{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9998731447418495,
  "eval_steps": 500,
  "global_step": 3941,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002537105163009007,
      "grad_norm": 4.350327938684374,
      "learning_rate": 2.5316455696202533e-07,
      "loss": 1.8196,
      "step": 10
    },
    {
      "epoch": 0.005074210326018014,
      "grad_norm": 3.678530456221003,
      "learning_rate": 5.063291139240507e-07,
      "loss": 1.8108,
      "step": 20
    },
    {
      "epoch": 0.00761131548902702,
      "grad_norm": 2.945036910011768,
      "learning_rate": 7.59493670886076e-07,
      "loss": 1.8147,
      "step": 30
    },
    {
      "epoch": 0.010148420652036028,
      "grad_norm": 2.3190574498378447,
      "learning_rate": 1.0126582278481013e-06,
      "loss": 1.7548,
      "step": 40
    },
    {
      "epoch": 0.012685525815045033,
      "grad_norm": 2.074674879169714,
      "learning_rate": 1.2658227848101267e-06,
      "loss": 1.7204,
      "step": 50
    },
    {
      "epoch": 0.01522263097805404,
      "grad_norm": 1.748653753253889,
      "learning_rate": 1.518987341772152e-06,
      "loss": 1.6762,
      "step": 60
    },
    {
      "epoch": 0.01775973614106305,
      "grad_norm": 1.6751399170091914,
      "learning_rate": 1.7721518987341774e-06,
      "loss": 1.6487,
      "step": 70
    },
    {
      "epoch": 0.020296841304072055,
      "grad_norm": 1.6363195765660283,
      "learning_rate": 2.0253164556962026e-06,
      "loss": 1.6116,
      "step": 80
    },
    {
      "epoch": 0.022833946467081062,
      "grad_norm": 1.5794976909544993,
      "learning_rate": 2.278481012658228e-06,
      "loss": 1.5916,
      "step": 90
    },
    {
      "epoch": 0.025371051630090066,
      "grad_norm": 1.6233632992191482,
      "learning_rate": 2.5316455696202535e-06,
      "loss": 1.598,
      "step": 100
    },
    {
      "epoch": 0.027908156793099072,
      "grad_norm": 1.5800661094425872,
      "learning_rate": 2.7848101265822785e-06,
      "loss": 1.5626,
      "step": 110
    },
    {
      "epoch": 0.03044526195610808,
      "grad_norm": 1.6072050143283245,
      "learning_rate": 3.037974683544304e-06,
      "loss": 1.5457,
      "step": 120
    },
    {
      "epoch": 0.03298236711911709,
      "grad_norm": 1.6572306247078625,
      "learning_rate": 3.2911392405063294e-06,
      "loss": 1.5391,
      "step": 130
    },
    {
      "epoch": 0.0355194722821261,
      "grad_norm": 1.586848380490154,
      "learning_rate": 3.544303797468355e-06,
      "loss": 1.5125,
      "step": 140
    },
    {
      "epoch": 0.038056577445135104,
      "grad_norm": 1.638449311664989,
      "learning_rate": 3.7974683544303802e-06,
      "loss": 1.5243,
      "step": 150
    },
    {
      "epoch": 0.04059368260814411,
      "grad_norm": 1.6031920359772533,
      "learning_rate": 4.050632911392405e-06,
      "loss": 1.4856,
      "step": 160
    },
    {
      "epoch": 0.04313078777115312,
      "grad_norm": 1.5967439995800559,
      "learning_rate": 4.303797468354431e-06,
      "loss": 1.5129,
      "step": 170
    },
    {
      "epoch": 0.045667892934162124,
      "grad_norm": 1.6075212497819606,
      "learning_rate": 4.556962025316456e-06,
      "loss": 1.4961,
      "step": 180
    },
    {
      "epoch": 0.04820499809717113,
      "grad_norm": 1.6791238021539772,
      "learning_rate": 4.8101265822784815e-06,
      "loss": 1.4893,
      "step": 190
    },
    {
      "epoch": 0.05074210326018013,
      "grad_norm": 1.6210921331693446,
      "learning_rate": 5.063291139240507e-06,
      "loss": 1.5013,
      "step": 200
    },
    {
      "epoch": 0.05327920842318914,
      "grad_norm": 1.7652729374280518,
      "learning_rate": 5.3164556962025316e-06,
      "loss": 1.473,
      "step": 210
    },
    {
      "epoch": 0.055816313586198145,
      "grad_norm": 1.6723244529240142,
      "learning_rate": 5.569620253164557e-06,
      "loss": 1.4791,
      "step": 220
    },
    {
      "epoch": 0.05835341874920715,
      "grad_norm": 1.825326138794735,
      "learning_rate": 5.8227848101265824e-06,
      "loss": 1.4761,
      "step": 230
    },
    {
      "epoch": 0.06089052391221616,
      "grad_norm": 1.9131148271572453,
      "learning_rate": 6.075949367088608e-06,
      "loss": 1.4626,
      "step": 240
    },
    {
      "epoch": 0.06342762907522517,
      "grad_norm": 1.6613770739809675,
      "learning_rate": 6.329113924050634e-06,
      "loss": 1.4601,
      "step": 250
    },
    {
      "epoch": 0.06596473423823418,
      "grad_norm": 1.6666458237214428,
      "learning_rate": 6.582278481012659e-06,
      "loss": 1.4686,
      "step": 260
    },
    {
      "epoch": 0.06850183940124319,
      "grad_norm": 1.5745675069520453,
      "learning_rate": 6.835443037974684e-06,
      "loss": 1.461,
      "step": 270
    },
    {
      "epoch": 0.0710389445642522,
      "grad_norm": 1.6507776778175596,
      "learning_rate": 7.08860759493671e-06,
      "loss": 1.47,
      "step": 280
    },
    {
      "epoch": 0.0735760497272612,
      "grad_norm": 1.6009958375778823,
      "learning_rate": 7.341772151898735e-06,
      "loss": 1.4526,
      "step": 290
    },
    {
      "epoch": 0.07611315489027021,
      "grad_norm": 1.6786912574149853,
      "learning_rate": 7.5949367088607605e-06,
      "loss": 1.4501,
      "step": 300
    },
    {
      "epoch": 0.07865026005327921,
      "grad_norm": 1.6698693144659327,
      "learning_rate": 7.848101265822786e-06,
      "loss": 1.4483,
      "step": 310
    },
    {
      "epoch": 0.08118736521628822,
      "grad_norm": 1.7393580296857223,
      "learning_rate": 8.10126582278481e-06,
      "loss": 1.4252,
      "step": 320
    },
    {
      "epoch": 0.08372447037929723,
      "grad_norm": 1.6124831573952214,
      "learning_rate": 8.354430379746837e-06,
      "loss": 1.4274,
      "step": 330
    },
    {
      "epoch": 0.08626157554230623,
      "grad_norm": 1.6899774259466704,
      "learning_rate": 8.607594936708861e-06,
      "loss": 1.437,
      "step": 340
    },
    {
      "epoch": 0.08879868070531524,
      "grad_norm": 1.6821954539953226,
      "learning_rate": 8.860759493670886e-06,
      "loss": 1.4388,
      "step": 350
    },
    {
      "epoch": 0.09133578586832425,
      "grad_norm": 1.8121412852354848,
      "learning_rate": 9.113924050632912e-06,
      "loss": 1.4151,
      "step": 360
    },
    {
      "epoch": 0.09387289103133326,
      "grad_norm": 1.5407350947949157,
      "learning_rate": 9.367088607594937e-06,
      "loss": 1.4274,
      "step": 370
    },
    {
      "epoch": 0.09640999619434226,
      "grad_norm": 1.7381357929095853,
      "learning_rate": 9.620253164556963e-06,
      "loss": 1.4309,
      "step": 380
    },
    {
      "epoch": 0.09894710135735126,
      "grad_norm": 1.6085237968347799,
      "learning_rate": 9.87341772151899e-06,
      "loss": 1.4173,
      "step": 390
    },
    {
      "epoch": 0.10148420652036026,
      "grad_norm": 1.728407830056737,
      "learning_rate": 9.999950942931784e-06,
      "loss": 1.4312,
      "step": 400
    },
    {
      "epoch": 0.10402131168336927,
      "grad_norm": 1.6972786696047149,
      "learning_rate": 9.999558492161865e-06,
      "loss": 1.422,
      "step": 410
    },
    {
      "epoch": 0.10655841684637828,
      "grad_norm": 1.6297612720977512,
      "learning_rate": 9.998773621425852e-06,
      "loss": 1.3892,
      "step": 420
    },
    {
      "epoch": 0.10909552200938728,
      "grad_norm": 1.7368434280409393,
      "learning_rate": 9.997596392328971e-06,
      "loss": 1.4368,
      "step": 430
    },
    {
      "epoch": 0.11163262717239629,
      "grad_norm": 1.8196387241516612,
      "learning_rate": 9.996026897273024e-06,
      "loss": 1.4129,
      "step": 440
    },
    {
      "epoch": 0.1141697323354053,
      "grad_norm": 1.6083388947957715,
      "learning_rate": 9.994065259449128e-06,
      "loss": 1.4181,
      "step": 450
    },
    {
      "epoch": 0.1167068374984143,
      "grad_norm": 1.7315485017229137,
      "learning_rate": 9.991711632828049e-06,
      "loss": 1.4107,
      "step": 460
    },
    {
      "epoch": 0.11924394266142331,
      "grad_norm": 1.6192259621686464,
      "learning_rate": 9.988966202148115e-06,
      "loss": 1.3933,
      "step": 470
    },
    {
      "epoch": 0.12178104782443232,
      "grad_norm": 1.685661277294985,
      "learning_rate": 9.985829182900717e-06,
      "loss": 1.4305,
      "step": 480
    },
    {
      "epoch": 0.12431815298744132,
      "grad_norm": 1.7673237121898477,
      "learning_rate": 9.982300821313394e-06,
      "loss": 1.407,
