trainer_state.json

{
  "best_metric": 0.9165333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0019/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": 0.7897019982337952,
      "learning_rate": 0.00042857142857142855,
      "loss": 0.8052,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8184,
      "eval_loss": 0.6107335090637207,
      "eval_runtime": 12.5165,
      "eval_samples_per_second": 299.605,
      "eval_steps_per_second": 4.714,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.8987799286842346,
      "learning_rate": 0.00035714285714285714,
      "loss": 0.411,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8413333333333334,
      "eval_loss": 0.5304506421089172,
      "eval_runtime": 12.4614,
      "eval_samples_per_second": 300.928,
      "eval_steps_per_second": 4.735,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05725838616490364,
      "learning_rate": 0.0002857142857142857,
      "loss": 0.2472,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8578666666666667,
      "eval_loss": 0.5002932548522949,
      "eval_runtime": 12.7261,
      "eval_samples_per_second": 294.67,
      "eval_steps_per_second": 4.636,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0733546018600464,
      "learning_rate": 0.00021428571428571427,
      "loss": 0.1433,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8621333333333333,
      "eval_loss": 0.5239591598510742,
      "eval_runtime": 12.3106,
      "eval_samples_per_second": 304.615,
      "eval_steps_per_second": 4.793,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.7391700744628906,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.0748,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8912,
      "eval_loss": 0.44313451647758484,
      "eval_runtime": 11.9222,
      "eval_samples_per_second": 314.538,
      "eval_steps_per_second": 4.949,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.011820327490568161,
      "learning_rate": 7.142857142857142e-05,
      "loss": 0.029,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9048,
      "eval_loss": 0.39284420013427734,
      "eval_runtime": 12.0216,
      "eval_samples_per_second": 311.938,
      "eval_steps_per_second": 4.908,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 2.1450021450021452e-07,
      "loss": 0.0094,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9165333333333333,
      "eval_loss": 0.36500659584999084,
      "eval_runtime": 12.5339,
      "eval_samples_per_second": 299.188,
      "eval_steps_per_second": 4.707,
      "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
}