Repository navigation

FoldQuant W8A8/W4A4 inference and FoldQuantVLA quantized-model converter - #33

Merged

hungho77 merged 12 commits into

mainfrom

feat/quantized-inference

Oct 5, 2026

Commits on Sep 30, 2026
- FoldQuant W8A8/W4A4 inference and FoldQuantVLA quantized-model converter
  hungho77
  committed
  16c0ce1
Commits on Oct 2, 2026
Commits on Oct 3, 2026
- CI: build and test the SYCL and OpenVINO backends
  hungho77
  committed
  48fca62
- docs: FoldQuant W8A8 numbers from a calibrated FoldQuantVLA checkpoint
  hungho77
  committed
  b32ecda
Commits on Oct 5, 2026
- CI fixes for the SYCL and OpenVINO jobs
  hungho77
  committed
  870786b
- FoldQuant SYCL: one device image per kernel
  hungho77
  committed
  bea4698
- FoldQuant SYCL: VLA_FQ_CHECK re-checks every node in a model run
  hungho77
  committed
  68c1f81