table-cleanup-policy-lora-20k / trainer_state.json
tyrleng's picture
Upload folder using huggingface_hub
25801f5 verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 13.812154696132596,
"eval_steps": 500,
"global_step": 20000,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.006906077348066298,
"grad_norm": 2.1743838787078857,
"learning_rate": 9e-07,
"loss": 0.9948,
"step": 10
},
{
"epoch": 0.013812154696132596,
"grad_norm": 1.8689522743225098,
"learning_rate": 1.9e-06,
"loss": 0.9971,
"step": 20
},
{
"epoch": 0.020718232044198894,
"grad_norm": 1.4171581268310547,
"learning_rate": 2.9e-06,
"loss": 0.9942,
"step": 30
},
{
"epoch": 0.027624309392265192,
"grad_norm": 1.5665574073791504,
"learning_rate": 3.9e-06,
"loss": 0.954,
"step": 40
},
{
"epoch": 0.034530386740331494,
"grad_norm": 1.5501309633255005,
"learning_rate": 4.9000000000000005e-06,
"loss": 0.8629,
"step": 50
},
{
"epoch": 0.04143646408839779,
"grad_norm": 1.1840782165527344,
"learning_rate": 5.9e-06,
"loss": 0.8254,
"step": 60
},
{
"epoch": 0.04834254143646409,
"grad_norm": 1.3108546733856201,
"learning_rate": 6.900000000000001e-06,
"loss": 0.6742,
"step": 70
},
{
"epoch": 0.055248618784530384,
"grad_norm": 0.9183031320571899,
"learning_rate": 7.9e-06,
"loss": 0.5774,
"step": 80
},
{
"epoch": 0.062154696132596686,
"grad_norm": 0.5960970520973206,
"learning_rate": 8.9e-06,
"loss": 0.4576,
"step": 90
},
{
"epoch": 0.06906077348066299,
"grad_norm": 0.5846773982048035,
"learning_rate": 9.900000000000002e-06,
"loss": 0.4257,
"step": 100
},
{
"epoch": 0.07596685082872928,
"grad_norm": 0.43746134638786316,
"learning_rate": 1.09e-05,
"loss": 0.3877,
"step": 110
},
{
"epoch": 0.08287292817679558,
"grad_norm": 0.3833974003791809,
"learning_rate": 1.19e-05,
"loss": 0.3288,
"step": 120
},
{
"epoch": 0.08977900552486189,
"grad_norm": 0.4290325939655304,
"learning_rate": 1.29e-05,
"loss": 0.3196,
"step": 130
},
{
"epoch": 0.09668508287292818,
"grad_norm": 0.4357810318470001,
"learning_rate": 1.3900000000000002e-05,
"loss": 0.3117,
"step": 140
},
{
"epoch": 0.10359116022099447,
"grad_norm": 0.33745425939559937,
"learning_rate": 1.49e-05,
"loss": 0.2888,
"step": 150
},
{
"epoch": 0.11049723756906077,
"grad_norm": 0.3580892086029053,
"learning_rate": 1.59e-05,
"loss": 0.2803,
"step": 160
},
{
"epoch": 0.11740331491712708,
"grad_norm": 0.30870571732521057,
"learning_rate": 1.69e-05,
"loss": 0.2556,
"step": 170
},
{
"epoch": 0.12430939226519337,
"grad_norm": 0.4122651219367981,
"learning_rate": 1.79e-05,
"loss": 0.253,
"step": 180
},
{
"epoch": 0.13121546961325967,
"grad_norm": 0.46398815512657166,
"learning_rate": 1.8900000000000002e-05,
"loss": 0.2369,
"step": 190
},
{
"epoch": 0.13812154696132597,
"grad_norm": 0.497284859418869,
"learning_rate": 1.9900000000000003e-05,
"loss": 0.2358,
"step": 200
},
{
"epoch": 0.14502762430939226,
"grad_norm": 0.28150540590286255,
"learning_rate": 2.09e-05,
"loss": 0.2371,
"step": 210
},
{
"epoch": 0.15193370165745856,
"grad_norm": 0.3314034938812256,
"learning_rate": 2.19e-05,
"loss": 0.2405,
"step": 220
},
{
"epoch": 0.15883977900552487,
"grad_norm": 0.4119907021522522,
"learning_rate": 2.29e-05,
"loss": 0.225,
"step": 230
},
{
"epoch": 0.16574585635359115,
"grad_norm": 0.5320000648498535,
"learning_rate": 2.39e-05,
"loss": 0.2165,
"step": 240
},
{
"epoch": 0.17265193370165746,
"grad_norm": 0.7389222979545593,
"learning_rate": 2.4900000000000002e-05,
"loss": 0.2045,
"step": 250
},
{
"epoch": 0.17955801104972377,
"grad_norm": 0.3889163136482239,
"learning_rate": 2.5900000000000003e-05,
"loss": 0.2135,
"step": 260
},
{
"epoch": 0.18646408839779005,
"grad_norm": 0.3049943149089813,
"learning_rate": 2.6900000000000003e-05,
"loss": 0.1976,
"step": 270
},
{
"epoch": 0.19337016574585636,
"grad_norm": 0.3817075490951538,
"learning_rate": 2.7900000000000004e-05,
"loss": 0.1823,
"step": 280
},
{
"epoch": 0.20027624309392264,
"grad_norm": 0.605591356754303,
"learning_rate": 2.8899999999999998e-05,
"loss": 0.1997,
"step": 290
},
{
"epoch": 0.20718232044198895,
"grad_norm": 0.36424922943115234,
"learning_rate": 2.9900000000000002e-05,
"loss": 0.18,
"step": 300
},
{
"epoch": 0.21408839779005526,
"grad_norm": 0.3346197307109833,
"learning_rate": 3.09e-05,
"loss": 0.1868,
"step": 310
},
{
"epoch": 0.22099447513812154,
"grad_norm": 0.559263288974762,
"learning_rate": 3.19e-05,
"loss": 0.185,
"step": 320
},
{
"epoch": 0.22790055248618785,
"grad_norm": 0.6093797087669373,
"learning_rate": 3.29e-05,
"loss": 0.1803,
"step": 330
},
{
"epoch": 0.23480662983425415,
"grad_norm": 0.3911588191986084,
"learning_rate": 3.3900000000000004e-05,
"loss": 0.1694,
"step": 340
},
{
"epoch": 0.24171270718232044,
"grad_norm": 0.47210046648979187,
"learning_rate": 3.49e-05,
"loss": 0.17,
"step": 350
},
{
"epoch": 0.24861878453038674,
"grad_norm": 0.7929732799530029,
"learning_rate": 3.59e-05,
"loss": 0.1754,
"step": 360
},
{
"epoch": 0.255524861878453,
"grad_norm": 0.5116897225379944,
"learning_rate": 3.69e-05,
"loss": 0.1679,
"step": 370
},
{
"epoch": 0.26243093922651933,
"grad_norm": 0.4088026285171509,
"learning_rate": 3.79e-05,
"loss": 0.1512,
"step": 380
},
{
"epoch": 0.26933701657458564,
"grad_norm": 0.44423434138298035,
"learning_rate": 3.8900000000000004e-05,
"loss": 0.1454,
"step": 390
},
{
"epoch": 0.27624309392265195,
"grad_norm": 0.4872801899909973,
"learning_rate": 3.99e-05,
"loss": 0.1506,
"step": 400
},
{
"epoch": 0.28314917127071826,
"grad_norm": 0.5234099626541138,
"learning_rate": 4.09e-05,
"loss": 0.148,
"step": 410
},
{
"epoch": 0.2900552486187845,
"grad_norm": 0.4611225724220276,
"learning_rate": 4.19e-05,
"loss": 0.1454,
"step": 420
},
{
"epoch": 0.2969613259668508,
"grad_norm": 0.42924195528030396,
"learning_rate": 4.29e-05,
"loss": 0.1442,
"step": 430
},
{
"epoch": 0.30386740331491713,
"grad_norm": 0.6093170642852783,
"learning_rate": 4.39e-05,
"loss": 0.1345,
"step": 440
},
{
"epoch": 0.31077348066298344,
"grad_norm": 0.403637170791626,
"learning_rate": 4.49e-05,
"loss": 0.1411,
"step": 450
},
{
"epoch": 0.31767955801104975,
"grad_norm": 0.48455968499183655,
"learning_rate": 4.5900000000000004e-05,
"loss": 0.1426,
"step": 460
},
{
"epoch": 0.324585635359116,
"grad_norm": 0.8913393616676331,
"learning_rate": 4.69e-05,
"loss": 0.1481,
"step": 470
},
{
"epoch": 0.3314917127071823,
"grad_norm": 0.48442342877388,
"learning_rate": 4.79e-05,
"loss": 0.1277,
"step": 480
},
{
"epoch": 0.3383977900552486,
"grad_norm": 0.5443485379219055,
"learning_rate": 4.89e-05,
"loss": 0.1337,
"step": 490
},
{
"epoch": 0.3453038674033149,
"grad_norm": 0.5175359845161438,
"learning_rate": 4.99e-05,
"loss": 0.122,
"step": 500
},
{
"epoch": 0.35220994475138123,
"grad_norm": 0.5098366141319275,
"learning_rate": 5.0900000000000004e-05,
"loss": 0.1304,
"step": 510
},
{
"epoch": 0.35911602209944754,
"grad_norm": 0.36380815505981445,
"learning_rate": 5.19e-05,
"loss": 0.1314,
"step": 520
},
{
"epoch": 0.3660220994475138,
"grad_norm": 0.5042738318443298,
"learning_rate": 5.2900000000000005e-05,
"loss": 0.128,
"step": 530
},
{
"epoch": 0.3729281767955801,
"grad_norm": 0.4525247812271118,
"learning_rate": 5.390000000000001e-05,
"loss": 0.1205,
"step": 540
},
{
"epoch": 0.3798342541436464,
"grad_norm": 0.5141445398330688,
"learning_rate": 5.4900000000000006e-05,
"loss": 0.1265,
"step": 550
},
{
"epoch": 0.3867403314917127,
"grad_norm": 0.4617238938808441,
"learning_rate": 5.590000000000001e-05,
"loss": 0.1219,
"step": 560
},
{
"epoch": 0.393646408839779,
"grad_norm": 0.7641685605049133,
"learning_rate": 5.69e-05,
"loss": 0.1315,
"step": 570
},
{
"epoch": 0.4005524861878453,
"grad_norm": 0.3960668742656708,
"learning_rate": 5.79e-05,
"loss": 0.1272,
"step": 580
},
{
"epoch": 0.4074585635359116,
"grad_norm": 0.35828521847724915,
"learning_rate": 5.89e-05,
"loss": 0.128,
"step": 590
},
{
"epoch": 0.4143646408839779,
"grad_norm": 0.42251715064048767,
"learning_rate": 5.99e-05,
"loss": 0.1267,
"step": 600
},
{
"epoch": 0.4212707182320442,
"grad_norm": 0.5093334913253784,
"learning_rate": 6.09e-05,
"loss": 0.1359,
"step": 610
},
{
"epoch": 0.4281767955801105,
"grad_norm": 0.7192143797874451,
"learning_rate": 6.19e-05,
"loss": 0.1205,
"step": 620
},
{
"epoch": 0.4350828729281768,
"grad_norm": 0.8674402236938477,
"learning_rate": 6.29e-05,
"loss": 0.1156,
"step": 630
},
{
"epoch": 0.4419889502762431,
"grad_norm": 0.31834760308265686,
"learning_rate": 6.390000000000001e-05,
"loss": 0.1116,
"step": 640
},
{
"epoch": 0.4488950276243094,
"grad_norm": 0.3638267517089844,
"learning_rate": 6.49e-05,
"loss": 0.1142,
"step": 650
},
{
"epoch": 0.4558011049723757,
"grad_norm": 0.48287007212638855,
"learning_rate": 6.59e-05,
"loss": 0.1179,
"step": 660
},
{
"epoch": 0.462707182320442,
"grad_norm": 0.42799556255340576,
"learning_rate": 6.690000000000001e-05,
"loss": 0.1155,
"step": 670
},
{
"epoch": 0.4696132596685083,
"grad_norm": 0.5267932415008545,
"learning_rate": 6.790000000000001e-05,
"loss": 0.1122,
"step": 680
},
{
"epoch": 0.47651933701657456,
"grad_norm": 0.3648410737514496,
"learning_rate": 6.89e-05,
"loss": 0.1119,
"step": 690
},
{
"epoch": 0.48342541436464087,
"grad_norm": 0.40923115611076355,
"learning_rate": 6.99e-05,
"loss": 0.1155,
"step": 700
},
{
"epoch": 0.4903314917127072,
"grad_norm": 0.43576744198799133,
"learning_rate": 7.09e-05,
"loss": 0.1262,
"step": 710
},
{
"epoch": 0.4972375690607735,
"grad_norm": 0.5781348347663879,
"learning_rate": 7.19e-05,
"loss": 0.1148,
"step": 720
},
{
"epoch": 0.5041436464088398,
"grad_norm": 0.4100674092769623,
"learning_rate": 7.29e-05,
"loss": 0.1098,
"step": 730
},
{
"epoch": 0.511049723756906,
"grad_norm": 0.3947247266769409,
"learning_rate": 7.390000000000001e-05,
"loss": 0.1022,
"step": 740
},
{
"epoch": 0.5179558011049724,
"grad_norm": 0.437826007604599,
"learning_rate": 7.49e-05,
"loss": 0.1135,
"step": 750
},
{
"epoch": 0.5248618784530387,
"grad_norm": 0.362348735332489,
"learning_rate": 7.59e-05,
"loss": 0.1009,
"step": 760
},
{
"epoch": 0.5317679558011049,
"grad_norm": 0.5351320505142212,
"learning_rate": 7.69e-05,
"loss": 0.1088,
"step": 770
},
{
"epoch": 0.5386740331491713,
"grad_norm": 0.5323932766914368,
"learning_rate": 7.790000000000001e-05,
"loss": 0.1033,
"step": 780
},
{
"epoch": 0.5455801104972375,
"grad_norm": 0.4217228889465332,
"learning_rate": 7.890000000000001e-05,
"loss": 0.1172,
"step": 790
},
{
"epoch": 0.5524861878453039,
"grad_norm": 0.43072474002838135,
"learning_rate": 7.99e-05,
"loss": 0.1109,
"step": 800
},
{
"epoch": 0.5593922651933702,
"grad_norm": 0.34974709153175354,
"learning_rate": 8.090000000000001e-05,
"loss": 0.1181,
"step": 810
},
{
"epoch": 0.5662983425414365,
"grad_norm": 0.35198482871055603,
"learning_rate": 8.19e-05,
"loss": 0.1067,
"step": 820
},
{
"epoch": 0.5732044198895028,
"grad_norm": 0.36193737387657166,
"learning_rate": 8.29e-05,
"loss": 0.1066,
"step": 830
},
{
"epoch": 0.580110497237569,
"grad_norm": 0.5137208104133606,
"learning_rate": 8.39e-05,
"loss": 0.1008,
"step": 840
},
{
"epoch": 0.5870165745856354,
"grad_norm": 0.4571826457977295,
"learning_rate": 8.49e-05,
"loss": 0.1187,
"step": 850
},
{
"epoch": 0.5939226519337016,
"grad_norm": 0.4294939339160919,
"learning_rate": 8.59e-05,
"loss": 0.1051,
"step": 860
},
{
"epoch": 0.600828729281768,
"grad_norm": 0.5359936952590942,
"learning_rate": 8.69e-05,
"loss": 0.1094,
"step": 870
},
{
"epoch": 0.6077348066298343,
"grad_norm": 0.31322285532951355,
"learning_rate": 8.790000000000001e-05,
"loss": 0.1003,
"step": 880
},
{
"epoch": 0.6146408839779005,
"grad_norm": 0.27965644001960754,
"learning_rate": 8.89e-05,
"loss": 0.0963,
"step": 890
},
{
"epoch": 0.6215469613259669,
"grad_norm": 0.2999873757362366,
"learning_rate": 8.99e-05,
"loss": 0.1014,
"step": 900
},
{
"epoch": 0.6284530386740331,
"grad_norm": 0.47803860902786255,
"learning_rate": 9.090000000000001e-05,
"loss": 0.0926,
"step": 910
},
{
"epoch": 0.6353591160220995,
"grad_norm": 0.44106921553611755,
"learning_rate": 9.190000000000001e-05,
"loss": 0.1029,
"step": 920
},
{
"epoch": 0.6422651933701657,
"grad_norm": 0.3900652229785919,
"learning_rate": 9.290000000000001e-05,
"loss": 0.0997,
"step": 930
},
{
"epoch": 0.649171270718232,
"grad_norm": 0.4396621286869049,
"learning_rate": 9.39e-05,
"loss": 0.1,
"step": 940
},
{
"epoch": 0.6560773480662984,
"grad_norm": 0.43293172121047974,
"learning_rate": 9.49e-05,
"loss": 0.1092,
"step": 950
},
{
"epoch": 0.6629834254143646,
"grad_norm": 0.4162095785140991,
"learning_rate": 9.59e-05,
"loss": 0.0911,
"step": 960
},
{
"epoch": 0.669889502762431,
"grad_norm": 0.385890930891037,
"learning_rate": 9.69e-05,
"loss": 0.0984,
"step": 970
},
{
"epoch": 0.6767955801104972,
"grad_norm": 0.36000925302505493,
"learning_rate": 9.790000000000001e-05,
"loss": 0.0941,
"step": 980
},
{
"epoch": 0.6837016574585635,
"grad_norm": 0.28026294708251953,
"learning_rate": 9.89e-05,
"loss": 0.1002,
"step": 990
},
{
"epoch": 0.6906077348066298,
"grad_norm": 0.3528141379356384,
"learning_rate": 9.99e-05,
"loss": 0.0872,
"step": 1000
},
{
"epoch": 0.6975138121546961,
"grad_norm": 0.500379741191864,
"learning_rate": 9.999994463727085e-05,
"loss": 0.0938,
"step": 1010
},
{
"epoch": 0.7044198895027625,
"grad_norm": 0.42596903443336487,
"learning_rate": 9.999975326009292e-05,
"loss": 0.0959,
"step": 1020
},
{
"epoch": 0.7113259668508287,
"grad_norm": 0.2438584268093109,
"learning_rate": 9.999942518549879e-05,
"loss": 0.0963,
"step": 1030
},
{
"epoch": 0.7182320441988951,
"grad_norm": 0.3988669514656067,
"learning_rate": 9.999896041438544e-05,
"loss": 0.0932,
"step": 1040
},
{
"epoch": 0.7251381215469613,
"grad_norm": 0.28523698449134827,
"learning_rate": 9.999835894802353e-05,
"loss": 0.0888,
"step": 1050
},
{
"epoch": 0.7320441988950276,
"grad_norm": 0.511328399181366,
"learning_rate": 9.999762078805743e-05,
"loss": 0.104,
"step": 1060
},
{
"epoch": 0.738950276243094,
"grad_norm": 0.2832479476928711,
"learning_rate": 9.999674593650526e-05,
"loss": 0.0905,
"step": 1070
},
{
"epoch": 0.7458563535911602,
"grad_norm": 0.32691365480422974,
"learning_rate": 9.99957343957588e-05,
"loss": 0.0978,
"step": 1080
},
{
"epoch": 0.7527624309392266,
"grad_norm": 0.349331259727478,
"learning_rate": 9.99945861685836e-05,
"loss": 0.0911,
"step": 1090
},
{
"epoch": 0.7596685082872928,
"grad_norm": 0.4922757148742676,
"learning_rate": 9.999330125811884e-05,
"loss": 0.0881,
"step": 1100
},
{
"epoch": 0.7665745856353591,
"grad_norm": 0.26750776171684265,
"learning_rate": 9.999187966787744e-05,
"loss": 0.0856,
"step": 1110
},
{
"epoch": 0.7734806629834254,
"grad_norm": 0.298186719417572,
"learning_rate": 9.999032140174595e-05,
"loss": 0.0875,
"step": 1120
},
{
"epoch": 0.7803867403314917,
"grad_norm": 0.3135933578014374,
"learning_rate": 9.998862646398464e-05,
"loss": 0.0935,
"step": 1130
},
{
"epoch": 0.787292817679558,
"grad_norm": 0.37238550186157227,
"learning_rate": 9.998679485922739e-05,
"loss": 0.0975,
"step": 1140
},
{
"epoch": 0.7941988950276243,
"grad_norm": 0.3704855442047119,
"learning_rate": 9.998482659248174e-05,
"loss": 0.0886,
"step": 1150
},
{
"epoch": 0.8011049723756906,
"grad_norm": 0.3354080617427826,
"learning_rate": 9.998272166912883e-05,
"loss": 0.0923,
"step": 1160
},
{
"epoch": 0.8080110497237569,
"grad_norm": 0.2832203209400177,
"learning_rate": 9.998048009492347e-05,
"loss": 0.0905,
"step": 1170
},
{
"epoch": 0.8149171270718232,
"grad_norm": 0.3106226921081543,
"learning_rate": 9.997810187599403e-05,
"loss": 0.0874,
"step": 1180
},
{
"epoch": 0.8218232044198895,
"grad_norm": 0.26605871319770813,
"learning_rate": 9.997558701884249e-05,
"loss": 0.0883,
"step": 1190
},
{
"epoch": 0.8287292817679558,
"grad_norm": 0.23787857592105865,
"learning_rate": 9.997293553034433e-05,
"loss": 0.0851,
"step": 1200
},
{
"epoch": 0.835635359116022,
"grad_norm": 0.2726190686225891,
"learning_rate": 9.997014741774866e-05,
"loss": 0.0912,
"step": 1210
},
{
"epoch": 0.8425414364640884,
"grad_norm": 0.3378348648548126,
"learning_rate": 9.996722268867803e-05,
"loss": 0.0837,
"step": 1220
},
{
"epoch": 0.8494475138121547,
"grad_norm": 0.2348347008228302,
"learning_rate": 9.996416135112858e-05,
"loss": 0.095,
"step": 1230
},
{
"epoch": 0.856353591160221,
"grad_norm": 0.29881569743156433,
"learning_rate": 9.996096341346988e-05,
"loss": 0.0807,
"step": 1240
},
{
"epoch": 0.8632596685082873,
"grad_norm": 0.278564989566803,
"learning_rate": 9.995762888444495e-05,
"loss": 0.0941,
"step": 1250
},
{
"epoch": 0.8701657458563536,
"grad_norm": 0.26641955971717834,
"learning_rate": 9.995415777317027e-05,
"loss": 0.0813,
"step": 1260
},
{
"epoch": 0.8770718232044199,
"grad_norm": 0.2798615097999573,
"learning_rate": 9.995055008913574e-05,
"loss": 0.0885,
"step": 1270
},
{
"epoch": 0.8839779005524862,
"grad_norm": 0.35161152482032776,
"learning_rate": 9.994680584220463e-05,
"loss": 0.0929,
"step": 1280
},
{
"epoch": 0.8908839779005525,
"grad_norm": 0.3145495653152466,
"learning_rate": 9.994292504261355e-05,
"loss": 0.0903,
"step": 1290
},
{
"epoch": 0.8977900552486188,
"grad_norm": 0.2972181439399719,
"learning_rate": 9.993890770097247e-05,
"loss": 0.0806,
"step": 1300
},
{
"epoch": 0.9046961325966851,
"grad_norm": 0.45706817507743835,
"learning_rate": 9.993475382826467e-05,
"loss": 0.088,
"step": 1310
},
{
"epoch": 0.9116022099447514,
"grad_norm": 0.35890793800354004,
"learning_rate": 9.993046343584664e-05,
"loss": 0.0869,
"step": 1320
},
{
"epoch": 0.9185082872928176,
"grad_norm": 0.3728267252445221,
"learning_rate": 9.992603653544816e-05,
"loss": 0.077,
"step": 1330
},
{
"epoch": 0.925414364640884,
"grad_norm": 0.3744541108608246,
"learning_rate": 9.992147313917222e-05,
"loss": 0.0818,
"step": 1340
},
{
"epoch": 0.9323204419889503,
"grad_norm": 0.3192994296550751,
"learning_rate": 9.991677325949497e-05,
"loss": 0.0793,
"step": 1350
},
{
"epoch": 0.9392265193370166,
"grad_norm": 0.31171876192092896,
"learning_rate": 9.991193690926568e-05,
"loss": 0.0832,
"step": 1360
},
{
"epoch": 0.9461325966850829,
"grad_norm": 0.24060353636741638,
"learning_rate": 9.990696410170678e-05,
"loss": 0.0721,
"step": 1370
},
{
"epoch": 0.9530386740331491,
"grad_norm": 0.2516084909439087,
"learning_rate": 9.990185485041371e-05,
"loss": 0.0782,
"step": 1380
},
{
"epoch": 0.9599447513812155,
"grad_norm": 0.2999783158302307,
"learning_rate": 9.989660916935498e-05,
"loss": 0.0724,
"step": 1390
},
{
"epoch": 0.9668508287292817,
"grad_norm": 0.3542238473892212,
"learning_rate": 9.989122707287208e-05,
"loss": 0.0844,
"step": 1400
},
{
"epoch": 0.9737569060773481,
"grad_norm": 0.3433320224285126,
"learning_rate": 9.988570857567945e-05,
"loss": 0.0772,
"step": 1410
},
{
"epoch": 0.9806629834254144,
"grad_norm": 0.3041175603866577,
"learning_rate": 9.988005369286446e-05,
"loss": 0.0891,
"step": 1420
},
{
"epoch": 0.9875690607734806,
"grad_norm": 0.2909347712993622,
"learning_rate": 9.987426243988734e-05,
"loss": 0.079,
"step": 1430
},
{
"epoch": 0.994475138121547,
"grad_norm": 0.4544817805290222,
"learning_rate": 9.986833483258114e-05,
"loss": 0.0838,
"step": 1440
},
{
"epoch": 1.0013812154696133,
"grad_norm": 0.2458307147026062,
"learning_rate": 9.986227088715173e-05,
"loss": 0.0803,
"step": 1450
},
{
"epoch": 1.0082872928176796,
"grad_norm": 0.4468945562839508,
"learning_rate": 9.98560706201777e-05,
"loss": 0.0841,
"step": 1460
},
{
"epoch": 1.0151933701657458,
"grad_norm": 0.3525172472000122,
"learning_rate": 9.984973404861036e-05,
"loss": 0.0851,
"step": 1470
},
{
"epoch": 1.022099447513812,
"grad_norm": 0.45504671335220337,
"learning_rate": 9.984326118977361e-05,
"loss": 0.0823,
"step": 1480
},
{
"epoch": 1.0290055248618784,
"grad_norm": 0.28854405879974365,
"learning_rate": 9.983665206136406e-05,
"loss": 0.0798,
"step": 1490
},
{
"epoch": 1.0359116022099448,
"grad_norm": 0.317660927772522,
"learning_rate": 9.982990668145075e-05,
"loss": 0.0849,
"step": 1500
},
{
"epoch": 1.042817679558011,
"grad_norm": 0.37529265880584717,
"learning_rate": 9.982302506847534e-05,
"loss": 0.0771,
"step": 1510
},
{
"epoch": 1.0497237569060773,
"grad_norm": 0.34826648235321045,
"learning_rate": 9.981600724125189e-05,
"loss": 0.0779,
"step": 1520
},
{
"epoch": 1.0566298342541436,
"grad_norm": 0.2414906919002533,
"learning_rate": 9.980885321896685e-05,
"loss": 0.079,
"step": 1530
},
{
"epoch": 1.06353591160221,
"grad_norm": 0.23677992820739746,
"learning_rate": 9.980156302117905e-05,
"loss": 0.0781,
"step": 1540
},
{
"epoch": 1.0704419889502763,
"grad_norm": 0.3170182704925537,
"learning_rate": 9.979413666781963e-05,
"loss": 0.0767,
"step": 1550
},
{
"epoch": 1.0773480662983426,
"grad_norm": 0.18258549273014069,
"learning_rate": 9.978657417919193e-05,
"loss": 0.0747,
"step": 1560
},
{
"epoch": 1.0842541436464088,
"grad_norm": 0.2986181378364563,
"learning_rate": 9.977887557597153e-05,
"loss": 0.0688,
"step": 1570
},
{
"epoch": 1.091160220994475,
"grad_norm": 0.3077967166900635,
"learning_rate": 9.97710408792061e-05,
"loss": 0.0732,
"step": 1580
},
{
"epoch": 1.0980662983425415,
"grad_norm": 0.2392505258321762,
"learning_rate": 9.976307011031542e-05,
"loss": 0.0741,
"step": 1590
},
{
"epoch": 1.1049723756906078,
"grad_norm": 0.40254053473472595,
"learning_rate": 9.975496329109126e-05,
"loss": 0.0773,
"step": 1600
},
{
"epoch": 1.111878453038674,
"grad_norm": 0.2991298735141754,
"learning_rate": 9.974672044369732e-05,
"loss": 0.0673,
"step": 1610
},
{
"epoch": 1.1187845303867403,
"grad_norm": 0.2567093074321747,
"learning_rate": 9.97383415906693e-05,
"loss": 0.0661,
"step": 1620
},
{
"epoch": 1.1256906077348066,
"grad_norm": 0.250507652759552,
"learning_rate": 9.97298267549146e-05,
"loss": 0.0696,
"step": 1630
},
{
"epoch": 1.132596685082873,
"grad_norm": 0.35195478796958923,
"learning_rate": 9.972117595971249e-05,
"loss": 0.0754,
"step": 1640
},
{
"epoch": 1.1395027624309393,
"grad_norm": 0.27800172567367554,
"learning_rate": 9.971238922871391e-05,
"loss": 0.0821,
"step": 1650
},
{
"epoch": 1.1464088397790055,
"grad_norm": 0.17888322472572327,
"learning_rate": 9.970346658594142e-05,
"loss": 0.0699,
"step": 1660
},
{
"epoch": 1.1533149171270718,
"grad_norm": 0.19279469549655914,
"learning_rate": 9.969440805578923e-05,
"loss": 0.0712,
"step": 1670
},
{
"epoch": 1.160220994475138,
"grad_norm": 0.3412877917289734,
"learning_rate": 9.968521366302298e-05,
"loss": 0.0764,
"step": 1680
},
{
"epoch": 1.1671270718232045,
"grad_norm": 0.3297848403453827,
"learning_rate": 9.967588343277981e-05,
"loss": 0.0758,
"step": 1690
},
{
"epoch": 1.1740331491712708,
"grad_norm": 0.3466518521308899,
"learning_rate": 9.966641739056818e-05,
"loss": 0.0765,
"step": 1700
},
{
"epoch": 1.180939226519337,
"grad_norm": 0.2665199637413025,
"learning_rate": 9.965681556226793e-05,
"loss": 0.076,
"step": 1710
},
{
"epoch": 1.1878453038674033,
"grad_norm": 0.28554102778434753,
"learning_rate": 9.964707797413006e-05,
"loss": 0.078,
"step": 1720
},
{
"epoch": 1.1947513812154695,
"grad_norm": 0.256296843290329,
"learning_rate": 9.963720465277679e-05,
"loss": 0.0744,
"step": 1730
},
{
"epoch": 1.201657458563536,
"grad_norm": 0.42945459485054016,
"learning_rate": 9.96271956252014e-05,
"loss": 0.0759,
"step": 1740
},
{
"epoch": 1.2085635359116023,
"grad_norm": 0.29390648007392883,
"learning_rate": 9.961705091876816e-05,
"loss": 0.0736,
"step": 1750
},
{
"epoch": 1.2154696132596685,
"grad_norm": 0.1821282058954239,
"learning_rate": 9.960677056121235e-05,
"loss": 0.0725,
"step": 1760
},
{
"epoch": 1.2223756906077348,
"grad_norm": 0.29151904582977295,
"learning_rate": 9.959635458064005e-05,
"loss": 0.0715,
"step": 1770
},
{
"epoch": 1.229281767955801,
"grad_norm": 0.1900283247232437,
"learning_rate": 9.958580300552815e-05,
"loss": 0.0696,
"step": 1780
},
{
"epoch": 1.2361878453038675,
"grad_norm": 0.19765739142894745,
"learning_rate": 9.957511586472426e-05,
"loss": 0.0755,
"step": 1790
},
{
"epoch": 1.2430939226519337,
"grad_norm": 0.3573096990585327,
"learning_rate": 9.956429318744662e-05,
"loss": 0.0754,
"step": 1800
},
{
"epoch": 1.25,
"grad_norm": 0.36740875244140625,
"learning_rate": 9.955333500328404e-05,
"loss": 0.0706,
"step": 1810
},
{
"epoch": 1.2569060773480663,
"grad_norm": 0.5108208060264587,
"learning_rate": 9.95422413421957e-05,
"loss": 0.0708,
"step": 1820
},
{
"epoch": 1.2638121546961325,
"grad_norm": 0.3641519546508789,
"learning_rate": 9.953101223451133e-05,
"loss": 0.0709,
"step": 1830
},
{
"epoch": 1.270718232044199,
"grad_norm": 0.27009454369544983,
"learning_rate": 9.951964771093085e-05,
"loss": 0.0702,
"step": 1840
},
{
"epoch": 1.2776243093922652,
"grad_norm": 0.2130231261253357,
"learning_rate": 9.950814780252442e-05,
"loss": 0.0639,
"step": 1850
},
{
"epoch": 1.2845303867403315,
"grad_norm": 0.286380797624588,
"learning_rate": 9.949651254073236e-05,
"loss": 0.0764,
"step": 1860
},
{
"epoch": 1.2914364640883977,
"grad_norm": 0.309061735868454,
"learning_rate": 9.948474195736504e-05,
"loss": 0.0689,
"step": 1870
},
{
"epoch": 1.298342541436464,
"grad_norm": 0.3645918667316437,
"learning_rate": 9.947283608460277e-05,
"loss": 0.0716,
"step": 1880
},
{
"epoch": 1.3052486187845305,
"grad_norm": 0.30466482043266296,
"learning_rate": 9.946079495499577e-05,
"loss": 0.0658,
"step": 1890
},
{
"epoch": 1.3121546961325967,
"grad_norm": 0.24554474651813507,
"learning_rate": 9.944861860146401e-05,
"loss": 0.0723,
"step": 1900
},
{
"epoch": 1.319060773480663,
"grad_norm": 0.22281047701835632,
"learning_rate": 9.943630705729719e-05,
"loss": 0.0743,
"step": 1910
},
{
"epoch": 1.3259668508287292,
"grad_norm": 0.22427059710025787,
"learning_rate": 9.942386035615459e-05,
"loss": 0.0688,
"step": 1920
},
{
"epoch": 1.3328729281767955,
"grad_norm": 0.3036755621433258,
"learning_rate": 9.941127853206503e-05,
"loss": 0.0707,
"step": 1930
},
{
"epoch": 1.339779005524862,
"grad_norm": 0.23490217328071594,
"learning_rate": 9.939856161942673e-05,
"loss": 0.0701,
"step": 1940
},
{
"epoch": 1.3466850828729282,
"grad_norm": 0.2540903389453888,
"learning_rate": 9.938570965300724e-05,
"loss": 0.0668,
"step": 1950
},
{
"epoch": 1.3535911602209945,
"grad_norm": 0.351299911737442,
"learning_rate": 9.937272266794335e-05,
"loss": 0.0603,
"step": 1960
},
{
"epoch": 1.3604972375690607,
"grad_norm": 0.30575865507125854,
"learning_rate": 9.935960069974096e-05,
"loss": 0.0735,
"step": 1970
},
{
"epoch": 1.367403314917127,
"grad_norm": 0.2758401930332184,
"learning_rate": 9.934634378427506e-05,
"loss": 0.0723,
"step": 1980
},
{
"epoch": 1.3743093922651934,
"grad_norm": 0.42013078927993774,
"learning_rate": 9.933295195778954e-05,
"loss": 0.0671,
"step": 1990
},
{
"epoch": 1.3812154696132597,
"grad_norm": 0.30824539065361023,
"learning_rate": 9.931942525689715e-05,
"loss": 0.0636,
"step": 2000
},
{
"epoch": 1.388121546961326,
"grad_norm": 0.2764925956726074,
"learning_rate": 9.930576371857936e-05,
"loss": 0.0655,
"step": 2010
},
{
"epoch": 1.3950276243093922,
"grad_norm": 0.25183069705963135,
"learning_rate": 9.929196738018629e-05,
"loss": 0.0655,
"step": 2020
},
{
"epoch": 1.4019337016574585,
"grad_norm": 0.18897801637649536,
"learning_rate": 9.927803627943662e-05,
"loss": 0.068,
"step": 2030
},
{
"epoch": 1.408839779005525,
"grad_norm": 0.4160577356815338,
"learning_rate": 9.926397045441744e-05,
"loss": 0.0731,
"step": 2040
},
{
"epoch": 1.4157458563535912,
"grad_norm": 0.25782227516174316,
"learning_rate": 9.924976994358417e-05,
"loss": 0.065,
"step": 2050
},
{
"epoch": 1.4226519337016574,
"grad_norm": 0.42086145281791687,
"learning_rate": 9.923543478576048e-05,
"loss": 0.0637,
"step": 2060
},
{
"epoch": 1.4295580110497237,
"grad_norm": 0.25658249855041504,
"learning_rate": 9.922096502013813e-05,
"loss": 0.0609,
"step": 2070
},
{
"epoch": 1.43646408839779,
"grad_norm": 0.31206923723220825,
"learning_rate": 9.92063606862769e-05,
"loss": 0.0671,
"step": 2080
},
{
"epoch": 1.4433701657458564,
"grad_norm": 0.2223379760980606,
"learning_rate": 9.919162182410453e-05,
"loss": 0.0604,
"step": 2090
},
{
"epoch": 1.4502762430939227,
"grad_norm": 0.3089164197444916,
"learning_rate": 9.917674847391645e-05,
"loss": 0.0647,
"step": 2100
},
{
"epoch": 1.457182320441989,
"grad_norm": 0.28447818756103516,
"learning_rate": 9.916174067637584e-05,
"loss": 0.0625,
"step": 2110
},
{
"epoch": 1.4640883977900552,
"grad_norm": 0.3222653567790985,
"learning_rate": 9.914659847251348e-05,
"loss": 0.0592,
"step": 2120
},
{
"epoch": 1.4709944751381214,
"grad_norm": 0.17972660064697266,
"learning_rate": 9.913132190372753e-05,
"loss": 0.0671,
"step": 2130
},
{
"epoch": 1.477900552486188,
"grad_norm": 0.3882119357585907,
"learning_rate": 9.911591101178359e-05,
"loss": 0.0633,
"step": 2140
},
{
"epoch": 1.4848066298342542,
"grad_norm": 0.2396843284368515,
"learning_rate": 9.910036583881443e-05,
"loss": 0.0699,
"step": 2150
},
{
"epoch": 1.4917127071823204,
"grad_norm": 0.2941199839115143,
"learning_rate": 9.908468642731995e-05,
"loss": 0.0696,
"step": 2160
},
{
"epoch": 1.4986187845303867,
"grad_norm": 0.28291407227516174,
"learning_rate": 9.906887282016707e-05,
"loss": 0.0654,
"step": 2170
},
{
"epoch": 1.505524861878453,
"grad_norm": 0.21583221852779388,
"learning_rate": 9.90529250605896e-05,
"loss": 0.0638,
"step": 2180
},
{
"epoch": 1.5124309392265194,
"grad_norm": 0.20484983921051025,
"learning_rate": 9.903684319218809e-05,
"loss": 0.0663,
"step": 2190
},
{
"epoch": 1.5193370165745856,
"grad_norm": 0.3546452820301056,
"learning_rate": 9.902062725892976e-05,
"loss": 0.065,
"step": 2200
},
{
"epoch": 1.526243093922652,
"grad_norm": 0.31812551617622375,
"learning_rate": 9.900427730514834e-05,
"loss": 0.058,
"step": 2210
},
{
"epoch": 1.5331491712707184,
"grad_norm": 0.24564820528030396,
"learning_rate": 9.8987793375544e-05,
"loss": 0.0656,
"step": 2220
},
{
"epoch": 1.5400552486187844,
"grad_norm": 0.3509599566459656,
"learning_rate": 9.897117551518318e-05,
"loss": 0.0667,
"step": 2230
},
{
"epoch": 1.5469613259668509,
"grad_norm": 0.1958969682455063,
"learning_rate": 9.895442376949844e-05,
"loss": 0.0627,
"step": 2240
},
{
"epoch": 1.5538674033149171,
"grad_norm": 0.2799903154373169,
"learning_rate": 9.893753818428845e-05,
"loss": 0.0649,
"step": 2250
},
{
"epoch": 1.5607734806629834,
"grad_norm": 0.27472683787345886,
"learning_rate": 9.892051880571773e-05,
"loss": 0.0624,
"step": 2260
},
{
"epoch": 1.5676795580110499,
"grad_norm": 0.3503767251968384,
"learning_rate": 9.890336568031663e-05,
"loss": 0.0618,
"step": 2270
},
{
"epoch": 1.5745856353591159,
"grad_norm": 0.25655266642570496,
"learning_rate": 9.888607885498113e-05,
"loss": 0.0654,
"step": 2280
},
{
"epoch": 1.5814917127071824,
"grad_norm": 0.19729627668857574,
"learning_rate": 9.886865837697275e-05,
"loss": 0.0592,
"step": 2290
},
{
"epoch": 1.5883977900552486,
"grad_norm": 0.46757638454437256,
"learning_rate": 9.88511042939184e-05,
"loss": 0.0591,
"step": 2300
},
{
"epoch": 1.5953038674033149,
"grad_norm": 0.3462216556072235,
"learning_rate": 9.883341665381028e-05,
"loss": 0.0654,
"step": 2310
},
{
"epoch": 1.6022099447513813,
"grad_norm": 0.2920167148113251,
"learning_rate": 9.881559550500575e-05,
"loss": 0.0608,
"step": 2320
},
{
"epoch": 1.6091160220994474,
"grad_norm": 0.22915363311767578,
"learning_rate": 9.879764089622712e-05,
"loss": 0.0705,
"step": 2330
},
{
"epoch": 1.6160220994475138,
"grad_norm": 0.28002291917800903,
"learning_rate": 9.87795528765616e-05,
"loss": 0.0621,
"step": 2340
},
{
"epoch": 1.62292817679558,
"grad_norm": 0.24930529296398163,
"learning_rate": 9.876133149546118e-05,
"loss": 0.0689,
"step": 2350
},
{
"epoch": 1.6298342541436464,
"grad_norm": 0.33754250407218933,
"learning_rate": 9.874297680274238e-05,
"loss": 0.0596,
"step": 2360
},
{
"epoch": 1.6367403314917128,
"grad_norm": 0.3058784306049347,
"learning_rate": 9.872448884858624e-05,
"loss": 0.0614,
"step": 2370
},
{
"epoch": 1.6436464088397789,
"grad_norm": 0.2852232754230499,
"learning_rate": 9.870586768353815e-05,
"loss": 0.0549,
"step": 2380
},
{
"epoch": 1.6505524861878453,
"grad_norm": 0.28618133068084717,
"learning_rate": 9.868711335850764e-05,
"loss": 0.0609,
"step": 2390
},
{
"epoch": 1.6574585635359116,
"grad_norm": 0.19881612062454224,
"learning_rate": 9.866822592476833e-05,
"loss": 0.0635,
"step": 2400
},
{
"epoch": 1.6643646408839778,
"grad_norm": 0.2809147834777832,
"learning_rate": 9.86492054339577e-05,
"loss": 0.058,
"step": 2410
},
{
"epoch": 1.6712707182320443,
"grad_norm": 0.2217167168855667,
"learning_rate": 9.863005193807711e-05,
"loss": 0.061,
"step": 2420
},
{
"epoch": 1.6781767955801103,
"grad_norm": 0.24295663833618164,
"learning_rate": 9.861076548949143e-05,
"loss": 0.0606,
"step": 2430
},
{
"epoch": 1.6850828729281768,
"grad_norm": 0.3120078146457672,
"learning_rate": 9.859134614092912e-05,
"loss": 0.063,
"step": 2440
},
{
"epoch": 1.691988950276243,
"grad_norm": 0.2926420271396637,
"learning_rate": 9.857179394548191e-05,
"loss": 0.0628,
"step": 2450
},
{
"epoch": 1.6988950276243093,
"grad_norm": 0.2582651376724243,
"learning_rate": 9.855210895660477e-05,
"loss": 0.0615,
"step": 2460
},
{
"epoch": 1.7058011049723758,
"grad_norm": 0.18011657893657684,
"learning_rate": 9.853229122811568e-05,
"loss": 0.0558,
"step": 2470
},
{
"epoch": 1.7127071823204418,
"grad_norm": 0.29458683729171753,
"learning_rate": 9.851234081419559e-05,
"loss": 0.059,
"step": 2480
},
{
"epoch": 1.7196132596685083,
"grad_norm": 0.16498754918575287,
"learning_rate": 9.849225776938814e-05,
"loss": 0.053,
"step": 2490
},
{
"epoch": 1.7265193370165746,
"grad_norm": 0.31275665760040283,
"learning_rate": 9.847204214859964e-05,
"loss": 0.0613,
"step": 2500
},
{
"epoch": 1.7334254143646408,
"grad_norm": 0.18819302320480347,
"learning_rate": 9.845169400709879e-05,
"loss": 0.0597,
"step": 2510
},
{
"epoch": 1.7403314917127073,
"grad_norm": 0.278097003698349,
"learning_rate": 9.843121340051664e-05,
"loss": 0.0525,
"step": 2520
},
{
"epoch": 1.7472375690607733,
"grad_norm": 0.28996750712394714,
"learning_rate": 9.841060038484641e-05,
"loss": 0.0574,
"step": 2530
},
{
"epoch": 1.7541436464088398,
"grad_norm": 0.2054147720336914,
"learning_rate": 9.838985501644328e-05,
"loss": 0.0566,
"step": 2540
},
{
"epoch": 1.761049723756906,
"grad_norm": 0.23275215923786163,
"learning_rate": 9.83689773520243e-05,
"loss": 0.0687,
"step": 2550
},
{
"epoch": 1.7679558011049723,
"grad_norm": 0.30844447016716003,
"learning_rate": 9.834796744866819e-05,
