{
  "best_global_step": 500,
  "best_metric": 2.0110304026567256,
  "best_model_checkpoint": "/content/cot-split2/checkpoint-500",
  "epoch": 0.39157334168689795,
  "eval_steps": 250,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0,
      "eval_bpd": 5.449506205976702,
      "eval_entropy": 3.8259646681880883,
      "eval_loss": 3.7819511890411377,
      "eval_mean_token_accuracy": 0.3298281561563004,
      "eval_nll": 3.777309862116675,
      "eval_num_tokens": 0.0,
      "eval_ppl": 43.69832894461198,
      "eval_runtime": 318.4675,
      "eval_samples_per_second": 570.259,
      "eval_steps_per_second": 8.911,
      "step": 0
    },
    {
      "entropy": 4.035161377492861,
      "epoch": 0.0007831466833737959,
      "grad_norm": 2.9375,
      "learning_rate": 0.0,
      "loss": 3.6375203132629395,
      "mean_token_accuracy": 0.3242907148120855,
      "num_tokens": 156431.0,
      "step": 1
    },
    {
      "entropy": 4.157279884560115,
      "epoch": 0.019578667084344897,
      "grad_norm": 1.8515625,
      "learning_rate": 9.375e-05,
      "loss": 3.7293694814046225,
      "mean_token_accuracy": 0.3223868495477843,
      "num_tokens": 3882398.0,
      "step": 25
    },
    {
      "entropy": 4.069613211207327,
      "epoch": 0.03915733416868979,
      "grad_norm": 6.65625,
      "learning_rate": 0.00019140625,
      "loss": 3.286188659667969,
      "mean_token_accuracy": 0.3470086962011918,
      "num_tokens": 7771579.0,
      "step": 50
    },
    {
      "entropy": 3.712915897718112,
      "epoch": 0.05873600125303469,
      "grad_norm": 1.078125,
      "learning_rate": 0.0002890625,
      "loss": 2.9255511474609377,
      "mean_token_accuracy": 0.37689020381328076,
      "num_tokens": 11677887.0,
      "step": 75
    },
    {
      "entropy": 3.4658741959475385,
      "epoch": 0.07831466833737959,
      "grad_norm": 1.671875,
      "learning_rate": 0.00038671875,
      "loss": 2.5379095458984375,
      "mean_token_accuracy": 0.41490777290923164,
      "num_tokens": 15584019.0,
      "step": 100
    },
    {
      "entropy": 3.295554994185651,
      "epoch": 0.09789333542172449,
      "grad_norm": 0.59765625,
      "learning_rate": 0.000484375,
      "loss": 2.405628204345703,
      "mean_token_accuracy": 0.4313112262648087,
      "num_tokens": 19484337.0,
      "step": 125
    },
    {
      "entropy": 3.1617948360015533,
      "epoch": 0.11747200250606939,
      "grad_norm": 0.80078125,
      "learning_rate": 0.0004999075642455791,
      "loss": 2.3207965087890625,
      "mean_token_accuracy": 0.4436994497174724,
      "num_tokens": 23377294.0,
      "step": 150
    },
    {
      "entropy": 3.0947463825081942,
      "epoch": 0.1370506695904143,
      "grad_norm": 0.6640625,
      "learning_rate": 0.0004995565798713207,
      "loss": 2.2583859252929686,
      "mean_token_accuracy": 0.45017372579314874,
      "num_tokens": 27287783.0,
      "step": 175
    },
    {
      "entropy": 3.045135160417696,
      "epoch": 0.15662933667475917,
      "grad_norm": 0.65625,
      "learning_rate": 0.0004989440608306114,
      "loss": 2.230653076171875,
      "mean_token_accuracy": 0.4551180636124264,
      "num_tokens": 31205249.0,
      "step": 200
    },
    {
      "entropy": 3.0288844077617805,
      "epoch": 0.1762080037591041,
      "grad_norm": 0.625,
      "learning_rate": 0.0004980706490418613,
      "loss": 2.214619140625,
      "mean_token_accuracy": 0.4571791829436394,
      "num_tokens": 35081817.0,
      "step": 225
    },
    {
      "entropy": 2.9857863094922594,
      "epoch": 0.19578667084344897,
      "grad_norm": 0.6015625,
      "learning_rate": 0.0004969372598384225,
      "loss": 2.183477783203125,
      "mean_token_accuracy": 0.4623076980301137,
      "num_tokens": 38961791.0,
      "step": 250
    },
    {
      "epoch": 0.19578667084344897,
      "eval_bpd": 3.0542670212451286,
      "eval_entropy": 2.831413693293467,
