multilingual-Davlan/afro-xlmr-base-lumasaba-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: 0.3151
- Precision: 0.9679
- Recall: 0.9442
- F1: 0.9559
- Accuracy: 0.9500
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 | 398 | 0.6183 | 0.8545 | 0.7956 | 0.8240 | 0.8070 |
| 1.2369 | 2.0 | 796 | 0.4397 | 0.8643 | 0.8974 | 0.8805 | 0.8829 |
| 0.434 | 3.0 | 1194 | 0.3404 | 0.9291 | 0.9131 | 0.9210 | 0.9181 |
| 0.2226 | 4.0 | 1592 | 0.3327 | 0.9384 | 0.9178 | 0.9279 | 0.9305 |
| 0.2226 | 5.0 | 1990 | 0.3813 | 0.9310 | 0.9303 | 0.9307 | 0.9344 |
| 0.1331 | 6.0 | 2388 | 0.3147 | 0.9272 | 0.9374 | 0.9322 | 0.9395 |
| 0.0956 | 7.0 | 2786 | 0.3120 | 0.9489 | 0.9593 | 0.9540 | 0.9503 |
| 0.0658 | 8.0 | 3184 | 0.4163 | 0.9489 | 0.9460 | 0.9475 | 0.9455 |
| 0.0433 | 9.0 | 3582 | 0.4144 | 0.9493 | 0.9381 | 0.9437 | 0.9408 |
| 0.0433 | 10.0 | 3980 | 0.3658 | 0.9470 | 0.9522 | 0.9496 | 0.9451 |
Framework versions
- Transformers 4.55.4
- Pytorch 2.8.0+cu126
- Datasets 4.0.0
- Tokenizers 0.21.4