{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9024390243902438,
  "eval_steps": 3,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0975609756097561,
      "grad_norm": 0.13416488468647003,
      "learning_rate": 2e-05,
      "loss": 1.1198,
      "step": 1
    },
    {
      "epoch": 0.0975609756097561,
      "eval_loss": 1.0946272611618042,
      "eval_runtime": 105.1442,
      "eval_samples_per_second": 0.951,
      "eval_steps_per_second": 0.476,
      "step": 1
    },
    {
      "epoch": 0.1951219512195122,
      "grad_norm": 0.14614562690258026,
      "learning_rate": 4e-05,
      "loss": 1.1433,
      "step": 2
    },
    {
      "epoch": 0.2926829268292683,
      "grad_norm": 0.12903447449207306,
      "learning_rate": 6e-05,
      "loss": 1.0338,
      "step": 3
    },
    {
      "epoch": 0.2926829268292683,
      "eval_loss": 1.0899513959884644,
      "eval_runtime": 105.534,
      "eval_samples_per_second": 0.948,
      "eval_steps_per_second": 0.474,
      "step": 3
    },
    {
      "epoch": 0.3902439024390244,
      "grad_norm": 0.12645691633224487,
      "learning_rate": 8e-05,
      "loss": 1.0196,
      "step": 4
    },
    {
      "epoch": 0.4878048780487805,
      "grad_norm": 0.15484200417995453,
      "learning_rate": 0.0001,
      "loss": 1.0868,
      "step": 5
    },
    {
      "epoch": 0.5853658536585366,
      "grad_norm": 0.13987194001674652,
      "learning_rate": 0.00012,
      "loss": 1.0658,
      "step": 6
    },
    {
      "epoch": 0.5853658536585366,
      "eval_loss": 1.0339258909225464,
      "eval_runtime": 105.5831,
      "eval_samples_per_second": 0.947,
      "eval_steps_per_second": 0.474,
      "step": 6
    },
    {
      "epoch": 0.6829268292682927,
      "grad_norm": 0.10199978202581406,
      "learning_rate": 0.00014,
      "loss": 1.0156,
      "step": 7
    },
    {
      "epoch": 0.7804878048780488,
      "grad_norm": 0.12288854271173477,
      "learning_rate": 0.00016,
      "loss": 1.0503,
      "step": 8
    },
    {
      "epoch": 0.8780487804878049,
      "grad_norm": 0.12781274318695068,
      "learning_rate": 0.00018,
      "loss": 1.0521,
      "step": 9
    },
    {
      "epoch": 0.8780487804878049,
      "eval_loss": 0.9954264163970947,
      "eval_runtime": 105.5993,
      "eval_samples_per_second": 0.947,
      "eval_steps_per_second": 0.473,
      "step": 9
    },
    {
      "epoch": 0.975609756097561,
      "grad_norm": 0.11168772727251053,
      "learning_rate": 0.0002,
      "loss": 1.0098,
      "step": 10
    },
    {
      "epoch": 1.024390243902439,
      "grad_norm": 1.0994621515274048,
      "learning_rate": 0.00019945218953682734,
      "loss": 1.0201,
      "step": 11
    },
    {
      "epoch": 1.1219512195121952,
      "grad_norm": 0.13263079524040222,
      "learning_rate": 0.00019781476007338058,
      "loss": 1.0211,
      "step": 12
    },
    {
      "epoch": 1.1219512195121952,
      "eval_loss": 0.980489194393158,
      "eval_runtime": 105.5984,
      "eval_samples_per_second": 0.947,
      "eval_steps_per_second": 0.473,
      "step": 12
    },
    {
      "epoch": 1.2195121951219512,
      "grad_norm": 0.10054657608270645,
      "learning_rate": 0.00019510565162951537,
      "loss": 0.9632,
      "step": 13
    },
    {
      "epoch": 1.3170731707317074,
      "grad_norm": 0.09211089462041855,
      "learning_rate": 0.0001913545457642601,
      "loss": 0.9381,
      "step": 14
    },
    {
      "epoch": 1.4146341463414633,
      "grad_norm": 0.0957920104265213,
      "learning_rate": 0.00018660254037844388,
      "loss": 0.9945,
      "step": 15
    },
    {
      "epoch": 1.4146341463414633,
      "eval_loss": 0.9690712094306946,
      "eval_runtime": 105.5491,
      "eval_samples_per_second": 0.947,
      "eval_steps_per_second": 0.474,
      "step": 15
    },
    {
      "epoch": 1.5121951219512195,
      "grad_norm": 0.11032579094171524,
      "learning_rate": 0.00018090169943749476,
      "loss": 0.9766,
      "step": 16
    },
    {
      "epoch": 1.6097560975609757,
      "grad_norm": 0.08608294278383255,
      "learning_rate": 0.00017431448254773944,
      "loss": 0.9178,
      "step": 17
    },
    {
      "epoch": 1.7073170731707317,
      "grad_norm": 0.08259813487529755,
      "learning_rate": 0.00016691306063588583,
      "loss": 0.967,
      "step": 18
    },
    {
      "epoch": 1.7073170731707317,
      "eval_loss": 0.9576988220214844,
      "eval_runtime": 105.558,
      "eval_samples_per_second": 0.947,
      "eval_steps_per_second": 0.474,
      "step": 18
    },
    {
      "epoch": 1.8048780487804879,
      "grad_norm": 0.08044092357158661,
      "learning_rate": 0.00015877852522924732,
      "loss": 1.0114,
      "step": 19
    },
    {
      "epoch": 1.9024390243902438,
      "grad_norm": 0.08861377835273743,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.0564,
      "step": 20
    }
  ],
  "logging_steps": 1,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 10,
  "total_flos": 2.965391547354317e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}