bge-small-ner-hw2
This model is a fine-tuned version of BAAI/bge-small-en-v1.5 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0860
- Precision: 0.9020
- Recall: 0.9276
- F1: 0.9146
- Accuracy: 0.9822
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1471 | 1.0 | 1252 | 0.1270 | 0.8208 | 0.8750 | 0.8470 | 0.9714 |
| 0.0896 | 2.0 | 2504 | 0.0899 | 0.8648 | 0.9096 | 0.8866 | 0.9787 |
| 0.0586 | 3.0 | 3756 | 0.0807 | 0.8894 | 0.9159 | 0.9024 | 0.9805 |
| 0.0411 | 4.0 | 5008 | 0.0838 | 0.8863 | 0.9260 | 0.9057 | 0.9802 |
| 0.0412 | 5.0 | 6260 | 0.0824 | 0.8909 | 0.9246 | 0.9074 | 0.9804 |
| 0.0309 | 6.0 | 7512 | 0.0832 | 0.9002 | 0.9260 | 0.9129 | 0.9818 |
| 0.0147 | 7.0 | 8764 | 0.0850 | 0.9090 | 0.9276 | 0.9182 | 0.9827 |
| 0.03 | 8.0 | 10016 | 0.0856 | 0.9089 | 0.9268 | 0.9178 | 0.9822 |
| 0.0185 | 9.0 | 11268 | 0.0860 | 0.9020 | 0.9276 | 0.9146 | 0.9822 |
Framework versions
- Transformers 4.50.0
- Pytorch 2.14.1+cu130
- Datasets 3.4.1
- Tokenizers 0.21.4