trainer_state.json

{
  "best_metric": 0.8968,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0038/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": 8.584298584298584e-05,
      "loss": 0.9161,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7885333333333333,
      "eval_loss": 0.7314623594284058,
      "eval_runtime": 10.7113,
      "eval_samples_per_second": 350.096,
      "eval_steps_per_second": 5.508,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.5318008065223694,
      "learning_rate": 7.155727155727156e-05,
      "loss": 0.3064,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8328,
      "eval_loss": 0.5765891671180725,
      "eval_runtime": 10.912,
      "eval_samples_per_second": 343.659,
      "eval_steps_per_second": 5.407,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.46010622382164,
      "learning_rate": 5.727155727155728e-05,
      "loss": 0.1455,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.848,
      "eval_loss": 0.5568447709083557,
      "eval_runtime": 10.9773,
      "eval_samples_per_second": 341.615,
      "eval_steps_per_second": 5.375,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.01287461631000042,
      "learning_rate": 4.298584298584299e-05,
      "loss": 0.073,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8648,
      "eval_loss": 0.5423625111579895,
      "eval_runtime": 11.6345,
      "eval_samples_per_second": 322.319,
      "eval_steps_per_second": 5.071,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.030072396621108055,
      "learning_rate": 2.8700128700128703e-05,
      "loss": 0.0309,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8858666666666667,
      "eval_loss": 0.4776589870452881,
      "eval_runtime": 10.75,
      "eval_samples_per_second": 348.839,
      "eval_steps_per_second": 5.488,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.0012541046598926187,
      "learning_rate": 1.4414414414414416e-05,
      "loss": 0.0081,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8909333333333334,
      "eval_loss": 0.4535156786441803,
      "eval_runtime": 10.6518,
      "eval_samples_per_second": 352.054,
      "eval_steps_per_second": 5.539,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.0004397554730530828,
      "learning_rate": 1.287001287001287e-07,
      "loss": 0.0027,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8968,
      "eval_loss": 0.44848182797431946,
      "eval_runtime": 10.7165,
      "eval_samples_per_second": 349.927,
      "eval_steps_per_second": 5.506,
      "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
}