      "eval_loss": 2.114036798477173,
      "eval_mean_token_accuracy": 0.471896302096762,
      "eval_nll": 2.1170565744532834,
      "eval_num_tokens": 38961791.0,
      "eval_ppl": 8.306651458853036,
      "eval_runtime": 317.7783,
      "eval_samples_per_second": 571.496,
      "eval_steps_per_second": 8.931,
      "step": 250
    },
    {
      "entropy": 2.960300336004756,
      "epoch": 0.2153653379277939,
      "grad_norm": 0.59765625,
      "learning_rate": 0.0004955450810093227,
      "loss": 2.1587040710449217,
      "mean_token_accuracy": 0.46429396638467557,
      "num_tokens": 42868157.0,
      "step": 275
    },
    {
      "entropy": 2.9496937473585083,
      "epoch": 0.23494400501213877,
      "grad_norm": 0.6953125,
      "learning_rate": 0.0004938955715544644,
      "loss": 2.1702507019042967,
      "mean_token_accuracy": 0.4644734879386833,
      "num_tokens": 46774567.0,
      "step": 300
    },
    {
      "entropy": 2.925756508647974,
      "epoch": 0.25452267209648366,
      "grad_norm": 0.61328125,
      "learning_rate": 0.0004919904601555969,
      "loss": 2.1320841979980467,
      "mean_token_accuracy": 0.4678374269200783,
      "num_tokens": 50651509.0,
      "step": 325
    },
    {
      "entropy": 2.916845726644516,
      "epoch": 0.2741013391808286,
      "grad_norm": 0.60546875,
      "learning_rate": 0.0004898317433646626,
      "loss": 2.149468994140625,
      "mean_token_accuracy": 0.46948018886239345,
      "num_tokens": 54545113.0,
      "step": 350
    },
    {
      "entropy": 2.8672960236041836,
      "epoch": 0.2936800062651735,
      "grad_norm": 0.59765625,
      "learning_rate": 0.00048742168351141525,
      "loss": 2.095373077392578,
      "mean_token_accuracy": 0.47492304637997923,
      "num_tokens": 58450259.0,
      "step": 375
    },
    {
      "entropy": 2.866450867223391,
      "epoch": 0.31325867334951835,
      "grad_norm": 0.609375,
      "learning_rate": 0.00048476280633250404,
      "loss": 2.093819122314453,
      "mean_token_accuracy": 0.4762016071737522,
      "num_tokens": 62334830.0,
      "step": 400
    },
    {
      "entropy": 2.8444264384733913,
      "epoch": 0.33283734043386326,
      "grad_norm": 0.671875,
      "learning_rate": 0.00048185789832450813,
      "loss": 2.07142822265625,
      "mean_token_accuracy": 0.47914267030296726,
      "num_tokens": 66254709.0,
      "step": 425
    },
    {
      "entropy": 2.8643093723566215,
      "epoch": 0.3524160075182082,
      "grad_norm": 0.578125,
      "learning_rate": 0.00047871000382369524,
      "loss": 2.0907044982910157,
      "mean_token_accuracy": 0.475262880935016,
      "num_tokens": 70149264.0,
      "step": 450
    },
    {
      "entropy": 2.8425666356253885,
      "epoch": 0.37199467460255303,
      "grad_norm": 1.453125,
      "learning_rate": 0.0004753224218155648,
      "loss": 2.0714340209960938,
      "mean_token_accuracy": 0.47639239522332366,
      "num_tokens": 74036195.0,
      "step": 475
    },
    {
      "entropy": 2.8495917513303857,
      "epoch": 0.39157334168689795,
      "grad_norm": 0.58203125,
      "learning_rate": 0.0004716987024775197,
      "loss": 2.0957228088378907,
      "mean_token_accuracy": 0.4775797151536097,
      "num_tokens": 77964744.0,
      "step": 500
    },
    {
      "epoch": 0.39157334168689795,
      "eval_bpd": 2.9013035889897933,
      "eval_entropy": 2.6662824092598547,
      "eval_loss": 2.007856607437134,
      "eval_mean_token_accuracy": 0.48950937525505117,
      "eval_nll": 2.0110304026567256,
      "eval_num_tokens": 77964744.0,
      "eval_ppl": 7.471011532715633,
      "eval_runtime": 317.7246,
      "eval_samples_per_second": 571.593,
      "eval_steps_per_second": 8.932,
      "step": 500
    }
  ],
  "logging_steps": 25,
  "max_steps": 2554,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.5435666628026368e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}