You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository has been archived by the owner on Mar 18, 2023. It is now read-only.
Currently we say things like "<8 TPU chips in 2x2x2 topology>", which is informative, but doesn't match Google terminology. We could try printing something like "<TPUv3-8 in 2x2 topology>". We don't get the chip kind information from the topology, but we could for now just determine it from the memory information.
The text was updated successfully, but these errors were encountered:
Turns out the memory information does not reflect the actual TPU memory, so we're gonna have to figure out something else here. It is possible to get this from the profiler, but that's on a separate port, so it's a bit annoying.
Sign up for freeto subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Currently we say things like "<8 TPU chips in 2x2x2 topology>", which is informative, but doesn't match Google terminology. We could try printing something like "<TPUv3-8 in 2x2 topology>". We don't get the chip kind information from the topology, but we could for now just determine it from the memory information.
The text was updated successfully, but these errors were encountered: