Skip to content

[BugFix] fix eb5 prefix bug#5879

Merged
yuanlehome merged 10 commits into
PaddlePaddle:developfrom
kevincheng2:fix_eb5_prefix_cache
Jan 7, 2026
Merged

[BugFix] fix eb5 prefix bug#5879
yuanlehome merged 10 commits into
PaddlePaddle:developfrom
kevincheng2:fix_eb5_prefix_cache

Conversation

@kevincheng2

@kevincheng2 kevincheng2 commented Jan 5, 2026

Copy link
Copy Markdown
Collaborator

Motivation

💡 If this PR is a Cherry Pick, the PR title needs to follow the format by adding the [Cherry-Pick] label at the very beginning and appending the original PR ID at the end. For example, [Cherry-Pick][CI] Add check trigger and logic(#5191)

💡 如若此PR是Cherry Pick,PR标题需遵循格式,在最开始加上[Cherry-Pick]标签,以及最后面加上原PR ID,例如[Cherry-Pick][CI] Add check trigger and logic(#5191)

当上下文缓存完全命中时,会revert 一个block的token,保证输入非空。但在 eb5 多模场景下,会导致输入缺少多模信息

Modifications

  1. eb5 多模场景完全命中时,提前revert 掉一个block的token,再去执行 revert_chunked_mm_input,避免命中部分图片
  2. 将 prefix_cache_manager.py 中的 revert 逻辑移动到 resource_manager_v1.py 中,减少和匹配逻辑的耦合

Usage or Command

Accuracy Tests

Checklist

  • Add at least a tag in the PR title.
    • Tag list: [[FDConfig],[APIServer],[Engine], [Scheduler], [PD Disaggregation], [Executor], [Graph Optimization], [Speculative Decoding], [RL], [Models], [Quantization], [Loader], [OP], [KVCache], [DataProcessor], [BugFix], [Docs], [CI], [Optimization], [Feature], [Benchmark], [Others], [XPU], [HPU], [GCU], [DCU], [Iluvatar], [Metax]]
    • You can add new tags based on the PR content, but the semantics must be clear.
  • Format your code, run pre-commit before commit.
  • Add unit tests. Please write the reason in this PR if no unit tests.
  • Provide accuracy results.
  • If the current PR is submitting to the release branch, make sure the PR has been submitted to the develop branch, then cherry-pick it to the release branch with the [Cherry-Pick] PR tag.

@paddle-bot

paddle-bot Bot commented Jan 5, 2026

Copy link
Copy Markdown

Thanks for your contribution!

@codecov-commenter

codecov-commenter commented Jan 5, 2026

Copy link
Copy Markdown

Codecov Report

❌ Patch coverage is 50.00000% with 10 lines in your changes missing coverage. Please review.
⚠️ Please upload report for BASE (develop@7ad5737). Learn more about missing BASE report.

Files with missing lines Patch % Lines
fastdeploy/engine/sched/resource_manager_v1.py 50.00% 10 Missing ⚠️
Additional details and impacted files
@@            Coverage Diff             @@
##             develop    #5879   +/-   ##
==========================================
  Coverage           ?   66.63%           
==========================================
  Files              ?      347           
  Lines              ?    44421           
  Branches           ?     6825           
==========================================
  Hits               ?    29599           
  Misses             ?    12640           
  Partials           ?     2182           
Flag Coverage Δ
GPU 66.63% <50.00%> (?)

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:
  • ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

Comment thread fastdeploy/cache_manager/prefix_cache_manager.py Outdated

@rainyfly rainyfly left a comment

Copy link
Copy Markdown
Collaborator

Choose a reason for hiding this comment

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

如果是从前缀树里把这个 node 给摘除,是否需要加一个 从node.parent的孩子节点里移除

@kevincheng2

Copy link
Copy Markdown
Collaborator Author

如果是从前缀树里把这个 node 给摘除,是否需要加一个 从node.parent的孩子节点里移除

建树的时候,增加了判断是否已存在相同hash的block,如果有:保留旧block,跳过创建新block

yuanlehome pushed a commit that referenced this pull request Jan 7, 2026
* fix eb5 prefix bug

* update code

* update code

* update code

* update code
@yuanlehome yuanlehome merged commit eabd01c into PaddlePaddle:develop Jan 7, 2026
16 of 20 checks passed
@kevincheng2 kevincheng2 deleted the fix_eb5_prefix_cache branch January 19, 2026 03:42
chang-wenbin pushed a commit to chang-wenbin/FastDeploy that referenced this pull request Mar 2, 2026
* fix eb5 prefix bug

* update ci test

* update code

* update code

* update code

* update code

* update code

* update code

* update code
xiaoguoguo626807 pushed a commit to xiaoguoguo626807/FastDeploy that referenced this pull request May 7, 2026
* fix eb5 prefix bug

* update ci test

* update code

* update code

* update code

* update code

* update code

* update code

* update code
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.

4 participants