Whisper Base Bengali
This model is a fine-tuned version of openai/whisper-tiny on the mozilla-foundation/common_voice_16_0 bn dataset. It achieves the following results on the evaluation set:
- Loss: 0.4033
- Wer: 48.4864
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-06
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- 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 |
|---|---|---|---|---|
| 1.084 | 3.03 | 1000 | 1.0815 | 89.0822 |
| 0.4981 | 6.05 | 2000 | 0.5253 | 58.1107 |
| 0.417 | 10.02 | 3000 | 0.4404 | 51.4890 |
| 0.3759 | 13.04 | 4000 | 0.4114 | 49.1128 |
| 0.3841 | 17.01 | 5000 | 0.4033 | 48.4864 |
Framework versions
- Transformers 4.38.0.dev0
- Pytorch 2.1.2+cu121
- Datasets 2.16.2.dev0
- Tokenizers 0.15.0