884497145366712,
+      "grad_norm": 0.13596096634864807,
+      "learning_rate": 1.6172192672715525e-07,
+      "loss": 0.0886,
+      "step": 7525
+    },
+    {
+      "epoch": 1.8897044984001505,
+      "grad_norm": 0.22909042239189148,
+      "learning_rate": 1.5810151641489912e-07,
+      "loss": 0.0846,
+      "step": 7530
+    },
+    {
+      "epoch": 1.89095928226363,
+      "grad_norm": 0.24221235513687134,
+      "learning_rate": 1.545217676834554e-07,
+      "loss": 0.0815,
+      "step": 7535
+    },
+    {
+      "epoch": 1.8922140661271096,
+      "grad_norm": 0.14234666526317596,
+      "learning_rate": 1.5098269532247357e-07,
+      "loss": 0.0863,
+      "step": 7540
+    },
+    {
+      "epoch": 1.8934688499905892,
+      "grad_norm": 0.24849990010261536,
+      "learning_rate": 1.4748431395355088e-07,
+      "loss": 0.0819,
+      "step": 7545
+    },
+    {
+      "epoch": 1.8947236338540687,
+      "grad_norm": 0.2965373992919922,
+      "learning_rate": 1.4402663803017249e-07,
+      "loss": 0.0795,
+      "step": 7550
+    },
+    {
+      "epoch": 1.895978417717548,
+      "grad_norm": 0.3867776691913605,
+      "learning_rate": 1.4060968183764678e-07,
+      "loss": 0.0904,
+      "step": 7555
+    },
+    {
+      "epoch": 1.8972332015810278,
+      "grad_norm": 0.31307804584503174,
+      "learning_rate": 1.3723345949305245e-07,
+      "loss": 0.0798,
+      "step": 7560
+    },
+    {
+      "epoch": 1.8984879854445071,
+      "grad_norm": 0.3106546700000763,
+      "learning_rate": 1.338979849451738e-07,
+      "loss": 0.0864,
+      "step": 7565
+    },
+    {
+      "epoch": 1.8997427693079867,
+      "grad_norm": 0.22093643248081207,
+      "learning_rate": 1.3060327197444767e-07,
+      "loss": 0.0776,
+      "step": 7570
+    },
+    {
+      "epoch": 1.9009975531714662,
+      "grad_norm": 0.1310010552406311,
+      "learning_rate": 1.2734933419290996e-07,
+      "loss": 0.077,
+      "step": 7575
+    },
+    {
+      "epoch": 1.9022523370349458,
+      "grad_norm": 0.23409627377986908,
+      "learning_rate": 1.2413618504412806e-07,
+      "loss": 0.0777,
+      "step": 7580
+    },
+    {
+      "epoch": 1.9035071208984253,
+      "grad_norm": 0.2928149402141571,
+      "learning_rate": 1.2096383780315411e-07,
+      "loss": 0.0899,
+      "step": 7585
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.26420190930366516,
+      "learning_rate": 1.1783230557647075e-07,
+      "loss": 0.0875,
+      "step": 7590
+    },
+    {
+      "epoch": 1.9060166886253844,
+      "grad_norm": 0.17748254537582397,
+      "learning_rate": 1.1474160130193313e-07,
+      "loss": 0.0878,
+      "step": 7595
+    },
+    {
+      "epoch": 1.9072714724888638,
+      "grad_norm": 0.41803818941116333,
+      "learning_rate": 1.1169173774871478e-07,
+      "loss": 0.0881,
+      "step": 7600
+    },
+    {
+      "epoch": 1.9085262563523433,
+      "grad_norm": 0.22346888482570648,
+      "learning_rate": 1.086827275172575e-07,
+      "loss": 0.0794,
+      "step": 7605
+    },
+    {
+      "epoch": 1.9097810402158228,
+      "grad_norm": 0.1962645947933197,
+      "learning_rate": 1.0571458303922033e-07,
+      "loss": 0.0811,
+      "step": 7610
+    },
+    {
+      "epoch": 1.9110358240793024,
