{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 1000,
  "global_step": 5951,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00016803898504453034,
      "grad_norm": 6.990426540374756,
      "learning_rate": 0.0,
      "loss": 2.5326,
      "step": 1
    },
    {
      "epoch": 0.16803898504453033,
      "grad_norm": 4.953593730926514,
      "learning_rate": 4.196597353497165e-06,
      "loss": 2.6779,
      "step": 1000
    },
    {
      "epoch": 0.16803898504453033,
      "eval_cosine_accuracy": 0.9457356333732605,
      "eval_loss": 1.3122717142105103,
      "eval_runtime": 21.2029,
      "eval_samples_per_second": 448.476,
      "eval_steps_per_second": 3.537,
      "step": 1000
    },
    {
      "epoch": 0.33607797008906065,
      "grad_norm": 5.172318458557129,
      "learning_rate": 8.397395505145979e-06,
      "loss": 2.1925,
      "step": 2000
    },
    {
      "epoch": 0.33607797008906065,
      "eval_cosine_accuracy": 0.951729953289032,
      "eval_loss": 1.2547080516815186,
      "eval_runtime": 21.0477,
      "eval_samples_per_second": 451.783,
      "eval_steps_per_second": 3.563,
      "step": 2000
    },
    {
      "epoch": 0.504116955133591,
      "grad_norm": 5.013792514801025,
      "learning_rate": 1.2598193656794792e-05,
      "loss": 1.9607,
      "step": 3000
    },
    {
      "epoch": 0.504116955133591,
      "eval_cosine_accuracy": 0.9543590545654297,
      "eval_loss": 1.2117713689804077,
      "eval_runtime": 21.1746,
      "eval_samples_per_second": 449.076,
      "eval_steps_per_second": 3.542,
      "step": 3000
    },
    {
      "epoch": 0.6721559401781213,
      "grad_norm": 5.219577312469482,
      "learning_rate": 1.6798991808443605e-05,
      "loss": 1.7977,
      "step": 4000
    },
    {
      "epoch": 0.6721559401781213,
      "eval_cosine_accuracy": 0.9588810801506042,
      "eval_loss": 1.1573827266693115,
      "eval_runtime": 20.9565,
      "eval_samples_per_second": 453.749,
      "eval_steps_per_second": 3.579,
      "step": 4000
    },
    {
      "epoch": 0.8401949252226516,
      "grad_norm": 5.110533714294434,
      "learning_rate": 1.9751089639237518e-05,
      "loss": 1.6686,
      "step": 5000
    },
    {
      "epoch": 0.8401949252226516,
      "eval_cosine_accuracy": 0.9588810801506042,
      "eval_loss": 1.1438082456588745,
      "eval_runtime": 20.8496,
      "eval_samples_per_second": 456.075,
      "eval_steps_per_second": 3.597,
      "step": 5000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 23804,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 0.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}