trainer_state.json

{
  "best_metric": 0.43493333333333334,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0736/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": 24.237442016601562,
      "learning_rate": 0.0003,
      "loss": 3.2792,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.22106666666666666,
      "eval_loss": 2.9511377811431885,
      "eval_runtime": 13.6998,
      "eval_samples_per_second": 273.727,
      "eval_steps_per_second": 4.307,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 22.33233642578125,
      "learning_rate": 0.0003,
      "loss": 2.7097,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2976,
      "eval_loss": 2.5777394771575928,
      "eval_runtime": 13.3491,
      "eval_samples_per_second": 280.918,
      "eval_steps_per_second": 4.42,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 26.107521057128906,
      "learning_rate": 0.0003,
      "loss": 2.4778,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3248,
      "eval_loss": 2.5079078674316406,
      "eval_runtime": 13.4603,
      "eval_samples_per_second": 278.597,
      "eval_steps_per_second": 4.383,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 20.804536819458008,
      "learning_rate": 0.0003,
      "loss": 2.2858,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3549333333333333,
      "eval_loss": 2.362882137298584,
      "eval_runtime": 13.5921,
      "eval_samples_per_second": 275.895,
      "eval_steps_per_second": 4.341,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 29.062177658081055,
      "learning_rate": 0.0003,
      "loss": 2.1515,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3776,
      "eval_loss": 2.250504970550537,
      "eval_runtime": 13.3667,
      "eval_samples_per_second": 280.549,
      "eval_steps_per_second": 4.414,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 18.384132385253906,
      "learning_rate": 0.0003,
      "loss": 2.0008,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4008,
      "eval_loss": 2.1929521560668945,
      "eval_runtime": 13.4232,
      "eval_samples_per_second": 279.367,
      "eval_steps_per_second": 4.395,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0003,
      "loss": 1.9004,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.43493333333333334,
      "eval_loss": 2.021519422531128,
      "eval_runtime": 13.6819,
      "eval_samples_per_second": 274.085,
      "eval_steps_per_second": 4.312,
      "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
}