trainer_state.json

{
  "best_metric": 0.9136,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0497/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": 7.190426349639893,
      "learning_rate": 3e-05,
      "loss": 0.7736,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8866666666666667,
      "eval_loss": 0.3949911594390869,
      "eval_runtime": 7.733,
      "eval_samples_per_second": 484.936,
      "eval_steps_per_second": 7.63,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.8067626953125,
      "learning_rate": 3e-05,
      "loss": 0.1934,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9056,
      "eval_loss": 0.3246007561683655,
      "eval_runtime": 7.9388,
      "eval_samples_per_second": 472.361,
      "eval_steps_per_second": 7.432,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0377354621887207,
      "learning_rate": 3e-05,
      "loss": 0.0951,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9,
      "eval_loss": 0.37464213371276855,
      "eval_runtime": 8.0772,
      "eval_samples_per_second": 464.27,
      "eval_steps_per_second": 7.305,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 24.929906845092773,
      "learning_rate": 3e-05,
      "loss": 0.0628,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9061333333333333,
      "eval_loss": 0.366137832403183,
      "eval_runtime": 7.6993,
      "eval_samples_per_second": 487.054,
      "eval_steps_per_second": 7.663,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.3449089527130127,
      "learning_rate": 3e-05,
      "loss": 0.043,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9117333333333333,
      "eval_loss": 0.3587650954723358,
      "eval_runtime": 7.8163,
      "eval_samples_per_second": 479.764,
      "eval_steps_per_second": 7.548,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 2.703911304473877,
      "learning_rate": 3e-05,
      "loss": 0.0367,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.904,
      "eval_loss": 0.4143882393836975,
      "eval_runtime": 7.8523,
      "eval_samples_per_second": 477.566,
      "eval_steps_per_second": 7.514,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0037026633508503437,
      "learning_rate": 3e-05,
      "loss": 0.0272,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9136,
      "eval_loss": 0.4156356751918793,
      "eval_runtime": 7.679,
      "eval_samples_per_second": 488.342,
      "eval_steps_per_second": 7.683,
      "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
}