sparse_broadcast_to: less memory footprint, fewer kernel launches #142364
Conversation
This file contains hidden or 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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/142364
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit 1c535d6 with merge base c29b4ed ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
nikitaved
added
module: sparse
nikitaved
force-pushed
the
nikitaved/coo_broadcast_less_memory
branch
3 times, most recently
from
9db7b0a to
fae893d
Compare
janeyx99
added
the
triaged
amjames
approved these changes
Dec 10, 2024
Collaborator
amjames
left a comment
amjames
left a comment
There was a problem hiding this comment.
Nice! This looks good, minor nit in the CHECK message.
aten/src/ATen/native/TensorShape.cpp
Outdated
|
|
||
| for (int64_t i = 0; i < self.dim(); ++i) { | ||
| TORCH_CHECK(self_size[i] == 1 || self_size[i] == size[i + new_sparse_dims], | ||
| "The input's lenght ", self_size[i], " at dimension ", i, |
Collaborator
There was a problem hiding this comment.
Suggested change
| "The input's lenght ", self_size[i], " at dimension ", i, | |
| "The input's length ", self_size[i], " at dimension ", i, |
aten/src/ATen/native/TensorShape.cpp
Outdated
| for (int64_t i = 0; i < self.dim(); ++i) { | ||
| TORCH_CHECK(self_size[i] == 1 || self_size[i] == size[i + new_sparse_dims], | ||
| "The input's lenght ", self_size[i], " at dimension ", i, | ||
| " does not broadcast over the requested shape of lenght ", size[i + new_sparse_dims], |
Collaborator
There was a problem hiding this comment.
Suggested change
| " does not broadcast over the requested shape of lenght ", size[i + new_sparse_dims], | |
| " does not broadcast over the requested shape of length ", size[i + new_sparse_dims], |
nikitaved
force-pushed
the
nikitaved/coo_broadcast_less_memory
branch
from
fae893d to
1c535d6
Compare
Collaborator
Author
|
Thank you for your reviews, guys! @pytorchbot merge |
Collaborator
Merge startedYour change will be merged once all checks pass (ETA 0-4 Hours). Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
6 participants
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.
As per title.
The following implementation removes the usage of
repeat_interleave, tileandfull_coo_indicesand replaces them with broadcasting. That way we reduce memory traffic (and are likely to hit cache a lot) and the total number of launched kernels.cc @alexsamardzic @pearu @cpuhrsch @amjames @bhosmer @jcaip