Whisper Small Xhosa
This model is a fine-tuned version of openai/whisper-small on the google/fleurs xh_za dataset. It achieves the following results on the evaluation set:
- Loss: 1.1217
- Wer: 65.9512
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: 1e-05
- train_batch_size: 64
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- training_steps: 200
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.583 | 7.01 | 100 | 1.0822 | 67.3176 |
| 0.1146 | 14.02 | 200 | 1.1217 | 65.9512 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.8.1.dev0
- Tokenizers 0.13.2
- Downloads last month
- 10
Dataset used to train julie200/whisper-small-xh_za
Evaluation results
- Wer on google/fleurs xh_zatest set self-reported65.951