ner_peoples_daily
This model is a fine-tuned version of hfl/rbt6 on the peoples_daily_ner dataset. It achieves the following results on the evaluation set:
- Loss: 0.0249
- Precision: 0.9205
- Recall: 0.9365
- F1: 0.9285
- Accuracy: 0.9930
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: 128
- eval_batch_size: 128
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 8
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.3154 | 1.0 | 164 | 0.0410 | 0.8258 | 0.8684 | 0.8466 | 0.9868 |
| 0.0394 | 2.0 | 328 | 0.0287 | 0.8842 | 0.9070 | 0.8954 | 0.9905 |
| 0.0293 | 3.0 | 492 | 0.0264 | 0.8978 | 0.9168 | 0.9072 | 0.9916 |
| 0.02 | 4.0 | 656 | 0.0254 | 0.9149 | 0.9226 | 0.9188 | 0.9923 |
| 0.016 | 5.0 | 820 | 0.0250 | 0.9167 | 0.9281 | 0.9224 | 0.9927 |
| 0.0124 | 6.0 | 984 | 0.0252 | 0.9114 | 0.9328 | 0.9220 | 0.9928 |
| 0.0108 | 7.0 | 1148 | 0.0249 | 0.9169 | 0.9339 | 0.9254 | 0.9928 |
| 0.0097 | 8.0 | 1312 | 0.0249 | 0.9205 | 0.9365 | 0.9285 | 0.9930 |
Framework versions
- Transformers 4.23.1
- Pytorch 1.12.1+cu113
- Datasets 2.5.2
- Tokenizers 0.13.1