trainer_state.json

{
  "best_metric": 0.8586666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0663/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": 0.0001,
      "loss": 1.754,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7792,
      "eval_loss": 0.7989434003829956,
      "eval_runtime": 12.5643,
      "eval_samples_per_second": 298.466,
      "eval_steps_per_second": 4.696,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.993532180786133,
      "learning_rate": 0.0001,
      "loss": 0.5466,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8274666666666667,
      "eval_loss": 0.5970971584320068,
      "eval_runtime": 11.951,
      "eval_samples_per_second": 313.782,
      "eval_steps_per_second": 4.937,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.3393,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8464,
      "eval_loss": 0.5140204429626465,
      "eval_runtime": 12.1887,
      "eval_samples_per_second": 307.661,
      "eval_steps_per_second": 4.841,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.20600463449954987,
      "learning_rate": 0.0001,
      "loss": 0.2145,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8554666666666667,
      "eval_loss": 0.4981195628643036,
      "eval_runtime": 12.2452,
      "eval_samples_per_second": 306.242,
      "eval_steps_per_second": 4.818,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.04584474489092827,
      "learning_rate": 0.0001,
      "loss": 0.1613,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.828,
      "eval_loss": 0.6768359541893005,
      "eval_runtime": 11.8707,
      "eval_samples_per_second": 315.903,
      "eval_steps_per_second": 4.97,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 35.88604736328125,
      "learning_rate": 0.0001,
      "loss": 0.1227,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.5549832582473755,
      "eval_runtime": 12.3057,
      "eval_samples_per_second": 304.737,
      "eval_steps_per_second": 4.795,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03892930969595909,
      "learning_rate": 0.0001,
      "loss": 0.0976,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8586666666666667,
      "eval_loss": 0.5953060984611511,
      "eval_runtime": 11.8814,
      "eval_samples_per_second": 315.619,
      "eval_steps_per_second": 4.966,
      "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
}