Lower aligment requirements for shallow conv2d #17711
Open
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
force-pushed
the
pjosipovic/shallow_conv_8
branch
from
21a66d9 to
2af90f1
Compare
s-jovic
approved these changes
Feb 7, 2025
| @@ -11,6 +11,7 @@ | |||
| #include "conv2d_utils.hpp" | |||
| #include <tt-metalium/buffer_constants.hpp> | |||
| #include "tt-metalium/hal.hpp" | |||
| #include "tt-metalium/math.hpp" | |||
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Outdated suggestions cannot be applied.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.
Conv2d aligns input channels to 32.
Currently for height sharded convs we user to relax this and to set 16 as alignment for input channels
This changes relaxes this further, and user can set 8 as min channel alignment for height shareded convs.
This was enabled by #17122
Auto-shard codepath now uses 8 as channel alignment for shallow convs.
This enables more pytorch traces to pass as it saves on L1 memory usage.
This change was manually tested on BH as well.
Checklist