Revert "fix new_page_count_next_decode" #7410
Conversation
This reverts commit 27e327b.
There was a problem hiding this comment.
Summary of Changes
Hello @hnyls2002, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!
This pull request serves as a direct revert of a prior change (#6671), specifically targeting the logic for calculating KV cache page counts in the schedule_batch.py module. Its purpose is to restore the original behavior of the new_page_count_next_decode function, effectively undoing an adjustment made to sequence length calculations and removing the associated explanatory comment.
Highlights
- Revert Logic: Undid a previous modification in the
new_page_count_next_decodefunction withinschedule_batch.py. This reverts the condition for calculating KV cache page counts from(req.seqlen - 1) % page_size == 0back toreq.seqlen % page_size == 0. - Code Cleanup: Removed a two-line comment explaining the previous logic related to the
seqlen - 1adjustment, as that logic has now been reverted.
Using Gemini Code Assist
The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.
Invoking Gemini
You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.
| Feature | Command | Description |
|---|---|---|
| Code Review | /gemini review |
Performs a code review for the current pull request in its current state. |
| Pull Request Summary | /gemini summary |
Provides a summary of the current pull request in its current state. |
| Comment | @gemini-code-assist | Responds in comments when explicitly tagged, both in issue comments and review comments. |
| Help | /gemini help |
Displays a list of available commands. |
Customization
To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.
Limitations & Feedback
Gemini Code Assist is currently in preview and may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments to provide feedback.
You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.
Footnotes
-
Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩
|
@pansicheng, please give your test scenarios and reproduction scripts. After this PR, we will probably encounter a decode OOM. |
There was a problem hiding this comment.
Code Review
The code reverts a change in the new_page_count_next_decode function in schedule_batch.py. The change removes the subtraction of 1 from req.seqlen when calculating the sum, which affects how the length of a request's KV cache is determined during the decoding phase.
Reverts #6671