llama.cpp
webgpu : use f32 while dequantizing to avoid FTZ
#32140
Merged

webgpu : use f32 while dequantizing to avoid FTZ #32140

ggerganov merged 1 commit into ggml-org:master from tokinasin:fix/arc-ftz
tokinasin
tokinasin tokinasin requested a review 6 days ago
github-actions github-actions added ggml
github-actions github-actions added WebGPU
yomaytk
yomaytk approved these changes on 2026-11-15
yomaytk
yomaytk
tokinasin webgpu : use f32 while dequantizing to avoid FTZ
6bc23bba
tokinasin tokinasin force pushed from 749de61c to 6bc23bba 2 days ago
yomaytk
yomaytk approved these changes on 2026-11-16
yomaytk yomaytk added merge ready
ggerganov ggerganov merged c145db27 into master 1 day ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
Milestone