Hacker News
new
past
comments
ask
show
jobs
points
by
wyager
4 days ago
|
comments
by
kurisufag
4 days ago
|
[-]
It's certainly large enough for trillion-param frontier-tier trainings, which will likely result in capable open-weight models, the thing you just wished for.
reply