trainer_state.json

{
  "best_metric": 0.8722666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0253/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": 0.00025727155727155724,
      "loss": 2.0918,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.672,
      "eval_loss": 1.125277042388916,
      "eval_runtime": 7.6522,
      "eval_samples_per_second": 490.053,
      "eval_steps_per_second": 7.71,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0002145431145431145,
      "loss": 0.7694,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7762666666666667,
      "eval_loss": 0.7675689458847046,
      "eval_runtime": 7.192,
      "eval_samples_per_second": 521.412,
      "eval_steps_per_second": 8.204,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.2432502657175064,
      "learning_rate": 0.00017168597168597168,
      "loss": 0.42,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8213333333333334,
      "eval_loss": 0.6267414093017578,
      "eval_runtime": 7.2462,
      "eval_samples_per_second": 517.511,
      "eval_steps_per_second": 8.142,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.04791967198252678,
      "learning_rate": 0.0001288288288288288,
      "loss": 0.2123,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8117333333333333,
      "eval_loss": 0.6895110011100769,
      "eval_runtime": 7.2578,
      "eval_samples_per_second": 516.683,
      "eval_steps_per_second": 8.129,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.398215651512146,
      "learning_rate": 8.597168597168596e-05,
      "loss": 0.1018,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8496,
      "eval_loss": 0.578416645526886,
      "eval_runtime": 6.7473,
      "eval_samples_per_second": 555.776,
      "eval_steps_per_second": 8.744,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.015443765558302402,
      "learning_rate": 4.311454311454311e-05,
      "loss": 0.0385,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8642666666666666,
      "eval_loss": 0.5297573208808899,
      "eval_runtime": 7.1243,
      "eval_samples_per_second": 526.368,
      "eval_steps_per_second": 8.282,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.010387449525296688,
      "learning_rate": 2.5740025740025737e-07,
      "loss": 0.0121,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8722666666666666,
      "eval_loss": 0.5131955146789551,
      "eval_runtime": 8.3389,
      "eval_samples_per_second": 449.701,
      "eval_steps_per_second": 7.075,
      "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
}