Lightning-AI / Lightning-AI/lightning-thunder

Support the `trtllm autodeploy` flashinfer KV-Cached attention in Thunder

Open
#2,472 0 comments 0 reactions 1 assignee View on GitHub

@kiya00 is already working on this.

Since Aug 26, 2025.

enhancement
Dominant language
Python
Stars
1.5k
Forks
121
PR merge metrics
No merged PRs in 30d

Description

🚀 Feature

This issue tracks the progress of supporting Flashinfer KV-cached attention that used in TRTLLM autodeploy in Thunder.

We should implement a transformation that replaces the original SDPA with a KV-cached attention.
Additionally, we need to design a way to pass the KV-cache and other required information—such as page numbers—into the computation graph.

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.