trainer_state.json

{
  "best_metric": 0.9498666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0305/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": 2.9523117542266846,
      "learning_rate": 8.728616793536588e-05,
      "loss": 0.6312,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9232,
      "eval_loss": 0.2750612199306488,
      "eval_runtime": 7.4788,
      "eval_samples_per_second": 501.416,
      "eval_steps_per_second": 7.889,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.05039806663990021,
      "learning_rate": 7.947199994035401e-05,
      "loss": 0.1401,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9365333333333333,
      "eval_loss": 0.23802122473716736,
      "eval_runtime": 6.7868,
      "eval_samples_per_second": 552.546,
      "eval_steps_per_second": 8.693,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 6.118083477020264,
      "learning_rate": 6.75e-05,
      "loss": 0.0652,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.2642131745815277,
      "eval_runtime": 6.7839,
      "eval_samples_per_second": 552.779,
      "eval_steps_per_second": 8.697,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2532743513584137,
      "learning_rate": 5.2814167995011876e-05,
      "loss": 0.0269,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9370666666666667,
      "eval_loss": 0.25039756298065186,
      "eval_runtime": 6.7973,
      "eval_samples_per_second": 551.691,
      "eval_steps_per_second": 8.68,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.01434257347136736,
      "learning_rate": 3.7185832004988136e-05,
      "loss": 0.0107,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.944,
      "eval_loss": 0.25416091084480286,
      "eval_runtime": 6.7115,
      "eval_samples_per_second": 558.744,
      "eval_steps_per_second": 8.791,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.02742774598300457,
      "learning_rate": 2.250000000000001e-05,
      "loss": 0.0037,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9445333333333333,
      "eval_loss": 0.2494526356458664,
      "eval_runtime": 7.5354,
      "eval_samples_per_second": 497.648,
      "eval_steps_per_second": 7.83,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 1.055833996679956e-05,
      "loss": 0.0016,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9498666666666666,
      "eval_loss": 0.23527829349040985,
      "eval_runtime": 6.6732,
      "eval_samples_per_second": 561.953,
      "eval_steps_per_second": 8.841,
      "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
}