whisper-small-finetuned-iwslt-irish-welsh-curriculum-synthetic-b
This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.4918
- Wer: 100.2168
- Bleu: 6.8330
- Chrf++: 27.4191
- Comet: 0.5177
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: 5e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 1000
- training_steps: 15000
- mixed_precision_training: Native AMP
- label_smoothing_factor: 0.1
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Bleu | Chrf++ | Comet |
|---|---|---|---|---|---|---|---|
| 2.2968 | 0.4264 | 2500 | 3.6900 | 119.2808 | 3.4280 | 20.6598 | 0.4696 |
| 2.0666 | 0.8528 | 5000 | 3.5337 | 119.8048 | 4.5093 | 23.0651 | 0.4885 |
| 2.3623 | 1.2792 | 7500 | 3.4509 | 102.3130 | 6.5378 | 26.0156 | 0.5112 |
| 2.1774 | 1.7056 | 10000 | 3.4396 | 109.8392 | 6.3905 | 26.5266 | 0.5161 |
| 2.0625 | 2.1320 | 12500 | 3.4622 | 98.0033 | 7.2093 | 27.2407 | 0.5218 |
| 2.0016 | 2.5584 | 15000 | 3.4918 | 100.2168 | 6.8330 | 27.4191 | 0.5177 |
Framework versions
- Transformers 4.52.0
- Pytorch 2.12.1+cu130
- Datasets 2.21.0
- Tokenizers 0.21.4