trainer_state.json

{
  "best_metric": 0.9576,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0844/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": 1.1076105833053589,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.5862,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9296,
      "eval_loss": 0.25066375732421875,
      "eval_runtime": 11.2201,
      "eval_samples_per_second": 334.223,
      "eval_steps_per_second": 5.258,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.04296064376831055,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.1184,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9458666666666666,
      "eval_loss": 0.20790015161037445,
      "eval_runtime": 11.4384,
      "eval_samples_per_second": 327.843,
      "eval_steps_per_second": 5.158,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.036052290350198746,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.054,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9464,
      "eval_loss": 0.21512532234191895,
      "eval_runtime": 11.0079,
      "eval_samples_per_second": 340.664,
      "eval_steps_per_second": 5.36,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.01968609169125557,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 0.0242,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9525333333333333,
      "eval_loss": 0.18875540792942047,
      "eval_runtime": 11.841,
      "eval_samples_per_second": 316.696,
      "eval_steps_per_second": 4.983,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.04538640379905701,
      "learning_rate": 2.857142857142857e-05,
      "loss": 0.012,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9525333333333333,
      "eval_loss": 0.19517841935157776,
      "eval_runtime": 11.3115,
      "eval_samples_per_second": 331.52,
      "eval_steps_per_second": 5.216,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.00814194604754448,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.0044,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9538666666666666,
      "eval_loss": 0.17996105551719666,
      "eval_runtime": 11.5775,
      "eval_samples_per_second": 323.903,
      "eval_steps_per_second": 5.096,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 4.2900042900042903e-08,
      "loss": 0.002,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9576,
      "eval_loss": 0.17864052951335907,
      "eval_runtime": 11.2349,
      "eval_samples_per_second": 333.781,
      "eval_steps_per_second": 5.251,
      "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
}