trainer_state.json

{
  "best_metric": 0.336,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0183/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": 3.404,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.2184,
      "eval_loss": 2.9787416458129883,
      "eval_runtime": 6.2497,
      "eval_samples_per_second": 600.029,
      "eval_steps_per_second": 9.44,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.253387451171875,
      "learning_rate": 0.0005,
      "loss": 2.9227,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.23226666666666668,
      "eval_loss": 2.882350444793701,
      "eval_runtime": 6.193,
      "eval_samples_per_second": 605.524,
      "eval_steps_per_second": 9.527,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 21.656845092773438,
      "learning_rate": 0.0005,
      "loss": 2.7779,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2272,
      "eval_loss": 2.853177070617676,
      "eval_runtime": 6.2204,
      "eval_samples_per_second": 602.858,
      "eval_steps_per_second": 9.485,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 12.97970199584961,
      "learning_rate": 0.0005,
      "loss": 2.6405,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2786666666666667,
      "eval_loss": 2.7201051712036133,
      "eval_runtime": 6.2548,
      "eval_samples_per_second": 599.536,
      "eval_steps_per_second": 9.433,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 19.779897689819336,
      "learning_rate": 0.0005,
      "loss": 2.5395,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3056,
      "eval_loss": 2.5466363430023193,
      "eval_runtime": 6.1491,
      "eval_samples_per_second": 609.841,
      "eval_steps_per_second": 9.595,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 19.070899963378906,
      "learning_rate": 0.0005,
      "loss": 2.4269,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3253333333333333,
      "eval_loss": 2.4592041969299316,
      "eval_runtime": 6.1978,
      "eval_samples_per_second": 605.052,
      "eval_steps_per_second": 9.519,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 17.459861755371094,
      "learning_rate": 0.0005,
      "loss": 2.3641,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.336,
      "eval_loss": 2.420427083969116,
      "eval_runtime": 6.3034,
      "eval_samples_per_second": 594.916,
      "eval_steps_per_second": 9.36,
      "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
}