      "step": 490
    },
    {
      "epoch": 0.12685525815045035,
      "grad_norm": 1.6729684650970384,
      "learning_rate": 9.978381394330509e-06,
      "loss": 1.3941,
      "step": 500
    },
    {
      "epoch": 0.12939236331345935,
      "grad_norm": 1.6737204711157692,
      "learning_rate": 9.974071209591507e-06,
      "loss": 1.4083,
      "step": 510
    },
    {
      "epoch": 0.13192946847646836,
      "grad_norm": 1.5846450496238496,
      "learning_rate": 9.96937060540677e-06,
      "loss": 1.3913,
      "step": 520
    },
    {
      "epoch": 0.13446657363947737,
      "grad_norm": 1.6440175318683266,
      "learning_rate": 9.964279950731066e-06,
      "loss": 1.4141,
      "step": 530
    },
    {
      "epoch": 0.13700367880248637,
      "grad_norm": 1.5435352480418292,
      "learning_rate": 9.958799645134585e-06,
      "loss": 1.3923,
      "step": 540
    },
    {
      "epoch": 0.13954078396549538,
      "grad_norm": 1.6806917695478834,
      "learning_rate": 9.952930118771576e-06,
      "loss": 1.3882,
      "step": 550
    },
    {
      "epoch": 0.1420778891285044,
      "grad_norm": 1.6991483906725386,
      "learning_rate": 9.946671832346588e-06,
      "loss": 1.3806,
      "step": 560
    },
    {
      "epoch": 0.1446149942915134,
      "grad_norm": 1.6444779930069549,
      "learning_rate": 9.940025277078304e-06,
      "loss": 1.3877,
      "step": 570
    },
    {
      "epoch": 0.1471520994545224,
      "grad_norm": 1.584958906864304,
      "learning_rate": 9.932990974660992e-06,
      "loss": 1.3758,
      "step": 580
    },
    {
      "epoch": 0.1496892046175314,
      "grad_norm": 1.6339337045637337,
      "learning_rate": 9.925569477223549e-06,
      "loss": 1.3942,
      "step": 590
    },
    {
      "epoch": 0.15222630978054041,
      "grad_norm": 1.6782688039697937,
      "learning_rate": 9.917761367286164e-06,
      "loss": 1.3997,
      "step": 600
    },
    {
      "epoch": 0.15476341494354942,
      "grad_norm": 1.722985917307532,
      "learning_rate": 9.909567257714605e-06,
      "loss": 1.3902,
      "step": 610
    },
    {
      "epoch": 0.15730052010655843,
      "grad_norm": 1.515629790408513,
      "learning_rate": 9.9009877916721e-06,
      "loss": 1.3906,
      "step": 620
    },
    {
      "epoch": 0.15983762526956743,
      "grad_norm": 1.6832684084973726,
      "learning_rate": 9.892023642568871e-06,
      "loss": 1.3644,
      "step": 630
    },
    {
      "epoch": 0.16237473043257644,
      "grad_norm": 1.7217857604177804,
      "learning_rate": 9.882675514009262e-06,
      "loss": 1.3673,
      "step": 640
    },
    {
      "epoch": 0.16491183559558545,
      "grad_norm": 1.772784930329774,
      "learning_rate": 9.872944139736523e-06,
      "loss": 1.3751,
      "step": 650
    },
    {
      "epoch": 0.16744894075859446,
      "grad_norm": 1.528731449667675,
      "learning_rate": 9.862830283575215e-06,
      "loss": 1.3678,
      "step": 660
    },
    {
      "epoch": 0.16998604592160346,
      "grad_norm": 1.6156407009731812,
      "learning_rate": 9.852334739371252e-06,
      "loss": 1.3825,
      "step": 670
    },
    {
      "epoch": 0.17252315108461247,
      "grad_norm": 1.6415774929326135,
      "learning_rate": 9.841458330929598e-06,
      "loss": 1.3884,
      "step": 680
    },
    {
      "epoch": 0.17506025624762148,
      "grad_norm": 1.6070221223746397,
      "learning_rate": 9.830201911949604e-06,
      "loss": 1.3934,
      "step": 690
    },
    {
      "epoch": 0.17759736141063048,
      "grad_norm": 1.562454926578275,
      "learning_rate": 9.818566365957996e-06,
      "loss": 1.3645,
      "step": 700
    },
    {
      "epoch": 0.1801344665736395,
      "grad_norm": 1.5996614008577792,
      "learning_rate": 9.80655260623953e-06,
      "loss": 1.3708,
      "step": 710
    },
    {
      "epoch": 0.1826715717366485,
      "grad_norm": 1.5048794279696338,
      "learning_rate": 9.794161575765311e-06,
      "loss": 1.3749,
      "step": 720
    },
    {
      "epoch": 0.1852086768996575,
      "grad_norm": 1.5935516523984996,
      "learning_rate": 9.78139424711877e-06,
      "loss": 1.3886,
      "step": 730
    },
    {
      "epoch": 0.1877457820626665,
      "grad_norm": 1.5282445861981415,
      "learning_rate": 9.76825162241933e-06,
      "loss": 1.373,
      "step": 740
    },
    {
      "epoch": 0.19028288722567552,
      "grad_norm": 1.6358363561782086,
      "learning_rate": 9.754734733243749e-06,
      "loss": 1.3742,
      "step": 750
    },
    {
      "epoch": 0.19281999238868452,
      "grad_norm": 1.587601691095452,
      "learning_rate": 9.740844640545151e-06,
      "loss": 1.3603,
      "step": 760
    },
    {
      "epoch": 0.19535709755169353,
      "grad_norm": 1.5280980736395107,
      "learning_rate": 9.726582434569744e-06,
      "loss": 1.3636,
      "step": 770
    },
    {
      "epoch": 0.1978942027147025,
      "grad_norm": 1.5821827900533842,
      "learning_rate": 9.711949234771258e-06,
      "loss": 1.3536,
      "step": 780
    },
    {
      "epoch": 0.20043130787771152,
      "grad_norm": 1.6354511380141648,
      "learning_rate": 9.696946189723067e-06,
      "loss": 1.3777,
      "step": 790
    },
    {
      "epoch": 0.20296841304072052,
      "grad_norm": 1.4961728532518945,
      "learning_rate": 9.681574477028039e-06,
      "loss": 1.3555,
      "step": 800
    },
    {
      "epoch": 0.20550551820372953,
      "grad_norm": 1.6165039622149184,
      "learning_rate": 9.66583530322611e-06,
      "loss": 1.3736,
      "step": 810
    },
    {
      "epoch": 0.20804262336673854,
      "grad_norm": 1.6085071421673924,
      "learning_rate": 9.649729903699575e-06,
      "loss": 1.3685,
      "step": 820
    },
    {
      "epoch": 0.21057972852974755,
      "grad_norm": 1.674024959469941,
      "learning_rate": 9.633259542576127e-06,
      "loss": 1.3516,
      "step": 830
    },
    {
      "epoch": 0.21311683369275655,
      "grad_norm": 1.5457350813422102,
      "learning_rate": 9.61642551262963e-06,
      "loss": 1.3433,
      "step": 840
    },
    {
      "epoch": 0.21565393885576556,
      "grad_norm": 1.5638603110160445,
      "learning_rate": 9.599229135178651e-06,
      "loss": 1.3596,
      "step": 850
    },
    {
      "epoch": 0.21819104401877457,
      "grad_norm": 1.5845605847496684,
      "learning_rate": 9.581671759982747e-06,
      "loss": 1.3821,
      "step": 860
    },
    {
      "epoch": 0.22072814918178357,
      "grad_norm": 1.56321358587459,
      "learning_rate": 9.563754765136522e-06,
      "loss": 1.3568,
      "step": 870
    },
    {
      "epoch": 0.22326525434479258,
      "grad_norm": 1.621089242536098,
      "learning_rate": 9.545479556961457e-06,
      "loss": 1.3614,
      "step": 880
    },
    {
      "epoch": 0.2258023595078016,
      "grad_norm": 1.5746598904044478,
      "learning_rate": 9.526847569895529e-06,
      "loss": 1.3536,
      "step": 890
    },
    {
      "epoch": 0.2283394646708106,
      "grad_norm": 1.632144120218129,
      "learning_rate": 9.507860266380625e-06,
      "loss": 1.3521,
      "step": 900
    },
    {
      "epoch": 0.2308765698338196,
      "grad_norm": 1.6660492103415234,
      "learning_rate": 9.488519136747741e-06,
      "loss": 1.3455,
      "step": 910
    },
    {
      "epoch": 0.2334136749968286,
      "grad_norm": 1.6307872469664786,
      "learning_rate": 9.468825699100013e-06,
      "loss": 1.3388,
      "step": 920
    },
    {
      "epoch": 0.23595078015983761,
      "grad_norm": 1.51760811186189,
      "learning_rate": 9.448781499193563e-06,
      "loss": 1.36,
      "step": 930
    },
    {
      "epoch": 0.23848788532284662,
      "grad_norm": 1.6298958079571104,
      "learning_rate": 9.428388110316165e-06,
      "loss": 1.346,
      "step": 940
    },
    {
      "epoch": 0.24102499048585563,
      "grad_norm": 1.6241168589647443,
      "learning_rate": 9.407647133163754e-06,
      "loss": 1.3565,
      "step": 950
    },
    {
      "epoch": 0.24356209564886463,
      "grad_norm": 1.6330870068463266,
      "learning_rate": 9.386560195714796e-06,
      "loss": 1.3539,
      "step": 960
    },
    {
      "epoch": 0.24609920081187364,
      "grad_norm": 1.5846187793083721,
      "learning_rate": 9.365128953102495e-06,
      "loss": 1.3443,
      "step": 970
    },
    {
      "epoch": 0.24863630597488265,
      "grad_norm": 1.5816319458789425,
      "learning_rate": 9.343355087484893e-06,
      "loss": 1.3449,
      "step": 980
    },
    {
      "epoch": 0.25117341113789166,
      "grad_norm": 1.608667389007063,
      "learning_rate": 9.321240307912818e-06,
