trainer_state.json

{
  "best_metric": 0.9298666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0744/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": Infinity,
      "learning_rate": 2.574002574002574e-05,
      "loss": 0.7583,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8885333333333333,
      "eval_loss": 0.35074418783187866,
      "eval_runtime": 12.7557,
      "eval_samples_per_second": 293.987,
      "eval_steps_per_second": 4.625,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 2.146718146718147e-05,
      "loss": 0.17,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9002666666666667,
      "eval_loss": 0.3317021429538727,
      "eval_runtime": 12.5927,
      "eval_samples_per_second": 297.79,
      "eval_steps_per_second": 4.685,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.13327567279338837,
      "learning_rate": 1.718146718146718e-05,
      "loss": 0.0638,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9104,
      "eval_loss": 0.30700618028640747,
      "eval_runtime": 12.3777,
      "eval_samples_per_second": 302.964,
      "eval_steps_per_second": 4.767,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.2560471296310425,
      "learning_rate": 1.2895752895752896e-05,
      "loss": 0.0242,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9170666666666667,
      "eval_loss": 0.31133726239204407,
      "eval_runtime": 12.2997,
      "eval_samples_per_second": 304.887,
      "eval_steps_per_second": 4.797,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.0822438895702362,
      "learning_rate": 8.61003861003861e-06,
      "loss": 0.006,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.924,
      "eval_loss": 0.31292369961738586,
      "eval_runtime": 12.5198,
      "eval_samples_per_second": 299.526,
      "eval_steps_per_second": 4.713,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.007162551395595074,
      "learning_rate": 4.324324324324325e-06,
      "loss": 0.0012,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9274666666666667,
      "eval_loss": 0.2992449998855591,
      "eval_runtime": 12.3741,
      "eval_samples_per_second": 303.053,
      "eval_steps_per_second": 4.768,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.018180157989263535,
      "learning_rate": 3.861003861003861e-08,
      "loss": 0.0003,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9298666666666666,
      "eval_loss": 0.30145376920700073,
      "eval_runtime": 11.5876,
      "eval_samples_per_second": 323.62,
      "eval_steps_per_second": 5.092,
      "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
}