{
  "best_global_step": 400,
  "best_metric": 0.38200808,
  "best_model_checkpoint": "/home/ubuntu/output/v1-20250507-032200/checkpoint-400",
  "epoch": 4.123711340206185,
  "eval_steps": 50,
  "global_step": 400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.010309278350515464,
      "grad_norm": 2.8573532540412123,
      "learning_rate": 4e-08,
      "loss": 0.4424864649772644,
      "memory(GiB)": 37.04,
      "step": 1,
      "token_acc": 0.8862085383272497,
      "train_speed(iter/s)": 0.083623
    },
    {
      "epoch": 0.05154639175257732,
      "grad_norm": 2.8403080546549457,
      "learning_rate": 2e-07,
      "loss": 0.41638126969337463,
      "memory(GiB)": 37.04,
      "step": 5,
      "token_acc": 0.848388504457328,
      "train_speed(iter/s)": 0.199696
    },
    {
      "epoch": 0.10309278350515463,
      "grad_norm": 2.6629741598617374,
      "learning_rate": 4e-07,
      "loss": 0.4429419994354248,
      "memory(GiB)": 37.04,
      "step": 10,
      "token_acc": 0.8634871926011166,
      "train_speed(iter/s)": 0.246696
    },
    {
      "epoch": 0.15463917525773196,
      "grad_norm": 3.095692053663771,
      "learning_rate": 6e-07,
      "loss": 0.45945310592651367,
      "memory(GiB)": 37.04,
      "step": 15,
      "token_acc": 0.8555051349414855,
      "train_speed(iter/s)": 0.275511
    },
    {
      "epoch": 0.20618556701030927,
      "grad_norm": 2.637907165803134,
      "learning_rate": 8e-07,
      "loss": 0.4365420341491699,
      "memory(GiB)": 37.04,
      "step": 20,
      "token_acc": 0.8420170882894957,
      "train_speed(iter/s)": 0.28272
    },
    {
      "epoch": 0.25773195876288657,
      "grad_norm": 2.476440678075636,
      "learning_rate": 1e-06,
      "loss": 0.4216726779937744,
      "memory(GiB)": 37.04,
      "step": 25,
      "token_acc": 0.8617449014193673,
      "train_speed(iter/s)": 0.284174
    },
    {
      "epoch": 0.30927835051546393,
      "grad_norm": 2.2717829784567276,
      "learning_rate": 9.99708511183087e-07,
      "loss": 0.4195085048675537,
      "memory(GiB)": 37.04,
      "step": 30,
      "token_acc": 0.8650009974067425,
      "train_speed(iter/s)": 0.28951
    },
    {
      "epoch": 0.36082474226804123,
      "grad_norm": 2.077114274728423,
      "learning_rate": 9.988343845952696e-07,
      "loss": 0.3923480987548828,
      "memory(GiB)": 37.04,
      "step": 35,
      "token_acc": 0.85941012498876,
      "train_speed(iter/s)": 0.288231
    },
    {
      "epoch": 0.41237113402061853,
      "grad_norm": 1.734736391920931,
      "learning_rate": 9.973786394290473e-07,
      "loss": 0.3943674564361572,
      "memory(GiB)": 37.04,
      "step": 40,
      "token_acc": 0.8583430377640991,
      "train_speed(iter/s)": 0.28752
    },
    {
      "epoch": 0.4639175257731959,
      "grad_norm": 1.3825734749873666,
      "learning_rate": 9.953429730181652e-07,
      "loss": 0.39246172904968263,
      "memory(GiB)": 37.04,
      "step": 45,
      "token_acc": 0.86418609718893,
      "train_speed(iter/s)": 0.291708
    },
    {
      "epoch": 0.5154639175257731,
      "grad_norm": 0.8751053891908908,
      "learning_rate": 9.927297588585983e-07,
      "loss": 0.39423561096191406,
      "memory(GiB)": 37.04,
