trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1755,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 0.87109375,
      "learning_rate": 0.00018,
      "loss": 1.2314,
      "step": 585
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.166209101676941,
      "eval_runtime": 30.6211,
      "eval_samples_per_second": 4.245,
      "eval_steps_per_second": 4.245,
      "step": 585
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.90625,
      "learning_rate": 0.00016,
      "loss": 0.9796,
      "step": 1170
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.1119043827056885,
      "eval_runtime": 30.6112,
      "eval_samples_per_second": 4.247,
      "eval_steps_per_second": 4.247,
      "step": 1170
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8125,
      "learning_rate": 0.00014,
      "loss": 0.7218,
      "step": 1755
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.101173996925354,
      "eval_runtime": 30.6537,
      "eval_samples_per_second": 4.241,
      "eval_steps_per_second": 4.241,
      "step": 1755
    }
  ],
  "logging_steps": 500,
  "max_steps": 5850,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 1.4548022929118822e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}