ai-models/WanVideo_comfy_fp8_scaled
2
0
Fork 0
mirror of https://www.modelscope.cn/Kijai/WanVideo_comfy_fp8_scaled.git synced 2026-04-02 14:02:53 +08:00
Code Issues Packages Projects Releases Wiki Activity

Create README.md

Browse Source
This commit is contained in:
Cherrytest
2025-10-06 16:17:10 +00:00
parent 806c9b24a1
commit 94b573b0bd
1 changed files with 10 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
Wan22Animate/README.md Normal file
View File

@ -0,0 +1,10 @@
# WanAnimate fp8_scaled models
# v2 bugfix update for native workflows
The initial uploads has the face encoder layers in bf16, as back then I was unsure if they can be quantized.
When running these models in the WanVideoWrapper those layers are used in the base_precision, but when running these
in native ComfyUI, they get downscaled to normal fp8, which seems to introduce pretty bad noise in the outputs in a grid pattern.
The v2 update addresses this by using fp8 scaled on the face encoder layers as well. For the wrapper you can keep using the initial models.