      "step": 50,
      "token_acc": 0.8659434116556725,
      "train_speed(iter/s)": 0.294091
    },
    {
      "epoch": 0.5154639175257731,
      "eval_loss": 0.41471877694129944,
      "eval_runtime": 0.9078,
      "eval_samples_per_second": 16.524,
      "eval_steps_per_second": 2.203,
      "eval_token_acc": 0.8572620288991509,
      "step": 50
    },
    {
      "epoch": 0.5670103092783505,
      "grad_norm": 0.8663484008448668,
      "learning_rate": 9.895420438411615e-07,
      "loss": 0.3726518154144287,
      "memory(GiB)": 37.04,
      "step": 55,
      "token_acc": 0.8743935309973045,
      "train_speed(iter/s)": 0.230989
    },
    {
      "epoch": 0.6185567010309279,
      "grad_norm": 0.667816147564415,
      "learning_rate": 9.857835446989707e-07,
      "loss": 0.36932010650634767,
      "memory(GiB)": 37.04,
      "step": 60,
      "token_acc": 0.8602125147579693,
      "train_speed(iter/s)": 0.23604
    },
    {
      "epoch": 0.6701030927835051,
      "grad_norm": 0.8293459879499964,
      "learning_rate": 9.814586436738997e-07,
      "loss": 0.3804448366165161,
      "memory(GiB)": 37.04,
      "step": 65,
      "token_acc": 0.8690644468875975,
      "train_speed(iter/s)": 0.241146
    },
    {
      "epoch": 0.7216494845360825,
      "grad_norm": 0.7787810832421912,
      "learning_rate": 9.765723834070804e-07,
      "loss": 0.36735010147094727,
      "memory(GiB)": 37.04,
      "step": 70,
      "token_acc": 0.8663579855385005,
      "train_speed(iter/s)": 0.243142
    },
    {
      "epoch": 0.7731958762886598,
      "grad_norm": 0.7336453374485993,
      "learning_rate": 9.711304610594102e-07,
      "loss": 0.3681507110595703,
      "memory(GiB)": 37.04,
      "step": 75,
      "token_acc": 0.8693066207544217,
      "train_speed(iter/s)": 0.245585
    },
    {
      "epoch": 0.8247422680412371,
      "grad_norm": 0.8189938563466513,
      "learning_rate": 9.651392216689165e-07,
      "loss": 0.3353311061859131,
      "memory(GiB)": 37.04,
      "step": 80,
      "token_acc": 0.8910798995442284,
      "train_speed(iter/s)": 0.249085
    },
    {
      "epoch": 0.8762886597938144,
      "grad_norm": 0.6779929134307079,
      "learning_rate": 9.586056507527264e-07,
      "loss": 0.34499826431274416,
      "memory(GiB)": 37.04,
      "step": 85,
      "token_acc": 0.8659364094610315,
      "train_speed(iter/s)": 0.252498
    },
    {
      "epoch": 0.9278350515463918,
      "grad_norm": 0.7719151294405667,
      "learning_rate": 9.515373661622663e-07,
      "loss": 0.3747685432434082,
      "memory(GiB)": 37.04,
      "step": 90,
      "token_acc": 0.8549843014128728,
      "train_speed(iter/s)": 0.253844
    },
    {
      "epoch": 0.979381443298969,
      "grad_norm": 0.6973169912038593,
      "learning_rate": 9.439426092011875e-07,
      "loss": 0.36587209701538087,
      "memory(GiB)": 37.04,
      "step": 95,
      "token_acc": 0.8687297229847767,
      "train_speed(iter/s)": 0.257063
    },
    {
      "epoch": 1.0309278350515463,
      "grad_norm": 0.692769567261216,
      "learning_rate": 9.358302350163756e-07,
      "loss": 0.3589401006698608,
      "memory(GiB)": 37.04,
      "step": 100,
      "token_acc": 0.8745442333525235,
      "train_speed(iter/s)": 0.259875
