Skip to content

Conversation

danielhua23
Copy link
Contributor

@danielhua23 danielhua23 commented Aug 20, 2025

This PR included reference.py and submission.py and task description

I initially added some shapes which all from popular models like deepseek series, gpt-oss, mixtral. You can check if they work with the reference.py and submission.py on your end.

Thanks

cfg, all_rank_data = data
world_size = 8

mp.set_start_method("spawn", force=True)
Copy link
Member

Choose a reason for hiding this comment

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

we were doing this on our end, I'm still unsure whether we should do this or make users do this. It seems "fine" if we can just provide a helper or copy pastable code in template for folks -cc @ngc92

# num experts per rank
self.num_local_experts = cfg.num_experts // world_size
# max recv tokens per rank
self.max_recv = cfg.max_num_tokens * cfg.experts_per_token
Copy link
Member

Choose a reason for hiding this comment

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

n00b q is reasonable loading balancing guaranteed?

@S1ro1 S1ro1 merged commit a38f4e8 into gpu-mode:main Aug 31, 2025
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.

3 participants