{
  "best_metric": 0.9346915483474731,
  "best_model_checkpoint": "outputs/checkpoint-185",
  "epoch": 3.9946018893387314,
  "eval_steps": 500,
  "global_step": 185,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.4318488529014845,
      "grad_norm": 1.8083888292312622,
      "learning_rate": 4e-05,
      "loss": 2.7223,
      "step": 20
    },
    {
      "epoch": 0.863697705802969,
      "grad_norm": 0.509838342666626,
      "learning_rate": 8e-05,
      "loss": 2.063,
      "step": 40
    },
    {
      "epoch": 0.9932523616734144,
      "eval_loss": 1.6813331842422485,
      "eval_runtime": 12.0916,
      "eval_samples_per_second": 30.765,
      "eval_steps_per_second": 3.887,
      "step": 46
    },
    {
      "epoch": 1.2955465587044535,
      "grad_norm": 0.6181766390800476,
      "learning_rate": 0.00012,
      "loss": 1.8828,
      "step": 60
    },
    {
      "epoch": 1.7273954116059378,
      "grad_norm": 0.8454991579055786,
      "learning_rate": 0.00016,
      "loss": 1.6444,
      "step": 80
    },
    {
      "epoch": 1.9865047233468287,
      "eval_loss": 1.2503478527069092,
      "eval_runtime": 12.1403,
      "eval_samples_per_second": 30.642,
      "eval_steps_per_second": 3.871,
      "step": 92
    },
    {
      "epoch": 2.1592442645074224,
      "grad_norm": 1.1776914596557617,
      "learning_rate": 0.0002,
      "loss": 1.5242,
      "step": 100
    },
    {
      "epoch": 2.591093117408907,
      "grad_norm": 1.091369390487671,
      "learning_rate": 0.00019943348002101371,
      "loss": 1.3296,
      "step": 120
    },
    {
      "epoch": 2.979757085020243,
      "eval_loss": 1.028580904006958,
      "eval_runtime": 12.1164,
      "eval_samples_per_second": 30.702,
      "eval_steps_per_second": 3.879,
      "step": 138
    },
    {
      "epoch": 3.0229419703103915,
      "grad_norm": 1.0799137353897095,
      "learning_rate": 0.00019774033898178667,
      "loss": 1.2821,
      "step": 140
    },
    {
      "epoch": 3.454790823211876,
      "grad_norm": 1.2310456037521362,
      "learning_rate": 0.00019493976084683813,
      "loss": 1.0677,
      "step": 160
    },
    {
      "epoch": 3.8866396761133606,
      "grad_norm": 1.1375850439071655,
      "learning_rate": 0.00019106347728549135,
      "loss": 1.0588,
      "step": 180
    },
    {
      "epoch": 3.9946018893387314,
      "eval_loss": 0.9346915483474731,
      "eval_runtime": 12.162,
      "eval_samples_per_second": 30.587,
      "eval_steps_per_second": 3.864,
      "step": 185
    }
  ],
  "logging_steps": 20,
  "max_steps": 690,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 15,
  "save_steps": 500,
  "total_flos": 1.64514233315328e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}