{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.91616766467066,
  "eval_steps": 500,
  "global_step": 1110,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08982035928143713,
      "grad_norm": 49.68221664428711,
      "learning_rate": 1.4414414414414416e-06,
      "loss": 8.7467,
      "step": 10
    },
    {
      "epoch": 0.17964071856287425,
      "grad_norm": 22.61308479309082,
      "learning_rate": 3.063063063063063e-06,
      "loss": 5.5972,
      "step": 20
    },
    {
      "epoch": 0.2694610778443114,
      "grad_norm": 11.650931358337402,
      "learning_rate": 4.864864864864866e-06,
      "loss": 4.1239,
      "step": 30
    },
    {
      "epoch": 0.3592814371257485,
      "grad_norm": 11.040465354919434,
      "learning_rate": 6.666666666666667e-06,
      "loss": 3.3473,
      "step": 40
    },
    {
      "epoch": 0.4491017964071856,
      "grad_norm": 21.185564041137695,
      "learning_rate": 8.46846846846847e-06,
      "loss": 2.843,
      "step": 50
    },
    {
      "epoch": 0.5389221556886228,
      "grad_norm": 6.222358226776123,
      "learning_rate": 1.027027027027027e-05,
      "loss": 2.5158,
      "step": 60
    },
    {
      "epoch": 0.6287425149700598,
      "grad_norm": 12.189018249511719,
      "learning_rate": 1.2072072072072074e-05,
      "loss": 2.1371,
      "step": 70
    },
    {
      "epoch": 0.718562874251497,
      "grad_norm": 17.56637191772461,
      "learning_rate": 1.3873873873873875e-05,
      "loss": 1.9438,
      "step": 80
    },
    {
      "epoch": 0.8083832335329342,
      "grad_norm": 5.583638668060303,
      "learning_rate": 1.5675675675675676e-05,
      "loss": 1.7556,
      "step": 90
    },
    {
      "epoch": 0.8982035928143712,
      "grad_norm": 3.642561197280884,
      "learning_rate": 1.7477477477477477e-05,
      "loss": 1.6005,
      "step": 100
    },
    {
      "epoch": 0.9880239520958084,
      "grad_norm": 2.80586314201355,
      "learning_rate": 1.927927927927928e-05,
      "loss": 1.5588,
      "step": 110
    },
    {
      "epoch": 1.0718562874251496,
      "grad_norm": 2.6551783084869385,
      "learning_rate": 1.9998219965624736e-05,
      "loss": 1.4294,
      "step": 120
    },
    {
      "epoch": 1.1616766467065869,
      "grad_norm": 2.895826578140259,
      "learning_rate": 1.9987344272588007e-05,
      "loss": 1.4837,
      "step": 130
    },
    {
      "epoch": 1.251497005988024,
      "grad_norm": 2.8669326305389404,
      "learning_rate": 1.9966592535953532e-05,
      "loss": 1.453,
      "step": 140
    },
    {
      "epoch": 1.341317365269461,
      "grad_norm": 2.1909446716308594,
      "learning_rate": 1.9935985276197033e-05,
      "loss": 1.3837,
      "step": 150
    },
    {
      "epoch": 1.4311377245508983,
      "grad_norm": 3.2299187183380127,
      "learning_rate": 1.989555275948572e-05,
      "loss": 1.3152,
      "step": 160
    },
    {
      "epoch": 1.5209580838323353,
      "grad_norm": 1.9108384847640991,
      "learning_rate": 1.984533496774942e-05,
      "loss": 1.1961,
      "step": 170
    },
    {
      "epoch": 1.6107784431137726,
      "grad_norm": 1.1921658515930176,
      "learning_rate": 1.9785381559144196e-05,
      "loss": 1.1209,
      "step": 180
    },
    {
      "epoch": 1.7005988023952097,