    },
    {
      "epoch": 1.0309278350515463,
      "eval_loss": 0.3933715224266052,
      "eval_runtime": 0.9247,
      "eval_samples_per_second": 16.222,
      "eval_steps_per_second": 2.163,
      "eval_token_acc": 0.862684343810517,
      "step": 100
    },
    {
      "epoch": 1.0824742268041236,
      "grad_norm": 0.6544219439665868,
      "learning_rate": 9.272097022732443e-07,
      "loss": 0.33444135189056395,
      "memory(GiB)": 37.04,
      "step": 105,
      "token_acc": 0.8807363927427961,
      "train_speed(iter/s)": 0.232768
    },
    {
      "epoch": 1.134020618556701,
      "grad_norm": 0.6367223115088674,
      "learning_rate": 9.180910621273555e-07,
      "loss": 0.34373011589050295,
      "memory(GiB)": 37.04,
      "step": 110,
      "token_acc": 0.8613336475023563,
      "train_speed(iter/s)": 0.235371
    },
    {
      "epoch": 1.1855670103092784,
      "grad_norm": 0.7492955478131063,
      "learning_rate": 9.084849465052209e-07,
      "loss": 0.3586832046508789,
      "memory(GiB)": 37.04,
      "step": 115,
      "token_acc": 0.8713512419689381,
      "train_speed(iter/s)": 0.237153
    },
    {
      "epoch": 1.2371134020618557,
      "grad_norm": 0.7071477506615176,
      "learning_rate": 8.984025557079522e-07,
      "loss": 0.3494635343551636,
      "memory(GiB)": 37.04,
      "step": 120,
      "token_acc": 0.8666221703098059,
      "train_speed(iter/s)": 0.239655
    },
    {
      "epoch": 1.2886597938144329,
      "grad_norm": 0.6092307857366024,
      "learning_rate": 8.878556453522099e-07,
      "loss": 0.3500272512435913,
      "memory(GiB)": 37.04,
      "step": 125,
      "token_acc": 0.8708257830701527,
      "train_speed(iter/s)": 0.241223
    },
    {
      "epoch": 1.3402061855670104,
      "grad_norm": 0.6624451351751124,
      "learning_rate": 8.768565126636805e-07,
      "loss": 0.3489999294281006,
      "memory(GiB)": 37.04,
      "step": 130,
      "token_acc": 0.8813361936867913,
      "train_speed(iter/s)": 0.243732
    },
    {
      "epoch": 1.3917525773195876,
      "grad_norm": 0.7828961232174202,
      "learning_rate": 8.654179821390621e-07,
      "loss": 0.3605631113052368,
      "memory(GiB)": 37.04,
      "step": 135,
      "token_acc": 0.8725261932479628,
      "train_speed(iter/s)": 0.245988
    },
    {
      "epoch": 1.443298969072165,
      "grad_norm": 0.7711241338182571,
      "learning_rate": 8.535533905932737e-07,
      "loss": 0.34404377937316893,
      "memory(GiB)": 37.04,
      "step": 140,
      "token_acc": 0.8685788787483703,
      "train_speed(iter/s)": 0.247325
    },
    {
      "epoch": 1.4948453608247423,
      "grad_norm": 0.6706712637235819,
      "learning_rate": 8.41276571609327e-07,
      "loss": 0.34629359245300295,
      "memory(GiB)": 37.04,
      "step": 145,
      "token_acc": 0.8681879534334402,
      "train_speed(iter/s)": 0.248924
    },
    {
      "epoch": 1.5463917525773194,
      "grad_norm": 0.7453364306944348,
      "learning_rate": 8.286018394089863e-07,
      "loss": 0.35725903511047363,
      "memory(GiB)": 37.04,
      "step": 150,
      "token_acc": 0.8822486705495062,
      "train_speed(iter/s)": 0.250797
    },
    {
      "epoch": 1.5463917525773194,
