Skip to content

fix: skip SP for gemma and qwen plans #859

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
Draft
akoumpa wants to merge 3 commits into
base: main
Choose a base branch
from
Draft

fix: skip SP for gemma and qwen plans #859

akoumpa wants to merge 3 commits into from

Conversation

@akoumpa
Copy link
Contributor

@akoumpa akoumpa commented Nov 21, 2025

No description provided.

@akoumpa
skip SP for gemma and qwen plans
830b528 
Signed-off-by: Alexandros Koumparoulis <[email protected]>
@copy-pr-bot
Copy link

copy-pr-bot bot commented Nov 21, 2025

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@akoumpa akoumpa linked an issue Nov 21, 2025 that may be closed by this pull request
Qwen parallelizer with sequence parallelism #652
Open
@akoumpa
Copy link
Contributor Author

akoumpa commented Nov 21, 2025

/ok to test 830b528

@akoumpa akoumpa added the r0.2.0 Add for cherry-pick into release branch 0.2.0 label Nov 21, 2025
@ZhiyuLi-Nvidia
Copy link
Contributor

ZhiyuLi-Nvidia commented Nov 21, 2025

@akoumpa

As discussed offline:

  • qwen3(Qwen3ForCausalLM) --> fails with SP
  • qwen2(Qwen2ForCausalLM) --> works with SP

We can either keep qwen2 or bypass the following functional test
https://github.com/NVIDIA-NeMo/Automodel/blob/main/tests/functional_tests/hf_transformer_finetune/L2_HF_Transformer_PEFT_Benchmark_qwen2_custom.sh

@akoumpa
skip tests
85c188e 
Signed-off-by: Alexandros Koumparoulis <[email protected]>
@akoumpa akoumpa removed the r0.2.0 Add for cherry-pick into release branch 0.2.0 label Nov 21, 2025
@akoumpa
skip sp plans
229bff5 
Signed-off-by: Alexandros Koumparoulis <[email protected]>
@akoumpa
Copy link
Contributor Author

akoumpa commented Nov 21, 2025

/ok to test 229bff5

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

Reviewers

@HuiyingLi HuiyingLi Awaiting requested review from HuiyingLi HuiyingLi will be requested when the pull request is marked ready for review HuiyingLi is a code owner

@adil-a adil-a Awaiting requested review from adil-a adil-a will be requested when the pull request is marked ready for review adil-a is a code owner

@hemildesai hemildesai Awaiting requested review from hemildesai hemildesai will be requested when the pull request is marked ready for review hemildesai is a code owner

@ybabakhin ybabakhin Awaiting requested review from ybabakhin ybabakhin will be requested when the pull request is marked ready for review ybabakhin is a code owner

@shan-nvidia shan-nvidia Awaiting requested review from shan-nvidia shan-nvidia will be requested when the pull request is marked ready for review shan-nvidia is a code owner

At least 1 approving review is required to merge this pull request.

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Qwen parallelizer with sequence parallelism

3 participants

@akoumpa @ZhiyuLi-Nvidia