      "grad_norm": 0.8914986848831177,
      "learning_rate": 1.9715751818947603e-05,
      "loss": 1.1056,
      "step": 190
    },
    {
      "epoch": 1.7904191616766467,
      "grad_norm": 0.9408266544342041,
      "learning_rate": 1.963651460093409e-05,
      "loss": 1.0827,
      "step": 200
    },
    {
      "epoch": 1.8802395209580838,
      "grad_norm": 0.8248458504676819,
      "learning_rate": 1.9547748259288536e-05,
      "loss": 1.0787,
      "step": 210
    },
    {
      "epoch": 1.9700598802395208,
      "grad_norm": 0.9938483238220215,
      "learning_rate": 1.9449540571125284e-05,
      "loss": 1.0597,
      "step": 220
    },
    {
      "epoch": 2.053892215568862,
      "grad_norm": 0.9142336249351501,
      "learning_rate": 1.93419886496892e-05,
      "loss": 0.9906,
      "step": 230
    },
    {
      "epoch": 2.143712574850299,
      "grad_norm": 0.8745118379592896,
      "learning_rate": 1.9225198848324687e-05,
      "loss": 1.05,
      "step": 240
    },
    {
      "epoch": 2.2335329341317367,
      "grad_norm": 0.7089764475822449,
      "learning_rate": 1.909928665530757e-05,
      "loss": 1.043,
      "step": 250
    },
    {
      "epoch": 2.3233532934131738,
      "grad_norm": 0.4864867031574249,
      "learning_rate": 1.896437657964382e-05,
      "loss": 1.0463,
      "step": 260
    },
    {
      "epoch": 2.413173652694611,
      "grad_norm": 0.971626341342926,
      "learning_rate": 1.8820602027948112e-05,
      "loss": 1.0418,
      "step": 270
    },
    {
      "epoch": 2.502994011976048,
      "grad_norm": 0.699500322341919,
      "learning_rate": 1.866810517252393e-05,
      "loss": 1.038,
      "step": 280
    },
    {
      "epoch": 2.592814371257485,
      "grad_norm": 0.9161490797996521,
      "learning_rate": 1.8507036810775617e-05,
      "loss": 1.0338,
      "step": 290
    },
    {
      "epoch": 2.682634730538922,
      "grad_norm": 0.6857994198799133,
      "learning_rate": 1.833755621609152e-05,
      "loss": 1.0231,
      "step": 300
    },
    {
      "epoch": 2.772455089820359,
      "grad_norm": 0.7041919827461243,
      "learning_rate": 1.815983098034555e-05,
      "loss": 1.0218,
      "step": 310
    },
    {
      "epoch": 2.8622754491017965,
      "grad_norm": 0.9268773794174194,
      "learning_rate": 1.7974036848172992e-05,
      "loss": 1.0229,
      "step": 320
    },
    {
      "epoch": 2.9520958083832336,
      "grad_norm": 0.7882820963859558,
      "learning_rate": 1.7780357543184396e-05,
      "loss": 1.0199,
      "step": 330
    },
    {
      "epoch": 3.035928143712575,
      "grad_norm": 0.7048642635345459,
      "learning_rate": 1.757898458628941e-05,
      "loss": 0.9551,
      "step": 340
    },
    {
      "epoch": 3.125748502994012,
      "grad_norm": 0.5380260348320007,
      "learning_rate": 1.7370117106310216e-05,
      "loss": 1.0078,
      "step": 350
    },
    {
      "epoch": 3.215568862275449,
      "grad_norm": 0.9133287668228149,
      "learning_rate": 1.715396164307182e-05,
      "loss": 1.0026,
      "step": 360
    },
    {
      "epoch": 3.305389221556886,
      "grad_norm": 0.8086357712745667,
      "learning_rate": 1.6930731943163975e-05,
      "loss": 1.0082,
      "step": 370
    },
    {
      "epoch": 3.3952095808383236,
