wav2vec2-xlsr-dora-afrispeech-clinical
This model is a fine-tuned version of jonatasgrosman/wav2vec2-large-xlsr-53-english on the afrispeech-200 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4738
- Wer: 0.34
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- 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: 2239
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.5498 | 1.0 | 2240 | 0.6381 | 0.423 |
| 1.4043 | 2.0 | 4480 | 0.5663 | 0.392 |
| 1.4029 | 3.0 | 6720 | 0.5337 | 0.372 |
| 1.3421 | 4.0 | 8960 | 0.5182 | 0.364 |
| 1.2465 | 5.0 | 11200 | 0.4989 | 0.353 |
| 1.2461 | 6.0 | 13440 | 0.4914 | 0.349 |
| 1.1286 | 7.0 | 15680 | 0.4825 | 0.346 |
| 1.1365 | 8.0 | 17920 | 0.4766 | 0.343 |
| 1.0916 | 9.0 | 20160 | 0.4756 | 0.341 |
| 1.1049 | 10.0 | 22400 | 0.4738 | 0.34 |
Framework versions
- PEFT 0.18.1
- Transformers 5.3.0
- Pytorch 2.12.0.dev20260307+cu128
- Datasets 2.19.0
- Tokenizers 0.22.2