      "loss": 1.3503,
      "step": 990
    },
    {
      "epoch": 0.2537105163009007,
      "grad_norm": 1.535736158897923,
      "learning_rate": 9.298786350195758e-06,
      "loss": 1.3504,
      "step": 1000
    },
    {
      "epoch": 0.25624762146390967,
      "grad_norm": 1.6105502703548435,
      "learning_rate": 9.275994976765602e-06,
      "loss": 1.3512,
      "step": 1010
    },
    {
      "epoch": 0.2587847266269187,
      "grad_norm": 1.457234439212148,
      "learning_rate": 9.252867976538312e-06,
      "loss": 1.3447,
      "step": 1020
    },
    {
      "epoch": 0.2613218317899277,
      "grad_norm": 1.632312084639862,
      "learning_rate": 9.22940716477351e-06,
      "loss": 1.3451,
      "step": 1030
    },
    {
      "epoch": 0.2638589369529367,
      "grad_norm": 1.5657163405769847,
      "learning_rate": 9.205614382931986e-06,
      "loss": 1.3678,
      "step": 1040
    },
    {
      "epoch": 0.2663960421159457,
      "grad_norm": 1.523325498659843,
      "learning_rate": 9.181491498531179e-06,
      "loss": 1.355,
      "step": 1050
    },
    {
      "epoch": 0.26893314727895473,
      "grad_norm": 1.5647021825494114,
      "learning_rate": 9.157040404998572e-06,
      "loss": 1.3455,
      "step": 1060
    },
    {
      "epoch": 0.2714702524419637,
      "grad_norm": 1.581907280598391,
      "learning_rate": 9.132263021523096e-06,
      "loss": 1.353,
      "step": 1070
    },
    {
      "epoch": 0.27400735760497275,
      "grad_norm": 1.4861566014453274,
      "learning_rate": 9.107161292904476e-06,
      "loss": 1.3428,
      "step": 1080
    },
    {
      "epoch": 0.2765444627679817,
      "grad_norm": 1.6256210181495103,
      "learning_rate": 9.081737189400583e-06,
      "loss": 1.3421,
      "step": 1090
    },
    {
      "epoch": 0.27908156793099076,
      "grad_norm": 1.4876360574590954,
      "learning_rate": 9.0559927065728e-06,
      "loss": 1.3377,
      "step": 1100
    },
    {
      "epoch": 0.28161867309399974,
      "grad_norm": 1.5204275847901962,
      "learning_rate": 9.029929865129375e-06,
      "loss": 1.349,
      "step": 1110
    },
    {
      "epoch": 0.2841557782570088,
      "grad_norm": 1.5624633516357405,
      "learning_rate": 9.003550710766813e-06,
      "loss": 1.3552,
      "step": 1120
    },
    {
      "epoch": 0.28669288342001775,
      "grad_norm": 1.5591567540085947,
      "learning_rate": 8.97685731400932e-06,
      "loss": 1.3209,
      "step": 1130
    },
    {
      "epoch": 0.2892299885830268,
      "grad_norm": 1.5373317845285133,
      "learning_rate": 8.949851770046272e-06,
      "loss": 1.3267,
      "step": 1140
    },
    {
      "epoch": 0.29176709374603577,
      "grad_norm": 1.5556061129094692,
      "learning_rate": 8.922536198567772e-06,
      "loss": 1.3379,
      "step": 1150
    },
    {
      "epoch": 0.2943041989090448,
      "grad_norm": 1.971486780664198,
      "learning_rate": 8.894912743598269e-06,
      "loss": 1.3272,
      "step": 1160
    },
    {
      "epoch": 0.2968413040720538,
      "grad_norm": 1.5365700226491938,
      "learning_rate": 8.866983573328267e-06,
      "loss": 1.333,
      "step": 1170
    },
    {
      "epoch": 0.2993784092350628,
      "grad_norm": 1.6217713070921793,
      "learning_rate": 8.83875087994415e-06,
      "loss": 1.3497,
      "step": 1180
    },
    {
      "epoch": 0.3019155143980718,
      "grad_norm": 1.4917017043884344,
      "learning_rate": 8.810216879456114e-06,
      "loss": 1.3355,
      "step": 1190
    },
    {
      "epoch": 0.30445261956108083,
      "grad_norm": 1.5427563058731948,
      "learning_rate": 8.781383811524222e-06,
      "loss": 1.3339,
      "step": 1200
    },
    {
      "epoch": 0.3069897247240898,
      "grad_norm": 1.5666645778409243,
      "learning_rate": 8.752253939282622e-06,
      "loss": 1.332,
      "step": 1210
    },
    {
      "epoch": 0.30952682988709884,
      "grad_norm": 1.5940427272465527,
      "learning_rate": 8.722829549161904e-06,
      "loss": 1.3411,
      "step": 1220
    },
    {
      "epoch": 0.3120639350501078,
      "grad_norm": 1.569355522659196,
      "learning_rate": 8.69311295070964e-06,
      "loss": 1.321,
      "step": 1230
    },
    {
      "epoch": 0.31460104021311686,
      "grad_norm": 1.5823744419831982,
      "learning_rate": 8.663106476409107e-06,
      "loss": 1.3511,
      "step": 1240
    },
    {
      "epoch": 0.31713814537612584,
      "grad_norm": 1.5626340370876246,
      "learning_rate": 8.632812481496195e-06,
      "loss": 1.3491,
      "step": 1250
    },
    {
      "epoch": 0.31967525053913487,
      "grad_norm": 1.6216546055767536,
      "learning_rate": 8.602233343774562e-06,
      "loss": 1.3294,
      "step": 1260
    },
    {
      "epoch": 0.32221235570214385,
      "grad_norm": 1.4885399811487754,
      "learning_rate": 8.571371463428986e-06,
      "loss": 1.3419,
      "step": 1270
    },
    {
      "epoch": 0.3247494608651529,
      "grad_norm": 1.597124872589071,
      "learning_rate": 8.540229262836974e-06,
      "loss": 1.3245,
      "step": 1280
    },
    {
      "epoch": 0.32728656602816186,
      "grad_norm": 1.5069638761813242,
      "learning_rate": 8.508809186378631e-06,
      "loss": 1.3357,
      "step": 1290
    },
    {
      "epoch": 0.3298236711911709,
      "grad_norm": 1.5496475251999724,
      "learning_rate": 8.477113700244788e-06,
      "loss": 1.3297,
      "step": 1300
    },
    {
      "epoch": 0.3323607763541799,
      "grad_norm": 1.5177410295586948,
      "learning_rate": 8.445145292243446e-06,
      "loss": 1.3361,
      "step": 1310
    },
    {
      "epoch": 0.3348978815171889,
      "grad_norm": 1.4375424317665,
      "learning_rate": 8.412906471604489e-06,
      "loss": 1.3365,
      "step": 1320
    },
    {
      "epoch": 0.3374349866801979,
      "grad_norm": 1.4733958562961815,
      "learning_rate": 8.380399768782742e-06,
      "loss": 1.3364,
      "step": 1330
    },
    {
      "epoch": 0.3399720918432069,
      "grad_norm": 1.5665888162471464,
      "learning_rate": 8.347627735259344e-06,
      "loss": 1.3572,
      "step": 1340
    },
    {
      "epoch": 0.3425091970062159,
      "grad_norm": 1.5175787042273947,
      "learning_rate": 8.314592943341494e-06,
      "loss": 1.311,
      "step": 1350
    },
    {
      "epoch": 0.34504630216922494,
      "grad_norm": 1.5210307965368668,
      "learning_rate": 8.281297985960538e-06,
      "loss": 1.3261,
      "step": 1360
    },
    {
      "epoch": 0.3475834073322339,
      "grad_norm": 1.5365431443148119,
      "learning_rate": 8.247745476468449e-06,
      "loss": 1.3433,
      "step": 1370
    },
    {
      "epoch": 0.35012051249524295,
      "grad_norm": 1.5548012069585933,
      "learning_rate": 8.213938048432697e-06,
      "loss": 1.3134,
      "step": 1380
    },
    {
      "epoch": 0.35265761765825193,
      "grad_norm": 1.4642811591908687,
      "learning_rate": 8.179878355429556e-06,
      "loss": 1.3159,
      "step": 1390
    },
    {
      "epoch": 0.35519472282126097,
      "grad_norm": 1.6713134353309254,
      "learning_rate": 8.145569070835799e-06,
      "loss": 1.3285,
      "step": 1400
    },
    {
      "epoch": 0.35773182798426995,
      "grad_norm": 1.5444628338197106,
      "learning_rate": 8.111012887618882e-06,
      "loss": 1.344,
      "step": 1410
    },
    {
      "epoch": 0.360268933147279,
      "grad_norm": 1.5042040298049457,
      "learning_rate": 8.076212518125556e-06,
      "loss": 1.3217,
      "step": 1420
    },
    {
      "epoch": 0.36280603831028796,
      "grad_norm": 1.5827643194628298,
      "learning_rate": 8.041170693868985e-06,
      "loss": 1.3284,
      "step": 1430
    },
    {
      "epoch": 0.365343143473297,
      "grad_norm": 1.4314485322723574,
      "learning_rate": 8.005890165314334e-06,
      "loss": 1.3188,
      "step": 1440
    },
    {
      "epoch": 0.367880248636306,
      "grad_norm": 1.5452457890288078,
      "learning_rate": 7.970373701662892e-06,
      "loss": 1.3123,
      "step": 1450
    },
    {
      "epoch": 0.370417353799315,
      "grad_norm": 1.5944938106930338,
      "learning_rate": 7.934624090634713e-06,
      "loss": 1.3131,
      "step": 1460
    },
    {
      "epoch": 0.372954458962324,
      "grad_norm": 1.5553727991379855,
      "learning_rate": 7.8986441382498e-06,
      "loss": 1.3318,
      "step": 1470
    },
    {
      "epoch": 0.375491564125333,
      "grad_norm": 1.5196578480754726,
      "learning_rate": 7.862436668607865e-06,
      "loss": 1.3164,
      "step": 1480
    },
    {
      "epoch": 0.378028669288342,
      "grad_norm": 1.5354385242535227,
