xinyun/ComfyUI
1
0
Fork 0
mirror of https://git.datalinker.icu/comfyanonymous/ComfyUI synced 2026-08-14 14:30:11 +08:00
Code Issues Packages Projects Releases Wiki Activity
ComfyUI/comfy/ldm/wan
History
Rattus 39eb74cabe wan: Optimize qkv intermediates on attention 
As commented. The former logic computed independent pieces of QKV in
parallel which help more inference intermediates in VRAM spiking
VRAM usage. Fully roping Q and garbage collecting the intermediates
before touching K reduces the peak inference VRAM usage.
2025-09-16 22:54:01 +10:00
..
model.py
wan: Optimize qkv intermediates on attention
2025-09-16 22:54:01 +10:00
vae2_2.py
Tiny wan vae optimizations. (#9136)
2025-08-01 05:25:38 -04:00
vae.py
Tiny wan vae optimizations. (#9136)
2025-08-01 05:25:38 -04:00