+      "grad_norm": 0.3406698703765869,
+      "learning_rate": 1.0278731657742292e-07,
+      "loss": 0.0816,
+      "step": 7615
+    },
+    {
+      "epoch": 1.912290607942782,
+      "grad_norm": 0.14449894428253174,
+      "learning_rate": 9.990094022580332e-08,
+      "loss": 0.0856,
+      "step": 7620
+    },
+    {
+      "epoch": 1.9135453918062613,
+      "grad_norm": 0.1097446084022522,
+      "learning_rate": 9.70554659093581e-08,
+      "loss": 0.0911,
+      "step": 7625
+    },
+    {
+      "epoch": 1.914800175669741,
+      "grad_norm": 0.43302279710769653,
+      "learning_rate": 9.425090538409898e-08,
+      "loss": 0.0877,
+      "step": 7630
+    },
+    {
+      "epoch": 1.9160549595332204,
+      "grad_norm": 0.26026082038879395,
+      "learning_rate": 9.148727023700731e-08,
+      "loss": 0.0803,
+      "step": 7635
+    },
+    {
+      "epoch": 1.9173097433967,
+      "grad_norm": 0.2453165352344513,
+      "learning_rate": 8.876457188597642e-08,
+      "loss": 0.0843,
+      "step": 7640
+    },
+    {
+      "epoch": 1.9185645272601795,
+      "grad_norm": 0.4708387851715088,
+      "learning_rate": 8.608282157977488e-08,
+      "loss": 0.0788,
+      "step": 7645
+    },
+    {
+      "epoch": 1.919819311123659,
+      "grad_norm": 0.12131650745868683,
+      "learning_rate": 8.344203039799214e-08,
+      "loss": 0.0806,
+      "step": 7650
+    },
+    {
+      "epoch": 1.9210740949871385,
+      "grad_norm": 0.21256065368652344,
+      "learning_rate": 8.084220925099751e-08,
+      "loss": 0.087,
+      "step": 7655
+    },
+    {
+      "epoch": 1.9223288788506179,
+      "grad_norm": 0.2987827658653259,
+      "learning_rate": 7.82833688798934e-08,
+      "loss": 0.0807,
+      "step": 7660
+    },
+    {
+      "epoch": 1.9235836627140976,
+      "grad_norm": 0.12467091530561447,
+      "learning_rate": 7.576551985647107e-08,
+      "loss": 0.0995,
+      "step": 7665
+    },
+    {
+      "epoch": 1.924838446577577,
+      "grad_norm": 0.15592101216316223,
+      "learning_rate": 7.328867258316608e-08,
+      "loss": 0.0881,
+      "step": 7670
+    },
+    {
+      "epoch": 1.9260932304410565,
+      "grad_norm": 0.1511273980140686,
+      "learning_rate": 7.085283729301728e-08,
+      "loss": 0.0847,
+      "step": 7675
+    },
+    {
+      "epoch": 1.927348014304536,
+      "grad_norm": 0.13429687917232513,
+      "learning_rate": 6.845802404962243e-08,
+      "loss": 0.0766,
+      "step": 7680
+    },
+    {
+      "epoch": 1.9286027981680156,
+      "grad_norm": 0.10453824698925018,
+      "learning_rate": 6.610424274710037e-08,
+      "loss": 0.077,
+      "step": 7685
+    },
+    {
+      "epoch": 1.9298575820314952,
+      "grad_norm": 0.18403221666812897,
+      "learning_rate": 6.379150311004224e-08,
+      "loss": 0.0802,
+      "step": 7690
+    },
+    {
+      "epoch": 1.9311123658949745,
+      "grad_norm": 0.09399021416902542,
+      "learning_rate": 6.151981469348034e-08,
+      "loss": 0.0947,
+      "step": 7695
+    },
+    {
+      "epoch": 1.9323671497584543,
+      "grad_norm": 0.19107823073863983,
+      "learning_rate": 5.928918688284602e-08,
+      "loss": 0.0799,
+      "step": 7700
+    },
+    {
+      "epoch": 1.9336219336219336,
+      "grad_norm": 0.18196965754032135,
