trainer_state.json

{
  "best_metric": 0.42746666666666666,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0599/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": 20.491596221923828,
      "learning_rate": 0.00028540736881871185,
      "loss": 3.4055,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.20373333333333332,
      "eval_loss": 3.0646307468414307,
      "eval_runtime": 5.9004,
      "eval_samples_per_second": 635.545,
      "eval_steps_per_second": 9.999,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 27.480602264404297,
      "learning_rate": 0.00024399687365927005,
      "loss": 2.8385,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.25466666666666665,
      "eval_loss": 2.7812557220458984,
      "eval_runtime": 5.9199,
      "eval_samples_per_second": 633.458,
      "eval_steps_per_second": 9.966,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 27.596248626708984,
      "learning_rate": 0.00018396914487560668,
      "loss": 2.5445,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.3144,
      "eval_loss": 2.5256974697113037,
      "eval_runtime": 5.7937,
      "eval_samples_per_second": 647.25,
      "eval_steps_per_second": 10.183,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 24.03362274169922,
      "learning_rate": 0.00011721341034510718,
      "loss": 2.2913,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3496,
      "eval_loss": 2.396533966064453,
      "eval_runtime": 5.7432,
      "eval_samples_per_second": 652.941,
      "eval_steps_per_second": 10.273,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 4.124096870422363,
      "learning_rate": 5.695146199689343e-05,
      "loss": 2.0183,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.40826666666666667,
      "eval_loss": 2.2186191082000732,
      "eval_runtime": 5.9469,
      "eval_samples_per_second": 630.582,
      "eval_steps_per_second": 9.921,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 40.79184341430664,
      "learning_rate": 1.5118917765624467e-05,
      "loss": 1.7695,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.42186666666666667,
      "eval_loss": 2.140331268310547,
      "eval_runtime": 5.9624,
      "eval_samples_per_second": 628.946,
      "eval_steps_per_second": 9.895,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": Infinity,
      "learning_rate": 2.1796933167061993e-09,
      "loss": 1.586,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.42746666666666666,
      "eval_loss": 2.1614794731140137,
      "eval_runtime": 5.7341,
      "eval_samples_per_second": 653.981,
      "eval_steps_per_second": 10.289,
      "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
}