Cbert_base_ws-finetuned-ner
This model is a fine-tuned version of ckiplab/bert-base-chinese-ws on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0582
- Precision: 0.9602
- Recall: 0.9633
- F1: 0.9617
- Accuracy: 0.9827
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: 18
- eval_batch_size: 18
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
0.0482 | 0.64 | 1000 | 0.0509 | 0.9601 | 0.9582 | 0.9592 | 0.9817 |
0.0364 | 1.28 | 2000 | 0.0521 | 0.9590 | 0.9615 | 0.9602 | 0.9820 |
0.0341 | 1.92 | 3000 | 0.0548 | 0.9546 | 0.9625 | 0.9585 | 0.9812 |
0.0264 | 2.56 | 4000 | 0.0550 | 0.9593 | 0.9623 | 0.9608 | 0.9822 |
0.0227 | 3.19 | 5000 | 0.0582 | 0.9602 | 0.9633 | 0.9617 | 0.9827 |
0.021 | 3.83 | 6000 | 0.0595 | 0.9581 | 0.9624 | 0.9603 | 0.9820 |
0.0162 | 4.47 | 7000 | 0.0686 | 0.9574 | 0.9626 | 0.9600 | 0.9819 |
0.0159 | 5.11 | 8000 | 0.0719 | 0.9596 | 0.9614 | 0.9605 | 0.9822 |
0.0144 | 5.75 | 9000 | 0.0732 | 0.9590 | 0.9620 | 0.9605 | 0.9822 |
0.0109 | 6.39 | 10000 | 0.0782 | 0.9599 | 0.9626 | 0.9612 | 0.9824 |
0.0122 | 7.03 | 11000 | 0.0803 | 0.9605 | 0.9620 | 0.9612 | 0.9825 |
0.0097 | 7.67 | 12000 | 0.0860 | 0.9591 | 0.9620 | 0.9605 | 0.9822 |
0.0087 | 8.31 | 13000 | 0.0877 | 0.9591 | 0.9616 | 0.9603 | 0.9821 |
0.0087 | 8.95 | 14000 | 0.0902 | 0.9585 | 0.9630 | 0.9607 | 0.9823 |
0.0078 | 9.58 | 15000 | 0.0929 | 0.9589 | 0.9621 | 0.9605 | 0.9821 |
Framework versions
- Transformers 4.13.0
- Pytorch 1.8.0+cu111
- Datasets 2.4.0
- Tokenizers 0.10.3
- Downloads last month
- 14
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.