+      "learning_rate": 5.709962889392628e-08,
+      "loss": 0.0696,
+      "step": 7705
+    },
+    {
+      "epoch": 1.9348767174854131,
+      "grad_norm": 0.32829707860946655,
+      "learning_rate": 5.495114977282945e-08,
+      "loss": 0.0825,
+      "step": 7710
+    },
+    {
+      "epoch": 1.9361315013488927,
+      "grad_norm": 0.21754246950149536,
+      "learning_rate": 5.284375839594958e-08,
+      "loss": 0.0936,
+      "step": 7715
+    },
+    {
+      "epoch": 1.9373862852123722,
+      "grad_norm": 0.13130666315555573,
+      "learning_rate": 5.0777463469925406e-08,
+      "loss": 0.089,
+      "step": 7720
+    },
+    {
+      "epoch": 1.9386410690758518,
+      "grad_norm": 0.1840493381023407,
+      "learning_rate": 4.8752273531609276e-08,
+      "loss": 0.0847,
+      "step": 7725
+    },
+    {
+      "epoch": 1.939895852939331,
+      "grad_norm": 0.12967444956302643,
+      "learning_rate": 4.676819694802604e-08,
+      "loss": 0.0782,
+      "step": 7730
+    },
+    {
+      "epoch": 1.9411506368028109,
+      "grad_norm": 0.1979934573173523,
+      "learning_rate": 4.4825241916344184e-08,
+      "loss": 0.0797,
+      "step": 7735
+    },
+    {
+      "epoch": 1.9424054206662902,
+      "grad_norm": 0.12666776776313782,
+      "learning_rate": 4.292341646383813e-08,
+      "loss": 0.0794,
+      "step": 7740
+    },
+    {
+      "epoch": 1.9436602045297697,
+      "grad_norm": 0.11725469678640366,
+      "learning_rate": 4.106272844785486e-08,
+      "loss": 0.0777,
+      "step": 7745
+    },
+    {
+      "epoch": 1.9449149883932493,
+      "grad_norm": 0.4441820979118347,
+      "learning_rate": 3.924318555578843e-08,
+      "loss": 0.0829,
+      "step": 7750
+    },
+    {
+      "epoch": 1.9461697722567288,
+      "grad_norm": 0.26033830642700195,
+      "learning_rate": 3.7464795305036664e-08,
+      "loss": 0.0806,
+      "step": 7755
+    },
+    {
+      "epoch": 1.9474245561202084,
+      "grad_norm": 0.19797271490097046,
+      "learning_rate": 3.572756504297892e-08,
+      "loss": 0.0681,
+      "step": 7760
+    },
+    {
+      "epoch": 1.9486793399836877,
+      "grad_norm": 0.15105128288269043,
+      "learning_rate": 3.4031501946942826e-08,
+      "loss": 0.0868,
+      "step": 7765
+    },
+    {
+      "epoch": 1.9499341238471675,
+      "grad_norm": 0.2811802625656128,
+      "learning_rate": 3.2376613024175384e-08,
+      "loss": 0.0895,
+      "step": 7770
+    },
+    {
+      "epoch": 1.9511889077106468,
+      "grad_norm": 0.3569127917289734,
+      "learning_rate": 3.0762905111811904e-08,
+      "loss": 0.0927,
+      "step": 7775
+    },
+    {
+      "epoch": 1.9524436915741263,
+      "grad_norm": 0.5091187953948975,
+      "learning_rate": 2.9190384876849333e-08,
+      "loss": 0.0864,
+      "step": 7780
+    },
+    {
+      "epoch": 1.9536984754376059,
+      "grad_norm": 0.228666290640831,
+      "learning_rate": 2.7659058816121855e-08,
+      "loss": 0.0897,
+      "step": 7785
+    },
+    {
+      "epoch": 1.9549532593010854,
+      "grad_norm": 0.07750795036554337,
+      "learning_rate": 2.616893325626646e-08,
+      "loss": 0.0733,
+      "step": 7790
+    },
+    {
+      "epoch": 1.956208043164565,
+      "grad_norm": 0.2381664514541626,
+      "learning_rate": 2.472001435370297e-08,
