Hacker News new | ask | show | jobs
by riku_iki 1203 days ago
perhaps this approach can lead to better training of foundational models?..
1 comments
vov_or 1203 days ago
More efficient - for sure!
link