      "learning_rate": 7.826004523666661e-06,
      "loss": 1.3292,
      "step": 1490
    },
    {
      "epoch": 0.38056577445135104,
      "grad_norm": 1.5449910825994637,
      "learning_rate": 7.78935056301891e-06,
      "loss": 1.3272,
      "step": 1500
    },
    {
      "epoch": 0.38310287961436,
      "grad_norm": 1.4946907973724173,
      "learning_rate": 7.752477663667854e-06,
      "loss": 1.3391,
      "step": 1510
    },
    {
      "epoch": 0.38563998477736905,
      "grad_norm": 1.5791940161814702,
      "learning_rate": 7.715388719801437e-06,
      "loss": 1.3392,
      "step": 1520
    },
    {
      "epoch": 0.38817708994037803,
      "grad_norm": 1.4567702862839176,
      "learning_rate": 7.67808664256514e-06,
      "loss": 1.2971,
      "step": 1530
    },
    {
      "epoch": 0.39071419510338706,
      "grad_norm": 1.4605769814867744,
      "learning_rate": 7.640574359833472e-06,
      "loss": 1.3148,
      "step": 1540
    },
    {
      "epoch": 0.39325130026639604,
      "grad_norm": 1.5566796816874888,
      "learning_rate": 7.6028548159801685e-06,
      "loss": 1.3315,
      "step": 1550
    },
    {
      "epoch": 0.395788405429405,
      "grad_norm": 1.5768032029757384,
      "learning_rate": 7.564930971647087e-06,
      "loss": 1.3238,
      "step": 1560
    },
    {
      "epoch": 0.39832551059241406,
      "grad_norm": 1.5702550171255043,
      "learning_rate": 7.52680580351181e-06,
      "loss": 1.3175,
      "step": 1570
    },
    {
      "epoch": 0.40086261575542304,
      "grad_norm": 1.5687466837527182,
      "learning_rate": 7.488482304054019e-06,
      "loss": 1.3104,
      "step": 1580
    },
    {
      "epoch": 0.40339972091843207,
      "grad_norm": 1.5897550883645912,
      "learning_rate": 7.449963481320599e-06,
      "loss": 1.316,
      "step": 1590
    },
    {
      "epoch": 0.40593682608144105,
      "grad_norm": 1.5236147067965886,
      "learning_rate": 7.411252358689541e-06,
      "loss": 1.3273,
      "step": 1600
    },
    {
      "epoch": 0.4084739312444501,
      "grad_norm": 1.5469446528938424,
      "learning_rate": 7.372351974632634e-06,
      "loss": 1.3119,
      "step": 1610
    },
    {
      "epoch": 0.41101103640745906,
      "grad_norm": 1.4722026799112722,
      "learning_rate": 7.333265382476971e-06,
      "loss": 1.3151,
      "step": 1620
    },
    {
      "epoch": 0.4135481415704681,
      "grad_norm": 1.5178886141824586,
      "learning_rate": 7.293995650165287e-06,
      "loss": 1.3245,
      "step": 1630
    },
    {
      "epoch": 0.4160852467334771,
      "grad_norm": 1.5308435376939995,
      "learning_rate": 7.2545458600151615e-06,
      "loss": 1.3317,
      "step": 1640
    },
    {
      "epoch": 0.4186223518964861,
      "grad_norm": 1.5091424984828243,
      "learning_rate": 7.214919108477077e-06,
      "loss": 1.3044,
      "step": 1650
    },
    {
      "epoch": 0.4211594570594951,
      "grad_norm": 1.457202507709852,
      "learning_rate": 7.175118505891385e-06,
      "loss": 1.3339,
      "step": 1660
    },
    {
      "epoch": 0.4236965622225041,
      "grad_norm": 1.530896247556501,
      "learning_rate": 7.135147176244158e-06,
      "loss": 1.3044,
      "step": 1670
    },
    {
      "epoch": 0.4262336673855131,
      "grad_norm": 1.5274463812149695,
      "learning_rate": 7.0950082569219955e-06,
      "loss": 1.3048,
      "step": 1680
    },
    {
      "epoch": 0.42877077254852214,
      "grad_norm": 1.507428973101804,
      "learning_rate": 7.054704898465772e-06,
      "loss": 1.3069,
      "step": 1690
    },
    {
      "epoch": 0.4313078777115311,
      "grad_norm": 1.5716469315983397,
      "learning_rate": 7.0142402643233346e-06,
      "loss": 1.3136,
      "step": 1700
    },
    {
      "epoch": 0.43384498287454015,
      "grad_norm": 1.4220881687524514,
      "learning_rate": 6.973617530601209e-06,
      "loss": 1.3165,
      "step": 1710
    },
    {
      "epoch": 0.43638208803754913,
      "grad_norm": 1.5926945403384438,
      "learning_rate": 6.932839885815304e-06,
      "loss": 1.3301,
      "step": 1720
    },
    {
      "epoch": 0.43891919320055817,
      "grad_norm": 1.4527595611730801,
      "learning_rate": 6.891910530640642e-06,
      "loss": 1.3145,
      "step": 1730
    },
    {
      "epoch": 0.44145629836356715,
      "grad_norm": 1.5069254389998272,
      "learning_rate": 6.850832677660134e-06,
      "loss": 1.3139,
      "step": 1740
    },
    {
      "epoch": 0.4439934035265762,
      "grad_norm": 1.4587280578384394,
      "learning_rate": 6.809609551112419e-06,
      "loss": 1.3085,
      "step": 1750
    },
    {
      "epoch": 0.44653050868958516,
      "grad_norm": 1.5122830472595903,
      "learning_rate": 6.768244386638793e-06,
      "loss": 1.3158,
      "step": 1760
    },
    {
      "epoch": 0.4490676138525942,
      "grad_norm": 1.4912245201929943,
      "learning_rate": 6.726740431029243e-06,
      "loss": 1.3167,
      "step": 1770
    },
    {
      "epoch": 0.4516047190156032,
      "grad_norm": 1.5574941259720791,
      "learning_rate": 6.685100941967596e-06,
      "loss": 1.3118,
      "step": 1780
    },
    {
      "epoch": 0.4541418241786122,
      "grad_norm": 1.4994130740882026,
      "learning_rate": 6.643329187775827e-06,
      "loss": 1.307,
      "step": 1790
    },
    {
      "epoch": 0.4566789293416212,
      "grad_norm": 1.5791237950971593,
      "learning_rate": 6.601428447157525e-06,
      "loss": 1.3086,
      "step": 1800
    },
    {
      "epoch": 0.4592160345046302,
      "grad_norm": 1.5319564794342408,
      "learning_rate": 6.559402008940539e-06,
      "loss": 1.3025,
      "step": 1810
    },
    {
      "epoch": 0.4617531396676392,
      "grad_norm": 1.5560620624811086,
      "learning_rate": 6.517253171818844e-06,
      "loss": 1.3146,
      "step": 1820
    },
    {
      "epoch": 0.46429024483064824,
      "grad_norm": 1.5762189341956727,
      "learning_rate": 6.474985244093613e-06,
      "loss": 1.307,
      "step": 1830
    },
    {
      "epoch": 0.4668273499936572,
      "grad_norm": 1.568824162809672,
      "learning_rate": 6.432601543413552e-06,
      "loss": 1.2996,
      "step": 1840
    },
    {
      "epoch": 0.46936445515666625,
      "grad_norm": 1.461712822890638,
      "learning_rate": 6.390105396514497e-06,
      "loss": 1.3013,
      "step": 1850
    },
    {
      "epoch": 0.47190156031967523,
      "grad_norm": 1.4727912142252813,
      "learning_rate": 6.347500138958285e-06,
      "loss": 1.3086,
      "step": 1860
    },
    {
      "epoch": 0.47443866548268426,
      "grad_norm": 1.4842630358439066,
      "learning_rate": 6.304789114870953e-06,
      "loss": 1.3121,
      "step": 1870
    },
    {
      "epoch": 0.47697577064569324,
      "grad_norm": 1.5147058669468259,
      "learning_rate": 6.261975676680252e-06,
      "loss": 1.3109,
      "step": 1880
    },
    {
      "epoch": 0.4795128758087023,
      "grad_norm": 1.5879467208142688,
      "learning_rate": 6.219063184852509e-06,
      "loss": 1.3057,
      "step": 1890
    },
    {
      "epoch": 0.48204998097171126,
      "grad_norm": 1.4622817504218393,
      "learning_rate": 6.176055007628859e-06,
      "loss": 1.2978,
      "step": 1900
    },
    {
      "epoch": 0.4845870861347203,
      "grad_norm": 1.4651555100721898,
      "learning_rate": 6.132954520760882e-06,
      "loss": 1.2936,
      "step": 1910
    },
    {
      "epoch": 0.48712419129772927,
      "grad_norm": 1.4242680820832143,
      "learning_rate": 6.089765107245616e-06,
      "loss": 1.311,
      "step": 1920
    },
    {
      "epoch": 0.4896612964607383,
      "grad_norm": 1.4510357489546541,
      "learning_rate": 6.046490157060041e-06,
      "loss": 1.2917,
      "step": 1930
    },
    {
      "epoch": 0.4921984016237473,
      "grad_norm": 1.5389362630585735,
      "learning_rate": 6.003133066894987e-06,
      "loss": 1.3173,
      "step": 1940
    },
    {
      "epoch": 0.4947355067867563,
      "grad_norm": 1.5597918071325416,
      "learning_rate": 5.959697239888525e-06,
      "loss": 1.2978,
      "step": 1950
    },
    {
      "epoch": 0.4972726119497653,
      "grad_norm": 1.481163850939429,
      "learning_rate": 5.916186085358858e-06,
      "loss": 1.3125,
      "step": 1960
    },
    {
      "epoch": 0.49980971711277433,
      "grad_norm": 1.5172196100773179,
      "learning_rate": 5.872603018536713e-06,
      "loss": 1.3035,
      "step": 1970
    },
    {
      "epoch": 0.5023468222757833,
      "grad_norm": 1.520182324070576,
      "learning_rate": 5.828951460297277e-06,
      "loss": 1.2943,
      "step": 1980