      "eval_loss": 0.38744473457336426,
      "eval_runtime": 0.9547,
      "eval_samples_per_second": 15.712,
      "eval_steps_per_second": 2.095,
      "eval_token_acc": 0.8649188142410249,
      "step": 150
    },
    {
      "epoch": 1.597938144329897,
      "grad_norm": 0.6717542898329211,
      "learning_rate": 8.155439721630264e-07,
      "loss": 0.3512057065963745,
      "memory(GiB)": 37.04,
      "step": 155,
      "token_acc": 0.8510025260498895,
      "train_speed(iter/s)": 0.133083
    },
    {
      "epoch": 1.6494845360824741,
      "grad_norm": 0.6040559170916198,
      "learning_rate": 8.021181947605472e-07,
      "loss": 0.34543561935424805,
      "memory(GiB)": 37.23,
      "step": 160,
      "token_acc": 0.8643211778804999,
      "train_speed(iter/s)": 0.135409
    },
    {
      "epoch": 1.7010309278350515,
      "grad_norm": 0.6664553249603009,
      "learning_rate": 7.883401610574336e-07,
      "loss": 0.34488234519958494,
      "memory(GiB)": 37.23,
      "step": 165,
      "token_acc": 0.8770163214660686,
      "train_speed(iter/s)": 0.137732
    },
    {
      "epoch": 1.7525773195876289,
      "grad_norm": 0.8055880341832813,
      "learning_rate": 7.742259356246594e-07,
      "loss": 0.34050488471984863,
      "memory(GiB)": 37.23,
      "step": 170,
      "token_acc": 0.8739508700102354,
      "train_speed(iter/s)": 0.139961
    },
    {
      "epoch": 1.8041237113402062,
      "grad_norm": 0.7026822100715953,
      "learning_rate": 7.597919750177168e-07,
      "loss": 0.35553760528564454,
      "memory(GiB)": 37.23,
      "step": 175,
      "token_acc": 0.8798612963456922,
      "train_speed(iter/s)": 0.142407
    },
    {
      "epoch": 1.8556701030927836,
      "grad_norm": 0.6296090546523304,
      "learning_rate": 7.450551085890087e-07,
      "loss": 0.3472688913345337,
      "memory(GiB)": 37.23,
      "step": 180,
      "token_acc": 0.8816289399047467,
      "train_speed(iter/s)": 0.144588
    },
    {
      "epoch": 1.9072164948453607,
      "grad_norm": 0.7100452196024024,
      "learning_rate": 7.30032518865576e-07,
      "loss": 0.342409086227417,
      "memory(GiB)": 37.23,
      "step": 185,
      "token_acc": 0.8932488313556275,
      "train_speed(iter/s)": 0.146787
    },
    {
      "epoch": 1.9587628865979383,
      "grad_norm": 0.6751616783204145,
      "learning_rate": 7.14741721515041e-07,
      "loss": 0.3448726892471313,
      "memory(GiB)": 37.23,
      "step": 190,
      "token_acc": 0.87621425082759,
      "train_speed(iter/s)": 0.148964
    },
    {
      "epoch": 2.0103092783505154,
      "grad_norm": 0.683041065571367,
      "learning_rate": 6.992005449231207e-07,
      "loss": 0.3384934663772583,
      "memory(GiB)": 37.23,
      "step": 195,
      "token_acc": 0.8557946771857992,
      "train_speed(iter/s)": 0.15077
    },
    {
      "epoch": 2.0618556701030926,
      "grad_norm": 0.6516412497661351,
      "learning_rate": 6.834271094065282e-07,
      "loss": 0.34560070037841795,
      "memory(GiB)": 37.23,
      "step": 200,
      "token_acc": 0.8638397328881469,
      "train_speed(iter/s)": 0.152746
    },
    {
      "epoch": 2.0618556701030926,
      "eval_loss": 0.38463127613067627,
      "eval_runtime": 0.9665,
