bge-small-en-v1.5-conll2003-ner
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: 0.1579
- Precision: 0.8891
- Recall: 0.9211
- F1: 0.9048
- Accuracy: 0.9809
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: 64
- eval_batch_size: 128
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 15
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.2045 | 1.0 | 157 | 0.1888 | 0.8739 | 0.8992 | 0.8864 | 0.9784 |
| 0.1609 | 2.0 | 314 | 0.1750 | 0.8731 | 0.9041 | 0.8883 | 0.9789 |
| 0.1348 | 3.0 | 471 | 0.1671 | 0.8747 | 0.9073 | 0.8907 | 0.9790 |
| 0.1179 | 4.0 | 628 | 0.1636 | 0.8885 | 0.9128 | 0.9005 | 0.9800 |
| 0.0951 | 5.0 | 785 | 0.1591 | 0.8823 | 0.9180 | 0.8998 | 0.9805 |
| 0.0897 | 6.0 | 942 | 0.1563 | 0.8826 | 0.9219 | 0.9018 | 0.9801 |
| 0.0768 | 7.0 | 1099 | 0.1588 | 0.8884 | 0.9207 | 0.9043 | 0.9806 |
| 0.0662 | 8.0 | 1256 | 0.1577 | 0.8937 | 0.9211 | 0.9072 | 0.9813 |
| 0.0618 | 9.0 | 1413 | 0.1580 | 0.8913 | 0.9207 | 0.9058 | 0.9811 |
| 0.0637 | 10.0 | 1570 | 0.1558 | 0.8910 | 0.9219 | 0.9062 | 0.9810 |
| 0.0509 | 11.0 | 1727 | 0.1579 | 0.8891 | 0.9211 | 0.9048 | 0.9809 |
Framework versions
- Transformers 5.16.1
- Pytorch 2.11.0+cu128
- Datasets 4.8.5
- Tokenizers 0.23.1