Downloads · 30 days
3
9% of all-time downloads
whisper-medium-korean-20epoch
NgQuocThai/whisper-medium-korean-20epoch
whisper-medium-korean-20epoch is a automatic speech recognition model from NgQuocThai. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
33
Public
Parameters
764M
3.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors3.1 GB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- whisper
- Access
- Public
- Created
- Aug 29, 2025
- Updated
- Aug 30, 2025
- SHA
- 843e500e
Base models
Model card
From the Hugging Face model README
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
whisper-medium-korean-20epoch
This model is a fine-tuned version of openai/whisper-medium on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.8196
- Cer: 16.1389
- Wer: 28.3068
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: 4
- eval_batch_size: 8
- 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: 500
- num_epochs: 25
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer | Wer |
|---|---|---|---|---|---|
| 1.8837 | 1.0 | 2702 | 0.7118 | 27.2803 | 44.3503 |
| 0.9145 | 2.0 | 5404 | 0.6716 | 22.7087 | 39.6063 |
| 0.7338 | 3.0 | 8106 | 0.6486 | 22.8998 | 37.5077 |
| 0.6122 | 4.0 | 10808 | 0.6450 | 20.8439 | 34.0215 |
| 0.5172 | 5.0 | 13510 | 0.6406 | 19.6020 | 32.7979 |
| 0.439 | 6.0 | 16212 | 0.6463 | 18.9193 | 32.7842 |
| 0.3738 | 7.0 | 18914 | 0.6604 | 20.3980 | 33.4404 |
| 0.3187 | 8.0 | 21616 | 0.6605 | 19.5522 | 32.4014 |
| 0.2709 | 9.0 | 24318 | 0.6820 | 17.1221 | 29.7355 |
| 0.2331 | 10.0 | 27020 | 0.6895 | 17.9182 | 30.6036 |
| 0.1975 | 11.0 | 29722 | 0.7089 | 16.7758 | 29.1339 |
| 0.1703 | 12.0 | 32424 | 0.7093 | 17.2415 | 29.4894 |
| 0.1463 | 13.0 | 35126 | 0.7251 | 16.8932 | 29.3390 |
| 0.1253 | 14.0 | 37828 | 0.7343 | 16.8853 | 29.3800 |
| 0.1091 | 15.0 | 40530 | 0.7537 | 17.1321 | 29.5509 |
| 0.0949 | 16.0 | 43232 | 0.7707 | 16.5390 | 29.0382 |
| 0.0829 | 17.0 | 45934 | 0.7829 | 16.5668 | 29.0792 |
| 0.0732 | 18.0 | 48636 | 0.7883 | 16.4633 | 28.8947 |
| 0.0653 | 19.0 | 51338 | 0.8004 | 16.2544 | 28.6896 |
| 0.0584 | 20.0 | 54040 | 0.8129 | 16.2643 | 28.5460 |
| 0.0535 | 21.0 | 56742 | 0.8192 | 16.1946 | 28.5665 |
| 0.0495 | 22.0 | 59444 | 0.8196 | 16.1389 | 28.3068 |
| 0.0459 | 23.0 | 62146 | 0.8238 | 16.7320 | 29.3321 |
| 0.0429 | 24.0 | 64848 | 0.8287 | 16.6564 | 29.1339 |
| 0.0414 | 25.0 | 67550 | 0.8304 | 16.3260 | 28.7306 |
Framework versions
- Transformers 4.55.2
- Pytorch 2.7.1+cu126
- Datasets 3.6.0
- Tokenizers 0.21.2
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- License
- apache-2.0
- Created
- Aug 29, 2025
- Updated
- Aug 30, 2025