llama.cpp
ggml: skip 0-sized ids tensor when offloading selected experts
#28739

Merged

ggml: skip 0-sized ids tensor when offloading selected experts #28739

pwilkin merged 1 commit into master from aman/fix-offload-bug

ggml: skip 0-sized ids tensor when offloading selected experts

3c640992

github-actions added ggml

ggerganov approved these changes on 2026-09-11

am17an added merge ready

pwilkin approved these changes on 2026-09-11

pwilkin merged 43f3dda6 into master 11 days ago

Reviewers

pwilkin

ggerganov

Assignees

No one assigned

Labels

ggml merge ready

Milestone

No milestone