afro-xlmr-base-finetuned-augmentation
This model is a fine-tuned version of Davlan/afro-xlmr-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.3150
- F1: 0.4562
- Roc Auc: 0.6873
- Accuracy: 0.5316
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 100
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Roc Auc | Accuracy |
|---|---|---|---|---|---|---|
| 0.3329 | 1.0 | 198 | 0.3191 | 0.0 | 0.5 | 0.3939 |
| 0.2856 | 2.0 | 396 | 0.2775 | 0.2481 | 0.6027 | 0.4874 |
| 0.255 | 3.0 | 594 | 0.2676 | 0.2843 | 0.6146 | 0.5076 |
| 0.2259 | 4.0 | 792 | 0.2701 | 0.3125 | 0.6277 | 0.5227 |
| 0.1698 | 5.0 | 990 | 0.2734 | 0.3600 | 0.6396 | 0.5303 |
| 0.1543 | 6.0 | 1188 | 0.2924 | 0.3801 | 0.6487 | 0.5278 |
| 0.1171 | 7.0 | 1386 | 0.2982 | 0.3969 | 0.6589 | 0.5379 |
| 0.1093 | 8.0 | 1584 | 0.3115 | 0.4478 | 0.6848 | 0.5202 |
| 0.0934 | 9.0 | 1782 | 0.3150 | 0.4562 | 0.6873 | 0.5316 |
| 0.0791 | 10.0 | 1980 | 0.3377 | 0.4285 | 0.6796 | 0.5189 |
| 0.0594 | 11.0 | 2178 | 0.3463 | 0.4420 | 0.6746 | 0.5379 |
| 0.0571 | 12.0 | 2376 | 0.3617 | 0.4504 | 0.6866 | 0.5366 |
| 0.0412 | 13.0 | 2574 | 0.3723 | 0.4357 | 0.6780 | 0.5328 |
Framework versions
- Transformers 4.45.1
- Pytorch 2.4.0
- Datasets 3.0.1
- Tokenizers 0.20.0