trainer_state.json

{
  "best_metric": 0.8917333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0012/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": 4.2878592878592884e-05,
      "loss": 1.6794,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8133333333333334,
      "eval_loss": 0.720173716545105,
      "eval_runtime": 11.4049,
      "eval_samples_per_second": 328.806,
      "eval_steps_per_second": 5.173,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.720415115356445,
      "learning_rate": 3.573573573573574e-05,
      "loss": 0.4803,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8565333333333334,
      "eval_loss": 0.4950363337993622,
      "eval_runtime": 11.3945,
      "eval_samples_per_second": 329.107,
      "eval_steps_per_second": 5.178,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.117828845977783,
      "learning_rate": 2.8592878592878596e-05,
      "loss": 0.2412,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8781333333333333,
      "eval_loss": 0.4074266850948334,
      "eval_runtime": 10.9383,
      "eval_samples_per_second": 342.831,
      "eval_steps_per_second": 5.394,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": Infinity,
      "learning_rate": 2.147147147147147e-05,
      "loss": 0.1232,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8754666666666666,
      "eval_loss": 0.42447948455810547,
      "eval_runtime": 11.1987,
      "eval_samples_per_second": 334.861,
      "eval_steps_per_second": 5.268,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.0121376514434814,
      "learning_rate": 1.4328614328614328e-05,
      "loss": 0.0566,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8818666666666667,
      "eval_loss": 0.4186410903930664,
      "eval_runtime": 11.1072,
      "eval_samples_per_second": 337.62,
      "eval_steps_per_second": 5.312,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.9164219498634338,
      "learning_rate": 7.1857571857571855e-06,
      "loss": 0.0245,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8882666666666666,
      "eval_loss": 0.40363892912864685,
      "eval_runtime": 10.8919,
      "eval_samples_per_second": 344.294,
      "eval_steps_per_second": 5.417,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.38206425309181213,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.0116,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8917333333333334,
      "eval_loss": 0.4002712666988373,
      "eval_runtime": 10.9283,
      "eval_samples_per_second": 343.144,
      "eval_steps_per_second": 5.399,
      "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
}