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.

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

Add StepResources and step replicas in Pipeline #750

Merged
merged 14 commits into from
Jun 27, 2024
Merged

Conversation

gabrielmbmb
Copy link
Member

@gabrielmbmb gabrielmbmb commented Jun 21, 2024

Description

This PR adds a new StepResource class that allow defining the resources and replicas required for each Step. It also adds a resources attribute to the Step class, and updates the _run_steps methods to create the number of replicas specified. In addition, the _BatchManager has been updated to handle receiving unsorted batches by its seq_no due to the Step replicas.

Copy link

codspeed-hq bot commented Jun 21, 2024

CodSpeed Performance Report

Merging #750 will degrade performances by 13.44%

Comparing multiple-workers (fb01893) with develop (63ee8c5)

Summary

❌ 1 (👁 1) regressions

Benchmarks breakdown

Benchmark develop multiple-workers Change
👁 test_cache_time 294.3 ms 340 ms -13.44%

@gabrielmbmb gabrielmbmb requested a review from plaguss June 26, 2024 09:32
@gabrielmbmb gabrielmbmb self-assigned this Jun 26, 2024
@gabrielmbmb gabrielmbmb added the enhancement New feature or request label Jun 26, 2024
@gabrielmbmb gabrielmbmb added this to the 1.3.0 milestone Jun 26, 2024
@gabrielmbmb gabrielmbmb linked an issue Jun 26, 2024 that may be closed by this pull request
src/distilabel/pipeline/batch_manager.py Outdated Show resolved Hide resolved
src/distilabel/pipeline/batch_manager.py Outdated Show resolved Hide resolved
@gabrielmbmb gabrielmbmb merged commit e040dec into develop Jun 27, 2024
7 checks passed
@gabrielmbmb gabrielmbmb deleted the multiple-workers branch June 27, 2024 10:39
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
enhancement New feature or request
Projects
None yet
Development

Successfully merging this pull request may close these issues.

[FEATURE] Multiprocessing at Step level for Pipeline
2 participants