Fix DLRMv3 fbgemm_gpu loading with hardcoded Bazel paths#2439
Open
ssam18 wants to merge 5 commits intomlcommons:masterfrom
Open
Fix DLRMv3 fbgemm_gpu loading with hardcoded Bazel paths#2439ssam18 wants to merge 5 commits intomlcommons:masterfrom
ssam18 wants to merge 5 commits intomlcommons:masterfrom
Conversation
Replace internal Bazel build paths with proper Python imports to fix library loading failures outside Meta's build environment. Fixes mlcommons#2429
Contributor
|
MLCommons CLA bot All contributors have signed the MLCommons CLA ✍️ ✅ |
Contributor
Author
|
Signing the CLA now via the Google form. |
Contributor
Author
|
recheck |
LinjianMa
approved these changes
Jan 19, 2026
Contributor
|
I have verified the PR works as expected. @hanyunfan @mrmhodak can we merge the PR? |
mrmhodak
previously approved these changes
Feb 9, 2026
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Problem
The DLRMv3 harness was using hardcoded Bazel build paths specific to Meta's internal build system:
torch.ops.load_library("//deeplearning/fbgemm/fbgemm_gpu:sparse_ops")
torch.ops.load_library("//deeplearning/fbgemm/fbgemm_gpu:sparse_ops_cpu")
This caused failures when running outside that environment:
FAILED to load sparse_ops_cpu in position: Could not load this library: /deeplearning/fbgemm/fbgemm_gpu:sparse_ops
FAILED to load sparse_ops_cpu in jagged: Could not load this library: /deeplearning/fbgemm/fbgemm_gpu:sparse_ops
FAILED to load sparse_ops_cpu in jagged tensors: Could not load this library: /deeplearning/fbgemm/fbgemm_gpu:sparse_ops
FAILED to load sparse_ops_cpu in hstu attention: Could not load this library: /deeplearning/fbgemm/fbgemm_gpu:sparse_ops
Fixes #2429