mn-twhin-bert-named-entity
This model is a fine-tuned version of Twitter/twhin-bert-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1591
- Precision: 0.9068
- Recall: 0.9199
- F1: 0.9133
- Accuracy: 0.9728
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: 32
- 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 |
|---|---|---|---|---|---|---|---|
| 0.1901 | 1.0 | 477 | 0.1052 | 0.8528 | 0.8872 | 0.8697 | 0.9666 |
| 0.0853 | 2.0 | 954 | 0.1220 | 0.8731 | 0.8963 | 0.8845 | 0.9666 |
| 0.0577 | 3.0 | 1431 | 0.1109 | 0.8889 | 0.9082 | 0.8984 | 0.9696 |
| 0.0396 | 4.0 | 1908 | 0.1172 | 0.9006 | 0.9175 | 0.9090 | 0.9724 |
| 0.0287 | 5.0 | 2385 | 0.1314 | 0.9002 | 0.9169 | 0.9085 | 0.9720 |
| 0.0213 | 6.0 | 2862 | 0.1363 | 0.9051 | 0.9181 | 0.9116 | 0.9720 |
| 0.0158 | 7.0 | 3339 | 0.1437 | 0.9114 | 0.9221 | 0.9167 | 0.9732 |
| 0.011 | 8.0 | 3816 | 0.1517 | 0.9091 | 0.9202 | 0.9146 | 0.9726 |
| 0.0077 | 9.0 | 4293 | 0.1570 | 0.9070 | 0.9199 | 0.9134 | 0.9728 |
| 0.0059 | 10.0 | 4770 | 0.1591 | 0.9068 | 0.9199 | 0.9133 | 0.9728 |
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.0+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3