trainer_state.json

{
  "best_metric": 0.5776,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0795/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": 0.0004287859287859288,
      "loss": 3.1134,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2616,
      "eval_loss": 2.6914913654327393,
      "eval_runtime": 6.2995,
      "eval_samples_per_second": 595.285,
      "eval_steps_per_second": 9.366,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00035757185757185754,
      "loss": 2.4664,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.36533333333333334,
      "eval_loss": 2.2695083618164062,
      "eval_runtime": 6.4589,
      "eval_samples_per_second": 580.591,
      "eval_steps_per_second": 9.135,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0002863577863577864,
      "loss": 2.0267,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4381333333333333,
      "eval_loss": 1.9553192853927612,
      "eval_runtime": 8.6003,
      "eval_samples_per_second": 436.033,
      "eval_steps_per_second": 6.86,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.542957305908203,
      "learning_rate": 0.00021492921492921493,
      "loss": 1.6662,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5,
      "eval_loss": 1.7570838928222656,
      "eval_runtime": 6.3157,
      "eval_samples_per_second": 593.756,
      "eval_steps_per_second": 9.342,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 11.173380851745605,
      "learning_rate": 0.00014350064350064352,
      "loss": 1.2754,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5336,
      "eval_loss": 1.6374354362487793,
      "eval_runtime": 6.3429,
      "eval_samples_per_second": 591.211,
      "eval_steps_per_second": 9.302,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 16.842491149902344,
      "learning_rate": 7.207207207207208e-05,
      "loss": 0.8598,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.56,
      "eval_loss": 1.5447783470153809,
      "eval_runtime": 6.3239,
      "eval_samples_per_second": 592.986,
      "eval_steps_per_second": 9.33,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 8.580008580008581e-07,
      "loss": 0.4789,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5776,
      "eval_loss": 1.5153086185455322,
      "eval_runtime": 6.3415,
      "eval_samples_per_second": 591.346,
      "eval_steps_per_second": 9.304,
      "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
}