"loss": 0.0573,
"step": 2560
},
{
"epoch": 1.7748618784530388,
"grad_norm": 0.22533856332302094,
"learning_rate": 9.832682536381525e-05,
"loss": 0.0612,
"step": 2570
},
{
"epoch": 1.7817679558011048,
"grad_norm": 0.2721378803253174,
"learning_rate": 9.830555115526711e-05,
"loss": 0.055,
"step": 2580
},
{
"epoch": 1.7886740331491713,
"grad_norm": 0.2064887285232544,
"learning_rate": 9.828414488118667e-05,
"loss": 0.0601,
"step": 2590
},
{
"epoch": 1.7955801104972375,
"grad_norm": 0.28619638085365295,
"learning_rate": 9.826260660009785e-05,
"loss": 0.0589,
"step": 2600
},
{
"epoch": 1.8024861878453038,
"grad_norm": 0.14919589459896088,
"learning_rate": 9.824093637088547e-05,
"loss": 0.0538,
"step": 2610
},
{
"epoch": 1.8093922651933703,
"grad_norm": 0.28396356105804443,
"learning_rate": 9.821913425279514e-05,
"loss": 0.0627,
"step": 2620
},
{
"epoch": 1.8162983425414365,
"grad_norm": 0.22441565990447998,
"learning_rate": 9.8197200305433e-05,
"loss": 0.0487,
"step": 2630
},
{
"epoch": 1.8232044198895028,
"grad_norm": 0.18380586802959442,
"learning_rate": 9.817513458876564e-05,
"loss": 0.0546,
"step": 2640
},
{
"epoch": 1.830110497237569,
"grad_norm": 0.24778667092323303,
"learning_rate": 9.815293716311987e-05,
"loss": 0.063,
"step": 2650
},
{
"epoch": 1.8370165745856353,
"grad_norm": 0.22658787667751312,
"learning_rate": 9.813060808918262e-05,
"loss": 0.0665,
"step": 2660
},
{
"epoch": 1.8439226519337018,
"grad_norm": 0.2324257493019104,
"learning_rate": 9.810814742800069e-05,
"loss": 0.0576,
"step": 2670
},
{
"epoch": 1.850828729281768,
"grad_norm": 0.3127438724040985,
"learning_rate": 9.808555524098074e-05,
"loss": 0.0606,
"step": 2680
},
{
"epoch": 1.8577348066298343,
"grad_norm": 0.27944085001945496,
"learning_rate": 9.806283158988887e-05,
"loss": 0.0538,
"step": 2690
},
{
"epoch": 1.8646408839779005,
"grad_norm": 0.2294415533542633,
"learning_rate": 9.803997653685072e-05,
"loss": 0.0608,
"step": 2700
},
{
"epoch": 1.8715469613259668,
"grad_norm": 0.280841201543808,
"learning_rate": 9.801699014435112e-05,
"loss": 0.0614,
"step": 2710
},
{
"epoch": 1.8784530386740332,
"grad_norm": 0.36730295419692993,
"learning_rate": 9.799387247523398e-05,
"loss": 0.0551,
"step": 2720
},
{
"epoch": 1.8853591160220995,
"grad_norm": 0.27677783370018005,
"learning_rate": 9.797062359270215e-05,
"loss": 0.0606,
"step": 2730
},
{
"epoch": 1.8922651933701657,
"grad_norm": 0.23716320097446442,
"learning_rate": 9.794724356031715e-05,
"loss": 0.0591,
"step": 2740
},
{
"epoch": 1.899171270718232,
"grad_norm": 0.22223462164402008,
"learning_rate": 9.792373244199913e-05,
"loss": 0.052,
"step": 2750
},
{
"epoch": 1.9060773480662982,
"grad_norm": 0.24536921083927155,
"learning_rate": 9.790009030202658e-05,
"loss": 0.0564,
"step": 2760
},
{
"epoch": 1.9129834254143647,
"grad_norm": 0.25623899698257446,
"learning_rate": 9.78763172050362e-05,
"loss": 0.0566,
"step": 2770
},
{
"epoch": 1.919889502762431,
"grad_norm": 0.20985721051692963,
"learning_rate": 9.785241321602274e-05,
"loss": 0.0594,
"step": 2780
},
{
"epoch": 1.9267955801104972,
"grad_norm": 0.22064441442489624,
"learning_rate": 9.782837840033879e-05,
"loss": 0.0551,
"step": 2790
},
{
"epoch": 1.9337016574585635,
"grad_norm": 0.19597521424293518,
"learning_rate": 9.780421282369461e-05,
"loss": 0.0541,
"step": 2800
},
{
"epoch": 1.9406077348066297,
"grad_norm": 0.22597678005695343,
"learning_rate": 9.777991655215797e-05,
"loss": 0.0551,
"step": 2810
},
{
"epoch": 1.9475138121546962,
"grad_norm": 0.2016141563653946,
"learning_rate": 9.775548965215394e-05,
"loss": 0.047,
"step": 2820
},
{
"epoch": 1.9544198895027625,
"grad_norm": 0.20736120641231537,
"learning_rate": 9.773093219046474e-05,
"loss": 0.0581,
"step": 2830
},
{
"epoch": 1.9613259668508287,
"grad_norm": 0.2478555291891098,
"learning_rate": 9.770624423422954e-05,
"loss": 0.0574,
"step": 2840
},
{
"epoch": 1.9682320441988952,
"grad_norm": 0.2794519066810608,
"learning_rate": 9.768142585094426e-05,
"loss": 0.0526,
"step": 2850
},
{
"epoch": 1.9751381215469612,
"grad_norm": 0.24204091727733612,
"learning_rate": 9.765647710846142e-05,
"loss": 0.0562,
"step": 2860
},
{
"epoch": 1.9820441988950277,
"grad_norm": 0.2637561857700348,
"learning_rate": 9.763139807498991e-05,
"loss": 0.0565,
"step": 2870
},
{
"epoch": 1.988950276243094,
"grad_norm": 0.18525560200214386,
"learning_rate": 9.760618881909487e-05,
"loss": 0.0551,
"step": 2880
},
{
"epoch": 1.9958563535911602,
"grad_norm": 0.18425393104553223,
"learning_rate": 9.758084940969744e-05,
"loss": 0.0536,
"step": 2890
},
{
"epoch": 2.0027624309392267,
"grad_norm": 0.15110640227794647,
"learning_rate": 9.755537991607459e-05,
"loss": 0.0589,
"step": 2900
},
{
"epoch": 2.0096685082872927,
"grad_norm": 0.2911590039730072,
"learning_rate": 9.752978040785895e-05,
"loss": 0.0641,
"step": 2910
},
{
"epoch": 2.016574585635359,
"grad_norm": 0.4290618300437927,
"learning_rate": 9.750405095503859e-05,
"loss": 0.0574,
"step": 2920
},
{
"epoch": 2.023480662983425,
"grad_norm": 0.13661666214466095,
"learning_rate": 9.747819162795686e-05,
"loss": 0.0551,
"step": 2930
},
{
"epoch": 2.0303867403314917,
"grad_norm": 0.18142226338386536,
"learning_rate": 9.745220249731217e-05,
"loss": 0.0515,
"step": 2940
},
{
"epoch": 2.037292817679558,
"grad_norm": 0.2540971040725708,
"learning_rate": 9.742608363415781e-05,
"loss": 0.0563,
"step": 2950
},
{
"epoch": 2.044198895027624,
"grad_norm": 0.1666320562362671,
"learning_rate": 9.739983510990176e-05,
"loss": 0.0616,
"step": 2960
},
{
"epoch": 2.0511049723756907,
"grad_norm": 0.20273157954216003,
"learning_rate": 9.737345699630647e-05,
"loss": 0.0586,
"step": 2970
},
{
"epoch": 2.0580110497237567,
"grad_norm": 0.3235710859298706,
"learning_rate": 9.734694936548869e-05,
"loss": 0.0554,
"step": 2980
},
{
"epoch": 2.064917127071823,
"grad_norm": 0.23521968722343445,
"learning_rate": 9.732031228991932e-05,
"loss": 0.0533,
"step": 2990
},
{
"epoch": 2.0718232044198897,
"grad_norm": 0.17394329607486725,
"learning_rate": 9.729354584242302e-05,
"loss": 0.0567,
"step": 3000
},
{
"epoch": 2.0787292817679557,
"grad_norm": 0.1692623794078827,
"learning_rate": 9.726665009617832e-05,
"loss": 0.0512,
"step": 3010
},
{
"epoch": 2.085635359116022,
"grad_norm": 0.30398163199424744,
"learning_rate": 9.723962512471714e-05,
"loss": 0.0581,
"step": 3020
},
{
"epoch": 2.092541436464088,
"grad_norm": 0.3168894648551941,
"learning_rate": 9.72124710019247e-05,
"loss": 0.0506,
"step": 3030
},
{
"epoch": 2.0994475138121547,
"grad_norm": 0.22003120183944702,
"learning_rate": 9.718518780203934e-05,
"loss": 0.0504,
"step": 3040
},
{
"epoch": 2.106353591160221,
"grad_norm": 0.21898090839385986,
"learning_rate": 9.715777559965228e-05,
"loss": 0.0543,
"step": 3050
},
{
"epoch": 2.113259668508287,
"grad_norm": 0.17014583945274353,
"learning_rate": 9.713023446970746e-05,
"loss": 0.0552,
"step": 3060
},
{
"epoch": 2.1201657458563536,
"grad_norm": 0.2032182514667511,
"learning_rate": 9.710256448750126e-05,
"loss": 0.0566,
"step": 3070
},
{
"epoch": 2.12707182320442,
"grad_norm": 0.14699594676494598,
"learning_rate": 9.707476572868235e-05,
"loss": 0.0503,
"step": 3080
},
{
"epoch": 2.133977900552486,
"grad_norm": 0.28991562128067017,
"learning_rate": 9.704683826925149e-05,
"loss": 0.0569,
"step": 3090
},
{
"epoch": 2.1408839779005526,
"grad_norm": 0.303592324256897,
"learning_rate": 9.701878218556129e-05,
"loss": 0.0534,
"step": 3100
},
{
"epoch": 2.1477900552486187,
"grad_norm": 0.2398727685213089,
"learning_rate": 9.699059755431598e-05,
"loss": 0.0602,
"step": 3110
},
{
"epoch": 2.154696132596685,
"grad_norm": 0.19923605024814606,
"learning_rate": 9.696228445257132e-05,
"loss": 0.0508,
"step": 3120
},
{
"epoch": 2.1616022099447516,
"grad_norm": 0.23763804137706757,
"learning_rate": 9.693384295773419e-05,
"loss": 0.0558,
"step": 3130
},
{
"epoch": 2.1685082872928176,
"grad_norm": 0.1759253740310669,
"learning_rate": 9.690527314756259e-05,
"loss": 0.0549,
"step": 3140
},
{
"epoch": 2.175414364640884,
"grad_norm": 0.425621896982193,
"learning_rate": 9.687657510016527e-05,
"loss": 0.059,
"step": 3150
},
{
"epoch": 2.18232044198895,
"grad_norm": 0.3176600933074951,
"learning_rate": 9.684774889400161e-05,
"loss": 0.0525,
"step": 3160
},
{
"epoch": 2.1892265193370166,
"grad_norm": 0.2451590746641159,
"learning_rate": 9.681879460788135e-05,
"loss": 0.0563,
"step": 3170
},
{
"epoch": 2.196132596685083,
"grad_norm": 0.1554330587387085,
"learning_rate": 9.67897123209644e-05,
"loss": 0.0477,
"step": 3180
},
{
"epoch": 2.203038674033149,
"grad_norm": 0.1610517054796219,
"learning_rate": 9.676050211276062e-05,
"loss": 0.0485,
"step": 3190
},
{
"epoch": 2.2099447513812156,
"grad_norm": 0.23754076659679413,
"learning_rate": 9.673116406312962e-05,
"loss": 0.0573,
"step": 3200
},
{
"epoch": 2.2168508287292816,
"grad_norm": 0.24069735407829285,
"learning_rate": 9.67016982522805e-05,
"loss": 0.0533,
"step": 3210
},
{
"epoch": 2.223756906077348,
"grad_norm": 0.25840380787849426,
"learning_rate": 9.667210476077164e-05,
"loss": 0.0488,
"step": 3220
},
{
"epoch": 2.2306629834254146,
"grad_norm": 0.24195915460586548,
"learning_rate": 9.664238366951055e-05,
"loss": 0.051,
"step": 3230
},
{
"epoch": 2.2375690607734806,
"grad_norm": 0.3614242970943451,
"learning_rate": 9.661253505975355e-05,
"loss": 0.0475,
"step": 3240
},
{
"epoch": 2.244475138121547,
"grad_norm": 0.18738463521003723,
"learning_rate": 9.658255901310557e-05,
"loss": 0.0496,
"step": 3250
},
{
"epoch": 2.251381215469613,
"grad_norm": 0.18590529263019562,
"learning_rate": 9.655245561152e-05,
"loss": 0.0555,
"step": 3260
},
{
"epoch": 2.2582872928176796,
"grad_norm": 0.16456319391727448,
"learning_rate": 9.65222249372984e-05,
"loss": 0.0497,
"step": 3270
},
{
"epoch": 2.265193370165746,
"grad_norm": 0.2985817790031433,
"learning_rate": 9.649186707309026e-05,
"loss": 0.0483,
"step": 3280
},
{
"epoch": 2.272099447513812,
"grad_norm": 0.3699333667755127,
"learning_rate": 9.646138210189283e-05,
"loss": 0.0534,
"step": 3290
},
{
"epoch": 2.2790055248618786,
"grad_norm": 0.2080710232257843,
"learning_rate": 9.643077010705087e-05,
"loss": 0.0464,
"step": 3300
},
{
"epoch": 2.2859116022099446,
"grad_norm": 0.1790868490934372,
"learning_rate": 9.640003117225637e-05,
"loss": 0.0476,
"step": 3310
},
{
"epoch": 2.292817679558011,
"grad_norm": 0.21831649541854858,
"learning_rate": 9.636916538154846e-05,
"loss": 0.0504,
"step": 3320
},
{
"epoch": 2.2997237569060776,
"grad_norm": 0.29648539423942566,
"learning_rate": 9.633817281931296e-05,
"loss": 0.0505,
"step": 3330
},
{
"epoch": 2.3066298342541436,
"grad_norm": 0.2649894952774048,
"learning_rate": 9.630705357028242e-05,
"loss": 0.0559,
"step": 3340
},
{
"epoch": 2.31353591160221,
"grad_norm": 0.3060481548309326,
"learning_rate": 9.627580771953563e-05,
"loss": 0.0526,
"step": 3350
},
{
"epoch": 2.320441988950276,
"grad_norm": 0.20820169150829315,
"learning_rate": 9.624443535249759e-05,
"loss": 0.0494,
"step": 3360
},
{
"epoch": 2.3273480662983426,
"grad_norm": 0.16137447953224182,
"learning_rate": 9.621293655493913e-05,
"loss": 0.0538,
"step": 3370
},
{
"epoch": 2.334254143646409,
"grad_norm": 0.18672014772891998,
"learning_rate": 9.618131141297675e-05,
"loss": 0.0523,
"step": 3380
},
{
"epoch": 2.341160220994475,
"grad_norm": 0.2763626277446747,
"learning_rate": 9.614956001307242e-05,
"loss": 0.0512,
"step": 3390
},
{
"epoch": 2.3480662983425415,
"grad_norm": 0.27899008989334106,
"learning_rate": 9.611768244203321e-05,
"loss": 0.0558,
"step": 3400
},
{
"epoch": 2.3549723756906076,
"grad_norm": 0.20031829178333282,
"learning_rate": 9.60856787870112e-05,
"loss": 0.0458,
"step": 3410
},
{
"epoch": 2.361878453038674,
"grad_norm": 0.26932966709136963,
"learning_rate": 9.605354913550318e-05,
"loss": 0.0564,
"step": 3420
},
{
"epoch": 2.3687845303867405,
"grad_norm": 0.17095929384231567,
"learning_rate": 9.602129357535037e-05,
"loss": 0.0479,
"step": 3430
},
{
"epoch": 2.3756906077348066,
"grad_norm": 0.18501821160316467,
"learning_rate": 9.598891219473825e-05,
"loss": 0.0479,
"step": 3440
},
{
"epoch": 2.382596685082873,
"grad_norm": 0.1560228317975998,
"learning_rate": 9.595640508219625e-05,
"loss": 0.0504,
"step": 3450
},
{
"epoch": 2.389502762430939,
"grad_norm": 0.24531404674053192,
"learning_rate": 9.592377232659761e-05,
"loss": 0.0566,
"step": 3460
},
{
"epoch": 2.3964088397790055,
"grad_norm": 0.20866815745830536,
"learning_rate": 9.589101401715904e-05,
"loss": 0.0444,
"step": 3470
},
{
"epoch": 2.403314917127072,
"grad_norm": 0.20980527997016907,
"learning_rate": 9.585813024344045e-05,
"loss": 0.0482,
"step": 3480
},
{
"epoch": 2.410220994475138,
"grad_norm": 0.20052018761634827,
"learning_rate": 9.58251210953449e-05,
"loss": 0.0472,
"step": 3490
},
{
"epoch": 2.4171270718232045,
"grad_norm": 0.20939862728118896,
"learning_rate": 9.579198666311809e-05,
"loss": 0.0556,
"step": 3500
},
{
"epoch": 2.4240331491712706,
"grad_norm": 0.15571282804012299,
"learning_rate": 9.575872703734832e-05,
"loss": 0.0504,
"step": 3510
},
{
"epoch": 2.430939226519337,
"grad_norm": 0.12131053954362869,
"learning_rate": 9.572534230896611e-05,
"loss": 0.0456,
"step": 3520
},
{
"epoch": 2.4378453038674035,
"grad_norm": 0.2301919013261795,
"learning_rate": 9.569183256924403e-05,
"loss": 0.0487,
"step": 3530
},
{
"epoch": 2.4447513812154695,
"grad_norm": 0.24703747034072876,
"learning_rate": 9.565819790979646e-05,
"loss": 0.051,
"step": 3540
},
{
"epoch": 2.451657458563536,
"grad_norm": 0.23158280551433563,
"learning_rate": 9.562443842257925e-05,
"loss": 0.051,
"step": 3550
},
{
"epoch": 2.458563535911602,
"grad_norm": 0.25929200649261475,
"learning_rate": 9.559055419988956e-05,
"loss": 0.0515,
"step": 3560
},
{
"epoch": 2.4654696132596685,
"grad_norm": 0.17875461280345917,
"learning_rate": 9.555654533436557e-05,
"loss": 0.0486,
"step": 3570
},
{
"epoch": 2.472375690607735,
"grad_norm": 0.17399534583091736,
"learning_rate": 9.552241191898621e-05,
"loss": 0.0505,
"step": 3580
},
{
"epoch": 2.479281767955801,
"grad_norm": 0.19749684631824493,
"learning_rate": 9.548815404707092e-05,
"loss": 0.0461,
"step": 3590
},
{
"epoch": 2.4861878453038675,
"grad_norm": 0.24137671291828156,
"learning_rate": 9.545377181227942e-05,
"loss": 0.05,
"step": 3600
},
{
"epoch": 2.4930939226519335,
"grad_norm": 0.21367807686328888,
"learning_rate": 9.541926530861145e-05,
"loss": 0.0445,
"step": 3610
},
{
"epoch": 2.5,
"grad_norm": 0.21933801472187042,
"learning_rate": 9.538463463040645e-05,
"loss": 0.0474,
"step": 3620
},
{
"epoch": 2.5069060773480665,
"grad_norm": 0.14489497244358063,
"learning_rate": 9.534987987234337e-05,
"loss": 0.0489,
"step": 3630
},
{
"epoch": 2.5138121546961325,
"grad_norm": 0.20172163844108582,
"learning_rate": 9.53150011294404e-05,
"loss": 0.0501,
"step": 3640
},
{
"epoch": 2.520718232044199,
"grad_norm": 0.24196617305278778,
"learning_rate": 9.527999849705471e-05,
"loss": 0.0503,
"step": 3650
},
{
"epoch": 2.527624309392265,
"grad_norm": 0.15521635115146637,
"learning_rate": 9.524487207088213e-05,
"loss": 0.0413,
"step": 3660
},
{
"epoch": 2.5345303867403315,
"grad_norm": 0.17045432329177856,
"learning_rate": 9.520962194695698e-05,
"loss": 0.0507,
"step": 3670
},
{
"epoch": 2.541436464088398,
"grad_norm": 0.21607710421085358,
"learning_rate": 9.517424822165175e-05,
"loss": 0.0561,
"step": 3680
},
{
"epoch": 2.548342541436464,
"grad_norm": 0.22375300526618958,
"learning_rate": 9.513875099167685e-05,
"loss": 0.0475,
"step": 3690
},
{
"epoch": 2.5552486187845305,
"grad_norm": 0.20457430183887482,
"learning_rate": 9.510313035408035e-05,
"loss": 0.0452,
"step": 3700
},
{
"epoch": 2.5621546961325965,
"grad_norm": 0.16741271317005157,
"learning_rate": 9.506738640624775e-05,
"loss": 0.0499,
"step": 3710
},
{
"epoch": 2.569060773480663,
"grad_norm": 0.21843063831329346,
"learning_rate": 9.50315192459016e-05,
"loss": 0.0466,
"step": 3720
},
{
"epoch": 2.5759668508287294,
"grad_norm": 0.2357361763715744,
"learning_rate": 9.499552897110136e-05,
"loss": 0.0448,
"step": 3730
},
{
"epoch": 2.5828729281767955,
"grad_norm": 0.19706009328365326,
"learning_rate": 9.495941568024304e-05,
"loss": 0.0473,
"step": 3740
},
{
"epoch": 2.589779005524862,
"grad_norm": 0.1979895979166031,
"learning_rate": 9.492317947205904e-05,
"loss": 0.0484,
"step": 3750
},
{
"epoch": 2.596685082872928,
"grad_norm": 0.21522966027259827,
"learning_rate": 9.488682044561775e-05,
"loss": 0.0475,
"step": 3760
},
{
"epoch": 2.6035911602209945,
"grad_norm": 0.21937403082847595,
"learning_rate": 9.485033870032335e-05,
"loss": 0.049,
"step": 3770
},
{
"epoch": 2.610497237569061,
"grad_norm": 0.2256457507610321,
"learning_rate": 9.481373433591556e-05,
"loss": 0.049,
"step": 3780
},
{
"epoch": 2.617403314917127,
"grad_norm": 0.17436626553535461,
"learning_rate": 9.47770074524693e-05,
"loss": 0.0413,
"step": 3790
},
{
"epoch": 2.6243093922651934,
"grad_norm": 0.19035008549690247,
"learning_rate": 9.474015815039446e-05,
"loss": 0.0464,
"step": 3800
},
{
"epoch": 2.6312154696132595,
"grad_norm": 0.1901327669620514,
"learning_rate": 9.470318653043565e-05,
"loss": 0.0404,
"step": 3810
},
{
"epoch": 2.638121546961326,
"grad_norm": 0.22293877601623535,
"learning_rate": 9.466609269367185e-05,
"loss": 0.049,
"step": 3820
},
{
"epoch": 2.6450276243093924,
"grad_norm": 0.15183129906654358,
"learning_rate": 9.46288767415162e-05,
"loss": 0.0499,
"step": 3830
},
{
"epoch": 2.6519337016574585,
"grad_norm": 0.25658154487609863,
"learning_rate": 9.459153877571567e-05,
"loss": 0.046,
"step": 3840
},
{
"epoch": 2.658839779005525,
"grad_norm": 0.15512892603874207,
"learning_rate": 9.455407889835087e-05,
"loss": 0.0425,
"step": 3850
},
{
"epoch": 2.665745856353591,
"grad_norm": 0.28329044580459595,
"learning_rate": 9.451649721183564e-05,
"loss": 0.0489,
"step": 3860
},
{
"epoch": 2.6726519337016574,
"grad_norm": 0.23139910399913788,
"learning_rate": 9.447879381891692e-05,
"loss": 0.046,
"step": 3870
},
{
"epoch": 2.679558011049724,
"grad_norm": 0.16557002067565918,
"learning_rate": 9.444096882267428e-05,
"loss": 0.0493,
"step": 3880
},
{
"epoch": 2.68646408839779,
"grad_norm": 0.21604911983013153,
"learning_rate": 9.440302232651988e-05,
"loss": 0.0466,
"step": 3890
},
{
"epoch": 2.6933701657458564,
"grad_norm": 0.17157138884067535,
"learning_rate": 9.436495443419795e-05,
"loss": 0.0507,
"step": 3900
},
{
"epoch": 2.7002762430939224,
"grad_norm": 0.22608821094036102,
"learning_rate": 9.432676524978466e-05,
"loss": 0.0438,
"step": 3910
},
{
"epoch": 2.707182320441989,
"grad_norm": 0.21017669141292572,
"learning_rate": 9.42884548776878e-05,
"loss": 0.0474,
"step": 3920
},
{
"epoch": 2.7140883977900554,
"grad_norm": 0.1866975575685501,
"learning_rate": 9.425002342264646e-05,
"loss": 0.0472,
"step": 3930
},
{
"epoch": 2.7209944751381214,
"grad_norm": 0.24714712798595428,
"learning_rate": 9.421147098973077e-05,
"loss": 0.0484,
"step": 3940
},
{
"epoch": 2.727900552486188,
"grad_norm": 0.19898390769958496,
"learning_rate": 9.41727976843416e-05,
"loss": 0.0532,
"step": 3950
},
{
"epoch": 2.734806629834254,
"grad_norm": 0.22621652483940125,
"learning_rate": 9.413400361221029e-05,
"loss": 0.0465,
"step": 3960
},
{
"epoch": 2.7417127071823204,
"grad_norm": 0.17821535468101501,
"learning_rate": 9.409508887939835e-05,
"loss": 0.0435,
"step": 3970
},
{
"epoch": 2.748618784530387,
"grad_norm": 0.21015261113643646,
"learning_rate": 9.40560535922972e-05,
"loss": 0.0441,
"step": 3980
},
{
"epoch": 2.755524861878453,
"grad_norm": 0.1606256365776062,
"learning_rate": 9.40168978576278e-05,
"loss": 0.0464,
"step": 3990
},
{
"epoch": 2.7624309392265194,
"grad_norm": 0.19328652322292328,
"learning_rate": 9.397762178244043e-05,
"loss": 0.0486,
"step": 4000
},
{
"epoch": 2.7693370165745854,
"grad_norm": 0.3121219575405121,
"learning_rate": 9.393822547411439e-05,
"loss": 0.0456,
"step": 4010
},
{
"epoch": 2.776243093922652,
"grad_norm": 0.2536201477050781,
"learning_rate": 9.389870904035769e-05,
"loss": 0.045,
"step": 4020
},
{
"epoch": 2.7831491712707184,
"grad_norm": 0.24330967664718628,
"learning_rate": 9.385907258920672e-05,
"loss": 0.0495,
"step": 4030
},
{
"epoch": 2.7900552486187844,
"grad_norm": 0.1849660575389862,
"learning_rate": 9.381931622902607e-05,
"loss": 0.0489,
"step": 4040
},
{
"epoch": 2.796961325966851,
"grad_norm": 0.19356662034988403,
"learning_rate": 9.377944006850807e-05,
"loss": 0.0391,
"step": 4050
},
{
"epoch": 2.803867403314917,
"grad_norm": 0.18262770771980286,
"learning_rate": 9.373944421667265e-05,
"loss": 0.0478,
"step": 4060
},
{
"epoch": 2.8107734806629834,
"grad_norm": 0.16234233975410461,
"learning_rate": 9.369932878286691e-05,
"loss": 0.0445,
"step": 4070
},
{
"epoch": 2.81767955801105,
"grad_norm": 0.21413561701774597,
"learning_rate": 9.365909387676494e-05,
"loss": 0.0452,
"step": 4080
},
{
"epoch": 2.824585635359116,
"grad_norm": 0.17325039207935333,
"learning_rate": 9.361873960836744e-05,
"loss": 0.0451,
"step": 4090
},
{
"epoch": 2.8314917127071824,
"grad_norm": 0.17369811236858368,
"learning_rate": 9.357826608800142e-05,
"loss": 0.0476,
"step": 4100
},
{
"epoch": 2.8383977900552484,
"grad_norm": 0.1543487012386322,
"learning_rate": 9.353767342631994e-05,
"loss": 0.0439,
"step": 4110
},
{
"epoch": 2.845303867403315,
"grad_norm": 0.20994071662425995,
"learning_rate": 9.34969617343018e-05,
"loss": 0.0432,
"step": 4120
},
{
"epoch": 2.8522099447513813,
"grad_norm": 0.23724523186683655,
"learning_rate": 9.345613112325122e-05,
"loss": 0.0425,
"step": 4130
},
{
"epoch": 2.8591160220994474,
"grad_norm": 0.24331454932689667,
"learning_rate": 9.34151817047975e-05,
"loss": 0.0427,
"step": 4140
},
{
"epoch": 2.866022099447514,
"grad_norm": 0.19277483224868774,
"learning_rate": 9.33741135908948e-05,
"loss": 0.0452,
"step": 4150
},
{
"epoch": 2.87292817679558,
"grad_norm": 0.17652346193790436,
"learning_rate": 9.33329268938218e-05,
"loss": 0.0337,
"step": 4160
},
{
"epoch": 2.8798342541436464,
"grad_norm": 0.1750505417585373,
"learning_rate": 9.329162172618132e-05,
"loss": 0.0461,
"step": 4170
},
{
"epoch": 2.886740331491713,
"grad_norm": 0.20019155740737915,
"learning_rate": 9.325019820090013e-05,
"loss": 0.0496,
"step": 4180
},
{
"epoch": 2.893646408839779,
"grad_norm": 0.18941523134708405,
"learning_rate": 9.320865643122855e-05,
"loss": 0.0461,
"step": 4190
},
{
"epoch": 2.9005524861878453,
"grad_norm": 0.1515769213438034,
"learning_rate": 9.316699653074023e-05,
"loss": 0.0427,
"step": 4200
},
{
"epoch": 2.9074585635359114,
"grad_norm": 0.1701519638299942,
"learning_rate": 9.312521861333172e-05,
"loss": 0.0434,
"step": 4210
},
{
"epoch": 2.914364640883978,
"grad_norm": 0.23309242725372314,
"learning_rate": 9.308332279322224e-05,
"loss": 0.0456,
"step": 4220
},
{
"epoch": 2.9212707182320443,
"grad_norm": 0.2197219729423523,
"learning_rate": 9.304130918495338e-05,
"loss": 0.0428,
"step": 4230
},
{
"epoch": 2.9281767955801103,
"grad_norm": 0.16014282405376434,
"learning_rate": 9.299917790338874e-05,
"loss": 0.0436,
"step": 4240
},
{
"epoch": 2.935082872928177,
"grad_norm": 0.18948529660701752,
"learning_rate": 9.295692906371363e-05,
"loss": 0.0487,
"step": 4250
},
{
"epoch": 2.941988950276243,
"grad_norm": 0.1489516645669937,
"learning_rate": 9.291456278143476e-05,
"loss": 0.0446,
"step": 4260
},
{
"epoch": 2.9488950276243093,
"grad_norm": 0.1619735211133957,
"learning_rate": 9.287207917237994e-05,
"loss": 0.0399,
"step": 4270
},
{
"epoch": 2.955801104972376,
"grad_norm": 0.20590822398662567,
"learning_rate": 9.282947835269773e-05,
"loss": 0.0452,
"step": 4280
},
{
"epoch": 2.962707182320442,
"grad_norm": 0.16909535229206085,
"learning_rate": 9.278676043885715e-05,
"loss": 0.0384,
"step": 4290
},
{
"epoch": 2.9696132596685083,
"grad_norm": 0.184917613863945,
"learning_rate": 9.274392554764733e-05,
"loss": 0.0393,
"step": 4300
},
{
"epoch": 2.9765193370165743,
"grad_norm": 0.2012282758951187,
"learning_rate": 9.270097379617723e-05,
"loss": 0.0419,
"step": 4310
},
{
"epoch": 2.983425414364641,
"grad_norm": 0.188528373837471,
"learning_rate": 9.26579053018753e-05,
"loss": 0.0462,
"step": 4320
},
{
"epoch": 2.9903314917127073,
"grad_norm": 0.21025685966014862,
"learning_rate": 9.261472018248918e-05,
"loss": 0.0403,
"step": 4330
},
{
"epoch": 2.9972375690607733,
"grad_norm": 0.2396075278520584,
"learning_rate": 9.25714185560853e-05,
"loss": 0.0398,
"step": 4340
},
{
"epoch": 3.00414364640884,
"grad_norm": 0.22464925050735474,
"learning_rate": 9.252800054104868e-05,
"loss": 0.0449,
"step": 4350
},
{
"epoch": 3.0110497237569063,
"grad_norm": 0.1573522537946701,
"learning_rate": 9.248446625608252e-05,
"loss": 0.0373,
"step": 4360
},
{
"epoch": 3.0179558011049723,
"grad_norm": 0.1664257049560547,
"learning_rate": 9.244081582020789e-05,
"loss": 0.0428,
"step": 4370
},
{
"epoch": 3.0248618784530388,
"grad_norm": 0.219674214720726,
"learning_rate": 9.239704935276339e-05,
"loss": 0.0433,
"step": 4380
},
{
"epoch": 3.031767955801105,
"grad_norm": 0.18510626256465912,
"learning_rate": 9.235316697340489e-05,
"loss": 0.0425,
"step": 4390
},
{
"epoch": 3.0386740331491713,
"grad_norm": 0.14783309400081635,
"learning_rate": 9.230916880210512e-05,
"loss": 0.0501,
"step": 4400
},
{
"epoch": 3.0455801104972378,
"grad_norm": 0.20491963624954224,
"learning_rate": 9.226505495915342e-05,
"loss": 0.047,
"step": 4410
},
{
"epoch": 3.052486187845304,
"grad_norm": 0.19762618839740753,
"learning_rate": 9.222082556515536e-05,
"loss": 0.0408,
"step": 4420
},
{
"epoch": 3.0593922651933703,
"grad_norm": 0.1874648630619049,
"learning_rate": 9.217648074103242e-05,
"loss": 0.0438,
"step": 4430
},
{
"epoch": 3.0662983425414363,
"grad_norm": 0.15683989226818085,
"learning_rate": 9.213202060802161e-05,
"loss": 0.0471,
"step": 4440
},
{
"epoch": 3.0732044198895028,
"grad_norm": 0.19267770648002625,
"learning_rate": 9.208744528767528e-05,
"loss": 0.0424,
"step": 4450
},
{
"epoch": 3.0801104972375692,
"grad_norm": 0.20659422874450684,
"learning_rate": 9.204275490186064e-05,
"loss": 0.0431,
"step": 4460
},
{
"epoch": 3.0870165745856353,
"grad_norm": 0.1620733141899109,
"learning_rate": 9.199794957275949e-05,
"loss": 0.04,
"step": 4470
},
{
"epoch": 3.0939226519337018,
"grad_norm": 0.21852877736091614,
"learning_rate": 9.19530294228679e-05,
"loss": 0.0434,
"step": 4480
},
{
"epoch": 3.100828729281768,
"grad_norm": 0.18497471511363983,
"learning_rate": 9.190799457499583e-05,
"loss": 0.0427,
"step": 4490
},
{
"epoch": 3.1077348066298343,
"grad_norm": 0.20119594037532806,
"learning_rate": 9.186284515226686e-05,
"loss": 0.0445,
"step": 4500
},
{
"epoch": 3.1146408839779007,
"grad_norm": 0.16487465798854828,
"learning_rate": 9.181758127811777e-05,
"loss": 0.0463,
"step": 4510
},
{
"epoch": 3.1215469613259668,
"grad_norm": 0.18515750765800476,
"learning_rate": 9.177220307629825e-05,
"loss": 0.0434,
"step": 4520
},
{
"epoch": 3.1284530386740332,
"grad_norm": 0.15298950672149658,
"learning_rate": 9.172671067087059e-05,
"loss": 0.0426,
"step": 4530
},
{
"epoch": 3.1353591160220993,
"grad_norm": 0.2684433162212372,
"learning_rate": 9.16811041862093e-05,
"loss": 0.0368,
"step": 4540
},
{
"epoch": 3.1422651933701657,
"grad_norm": 0.16758117079734802,
"learning_rate": 9.163538374700076e-05,
"loss": 0.0406,
"step": 4550
},
{
"epoch": 3.149171270718232,
"grad_norm": 0.16840502619743347,
"learning_rate": 9.158954947824287e-05,
"loss": 0.0366,
"step": 4560
},
{
"epoch": 3.1560773480662982,
"grad_norm": 0.11108331382274628,
"learning_rate": 9.154360150524482e-05,
"loss": 0.0367,
"step": 4570
},
{
"epoch": 3.1629834254143647,
"grad_norm": 0.189457967877388,
"learning_rate": 9.14975399536266e-05,
"loss": 0.0422,
"step": 4580
},
{
"epoch": 3.1698895027624308,
"grad_norm": 0.19346898794174194,
"learning_rate": 9.14513649493187e-05,
"loss": 0.0437,
"step": 4590
},
{
"epoch": 3.1767955801104972,
"grad_norm": 0.3062164783477783,
"learning_rate": 9.140507661856187e-05,
"loss": 0.0427,
"step": 4600
},
{
"epoch": 3.1837016574585637,
"grad_norm": 0.18615876138210297,
"learning_rate": 9.135867508790661e-05,
"loss": 0.0413,
"step": 4610
},
{
"epoch": 3.1906077348066297,
"grad_norm": 0.18920812010765076,
"learning_rate": 9.131216048421291e-05,
"loss": 0.0466,
"step": 4620
},
{
"epoch": 3.197513812154696,
"grad_norm": 0.1465039700269699,
"learning_rate": 9.126553293464998e-05,
"loss": 0.0442,
"step": 4630
},
{
"epoch": 3.2044198895027622,
"grad_norm": 0.2439999282360077,
"learning_rate": 9.121879256669572e-05,
"loss": 0.0458,
"step": 4640
},
{
"epoch": 3.2113259668508287,
"grad_norm": 0.40056276321411133,
"learning_rate": 9.117193950813652e-05,
"loss": 0.0454,
"step": 4650
},
{
"epoch": 3.218232044198895,
"grad_norm": 0.18807795643806458,
"learning_rate": 9.112497388706685e-05,
"loss": 0.0508,
"step": 4660
},
{
"epoch": 3.2251381215469612,
"grad_norm": 0.18762271106243134,
"learning_rate": 9.10778958318889e-05,
"loss": 0.0425,
"step": 4670
},
{
"epoch": 3.2320441988950277,
"grad_norm": 0.19282494485378265,
"learning_rate": 9.103070547131232e-05,
"loss": 0.0353,
"step": 4680
},
{
"epoch": 3.2389502762430937,
"grad_norm": 0.2008621245622635,
"learning_rate": 9.098340293435375e-05,
"loss": 0.0406,
"step": 4690
},
{
"epoch": 3.24585635359116,
"grad_norm": 0.1510397046804428,
"learning_rate": 9.093598835033649e-05,
"loss": 0.0439,
"step": 4700
},
{
"epoch": 3.2527624309392267,
"grad_norm": 0.22977766394615173,
"learning_rate": 9.088846184889021e-05,
"loss": 0.0414,
"step": 4710
},
{
"epoch": 3.2596685082872927,
"grad_norm": 0.17466506361961365,
"learning_rate": 9.084082355995057e-05,
"loss": 0.0443,
"step": 4720
},
{
"epoch": 3.266574585635359,
"grad_norm": 0.17879237234592438,
"learning_rate": 9.079307361375882e-05,
"loss": 0.0425,
"step": 4730
},
{
"epoch": 3.2734806629834257,
"grad_norm": 0.20717912912368774,
"learning_rate": 9.074521214086149e-05,
"loss": 0.0373,
"step": 4740
},
{
"epoch": 3.2803867403314917,
"grad_norm": 0.19752268493175507,
"learning_rate": 9.069723927211001e-05,
"loss": 0.0391,
"step": 4750
},
{
"epoch": 3.287292817679558,
"grad_norm": 0.19491535425186157,
"learning_rate": 9.064915513866037e-05,
"loss": 0.0434,
"step": 4760
},
{
"epoch": 3.294198895027624,
"grad_norm": 0.21187685430049896,
"learning_rate": 9.060095987197279e-05,
"loss": 0.0402,
"step": 4770
},
{
"epoch": 3.3011049723756907,
"grad_norm": 0.15764431655406952,
"learning_rate": 9.055265360381126e-05,
"loss": 0.0423,
"step": 4780
},
{
"epoch": 3.308011049723757,
"grad_norm": 0.15811102092266083,
"learning_rate": 9.050423646624326e-05,
"loss": 0.0385,
"step": 4790
},
{
"epoch": 3.314917127071823,
"grad_norm": 0.16640017926692963,
"learning_rate": 9.045570859163943e-05,
"loss": 0.0427,
"step": 4800
},
{
"epoch": 3.3218232044198897,
"grad_norm": 0.15784433484077454,
"learning_rate": 9.04070701126731e-05,
"loss": 0.0415,
"step": 4810
},
{
"epoch": 3.3287292817679557,
"grad_norm": 0.1673387885093689,
"learning_rate": 9.035832116232001e-05,
"loss": 0.0386,
"step": 4820
},
{
"epoch": 3.335635359116022,
"grad_norm": 0.2144685834646225,
"learning_rate": 9.030946187385796e-05,
"loss": 0.0497,
"step": 4830
},
{
"epoch": 3.3425414364640886,
"grad_norm": 0.23754587769508362,
"learning_rate": 9.026049238086635e-05,
"loss": 0.048,
"step": 4840
},
{
"epoch": 3.3494475138121547,
"grad_norm": 0.1485087275505066,
"learning_rate": 9.021141281722591e-05,
"loss": 0.0363,
"step": 4850
},
{
"epoch": 3.356353591160221,
"grad_norm": 0.19838620722293854,
"learning_rate": 9.01622233171183e-05,
"loss": 0.0398,
"step": 4860
},
{
"epoch": 3.363259668508287,
"grad_norm": 0.22358939051628113,
"learning_rate": 9.011292401502574e-05,
"loss": 0.044,
"step": 4870
},
{
"epoch": 3.3701657458563536,
"grad_norm": 0.21143372356891632,
"learning_rate": 9.006351504573063e-05,
"loss": 0.042,
"step": 4880
},
{
"epoch": 3.37707182320442,
"grad_norm": 0.19049137830734253,
"learning_rate": 9.001399654431519e-05,
"loss": 0.0403,
"step": 4890
},
{
"epoch": 3.383977900552486,
"grad_norm": 0.20016619563102722,
"learning_rate": 8.996436864616116e-05,
"loss": 0.042,
"step": 4900
},
{
"epoch": 3.3908839779005526,
"grad_norm": 0.16115543246269226,
"learning_rate": 8.991463148694925e-05,
"loss": 0.0387,
"step": 4910
},
{
"epoch": 3.3977900552486187,
"grad_norm": 0.18714101612567902,
"learning_rate": 8.986478520265902e-05,
"loss": 0.0357,
"step": 4920
},
{
"epoch": 3.404696132596685,
"grad_norm": 0.23854629695415497,
"learning_rate": 8.981482992956827e-05,
"loss": 0.0414,
"step": 4930
},
{
"epoch": 3.4116022099447516,
"grad_norm": 0.18258176743984222,
"learning_rate": 8.976476580425282e-05,
"loss": 0.0406,
"step": 4940
},
{
"epoch": 3.4185082872928176,
"grad_norm": 0.1593986600637436,
"learning_rate": 8.971459296358606e-05,
"loss": 0.0413,
"step": 4950
},
{
"epoch": 3.425414364640884,
"grad_norm": 0.2746127247810364,
"learning_rate": 8.966431154473864e-05,
"loss": 0.0399,
"step": 4960
},
{
"epoch": 3.43232044198895,
"grad_norm": 0.23390236496925354,
"learning_rate": 8.961392168517803e-05,
"loss": 0.0413,
"step": 4970
},
{
"epoch": 3.4392265193370166,
"grad_norm": 0.1735554188489914,
"learning_rate": 8.956342352266821e-05,
"loss": 0.0452,
"step": 4980
},
{
"epoch": 3.446132596685083,
"grad_norm": 0.15287978947162628,
"learning_rate": 8.95128171952692e-05,
"loss": 0.0384,
"step": 4990
},
{
"epoch": 3.453038674033149,
"grad_norm": 0.16149814426898956,
"learning_rate": 8.946210284133676e-05,
"loss": 0.0414,
"step": 5000
},
{
"epoch": 3.4599447513812156,
"grad_norm": 0.1760193556547165,
"learning_rate": 8.941128059952201e-05,
"loss": 0.0426,
"step": 5010
},
{
"epoch": 3.4668508287292816,
"grad_norm": 0.21478940546512604,
"learning_rate": 8.936035060877102e-05,
"loss": 0.041,
"step": 5020
},
{
"epoch": 3.473756906077348,
"grad_norm": 0.20234301686286926,
"learning_rate": 8.930931300832443e-05,
"loss": 0.0413,
"step": 5030
},
{
"epoch": 3.4806629834254146,
"grad_norm": 0.1900159865617752,
"learning_rate": 8.925816793771711e-05,
"loss": 0.0405,
"step": 5040
},
{
"epoch": 3.4875690607734806,
"grad_norm": 0.15434470772743225,
"learning_rate": 8.92069155367777e-05,
"loss": 0.0383,
"step": 5050
},
{
"epoch": 3.494475138121547,
"grad_norm": 0.14990466833114624,
"learning_rate": 8.915555594562834e-05,
"loss": 0.04,
"step": 5060
},
{
"epoch": 3.501381215469613,
"grad_norm": 0.14787523448467255,
