Support dim order in Arm backend #5576
Conversation
Add both ahead-of-time and runtime guards to make sure we don't accept inputs with any other memory-format than contiguous format. Signed-off-by: Oscar Andersson <oscar.andersson@arm.com> Change-Id: I9e29badbcf238d458e12f0d62394abae66d421b7
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/5576
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 7812e34 with merge base df72b8c ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
facebook-github-bot
added
the
CLA Signed
oscarandersson8218
added
partner: arm
facebook-github-bot
added
the
CLA Signed
|
Addresses #4873 @digantdesai |
|
@oscarandersson8218 can you comment on 4873 saying it is resolved by this, just for visibility? Thanks again. |
There was a problem hiding this comment.
Since you don't lower to_copy or other core aten ops which takes torch.memory_format as an arg, this should be sufficient.
I would say please run ET tests locally with this switch flipped just to make sure.
|
@digantdesai has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
|
@digantdesai merged this pull request in 2dd88fb. |
Great! I gave it a go and everything passed @digantdesai |
3 participants
Add both ahead-of-time and runtime guards to make sure we don't accept inputs with any other memory-format than contiguous format.
Change-Id: I9e29badbcf238d458e12f0d62394abae66d421b7