Qwen2.5-0.5B-Open-R1-Distill / trainer_state.json
rhinocamp's picture
Model save
971e654 verified
{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.0,
"eval_steps": 500,
"global_step": 2144,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0023320895522388058,
"grad_norm": 3.7231960217328877,
"learning_rate": 2.3148148148148148e-06,
"loss": 1.0047,
"step": 5
},
{
"epoch": 0.0046641791044776115,
"grad_norm": 2.8583368538866347,
"learning_rate": 4.6296296296296296e-06,
"loss": 0.9757,
"step": 10
},
{
"epoch": 0.006996268656716418,
"grad_norm": 2.8571738838176763,
"learning_rate": 6.944444444444445e-06,
"loss": 0.9678,
"step": 15
},
{
"epoch": 0.009328358208955223,
"grad_norm": 1.7369287409068157,
"learning_rate": 9.259259259259259e-06,
"loss": 0.9001,
"step": 20
},
{
"epoch": 0.01166044776119403,
"grad_norm": 1.1660883689238433,
"learning_rate": 1.1574074074074075e-05,
"loss": 0.8759,
"step": 25
},
{
"epoch": 0.013992537313432836,
"grad_norm": 1.1321862136040273,
"learning_rate": 1.388888888888889e-05,
"loss": 0.8381,
"step": 30
},
{
"epoch": 0.01632462686567164,
"grad_norm": 1.024631758461693,
"learning_rate": 1.6203703703703704e-05,
"loss": 0.834,
"step": 35
},
{
"epoch": 0.018656716417910446,
"grad_norm": 0.9053510661908479,
"learning_rate": 1.8518518518518518e-05,
"loss": 0.8172,
"step": 40
},
{
"epoch": 0.020988805970149255,
"grad_norm": 1.0170230987058317,
"learning_rate": 2.0833333333333336e-05,
"loss": 0.8146,
"step": 45
},
{
"epoch": 0.02332089552238806,
"grad_norm": 0.9764899017407078,
"learning_rate": 2.314814814814815e-05,
"loss": 0.7891,
"step": 50
},
{
"epoch": 0.025652985074626867,
"grad_norm": 1.0251102524307585,
"learning_rate": 2.5462962962962965e-05,
"loss": 0.7814,
"step": 55
},
{
"epoch": 0.027985074626865673,
"grad_norm": 1.1022491122388751,
"learning_rate": 2.777777777777778e-05,
"loss": 0.7872,
"step": 60
},
{
"epoch": 0.03031716417910448,
"grad_norm": 1.0992277107302175,
"learning_rate": 3.0092592592592593e-05,
"loss": 0.7726,
"step": 65
},
{
"epoch": 0.03264925373134328,
"grad_norm": 1.0365397617690528,
"learning_rate": 3.240740740740741e-05,
"loss": 0.7455,
"step": 70
},
{
"epoch": 0.034981343283582086,
"grad_norm": 1.195178900196762,
"learning_rate": 3.472222222222222e-05,
"loss": 0.756,
"step": 75
},
{
"epoch": 0.03731343283582089,
"grad_norm": 1.1412528086977047,
"learning_rate": 3.7037037037037037e-05,
"loss": 0.7633,
"step": 80
},
{
"epoch": 0.039645522388059705,
"grad_norm": 1.2472194183265157,
"learning_rate": 3.935185185185186e-05,
"loss": 0.7557,
"step": 85
},
{
"epoch": 0.04197761194029851,
"grad_norm": 1.0570653948605382,
"learning_rate": 4.166666666666667e-05,
"loss": 0.7601,
"step": 90
},
{
"epoch": 0.044309701492537316,
"grad_norm": 1.1587882371432328,
"learning_rate": 4.3981481481481486e-05,
"loss": 0.7417,
"step": 95
},
{
"epoch": 0.04664179104477612,
"grad_norm": 1.3162397865412059,
"learning_rate": 4.62962962962963e-05,
"loss": 0.739,
"step": 100
},
{
"epoch": 0.04897388059701493,
"grad_norm": 0.9985279247385236,
"learning_rate": 4.8611111111111115e-05,
"loss": 0.7687,
"step": 105
},
{
"epoch": 0.051305970149253734,
"grad_norm": 1.2887741163380346,
"learning_rate": 4.999989285883431e-05,
"loss": 0.7664,
"step": 110
},
{
"epoch": 0.05363805970149254,
"grad_norm": 1.2392518485033788,
"learning_rate": 4.999868753243865e-05,
"loss": 0.7675,
"step": 115
},
{
"epoch": 0.055970149253731345,
"grad_norm": 1.8916579665083808,
"learning_rate": 4.999614302517356e-05,
"loss": 0.7598,
"step": 120
},
{
"epoch": 0.05830223880597015,
"grad_norm": 1.5520377612517253,
"learning_rate": 4.9992259488494784e-05,
"loss": 0.7424,
"step": 125
},
{
"epoch": 0.06063432835820896,
"grad_norm": 1.239718216272371,
"learning_rate": 4.99870371535606e-05,
"loss": 0.7366,
"step": 130
},
{
"epoch": 0.06296641791044776,
"grad_norm": 0.9825341780149127,
"learning_rate": 4.9980476331218066e-05,
"loss": 0.786,
"step": 135
},
{
"epoch": 0.06529850746268656,
"grad_norm": 0.9155268841122092,
"learning_rate": 4.997257741198456e-05,
"loss": 0.7505,
"step": 140
},
{
"epoch": 0.06763059701492537,
"grad_norm": 1.082193107284934,
"learning_rate": 4.996334086602444e-05,
"loss": 0.7351,
"step": 145
},
{
"epoch": 0.06996268656716417,
"grad_norm": 1.0423700442715462,
"learning_rate": 4.9952767243121146e-05,
"loss": 0.7551,
"step": 150
},
{
"epoch": 0.07229477611940298,
"grad_norm": 0.9000144515873072,
"learning_rate": 4.994085717264443e-05,
"loss": 0.7368,
"step": 155
},
{
"epoch": 0.07462686567164178,
"grad_norm": 1.044303554591329,
"learning_rate": 4.992761136351291e-05,
"loss": 0.7436,
"step": 160
},
{
"epoch": 0.07695895522388059,
"grad_norm": 1.0470880900421262,
"learning_rate": 4.99130306041519e-05,
"loss": 0.7446,
"step": 165
},
{
"epoch": 0.07929104477611941,
"grad_norm": 1.1323111761895537,
"learning_rate": 4.989711576244639e-05,
"loss": 0.7226,
"step": 170
},
{
"epoch": 0.08162313432835822,
"grad_norm": 1.0668807832388134,
"learning_rate": 4.9879867785689506e-05,
"loss": 0.7107,
"step": 175
},
{
"epoch": 0.08395522388059702,
"grad_norm": 0.9414033524848218,
"learning_rate": 4.986128770052603e-05,
"loss": 0.7315,
"step": 180
},
{
"epoch": 0.08628731343283583,
"grad_norm": 0.9861314941949435,
"learning_rate": 4.9841376612891365e-05,
"loss": 0.7226,
"step": 185
},
{
"epoch": 0.08861940298507463,
"grad_norm": 0.9154465133645083,
"learning_rate": 4.9820135707945634e-05,
"loss": 0.719,
"step": 190
},
{
"epoch": 0.09095149253731344,
"grad_norm": 1.0049244209561472,
"learning_rate": 4.97975662500032e-05,
"loss": 0.7474,
"step": 195
},
{
"epoch": 0.09328358208955224,
"grad_norm": 0.9649064226232575,
"learning_rate": 4.9773669582457364e-05,
"loss": 0.7273,
"step": 200
},
{
"epoch": 0.09561567164179105,
"grad_norm": 1.0099484363867934,
"learning_rate": 4.974844712770044e-05,
"loss": 0.7456,
"step": 205
},
{
"epoch": 0.09794776119402986,
"grad_norm": 0.9092588577896343,
"learning_rate": 4.972190038703905e-05,
"loss": 0.7389,
"step": 210
},
{
"epoch": 0.10027985074626866,
"grad_norm": 1.0001019634384871,
"learning_rate": 4.96940309406048e-05,
"loss": 0.7334,
"step": 215
},
{
"epoch": 0.10261194029850747,
"grad_norm": 1.0388048983714429,
"learning_rate": 4.966484044726024e-05,
"loss": 0.7246,
"step": 220
},
{
"epoch": 0.10494402985074627,
"grad_norm": 1.0827378138832435,
"learning_rate": 4.963433064450002e-05,
"loss": 0.7208,
"step": 225
},
{
"epoch": 0.10727611940298508,
"grad_norm": 1.0427299412602653,
"learning_rate": 4.9602503348347625e-05,
"loss": 0.7133,
"step": 230
},
{
"epoch": 0.10960820895522388,
"grad_norm": 0.9388505104499245,
"learning_rate": 4.956936045324716e-05,
"loss": 0.6982,
"step": 235
},
{
"epoch": 0.11194029850746269,
"grad_norm": 0.8567679904473422,
"learning_rate": 4.953490393195063e-05,
"loss": 0.7518,
"step": 240
},
{
"epoch": 0.1142723880597015,
"grad_norm": 0.9211748764639448,
"learning_rate": 4.9499135835400526e-05,
"loss": 0.7287,
"step": 245
},
{
"epoch": 0.1166044776119403,
"grad_norm": 1.0303852066380612,
"learning_rate": 4.9462058292607735e-05,
"loss": 0.7134,
"step": 250
},
{
"epoch": 0.11893656716417911,
"grad_norm": 1.2197325105359402,
"learning_rate": 4.942367351052481e-05,
"loss": 0.7086,
"step": 255
},
{
"epoch": 0.12126865671641791,
"grad_norm": 0.8299444059232342,
"learning_rate": 4.938398377391461e-05,
"loss": 0.7187,
"step": 260
},
{
"epoch": 0.12360074626865672,
"grad_norm": 0.9224154095712178,
"learning_rate": 4.9342991445214334e-05,
"loss": 0.7126,
"step": 265
},
{
"epoch": 0.1259328358208955,
"grad_norm": 0.88090785003352,
"learning_rate": 4.930069896439485e-05,
"loss": 0.7105,
"step": 270
},
{
"epoch": 0.12826492537313433,
"grad_norm": 0.859476692536741,
"learning_rate": 4.925710884881551e-05,
"loss": 0.7071,
"step": 275
},
{
"epoch": 0.13059701492537312,
"grad_norm": 1.013753901988807,
"learning_rate": 4.921222369307427e-05,
"loss": 0.7205,
"step": 280
},
{
"epoch": 0.13292910447761194,
"grad_norm": 0.940182496551757,
"learning_rate": 4.916604616885328e-05,
"loss": 0.7139,
"step": 285
},
{
"epoch": 0.13526119402985073,
"grad_norm": 1.0259898104670397,
"learning_rate": 4.9118579024759854e-05,
"loss": 0.7194,
"step": 290
},
{
"epoch": 0.13759328358208955,
"grad_norm": 0.941096247874784,
"learning_rate": 4.9069825086162865e-05,
"loss": 0.694,
"step": 295
},
{
"epoch": 0.13992537313432835,
"grad_norm": 0.8605571897927781,
"learning_rate": 4.901978725502454e-05,
"loss": 0.7021,
"step": 300
},
{
"epoch": 0.14225746268656717,
"grad_norm": 1.012855599597888,
"learning_rate": 4.89684685097278e-05,
"loss": 0.6889,
"step": 305
},
{
"epoch": 0.14458955223880596,
"grad_norm": 0.9877119462322568,
"learning_rate": 4.891587190489891e-05,
"loss": 0.7154,
"step": 310
},
{
"epoch": 0.14692164179104478,
"grad_norm": 1.090755921619965,
"learning_rate": 4.886200057122568e-05,
"loss": 0.7195,
"step": 315
},
{
"epoch": 0.14925373134328357,
"grad_norm": 0.9811029143694675,
"learning_rate": 4.880685771527114e-05,
"loss": 0.7055,
"step": 320
},
{
"epoch": 0.1515858208955224,
"grad_norm": 0.8841142544985066,
"learning_rate": 4.8750446619282646e-05,
"loss": 0.7029,
"step": 325
},
{
"epoch": 0.15391791044776118,
"grad_norm": 0.8332150176518437,
"learning_rate": 4.869277064099654e-05,
"loss": 0.7093,
"step": 330
},
{
"epoch": 0.15625,
"grad_norm": 0.8744887046543461,
"learning_rate": 4.8633833213438254e-05,
"loss": 0.7065,
"step": 335
},
{
"epoch": 0.15858208955223882,
"grad_norm": 0.837397759622329,
"learning_rate": 4.8573637844718e-05,
"loss": 0.7325,
"step": 340
},
{
"epoch": 0.1609141791044776,
"grad_norm": 0.9698882033094054,
"learning_rate": 4.851218811782195e-05,
"loss": 0.7209,
"step": 345
},
{
"epoch": 0.16324626865671643,
"grad_norm": 0.8572480215660884,
"learning_rate": 4.844948769039896e-05,
"loss": 0.7117,
"step": 350
},
{
"epoch": 0.16557835820895522,
"grad_norm": 0.8647940595338452,
"learning_rate": 4.8385540294542855e-05,
"loss": 0.7059,
"step": 355
},
{
"epoch": 0.16791044776119404,
"grad_norm": 0.863764155230387,
"learning_rate": 4.83203497365703e-05,
"loss": 0.7114,
"step": 360
},
{
"epoch": 0.17024253731343283,
"grad_norm": 0.8063210763659003,
"learning_rate": 4.825391989679422e-05,
"loss": 0.7034,
"step": 365
},
{
"epoch": 0.17257462686567165,
"grad_norm": 0.8575257639465484,
"learning_rate": 4.818625472929286e-05,
"loss": 0.6947,
"step": 370
},
{
"epoch": 0.17490671641791045,
"grad_norm": 0.8200838613816108,
"learning_rate": 4.811735826167436e-05,
"loss": 0.681,
"step": 375
},
{
"epoch": 0.17723880597014927,
"grad_norm": 0.7823498981274288,
"learning_rate": 4.8047234594837143e-05,
"loss": 0.6958,
"step": 380
},
{
"epoch": 0.17957089552238806,
"grad_norm": 0.8190018216789188,
"learning_rate": 4.7975887902725696e-05,
"loss": 0.6835,
"step": 385
},
{
"epoch": 0.18190298507462688,
"grad_norm": 0.9214403517298152,
"learning_rate": 4.7903322432082185e-05,
"loss": 0.6984,
"step": 390
},
{
"epoch": 0.18423507462686567,
"grad_norm": 0.8273600510500148,
"learning_rate": 4.78295425021937e-05,
"loss": 0.6753,
"step": 395
},
{
"epoch": 0.1865671641791045,
"grad_norm": 0.9257746216157634,
"learning_rate": 4.775455250463507e-05,
"loss": 0.7063,
"step": 400
},
{
"epoch": 0.18889925373134328,
"grad_norm": 0.8553399615781841,
"learning_rate": 4.767835690300759e-05,
"loss": 0.7031,
"step": 405
},
{
"epoch": 0.1912313432835821,
"grad_norm": 0.8035273180959155,
"learning_rate": 4.760096023267322e-05,
"loss": 0.7098,
"step": 410
},
{
"epoch": 0.1935634328358209,
"grad_norm": 0.8691622488936168,
"learning_rate": 4.752236710048472e-05,
"loss": 0.6861,
"step": 415
},
{
"epoch": 0.1958955223880597,
"grad_norm": 0.9841920794382144,
"learning_rate": 4.744258218451135e-05,
"loss": 0.7031,
"step": 420
},
{
"epoch": 0.1982276119402985,
"grad_norm": 0.9498479171681536,
"learning_rate": 4.736161023376051e-05,
"loss": 0.6726,
"step": 425
},
{
"epoch": 0.20055970149253732,
"grad_norm": 0.8700794084478172,
"learning_rate": 4.7279456067895e-05,
"loss": 0.7153,
"step": 430
},
{
"epoch": 0.20289179104477612,
"grad_norm": 0.8127172667880727,
"learning_rate": 4.719612457694616e-05,
"loss": 0.6963,
"step": 435
},
{
"epoch": 0.20522388059701493,
"grad_norm": 0.7902973875049144,
"learning_rate": 4.71116207210228e-05,
"loss": 0.6773,
"step": 440
},
{
"epoch": 0.20755597014925373,
"grad_norm": 0.7788306399019677,
"learning_rate": 4.7025949530016e-05,
"loss": 0.6684,
"step": 445
},
{
"epoch": 0.20988805970149255,
"grad_norm": 0.7944637728792737,
"learning_rate": 4.6939116103299655e-05,
"loss": 0.6905,
"step": 450
},
{
"epoch": 0.21222014925373134,
"grad_norm": 0.8571602340727155,
"learning_rate": 4.685112560942699e-05,
"loss": 0.6881,
"step": 455
},
{
"epoch": 0.21455223880597016,
"grad_norm": 0.8780895287743348,
"learning_rate": 4.676198328582288e-05,
"loss": 0.6927,
"step": 460
},
{
"epoch": 0.21688432835820895,
"grad_norm": 0.9388682206384781,
"learning_rate": 4.6671694438472154e-05,
"loss": 0.7056,
"step": 465
},
{
"epoch": 0.21921641791044777,
"grad_norm": 0.8229322015185289,
"learning_rate": 4.6580264441603724e-05,
"loss": 0.6894,
"step": 470
},
{
"epoch": 0.22154850746268656,
"grad_norm": 0.9117054400216464,
"learning_rate": 4.648769873737071e-05,
"loss": 0.6972,
"step": 475
},
{
"epoch": 0.22388059701492538,
"grad_norm": 0.7774500802624019,
"learning_rate": 4.6394002835526535e-05,
"loss": 0.6969,
"step": 480
},
{
"epoch": 0.22621268656716417,
"grad_norm": 0.9974671450500393,
"learning_rate": 4.6299182313096916e-05,
"loss": 0.7168,
"step": 485
},
{
"epoch": 0.228544776119403,
"grad_norm": 0.7745214676810972,
"learning_rate": 4.6203242814047946e-05,
"loss": 0.6388,
"step": 490
},
{
"epoch": 0.23087686567164178,
"grad_norm": 0.7946997637352378,
"learning_rate": 4.610619004895017e-05,
"loss": 0.6932,
"step": 495
},
{
"epoch": 0.2332089552238806,
"grad_norm": 0.8376007460886711,
"learning_rate": 4.6008029794638596e-05,
"loss": 0.6626,
"step": 500
},
{
"epoch": 0.2355410447761194,
"grad_norm": 0.7750371283803145,
"learning_rate": 4.590876789386893e-05,
"loss": 0.6681,
"step": 505
},
{
"epoch": 0.23787313432835822,
"grad_norm": 0.8455168484121804,
"learning_rate": 4.580841025496974e-05,
"loss": 0.6786,
"step": 510
},
{
"epoch": 0.240205223880597,
"grad_norm": 0.7240530893309742,
"learning_rate": 4.570696285149084e-05,
"loss": 0.6825,
"step": 515
},
{
"epoch": 0.24253731343283583,
"grad_norm": 0.7536963649856001,
"learning_rate": 4.560443172184763e-05,
"loss": 0.6856,
"step": 520
},
{
"epoch": 0.24486940298507462,
"grad_norm": 0.7793205383613011,
"learning_rate": 4.55008229689618e-05,
"loss": 0.6667,
"step": 525
},
{
"epoch": 0.24720149253731344,
"grad_norm": 0.7157463633792304,
"learning_rate": 4.539614275989793e-05,
"loss": 0.6754,
"step": 530
},
{
"epoch": 0.24953358208955223,
"grad_norm": 0.7890577436144398,
