wangbo/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2026-07-20 13:28:19 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

fix: filter_fn returns True to quantize, not skip

Browse Source
This commit is contained in:
JWLHS 2026-07-10 12:10:46 +08:00
parent 2d0bb38f4f
commit 015e8a2676
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
comfy/quantization/torchao/quantize.py
View File

@ -24,7 +24,7 @@ def quantize_model(
model: PyTorch model to quantize. model: PyTorch model to quantize.
group_size: Quantization group size (32/64/128/256). group_size: Quantization group size (32/64/128/256).
filter_fn: Optional callable(module, name) → bool. filter_fn: Optional callable(module, name) → bool.
Return True to skip that module. Return True to quantize that module.
Returns: Returns:
Same model (quantized in-place). Same model (quantized in-place).