    },
    {
      "epoch": 0.5048839274387923,
      "grad_norm": 1.374020881318329,
      "learning_rate": 5.785234836891697e-06,
      "loss": 1.3019,
      "step": 1990
    },
    {
      "epoch": 0.5074210326018014,
      "grad_norm": 1.609172422257604,
      "learning_rate": 5.741456579678141e-06,
      "loss": 1.2929,
      "step": 2000
    },
    {
      "epoch": 0.5099581377648104,
      "grad_norm": 1.451921659432821,
      "learning_rate": 5.697620124852472e-06,
      "loss": 1.2868,
      "step": 2010
    },
    {
      "epoch": 0.5124952429278193,
      "grad_norm": 1.531522896512812,
      "learning_rate": 5.65372891317854e-06,
      "loss": 1.2875,
      "step": 2020
    },
    {
      "epoch": 0.5150323480908283,
      "grad_norm": 1.443649652350418,
      "learning_rate": 5.6097863897181075e-06,
      "loss": 1.2963,
      "step": 2030
    },
    {
      "epoch": 0.5175694532538374,
      "grad_norm": 1.5591743411035264,
      "learning_rate": 5.565796003560447e-06,
      "loss": 1.3121,
      "step": 2040
    },
    {
      "epoch": 0.5201065584168464,
      "grad_norm": 1.428229068798765,
      "learning_rate": 5.521761207551622e-06,
      "loss": 1.2979,
      "step": 2050
    },
    {
      "epoch": 0.5226436635798554,
      "grad_norm": 1.5164415865949983,
      "learning_rate": 5.47768545802346e-06,
      "loss": 1.3107,
      "step": 2060
    },
    {
      "epoch": 0.5251807687428643,
      "grad_norm": 1.5292361648846982,
      "learning_rate": 5.433572214522275e-06,
      "loss": 1.2952,
      "step": 2070
    },
    {
      "epoch": 0.5277178739058734,
      "grad_norm": 1.4451039662214231,
      "learning_rate": 5.389424939537311e-06,
      "loss": 1.2922,
      "step": 2080
    },
    {
      "epoch": 0.5302549790688824,
      "grad_norm": 1.558654012548035,
      "learning_rate": 5.345247098228977e-06,
      "loss": 1.2942,
      "step": 2090
    },
    {
      "epoch": 0.5327920842318914,
      "grad_norm": 1.5393309134302235,
      "learning_rate": 5.301042158156866e-06,
      "loss": 1.2898,
      "step": 2100
    },
    {
      "epoch": 0.5353291893949004,
      "grad_norm": 1.5206662969722375,
      "learning_rate": 5.256813589007571e-06,
      "loss": 1.2967,
      "step": 2110
    },
    {
      "epoch": 0.5378662945579095,
      "grad_norm": 1.5295277898061372,
      "learning_rate": 5.212564862322355e-06,
      "loss": 1.2987,
      "step": 2120
    },
    {
      "epoch": 0.5404033997209184,
      "grad_norm": 1.5121887795702076,
      "learning_rate": 5.168299451224665e-06,
      "loss": 1.2859,
      "step": 2130
    },
    {
      "epoch": 0.5429405048839274,
      "grad_norm": 1.5405224763949017,
      "learning_rate": 5.124020830147525e-06,
      "loss": 1.2942,
      "step": 2140
    },
    {
      "epoch": 0.5454776100469364,
      "grad_norm": 1.5241647102261355,
      "learning_rate": 5.079732474560821e-06,
      "loss": 1.2967,
      "step": 2150
    },
    {
      "epoch": 0.5480147152099455,
      "grad_norm": 1.5740459163455902,
      "learning_rate": 5.035437860698508e-06,
      "loss": 1.2792,
      "step": 2160
    },
    {
      "epoch": 0.5505518203729545,
      "grad_norm": 1.455514390960437,
      "learning_rate": 4.991140465285762e-06,
      "loss": 1.2722,
      "step": 2170
    },
    {
      "epoch": 0.5530889255359634,
      "grad_norm": 1.4543563727275153,
      "learning_rate": 4.94684376526608e-06,
      "loss": 1.294,
      "step": 2180
    },
    {
      "epoch": 0.5556260306989724,
      "grad_norm": 1.415880887469612,
      "learning_rate": 4.902551237528387e-06,
      "loss": 1.2898,
      "step": 2190
    },
    {
      "epoch": 0.5581631358619815,
      "grad_norm": 1.5027054686198038,
      "learning_rate": 4.858266358634109e-06,
      "loss": 1.2943,
      "step": 2200
    },
    {
      "epoch": 0.5607002410249905,
      "grad_norm": 1.495415983271707,
      "learning_rate": 4.813992604544319e-06,
      "loss": 1.309,
      "step": 2210
    },
    {
      "epoch": 0.5632373461879995,
      "grad_norm": 1.5256997169566149,
      "learning_rate": 4.769733450346885e-06,
      "loss": 1.2941,
      "step": 2220
    },
    {
      "epoch": 0.5657744513510085,
      "grad_norm": 1.4207029137255274,
      "learning_rate": 4.725492369983721e-06,
      "loss": 1.2808,
      "step": 2230
    },
    {
      "epoch": 0.5683115565140175,
      "grad_norm": 1.5127789303300487,
      "learning_rate": 4.6812728359781064e-06,
      "loss": 1.2886,
      "step": 2240
    },
    {
      "epoch": 0.5708486616770265,
      "grad_norm": 1.4480660719145084,
      "learning_rate": 4.637078319162127e-06,
      "loss": 1.2848,
      "step": 2250
    },
    {
      "epoch": 0.5733857668400355,
      "grad_norm": 1.4818074524822986,
      "learning_rate": 4.592912288404251e-06,
      "loss": 1.2747,
      "step": 2260
    },
    {
      "epoch": 0.5759228720030445,
      "grad_norm": 1.496021447098999,
      "learning_rate": 4.5487782103370445e-06,
      "loss": 1.2889,
      "step": 2270
    },
    {
      "epoch": 0.5784599771660536,
      "grad_norm": 1.4726400774082267,
      "learning_rate": 4.504679549085077e-06,
      "loss": 1.2956,
      "step": 2280
    },
    {
      "epoch": 0.5809970823290626,
      "grad_norm": 1.492109044123467,
      "learning_rate": 4.460619765993025e-06,
      "loss": 1.2974,
      "step": 2290
    },
    {
      "epoch": 0.5835341874920715,
      "grad_norm": 1.4567515467141523,
      "learning_rate": 4.416602319353974e-06,
      "loss": 1.29,
      "step": 2300
    },
    {
      "epoch": 0.5860712926550805,
      "grad_norm": 1.460535915347314,
      "learning_rate": 4.3726306641379915e-06,
      "loss": 1.2745,
      "step": 2310
    },
    {
      "epoch": 0.5886083978180896,
      "grad_norm": 1.4651576736560898,
      "learning_rate": 4.328708251720924e-06,
      "loss": 1.2739,
      "step": 2320
    },
    {
      "epoch": 0.5911455029810986,
      "grad_norm": 1.6196158147206026,
      "learning_rate": 4.2848385296135165e-06,
      "loss": 1.3101,
      "step": 2330
    },
    {
      "epoch": 0.5936826081441076,
      "grad_norm": 1.527439804056797,
      "learning_rate": 4.241024941190792e-06,
      "loss": 1.2771,
      "step": 2340
    },
    {
      "epoch": 0.5962197133071165,
      "grad_norm": 1.4872645401772542,
      "learning_rate": 4.197270925421796e-06,
      "loss": 1.2877,
      "step": 2350
    },
    {
      "epoch": 0.5987568184701256,
      "grad_norm": 1.4908027336325684,
      "learning_rate": 4.153579916599659e-06,
      "loss": 1.2969,
      "step": 2360
    },
    {
      "epoch": 0.6012939236331346,
      "grad_norm": 1.370441167203172,
      "learning_rate": 4.109955344072036e-06,
      "loss": 1.2745,
      "step": 2370
    },
    {
      "epoch": 0.6038310287961436,
      "grad_norm": 1.457801692594122,
      "learning_rate": 4.066400631971938e-06,
      "loss": 1.2714,
      "step": 2380
    },
    {
      "epoch": 0.6063681339591526,
      "grad_norm": 1.5047248748403204,
      "learning_rate": 4.022919198948966e-06,
      "loss": 1.2759,
      "step": 2390
    },
    {
      "epoch": 0.6089052391221617,
      "grad_norm": 1.5232259549425642,
      "learning_rate": 3.979514457900982e-06,
      "loss": 1.2845,
      "step": 2400
    },
    {
      "epoch": 0.6114423442851706,
      "grad_norm": 1.4170452963382303,
      "learning_rate": 3.936189815706219e-06,
      "loss": 1.2833,
      "step": 2410
    },
    {
      "epoch": 0.6139794494481796,
      "grad_norm": 1.5010818180720833,
      "learning_rate": 3.8929486729558775e-06,
      "loss": 1.2941,
      "step": 2420
    },
    {
      "epoch": 0.6165165546111886,
      "grad_norm": 1.4420347497785075,
      "learning_rate": 3.849794423687212e-06,
      "loss": 1.2775,
      "step": 2430
    },
    {
      "epoch": 0.6190536597741977,
      "grad_norm": 1.520468191298721,
      "learning_rate": 3.8067304551171247e-06,
      "loss": 1.2627,
      "step": 2440
    },
    {
      "epoch": 0.6215907649372067,
      "grad_norm": 1.4753704862458017,
      "learning_rate": 3.7637601473763035e-06,
      "loss": 1.284,
      "step": 2450
    },
    {
      "epoch": 0.6241278701002156,
      "grad_norm": 1.469877746697786,
      "learning_rate": 3.7208868732439145e-06,
      "loss": 1.2927,
      "step": 2460
    },
    {
      "epoch": 0.6266649752632246,
      "grad_norm": 1.4601548141707599,
      "learning_rate": 3.6781139978828606e-06,
      "loss": 1.2947,
      "step": 2470
    },
    {
      "epoch": 0.6292020804262337,
      "grad_norm": 1.5092438879342172,
      "learning_rate": 3.6354448785756558e-06,
