trainer_state.json

{
  "best_metric": 0.8416,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0898/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "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": 0.0001,
      "loss": 1.5252,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7874666666666666,
      "eval_loss": 0.7149600982666016,
      "eval_runtime": 13.4649,
      "eval_samples_per_second": 278.503,
      "eval_steps_per_second": 4.382,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.4914,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8272,
      "eval_loss": 0.5701954364776611,
      "eval_runtime": 12.9454,
      "eval_samples_per_second": 289.678,
      "eval_steps_per_second": 4.558,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.2658,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8405333333333334,
      "eval_loss": 0.5462485551834106,
      "eval_runtime": 12.294,
      "eval_samples_per_second": 305.026,
      "eval_steps_per_second": 4.799,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.3115854263305664,
      "learning_rate": 0.0001,
      "loss": 0.1714,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8341333333333333,
      "eval_loss": 0.5900708436965942,
      "eval_runtime": 12.3004,
      "eval_samples_per_second": 304.869,
      "eval_steps_per_second": 4.797,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.15606819093227386,
      "learning_rate": 0.0001,
      "loss": 0.1231,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.832,
      "eval_loss": 0.6140334010124207,
      "eval_runtime": 13.2137,
      "eval_samples_per_second": 283.797,
      "eval_steps_per_second": 4.465,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.12066475301980972,
      "learning_rate": 0.0001,
      "loss": 0.0948,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8416,
      "eval_loss": 0.6247767210006714,
      "eval_runtime": 12.418,
      "eval_samples_per_second": 301.98,
      "eval_steps_per_second": 4.751,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}