"learning_rate": 8.910408930468416e-05,
"loss": 0.0332,
"step": 5070
},
{
"epoch": 3.5082872928176796,
"grad_norm": 0.17466671764850616,
"learning_rate": 8.905251575465303e-05,
"loss": 0.0406,
"step": 5080
},
{
"epoch": 3.515193370165746,
"grad_norm": 0.20608386397361755,
"learning_rate": 8.900083543653502e-05,
"loss": 0.0432,
"step": 5090
},
{
"epoch": 3.522099447513812,
"grad_norm": 0.15128932893276215,
"learning_rate": 8.894904849162218e-05,
"loss": 0.0392,
"step": 5100
},
{
"epoch": 3.5290055248618786,
"grad_norm": 0.21203255653381348,
"learning_rate": 8.889715506149802e-05,
"loss": 0.0376,
"step": 5110
},
{
"epoch": 3.5359116022099446,
"grad_norm": 0.18824951350688934,
"learning_rate": 8.884515528803722e-05,
"loss": 0.0444,
"step": 5120
},
{
"epoch": 3.542817679558011,
"grad_norm": 0.18114745616912842,
"learning_rate": 8.879304931340517e-05,
"loss": 0.0354,
"step": 5130
},
{
"epoch": 3.5497237569060776,
"grad_norm": 0.16870130598545074,
"learning_rate": 8.874083728005759e-05,
"loss": 0.036,
"step": 5140
},
{
"epoch": 3.5566298342541436,
"grad_norm": 0.17004938423633575,
"learning_rate": 8.868851933074021e-05,
"loss": 0.0384,
"step": 5150
},
{
"epoch": 3.56353591160221,
"grad_norm": 0.21150527894496918,
"learning_rate": 8.863609560848829e-05,
"loss": 0.0415,
"step": 5160
},
{
"epoch": 3.570441988950276,
"grad_norm": 0.2698124945163727,
"learning_rate": 8.85835662566263e-05,
"loss": 0.0314,
"step": 5170
},
{
"epoch": 3.5773480662983426,
"grad_norm": 0.2014392763376236,
"learning_rate": 8.853093141876747e-05,
"loss": 0.0394,
"step": 5180
},
{
"epoch": 3.584254143646409,
"grad_norm": 0.18041956424713135,
"learning_rate": 8.847819123881343e-05,
"loss": 0.0424,
"step": 5190
},
{
"epoch": 3.591160220994475,
"grad_norm": 0.15482786297798157,
"learning_rate": 8.842534586095383e-05,
"loss": 0.0444,
"step": 5200
},
{
"epoch": 3.5980662983425415,
"grad_norm": 0.1852821409702301,
"learning_rate": 8.837239542966593e-05,
"loss": 0.0359,
"step": 5210
},
{
"epoch": 3.6049723756906076,
"grad_norm": 0.21371380984783173,
"learning_rate": 8.831934008971417e-05,
"loss": 0.0416,
"step": 5220
},
{
"epoch": 3.611878453038674,
"grad_norm": 0.11043145507574081,
"learning_rate": 8.826617998614982e-05,
"loss": 0.0354,
"step": 5230
},
{
"epoch": 3.6187845303867405,
"grad_norm": 0.15232881903648376,
"learning_rate": 8.821291526431056e-05,
"loss": 0.0354,
"step": 5240
},
{
"epoch": 3.6256906077348066,
"grad_norm": 0.14688993990421295,
"learning_rate": 8.815954606982015e-05,
"loss": 0.034,
"step": 5250
},
{
"epoch": 3.632596685082873,
"grad_norm": 0.1413627713918686,
"learning_rate": 8.810607254858789e-05,
"loss": 0.0399,
"step": 5260
},
{
"epoch": 3.639502762430939,
"grad_norm": 0.18085753917694092,
"learning_rate": 8.805249484680838e-05,
"loss": 0.0412,
"step": 5270
},
{
"epoch": 3.6464088397790055,
"grad_norm": 0.17890624701976776,
"learning_rate": 8.799881311096096e-05,
"loss": 0.0399,
"step": 5280
},
{
"epoch": 3.653314917127072,
"grad_norm": 0.1451053023338318,
"learning_rate": 8.794502748780949e-05,
"loss": 0.0406,
"step": 5290
},
{
"epoch": 3.660220994475138,
"grad_norm": 0.14209549129009247,
"learning_rate": 8.78911381244018e-05,
"loss": 0.0376,
"step": 5300
},
{
"epoch": 3.6671270718232045,
"grad_norm": 0.17420484125614166,
"learning_rate": 8.783714516806933e-05,
"loss": 0.0361,
"step": 5310
},
{
"epoch": 3.6740331491712706,
"grad_norm": 0.20409078896045685,
"learning_rate": 8.77830487664268e-05,
"loss": 0.0359,
"step": 5320
},
{
"epoch": 3.680939226519337,
"grad_norm": 0.13683636486530304,
"learning_rate": 8.772884906737167e-05,
"loss": 0.0395,
"step": 5330
},
{
"epoch": 3.6878453038674035,
"grad_norm": 0.16172552108764648,
"learning_rate": 8.767454621908387e-05,
"loss": 0.0359,
"step": 5340
},
{
"epoch": 3.6947513812154695,
"grad_norm": 0.1768387109041214,
"learning_rate": 8.76201403700253e-05,
"loss": 0.036,
"step": 5350
},
{
"epoch": 3.701657458563536,
"grad_norm": 0.22492876648902893,
"learning_rate": 8.756563166893949e-05,
"loss": 0.0379,
"step": 5360
},
{
"epoch": 3.708563535911602,
"grad_norm": 0.16830484569072723,
"learning_rate": 8.751102026485113e-05,
"loss": 0.0436,
"step": 5370
},
{
"epoch": 3.7154696132596685,
"grad_norm": 0.17052343487739563,
"learning_rate": 8.745630630706571e-05,
"loss": 0.0385,
"step": 5380
},
{
"epoch": 3.722375690607735,
"grad_norm": 0.22967390716075897,
"learning_rate": 8.740148994516912e-05,
"loss": 0.0383,
"step": 5390
},
{
"epoch": 3.729281767955801,
"grad_norm": 0.2420748621225357,
"learning_rate": 8.73465713290272e-05,
"loss": 0.0377,
"step": 5400
},
{
"epoch": 3.7361878453038675,
"grad_norm": 0.20598067343235016,
"learning_rate": 8.729155060878533e-05,
"loss": 0.0394,
"step": 5410
},
{
"epoch": 3.7430939226519335,
"grad_norm": 0.17275604605674744,
"learning_rate": 8.723642793486809e-05,
"loss": 0.0336,
"step": 5420
},
{
"epoch": 3.75,
"grad_norm": 0.26493966579437256,
"learning_rate": 8.718120345797873e-05,
"loss": 0.0392,
"step": 5430
},
{
"epoch": 3.7569060773480665,
"grad_norm": 0.21323080360889435,
"learning_rate": 8.712587732909889e-05,
"loss": 0.0373,
"step": 5440
},
{
"epoch": 3.7638121546961325,
"grad_norm": 0.2604227364063263,
"learning_rate": 8.707044969948806e-05,
"loss": 0.0406,
"step": 5450
},
{
"epoch": 3.770718232044199,
"grad_norm": 0.14074908196926117,
"learning_rate": 8.701492072068329e-05,
"loss": 0.0366,
"step": 5460
},
{
"epoch": 3.777624309392265,
"grad_norm": 0.15715032815933228,
"learning_rate": 8.695929054449869e-05,
"loss": 0.0371,
"step": 5470
},
{
"epoch": 3.7845303867403315,
"grad_norm": 0.1916290670633316,
"learning_rate": 8.690355932302501e-05,
"loss": 0.0323,
"step": 5480
},
{
"epoch": 3.791436464088398,
"grad_norm": 0.22312071919441223,
"learning_rate": 8.684772720862931e-05,
"loss": 0.0412,
"step": 5490
},
{
"epoch": 3.798342541436464,
"grad_norm": 0.19781441986560822,
"learning_rate": 8.679179435395446e-05,
"loss": 0.0358,
"step": 5500
},
{
"epoch": 3.8052486187845305,
"grad_norm": 0.1877889633178711,
"learning_rate": 8.673576091191874e-05,
"loss": 0.0359,
"step": 5510
},
{
"epoch": 3.8121546961325965,
"grad_norm": 0.17329281568527222,
"learning_rate": 8.667962703571541e-05,
"loss": 0.0331,
"step": 5520
},
{
"epoch": 3.819060773480663,
"grad_norm": 0.23658014833927155,
"learning_rate": 8.662339287881238e-05,
"loss": 0.0427,
"step": 5530
},
{
"epoch": 3.8259668508287294,
"grad_norm": 0.20422625541687012,
"learning_rate": 8.656705859495169e-05,
"loss": 0.041,
"step": 5540
},
{
"epoch": 3.8328729281767955,
"grad_norm": 0.17714661359786987,
"learning_rate": 8.651062433814912e-05,
"loss": 0.0318,
"step": 5550
},
{
"epoch": 3.839779005524862,
"grad_norm": 0.18044039607048035,
"learning_rate": 8.645409026269375e-05,
"loss": 0.0389,
"step": 5560
},
{
"epoch": 3.846685082872928,
"grad_norm": 0.14779943227767944,
"learning_rate": 8.639745652314759e-05,
"loss": 0.0314,
"step": 5570
},
{
"epoch": 3.8535911602209945,
"grad_norm": 0.16303597390651703,
"learning_rate": 8.634072327434515e-05,
"loss": 0.037,
"step": 5580
},
{
"epoch": 3.860497237569061,
"grad_norm": 0.17153339087963104,
"learning_rate": 8.628389067139294e-05,
"loss": 0.0386,
"step": 5590
},
{
"epoch": 3.867403314917127,
"grad_norm": 0.2033110409975052,
"learning_rate": 8.622695886966911e-05,
"loss": 0.0354,
"step": 5600
},
{
"epoch": 3.8743093922651934,
"grad_norm": 0.16191303730010986,
"learning_rate": 8.616992802482308e-05,
"loss": 0.0403,
"step": 5610
},
{
"epoch": 3.8812154696132595,
"grad_norm": 0.1545790135860443,
"learning_rate": 8.611279829277496e-05,
"loss": 0.038,
"step": 5620
},
{
"epoch": 3.888121546961326,
"grad_norm": 0.18758025765419006,
"learning_rate": 8.605556982971528e-05,
"loss": 0.0355,
"step": 5630
},
{
"epoch": 3.8950276243093924,
"grad_norm": 0.16228602826595306,
"learning_rate": 8.599824279210447e-05,
"loss": 0.0389,
"step": 5640
},
{
"epoch": 3.9019337016574585,
"grad_norm": 0.1846267133951187,
"learning_rate": 8.594081733667243e-05,
"loss": 0.0343,
"step": 5650
},
{
"epoch": 3.908839779005525,
"grad_norm": 0.22637325525283813,
"learning_rate": 8.58832936204182e-05,
"loss": 0.0358,
"step": 5660
},
{
"epoch": 3.915745856353591,
"grad_norm": 0.15883108973503113,
"learning_rate": 8.582567180060942e-05,
"loss": 0.037,
"step": 5670
},
{
"epoch": 3.9226519337016574,
"grad_norm": 0.20179514586925507,
"learning_rate": 8.576795203478194e-05,
"loss": 0.0367,
"step": 5680
},
{
"epoch": 3.929558011049724,
"grad_norm": 0.226895272731781,
"learning_rate": 8.571013448073939e-05,
"loss": 0.0422,
"step": 5690
},
{
"epoch": 3.93646408839779,
"grad_norm": 0.1747041642665863,
"learning_rate": 8.565221929655275e-05,
"loss": 0.0324,
"step": 5700
},
{
"epoch": 3.9433701657458564,
"grad_norm": 0.22531364858150482,
"learning_rate": 8.559420664055992e-05,
"loss": 0.0398,
"step": 5710
},
{
"epoch": 3.9502762430939224,
"grad_norm": 0.2313089370727539,
"learning_rate": 8.553609667136532e-05,
"loss": 0.0418,
"step": 5720
},
{
"epoch": 3.957182320441989,
"grad_norm": 0.2609806954860687,
"learning_rate": 8.547788954783936e-05,
"loss": 0.0445,
"step": 5730
},
{
"epoch": 3.9640883977900554,
"grad_norm": 0.18635094165802002,
"learning_rate": 8.541958542911808e-05,
"loss": 0.0337,
"step": 5740
},
{
"epoch": 3.9709944751381214,
"grad_norm": 0.20794489979743958,
"learning_rate": 8.536118447460275e-05,
"loss": 0.0365,
"step": 5750
},
{
"epoch": 3.977900552486188,
"grad_norm": 0.19042007625102997,
"learning_rate": 8.530268684395932e-05,
"loss": 0.0357,
"step": 5760
},
{
"epoch": 3.984806629834254,
"grad_norm": 0.30799993872642517,
"learning_rate": 8.524409269711807e-05,
"loss": 0.0402,
"step": 5770
},
{
"epoch": 3.9917127071823204,
"grad_norm": 0.19904528558254242,
"learning_rate": 8.51854021942732e-05,
"loss": 0.0394,
"step": 5780
},
{
"epoch": 3.998618784530387,
"grad_norm": 0.16762062907218933,
"learning_rate": 8.512661549588227e-05,
"loss": 0.038,
"step": 5790
},
{
"epoch": 4.005524861878453,
"grad_norm": 0.1410869061946869,
"learning_rate": 8.506773276266588e-05,
"loss": 0.0316,
"step": 5800
},
{
"epoch": 4.012430939226519,
"grad_norm": 0.17487668991088867,
"learning_rate": 8.500875415560721e-05,
"loss": 0.034,
"step": 5810
},
{
"epoch": 4.019337016574585,
"grad_norm": 0.15733812749385834,
"learning_rate": 8.494967983595144e-05,
"loss": 0.0327,
"step": 5820
},
{
"epoch": 4.026243093922652,
"grad_norm": 0.15330734848976135,
"learning_rate": 8.489050996520558e-05,
"loss": 0.038,
"step": 5830
},
{
"epoch": 4.033149171270718,
"grad_norm": 0.10522384196519852,
"learning_rate": 8.483124470513775e-05,
"loss": 0.0361,
"step": 5840
},
{
"epoch": 4.040055248618785,
"grad_norm": 0.18659597635269165,
"learning_rate": 8.477188421777692e-05,
"loss": 0.0398,
"step": 5850
},
{
"epoch": 4.04696132596685,
"grad_norm": 0.13936083018779755,
"learning_rate": 8.47124286654124e-05,
"loss": 0.0382,
"step": 5860
},
{
"epoch": 4.053867403314917,
"grad_norm": 0.1510729044675827,
"learning_rate": 8.465287821059341e-05,
"loss": 0.0345,
"step": 5870
},
{
"epoch": 4.060773480662983,
"grad_norm": 0.18497096002101898,
"learning_rate": 8.45932330161286e-05,
"loss": 0.0408,
"step": 5880
},
{
"epoch": 4.06767955801105,
"grad_norm": 0.1754857897758484,
"learning_rate": 8.453349324508567e-05,
"loss": 0.0367,
"step": 5890
},
{
"epoch": 4.074585635359116,
"grad_norm": 0.18897674977779388,
"learning_rate": 8.447365906079088e-05,
"loss": 0.0379,
"step": 5900
},
{
"epoch": 4.081491712707182,
"grad_norm": 0.1608031988143921,
"learning_rate": 8.441373062682856e-05,
"loss": 0.0343,
"step": 5910
},
{
"epoch": 4.088397790055248,
"grad_norm": 0.17947819828987122,
"learning_rate": 8.43537081070408e-05,
"loss": 0.0346,
"step": 5920
},
{
"epoch": 4.095303867403315,
"grad_norm": 0.15115374326705933,
"learning_rate": 8.429359166552689e-05,
"loss": 0.0319,
"step": 5930
},
{
"epoch": 4.102209944751381,
"grad_norm": 0.18776749074459076,
"learning_rate": 8.423338146664284e-05,
"loss": 0.0361,
"step": 5940
},
{
"epoch": 4.109116022099448,
"grad_norm": 0.1795329600572586,
"learning_rate": 8.417307767500107e-05,
"loss": 0.0346,
"step": 5950
},
{
"epoch": 4.116022099447513,
"grad_norm": 0.17658886313438416,
"learning_rate": 8.411268045546983e-05,
"loss": 0.0318,
"step": 5960
},
{
"epoch": 4.12292817679558,
"grad_norm": 0.18604636192321777,
"learning_rate": 8.405218997317281e-05,
"loss": 0.0332,
"step": 5970
},
{
"epoch": 4.129834254143646,
"grad_norm": 0.18217170238494873,
"learning_rate": 8.399160639348869e-05,
"loss": 0.0311,
"step": 5980
},
{
"epoch": 4.136740331491713,
"grad_norm": 0.16615109145641327,
"learning_rate": 8.393092988205065e-05,
"loss": 0.0324,
"step": 5990
},
{
"epoch": 4.143646408839779,
"grad_norm": 0.2016613483428955,
"learning_rate": 8.387016060474597e-05,
"loss": 0.0349,
"step": 6000
},
{
"epoch": 4.150552486187845,
"grad_norm": 0.12477770447731018,
"learning_rate": 8.380929872771551e-05,
"loss": 0.0316,
"step": 6010
},
{
"epoch": 4.157458563535911,
"grad_norm": 0.1528131514787674,
"learning_rate": 8.374834441735335e-05,
"loss": 0.0354,
"step": 6020
},
{
"epoch": 4.164364640883978,
"grad_norm": 0.10325898975133896,
"learning_rate": 8.368729784030622e-05,
"loss": 0.0299,
"step": 6030
},
{
"epoch": 4.171270718232044,
"grad_norm": 0.17199422419071198,
"learning_rate": 8.362615916347315e-05,
"loss": 0.0365,
"step": 6040
},
{
"epoch": 4.178176795580111,
"grad_norm": 0.1270386278629303,
"learning_rate": 8.356492855400493e-05,
"loss": 0.0311,
"step": 6050
},
{
"epoch": 4.185082872928176,
"grad_norm": 0.14681442081928253,
"learning_rate": 8.350360617930371e-05,
"loss": 0.0343,
"step": 6060
},
{
"epoch": 4.191988950276243,
"grad_norm": 0.14189831912517548,
"learning_rate": 8.344219220702255e-05,
"loss": 0.0326,
"step": 6070
},
{
"epoch": 4.198895027624309,
"grad_norm": 0.22745615243911743,
"learning_rate": 8.338068680506485e-05,
"loss": 0.037,
"step": 6080
},
{
"epoch": 4.205801104972376,
"grad_norm": 0.1944071650505066,
"learning_rate": 8.33190901415841e-05,
"loss": 0.0377,
"step": 6090
},
{
"epoch": 4.212707182320442,
"grad_norm": 0.14716528356075287,
"learning_rate": 8.325740238498317e-05,
"loss": 0.0284,
"step": 6100
},
{
"epoch": 4.219613259668508,
"grad_norm": 0.14018702507019043,
"learning_rate": 8.319562370391406e-05,
"loss": 0.0345,
"step": 6110
},
{
"epoch": 4.226519337016574,
"grad_norm": 0.12675043940544128,
"learning_rate": 8.31337542672773e-05,
"loss": 0.033,
"step": 6120
},
{
"epoch": 4.233425414364641,
"grad_norm": 0.1785709261894226,
"learning_rate": 8.307179424422158e-05,
"loss": 0.0378,
"step": 6130
},
{
"epoch": 4.240331491712707,
"grad_norm": 0.15371115505695343,
"learning_rate": 8.300974380414327e-05,
"loss": 0.0299,
"step": 6140
},
{
"epoch": 4.247237569060774,
"grad_norm": 0.21764731407165527,
"learning_rate": 8.294760311668586e-05,
"loss": 0.0309,
"step": 6150
},
{
"epoch": 4.25414364640884,
"grad_norm": 0.24038873612880707,
"learning_rate": 8.288537235173961e-05,
"loss": 0.0353,
"step": 6160
},
{
"epoch": 4.261049723756906,
"grad_norm": 0.17432773113250732,
"learning_rate": 8.282305167944108e-05,
"loss": 0.0324,
"step": 6170
},
{
"epoch": 4.267955801104972,
"grad_norm": 0.1402132213115692,
"learning_rate": 8.276064127017262e-05,
"loss": 0.0361,
"step": 6180
},
{
"epoch": 4.274861878453039,
"grad_norm": 0.13318690657615662,
"learning_rate": 8.269814129456189e-05,
"loss": 0.0369,
"step": 6190
},
{
"epoch": 4.281767955801105,
"grad_norm": 0.11651228368282318,
"learning_rate": 8.263555192348143e-05,
"loss": 0.0363,
"step": 6200
},
{
"epoch": 4.288674033149171,
"grad_norm": 0.21030980348587036,
"learning_rate": 8.257287332804819e-05,
"loss": 0.0348,
"step": 6210
},
{
"epoch": 4.295580110497237,
"grad_norm": 0.17713765799999237,
"learning_rate": 8.251010567962307e-05,
"loss": 0.0339,
"step": 6220
},
{
"epoch": 4.302486187845304,
"grad_norm": 0.1809339076280594,
"learning_rate": 8.244724914981041e-05,
"loss": 0.0341,
"step": 6230
},
{
"epoch": 4.30939226519337,
"grad_norm": 0.18925215303897858,
"learning_rate": 8.238430391045757e-05,
"loss": 0.032,
"step": 6240
},
{
"epoch": 4.316298342541437,
"grad_norm": 0.21282361447811127,
"learning_rate": 8.232127013365445e-05,
"loss": 0.0334,
"step": 6250
},
{
"epoch": 4.323204419889503,
"grad_norm": 0.17454232275485992,
"learning_rate": 8.225814799173295e-05,
"loss": 0.0367,
"step": 6260
},
{
"epoch": 4.330110497237569,
"grad_norm": 0.22003421187400818,
"learning_rate": 8.219493765726663e-05,
"loss": 0.0356,
"step": 6270
},
{
"epoch": 4.337016574585635,
"grad_norm": 0.11667243391275406,
"learning_rate": 8.21316393030701e-05,
"loss": 0.0381,
"step": 6280
},
{
"epoch": 4.343922651933702,
"grad_norm": 0.17640861868858337,
"learning_rate": 8.206825310219865e-05,
"loss": 0.041,
"step": 6290
},
{
"epoch": 4.350828729281768,
"grad_norm": 0.18771280348300934,
"learning_rate": 8.200477922794776e-05,
"loss": 0.0297,
"step": 6300
},
{
"epoch": 4.357734806629834,
"grad_norm": 0.15554696321487427,
"learning_rate": 8.194121785385256e-05,
"loss": 0.035,
"step": 6310
},
{
"epoch": 4.3646408839779,
"grad_norm": 0.1682814359664917,
"learning_rate": 8.187756915368741e-05,
"loss": 0.0287,
"step": 6320
},
{
"epoch": 4.371546961325967,
"grad_norm": 0.12980349361896515,
"learning_rate": 8.181383330146544e-05,
"loss": 0.03,
"step": 6330
},
{
"epoch": 4.378453038674033,
"grad_norm": 0.1720675230026245,
"learning_rate": 8.175001047143804e-05,
"loss": 0.0339,
"step": 6340
},
{
"epoch": 4.3853591160221,
"grad_norm": 0.18164749443531036,
"learning_rate": 8.168610083809438e-05,
"loss": 0.0302,
"step": 6350
},
{
"epoch": 4.392265193370166,
"grad_norm": 0.1585766226053238,
"learning_rate": 8.162210457616095e-05,
"loss": 0.0341,
"step": 6360
},
{
"epoch": 4.399171270718232,
"grad_norm": 0.16067397594451904,
"learning_rate": 8.155802186060109e-05,
"loss": 0.0315,
"step": 6370
},
{
"epoch": 4.406077348066298,
"grad_norm": 0.15979492664337158,
"learning_rate": 8.149385286661453e-05,
"loss": 0.0321,
"step": 6380
},
{
"epoch": 4.412983425414365,
"grad_norm": 0.11252643913030624,
"learning_rate": 8.14295977696368e-05,
"loss": 0.0343,
"step": 6390
},
{
"epoch": 4.419889502762431,
"grad_norm": 0.19719359278678894,
"learning_rate": 8.13652567453389e-05,
"loss": 0.0335,
"step": 6400
},
{
"epoch": 4.426795580110497,
"grad_norm": 0.17754772305488586,
"learning_rate": 8.130082996962676e-05,
"loss": 0.0354,
"step": 6410
},
{
"epoch": 4.433701657458563,
"grad_norm": 0.15615013241767883,
"learning_rate": 8.123631761864068e-05,
"loss": 0.0347,
"step": 6420
},
{
"epoch": 4.44060773480663,
"grad_norm": 0.1586446464061737,
"learning_rate": 8.1171719868755e-05,
"loss": 0.032,
"step": 6430
},
{
"epoch": 4.447513812154696,
"grad_norm": 0.1573350876569748,
"learning_rate": 8.110703689657748e-05,
"loss": 0.0321,
"step": 6440
},
{
"epoch": 4.454419889502763,
"grad_norm": 0.14814695715904236,
"learning_rate": 8.104226887894892e-05,
"loss": 0.0303,
"step": 6450
},
{
"epoch": 4.461325966850829,
"grad_norm": 0.16024483740329742,
"learning_rate": 8.097741599294257e-05,
"loss": 0.031,
"step": 6460
},
{
"epoch": 4.468232044198895,
"grad_norm": 0.19926926493644714,
"learning_rate": 8.091247841586378e-05,
"loss": 0.0345,
"step": 6470
},
{
"epoch": 4.475138121546961,
"grad_norm": 0.22303733229637146,
"learning_rate": 8.084745632524939e-05,
"loss": 0.0308,
"step": 6480
},
{
"epoch": 4.482044198895028,
"grad_norm": 0.20035359263420105,
"learning_rate": 8.07823498988673e-05,
"loss": 0.0317,
"step": 6490
},
{
"epoch": 4.488950276243094,
"grad_norm": 0.17351382970809937,
"learning_rate": 8.071715931471602e-05,
"loss": 0.032,
"step": 6500
},
{
"epoch": 4.49585635359116,
"grad_norm": 0.2802712619304657,
"learning_rate": 8.06518847510241e-05,
"loss": 0.0326,
"step": 6510
},
{
"epoch": 4.502762430939226,
"grad_norm": 0.2340390384197235,
"learning_rate": 8.058652638624971e-05,
"loss": 0.033,
"step": 6520
},
{
"epoch": 4.509668508287293,
"grad_norm": 0.12652724981307983,
"learning_rate": 8.052108439908013e-05,
"loss": 0.0391,
"step": 6530
},
{
"epoch": 4.516574585635359,
"grad_norm": 0.15997207164764404,
"learning_rate": 8.045555896843125e-05,
"loss": 0.0332,
"step": 6540
},
{
"epoch": 4.523480662983426,
"grad_norm": 0.19305022060871124,
"learning_rate": 8.03899502734471e-05,
"loss": 0.0334,
"step": 6550
},
{
"epoch": 4.530386740331492,
"grad_norm": 0.18047112226486206,
"learning_rate": 8.032425849349931e-05,
"loss": 0.0363,
"step": 6560
},
{
"epoch": 4.537292817679558,
"grad_norm": 0.171234130859375,
"learning_rate": 8.025848380818674e-05,
"loss": 0.0386,
"step": 6570
},
{
"epoch": 4.544198895027624,
"grad_norm": 0.16757138073444366,
"learning_rate": 8.019262639733487e-05,
"loss": 0.0402,
"step": 6580
},
{
"epoch": 4.551104972375691,
"grad_norm": 0.18011297285556793,
"learning_rate": 8.012668644099531e-05,
"loss": 0.0336,
"step": 6590
},
{
"epoch": 4.558011049723757,
"grad_norm": 0.15435227751731873,
"learning_rate": 8.006066411944542e-05,
"loss": 0.032,
"step": 6600
},
{
"epoch": 4.564917127071823,
"grad_norm": 0.16600175201892853,
"learning_rate": 7.999455961318769e-05,
"loss": 0.0338,
"step": 6610
},
{
"epoch": 4.571823204419889,
"grad_norm": 0.18113788962364197,
"learning_rate": 7.992837310294932e-05,
"loss": 0.0306,
"step": 6620
},
{
"epoch": 4.578729281767956,
"grad_norm": 0.15672174096107483,
"learning_rate": 7.986210476968167e-05,
"loss": 0.0387,
"step": 6630
},
{
"epoch": 4.585635359116022,
"grad_norm": 0.14913024008274078,
"learning_rate": 7.97957547945599e-05,
"loss": 0.0355,
"step": 6640
},
{
"epoch": 4.592541436464089,
"grad_norm": 0.17059262096881866,
"learning_rate": 7.972932335898226e-05,
"loss": 0.0314,
"step": 6650
},
{
"epoch": 4.599447513812155,
"grad_norm": 0.15826822817325592,
"learning_rate": 7.966281064456975e-05,
"loss": 0.0325,
"step": 6660
},
{
"epoch": 4.606353591160221,
"grad_norm": 0.15611670911312103,
"learning_rate": 7.959621683316563e-05,
"loss": 0.0291,
"step": 6670
},
{
"epoch": 4.613259668508287,
"grad_norm": 0.16045376658439636,
"learning_rate": 7.952954210683481e-05,
"loss": 0.0356,
"step": 6680
},
{
"epoch": 4.620165745856354,
"grad_norm": 0.11433514207601547,
"learning_rate": 7.946278664786345e-05,
"loss": 0.031,
"step": 6690
},
{
"epoch": 4.62707182320442,
"grad_norm": 0.18826702237129211,
"learning_rate": 7.939595063875842e-05,
"loss": 0.0351,
"step": 6700
},
{
"epoch": 4.633977900552486,
"grad_norm": 0.15050998330116272,
"learning_rate": 7.932903426224683e-05,
"loss": 0.0291,
"step": 6710
},
{
"epoch": 4.640883977900552,
"grad_norm": 0.22147364914417267,
"learning_rate": 7.926203770127552e-05,
"loss": 0.0337,
"step": 6720
},
{
"epoch": 4.647790055248619,
"grad_norm": 0.22360795736312866,
"learning_rate": 7.919496113901046e-05,
"loss": 0.0339,
"step": 6730
},
{
"epoch": 4.654696132596685,
"grad_norm": 0.13165533542633057,
"learning_rate": 7.912780475883649e-05,
"loss": 0.0379,
"step": 6740
},
{
"epoch": 4.661602209944752,
"grad_norm": 0.16516909003257751,
"learning_rate": 7.906056874435652e-05,
"loss": 0.0296,
"step": 6750
},
{
"epoch": 4.668508287292818,
"grad_norm": 0.1891750693321228,
"learning_rate": 7.899325327939131e-05,
"loss": 0.028,
"step": 6760
},
{
"epoch": 4.675414364640884,
"grad_norm": 0.16898000240325928,
"learning_rate": 7.892585854797872e-05,
"loss": 0.031,
"step": 6770
},
{
"epoch": 4.68232044198895,
"grad_norm": 0.1583898961544037,
"learning_rate": 7.88583847343734e-05,
"loss": 0.0301,
"step": 6780
},
{
"epoch": 4.689226519337017,
"grad_norm": 0.13791148364543915,
"learning_rate": 7.879083202304616e-05,
"loss": 0.0324,
"step": 6790
},
{
"epoch": 4.696132596685083,
"grad_norm": 0.16449858248233795,
"learning_rate": 7.872320059868355e-05,
"loss": 0.0373,
"step": 6800
},
{
"epoch": 4.703038674033149,
"grad_norm": 0.12947428226470947,
"learning_rate": 7.865549064618729e-05,
"loss": 0.0304,
"step": 6810
},
{
"epoch": 4.709944751381215,
"grad_norm": 0.2018185257911682,
"learning_rate": 7.858770235067381e-05,
"loss": 0.0287,
"step": 6820
},
{
"epoch": 4.716850828729282,
"grad_norm": 0.1298319399356842,
"learning_rate": 7.851983589747374e-05,
"loss": 0.0275,
"step": 6830
},
{
"epoch": 4.723756906077348,
"grad_norm": 0.13833770155906677,
"learning_rate": 7.845189147213133e-05,
"loss": 0.0289,
"step": 6840
},
{
"epoch": 4.730662983425415,
"grad_norm": 0.17239126563072205,
"learning_rate": 7.838386926040407e-05,
"loss": 0.0295,
"step": 6850
},
{
"epoch": 4.737569060773481,
"grad_norm": 0.16230639815330505,
"learning_rate": 7.83157694482621e-05,
"loss": 0.0323,
"step": 6860
},
{
"epoch": 4.744475138121547,
"grad_norm": 0.14917680621147156,
"learning_rate": 7.824759222188768e-05,
"loss": 0.0358,
"step": 6870
},
{
"epoch": 4.751381215469613,
"grad_norm": 0.1046120896935463,
"learning_rate": 7.817933776767478e-05,
"loss": 0.0317,
"step": 6880
},
{
"epoch": 4.75828729281768,
"grad_norm": 0.08659262955188751,
"learning_rate": 7.811100627222842e-05,
"loss": 0.0349,
"step": 6890
},
{
"epoch": 4.765193370165746,
"grad_norm": 0.16613365709781647,
"learning_rate": 7.804259792236435e-05,
"loss": 0.0319,
"step": 6900
},
{
"epoch": 4.7720994475138125,
"grad_norm": 0.14178809523582458,
"learning_rate": 7.797411290510835e-05,
"loss": 0.0359,
"step": 6910
},
{
"epoch": 4.779005524861878,
"grad_norm": 0.13307057321071625,
"learning_rate": 7.790555140769586e-05,
"loss": 0.0369,
"step": 6920
},
{
"epoch": 4.785911602209945,
"grad_norm": 0.13005660474300385,
"learning_rate": 7.78369136175714e-05,
"loss": 0.0301,
"step": 6930
},
{
"epoch": 4.792817679558011,
"grad_norm": 0.15029692649841309,
"learning_rate": 7.776819972238806e-05,
"loss": 0.0288,
"step": 6940
},
{
"epoch": 4.7997237569060776,
"grad_norm": 0.1589047908782959,
"learning_rate": 7.7699409910007e-05,
"loss": 0.0338,
"step": 6950
},
{
"epoch": 4.806629834254144,
"grad_norm": 0.16635844111442566,
"learning_rate": 7.763054436849694e-05,
"loss": 0.032,
"step": 6960
},
{
"epoch": 4.81353591160221,
"grad_norm": 0.1771904081106186,
"learning_rate": 7.756160328613364e-05,
"loss": 0.0293,
"step": 6970
},
{
"epoch": 4.820441988950276,
"grad_norm": 0.1983332484960556,
"learning_rate": 7.749258685139942e-05,
"loss": 0.0354,
"step": 6980
},
{
"epoch": 4.827348066298343,
"grad_norm": 0.16270354390144348,
"learning_rate": 7.742349525298253e-05,
"loss": 0.0304,
"step": 6990
},
{
"epoch": 4.834254143646409,
"grad_norm": 0.22154197096824646,
"learning_rate": 7.735432867977679e-05,
"loss": 0.0327,
"step": 7000
},
{
"epoch": 4.8411602209944755,
"grad_norm": 0.1402016133069992,
"learning_rate": 7.728508732088096e-05,
"loss": 0.0337,
"step": 7010
},
{
"epoch": 4.848066298342541,
"grad_norm": 0.07997703552246094,
"learning_rate": 7.721577136559825e-05,
"loss": 0.0287,
"step": 7020
},
{
"epoch": 4.854972375690608,
"grad_norm": 0.16968661546707153,
"learning_rate": 7.714638100343588e-05,
"loss": 0.0316,
"step": 7030
},
{
"epoch": 4.861878453038674,
"grad_norm": 0.1546269953250885,
"learning_rate": 7.707691642410444e-05,
"loss": 0.0322,
"step": 7040
},
{
"epoch": 4.8687845303867405,
"grad_norm": 0.13078097999095917,
"learning_rate": 7.70073778175174e-05,
"loss": 0.0358,
"step": 7050
},
{
"epoch": 4.875690607734807,
"grad_norm": 0.1726975291967392,
"learning_rate": 7.69377653737907e-05,
"loss": 0.0317,
"step": 7060
},
{
"epoch": 4.882596685082873,
"grad_norm": 0.17169572412967682,
"learning_rate": 7.686807928324209e-05,
"loss": 0.0319,
"step": 7070
},
{
"epoch": 4.889502762430939,
"grad_norm": 0.17972759902477264,
"learning_rate": 7.679831973639065e-05,
"loss": 0.031,
"step": 7080
},
{
"epoch": 4.8964088397790055,
"grad_norm": 0.16098080575466156,
"learning_rate": 7.672848692395637e-05,
"loss": 0.0279,
"step": 7090
},
{
"epoch": 4.903314917127072,
"grad_norm": 0.18605658411979675,
"learning_rate": 7.665858103685944e-05,
"loss": 0.0316,
"step": 7100
},
{
"epoch": 4.9102209944751385,
"grad_norm": 0.16000963747501373,
"learning_rate": 7.658860226621991e-05,
"loss": 0.0302,
"step": 7110
},
{
"epoch": 4.917127071823204,
"grad_norm": 0.17880304157733917,
"learning_rate": 7.651855080335708e-05,
"loss": 0.031,
"step": 7120
},
{
"epoch": 4.9240331491712706,
"grad_norm": 0.15252450108528137,
"learning_rate": 7.644842683978896e-05,
"loss": 0.0286,
"step": 7130
},
{
"epoch": 4.930939226519337,
"grad_norm": 0.1624738872051239,
"learning_rate": 7.63782305672318e-05,
"loss": 0.0343,
"step": 7140
},
{
"epoch": 4.9378453038674035,
"grad_norm": 0.1616460680961609,
"learning_rate": 7.63079621775995e-05,
"loss": 0.0319,
"step": 7150
},
{
"epoch": 4.94475138121547,
"grad_norm": 0.1971554160118103,
"learning_rate": 7.623762186300319e-05,
"loss": 0.0362,
"step": 7160
},
{
"epoch": 4.951657458563536,
"grad_norm": 0.1854974925518036,
"learning_rate": 7.616720981575057e-05,
"loss": 0.0321,
"step": 7170
},
{
"epoch": 4.958563535911602,
"grad_norm": 0.1670568883419037,
"learning_rate": 7.609672622834552e-05,
"loss": 0.029,
"step": 7180
},
{
"epoch": 4.9654696132596685,
"grad_norm": 0.20243513584136963,
"learning_rate": 7.602617129348747e-05,
"loss": 0.0357,
"step": 7190
},
{
"epoch": 4.972375690607735,
"grad_norm": 0.13823777437210083,
"learning_rate": 7.595554520407088e-05,
"loss": 0.0307,
"step": 7200
},
{
"epoch": 4.9792817679558015,
"grad_norm": 0.24209170043468475,
"learning_rate": 7.588484815318484e-05,
"loss": 0.0281,
"step": 7210
},
{
"epoch": 4.986187845303867,
"grad_norm": 0.16390523314476013,
"learning_rate": 7.581408033411234e-05,
"loss": 0.0368,
"step": 7220
},
{
"epoch": 4.9930939226519335,
"grad_norm": 0.22568488121032715,
"learning_rate": 7.574324194032995e-05,
"loss": 0.0335,
"step": 7230
},
{
"epoch": 5.0,
"grad_norm": 0.17507368326187134,
"learning_rate": 7.567233316550705e-05,
"loss": 0.0314,
"step": 7240
},
{
"epoch": 5.0069060773480665,
"grad_norm": 0.1915036290884018,
"learning_rate": 7.560135420350562e-05,
"loss": 0.0275,
"step": 7250
},
{
"epoch": 5.013812154696133,
"grad_norm": 0.13837654888629913,
"learning_rate": 7.553030524837935e-05,
"loss": 0.0292,
"step": 7260
},
{
"epoch": 5.0207182320441985,
"grad_norm": 0.11749670654535294,
"learning_rate": 7.545918649437341e-05,
"loss": 0.0328,
"step": 7270
},
{
"epoch": 5.027624309392265,
"grad_norm": 0.11151426285505295,
"learning_rate": 7.538799813592377e-05,
"loss": 0.0277,
"step": 7280
},
{
"epoch": 5.0345303867403315,
"grad_norm": 0.13570214807987213,
"learning_rate": 7.531674036765662e-05,
"loss": 0.0298,
"step": 7290
},
{
"epoch": 5.041436464088398,
"grad_norm": 0.15404485166072845,
"learning_rate": 7.524541338438807e-05,
"loss": 0.0336,
"step": 7300
},
{
"epoch": 5.048342541436464,
"grad_norm": 0.1855136603116989,
"learning_rate": 7.517401738112328e-05,
"loss": 0.0326,
"step": 7310
},
{
"epoch": 5.05524861878453,
"grad_norm": 0.1743932068347931,
"learning_rate": 7.510255255305628e-05,
"loss": 0.0272,
"step": 7320
},
{
"epoch": 5.0621546961325965,
"grad_norm": 0.1550290435552597,
"learning_rate": 7.503101909556911e-05,
"loss": 0.0314,
"step": 7330
},
{
"epoch": 5.069060773480663,
"grad_norm": 0.14492987096309662,
"learning_rate": 7.495941720423154e-05,
"loss": 0.0304,
"step": 7340
},
{
"epoch": 5.0759668508287294,
"grad_norm": 0.16522392630577087,
"learning_rate": 7.488774707480042e-05,
"loss": 0.0296,
"step": 7350
},
{
"epoch": 5.082872928176796,
"grad_norm": 0.2069702297449112,
"learning_rate": 7.481600890321911e-05,
"loss": 0.0261,
"step": 7360
},
{
"epoch": 5.0897790055248615,
"grad_norm": 0.17909006774425507,
"learning_rate": 7.474420288561708e-05,
"loss": 0.0306,
"step": 7370
},
{
"epoch": 5.096685082872928,
"grad_norm": 0.1516609787940979,
"learning_rate": 7.467232921830921e-05,
"loss": 0.0308,
"step": 7380
},
{
"epoch": 5.1035911602209945,
"grad_norm": 0.12364234030246735,
"learning_rate": 7.460038809779537e-05,
"loss": 0.0321,
"step": 7390
},
{
"epoch": 5.110497237569061,
"grad_norm": 0.16813722252845764,
"learning_rate": 7.452837972075983e-05,
"loss": 0.0293,
"step": 7400
},
{
"epoch": 5.117403314917127,
"grad_norm": 0.21152564883232117,
"learning_rate": 7.445630428407074e-05,
"loss": 0.0256,
"step": 7410
},
{
"epoch": 5.124309392265193,
"grad_norm": 0.233183816075325,
"learning_rate": 7.43841619847796e-05,
"loss": 0.0328,
"step": 7420
},
{
"epoch": 5.1312154696132595,
"grad_norm": 0.16253533959388733,
"learning_rate": 7.431195302012072e-05,
"loss": 0.028,
"step": 7430
},
{
"epoch": 5.138121546961326,
"grad_norm": 0.11992829293012619,
"learning_rate": 7.423967758751061e-05,
"loss": 0.0327,
"step": 7440
},
{
"epoch": 5.145027624309392,
"grad_norm": 0.08840538561344147,
"learning_rate": 7.416733588454758e-05,
"loss": 0.0236,
"step": 7450
},
{
"epoch": 5.151933701657459,
"grad_norm": 0.18914037942886353,
"learning_rate": 7.409492810901106e-05,
"loss": 0.0306,
"step": 7460
},
{
"epoch": 5.1588397790055245,
"grad_norm": 0.203793466091156,
"learning_rate": 7.402245445886116e-05,
"loss": 0.0298,
"step": 7470
},
{
"epoch": 5.165745856353591,
"grad_norm": 0.15535494685173035,
"learning_rate": 7.394991513223806e-05,
"loss": 0.0328,
"step": 7480
},
{
"epoch": 5.172651933701657,
"grad_norm": 0.1621917188167572,
"learning_rate": 7.38773103274615e-05,
"loss": 0.0327,
"step": 7490
},
{
"epoch": 5.179558011049724,
"grad_norm": 0.15905600786209106,
"learning_rate": 7.380464024303028e-05,
"loss": 0.0306,
"step": 7500
},
{
"epoch": 5.18646408839779,
"grad_norm": 0.12775392830371857,
"learning_rate": 7.373190507762162e-05,
"loss": 0.0277,
"step": 7510
},
{
"epoch": 5.193370165745856,
"grad_norm": 0.20729433000087738,
"learning_rate": 7.365910503009066e-05,
"loss": 0.0305,
"step": 7520
},
{
"epoch": 5.2002762430939224,
"grad_norm": 0.1747429072856903,
"learning_rate": 7.358624029946996e-05,
"loss": 0.0318,
"step": 7530
},
{
"epoch": 5.207182320441989,
"grad_norm": 0.12930446863174438,
"learning_rate": 7.351331108496893e-05,
"loss": 0.0264,
"step": 7540
},
{
"epoch": 5.214088397790055,
"grad_norm": 0.16786518692970276,
"learning_rate": 7.344031758597325e-05,
"loss": 0.0276,
"step": 7550
},
{
"epoch": 5.220994475138122,
"grad_norm": 0.15410704910755157,
"learning_rate": 7.336726000204435e-05,
"loss": 0.0274,
"step": 7560
},
{
"epoch": 5.2279005524861875,
"grad_norm": 0.11539943516254425,
"learning_rate": 7.32941385329189e-05,
"loss": 0.0278,
"step": 7570
},
{
"epoch": 5.234806629834254,
"grad_norm": 0.2146443873643875,
"learning_rate": 7.322095337850816e-05,
"loss": 0.035,
