bert-base-chinese-finetuned-ner_0220_J_ORIDATA_FULL_NOMOD
This model is a fine-tuned version of ckiplab/bert-base-chinese-ner on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0522
- Precision: 0.9728
- Recall: 0.9739
- F1: 0.9733
- Accuracy: 0.9954
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: 2
- eval_batch_size: 2
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 12
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.3616 | 1.0 | 705 | 0.0914 | 0.8789 | 0.9239 | 0.9008 | 0.9821 |
| 0.0643 | 2.0 | 1410 | 0.0602 | 0.9242 | 0.9420 | 0.9330 | 0.9912 |
| 0.0339 | 3.0 | 2115 | 0.0533 | 0.9385 | 0.9545 | 0.9465 | 0.9910 |
| 0.024 | 4.0 | 2820 | 0.0558 | 0.9595 | 0.9693 | 0.9644 | 0.9932 |
| 0.0145 | 5.0 | 3525 | 0.0584 | 0.9484 | 0.9614 | 0.9549 | 0.9921 |
| 0.007 | 6.0 | 4230 | 0.0535 | 0.9637 | 0.9648 | 0.9642 | 0.9940 |
| 0.0145 | 7.0 | 4935 | 0.0492 | 0.9573 | 0.9682 | 0.9627 | 0.9942 |
| 0.0091 | 8.0 | 5640 | 0.0486 | 0.9694 | 0.9716 | 0.9705 | 0.9957 |
| 0.0049 | 9.0 | 6345 | 0.0526 | 0.9727 | 0.9727 | 0.9727 | 0.9950 |
| 0.0033 | 10.0 | 7050 | 0.0515 | 0.9661 | 0.9727 | 0.9694 | 0.9949 |
| 0.0023 | 11.0 | 7755 | 0.0523 | 0.9661 | 0.9716 | 0.9688 | 0.9950 |
| 0.0019 | 12.0 | 8460 | 0.0522 | 0.9728 | 0.9739 | 0.9733 | 0.9954 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.13.0+cu117
- Datasets 2.8.0
- Tokenizers 0.12.1