llama.cpp
convert : force f16 or f32 on step3-vl conv weights
#21646

Merged

convert : force f16 or f32 on step3-vl conv weights #21646

CISC merged 1 commit into master from cisc/convert-step3vl-conv-tensors-fix

force f16 or f32 on step3-vl conv weights

f5059efa

CISC requested a review from

ngxson 54 days ago

ngxson approved these changes on 2026-04-12

ServeurpersoCom approved these changes on 2026-04-12

CISC merged 1e9d771e into master 50 days ago

CISC deleted the cisc/convert-step3vl-conv-tensors-fix branch 50 days ago

Reviewers

ServeurpersoCom

ngxson

Assignees

No one assigned

Labels

None yet

Milestone

No milestone