trainer_state.json

{
  "best_metric": 0.8690666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0027/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": 0.00026261261261261256,
      "loss": 1.813,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7168,
      "eval_loss": 0.9322386980056763,
      "eval_runtime": 8.5404,
      "eval_samples_per_second": 439.087,
      "eval_steps_per_second": 6.908,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.814764976501465,
      "learning_rate": 0.0002251126126126126,
      "loss": 0.6817,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7829333333333334,
      "eval_loss": 0.6971080303192139,
      "eval_runtime": 7.7156,
      "eval_samples_per_second": 486.029,
      "eval_steps_per_second": 7.647,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.128695487976074,
      "learning_rate": 0.00018761261261261258,
      "loss": 0.4041,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8125333333333333,
      "eval_loss": 0.6166936755180359,
      "eval_runtime": 6.5621,
      "eval_samples_per_second": 571.463,
      "eval_steps_per_second": 8.991,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001502252252252252,
      "loss": 0.24,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8373333333333334,
      "eval_loss": 0.5452679991722107,
      "eval_runtime": 6.585,
      "eval_samples_per_second": 569.473,
      "eval_steps_per_second": 8.96,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00011283783783783782,
      "loss": 0.1382,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8384,
      "eval_loss": 0.5649423003196716,
      "eval_runtime": 6.8896,
      "eval_samples_per_second": 544.296,
      "eval_steps_per_second": 8.564,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.010372706688940525,
      "learning_rate": 7.533783783783783e-05,
      "loss": 0.0636,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8557333333333333,
      "eval_loss": 0.5369786620140076,
      "eval_runtime": 7.0319,
      "eval_samples_per_second": 533.283,
      "eval_steps_per_second": 8.39,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03680369630455971,
      "learning_rate": 3.783783783783783e-05,
      "loss": 0.0247,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8690666666666667,
      "eval_loss": 0.5272208452224731,
      "eval_runtime": 7.1356,
      "eval_samples_per_second": 525.536,
      "eval_steps_per_second": 8.268,
      "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
}