llama.cpp
convert : force patch_embd weights to F16 or F32 to avoid broken GGUFs
#15367
Merged
Overview
Commits
2
Changes
View On GitHub

convert : force patch_embd weights to F16 or F32 to avoid broken GGUFs #15367

CISC merged 2 commits into master from cisc/fix-convert-mmproj-patch-embd
CISC
CISC force patch_embd weights to f32
fe625c62
CISC CISC requested a review from ngxson ngxson 159 days ago
github-actions github-actions added python
ngxson
ngxson commented on 2025-08-17
CISC use MmprojModel base tensor_force_quant instead
28722d16
CISC CISC requested a review from ngxson ngxson 159 days ago
CISC CISC changed the title convert : force patch_embd weights to F32 to avoid broken GGUFs convert : force patch_embd weights to F16 or F32 to avoid broken GGUFs 159 days ago
ngxson
CISC
ngxson
CISC
ngxson
ngxson
ngxson approved these changes on 2025-08-17
CISC CISC merged 4d196981 into master 159 days ago
CISC CISC deleted the cisc/fix-convert-mmproj-patch-embd branch 159 days ago

Login to write a write a comment.

Login via GitHub

Reviewers
Assignees
No one assigned
Labels
python
Milestone
No milestone
Terms Of Use Privacy Policy Refund Policy Impressum