Whisper Large-V3 Ko to En
This model is a fine-tuned version of openai/whisper-large-v3 on the Gentlemen dataset. It achieves the following results on the evaluation set:
- Loss: 1.1304
- Wer: 0.7630
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-06
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.2511 | 0.4507 | 1000 | 1.2619 | 0.7806 |
| 1.1681 | 0.9013 | 2000 | 1.1703 | 0.7751 |
| 0.9217 | 1.3520 | 3000 | 1.1486 | 0.7768 |
| 0.9093 | 1.8026 | 4000 | 1.1304 | 0.7630 |
Framework versions
- Transformers 4.45.2
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.20.3