ggml-webgpu: address quantization precision and backend lifecycle managment#21521

Merged

reeselevine merged 26 commits intoggml-org:masterfrom

noumena-labs:master

Apr 10, 2026

Commits on Mar 23, 2026

ggml(webgpu): fix the busy-polls in Emscripten in the waitAny after #20618, and remove the busy webgpu log
Constannnnnt
committed

Commits on Apr 2, 2026

Merge branch 'master' of github.com:ggml-org/llama.cpp into constant/dev2
jjhartmann
committed

Commits on Apr 3, 2026

Commits on Apr 4, 2026

Fix GET_ROWS packed integer NaN when using f16 as memory buffer in shader quants
jjhartmann
committed

Commits on Apr 6, 2026

merge: merge with upstream master
Constannnnnt
committed

Commits on Apr 7, 2026

Commits on Apr 9, 2026