speed up reference resize kernel #8592
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
|
We could try to use https://github.com/apache/tvm/blob/main/tests/python/topi/python/test_topi_clip.py#L35 memoize to pre-cache computation if the reference is too slow |
|
With this change, the reference is no longer the bottleneck, the issue is just running a lot of tests. |
|
Thanks @mbrookhart @jwfromm ! |
ylc
pushed a commit
to ylc/tvm
that referenced
this pull request
Sep 29, 2021
ylc
pushed a commit
to ylc/tvm
that referenced
this pull request
Jan 13, 2022
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Outdated suggestions cannot be applied.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.
@jwfromm this refactors the resize reference kernel to use numpy array manipulation. It achieves a ~5-8x speedup on the reference kernel which leads to a ~2x speedup in the numerous places we test resize. for relay/test_op_level5.py:test_resize2d, this takes the test time on my machine from 194 seconds to 93 seconds.