trainer_state.json

{
  "best_metric": 0.7832,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0546/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": 0.0003,
      "loss": 2.2691,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6301333333333333,
      "eval_loss": 1.21942937374115,
      "eval_runtime": 5.812,
      "eval_samples_per_second": 645.214,
      "eval_steps_per_second": 10.151,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.0552810430526733,
      "learning_rate": 0.0003,
      "loss": 0.8853,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7184,
      "eval_loss": 0.9439092874526978,
      "eval_runtime": 5.816,
      "eval_samples_per_second": 644.772,
      "eval_steps_per_second": 10.144,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.5895276069641113,
      "learning_rate": 0.0003,
      "loss": 0.5815,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7365333333333334,
      "eval_loss": 0.8514481782913208,
      "eval_runtime": 5.8187,
      "eval_samples_per_second": 644.473,
      "eval_steps_per_second": 10.14,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.896550178527832,
      "learning_rate": 0.0003,
      "loss": 0.4158,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.764,
      "eval_loss": 0.8301653861999512,
      "eval_runtime": 6.2428,
      "eval_samples_per_second": 600.692,
      "eval_steps_per_second": 9.451,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.022973664104938507,
      "learning_rate": 0.0003,
      "loss": 0.3073,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7754666666666666,
      "eval_loss": 0.773627519607544,
      "eval_runtime": 5.9015,
      "eval_samples_per_second": 635.431,
      "eval_steps_per_second": 9.997,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.08008267730474472,
      "learning_rate": 0.0003,
      "loss": 0.2394,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7808,
      "eval_loss": 0.804603099822998,
      "eval_runtime": 5.9281,
      "eval_samples_per_second": 632.577,
      "eval_steps_per_second": 9.953,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.3888905644416809,
      "learning_rate": 0.0003,
      "loss": 0.2033,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.7832,
      "eval_loss": 0.8266289234161377,
      "eval_runtime": 5.794,
      "eval_samples_per_second": 647.225,
      "eval_steps_per_second": 10.183,
      "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
}