      "grad_norm": 0.7032626271247864,
      "learning_rate": 1.6700648748576577e-05,
      "loss": 1.0025,
      "step": 380
    },
    {
      "epoch": 3.4850299401197606,
      "grad_norm": 0.7610228657722473,
      "learning_rate": 1.6463939578417695e-05,
      "loss": 0.9943,
      "step": 390
    },
    {
      "epoch": 3.5748502994011977,
      "grad_norm": 0.7735270261764526,
      "learning_rate": 1.622083850392996e-05,
      "loss": 1.0014,
      "step": 400
    },
    {
      "epoch": 3.6646706586826348,
      "grad_norm": 0.5382483005523682,
      "learning_rate": 1.5971585917027864e-05,
      "loss": 0.9961,
      "step": 410
    },
    {
      "epoch": 3.754491017964072,
      "grad_norm": 0.6216614246368408,
      "learning_rate": 1.5716428292584788e-05,
      "loss": 0.992,
      "step": 420
    },
    {
      "epoch": 3.844311377245509,
      "grad_norm": 0.5743616819381714,
      "learning_rate": 1.545561794470492e-05,
      "loss": 0.9861,
      "step": 430
    },
    {
      "epoch": 3.934131736526946,
      "grad_norm": 0.7634334564208984,
      "learning_rate": 1.518941277722096e-05,
      "loss": 0.9881,
      "step": 440
    },
    {
      "epoch": 4.017964071856287,
      "grad_norm": 0.7453029751777649,
      "learning_rate": 1.491807602866442e-05,
      "loss": 0.9172,
      "step": 450
    },
    {
      "epoch": 4.107784431137724,
      "grad_norm": 1.0462350845336914,
      "learning_rate": 1.4641876011960661e-05,
      "loss": 0.9798,
      "step": 460
    },
    {
      "epoch": 4.197604790419161,
      "grad_norm": 0.8113179206848145,
      "learning_rate": 1.436108584910611e-05,
      "loss": 0.9799,
      "step": 470
    },
    {
      "epoch": 4.287425149700598,
      "grad_norm": 0.6361674070358276,
      "learning_rate": 1.4075983201089964e-05,
      "loss": 0.9654,
      "step": 480
    },
    {
      "epoch": 4.3772455089820355,
      "grad_norm": 0.6291260123252869,
      "learning_rate": 1.3786849993327503e-05,
      "loss": 0.9631,
      "step": 490
    },
    {
      "epoch": 4.467065868263473,
      "grad_norm": 0.988298773765564,
      "learning_rate": 1.349397213687651e-05,
      "loss": 0.9729,
      "step": 500
    },
    {
      "epoch": 4.5568862275449105,
      "grad_norm": 1.0734843015670776,
      "learning_rate": 1.3197639245712454e-05,
      "loss": 0.963,
      "step": 510
    },
    {
      "epoch": 4.6467065868263475,
      "grad_norm": 0.834682047367096,
      "learning_rate": 1.2898144350342015e-05,
      "loss": 0.9538,
      "step": 520
    },
    {
      "epoch": 4.736526946107785,
      "grad_norm": 0.5939741134643555,
      "learning_rate": 1.2595783608038157e-05,
      "loss": 0.9623,
      "step": 530
    },
    {
      "epoch": 4.826347305389222,
      "grad_norm": 0.8625423908233643,
      "learning_rate": 1.22908560099833e-05,
      "loss": 0.9557,
      "step": 540
    },
    {
      "epoch": 4.916167664670659,
      "grad_norm": 0.8925888538360596,
      "learning_rate": 1.198366308561013e-05,
      "loss": 0.9419,
      "step": 550
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3238130807876587,
      "learning_rate": 1.1674508604432464e-05,
      "loss": 0.8782,
      "step": 560
    },
    {
      "epoch": 5.089820359281437,
      "grad_norm": 0.5171638131141663,
