nyansapo_model_v2
This model is a fine-tuned version of Nzyoka19/nyansapo_model_v1 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0302
- Wer: 17.4747
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: 3e-06
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 16
- 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: cosine
- lr_scheduler_warmup_steps: 200
- num_epochs: 8
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1565 | 0.6211 | 100 | 0.0309 | 19.6970 |
| 0.1518 | 1.2422 | 200 | 0.0328 | 21.5152 |
| 0.1585 | 1.8634 | 300 | 0.0293 | 19.7980 |
| 0.1341 | 2.4845 | 400 | 0.0289 | 18.0808 |
| 0.1516 | 3.1056 | 500 | 0.0302 | 18.0808 |
| 0.1196 | 3.7267 | 600 | 0.0298 | 18.6869 |
| 0.1494 | 4.3478 | 700 | 0.0284 | 18.1818 |
| 0.1337 | 4.9689 | 800 | 0.0302 | 17.2727 |
| 0.139 | 5.5901 | 900 | 0.0297 | 17.8788 |
| 0.1356 | 6.2112 | 1000 | 0.0299 | 17.7778 |
| 0.121 | 6.8323 | 1100 | 0.0302 | 17.5758 |
| 0.091 | 7.4534 | 1200 | 0.0302 | 17.4747 |
Framework versions
- Transformers 4.57.3
- Pytorch 2.9.0+cu126
- Datasets 4.4.1
- Tokenizers 0.22.1
- Downloads last month
- 43