+      "loss": 0.0729,
+      "step": 7795
+    },
+    {
+      "epoch": 1.9574628270280443,
+      "grad_norm": 0.27068066596984863,
+      "learning_rate": 2.3312308094607382e-08,
+      "loss": 0.0957,
+      "step": 7800
+    },
+    {
+      "epoch": 1.958717610891524,
+      "grad_norm": 0.37700343132019043,
+      "learning_rate": 2.1945820294888564e-08,
+      "loss": 0.0766,
+      "step": 7805
+    },
+    {
+      "epoch": 1.9599723947550034,
+      "grad_norm": 0.32152286171913147,
+      "learning_rate": 2.062055660015716e-08,
+      "loss": 0.0764,
+      "step": 7810
+    },
+    {
+      "epoch": 1.961227178618483,
+      "grad_norm": 0.164317324757576,
+      "learning_rate": 1.9336522485710053e-08,
+      "loss": 0.0825,
+      "step": 7815
+    },
+    {
+      "epoch": 1.9624819624819625,
+      "grad_norm": 0.1482664942741394,
+      "learning_rate": 1.8093723256507044e-08,
+      "loss": 0.0761,
+      "step": 7820
+    },
+    {
+      "epoch": 1.963736746345442,
+      "grad_norm": 0.15529005229473114,
+      "learning_rate": 1.689216404714311e-08,
+      "loss": 0.0879,
+      "step": 7825
+    },
+    {
+      "epoch": 1.9649915302089216,
+      "grad_norm": 0.2088499665260315,
+      "learning_rate": 1.5731849821833955e-08,
+      "loss": 0.0852,
+      "step": 7830
+    },
+    {
+      "epoch": 1.966246314072401,
+      "grad_norm": 0.2643318176269531,
+      "learning_rate": 1.4612785374392701e-08,
+      "loss": 0.0798,
+      "step": 7835
+    },
+    {
+      "epoch": 1.9675010979358807,
+      "grad_norm": 0.16332080960273743,
+      "learning_rate": 1.3534975328205468e-08,
+      "loss": 0.0802,
+      "step": 7840
+    },
+    {
+      "epoch": 1.96875588179936,
+      "grad_norm": 0.1658352017402649,
+      "learning_rate": 1.2498424136223597e-08,
+      "loss": 0.0821,
+      "step": 7845
+    },
+    {
+      "epoch": 1.9700106656628396,
+      "grad_norm": 0.35251694917678833,
+      "learning_rate": 1.1503136080932565e-08,
+      "loss": 0.0822,
+      "step": 7850
+    },
+    {
+      "epoch": 1.971265449526319,
+      "grad_norm": 0.1039208248257637,
+      "learning_rate": 1.0549115274344213e-08,
+      "loss": 0.0847,
+      "step": 7855
+    },
+    {
+      "epoch": 1.9725202333897986,
+      "grad_norm": 0.3832782208919525,
+      "learning_rate": 9.636365657971215e-09,
+      "loss": 0.0732,
+      "step": 7860
+    },
+    {
+      "epoch": 1.9737750172532782,
+      "grad_norm": 0.35436856746673584,
+      "learning_rate": 8.764891002821519e-09,
+      "loss": 0.0796,
+      "step": 7865
+    },
+    {
+      "epoch": 1.9750298011167575,
+      "grad_norm": 0.3917001187801361,
+      "learning_rate": 7.93469490936949e-09,
+      "loss": 0.0747,
+      "step": 7870
+    },
+    {
+      "epoch": 1.9762845849802373,
+      "grad_norm": 0.20635858178138733,
+      "learning_rate": 7.145780807553681e-09,
+      "loss": 0.0804,
+      "step": 7875
+    },
+    {
+      "epoch": 1.9775393688437166,
+      "grad_norm": 0.12219113856554031,
+      "learning_rate": 6.398151956754639e-09,
+      "loss": 0.0854,
+      "step": 7880
+    },
+    {
+      "epoch": 1.9787941527071962,
+      "grad_norm": 0.08329229801893234,
+      "learning_rate": 5.69181144578268e-09,
+      "loss": 0.0863,
