trainer_state.json

{
  "best_metric": 0.4352,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0597/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": 22.404211044311523,
      "learning_rate": 0.00028540736881871185,
      "loss": 3.3205,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20613333333333334,
      "eval_loss": 3.0170035362243652,
      "eval_runtime": 6.2458,
      "eval_samples_per_second": 600.404,
      "eval_steps_per_second": 9.446,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 26.09221839904785,
      "learning_rate": 0.00024399687365927005,
      "loss": 2.8176,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2594666666666667,
      "eval_loss": 2.796846389770508,
      "eval_runtime": 8.4258,
      "eval_samples_per_second": 445.059,
      "eval_steps_per_second": 7.002,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 21.484790802001953,
      "learning_rate": 0.00018396914487560668,
      "loss": 2.5458,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3154666666666667,
      "eval_loss": 2.529283285140991,
      "eval_runtime": 6.1626,
      "eval_samples_per_second": 608.51,
      "eval_steps_per_second": 9.574,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 22.958396911621094,
      "learning_rate": 0.00011721341034510718,
      "loss": 2.2822,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.35973333333333335,
      "eval_loss": 2.3308796882629395,
      "eval_runtime": 6.192,
      "eval_samples_per_second": 605.623,
      "eval_steps_per_second": 9.528,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 34.702022552490234,
      "learning_rate": 5.695146199689343e-05,
      "loss": 1.9977,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4013333333333333,
      "eval_loss": 2.166421890258789,
      "eval_runtime": 6.1633,
      "eval_samples_per_second": 608.445,
      "eval_steps_per_second": 9.573,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 30.285245895385742,
      "learning_rate": 1.5118917765624467e-05,
      "loss": 1.7127,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4242666666666667,
      "eval_loss": 2.09976863861084,
      "eval_runtime": 6.4583,
      "eval_samples_per_second": 580.647,
      "eval_steps_per_second": 9.136,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 2.1796933167061993e-09,
      "loss": 1.5058,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4352,
      "eval_loss": 2.1061275005340576,
      "eval_runtime": 6.3591,
      "eval_samples_per_second": 589.708,
      "eval_steps_per_second": 9.278,
      "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
}