      "learning_rate": 1.1363698275661002e-05,
      "loss": 0.9256,
      "step": 570
    },
    {
      "epoch": 5.179640718562874,
      "grad_norm": 1.136551022529602,
      "learning_rate": 1.1051539445900982e-05,
      "loss": 0.9282,
      "step": 580
    },
    {
      "epoch": 5.269461077844311,
      "grad_norm": 0.9350560307502747,
      "learning_rate": 1.0738340795230722e-05,
      "loss": 0.9316,
      "step": 590
    },
    {
      "epoch": 5.359281437125748,
      "grad_norm": 1.0158970355987549,
      "learning_rate": 1.0424412031961485e-05,
      "loss": 0.9381,
      "step": 600
    },
    {
      "epoch": 5.449101796407185,
      "grad_norm": 1.0037592649459839,
      "learning_rate": 1.0110063586380647e-05,
      "loss": 0.9288,
      "step": 610
    },
    {
      "epoch": 5.538922155688622,
      "grad_norm": 0.6334489583969116,
      "learning_rate": 9.795606303780885e-06,
      "loss": 0.9272,
      "step": 620
    },
    {
      "epoch": 5.62874251497006,
      "grad_norm": 0.7240671515464783,
      "learning_rate": 9.48135113707899e-06,
      "loss": 0.9318,
      "step": 630
    },
    {
      "epoch": 5.718562874251497,
      "grad_norm": 0.9315741658210754,
      "learning_rate": 9.167608839328273e-06,
      "loss": 0.9297,
      "step": 640
    },
    {
      "epoch": 5.808383233532934,
      "grad_norm": 0.7542333602905273,
      "learning_rate": 8.854689656428591e-06,
      "loss": 0.9119,
      "step": 650
    },
    {
      "epoch": 5.8982035928143715,
      "grad_norm": 0.864987313747406,
      "learning_rate": 8.542903020337887e-06,
      "loss": 0.919,
      "step": 660
    },
    {
      "epoch": 5.9880239520958085,
      "grad_norm": 0.5868083834648132,
      "learning_rate": 8.232557243088585e-06,
      "loss": 0.9136,
      "step": 670
    },
    {
      "epoch": 6.07185628742515,
      "grad_norm": 0.8045146465301514,
      "learning_rate": 7.923959211911448e-06,
      "loss": 0.8452,
      "step": 680
    },
    {
      "epoch": 6.161676646706587,
      "grad_norm": 0.9145955443382263,
      "learning_rate": 7.617414085768352e-06,
      "loss": 0.9,
      "step": 690
    },
    {
      "epoch": 6.251497005988024,
      "grad_norm": 0.6515654921531677,
      "learning_rate": 7.313224993594057e-06,
      "loss": 0.8946,
      "step": 700
    },
    {
      "epoch": 6.341317365269461,
      "grad_norm": 0.6843001246452332,
      "learning_rate": 7.011692734545403e-06,
      "loss": 0.8994,
      "step": 710
    },
    {
      "epoch": 6.431137724550898,
      "grad_norm": 0.8299969434738159,
      "learning_rate": 6.713115480554313e-06,
      "loss": 0.894,
      "step": 720
    },
    {
      "epoch": 6.520958083832335,
      "grad_norm": 1.121025800704956,
      "learning_rate": 6.4177884814787284e-06,
      "loss": 0.8982,
      "step": 730
    },
    {
      "epoch": 6.610778443113772,
      "grad_norm": 1.1040045022964478,
      "learning_rate": 6.126003773143072e-06,
      "loss": 0.8985,
      "step": 740
    },
    {
      "epoch": 6.700598802395209,
      "grad_norm": 0.711075484752655,
      "learning_rate": 5.8380498885569246e-06,
      "loss": 0.8923,
      "step": 750
    },
    {
      "epoch": 6.790419161676647,
      "grad_norm": 0.8444594740867615,
