trainer_state.json

{
  "best_metric": 0.904,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0274/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": 8.55698766265509e-05,
      "loss": 1.5446,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8010666666666667,
      "eval_loss": 0.692766547203064,
      "eval_runtime": 7.7485,
      "eval_samples_per_second": 483.964,
      "eval_steps_per_second": 7.614,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7.31517728663805e-05,
      "loss": 0.536,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8472,
      "eval_loss": 0.5134183764457703,
      "eval_runtime": 6.7294,
      "eval_samples_per_second": 557.26,
      "eval_steps_per_second": 8.768,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4285467863082886,
      "learning_rate": 5.5131661351986275e-05,
      "loss": 0.3161,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8725333333333334,
      "eval_loss": 0.44778671860694885,
      "eval_runtime": 6.6476,
      "eval_samples_per_second": 564.11,
      "eval_steps_per_second": 8.875,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.19927065074443817,
      "learning_rate": 3.5104850050159014e-05,
      "loss": 0.1896,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.45150625705718994,
      "eval_runtime": 7.0167,
      "eval_samples_per_second": 534.442,
      "eval_steps_per_second": 8.409,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.21656665205955505,
      "learning_rate": 1.703789455194791e-05,
      "loss": 0.1023,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8888,
      "eval_loss": 0.39833390712738037,
      "eval_runtime": 6.996,
      "eval_samples_per_second": 536.017,
      "eval_steps_per_second": 8.433,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.04831075668334961,
      "learning_rate": 4.509176970441849e-06,
      "loss": 0.0544,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8976,
      "eval_loss": 0.37391453981399536,
      "eval_runtime": 6.9129,
      "eval_samples_per_second": 542.465,
      "eval_steps_per_second": 8.535,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.1822198629379272,
      "learning_rate": 1.6347729569543557e-10,
      "loss": 0.0343,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.36150142550468445,
      "eval_runtime": 7.1169,
      "eval_samples_per_second": 526.918,
      "eval_steps_per_second": 8.29,
      "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
}