"step": 7580
},
{
"epoch": 5.24171270718232,
"grad_norm": 0.17688392102718353,
"learning_rate": 7.314770473889758e-05,
"loss": 0.0299,
"step": 7590
},
{
"epoch": 5.248618784530387,
"grad_norm": 0.1647810935974121,
"learning_rate": 7.307439281434615e-05,
"loss": 0.034,
"step": 7600
},
{
"epoch": 5.255524861878453,
"grad_norm": 0.14570735394954681,
"learning_rate": 7.300101780528585e-05,
"loss": 0.0278,
"step": 7610
},
{
"epoch": 5.262430939226519,
"grad_norm": 0.14859934151172638,
"learning_rate": 7.292757991232117e-05,
"loss": 0.0299,
"step": 7620
},
{
"epoch": 5.269337016574585,
"grad_norm": 0.19481755793094635,
"learning_rate": 7.285407933622848e-05,
"loss": 0.0295,
"step": 7630
},
{
"epoch": 5.276243093922652,
"grad_norm": 0.18291620910167694,
"learning_rate": 7.278051627795557e-05,
"loss": 0.0309,
"step": 7640
},
{
"epoch": 5.283149171270718,
"grad_norm": 0.135209321975708,
"learning_rate": 7.270689093862105e-05,
"loss": 0.0291,
"step": 7650
},
{
"epoch": 5.290055248618785,
"grad_norm": 0.15347661077976227,
"learning_rate": 7.263320351951374e-05,
"loss": 0.029,
"step": 7660
},
{
"epoch": 5.29696132596685,
"grad_norm": 0.17232327163219452,
"learning_rate": 7.255945422209227e-05,
"loss": 0.0304,
"step": 7670
},
{
"epoch": 5.303867403314917,
"grad_norm": 0.10398121923208237,
"learning_rate": 7.248564324798437e-05,
"loss": 0.0269,
"step": 7680
},
{
"epoch": 5.310773480662983,
"grad_norm": 0.20927193760871887,
"learning_rate": 7.241177079898644e-05,
"loss": 0.0303,
"step": 7690
},
{
"epoch": 5.31767955801105,
"grad_norm": 0.15718117356300354,
"learning_rate": 7.233783707706295e-05,
"loss": 0.0303,
"step": 7700
},
{
"epoch": 5.324585635359116,
"grad_norm": 0.15242931246757507,
"learning_rate": 7.226384228434586e-05,
"loss": 0.0303,
"step": 7710
},
{
"epoch": 5.331491712707182,
"grad_norm": 0.2130342721939087,
"learning_rate": 7.21897866231341e-05,
"loss": 0.0268,
"step": 7720
},
{
"epoch": 5.338397790055248,
"grad_norm": 0.1480337530374527,
"learning_rate": 7.211567029589303e-05,
"loss": 0.0312,
"step": 7730
},
{
"epoch": 5.345303867403315,
"grad_norm": 0.12969496846199036,
"learning_rate": 7.204149350525387e-05,
"loss": 0.0276,
"step": 7740
},
{
"epoch": 5.352209944751381,
"grad_norm": 0.1786733865737915,
"learning_rate": 7.196725645401309e-05,
"loss": 0.0285,
"step": 7750
},
{
"epoch": 5.359116022099448,
"grad_norm": 0.1749386489391327,
"learning_rate": 7.1892959345132e-05,
"loss": 0.0262,
"step": 7760
},
{
"epoch": 5.366022099447513,
"grad_norm": 0.11476799100637436,
"learning_rate": 7.181860238173605e-05,
"loss": 0.0267,
"step": 7770
},
{
"epoch": 5.37292817679558,
"grad_norm": 0.17516183853149414,
"learning_rate": 7.174418576711432e-05,
"loss": 0.0302,
"step": 7780
},
{
"epoch": 5.379834254143646,
"grad_norm": 0.18794813752174377,
"learning_rate": 7.1669709704719e-05,
"loss": 0.0285,
"step": 7790
},
{
"epoch": 5.386740331491713,
"grad_norm": 0.12954457104206085,
"learning_rate": 7.159517439816481e-05,
"loss": 0.0313,
"step": 7800
},
{
"epoch": 5.393646408839779,
"grad_norm": 0.14184238016605377,
"learning_rate": 7.152058005122842e-05,
"loss": 0.0278,
"step": 7810
},
{
"epoch": 5.400552486187845,
"grad_norm": 0.1285337507724762,
"learning_rate": 7.144592686784793e-05,
"loss": 0.0306,
"step": 7820
},
{
"epoch": 5.407458563535911,
"grad_norm": 0.14311516284942627,
"learning_rate": 7.137121505212229e-05,
"loss": 0.039,
"step": 7830
},
{
"epoch": 5.414364640883978,
"grad_norm": 0.17362003028392792,
"learning_rate": 7.129644480831077e-05,
"loss": 0.0284,
"step": 7840
},
{
"epoch": 5.421270718232044,
"grad_norm": 0.10782770067453384,
"learning_rate": 7.122161634083234e-05,
"loss": 0.0284,
"step": 7850
},
{
"epoch": 5.428176795580111,
"grad_norm": 0.1495029479265213,
"learning_rate": 7.114672985426516e-05,
"loss": 0.0274,
"step": 7860
},
{
"epoch": 5.435082872928177,
"grad_norm": 0.14453409612178802,
"learning_rate": 7.107178555334606e-05,
"loss": 0.0274,
"step": 7870
},
{
"epoch": 5.441988950276243,
"grad_norm": 0.2002222239971161,
"learning_rate": 7.099678364296989e-05,
"loss": 0.0337,
"step": 7880
},
{
"epoch": 5.448895027624309,
"grad_norm": 0.1697368025779724,
"learning_rate": 7.0921724328189e-05,
"loss": 0.0303,
"step": 7890
},
{
"epoch": 5.455801104972376,
"grad_norm": 0.14910706877708435,
"learning_rate": 7.084660781421268e-05,
"loss": 0.0299,
"step": 7900
},
{
"epoch": 5.462707182320442,
"grad_norm": 0.1361190676689148,
"learning_rate": 7.077143430640662e-05,
"loss": 0.0316,
"step": 7910
},
{
"epoch": 5.469613259668508,
"grad_norm": 0.172740176320076,
"learning_rate": 7.069620401029232e-05,
"loss": 0.0323,
"step": 7920
},
{
"epoch": 5.476519337016574,
"grad_norm": 0.2003297507762909,
"learning_rate": 7.062091713154655e-05,
"loss": 0.0281,
"step": 7930
},
{
"epoch": 5.483425414364641,
"grad_norm": 0.19455312192440033,
"learning_rate": 7.054557387600075e-05,
"loss": 0.0303,
"step": 7940
},
{
"epoch": 5.490331491712707,
"grad_norm": 0.13216443359851837,
"learning_rate": 7.04701744496405e-05,
"loss": 0.036,
"step": 7950
},
{
"epoch": 5.497237569060774,
"grad_norm": 0.1445705145597458,
"learning_rate": 7.039471905860495e-05,
"loss": 0.0339,
"step": 7960
},
{
"epoch": 5.50414364640884,
"grad_norm": 0.15850737690925598,
"learning_rate": 7.031920790918628e-05,
"loss": 0.0286,
"step": 7970
},
{
"epoch": 5.511049723756906,
"grad_norm": 0.18733613193035126,
"learning_rate": 7.024364120782906e-05,
"loss": 0.0377,
"step": 7980
},
{
"epoch": 5.517955801104972,
"grad_norm": 0.13512170314788818,
"learning_rate": 7.016801916112978e-05,
"loss": 0.031,
"step": 7990
},
{
"epoch": 5.524861878453039,
"grad_norm": 0.14883825182914734,
"learning_rate": 7.009234197583623e-05,
"loss": 0.0305,
"step": 8000
},
{
"epoch": 5.531767955801105,
"grad_norm": 0.14048552513122559,
"learning_rate": 7.001660985884692e-05,
"loss": 0.0287,
"step": 8010
},
{
"epoch": 5.538674033149171,
"grad_norm": 0.14698736369609833,
"learning_rate": 6.994082301721063e-05,
"loss": 0.0333,
"step": 8020
},
{
"epoch": 5.545580110497237,
"grad_norm": 0.26427164673805237,
"learning_rate": 6.986498165812563e-05,
"loss": 0.0319,
"step": 8030
},
{
"epoch": 5.552486187845304,
"grad_norm": 0.12380706518888474,
"learning_rate": 6.978908598893932e-05,
"loss": 0.0258,
"step": 8040
},
{
"epoch": 5.55939226519337,
"grad_norm": 0.17819596827030182,
"learning_rate": 6.971313621714756e-05,
"loss": 0.0275,
"step": 8050
},
{
"epoch": 5.566298342541437,
"grad_norm": 0.1824468970298767,
"learning_rate": 6.96371325503941e-05,
"loss": 0.0293,
"step": 8060
},
{
"epoch": 5.573204419889503,
"grad_norm": 0.12114512920379639,
"learning_rate": 6.956107519647014e-05,
"loss": 0.0284,
"step": 8070
},
{
"epoch": 5.580110497237569,
"grad_norm": 0.12274518609046936,
"learning_rate": 6.94849643633135e-05,
"loss": 0.028,
"step": 8080
},
{
"epoch": 5.587016574585635,
"grad_norm": 0.1467324197292328,
"learning_rate": 6.940880025900834e-05,
"loss": 0.0266,
"step": 8090
},
{
"epoch": 5.593922651933702,
"grad_norm": 0.18122711777687073,
"learning_rate": 6.933258309178438e-05,
"loss": 0.0239,
"step": 8100
},
{
"epoch": 5.600828729281768,
"grad_norm": 0.14374437928199768,
"learning_rate": 6.925631307001646e-05,
"loss": 0.0296,
"step": 8110
},
{
"epoch": 5.607734806629834,
"grad_norm": 0.16520264744758606,
"learning_rate": 6.91799904022239e-05,
"loss": 0.028,
"step": 8120
},
{
"epoch": 5.6146408839779,
"grad_norm": 0.13161131739616394,
"learning_rate": 6.910361529706997e-05,
"loss": 0.0245,
"step": 8130
},
{
"epoch": 5.621546961325967,
"grad_norm": 0.13930374383926392,
"learning_rate": 6.902718796336131e-05,
"loss": 0.0299,
"step": 8140
},
{
"epoch": 5.628453038674033,
"grad_norm": 0.12098735570907593,
"learning_rate": 6.895070861004729e-05,
"loss": 0.0308,
"step": 8150
},
{
"epoch": 5.6353591160221,
"grad_norm": 0.19493934512138367,
"learning_rate": 6.887417744621956e-05,
"loss": 0.0288,
"step": 8160
},
{
"epoch": 5.642265193370166,
"grad_norm": 0.10348235070705414,
"learning_rate": 6.87975946811114e-05,
"loss": 0.0274,
"step": 8170
},
{
"epoch": 5.649171270718232,
"grad_norm": 0.19127143919467926,
"learning_rate": 6.872096052409718e-05,
"loss": 0.0284,
"step": 8180
},
{
"epoch": 5.656077348066298,
"grad_norm": 0.17267738282680511,
"learning_rate": 6.864427518469174e-05,
"loss": 0.0303,
"step": 8190
},
{
"epoch": 5.662983425414365,
"grad_norm": 0.14508427679538727,
"learning_rate": 6.856753887254986e-05,
"loss": 0.0262,
"step": 8200
},
{
"epoch": 5.669889502762431,
"grad_norm": 0.15535277128219604,
"learning_rate": 6.849075179746572e-05,
"loss": 0.0304,
"step": 8210
},
{
"epoch": 5.676795580110497,
"grad_norm": 0.1699511557817459,
"learning_rate": 6.841391416937221e-05,
"loss": 0.0321,
"step": 8220
},
{
"epoch": 5.683701657458563,
"grad_norm": 0.17711235582828522,
"learning_rate": 6.833702619834053e-05,
"loss": 0.0276,
"step": 8230
},
{
"epoch": 5.69060773480663,
"grad_norm": 0.21901249885559082,
"learning_rate": 6.82600880945794e-05,
"loss": 0.029,
"step": 8240
},
{
"epoch": 5.697513812154696,
"grad_norm": 0.11280661821365356,
"learning_rate": 6.818310006843468e-05,
"loss": 0.0237,
"step": 8250
},
{
"epoch": 5.704419889502763,
"grad_norm": 0.12669578194618225,
"learning_rate": 6.810606233038868e-05,
"loss": 0.0273,
"step": 8260
},
{
"epoch": 5.711325966850829,
"grad_norm": 0.15255232155323029,
"learning_rate": 6.802897509105966e-05,
"loss": 0.0331,
"step": 8270
},
{
"epoch": 5.718232044198895,
"grad_norm": 0.17251603305339813,
"learning_rate": 6.79518385612012e-05,
"loss": 0.0318,
"step": 8280
},
{
"epoch": 5.725138121546961,
"grad_norm": 0.15754754841327667,
"learning_rate": 6.787465295170157e-05,
"loss": 0.0272,
"step": 8290
},
{
"epoch": 5.732044198895028,
"grad_norm": 0.13025134801864624,
"learning_rate": 6.779741847358332e-05,
"loss": 0.0246,
"step": 8300
},
{
"epoch": 5.738950276243094,
"grad_norm": 0.1261071413755417,
"learning_rate": 6.772013533800256e-05,
"loss": 0.0305,
"step": 8310
},
{
"epoch": 5.74585635359116,
"grad_norm": 0.15836410224437714,
"learning_rate": 6.764280375624843e-05,
"loss": 0.0307,
"step": 8320
},
{
"epoch": 5.752762430939226,
"grad_norm": 0.21966996788978577,
"learning_rate": 6.756542393974252e-05,
"loss": 0.0309,
"step": 8330
},
{
"epoch": 5.759668508287293,
"grad_norm": 0.16588178277015686,
"learning_rate": 6.748799610003828e-05,
"loss": 0.0266,
"step": 8340
},
{
"epoch": 5.766574585635359,
"grad_norm": 0.11811785399913788,
"learning_rate": 6.741052044882048e-05,
"loss": 0.0301,
"step": 8350
},
{
"epoch": 5.773480662983426,
"grad_norm": 0.09797261655330658,
"learning_rate": 6.73329971979046e-05,
"loss": 0.0263,
"step": 8360
},
{
"epoch": 5.780386740331492,
"grad_norm": 0.12955108284950256,
"learning_rate": 6.725542655923625e-05,
"loss": 0.0262,
"step": 8370
},
{
"epoch": 5.787292817679558,
"grad_norm": 0.14564071595668793,
"learning_rate": 6.717780874489057e-05,
"loss": 0.0287,
"step": 8380
},
{
"epoch": 5.794198895027624,
"grad_norm": 0.19383345544338226,
"learning_rate": 6.710014396707172e-05,
"loss": 0.028,
"step": 8390
},
{
"epoch": 5.801104972375691,
"grad_norm": 0.11382323503494263,
"learning_rate": 6.702243243811221e-05,
"loss": 0.0218,
"step": 8400
},
{
"epoch": 5.808011049723757,
"grad_norm": 0.15130683779716492,
"learning_rate": 6.694467437047244e-05,
"loss": 0.0265,
"step": 8410
},
{
"epoch": 5.814917127071823,
"grad_norm": 0.15816499292850494,
"learning_rate": 6.686686997673997e-05,
"loss": 0.0277,
"step": 8420
},
{
"epoch": 5.821823204419889,
"grad_norm": 0.1593073159456253,
"learning_rate": 6.678901946962903e-05,
"loss": 0.0262,
"step": 8430
},
{
"epoch": 5.828729281767956,
"grad_norm": 0.11927701532840729,
"learning_rate": 6.671112306197996e-05,
"loss": 0.024,
"step": 8440
},
{
"epoch": 5.835635359116022,
"grad_norm": 0.1482495218515396,
"learning_rate": 6.663318096675854e-05,
"loss": 0.0269,
"step": 8450
},
{
"epoch": 5.842541436464089,
"grad_norm": 0.1422957479953766,
"learning_rate": 6.655519339705552e-05,
"loss": 0.0229,
"step": 8460
},
{
"epoch": 5.849447513812155,
"grad_norm": 0.12634189426898956,
"learning_rate": 6.647716056608588e-05,
"loss": 0.026,
"step": 8470
},
{
"epoch": 5.856353591160221,
"grad_norm": 0.21049568057060242,
"learning_rate": 6.639908268718843e-05,
"loss": 0.0302,
"step": 8480
},
{
"epoch": 5.863259668508287,
"grad_norm": 0.12989932298660278,
"learning_rate": 6.632095997382514e-05,
"loss": 0.0301,
"step": 8490
},
{
"epoch": 5.870165745856354,
"grad_norm": 0.2166806012392044,
"learning_rate": 6.624279263958047e-05,
"loss": 0.0226,
"step": 8500
},
{
"epoch": 5.87707182320442,
"grad_norm": 0.2074391394853592,
"learning_rate": 6.616458089816097e-05,
"loss": 0.0307,
"step": 8510
},
{
"epoch": 5.883977900552486,
"grad_norm": 0.15048719942569733,
"learning_rate": 6.608632496339454e-05,
"loss": 0.0283,
"step": 8520
},
{
"epoch": 5.890883977900552,
"grad_norm": 0.12132376432418823,
"learning_rate": 6.600802504922988e-05,
"loss": 0.0273,
"step": 8530
},
{
"epoch": 5.897790055248619,
"grad_norm": 0.13046051561832428,
"learning_rate": 6.592968136973604e-05,
"loss": 0.0263,
"step": 8540
},
{
"epoch": 5.904696132596685,
"grad_norm": 0.16431337594985962,
"learning_rate": 6.585129413910159e-05,
"loss": 0.0312,
"step": 8550
},
{
"epoch": 5.911602209944752,
"grad_norm": 0.15263038873672485,
"learning_rate": 6.577286357163424e-05,
"loss": 0.0292,
"step": 8560
},
{
"epoch": 5.918508287292818,
"grad_norm": 0.1429017037153244,
"learning_rate": 6.569438988176018e-05,
"loss": 0.0295,
"step": 8570
},
{
"epoch": 5.925414364640884,
"grad_norm": 0.1810126006603241,
"learning_rate": 6.561587328402347e-05,
"loss": 0.0264,
"step": 8580
},
{
"epoch": 5.93232044198895,
"grad_norm": 0.1651168018579483,
"learning_rate": 6.553731399308549e-05,
"loss": 0.0264,
"step": 8590
},
{
"epoch": 5.939226519337017,
"grad_norm": 0.13501553237438202,
"learning_rate": 6.545871222372436e-05,
"loss": 0.032,
"step": 8600
},
{
"epoch": 5.946132596685083,
"grad_norm": 0.1565493941307068,
"learning_rate": 6.538006819083426e-05,
"loss": 0.0269,
"step": 8610
},
{
"epoch": 5.953038674033149,
"grad_norm": 0.16880425810813904,
"learning_rate": 6.530138210942505e-05,
"loss": 0.0304,
"step": 8620
},
{
"epoch": 5.959944751381215,
"grad_norm": 0.12984171509742737,
"learning_rate": 6.522265419462141e-05,
"loss": 0.0264,
"step": 8630
},
{
"epoch": 5.966850828729282,
"grad_norm": 0.11698273569345474,
"learning_rate": 6.514388466166248e-05,
"loss": 0.0278,
"step": 8640
},
{
"epoch": 5.973756906077348,
"grad_norm": 0.20552147924900055,
"learning_rate": 6.506507372590119e-05,
"loss": 0.0288,
"step": 8650
},
{
"epoch": 5.980662983425415,
"grad_norm": 0.14010268449783325,
"learning_rate": 6.498622160280355e-05,
"loss": 0.0235,
"step": 8660
},
{
"epoch": 5.987569060773481,
"grad_norm": 0.17830894887447357,
"learning_rate": 6.490732850794832e-05,
"loss": 0.0301,
"step": 8670
},
{
"epoch": 5.994475138121547,
"grad_norm": 0.1695212870836258,
"learning_rate": 6.482839465702616e-05,
"loss": 0.0315,
"step": 8680
},
{
"epoch": 6.001381215469613,
"grad_norm": 0.17388391494750977,
"learning_rate": 6.474942026583923e-05,
"loss": 0.0276,
"step": 8690
},
{
"epoch": 6.00828729281768,
"grad_norm": 0.16767103970050812,
"learning_rate": 6.467040555030052e-05,
"loss": 0.0259,
"step": 8700
},
{
"epoch": 6.015193370165746,
"grad_norm": 0.13405068218708038,
"learning_rate": 6.459135072643321e-05,
"loss": 0.0223,
"step": 8710
},
{
"epoch": 6.0220994475138125,
"grad_norm": 0.11968281120061874,
"learning_rate": 6.451225601037019e-05,
"loss": 0.028,
"step": 8720
},
{
"epoch": 6.029005524861878,
"grad_norm": 0.13727779686450958,
"learning_rate": 6.443312161835338e-05,
"loss": 0.0285,
"step": 8730
},
{
"epoch": 6.035911602209945,
"grad_norm": 0.140719473361969,
"learning_rate": 6.43539477667332e-05,
"loss": 0.027,
"step": 8740
},
{
"epoch": 6.042817679558011,
"grad_norm": 0.11544650793075562,
"learning_rate": 6.427473467196793e-05,
"loss": 0.0303,
"step": 8750
},
{
"epoch": 6.0497237569060776,
"grad_norm": 0.1590787172317505,
"learning_rate": 6.419548255062315e-05,
"loss": 0.0281,
"step": 8760
},
{
"epoch": 6.056629834254144,
"grad_norm": 0.1701236516237259,
"learning_rate": 6.411619161937112e-05,
"loss": 0.0282,
"step": 8770
},
{
"epoch": 6.06353591160221,
"grad_norm": 0.19218502938747406,
"learning_rate": 6.403686209499022e-05,
"loss": 0.0308,
"step": 8780
},
{
"epoch": 6.070441988950276,
"grad_norm": 0.18104462325572968,
"learning_rate": 6.395749419436437e-05,
"loss": 0.0299,
"step": 8790
},
{
"epoch": 6.077348066298343,
"grad_norm": 0.11846525967121124,
"learning_rate": 6.387808813448234e-05,
"loss": 0.0236,
"step": 8800
},
{
"epoch": 6.084254143646409,
"grad_norm": 0.11504874378442764,
"learning_rate": 6.37986441324373e-05,
"loss": 0.0322,
"step": 8810
},
{
"epoch": 6.0911602209944755,
"grad_norm": 0.15801477432250977,
"learning_rate": 6.37191624054261e-05,
"loss": 0.0251,
"step": 8820
},
{
"epoch": 6.098066298342541,
"grad_norm": 0.2594138979911804,
"learning_rate": 6.363964317074872e-05,
"loss": 0.0332,
"step": 8830
},
{
"epoch": 6.104972375690608,
"grad_norm": 0.1887638121843338,
"learning_rate": 6.356008664580776e-05,
"loss": 0.0251,
"step": 8840
},
{
"epoch": 6.111878453038674,
"grad_norm": 0.19078470766544342,
"learning_rate": 6.348049304810771e-05,
"loss": 0.0272,
"step": 8850
},
{
"epoch": 6.1187845303867405,
"grad_norm": 0.17975382506847382,
"learning_rate": 6.340086259525442e-05,
"loss": 0.0314,
"step": 8860
},
{
"epoch": 6.125690607734807,
"grad_norm": 0.1428086757659912,
"learning_rate": 6.332119550495448e-05,
"loss": 0.0259,
"step": 8870
},
{
"epoch": 6.132596685082873,
"grad_norm": 0.14597752690315247,
"learning_rate": 6.324149199501473e-05,
"loss": 0.0264,
"step": 8880
},
{
"epoch": 6.139502762430939,
"grad_norm": 0.15924973785877228,
"learning_rate": 6.316175228334146e-05,
"loss": 0.0258,
"step": 8890
},
{
"epoch": 6.1464088397790055,
"grad_norm": 0.15013249218463898,
"learning_rate": 6.308197658794003e-05,
"loss": 0.0279,
"step": 8900
},
{
"epoch": 6.153314917127072,
"grad_norm": 0.17688699066638947,
"learning_rate": 6.300216512691417e-05,
"loss": 0.0262,
"step": 8910
},
{
"epoch": 6.1602209944751385,
"grad_norm": 0.15361438691616058,
"learning_rate": 6.292231811846532e-05,
"loss": 0.0242,
"step": 8920
},
{
"epoch": 6.167127071823204,
"grad_norm": 0.10727053880691528,
"learning_rate": 6.284243578089217e-05,
"loss": 0.0242,
"step": 8930
},
{
"epoch": 6.1740331491712706,
"grad_norm": 0.17150332033634186,
"learning_rate": 6.276251833258999e-05,
"loss": 0.0284,
"step": 8940
},
{
"epoch": 6.180939226519337,
"grad_norm": 0.2145872861146927,
"learning_rate": 6.268256599205003e-05,
"loss": 0.026,
"step": 8950
},
{
"epoch": 6.1878453038674035,
"grad_norm": 0.13672959804534912,
"learning_rate": 6.260257897785892e-05,
"loss": 0.0257,
"step": 8960
},
{
"epoch": 6.19475138121547,
"grad_norm": 0.1518806666135788,
"learning_rate": 6.252255750869811e-05,
"loss": 0.0256,
"step": 8970
},
{
"epoch": 6.201657458563536,
"grad_norm": 0.13806003332138062,
"learning_rate": 6.244250180334325e-05,
"loss": 0.0281,
"step": 8980
},
{
"epoch": 6.208563535911602,
"grad_norm": 0.1294434815645218,
"learning_rate": 6.236241208066356e-05,
"loss": 0.0294,
"step": 8990
},
{
"epoch": 6.2154696132596685,
"grad_norm": 0.13486751914024353,
"learning_rate": 6.228228855962133e-05,
"loss": 0.033,
"step": 9000
},
{
"epoch": 6.222375690607735,
"grad_norm": 0.13604870438575745,
"learning_rate": 6.220213145927115e-05,
"loss": 0.0247,
"step": 9010
},
{
"epoch": 6.2292817679558015,
"grad_norm": 0.22754208743572235,
"learning_rate": 6.212194099875951e-05,
"loss": 0.0239,
"step": 9020
},
{
"epoch": 6.236187845303867,
"grad_norm": 0.18670392036437988,
"learning_rate": 6.204171739732405e-05,
"loss": 0.0267,
"step": 9030
},
{
"epoch": 6.2430939226519335,
"grad_norm": 0.16089409589767456,
"learning_rate": 6.196146087429303e-05,
"loss": 0.0236,
"step": 9040
},
{
"epoch": 6.25,
"grad_norm": 0.11962158232927322,
"learning_rate": 6.188117164908474e-05,
"loss": 0.028,
"step": 9050
},
{
"epoch": 6.2569060773480665,
"grad_norm": 0.15136407315731049,
"learning_rate": 6.180084994120684e-05,
"loss": 0.0258,
"step": 9060
},
{
"epoch": 6.263812154696133,
"grad_norm": 0.17362581193447113,
"learning_rate": 6.17204959702558e-05,
"loss": 0.0235,
"step": 9070
},
{
"epoch": 6.2707182320441985,
"grad_norm": 0.12609721720218658,
"learning_rate": 6.164010995591635e-05,
"loss": 0.0274,
"step": 9080
},
{
"epoch": 6.277624309392265,
"grad_norm": 0.13596007227897644,
"learning_rate": 6.155969211796076e-05,
"loss": 0.0263,
"step": 9090
},
{
"epoch": 6.2845303867403315,
"grad_norm": 0.12467209994792938,
"learning_rate": 6.147924267624829e-05,
"loss": 0.0227,
"step": 9100
},
{
"epoch": 6.291436464088398,
"grad_norm": 0.2607831656932831,
"learning_rate": 6.13987618507247e-05,
"loss": 0.0256,
"step": 9110
},
{
"epoch": 6.298342541436464,
"grad_norm": 0.12686564028263092,
"learning_rate": 6.131824986142147e-05,
"loss": 0.0269,
"step": 9120
},
{
"epoch": 6.30524861878453,
"grad_norm": 0.1370326429605484,
"learning_rate": 6.123770692845529e-05,
"loss": 0.0205,
"step": 9130
},
{
"epoch": 6.3121546961325965,
"grad_norm": 0.18716467916965485,
"learning_rate": 6.11571332720275e-05,
"loss": 0.024,
"step": 9140
},
{
"epoch": 6.319060773480663,
"grad_norm": 0.1517820656299591,
"learning_rate": 6.107652911242336e-05,
"loss": 0.0265,
"step": 9150
},
{
"epoch": 6.3259668508287294,
"grad_norm": 0.10027366876602173,
"learning_rate": 6.0995894670011586e-05,
"loss": 0.03,
"step": 9160
},
{
"epoch": 6.332872928176796,
"grad_norm": 0.16794605553150177,
"learning_rate": 6.091523016524368e-05,
"loss": 0.0276,
"step": 9170
},
{
"epoch": 6.3397790055248615,
"grad_norm": 0.1418285071849823,
"learning_rate": 6.083453581865328e-05,
"loss": 0.0284,
"step": 9180
},
{
"epoch": 6.346685082872928,
"grad_norm": 0.15791954100131989,
"learning_rate": 6.075381185085568e-05,
"loss": 0.0271,
"step": 9190
},
{
"epoch": 6.3535911602209945,
"grad_norm": 0.1225113794207573,
"learning_rate": 6.067305848254709e-05,
"loss": 0.0258,
"step": 9200
},
{
"epoch": 6.360497237569061,
"grad_norm": 0.08965400606393814,
"learning_rate": 6.059227593450418e-05,
"loss": 0.0257,
"step": 9210
},
{
"epoch": 6.367403314917127,
"grad_norm": 0.12374383211135864,
"learning_rate": 6.051146442758333e-05,
"loss": 0.0289,
"step": 9220
},
{
"epoch": 6.374309392265193,
"grad_norm": 0.13927176594734192,
"learning_rate": 6.043062418272012e-05,
"loss": 0.0249,
"step": 9230
},
{
"epoch": 6.3812154696132595,
"grad_norm": 0.16667596995830536,
"learning_rate": 6.0349755420928666e-05,
"loss": 0.0257,
"step": 9240
},
{
"epoch": 6.388121546961326,
"grad_norm": 0.10272025316953659,
"learning_rate": 6.0268858363301105e-05,
"loss": 0.0263,
"step": 9250
},
{
"epoch": 6.395027624309392,
"grad_norm": 0.14296506345272064,
"learning_rate": 6.018793323100689e-05,
"loss": 0.0266,
"step": 9260
},
{
"epoch": 6.401933701657459,
"grad_norm": 0.19271421432495117,
"learning_rate": 6.0106980245292255e-05,
"loss": 0.028,
"step": 9270
},
{
"epoch": 6.4088397790055245,
"grad_norm": 0.13540977239608765,
"learning_rate": 6.002599962747957e-05,
"loss": 0.0247,
"step": 9280
},
{
"epoch": 6.415745856353591,
"grad_norm": 0.18024395406246185,
"learning_rate": 5.994499159896673e-05,
"loss": 0.0256,
"step": 9290
},
{
"epoch": 6.422651933701657,
"grad_norm": 0.18153737485408783,
"learning_rate": 5.9863956381226607e-05,
"loss": 0.0278,
"step": 9300
},
{
"epoch": 6.429558011049724,
"grad_norm": 0.12531007826328278,
"learning_rate": 5.9782894195806394e-05,
"loss": 0.0265,
"step": 9310
},
{
"epoch": 6.43646408839779,
"grad_norm": 0.13979461789131165,
"learning_rate": 5.9701805264327004e-05,
"loss": 0.0275,
"step": 9320
},
{
"epoch": 6.443370165745856,
"grad_norm": 0.1680983603000641,
"learning_rate": 5.96206898084825e-05,
"loss": 0.0283,
"step": 9330
},
{
"epoch": 6.4502762430939224,
"grad_norm": 0.1745212972164154,
"learning_rate": 5.953954805003942e-05,
"loss": 0.0305,
"step": 9340
},
{
"epoch": 6.457182320441989,
"grad_norm": 0.10051245987415314,
"learning_rate": 5.945838021083623e-05,
"loss": 0.0239,
"step": 9350
},
{
"epoch": 6.464088397790055,
"grad_norm": 0.1596742868423462,
"learning_rate": 5.9377186512782714e-05,
"loss": 0.0287,
"step": 9360
},
{
"epoch": 6.470994475138122,
"grad_norm": 0.16254925727844238,
"learning_rate": 5.929596717785935e-05,
"loss": 0.0314,
"step": 9370
},
{
"epoch": 6.4779005524861875,
"grad_norm": 0.17129167914390564,
"learning_rate": 5.921472242811668e-05,
"loss": 0.029,
"step": 9380
},
{
"epoch": 6.484806629834254,
"grad_norm": 0.1531635820865631,
"learning_rate": 5.913345248567475e-05,
"loss": 0.0289,
"step": 9390
},
{
"epoch": 6.49171270718232,
"grad_norm": 0.15288211405277252,
"learning_rate": 5.905215757272248e-05,
"loss": 0.0236,
"step": 9400
},
{
"epoch": 6.498618784530387,
"grad_norm": 0.1294633448123932,
"learning_rate": 5.897083791151706e-05,
"loss": 0.0247,
"step": 9410
},
{
"epoch": 6.505524861878453,
"grad_norm": 0.12580148875713348,
"learning_rate": 5.888949372438336e-05,
"loss": 0.0282,
"step": 9420
},
{
"epoch": 6.512430939226519,
"grad_norm": 0.14940418303012848,
"learning_rate": 5.8808125233713255e-05,
"loss": 0.0218,
"step": 9430
},
{
"epoch": 6.519337016574585,
"grad_norm": 0.17906662821769714,
"learning_rate": 5.872673266196509e-05,
"loss": 0.0281,
"step": 9440
},
{
"epoch": 6.526243093922652,
"grad_norm": 0.11938751488924026,
"learning_rate": 5.864531623166305e-05,
"loss": 0.0257,
"step": 9450
},
{
"epoch": 6.533149171270718,
"grad_norm": 0.14027191698551178,
"learning_rate": 5.856387616539656e-05,
"loss": 0.0276,
"step": 9460
},
{
"epoch": 6.540055248618785,
"grad_norm": 0.1458040028810501,
"learning_rate": 5.848241268581967e-05,
"loss": 0.024,
"step": 9470
},
{
"epoch": 6.546961325966851,
"grad_norm": 0.1046091765165329,
"learning_rate": 5.840092601565037e-05,
"loss": 0.0291,
"step": 9480
},
{
"epoch": 6.553867403314917,
"grad_norm": 0.17765271663665771,
"learning_rate": 5.8319416377670144e-05,
"loss": 0.0265,
"step": 9490
},
{
"epoch": 6.560773480662983,
"grad_norm": 0.11805755645036697,
"learning_rate": 5.82378839947232e-05,
"loss": 0.0237,
"step": 9500
},
{
"epoch": 6.56767955801105,
"grad_norm": 0.10168580710887909,
"learning_rate": 5.815632908971599e-05,
"loss": 0.0266,
"step": 9510
},
{
"epoch": 6.574585635359116,
"grad_norm": 0.1571616679430008,
"learning_rate": 5.80747518856165e-05,
"loss": 0.0281,
"step": 9520
},
{
"epoch": 6.581491712707182,
"grad_norm": 0.11476701498031616,
"learning_rate": 5.799315260545367e-05,
"loss": 0.0301,
"step": 9530
},
{
"epoch": 6.588397790055248,
"grad_norm": 0.14077112078666687,
"learning_rate": 5.791153147231686e-05,
"loss": 0.0326,
"step": 9540
},
{
"epoch": 6.595303867403315,
"grad_norm": 0.10888878256082535,
"learning_rate": 5.782988870935509e-05,
"loss": 0.027,
"step": 9550
},
{
"epoch": 6.602209944751381,
"grad_norm": 0.18977871537208557,
"learning_rate": 5.774822453977657e-05,
"loss": 0.0281,
"step": 9560
},
{
"epoch": 6.609116022099448,
"grad_norm": 0.16340450942516327,
"learning_rate": 5.7666539186848036e-05,
"loss": 0.0274,
"step": 9570
},
{
"epoch": 6.616022099447514,
"grad_norm": 0.16446462273597717,
"learning_rate": 5.758483287389411e-05,
"loss": 0.0276,
"step": 9580
},
{
"epoch": 6.62292817679558,
"grad_norm": 0.11147239059209824,
"learning_rate": 5.7503105824296735e-05,
"loss": 0.0248,
"step": 9590
},
{
"epoch": 6.629834254143646,
"grad_norm": 0.17083384096622467,
"learning_rate": 5.742135826149453e-05,
"loss": 0.0254,
"step": 9600
},
{
"epoch": 6.636740331491713,
"grad_norm": 0.09648158401250839,
"learning_rate": 5.7339590408982223e-05,
"loss": 0.0284,
"step": 9610
},
{
"epoch": 6.643646408839779,
"grad_norm": 0.18214352428913116,
"learning_rate": 5.725780249031e-05,
"loss": 0.0354,
"step": 9620
},
{
"epoch": 6.650552486187845,
"grad_norm": 0.12843012809753418,
"learning_rate": 5.717599472908292e-05,
"loss": 0.0308,
"step": 9630
},
{
"epoch": 6.657458563535911,
"grad_norm": 0.1309477984905243,
"learning_rate": 5.7094167348960237e-05,
"loss": 0.024,
"step": 9640
},
{
"epoch": 6.664364640883978,
"grad_norm": 0.15698851644992828,
"learning_rate": 5.7012320573654945e-05,
"loss": 0.0267,
"step": 9650
},
{
"epoch": 6.671270718232044,
"grad_norm": 0.14546315371990204,
"learning_rate": 5.693045462693295e-05,
"loss": 0.0273,
"step": 9660
},
{
"epoch": 6.678176795580111,
"grad_norm": 0.09835139662027359,
"learning_rate": 5.684856973261266e-05,
"loss": 0.0244,
"step": 9670
},
{
"epoch": 6.685082872928177,
"grad_norm": 0.16771312057971954,
"learning_rate": 5.6766666114564215e-05,
"loss": 0.0289,
"step": 9680
},
{
"epoch": 6.691988950276243,
"grad_norm": 0.15076804161071777,
"learning_rate": 5.668474399670899e-05,
"loss": 0.0245,
"step": 9690
},
{
"epoch": 6.698895027624309,
"grad_norm": 0.2206779420375824,
"learning_rate": 5.660280360301896e-05,
"loss": 0.0242,
"step": 9700
},
{
"epoch": 6.705801104972376,
"grad_norm": 0.15940850973129272,
"learning_rate": 5.652084515751599e-05,
"loss": 0.0291,
"step": 9710
},
{
"epoch": 6.712707182320442,
"grad_norm": 0.1726212501525879,
"learning_rate": 5.643886888427137e-05,
"loss": 0.0271,
"step": 9720
},
{
"epoch": 6.719613259668508,
"grad_norm": 0.15121586620807648,
"learning_rate": 5.6356875007405074e-05,
"loss": 0.0256,
"step": 9730
},
{
"epoch": 6.726519337016574,
"grad_norm": 0.13991855084896088,
"learning_rate": 5.627486375108525e-05,
"loss": 0.0272,
"step": 9740
},
{
"epoch": 6.733425414364641,
"grad_norm": 0.1048620268702507,
"learning_rate": 5.619283533952754e-05,
"loss": 0.0253,
"step": 9750
},
{
"epoch": 6.740331491712707,
"grad_norm": 0.15269067883491516,
"learning_rate": 5.6110789996994474e-05,
"loss": 0.0256,
"step": 9760
},
{
"epoch": 6.747237569060774,
"grad_norm": 0.1541956663131714,
"learning_rate": 5.602872794779491e-05,
"loss": 0.0259,
"step": 9770
},
{
"epoch": 6.75414364640884,
"grad_norm": 0.1253366470336914,
"learning_rate": 5.594664941628334e-05,
"loss": 0.0244,
"step": 9780
},
{
"epoch": 6.761049723756906,
"grad_norm": 0.1777985841035843,
"learning_rate": 5.5864554626859324e-05,
"loss": 0.0278,
"step": 9790
},
{
"epoch": 6.767955801104972,
"grad_norm": 0.17948810756206512,
"learning_rate": 5.578244380396691e-05,
"loss": 0.0232,
"step": 9800
},
{
"epoch": 6.774861878453039,
"grad_norm": 0.1327214539051056,
"learning_rate": 5.570031717209394e-05,
"loss": 0.0261,
"step": 9810
},
{
"epoch": 6.781767955801105,
"grad_norm": 0.22230222821235657,
"learning_rate": 5.561817495577147e-05,
"loss": 0.0244,
"step": 9820
},
{
"epoch": 6.788674033149171,
"grad_norm": 0.15005578100681305,
"learning_rate": 5.5536017379573215e-05,
"loss": 0.0288,
"step": 9830
},
{
"epoch": 6.795580110497237,
"grad_norm": 0.1266830712556839,
"learning_rate": 5.545384466811483e-05,
"loss": 0.0277,
"step": 9840
},
{
"epoch": 6.802486187845304,
"grad_norm": 0.16804485023021698,
"learning_rate": 5.5371657046053384e-05,
"loss": 0.0296,
"step": 9850
},
{
"epoch": 6.80939226519337,
"grad_norm": 0.16403374075889587,
"learning_rate": 5.528945473808669e-05,
"loss": 0.0279,
"step": 9860
},
{
"epoch": 6.816298342541437,
"grad_norm": 0.13253077864646912,
"learning_rate": 5.520723796895272e-05,
"loss": 0.0211,
"step": 9870
},
{
"epoch": 6.823204419889503,
"grad_norm": 0.13611510396003723,
"learning_rate": 5.512500696342897e-05,
"loss": 0.0246,
"step": 9880
},
{
"epoch": 6.830110497237569,
"grad_norm": 0.16444122791290283,
"learning_rate": 5.504276194633188e-05,
"loss": 0.0293,
"step": 9890
},
{
"epoch": 6.837016574585635,
"grad_norm": 0.1384178102016449,
"learning_rate": 5.49605031425162e-05,
"loss": 0.027,
"step": 9900
},
{
"epoch": 6.843922651933702,
"grad_norm": 0.11369351297616959,
"learning_rate": 5.487823077687434e-05,
"loss": 0.0261,
"step": 9910
},
{
"epoch": 6.850828729281768,
"grad_norm": 0.1023038923740387,
"learning_rate": 5.4795945074335806e-05,
"loss": 0.0307,
"step": 9920
},
{
"epoch": 6.857734806629834,
"grad_norm": 0.10871675610542297,
"learning_rate": 5.471364625986657e-05,
"loss": 0.0215,
"step": 9930
},
{
"epoch": 6.8646408839779,
"grad_norm": 0.16438175737857819,
"learning_rate": 5.463133455846845e-05,
"loss": 0.0291,
"step": 9940
},
{
"epoch": 6.871546961325967,
"grad_norm": 0.10552588105201721,
"learning_rate": 5.4549010195178505e-05,
"loss": 0.0283,
"step": 9950
},
{
"epoch": 6.878453038674033,
"grad_norm": 0.15607544779777527,
"learning_rate": 5.446667339506838e-05,
"loss": 0.0272,
"step": 9960
},
{
"epoch": 6.8853591160221,
"grad_norm": 0.17612875998020172,
"learning_rate": 5.4384324383243756e-05,
"loss": 0.0251,
"step": 9970
},
{
"epoch": 6.892265193370166,
"grad_norm": 0.1139112040400505,
"learning_rate": 5.430196338484368e-05,
"loss": 0.0251,
"step": 9980
},
{
"epoch": 6.899171270718232,
"grad_norm": 0.17247068881988525,
"learning_rate": 5.4219590625039975e-05,
"loss": 0.0265,
"step": 9990
},
{
"epoch": 6.906077348066298,
"grad_norm": 0.22626756131649017,
"learning_rate": 5.413720632903664e-05,
"loss": 0.029,
"step": 10000
},
{
"epoch": 6.912983425414365,
"grad_norm": 0.1493772268295288,
"learning_rate": 5.405481072206917e-05,
"loss": 0.0247,
"step": 10010
},
{
"epoch": 6.919889502762431,
"grad_norm": 0.11201240867376328,
"learning_rate": 5.397240402940402e-05,
"loss": 0.0267,
"step": 10020
},
{
"epoch": 6.926795580110497,
"grad_norm": 0.12302827835083008,
"learning_rate": 5.388998647633794e-05,
"loss": 0.024,
"step": 10030
},
{
"epoch": 6.933701657458563,
"grad_norm": 0.15909771621227264,
"learning_rate": 5.380755828819737e-05,
"loss": 0.0279,
"step": 10040
},
{
"epoch": 6.94060773480663,
"grad_norm": 0.1554439663887024,
"learning_rate": 5.3725119690337846e-05,
"loss": 0.0276,
"step": 10050
},
{
"epoch": 6.947513812154696,
"grad_norm": 0.15158556401729584,
"learning_rate": 5.3642670908143324e-05,
"loss": 0.0248,
"step": 10060
},
{
"epoch": 6.954419889502763,
"grad_norm": 0.1534639447927475,
"learning_rate": 5.356021216702562e-05,
"loss": 0.0274,
"step": 10070
},
{
"epoch": 6.961325966850829,
"grad_norm": 0.16105881333351135,
"learning_rate": 5.347774369242381e-05,
"loss": 0.0264,
"step": 10080
},
{
"epoch": 6.968232044198895,
"grad_norm": 0.11113212257623672,
"learning_rate": 5.3395265709803545e-05,
"loss": 0.0216,
"step": 10090
},
{
"epoch": 6.975138121546961,