      "eval_samples_per_second": 15.52,
      "eval_steps_per_second": 2.069,
      "eval_token_acc": 0.865127364814539,
      "step": 200
    },
    {
      "epoch": 2.11340206185567,
      "grad_norm": 0.7045810824940504,
      "learning_rate": 6.67439806085493e-07,
      "loss": 0.33291826248168943,
      "memory(GiB)": 37.77,
      "step": 205,
      "token_acc": 0.8808652368055048,
      "train_speed(iter/s)": 0.103353
    },
    {
      "epoch": 2.1649484536082473,
      "grad_norm": 0.744342227263682,
      "learning_rate": 6.512572754405379e-07,
      "loss": 0.3307349681854248,
      "memory(GiB)": 37.77,
      "step": 210,
      "token_acc": 0.8755916159567275,
      "train_speed(iter/s)": 0.104962
    },
    {
      "epoch": 2.216494845360825,
      "grad_norm": 0.6323077432320735,
      "learning_rate": 6.348983855785121e-07,
      "loss": 0.3156585216522217,
      "memory(GiB)": 37.77,
      "step": 215,
      "token_acc": 0.8785982478097623,
      "train_speed(iter/s)": 0.106558
    },
    {
      "epoch": 2.268041237113402,
      "grad_norm": 0.6915848622598538,
      "learning_rate": 6.183822102332234e-07,
      "loss": 0.31838016510009765,
      "memory(GiB)": 37.77,
      "step": 220,
      "token_acc": 0.8778223495702006,
      "train_speed(iter/s)": 0.108175
    },
    {
      "epoch": 2.319587628865979,
      "grad_norm": 0.5308281735282047,
      "learning_rate": 6.01728006526317e-07,
      "loss": 0.3378974437713623,
      "memory(GiB)": 37.77,
      "step": 225,
      "token_acc": 0.871986195961386,
      "train_speed(iter/s)": 0.109623
    },
    {
      "epoch": 2.3711340206185567,
      "grad_norm": 0.6489926882668993,
      "learning_rate": 5.849551925143333e-07,
      "loss": 0.32499127388000487,
      "memory(GiB)": 37.77,
      "step": 230,
      "token_acc": 0.8704572098475967,
      "train_speed(iter/s)": 0.111249
    },
    {
      "epoch": 2.422680412371134,
      "grad_norm": 0.5977786295969187,
      "learning_rate": 5.680833245481234e-07,
      "loss": 0.3224591493606567,
      "memory(GiB)": 37.77,
      "step": 235,
      "token_acc": 0.8655716346377553,
      "train_speed(iter/s)": 0.112868
    },
    {
      "epoch": 2.4742268041237114,
      "grad_norm": 0.6356644743549772,
      "learning_rate": 5.51132074471017e-07,
      "loss": 0.3337638139724731,
      "memory(GiB)": 37.77,
      "step": 240,
      "token_acc": 0.8858976428802067,
      "train_speed(iter/s)": 0.114243
    },
    {
      "epoch": 2.5257731958762886,
      "grad_norm": 0.6955635518347525,
      "learning_rate": 5.341212066823355e-07,
      "loss": 0.34955778121948244,
      "memory(GiB)": 37.77,
      "step": 245,
      "token_acc": 0.875490146152671,
      "train_speed(iter/s)": 0.115731
    },
    {
      "epoch": 2.5773195876288657,
      "grad_norm": 0.6678415617901132,
      "learning_rate": 5.170705550929839e-07,
      "loss": 0.33729376792907717,
      "memory(GiB)": 37.77,
      "step": 250,
      "token_acc": 0.875747248449634,
      "train_speed(iter/s)": 0.1172
    },
    {
      "epoch": 2.5773195876288657,
      "eval_loss": 0.38304245471954346,
      "eval_runtime": 0.9688,
      "eval_samples_per_second": 15.482,
      "eval_steps_per_second": 2.064,