+      "step": 7885
+    },
+    {
+      "epoch": 1.9800489365706757,
+      "grad_norm": 0.2005588710308075,
+      "learning_rate": 5.026762192870127e-09,
+      "loss": 0.077,
+      "step": 7890
+    },
+    {
+      "epoch": 1.9813037204341553,
+      "grad_norm": 0.1955815702676773,
+      "learning_rate": 4.403006945650212e-09,
+      "loss": 0.0843,
+      "step": 7895
+    },
+    {
+      "epoch": 1.9825585042976348,
+      "grad_norm": 0.17901791632175446,
+      "learning_rate": 3.820548281154857e-09,
+      "loss": 0.0731,
+      "step": 7900
+    },
+    {
+      "epoch": 1.9838132881611141,
+      "grad_norm": 0.1416298747062683,
+      "learning_rate": 3.2793886057991277e-09,
+      "loss": 0.0729,
+      "step": 7905
+    },
+    {
+      "epoch": 1.985068072024594,
+      "grad_norm": 0.3370745778083801,
+      "learning_rate": 2.7795301553712463e-09,
+      "loss": 0.0794,
+      "step": 7910
+    },
+    {
+      "epoch": 1.9863228558880732,
+      "grad_norm": 0.16576644778251648,
+      "learning_rate": 2.3209749950259264e-09,
+      "loss": 0.075,
+      "step": 7915
+    },
+    {
+      "epoch": 1.9875776397515528,
+      "grad_norm": 0.2831018567085266,
+      "learning_rate": 1.9037250192732728e-09,
+      "loss": 0.0754,
+      "step": 7920
+    },
+    {
+      "epoch": 1.9888324236150323,
+      "grad_norm": 0.1539755016565323,
+      "learning_rate": 1.527781951971008e-09,
+      "loss": 0.0859,
+      "step": 7925
+    },
+    {
+      "epoch": 1.9900872074785119,
+      "grad_norm": 0.18794752657413483,
+      "learning_rate": 1.1931473463200339e-09,
+      "loss": 0.0924,
+      "step": 7930
+    },
+    {
+      "epoch": 1.9913419913419914,
+      "grad_norm": 0.1152065321803093,
+      "learning_rate": 8.998225848566577e-10,
+      "loss": 0.0793,
+      "step": 7935
+    },
+    {
+      "epoch": 1.9925967752054707,
+      "grad_norm": 0.36054694652557373,
+      "learning_rate": 6.478088794448223e-10,
+      "loss": 0.0732,
+      "step": 7940
+    },
+    {
+      "epoch": 1.9938515590689505,
+      "grad_norm": 0.05416973680257797,
+      "learning_rate": 4.3710727127277417e-10,
+      "loss": 0.0812,
+      "step": 7945
+    },
+    {
+      "epoch": 1.9951063429324298,
+      "grad_norm": 0.19745150208473206,
+      "learning_rate": 2.677186308497337e-10,
+      "loss": 0.0805,
+      "step": 7950
+    },
+    {
+      "epoch": 1.9963611267959094,
+      "grad_norm": 0.11429372429847717,
+      "learning_rate": 1.3964365800145374e-10,
+      "loss": 0.0796,
+      "step": 7955
+    },
+    {
+      "epoch": 1.997615910659389,
+      "grad_norm": 0.16175100207328796,
+      "learning_rate": 5.288288186688917e-11,
+      "loss": 0.0791,
+      "step": 7960
+    },
+    {
+      "epoch": 1.9988706945228685,
+      "grad_norm": 0.3085331618785858,
+      "learning_rate": 7.436660894866165e-12,
+      "loss": 0.0807,
+      "step": 7965
+    },
+    {
+      "epoch": 1.9996235648409562,
+      "step": 7968,
+      "total_flos": 0.0,
+      "train_loss": 0.0878100987938962,
+      "train_runtime": 323838.4314,
+      "train_samples_per_second": 1.575,
+      "train_steps_per_second": 0.025
+    }
+  ],
+  "logging_steps": 5,
+  "max_steps": 7968,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}