      "loss": 1.2843,
      "step": 2480
    },
    {
      "epoch": 0.6317391855892427,
      "grad_norm": 1.4368007055488876,
      "learning_rate": 3.592882864460905e-06,
      "loss": 1.265,
      "step": 2490
    },
    {
      "epoch": 0.6342762907522517,
      "grad_norm": 1.4672055312297339,
      "learning_rate": 3.5504312962704245e-06,
      "loss": 1.2709,
      "step": 2500
    },
    {
      "epoch": 0.6368133959152606,
      "grad_norm": 1.4995451462382032,
      "learning_rate": 3.5080935060670345e-06,
      "loss": 1.2679,
      "step": 2510
    },
    {
      "epoch": 0.6393505010782697,
      "grad_norm": 1.458116276283539,
      "learning_rate": 3.465872816983008e-06,
      "loss": 1.2821,
      "step": 2520
    },
    {
      "epoch": 0.6418876062412787,
      "grad_norm": 1.4447640379158275,
      "learning_rate": 3.4237725429592507e-06,
      "loss": 1.2865,
      "step": 2530
    },
    {
      "epoch": 0.6444247114042877,
      "grad_norm": 1.3965736731366891,
      "learning_rate": 3.3817959884851735e-06,
      "loss": 1.2698,
      "step": 2540
    },
    {
      "epoch": 0.6469618165672967,
      "grad_norm": 1.4648194884238146,
      "learning_rate": 3.3399464483393272e-06,
      "loss": 1.291,
      "step": 2550
    },
    {
      "epoch": 0.6494989217303058,
      "grad_norm": 1.4271493727093771,
      "learning_rate": 3.298227207330792e-06,
      "loss": 1.2765,
      "step": 2560
    },
    {
      "epoch": 0.6520360268933147,
      "grad_norm": 1.5962462881292958,
      "learning_rate": 3.256641540041346e-06,
      "loss": 1.2905,
      "step": 2570
    },
    {
      "epoch": 0.6545731320563237,
      "grad_norm": 1.4501719681830862,
      "learning_rate": 3.2151927105684423e-06,
      "loss": 1.298,
      "step": 2580
    },
    {
      "epoch": 0.6571102372193327,
      "grad_norm": 1.5186349976521718,
      "learning_rate": 3.1738839722690085e-06,
      "loss": 1.2742,
      "step": 2590
    },
    {
      "epoch": 0.6596473423823418,
      "grad_norm": 1.3901740398219145,
      "learning_rate": 3.1327185675040907e-06,
      "loss": 1.2769,
      "step": 2600
    },
    {
      "epoch": 0.6621844475453508,
      "grad_norm": 1.4618375024699428,
      "learning_rate": 3.0916997273843454e-06,
      "loss": 1.2938,
      "step": 2610
    },
    {
      "epoch": 0.6647215527083598,
      "grad_norm": 1.4675982361039484,
      "learning_rate": 3.0508306715164416e-06,
      "loss": 1.2913,
      "step": 2620
    },
    {
      "epoch": 0.6672586578713687,
      "grad_norm": 1.5086185778550512,
      "learning_rate": 3.0101146077503386e-06,
      "loss": 1.2777,
      "step": 2630
    },
    {
      "epoch": 0.6697957630343778,
      "grad_norm": 1.4573487737483761,
      "learning_rate": 2.9695547319275093e-06,
      "loss": 1.2633,
      "step": 2640
    },
    {
      "epoch": 0.6723328681973868,
      "grad_norm": 1.43323809832072,
      "learning_rate": 2.9291542276300866e-06,
      "loss": 1.289,
      "step": 2650
    },
    {
      "epoch": 0.6748699733603958,
      "grad_norm": 1.420082813628849,
      "learning_rate": 2.8889162659309832e-06,
      "loss": 1.2729,
      "step": 2660
    },
    {
      "epoch": 0.6774070785234048,
      "grad_norm": 1.424990219399345,
      "learning_rate": 2.848844005145004e-06,
      "loss": 1.3024,
      "step": 2670
    },
    {
      "epoch": 0.6799441836864138,
      "grad_norm": 1.4395745448115305,
      "learning_rate": 2.808940590580922e-06,
      "loss": 1.2845,
      "step": 2680
    },
    {
      "epoch": 0.6824812888494228,
      "grad_norm": 1.4802086998925903,
      "learning_rate": 2.769209154294623e-06,
      "loss": 1.2844,
      "step": 2690
    },
    {
      "epoch": 0.6850183940124318,
      "grad_norm": 1.491623196795251,
      "learning_rate": 2.7296528148432565e-06,
      "loss": 1.2683,
      "step": 2700
    },
    {
      "epoch": 0.6875554991754408,
      "grad_norm": 1.416764375906272,
      "learning_rate": 2.690274677040462e-06,
      "loss": 1.2776,
      "step": 2710
    },
    {
      "epoch": 0.6900926043384499,
      "grad_norm": 1.519033593874162,
      "learning_rate": 2.6510778317126597e-06,
      "loss": 1.2807,
      "step": 2720
    },
    {
      "epoch": 0.6926297095014589,
      "grad_norm": 1.3894691132515595,
      "learning_rate": 2.6120653554564624e-06,
      "loss": 1.2777,
      "step": 2730
    },
    {
      "epoch": 0.6951668146644678,
      "grad_norm": 1.4049713206074572,
      "learning_rate": 2.573240310397187e-06,
      "loss": 1.2736,
      "step": 2740
    },
    {
      "epoch": 0.6977039198274768,
      "grad_norm": 1.4357642101900112,
      "learning_rate": 2.5346057439484923e-06,
      "loss": 1.2803,
      "step": 2750
    },
    {
      "epoch": 0.7002410249904859,
      "grad_norm": 1.490167340198777,
      "learning_rate": 2.4961646885732034e-06,
      "loss": 1.2744,
      "step": 2760
    },
    {
      "epoch": 0.7027781301534949,
      "grad_norm": 1.4179312953545702,
      "learning_rate": 2.4579201615452812e-06,
      "loss": 1.2842,
      "step": 2770
    },
    {
      "epoch": 0.7053152353165039,
      "grad_norm": 1.6140649717523825,
      "learning_rate": 2.4198751647129896e-06,
      "loss": 1.2963,
      "step": 2780
    },
    {
      "epoch": 0.7078523404795128,
      "grad_norm": 1.530468810042779,
      "learning_rate": 2.3820326842632894e-06,
      "loss": 1.2637,
      "step": 2790
    },
    {
      "epoch": 0.7103894456425219,
      "grad_norm": 1.412588711043796,
      "learning_rate": 2.344395690487441e-06,
      "loss": 1.2856,
      "step": 2800
    },
    {
      "epoch": 0.7129265508055309,
      "grad_norm": 1.5447254908338892,
      "learning_rate": 2.3069671375478645e-06,
      "loss": 1.2848,
      "step": 2810
    },
    {
      "epoch": 0.7154636559685399,
      "grad_norm": 1.43691808431636,
      "learning_rate": 2.2697499632462695e-06,
      "loss": 1.2536,
      "step": 2820
    },
    {
      "epoch": 0.7180007611315489,
      "grad_norm": 1.5560428170621574,
      "learning_rate": 2.2327470887930595e-06,
      "loss": 1.3015,
      "step": 2830
    },
    {
      "epoch": 0.720537866294558,
      "grad_norm": 1.450374747082515,
      "learning_rate": 2.195961418578041e-06,
      "loss": 1.2744,
      "step": 2840
    },
    {
      "epoch": 0.7230749714575669,
      "grad_norm": 1.484538648746269,
      "learning_rate": 2.159395839942464e-06,
      "loss": 1.2664,
      "step": 2850
    },
    {
      "epoch": 0.7256120766205759,
      "grad_norm": 1.3953379506558543,
      "learning_rate": 2.1230532229523865e-06,
      "loss": 1.2489,
      "step": 2860
    },
    {
      "epoch": 0.7281491817835849,
      "grad_norm": 1.4415654155573785,
      "learning_rate": 2.086936420173399e-06,
      "loss": 1.2719,
      "step": 2870
    },
    {
      "epoch": 0.730686286946594,
      "grad_norm": 1.4271516629005172,
      "learning_rate": 2.051048266446727e-06,
      "loss": 1.2652,
      "step": 2880
    },
    {
      "epoch": 0.733223392109603,
      "grad_norm": 1.4951992832082914,
      "learning_rate": 2.0153915786667203e-06,
      "loss": 1.26,
      "step": 2890
    },
    {
      "epoch": 0.735760497272612,
      "grad_norm": 1.4351479751585414,
      "learning_rate": 1.9799691555597555e-06,
      "loss": 1.2881,
      "step": 2900
    },
    {
      "epoch": 0.7382976024356209,
      "grad_norm": 1.474899241565124,
      "learning_rate": 1.9447837774645513e-06,
      "loss": 1.2702,
      "step": 2910
    },
    {
      "epoch": 0.74083470759863,
      "grad_norm": 1.4426835070499822,
      "learning_rate": 1.9098382061139503e-06,
      "loss": 1.2699,
      "step": 2920
    },
    {
      "epoch": 0.743371812761639,
      "grad_norm": 1.4876818985570295,
      "learning_rate": 1.8751351844181414e-06,
      "loss": 1.2612,
      "step": 2930
    },
    {
      "epoch": 0.745908917924648,
      "grad_norm": 1.4360645410392319,
      "learning_rate": 1.8406774362493662e-06,
      "loss": 1.2754,
      "step": 2940
    },
    {
      "epoch": 0.748446023087657,
      "grad_norm": 1.4473888665732064,
      "learning_rate": 1.8064676662281206e-06,
      "loss": 1.2902,
      "step": 2950
    },
    {
      "epoch": 0.750983128250666,
      "grad_norm": 1.4434612838312966,
      "learning_rate": 1.7725085595108682e-06,
      "loss": 1.273,
      "step": 2960
    },
    {
      "epoch": 0.753520233413675,
      "grad_norm": 1.558136105535075,
      "learning_rate": 1.7388027815792725e-06,
      "loss": 1.2787,
      "step": 2970
    },
    {
      "epoch": 0.756057338576684,
