trainer_state.json

{
  "best_metric": 0.9082666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0131/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": 7.71814671814672e-05,
      "loss": 1.5123,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.804,
      "eval_loss": 0.7046989798545837,
      "eval_runtime": 8.0865,
      "eval_samples_per_second": 463.736,
      "eval_steps_per_second": 7.296,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 6.436293436293436e-05,
      "loss": 0.5128,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8637333333333334,
      "eval_loss": 0.47097817063331604,
      "eval_runtime": 7.1366,
      "eval_samples_per_second": 525.463,
      "eval_steps_per_second": 8.267,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4535754323005676,
      "learning_rate": 5.1505791505791514e-05,
      "loss": 0.3027,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8784,
      "eval_loss": 0.41604697704315186,
      "eval_runtime": 6.857,
      "eval_samples_per_second": 546.889,
      "eval_steps_per_second": 8.604,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.6789469122886658,
      "learning_rate": 3.864864864864865e-05,
      "loss": 0.1852,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.4209411144256592,
      "eval_runtime": 6.8366,
      "eval_samples_per_second": 548.516,
      "eval_steps_per_second": 8.63,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.044501781463623,
      "learning_rate": 2.5791505791505793e-05,
      "loss": 0.1118,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.888,
      "eval_loss": 0.3767320215702057,
      "eval_runtime": 6.8978,
      "eval_samples_per_second": 543.652,
      "eval_steps_per_second": 8.553,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.07354779541492462,
      "learning_rate": 1.2934362934362936e-05,
      "loss": 0.0589,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8994666666666666,
      "eval_loss": 0.36963483691215515,
      "eval_runtime": 6.8763,
      "eval_samples_per_second": 545.355,
      "eval_steps_per_second": 8.58,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.17848744988441467,
      "learning_rate": 7.722007722007723e-08,
      "loss": 0.0309,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9082666666666667,
      "eval_loss": 0.3393380343914032,
      "eval_runtime": 6.8229,
      "eval_samples_per_second": 549.616,
      "eval_steps_per_second": 8.647,
      "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
}