Downloads · 30 days
94
7% of all-time downloads
wav2vec2-large-xlsr-persian-v3
Muayad2025/wav2vec2-large-xlsr-persian-v3
wav2vec2-large-xlsr-persian-v3 is a automatic speech recognition model from Muayad2025. 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
1.4K
Public
Parameters
315M
349 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors2.5 GB · 50%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- Model type
- wav2vec2
- Access
- Public
- Created
- Apr 18, 2026
- Updated
- Apr 25, 2026
- SHA
- 72504e2a
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. -->
wav2vec2-large-xlsr-persian-v3
This model is a fine-tuned version of m3hrdadfi/wav2vec2-large-xlsr-persian-v3 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.9866
- Wer: 0.4140
- Cer: 0.0857
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED 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: 1000
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer |
|---|---|---|---|---|---|
| 3.0631 | 1.0246 | 500 | 1.1507 | 0.7601 | 0.1762 |
| 2.2749 | 2.0492 | 1000 | 0.9054 | 0.6031 | 0.1372 |
| 1.7436 | 3.0738 | 1500 | 0.8125 | 0.5660 | 0.1249 |
| 1.3863 | 4.0985 | 2000 | 0.8070 | 0.5234 | 0.1149 |
| 1.1449 | 5.1231 | 2500 | 0.7524 | 0.4885 | 0.1062 |
| 0.9244 | 6.1477 | 3000 | 0.7877 | 0.4952 | 0.1079 |
| 0.8508 | 7.1723 | 3500 | 0.8065 | 0.4728 | 0.1023 |
| 0.7046 | 8.1969 | 4000 | 0.8046 | 0.4640 | 0.0996 |
| 0.6719 | 9.2215 | 4500 | 0.8432 | 0.4589 | 0.0972 |
| 0.6039 | 10.2462 | 5000 | 0.8846 | 0.4534 | 0.0954 |
| 0.5592 | 11.2708 | 5500 | 0.8684 | 0.4483 | 0.0957 |
| 0.4851 | 12.2954 | 6000 | 0.8910 | 0.4372 | 0.0924 |
| 0.4627 | 13.32 | 6500 | 0.9257 | 0.4399 | 0.0916 |
| 0.4216 | 14.3446 | 7000 | 0.9784 | 0.4317 | 0.0900 |
| 0.4046 | 15.3692 | 7500 | 0.9719 | 0.4252 | 0.0886 |
| 0.3820 | 16.3938 | 8000 | 0.9712 | 0.4277 | 0.0893 |
| 0.3696 | 17.4185 | 8500 | 0.9684 | 0.4214 | 0.0880 |
| 0.3592 | 18.4431 | 9000 | 0.9757 | 0.4176 | 0.0869 |
| 0.3343 | 19.4677 | 9500 | 0.9866 | 0.4140 | 0.0857 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 4.8.3
- Tokenizers 0.22.2
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- Created
- Apr 18, 2026
- Updated
- Apr 25, 2026