"learning_rate": 4.529039732549653e-05,
"loss": 0.675,
"step": 535
},
{
"epoch": 0.251865671641791,
"grad_norm": 0.7846004685785093,
"learning_rate": 4.5183592960003104e-05,
"loss": 0.6635,
"step": 540
},
{
"epoch": 0.25419776119402987,
"grad_norm": 0.8071994271154869,
"learning_rate": 4.507573602069351e-05,
"loss": 0.6773,
"step": 545
},
{
"epoch": 0.25652985074626866,
"grad_norm": 0.8233537619340792,
"learning_rate": 4.496683292749555e-05,
"loss": 0.6719,
"step": 550
},
{
"epoch": 0.25886194029850745,
"grad_norm": 0.7715749039189433,
"learning_rate": 4.485689016260686e-05,
"loss": 0.6643,
"step": 555
},
{
"epoch": 0.26119402985074625,
"grad_norm": 0.7396872015907794,
"learning_rate": 4.4745914270109055e-05,
"loss": 0.6943,
"step": 560
},
{
"epoch": 0.2635261194029851,
"grad_norm": 0.7904102147449842,
"learning_rate": 4.463391185557822e-05,
"loss": 0.6819,
"step": 565
},
{
"epoch": 0.2658582089552239,
"grad_norm": 0.7165693906401905,
"learning_rate": 4.4520889585691705e-05,
"loss": 0.6865,
"step": 570
},
{
"epoch": 0.2681902985074627,
"grad_norm": 0.8804513560159245,
"learning_rate": 4.440685418783135e-05,
"loss": 0.6631,
"step": 575
},
{
"epoch": 0.27052238805970147,
"grad_norm": 0.7607710965405562,
"learning_rate": 4.429181244968301e-05,
"loss": 0.6699,
"step": 580
},
{
"epoch": 0.2728544776119403,
"grad_norm": 0.7620322976601013,
"learning_rate": 4.417577121883256e-05,
"loss": 0.669,
"step": 585
},
{
"epoch": 0.2751865671641791,
"grad_norm": 0.7328194761484526,
"learning_rate": 4.4058737402358295e-05,
"loss": 0.6488,
"step": 590
},
{
"epoch": 0.2775186567164179,
"grad_norm": 0.7634679687008787,
"learning_rate": 4.394071796641983e-05,
"loss": 0.6729,
"step": 595
},
{
"epoch": 0.2798507462686567,
"grad_norm": 0.8531001469934757,
"learning_rate": 4.38217199358434e-05,
"loss": 0.685,
"step": 600
},
{
"epoch": 0.28218283582089554,
"grad_norm": 0.7264481750331164,
"learning_rate": 4.3701750393703786e-05,
"loss": 0.6461,
"step": 605
},
{
"epoch": 0.28451492537313433,
"grad_norm": 0.7533827654944563,
"learning_rate": 4.3580816480902656e-05,
"loss": 0.6745,
"step": 610
},
{
"epoch": 0.2868470149253731,
"grad_norm": 0.7458225351946263,
"learning_rate": 4.345892539574359e-05,
"loss": 0.6831,
"step": 615
},
{
"epoch": 0.2891791044776119,
"grad_norm": 0.724365300513045,
"learning_rate": 4.3336084393503545e-05,
"loss": 0.6546,
"step": 620
},
{
"epoch": 0.29151119402985076,
"grad_norm": 0.7838535311430127,
"learning_rate": 4.3212300786001045e-05,
"loss": 0.6577,
"step": 625
},
{
"epoch": 0.29384328358208955,
"grad_norm": 0.7715960756766049,
"learning_rate": 4.308758194116094e-05,
"loss": 0.684,
"step": 630
},
{
"epoch": 0.29617537313432835,
"grad_norm": 0.7160624711169831,
"learning_rate": 4.296193528257586e-05,
"loss": 0.6893,
"step": 635
},
{
"epoch": 0.29850746268656714,
"grad_norm": 0.7745707050758446,
"learning_rate": 4.283536828906436e-05,
"loss": 0.6789,
"step": 640
},
{
"epoch": 0.300839552238806,
"grad_norm": 0.8251172241406618,
"learning_rate": 4.270788849422572e-05,
"loss": 0.6623,
"step": 645
},
{
"epoch": 0.3031716417910448,
"grad_norm": 0.8355188301078915,
"learning_rate": 4.2579503485991567e-05,
"loss": 0.6724,
"step": 650
},
{
"epoch": 0.30550373134328357,
"grad_norm": 0.8380841544851293,
"learning_rate": 4.245022090617418e-05,
"loss": 0.6765,
"step": 655
},
{
"epoch": 0.30783582089552236,
"grad_norm": 0.7456377128567805,
"learning_rate": 4.2320048450011684e-05,
"loss": 0.6741,
"step": 660
},
{
"epoch": 0.3101679104477612,
"grad_norm": 0.6884827740191439,
"learning_rate": 4.218899386570994e-05,
"loss": 0.6556,
"step": 665
},
{
"epoch": 0.3125,
"grad_norm": 0.6695176963355354,
"learning_rate": 4.205706495398143e-05,
"loss": 0.6824,
"step": 670
},
{
"epoch": 0.3148320895522388,
"grad_norm": 0.6518330480049027,
"learning_rate": 4.192426956758085e-05,
"loss": 0.6452,
"step": 675
},
{
"epoch": 0.31716417910447764,
"grad_norm": 0.7023106383456971,
"learning_rate": 4.179061561083777e-05,
"loss": 0.6573,
"step": 680
},
{
"epoch": 0.31949626865671643,
"grad_norm": 0.6796008979702891,
"learning_rate": 4.165611103918612e-05,
"loss": 0.6813,
"step": 685
},
{
"epoch": 0.3218283582089552,
"grad_norm": 0.7234212222083802,
"learning_rate": 4.1520763858690644e-05,
"loss": 0.6563,
"step": 690
},
{
"epoch": 0.324160447761194,
"grad_norm": 0.8232766133796315,
"learning_rate": 4.138458212557038e-05,
"loss": 0.6789,
"step": 695
},
{
"epoch": 0.32649253731343286,
"grad_norm": 0.892744295966519,
"learning_rate": 4.124757394571914e-05,
"loss": 0.6511,
"step": 700
},
{
"epoch": 0.32882462686567165,
"grad_norm": 0.7465253057638759,
"learning_rate": 4.110974747422299e-05,
"loss": 0.6609,
"step": 705
},
{
"epoch": 0.33115671641791045,
"grad_norm": 0.9710864864368132,
"learning_rate": 4.097111091487486e-05,
"loss": 0.6567,
"step": 710
},
{
"epoch": 0.33348880597014924,
"grad_norm": 0.7586336649039006,
"learning_rate": 4.083167251968625e-05,
"loss": 0.6868,
"step": 715
},
{
"epoch": 0.3358208955223881,
"grad_norm": 0.7411454559784593,
"learning_rate": 4.069144058839605e-05,
"loss": 0.6715,
"step": 720
},
{
"epoch": 0.3381529850746269,
"grad_norm": 0.7406787177182912,
"learning_rate": 4.055042346797643e-05,
"loss": 0.6655,
"step": 725
},
{
"epoch": 0.34048507462686567,
"grad_norm": 0.806810167383283,
"learning_rate": 4.040862955213615e-05,
"loss": 0.6792,
"step": 730
},
{
"epoch": 0.34281716417910446,
"grad_norm": 0.684339497670028,
"learning_rate": 4.026606728082082e-05,
"loss": 0.6643,
"step": 735
},
{
"epoch": 0.3451492537313433,
"grad_norm": 0.7645437701144697,
"learning_rate": 4.012274513971061e-05,
"loss": 0.6839,
"step": 740
},
{
"epoch": 0.3474813432835821,
"grad_norm": 0.6793956256373396,
"learning_rate": 3.997867165971512e-05,
"loss": 0.6507,
"step": 745
},
{
"epoch": 0.3498134328358209,
"grad_norm": 0.720694420633502,
"learning_rate": 3.9833855416465624e-05,
"loss": 0.6643,
"step": 750
},
{
"epoch": 0.3521455223880597,
"grad_norm": 0.7523164344936248,
"learning_rate": 3.968830502980459e-05,
"loss": 0.6591,
"step": 755
},
{
"epoch": 0.35447761194029853,
"grad_norm": 0.7805853823171353,
"learning_rate": 3.954202916327264e-05,
"loss": 0.6549,
"step": 760
},
{
"epoch": 0.3568097014925373,
"grad_norm": 0.7234820701835654,
"learning_rate": 3.939503652359287e-05,
"loss": 0.6572,
"step": 765
},
{
"epoch": 0.3591417910447761,
"grad_norm": 0.7688015369932021,
"learning_rate": 3.924733586015257e-05,
"loss": 0.6533,
"step": 770
},
{
"epoch": 0.3614738805970149,
"grad_norm": 0.708671520124952,
"learning_rate": 3.9098935964482476e-05,
"loss": 0.6551,
"step": 775
},
{
"epoch": 0.36380597014925375,
"grad_norm": 0.7183227720244021,
"learning_rate": 3.894984566973346e-05,
"loss": 0.6646,
"step": 780
},
{
"epoch": 0.36613805970149255,
"grad_norm": 0.6737145813609308,
"learning_rate": 3.880007385015075e-05,
"loss": 0.6511,
"step": 785
},
{
"epoch": 0.36847014925373134,
"grad_norm": 0.6920229984205842,
"learning_rate": 3.864962942054572e-05,
"loss": 0.6818,
"step": 790
},
{
"epoch": 0.37080223880597013,
"grad_norm": 0.775113876143003,
"learning_rate": 3.849852133576527e-05,
"loss": 0.6435,
"step": 795
},
{
"epoch": 0.373134328358209,
"grad_norm": 0.7394720394307068,
"learning_rate": 3.834675859015876e-05,
"loss": 0.671,
"step": 800
},
{
"epoch": 0.37546641791044777,
"grad_norm": 0.7324224028801083,
"learning_rate": 3.819435021704274e-05,