      "learning_rate": 5.554211572597477e-06,
      "loss": 0.8989,
      "step": 760
    },
    {
      "epoch": 6.880239520958084,
      "grad_norm": 0.9748265743255615,
      "learning_rate": 5.274769500437882e-06,
      "loss": 0.8956,
      "step": 770
    },
    {
      "epoch": 6.970059880239521,
      "grad_norm": 0.8513533473014832,
      "learning_rate": 5.000000000000003e-06,
      "loss": 0.8906,
      "step": 780
    },
    {
      "epoch": 7.053892215568863,
      "grad_norm": 1.3123410940170288,
      "learning_rate": 4.730174778705909e-06,
      "loss": 0.8249,
      "step": 790
    },
    {
      "epoch": 7.1437125748503,
      "grad_norm": 1.0771753787994385,
      "learning_rate": 4.4655606547984165e-06,
      "loss": 0.8777,
      "step": 800
    },
    {
      "epoch": 7.233532934131737,
      "grad_norm": 0.8778141736984253,
      "learning_rate": 4.206419293496333e-06,
      "loss": 0.8832,
      "step": 810
    },
    {
      "epoch": 7.323353293413174,
      "grad_norm": 1.0552406311035156,
      "learning_rate": 3.953006948245247e-06,
      "loss": 0.872,
      "step": 820
    },
    {
      "epoch": 7.413173652694611,
      "grad_norm": 0.6986867785453796,
      "learning_rate": 3.705574207319844e-06,
      "loss": 0.8807,
      "step": 830
    },
    {
      "epoch": 7.502994011976048,
      "grad_norm": 0.7952435612678528,
      "learning_rate": 3.4643657460282078e-06,
      "loss": 0.8793,
      "step": 840
    },
    {
      "epoch": 7.592814371257485,
      "grad_norm": 0.9549069404602051,
      "learning_rate": 3.2296200847632376e-06,
      "loss": 0.8751,
      "step": 850
    },
    {
      "epoch": 7.682634730538922,
      "grad_norm": 0.6770684719085693,
      "learning_rate": 3.001569353140347e-06,
      "loss": 0.8772,
      "step": 860
    },
    {
      "epoch": 7.772455089820359,
      "grad_norm": 0.8119450211524963,
      "learning_rate": 2.780439060454756e-06,
      "loss": 0.8721,
      "step": 870
    },
    {
      "epoch": 7.862275449101796,
      "grad_norm": 0.7255963087081909,
      "learning_rate": 2.566447872685298e-06,
      "loss": 0.8734,
      "step": 880
    },
    {
      "epoch": 7.952095808383233,
      "grad_norm": 0.6024225950241089,
      "learning_rate": 2.359807396265307e-06,
      "loss": 0.8779,
      "step": 890
    },
    {
      "epoch": 8.035928143712574,
      "grad_norm": 0.6504734754562378,
      "learning_rate": 2.160721968834344e-06,
      "loss": 0.8174,
      "step": 900
    },
    {
      "epoch": 8.125748502994012,
      "grad_norm": 0.8142725825309753,
      "learning_rate": 1.969388457177743e-06,
      "loss": 0.863,
      "step": 910
    },
    {
      "epoch": 8.215568862275449,
      "grad_norm": 0.8136929869651794,
      "learning_rate": 1.7859960625537476e-06,
      "loss": 0.8677,
      "step": 920
    },
    {
      "epoch": 8.305389221556887,
      "grad_norm": 0.6553688049316406,
      "learning_rate": 1.6107261336007284e-06,
      "loss": 0.853,
      "step": 930
    },
    {
      "epoch": 8.395209580838323,
      "grad_norm": 0.8645008206367493,
      "learning_rate": 1.443751987009533e-06,
      "loss": 0.8569,
      "step": 940
    },
    {
      "epoch": 8.48502994011976,
      "grad_norm": 0.7802151441574097,
      "learning_rate": 1.2852387361382768e-06,