      "eval_token_acc": 0.8655742589006405,
      "step": 250
    },
    {
      "epoch": 2.6288659793814433,
      "grad_norm": 0.7100665104948269,
      "learning_rate": 5e-07,
      "loss": 0.3353480339050293,
      "memory(GiB)": 37.77,
      "step": 255,
      "token_acc": 0.877640879069947,
      "train_speed(iter/s)": 0.09114
    },
    {
      "epoch": 2.680412371134021,
      "grad_norm": 0.718021243798035,
      "learning_rate": 4.829294449070161e-07,
      "loss": 0.35503473281860354,
      "memory(GiB)": 37.77,
      "step": 260,
      "token_acc": 0.8858998533417138,
      "train_speed(iter/s)": 0.092445
    },
    {
      "epoch": 2.731958762886598,
      "grad_norm": 0.7452764083029736,
      "learning_rate": 4.6587879331766457e-07,
      "loss": 0.3445273876190186,
      "memory(GiB)": 37.77,
      "step": 265,
      "token_acc": 0.88517933679224,
      "train_speed(iter/s)": 0.093659
    },
    {
      "epoch": 2.783505154639175,
      "grad_norm": 0.7104117036082537,
      "learning_rate": 4.4886792552898283e-07,
      "loss": 0.33231029510498045,
      "memory(GiB)": 37.77,
      "step": 270,
      "token_acc": 0.8740161938735066,
      "train_speed(iter/s)": 0.094846
    },
    {
      "epoch": 2.8350515463917527,
      "grad_norm": 0.6272461565144517,
      "learning_rate": 4.3191667545187675e-07,
      "loss": 0.3508894920349121,
      "memory(GiB)": 37.77,
      "step": 275,
      "token_acc": 0.8654118758859662,
      "train_speed(iter/s)": 0.0961
    },
    {
      "epoch": 2.88659793814433,
      "grad_norm": 0.6361538625329014,
      "learning_rate": 4.150448074856667e-07,
      "loss": 0.3445762634277344,
      "memory(GiB)": 37.77,
      "step": 280,
      "token_acc": 0.8735829231875242,
      "train_speed(iter/s)": 0.097284
    },
    {
      "epoch": 2.9381443298969074,
      "grad_norm": 0.6794447157660283,
      "learning_rate": 3.9827199347368317e-07,
      "loss": 0.3385981321334839,
      "memory(GiB)": 37.77,
      "step": 285,
      "token_acc": 0.8693530615457264,
      "train_speed(iter/s)": 0.098521
    },
    {
      "epoch": 2.9896907216494846,
      "grad_norm": 0.5858240811392458,
      "learning_rate": 3.816177897667766e-07,
      "loss": 0.3372650623321533,
      "memory(GiB)": 37.77,
      "step": 290,
      "token_acc": 0.8661933904528764,
      "train_speed(iter/s)": 0.099708
    },
    {
      "epoch": 3.0412371134020617,
      "grad_norm": 0.6492571076999712,
      "learning_rate": 3.651016144214878e-07,
      "loss": 0.3159039974212646,
      "memory(GiB)": 37.77,
      "step": 295,
      "token_acc": 0.8897466689540802,
      "train_speed(iter/s)": 0.100872
    },
    {
      "epoch": 3.0927835051546393,
      "grad_norm": 0.5938135529575547,
      "learning_rate": 3.4874272455946216e-07,
      "loss": 0.3142085075378418,
      "memory(GiB)": 37.77,
      "step": 300,
      "token_acc": 0.898450120061122,
      "train_speed(iter/s)": 0.102031
    },
    {
      "epoch": 3.0927835051546393,
      "eval_loss": 0.3824613094329834,
      "eval_runtime": 0.9688,
      "eval_samples_per_second": 15.484,
      "eval_steps_per_second": 2.064,
      "eval_token_acc": 0.8658126024132281,
      "step": 300
    },
    {
      "epoch": 3.1443298969072164,