"grad_norm": 0.11615071445703506,
"learning_rate": 5.331277844465647e-05,
"loss": 0.0225,
"step": 10100
},
{
"epoch": 6.982044198895028,
"grad_norm": 0.12254151701927185,
"learning_rate": 5.323028212249963e-05,
"loss": 0.0269,
"step": 10110
},
{
"epoch": 6.988950276243094,
"grad_norm": 0.1819300502538681,
"learning_rate": 5.314777696887481e-05,
"loss": 0.0212,
"step": 10120
},
{
"epoch": 6.99585635359116,
"grad_norm": 0.1481214165687561,
"learning_rate": 5.306526320934796e-05,
"loss": 0.0254,
"step": 10130
},
{
"epoch": 7.002762430939226,
"grad_norm": 0.12868042290210724,
"learning_rate": 5.298274106950854e-05,
"loss": 0.0229,
"step": 10140
},
{
"epoch": 7.009668508287293,
"grad_norm": 0.11328385025262833,
"learning_rate": 5.290021077496893e-05,
"loss": 0.0232,
"step": 10150
},
{
"epoch": 7.016574585635359,
"grad_norm": 0.18687033653259277,
"learning_rate": 5.2817672551363816e-05,
"loss": 0.0234,
"step": 10160
},
{
"epoch": 7.023480662983426,
"grad_norm": 0.12744495272636414,
"learning_rate": 5.273512662434952e-05,
"loss": 0.0261,
"step": 10170
},
{
"epoch": 7.030386740331492,
"grad_norm": 0.14903679490089417,
"learning_rate": 5.265257321960349e-05,
"loss": 0.0223,
"step": 10180
},
{
"epoch": 7.037292817679558,
"grad_norm": 0.17800362408161163,
"learning_rate": 5.257001256282357e-05,
"loss": 0.0285,
"step": 10190
},
{
"epoch": 7.044198895027624,
"grad_norm": 0.16140541434288025,
"learning_rate": 5.248744487972742e-05,
"loss": 0.0254,
"step": 10200
},
{
"epoch": 7.051104972375691,
"grad_norm": 0.15151412785053253,
"learning_rate": 5.240487039605196e-05,
"loss": 0.0239,
"step": 10210
},
{
"epoch": 7.058011049723757,
"grad_norm": 0.08070018887519836,
"learning_rate": 5.232228933755267e-05,
"loss": 0.0234,
"step": 10220
},
{
"epoch": 7.064917127071824,
"grad_norm": 0.148824080824852,
"learning_rate": 5.2239701930003006e-05,
"loss": 0.0277,
"step": 10230
},
{
"epoch": 7.071823204419889,
"grad_norm": 0.14293543994426727,
"learning_rate": 5.215710839919379e-05,
"loss": 0.0253,
"step": 10240
},
{
"epoch": 7.078729281767956,
"grad_norm": 0.14560624957084656,
"learning_rate": 5.207450897093257e-05,
"loss": 0.0224,
"step": 10250
},
{
"epoch": 7.085635359116022,
"grad_norm": 0.15720799565315247,
"learning_rate": 5.1991903871043046e-05,
"loss": 0.0254,
"step": 10260
},
{
"epoch": 7.092541436464089,
"grad_norm": 0.11100403964519501,
"learning_rate": 5.190929332536439e-05,
"loss": 0.0287,
"step": 10270
},
{
"epoch": 7.099447513812155,
"grad_norm": 0.15021225810050964,
"learning_rate": 5.182667755975071e-05,
"loss": 0.0278,
"step": 10280
},
{
"epoch": 7.106353591160221,
"grad_norm": 0.13029877841472626,
"learning_rate": 5.1744056800070315e-05,
"loss": 0.0257,
"step": 10290
},
{
"epoch": 7.113259668508287,
"grad_norm": 0.22811639308929443,
"learning_rate": 5.166143127220524e-05,
"loss": 0.0262,
"step": 10300
},
{
"epoch": 7.120165745856354,
"grad_norm": 0.15374548733234406,
"learning_rate": 5.1578801202050485e-05,
"loss": 0.0239,
"step": 10310
},
{
"epoch": 7.12707182320442,
"grad_norm": 0.13846486806869507,
"learning_rate": 5.149616681551355e-05,
"loss": 0.0285,
"step": 10320
},
{
"epoch": 7.133977900552487,
"grad_norm": 0.13904505968093872,
"learning_rate": 5.141352833851367e-05,
"loss": 0.0242,
"step": 10330
},
{
"epoch": 7.140883977900552,
"grad_norm": 0.1465989053249359,
"learning_rate": 5.1330885996981285e-05,
"loss": 0.0227,
"step": 10340
},
{
"epoch": 7.147790055248619,
"grad_norm": 0.1020495742559433,
"learning_rate": 5.124824001685741e-05,
"loss": 0.0236,
"step": 10350
},
{
"epoch": 7.154696132596685,
"grad_norm": 0.10960061103105545,
"learning_rate": 5.116559062409298e-05,
"loss": 0.0228,
"step": 10360
},
{
"epoch": 7.161602209944752,
"grad_norm": 0.1321902722120285,
"learning_rate": 5.10829380446483e-05,
"loss": 0.0222,
"step": 10370
},
{
"epoch": 7.168508287292818,
"grad_norm": 0.12457681447267532,
"learning_rate": 5.100028250449235e-05,
"loss": 0.0239,
"step": 10380
},
{
"epoch": 7.175414364640884,
"grad_norm": 0.20130808651447296,
"learning_rate": 5.0917624229602234e-05,
"loss": 0.0249,
"step": 10390
},
{
"epoch": 7.18232044198895,
"grad_norm": 0.19484300911426544,
"learning_rate": 5.0834963445962524e-05,
"loss": 0.0261,
"step": 10400
},
{
"epoch": 7.189226519337017,
"grad_norm": 0.12139289826154709,
"learning_rate": 5.075230037956461e-05,
"loss": 0.0227,
"step": 10410
},
{
"epoch": 7.196132596685083,
"grad_norm": 0.1364196538925171,
"learning_rate": 5.0669635256406213e-05,
"loss": 0.0229,
"step": 10420
},
{
"epoch": 7.20303867403315,
"grad_norm": 0.12963388860225677,
"learning_rate": 5.058696830249058e-05,
"loss": 0.0203,
"step": 10430
},
{
"epoch": 7.209944751381215,
"grad_norm": 0.14057807624340057,
"learning_rate": 5.050429974382602e-05,
"loss": 0.0254,
"step": 10440
},
{
"epoch": 7.216850828729282,
"grad_norm": 0.14869332313537598,
"learning_rate": 5.042162980642523e-05,
"loss": 0.0235,
"step": 10450
},
{
"epoch": 7.223756906077348,
"grad_norm": 0.17959578335285187,
"learning_rate": 5.033895871630462e-05,
"loss": 0.0227,
"step": 10460
},
{
"epoch": 7.230662983425415,
"grad_norm": 0.12166395783424377,
"learning_rate": 5.025628669948386e-05,
"loss": 0.0259,
"step": 10470
},
{
"epoch": 7.237569060773481,
"grad_norm": 0.13333919644355774,
"learning_rate": 5.017361398198502e-05,
"loss": 0.0239,
"step": 10480
},
{
"epoch": 7.244475138121547,
"grad_norm": 0.17320550978183746,
"learning_rate": 5.009094078983221e-05,
"loss": 0.0228,
"step": 10490
},
{
"epoch": 7.251381215469613,
"grad_norm": 0.11817017942667007,
"learning_rate": 5.000826734905073e-05,
"loss": 0.0207,
"step": 10500
},
{
"epoch": 7.25828729281768,
"grad_norm": 0.15460477769374847,
"learning_rate": 4.9925593885666645e-05,
"loss": 0.0262,
"step": 10510
},
{
"epoch": 7.265193370165746,
"grad_norm": 0.19087368249893188,
"learning_rate": 4.984292062570602e-05,
"loss": 0.0266,
"step": 10520
},
{
"epoch": 7.2720994475138125,
"grad_norm": 0.12094233930110931,
"learning_rate": 4.976024779519442e-05,
"loss": 0.0243,
"step": 10530
},
{
"epoch": 7.279005524861878,
"grad_norm": 0.14833706617355347,
"learning_rate": 4.9677575620156194e-05,
"loss": 0.0244,
"step": 10540
},
{
"epoch": 7.285911602209945,
"grad_norm": 0.1710749715566635,
"learning_rate": 4.959490432661391e-05,
"loss": 0.0222,
"step": 10550
},
{
"epoch": 7.292817679558011,
"grad_norm": 0.19592711329460144,
"learning_rate": 4.9512234140587726e-05,
"loss": 0.0231,
"step": 10560
},
{
"epoch": 7.2997237569060776,
"grad_norm": 0.1926083117723465,
"learning_rate": 4.942956528809477e-05,
"loss": 0.0262,
"step": 10570
},
{
"epoch": 7.306629834254144,
"grad_norm": 0.11659876257181168,
"learning_rate": 4.934689799514854e-05,
"loss": 0.0236,
"step": 10580
},
{
"epoch": 7.31353591160221,
"grad_norm": 0.1698133498430252,
"learning_rate": 4.926423248775827e-05,
"loss": 0.0256,
"step": 10590
},
{
"epoch": 7.320441988950276,
"grad_norm": 0.12696529924869537,
"learning_rate": 4.918156899192826e-05,
"loss": 0.0256,
"step": 10600
},
{
"epoch": 7.327348066298343,
"grad_norm": 0.18490107357501984,
"learning_rate": 4.909890773365738e-05,
"loss": 0.0257,
"step": 10610
},
{
"epoch": 7.334254143646409,
"grad_norm": 0.1368025690317154,
"learning_rate": 4.9016248938938344e-05,
"loss": 0.0219,
"step": 10620
},
{
"epoch": 7.3411602209944755,
"grad_norm": 0.17904599010944366,
"learning_rate": 4.8933592833757156e-05,
"loss": 0.0259,
"step": 10630
},
{
"epoch": 7.348066298342541,
"grad_norm": 0.1488611102104187,
"learning_rate": 4.8850939644092435e-05,
"loss": 0.023,
"step": 10640
},
{
"epoch": 7.354972375690608,
"grad_norm": 0.16605186462402344,
"learning_rate": 4.876828959591485e-05,
"loss": 0.0236,
"step": 10650
},
{
"epoch": 7.361878453038674,
"grad_norm": 0.1492478996515274,
"learning_rate": 4.8685642915186474e-05,
"loss": 0.0237,
"step": 10660
},
{
"epoch": 7.3687845303867405,
"grad_norm": 0.18098857998847961,
"learning_rate": 4.860299982786018e-05,
"loss": 0.0228,
"step": 10670
},
{
"epoch": 7.375690607734807,
"grad_norm": 0.19155803322792053,
"learning_rate": 4.852036055987901e-05,
"loss": 0.0265,
"step": 10680
},
{
"epoch": 7.382596685082873,
"grad_norm": 0.11117758601903915,
"learning_rate": 4.843772533717558e-05,
"loss": 0.0202,
"step": 10690
},
{
"epoch": 7.389502762430939,
"grad_norm": 0.16134271025657654,
"learning_rate": 4.835509438567142e-05,
"loss": 0.026,
"step": 10700
},
{
"epoch": 7.3964088397790055,
"grad_norm": 0.1662876456975937,
"learning_rate": 4.827246793127639e-05,
"loss": 0.0229,
"step": 10710
},
{
"epoch": 7.403314917127072,
"grad_norm": 0.17498347163200378,
"learning_rate": 4.818984619988807e-05,
"loss": 0.0271,
"step": 10720
},
{
"epoch": 7.4102209944751385,
"grad_norm": 0.14823921024799347,
"learning_rate": 4.810722941739115e-05,
"loss": 0.0209,
"step": 10730
},
{
"epoch": 7.417127071823204,
"grad_norm": 0.15792728960514069,
"learning_rate": 4.8024617809656684e-05,
"loss": 0.0223,
"step": 10740
},
{
"epoch": 7.4240331491712706,
"grad_norm": 0.1337282657623291,
"learning_rate": 4.794201160254171e-05,
"loss": 0.0247,
"step": 10750
},
{
"epoch": 7.430939226519337,
"grad_norm": 0.17503942549228668,
"learning_rate": 4.785941102188844e-05,
"loss": 0.0239,
"step": 10760
},
{
"epoch": 7.4378453038674035,
"grad_norm": 0.1897234469652176,
"learning_rate": 4.7776816293523686e-05,
"loss": 0.0256,
"step": 10770
},
{
"epoch": 7.44475138121547,
"grad_norm": 0.1384185552597046,
"learning_rate": 4.769422764325832e-05,
"loss": 0.0254,
"step": 10780
},
{
"epoch": 7.451657458563536,
"grad_norm": 0.1439378261566162,
"learning_rate": 4.76116452968865e-05,
"loss": 0.0263,
"step": 10790
},
{
"epoch": 7.458563535911602,
"grad_norm": 0.1375391185283661,
"learning_rate": 4.752906948018525e-05,
"loss": 0.0231,
"step": 10800
},
{
"epoch": 7.4654696132596685,
"grad_norm": 0.0991685688495636,
"learning_rate": 4.7446500418913684e-05,
"loss": 0.0238,
"step": 10810
},
{
"epoch": 7.472375690607735,
"grad_norm": 0.12018142640590668,
"learning_rate": 4.736393833881247e-05,
"loss": 0.0249,
"step": 10820
},
{
"epoch": 7.4792817679558015,
"grad_norm": 0.15268614888191223,
"learning_rate": 4.7281383465603194e-05,
"loss": 0.0278,
"step": 10830
},
{
"epoch": 7.486187845303867,
"grad_norm": 0.15701478719711304,
"learning_rate": 4.71988360249877e-05,
"loss": 0.0267,
"step": 10840
},
{
"epoch": 7.4930939226519335,
"grad_norm": 0.18149255216121674,
"learning_rate": 4.7116296242647554e-05,
"loss": 0.0252,
"step": 10850
},
{
"epoch": 7.5,
"grad_norm": 0.17894303798675537,
"learning_rate": 4.703376434424336e-05,
"loss": 0.0192,
"step": 10860
},
{
"epoch": 7.5069060773480665,
"grad_norm": 0.11874563992023468,
"learning_rate": 4.695124055541421e-05,
"loss": 0.0235,
"step": 10870
},
{
"epoch": 7.513812154696133,
"grad_norm": 0.12064361572265625,
"learning_rate": 4.6868725101776934e-05,
"loss": 0.0219,
"step": 10880
},
{
"epoch": 7.5207182320441985,
"grad_norm": 0.1320265680551529,
"learning_rate": 4.678621820892567e-05,
"loss": 0.0225,
"step": 10890
},
{
"epoch": 7.527624309392265,
"grad_norm": 0.13714221119880676,
"learning_rate": 4.670372010243111e-05,
"loss": 0.0249,
"step": 10900
},
{
"epoch": 7.5345303867403315,
"grad_norm": 0.11703638732433319,
"learning_rate": 4.662123100783992e-05,
"loss": 0.0236,
"step": 10910
},
{
"epoch": 7.541436464088398,
"grad_norm": 0.15061037242412567,
"learning_rate": 4.653875115067415e-05,
"loss": 0.0216,
"step": 10920
},
{
"epoch": 7.548342541436464,
"grad_norm": 0.16458094120025635,
"learning_rate": 4.6456280756430545e-05,
"loss": 0.0263,
"step": 10930
},
{
"epoch": 7.55524861878453,
"grad_norm": 0.13051296770572662,
"learning_rate": 4.637382005058004e-05,
"loss": 0.0231,
"step": 10940
},
{
"epoch": 7.5621546961325965,
"grad_norm": 0.09485608339309692,
"learning_rate": 4.629136925856705e-05,
"loss": 0.0235,
"step": 10950
},
{
"epoch": 7.569060773480663,
"grad_norm": 0.18857619166374207,
"learning_rate": 4.6208928605808895e-05,
"loss": 0.0229,
"step": 10960
},
{
"epoch": 7.5759668508287294,
"grad_norm": 0.16706739366054535,
"learning_rate": 4.612649831769519e-05,
"loss": 0.0256,
"step": 10970
},
{
"epoch": 7.582872928176796,
"grad_norm": 0.23252038657665253,
"learning_rate": 4.604407861958715e-05,
"loss": 0.0245,
"step": 10980
},
{
"epoch": 7.5897790055248615,
"grad_norm": 0.12180530279874802,
"learning_rate": 4.5961669736817114e-05,
"loss": 0.0258,
"step": 10990
},
{
"epoch": 7.596685082872928,
"grad_norm": 0.2081458866596222,
"learning_rate": 4.5879271894687814e-05,
"loss": 0.0291,
"step": 11000
},
{
"epoch": 7.6035911602209945,
"grad_norm": 0.14544755220413208,
"learning_rate": 4.5796885318471826e-05,
"loss": 0.0234,
"step": 11010
},
{
"epoch": 7.610497237569061,
"grad_norm": 0.11017230153083801,
"learning_rate": 4.571451023341086e-05,
"loss": 0.022,
"step": 11020
},
{
"epoch": 7.617403314917127,
"grad_norm": 0.11155513674020767,
"learning_rate": 4.563214686471527e-05,
"loss": 0.0311,
"step": 11030
},
{
"epoch": 7.624309392265193,
"grad_norm": 0.16592305898666382,
"learning_rate": 4.5549795437563365e-05,
"loss": 0.0239,
"step": 11040
},
{
"epoch": 7.6312154696132595,
"grad_norm": 0.15650154650211334,
"learning_rate": 4.546745617710081e-05,
"loss": 0.0225,
"step": 11050
},
{
"epoch": 7.638121546961326,
"grad_norm": 0.13385139405727386,
"learning_rate": 4.5385129308440014e-05,
"loss": 0.0218,
"step": 11060
},
{
"epoch": 7.645027624309392,
"grad_norm": 0.1123754158616066,
"learning_rate": 4.530281505665944e-05,
"loss": 0.0237,
"step": 11070
},
{
"epoch": 7.651933701657459,
"grad_norm": 0.13245031237602234,
"learning_rate": 4.5220513646803134e-05,
"loss": 0.0259,
"step": 11080
},
{
"epoch": 7.6588397790055245,
"grad_norm": 0.15437833964824677,
"learning_rate": 4.513822530388003e-05,
"loss": 0.0244,
"step": 11090
},
{
"epoch": 7.665745856353591,
"grad_norm": 0.15100827813148499,
"learning_rate": 4.5055950252863296e-05,
"loss": 0.0242,
"step": 11100
},
{
"epoch": 7.672651933701657,
"grad_norm": 0.13512608408927917,
"learning_rate": 4.4973688718689803e-05,
"loss": 0.0262,
"step": 11110
},
{
"epoch": 7.679558011049724,
"grad_norm": 0.17417916655540466,
"learning_rate": 4.4891440926259406e-05,
"loss": 0.0243,
"step": 11120
},
{
"epoch": 7.68646408839779,
"grad_norm": 0.14423848688602448,
"learning_rate": 4.480920710043443e-05,
"loss": 0.0264,
"step": 11130
},
{
"epoch": 7.693370165745856,
"grad_norm": 0.16867166757583618,
"learning_rate": 4.4726987466039044e-05,
"loss": 0.0282,
"step": 11140
},
{
"epoch": 7.7002762430939224,
"grad_norm": 0.19301173090934753,
"learning_rate": 4.46447822478586e-05,
"loss": 0.0311,
"step": 11150
},
{
"epoch": 7.707182320441989,
"grad_norm": 0.14649568498134613,
"learning_rate": 4.4562591670638974e-05,
"loss": 0.0242,
"step": 11160
},
{
"epoch": 7.714088397790055,
"grad_norm": 0.12760896980762482,
"learning_rate": 4.4480415959086105e-05,
"loss": 0.0241,
"step": 11170
},
{
"epoch": 7.720994475138122,
"grad_norm": 0.09078028798103333,
"learning_rate": 4.439825533786522e-05,
"loss": 0.0258,
"step": 11180
},
{
"epoch": 7.7279005524861875,
"grad_norm": 0.18043041229248047,
"learning_rate": 4.431611003160035e-05,
"loss": 0.0231,
"step": 11190
},
{
"epoch": 7.734806629834254,
"grad_norm": 0.1317226141691208,
"learning_rate": 4.4233980264873636e-05,
"loss": 0.0272,
"step": 11200
},
{
"epoch": 7.74171270718232,
"grad_norm": 0.09718924015760422,
"learning_rate": 4.4151866262224684e-05,
"loss": 0.0272,
"step": 11210
},
{
"epoch": 7.748618784530387,
"grad_norm": 0.13303658366203308,
"learning_rate": 4.406976824815006e-05,
"loss": 0.0232,
"step": 11220
},
{
"epoch": 7.755524861878453,
"grad_norm": 0.13181060552597046,
"learning_rate": 4.3987686447102595e-05,
"loss": 0.0238,
"step": 11230
},
{
"epoch": 7.762430939226519,
"grad_norm": 0.14495405554771423,
"learning_rate": 4.3905621083490804e-05,
"loss": 0.0266,
"step": 11240
},
{
"epoch": 7.769337016574585,
"grad_norm": 0.16467227041721344,
"learning_rate": 4.3823572381678286e-05,
"loss": 0.025,
"step": 11250
},
{
"epoch": 7.776243093922652,
"grad_norm": 0.16528834402561188,
"learning_rate": 4.374154056598301e-05,
"loss": 0.0235,
"step": 11260
},
{
"epoch": 7.783149171270718,
"grad_norm": 0.14273691177368164,
"learning_rate": 4.3659525860676845e-05,
"loss": 0.0218,
"step": 11270
},
{
"epoch": 7.790055248618785,
"grad_norm": 0.11080687493085861,
"learning_rate": 4.3577528489984854e-05,
"loss": 0.0231,
"step": 11280
},
{
"epoch": 7.796961325966851,
"grad_norm": 0.15187494456768036,
"learning_rate": 4.349554867808476e-05,
"loss": 0.0244,
"step": 11290
},
{
"epoch": 7.803867403314917,
"grad_norm": 0.139710932970047,
"learning_rate": 4.34135866491062e-05,
"loss": 0.028,
"step": 11300
},
{
"epoch": 7.810773480662983,
"grad_norm": 0.16394203901290894,
"learning_rate": 4.333164262713022e-05,
"loss": 0.0237,
"step": 11310
},
{
"epoch": 7.81767955801105,
"grad_norm": 0.12090770900249481,
"learning_rate": 4.324971683618868e-05,
"loss": 0.0224,
"step": 11320
},
{
"epoch": 7.824585635359116,
"grad_norm": 0.11589226126670837,
"learning_rate": 4.316780950026354e-05,
"loss": 0.0259,
"step": 11330
},
{
"epoch": 7.831491712707182,
"grad_norm": 0.16017724573612213,
"learning_rate": 4.308592084328637e-05,
"loss": 0.0226,
"step": 11340
},
{
"epoch": 7.838397790055248,
"grad_norm": 0.14901770651340485,
"learning_rate": 4.3004051089137576e-05,
"loss": 0.0266,
"step": 11350
},
{
"epoch": 7.845303867403315,
"grad_norm": 0.1358511745929718,
"learning_rate": 4.292220046164597e-05,
"loss": 0.0197,
"step": 11360
},
{
"epoch": 7.852209944751381,
"grad_norm": 0.1269238144159317,
"learning_rate": 4.2840369184588035e-05,
"loss": 0.0255,
"step": 11370
},
{
"epoch": 7.859116022099448,
"grad_norm": 0.16940242052078247,
"learning_rate": 4.2758557481687345e-05,
"loss": 0.022,
"step": 11380
},
{
"epoch": 7.866022099447514,
"grad_norm": 0.1460244357585907,
"learning_rate": 4.267676557661403e-05,
"loss": 0.0194,
"step": 11390
},
{
"epoch": 7.87292817679558,
"grad_norm": 0.13890156149864197,
"learning_rate": 4.2594993692983955e-05,
"loss": 0.0199,
"step": 11400
},
{
"epoch": 7.879834254143646,
"grad_norm": 0.14905396103858948,
"learning_rate": 4.251324205435837e-05,
"loss": 0.0225,
"step": 11410
},
{
"epoch": 7.886740331491713,
"grad_norm": 0.12149365246295929,
"learning_rate": 4.243151088424312e-05,
"loss": 0.0243,
"step": 11420
},
{
"epoch": 7.893646408839779,
"grad_norm": 0.11986390501260757,
"learning_rate": 4.234980040608813e-05,
"loss": 0.023,
"step": 11430
},
{
"epoch": 7.900552486187845,
"grad_norm": 0.08489210903644562,
"learning_rate": 4.22681108432867e-05,
"loss": 0.0201,
"step": 11440
},
{
"epoch": 7.907458563535911,
"grad_norm": 0.15432953834533691,
"learning_rate": 4.2186442419174984e-05,
"loss": 0.0275,
"step": 11450
},
{
"epoch": 7.914364640883978,
"grad_norm": 0.1917029619216919,
"learning_rate": 4.210479535703133e-05,
"loss": 0.0274,
"step": 11460
},
{
"epoch": 7.921270718232044,
"grad_norm": 0.15083684027194977,
"learning_rate": 4.202316988007567e-05,
"loss": 0.0278,
"step": 11470
},
{
"epoch": 7.928176795580111,
"grad_norm": 0.14471709728240967,
"learning_rate": 4.194156621146901e-05,
"loss": 0.0239,
"step": 11480
},
{
"epoch": 7.935082872928177,
"grad_norm": 0.18719494342803955,
"learning_rate": 4.1859984574312596e-05,
"loss": 0.0208,
"step": 11490
},
{
"epoch": 7.941988950276243,
"grad_norm": 0.09603523463010788,
"learning_rate": 4.177842519164752e-05,
"loss": 0.0254,
"step": 11500
},
{
"epoch": 7.948895027624309,
"grad_norm": 0.18613314628601074,
"learning_rate": 4.169688828645404e-05,
"loss": 0.0214,
"step": 11510
},
{
"epoch": 7.955801104972376,
"grad_norm": 0.1553083211183548,
"learning_rate": 4.161537408165092e-05,
"loss": 0.0256,
"step": 11520
},
{
"epoch": 7.962707182320442,
"grad_norm": 0.19505159556865692,
"learning_rate": 4.1533882800094924e-05,
"loss": 0.0243,
"step": 11530
},
{
"epoch": 7.969613259668508,
"grad_norm": 0.21026429533958435,
"learning_rate": 4.145241466458005e-05,
"loss": 0.0217,
"step": 11540
},
{
"epoch": 7.976519337016574,
"grad_norm": 0.13795223832130432,
"learning_rate": 4.13709698978371e-05,
"loss": 0.0227,
"step": 11550
},
{
"epoch": 7.983425414364641,
"grad_norm": 0.16697093844413757,
"learning_rate": 4.1289548722532944e-05,
"loss": 0.0228,
"step": 11560
},
{
"epoch": 7.990331491712707,
"grad_norm": 0.18963934481143951,
"learning_rate": 4.120815136126999e-05,
"loss": 0.0238,
"step": 11570
},
{
"epoch": 7.997237569060774,
"grad_norm": 0.13078151643276215,
"learning_rate": 4.112677803658548e-05,
"loss": 0.0196,
"step": 11580
},
{
"epoch": 8.00414364640884,
"grad_norm": 0.12813352048397064,
"learning_rate": 4.1045428970951e-05,
"loss": 0.0221,
"step": 11590
},
{
"epoch": 8.011049723756907,
"grad_norm": 0.125010147690773,
"learning_rate": 4.0964104386771785e-05,
"loss": 0.0223,
"step": 11600
},
{
"epoch": 8.017955801104973,
"grad_norm": 0.10712146013975143,
"learning_rate": 4.0882804506386144e-05,
"loss": 0.0218,
"step": 11610
},
{
"epoch": 8.024861878453038,
"grad_norm": 0.17532770335674286,
"learning_rate": 4.080152955206485e-05,
"loss": 0.0243,
"step": 11620
},
{
"epoch": 8.031767955801104,
"grad_norm": 0.14849954843521118,
"learning_rate": 4.0720279746010505e-05,
"loss": 0.0225,
"step": 11630
},
{
"epoch": 8.03867403314917,
"grad_norm": 0.1307544708251953,
"learning_rate": 4.063905531035699e-05,
"loss": 0.0251,
"step": 11640
},
{
"epoch": 8.045580110497237,
"grad_norm": 0.11142382770776749,
"learning_rate": 4.055785646716882e-05,
"loss": 0.0226,
"step": 11650
},
{
"epoch": 8.052486187845304,
"grad_norm": 0.09290549904108047,
"learning_rate": 4.047668343844051e-05,
"loss": 0.0216,
"step": 11660
},
{
"epoch": 8.05939226519337,
"grad_norm": 0.14706315100193024,
"learning_rate": 4.039553644609604e-05,
"loss": 0.0217,
"step": 11670
},
{
"epoch": 8.066298342541437,
"grad_norm": 0.16171616315841675,
"learning_rate": 4.0314415711988176e-05,
"loss": 0.0213,
"step": 11680
},
{
"epoch": 8.073204419889503,
"grad_norm": 0.09882531315088272,
"learning_rate": 4.023332145789792e-05,
"loss": 0.0197,
"step": 11690
},
{
"epoch": 8.08011049723757,
"grad_norm": 0.13337063789367676,
"learning_rate": 4.015225390553385e-05,
"loss": 0.0249,
"step": 11700
},
{
"epoch": 8.087016574585636,
"grad_norm": 0.1553422063589096,
"learning_rate": 4.007121327653158e-05,
"loss": 0.0294,
"step": 11710
},
{
"epoch": 8.0939226519337,
"grad_norm": 0.1073046326637268,
"learning_rate": 3.9990199792453064e-05,
"loss": 0.0198,
"step": 11720
},
{
"epoch": 8.100828729281767,
"grad_norm": 0.11477019637823105,
"learning_rate": 3.9909213674786103e-05,
"loss": 0.0207,
"step": 11730
},
{
"epoch": 8.107734806629834,
"grad_norm": 0.10524137318134308,
"learning_rate": 3.982825514494363e-05,
"loss": 0.023,
"step": 11740
},
{
"epoch": 8.1146408839779,
"grad_norm": 0.15390941500663757,
"learning_rate": 3.974732442426319e-05,
"loss": 0.0225,
"step": 11750
},
{
"epoch": 8.121546961325967,
"grad_norm": 0.09843657910823822,
"learning_rate": 3.966642173400629e-05,
"loss": 0.0201,
"step": 11760
},
{
"epoch": 8.128453038674033,
"grad_norm": 0.17840729653835297,
"learning_rate": 3.9585547295357764e-05,
"loss": 0.0191,
"step": 11770
},
{
"epoch": 8.1353591160221,
"grad_norm": 0.11976867914199829,
"learning_rate": 3.950470132942526e-05,
"loss": 0.0208,
"step": 11780
},
{
"epoch": 8.142265193370166,
"grad_norm": 0.16819389164447784,
"learning_rate": 3.942388405723856e-05,
"loss": 0.0212,
"step": 11790
},
{
"epoch": 8.149171270718233,
"grad_norm": 0.13980218768119812,
"learning_rate": 3.9343095699749e-05,
"loss": 0.0252,
"step": 11800
},
{
"epoch": 8.1560773480663,
"grad_norm": 0.15626643598079681,
"learning_rate": 3.9262336477828874e-05,
"loss": 0.0205,
"step": 11810
},
{
"epoch": 8.162983425414364,
"grad_norm": 0.2204706221818924,
"learning_rate": 3.9181606612270794e-05,
"loss": 0.0256,
"step": 11820
},
{
"epoch": 8.16988950276243,
"grad_norm": 0.19606149196624756,
"learning_rate": 3.910090632378713e-05,
"loss": 0.0247,
"step": 11830
},
{
"epoch": 8.176795580110497,
"grad_norm": 0.1758870631456375,
"learning_rate": 3.90202358330094e-05,
"loss": 0.0199,
"step": 11840
},
{
"epoch": 8.183701657458563,
"grad_norm": 0.07964209467172623,
"learning_rate": 3.8939595360487656e-05,
"loss": 0.0221,
"step": 11850
},
{
"epoch": 8.19060773480663,
"grad_norm": 0.14514313638210297,
"learning_rate": 3.885898512668984e-05,
"loss": 0.0255,
"step": 11860
},
{
"epoch": 8.197513812154696,
"grad_norm": 0.1320313960313797,
"learning_rate": 3.877840535200127e-05,
"loss": 0.0274,
"step": 11870
},
{
"epoch": 8.204419889502763,
"grad_norm": 0.1110544502735138,
"learning_rate": 3.869785625672397e-05,
"loss": 0.019,
"step": 11880
},
{
"epoch": 8.21132596685083,
"grad_norm": 0.11055376380681992,
"learning_rate": 3.8617338061076094e-05,
"loss": 0.0198,
"step": 11890
},
{
"epoch": 8.218232044198896,
"grad_norm": 0.15744233131408691,
"learning_rate": 3.853685098519132e-05,
"loss": 0.0253,
"step": 11900
},
{
"epoch": 8.225138121546962,
"grad_norm": 0.14390379190444946,
"learning_rate": 3.845639524911823e-05,
"loss": 0.0264,
"step": 11910
},
{
"epoch": 8.232044198895027,
"grad_norm": 0.09943123161792755,
"learning_rate": 3.837597107281974e-05,
"loss": 0.0178,
"step": 11920
},
{
"epoch": 8.238950276243093,
"grad_norm": 0.15828579664230347,
"learning_rate": 3.829557867617247e-05,
"loss": 0.0191,
"step": 11930
},
{
"epoch": 8.24585635359116,
"grad_norm": 0.12498103082180023,
"learning_rate": 3.821521827896618e-05,
"loss": 0.0232,
"step": 11940
},
{
"epoch": 8.252762430939226,
"grad_norm": 0.12863773107528687,
"learning_rate": 3.81348901009031e-05,
"loss": 0.0221,
"step": 11950
},
{
"epoch": 8.259668508287293,
"grad_norm": 0.15252485871315002,
"learning_rate": 3.805459436159741e-05,
"loss": 0.0219,
"step": 11960
},
{
"epoch": 8.26657458563536,
"grad_norm": 0.09118683636188507,
"learning_rate": 3.797433128057461e-05,
"loss": 0.025,
"step": 11970
},
{
"epoch": 8.273480662983426,
"grad_norm": 0.17367224395275116,
"learning_rate": 3.789410107727089e-05,
"loss": 0.0216,
"step": 11980
},
{
"epoch": 8.280386740331492,
"grad_norm": 0.10616571456193924,
"learning_rate": 3.781390397103257e-05,
"loss": 0.0209,
"step": 11990
},
{
"epoch": 8.287292817679559,
"grad_norm": 0.19001731276512146,
"learning_rate": 3.7733740181115455e-05,
"loss": 0.0228,
"step": 12000
},
{
"epoch": 8.294198895027625,
"grad_norm": 0.16032475233078003,
"learning_rate": 3.7653609926684306e-05,
"loss": 0.0218,
"step": 12010
},
{
"epoch": 8.30110497237569,
"grad_norm": 0.12353052943944931,
"learning_rate": 3.757351342681217e-05,
"loss": 0.0237,
"step": 12020
},
{
"epoch": 8.308011049723756,
"grad_norm": 0.15098872780799866,
"learning_rate": 3.749345090047982e-05,
"loss": 0.0244,
"step": 12030
},
{
"epoch": 8.314917127071823,
"grad_norm": 0.13546410202980042,
"learning_rate": 3.741342256657515e-05,
"loss": 0.0206,
"step": 12040
},
{
"epoch": 8.32182320441989,
"grad_norm": 0.13305281102657318,
"learning_rate": 3.7333428643892567e-05,
"loss": 0.0251,
"step": 12050
},
{
"epoch": 8.328729281767956,
"grad_norm": 0.11955028027296066,
"learning_rate": 3.725346935113239e-05,
"loss": 0.0247,
"step": 12060
},
{
"epoch": 8.335635359116022,
"grad_norm": 0.16611330211162567,
"learning_rate": 3.717354490690029e-05,
"loss": 0.0263,
"step": 12070
},
{
"epoch": 8.342541436464089,
"grad_norm": 0.16200731694698334,
"learning_rate": 3.709365552970664e-05,
"loss": 0.0217,
"step": 12080
},
{
"epoch": 8.349447513812155,
"grad_norm": 0.12580612301826477,
"learning_rate": 3.7013801437965945e-05,
"loss": 0.0222,
"step": 12090
},
{
"epoch": 8.356353591160222,
"grad_norm": 0.15770767629146576,
"learning_rate": 3.693398284999623e-05,
"loss": 0.0224,
"step": 12100
},
{
"epoch": 8.363259668508288,
"grad_norm": 0.15847472846508026,
"learning_rate": 3.6854199984018484e-05,
"loss": 0.0219,
"step": 12110
},
{
"epoch": 8.370165745856353,
"grad_norm": 0.16961674392223358,
"learning_rate": 3.677445305815601e-05,
"loss": 0.0234,
"step": 12120
},
{
"epoch": 8.37707182320442,
"grad_norm": 0.14952635765075684,
"learning_rate": 3.669474229043387e-05,
"loss": 0.0213,
"step": 12130
},
{
"epoch": 8.383977900552486,
"grad_norm": 0.1705881506204605,
"learning_rate": 3.6615067898778235e-05,
"loss": 0.0221,
"step": 12140
},
{
"epoch": 8.390883977900552,
"grad_norm": 0.09255818277597427,
"learning_rate": 3.6535430101015866e-05,
"loss": 0.0235,
"step": 12150
},
{
"epoch": 8.397790055248619,
"grad_norm": 0.12237586826086044,
"learning_rate": 3.645582911487345e-05,
"loss": 0.0297,
"step": 12160
},
{
"epoch": 8.404696132596685,
"grad_norm": 0.1676071137189865,
"learning_rate": 3.637626515797706e-05,
"loss": 0.0222,
"step": 12170
},
{
"epoch": 8.411602209944752,
"grad_norm": 0.16139745712280273,
"learning_rate": 3.629673844785152e-05,
"loss": 0.0226,
"step": 12180
},
{
"epoch": 8.418508287292818,
"grad_norm": 0.12163612991571426,
"learning_rate": 3.621724920191979e-05,
"loss": 0.0242,
"step": 12190
},
{
"epoch": 8.425414364640885,
"grad_norm": 0.13052451610565186,
"learning_rate": 3.6137797637502444e-05,
"loss": 0.0196,
"step": 12200
},
{
"epoch": 8.432320441988951,
"grad_norm": 0.16374514997005463,
"learning_rate": 3.6058383971817035e-05,
"loss": 0.0235,
"step": 12210
},
{
"epoch": 8.439226519337016,
"grad_norm": 0.14451929926872253,
"learning_rate": 3.59790084219775e-05,
"loss": 0.022,
"step": 12220
},
{
"epoch": 8.446132596685082,
"grad_norm": 0.16537247598171234,
"learning_rate": 3.589967120499353e-05,
"loss": 0.0278,
"step": 12230
},
{
"epoch": 8.453038674033149,
"grad_norm": 0.17903123795986176,
"learning_rate": 3.5820372537770075e-05,
"loss": 0.0212,
"step": 12240
},
{
"epoch": 8.459944751381215,
"grad_norm": 0.1467219591140747,
"learning_rate": 3.5741112637106655e-05,
"loss": 0.0228,
"step": 12250
},
{
"epoch": 8.466850828729282,
"grad_norm": 0.14892370998859406,
"learning_rate": 3.5661891719696804e-05,
"loss": 0.0264,
"step": 12260
},
{
"epoch": 8.473756906077348,
"grad_norm": 0.10842717438936234,
"learning_rate": 3.5582710002127504e-05,
"loss": 0.0236,
"step": 12270
},
{
"epoch": 8.480662983425415,
"grad_norm": 0.12782053649425507,
"learning_rate": 3.550356770087853e-05,
"loss": 0.0221,
"step": 12280
},
{
"epoch": 8.487569060773481,
"grad_norm": 0.15519659221172333,
"learning_rate": 3.5424465032321914e-05,
"loss": 0.0235,
"step": 12290
},
{
"epoch": 8.494475138121548,
"grad_norm": 0.16490302979946136,
"learning_rate": 3.5345402212721335e-05,
"loss": 0.0188,
"step": 12300
},
{
"epoch": 8.501381215469614,
"grad_norm": 0.150857076048851,
"learning_rate": 3.526637945823152e-05,
"loss": 0.0189,
"step": 12310
},
{
"epoch": 8.50828729281768,
"grad_norm": 0.12046564370393753,
"learning_rate": 3.518739698489767e-05,
"loss": 0.0211,
"step": 12320
},
{
"epoch": 8.515193370165745,
"grad_norm": 0.16205374896526337,
"learning_rate": 3.510845500865485e-05,
"loss": 0.0225,
"step": 12330
},
{
"epoch": 8.522099447513812,
"grad_norm": 0.15190370380878448,
"learning_rate": 3.502955374532739e-05,
"loss": 0.0226,
"step": 12340
},
{
"epoch": 8.529005524861878,
"grad_norm": 0.16450901329517365,
"learning_rate": 3.495069341062836e-05,
"loss": 0.0206,
"step": 12350
},
{
"epoch": 8.535911602209945,
"grad_norm": 0.11976935714483261,
"learning_rate": 3.4871874220158896e-05,
"loss": 0.0214,
"step": 12360
},
{
"epoch": 8.542817679558011,
"grad_norm": 0.13164155185222626,
"learning_rate": 3.479309638940762e-05,
"loss": 0.0197,
"step": 12370
},
{
"epoch": 8.549723756906078,
"grad_norm": 0.13826210796833038,
"learning_rate": 3.4714360133750146e-05,
"loss": 0.025,
"step": 12380
},
{
"epoch": 8.556629834254144,
"grad_norm": 0.16191700100898743,
"learning_rate": 3.463566566844839e-05,
"loss": 0.0222,
"step": 12390
},
{
"epoch": 8.56353591160221,
"grad_norm": 0.14451029896736145,
"learning_rate": 3.4557013208650016e-05,
"loss": 0.0219,
"step": 12400
},
{
"epoch": 8.570441988950277,
"grad_norm": 0.1787310689687729,
"learning_rate": 3.4478402969387857e-05,
"loss": 0.024,
"step": 12410
},
{
"epoch": 8.577348066298342,
"grad_norm": 0.15901915729045868,
"learning_rate": 3.4399835165579266e-05,
"loss": 0.0233,
"step": 12420
},
{
"epoch": 8.584254143646408,
"grad_norm": 0.1560279279947281,
"learning_rate": 3.4321310012025645e-05,
"loss": 0.0188,
"step": 12430
},
{
"epoch": 8.591160220994475,
"grad_norm": 0.15497836470603943,
"learning_rate": 3.424282772341176e-05,
"loss": 0.0238,
"step": 12440
},
{
"epoch": 8.598066298342541,
"grad_norm": 0.13106022775173187,
"learning_rate": 3.416438851430519e-05,
"loss": 0.019,
"step": 12450
},
{
"epoch": 8.604972375690608,
"grad_norm": 0.13551107048988342,
"learning_rate": 3.408599259915577e-05,
"loss": 0.0207,
"step": 12460
},
{
"epoch": 8.611878453038674,
"grad_norm": 0.15204903483390808,
"learning_rate": 3.400764019229487e-05,
"loss": 0.0216,
"step": 12470
},
{
"epoch": 8.61878453038674,
"grad_norm": 0.16227421164512634,
"learning_rate": 3.3929331507935035e-05,
"loss": 0.0237,
"step": 12480
},
{
"epoch": 8.625690607734807,
"grad_norm": 0.1513846218585968,
"learning_rate": 3.3851066760169196e-05,
"loss": 0.0221,
"step": 12490
},
{
"epoch": 8.632596685082873,
"grad_norm": 0.16585606336593628,
"learning_rate": 3.377284616297021e-05,
"loss": 0.0205,
"step": 12500
},
{
"epoch": 8.63950276243094,
"grad_norm": 0.1781190186738968,
"learning_rate": 3.3694669930190166e-05,
"loss": 0.0196,
"step": 12510
},
{
"epoch": 8.646408839779006,
"grad_norm": 0.1248004138469696,
"learning_rate": 3.36165382755599e-05,
"loss": 0.0198,
"step": 12520
},
{
"epoch": 8.653314917127071,
"grad_norm": 0.14076818525791168,
"learning_rate": 3.35384514126884e-05,
"loss": 0.0184,
"step": 12530
},
{
"epoch": 8.660220994475138,
"grad_norm": 0.14751794934272766,
"learning_rate": 3.3460409555062154e-05,
"loss": 0.019,
"step": 12540
},
{
"epoch": 8.667127071823204,
"grad_norm": 0.17103084921836853,
"learning_rate": 3.3382412916044645e-05,
"loss": 0.0224,
"step": 12550
},
{
"epoch": 8.67403314917127,
"grad_norm": 0.1588471233844757,
"learning_rate": 3.330446170887566e-05,
"loss": 0.0216,
"step": 12560
},
{
"epoch": 8.680939226519337,
"grad_norm": 0.13995952904224396,
"learning_rate": 3.3226556146670834e-05,
"loss": 0.0202,
"step": 12570
},
{
"epoch": 8.687845303867404,
"grad_norm": 0.17354713380336761,
"learning_rate": 3.314869644242102e-05,
"loss": 0.0255,
"step": 12580
},
{
"epoch": 8.69475138121547,
