trainer_state.json

{
  "best_metric": 0.9248,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0686/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": 35.76716613769531,
      "learning_rate": 8.559608050575152e-05,
      "loss": 0.7984,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8317333333333333,
      "eval_loss": 0.532991349697113,
      "eval_runtime": 12.2964,
      "eval_samples_per_second": 304.967,
      "eval_steps_per_second": 4.798,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 33.66744613647461,
      "learning_rate": 7.31517728663805e-05,
      "loss": 0.3109,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8664,
      "eval_loss": 0.44075509905815125,
      "eval_runtime": 12.2012,
      "eval_samples_per_second": 307.346,
      "eval_steps_per_second": 4.836,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.826061248779297,
      "learning_rate": 5.5131661351986275e-05,
      "loss": 0.1637,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8768,
      "eval_loss": 0.4209848642349243,
      "eval_runtime": 12.2213,
      "eval_samples_per_second": 306.841,
      "eval_steps_per_second": 4.828,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.1940951347351074,
      "learning_rate": 3.5104850050159014e-05,
      "loss": 0.0883,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9,
      "eval_loss": 0.373271644115448,
      "eval_runtime": 12.2563,
      "eval_samples_per_second": 305.966,
      "eval_steps_per_second": 4.814,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.07668835669755936,
      "learning_rate": 1.703789455194791e-05,
      "loss": 0.0365,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9178666666666667,
      "eval_loss": 0.3285280168056488,
      "eval_runtime": 12.2721,
      "eval_samples_per_second": 305.571,
      "eval_steps_per_second": 4.808,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.65816068649292,
      "learning_rate": 4.509176970441849e-06,
      "loss": 0.0102,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9218666666666666,
      "eval_loss": 0.31317636370658875,
      "eval_runtime": 12.2158,
      "eval_samples_per_second": 306.979,
      "eval_steps_per_second": 4.83,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.008597261272370815,
      "learning_rate": 1.6347729569543557e-10,
      "loss": 0.0024,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9248,
      "eval_loss": 0.2954985797405243,
      "eval_runtime": 12.4327,
      "eval_samples_per_second": 301.624,
      "eval_steps_per_second": 4.746,
      "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
}