      "grad_norm": 1.4724878594646564,
      "learning_rate": 1.705352978030993e-06,
      "loss": 1.2627,
      "step": 2980
    },
    {
      "epoch": 0.758594443739693,
      "grad_norm": 1.4768497018650097,
      "learning_rate": 1.672161774372022e-06,
      "loss": 1.2911,
      "step": 2990
    },
    {
      "epoch": 0.7611315489027021,
      "grad_norm": 1.4598692131173956,
      "learning_rate": 1.639231775810602e-06,
      "loss": 1.2907,
      "step": 3000
    },
    {
      "epoch": 0.763668654065711,
      "grad_norm": 1.3971487709781405,
      "learning_rate": 1.6065655670527546e-06,
      "loss": 1.2632,
      "step": 3010
    },
    {
      "epoch": 0.76620575922872,
      "grad_norm": 1.4196228285690422,
      "learning_rate": 1.574165712099392e-06,
      "loss": 1.2542,
      "step": 3020
    },
    {
      "epoch": 0.768742864391729,
      "grad_norm": 1.4395590200787511,
      "learning_rate": 1.542034754045067e-06,
      "loss": 1.2693,
      "step": 3030
    },
    {
      "epoch": 0.7712799695547381,
      "grad_norm": 1.4538143237649903,
      "learning_rate": 1.5101752148783705e-06,
      "loss": 1.2728,
      "step": 3040
    },
    {
      "epoch": 0.7738170747177471,
      "grad_norm": 1.4483981816763403,
      "learning_rate": 1.4785895952839735e-06,
      "loss": 1.2671,
      "step": 3050
    },
    {
      "epoch": 0.7763541798807561,
      "grad_norm": 1.5335192207213328,
      "learning_rate": 1.447280374446346e-06,
      "loss": 1.2778,
      "step": 3060
    },
    {
      "epoch": 0.778891285043765,
      "grad_norm": 1.4504666284348766,
      "learning_rate": 1.4162500098551608e-06,
      "loss": 1.276,
      "step": 3070
    },
    {
      "epoch": 0.7814283902067741,
      "grad_norm": 1.454412830474016,
      "learning_rate": 1.385500937112415e-06,
      "loss": 1.2804,
      "step": 3080
    },
    {
      "epoch": 0.7839654953697831,
      "grad_norm": 1.462536001446098,
      "learning_rate": 1.3550355697412386e-06,
      "loss": 1.2586,
      "step": 3090
    },
    {
      "epoch": 0.7865026005327921,
      "grad_norm": 1.4861860594882876,
      "learning_rate": 1.3248562989964719e-06,
      "loss": 1.2843,
      "step": 3100
    },
    {
      "epoch": 0.7890397056958011,
      "grad_norm": 1.391241218546658,
      "learning_rate": 1.2949654936769622e-06,
      "loss": 1.2723,
      "step": 3110
    },
    {
      "epoch": 0.79157681085881,
      "grad_norm": 1.412674356321388,
      "learning_rate": 1.2653654999396436e-06,
      "loss": 1.2621,
      "step": 3120
    },
    {
      "epoch": 0.7941139160218191,
      "grad_norm": 1.406341007739084,
      "learning_rate": 1.2360586411153747e-06,
      "loss": 1.2897,
      "step": 3130
    },
    {
      "epoch": 0.7966510211848281,
      "grad_norm": 1.4125498017483746,
      "learning_rate": 1.2070472175265857e-06,
      "loss": 1.2657,
      "step": 3140
    },
    {
      "epoch": 0.7991881263478371,
      "grad_norm": 1.575395352386111,
      "learning_rate": 1.1783335063067286e-06,
      "loss": 1.2974,
      "step": 3150
    },
    {
      "epoch": 0.8017252315108461,
      "grad_norm": 1.4632409776646316,
      "learning_rate": 1.1499197612215269e-06,
      "loss": 1.2914,
      "step": 3160
    },
    {
      "epoch": 0.8042623366738552,
      "grad_norm": 1.420307782085356,
      "learning_rate": 1.1218082124920903e-06,
      "loss": 1.2583,
      "step": 3170
    },
    {
      "epoch": 0.8067994418368641,
      "grad_norm": 1.4073518626370982,
      "learning_rate": 1.0940010666198575e-06,
      "loss": 1.2588,
      "step": 3180
    },
    {
      "epoch": 0.8093365469998731,
      "grad_norm": 1.427712685491864,
      "learning_rate": 1.0665005062134015e-06,
      "loss": 1.2641,
      "step": 3190
    },
    {
      "epoch": 0.8118736521628821,
      "grad_norm": 1.4042233353051128,
      "learning_rate": 1.0393086898171234e-06,
      "loss": 1.2623,
      "step": 3200
    },
    {
      "epoch": 0.8144107573258912,
      "grad_norm": 1.4353551061219325,
      "learning_rate": 1.0124277517418196e-06,
      "loss": 1.2701,
      "step": 3210
    },
    {
      "epoch": 0.8169478624889002,
      "grad_norm": 1.4714738106408498,
      "learning_rate": 9.858598018971599e-07,
      "loss": 1.2665,
      "step": 3220
    },
    {
      "epoch": 0.8194849676519091,
      "grad_norm": 1.3867710691517015,
      "learning_rate": 9.596069256260792e-07,
      "loss": 1.2811,
      "step": 3230
    },
    {
      "epoch": 0.8220220728149181,
      "grad_norm": 1.5556697110120234,
      "learning_rate": 9.336711835410972e-07,
      "loss": 1.2577,
      "step": 3240
    },
    {
      "epoch": 0.8245591779779272,
      "grad_norm": 1.4677232808017586,
      "learning_rate": 9.080546113625738e-07,
      "loss": 1.2675,
      "step": 3250
    },
    {
      "epoch": 0.8270962831409362,
      "grad_norm": 1.4009076826572764,
      "learning_rate": 8.827592197589341e-07,
      "loss": 1.2573,
      "step": 3260
    },
    {
      "epoch": 0.8296333883039452,
      "grad_norm": 1.3584163410682717,
      "learning_rate": 8.577869941888389e-07,
      "loss": 1.2654,
      "step": 3270
    },
    {
      "epoch": 0.8321704934669542,
      "grad_norm": 1.4763004523041792,
      "learning_rate": 8.331398947453512e-07,
      "loss": 1.271,
      "step": 3280
    },
    {
      "epoch": 0.8347075986299632,
      "grad_norm": 1.432107775038367,
      "learning_rate": 8.08819856002081e-07,
      "loss": 1.2771,
      "step": 3290
    },
    {
      "epoch": 0.8372447037929722,
      "grad_norm": 1.4636493219573536,
      "learning_rate": 7.848287868613441e-07,
      "loss": 1.2511,
      "step": 3300
    },
    {
      "epoch": 0.8397818089559812,
      "grad_norm": 1.501456048501624,
      "learning_rate": 7.611685704043281e-07,
      "loss": 1.2724,
      "step": 3310
    },
    {
      "epoch": 0.8423189141189902,
      "grad_norm": 1.4533194309629769,
      "learning_rate": 7.378410637432848e-07,
      "loss": 1.2761,
      "step": 3320
    },
    {
      "epoch": 0.8448560192819993,
      "grad_norm": 1.4930040288043631,
      "learning_rate": 7.148480978757694e-07,
      "loss": 1.2808,
      "step": 3330
    },
    {
      "epoch": 0.8473931244450083,
      "grad_norm": 1.4327631888495864,
      "learning_rate": 6.921914775409211e-07,
      "loss": 1.2764,
      "step": 3340
    },
    {
      "epoch": 0.8499302296080172,
      "grad_norm": 1.4066505744498654,
      "learning_rate": 6.698729810778065e-07,
      "loss": 1.2724,
      "step": 3350
    },
    {
      "epoch": 0.8524673347710262,
      "grad_norm": 1.4214214237910756,
      "learning_rate": 6.478943602858373e-07,
      "loss": 1.2703,
      "step": 3360
    },
    {
      "epoch": 0.8550044399340353,
      "grad_norm": 1.4609797404161982,
      "learning_rate": 6.262573402872707e-07,
      "loss": 1.2702,
      "step": 3370
    },
    {
      "epoch": 0.8575415450970443,
      "grad_norm": 1.4426076897314533,
      "learning_rate": 6.04963619391799e-07,
      "loss": 1.2652,
      "step": 3380
    },
    {
      "epoch": 0.8600786502600533,
      "grad_norm": 1.4557782632700174,
      "learning_rate": 5.840148689632536e-07,
      "loss": 1.2628,
      "step": 3390
    },
    {
      "epoch": 0.8626157554230622,
      "grad_norm": 1.4465949605683495,
      "learning_rate": 5.634127332884143e-07,
      "loss": 1.2649,
      "step": 3400
    },
    {
      "epoch": 0.8651528605860713,
      "grad_norm": 1.4543385042228827,
      "learning_rate": 5.431588294479479e-07,
      "loss": 1.2863,
      "step": 3410
    },
    {
      "epoch": 0.8676899657490803,
      "grad_norm": 1.4211990179028964,
      "learning_rate": 5.232547471894839e-07,
      "loss": 1.2603,
      "step": 3420
    },
    {
      "epoch": 0.8702270709120893,
      "grad_norm": 1.4970252757505178,
      "learning_rate": 5.037020488028322e-07,
      "loss": 1.2659,
      "step": 3430
    },
    {
      "epoch": 0.8727641760750983,
      "grad_norm": 1.476846612856639,
      "learning_rate": 4.845022689973567e-07,
      "loss": 1.2622,
      "step": 3440
    },
    {
      "epoch": 0.8753012812381074,
      "grad_norm": 1.3975933791175674,
      "learning_rate": 4.656569147815171e-07,
      "loss": 1.2675,
      "step": 3450
    },
    {
      "epoch": 0.8778383864011163,
      "grad_norm": 1.4066115466115592,
      "learning_rate": 4.471674653445801e-07,
      "loss": 1.2657,
      "step": 3460
    },
    {
      "epoch": 0.8803754915641253,
      "grad_norm": 1.4093052515025426,
      "learning_rate": 4.290353719405199e-07,
      "loss": 1.2622,
      "step": 3470
