rednote-machine-learning /

RedKnot

Efficient Long-Context LLM Serving with Head-Aware KV Reuse and SegPagedAttention

#24 in LLM serving Python 3,055 stars

Updated 42 min ago

View on GitHub

Why now

+328 stars gained this week

3× its usual pace

Stars gained each day, last 30 days

Usual pace · 15 / day

Oct 4Oct 19Today

This week Before Usual pace

In short

The README is mostly about a development branch and benchmark harness; it notes that GPU execution targets NVIDIA SM100, with an H200/SM90 port selected via a runtime-root flag, and that the optimized native execution is still pending.

Summary written by AI from the repository’s description and README.

rednote-machine-learning / RedKnot