Repository navigation

Do not delegate quantized linears that have no Vulkan kernel (#24971) - #24971

Merged

psiddh merged 3 commits into

pytorch:mainfrom

psiddh:export-D131277775

Nov 14, 2026

Merged

Do not delegate quantized linears that have no Vulkan kernel (#24971)#24971
psiddh merged 3 commits into
pytorch:mainfrom
psiddh:export-D131277775

Commits on Nov 14, 2026
- Phonon: streaming output and per-backend weight quantization (#24929)
  
  psiddh
  authored and
  facebook-github-bot
  committed
  8f30788
- Fix the linear_q4gsw reference implementation for output channels not a multiple of 4 (#24930)
  
  psiddh
  authored and
  facebook-github-bot
  committed
  47de901
- Do not delegate quantized linears that have no Vulkan kernel (#24971)
  
  psiddh
  authored and
  facebook-github-bot
  committed
  003c440