Downloads · 30 days
11
15% of all-time downloads
Magadhi-Kumaoni-Marathi-Rajasthani-w2v-bert-2.0
dhasmana/Magadhi-Kumaoni-Marathi-Rajasthani-w2v-bert-2.0
Magadhi-Kumaoni-Marathi-Rajasthani-w2v-bert-2.0 is a automatic speech recognition model from dhasmana. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as mit.
should probably proofread and complete it, then remove this comment. --
All-time downloads
74
Public
Parameters
606M
17 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors2.4 GB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- mit
- Model type
- wav2vec2-bert
- Access
- Public
- Created
- Feb 13, 2026
- Updated
- Feb 13, 2026
- SHA
- d1f31ef4
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. -->
Magadhi-Kumaoni-Marathi-Rajasthani-w2v-bert-2.0
This model is a fine-tuned version of facebook/w2v-bert-2.0 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.7012
- Cer: 0.1706
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-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- 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: 500
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Cer |
|---|---|---|---|---|
| 5.1348 | 0.7519 | 300 | 1.0405 | 0.2238 |
| 1.7498 | 1.5038 | 600 | 0.8808 | 0.2111 |
| 1.3650 | 2.2556 | 900 | 0.8322 | 0.1905 |
| 1.0378 | 3.0075 | 1200 | 0.8499 | 0.1965 |
| 0.7725 | 3.7594 | 1500 | 0.8474 | 0.1774 |
| 0.5685 | 4.5113 | 1800 | 0.9253 | 0.1735 |
| 0.3957 | 5.2632 | 2100 | 1.0293 | 0.1746 |
| 0.2809 | 6.0150 | 2400 | 1.0283 | 0.1754 |
| 0.1764 | 6.7669 | 2700 | 1.2738 | 0.1727 |
| 0.1236 | 7.5188 | 3000 | 1.3591 | 0.1733 |
| 0.0829 | 8.2707 | 3300 | 1.5282 | 0.1734 |
| 0.0519 | 9.0226 | 3600 | 1.6169 | 0.1708 |
| 0.0307 | 9.7744 | 3900 | 1.7012 | 0.1706 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.9.0+cu128
- Datasets 4.0.0
- Tokenizers 0.22.2
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2-bert
- License
- mit
- Created
- Feb 13, 2026
- Updated
- Feb 13, 2026