multilingual-Davlan/afro-xlmr-base-kanuri-ner-v1
This model is a fine-tuned version of Davlan/afro-xlmr-base on the Beijuka/Multilingual_PII_NER_dataset dataset. It achieves the following results on the evaluation set:
- Loss: 0.0769
- Precision: 0.9328
- Recall: 0.9530
- F1: 0.9428
- Accuracy: 0.9857
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 | 0.1158 | 0.8646 | 0.8610 | 0.8628 | 0.9683 |
| 0.2058 | 2.0 | 602 | 0.0876 | 0.8848 | 0.9431 | 0.9130 | 0.9751 |
| 0.2058 | 3.0 | 903 | 0.0854 | 0.9078 | 0.9143 | 0.9110 | 0.9783 |
| 0.0658 | 4.0 | 1204 | 0.1092 | 0.8847 | 0.9383 | 0.9107 | 0.9755 |
| 0.0491 | 5.0 | 1505 | 0.0881 | 0.9046 | 0.9431 | 0.9234 | 0.9782 |
| 0.0491 | 6.0 | 1806 | 0.1227 | 0.9015 | 0.9323 | 0.9166 | 0.9770 |
| 0.0298 | 7.0 | 2107 | 0.1005 | 0.9218 | 0.9461 | 0.9338 | 0.9805 |
| 0.0298 | 8.0 | 2408 | 0.1454 | 0.8970 | 0.9395 | 0.9178 | 0.9774 |
| 0.0164 | 9.0 | 2709 | 0.1301 | 0.9146 | 0.9305 | 0.9225 | 0.9789 |
| 0.0089 | 10.0 | 3010 | 0.1297 | 0.9215 | 0.9425 | 0.9319 | 0.9806 |
Framework versions
- Transformers 4.55.4
- Pytorch 2.8.0+cu126
- Datasets 4.0.0
- Tokenizers 0.21.4