whisper-small-tamil
This model is a fine-tuned version of openai/whisper-small on the google/fleurs dataset for Tamil. It achieves the following results on the evaluation set:
- Loss: 0.42
- Wer: 15.02
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 5000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0882 | 2.27 | 500 | 0.2674 | 16.7354 |
| 0.0026 | 11.76 | 1000 | 0.3508 | 15.3720 |
| 0.0012 | 17.64 | 1500 | 0.3920 | 15.6156 |
| 0.0009 | 23.53 | 2000 | 0.4076 | 15.4284 |
| 0.0002 | 29.41 | 2500 | 0.4268 | 15.0215 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1
- Tokenizers 0.13.2