Downloads · 30 days
4
5% of all-time downloads
wav2vec2-large-mms-1b-aft-lth
robertp408/wav2vec2-large-mms-1b-aft-lth
wav2vec2-large-mms-1b-aft-lth is a automatic speech recognition model from robertp408. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as cc-by-nc-4.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
82
Public
Parameters
965M
69.5 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors3.9 GB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- cc-by-nc-4.0
- Model type
- wav2vec2
- Access
- Public
- Created
- Oct 7, 2025
- Updated
- Oct 10, 2025
- SHA
- 612f9f04
Base models
Datasets
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-mms-1b-aft-lth
This model is a fine-tuned version of facebook/mms-1b-all on the audiofolder dataset. It achieves the following results on the evaluation set:
- Loss: 0.5179
- Wer: 0.3647
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: 0.001
- train_batch_size: 2
- eval_batch_size: 8
- seed: 42
- 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: 100
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 5.0379 | 0.1427 | 100 | 0.8269 | 0.6958 |
| 0.9021 | 0.2853 | 200 | 0.7880 | 0.5667 |
| 0.7282 | 0.4280 | 300 | 0.6612 | 0.5155 |
| 0.6976 | 0.5706 | 400 | 0.7865 | 0.4880 |
| 0.6739 | 0.7133 | 500 | 0.6663 | 0.4963 |
| 0.6461 | 0.8559 | 600 | 0.6165 | 0.5272 |
| 0.6579 | 0.9986 | 700 | 0.5786 | 0.4675 |
| 0.6791 | 1.1412 | 800 | 0.6118 | 0.4647 |
| 0.5954 | 1.2839 | 900 | 0.5795 | 0.4246 |
| 0.6232 | 1.4265 | 1000 | 0.6284 | 0.4412 |
| 0.5789 | 1.5692 | 1100 | 0.5952 | 0.4333 |
| 0.5977 | 1.7118 | 1200 | 0.5452 | 0.4211 |
| 0.5746 | 1.8545 | 1300 | 0.5753 | 0.4207 |
| 0.5984 | 1.9971 | 1400 | 0.5416 | 0.4069 |
| 0.5468 | 2.1398 | 1500 | 0.5811 | 0.3996 |
| 0.5742 | 2.2825 | 1600 | 0.5287 | 0.4280 |
| 0.5633 | 2.4251 | 1700 | 0.5436 | 0.4238 |
| 0.5493 | 2.5678 | 1800 | 0.6521 | 0.4069 |
| 0.556 | 2.7104 | 1900 | 0.5192 | 0.4191 |
| 0.5818 | 2.8531 | 2000 | 0.5227 | 0.3911 |
| 0.5501 | 2.9957 | 2100 | 0.5712 | 0.3880 |
| 0.5122 | 3.1384 | 2200 | 0.5706 | 0.3914 |
| 0.5506 | 3.2810 | 2300 | 0.5104 | 0.3931 |
| 0.4953 | 3.4237 | 2400 | 0.5773 | 0.3728 |
| 0.52 | 3.5663 | 2500 | 0.5611 | 0.3903 |
| 0.5139 | 3.7090 | 2600 | 0.5188 | 0.3724 |
| 0.5062 | 3.8516 | 2700 | 0.5030 | 0.4151 |
| 0.5151 | 3.9943 | 2800 | 0.5060 | 0.3950 |
| 0.5107 | 4.1369 | 2900 | 0.4947 | 0.3884 |
| 0.4942 | 4.2796 | 3000 | 0.5196 | 0.3714 |
| 0.4899 | 4.4223 | 3100 | 0.5121 | 0.3714 |
| 0.4778 | 4.5649 | 3200 | 0.5128 | 0.3625 |
| 0.4619 | 4.7076 | 3300 | 0.5054 | 0.3648 |
| 0.4472 | 4.8502 | 3400 | 0.5040 | 0.3652 |
| 0.4577 | 4.9929 | 3500 | 0.5179 | 0.3647 |
Framework versions
- Transformers 4.57.0
- Pytorch 2.8.0+cu128
- Datasets 4.1.1
- Tokenizers 0.22.1
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- cc-by-nc-4.0
- Created
- Oct 7, 2025
- Updated
- Oct 10, 2025