      "loss": 0.8608,
      "step": 950
    },
    {
      "epoch": 8.574850299401197,
      "grad_norm": 0.8659719824790955,
      "learning_rate": 1.1353431277390125e-06,
      "loss": 0.8723,
      "step": 960
    },
    {
      "epoch": 8.664670658682635,
      "grad_norm": 1.073027491569519,
      "learning_rate": 9.942133869578164e-07,
      "loss": 0.8538,
      "step": 970
    },
    {
      "epoch": 8.754491017964071,
      "grad_norm": 0.8132328391075134,
      "learning_rate": 8.619890707614687e-07,
      "loss": 0.8552,
      "step": 980
    },
    {
      "epoch": 8.844311377245509,
      "grad_norm": 0.959922730922699,
      "learning_rate": 7.388009299357412e-07,
      "loss": 0.8658,
      "step": 990
    },
    {
      "epoch": 8.934131736526947,
      "grad_norm": 0.8655376434326172,
      "learning_rate": 6.247707797917257e-07,
      "loss": 0.8522,
      "step": 1000
    },
    {
      "epoch": 9.017964071856287,
      "grad_norm": 0.8419204950332642,
      "learning_rate": 5.200113797080464e-07,
      "loss": 0.7919,
      "step": 1010
    },
    {
      "epoch": 9.107784431137725,
      "grad_norm": 0.9591242074966431,
      "learning_rate": 4.2462632162809103e-07,
      "loss": 0.857,
      "step": 1020
    },
    {
      "epoch": 9.197604790419161,
      "grad_norm": 0.4990997016429901,
      "learning_rate": 3.387099276225214e-07,
      "loss": 0.8503,
      "step": 1030
    },
    {
      "epoch": 9.2874251497006,
      "grad_norm": 0.9034556150436401,
      "learning_rate": 2.623471566183322e-07,
      "loss": 0.8465,
      "step": 1040
    },
    {
      "epoch": 9.377245508982035,
      "grad_norm": 0.8334706425666809,
      "learning_rate": 1.9561352038673264e-07,
      "loss": 0.8576,
      "step": 1050
    },
    {
      "epoch": 9.467065868263473,
      "grad_norm": 0.6150113344192505,
      "learning_rate": 1.3857500887288544e-07,
      "loss": 0.8544,
      "step": 1060
    },
    {
      "epoch": 9.55688622754491,
      "grad_norm": 0.7956252098083496,
      "learning_rate": 9.12880249413628e-08,
      "loss": 0.8488,
      "step": 1070
    },
    {
      "epoch": 9.646706586826348,
      "grad_norm": 0.755624532699585,
      "learning_rate": 5.379932860185122e-08,
      "loss": 0.8512,
      "step": 1080
    },
    {
      "epoch": 9.736526946107784,
      "grad_norm": 0.6398297548294067,
      "learning_rate": 2.6145990770238827e-08,
      "loss": 0.8559,
      "step": 1090
    },
    {
      "epoch": 9.826347305389222,
      "grad_norm": 0.9212434887886047,
      "learning_rate": 8.355356610822984e-09,
      "loss": 0.8499,
      "step": 1100
    },
    {
      "epoch": 9.91616766467066,
      "grad_norm": 0.868816614151001,
      "learning_rate": 4.4501849589040357e-10,
      "loss": 0.8618,
      "step": 1110
    },
    {
      "epoch": 9.91616766467066,
      "step": 1110,
      "total_flos": 3.0002693815874355e+17,
      "train_loss": 1.1883513411960087,
      "train_runtime": 36141.346,
      "train_samples_per_second": 0.553,
      "train_steps_per_second": 0.031
    }
  ],
  "logging_steps": 10,
  "max_steps": 1110,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.0002693815874355e+17,
  "train_batch_size": 3,
  "trial_name": null,
  "trial_params": null
}