trainer_state.json

{
  "best_metric": 0.8928,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0899/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": 3.867995023727417,
      "learning_rate": 3e-05,
      "loss": 2.2633,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7797333333333333,
      "eval_loss": 1.0576770305633545,
      "eval_runtime": 12.3578,
      "eval_samples_per_second": 303.452,
      "eval_steps_per_second": 4.774,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 3e-05,
      "loss": 0.7106,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8493333333333334,
      "eval_loss": 0.5892411470413208,
      "eval_runtime": 12.3568,
      "eval_samples_per_second": 303.477,
      "eval_steps_per_second": 4.775,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.1499887704849243,
      "learning_rate": 3e-05,
      "loss": 0.3778,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8829333333333333,
      "eval_loss": 0.4373627007007599,
      "eval_runtime": 12.7594,
      "eval_samples_per_second": 293.9,
      "eval_steps_per_second": 4.624,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.433309406042099,
      "learning_rate": 3e-05,
      "loss": 0.2385,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.39849430322647095,
      "eval_runtime": 12.6125,
      "eval_samples_per_second": 297.324,
      "eval_steps_per_second": 4.678,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 16.405712127685547,
      "learning_rate": 3e-05,
      "loss": 0.1539,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8869333333333334,
      "eval_loss": 0.4000135064125061,
      "eval_runtime": 12.5588,
      "eval_samples_per_second": 298.595,
      "eval_steps_per_second": 4.698,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 39.58985137939453,
      "learning_rate": 3e-05,
      "loss": 0.1033,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8922666666666667,
      "eval_loss": 0.3867323100566864,
      "eval_runtime": 12.7506,
      "eval_samples_per_second": 294.104,
      "eval_steps_per_second": 4.627,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.19844813644886017,
      "learning_rate": 3e-05,
      "loss": 0.0699,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8928,
      "eval_loss": 0.3923831582069397,
      "eval_runtime": 12.5088,
      "eval_samples_per_second": 299.789,
      "eval_steps_per_second": 4.717,
      "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
}