trainer_state.json

{
  "best_metric": 0.42373333333333335,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0347/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": 0.0003,
      "loss": 3.2898,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20933333333333334,
      "eval_loss": 2.9450953006744385,
      "eval_runtime": 12.1508,
      "eval_samples_per_second": 308.62,
      "eval_steps_per_second": 4.856,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 23.65144920349121,
      "learning_rate": 0.0003,
      "loss": 2.7173,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.29013333333333335,
      "eval_loss": 2.674659013748169,
      "eval_runtime": 11.3638,
      "eval_samples_per_second": 329.995,
      "eval_steps_per_second": 5.192,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 26.240955352783203,
      "learning_rate": 0.0003,
      "loss": 2.4685,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3224,
      "eval_loss": 2.4599223136901855,
      "eval_runtime": 12.3556,
      "eval_samples_per_second": 303.507,
      "eval_steps_per_second": 4.775,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 17.14536476135254,
      "learning_rate": 0.0003,
      "loss": 2.2599,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3722666666666667,
      "eval_loss": 2.2675883769989014,
      "eval_runtime": 11.7049,
      "eval_samples_per_second": 320.379,
      "eval_steps_per_second": 5.041,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 27.51565933227539,
      "learning_rate": 0.0003,
      "loss": 2.1157,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4008,
      "eval_loss": 2.1735033988952637,
      "eval_runtime": 11.6038,
      "eval_samples_per_second": 323.169,
      "eval_steps_per_second": 5.085,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 23.350780487060547,
      "learning_rate": 0.0003,
      "loss": 1.9807,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.40613333333333335,
      "eval_loss": 2.125572919845581,
      "eval_runtime": 11.2883,
      "eval_samples_per_second": 332.203,
      "eval_steps_per_second": 5.227,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 18.109786987304688,
      "learning_rate": 0.0003,
      "loss": 1.864,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.42373333333333335,
      "eval_loss": 2.0726726055145264,
      "eval_runtime": 11.5163,
      "eval_samples_per_second": 325.624,
      "eval_steps_per_second": 5.123,
      "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
}