llama.cpp
ggml webgpu: fix workgroup dispatch limit for large batch sizes
#19965
Merged
Merged
ggml webgpu: fix workgroup dispatch limit for large batch sizes #19965
reeselevine
merged
5 commits into
ggml-org:master
from
abhijitramesh:abhijit/webgpu-matmul-workgroup-limit
ggml-webgpu: fix workgroup dispatch limit for large batch sizes
b0169df5
ggml-webgpu: add bounds checking for over-dispatched workgroups
11253444
reeselevine
commented
on 2026-02-28
ggml-webgpu: add back TODO for spliting large sizes into batches
b92eee7d
Optimize 2d workgroup provisioning
57e9ed26
Set some parameters that increase speed
023b5658
reeselevine
approved these changes
on 2026-03-03
Login to write a write a comment.
Login via GitHub