Whisper Small Korean
This model is a fine-tuned version of openai/whisper-small on the Zeroth Korean dataset. It achieves the following results on the evaluation set:
- Loss: 0.0899
- Wer: 6.7610
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: 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 |
|---|---|---|---|---|
| 0.1277 | 0.72 | 1000 | 0.1489 | 12.2271 |
| 0.0379 | 1.44 | 2000 | 0.1053 | 6.7159 |
| 0.0138 | 2.16 | 3000 | 0.0918 | 6.0382 |
| 0.0141 | 2.87 | 4000 | 0.0899 | 6.7610 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0a0+d0d6b1f
- Datasets 2.7.1
- Tokenizers 0.13.2
- Downloads last month
- 69
Dataset used to train seastar105/whisper-small-ko-zeroth
Evaluation results
- Wer on Zeroth Koreantest set self-reported6.761