Downloads · 30 days
1
3% of all-time downloads
whisper-turbo-fr-pho-large
Ycharo/whisper-turbo-fr-pho-large
whisper-turbo-fr-pho-large is a automatic speech recognition model from Ycharo. Use it when you need speech turned into text. It is set up for transformers.
should probably proofread and complete it, then remove this comment. --
All-time downloads
40
Public
Parameters
809M
97.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors3.2 GB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- Model type
- whisper
- Access
- Public
- Created
- May 9, 2025
- Updated
- May 12, 2025
- SHA
- 1b503c07
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-turbo-fr-pho-large
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0720
- Wer Ortho: 137.5675
- Wer: 130.9970
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: 16
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 50
- num_epochs: 3.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
|---|---|---|---|---|---|
| 0.1324 | 0.1 | 500 | 0.1401 | 28.4331 | 14.2598 |
| 0.1154 | 0.2 | 1000 | 0.1158 | 78.2105 | 68.0602 |
| 0.106 | 0.3 | 1500 | 0.1073 | 44.4435 | 30.8810 |
| 0.0878 | 0.4 | 2000 | 0.0916 | 137.0165 | 128.9766 |
| 0.0982 | 0.5 | 2500 | 0.1022 | 181.8600 | 176.1832 |
| 0.0784 | 0.6 | 3000 | 0.0889 | 101.1986 | 91.6390 |
| 0.0802 | 0.7 | 3500 | 0.0814 | 56.4498 | 44.5975 |
| 0.0705 | 0.8 | 4000 | 0.0834 | 52.7781 | 41.1760 |
| 0.0753 | 0.9 | 4500 | 0.0807 | 210.9888 | 206.4460 |
| 0.0778 | 1.0 | 5000 | 0.0764 | 277.0980 | 275.3281 |
| 0.0534 | 1.1 | 5500 | 0.0796 | 248.0980 | 245.4206 |
| 0.0574 | 1.2 | 6000 | 0.0828 | 297.6858 | 297.2156 |
| 0.0553 | 1.3 | 6500 | 0.0755 | 246.3392 | 244.6851 |
| 0.0548 | 1.4 | 7000 | 0.0742 | 383.3272 | 386.8018 |
| 0.049 | 1.5 | 7500 | 0.0742 | 187.3127 | 182.0145 |
| 0.0549 | 1.6 | 8000 | 0.0759 | 250.2355 | 247.9754 |
| 0.0514 | 1.7 | 8500 | 0.0738 | 308.1355 | 308.5257 |
| 0.0565 | 1.8 | 9000 | 0.0739 | 153.9500 | 147.8599 |
| 0.0587 | 1.9 | 9500 | 0.0715 | 282.6388 | 281.4569 |
| 0.0575 | 2.0 | 10000 | 0.0709 | 286.4367 | 285.5370 |
| 0.0429 | 2.1 | 10500 | 0.0753 | 74.2237 | 64.2203 |
| 0.0391 | 2.2 | 11000 | 0.0768 | 369.1016 | 372.4500 |
| 0.0456 | 2.3 | 11500 | 0.0752 | 84.3226 | 75.1626 |
| 0.0373 | 2.4 | 12000 | 0.0757 | 127.9561 | 120.1080 |
| 0.0401 | 2.5 | 12500 | 0.0717 | 201.8650 | 197.4975 |
| 0.0381 | 2.6 | 13000 | 0.0742 | 90.5474 | 80.6946 |
| 0.0437 | 2.7 | 13500 | 0.0725 | 402.4980 | 406.9978 |
| 0.039 | 2.8 | 14000 | 0.0705 | 134.6089 | 126.1199 |
| 0.0354 | 2.9 | 14500 | 0.0701 | 190.7826 | 187.0045 |
| 0.0414 | 3.0 | 15000 | 0.0720 | 137.5675 | 130.9970 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.5.0
- Tokenizers 0.21.1
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- whisper
- Created
- May 9, 2025
- Updated
- May 12, 2025