Stuck in "Loading Yolo model"

[PEACEandSAFE]

when run demo_inference.py , it is stuck in "Loading Yolo model" and ''loading pose model'' shows 0%.

[hsauod]

same error anyone could help? thank you.

[korin-lf]

same problem: stock on Loading pose model

[cmc370]

@korin-lf How did you solve it?

[EmbeddedAndMore]

download it from: https://pjreddie.com/media/files/yolov3-spp.weights and put it in detector/yolo/data

[WyattAutomation]

It wasn't clear on local setup instructions that vanilla yolov3 weights are a dependency--I saw this error but assumed it was a module the demo script uses that just failed to download it (not seeing any docs on it on it made me think it was a download script or path error). Sorry if it is in the docs and I just didn't see it, hopefully an easy fix.

[lawo123]

download it from: https://pjreddie.com/media/files/yolov3-spp.weights and put it in detector/yolo/data

放到下面之后会报这个错： File "/data/zhoumeng/pose/AlphaPose/detector/yolo/darknet.py", line 488, in load_weights conv_weights = conv_weights.view_as(conv.weight.data) RuntimeError: shape '[512, 256, 3, 3]' is invalid for input of size 56654

[SMY19999]

I have solve this problem by adding the root directory. Locate and Go to the code by terminal output , then add code to complete the file_path with the absolute one. cfgfile_root = '/xxx/xxx/xxx/Alphapose' file = open(os.path.join(cfgfile_root,cfgfile),'r') which can solve any problem like this.