Hi all,
at 
Hugging Face, https://huggingface.co/, we are currently organizing an event where Facebook’s multi-lingual Wav2Vec2 model: Wav2vec 2.0: Learning the structure of speech from raw audio will be fine-tuned on all 60 languages provided by Common Voice 6.1.
Here is a blog post that describes in detail how to participate:
If you’re keen on helping the community have access to state-of-the-art Speech models, we would be more than happy to include you in the event!