nodes/chflame163-ComfyUI_LayerStyle_Advance
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Reorder execution providers in model initialization

Browse Source 
When TensorRT libraries are missing, forcibly using the TensorRT Execution Provider can crash the entire ComfyUI Python process.
This commit is contained in:
hben35096
2026-09-18 09:07:50 +08:00
committed by GitHub
parent 7b678b401a
commit 5f68ae0de7
1 changed files with 1 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
py/human_parts_ultra.py
+1 -1
View File
@@ -73,7 +73,7 @@ class LS_HumanPartsUltra:
""" """
import onnxruntime as ort import onnxruntime as ort
model = ort.InferenceSession(model_path, providers=['TensorrtExecutionProvider', 'CUDAExecutionProvider', 'CPUExecutionProvider']) model = ort.InferenceSession(model_path, providers=['CUDAExecutionProvider', 'CPUExecutionProvider', 'TensorrtExecutionProvider'])
ret_images = [] ret_images = []
ret_masks = [] ret_masks = []
for img in image: for img in image: