trainer_state.json

{
  "best_metric": 0.38693333333333335,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0698/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": 19.0509090423584,
      "learning_rate": 0.0004756789480311864,
      "loss": 3.469,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.17813333333333334,
      "eval_loss": 3.1020214557647705,
      "eval_runtime": 12.3708,
      "eval_samples_per_second": 303.134,
      "eval_steps_per_second": 4.769,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.00040692388948875507,
      "loss": 2.9373,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.20933333333333334,
      "eval_loss": 2.949162483215332,
      "eval_runtime": 12.3523,
      "eval_samples_per_second": 303.586,
      "eval_steps_per_second": 4.776,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 23.701595306396484,
      "learning_rate": 0.0003069433725708678,
      "loss": 2.7371,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2885333333333333,
      "eval_loss": 2.69112229347229,
      "eval_runtime": 12.4207,
      "eval_samples_per_second": 301.915,
      "eval_steps_per_second": 4.75,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 21.173187255859375,
      "learning_rate": 0.0001956845223506859,
      "loss": 2.5071,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3274666666666667,
      "eval_loss": 2.5015718936920166,
      "eval_runtime": 12.4135,
      "eval_samples_per_second": 302.091,
      "eval_steps_per_second": 4.753,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 25.098468780517578,
      "learning_rate": 9.518351861456881e-05,
      "loss": 2.2716,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3605333333333333,
      "eval_loss": 2.3386037349700928,
      "eval_runtime": 12.3326,
      "eval_samples_per_second": 304.071,
      "eval_steps_per_second": 4.784,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 29.510316848754883,
      "learning_rate": 2.5345817716378312e-05,
      "loss": 2.036,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3816,
      "eval_loss": 2.2853095531463623,
      "eval_runtime": 12.3344,
      "eval_samples_per_second": 304.028,
      "eval_steps_per_second": 4.783,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 29.24658203125,
      "learning_rate": 3.6328221945103325e-09,
      "loss": 1.8647,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.38693333333333335,
      "eval_loss": 2.2928144931793213,
      "eval_runtime": 12.4659,
      "eval_samples_per_second": 300.82,
      "eval_steps_per_second": 4.733,
      "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
}