Downloads · 30 days
6
1% of all-time downloads
wav2vec2-xlsr-romansh_vallader
sammy786/wav2vec2-xlsr-romansh_vallader
wav2vec2-xlsr-romansh_vallader is a automatic speech recognition model from sammy786. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the MOZILLA-FOUNDATION/COMMONVOICE80 - rm-vallader dataset. It achieves the following results on evaluation set (which is 10 percent of train data se…
All-time downloads
416
Public
Repo size
7.7 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin3.9 GB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- wav2vec2
- Access
- Public
- Created
- Mar 2, 2022
- Updated
- Mar 23, 2022
- SHA
- c1ce8277
Model card
From the Hugging Face model README
sammy786/wav2vec2-xlsr-romansh_vallader
This model is a fine-tuned version of facebook/wav2vec2-xls-r-1b on the MOZILLA-FOUNDATION/COMMON_VOICE_8_0 - rm-vallader dataset. It achieves the following results on evaluation set (which is 10 percent of train data set merged with other and dev datasets):
- Loss: 30.31
- Wer: 26.32
Model description
"facebook/wav2vec2-xls-r-1b" was finetuned.
Intended uses & limitations
More information needed
Training and evaluation data
Training data - Common voice Finnish train.tsv, dev.tsv and other.tsv
Training procedure
For creating the train dataset, all possible datasets were appended and 90-10 split was used.
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.000045637994662983496
- train_batch_size: 16
- eval_batch_size: 16
- seed: 13
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine_with_restarts
- lr_scheduler_warmup_steps: 500
- num_epochs: 40
- mixed_precision_training: Native AMP
Training results
| Step | Training Loss | Validation Loss | Wer |
|---|---|---|---|
| 200 | 5.895100 | 3.136624 | 0.999713 |
| 400 | 1.545700 | 0.445069 | 0.471584 |
| 600 | 0.693900 | 0.340700 | 0.363088 |
| 800 | 0.510600 | 0.295432 | 0.289610 |
| 1000 | 0.318800 | 0.286795 | 0.281860 |
| 1200 | 0.194000 | 0.307468 | 0.274110 |
| 1400 | 0.151800 | 0.304849 | 0.264351 |
| 1600 | 0.148300 | 0.303112 | 0.263203 |
Framework versions
- Transformers 4.16.0.dev0
- Pytorch 1.10.0+cu102
- Datasets 1.17.1.dev0
- Tokenizers 0.10.3
Evaluation Commands
- To evaluate on
mozilla-foundation/common_voice_8_0with splittest
python eval.py --model_id sammy786/wav2vec2-xlsr-romansh_vallader --dataset mozilla-foundation/common_voice_8_0 --config rm-vallader --split test
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- apache-2.0
- Created
- Mar 2, 2022
- Updated
- Mar 23, 2022