"grad_norm": 0.1409205198287964,
"learning_rate": 3.3070882808991674e-05,
"loss": 0.0216,
"step": 12590
},
{
"epoch": 8.701657458563536,
"grad_norm": 0.15135961771011353,
"learning_rate": 3.2993115459122305e-05,
"loss": 0.0229,
"step": 12600
},
{
"epoch": 8.708563535911603,
"grad_norm": 0.1787615269422531,
"learning_rate": 3.2915394605425835e-05,
"loss": 0.0204,
"step": 12610
},
{
"epoch": 8.715469613259668,
"grad_norm": 0.1657034307718277,
"learning_rate": 3.283772046038816e-05,
"loss": 0.021,
"step": 12620
},
{
"epoch": 8.722375690607734,
"grad_norm": 0.14173857867717743,
"learning_rate": 3.276009323636739e-05,
"loss": 0.0189,
"step": 12630
},
{
"epoch": 8.7292817679558,
"grad_norm": 0.1563398241996765,
"learning_rate": 3.268251314559344e-05,
"loss": 0.0205,
"step": 12640
},
{
"epoch": 8.736187845303867,
"grad_norm": 0.1378328651189804,
"learning_rate": 3.2604980400167254e-05,
"loss": 0.02,
"step": 12650
},
{
"epoch": 8.743093922651934,
"grad_norm": 0.20059895515441895,
"learning_rate": 3.252749521206042e-05,
"loss": 0.0211,
"step": 12660
},
{
"epoch": 8.75,
"grad_norm": 0.11326922476291656,
"learning_rate": 3.2450057793114494e-05,
"loss": 0.0238,
"step": 12670
},
{
"epoch": 8.756906077348066,
"grad_norm": 0.17236323654651642,
"learning_rate": 3.2372668355040435e-05,
"loss": 0.0204,
"step": 12680
},
{
"epoch": 8.763812154696133,
"grad_norm": 0.10979443788528442,
"learning_rate": 3.2295327109418005e-05,
"loss": 0.0238,
"step": 12690
},
{
"epoch": 8.7707182320442,
"grad_norm": 0.11837854981422424,
"learning_rate": 3.221803426769518e-05,
"loss": 0.024,
"step": 12700
},
{
"epoch": 8.777624309392266,
"grad_norm": 0.10363747179508209,
"learning_rate": 3.214079004118768e-05,
"loss": 0.0213,
"step": 12710
},
{
"epoch": 8.784530386740332,
"grad_norm": 0.1248355507850647,
"learning_rate": 3.2063594641078234e-05,
"loss": 0.0223,
"step": 12720
},
{
"epoch": 8.791436464088397,
"grad_norm": 0.09620241075754166,
"learning_rate": 3.198644827841616e-05,
"loss": 0.0212,
"step": 12730
},
{
"epoch": 8.798342541436464,
"grad_norm": 0.16484878957271576,
"learning_rate": 3.1909351164116654e-05,
"loss": 0.0198,
"step": 12740
},
{
"epoch": 8.80524861878453,
"grad_norm": 0.15809795260429382,
"learning_rate": 3.183230350896026e-05,
"loss": 0.0216,
"step": 12750
},
{
"epoch": 8.812154696132596,
"grad_norm": 0.07511230558156967,
"learning_rate": 3.1755305523592337e-05,
"loss": 0.0226,
"step": 12760
},
{
"epoch": 8.819060773480663,
"grad_norm": 0.0902475044131279,
"learning_rate": 3.167835741852245e-05,
"loss": 0.0209,
"step": 12770
},
{
"epoch": 8.82596685082873,
"grad_norm": 0.16243767738342285,
"learning_rate": 3.160145940412378e-05,
"loss": 0.0254,
"step": 12780
},
{
"epoch": 8.832872928176796,
"grad_norm": 0.17259711027145386,
"learning_rate": 3.1524611690632545e-05,
"loss": 0.0239,
"step": 12790
},
{
"epoch": 8.839779005524862,
"grad_norm": 0.11631940305233002,
"learning_rate": 3.144781448814746e-05,
"loss": 0.022,
"step": 12800
},
{
"epoch": 8.846685082872929,
"grad_norm": 0.14422062039375305,
"learning_rate": 3.1371068006629145e-05,
"loss": 0.0266,
"step": 12810
},
{
"epoch": 8.853591160220994,
"grad_norm": 0.11181170493364334,
"learning_rate": 3.129437245589956e-05,
"loss": 0.0235,
"step": 12820
},
{
"epoch": 8.86049723756906,
"grad_norm": 0.168172687292099,
"learning_rate": 3.121772804564143e-05,
"loss": 0.0215,
"step": 12830
},
{
"epoch": 8.867403314917127,
"grad_norm": 0.14143504202365875,
"learning_rate": 3.11411349853976e-05,
"loss": 0.0245,
"step": 12840
},
{
"epoch": 8.874309392265193,
"grad_norm": 0.12448939681053162,
"learning_rate": 3.10645934845706e-05,
"loss": 0.0194,
"step": 12850
},
{
"epoch": 8.88121546961326,
"grad_norm": 0.11174539476633072,
"learning_rate": 3.098810375242196e-05,
"loss": 0.0196,
"step": 12860
},
{
"epoch": 8.888121546961326,
"grad_norm": 0.15435843169689178,
"learning_rate": 3.0911665998071704e-05,
"loss": 0.0231,
"step": 12870
},
{
"epoch": 8.895027624309392,
"grad_norm": 0.14516079425811768,
"learning_rate": 3.083528043049774e-05,
"loss": 0.021,
"step": 12880
},
{
"epoch": 8.901933701657459,
"grad_norm": 0.17633748054504395,
"learning_rate": 3.0758947258535255e-05,
"loss": 0.0216,
"step": 12890
},
{
"epoch": 8.908839779005525,
"grad_norm": 0.112397700548172,
"learning_rate": 3.068266669087625e-05,
"loss": 0.0197,
"step": 12900
},
{
"epoch": 8.915745856353592,
"grad_norm": 0.27577459812164307,
"learning_rate": 3.060643893606887e-05,
"loss": 0.0217,
"step": 12910
},
{
"epoch": 8.922651933701658,
"grad_norm": 0.15804259479045868,
"learning_rate": 3.053026420251693e-05,
"loss": 0.0202,
"step": 12920
},
{
"epoch": 8.929558011049723,
"grad_norm": 0.14614105224609375,
"learning_rate": 3.0454142698479183e-05,
"loss": 0.0241,
"step": 12930
},
{
"epoch": 8.93646408839779,
"grad_norm": 0.1321985125541687,
"learning_rate": 3.0378074632068954e-05,
"loss": 0.0193,
"step": 12940
},
{
"epoch": 8.943370165745856,
"grad_norm": 0.17028087377548218,
"learning_rate": 3.0302060211253408e-05,
"loss": 0.0222,
"step": 12950
},
{
"epoch": 8.950276243093922,
"grad_norm": 0.14149953424930573,
"learning_rate": 3.0226099643853073e-05,
"loss": 0.0193,
"step": 12960
},
{
"epoch": 8.957182320441989,
"grad_norm": 0.16676829755306244,
"learning_rate": 3.0150193137541283e-05,
"loss": 0.0244,
"step": 12970
},
{
"epoch": 8.964088397790055,
"grad_norm": 0.13386687636375427,
"learning_rate": 3.0074340899843467e-05,
"loss": 0.022,
"step": 12980
},
{
"epoch": 8.970994475138122,
"grad_norm": 0.13977794349193573,
"learning_rate": 2.999854313813677e-05,
"loss": 0.0204,
"step": 12990
},
{
"epoch": 8.977900552486188,
"grad_norm": 0.14581474661827087,
"learning_rate": 2.9922800059649382e-05,
"loss": 0.0211,
"step": 13000
},
{
"epoch": 8.984806629834255,
"grad_norm": 0.15299993753433228,
"learning_rate": 2.9847111871459976e-05,
"loss": 0.0245,
"step": 13010
},
{
"epoch": 8.99171270718232,
"grad_norm": 0.13112711906433105,
"learning_rate": 2.977147878049721e-05,
"loss": 0.0226,
"step": 13020
},
{
"epoch": 8.998618784530386,
"grad_norm": 0.16714787483215332,
"learning_rate": 2.9695900993539006e-05,
"loss": 0.0233,
"step": 13030
},
{
"epoch": 9.005524861878452,
"grad_norm": 0.1377313733100891,
"learning_rate": 2.9620378717212183e-05,
"loss": 0.0226,
"step": 13040
},
{
"epoch": 9.012430939226519,
"grad_norm": 0.15656685829162598,
"learning_rate": 2.9544912157991745e-05,
"loss": 0.0218,
"step": 13050
},
{
"epoch": 9.019337016574585,
"grad_norm": 0.2332046627998352,
"learning_rate": 2.9469501522200405e-05,
"loss": 0.0247,
"step": 13060
},
{
"epoch": 9.026243093922652,
"grad_norm": 0.14399616420269012,
"learning_rate": 2.9394147016007946e-05,
"loss": 0.0177,
"step": 13070
},
{
"epoch": 9.033149171270718,
"grad_norm": 0.12218581885099411,
"learning_rate": 2.9318848845430702e-05,
"loss": 0.0208,
"step": 13080
},
{
"epoch": 9.040055248618785,
"grad_norm": 0.1258987933397293,
"learning_rate": 2.9243607216331013e-05,
"loss": 0.0229,
"step": 13090
},
{
"epoch": 9.046961325966851,
"grad_norm": 0.11139830201864243,
"learning_rate": 2.916842233441661e-05,
"loss": 0.0196,
"step": 13100
},
{
"epoch": 9.053867403314918,
"grad_norm": 0.18543629348278046,
"learning_rate": 2.90932944052401e-05,
"loss": 0.0215,
"step": 13110
},
{
"epoch": 9.060773480662984,
"grad_norm": 0.17765985429286957,
"learning_rate": 2.9018223634198354e-05,
"loss": 0.0203,
"step": 13120
},
{
"epoch": 9.067679558011049,
"grad_norm": 0.1288689821958542,
"learning_rate": 2.8943210226532025e-05,
"loss": 0.0256,
"step": 13130
},
{
"epoch": 9.074585635359115,
"grad_norm": 0.12179256230592728,
"learning_rate": 2.8868254387324857e-05,
"loss": 0.0204,
"step": 13140
},
{
"epoch": 9.081491712707182,
"grad_norm": 0.1416948288679123,
"learning_rate": 2.8793356321503306e-05,
"loss": 0.0211,
"step": 13150
},
{
"epoch": 9.088397790055248,
"grad_norm": 0.1624247431755066,
"learning_rate": 2.87185162338358e-05,
"loss": 0.0184,
"step": 13160
},
{
"epoch": 9.095303867403315,
"grad_norm": 0.19820043444633484,
"learning_rate": 2.8643734328932253e-05,
"loss": 0.021,
"step": 13170
},
{
"epoch": 9.102209944751381,
"grad_norm": 0.1763990819454193,
"learning_rate": 2.856901081124359e-05,
"loss": 0.0204,
"step": 13180
},
{
"epoch": 9.109116022099448,
"grad_norm": 0.09194623678922653,
"learning_rate": 2.8494345885061002e-05,
"loss": 0.019,
"step": 13190
},
{
"epoch": 9.116022099447514,
"grad_norm": 0.12375927716493607,
"learning_rate": 2.8419739754515616e-05,
"loss": 0.025,
"step": 13200
},
{
"epoch": 9.12292817679558,
"grad_norm": 0.18553689122200012,
"learning_rate": 2.8345192623577666e-05,
"loss": 0.0233,
"step": 13210
},
{
"epoch": 9.129834254143647,
"grad_norm": 0.14989089965820312,
"learning_rate": 2.8270704696056193e-05,
"loss": 0.0212,
"step": 13220
},
{
"epoch": 9.136740331491712,
"grad_norm": 0.13519397377967834,
"learning_rate": 2.8196276175598367e-05,
"loss": 0.0234,
"step": 13230
},
{
"epoch": 9.143646408839778,
"grad_norm": 0.09733614325523376,
"learning_rate": 2.8121907265688884e-05,
"loss": 0.0271,
"step": 13240
},
{
"epoch": 9.150552486187845,
"grad_norm": 0.12137151509523392,
"learning_rate": 2.804759816964957e-05,
"loss": 0.0199,
"step": 13250
},
{
"epoch": 9.157458563535911,
"grad_norm": 0.1408531367778778,
"learning_rate": 2.797334909063857e-05,
"loss": 0.0218,
"step": 13260
},
{
"epoch": 9.164364640883978,
"grad_norm": 0.1920802891254425,
"learning_rate": 2.7899160231650056e-05,
"loss": 0.0229,
"step": 13270
},
{
"epoch": 9.171270718232044,
"grad_norm": 0.10271954536437988,
"learning_rate": 2.7825031795513585e-05,
"loss": 0.0242,
"step": 13280
},
{
"epoch": 9.17817679558011,
"grad_norm": 0.16070392727851868,
"learning_rate": 2.775096398489341e-05,
"loss": 0.0214,
"step": 13290
},
{
"epoch": 9.185082872928177,
"grad_norm": 0.1328756958246231,
"learning_rate": 2.7676957002288163e-05,
"loss": 0.0268,
"step": 13300
},
{
"epoch": 9.191988950276244,
"grad_norm": 0.11246970295906067,
"learning_rate": 2.760301105003003e-05,
"loss": 0.0208,
"step": 13310
},
{
"epoch": 9.19889502762431,
"grad_norm": 0.149558886885643,
"learning_rate": 2.752912633028446e-05,
"loss": 0.019,
"step": 13320
},
{
"epoch": 9.205801104972375,
"grad_norm": 0.14351671934127808,
"learning_rate": 2.7455303045049474e-05,
"loss": 0.0237,
"step": 13330
},
{
"epoch": 9.212707182320441,
"grad_norm": 0.15685635805130005,
"learning_rate": 2.7381541396155098e-05,
"loss": 0.0219,
"step": 13340
},
{
"epoch": 9.219613259668508,
"grad_norm": 0.19899433851242065,
"learning_rate": 2.730784158526286e-05,
"loss": 0.022,
"step": 13350
},
{
"epoch": 9.226519337016574,
"grad_norm": 0.13601619005203247,
"learning_rate": 2.723420381386521e-05,
"loss": 0.0208,
"step": 13360
},
{
"epoch": 9.23342541436464,
"grad_norm": 0.14668749272823334,
"learning_rate": 2.7160628283285018e-05,
"loss": 0.0205,
"step": 13370
},
{
"epoch": 9.240331491712707,
"grad_norm": 0.15259882807731628,
"learning_rate": 2.7087115194675007e-05,
"loss": 0.0209,
"step": 13380
},
{
"epoch": 9.247237569060774,
"grad_norm": 0.14573076367378235,
"learning_rate": 2.701366474901712e-05,
"loss": 0.0216,
"step": 13390
},
{
"epoch": 9.25414364640884,
"grad_norm": 0.12216416746377945,
"learning_rate": 2.6940277147122085e-05,
"loss": 0.0176,
"step": 13400
},
{
"epoch": 9.261049723756907,
"grad_norm": 0.16968750953674316,
"learning_rate": 2.686695258962878e-05,
"loss": 0.0211,
"step": 13410
},
{
"epoch": 9.267955801104973,
"grad_norm": 0.19388921558856964,
"learning_rate": 2.679369127700375e-05,
"loss": 0.0219,
"step": 13420
},
{
"epoch": 9.274861878453038,
"grad_norm": 0.15601925551891327,
"learning_rate": 2.672049340954067e-05,
"loss": 0.0254,
"step": 13430
},
{
"epoch": 9.281767955801104,
"grad_norm": 0.14723803102970123,
"learning_rate": 2.6647359187359676e-05,
"loss": 0.021,
"step": 13440
},
{
"epoch": 9.28867403314917,
"grad_norm": 0.1670716404914856,
"learning_rate": 2.6574288810406946e-05,
"loss": 0.0259,
"step": 13450
},
{
"epoch": 9.295580110497237,
"grad_norm": 0.1267160326242447,
"learning_rate": 2.6501282478454083e-05,
"loss": 0.0196,
"step": 13460
},
{
"epoch": 9.302486187845304,
"grad_norm": 0.18623648583889008,
"learning_rate": 2.6428340391097618e-05,
"loss": 0.0249,
"step": 13470
},
{
"epoch": 9.30939226519337,
"grad_norm": 0.15965907275676727,
"learning_rate": 2.6355462747758485e-05,
"loss": 0.0171,
"step": 13480
},
{
"epoch": 9.316298342541437,
"grad_norm": 0.13149818778038025,
"learning_rate": 2.6282649747681304e-05,
"loss": 0.0229,
"step": 13490
},
{
"epoch": 9.323204419889503,
"grad_norm": 0.11458833515644073,
"learning_rate": 2.620990158993406e-05,
"loss": 0.0257,
"step": 13500
},
{
"epoch": 9.33011049723757,
"grad_norm": 0.12951171398162842,
"learning_rate": 2.6137218473407477e-05,
"loss": 0.0231,
"step": 13510
},
{
"epoch": 9.337016574585636,
"grad_norm": 0.14793288707733154,
"learning_rate": 2.606460059681436e-05,
"loss": 0.0185,
"step": 13520
},
{
"epoch": 9.3439226519337,
"grad_norm": 0.12049033492803574,
"learning_rate": 2.599204815868928e-05,
"loss": 0.0178,
"step": 13530
},
{
"epoch": 9.350828729281767,
"grad_norm": 0.17162886261940002,
"learning_rate": 2.5919561357387756e-05,
"loss": 0.023,
"step": 13540
},
{
"epoch": 9.357734806629834,
"grad_norm": 0.12784846127033234,
"learning_rate": 2.5847140391085972e-05,
"loss": 0.0243,
"step": 13550
},
{
"epoch": 9.3646408839779,
"grad_norm": 0.12364140897989273,
"learning_rate": 2.5774785457780103e-05,
"loss": 0.0229,
"step": 13560
},
{
"epoch": 9.371546961325967,
"grad_norm": 0.0785508081316948,
"learning_rate": 2.5702496755285753e-05,
"loss": 0.0194,
"step": 13570
},
{
"epoch": 9.378453038674033,
"grad_norm": 0.15732654929161072,
"learning_rate": 2.5630274481237483e-05,
"loss": 0.0246,
"step": 13580
},
{
"epoch": 9.3853591160221,
"grad_norm": 0.12818360328674316,
"learning_rate": 2.5558118833088197e-05,
"loss": 0.0222,
"step": 13590
},
{
"epoch": 9.392265193370166,
"grad_norm": 0.10840357095003128,
"learning_rate": 2.548603000810872e-05,
"loss": 0.023,
"step": 13600
},
{
"epoch": 9.399171270718233,
"grad_norm": 0.12543641030788422,
"learning_rate": 2.5414008203387152e-05,
"loss": 0.0198,
"step": 13610
},
{
"epoch": 9.4060773480663,
"grad_norm": 0.12083574384450912,
"learning_rate": 2.534205361582834e-05,
"loss": 0.0237,
"step": 13620
},
{
"epoch": 9.412983425414364,
"grad_norm": 0.09807329624891281,
"learning_rate": 2.527016644215338e-05,
"loss": 0.0224,
"step": 13630
},
{
"epoch": 9.41988950276243,
"grad_norm": 0.16850584745407104,
"learning_rate": 2.519834687889905e-05,
"loss": 0.0248,
"step": 13640
},
{
"epoch": 9.426795580110497,
"grad_norm": 0.11073677241802216,
"learning_rate": 2.5126595122417295e-05,
"loss": 0.0183,
"step": 13650
},
{
"epoch": 9.433701657458563,
"grad_norm": 0.12772704660892487,
"learning_rate": 2.5054911368874713e-05,
"loss": 0.0248,
"step": 13660
},
{
"epoch": 9.44060773480663,
"grad_norm": 0.13772784173488617,
"learning_rate": 2.4983295814251916e-05,
"loss": 0.0179,
"step": 13670
},
{
"epoch": 9.447513812154696,
"grad_norm": 0.1286805272102356,
"learning_rate": 2.4911748654343105e-05,
"loss": 0.021,
"step": 13680
},
{
"epoch": 9.454419889502763,
"grad_norm": 0.14480936527252197,
"learning_rate": 2.4840270084755463e-05,
"loss": 0.0242,
"step": 13690
},
{
"epoch": 9.46132596685083,
"grad_norm": 0.22492210566997528,
"learning_rate": 2.4768860300908685e-05,
"loss": 0.0227,
"step": 13700
},
{
"epoch": 9.468232044198896,
"grad_norm": 0.15835162997245789,
"learning_rate": 2.469751949803443e-05,
"loss": 0.0222,
"step": 13710
},
{
"epoch": 9.475138121546962,
"grad_norm": 0.1333426535129547,
"learning_rate": 2.4626247871175666e-05,
"loss": 0.018,
"step": 13720
},
{
"epoch": 9.482044198895027,
"grad_norm": 0.06573522090911865,
"learning_rate": 2.4555045615186346e-05,
"loss": 0.0217,
"step": 13730
},
{
"epoch": 9.488950276243093,
"grad_norm": 0.13375243544578552,
"learning_rate": 2.4483912924730677e-05,
"loss": 0.0205,
"step": 13740
},
{
"epoch": 9.49585635359116,
"grad_norm": 0.14595822989940643,
"learning_rate": 2.4412849994282742e-05,
"loss": 0.0242,
"step": 13750
},
{
"epoch": 9.502762430939226,
"grad_norm": 0.16502094268798828,
"learning_rate": 2.434185701812592e-05,
"loss": 0.0196,
"step": 13760
},
{
"epoch": 9.509668508287293,
"grad_norm": 0.15429244935512543,
"learning_rate": 2.4270934190352218e-05,
"loss": 0.0221,
"step": 13770
},
{
"epoch": 9.51657458563536,
"grad_norm": 0.12383496761322021,
"learning_rate": 2.4200081704861998e-05,
"loss": 0.0203,
"step": 13780
},
{
"epoch": 9.523480662983426,
"grad_norm": 0.1347363144159317,
"learning_rate": 2.412929975536321e-05,
"loss": 0.0205,
"step": 13790
},
{
"epoch": 9.530386740331492,
"grad_norm": 0.17767280340194702,
"learning_rate": 2.4058588535371017e-05,
"loss": 0.0241,
"step": 13800
},
{
"epoch": 9.537292817679559,
"grad_norm": 0.16614821553230286,
"learning_rate": 2.3987948238207243e-05,
"loss": 0.0261,
"step": 13810
},
{
"epoch": 9.544198895027625,
"grad_norm": 0.15801551938056946,
"learning_rate": 2.3917379056999678e-05,
"loss": 0.0212,
"step": 13820
},
{
"epoch": 9.55110497237569,
"grad_norm": 0.16081950068473816,
"learning_rate": 2.3846881184681824e-05,
"loss": 0.022,
"step": 13830
},
{
"epoch": 9.558011049723756,
"grad_norm": 0.18439306318759918,
"learning_rate": 2.377645481399214e-05,
"loss": 0.0217,
"step": 13840
},
{
"epoch": 9.564917127071823,
"grad_norm": 0.10785002261400223,
"learning_rate": 2.3706100137473667e-05,
"loss": 0.0198,
"step": 13850
},
{
"epoch": 9.57182320441989,
"grad_norm": 0.12533661723136902,
"learning_rate": 2.3635817347473394e-05,
"loss": 0.0208,
"step": 13860
},
{
"epoch": 9.578729281767956,
"grad_norm": 0.1306634098291397,
"learning_rate": 2.3565606636141757e-05,
"loss": 0.0175,
"step": 13870
},
{
"epoch": 9.585635359116022,
"grad_norm": 0.1544395536184311,
"learning_rate": 2.3495468195432203e-05,
"loss": 0.0192,
"step": 13880
},
{
"epoch": 9.592541436464089,
"grad_norm": 0.0991256982088089,
"learning_rate": 2.3425402217100507e-05,
"loss": 0.02,
"step": 13890
},
{
"epoch": 9.599447513812155,
"grad_norm": 0.09305687248706818,
"learning_rate": 2.3355408892704424e-05,
"loss": 0.021,
"step": 13900
},
{
"epoch": 9.606353591160222,
"grad_norm": 0.14391951262950897,
"learning_rate": 2.3285488413603003e-05,
"loss": 0.0193,
"step": 13910
},
{
"epoch": 9.613259668508288,
"grad_norm": 0.17865988612174988,
"learning_rate": 2.321564097095615e-05,
"loss": 0.0215,
"step": 13920
},
{
"epoch": 9.620165745856355,
"grad_norm": 0.10739440470933914,
"learning_rate": 2.3145866755724142e-05,
"loss": 0.0228,
"step": 13930
},
{
"epoch": 9.62707182320442,
"grad_norm": 0.14161740243434906,
"learning_rate": 2.307616595866699e-05,
"loss": 0.0189,
"step": 13940
},
{
"epoch": 9.633977900552486,
"grad_norm": 0.21180714666843414,
"learning_rate": 2.3006538770344032e-05,
"loss": 0.0233,
"step": 13950
},
{
"epoch": 9.640883977900552,
"grad_norm": 0.11193030327558517,
"learning_rate": 2.293698538111334e-05,
"loss": 0.0205,
"step": 13960
},
{
"epoch": 9.647790055248619,
"grad_norm": 0.2148815244436264,
"learning_rate": 2.28675059811312e-05,
"loss": 0.0193,
"step": 13970
},
{
"epoch": 9.654696132596685,
"grad_norm": 0.13390229642391205,
"learning_rate": 2.279810076035167e-05,
"loss": 0.02,
"step": 13980
},
{
"epoch": 9.661602209944752,
"grad_norm": 0.0927250012755394,
"learning_rate": 2.272876990852596e-05,
"loss": 0.0197,
"step": 13990
},
{
"epoch": 9.668508287292818,
"grad_norm": 0.12975889444351196,
"learning_rate": 2.265951361520195e-05,
"loss": 0.0203,
"step": 14000
},
{
"epoch": 9.675414364640885,
"grad_norm": 0.1706344038248062,
"learning_rate": 2.2590332069723748e-05,
"loss": 0.019,
"step": 14010
},
{
"epoch": 9.682320441988951,
"grad_norm": 0.21264302730560303,
"learning_rate": 2.2521225461231004e-05,
"loss": 0.0209,
"step": 14020
},
{
"epoch": 9.689226519337016,
"grad_norm": 0.14462417364120483,
"learning_rate": 2.2452193978658597e-05,
"loss": 0.0222,
"step": 14030
},
{
"epoch": 9.696132596685082,
"grad_norm": 0.12847472727298737,
"learning_rate": 2.238323781073594e-05,
"loss": 0.0194,
"step": 14040
},
{
"epoch": 9.703038674033149,
"grad_norm": 0.1844336837530136,
"learning_rate": 2.2314357145986552e-05,
"loss": 0.0198,
"step": 14050
},
{
"epoch": 9.709944751381215,
"grad_norm": 0.16759341955184937,
"learning_rate": 2.224555217272757e-05,
"loss": 0.0194,
"step": 14060
},
{
"epoch": 9.716850828729282,
"grad_norm": 0.10547558218240738,
"learning_rate": 2.2176823079069127e-05,
"loss": 0.0214,
"step": 14070
},
{
"epoch": 9.723756906077348,
"grad_norm": 0.11423495411872864,
"learning_rate": 2.210817005291398e-05,
"loss": 0.0208,
"step": 14080
},
{
"epoch": 9.730662983425415,
"grad_norm": 0.13519664108753204,
"learning_rate": 2.203959328195686e-05,
"loss": 0.0195,
"step": 14090
},
{
"epoch": 9.737569060773481,
"grad_norm": 0.14309315383434296,
"learning_rate": 2.1971092953684026e-05,
"loss": 0.0192,
"step": 14100
},
{
"epoch": 9.744475138121548,
"grad_norm": 0.10871468484401703,
"learning_rate": 2.1902669255372788e-05,
"loss": 0.0171,
"step": 14110
},
{
"epoch": 9.751381215469614,
"grad_norm": 0.15495355427265167,
"learning_rate": 2.1834322374090897e-05,
"loss": 0.0203,
"step": 14120
},
{
"epoch": 9.75828729281768,
"grad_norm": 0.10269203782081604,
"learning_rate": 2.1766052496696153e-05,
"loss": 0.0189,
"step": 14130
},
{
"epoch": 9.765193370165745,
"grad_norm": 0.17774659395217896,
"learning_rate": 2.169785980983577e-05,
"loss": 0.0239,
"step": 14140
},
{
"epoch": 9.772099447513812,
"grad_norm": 0.12673352658748627,
"learning_rate": 2.162974449994593e-05,
"loss": 0.02,
"step": 14150
},
{
"epoch": 9.779005524861878,
"grad_norm": 0.12161792069673538,
"learning_rate": 2.1561706753251337e-05,
"loss": 0.0195,
"step": 14160
},
{
"epoch": 9.785911602209945,
"grad_norm": 0.08473629504442215,
"learning_rate": 2.1493746755764544e-05,
"loss": 0.0208,
"step": 14170
},
{
"epoch": 9.792817679558011,
"grad_norm": 0.16699481010437012,
"learning_rate": 2.1425864693285635e-05,
"loss": 0.0224,
"step": 14180
},
{
"epoch": 9.799723756906078,
"grad_norm": 0.11192629486322403,
"learning_rate": 2.1358060751401547e-05,
"loss": 0.0211,
"step": 14190
},
{
"epoch": 9.806629834254144,
"grad_norm": 0.16285032033920288,
"learning_rate": 2.129033511548566e-05,
"loss": 0.0197,
"step": 14200
},
{
"epoch": 9.81353591160221,
"grad_norm": 0.117989681661129,
"learning_rate": 2.1222687970697315e-05,
"loss": 0.0196,
"step": 14210
},
{
"epoch": 9.820441988950277,
"grad_norm": 0.13115745782852173,
"learning_rate": 2.1155119501981173e-05,
"loss": 0.0244,
"step": 14220
},
{
"epoch": 9.827348066298342,
"grad_norm": 0.2276378720998764,
"learning_rate": 2.1087629894066895e-05,
"loss": 0.0225,
"step": 14230
},
{
"epoch": 9.834254143646408,
"grad_norm": 0.09082102030515671,
"learning_rate": 2.1020219331468473e-05,
"loss": 0.0177,
"step": 14240
},
{
"epoch": 9.841160220994475,
"grad_norm": 0.17138603329658508,
"learning_rate": 2.095288799848379e-05,
"loss": 0.0239,
"step": 14250
},
{
"epoch": 9.848066298342541,
"grad_norm": 0.07798614352941513,
"learning_rate": 2.088563607919417e-05,
"loss": 0.0206,
"step": 14260
},
{
"epoch": 9.854972375690608,
"grad_norm": 0.14047075808048248,
"learning_rate": 2.0818463757463786e-05,
"loss": 0.018,
"step": 14270
},
{
"epoch": 9.861878453038674,
"grad_norm": 0.17085087299346924,
"learning_rate": 2.0751371216939175e-05,
"loss": 0.021,
"step": 14280
},
{
"epoch": 9.86878453038674,
"grad_norm": 0.14489038288593292,
"learning_rate": 2.068435864104882e-05,
"loss": 0.0159,
"step": 14290
},
{
"epoch": 9.875690607734807,
"grad_norm": 0.11485548317432404,
"learning_rate": 2.0617426213002506e-05,
"loss": 0.0179,
"step": 14300
},
{
"epoch": 9.882596685082873,
"grad_norm": 0.12750490009784698,
"learning_rate": 2.055057411579097e-05,
"loss": 0.017,
"step": 14310
},
{
"epoch": 9.88950276243094,
"grad_norm": 0.17045387625694275,
"learning_rate": 2.0483802532185286e-05,
"loss": 0.0238,
"step": 14320
},
{
"epoch": 9.896408839779006,
"grad_norm": 0.1332184374332428,
"learning_rate": 2.041711164473638e-05,
"loss": 0.0226,
"step": 14330
},
{
"epoch": 9.903314917127071,
"grad_norm": 0.16046547889709473,
"learning_rate": 2.0350501635774637e-05,
"loss": 0.0191,
"step": 14340
},
{
"epoch": 9.910220994475138,
"grad_norm": 0.15064607560634613,
"learning_rate": 2.0283972687409247e-05,
"loss": 0.0225,
"step": 14350
},
{
"epoch": 9.917127071823204,
"grad_norm": 0.11720279604196548,
"learning_rate": 2.021752498152784e-05,
"loss": 0.0215,
"step": 14360
},
{
"epoch": 9.92403314917127,
"grad_norm": 0.16652604937553406,
"learning_rate": 2.015115869979589e-05,
"loss": 0.0258,
"step": 14370
},
{
"epoch": 9.930939226519337,
"grad_norm": 0.12095716595649719,
"learning_rate": 2.0084874023656265e-05,
"loss": 0.022,
"step": 14380
},
{
"epoch": 9.937845303867404,
"grad_norm": 0.0894702821969986,
"learning_rate": 2.001867113432877e-05,
"loss": 0.0218,
"step": 14390
},
{
"epoch": 9.94475138121547,
"grad_norm": 0.20566357672214508,
"learning_rate": 1.995255021280954e-05,
"loss": 0.0187,
"step": 14400
},
{
"epoch": 9.951657458563536,
"grad_norm": 0.11899957805871964,
"learning_rate": 1.9886511439870688e-05,
"loss": 0.0217,
"step": 14410
},
{
"epoch": 9.958563535911603,
"grad_norm": 0.1229172870516777,
"learning_rate": 1.9820554996059675e-05,
"loss": 0.0197,
"step": 14420
},
{
"epoch": 9.965469613259668,
"grad_norm": 0.16941682994365692,
"learning_rate": 1.9754681061698893e-05,
"loss": 0.018,
"step": 14430
},
{
"epoch": 9.972375690607734,
"grad_norm": 0.13610994815826416,
"learning_rate": 1.9688889816885185e-05,
"loss": 0.0182,
"step": 14440
},
{
"epoch": 9.9792817679558,
"grad_norm": 0.14665129780769348,
"learning_rate": 1.962318144148928e-05,
"loss": 0.02,
"step": 14450
},
{
"epoch": 9.986187845303867,
"grad_norm": 0.13639798760414124,
"learning_rate": 1.955755611515539e-05,
"loss": 0.0269,
"step": 14460
},
{
"epoch": 9.993093922651934,
"grad_norm": 0.1341027170419693,
"learning_rate": 1.9492014017300642e-05,
"loss": 0.0203,
"step": 14470
},
{
"epoch": 10.0,
"grad_norm": 0.20185670256614685,
"learning_rate": 1.942655532711461e-05,
"loss": 0.0208,
"step": 14480
},
{
"epoch": 10.006906077348066,
"grad_norm": 0.09617722779512405,
"learning_rate": 1.9361180223558882e-05,
"loss": 0.0225,
"step": 14490
},
{
"epoch": 10.013812154696133,
"grad_norm": 0.08813314139842987,
"learning_rate": 1.929588888536647e-05,
"loss": 0.0171,
"step": 14500
},
{
"epoch": 10.0207182320442,
"grad_norm": 0.10605496168136597,
"learning_rate": 1.9230681491041425e-05,
"loss": 0.0175,
"step": 14510
},
{
"epoch": 10.027624309392266,
"grad_norm": 0.12312556058168411,
"learning_rate": 1.9165558218858264e-05,
"loss": 0.0174,
"step": 14520
},
{
"epoch": 10.034530386740332,
"grad_norm": 0.13579706847667694,
"learning_rate": 1.9100519246861505e-05,
"loss": 0.0179,
"step": 14530
},
{
"epoch": 10.041436464088397,
"grad_norm": 0.15092580020427704,
"learning_rate": 1.9035564752865248e-05,
"loss": 0.0209,
"step": 14540
},
{
"epoch": 10.048342541436464,
"grad_norm": 0.15202099084854126,
"learning_rate": 1.897069491445258e-05,
"loss": 0.018,
"step": 14550
},
{
"epoch": 10.05524861878453,
"grad_norm": 0.1487421989440918,
"learning_rate": 1.890590990897515e-05,
"loss": 0.0234,
"step": 14560
},
{
"epoch": 10.062154696132596,
"grad_norm": 0.14006099104881287,
"learning_rate": 1.884120991355272e-05,
"loss": 0.0223,
"step": 14570
},
{
"epoch": 10.069060773480663,
"grad_norm": 0.09489653259515762,
"learning_rate": 1.8776595105072576e-05,
"loss": 0.0203,
"step": 14580
},
{
"epoch": 10.07596685082873,
"grad_norm": 0.09173377603292465,
"learning_rate": 1.8712065660189166e-05,
"loss": 0.0186,
"step": 14590
},
{
"epoch": 10.082872928176796,
"grad_norm": 0.17884795367717743,
"learning_rate": 1.8647621755323513e-05,
"loss": 0.0182,
"step": 14600
},
{
"epoch": 10.089779005524862,
"grad_norm": 0.1608448177576065,
"learning_rate": 1.858326356666278e-05,
"loss": 0.0194,
"step": 14610
},
{
"epoch": 10.096685082872929,
"grad_norm": 0.11083773523569107,
"learning_rate": 1.851899127015983e-05,
"loss": 0.0197,
"step": 14620
},
{
"epoch": 10.103591160220995,
"grad_norm": 0.1610211730003357,
"learning_rate": 1.8454805041532626e-05,
"loss": 0.0244,
"step": 14630
},
{
"epoch": 10.11049723756906,
"grad_norm": 0.16736485064029694,
"learning_rate": 1.8390705056263906e-05,
"loss": 0.0214,
"step": 14640
},
{
"epoch": 10.117403314917127,
"grad_norm": 0.16821861267089844,
"learning_rate": 1.832669148960057e-05,
"loss": 0.0209,
"step": 14650
},
{
"epoch": 10.124309392265193,
"grad_norm": 0.14552289247512817,
"learning_rate": 1.8262764516553233e-05,
"loss": 0.0197,
"step": 14660
},
{
"epoch": 10.13121546961326,
"grad_norm": 0.24827535450458527,
"learning_rate": 1.8198924311895843e-05,
"loss": 0.0173,
"step": 14670
},
{
"epoch": 10.138121546961326,
"grad_norm": 0.19792486727237701,
"learning_rate": 1.813517105016505e-05,
"loss": 0.0222,
"step": 14680
},
{
"epoch": 10.145027624309392,
"grad_norm": 0.1277136504650116,
"learning_rate": 1.8071504905659888e-05,
"loss": 0.0168,
"step": 14690
},
{
"epoch": 10.151933701657459,
"grad_norm": 0.11872461438179016,
"learning_rate": 1.800792605244109e-05,
"loss": 0.0197,
"step": 14700
},
{
"epoch": 10.158839779005525,
"grad_norm": 0.16362428665161133,
"learning_rate": 1.7944434664330844e-05,
"loss": 0.0286,
"step": 14710
},
{
"epoch": 10.165745856353592,
"grad_norm": 0.1249469742178917,
"learning_rate": 1.7881030914912212e-05,
"loss": 0.0228,
"step": 14720
},
{
"epoch": 10.172651933701658,
"grad_norm": 0.1482723355293274,
"learning_rate": 1.7817714977528577e-05,
"loss": 0.0206,
"step": 14730
},
{
"epoch": 10.179558011049723,
"grad_norm": 0.12862233817577362,
"learning_rate": 1.7754487025283332e-05,
"loss": 0.0246,
"step": 14740
},
{
"epoch": 10.18646408839779,
"grad_norm": 0.14729537069797516,
"learning_rate": 1.7691347231039275e-05,
"loss": 0.0197,
"step": 14750
},
{
"epoch": 10.193370165745856,
"grad_norm": 0.16883131861686707,
"learning_rate": 1.7628295767418164e-05,
"loss": 0.0228,
"step": 14760
},
{
"epoch": 10.200276243093922,
"grad_norm": 0.15366175770759583,
"learning_rate": 1.7565332806800333e-05,
"loss": 0.0232,
"step": 14770
},
{
"epoch": 10.207182320441989,
"grad_norm": 0.15509574115276337,
"learning_rate": 1.750245852132408e-05,
"loss": 0.0203,
"step": 14780
},
{
"epoch": 10.214088397790055,
"grad_norm": 0.14112603664398193,
"learning_rate": 1.7439673082885323e-05,
"loss": 0.019,
"step": 14790
},
{
"epoch": 10.220994475138122,
"grad_norm": 0.09568461030721664,
"learning_rate": 1.7376976663137047e-05,
"loss": 0.0193,
"step": 14800
},
{
"epoch": 10.227900552486188,
"grad_norm": 0.12783320248126984,
"learning_rate": 1.7314369433488853e-05,
"loss": 0.0176,
"step": 14810
},
{
"epoch": 10.234806629834255,
"grad_norm": 0.09211356937885284,
"learning_rate": 1.7251851565106548e-05,
"loss": 0.0177,
"step": 14820
},
{
"epoch": 10.241712707182321,
"grad_norm": 0.15022572875022888,
"learning_rate": 1.7189423228911574e-05,
"loss": 0.0205,
"step": 14830
},
{
"epoch": 10.248618784530386,
"grad_norm": 0.09745365381240845,
"learning_rate": 1.7127084595580606e-05,
"loss": 0.0186,
"step": 14840
},
{
"epoch": 10.255524861878452,
"grad_norm": 0.16210335493087769,
"learning_rate": 1.706483583554513e-05,
"loss": 0.0207,
"step": 14850
},
{
"epoch": 10.262430939226519,
"grad_norm": 0.13680025935173035,
"learning_rate": 1.700267711899083e-05,
"loss": 0.0204,
"step": 14860
},
{
"epoch": 10.269337016574585,
"grad_norm": 0.11462069302797318,
"learning_rate": 1.69406086158573e-05,
"loss": 0.0181,
"step": 14870
},
{
"epoch": 10.276243093922652,
"grad_norm": 0.12634550034999847,
"learning_rate": 1.6878630495837455e-05,
"loss": 0.0207,
"step": 14880
},
{
"epoch": 10.283149171270718,
"grad_norm": 0.16681039333343506,
"learning_rate": 1.681674292837707e-05,
"loss": 0.02,
"step": 14890
},
{
"epoch": 10.290055248618785,
"grad_norm": 0.11668930947780609,
"learning_rate": 1.6754946082674444e-05,
"loss": 0.0203,
"step": 14900
},
{
"epoch": 10.296961325966851,
"grad_norm": 0.17742110788822174,
"learning_rate": 1.6693240127679748e-05,
"loss": 0.018,
"step": 14910
},
{
"epoch": 10.303867403314918,
"grad_norm": 0.13763536512851715,
"learning_rate": 1.663162523209475e-05,
"loss": 0.0254,
"step": 14920
},
{
"epoch": 10.310773480662984,
"grad_norm": 0.07589291781187057,
"learning_rate": 1.6570101564372193e-05,
"loss": 0.0219,
"step": 14930
},
{
"epoch": 10.317679558011049,
"grad_norm": 0.14768086373806,
"learning_rate": 1.650866929271543e-05,
"loss": 0.0206,
"step": 14940
},
{
"epoch": 10.324585635359115,
"grad_norm": 0.1339852213859558,
"learning_rate": 1.644732858507797e-05,
"loss": 0.0208,
"step": 14950
},
{
"epoch": 10.331491712707182,
"grad_norm": 0.15526872873306274,
"learning_rate": 1.6386079609162943e-05,
"loss": 0.0179,
"step": 14960
},
{
"epoch": 10.338397790055248,
"grad_norm": 0.08692698180675507,
"learning_rate": 1.6324922532422742e-05,
"loss": 0.015,
"step": 14970
},
{
"epoch": 10.345303867403315,
"grad_norm": 0.12334548681974411,
"learning_rate": 1.6263857522058434e-05,
"loss": 0.0179,
"step": 14980
},
{
"epoch": 10.352209944751381,
"grad_norm": 0.15830978751182556,
"learning_rate": 1.6202884745019443e-05,
"loss": 0.0196,
"step": 14990
},
{
"epoch": 10.359116022099448,
"grad_norm": 0.14715401828289032,
"learning_rate": 1.614200436800304e-05,
"loss": 0.0188,
"step": 15000
},
{
"epoch": 10.366022099447514,
"grad_norm": 0.13705182075500488,
"learning_rate": 1.6081216557453814e-05,
"loss": 0.0179,
"step": 15010
},
{
"epoch": 10.37292817679558,
"grad_norm": 0.09494274854660034,
"learning_rate": 1.6020521479563367e-05,
"loss": 0.0169,
"step": 15020
},
{
"epoch": 10.379834254143647,
"grad_norm": 0.16117554903030396,
"learning_rate": 1.5959919300269654e-05,
"loss": 0.0225,
"step": 15030
},
{
"epoch": 10.386740331491712,
"grad_norm": 0.12154057621955872,
"learning_rate": 1.5899410185256764e-05,
"loss": 0.0198,
"step": 15040
},
{
"epoch": 10.393646408839778,
"grad_norm": 0.15817032754421234,
"learning_rate": 1.583899429995431e-05,
"loss": 0.0231,
"step": 15050
},
{
"epoch": 10.400552486187845,
"grad_norm": 0.20743706822395325,
"learning_rate": 1.5778671809536993e-05,
"loss": 0.0198,
"step": 15060
},
{
"epoch": 10.407458563535911,
"grad_norm": 0.10531453043222427,
"learning_rate": 1.5718442878924246e-05,
