<!-- 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. -->
KR-Medium-NER-kluedata
This model is a fine-tuned version of snunlp/KR-Medium on the klue dataset. It achieves the following results on the evaluation set:
- Loss: 0.1809
- Precision: 0.7771
- Recall: 0.8131
- F1: 0.7947
- Accuracy: 0.9606
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: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 328
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.2442 | 1.0 | 165 | 0.2390 | 0.5542 | 0.6426 | 0.5952 | 0.9211 |
| 0.1419 | 2.0 | 330 | 0.1694 | 0.7017 | 0.7143 | 0.7080 | 0.9461 |
| 0.1067 | 3.0 | 495 | 0.1402 | 0.7220 | 0.7894 | 0.7542 | 0.9550 |
| 0.0682 | 4.0 | 660 | 0.1431 | 0.7468 | 0.7990 | 0.7721 | 0.9573 |
| 0.0544 | 5.0 | 825 | 0.1440 | 0.7679 | 0.8082 | 0.7876 | 0.9596 |
| 0.0408 | 6.0 | 990 | 0.1525 | 0.7706 | 0.8014 | 0.7857 | 0.9594 |
| 0.0292 | 7.0 | 1155 | 0.1608 | 0.7703 | 0.8148 | 0.7919 | 0.9603 |
| 0.0241 | 8.0 | 1320 | 0.1706 | 0.7767 | 0.8099 | 0.7930 | 0.9604 |
| 0.0206 | 9.0 | 1485 | 0.1769 | 0.7805 | 0.8174 | 0.7985 | 0.9611 |
| 0.0177 | 10.0 | 1650 | 0.1809 | 0.7771 | 0.8131 | 0.7947 | 0.9606 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1+cu116
- Datasets 2.9.0
- Tokenizers 0.13.2
