Skip to content
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

[Inference] Generate multiple sequences for single prompt #52

Closed
xwu99 opened this issue Jan 11, 2024 · 1 comment · Fixed by #62
Closed

[Inference] Generate multiple sequences for single prompt #52

xwu99 opened this issue Jan 11, 2024 · 1 comment · Fixed by #62

Comments

@xwu99
Copy link
Contributor

xwu99 commented Jan 11, 2024

No description provided.

@xwu99 xwu99 changed the title [Inference] generate multiple sequences for single prompt [Inference] Generate multiple sequences for single prompt Jan 11, 2024
@xwu99 xwu99 mentioned this issue Jan 17, 2024
Merged
@xwu99
Copy link
Contributor Author

xwu99 commented Jan 17, 2024 •
edited
Loading

@KepingYan could you update this to streamline vllm output after #20 is merged. There are two options:

Option 1:

  • request include single prompt: return single output
  • request include multiple prompts: return outputs as list

Option 2:

  • request include single prompt or multiple prompts both return output(s) as list.

Could you check how OpenAI deal with multiple prompts and we can align with it.?

@xwu99 xwu99 linked a pull request Jan 29, 2024 that will close this issue
[Serving] Support multiple prompts for generate #62
Merged
@xwu99 xwu99 closed this as completed Jan 29, 2024
zhangjian94cn pushed a commit to zhangjian94cn/llm-on-ray that referenced this issue Feb 4, 2024
@kira-lin
Refactor MegatronDataset and processer to add eval dataloader for Meg…
6c755f4 
…atron DDP pretrainers (intel#52)

* add eval loader

Signed-off-by: Zhi Lin <[email protected]>

* use step saved in checkpoint in megatron dataset

Signed-off-by: Zhi Lin <[email protected]>

---------

Signed-off-by: Zhi Lin <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

[Serving] Support multiple prompts for generate KepingYan/llm-on-ray
1 participant
@xwu99