trainer_state.json

{
  "best_metric": 0.8498666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0014/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": 9e-05,
      "loss": 1.6123,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7882666666666667,
      "eval_loss": 0.7811404466629028,
      "eval_runtime": 8.185,
      "eval_samples_per_second": 458.154,
      "eval_steps_per_second": 7.208,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.3528080880641937,
      "learning_rate": 9e-05,
      "loss": 0.5594,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8117333333333333,
      "eval_loss": 0.6570170521736145,
      "eval_runtime": 7.7819,
      "eval_samples_per_second": 481.888,
      "eval_steps_per_second": 7.582,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.8141839504241943,
      "learning_rate": 9e-05,
      "loss": 0.3418,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8256,
      "eval_loss": 0.586188018321991,
      "eval_runtime": 6.2323,
      "eval_samples_per_second": 601.708,
      "eval_steps_per_second": 9.467,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2663261592388153,
      "learning_rate": 9e-05,
      "loss": 0.2334,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.84,
      "eval_loss": 0.5555790066719055,
      "eval_runtime": 6.3034,
      "eval_samples_per_second": 594.917,
      "eval_steps_per_second": 9.36,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.21210797131061554,
      "learning_rate": 9e-05,
      "loss": 0.1659,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8474666666666667,
      "eval_loss": 0.5734807252883911,
      "eval_runtime": 6.2756,
      "eval_samples_per_second": 597.554,
      "eval_steps_per_second": 9.402,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.032158706337213516,
      "learning_rate": 9e-05,
      "loss": 0.1245,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.5758804082870483,
      "eval_runtime": 6.1752,
      "eval_samples_per_second": 607.27,
      "eval_steps_per_second": 9.554,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.8266268968582153,
      "learning_rate": 9e-05,
      "loss": 0.1023,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8498666666666667,
      "eval_loss": 0.547307014465332,
      "eval_runtime": 6.3518,
      "eval_samples_per_second": 590.383,
      "eval_steps_per_second": 9.289,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}