Skip to content

[ROCm][CI] Move Model Executor test step from MI250 to MI300 (gfx942) - #44370

Merged
AndreasKaratzas merged 2 commits into
vllm-project:mainfrom
JartX:ci/mi250-model-executor-parallelism
Jun 3, 2026
Merged

[ROCm][CI] Move Model Executor test step from MI250 to MI300 (gfx942)#44370
AndreasKaratzas merged 2 commits into
vllm-project:mainfrom
JartX:ci/mi250-model-executor-parallelism

Conversation

@JartX

@JartX JartX commented Jun 2, 2026

Copy link
Copy Markdown
Contributor

Set parallelism: 4 and shard with --shard-id/--num-shards (same pattern as LoRA %N). Gate the single-file tensorizer test to shard 0.

@AndreasKaratzas

@JartX
JartX requested review from Harry-Chen and khluu as code owners June 2, 2026 23:33
@JartX JartX changed the title [ROCM][CI] Shard MI250 Model Executor step across 4 jobs [ROCm][CI] Shard MI250 Model Executor step across 4 jobs Jun 2, 2026
@mergify mergify Bot added ci/build rocm Related to AMD ROCm labels Jun 2, 2026
@github-project-automation github-project-automation Bot moved this to Todo in AMD Jun 2, 2026

@AndreasKaratzas AndreasKaratzas left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

Thank you for the contribution, I think that the problem is different regarding the failure of this test group. Would be happy to merge this as soon as the review point has been addressed.

Comment thread .buildkite/test-amd.yaml Outdated
The Model Executor step's failures come from quantized (fp8) tests that
are unsupported on MI250 (gfx90a), and the step also takes ~38 min there.
Reassign the whole step to MI300 (gfx942, agent_pool mi300_1), where fp8
is natively supported, instead of sharding it on MI250.

Addresses review feedback from @AndreasKaratzas.

Signed-off-by: JartX <sagformas@epdcenter.es>
@JartX
JartX force-pushed the ci/mi250-model-executor-parallelism branch from 4045e46 to 9833177 Compare June 3, 2026 07:12
@JartX JartX changed the title [ROCm][CI] Shard MI250 Model Executor step across 4 jobs [ROCm][CI] Move Model Executor test step from MI250 to MI300 (gfx942) Jun 3, 2026

@AndreasKaratzas AndreasKaratzas left a comment

Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

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

LGTM

@AndreasKaratzas AndreasKaratzas added the ready ONLY add when PR is ready to merge/full CI is needed label Jun 3, 2026
@AndreasKaratzas
AndreasKaratzas merged commit 5b2a2be into vllm-project:main Jun 3, 2026
15 checks passed
@github-project-automation github-project-automation Bot moved this from Todo to Done in AMD Jun 3, 2026
mvanhorn pushed a commit to mvanhorn/vllm that referenced this pull request Jun 4, 2026
…vllm-project#44370)

Signed-off-by: JartX <sagformas@epdcenter.es>
Signed-off-by: Matt Van Horn <455140+mvanhorn@users.noreply.github.com>
JisoLya pushed a commit to JisoLya/vllm that referenced this pull request Jun 5, 2026
…vllm-project#44370)

Signed-off-by: JartX <sagformas@epdcenter.es>
Signed-off-by: JisoLya <523420504@qq.com>
knight0528 pushed a commit to knight0528/vllm that referenced this pull request Jun 8, 2026
waqahmed-amd-fi pushed a commit to waqahmed-amd-fi/vllm that referenced this pull request Jun 10, 2026
…vllm-project#44370)

Signed-off-by: JartX <sagformas@epdcenter.es>
Signed-off-by: Waqar Ahmed <waqar.ahmed@amd.com>
Saddss pushed a commit to Saddss/vllm that referenced this pull request Jun 14, 2026
divineearthly pushed a commit to divineearthly/vllm that referenced this pull request Jun 19, 2026
…vllm-project#44370)

Signed-off-by: JartX <sagformas@epdcenter.es>
Signed-off-by: divineearthly <divineearthly@gmail.com>
nkzhenhua pushed a commit to nkzhenhua/vllm that referenced this pull request Jun 24, 2026
philippesic pushed a commit to philippesic/vllm-semantic-cache that referenced this pull request Jul 19, 2026
plasticchris pushed a commit to plasticchris/vllm that referenced this pull request Jul 20, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

ci/build ready ONLY add when PR is ready to merge/full CI is needed rocm Related to AMD ROCm

Projects

Status: Done

Development

Successfully merging this pull request may close these issues.

2 participants