trainer_state.json

{
  "best_metric": 0.5282666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0717/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": 0.0005,
      "loss": 3.2316,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2384,
      "eval_loss": 2.903496265411377,
      "eval_runtime": 11.4053,
      "eval_samples_per_second": 328.795,
      "eval_steps_per_second": 5.173,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.6355,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3373333333333333,
      "eval_loss": 2.460258960723877,
      "eval_runtime": 12.7857,
      "eval_samples_per_second": 293.297,
      "eval_steps_per_second": 4.615,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.2707,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3824,
      "eval_loss": 2.2871828079223633,
      "eval_runtime": 11.2305,
      "eval_samples_per_second": 333.911,
      "eval_steps_per_second": 5.254,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.376530647277832,
      "learning_rate": 0.0005,
      "loss": 1.9735,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.43466666666666665,
      "eval_loss": 2.043107032775879,
      "eval_runtime": 11.4433,
      "eval_samples_per_second": 327.704,
      "eval_steps_per_second": 5.156,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 16.07221794128418,
      "learning_rate": 0.0005,
      "loss": 1.7341,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4706666666666667,
      "eval_loss": 1.8701802492141724,
      "eval_runtime": 11.0763,
      "eval_samples_per_second": 338.561,
      "eval_steps_per_second": 5.327,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 12.31285285949707,
      "learning_rate": 0.0005,
      "loss": 1.4894,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4898666666666667,
      "eval_loss": 1.7908234596252441,
      "eval_runtime": 11.1284,
      "eval_samples_per_second": 336.975,
      "eval_steps_per_second": 5.302,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 11.235486030578613,
      "learning_rate": 0.0005,
      "loss": 1.274,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5282666666666667,
      "eval_loss": 1.70403254032135,
      "eval_runtime": 11.4453,
      "eval_samples_per_second": 327.645,
      "eval_steps_per_second": 5.155,
      "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
}