trainer_state.json

{
  "best_metric": 0.9341333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0171/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": 0.7785678505897522,
      "learning_rate": 8.657457896300791e-05,
      "loss": 0.6609,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9074666666666666,
      "eval_loss": 0.31509101390838623,
      "eval_runtime": 6.9458,
      "eval_samples_per_second": 539.894,
      "eval_steps_per_second": 8.494,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 4.599603176116943,
      "learning_rate": 7.681980515339464e-05,
      "loss": 0.1643,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9138666666666667,
      "eval_loss": 0.30367863178253174,
      "eval_runtime": 6.5971,
      "eval_samples_per_second": 568.433,
      "eval_steps_per_second": 8.943,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.05541961267590523,
      "learning_rate": 6.222075445642905e-05,
      "loss": 0.0861,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9261333333333334,
      "eval_loss": 0.28062108159065247,
      "eval_runtime": 6.5083,
      "eval_samples_per_second": 576.187,
      "eval_steps_per_second": 9.065,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.025873763486742973,
      "learning_rate": 4.5e-05,
      "loss": 0.0438,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9274666666666667,
      "eval_loss": 0.2748910188674927,
      "eval_runtime": 6.6528,
      "eval_samples_per_second": 563.668,
      "eval_steps_per_second": 8.868,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.03435086831450462,
      "learning_rate": 2.7779245543570963e-05,
      "loss": 0.0167,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9336,
      "eval_loss": 0.2861184775829315,
      "eval_runtime": 6.2699,
      "eval_samples_per_second": 598.1,
      "eval_steps_per_second": 9.41,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.013767799362540245,
      "learning_rate": 1.3180194846605365e-05,
      "loss": 0.0077,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9325333333333333,
      "eval_loss": 0.2910551428794861,
      "eval_runtime": 6.7361,
      "eval_samples_per_second": 556.706,
      "eval_steps_per_second": 8.759,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 3.4457579724370674e-06,
      "loss": 0.0037,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9341333333333334,
      "eval_loss": 0.27261871099472046,
      "eval_runtime": 7.1186,
      "eval_samples_per_second": 526.787,
      "eval_steps_per_second": 8.288,
      "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
}