trainer_state.json

{
  "best_metric": 0.8797333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0521/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": 8.575718575718577e-05,
      "loss": 1.6354,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7826666666666666,
      "eval_loss": 0.7557437419891357,
      "eval_runtime": 5.8384,
      "eval_samples_per_second": 642.299,
      "eval_steps_per_second": 10.105,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 7.151437151437152e-05,
      "loss": 0.5337,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8189333333333333,
      "eval_loss": 0.6063522696495056,
      "eval_runtime": 7.0131,
      "eval_samples_per_second": 534.712,
      "eval_steps_per_second": 8.413,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.45273053646087646,
      "learning_rate": 5.7228657228657234e-05,
      "loss": 0.2664,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8373333333333334,
      "eval_loss": 0.5473006367683411,
      "eval_runtime": 5.9181,
      "eval_samples_per_second": 633.645,
      "eval_steps_per_second": 9.969,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.24570421874523163,
      "learning_rate": 4.294294294294294e-05,
      "loss": 0.1234,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8594666666666667,
      "eval_loss": 0.48796311020851135,
      "eval_runtime": 5.9228,
      "eval_samples_per_second": 633.143,
      "eval_steps_per_second": 9.961,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.2265641987323761,
      "learning_rate": 2.8657228657228656e-05,
      "loss": 0.0483,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8693333333333333,
      "eval_loss": 0.5005225539207458,
      "eval_runtime": 5.5127,
      "eval_samples_per_second": 680.247,
      "eval_steps_per_second": 10.703,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.05506483092904091,
      "learning_rate": 1.4371514371514371e-05,
      "loss": 0.0147,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8746666666666667,
      "eval_loss": 0.4959900975227356,
      "eval_runtime": 5.5462,
      "eval_samples_per_second": 676.141,
      "eval_steps_per_second": 10.638,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.15460924804210663,
      "learning_rate": 8.580008580008581e-08,
      "loss": 0.0062,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8797333333333334,
      "eval_loss": 0.4862971901893616,
      "eval_runtime": 5.5682,
      "eval_samples_per_second": 673.462,
      "eval_steps_per_second": 10.596,
      "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
}