trainer_state.json

{
  "best_metric": 0.904,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0610/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": 44.422279357910156,
      "learning_rate": 9.510675611750168e-05,
      "loss": 1.0669,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7981333333333334,
      "eval_loss": 0.6693850159645081,
      "eval_runtime": 11.435,
      "eval_samples_per_second": 327.94,
      "eval_steps_per_second": 5.16,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 8.13322912197567e-05,
      "loss": 0.3975,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8306666666666667,
      "eval_loss": 0.5525963306427002,
      "eval_runtime": 11.3071,
      "eval_samples_per_second": 331.649,
      "eval_steps_per_second": 5.218,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 40.11861038208008,
      "learning_rate": 6.13230482918689e-05,
      "loss": 0.1968,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8485333333333334,
      "eval_loss": 0.5176916718482971,
      "eval_runtime": 11.6669,
      "eval_samples_per_second": 321.423,
      "eval_steps_per_second": 5.057,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 31.281810760498047,
      "learning_rate": 3.90711367817024e-05,
      "loss": 0.0991,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8672,
      "eval_loss": 0.4857901334762573,
      "eval_runtime": 11.2961,
      "eval_samples_per_second": 331.974,
      "eval_steps_per_second": 5.223,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.04378271847963333,
      "learning_rate": 1.8983820665631146e-05,
      "loss": 0.043,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8890666666666667,
      "eval_loss": 0.4250566363334656,
      "eval_runtime": 10.8081,
      "eval_samples_per_second": 346.963,
      "eval_steps_per_second": 5.459,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0015784050337970257,
      "learning_rate": 5.0396392552081564e-06,
      "loss": 0.0128,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9013333333333333,
      "eval_loss": 0.37566953897476196,
      "eval_runtime": 10.9937,
      "eval_samples_per_second": 341.106,
      "eval_steps_per_second": 5.367,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.039483532309532166,
      "learning_rate": 4.0869292992629095e-10,
      "loss": 0.0035,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.37748754024505615,
      "eval_runtime": 11.1435,
      "eval_samples_per_second": 336.52,
      "eval_steps_per_second": 5.295,
      "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
}