checkpoints/checkpoint-20/trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.2306356132030487,
  "best_model_checkpoint": "./testing_checkpoint/checkpoint-20",
  "epoch": 5.0,
  "eval_steps": 20,
  "global_step": 20,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 5.0,
      "grad_norm": 0.37287092208862305,
      "learning_rate": 0.00025249999999999996,
      "loss": 2.0817,
      "step": 20
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.2306356132030487,
      "eval_runtime": 0.1699,
      "eval_samples_per_second": 29.43,
      "eval_steps_per_second": 5.886,
      "step": 20
    }
  ],
  "logging_steps": 20,
  "max_steps": 120,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 30,
  "save_steps": 20,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 30447894528000.0,
  "train_batch_size": 5,
  "trial_name": null,
  "trial_params": null
}