"loss": 0.66,
"step": 805
},
{
"epoch": 0.37779850746268656,
"grad_norm": 0.7482922307683427,
"learning_rate": 3.804130528816312e-05,
"loss": 0.6605,
"step": 810
},
{
"epoch": 0.38013059701492535,
"grad_norm": 0.693797019256493,
"learning_rate": 3.7887632913155355e-05,
"loss": 0.6459,
"step": 815
},
{
"epoch": 0.3824626865671642,
"grad_norm": 0.7332686963960172,
"learning_rate": 3.77333422390021e-05,
"loss": 0.6386,
"step": 820
},
{
"epoch": 0.384794776119403,
"grad_norm": 0.6722013452671626,
"learning_rate": 3.75784424494888e-05,
"loss": 0.6509,
"step": 825
},
{
"epoch": 0.3871268656716418,
"grad_norm": 0.6844841017486335,
"learning_rate": 3.7422942764657054e-05,
"loss": 0.6624,
"step": 830
},
{
"epoch": 0.3894589552238806,
"grad_norm": 0.709365449407489,
"learning_rate": 3.726685244025578e-05,
"loss": 0.6662,
"step": 835
},
{
"epoch": 0.3917910447761194,
"grad_norm": 0.7295699828504463,
"learning_rate": 3.711018076719034e-05,
"loss": 0.6578,
"step": 840
},
{
"epoch": 0.3941231343283582,
"grad_norm": 0.7263503777736142,
"learning_rate": 3.695293707096947e-05,
"loss": 0.6708,
"step": 845
},
{
"epoch": 0.396455223880597,
"grad_norm": 0.7812017284771229,
"learning_rate": 3.679513071115025e-05,
"loss": 0.6617,
"step": 850
},
{
"epoch": 0.3987873134328358,
"grad_norm": 0.6765328713271164,
"learning_rate": 3.663677108078094e-05,
"loss": 0.6674,
"step": 855
},
{
"epoch": 0.40111940298507465,
"grad_norm": 0.6874658416764338,
"learning_rate": 3.647786760584194e-05,
"loss": 0.657,
"step": 860
},
{
"epoch": 0.40345149253731344,
"grad_norm": 0.6396850329707013,
"learning_rate": 3.6318429744684676e-05,
"loss": 0.6728,
"step": 865
},
{
"epoch": 0.40578358208955223,
"grad_norm": 0.7376511879685435,
"learning_rate": 3.615846698746869e-05,
"loss": 0.6626,
"step": 870
},
{
"epoch": 0.408115671641791,
"grad_norm": 0.7037616699556043,
"learning_rate": 3.599798885559667e-05,
"loss": 0.677,
"step": 875
},
{
"epoch": 0.41044776119402987,
"grad_norm": 0.6519988831184151,
"learning_rate": 3.583700490114776e-05,
"loss": 0.6385,
"step": 880
},
{
"epoch": 0.41277985074626866,
"grad_norm": 0.6836720339790674,
"learning_rate": 3.5675524706309014e-05,
"loss": 0.6397,
"step": 885
},
{
"epoch": 0.41511194029850745,
"grad_norm": 0.6577966791961214,
"learning_rate": 3.5513557882805e-05,
"loss": 0.6718,
"step": 890
},
{
"epoch": 0.41744402985074625,
"grad_norm": 0.6676911965920004,
"learning_rate": 3.5351114071325696e-05,
"loss": 0.6552,
"step": 895
},
{
"epoch": 0.4197761194029851,
"grad_norm": 0.6775036363841046,
"learning_rate": 3.518820294095267e-05,
"loss": 0.66,
"step": 900
},
{
"epoch": 0.4221082089552239,
"grad_norm": 0.7274179785465997,
"learning_rate": 3.50248341885835e-05,
"loss": 0.6457,
"step": 905
},
{
"epoch": 0.4244402985074627,
"grad_norm": 0.6365285914594158,
"learning_rate": 3.486101753835468e-05,
"loss": 0.6549,
"step": 910
},
{
"epoch": 0.42677238805970147,
"grad_norm": 0.7015328692797856,
"learning_rate": 3.469676274106271e-05,
"loss": 0.6446,
"step": 915
},
{
"epoch": 0.4291044776119403,
"grad_norm": 0.6897225314313983,
"learning_rate": 3.453207957358377e-05,
"loss": 0.6481,
"step": 920
},
{
"epoch": 0.4314365671641791,
"grad_norm": 0.7018612270350465,
"learning_rate": 3.436697783829178e-05,
"loss": 0.6421,
"step": 925
},
{
"epoch": 0.4337686567164179,
"grad_norm": 0.683607419776046,
"learning_rate": 3.420146736247487e-05,
"loss": 0.6262,
"step": 930
},
{
"epoch": 0.4361007462686567,
"grad_norm": 0.6817625776138896,
"learning_rate": 3.4035557997750506e-05,
"loss": 0.6655,
"step": 935
},
{
"epoch": 0.43843283582089554,
"grad_norm": 0.7113749823554164,
"learning_rate": 3.386925961947906e-05,
"loss": 0.6589,
"step": 940
},
{
"epoch": 0.44076492537313433,
"grad_norm": 0.701429407427016,
"learning_rate": 3.370258212617602e-05,
"loss": 0.6756,
"step": 945
},
{
"epoch": 0.4430970149253731,
"grad_norm": 0.6831638841691235,
"learning_rate": 3.353553543892277e-05,
"loss": 0.645,
"step": 950
},
{
"epoch": 0.4454291044776119,
"grad_norm": 0.6951969526491697,
"learning_rate": 3.336812950077611e-05,
"loss": 0.653,
"step": 955
},
{
"epoch": 0.44776119402985076,
"grad_norm": 0.712881849067669,
"learning_rate": 3.320037427617639e-05,
"loss": 0.6503,
"step": 960
},
{
"epoch": 0.45009328358208955,
"grad_norm": 0.677080067582334,
"learning_rate": 3.30322797503544e-05,
"loss": 0.6634,
"step": 965
},
{
"epoch": 0.45242537313432835,
"grad_norm": 0.673026887857187,
"learning_rate": 3.2863855928737026e-05,
"loss": 0.6299,
"step": 970
},
{
"epoch": 0.45475746268656714,
"grad_norm": 0.6976163186906711,
"learning_rate": 3.2695112836351703e-05,
"loss": 0.6514,
"step": 975
},
{
"epoch": 0.457089552238806,
"grad_norm": 0.6835684109054025,
"learning_rate": 3.252606051722972e-05,
"loss": 0.6715,
"step": 980
},
{
"epoch": 0.4594216417910448,
"grad_norm": 0.634542987869755,
"learning_rate": 3.235670903380832e-05,
"loss": 0.6452,
"step": 985
},
{
"epoch": 0.46175373134328357,
"grad_norm": 0.6700081127629496,
"learning_rate": 3.218706846633183e-05,
"loss": 0.6424,
"step": 990
},
{
"epoch": 0.46408582089552236,
"grad_norm": 0.674844701562939,
"learning_rate": 3.201714891225156e-05,
"loss": 0.64,
"step": 995
},
{
"epoch": 0.4664179104477612,
"grad_norm": 0.6901945362990005,
"learning_rate": 3.1846960485624886e-05,
"loss": 0.6359,
"step": 1000
},
{
"epoch": 0.46875,
"grad_norm": 0.711302010737976,
"learning_rate": 3.1676513316513156e-05,
"loss": 0.6613,
"step": 1005
},
{
"epoch": 0.4710820895522388,
"grad_norm": 0.6415681002352712,
"learning_rate": 3.150581755037877e-05,
"loss": 0.6448,
"step": 1010
},
{
"epoch": 0.47341417910447764,
"grad_norm": 0.6730615759692896,
"learning_rate": 3.133488334748125e-05,
"loss": 0.6569,
"step": 1015
},
{
"epoch": 0.47574626865671643,
"grad_norm": 0.6995346562904475,
"learning_rate": 3.1163720882272516e-05,
"loss": 0.637,
"step": 1020
},
{
"epoch": 0.4780783582089552,
"grad_norm": 0.7048313864139962,
"learning_rate": 3.0992340342791246e-05,
"loss": 0.6496,
"step": 1025
},
{
"epoch": 0.480410447761194,
"grad_norm": 0.6599686029534936,
"learning_rate": 3.08207519300565e-05,
"loss": 0.642,
"step": 1030
},
{
"epoch": 0.48274253731343286,
"grad_norm": 0.676508421822385,
"learning_rate": 3.064896585746045e-05,
"loss": 0.6307,
"step": 1035
},
{
"epoch": 0.48507462686567165,
"grad_norm": 0.6701675191331543,
"learning_rate": 3.047699235016056e-05,
"loss": 0.6384,
"step": 1040
},
{
"epoch": 0.48740671641791045,
"grad_norm": 0.7258698708508784,
"learning_rate": 3.030484164447085e-05,
"loss": 0.6283,
"step": 1045
},
{
"epoch": 0.48973880597014924,
"grad_norm": 0.7407558467480371,
"learning_rate": 3.0132523987252658e-05,
"loss": 0.6495,
"step": 1050
},
{
"epoch": 0.4920708955223881,
"grad_norm": 0.689449626576069,
"learning_rate": 2.9960049635304755e-05,
"loss": 0.6219,
"step": 1055
},
{
"epoch": 0.4944029850746269,
"grad_norm": 0.6605235489630851,
"learning_rate": 2.9787428854752736e-05,
"loss": 0.6328,
"step": 1060
},
{
"epoch": 0.49673507462686567,
"grad_norm": 0.6521941117929255,
"learning_rate": 2.961467192043807e-05,
"loss": 0.6445,
"step": 1065
},
{
"epoch": 0.49906716417910446,
"grad_norm": 0.719415886577355,
"learning_rate": 2.9441789115306402e-05,
"loss": 0.6601,
"step": 1070
},
{
"epoch": 0.5013992537313433,
"grad_norm": 0.7143547584326713,
"learning_rate": 2.926879072979558e-05,
"loss": 0.6477,
"step": 1075
},
{
