Conv2dConfig reallocate_halo_output default to true #16185
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
pavlejosipovic
requested review from
nkpatel-tt,
mywoodstock,
shwetankTT and
sankarmanoj-tt
as code owners
mywoodstock
approved these changes
Dec 20, 2024
pavlejosipovic
force-pushed
the
pjosipovic/halo_reallocate
branch
2 times, most recently
from
ad4f79c to
b26c6c2
Compare
Idea behind this flag is to enable reallocation of halo output buffer in place of input buffer if case deallocate_activation flag is also set. In this case halo output is moved up and memory fragmentation is reduced. In case deallocate_activation is not set reallocate_halo_output won't have an effect as ttnn::move will be a no-op since input buffer is not deallocated. Current problem is that if user sets deallocate_activation it doesn't help them with memory issue unless they set reallocate_halo_output as well. This change is addressing this by setting reallocate_halo_output to true by default. This helps by increassing pass rate in ttnn torch traces.
pavlejosipovic
force-pushed
the
pjosipovic/halo_reallocate
branch
from
b26c6c2 to
51652b4
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Idea behind this flag is to enable reallocation of halo output buffer in place of input buffer if case deallocate_activation flag is also set.
In this case halo output is moved up and memory
fragmentation is reduced.
In case deallocate_activation is not set reallocate_halo_output won't have an effect as ttnn::move will be a no-op since input buffer is not deallocated.
Current problem is that if user sets deallocate_activation it doesn't help them with memory issue unless they set reallocate_halo_output as well.
This change is addressing this by setting reallocate_halo_output to true by default.
This helps by increasing pass rate in ttnn torch
traces.
Checklist