Skip to content

[model] fix: apply attention_value_scale to V in MiMo-V2-Flash#4155

Open
Eisenhower wants to merge 4 commits into
NVIDIA-NeMo:mainfrom
Eisenhower:fix/mimo-v2-flash-value-scale
Open

[model] fix: apply attention_value_scale to V in MiMo-V2-Flash#4155
Eisenhower wants to merge 4 commits into
NVIDIA-NeMo:mainfrom
Eisenhower:fix/mimo-v2-flash-value-scale

test: rewrite attention_value_scale UTs with real GPU forward pass

650e37b
Select commit
Loading
Failed to load commit list.
DCO / DCO succeeded Jun 16, 2026 in 0s

DCO

All commits are signed off!