Skip to content

Issue/168 InfiniLM接入paged attention接口 #173

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
Open
PanZezhong1725 wants to merge 5 commits into
base: main
Choose a base branch
from
Open

Issue/168 InfiniLM接入paged attention接口 #173

PanZezhong1725 wants to merge 5 commits into from

Conversation

@PanZezhong1725
Copy link
Collaborator

@PanZezhong1725 PanZezhong1725 commented Dec 30, 2025

No description provided.

@PanZezhong1725 PanZezhong1725 marked this pull request as ready for review January 8, 2026 09:14
@PanZezhong1725 PanZezhong1725 requested review from a team, spike-zhu and whjthu January 8, 2026 09:14
@PanZezhong1725
Copy link
Collaborator Author

PanZezhong1725 commented Jan 8, 2026

普通kv，9G7B 双卡 bs=4：
image

paged kv，9G7B 双卡 bs=4：
image

whjthu
whjthu requested changes Jan 8, 2026
View reviewed changes
examples/jiuge.py Outdated
max_new_tokens + len(input_ids_list[0]),
)
# ---------------------------------------------------------------------------- #
# 创建KVCache
Copy link

@whjthu whjthu Jan 8, 2026

Choose a reason for hiding this comment

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

后面所有新增代码不要再出现中文注释，之前的可以先不改，留着后面集中修改

Copy link
Collaborator Author

@PanZezhong1725 PanZezhong1725 Jan 9, 2026

Choose a reason for hiding this comment

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

已修改

@PanZezhong1725 PanZezhong1725 requested review from ma-hang and removed request for spike-zhu January 9, 2026 12:29
@PanZezhong1725
Copy link
Collaborator Author

PanZezhong1725 commented Jan 9, 2026 •
edited
Loading

最新代码依赖InfiniTensor/InfiniCore#905

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@whjthu whjthu whjthu requested changes

@ma-hang ma-hang Awaiting requested review from ma-hang

Requested changes must be addressed to merge this pull request.

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@PanZezhong1725 @whjthu