multilingual-roberta-base-hausa-ner-v1
This model is a fine-tuned version of roberta-base on the Beijuka/Multilingual_PII_NER_dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.0617
- Precision: 0.9420
- Recall: 0.9505
- F1: 0.9462
- Accuracy: 0.9879
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: 5e-05
- train_batch_size: 8
- eval_batch_size: 8
- 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: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 301 | 0.1093 | 0.8649 | 0.8711 | 0.8680 | 0.9689 |
| 0.1911 | 2.0 | 602 | 0.0992 | 0.8583 | 0.8817 | 0.8698 | 0.9736 |
| 0.1911 | 3.0 | 903 | 0.1074 | 0.8352 | 0.8644 | 0.8496 | 0.9735 |
| 0.0675 | 4.0 | 1204 | 0.0891 | 0.8734 | 0.9311 | 0.9013 | 0.9774 |
| 0.0459 | 5.0 | 1505 | 0.1135 | 0.8932 | 0.9339 | 0.9131 | 0.9747 |
| 0.0459 | 6.0 | 1806 | 0.0827 | 0.9133 | 0.9367 | 0.9248 | 0.9813 |
| 0.0322 | 7.0 | 2107 | 0.1116 | 0.8798 | 0.9311 | 0.9047 | 0.9777 |
| 0.0322 | 8.0 | 2408 | 0.1272 | 0.8842 | 0.9244 | 0.9039 | 0.9790 |
| 0.0223 | 9.0 | 2709 | 0.1036 | 0.8986 | 0.935 | 0.9164 | 0.9805 |
Framework versions
- Transformers 4.55.4
- Pytorch 2.8.0+cu126
- Datasets 4.0.0
- Tokenizers 0.21.4