llama.cpp
Fix granite speech model inference by applying embedding scale when deepstack is not used
#24357
Merged
Merged
Fix granite speech model inference by applying embedding scale when deepstack is not used #24357
ngxson
merged
3 commits into
ggml-org:master
from
arnu515:master
llama-graph : apply embedding scale when deepstack is not used
c4b22de6
gabe-l-hart
approved these changes
on 2026-06-09
CISC
approved these changes
on 2026-06-09
nits: remove non-existant hunyuan-vl from the tests
4584f943
apply suggestion from @gabe-l-hart
c28b9386
ServeurpersoCom
approved these changes
on 2026-06-09
Reviewers
Assignees
No one assigned
Labels
examples
Milestone
No milestone
Login to write a write a comment.
Login via GitHub