    },
    {
      "epoch": 0.8829125967271343,
      "grad_norm": 1.517434872609148,
      "learning_rate": 4.1126205777410054e-07,
      "loss": 1.2658,
      "step": 3480
    },
    {
      "epoch": 0.8854497018901434,
      "grad_norm": 1.3671982309013966,
      "learning_rate": 3.938489178891769e-07,
      "loss": 1.26,
      "step": 3490
    },
    {
      "epoch": 0.8879868070531524,
      "grad_norm": 1.3848746120915914,
      "learning_rate": 3.767973190591906e-07,
      "loss": 1.252,
      "step": 3500
    },
    {
      "epoch": 0.8905239122161613,
      "grad_norm": 1.404335518132526,
      "learning_rate": 3.6010859967988975e-07,
      "loss": 1.2684,
      "step": 3510
    },
    {
      "epoch": 0.8930610173791703,
      "grad_norm": 1.4696908362694405,
      "learning_rate": 3.437840696642797e-07,
      "loss": 1.28,
      "step": 3520
    },
    {
      "epoch": 0.8955981225421794,
      "grad_norm": 1.4394982066957633,
      "learning_rate": 3.2782501033980897e-07,
      "loss": 1.2596,
      "step": 3530
    },
    {
      "epoch": 0.8981352277051884,
      "grad_norm": 1.441827806292722,
      "learning_rate": 3.1223267434778934e-07,
      "loss": 1.2548,
      "step": 3540
    },
    {
      "epoch": 0.9006723328681974,
      "grad_norm": 1.4029572337771223,
      "learning_rate": 2.9700828554508175e-07,
      "loss": 1.2714,
      "step": 3550
    },
    {
      "epoch": 0.9032094380312063,
      "grad_norm": 1.456563644801128,
      "learning_rate": 2.82153038908034e-07,
      "loss": 1.271,
      "step": 3560
    },
    {
      "epoch": 0.9057465431942154,
      "grad_norm": 1.4854658320433165,
      "learning_rate": 2.6766810043867996e-07,
      "loss": 1.2636,
      "step": 3570
    },
    {
      "epoch": 0.9082836483572244,
      "grad_norm": 1.3976434876269141,
      "learning_rate": 2.53554607073227e-07,
      "loss": 1.2555,
      "step": 3580
    },
    {
      "epoch": 0.9108207535202334,
      "grad_norm": 1.4520566750739115,
      "learning_rate": 2.3981366659281135e-07,
      "loss": 1.2741,
      "step": 3590
    },
    {
      "epoch": 0.9133578586832424,
      "grad_norm": 1.431635462724296,
      "learning_rate": 2.2644635753654832e-07,
      "loss": 1.2641,
      "step": 3600
    },
    {
      "epoch": 0.9158949638462515,
      "grad_norm": 1.4459695881350807,
      "learning_rate": 2.1345372911687868e-07,
      "loss": 1.2719,
      "step": 3610
    },
    {
      "epoch": 0.9184320690092604,
      "grad_norm": 1.4523937909973577,
      "learning_rate": 2.008368011372136e-07,
      "loss": 1.2574,
      "step": 3620
    },
    {
      "epoch": 0.9209691741722694,
      "grad_norm": 1.448000545062192,
      "learning_rate": 1.8859656391188918e-07,
      "loss": 1.2678,
      "step": 3630
    },
    {
      "epoch": 0.9235062793352784,
      "grad_norm": 1.4574912967654818,
      "learning_rate": 1.7673397818843696e-07,
      "loss": 1.2631,
      "step": 3640
    },
    {
      "epoch": 0.9260433844982875,
      "grad_norm": 1.469570266612804,
      "learning_rate": 1.65249975072172e-07,
      "loss": 1.2676,
      "step": 3650
    },
    {
      "epoch": 0.9285804896612965,
      "grad_norm": 1.4705072480655184,
      "learning_rate": 1.5414545595311193e-07,
      "loss": 1.2363,
      "step": 3660
    },
    {
      "epoch": 0.9311175948243055,
      "grad_norm": 1.469943074464241,
      "learning_rate": 1.4342129243522241e-07,
      "loss": 1.2716,
      "step": 3670
    },
    {
      "epoch": 0.9336546999873144,
      "grad_norm": 1.4039691988667693,
      "learning_rate": 1.3307832626800966e-07,
      "loss": 1.2674,
      "step": 3680
    },
    {
      "epoch": 0.9361918051503235,
      "grad_norm": 1.4357184274036978,
      "learning_rate": 1.2311736928044437e-07,
      "loss": 1.2662,
      "step": 3690
    },
    {
      "epoch": 0.9387289103133325,
      "grad_norm": 1.4198942808558601,
      "learning_rate": 1.1353920331724666e-07,
      "loss": 1.2743,
      "step": 3700
    },
    {
      "epoch": 0.9412660154763415,
      "grad_norm": 1.472183542066415,
      "learning_rate": 1.0434458017751392e-07,
      "loss": 1.2505,
      "step": 3710
    },
    {
      "epoch": 0.9438031206393505,
      "grad_norm": 1.4530560276575668,
      "learning_rate": 9.553422155571257e-08,
      "loss": 1.2637,
      "step": 3720
    },
    {
      "epoch": 0.9463402258023595,
      "grad_norm": 1.4342800198901315,
      "learning_rate": 8.710881898503276e-08,
      "loss": 1.2706,
      "step": 3730
    },
    {
      "epoch": 0.9488773309653685,
      "grad_norm": 1.4769111711160674,
      "learning_rate": 7.906903378310738e-08,
      "loss": 1.2717,
      "step": 3740
    },
    {
      "epoch": 0.9514144361283775,
      "grad_norm": 1.4434198544006103,
      "learning_rate": 7.141549700010741e-08,
      "loss": 1.2764,
      "step": 3750
    },
    {
      "epoch": 0.9539515412913865,
      "grad_norm": 1.3967239883734883,
      "learning_rate": 6.414880936920665e-08,
      "loss": 1.2454,
      "step": 3760
    },
    {
      "epoch": 0.9564886464543956,
      "grad_norm": 1.4745766429828837,
      "learning_rate": 5.726954125943318e-08,
      "loss": 1.2747,
      "step": 3770
    },
    {
      "epoch": 0.9590257516174046,
      "grad_norm": 1.3354472782196753,
      "learning_rate": 5.0778232630897536e-08,
      "loss": 1.2717,
      "step": 3780
    },
    {
      "epoch": 0.9615628567804135,
      "grad_norm": 1.5048063707346293,
      "learning_rate": 4.4675392992412634e-08,
      "loss": 1.2728,
      "step": 3790
    },
    {
      "epoch": 0.9640999619434225,
      "grad_norm": 1.4399837174477832,
      "learning_rate": 3.896150136150134e-08,
      "loss": 1.2826,
      "step": 3800
    },
    {
      "epoch": 0.9666370671064316,
      "grad_norm": 1.4450380885385077,
      "learning_rate": 3.3637006226797665e-08,
      "loss": 1.2534,
      "step": 3810
    },
    {
      "epoch": 0.9691741722694406,
      "grad_norm": 1.35673950609508,
      "learning_rate": 2.8702325512844908e-08,
      "loss": 1.2609,
      "step": 3820
    },
    {
      "epoch": 0.9717112774324496,
      "grad_norm": 1.43190405721669,
      "learning_rate": 2.4157846547292473e-08,
      "loss": 1.2787,
      "step": 3830
    },
    {
      "epoch": 0.9742483825954585,
      "grad_norm": 1.4245067018508633,
      "learning_rate": 2.000392603049517e-08,
      "loss": 1.2665,
      "step": 3840
    },
    {
      "epoch": 0.9767854877584676,
      "grad_norm": 1.4112885660620007,
      "learning_rate": 1.6240890007510612e-08,
      "loss": 1.2785,
      "step": 3850
    },
    {
      "epoch": 0.9793225929214766,
      "grad_norm": 1.499341122900986,
      "learning_rate": 1.286903384251581e-08,
      "loss": 1.2539,
      "step": 3860
    },
    {
      "epoch": 0.9818596980844856,
      "grad_norm": 1.4494246005853764,
      "learning_rate": 9.888622195615705e-09,
      "loss": 1.2725,
      "step": 3870
    },
    {
      "epoch": 0.9843968032474946,
      "grad_norm": 1.421431046100693,
      "learning_rate": 7.299889002075344e-09,
      "loss": 1.2726,
      "step": 3880
    },
    {
      "epoch": 0.9869339084105037,
      "grad_norm": 1.4209346862475516,
      "learning_rate": 5.103037453954573e-09,
      "loss": 1.2548,
      "step": 3890
    },
    {
      "epoch": 0.9894710135735126,
      "grad_norm": 1.3751824191752333,
      "learning_rate": 3.2982399841618996e-09,
      "loss": 1.2859,
      "step": 3900
    },
    {
      "epoch": 0.9920081187365216,
      "grad_norm": 1.4627888810880714,
      "learning_rate": 1.8856382529192085e-09,
      "loss": 1.2842,
      "step": 3910
    },
    {
      "epoch": 0.9945452238995306,
      "grad_norm": 1.3905016842842302,
      "learning_rate": 8.653431366406617e-10,
      "loss": 1.2447,
      "step": 3920
    },
    {
      "epoch": 0.9970823290625397,
      "grad_norm": 1.5470180096733397,
      "learning_rate": 2.374347192335424e-10,
      "loss": 1.2707,
      "step": 3930
    },
    {
      "epoch": 0.9996194342255487,
      "grad_norm": 1.4506027900307656,
      "learning_rate": 1.9622858088430564e-12,
      "loss": 1.2737,
      "step": 3940
    },
    {
      "epoch": 0.9998731447418495,
      "step": 3941,
      "total_flos": 3.7575827488610714e+18,
      "train_loss": 1.3289946492206504,
      "train_runtime": 13442.1001,
      "train_samples_per_second": 37.53,
      "train_steps_per_second": 0.293
    }
  ],
  "logging_steps": 10,
  "max_steps": 3941,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.7575827488610714e+18,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}