trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 99,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.30303030303030304,
      "grad_norm": 1.5732058135306703,
      "learning_rate": 5e-06,
      "loss": 0.7145,
      "step": 10
    },
    {
      "epoch": 0.6060606060606061,
      "grad_norm": 5.477077382338012,
      "learning_rate": 5e-06,
      "loss": 0.6119,
      "step": 20
    },
    {
      "epoch": 0.9090909090909091,
      "grad_norm": 2.3219180114679547,
      "learning_rate": 5e-06,
      "loss": 0.5956,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.5682209730148315,
      "eval_runtime": 3.7403,
      "eval_samples_per_second": 235.813,
      "eval_steps_per_second": 1.069,
      "step": 33
    },
    {
      "epoch": 1.2121212121212122,
      "grad_norm": 0.6856267648851094,
      "learning_rate": 5e-06,
      "loss": 0.5598,
      "step": 40
    },
    {
      "epoch": 1.5151515151515151,
      "grad_norm": 1.530770850167668,
      "learning_rate": 5e-06,
      "loss": 0.5373,
      "step": 50
    },
    {
      "epoch": 1.8181818181818183,
      "grad_norm": 1.2960700013825661,
      "learning_rate": 5e-06,
      "loss": 0.5344,
      "step": 60
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.5334687829017639,
      "eval_runtime": 3.6091,
      "eval_samples_per_second": 244.383,
      "eval_steps_per_second": 1.108,
      "step": 66
    },
    {
      "epoch": 2.121212121212121,
      "grad_norm": 1.0364074808402117,
      "learning_rate": 5e-06,
      "loss": 0.5131,
      "step": 70
    },
    {
      "epoch": 2.4242424242424243,
      "grad_norm": 1.1516572484623333,
      "learning_rate": 5e-06,
      "loss": 0.4878,
      "step": 80
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 0.953669869689903,
      "learning_rate": 5e-06,
      "loss": 0.4874,
      "step": 90
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.5247477293014526,
      "eval_runtime": 3.6293,
      "eval_samples_per_second": 243.02,
      "eval_steps_per_second": 1.102,
      "step": 99
    },
    {
      "epoch": 3.0,
      "step": 99,
      "total_flos": 165828687298560.0,
      "train_loss": 0.5521433955491192,
      "train_runtime": 1113.3042,
      "train_samples_per_second": 45.112,
      "train_steps_per_second": 0.089
    }
  ],
  "logging_steps": 10,
  "max_steps": 99,
  "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": 165828687298560.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}