Skip to content

Conversation

@njhill
Copy link
Member

@njhill njhill commented Apr 25, 2024

As a superclass of RayGPUExecutor, in preparation for introducing a sibling MultiprocessingGPUExecutor.

There is no new logic here, it's just moving some methods into a the superclass.

As a superclass of RayGPUExecutor, in preparation for introducing a sibling MultiprocessingGPUExecutor
Copy link
Member

@youkaichao youkaichao left a comment

Choose a reason for hiding this comment

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

LGTM, please take a look at the review.

@youkaichao
Copy link
Member

Another concern is the naming. I suggest something like DistributedExecutor , distributed is better than multi in my opinion.

@njhill
Copy link
Member Author

njhill commented Apr 25, 2024

@youkaichao so you think DistributedGPUExectuor rather than MutliGPUExecutor?

@youkaichao
Copy link
Member

Yes, I think DistributedGPUExectuor is good.

@njhill njhill changed the title [Core] Introduce MultiGPUExecutor abstract class [Core] Introduce DistributedGPUExecutor abstract class Apr 25, 2024
@njhill njhill enabled auto-merge (squash) April 25, 2024 20:11
@njhill njhill merged commit 258a2c5 into vllm-project:main Apr 27, 2024
@njhill njhill deleted the gpu_executor_hierarchy branch April 27, 2024 16:48
robertgshaw2-redhat pushed a commit to neuralmagic/nm-vllm that referenced this pull request May 6, 2024
z103cb pushed a commit to z103cb/opendatahub_vllm that referenced this pull request May 7, 2024
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.

2 participants