HN
Hacker News
Top
New
Best
Ask
Show
Jobs
Comment by IanCal | Hacker News Reader
Parent
Full thread
IanCal
·
It's not quantising existing models, they're training new ones.
View on HN
leroman
·
I understand this part but it seemed that the 16->8->4 etc is similar to compression of the "net" and seemed to lower quality below 8.
Reply on news.ycombinator.com