trainer_state.json

{
  "best_metric": 0.8744,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0962/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": 0.0001,
      "loss": 1.5541,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8162666666666667,
      "eval_loss": 0.6354945302009583,
      "eval_runtime": 10.3658,
      "eval_samples_per_second": 361.765,
      "eval_steps_per_second": 5.692,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 3.364107608795166,
      "learning_rate": 0.0001,
      "loss": 0.4758,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8466666666666667,
      "eval_loss": 0.5057896375656128,
      "eval_runtime": 10.582,
      "eval_samples_per_second": 354.375,
      "eval_steps_per_second": 5.576,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.8339237570762634,
      "learning_rate": 0.0001,
      "loss": 0.2881,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.46251657605171204,
      "eval_runtime": 10.6105,
      "eval_samples_per_second": 353.423,
      "eval_steps_per_second": 5.561,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.2411420345306396,
      "learning_rate": 0.0001,
      "loss": 0.1926,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8661333333333333,
      "eval_loss": 0.461021363735199,
      "eval_runtime": 10.7983,
      "eval_samples_per_second": 347.278,
      "eval_steps_per_second": 5.464,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 0.1344,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8712,
      "eval_loss": 0.4466906189918518,
      "eval_runtime": 11.6381,
      "eval_samples_per_second": 322.218,
      "eval_steps_per_second": 5.07,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.7042985558509827,
      "learning_rate": 0.0001,
      "loss": 0.1074,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8730666666666667,
      "eval_loss": 0.4782746732234955,
      "eval_runtime": 12.0765,
      "eval_samples_per_second": 310.521,
      "eval_steps_per_second": 4.886,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.03806076943874359,
      "learning_rate": 0.0001,
      "loss": 0.0873,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8744,
      "eval_loss": 0.4938906133174896,
      "eval_runtime": 10.5631,
      "eval_samples_per_second": 355.009,
      "eval_steps_per_second": 5.585,
      "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
}