Starstruk Logo MarkStarstruk

v0.11.2

vllm-project/vllmv0.11.2Nov 20, 2025by khluu

AI Summary

A patch release fixing critical bugs related to Ray, speculative decoding, async-scheduling, and SM100 CUTLASS MoE macro.

Key Highlights

  • Bugfix for Ray with multiple nodes
  • Fix false assertion with spec-decode=[2,4,..] and TP>2
  • Fix async-scheduling + FlashAttn MLA
  • Guard SM100 CUTLASS MoE macro to SM100 builds

New Features

  • Bugfixes for distributed inference and spec decode

Full Release Notes

This release includes 4 bug fixes on top of `v0.11.1`:
- [BugFix] Ray with multiple nodes (https://github.com/vllm-project/vllm/pull/28873)
- [BugFix] Fix false assertion with spec-decode=[2,4,..] and TP>2 (https://github.com/vllm-project/vllm/pull/29036)
- [BugFix] Fix async-scheduling + FlashAttn MLA (https://github.com/vllm-project/vllm/pull/28990)
- [NVIDIA] Guard SM100 CUTLASS MoE macro to SM100 builds v2 (https://github.com/vllm-project/vllm/pull/28938)