whisper-small-sinhala-general-185k
This model is a fine-tuned version of openai/whisper-small on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0724
- Wer: 25.0275
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: 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
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1493 | 0.3837 | 1000 | 0.1486 | 43.1593 |
| 0.1053 | 0.7675 | 2000 | 0.1040 | 32.5843 |
| 0.0773 | 1.1512 | 3000 | 0.0883 | 29.3236 |
| 0.071 | 1.5349 | 4000 | 0.0818 | 28.5305 |
| 0.0666 | 1.9186 | 5000 | 0.0767 | 26.6799 |
| 0.052 | 2.3024 | 6000 | 0.0744 | 25.3139 |
| 0.0474 | 2.6861 | 7000 | 0.0724 | 25.0275 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.9.0+cu126
- Datasets 4.4.1
- Tokenizers 0.22.1