trainer_state.json

{
  "best_metric": 0.8896,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0511/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 9.507764069616767e-05,
      "loss": 1.5862,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7944,
      "eval_loss": 0.7263412475585938,
      "eval_runtime": 6.5408,
      "eval_samples_per_second": 573.321,
      "eval_steps_per_second": 9.02,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.127974762931168e-05,
      "loss": 0.4803,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8461333333333333,
      "eval_loss": 0.5262216329574585,
      "eval_runtime": 6.2407,
      "eval_samples_per_second": 600.895,
      "eval_steps_per_second": 9.454,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.0402252674102783,
      "learning_rate": 6.125740150220697e-05,
      "loss": 0.2108,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8541333333333333,
      "eval_loss": 0.48816099762916565,
      "eval_runtime": 5.9166,
      "eval_samples_per_second": 633.805,
      "eval_steps_per_second": 9.972,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.3281441628932953,
      "learning_rate": 3.900538894462112e-05,
      "loss": 0.0857,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.4595865309238434,
      "eval_runtime": 5.7762,
      "eval_samples_per_second": 649.211,
      "eval_steps_per_second": 10.214,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.3933327794075012,
      "learning_rate": 1.893099394660879e-05,
      "loss": 0.0278,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8821333333333333,
      "eval_loss": 0.4289211630821228,
      "eval_runtime": 9.2003,
      "eval_samples_per_second": 407.596,
      "eval_steps_per_second": 6.413,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.5105985999107361,
      "learning_rate": 5.0101966338242764e-06,
      "loss": 0.0078,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8893333333333333,
      "eval_loss": 0.41383644938468933,
      "eval_runtime": 5.7201,
      "eval_samples_per_second": 655.584,
      "eval_steps_per_second": 10.315,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.033168975263834,
      "learning_rate": 1.8164143966159507e-10,
      "loss": 0.0047,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8896,
      "eval_loss": 0.4122750759124756,
      "eval_runtime": 6.177,
      "eval_samples_per_second": 607.086,
      "eval_steps_per_second": 9.551,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}