xinyun/ComfyUI
1
0
Fork 0
mirror of https://git.datalinker.icu/comfyanonymous/ComfyUI synced 2026-08-16 04:43:30 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fix Hunyuan unet config detection for some models. (#6877)

Browse Source 
The change to support 32 channel hunyuan models is missing the `key_prefix` on the key.

This addresses a complain in the comments of 0b5852ed5dd28fdf2a3efd2b670c871d24deb7f9.
This commit is contained in:
maedtb 2025-02-19 07:14:45 -05:00 committed by GitHub
parent 65a8e324bd
commit 50912cb296
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/model_detection.py
View File

@ -136,7 +136,7 @@ def detect_unet_config(state_dict, key_prefix):
if '{}txt_in.individual_token_refiner.blocks.0.norm1.weight'.format(key_prefix) in state_dict_keys: #Hunyuan Video
dit_config = {}
dit_config["image_model"] = "hunyuan_video"
dit_config["in_channels"] = state_dict["img_in.proj.weight"].shape[1] #SkyReels img2video has 32 input channels
dit_config["in_channels"] = state_dict['{}img_in.proj.weight'.format(key_prefix)].shape[1] #SkyReels img2video has 32 input channels
dit_config["patch_size"] = [1, 2, 2]
dit_config["out_channels"] = 16
dit_config["vec_in_dim"] = 768