trainer_state.json

{
  "best_metric": 0.8864,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0367/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": 2.9100759384327822e-05,
      "loss": 2.3322,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7416,
      "eval_loss": 1.1514216661453247,
      "eval_runtime": 6.3291,
      "eval_samples_per_second": 592.498,
      "eval_steps_per_second": 9.322,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 2.6510855361961218e-05,
      "loss": 0.783,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8181333333333334,
      "eval_loss": 0.6727586984634399,
      "eval_runtime": 6.427,
      "eval_samples_per_second": 583.477,
      "eval_steps_per_second": 9.18,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.1406304836273193,
      "learning_rate": 2.2527217722455156e-05,
      "loss": 0.4521,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8512,
      "eval_loss": 0.49920570850372314,
      "eval_runtime": 6.3484,
      "eval_samples_per_second": 590.701,
      "eval_steps_per_second": 9.294,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 7.884013652801514,
      "learning_rate": 1.763568653571882e-05,
      "loss": 0.3083,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8709333333333333,
      "eval_loss": 0.4250776767730713,
      "eval_runtime": 6.8333,
      "eval_samples_per_second": 548.784,
      "eval_steps_per_second": 8.634,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.678415775299072,
      "learning_rate": 1.2426252654184347e-05,
      "loss": 0.2142,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8773333333333333,
      "eval_loss": 0.4003533720970154,
      "eval_runtime": 6.6845,
      "eval_samples_per_second": 560.997,
      "eval_steps_per_second": 8.826,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.858291149139404,
      "learning_rate": 7.5272506870206094e-06,
      "loss": 0.1547,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8837333333333334,
      "eval_loss": 0.3824154734611511,
      "eval_runtime": 6.3699,
      "eval_samples_per_second": 588.704,
      "eval_steps_per_second": 9.262,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 36.76485824584961,
      "learning_rate": 3.5295725730362483e-06,
      "loss": 0.1142,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8864,
      "eval_loss": 0.36872392892837524,
      "eval_runtime": 6.35,
      "eval_samples_per_second": 590.548,
      "eval_steps_per_second": 9.291,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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
}