CUDA, Fix CI: split the PAD_REFLECT_1D grid to stay within the HIP grid limit #32179
CUDA: split the PAD_REFLECT_1D grid to stay within the HIP grid limit
ff0e760c
danbev
approved these changes
on 2026-11-13
am17an
approved these changes
on 2026-11-13
IMbackK
approved these changes
on 2026-11-13
ggerganov
merged
2d8a3250
into master 1 day ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub