checkpoint_dir/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.010351966873706004,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.003450655624568668,
      "grad_norm": 0.7515325546264648,
      "learning_rate": 4.665063509461098e-06,
      "loss": 1.4845,
      "step": 20
    },
    {
      "epoch": 0.006901311249137336,
      "grad_norm": 0.6366741061210632,
      "learning_rate": 1.852952387243698e-06,
      "loss": 1.4653,
      "step": 40
    },
    {
      "epoch": 0.010351966873706004,
      "grad_norm": 0.38952797651290894,
      "learning_rate": 0.0,
      "loss": 1.4896,
      "step": 60
    },
    {
      "epoch": 0.010351966873706004,
      "step": 60,
      "total_flos": 5503081701703680.0,
      "train_loss": 1.4797897656758627,
      "train_runtime": 2523.2494,
      "train_samples_per_second": 0.095,
      "train_steps_per_second": 0.024
    }
  ],
  "logging_steps": 20,
  "max_steps": 60,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 60,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5503081701703680.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}