trainer_state.json

{
  "best_metric": 0.3621333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0056/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": 0.0005,
      "loss": 3.3286,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20213333333333333,
      "eval_loss": 2.9457755088806152,
      "eval_runtime": 6.7412,
      "eval_samples_per_second": 556.278,
      "eval_steps_per_second": 8.752,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0005,
      "loss": 2.8533,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.23973333333333333,
      "eval_loss": 2.7672083377838135,
      "eval_runtime": 6.7428,
      "eval_samples_per_second": 556.15,
      "eval_steps_per_second": 8.75,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.53487205505371,
      "learning_rate": 0.0005,
      "loss": 2.7143,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.26106666666666667,
      "eval_loss": 2.8085172176361084,
      "eval_runtime": 6.7458,
      "eval_samples_per_second": 555.901,
      "eval_steps_per_second": 8.746,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 14.45648193359375,
      "learning_rate": 0.0005,
      "loss": 2.5869,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2936,
      "eval_loss": 2.5875461101531982,
      "eval_runtime": 7.2467,
      "eval_samples_per_second": 517.478,
      "eval_steps_per_second": 8.142,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 18.865734100341797,
      "learning_rate": 0.0005,
      "loss": 2.4578,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.31866666666666665,
      "eval_loss": 2.499234437942505,
      "eval_runtime": 6.7697,
      "eval_samples_per_second": 553.942,
      "eval_steps_per_second": 8.715,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 16.561426162719727,
      "learning_rate": 0.0005,
      "loss": 2.3587,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.34426666666666667,
      "eval_loss": 2.4263532161712646,
      "eval_runtime": 6.9387,
      "eval_samples_per_second": 540.45,
      "eval_steps_per_second": 8.503,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 21.63570785522461,
      "learning_rate": 0.0005,
      "loss": 2.2822,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3621333333333333,
      "eval_loss": 2.341996908187866,
      "eval_runtime": 6.7465,
      "eval_samples_per_second": 555.842,
      "eval_steps_per_second": 8.745,
      "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
}