whisper-small-lora-afrispeech-general
This model is a fine-tuned version of openai/whisper-small on the afrispeech-200 dataset. It achieves the following results on the evaluation set:
- Loss: 0.5842
- Wer: 0.174
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: 0.0001
- 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: linear
- lr_scheduler_warmup_steps: 1349
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.5444 | 1.0 | 1350 | 0.6210 | 0.192 |
| 0.4563 | 2.0 | 2700 | 0.5523 | 0.178 |
| 0.3381 | 3.0 | 4050 | 0.5415 | 0.17 |
| 0.2655 | 4.0 | 5400 | 0.5451 | 0.174 |
| 0.1993 | 5.0 | 6750 | 0.5664 | 0.174 |
| 0.1559 | 6.0 | 8100 | 0.5842 | 0.174 |
Framework versions
- Transformers 4.44.0
- Pytorch 2.10.0+cu128
- Datasets 2.19.0
- Tokenizers 0.19.1