Downloads · 30 days
12
1% of all-time downloads
wav2vec2-large-xlsr-53-kalmyk
tugstugi/wav2vec2-large-xlsr-53-kalmyk
wav2vec2-large-xlsr-53-kalmyk is a automatic speech recognition model from tugstugi. 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 Wav2Vec2 model was first pretrained on 500 hours Kalmyk TV recordings and 1000 hours Mongolian speech recognition dataset. After that, the model was finetuned on a 300 hours Kalmyk synthetic STT dataset created b…
All-time downloads
1.3K
Public
Repo size
12.6 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin1.3 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
- Jul 25, 2021
- SHA
- 9bbe0c8f
Model card
From the Hugging Face model README
Info
This Wav2Vec2 model was first pretrained on 500 hours Kalmyk TV recordings and 1000 hours Mongolian speech recognition dataset. After that, the model was finetuned on a 300 hours Kalmyk synthetic STT dataset created by a voice conversion model.
- 50% WER on a private test set created from Kalmyk TV recordnings
- on clean voice recordings, the model should have much lower WER
- voice conversion info
- 300 hours Kalmyk synthetic STT dataset
- The source voice is a Kalmyk female voice TTS
- Target voices are from the VCTK dataset
- example data: https://twitter.com/tugstugi/status/1409111296897912835
- each WAV has a different text created from Kalmyk books
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- apache-2.0
- Languages
- xal
- Created
- Mar 2, 2022
- Updated
- Jul 25, 2021