trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 6486,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23126734505087881,
      "grad_norm": 0.5234603881835938,
      "learning_rate": 5e-05,
      "loss": 2.7489,
      "step": 500
    },
    {
      "epoch": 0.46253469010175763,
      "grad_norm": 0.7101452946662903,
      "learning_rate": 4.582358837287003e-05,
      "loss": 0.5566,
      "step": 1000
    },
    {
      "epoch": 0.6938020351526365,
      "grad_norm": 0.36853936314582825,
      "learning_rate": 4.164717674574006e-05,
      "loss": 0.4853,
      "step": 1500
    },
    {
      "epoch": 0.9250693802035153,
      "grad_norm": 0.4888463020324707,
      "learning_rate": 3.747076511861009e-05,
      "loss": 0.4564,
      "step": 2000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3744633197784424,
      "eval_runtime": 32.7403,
      "eval_samples_per_second": 29.352,
      "eval_steps_per_second": 3.696,
      "step": 2162
    },
    {
      "epoch": 1.1563367252543941,
      "grad_norm": 0.41663652658462524,
      "learning_rate": 3.329435349148012e-05,
      "loss": 0.4338,
      "step": 2500
    },
    {
      "epoch": 1.3876040703052728,
      "grad_norm": 0.4507719576358795,
      "learning_rate": 2.911794186435015e-05,
      "loss": 0.4154,
      "step": 3000
    },
    {
      "epoch": 1.6188714153561516,
      "grad_norm": 0.4204727113246918,
      "learning_rate": 2.4941530237220183e-05,
      "loss": 0.41,
      "step": 3500
    },
    {
      "epoch": 1.8501387604070305,
      "grad_norm": 0.3517729938030243,
      "learning_rate": 2.0765118610090213e-05,
      "loss": 0.3964,
      "step": 4000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.33997321128845215,
      "eval_runtime": 32.7708,
      "eval_samples_per_second": 29.325,
      "eval_steps_per_second": 3.692,
      "step": 4324
    },
    {
      "epoch": 2.0814061054579094,
      "grad_norm": 0.40001216530799866,
      "learning_rate": 1.6588706982960243e-05,
      "loss": 0.3875,
      "step": 4500
    },
    {
      "epoch": 2.3126734505087883,
      "grad_norm": 0.35182565450668335,
      "learning_rate": 1.2412295355830272e-05,
      "loss": 0.3867,
      "step": 5000
    },
    {
      "epoch": 2.543940795559667,
      "grad_norm": 0.4260983467102051,
      "learning_rate": 8.235883728700302e-06,
      "loss": 0.3792,
      "step": 5500
    },
    {
      "epoch": 2.7752081406105455,
      "grad_norm": 0.4158080518245697,
      "learning_rate": 4.059472101570331e-06,
      "loss": 0.373,
      "step": 6000
    }
  ],
  "logging_steps": 500,
  "max_steps": 6486,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.579514976534528e+16,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}