"loss": 0.02,
"step": 15070
},
{
"epoch": 10.414364640883978,
"grad_norm": 0.13929533958435059,
"learning_rate": 1.5658307672779593e-05,
"loss": 0.016,
"step": 15080
},
{
"epoch": 10.421270718232044,
"grad_norm": 0.13578106462955475,
"learning_rate": 1.5598266355510427e-05,
"loss": 0.0199,
"step": 15090
},
{
"epoch": 10.42817679558011,
"grad_norm": 0.19008758664131165,
"learning_rate": 1.553831909126744e-05,
"loss": 0.0187,
"step": 15100
},
{
"epoch": 10.435082872928177,
"grad_norm": 0.14955757558345795,
"learning_rate": 1.5478466043944135e-05,
"loss": 0.0212,
"step": 15110
},
{
"epoch": 10.441988950276244,
"grad_norm": 0.1234457865357399,
"learning_rate": 1.5418707377176468e-05,
"loss": 0.0266,
"step": 15120
},
{
"epoch": 10.44889502762431,
"grad_norm": 0.1302083134651184,
"learning_rate": 1.535904325434233e-05,
"loss": 0.0185,
"step": 15130
},
{
"epoch": 10.455801104972375,
"grad_norm": 0.09352464973926544,
"learning_rate": 1.529947383856118e-05,
"loss": 0.019,
"step": 15140
},
{
"epoch": 10.462707182320441,
"grad_norm": 0.20320549607276917,
"learning_rate": 1.5239999292693524e-05,
"loss": 0.0179,
"step": 15150
},
{
"epoch": 10.469613259668508,
"grad_norm": 0.13011567294597626,
"learning_rate": 1.5180619779340505e-05,
"loss": 0.017,
"step": 15160
},
{
"epoch": 10.476519337016574,
"grad_norm": 0.1492019146680832,
"learning_rate": 1.5121335460843428e-05,
"loss": 0.0189,
"step": 15170
},
{
"epoch": 10.48342541436464,
"grad_norm": 0.12916570901870728,
"learning_rate": 1.5062146499283347e-05,
"loss": 0.0194,
"step": 15180
},
{
"epoch": 10.490331491712707,
"grad_norm": 0.16759514808654785,
"learning_rate": 1.5003053056480643e-05,
"loss": 0.0205,
"step": 15190
},
{
"epoch": 10.497237569060774,
"grad_norm": 0.13533750176429749,
"learning_rate": 1.4944055293994551e-05,
"loss": 0.0231,
"step": 15200
},
{
"epoch": 10.50414364640884,
"grad_norm": 0.14893312752246857,
"learning_rate": 1.4885153373122656e-05,
"loss": 0.0172,
"step": 15210
},
{
"epoch": 10.511049723756907,
"grad_norm": 0.10746175050735474,
"learning_rate": 1.482634745490059e-05,
"loss": 0.0159,
"step": 15220
},
{
"epoch": 10.517955801104973,
"grad_norm": 0.14947925508022308,
"learning_rate": 1.4767637700101466e-05,
"loss": 0.0188,
"step": 15230
},
{
"epoch": 10.524861878453038,
"grad_norm": 0.14596012234687805,
"learning_rate": 1.4709024269235528e-05,
"loss": 0.0204,
"step": 15240
},
{
"epoch": 10.531767955801104,
"grad_norm": 0.11377096176147461,
"learning_rate": 1.4650507322549684e-05,
"loss": 0.0213,
"step": 15250
},
{
"epoch": 10.53867403314917,
"grad_norm": 0.2057703733444214,
"learning_rate": 1.4592087020026972e-05,
"loss": 0.0188,
"step": 15260
},
{
"epoch": 10.545580110497237,
"grad_norm": 0.11566067487001419,
"learning_rate": 1.4533763521386318e-05,
"loss": 0.0197,
"step": 15270
},
{
"epoch": 10.552486187845304,
"grad_norm": 0.15610185265541077,
"learning_rate": 1.44755369860819e-05,
"loss": 0.0197,
"step": 15280
},
{
"epoch": 10.55939226519337,
"grad_norm": 0.11242447793483734,
"learning_rate": 1.441740757330287e-05,
"loss": 0.0202,
"step": 15290
},
{
"epoch": 10.566298342541437,
"grad_norm": 0.18600213527679443,
"learning_rate": 1.4359375441972844e-05,
"loss": 0.0233,
"step": 15300
},
{
"epoch": 10.573204419889503,
"grad_norm": 0.19395959377288818,
"learning_rate": 1.4301440750749395e-05,
"loss": 0.0213,
"step": 15310
},
{
"epoch": 10.58011049723757,
"grad_norm": 0.14071722328662872,
"learning_rate": 1.4243603658023808e-05,
"loss": 0.024,
"step": 15320
},
{
"epoch": 10.587016574585636,
"grad_norm": 0.16488364338874817,
"learning_rate": 1.4185864321920444e-05,
"loss": 0.026,
"step": 15330
},
{
"epoch": 10.5939226519337,
"grad_norm": 0.1464218646287918,
"learning_rate": 1.4128222900296485e-05,
"loss": 0.0211,
"step": 15340
},
{
"epoch": 10.600828729281767,
"grad_norm": 0.15647467970848083,
"learning_rate": 1.407067955074135e-05,
"loss": 0.0197,
"step": 15350
},
{
"epoch": 10.607734806629834,
"grad_norm": 0.11905504763126373,
"learning_rate": 1.4013234430576356e-05,
"loss": 0.0173,
"step": 15360
},
{
"epoch": 10.6146408839779,
"grad_norm": 0.09630614519119263,
"learning_rate": 1.3955887696854286e-05,
"loss": 0.0178,
"step": 15370
},
{
"epoch": 10.621546961325967,
"grad_norm": 0.15738511085510254,
"learning_rate": 1.38986395063589e-05,
"loss": 0.0189,
"step": 15380
},
{
"epoch": 10.628453038674033,
"grad_norm": 0.07597813010215759,
"learning_rate": 1.3841490015604597e-05,
"loss": 0.0179,
"step": 15390
},
{
"epoch": 10.6353591160221,
"grad_norm": 0.15570463240146637,
"learning_rate": 1.3784439380835879e-05,
"loss": 0.0257,
"step": 15400
},
{
"epoch": 10.642265193370166,
"grad_norm": 0.09338031709194183,
"learning_rate": 1.3727487758026986e-05,
"loss": 0.02,
"step": 15410
},
{
"epoch": 10.649171270718233,
"grad_norm": 0.14082670211791992,
"learning_rate": 1.3670635302881525e-05,
"loss": 0.0232,
"step": 15420
},
{
"epoch": 10.6560773480663,
"grad_norm": 0.13374818861484528,
"learning_rate": 1.3613882170831888e-05,
"loss": 0.0169,
"step": 15430
},
{
"epoch": 10.662983425414364,
"grad_norm": 0.13222579658031464,
"learning_rate": 1.355722851703901e-05,
"loss": 0.0191,
"step": 15440
},
{
"epoch": 10.66988950276243,
"grad_norm": 0.12320950627326965,
"learning_rate": 1.3500674496391814e-05,
"loss": 0.0195,
"step": 15450
},
{
"epoch": 10.676795580110497,
"grad_norm": 0.09267281740903854,
"learning_rate": 1.3444220263506795e-05,
"loss": 0.018,
"step": 15460
},
{
"epoch": 10.683701657458563,
"grad_norm": 0.14765200018882751,
"learning_rate": 1.3387865972727714e-05,
"loss": 0.0206,
"step": 15470
},
{
"epoch": 10.69060773480663,
"grad_norm": 0.14930613338947296,
"learning_rate": 1.3331611778125036e-05,
"loss": 0.0188,
"step": 15480
},
{
"epoch": 10.697513812154696,
"grad_norm": 0.1355215162038803,
"learning_rate": 1.3275457833495564e-05,
"loss": 0.0217,
"step": 15490
},
{
"epoch": 10.704419889502763,
"grad_norm": 0.13247044384479523,
"learning_rate": 1.3219404292362065e-05,
"loss": 0.0172,
"step": 15500
},
{
"epoch": 10.71132596685083,
"grad_norm": 0.14117850363254547,
"learning_rate": 1.3163451307972751e-05,
"loss": 0.0184,
"step": 15510
},
{
"epoch": 10.718232044198896,
"grad_norm": 0.12379027158021927,
"learning_rate": 1.3107599033300977e-05,
"loss": 0.0191,
"step": 15520
},
{
"epoch": 10.725138121546962,
"grad_norm": 0.19423779845237732,
"learning_rate": 1.305184762104471e-05,
"loss": 0.0259,
"step": 15530
},
{
"epoch": 10.732044198895027,
"grad_norm": 0.14055989682674408,
"learning_rate": 1.2996197223626178e-05,
"loss": 0.0175,
"step": 15540
},
{
"epoch": 10.738950276243093,
"grad_norm": 0.11190267652273178,
"learning_rate": 1.2940647993191457e-05,
"loss": 0.019,
"step": 15550
},
{
"epoch": 10.74585635359116,
"grad_norm": 0.14761070907115936,
"learning_rate": 1.2885200081610005e-05,
"loss": 0.0192,
"step": 15560
},
{
"epoch": 10.752762430939226,
"grad_norm": 0.136946901679039,
"learning_rate": 1.2829853640474316e-05,
"loss": 0.0189,
"step": 15570
},
{
"epoch": 10.759668508287293,
"grad_norm": 0.13317197561264038,
"learning_rate": 1.2774608821099438e-05,
"loss": 0.0192,
"step": 15580
},
{
"epoch": 10.76657458563536,
"grad_norm": 0.1472024768590927,
"learning_rate": 1.2719465774522577e-05,
"loss": 0.0206,
"step": 15590
},
{
"epoch": 10.773480662983426,
"grad_norm": 0.15507635474205017,
"learning_rate": 1.2664424651502755e-05,
"loss": 0.0185,
"step": 15600
},
{
"epoch": 10.780386740331492,
"grad_norm": 0.1318316012620926,
"learning_rate": 1.260948560252026e-05,
"loss": 0.0171,
"step": 15610
},
{
"epoch": 10.787292817679559,
"grad_norm": 0.11615798622369766,
"learning_rate": 1.2554648777776396e-05,
"loss": 0.0209,
"step": 15620
},
{
"epoch": 10.794198895027625,
"grad_norm": 0.16774071753025055,
"learning_rate": 1.2499914327192919e-05,
"loss": 0.0166,
"step": 15630
},
{
"epoch": 10.80110497237569,
"grad_norm": 0.21441291272640228,
"learning_rate": 1.2445282400411722e-05,
"loss": 0.0197,
"step": 15640
},
{
"epoch": 10.808011049723756,
"grad_norm": 0.11902355402708054,
"learning_rate": 1.2390753146794437e-05,
"loss": 0.0225,
"step": 15650
},
{
"epoch": 10.814917127071823,
"grad_norm": 0.14514003694057465,
"learning_rate": 1.2336326715421925e-05,
"loss": 0.0187,
"step": 15660
},
{
"epoch": 10.82182320441989,
"grad_norm": 0.1624397486448288,
"learning_rate": 1.2282003255094005e-05,
"loss": 0.0223,
"step": 15670
},
{
"epoch": 10.828729281767956,
"grad_norm": 0.12542851269245148,
"learning_rate": 1.2227782914328928e-05,
"loss": 0.0193,
"step": 15680
},
{
"epoch": 10.835635359116022,
"grad_norm": 0.13802413642406464,
"learning_rate": 1.2173665841363018e-05,
"loss": 0.0179,
"step": 15690
},
{
"epoch": 10.842541436464089,
"grad_norm": 0.0999394953250885,
"learning_rate": 1.211965218415032e-05,
"loss": 0.0182,
"step": 15700
},
{
"epoch": 10.849447513812155,
"grad_norm": 0.1239611878991127,
"learning_rate": 1.2065742090362082e-05,
"loss": 0.022,
"step": 15710
},
{
"epoch": 10.856353591160222,
"grad_norm": 0.11930327117443085,
"learning_rate": 1.2011935707386457e-05,
"loss": 0.017,
"step": 15720
},
{
"epoch": 10.863259668508288,
"grad_norm": 0.12612630426883698,
"learning_rate": 1.1958233182328044e-05,
"loss": 0.0177,
"step": 15730
},
{
"epoch": 10.870165745856355,
"grad_norm": 0.16491909325122833,
"learning_rate": 1.1904634662007474e-05,
"loss": 0.0216,
"step": 15740
},
{
"epoch": 10.87707182320442,
"grad_norm": 0.141819030046463,
"learning_rate": 1.1851140292961088e-05,
"loss": 0.0179,
"step": 15750
},
{
"epoch": 10.883977900552486,
"grad_norm": 0.12252664566040039,
"learning_rate": 1.1797750221440424e-05,
"loss": 0.0189,
"step": 15760
},
{
"epoch": 10.890883977900552,
"grad_norm": 0.1347576379776001,
"learning_rate": 1.1744464593411897e-05,
"loss": 0.0229,
"step": 15770
},
{
"epoch": 10.897790055248619,
"grad_norm": 0.0848279520869255,
"learning_rate": 1.1691283554556399e-05,
"loss": 0.0179,
"step": 15780
},
{
"epoch": 10.904696132596685,
"grad_norm": 0.15215791761875153,
"learning_rate": 1.1638207250268834e-05,
"loss": 0.0205,
"step": 15790
},
{
"epoch": 10.911602209944752,
"grad_norm": 0.1396626979112625,
"learning_rate": 1.158523582565782e-05,
"loss": 0.0206,
"step": 15800
},
{
"epoch": 10.918508287292818,
"grad_norm": 0.1636439710855484,
"learning_rate": 1.1532369425545192e-05,
"loss": 0.0202,
"step": 15810
},
{
"epoch": 10.925414364640885,
"grad_norm": 0.13826753199100494,
"learning_rate": 1.1479608194465662e-05,
"loss": 0.0223,
"step": 15820
},
{
"epoch": 10.932320441988951,
"grad_norm": 0.14061783254146576,
"learning_rate": 1.1426952276666442e-05,
"loss": 0.0193,
"step": 15830
},
{
"epoch": 10.939226519337016,
"grad_norm": 0.08424236625432968,
"learning_rate": 1.1374401816106778e-05,
"loss": 0.0205,
"step": 15840
},
{
"epoch": 10.946132596685082,
"grad_norm": 0.1465301215648651,
"learning_rate": 1.1321956956457646e-05,
"loss": 0.0242,
"step": 15850
},
{
"epoch": 10.953038674033149,
"grad_norm": 0.10818585008382797,
"learning_rate": 1.1269617841101277e-05,
"loss": 0.0158,
"step": 15860
},
{
"epoch": 10.959944751381215,
"grad_norm": 0.11729062348604202,
"learning_rate": 1.1217384613130804e-05,
"loss": 0.0194,
"step": 15870
},
{
"epoch": 10.966850828729282,
"grad_norm": 0.13734790682792664,
"learning_rate": 1.11652574153499e-05,
"loss": 0.0211,
"step": 15880
},
{
"epoch": 10.973756906077348,
"grad_norm": 0.10207629203796387,
"learning_rate": 1.1113236390272303e-05,
"loss": 0.0191,
"step": 15890
},
{
"epoch": 10.980662983425415,
"grad_norm": 0.09973394870758057,
"learning_rate": 1.106132168012155e-05,
"loss": 0.0218,
"step": 15900
},
{
"epoch": 10.987569060773481,
"grad_norm": 0.178752601146698,
"learning_rate": 1.1009513426830448e-05,
"loss": 0.0224,
"step": 15910
},
{
"epoch": 10.994475138121548,
"grad_norm": 0.1685081571340561,
"learning_rate": 1.0957811772040777e-05,
"loss": 0.0186,
"step": 15920
},
{
"epoch": 11.001381215469614,
"grad_norm": 0.17715397477149963,
"learning_rate": 1.0906216857102913e-05,
"loss": 0.0268,
"step": 15930
},
{
"epoch": 11.008287292817679,
"grad_norm": 0.1387261301279068,
"learning_rate": 1.0854728823075355e-05,
"loss": 0.017,
"step": 15940
},
{
"epoch": 11.015193370165745,
"grad_norm": 0.15848971903324127,
"learning_rate": 1.0803347810724452e-05,
"loss": 0.0198,
"step": 15950
},
{
"epoch": 11.022099447513812,
"grad_norm": 0.16126196086406708,
"learning_rate": 1.0752073960523911e-05,
"loss": 0.0239,
"step": 15960
},
{
"epoch": 11.029005524861878,
"grad_norm": 0.1097811833024025,
"learning_rate": 1.070090741265447e-05,
"loss": 0.019,
"step": 15970
},
{
"epoch": 11.035911602209945,
"grad_norm": 0.10773305594921112,
"learning_rate": 1.0649848307003547e-05,
"loss": 0.0189,
"step": 15980
},
{
"epoch": 11.042817679558011,
"grad_norm": 0.13943029940128326,
"learning_rate": 1.0598896783164757e-05,
"loss": 0.0205,
"step": 15990
},
{
"epoch": 11.049723756906078,
"grad_norm": 0.16361400485038757,
"learning_rate": 1.0548052980437645e-05,
"loss": 0.0186,
"step": 16000
},
{
"epoch": 11.056629834254144,
"grad_norm": 0.18143020570278168,
"learning_rate": 1.049731703782722e-05,
"loss": 0.0164,
"step": 16010
},
{
"epoch": 11.06353591160221,
"grad_norm": 0.11211825162172318,
"learning_rate": 1.0446689094043587e-05,
"loss": 0.0148,
"step": 16020
},
{
"epoch": 11.070441988950277,
"grad_norm": 0.11881590634584427,
"learning_rate": 1.039616928750165e-05,
"loss": 0.0184,
"step": 16030
},
{
"epoch": 11.077348066298342,
"grad_norm": 0.19520686566829681,
"learning_rate": 1.0345757756320612e-05,
"loss": 0.0205,
"step": 16040
},
{
"epoch": 11.084254143646408,
"grad_norm": 0.1139756515622139,
"learning_rate": 1.0295454638323666e-05,
"loss": 0.019,
"step": 16050
},
{
"epoch": 11.091160220994475,
"grad_norm": 0.15939398109912872,
"learning_rate": 1.0245260071037632e-05,
"loss": 0.0199,
"step": 16060
},
{
"epoch": 11.098066298342541,
"grad_norm": 0.12053225189447403,
"learning_rate": 1.0195174191692518e-05,
"loss": 0.016,
"step": 16070
},
{
"epoch": 11.104972375690608,
"grad_norm": 0.1569368690252304,
"learning_rate": 1.014519713722124e-05,
"loss": 0.0212,
"step": 16080
},
{
"epoch": 11.111878453038674,
"grad_norm": 0.12836386263370514,
"learning_rate": 1.0095329044259132e-05,
"loss": 0.0215,
"step": 16090
},
{
"epoch": 11.11878453038674,
"grad_norm": 0.23310133814811707,
"learning_rate": 1.004557004914365e-05,
"loss": 0.0205,
"step": 16100
},
{
"epoch": 11.125690607734807,
"grad_norm": 0.1868610829114914,
"learning_rate": 9.995920287914007e-06,
"loss": 0.0196,
"step": 16110
},
{
"epoch": 11.132596685082873,
"grad_norm": 0.17721503973007202,
"learning_rate": 9.946379896310737e-06,
"loss": 0.0188,
"step": 16120
},
{
"epoch": 11.13950276243094,
"grad_norm": 0.14387953281402588,
"learning_rate": 9.896949009775396e-06,
"loss": 0.0185,
"step": 16130
},
{
"epoch": 11.146408839779005,
"grad_norm": 0.12724824249744415,
"learning_rate": 9.847627763450134e-06,
"loss": 0.0178,
"step": 16140
},
{
"epoch": 11.153314917127071,
"grad_norm": 0.13191691040992737,
"learning_rate": 9.798416292177337e-06,
"loss": 0.0178,
"step": 16150
},
{
"epoch": 11.160220994475138,
"grad_norm": 0.1462077796459198,
"learning_rate": 9.74931473049932e-06,
"loss": 0.0232,
"step": 16160
},
{
"epoch": 11.167127071823204,
"grad_norm": 0.12221027910709381,
"learning_rate": 9.700323212657847e-06,
"loss": 0.0185,
"step": 16170
},
{
"epoch": 11.17403314917127,
"grad_norm": 0.12442642450332642,
"learning_rate": 9.65144187259388e-06,
"loss": 0.0228,
"step": 16180
},
{
"epoch": 11.180939226519337,
"grad_norm": 0.16251784563064575,
"learning_rate": 9.602670843947132e-06,
"loss": 0.0192,
"step": 16190
},
{
"epoch": 11.187845303867404,
"grad_norm": 0.16168147325515747,
"learning_rate": 9.554010260055713e-06,
"loss": 0.0166,
"step": 16200
},
{
"epoch": 11.19475138121547,
"grad_norm": 0.13379383087158203,
"learning_rate": 9.505460253955834e-06,
"loss": 0.0158,
"step": 16210
},
{
"epoch": 11.201657458563536,
"grad_norm": 0.15277086198329926,
"learning_rate": 9.457020958381324e-06,
"loss": 0.018,
"step": 16220
},
{
"epoch": 11.208563535911603,
"grad_norm": 0.16304939985275269,
"learning_rate": 9.408692505763395e-06,
"loss": 0.0184,
"step": 16230
},
{
"epoch": 11.215469613259668,
"grad_norm": 0.16197776794433594,
"learning_rate": 9.360475028230181e-06,
"loss": 0.0188,
"step": 16240
},
{
"epoch": 11.222375690607734,
"grad_norm": 0.12607796490192413,
"learning_rate": 9.312368657606412e-06,
"loss": 0.0198,
"step": 16250
},
{
"epoch": 11.2292817679558,
"grad_norm": 0.14476558566093445,
"learning_rate": 9.264373525413096e-06,
"loss": 0.0201,
"step": 16260
},
{
"epoch": 11.236187845303867,
"grad_norm": 0.13373053073883057,
"learning_rate": 9.216489762867058e-06,
"loss": 0.0225,
"step": 16270
},
{
"epoch": 11.243093922651934,
"grad_norm": 0.10646121948957443,
"learning_rate": 9.168717500880708e-06,
"loss": 0.0126,
"step": 16280
},
{
"epoch": 11.25,
"grad_norm": 0.10445374995470047,
"learning_rate": 9.121056870061574e-06,
"loss": 0.0242,
"step": 16290
},
{
"epoch": 11.256906077348066,
"grad_norm": 0.13283097743988037,
"learning_rate": 9.073508000711983e-06,
"loss": 0.0163,
"step": 16300
},
{
"epoch": 11.263812154696133,
"grad_norm": 0.15861085057258606,
"learning_rate": 9.026071022828758e-06,
"loss": 0.0208,
"step": 16310
},
{
"epoch": 11.2707182320442,
"grad_norm": 0.14492949843406677,
"learning_rate": 8.978746066102771e-06,
"loss": 0.0201,
"step": 16320
},
{
"epoch": 11.277624309392266,
"grad_norm": 0.16569550335407257,
"learning_rate": 8.931533259918634e-06,
"loss": 0.0172,
"step": 16330
},
{
"epoch": 11.284530386740332,
"grad_norm": 0.11864589899778366,
"learning_rate": 8.884432733354382e-06,
"loss": 0.021,
"step": 16340
},
{
"epoch": 11.291436464088397,
"grad_norm": 0.13216151297092438,
"learning_rate": 8.837444615181029e-06,
"loss": 0.0171,
"step": 16350
},
{
"epoch": 11.298342541436464,
"grad_norm": 0.12307021021842957,
"learning_rate": 8.790569033862323e-06,
"loss": 0.0203,
"step": 16360
},
{
"epoch": 11.30524861878453,
"grad_norm": 0.09344703704118729,
"learning_rate": 8.7438061175543e-06,
"loss": 0.0186,
"step": 16370
},
{
"epoch": 11.312154696132596,
"grad_norm": 0.12735137343406677,
"learning_rate": 8.697155994104978e-06,
"loss": 0.0203,
"step": 16380
},
{
"epoch": 11.319060773480663,
"grad_norm": 0.12188874930143356,
"learning_rate": 8.650618791054033e-06,
"loss": 0.0208,
"step": 16390
},
{
"epoch": 11.32596685082873,
"grad_norm": 0.12098925560712814,
"learning_rate": 8.604194635632373e-06,
"loss": 0.0204,
"step": 16400
},
{
"epoch": 11.332872928176796,
"grad_norm": 0.13109688460826874,
"learning_rate": 8.557883654761906e-06,
"loss": 0.0196,
"step": 16410
},
{
"epoch": 11.339779005524862,
"grad_norm": 0.1311589777469635,
"learning_rate": 8.511685975055061e-06,
"loss": 0.0162,
"step": 16420
},
{
"epoch": 11.346685082872929,
"grad_norm": 0.09084175527095795,
"learning_rate": 8.46560172281452e-06,
"loss": 0.0147,
"step": 16430
},
{
"epoch": 11.353591160220994,
"grad_norm": 0.12907159328460693,
"learning_rate": 8.419631024032893e-06,
"loss": 0.0167,
"step": 16440
},
{
"epoch": 11.36049723756906,
"grad_norm": 0.10988034307956696,
"learning_rate": 8.373774004392293e-06,
"loss": 0.0174,
"step": 16450
},
{
"epoch": 11.367403314917127,
"grad_norm": 0.1319379210472107,
"learning_rate": 8.32803078926409e-06,
"loss": 0.0158,
"step": 16460
},
{
"epoch": 11.374309392265193,
"grad_norm": 0.11994582414627075,
"learning_rate": 8.282401503708454e-06,
"loss": 0.0187,
"step": 16470
},
{
"epoch": 11.38121546961326,
"grad_norm": 0.14029355347156525,
"learning_rate": 8.23688627247412e-06,
"loss": 0.0216,
"step": 16480
},
{
"epoch": 11.388121546961326,
"grad_norm": 0.16516615450382233,
"learning_rate": 8.191485219998007e-06,
"loss": 0.021,
"step": 16490
},
{
"epoch": 11.395027624309392,
"grad_norm": 0.1420423984527588,
"learning_rate": 8.146198470404843e-06,
"loss": 0.0181,
"step": 16500
},
{
"epoch": 11.401933701657459,
"grad_norm": 0.19425787031650543,
"learning_rate": 8.101026147506897e-06,
"loss": 0.0165,
"step": 16510
},
{
"epoch": 11.408839779005525,
"grad_norm": 0.10360033065080643,
"learning_rate": 8.05596837480353e-06,
"loss": 0.0187,
"step": 16520
},
{
"epoch": 11.415745856353592,
"grad_norm": 0.06478369981050491,
"learning_rate": 8.011025275480998e-06,
"loss": 0.0188,
"step": 16530
},
{
"epoch": 11.422651933701658,
"grad_norm": 0.170049786567688,
"learning_rate": 7.966196972412027e-06,
"loss": 0.0169,
"step": 16540
},
{
"epoch": 11.429558011049723,
"grad_norm": 0.11986824870109558,
"learning_rate": 7.92148358815547e-06,
"loss": 0.0199,
"step": 16550
},
{
"epoch": 11.43646408839779,
"grad_norm": 0.29806897044181824,
"learning_rate": 7.87688524495604e-06,
"loss": 0.0188,
"step": 16560
},
{
"epoch": 11.443370165745856,
"grad_norm": 0.1958770602941513,
"learning_rate": 7.83240206474386e-06,
"loss": 0.0198,
"step": 16570
},
{
"epoch": 11.450276243093922,
"grad_norm": 0.13644862174987793,
"learning_rate": 7.788034169134272e-06,
"loss": 0.0212,
"step": 16580
},
{
"epoch": 11.457182320441989,
"grad_norm": 0.11058889329433441,
"learning_rate": 7.743781679427414e-06,
"loss": 0.0193,
"step": 16590
},
{
"epoch": 11.464088397790055,
"grad_norm": 0.09904930740594864,
"learning_rate": 7.699644716607895e-06,
"loss": 0.0179,
"step": 16600
},
{
"epoch": 11.470994475138122,
"grad_norm": 0.09778125584125519,
"learning_rate": 7.655623401344486e-06,
"loss": 0.0195,
"step": 16610
},
{
"epoch": 11.477900552486188,
"grad_norm": 0.1029466763138771,
"learning_rate": 7.611717853989775e-06,
"loss": 0.018,
"step": 16620
},
{
"epoch": 11.484806629834255,
"grad_norm": 0.08880892395973206,
"learning_rate": 7.567928194579854e-06,
"loss": 0.0184,
"step": 16630
},
{
"epoch": 11.491712707182321,
"grad_norm": 0.11798331141471863,
"learning_rate": 7.524254542833997e-06,
"loss": 0.0174,
"step": 16640
},
{
"epoch": 11.498618784530386,
"grad_norm": 0.19138731062412262,
"learning_rate": 7.480697018154286e-06,
"loss": 0.0208,
"step": 16650
},
{
"epoch": 11.505524861878452,
"grad_norm": 0.11023187637329102,
"learning_rate": 7.437255739625332e-06,
"loss": 0.0205,
"step": 16660
},
{
"epoch": 11.512430939226519,
"grad_norm": 0.13642142713069916,
"learning_rate": 7.393930826013923e-06,
"loss": 0.0174,
"step": 16670
},
{
"epoch": 11.519337016574585,
"grad_norm": 0.1239386722445488,
"learning_rate": 7.350722395768722e-06,
"loss": 0.019,
"step": 16680
},
{
"epoch": 11.526243093922652,
"grad_norm": 0.13160714507102966,
"learning_rate": 7.307630567019963e-06,
"loss": 0.017,
"step": 16690
},
{
"epoch": 11.533149171270718,
"grad_norm": 0.1242450550198555,
"learning_rate": 7.264655457579e-06,
"loss": 0.0186,
"step": 16700
},
{
"epoch": 11.540055248618785,
"grad_norm": 0.12114911526441574,
"learning_rate": 7.221797184938184e-06,
"loss": 0.0228,
"step": 16710
},
{
"epoch": 11.546961325966851,
"grad_norm": 0.1373281180858612,
"learning_rate": 7.179055866270373e-06,
"loss": 0.0142,
"step": 16720
},
{
"epoch": 11.553867403314918,
"grad_norm": 0.1476965695619583,
"learning_rate": 7.136431618428707e-06,
"loss": 0.0238,
"step": 16730
},
{
"epoch": 11.560773480662984,
"grad_norm": 0.15891554951667786,
"learning_rate": 7.09392455794628e-06,
"loss": 0.0149,
"step": 16740
},
{
"epoch": 11.567679558011049,
"grad_norm": 0.11660894006490707,
"learning_rate": 7.051534801035725e-06,
"loss": 0.0148,
"step": 16750
},
{
"epoch": 11.574585635359115,
"grad_norm": 0.14604243636131287,
"learning_rate": 7.00926246358905e-06,
"loss": 0.0178,
"step": 16760
},
{
"epoch": 11.581491712707182,
"grad_norm": 0.11174402385950089,
"learning_rate": 6.967107661177191e-06,
"loss": 0.022,
"step": 16770
},
{
"epoch": 11.588397790055248,
"grad_norm": 0.18665166199207306,
"learning_rate": 6.925070509049786e-06,
"loss": 0.0176,
"step": 16780
},
{
"epoch": 11.595303867403315,
"grad_norm": 0.16662661731243134,
"learning_rate": 6.883151122134812e-06,
"loss": 0.0211,
"step": 16790
},
{
"epoch": 11.602209944751381,
"grad_norm": 0.13698987662792206,
"learning_rate": 6.8413496150382394e-06,
"loss": 0.0186,
"step": 16800
},
{
"epoch": 11.609116022099448,
"grad_norm": 0.14518332481384277,
"learning_rate": 6.7996661020438165e-06,
"loss": 0.019,
"step": 16810
},
{
"epoch": 11.616022099447514,
"grad_norm": 0.12339243292808533,
"learning_rate": 6.758100697112662e-06,
"loss": 0.0187,
"step": 16820
},
{
"epoch": 11.62292817679558,
"grad_norm": 0.16226792335510254,
"learning_rate": 6.716653513883026e-06,
"loss": 0.0182,
"step": 16830
},
{
"epoch": 11.629834254143645,
"grad_norm": 0.11799878627061844,
"learning_rate": 6.675324665669913e-06,
"loss": 0.0193,
"step": 16840
},
{
"epoch": 11.636740331491712,
"grad_norm": 0.11323769390583038,
"learning_rate": 6.634114265464803e-06,
"loss": 0.0205,
"step": 16850
},
{
"epoch": 11.643646408839778,
"grad_norm": 0.15336430072784424,
"learning_rate": 6.59302242593538e-06,
"loss": 0.018,
"step": 16860
},
{
"epoch": 11.650552486187845,
"grad_norm": 0.13874183595180511,
"learning_rate": 6.552049259425141e-06,
"loss": 0.0149,
"step": 16870
},
{
"epoch": 11.657458563535911,
"grad_norm": 0.12732809782028198,
"learning_rate": 6.511194877953181e-06,
"loss": 0.0209,
"step": 16880
},
{
"epoch": 11.664364640883978,
"grad_norm": 0.08383426815271378,
"learning_rate": 6.470459393213813e-06,
"loss": 0.0184,
"step": 16890
},
{
"epoch": 11.671270718232044,
"grad_norm": 0.10465892404317856,
"learning_rate": 6.429842916576279e-06,
"loss": 0.0255,
"step": 16900
},
{
"epoch": 11.67817679558011,
"grad_norm": 0.1163114532828331,
"learning_rate": 6.389345559084503e-06,
"loss": 0.0213,
"step": 16910
},
{
"epoch": 11.685082872928177,
"grad_norm": 0.13213935494422913,
"learning_rate": 6.348967431456682e-06,
"loss": 0.0184,
"step": 16920
},
{
"epoch": 11.691988950276244,
"grad_norm": 0.1505950689315796,
"learning_rate": 6.30870864408511e-06,
"loss": 0.0174,
"step": 16930
},
{
"epoch": 11.69889502762431,
"grad_norm": 0.18978534638881683,
"learning_rate": 6.268569307035754e-06,
"loss": 0.0193,
"step": 16940
},
{
"epoch": 11.705801104972375,
"grad_norm": 0.15616071224212646,
"learning_rate": 6.228549530048022e-06,
"loss": 0.0186,
"step": 16950
},
{
"epoch": 11.712707182320441,
"grad_norm": 0.12579162418842316,
"learning_rate": 6.1886494225344814e-06,
"loss": 0.0192,
"step": 16960
},
{
"epoch": 11.719613259668508,
"grad_norm": 0.11010239273309708,
"learning_rate": 6.148869093580479e-06,
"loss": 0.0213,
"step": 16970
},
{
"epoch": 11.726519337016574,
"grad_norm": 0.11935733258724213,
"learning_rate": 6.109208651943921e-06,
"loss": 0.0162,
"step": 16980
},
{
"epoch": 11.73342541436464,
"grad_norm": 0.1642450988292694,
"learning_rate": 6.069668206054946e-06,
"loss": 0.0209,
"step": 16990
},
{
"epoch": 11.740331491712707,
"grad_norm": 0.20178638398647308,
"learning_rate": 6.0302478640156145e-06,
"loss": 0.0217,
"step": 17000
},
{
"epoch": 11.747237569060774,
"grad_norm": 0.16644486784934998,
"learning_rate": 5.990947733599644e-06,
"loss": 0.0204,
"step": 17010
},
{
"epoch": 11.75414364640884,
"grad_norm": 0.15117019414901733,
"learning_rate": 5.951767922252105e-06,
"loss": 0.0184,
"step": 17020
},
{
"epoch": 11.761049723756907,
"grad_norm": 0.07056622207164764,
"learning_rate": 5.912708537089068e-06,
"loss": 0.0201,
"step": 17030
},
{
"epoch": 11.767955801104973,
"grad_norm": 0.10951358079910278,
"learning_rate": 5.873769684897434e-06,
"loss": 0.0167,
"step": 17040
},
{
"epoch": 11.774861878453038,
"grad_norm": 0.13333486020565033,
"learning_rate": 5.834951472134514e-06,
"loss": 0.0174,
"step": 17050
},
{
"epoch": 11.781767955801104,
"grad_norm": 0.20654167234897614,
"learning_rate": 5.796254004927832e-06,
"loss": 0.0205,
"step": 17060
},
{
"epoch": 11.78867403314917,
"grad_norm": 0.19155727326869965,
"learning_rate": 5.757677389074806e-06,
"loss": 0.0158,
"step": 17070
},
{
"epoch": 11.795580110497237,
"grad_norm": 0.1749805510044098,
"learning_rate": 5.719221730042385e-06,
"loss": 0.0256,
"step": 17080
},
{
"epoch": 11.802486187845304,
"grad_norm": 0.07431744039058685,
"learning_rate": 5.680887132966911e-06,
"loss": 0.0166,
"step": 17090
},
{
"epoch": 11.80939226519337,
"grad_norm": 0.10040867328643799,
"learning_rate": 5.642673702653683e-06,
"loss": 0.0192,
"step": 17100
},
{
"epoch": 11.816298342541437,
"grad_norm": 0.14079715311527252,
"learning_rate": 5.604581543576781e-06,
"loss": 0.0182,
"step": 17110
},
{
"epoch": 11.823204419889503,
"grad_norm": 0.19855310022830963,
"learning_rate": 5.566610759878704e-06,
"loss": 0.0231,
"step": 17120
},
{
"epoch": 11.83011049723757,
"grad_norm": 0.17754405736923218,
"learning_rate": 5.528761455370119e-06,
"loss": 0.0214,
"step": 17130
},
{
"epoch": 11.837016574585636,
"grad_norm": 0.1382066309452057,
"learning_rate": 5.491033733529594e-06,
"loss": 0.0272,
"step": 17140
},
{
"epoch": 11.8439226519337,
"grad_norm": 0.16806116700172424,
"learning_rate": 5.453427697503255e-06,
"loss": 0.0166,
"step": 17150
},
{
"epoch": 11.850828729281767,
"grad_norm": 0.14684727787971497,
"learning_rate": 5.415943450104599e-06,
"loss": 0.0184,
"step": 17160
},
{
"epoch": 11.857734806629834,
"grad_norm": 0.1558583527803421,
"learning_rate": 5.378581093814111e-06,
"loss": 0.0196,
"step": 17170
},
{
"epoch": 11.8646408839779,
"grad_norm": 0.2262316346168518,
"learning_rate": 5.3413407307790375e-06,
"loss": 0.0165,
"step": 17180
},
{
"epoch": 11.871546961325967,
"grad_norm": 0.11336057633161545,
"learning_rate": 5.30422246281313e-06,
"loss": 0.02,
"step": 17190
},
{
"epoch": 11.878453038674033,
"grad_norm": 0.17291094362735748,
"learning_rate": 5.267226391396296e-06,
"loss": 0.0171,
"step": 17200
},
{
"epoch": 11.8853591160221,
"grad_norm": 0.25549980998039246,
"learning_rate": 5.2303526176744e-06,
"loss": 0.0236,
"step": 17210
},
{
"epoch": 11.892265193370166,
"grad_norm": 0.16734923422336578,
"learning_rate": 5.193601242458929e-06,
"loss": 0.0178,
"step": 17220
},
{
"epoch": 11.899171270718233,
"grad_norm": 0.11904910951852798,
"learning_rate": 5.156972366226714e-06,
"loss": 0.0167,
"step": 17230
},
{
"epoch": 11.9060773480663,
"grad_norm": 0.12467493116855621,
"learning_rate": 5.120466089119735e-06,
"loss": 0.0164,
"step": 17240
},
{
"epoch": 11.912983425414364,
"grad_norm": 0.08813518285751343,
"learning_rate": 5.084082510944749e-06,
"loss": 0.0249,
"step": 17250
},
{
"epoch": 11.91988950276243,
"grad_norm": 0.11498162150382996,
"learning_rate": 5.047821731173058e-06,
"loss": 0.0175,
"step": 17260
},
{
"epoch": 11.926795580110497,
"grad_norm": 0.14780119061470032,
"learning_rate": 5.011683848940274e-06,
"loss": 0.0188,
"step": 17270
},
{
"epoch": 11.933701657458563,
"grad_norm": 0.15292783081531525,
"learning_rate": 4.975668963045954e-06,
"loss": 0.0187,
"step": 17280
},
{
"epoch": 11.94060773480663,
"grad_norm": 0.1259051412343979,
"learning_rate": 4.9397771719534525e-06,
"loss": 0.0245,
"step": 17290
},
{
"epoch": 11.947513812154696,
"grad_norm": 0.11878091841936111,
"learning_rate": 4.904008573789548e-06,
"loss": 0.0193,
"step": 17300
},
{
"epoch": 11.954419889502763,
"grad_norm": 0.15138937532901764,
"learning_rate": 4.8683632663442005e-06,
"loss": 0.0217,
"step": 17310
},
{
"epoch": 11.96132596685083,
"grad_norm": 0.12729822099208832,
"learning_rate": 4.832841347070343e-06,
"loss": 0.0194,
"step": 17320
},
{
"epoch": 11.968232044198896,
"grad_norm": 0.1367802619934082,
"learning_rate": 4.797442913083539e-06,
"loss": 0.018,
"step": 17330
},
{
"epoch": 11.975138121546962,
"grad_norm": 0.11630649864673615,
"learning_rate": 4.7621680611617596e-06,
"loss": 0.022,
"step": 17340
},
{
"epoch": 11.982044198895027,
"grad_norm": 0.1627459079027176,
"learning_rate": 4.727016887745095e-06,
"loss": 0.0198,
"step": 17350
},
{
"epoch": 11.988950276243093,
"grad_norm": 0.18625329434871674,
"learning_rate": 4.691989488935511e-06,
"loss": 0.0216,
"step": 17360
},
{
"epoch": 11.99585635359116,
"grad_norm": 0.19761201739311218,
"learning_rate": 4.657085960496588e-06,
"loss": 0.0172,
"step": 17370
},
{
"epoch": 12.002762430939226,
"grad_norm": 0.14513880014419556,
"learning_rate": 4.6223063978532265e-06,
"loss": 0.0151,
"step": 17380
},
{
"epoch": 12.009668508287293,
"grad_norm": 0.07700514793395996,
"learning_rate": 4.587650896091439e-06,
"loss": 0.0158,
"step": 17390
},
{
"epoch": 12.01657458563536,
"grad_norm": 0.15232037007808685,
"learning_rate": 4.553119549958035e-06,
"loss": 0.0158,
"step": 17400
},
{
"epoch": 12.023480662983426,
"grad_norm": 0.12302271276712418,
"learning_rate": 4.518712453860385e-06,
"loss": 0.018,
"step": 17410
},
{
"epoch": 12.030386740331492,
"grad_norm": 0.1230573058128357,
"learning_rate": 4.484429701866205e-06,
"loss": 0.0207,
"step": 17420
},
{
"epoch": 12.037292817679559,
"grad_norm": 0.13379715383052826,
"learning_rate": 4.4502713877031975e-06,
"loss": 0.0214,
"step": 17430
},
{
"epoch": 12.044198895027625,
"grad_norm": 0.13079579174518585,
"learning_rate": 4.416237604758911e-06,
"loss": 0.0191,
"step": 17440
},
{
"epoch": 12.05110497237569,
"grad_norm": 0.11686630547046661,
"learning_rate": 4.3823284460804025e-06,
"loss": 0.0165,
"step": 17450
},
{
"epoch": 12.058011049723756,
"grad_norm": 0.1403651237487793,
"learning_rate": 4.348544004374011e-06,
"loss": 0.0179,
"step": 17460
},
{
"epoch": 12.064917127071823,
"grad_norm": 0.17115865647792816,
"learning_rate": 4.314884372005123e-06,
"loss": 0.0196,
"step": 17470
},
{
"epoch": 12.07182320441989,
"grad_norm": 0.13972045481204987,
"learning_rate": 4.281349640997867e-06,
"loss": 0.0206,
"step": 17480
},
{
"epoch": 12.078729281767956,
"grad_norm": 0.1649554967880249,
"learning_rate": 4.247939903034942e-06,
"loss": 0.0195,
"step": 17490
},
{
"epoch": 12.085635359116022,
"grad_norm": 0.11629771441221237,
"learning_rate": 4.214655249457284e-06,
"loss": 0.0152,
"step": 17500
},
{
"epoch": 12.092541436464089,
"grad_norm": 0.17439007759094238,
"learning_rate": 4.181495771263855e-06,
"loss": 0.0158,
"step": 17510
},
{
"epoch": 12.099447513812155,
"grad_norm": 0.1433987021446228,
"learning_rate": 4.148461559111427e-06,
"loss": 0.02,
"step": 17520
},
{
"epoch": 12.106353591160222,
"grad_norm": 0.11897672712802887,
"learning_rate": 4.115552703314252e-06,
"loss": 0.0172,
"step": 17530
},
{
"epoch": 12.113259668508288,
"grad_norm": 0.1429292857646942,
"learning_rate": 4.082769293843886e-06,
"loss": 0.026,
"step": 17540
},
{
"epoch": 12.120165745856353,
