wangbo/ComfyUI
1
0
Fork 0
mirror of https://github.com/comfyanonymous/ComfyUI.git synced 2025-12-23 13:00:54 +08:00
Code Issues Actions 9 Packages Projects Releases Wiki Activity

fix: accelerate the first inference speed on low-level NPUs.

Browse Source
This commit is contained in:
muxuezzz 2025-09-27 15:23:09 +08:00
parent a9cf1cd249
commit 083d8aa330
1 changed files with 1 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
comfy/model_management.py
View File

@ -121,6 +121,7 @@ except:
try: try:
import torch_npu # noqa: F401 import torch_npu # noqa: F401
_ = torch.npu.device_count() _ = torch.npu.device_count()
torch_npu.npu.set_compile_mode(jit_compile = False)
npu_available = torch.npu.is_available() npu_available = torch.npu.is_available()
except: except:
npu_available = False npu_available = False