Skip to content

[feat] add memory saver support to sglang rollout#756

Merged
eric-haibin-lin merged 11 commits intoverl-project:mainfrom
SwordFaith:feat/sglang_with_enable_memory_saver
Mar 30, 2025
Merged

[feat] add memory saver support to sglang rollout#756
eric-haibin-lin merged 11 commits intoverl-project:mainfrom
SwordFaith:feat/sglang_with_enable_memory_saver

Conversation

@SwordFaith
Copy link
Copy Markdown
Collaborator

as title

@SwordFaith SwordFaith changed the title Add memory saver support to sglang rollout [feat] add memory saver support to sglang rollout Mar 25, 2025
Copy link
Copy Markdown
Collaborator

@eric-haibin-lin eric-haibin-lin left a comment

Choose a reason for hiding this comment

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

could u check the test failure?

@SwordFaith
Copy link
Copy Markdown
Collaborator Author

SwordFaith commented Mar 26, 2025

could u check the test failure?

It seems due to lack of new dependency torch-memory-saver>=0.0.3 in the docker. I think a possible way is add it in installation phase due to current --no-deps setting

@ocss884
Copy link
Copy Markdown
Collaborator

ocss884 commented Mar 27, 2025

Please set memory saver to >=0.0.5, we have a update yesterday. Otherwise the installation may fail.

@ocss884
Copy link
Copy Markdown
Collaborator

ocss884 commented Mar 27, 2025

could u check the test failure?

It seems due to lack of new dependency torch-memory-saver>=0.0.3 in the docker. I think a possible way is add it in installation phase due to current --no-deps setting

I will update the image today

@CLAassistant
Copy link
Copy Markdown

CLAassistant commented Mar 27, 2025

CLA assistant check
All committers have signed the CLA.

@ocss884
Copy link
Copy Markdown
Collaborator

ocss884 commented Mar 29, 2025

@vermouth1992 @PeterSH6 Hi team! I think we have the memory saver for SGLang ready to merge. Could you help trigger the CI? cc @zhaochenyang20


logger = init_logger(__name__)
"""
This version is strongly tied with Megatron to implement HybridEngine and weight sharing between vllm and Megatron.
Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

BTW i think the documentation requires updates too~
But it's okay to do it in upcoming PRs

@eric-haibin-lin eric-haibin-lin merged commit 5138a22 into verl-project:main Mar 30, 2025
22 checks passed
@zhaochenyang20
Copy link
Copy Markdown
Collaborator

thanks!

@eric-haibin-lin eric-haibin-lin mentioned this pull request Mar 30, 2025
eric-haibin-lin pushed a commit that referenced this pull request Apr 2, 2025
#756)

as title

---------

Co-authored-by: ocss884 <ocss.lin@gmail.com>
@SwordFaith SwordFaith deleted the feat/sglang_with_enable_memory_saver branch April 19, 2025 03:41
yuchenwang3 pushed a commit to yuchenwang3/verl that referenced this pull request Apr 25, 2025
verl-project#756)

as title

---------

Co-authored-by: ocss884 <ocss.lin@gmail.com>
histmeisah pushed a commit to SJTU-IAAR/verl that referenced this pull request Apr 27, 2025
verl-project#756)

as title

---------

Co-authored-by: ocss884 <ocss.lin@gmail.com>
chenjiaoAngel added a commit to chenjiaoAngel/verl that referenced this pull request Nov 14, 2025
verl-project#756)

as title

---------

Co-authored-by: ocss884 <ocss.lin@gmail.com>
TimurTaepov pushed a commit to giorgossideris/verl that referenced this pull request Dec 20, 2025
verl-project#756)

as title

---------

Co-authored-by: ocss884 <ocss.lin@gmail.com>
vyomakesh0728 added a commit to vyomakesh0728/verl that referenced this pull request Jan 22, 2026
verl-project#756)

as title

---------

Co-authored-by: ocss884 <ocss.lin@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants