Just what I've been waiting to see - serious progress on making large language models accessible to truly low-resource languages! This has huge potential to democratize access to language tech. https://www.reddit.com/user/GrowthExciting1126