<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->
distilbert-base-uncased-finetuned-ner
This model is a fine-tuned version of distilbert-base-uncased on the conll2003 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0602
- Precision: 0.9274
- Recall: 0.9370
- F1: 0.9322
- Accuracy: 0.9839
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
0.2431 | 1.0 | 878 | 0.0690 | 0.9174 | 0.9214 | 0.9194 | 0.9811 |
0.0525 | 2.0 | 1756 | 0.0606 | 0.9251 | 0.9348 | 0.9299 | 0.9830 |
0.0299 | 3.0 | 2634 | 0.0602 | 0.9274 | 0.9370 | 0.9322 | 0.9839 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.7.1
- Tokenizers 0.13.2