search

cfzd / Ultra-Fast-Lane-Detection

Ultra Fast Structure-aware Deep Lane Detection (ECCV 2020)
MIT License
1.76k stars 491 forks source link

Run in CPU without NVIDIA+CUDDA #337

Open jdanielhoyos opened 1 year ago

jdanielhoyos commented 1 year ago

Hello everyone, thanks beforehand When I run python demo.py configs/tusimple.py --test_model path_to_tusimple_18.pth

I got this error

AssertionError("Torch not compiled with CUDA enabled") AssertionError: Torch not compiled with CUDA enabled

I dont have a nvidia gpu so thats why I dont have CUDA. How can I run a pretrained network using CPU? Thanks

cfzd commented 1 year ago

@jdanielhoyos You might need to cancel all the .cuda() calls to make your model and data runs in cpu. Sorry I can't provide all the locations of .cuda(), but I can give you some example places: https://github.com/cfzd/Ultra-Fast-Lane-Detection/blob/353df107756b8c03c22c27201e33fc63d84ecfe6/demo.py#L28 https://github.com/cfzd/Ultra-Fast-Lane-Detection/blob/353df107756b8c03c22c27201e33fc63d84ecfe6/demo.py#L65