trainer_state.json

{
  "best_metric": 0.9464,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0168/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": 0.6775330305099487,
      "learning_rate": 5e-05,
      "loss": 0.8072,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9314666666666667,
      "eval_loss": 0.2504427433013916,
      "eval_runtime": 6.5015,
      "eval_samples_per_second": 576.791,
      "eval_steps_per_second": 9.075,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.31892624497413635,
      "learning_rate": 5e-05,
      "loss": 0.1251,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9424,
      "eval_loss": 0.2139415740966797,
      "eval_runtime": 5.8713,
      "eval_samples_per_second": 638.697,
      "eval_steps_per_second": 10.049,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.365393728017807,
      "learning_rate": 5e-05,
      "loss": 0.0515,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9330666666666667,
      "eval_loss": 0.24756506085395813,
      "eval_runtime": 5.8368,
      "eval_samples_per_second": 642.48,
      "eval_steps_per_second": 10.108,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.030193284153938293,
      "learning_rate": 5e-05,
      "loss": 0.0229,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9392,
      "eval_loss": 0.25154179334640503,
      "eval_runtime": 5.8759,
      "eval_samples_per_second": 638.195,
      "eval_steps_per_second": 10.041,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.011447733268141747,
      "learning_rate": 5e-05,
      "loss": 0.0175,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9405333333333333,
      "eval_loss": 0.2537371814250946,
      "eval_runtime": 6.24,
      "eval_samples_per_second": 600.957,
      "eval_steps_per_second": 9.455,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.018717225641012192,
      "learning_rate": 5e-05,
      "loss": 0.0138,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9450666666666667,
      "eval_loss": 0.2442488670349121,
      "eval_runtime": 5.9569,
      "eval_samples_per_second": 629.517,
      "eval_steps_per_second": 9.904,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 5e-05,
      "loss": 0.013,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9464,
      "eval_loss": 0.2539132535457611,
      "eval_runtime": 5.9555,
      "eval_samples_per_second": 629.667,
      "eval_steps_per_second": 9.907,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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
}