What's Changed
- Support Hugging Face config sources for local GGUF weights by @LxYuan0420 in #624
- Add M5 NAX prefill kernel for paged GQA attention by @WindChimeRan in #623
Full Changelog: v0.3.0.dev20260818075955...v0.3.0.dev20260819071047
Full Changelog: v0.3.0.dev20260818075955...v0.3.0.dev20260819071047
Don't miss a new vllm-metal release
NewReleases is sending notifications on new releases.