iHore/vLLM
8
0
Fork 0
Code Issues Pull Requests Actions 1 Packages Projects Releases Wiki Activity

[XPU] [Linear] add torch as xpu linear backend (#49664)

Browse Source 
Signed-off-by: Zhu, Zufang <zufang.zhu@intel.com>
Co-authored-by: Kunshang Ji <kunshang.ji@intel.com>
This commit is contained in:
zofia
2026-08-03 14:30:21 +08:00
committed by GitHub
co-authored by Kunshang Ji
parent 2755489a26
commit e481da9508
5 changed files with 15 additions and 8 deletions
Download Patch File Download Diff File Expand all files Collapse all files
docs/features/quantization/online.md
+1 -1
View File
@@ -62,7 +62,7 @@ weight name. Unset fields fall back to the `--quantization` shorthand's
defaults, or for already-quantized checkpoints to whatever the checkpoint
declares.
On XPU, non-block FP8 scaled-mm linear layers default to W8A16; setting `--linear-backend xpu` forces W8A8. Use `--linear-backend xpu_woq` to explicitly select weight-only quantization (W8A16).
On XPU, non-block FP8 scaled-mm linear layers default to W8A16; setting `--linear-backend xpu` forces W8A8. Use `--linear-backend xpu_woq` to explicitly select weight-only quantization (W8A16). Setting `--linear-backend torch` also forces W8A8 but runs the GEMM through `torch._scaled_mm` instead of the custom XPU kernel.
The CLI accepts the same shape as JSON or as dotted keys: