Hacker News
new
past
comments
ask
show
jobs
points
by
tripplyons
2 hours ago
|
comments
by
spwa4
1 hours ago
|
[-]
Is that really going to matter in FP32, FP16 or BF16? I would think models would be written so they'd be at least somewhat numerically stable.
Also if the inference provider guarantees specific hardware this shouldn't happen.
reply