Y
Hacker News
new
|
ask
|
show
|
jobs
by
pinstripes
1 day ago
Trying to scale text inference to 1 million tok/s on cheap hardware.