whisper-small-sw-common-voice-ndizi-158-1
This model is a fine-tuned version of smutuvi/whisper-small-sw-common-voice on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.7815
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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 1.0 | 18 | 1.7182 |
| 1.6241 | 2.0 | 36 | 1.5595 |
| 1.2418 | 3.0 | 54 | 1.5196 |
| 1.2418 | 4.0 | 72 | 1.4782 |
| 0.91 | 5.0 | 90 | 1.5149 |
| 0.6419 | 6.0 | 108 | 1.5551 |
| 0.4734 | 7.0 | 126 | 1.5569 |
| 0.4734 | 8.0 | 144 | 1.6284 |
| 0.3298 | 9.0 | 162 | 1.6445 |
| 0.213 | 10.0 | 180 | 1.6347 |
| 0.213 | 11.0 | 198 | 1.7054 |
| 0.1689 | 12.0 | 216 | 1.7045 |
| 0.1384 | 13.0 | 234 | 1.6981 |
| 0.0672 | 14.0 | 252 | 1.7077 |
| 0.0672 | 15.0 | 270 | 1.7382 |
| 0.0454 | 16.0 | 288 | 1.7670 |
| 0.0386 | 17.0 | 306 | 1.7628 |
| 0.0386 | 18.0 | 324 | 1.7687 |
| 0.0247 | 19.0 | 342 | 1.7854 |
| 0.021 | 20.0 | 360 | 1.7815 |
Framework versions
- PEFT 0.7.2.dev0
- Transformers 4.37.0.dev0
- Pytorch 2.0.0
- Datasets 2.16.1
- Tokenizers 0.15.0