{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 4000,
  "global_step": 29288,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00010245901639344262,
      "grad_norm": 6.315417766571045,
      "learning_rate": 0.0,
      "loss": 4.1362,
      "step": 1
    },
    {
      "epoch": 0.4098360655737705,
      "grad_norm": 4.62825345993042,
      "learning_rate": 1.536501024590164e-05,
      "loss": 3.1603,
      "step": 4000
    },
    {
      "epoch": 0.4098360655737705,
      "eval_cosine_accuracy": 0.953901469707489,
      "eval_loss": 1.0815402269363403,
      "eval_runtime": 24.8505,
      "eval_samples_per_second": 380.595,
      "eval_steps_per_second": 1.489,
      "step": 4000
    },
    {
      "epoch": 0.819672131147541,
      "grad_norm": 4.6219048500061035,
      "learning_rate": 2.999728935936485e-05,
      "loss": 2.7017,
      "step": 8000
    },
    {
      "epoch": 0.819672131147541,
      "eval_cosine_accuracy": 0.9081200957298279,
      "eval_loss": 1.3302178382873535,
      "eval_runtime": 22.5389,
      "eval_samples_per_second": 419.629,
      "eval_steps_per_second": 1.642,
      "step": 8000
    },
    {
      "epoch": 1.2295081967213115,
      "grad_norm": 4.111813068389893,
      "learning_rate": 2.868919465032591e-05,
      "loss": 2.4584,
      "step": 12000
    },
    {
      "epoch": 1.2295081967213115,
      "eval_cosine_accuracy": 0.957390546798706,
      "eval_loss": 0.9852072596549988,
      "eval_runtime": 25.1342,
      "eval_samples_per_second": 376.3,
      "eval_steps_per_second": 1.472,
      "step": 12000
    },
    {
      "epoch": 1.639344262295082,
      "grad_norm": 3.6833724975585938,
      "learning_rate": 2.51957420925816e-05,
      "loss": 2.2777,
      "step": 16000
    },
    {
      "epoch": 1.639344262295082,
      "eval_cosine_accuracy": 0.9606682062149048,
      "eval_loss": 0.9453343749046326,
      "eval_runtime": 22.7137,
      "eval_samples_per_second": 416.401,
      "eval_steps_per_second": 1.629,
      "step": 16000
    },
    {
      "epoch": 2.0491400491400493,
      "grad_norm": 7.341921329498291,
      "learning_rate": 2.0076071378946667e-05,
      "loss": 2.1676,
      "step": 20000
    },
    {
      "epoch": 2.0491400491400493,
      "eval_cosine_accuracy": 0.961091160774231,
      "eval_loss": 0.9533414840698242,
      "eval_runtime": 22.8729,
      "eval_samples_per_second": 413.502,
      "eval_steps_per_second": 1.618,
      "step": 20000
    },
    {
      "epoch": 2.458640458640459,
      "grad_norm": 4.12435245513916,
      "learning_rate": 1.4144955104259103e-05,
      "loss": 2.067,
      "step": 24000
    },
    {
      "epoch": 2.458640458640459,
      "eval_cosine_accuracy": 0.9617255330085754,
      "eval_loss": 0.9420561790466309,
      "eval_runtime": 22.7914,
      "eval_samples_per_second": 414.981,
      "eval_steps_per_second": 1.623,
      "step": 24000
    },
    {
      "epoch": 2.868140868140868,
      "grad_norm": 3.720942497253418,
      "learning_rate": 8.351689838906174e-06,
      "loss": 2.0149,
      "step": 28000
    },
    {
      "epoch": 2.868140868140868,
      "eval_cosine_accuracy": 0.9643687605857849,
      "eval_loss": 0.926335871219635,
      "eval_runtime": 22.8686,
      "eval_samples_per_second": 413.58,
      "eval_steps_per_second": 1.618,
      "step": 28000
    }
  ],
  "logging_steps": 4000,
  "max_steps": 39040,
  "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": 256,
  "trial_name": null,
  "trial_params": null
}