trainer_state.json

{
  "best_metric": 0.8792,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0293/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": 8.575718575718577e-05,
      "loss": 1.5919,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7749333333333334,
      "eval_loss": 0.7947626113891602,
      "eval_runtime": 8.3517,
      "eval_samples_per_second": 449.01,
      "eval_steps_per_second": 7.064,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7986994981765747,
      "learning_rate": 7.147147147147148e-05,
      "loss": 0.5378,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8232,
      "eval_loss": 0.6017781496047974,
      "eval_runtime": 6.0951,
      "eval_samples_per_second": 615.252,
      "eval_steps_per_second": 9.68,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 5.7228657228657234e-05,
      "loss": 0.3072,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8376,
      "eval_loss": 0.5527909994125366,
      "eval_runtime": 6.1438,
      "eval_samples_per_second": 610.371,
      "eval_steps_per_second": 9.603,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.72588711977005,
      "learning_rate": 4.294294294294294e-05,
      "loss": 0.1744,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.4833321273326874,
      "eval_runtime": 6.2167,
      "eval_samples_per_second": 603.209,
      "eval_steps_per_second": 9.49,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.219998836517334,
      "learning_rate": 2.8657228657228656e-05,
      "loss": 0.0922,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8666666666666667,
      "eval_loss": 0.47663554549217224,
      "eval_runtime": 6.7113,
      "eval_samples_per_second": 558.762,
      "eval_steps_per_second": 8.791,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.4801584482192993,
      "learning_rate": 1.4371514371514371e-05,
      "loss": 0.0447,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8706666666666667,
      "eval_loss": 0.4616999924182892,
      "eval_runtime": 6.0622,
      "eval_samples_per_second": 618.589,
      "eval_steps_per_second": 9.732,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.4213399589061737,
      "learning_rate": 8.580008580008581e-08,
      "loss": 0.0193,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8792,
      "eval_loss": 0.4375804662704468,
      "eval_runtime": 6.1011,
      "eval_samples_per_second": 614.639,
      "eval_steps_per_second": 9.67,
      "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
}