"epoch": 0.503731343283582,
"grad_norm": 0.6805215530249611,
"learning_rate": 2.9095687061223058e-05,
"loss": 0.6456,
"step": 1080
},
{
"epoch": 0.5060634328358209,
"grad_norm": 0.6951914121970127,
"learning_rate": 2.8922488413173053e-05,
"loss": 0.6441,
"step": 1085
},
{
"epoch": 0.5083955223880597,
"grad_norm": 0.706237459032283,
"learning_rate": 2.874920509488319e-05,
"loss": 0.6574,
"step": 1090
},
{
"epoch": 0.5107276119402985,
"grad_norm": 0.7346760125128584,
"learning_rate": 2.8575847420630887e-05,
"loss": 0.6314,
"step": 1095
},
{
"epoch": 0.5130597014925373,
"grad_norm": 0.6542912159857234,
"learning_rate": 2.8402425709119435e-05,
"loss": 0.659,
"step": 1100
},
{
"epoch": 0.5153917910447762,
"grad_norm": 0.6819702914670763,
"learning_rate": 2.8228950282863776e-05,
"loss": 0.6492,
"step": 1105
},
{
"epoch": 0.5177238805970149,
"grad_norm": 0.6770174696103765,
"learning_rate": 2.8055431467576106e-05,
"loss": 0.638,
"step": 1110
},
{
"epoch": 0.5200559701492538,
"grad_norm": 0.662818262088358,
"learning_rate": 2.788187959155124e-05,
"loss": 0.63,
"step": 1115
},
{
"epoch": 0.5223880597014925,
"grad_norm": 0.7072513898861241,
"learning_rate": 2.7708304985051868e-05,
"loss": 0.6623,
"step": 1120
},
{
"epoch": 0.5247201492537313,
"grad_norm": 0.711447160492985,
"learning_rate": 2.7534717979693647e-05,
"loss": 0.6344,
"step": 1125
},
{
"epoch": 0.5270522388059702,
"grad_norm": 0.6918500244833096,
"learning_rate": 2.7361128907830253e-05,
"loss": 0.6385,
"step": 1130
},
{
"epoch": 0.5293843283582089,
"grad_norm": 0.6722193101035042,
"learning_rate": 2.7187548101938353e-05,
"loss": 0.6256,
"step": 1135
},
{
"epoch": 0.5317164179104478,
"grad_norm": 0.6569908241938373,
"learning_rate": 2.7013985894002623e-05,
"loss": 0.6236,
"step": 1140
},
{
"epoch": 0.5340485074626866,
"grad_norm": 0.695528007373876,
"learning_rate": 2.6840452614900726e-05,
"loss": 0.6533,
"step": 1145
},
{
"epoch": 0.5363805970149254,
"grad_norm": 0.6182593924769839,
"learning_rate": 2.6666958593788405e-05,
"loss": 0.631,
"step": 1150
},
{
"epoch": 0.5387126865671642,
"grad_norm": 0.6254875427778447,
"learning_rate": 2.649351415748466e-05,
"loss": 0.6528,
"step": 1155
},
{
"epoch": 0.5410447761194029,
"grad_norm": 0.6195472352586545,
"learning_rate": 2.6320129629857093e-05,
"loss": 0.6421,
"step": 1160
},
{
"epoch": 0.5433768656716418,
"grad_norm": 0.6406233353457038,
"learning_rate": 2.6146815331207358e-05,
"loss": 0.6183,
"step": 1165
},
{
"epoch": 0.5457089552238806,
"grad_norm": 0.6456118637481933,
"learning_rate": 2.597358157765692e-05,
"loss": 0.6474,
"step": 1170
},
{
"epoch": 0.5480410447761194,
"grad_norm": 0.6505557028284191,
"learning_rate": 2.5800438680532974e-05,
"loss": 0.6307,
"step": 1175
},
{
"epoch": 0.5503731343283582,
"grad_norm": 0.6673856033395356,
"learning_rate": 2.56273969457547e-05,
"loss": 0.6305,
"step": 1180
},
{
"epoch": 0.5527052238805971,
"grad_norm": 0.6976453369750254,
"learning_rate": 2.545446667321984e-05,
"loss": 0.6511,
"step": 1185
},
{
"epoch": 0.5550373134328358,
"grad_norm": 0.6523315253145765,
"learning_rate": 2.528165815619162e-05,
"loss": 0.6326,
"step": 1190
},
{
"epoch": 0.5573694029850746,
"grad_norm": 0.6513887450128372,
"learning_rate": 2.5108981680686035e-05,
"loss": 0.6357,
"step": 1195
},
{
"epoch": 0.5597014925373134,
"grad_norm": 0.6686846233513335,
"learning_rate": 2.4936447524859625e-05,
"loss": 0.6243,
"step": 1200
},
{
"epoch": 0.5620335820895522,
"grad_norm": 0.6254057596195436,
"learning_rate": 2.4764065958397715e-05,
"loss": 0.6273,
"step": 1205
},
{
"epoch": 0.5643656716417911,
"grad_norm": 0.6496921416294746,
"learning_rate": 2.459184724190308e-05,
"loss": 0.6179,
"step": 1210
},
{
"epoch": 0.5666977611940298,
"grad_norm": 0.6568624775107711,
"learning_rate": 2.441980162628527e-05,
"loss": 0.6344,
"step": 1215
},
{
"epoch": 0.5690298507462687,
"grad_norm": 0.6567469938905354,
"learning_rate": 2.4247939352150386e-05,
"loss": 0.6449,
"step": 1220
},
{
"epoch": 0.5713619402985075,
"grad_norm": 0.61556362280737,
"learning_rate": 2.4076270649191573e-05,
"loss": 0.6555,
"step": 1225
},
{
"epoch": 0.5736940298507462,
"grad_norm": 0.6394081674254839,
"learning_rate": 2.390480573558012e-05,
"loss": 0.6363,
"step": 1230
},
{
"epoch": 0.5760261194029851,
"grad_norm": 0.6051477641244504,
"learning_rate": 2.3733554817357246e-05,
"loss": 0.6348,
"step": 1235
},
{
"epoch": 0.5783582089552238,
"grad_norm": 0.5930962238087756,
"learning_rate": 2.3562528087826573e-05,
"loss": 0.6206,
"step": 1240
},
{
"epoch": 0.5806902985074627,
"grad_norm": 0.698655785577389,
"learning_rate": 2.339173572694746e-05,
"loss": 0.6273,
"step": 1245
},
{
"epoch": 0.5830223880597015,
"grad_norm": 0.6327913417184158,
"learning_rate": 2.3221187900729003e-05,
"loss": 0.637,
"step": 1250
},
{
"epoch": 0.5853544776119403,
"grad_norm": 0.6339231296671493,
"learning_rate": 2.3050894760624982e-05,
"loss": 0.641,
"step": 1255
},
{
"epoch": 0.5876865671641791,
"grad_norm": 0.6051505920325168,
"learning_rate": 2.2880866442929544e-05,
"loss": 0.6265,
"step": 1260
},
{
"epoch": 0.590018656716418,
"grad_norm": 0.6637235396372423,
"learning_rate": 2.271111306817396e-05,
"loss": 0.6482,
"step": 1265
},
{
"epoch": 0.5923507462686567,
"grad_norm": 0.6469789610641435,
"learning_rate": 2.254164474052416e-05,
"loss": 0.6266,
"step": 1270
},
{
"epoch": 0.5946828358208955,
"grad_norm": 0.5897603262670164,
"learning_rate": 2.237247154717932e-05,
"loss": 0.6188,
"step": 1275
},
{
"epoch": 0.5970149253731343,
"grad_norm": 0.6115820772990138,
"learning_rate": 2.2203603557771447e-05,
"loss": 0.6198,
"step": 1280
},
{
"epoch": 0.5993470149253731,
"grad_norm": 0.6159243257321477,
"learning_rate": 2.2035050823766008e-05,
"loss": 0.6313,
"step": 1285
},
{
"epoch": 0.601679104477612,
"grad_norm": 0.6683096146103435,
"learning_rate": 2.186682337786365e-05,
"loss": 0.6369,
"step": 1290
},
{
"epoch": 0.6040111940298507,
"grad_norm": 0.603613728907185,
"learning_rate": 2.1698931233403013e-05,
"loss": 0.6151,
"step": 1295
},
{
"epoch": 0.6063432835820896,
"grad_norm": 0.6004781166951644,
"learning_rate": 2.153138438376473e-05,
"loss": 0.622,
"step": 1300
},
{
"epoch": 0.6086753731343284,
"grad_norm": 18.22261692025831,
"learning_rate": 2.136419280177655e-05,
"loss": 0.6161,
"step": 1305
},
{
"epoch": 0.6110074626865671,
"grad_norm": 0.5978990585115841,
"learning_rate": 2.119736643911979e-05,
"loss": 0.6416,
"step": 1310
},
{
"epoch": 0.613339552238806,
"grad_norm": 0.597454038032284,
"learning_rate": 2.1030915225736947e-05,
"loss": 0.6288,
"step": 1315
},
{
"epoch": 0.6156716417910447,
"grad_norm": 0.6430793847859734,
"learning_rate": 2.0864849069240645e-05,
"loss": 0.6383,
"step": 1320
},
{
"epoch": 0.6180037313432836,
"grad_norm": 0.5831365581687563,
"learning_rate": 2.0699177854323902e-05,
"loss": 0.6295,
"step": 1325
},
{
"epoch": 0.6203358208955224,
"grad_norm": 0.5998744003452269,
"learning_rate": 2.0533911442171805e-05,
"loss": 0.6287,
"step": 1330
},
{
"epoch": 0.6226679104477612,
"grad_norm": 0.6643837943700802,
"learning_rate": 2.036905966987449e-05,
"loss": 0.6343,
"step": 1335
},
{
"epoch": 0.625,
"grad_norm": 0.6438429216570205,
"learning_rate": 2.0204632349841667e-05,
"loss": 0.616,
"step": 1340
},
{
"epoch": 0.6273320895522388,
"grad_norm": 0.6110144914242343,
"learning_rate": 2.0040639269218532e-05,
"loss": 0.6212,
"step": 1345
},
{
"epoch": 0.6296641791044776,
"grad_norm": 0.6615405256014992,
"learning_rate": 1.9877090189303182e-05,
"loss": 0.6463,
"step": 1350
},
{
"epoch": 0.6319962686567164,
"grad_norm": 0.6183940678157902,
"learning_rate": 1.9713994844965657e-05,
"loss": 0.6183,
"step": 1355
},
{
"epoch": 0.6343283582089553,
"grad_norm": 0.6746036398298787,
"learning_rate": 1.9551362944068462e-05,
"loss": 0.6301,
"step": 1360
},
{
"epoch": 0.636660447761194,
"grad_norm": 0.648307373321323,
"learning_rate": 1.938920416688874e-05,
"loss": 0.6308,
"step": 1365
},
{
"epoch": 0.6389925373134329,
"grad_norm": 0.6069645644995993,
"learning_rate": 1.922752816554204e-05,
"loss": 0.6205,
"step": 1370
},
{
"epoch": 0.6413246268656716,
"grad_norm": 0.5822850576614598,
"learning_rate": 1.9066344563407856e-05,
"loss": 0.6295,
"step": 1375
},
{
"epoch": 0.6436567164179104,
"grad_norm": 0.6496746526806308,
"learning_rate": 1.890566295455678e-05,
"loss": 0.605,
"step": 1380
},
{
"epoch": 0.6459888059701493,
"grad_norm": 0.5866035567882427,
"learning_rate": 1.874549290317946e-05,
"loss": 0.608,
"step": 1385
},
{
"epoch": 0.648320895522388,
"grad_norm": 0.6041775047865613,
"learning_rate": 1.858584394301728e-05,
"loss": 0.6356,
"step": 1390
},
{
"epoch": 0.6506529850746269,
"grad_norm": 0.5932846773816811,
"learning_rate": 1.8426725576794918e-05,
"loss": 0.6576,
"step": 1395
},
{
"epoch": 0.6529850746268657,
"grad_norm": 0.6021211057686333,
"learning_rate": 1.8268147275654707e-05,
"loss": 0.6276,
"step": 1400
},
{
"epoch": 0.6553171641791045,
"grad_norm": 0.5706056220756198,
"learning_rate": 1.8110118478592915e-05,
"loss": 0.6129,
"step": 1405
},
{
"epoch": 0.6576492537313433,
"grad_norm": 0.6303230581411896,
"learning_rate": 1.7952648591897858e-05,
"loss": 0.6296,
"step": 1410
},
{
"epoch": 0.659981343283582,
"grad_norm": 0.6069553160577528,
"learning_rate": 1.7795746988590027e-05,
"loss": 0.6162,
"step": 1415
},
{
"epoch": 0.6623134328358209,
"grad_norm": 0.6233052721164215,
"learning_rate": 1.7639423007864252e-05,
"loss": 0.6071,
"step": 1420
},
{
"epoch": 0.6646455223880597,
"grad_norm": 0.5792817379292262,
"learning_rate": 1.7483685954533692e-05,
"loss": 0.6186,
"step": 1425
},
{
"epoch": 0.6669776119402985,
"grad_norm": 0.598455839355501,
"learning_rate": 1.7328545098476106e-05,
"loss": 0.6126,
"step": 1430
},
{
"epoch": 0.6693097014925373,
"grad_norm": 0.5887965955039507,
"learning_rate": 1.717400967408196e-05,
"loss": 0.6219,
"step": 1435
},
{
"epoch": 0.6716417910447762,
"grad_norm": 0.6037531009113533,
"learning_rate": 1.702008887970491e-05,
"loss": 0.6164,
"step": 1440
},
{
"epoch": 0.6739738805970149,
"grad_norm": 0.5608388614868615,
"learning_rate": 1.6866791877114165e-05,
"loss": 0.6246,
"step": 1445
},
{
"epoch": 0.6763059701492538,
"grad_norm": 0.5732233279593167,
"learning_rate": 1.671412779094926e-05,
"loss": 0.622,
"step": 1450
},
{
"epoch": 0.6786380597014925,
"grad_norm": 0.604276726888298,
"learning_rate": 1.656210570817685e-05,
"loss": 0.6227,
"step": 1455
},
{
"epoch": 0.6809701492537313,
"grad_norm": 0.6084928698482973,
"learning_rate": 1.6410734677549872e-05,
"loss": 0.6156,
"step": 1460
},
{
"epoch": 0.6833022388059702,
"grad_norm": 0.6167514863041356,
"learning_rate": 1.6260023709068932e-05,
"loss": 0.622,
"step": 1465
},
{
"epoch": 0.6856343283582089,
"grad_norm": 0.6360423400540859,
"learning_rate": 1.6109981773446036e-05,
"loss": 0.6183,
"step": 1470
},
{
"epoch": 0.6879664179104478,
"grad_norm": 0.6092909418703362,
"learning_rate": 1.5960617801570555e-05,
"loss": 0.6162,
"step": 1475
},
{
"epoch": 0.6902985074626866,
"grad_norm": 0.6033346353614264,
"learning_rate": 1.58119406839777e-05,
"loss": 0.6221,
"step": 1480
},
{
"epoch": 0.6926305970149254,
"grad_norm": 0.5497901412692066,
"learning_rate": 1.566395927031932e-05,
"loss": 0.6094,
"step": 1485
},
{
"epoch": 0.6949626865671642,
"grad_norm": 0.5726565331606571,
"learning_rate": 1.5516682368837133e-05,
"loss": 0.6165,
"step": 1490
},
{
"epoch": 0.6972947761194029,
"grad_norm": 0.6379827482434526,
"learning_rate": 1.5370118745838453e-05,
"loss": 0.6404,
"step": 1495
},
{
"epoch": 0.6996268656716418,
"grad_norm": 0.6035346967740655,
"learning_rate": 1.5224277125174388e-05,
"loss": 0.6238,
"step": 1500
},
{
"epoch": 0.7019589552238806,
"grad_norm": 0.6046299493263414,
"learning_rate": 1.5079166187720561e-05,
"loss": 0.6233,
"step": 1505
},
{
"epoch": 0.7042910447761194,
"grad_norm": 0.5813425376508258,
"learning_rate": 1.4934794570860416e-05,
"loss": 0.6406,
"step": 1510
},
{
"epoch": 0.7066231343283582,
"grad_norm": 0.6005101213362166,
"learning_rate": 1.4791170867971132e-05,
"loss": 0.6167,
"step": 1515
},
{
"epoch": 0.7089552238805971,
"grad_norm": 0.5968043199699391,
"learning_rate": 1.464830362791204e-05,
"loss": 0.6154,
"step": 1520
},
{
"epoch": 0.7112873134328358,
"grad_norm": 0.585701695377086,
"learning_rate": 1.450620135451585e-05,
"loss": 0.6225,
"step": 1525
},
{
"epoch": 0.7136194029850746,
"grad_norm": 0.5957247172214722,
"learning_rate": 1.4364872506082425e-05,
"loss": 0.6282,
"step": 1530
},
{
"epoch": 0.7159514925373134,
"grad_norm": 0.5967837516043032,
"learning_rate": 1.4224325494875385e-05,
"loss": 0.6399,
"step": 1535
},
{
"epoch": 0.7182835820895522,
"grad_norm": 0.587594664717925,
"learning_rate": 1.4084568686621314e-05,
"loss": 0.6201,
"step": 1540
},
{
"epoch": 0.7206156716417911,
"grad_norm": 0.5617516377764582,
"learning_rate": 1.3945610400011851e-05,
"loss": 0.633,
"step": 1545
},
{
"epoch": 0.7229477611940298,
"grad_norm": 0.6036131494081576,
"learning_rate": 1.3807458906208546e-05,
"loss": 0.6421,
"step": 1550
},
{
"epoch": 0.7252798507462687,
"grad_norm": 0.5619681271472718,
"learning_rate": 1.3670122428350521e-05,
"loss": 0.618,
"step": 1555
},
{
"epoch": 0.7276119402985075,
"grad_norm": 0.5870654465012264,
"learning_rate": 1.3533609141065008e-05,
"loss": 0.6101,
"step": 1560
},
{
"epoch": 0.7299440298507462,
"grad_norm": 0.6499759452581642,
"learning_rate": 1.3397927169980773e-05,
"loss": 0.6161,
"step": 1565
},
{
"epoch": 0.7322761194029851,
"grad_norm": 0.5802369273091457,
"learning_rate": 1.326308459124447e-05,
"loss": 0.6156,
"step": 1570
},
{
"epoch": 0.7346082089552238,
"grad_norm": 0.5862527640011228,
"learning_rate": 1.3129089431039931e-05,
"loss": 0.6218,
"step": 1575
},
{
"epoch": 0.7369402985074627,
"grad_norm": 0.6041233460220358,
"learning_rate": 1.299594966511038e-05,
"loss": 0.6344,
"step": 1580
},
{
"epoch": 0.7392723880597015,
"grad_norm": 0.5761625096468084,
"learning_rate": 1.2863673218283783e-05,
"loss": 0.617,
"step": 1585
},
{
"epoch": 0.7416044776119403,
"grad_norm": 0.6001779302881813,
"learning_rate": 1.2732267964001033e-05,
"loss": 0.6101,
"step": 1590
},
{
"epoch": 0.7439365671641791,
"grad_norm": 0.5964143297371681,
"learning_rate": 1.26017417238474e-05,
"loss": 0.6212,
"step": 1595
},
{
"epoch": 0.746268656716418,
"grad_norm": 0.5964222072799055,
"learning_rate": 1.2472102267086904e-05,
"loss": 0.6211,
"step": 1600
},
{
"epoch": 0.7486007462686567,
"grad_norm": 0.6476516452528284,
"learning_rate": 1.2343357310199925e-05,
"loss": 0.6389,
"step": 1605
},
{
"epoch": 0.7509328358208955,
"grad_norm": 0.6211226051748449,
"learning_rate": 1.2215514516423813e-05,
"loss": 0.6345,
"step": 1610
},
{
"epoch": 0.7532649253731343,
"grad_norm": 0.6009531989706945,
"learning_rate": 1.2088581495296852e-05,
"loss": 0.6106,
