trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.871794871794872,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9572649572649573,
      "eval_loss": 1.2024827003479004,
      "eval_runtime": 23.9779,
      "eval_samples_per_second": 8.216,
      "eval_steps_per_second": 0.167,
      "step": 7
    },
    {
      "epoch": 1.3675213675213675,
      "grad_norm": 20.20309688101842,
      "learning_rate": 5e-06,
      "loss": 1.3815,
      "step": 10
    },
    {
      "epoch": 1.9145299145299144,
      "eval_loss": 1.0915584564208984,
      "eval_runtime": 22.8554,
      "eval_samples_per_second": 8.619,
      "eval_steps_per_second": 0.175,
      "step": 14
    },
    {
      "epoch": 2.735042735042735,
      "grad_norm": 3.935326563247966,
      "learning_rate": 5e-06,
      "loss": 1.077,
      "step": 20
    },
    {
      "epoch": 2.871794871794872,
      "eval_loss": 1.0183525085449219,
      "eval_runtime": 22.6691,
      "eval_samples_per_second": 8.69,
      "eval_steps_per_second": 0.176,
      "step": 21
    },
    {
      "epoch": 2.871794871794872,
      "step": 21,
      "total_flos": 35123437240320.0,
      "train_loss": 1.2196591127486456,
      "train_runtime": 6047.8715,
      "train_samples_per_second": 1.855,
      "train_steps_per_second": 0.003
    }
  ],
  "logging_steps": 10,
  "max_steps": 21,
  "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": 35123437240320.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}