trainer_state.json

{
  "best_metric": 0.3469333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0773/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": 16.615421295166016,
      "learning_rate": 0.0005,
      "loss": 3.396,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20053333333333334,
      "eval_loss": 3.00124192237854,
      "eval_runtime": 12.0768,
      "eval_samples_per_second": 310.514,
      "eval_steps_per_second": 4.885,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 17.627836227416992,
      "learning_rate": 0.0005,
      "loss": 2.9055,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.22506666666666666,
      "eval_loss": 2.909101724624634,
      "eval_runtime": 12.0998,
      "eval_samples_per_second": 309.923,
      "eval_steps_per_second": 4.876,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.707172393798828,
      "learning_rate": 0.0005,
      "loss": 2.7375,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2424,
      "eval_loss": 2.7986056804656982,
      "eval_runtime": 12.0419,
      "eval_samples_per_second": 311.412,
      "eval_steps_per_second": 4.9,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 18.013710021972656,
      "learning_rate": 0.0005,
      "loss": 2.6059,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.28426666666666667,
      "eval_loss": 2.60630464553833,
      "eval_runtime": 12.0348,
      "eval_samples_per_second": 311.597,
      "eval_steps_per_second": 4.902,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 11.904728889465332,
      "learning_rate": 0.0005,
      "loss": 2.4776,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.304,
      "eval_loss": 2.520995855331421,
      "eval_runtime": 12.2807,
      "eval_samples_per_second": 305.358,
      "eval_steps_per_second": 4.804,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 22.757816314697266,
      "learning_rate": 0.0005,
      "loss": 2.3883,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.33786666666666665,
      "eval_loss": 2.4251351356506348,
      "eval_runtime": 12.0616,
      "eval_samples_per_second": 310.903,
      "eval_steps_per_second": 4.892,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.2942,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3469333333333333,
      "eval_loss": 2.3714711666107178,
      "eval_runtime": 12.0708,
      "eval_samples_per_second": 310.666,
      "eval_steps_per_second": 4.888,
      "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
}