dandankim's picture
dandankim/token_classifier2
9a5be97 verified
|
Raw History Blame Contribute Delete
2.24 kB
metadata
library_name: transformers
license: apache-2.0
base_model: distilbert-base-uncased
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: distilbert-ner-improved
    results: []

distilbert-ner-improved

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0471
  • Precision: 0.9588
  • Recall: 0.9758
  • F1: 0.9672
  • Accuracy: 0.9875

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 5e-05
  • train_batch_size: 32
  • eval_batch_size: 64
  • seed: 42
  • gradient_accumulation_steps: 2
  • total_train_batch_size: 64
  • optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
  • lr_scheduler_type: linear
  • num_epochs: 20
  • mixed_precision_training: Native AMP

Training results

Training Loss Epoch Step Validation Loss Precision Recall F1 Accuracy
1.0197 2.64 100 0.5256 0.4831 0.4319 0.4561 0.8719
0.1435 5.2667 200 0.1149 0.9453 0.9329 0.9390 0.9744
0.0655 7.9067 300 0.0673 0.9473 0.9688 0.9579 0.9809
0.0441 10.5333 400 0.0754 0.9708 0.9693 0.9700 0.9832
0.0236 13.16 500 0.0790 0.9697 0.9772 0.9734 0.9832
0.0171 15.8 600 0.0849 0.9670 0.9797 0.9733 0.9835

Framework versions

  • Transformers 4.48.3
  • Pytorch 2.5.1+cu124
  • Datasets 3.3.2
  • Tokenizers 0.21.0