[GH-ISSUE #336] How to reproduce deepseek R1-671b model performance on AIME2024 #260

Closed
opened 2026-10-26 21:18:06 +00:00 by bsakaguchi · 3 comments
Owner

Originally created by @xiaoycolor on GitHub (Feb 8, 2025).
Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-R1/issues/336

I am depoying deepseek R1-671b model on 2nodes(80G H100*8,currently, I can run sucfessfully model and get correct inference result,i plan to reproduce performance on AIME2024,but i dont know how to reproduce it,so whether there is guide for this,anyone can help me,thanks a lot!

Originally created by @xiaoycolor on GitHub (Feb 8, 2025). Original GitHub issue: https://github.com/deepseek-ai/DeepSeek-R1/issues/336 I am depoying deepseek R1-671b model on 2*nodes*(80G H100*8,currently, I can run sucfessfully model and get correct inference result,i plan to reproduce performance on AIME2024,but i dont know how to reproduce it,so whether there is guide for this,anyone can help me,thanks a lot!
bsakaguchi 2026-10-26 21:18:06 +00:00
Author
Owner

@zijianh4 commented on GitHub (Feb 20, 2025):

Same problem, did you solve it?

<!-- gh-comment-id:2670515251 --> @zijianh4 commented on GitHub (Feb 20, 2025): Same problem, did you solve it?
Author
Owner

@github-actions[bot] commented on GitHub (Mar 26, 2025):

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!

<!-- gh-comment-id:2752808552 --> @github-actions[bot] commented on GitHub (Mar 26, 2025): This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. If you believe this issue is still relevant, please leave a comment to keep it open. Thank you for your contributions!
Author
Owner

@github-actions[bot] commented on GitHub (May 1, 2025):

false

<!-- gh-comment-id:2843772290 --> @github-actions[bot] commented on GitHub (May 1, 2025): false
Sign in to join this conversation.
No milestone
No project
No assignees
1 participant
Notifications
Due date
The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference
deepseek-ai/DeepSeek-R1#260
No description provided.