"step": 1615
},
{
"epoch": 0.7555970149253731,
"grad_norm": 0.5457398269496171,
"learning_rate": 1.1962565802205255e-05,
"loss": 0.6107,
"step": 1620
},
{
"epoch": 0.757929104477612,
"grad_norm": 0.5485455623765262,
"learning_rate": 1.1837474937933464e-05,
"loss": 0.6194,
"step": 1625
},
{
"epoch": 0.7602611940298507,
"grad_norm": 0.5644743687543846,
"learning_rate": 1.1713316348217673e-05,
"loss": 0.6136,
"step": 1630
},
{
"epoch": 0.7625932835820896,
"grad_norm": 0.5944262234833658,
"learning_rate": 1.1590097423302684e-05,
"loss": 0.5942,
"step": 1635
},
{
"epoch": 0.7649253731343284,
"grad_norm": 0.5616590097486719,
"learning_rate": 1.1467825497501954e-05,
"loss": 0.6026,
"step": 1640
},
{
"epoch": 0.7672574626865671,
"grad_norm": 0.6146723556271906,
"learning_rate": 1.1346507848761077e-05,
"loss": 0.6156,
"step": 1645
},
{
"epoch": 0.769589552238806,
"grad_norm": 0.5965680064276904,
"learning_rate": 1.1226151698224597e-05,
"loss": 0.633,
"step": 1650
},
{
"epoch": 0.7719216417910447,
"grad_norm": 0.5614974770289655,
"learning_rate": 1.1106764209806127e-05,
"loss": 0.6194,
"step": 1655
},
{
"epoch": 0.7742537313432836,
"grad_norm": 0.6034383421575543,
"learning_rate": 1.0988352489762006e-05,
"loss": 0.6359,
"step": 1660
},
{
"epoch": 0.7765858208955224,
"grad_norm": 0.5568415876195435,
"learning_rate": 1.0870923586268245e-05,
"loss": 0.6143,
"step": 1665
},
{
"epoch": 0.7789179104477612,
"grad_norm": 0.5686287199324069,
"learning_rate": 1.0754484489001085e-05,
"loss": 0.6186,
"step": 1670
},
{
"epoch": 0.78125,
"grad_norm": 0.5739930134377701,
"learning_rate": 1.0639042128720847e-05,
"loss": 0.6235,
"step": 1675
},
{
"epoch": 0.7835820895522388,
"grad_norm": 0.5795603868856798,
"learning_rate": 1.052460337685951e-05,
"loss": 0.6049,
"step": 1680
},
{
"epoch": 0.7859141791044776,
"grad_norm": 0.56395308762799,
"learning_rate": 1.0411175045111602e-05,
"loss": 0.6032,
"step": 1685
},
{
"epoch": 0.7882462686567164,
"grad_norm": 0.58081174474857,
"learning_rate": 1.0298763885028839e-05,
"loss": 0.6178,
"step": 1690
},
{
"epoch": 0.7905783582089553,
"grad_norm": 0.5588983254320253,
"learning_rate": 1.018737658761817e-05,
"loss": 0.6105,
"step": 1695
},
{
"epoch": 0.792910447761194,
"grad_norm": 0.5364326303324629,
"learning_rate": 1.0077019782943584e-05,
"loss": 0.6237,
"step": 1700
},
{
"epoch": 0.7952425373134329,
"grad_norm": 0.5535399918900618,
"learning_rate": 9.967700039731427e-06,
"loss": 0.6129,
"step": 1705
},
{
"epoch": 0.7975746268656716,
"grad_norm": 0.5357340402470735,
"learning_rate": 9.859423864979441e-06,
"loss": 0.5947,
"step": 1710
},
{
"epoch": 0.7999067164179104,
"grad_norm": 0.6043622414636253,
"learning_rate": 9.752197703569422e-06,
"loss": 0.6341,
"step": 1715
},
{
"epoch": 0.8022388059701493,
"grad_norm": 0.5721045372713984,
"learning_rate": 9.646027937883622e-06,
"loss": 0.6161,
"step": 1720
},
{
"epoch": 0.804570895522388,
"grad_norm": 0.5312555490120068,
"learning_rate": 9.54092088742485e-06,
"loss": 0.64,
"step": 1725
},
{
"epoch": 0.8069029850746269,
"grad_norm": 0.5439729568974323,
"learning_rate": 9.436882808440334e-06,
"loss": 0.6165,
"step": 1730
},
{
"epoch": 0.8092350746268657,
"grad_norm": 0.5911843430242844,
"learning_rate": 9.333919893549294e-06,
"loss": 0.6099,
"step": 1735
},
{
"epoch": 0.8115671641791045,
"grad_norm": 0.5499239897213614,
"learning_rate": 9.232038271374377e-06,
"loss": 0.601,
"step": 1740
},
{
"epoch": 0.8138992537313433,
"grad_norm": 0.5917277497776632,
"learning_rate": 9.131244006176846e-06,
"loss": 0.6181,
"step": 1745
},
{
"epoch": 0.816231343283582,
"grad_norm": 0.5790530037092708,
"learning_rate": 9.031543097495638e-06,
"loss": 0.6033,
"step": 1750
},
{
"epoch": 0.8185634328358209,
"grad_norm": 0.556634899711528,
"learning_rate": 8.93294147979023e-06,
"loss": 0.5975,
"step": 1755
},
{
"epoch": 0.8208955223880597,
"grad_norm": 0.525574565943637,
"learning_rate": 8.835445022087426e-06,
"loss": 0.6194,
"step": 1760
},
{
"epoch": 0.8232276119402985,
"grad_norm": 0.542460572064194,
"learning_rate": 8.739059527631999e-06,
"loss": 0.6194,
"step": 1765
},
{
"epoch": 0.8255597014925373,
"grad_norm": 0.5961740451690235,
"learning_rate": 8.6437907335413e-06,
"loss": 0.6232,
"step": 1770
},
{
"epoch": 0.8278917910447762,
"grad_norm": 0.5156288871553688,
"learning_rate": 8.549644310463717e-06,
"loss": 0.625,
"step": 1775
},
{
"epoch": 0.8302238805970149,
"grad_norm": 0.5827304049610953,
"learning_rate": 8.456625862241193e-06,
"loss": 0.623,
"step": 1780
},
{
"epoch": 0.8325559701492538,
"grad_norm": 0.5473226394306203,
"learning_rate": 8.364740925575643e-06,
"loss": 0.6211,
"step": 1785
},
{
"epoch": 0.8348880597014925,
"grad_norm": 0.5663858177158172,
"learning_rate": 8.273994969699394e-06,
"loss": 0.6115,
"step": 1790
},
{
"epoch": 0.8372201492537313,
"grad_norm": 0.5667136142959207,
"learning_rate": 8.184393396049675e-06,
"loss": 0.6127,
"step": 1795
},
{
"epoch": 0.8395522388059702,
"grad_norm": 0.5603856211987771,
"learning_rate": 8.095941537947057e-06,
"loss": 0.6274,
"step": 1800
},
{
"epoch": 0.8418843283582089,
"grad_norm": 0.5594249365684848,
"learning_rate": 8.008644660278051e-06,
"loss": 0.5925,
"step": 1805
},
{
"epoch": 0.8442164179104478,
"grad_norm": 0.5635626874281433,
"learning_rate": 7.922507959181673e-06,
"loss": 0.6189,
"step": 1810
},
{
"epoch": 0.8465485074626866,
"grad_norm": 0.560313283634603,
"learning_rate": 7.837536561740225e-06,
"loss": 0.6097,
"step": 1815
},
{
"epoch": 0.8488805970149254,
"grad_norm": 0.6202519240674869,
"learning_rate": 7.753735525674059e-06,
"loss": 0.6174,
"step": 1820
},
{
"epoch": 0.8512126865671642,
"grad_norm": 0.5713772506301511,
"learning_rate": 7.671109839040547e-06,
"loss": 0.621,
"step": 1825
},
{
"epoch": 0.8535447761194029,
"grad_norm": 0.5237097428909497,
"learning_rate": 7.58966441993719e-06,
"loss": 0.5981,
"step": 1830
},
{
"epoch": 0.8558768656716418,
"grad_norm": 0.5483521318490681,
"learning_rate": 7.509404116208868e-06,
"loss": 0.6138,
"step": 1835
},
{
"epoch": 0.8582089552238806,
"grad_norm": 0.5778220550523655,
"learning_rate": 7.430333705159286e-06,
"loss": 0.6522,
"step": 1840
},
{
"epoch": 0.8605410447761194,
"grad_norm": 0.5486890952901237,
"learning_rate": 7.352457893266627e-06,
"loss": 0.6429,
"step": 1845
},
{
"epoch": 0.8628731343283582,
"grad_norm": 0.5693041316007504,
"learning_rate": 7.275781315903374e-06,
"loss": 0.6225,
"step": 1850
},
{
"epoch": 0.8652052238805971,
"grad_norm": 0.5594174594138699,
"learning_rate": 7.20030853706046e-06,
"loss": 0.6174,
"step": 1855
},
{
"epoch": 0.8675373134328358,
"grad_norm": 0.5408573140361542,
"learning_rate": 7.126044049075548e-06,
"loss": 0.6239,
"step": 1860
},
{
"epoch": 0.8698694029850746,
"grad_norm": 0.5625658917979015,
"learning_rate": 7.052992272365681e-06,
"loss": 0.6149,
"step": 1865
},
{
"epoch": 0.8722014925373134,
"grad_norm": 0.5795869183892941,
"learning_rate": 6.9811575551641224e-06,
"loss": 0.6014,
"step": 1870
},
{
"epoch": 0.8745335820895522,
"grad_norm": 0.5908193742703406,
"learning_rate": 6.910544173261588e-06,
"loss": 0.6047,
"step": 1875
},
{
"epoch": 0.8768656716417911,
"grad_norm": 0.5468229515625878,
"learning_rate": 6.8411563297516995e-06,
"loss": 0.5956,
"step": 1880
},
{
"epoch": 0.8791977611940298,
"grad_norm": 0.5746605159059298,
"learning_rate": 6.772998154780832e-06,
"loss": 0.612,
"step": 1885
},
{
"epoch": 0.8815298507462687,
"grad_norm": 0.5756230793946113,
"learning_rate": 6.706073705302254e-06,
"loss": 0.6145,
"step": 1890
},
{
"epoch": 0.8838619402985075,
"grad_norm": 0.5662725596320226,
"learning_rate": 6.6403869648346634e-06,
"loss": 0.6116,
"step": 1895
},
{
"epoch": 0.8861940298507462,
"grad_norm": 0.6638472478598192,
"learning_rate": 6.575941843225068e-06,
"loss": 0.5928,
"step": 1900
},
{
"epoch": 0.8885261194029851,
"grad_norm": 0.5392491951943847,
"learning_rate": 6.5127421764160685e-06,
"loss": 0.5994,
"step": 1905
},
{
"epoch": 0.8908582089552238,
"grad_norm": 0.5501297394583888,
"learning_rate": 6.450791726217538e-06,
"loss": 0.623,
"step": 1910
},
{
"epoch": 0.8931902985074627,
"grad_norm": 0.5561658935766048,
"learning_rate": 6.390094180082694e-06,
"loss": 0.6137,
"step": 1915
},
{
"epoch": 0.8955223880597015,
"grad_norm": 0.5486158970271028,
"learning_rate": 6.330653150888617e-06,
"loss": 0.6153,
"step": 1920
},
{
"epoch": 0.8978544776119403,
"grad_norm": 0.5624023243356174,
"learning_rate": 6.272472176721207e-06,
"loss": 0.6207,
"step": 1925
},
{
"epoch": 0.9001865671641791,
"grad_norm": 0.5469455949584299,
"learning_rate": 6.215554720664598e-06,
"loss": 0.6276,
"step": 1930
},
{
"epoch": 0.902518656716418,
"grad_norm": 0.5335851595481946,
"learning_rate": 6.159904170594982e-06,
"loss": 0.6151,
"step": 1935
},
{
"epoch": 0.9048507462686567,
"grad_norm": 0.5451933732807914,
"learning_rate": 6.105523838979022e-06,
"loss": 0.6302,
"step": 1940
},
{
"epoch": 0.9071828358208955,
"grad_norm": 0.5362545841210454,
"learning_rate": 6.052416962676621e-06,
"loss": 0.6058,
"step": 1945
},
{
"epoch": 0.9095149253731343,
"grad_norm": 0.5230617545385562,
"learning_rate": 6.000586702748301e-06,
"loss": 0.603,
"step": 1950
},
{
"epoch": 0.9118470149253731,
"grad_norm": 0.582021589520335,
"learning_rate": 5.950036144267021e-06,
"loss": 0.6111,
"step": 1955
},
{
"epoch": 0.914179104477612,
"grad_norm": 0.519728015454707,
"learning_rate": 5.900768296134551e-06,
"loss": 0.6049,
"step": 1960
},
{
"epoch": 0.9165111940298507,
"grad_norm": 0.592109342645659,
"learning_rate": 5.852786090902383e-06,
"loss": 0.6199,
"step": 1965
},
{
"epoch": 0.9188432835820896,
"grad_norm": 0.5517752644641193,
"learning_rate": 5.8060923845971825e-06,
"loss": 0.6113,
"step": 1970
},
{
"epoch": 0.9211753731343284,
"grad_norm": 0.5374604103547962,
"learning_rate": 5.760689956550763e-06,
"loss": 0.6051,
"step": 1975
},
{
"epoch": 0.9235074626865671,
"grad_norm": 0.5317691219584879,
"learning_rate": 5.7165815092346825e-06,
"loss": 0.5952,
"step": 1980
},
{
"epoch": 0.925839552238806,
"grad_norm": 0.5553747148697364,
"learning_rate": 5.673769668099364e-06,
"loss": 0.6088,
"step": 1985
},
{
"epoch": 0.9281716417910447,
"grad_norm": 0.5606224712873208,
"learning_rate": 5.632256981417845e-06,
"loss": 0.6134,
"step": 1990
},
{
"epoch": 0.9305037313432836,
"grad_norm": 0.5375797889998934,
"learning_rate": 5.59204592013407e-06,
"loss": 0.5973,
"step": 1995
},
{
"epoch": 0.9328358208955224,
"grad_norm": 0.5508292133907728,
"learning_rate": 5.553138877715833e-06,
"loss": 0.6298,
"step": 2000
},
{
"epoch": 0.9351679104477612,
"grad_norm": 0.5407565378019897,
"learning_rate": 5.515538170012309e-06,
"loss": 0.6085,
"step": 2005
},
{
"epoch": 0.9375,
"grad_norm": 0.5516140680428332,
"learning_rate": 5.479246035116201e-06,
"loss": 0.5998,
"step": 2010
},
{
"epoch": 0.9398320895522388,
"grad_norm": 0.5871082563627422,
"learning_rate": 5.444264633230531e-06,
"loss": 0.6146,
"step": 2015
},
{
"epoch": 0.9421641791044776,
"grad_norm": 0.5579174490831493,
"learning_rate": 5.410596046540051e-06,
"loss": 0.6008,
"step": 2020
},
{
"epoch": 0.9444962686567164,
"grad_norm": 0.5535279013398093,
"learning_rate": 5.378242279087314e-06,
"loss": 0.6071,
"step": 2025
},
{
"epoch": 0.9468283582089553,
"grad_norm": 0.5542818715981681,
"learning_rate": 5.347205256653387e-06,
"loss": 0.6131,
"step": 2030
},
{
"epoch": 0.949160447761194,
"grad_norm": 0.5507958590165508,
"learning_rate": 5.317486826643219e-06,
"loss": 0.5896,
"step": 2035
},
{
"epoch": 0.9514925373134329,
"grad_norm": 0.5386294413734031,
"learning_rate": 5.28908875797568e-06,
"loss": 0.6164,
"step": 2040
},
{
"epoch": 0.9538246268656716,
"grad_norm": 0.5657594223236935,
"learning_rate": 5.262012740978269e-06,
"loss": 0.6171,
"step": 2045
},
{
"epoch": 0.9561567164179104,
"grad_norm": 0.545809161265838,
"learning_rate": 5.236260387286509e-06,
"loss": 0.6014,
"step": 2050
},
{
"epoch": 0.9584888059701493,
"grad_norm": 0.5400541424481552,
"learning_rate": 5.2118332297480105e-06,
"loss": 0.6195,
"step": 2055
},
{
"epoch": 0.960820895522388,
"grad_norm": 0.5301287222550025,
"learning_rate": 5.1887327223312296e-06,
"loss": 0.5894,
"step": 2060
},
{
"epoch": 0.9631529850746269,
"grad_norm": 0.5826747467153147,
"learning_rate": 5.166960240038937e-06,
"loss": 0.6103,
"step": 2065
},
{
"epoch": 0.9654850746268657,
"grad_norm": 0.5624482325072565,
"learning_rate": 5.1465170788263595e-06,
"loss": 0.6205,
"step": 2070
},
{
"epoch": 0.9678171641791045,
"grad_norm": 0.5710832690138076,
"learning_rate": 5.1274044555240525e-06,
"loss": 0.6083,
"step": 2075
},
{
"epoch": 0.9701492537313433,
"grad_norm": 0.5427558552652122,
"learning_rate": 5.109623507765466e-06,
"loss": 0.6113,
"step": 2080
},
{
"epoch": 0.972481343283582,
"grad_norm": 0.5356447254957378,
"learning_rate": 5.093175293919228e-06,
"loss": 0.6138,
"step": 2085
},
{
"epoch": 0.9748134328358209,
"grad_norm": 0.554507085372338,
"learning_rate": 5.07806079302615e-06,
"loss": 0.6209,
"step": 2090
},
{
"epoch": 0.9771455223880597,
"grad_norm": 0.5347894717484287,
"learning_rate": 5.064280904740953e-06,
"loss": 0.6187,
"step": 2095
},
{
"epoch": 0.9794776119402985,
"grad_norm": 0.548171827658312,
"learning_rate": 5.051836449278715e-06,
"loss": 0.6069,
"step": 2100
},
{
"epoch": 0.9818097014925373,
"grad_norm": 0.562019041400098,
"learning_rate": 5.040728167366057e-06,
"loss": 0.6028,
"step": 2105
},
{
"epoch": 0.9841417910447762,
"grad_norm": 0.5554495773144189,
"learning_rate": 5.030956720197035e-06,
"loss": 0.6039,
"step": 2110
},
{
"epoch": 0.9864738805970149,
"grad_norm": 0.5847204114674615,
"learning_rate": 5.022522689393809e-06,
"loss": 0.6377,
"step": 2115
},
{
"epoch": 0.9888059701492538,
"grad_norm": 0.5293713892212091,
"learning_rate": 5.015426576972003e-06,
"loss": 0.6099,
"step": 2120
},
{
"epoch": 0.9911380597014925,
"grad_norm": 0.5377836319550625,
"learning_rate": 5.009668805310832e-06,
"loss": 0.6063,
"step": 2125
},
{
"epoch": 0.9934701492537313,
"grad_norm": 0.5661212299229181,
"learning_rate": 5.005249717127964e-06,
"loss": 0.6199,
"step": 2130
},
{
"epoch": 0.9958022388059702,
"grad_norm": 0.5383491265897701,
"learning_rate": 5.002169575459111e-06,
"loss": 0.6107,
"step": 2135
},
{
"epoch": 0.9981343283582089,
"grad_norm": 0.5598221695919163,
"learning_rate": 5.000428563642382e-06,
"loss": 0.6166,
"step": 2140
},
{
"epoch": 1.0,
"step": 2144,
"total_flos": 241268013662208.0,
"train_loss": 0.6583015036894314,
"train_runtime": 13168.5767,
"train_samples_per_second": 2.605,
"train_steps_per_second": 0.163
}
],
"logging_steps": 5,
"max_steps": 2144,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 100,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 241268013662208.0,
"train_batch_size": 16,
"trial_name": null,
"trial_params": null
}