I hope this doesn’t come off the wrong way, but why was there a BLOOM working group?
BLOOM models are under a proprietary license, so they are proprietary regardless of the debate around whether training data or preprocessing code needs to be provided.
Is there a Mistral working group?