trainer_state.json

{
  "best_metric": 0.9418666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0622/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": 27.891931533813477,
      "learning_rate": 6.127627627627627e-05,
      "loss": 0.6587,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8392,
      "eval_loss": 0.5312330722808838,
      "eval_runtime": 12.273,
      "eval_samples_per_second": 305.548,
      "eval_steps_per_second": 4.807,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.2303667068481445,
      "learning_rate": 5.252627627627627e-05,
      "loss": 0.1913,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8954666666666666,
      "eval_loss": 0.36240851879119873,
      "eval_runtime": 12.741,
      "eval_samples_per_second": 294.325,
      "eval_steps_per_second": 4.631,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 4.3802552552552546e-05,
      "loss": 0.0925,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9050666666666667,
      "eval_loss": 0.3750680387020111,
      "eval_runtime": 12.2526,
      "eval_samples_per_second": 306.059,
      "eval_steps_per_second": 4.815,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.0023060296662151814,
      "learning_rate": 3.5052552552552544e-05,
      "loss": 0.0511,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.916,
      "eval_loss": 0.3293105363845825,
      "eval_runtime": 11.9813,
      "eval_samples_per_second": 312.988,
      "eval_steps_per_second": 4.924,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.053233809769153595,
      "learning_rate": 2.630255255255255e-05,
      "loss": 0.0241,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9245333333333333,
      "eval_loss": 0.3067297339439392,
      "eval_runtime": 12.9907,
      "eval_samples_per_second": 288.668,
      "eval_steps_per_second": 4.542,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0063900575041770935,
      "learning_rate": 1.7552552552552552e-05,
      "loss": 0.0113,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9330666666666667,
      "eval_loss": 0.26043900847435,
      "eval_runtime": 12.3188,
      "eval_samples_per_second": 304.412,
      "eval_steps_per_second": 4.789,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.04197390004992485,
      "learning_rate": 8.802552552552553e-06,
      "loss": 0.0034,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9418666666666666,
      "eval_loss": 0.25523048639297485,
      "eval_runtime": 12.2167,
      "eval_samples_per_second": 306.958,
      "eval_steps_per_second": 4.829,
      "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
}