      "grad_norm": 0.6425324377834628,
      "learning_rate": 3.325601939145069e-07,
      "loss": 0.331402063369751,
      "memory(GiB)": 37.77,
      "step": 305,
      "token_acc": 0.8719620011787955,
      "train_speed(iter/s)": 0.084419
    },
    {
      "epoch": 3.195876288659794,
      "grad_norm": 0.6991897942082559,
      "learning_rate": 3.165728905934718e-07,
      "loss": 0.3265227794647217,
      "memory(GiB)": 37.77,
      "step": 310,
      "token_acc": 0.8806500761808025,
      "train_speed(iter/s)": 0.085442
    },
    {
      "epoch": 3.247422680412371,
      "grad_norm": 0.6735292235062921,
      "learning_rate": 3.007994550768793e-07,
      "loss": 0.3452199935913086,
      "memory(GiB)": 37.77,
      "step": 315,
      "token_acc": 0.8813902802737735,
      "train_speed(iter/s)": 0.086446
    },
    {
      "epoch": 3.2989690721649483,
      "grad_norm": 0.6808191589842796,
      "learning_rate": 2.852582784849591e-07,
      "loss": 0.32866055965423585,
      "memory(GiB)": 37.77,
      "step": 320,
      "token_acc": 0.882211663969091,
      "train_speed(iter/s)": 0.087454
    },
    {
      "epoch": 3.350515463917526,
      "grad_norm": 0.7081778457873291,
      "learning_rate": 2.699674811344239e-07,
      "loss": 0.32704184055328367,
      "memory(GiB)": 37.77,
      "step": 325,
      "token_acc": 0.8746191346739792,
      "train_speed(iter/s)": 0.08836
    },
    {
      "epoch": 3.402061855670103,
      "grad_norm": 0.6412966995516326,
      "learning_rate": 2.549448914109915e-07,
      "loss": 0.3270967960357666,
      "memory(GiB)": 37.77,
      "step": 330,
      "token_acc": 0.8627403990174346,
      "train_speed(iter/s)": 0.08934
    },
    {
      "epoch": 3.4536082474226806,
      "grad_norm": 0.5539028709876479,
      "learning_rate": 2.4020802498228334e-07,
      "loss": 0.3313950538635254,
      "memory(GiB)": 37.77,
      "step": 335,
      "token_acc": 0.8844105805446123,
      "train_speed(iter/s)": 0.090289
    },
    {
      "epoch": 3.5051546391752577,
      "grad_norm": 0.691681066452575,
      "learning_rate": 2.257740643753405e-07,
      "loss": 0.32365880012512205,
      "memory(GiB)": 37.77,
      "step": 340,
      "token_acc": 0.8830835117773019,
      "train_speed(iter/s)": 0.091196
    },
    {
      "epoch": 3.556701030927835,
      "grad_norm": 0.6193458212688764,
      "learning_rate": 2.1165983894256646e-07,
      "loss": 0.3258372783660889,
      "memory(GiB)": 37.77,
      "step": 345,
      "token_acc": 0.9008801723823089,
      "train_speed(iter/s)": 0.092053
    },
    {
      "epoch": 3.6082474226804124,
      "grad_norm": 0.6415244001554736,
      "learning_rate": 1.9788180523945275e-07,
      "loss": 0.33638935089111327,
      "memory(GiB)": 37.77,
      "step": 350,
      "token_acc": 0.8890733326823552,
      "train_speed(iter/s)": 0.093029
    },
    {
      "epoch": 3.6082474226804124,
      "eval_loss": 0.3821544945240021,
      "eval_runtime": 0.9819,
      "eval_samples_per_second": 15.277,
      "eval_steps_per_second": 2.037,
      "eval_token_acc": 0.8655742589006405,
      "step": 350
    },
    {
      "epoch": 3.6597938144329896,
      "grad_norm": 0.648833766542735,