"grad_norm": 0.07817108929157257,
"learning_rate": 4.050111420328939e-06,
"loss": 0.0207,
"step": 17550
},
{
"epoch": 12.12707182320442,
"grad_norm": 0.18502286076545715,
"learning_rate": 4.017579172054764e-06,
"loss": 0.0176,
"step": 17560
},
{
"epoch": 12.133977900552486,
"grad_norm": 0.17949388921260834,
"learning_rate": 3.985172637963308e-06,
"loss": 0.0222,
"step": 17570
},
{
"epoch": 12.140883977900552,
"grad_norm": 0.10429384559392929,
"learning_rate": 3.952891906652784e-06,
"loss": 0.0187,
"step": 17580
},
{
"epoch": 12.147790055248619,
"grad_norm": 0.14530150592327118,
"learning_rate": 3.920737066377478e-06,
"loss": 0.0199,
"step": 17590
},
{
"epoch": 12.154696132596685,
"grad_norm": 0.13472408056259155,
"learning_rate": 3.888708205047509e-06,
"loss": 0.0146,
"step": 17600
},
{
"epoch": 12.161602209944752,
"grad_norm": 0.12338479608297348,
"learning_rate": 3.856805410228542e-06,
"loss": 0.014,
"step": 17610
},
{
"epoch": 12.168508287292818,
"grad_norm": 0.12579374015331268,
"learning_rate": 3.82502876914162e-06,
"loss": 0.0186,
"step": 17620
},
{
"epoch": 12.175414364640885,
"grad_norm": 0.12865792214870453,
"learning_rate": 3.7933783686628586e-06,
"loss": 0.0176,
"step": 17630
},
{
"epoch": 12.182320441988951,
"grad_norm": 0.1593572199344635,
"learning_rate": 3.7618542953232306e-06,
"loss": 0.0205,
"step": 17640
},
{
"epoch": 12.189226519337016,
"grad_norm": 0.15046411752700806,
"learning_rate": 3.7304566353083658e-06,
"loss": 0.0187,
"step": 17650
},
{
"epoch": 12.196132596685082,
"grad_norm": 0.14371070265769958,
"learning_rate": 3.6991854744582555e-06,
"loss": 0.018,
"step": 17660
},
{
"epoch": 12.203038674033149,
"grad_norm": 0.2144523411989212,
"learning_rate": 3.6680408982670777e-06,
"loss": 0.0231,
"step": 17670
},
{
"epoch": 12.209944751381215,
"grad_norm": 0.13164383172988892,
"learning_rate": 3.637022991882899e-06,
"loss": 0.0152,
"step": 17680
},
{
"epoch": 12.216850828729282,
"grad_norm": 0.1407146155834198,
"learning_rate": 3.606131840107485e-06,
"loss": 0.019,
"step": 17690
},
{
"epoch": 12.223756906077348,
"grad_norm": 0.23521016538143158,
"learning_rate": 3.575367527396084e-06,
"loss": 0.0237,
"step": 17700
},
{
"epoch": 12.230662983425415,
"grad_norm": 0.15884459018707275,
"learning_rate": 3.5447301378571386e-06,
"loss": 0.0195,
"step": 17710
},
{
"epoch": 12.237569060773481,
"grad_norm": 0.16665038466453552,
"learning_rate": 3.514219755252113e-06,
"loss": 0.0179,
"step": 17720
},
{
"epoch": 12.244475138121548,
"grad_norm": 0.1557607352733612,
"learning_rate": 3.4838364629952213e-06,
"loss": 0.0166,
"step": 17730
},
{
"epoch": 12.251381215469614,
"grad_norm": 0.1939476877450943,
"learning_rate": 3.4535803441532123e-06,
"loss": 0.0215,
"step": 17740
},
{
"epoch": 12.258287292817679,
"grad_norm": 0.07367555052042007,
"learning_rate": 3.4234514814451836e-06,
"loss": 0.0182,
"step": 17750
},
{
"epoch": 12.265193370165745,
"grad_norm": 0.14954784512519836,
"learning_rate": 3.393449957242273e-06,
"loss": 0.0168,
"step": 17760
},
{
"epoch": 12.272099447513812,
"grad_norm": 0.11408301442861557,
"learning_rate": 3.363575853567524e-06,
"loss": 0.0234,
"step": 17770
},
{
"epoch": 12.279005524861878,
"grad_norm": 0.1830197125673294,
"learning_rate": 3.3338292520955826e-06,
"loss": 0.016,
"step": 17780
},
{
"epoch": 12.285911602209945,
"grad_norm": 0.12380652874708176,
"learning_rate": 3.304210234152516e-06,
"loss": 0.0213,
"step": 17790
},
{
"epoch": 12.292817679558011,
"grad_norm": 0.13707053661346436,
"learning_rate": 3.2747188807155993e-06,
"loss": 0.0191,
"step": 17800
},
{
"epoch": 12.299723756906078,
"grad_norm": 0.1347280889749527,
"learning_rate": 3.2453552724130643e-06,
"loss": 0.022,
"step": 17810
},
{
"epoch": 12.306629834254144,
"grad_norm": 0.13858382403850555,
"learning_rate": 3.216119489523889e-06,
"loss": 0.0165,
"step": 17820
},
{
"epoch": 12.31353591160221,
"grad_norm": 0.15664106607437134,
"learning_rate": 3.1870116119775917e-06,
"loss": 0.0201,
"step": 17830
},
{
"epoch": 12.320441988950277,
"grad_norm": 0.20320770144462585,
"learning_rate": 3.158031719353999e-06,
"loss": 0.0205,
"step": 17840
},
{
"epoch": 12.327348066298342,
"grad_norm": 0.1687484085559845,
"learning_rate": 3.1291798908830273e-06,
"loss": 0.0277,
"step": 17850
},
{
"epoch": 12.334254143646408,
"grad_norm": 0.1131737232208252,
"learning_rate": 3.1004562054444853e-06,
"loss": 0.0183,
"step": 17860
},
{
"epoch": 12.341160220994475,
"grad_norm": 0.16560691595077515,
"learning_rate": 3.071860741567806e-06,
"loss": 0.0184,
"step": 17870
},
{
"epoch": 12.348066298342541,
"grad_norm": 0.17285485565662384,
"learning_rate": 3.04339357743193e-06,
"loss": 0.0189,
"step": 17880
},
{
"epoch": 12.354972375690608,
"grad_norm": 0.16125869750976562,
"learning_rate": 3.0150547908649628e-06,
"loss": 0.019,
"step": 17890
},
{
"epoch": 12.361878453038674,
"grad_norm": 0.1124703586101532,
"learning_rate": 2.9868444593440957e-06,
"loss": 0.0184,
"step": 17900
},
{
"epoch": 12.36878453038674,
"grad_norm": 0.15693354606628418,
"learning_rate": 2.9587626599952846e-06,
"loss": 0.0149,
"step": 17910
},
{
"epoch": 12.375690607734807,
"grad_norm": 0.15678930282592773,
"learning_rate": 2.930809469593082e-06,
"loss": 0.0215,
"step": 17920
},
{
"epoch": 12.382596685082873,
"grad_norm": 0.13219571113586426,
"learning_rate": 2.9029849645604733e-06,
"loss": 0.0214,
"step": 17930
},
{
"epoch": 12.38950276243094,
"grad_norm": 0.14746655523777008,
"learning_rate": 2.8752892209685632e-06,
"loss": 0.0191,
"step": 17940
},
{
"epoch": 12.396408839779006,
"grad_norm": 0.12485410273075104,
"learning_rate": 2.847722314536483e-06,
"loss": 0.0185,
"step": 17950
},
{
"epoch": 12.403314917127071,
"grad_norm": 0.08592341095209122,
"learning_rate": 2.820284320631078e-06,
"loss": 0.0179,
"step": 17960
},
{
"epoch": 12.410220994475138,
"grad_norm": 0.1952495276927948,
"learning_rate": 2.792975314266788e-06,
"loss": 0.0234,
"step": 17970
},
{
"epoch": 12.417127071823204,
"grad_norm": 0.13520370423793793,
"learning_rate": 2.7657953701054007e-06,
"loss": 0.0202,
"step": 17980
},
{
"epoch": 12.42403314917127,
"grad_norm": 0.17327278852462769,
"learning_rate": 2.7387445624558306e-06,
"loss": 0.0201,
"step": 17990
},
{
"epoch": 12.430939226519337,
"grad_norm": 0.09471528232097626,
"learning_rate": 2.7118229652739747e-06,
"loss": 0.0185,
"step": 18000
},
{
"epoch": 12.437845303867404,
"grad_norm": 0.10891865938901901,
"learning_rate": 2.6850306521624236e-06,
"loss": 0.0218,
"step": 18010
},
{
"epoch": 12.44475138121547,
"grad_norm": 0.1492462456226349,
"learning_rate": 2.6583676963703507e-06,
"loss": 0.0221,
"step": 18020
},
{
"epoch": 12.451657458563536,
"grad_norm": 0.14195561408996582,
"learning_rate": 2.631834170793268e-06,
"loss": 0.0209,
"step": 18030
},
{
"epoch": 12.458563535911603,
"grad_norm": 0.11009178310632706,
"learning_rate": 2.6054301479728036e-06,
"loss": 0.0165,
"step": 18040
},
{
"epoch": 12.465469613259668,
"grad_norm": 0.12649135291576385,
"learning_rate": 2.579155700096575e-06,
"loss": 0.0212,
"step": 18050
},
{
"epoch": 12.472375690607734,
"grad_norm": 0.16060461103916168,
"learning_rate": 2.5530108989978873e-06,
"loss": 0.0217,
"step": 18060
},
{
"epoch": 12.4792817679558,
"grad_norm": 0.12160547077655792,
"learning_rate": 2.5269958161556416e-06,
"loss": 0.0165,
"step": 18070
},
{
"epoch": 12.486187845303867,
"grad_norm": 0.16475243866443634,
"learning_rate": 2.5011105226940888e-06,
"loss": 0.0145,
"step": 18080
},
{
"epoch": 12.493093922651934,
"grad_norm": 0.0938870906829834,
"learning_rate": 2.4753550893826248e-06,
"loss": 0.0179,
"step": 18090
},
{
"epoch": 12.5,
"grad_norm": 0.11797482520341873,
"learning_rate": 2.4497295866356296e-06,
"loss": 0.0199,
"step": 18100
},
{
"epoch": 12.506906077348066,
"grad_norm": 0.08225338906049728,
"learning_rate": 2.424234084512228e-06,
"loss": 0.0188,
"step": 18110
},
{
"epoch": 12.513812154696133,
"grad_norm": 0.14476081728935242,
"learning_rate": 2.3988686527161687e-06,
"loss": 0.0177,
"step": 18120
},
{
"epoch": 12.5207182320442,
"grad_norm": 0.15867789089679718,
"learning_rate": 2.373633360595573e-06,
"loss": 0.0168,
"step": 18130
},
{
"epoch": 12.527624309392266,
"grad_norm": 0.15915602445602417,
"learning_rate": 2.3485282771427585e-06,
"loss": 0.0187,
"step": 18140
},
{
"epoch": 12.534530386740332,
"grad_norm": 0.11553303152322769,
"learning_rate": 2.3235534709940665e-06,
"loss": 0.0191,
"step": 18150
},
{
"epoch": 12.541436464088397,
"grad_norm": 0.1287674605846405,
"learning_rate": 2.2987090104296617e-06,
"loss": 0.0169,
"step": 18160
},
{
"epoch": 12.548342541436464,
"grad_norm": 0.14101791381835938,
"learning_rate": 2.273994963373355e-06,
"loss": 0.0179,
"step": 18170
},
{
"epoch": 12.55524861878453,
"grad_norm": 0.2093939483165741,
"learning_rate": 2.249411397392409e-06,
"loss": 0.0206,
"step": 18180
},
{
"epoch": 12.562154696132596,
"grad_norm": 0.10685942322015762,
"learning_rate": 2.2249583796973506e-06,
"loss": 0.0208,
"step": 18190
},
{
"epoch": 12.569060773480663,
"grad_norm": 0.12277834117412567,
"learning_rate": 2.200635977141796e-06,
"loss": 0.0176,
"step": 18200
},
{
"epoch": 12.57596685082873,
"grad_norm": 0.12654268741607666,
"learning_rate": 2.17644425622226e-06,
"loss": 0.0152,
"step": 18210
},
{
"epoch": 12.582872928176796,
"grad_norm": 0.15644238889217377,
"learning_rate": 2.152383283077991e-06,
"loss": 0.0174,
"step": 18220
},
{
"epoch": 12.589779005524862,
"grad_norm": 0.21075870096683502,
"learning_rate": 2.128453123490781e-06,
"loss": 0.0198,
"step": 18230
},
{
"epoch": 12.596685082872929,
"grad_norm": 0.1122535914182663,
"learning_rate": 2.1046538428847462e-06,
"loss": 0.0194,
"step": 18240
},
{
"epoch": 12.603591160220994,
"grad_norm": 0.12407967448234558,
"learning_rate": 2.0809855063262273e-06,
"loss": 0.0192,
"step": 18250
},
{
"epoch": 12.61049723756906,
"grad_norm": 0.18401677906513214,
"learning_rate": 2.057448178523558e-06,
"loss": 0.0204,
"step": 18260
},
{
"epoch": 12.617403314917127,
"grad_norm": 0.10147448629140854,
"learning_rate": 2.034041923826885e-06,
"loss": 0.0185,
"step": 18270
},
{
"epoch": 12.624309392265193,
"grad_norm": 0.14004665613174438,
"learning_rate": 2.0107668062280204e-06,
"loss": 0.0179,
"step": 18280
},
{
"epoch": 12.63121546961326,
"grad_norm": 0.16008542478084564,
"learning_rate": 1.9876228893602357e-06,
"loss": 0.0177,
"step": 18290
},
{
"epoch": 12.638121546961326,
"grad_norm": 0.08414597809314728,
"learning_rate": 1.9646102364981266e-06,
"loss": 0.0222,
"step": 18300
},
{
"epoch": 12.645027624309392,
"grad_norm": 0.11705183237791061,
"learning_rate": 1.9417289105574053e-06,
"loss": 0.0197,
"step": 18310
},
{
"epoch": 12.651933701657459,
"grad_norm": 0.11046871542930603,
"learning_rate": 1.9189789740947427e-06,
"loss": 0.0169,
"step": 18320
},
{
"epoch": 12.658839779005525,
"grad_norm": 0.11994214355945587,
"learning_rate": 1.896360489307597e-06,
"loss": 0.0183,
"step": 18330
},
{
"epoch": 12.665745856353592,
"grad_norm": 0.13720695674419403,
"learning_rate": 1.8738735180340362e-06,
"loss": 0.0184,
"step": 18340
},
{
"epoch": 12.672651933701658,
"grad_norm": 0.14560860395431519,
"learning_rate": 1.8515181217525824e-06,
"loss": 0.0185,
"step": 18350
},
{
"epoch": 12.679558011049723,
"grad_norm": 0.12190794199705124,
"learning_rate": 1.8292943615820457e-06,
"loss": 0.0161,
"step": 18360
},
{
"epoch": 12.68646408839779,
"grad_norm": 0.18701349198818207,
"learning_rate": 1.8072022982813296e-06,
"loss": 0.0219,
"step": 18370
},
{
"epoch": 12.693370165745856,
"grad_norm": 0.11787126213312149,
"learning_rate": 1.7852419922492925e-06,
"loss": 0.0209,
"step": 18380
},
{
"epoch": 12.700276243093922,
"grad_norm": 0.12787750363349915,
"learning_rate": 1.763413503524569e-06,
"loss": 0.0195,
"step": 18390
},
{
"epoch": 12.707182320441989,
"grad_norm": 0.16412900388240814,
"learning_rate": 1.7417168917854165e-06,
"loss": 0.0235,
"step": 18400
},
{
"epoch": 12.714088397790055,
"grad_norm": 0.1265489161014557,
"learning_rate": 1.720152216349552e-06,
"loss": 0.0208,
"step": 18410
},
{
"epoch": 12.720994475138122,
"grad_norm": 0.16048400104045868,
"learning_rate": 1.6987195361739595e-06,
"loss": 0.0225,
"step": 18420
},
{
"epoch": 12.727900552486188,
"grad_norm": 0.14972177147865295,
"learning_rate": 1.6774189098547832e-06,
"loss": 0.018,
"step": 18430
},
{
"epoch": 12.734806629834255,
"grad_norm": 0.14532731473445892,
"learning_rate": 1.6562503956271069e-06,
"loss": 0.0162,
"step": 18440
},
{
"epoch": 12.74171270718232,
"grad_norm": 0.13556219637393951,
"learning_rate": 1.6352140513648417e-06,
"loss": 0.0217,
"step": 18450
},
{
"epoch": 12.748618784530386,
"grad_norm": 0.12861819565296173,
"learning_rate": 1.6143099345805712e-06,
"loss": 0.0186,
"step": 18460
},
{
"epoch": 12.755524861878452,
"grad_norm": 0.1650349497795105,
"learning_rate": 1.5935381024253293e-06,
"loss": 0.0201,
"step": 18470
},
{
"epoch": 12.762430939226519,
"grad_norm": 0.11116552352905273,
"learning_rate": 1.572898611688517e-06,
"loss": 0.0146,
"step": 18480
},
{
"epoch": 12.769337016574585,
"grad_norm": 0.13286760449409485,
"learning_rate": 1.5523915187977133e-06,
"loss": 0.0196,
"step": 18490
},
{
"epoch": 12.776243093922652,
"grad_norm": 0.15505313873291016,
"learning_rate": 1.532016879818532e-06,
"loss": 0.0186,
"step": 18500
},
{
"epoch": 12.783149171270718,
"grad_norm": 0.1274903118610382,
"learning_rate": 1.51177475045447e-06,
"loss": 0.0185,
"step": 18510
},
{
"epoch": 12.790055248618785,
"grad_norm": 0.14875926077365875,
"learning_rate": 1.4916651860467035e-06,
"loss": 0.0224,
"step": 18520
},
{
"epoch": 12.796961325966851,
"grad_norm": 0.20385435223579407,
"learning_rate": 1.471688241574043e-06,
"loss": 0.019,
"step": 18530
},
{
"epoch": 12.803867403314918,
"grad_norm": 0.19278424978256226,
"learning_rate": 1.451843971652672e-06,
"loss": 0.0249,
"step": 18540
},
{
"epoch": 12.810773480662984,
"grad_norm": 0.16716869175434113,
"learning_rate": 1.432132430536076e-06,
"loss": 0.0207,
"step": 18550
},
{
"epoch": 12.817679558011049,
"grad_norm": 0.1731618344783783,
"learning_rate": 1.412553672114869e-06,
"loss": 0.0194,
"step": 18560
},
{
"epoch": 12.824585635359115,
"grad_norm": 0.11455850303173065,
"learning_rate": 1.3931077499166056e-06,
"loss": 0.0196,
"step": 18570
},
{
"epoch": 12.831491712707182,
"grad_norm": 0.18242445588111877,
"learning_rate": 1.3737947171057085e-06,
"loss": 0.0171,
"step": 18580
},
{
"epoch": 12.838397790055248,
"grad_norm": 0.12524786591529846,
"learning_rate": 1.3546146264832582e-06,
"loss": 0.0191,
"step": 18590
},
{
"epoch": 12.845303867403315,
"grad_norm": 0.13134995102882385,
"learning_rate": 1.3355675304869086e-06,
"loss": 0.0213,
"step": 18600
},
{
"epoch": 12.852209944751381,
"grad_norm": 0.1714327186346054,
"learning_rate": 1.3166534811906827e-06,
"loss": 0.0208,
"step": 18610
},
{
"epoch": 12.859116022099448,
"grad_norm": 0.10196040570735931,
"learning_rate": 1.2978725303048666e-06,
"loss": 0.0164,
"step": 18620
},
{
"epoch": 12.866022099447514,
"grad_norm": 0.12005294859409332,
"learning_rate": 1.2792247291758762e-06,
"loss": 0.0185,
"step": 18630
},
{
"epoch": 12.87292817679558,
"grad_norm": 0.15142318606376648,
"learning_rate": 1.2607101287860635e-06,
"loss": 0.016,
"step": 18640
},
{
"epoch": 12.879834254143645,
"grad_norm": 0.1705733835697174,
"learning_rate": 1.2423287797536654e-06,
"loss": 0.0192,
"step": 18650
},
{
"epoch": 12.886740331491712,
"grad_norm": 0.13848072290420532,
"learning_rate": 1.2240807323325776e-06,
"loss": 0.0171,
"step": 18660
},
{
"epoch": 12.893646408839778,
"grad_norm": 0.17544282972812653,
"learning_rate": 1.205966036412254e-06,
"loss": 0.0215,
"step": 18670
},
{
"epoch": 12.900552486187845,
"grad_norm": 0.1576301008462906,
"learning_rate": 1.1879847415175949e-06,
"loss": 0.0258,
"step": 18680
},
{
"epoch": 12.907458563535911,
"grad_norm": 0.13134483993053436,
"learning_rate": 1.1701368968087712e-06,
"loss": 0.0171,
"step": 18690
},
{
"epoch": 12.914364640883978,
"grad_norm": 0.1861048936843872,
"learning_rate": 1.1524225510811116e-06,
"loss": 0.0216,
"step": 18700
},
{
"epoch": 12.921270718232044,
"grad_norm": 0.12704524397850037,
"learning_rate": 1.1348417527649535e-06,
"loss": 0.0207,
"step": 18710
},
{
"epoch": 12.92817679558011,
"grad_norm": 0.10087218880653381,
"learning_rate": 1.1173945499255268e-06,
"loss": 0.0166,
"step": 18720
},
{
"epoch": 12.935082872928177,
"grad_norm": 0.16379135847091675,
"learning_rate": 1.1000809902628307e-06,
"loss": 0.017,
"step": 18730
},
{
"epoch": 12.941988950276244,
"grad_norm": 0.12458409368991852,
"learning_rate": 1.082901121111468e-06,
"loss": 0.0183,
"step": 18740
},
{
"epoch": 12.94889502762431,
"grad_norm": 0.18087325990200043,
"learning_rate": 1.0658549894405456e-06,
"loss": 0.0168,
"step": 18750
},
{
"epoch": 12.955801104972375,
"grad_norm": 0.15164682269096375,
"learning_rate": 1.0489426418535342e-06,
"loss": 0.0187,
"step": 18760
},
{
"epoch": 12.962707182320441,
"grad_norm": 0.15599878132343292,
"learning_rate": 1.0321641245881474e-06,
"loss": 0.0175,
"step": 18770
},
{
"epoch": 12.969613259668508,
"grad_norm": 0.08068261295557022,
"learning_rate": 1.015519483516214e-06,
"loss": 0.0154,
"step": 18780
},
{
"epoch": 12.976519337016574,
"grad_norm": 0.14306342601776123,
"learning_rate": 9.990087641435443e-07,
"loss": 0.0202,
"step": 18790
},
{
"epoch": 12.98342541436464,
"grad_norm": 0.1224728599190712,
"learning_rate": 9.826320116098132e-07,
"loss": 0.019,
"step": 18800
},
{
"epoch": 12.990331491712707,
"grad_norm": 0.1325245350599289,
"learning_rate": 9.663892706884447e-07,
"loss": 0.0196,
"step": 18810
},
{
"epoch": 12.997237569060774,
"grad_norm": 0.12311689555644989,
"learning_rate": 9.502805857864616e-07,
"loss": 0.0162,
"step": 18820
},
{
"epoch": 13.00414364640884,
"grad_norm": 0.22265133261680603,
"learning_rate": 9.34306000944396e-07,
"loss": 0.0196,
"step": 18830
},
{
"epoch": 13.011049723756907,
"grad_norm": 0.11388003826141357,
"learning_rate": 9.184655598361624e-07,
"loss": 0.0151,
"step": 18840
},
{
"epoch": 13.017955801104973,
"grad_norm": 0.11467894166707993,
"learning_rate": 9.027593057689076e-07,
"loss": 0.0162,
"step": 18850
},
{
"epoch": 13.024861878453038,
"grad_norm": 0.17213301360607147,
"learning_rate": 8.871872816829441e-07,
"loss": 0.0164,
"step": 18860
},
{
"epoch": 13.031767955801104,
"grad_norm": 0.10395052284002304,
"learning_rate": 8.717495301515777e-07,
"loss": 0.0162,
"step": 18870
},
{
"epoch": 13.03867403314917,
"grad_norm": 0.14414413273334503,
"learning_rate": 8.564460933810415e-07,
"loss": 0.0185,
"step": 18880
},
{
"epoch": 13.045580110497237,
"grad_norm": 0.14760589599609375,
"learning_rate": 8.412770132103453e-07,
"loss": 0.0179,
"step": 18890
},
{
"epoch": 13.052486187845304,
"grad_norm": 0.14571161568164825,
"learning_rate": 8.262423311111711e-07,
"loss": 0.017,
"step": 18900
},
{
"epoch": 13.05939226519337,
"grad_norm": 0.1314871907234192,
"learning_rate": 8.113420881877665e-07,
"loss": 0.0184,
"step": 18910
},
{
"epoch": 13.066298342541437,
"grad_norm": 0.13378721475601196,
"learning_rate": 7.965763251768288e-07,
"loss": 0.0198,
"step": 18920
},
{
"epoch": 13.073204419889503,
"grad_norm": 0.16803504526615143,
"learning_rate": 7.819450824473995e-07,
"loss": 0.0212,
"step": 18930
},
{
"epoch": 13.08011049723757,
"grad_norm": 0.12649837136268616,
"learning_rate": 7.674484000007198e-07,
"loss": 0.0189,
"step": 18940
},
{
"epoch": 13.087016574585636,
"grad_norm": 0.13606299459934235,
"learning_rate": 7.530863174701752e-07,
"loss": 0.0178,
"step": 18950
},
{
"epoch": 13.0939226519337,
"grad_norm": 0.20246635377407074,
"learning_rate": 7.38858874121151e-07,
"loss": 0.0196,
"step": 18960
},
{
"epoch": 13.100828729281767,
"grad_norm": 0.10595327615737915,
"learning_rate": 7.247661088509328e-07,
"loss": 0.0189,
"step": 18970
},
{
"epoch": 13.107734806629834,
"grad_norm": 0.20289529860019684,
"learning_rate": 7.108080601886002e-07,
"loss": 0.0185,
"step": 18980
},
{
"epoch": 13.1146408839779,
"grad_norm": 0.16404929757118225,
"learning_rate": 6.969847662949336e-07,
"loss": 0.0158,
"step": 18990
},
{
"epoch": 13.121546961325967,
"grad_norm": 0.09362056851387024,
"learning_rate": 6.832962649622798e-07,
"loss": 0.0164,
"step": 19000
},
{
"epoch": 13.128453038674033,
"grad_norm": 0.09150226414203644,
"learning_rate": 6.697425936144863e-07,
"loss": 0.0161,
"step": 19010
},
{
"epoch": 13.1353591160221,
"grad_norm": 0.1991247683763504,
"learning_rate": 6.563237893067731e-07,
"loss": 0.0201,
"step": 19020
},
{
"epoch": 13.142265193370166,
"grad_norm": 0.1374427080154419,
"learning_rate": 6.430398887256328e-07,
"loss": 0.02,
"step": 19030
},
{
"epoch": 13.149171270718233,
"grad_norm": 0.200238436460495,
"learning_rate": 6.298909281887478e-07,
"loss": 0.0177,
"step": 19040
},
{
"epoch": 13.1560773480663,
"grad_norm": 0.14022763073444366,
"learning_rate": 6.168769436448673e-07,
"loss": 0.0177,
"step": 19050
},
{
"epoch": 13.162983425414364,
"grad_norm": 0.13676732778549194,
"learning_rate": 6.03997970673742e-07,
"loss": 0.0156,
"step": 19060
},
{
"epoch": 13.16988950276243,
"grad_norm": 0.12980788946151733,
"learning_rate": 5.912540444859782e-07,
"loss": 0.0182,
"step": 19070
},
{
"epoch": 13.176795580110497,
"grad_norm": 0.1635156124830246,
"learning_rate": 5.786451999229837e-07,
"loss": 0.0168,
"step": 19080
},
{
"epoch": 13.183701657458563,
"grad_norm": 0.1873101443052292,
"learning_rate": 5.661714714568722e-07,
"loss": 0.0174,
"step": 19090
},
{
"epoch": 13.19060773480663,
"grad_norm": 0.10086224973201752,
"learning_rate": 5.538328931903259e-07,
"loss": 0.0153,
"step": 19100
},
{
"epoch": 13.197513812154696,
"grad_norm": 0.12298750132322311,
"learning_rate": 5.416294988565551e-07,
"loss": 0.0207,
"step": 19110
},
{
"epoch": 13.204419889502763,
"grad_norm": 0.12365196645259857,
"learning_rate": 5.29561321819172e-07,
"loss": 0.0168,
"step": 19120
},
{
"epoch": 13.21132596685083,
"grad_norm": 0.10630401968955994,
"learning_rate": 5.176283950721061e-07,
"loss": 0.0172,
"step": 19130
},
{
"epoch": 13.218232044198896,
"grad_norm": 0.11978129297494888,
"learning_rate": 5.058307512395332e-07,
"loss": 0.0193,
"step": 19140
},
{
"epoch": 13.225138121546962,
"grad_norm": 0.11279390752315521,
"learning_rate": 4.941684225757526e-07,
"loss": 0.0184,
"step": 19150
},
{
"epoch": 13.232044198895027,
"grad_norm": 0.0903186947107315,
"learning_rate": 4.826414409651314e-07,
"loss": 0.0206,
"step": 19160
},
{
"epoch": 13.238950276243093,
"grad_norm": 0.15330806374549866,
"learning_rate": 4.712498379219943e-07,
"loss": 0.0202,
"step": 19170
},
{
"epoch": 13.24585635359116,
"grad_norm": 0.10040513426065445,
"learning_rate": 4.599936445905506e-07,
"loss": 0.0176,
"step": 19180
},
{
"epoch": 13.252762430939226,
"grad_norm": 0.14349868893623352,
"learning_rate": 4.4887289174480594e-07,
"loss": 0.0179,
"step": 19190
},
{
"epoch": 13.259668508287293,
"grad_norm": 0.14012572169303894,
"learning_rate": 4.378876097884621e-07,
"loss": 0.0196,
"step": 19200
},
{
"epoch": 13.26657458563536,
"grad_norm": 0.09931646287441254,
"learning_rate": 4.2703782875487264e-07,
"loss": 0.0158,
"step": 19210
},
{
"epoch": 13.273480662983426,
"grad_norm": 0.13279908895492554,
"learning_rate": 4.163235783069208e-07,
"loss": 0.0188,
"step": 19220
},
{
"epoch": 13.280386740331492,
"grad_norm": 0.15899476408958435,
"learning_rate": 4.057448877369585e-07,
"loss": 0.0192,
"step": 19230
},
{
"epoch": 13.287292817679559,
"grad_norm": 0.1579292118549347,
"learning_rate": 3.9530178596672295e-07,
"loss": 0.0159,
"step": 19240
},
{
"epoch": 13.294198895027625,
"grad_norm": 0.15862807631492615,
"learning_rate": 3.849943015472479e-07,
"loss": 0.0206,
"step": 19250
},
{
"epoch": 13.30110497237569,
"grad_norm": 0.09818954765796661,
"learning_rate": 3.748224626588137e-07,
"loss": 0.018,
"step": 19260
},
{
"epoch": 13.308011049723756,
"grad_norm": 0.1983523666858673,
"learning_rate": 3.647862971108307e-07,
"loss": 0.0163,
"step": 19270
},
{
"epoch": 13.314917127071823,
"grad_norm": 0.1573604941368103,
"learning_rate": 3.5488583234179473e-07,
"loss": 0.0191,
"step": 19280
},
{
"epoch": 13.32182320441989,
"grad_norm": 0.08943048864603043,
"learning_rate": 3.4512109541920413e-07,
"loss": 0.0204,
"step": 19290
},
{
"epoch": 13.328729281767956,
"grad_norm": 0.09052513539791107,
"learning_rate": 3.354921130394706e-07,
"loss": 0.0188,
"step": 19300
},
{
"epoch": 13.335635359116022,
"grad_norm": 0.14040173590183258,
"learning_rate": 3.259989115278639e-07,
"loss": 0.0168,
"step": 19310
},
{
"epoch": 13.342541436464089,
"grad_norm": 0.12269480526447296,
"learning_rate": 3.1664151683843403e-07,
"loss": 0.0168,
"step": 19320
},
{
"epoch": 13.349447513812155,
"grad_norm": 0.14032500982284546,
"learning_rate": 3.074199545539447e-07,
"loss": 0.0155,
"step": 19330
},
{
"epoch": 13.356353591160222,
"grad_norm": 0.1279202252626419,
"learning_rate": 2.983342498857955e-07,
"loss": 0.0197,
"step": 19340
},
{
"epoch": 13.363259668508288,
"grad_norm": 0.12618181109428406,
"learning_rate": 2.893844276739499e-07,
"loss": 0.0153,
"step": 19350
},
{
"epoch": 13.370165745856353,
"grad_norm": 0.12394572794437408,
"learning_rate": 2.8057051238688514e-07,
"loss": 0.0196,
"step": 19360
},
{
"epoch": 13.37707182320442,
"grad_norm": 0.09562724083662033,
"learning_rate": 2.71892528121509e-07,
"loss": 0.0157,
"step": 19370
},
{
"epoch": 13.383977900552486,
"grad_norm": 0.17502541840076447,
"learning_rate": 2.633504986030988e-07,
"loss": 0.0215,
"step": 19380
},
{
"epoch": 13.390883977900552,
"grad_norm": 0.16975905001163483,
"learning_rate": 2.549444471852347e-07,
"loss": 0.0159,
"step": 19390
},
{
"epoch": 13.397790055248619,
"grad_norm": 0.14912907779216766,
"learning_rate": 2.4667439684974423e-07,
"loss": 0.0195,
"step": 19400
},
{
"epoch": 13.404696132596685,
"grad_norm": 0.13770301640033722,
"learning_rate": 2.3854037020662467e-07,
"loss": 0.0183,
"step": 19410
},
{
"epoch": 13.411602209944752,
"grad_norm": 0.11400759220123291,
"learning_rate": 2.3054238949399288e-07,
"loss": 0.0208,
"step": 19420
},
{
"epoch": 13.418508287292818,
"grad_norm": 0.1524895429611206,
"learning_rate": 2.2268047657802993e-07,
"loss": 0.0164,
"step": 19430
},
{
"epoch": 13.425414364640885,
"grad_norm": 0.18355759978294373,
"learning_rate": 2.149546529529034e-07,
"loss": 0.0178,
"step": 19440
},
{
"epoch": 13.432320441988951,
"grad_norm": 0.1252148151397705,
"learning_rate": 2.0736493974071736e-07,
"loss": 0.021,
"step": 19450
},
{
"epoch": 13.439226519337016,
"grad_norm": 0.10146128386259079,
"learning_rate": 1.9991135769145686e-07,
"loss": 0.0129,
"step": 19460
},
{
"epoch": 13.446132596685082,
"grad_norm": 0.13486264646053314,
"learning_rate": 1.9259392718293245e-07,
"loss": 0.0182,
"step": 19470
},
{
"epoch": 13.453038674033149,
"grad_norm": 0.14005959033966064,
"learning_rate": 1.8541266822072467e-07,
"loss": 0.0211,
"step": 19480
},
{
"epoch": 13.459944751381215,
"grad_norm": 0.1532914638519287,
"learning_rate": 1.7836760043811184e-07,
"loss": 0.0231,
"step": 19490
},
{
"epoch": 13.466850828729282,
"grad_norm": 0.05971955880522728,
"learning_rate": 1.7145874309604792e-07,
"loss": 0.0143,
"step": 19500
},
{
"epoch": 13.473756906077348,
"grad_norm": 0.11946941912174225,
"learning_rate": 1.6468611508308474e-07,
"loss": 0.0172,
"step": 19510
},
{
"epoch": 13.480662983425415,
"grad_norm": 0.13063247501850128,
"learning_rate": 1.5804973491532204e-07,
"loss": 0.021,
"step": 19520
},
{
"epoch": 13.487569060773481,
"grad_norm": 0.1444508135318756,
"learning_rate": 1.5154962073637424e-07,
"loss": 0.017,
"step": 19530
},
{
"epoch": 13.494475138121548,
"grad_norm": 0.13403776288032532,
"learning_rate": 1.4518579031730372e-07,
"loss": 0.0168,
"step": 19540
},
{
"epoch": 13.501381215469614,
"grad_norm": 0.0963289737701416,
"learning_rate": 1.389582610565876e-07,
"loss": 0.0163,
"step": 19550
},
{
"epoch": 13.50828729281768,
"grad_norm": 0.07779234647750854,
"learning_rate": 1.3286704998003995e-07,
"loss": 0.0238,
"step": 19560
},
{
"epoch": 13.515193370165745,
"grad_norm": 0.12145277112722397,
"learning_rate": 1.2691217374080632e-07,
"loss": 0.0236,
"step": 19570
},
{
"epoch": 13.522099447513812,
"grad_norm": 0.1347256600856781,
"learning_rate": 1.2109364861929705e-07,
"loss": 0.0193,
"step": 19580
},
{
"epoch": 13.529005524861878,
"grad_norm": 0.15796039998531342,
"learning_rate": 1.1541149052312628e-07,
"loss": 0.0157,
"step": 19590
},
{
"epoch": 13.535911602209945,
"grad_norm": 0.15612980723381042,
"learning_rate": 1.0986571498710074e-07,
"loss": 0.0163,
"step": 19600
},
{
"epoch": 13.542817679558011,
"grad_norm": 0.13418418169021606,
"learning_rate": 1.0445633717316438e-07,
"loss": 0.0167,
"step": 19610
},
{
"epoch": 13.549723756906078,
"grad_norm": 0.21959972381591797,
"learning_rate": 9.918337187034277e-08,
"loss": 0.0156,
"step": 19620
},
{
"epoch": 13.556629834254144,
"grad_norm": 0.11513973772525787,
"learning_rate": 9.404683349472643e-08,
"loss": 0.0194,
"step": 19630
},
{
"epoch": 13.56353591160221,
"grad_norm": 0.13620199263095856,
"learning_rate": 8.904673608940983e-08,
"loss": 0.0164,
"step": 19640
},
{
"epoch": 13.570441988950277,
"grad_norm": 0.15609873831272125,
"learning_rate": 8.418309332447471e-08,
"loss": 0.0145,
"step": 19650
},
{
"epoch": 13.577348066298342,
"grad_norm": 0.1274515688419342,
"learning_rate": 7.945591849692902e-08,
"loss": 0.0162,
"step": 19660
},
{
"epoch": 13.584254143646408,
"grad_norm": 0.15051686763763428,
"learning_rate": 7.486522453069578e-08,
"loss": 0.0232,
"step": 19670
},
{
"epoch": 13.591160220994475,
"grad_norm": 0.17713497579097748,
"learning_rate": 7.041102397655208e-08,
"loss": 0.0163,
"step": 19680
},
{
"epoch": 13.598066298342541,
"grad_norm": 0.11196551471948624,
"learning_rate": 6.609332901210685e-08,
"loss": 0.0144,
"step": 19690
},
{
"epoch": 13.604972375690608,
"grad_norm": 0.1326775848865509,
"learning_rate": 6.191215144178419e-08,
"loss": 0.0188,
"step": 19700
},
{
"epoch": 13.611878453038674,
"grad_norm": 0.19205215573310852,
"learning_rate": 5.786750269675678e-08,
"loss": 0.0206,
"step": 19710
},
{
"epoch": 13.61878453038674,
"grad_norm": 0.18258792161941528,
"learning_rate": 5.395939383494031e-08,
"loss": 0.0212,
"step": 19720
},
{
"epoch": 13.625690607734807,
"grad_norm": 0.1334686279296875,
"learning_rate": 5.018783554095463e-08,
"loss": 0.0199,
"step": 19730
},
{
"epoch": 13.632596685082873,
"grad_norm": 0.08144468069076538,
"learning_rate": 4.655283812610156e-08,
"loss": 0.0158,
"step": 19740
},
{
"epoch": 13.63950276243094,
"grad_norm": 0.1487710326910019,
"learning_rate": 4.305441152831491e-08,
"loss": 0.0219,
"step": 19750
},
{
"epoch": 13.646408839779006,
"grad_norm": 0.10869629681110382,
"learning_rate": 3.9692565312171584e-08,
"loss": 0.0192,
"step": 19760
},
{
"epoch": 13.653314917127071,
"grad_norm": 0.12427249550819397,
"learning_rate": 3.6467308668824975e-08,
"loss": 0.0183,
"step": 19770
},
{
"epoch": 13.660220994475138,
"grad_norm": 0.1521531194448471,
"learning_rate": 3.3378650416004964e-08,
"loss": 0.0196,
"step": 19780
},
{
"epoch": 13.667127071823204,
"grad_norm": 0.14662829041481018,
"learning_rate": 3.042659899797906e-08,
"loss": 0.0215,
"step": 19790
},
{
"epoch": 13.67403314917127,
"grad_norm": 0.12476713210344315,
"learning_rate": 2.76111624855524e-08,
"loss": 0.0172,
"step": 19800
},
{
"epoch": 13.680939226519337,
"grad_norm": 0.15183356404304504,
"learning_rate": 2.4932348576017784e-08,
"loss": 0.0182,
"step": 19810
},
{
"epoch": 13.687845303867404,
"grad_norm": 0.1421179175376892,
"learning_rate": 2.239016459314458e-08,
"loss": 0.0157,
"step": 19820
},
{
"epoch": 13.69475138121547,
"grad_norm": 0.13849793374538422,
"learning_rate": 1.9984617487173174e-08,
"loss": 0.0213,
"step": 19830
},
{
"epoch": 13.701657458563536,
"grad_norm": 0.12221692502498627,
"learning_rate": 1.7715713834776105e-08,
"loss": 0.0177,
"step": 19840
},
{
"epoch": 13.708563535911603,
"grad_norm": 0.22249464690685272,
"learning_rate": 1.5583459839046964e-08,
"loss": 0.0211,
"step": 19850
},
{
"epoch": 13.715469613259668,
"grad_norm": 0.14573796093463898,
"learning_rate": 1.3587861329489304e-08,
"loss": 0.0241,
"step": 19860
},
{
"epoch": 13.722375690607734,
"grad_norm": 0.1393386274576187,
"learning_rate": 1.1728923761994415e-08,
"loss": 0.0183,
"step": 19870
},
{
"epoch": 13.7292817679558,
"grad_norm": 0.16851302981376648,
"learning_rate": 1.0006652218819135e-08,
"loss": 0.0194,
"step": 19880
},
{
"epoch": 13.736187845303867,
"grad_norm": 0.12711794674396515,
"learning_rate": 8.421051408596947e-09,
"loss": 0.0163,
"step": 19890
},
{
"epoch": 13.743093922651934,
"grad_norm": 0.16962414979934692,
"learning_rate": 6.972125666299123e-09,
"loss": 0.0194,
"step": 19900
},
{
"epoch": 13.75,
"grad_norm": 0.13897208869457245,
"learning_rate": 5.659878953229169e-09,
"loss": 0.0163,
"step": 19910
},
{
"epoch": 13.756906077348066,
"grad_norm": 0.1462239772081375,
"learning_rate": 4.48431485701728e-09,
"loss": 0.0155,
"step": 19920
},
{
"epoch": 13.763812154696133,
"grad_norm": 0.16802477836608887,
"learning_rate": 3.4454365916203322e-09,
"loss": 0.0153,
"step": 19930
},
{
"epoch": 13.7707182320442,
"grad_norm": 0.1017882451415062,
"learning_rate": 2.5432469972830332e-09,
"loss": 0.0157,
"step": 19940
},
{
"epoch": 13.777624309392266,
"grad_norm": 0.16575653851032257,
"learning_rate": 1.7777485405601203e-09,
"loss": 0.0203,
"step": 19950
},
{
"epoch": 13.784530386740332,
"grad_norm": 0.13925835490226746,
"learning_rate": 1.1489433142941597e-09,
"loss": 0.0181,
"step": 19960
},
{
"epoch": 13.791436464088397,
"grad_norm": 0.13744530081748962,
"learning_rate": 6.568330376210963e-10,
"loss": 0.0179,
"step": 19970
},
{
"epoch": 13.798342541436464,
"grad_norm": 0.09223198145627975,
"learning_rate": 3.0141905594249787e-10,
"loss": 0.0155,
"step": 19980
},
{
"epoch": 13.80524861878453,
"grad_norm": 0.11717510968446732,
"learning_rate": 8.270234094776008e-11,
"loss": 0.0192,
"step": 19990
},
{
"epoch": 13.812154696132596,
"grad_norm": 0.17134608328342438,
"learning_rate": 6.834906085551041e-13,
"loss": 0.0186,
"step": 20000
}
],
"logging_steps": 10,
"max_steps": 20000,
"num_input_tokens_seen": 0,
"num_train_epochs": 14,
"save_steps": 1000,
"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": 32,
"trial_name": null,
"trial_params": null
}