trainer_state.json

{
  "best_metric": 0.8824,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0874/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "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": 3e-05,
      "loss": 2.191,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7882666666666667,
      "eval_loss": 0.9892815351486206,
      "eval_runtime": 6.4226,
      "eval_samples_per_second": 583.876,
      "eval_steps_per_second": 9.186,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.6637,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8605333333333334,
      "eval_loss": 0.5470876693725586,
      "eval_runtime": 6.2925,
      "eval_samples_per_second": 595.952,
      "eval_steps_per_second": 9.376,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.874481201171875,
      "learning_rate": 3e-05,
      "loss": 0.3749,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8770666666666667,
      "eval_loss": 0.44349923729896545,
      "eval_runtime": 6.3133,
      "eval_samples_per_second": 593.987,
      "eval_steps_per_second": 9.345,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0958795547485352,
      "learning_rate": 3e-05,
      "loss": 0.2424,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8818666666666667,
      "eval_loss": 0.4190085530281067,
      "eval_runtime": 6.746,
      "eval_samples_per_second": 555.889,
      "eval_steps_per_second": 8.746,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.22354614734649658,
      "learning_rate": 3e-05,
      "loss": 0.1675,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8824,
      "eval_loss": 0.393738716840744,
      "eval_runtime": 6.3503,
      "eval_samples_per_second": 590.527,
      "eval_steps_per_second": 9.291,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}