      "learning_rate": 1.8445602783697373e-07,
      "loss": 0.31353211402893066,
      "memory(GiB)": 37.77,
      "step": 355,
      "token_acc": 0.8827717297390704,
      "train_speed(iter/s)": 0.080129
    },
    {
      "epoch": 3.711340206185567,
      "grad_norm": 0.6717087103762708,
      "learning_rate": 1.713981605910137e-07,
      "loss": 0.31485791206359864,
      "memory(GiB)": 37.77,
      "step": 360,
      "token_acc": 0.8871092587079488,
      "train_speed(iter/s)": 0.081028
    },
    {
      "epoch": 3.7628865979381443,
      "grad_norm": 0.6281626069775872,
      "learning_rate": 1.5872342839067304e-07,
      "loss": 0.34305672645568847,
      "memory(GiB)": 37.77,
      "step": 365,
      "token_acc": 0.874572343008498,
      "train_speed(iter/s)": 0.081796
    },
    {
      "epoch": 3.8144329896907214,
      "grad_norm": 0.6620673803452289,
      "learning_rate": 1.4644660940672627e-07,
      "loss": 0.33494589328765867,
      "memory(GiB)": 37.77,
      "step": 370,
      "token_acc": 0.889901752253621,
      "train_speed(iter/s)": 0.082621
    },
    {
      "epoch": 3.865979381443299,
      "grad_norm": 0.693656406725637,
      "learning_rate": 1.3458201786093794e-07,
      "loss": 0.3408153533935547,
      "memory(GiB)": 37.77,
      "step": 375,
      "token_acc": 0.876736032056828,
      "train_speed(iter/s)": 0.083425
    },
    {
      "epoch": 3.917525773195876,
      "grad_norm": 0.6371484991381379,
      "learning_rate": 1.2314348733631957e-07,
      "loss": 0.3388832092285156,
      "memory(GiB)": 37.77,
      "step": 380,
      "token_acc": 0.8669541981154189,
      "train_speed(iter/s)": 0.084189
    },
    {
      "epoch": 3.9690721649484537,
      "grad_norm": 0.6261422020208081,
      "learning_rate": 1.1214435464779003e-07,
      "loss": 0.32546391487121584,
      "memory(GiB)": 37.77,
      "step": 385,
      "token_acc": 0.877057210031348,
      "train_speed(iter/s)": 0.085015
    },
    {
      "epoch": 4.020618556701031,
      "grad_norm": 0.6597131788548194,
      "learning_rate": 1.0159744429204775e-07,
      "loss": 0.3102993965148926,
      "memory(GiB)": 37.77,
      "step": 390,
      "token_acc": 0.8759420669768662,
      "train_speed(iter/s)": 0.085822
    },
    {
      "epoch": 4.072164948453608,
      "grad_norm": 0.6279013012460978,
      "learning_rate": 9.1515053494779e-08,
      "loss": 0.33462517261505126,
      "memory(GiB)": 37.77,
      "step": 395,
      "token_acc": 0.8786680275580505,
      "train_speed(iter/s)": 0.086608
    },
    {
      "epoch": 4.123711340206185,
      "grad_norm": 0.7265548614470879,
      "learning_rate": 8.190893787264469e-08,
      "loss": 0.3322861433029175,
      "memory(GiB)": 37.77,
      "step": 400,
      "token_acc": 0.8789391575663027,
      "train_speed(iter/s)": 0.087381
    },
    {
      "epoch": 4.123711340206185,
      "eval_loss": 0.38200807571411133,
      "eval_runtime": 0.9767,
      "eval_samples_per_second": 15.358,
      "eval_steps_per_second": 2.048,
      "eval_token_acc": 0.8655444659615671,
      "step": 400
    }
  ],
  "logging_steps": 5,
  "max_steps": 485,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 29414501638144.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}