Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Support Multi-node Evaluation #132

Merged
merged 11 commits into from
Aug 26, 2023
Merged

Support Multi-node Evaluation #132

merged 11 commits into from
Aug 26, 2023

Conversation

NouamaneTazi
Copy link
Member

@NouamaneTazi NouamaneTazi commented Aug 7, 2023 •
edited
Loading

Fixes #103

Test using

pip install git+https://github.com/NouamaneTazi/beir@nouamane/better-multi-gpu
torchrun --nproc_per_node=2 scripts/retrieval_multigpu.py

@NouamaneTazi NouamaneTazi changed the title Multi-GPU support Support Multi-node Evaluation Aug 12, 2023
@NouamaneTazi NouamaneTazi marked this pull request as ready for review August 15, 2023 07:54
Muennighoff
Muennighoff approved these changes Aug 15, 2023
View reviewed changes
Copy link
Contributor

@Muennighoff Muennighoff left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Nice! I get the same results for 2 GPUs via Multi-GPU, 1 GPU via Multi-GPU & 1 GPU via default running ArguAna w/ e5-small on Nvidia Tesla T4 15109MiB

Times:
2 GPUs: 34.9s
1 GPU via Multi-GPU: 59.56s
1 GPU via default: 53.59s

mteb/abstasks/AbsTaskRetrieval.py Outdated Show resolved Hide resolved
@NouamaneTazi NouamaneTazi merged commit 0dd82a9 into main Aug 26, 2023
@KennethEnevoldsen KennethEnevoldsen deleted the nouamane/parallel branch March 20, 2024 17:03
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@Muennighoff Muennighoff Muennighoff approved these changes

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.

Status of multi-gpu support for BEIR evaluation
2 participants
@NouamaneTazi @Muennighoff