Skip to content

[devops] add large-scale distributed test marker #4451

Closed

Description

@ver217

Overview

Hybrid parallelism need at lease 8 processes to be fully tested. Our CI runner is run on a 4-gpu node and a 8-gpu node.

As the 8-gpu node is not always available, we run the most tests on 4-gpu node. And for those need 8-gpu, we run them regularly on 8-gpu node.

In short, we should add a marker to indicate which test should be run on 8-gpu node.

Activity

  1. converted this from a draft issue on Aug 16, 2023
  2. self-assigned this
    on Aug 16, 2023
  3. JazyJiang commented on Oct 16, 2026

    @JazyJiang
    Collaborator

    Closing as completed. The largedist marker is defined in pytest.ini ("at least 8 GPUs") and is used by the 8-GPU tests. The new three-layer CI (#6462) runs them as a dedicated batch: batch 10 selects -m largedist, while the 4-GPU batches use -m "not largedist". Batch membership is defined in .github/scripts/ci/run_colossalai_batch.sh.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Labels

No labels
No labels

Type

No type

Projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions