trainer_state.json

{
  "best_global_step": 8,
  "best_metric": 0.21261577308177948,
  "best_model_checkpoint": "./vit_focus/checkpoint-8",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 8,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "eval_loss": 0.1353742480278015,
      "eval_mae": 0.4155503511428833,
      "eval_mse": 0.21432310342788696,
      "eval_runtime": 1.4322,
      "eval_samples_per_second": 6.982,
      "eval_steps_per_second": 1.396,
      "step": 2
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.13430055975914001,
      "eval_mae": 0.41518500447273254,
      "eval_mse": 0.21401900053024292,
      "eval_runtime": 1.3272,
      "eval_samples_per_second": 7.535,
      "eval_steps_per_second": 1.507,
      "step": 4
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.13232454657554626,
      "eval_mae": 0.414475679397583,
      "eval_mse": 0.21343325078487396,
      "eval_runtime": 1.3189,
      "eval_samples_per_second": 7.582,
      "eval_steps_per_second": 1.516,
      "step": 6
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.12981335818767548,
      "eval_mae": 0.41351836919784546,
      "eval_mse": 0.21261577308177948,
      "eval_runtime": 1.4603,
      "eval_samples_per_second": 6.848,
      "eval_steps_per_second": 1.37,
      "step": 8
    },
    {
      "epoch": 4.0,
      "step": 8,
      "total_flos": 0.0,
      "train_loss": 0.24888712167739868,
      "train_runtime": 34.9883,
      "train_samples_per_second": 8.917,
      "train_steps_per_second": 0.229
    }
  ],
  "logging_steps": 50,
  "max_steps": 8,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 0.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}