search

duangenquan / YoloV2NCS

This project shows how to run tiny yolo v2 with movidius stick.
148 stars 63 forks source link

create_yolo_caffemodel.py error #33

Open sansoy opened 5 years ago

sansoy commented 5 years ago

I just checked out your repository and ran the shell script convertyo.sh using your example yolo .cfg & .weight files.

python create_yolo_caffemodel.py -m tiny-yolo-voc.prototxt -w tiny-yolo-voc.weights -o tiny-yolo-voc.caffemodel

I get the following error.

I1230 00:15:40.790437 11577 net.cpp:264] This network produces output conv9 I1230 00:15:40.790475 11577 net.cpp:284] Network initialization done. False (15867885,)

Total Net Layer 40

conv1(conv)-1-False bn1(batchnorm) Traceback (most recent call last): File "../python/create_yolo_caffemodel.py", line 139, in main(sys.argv[1:]) File "../python/create_yolo_caffemodel.py", line 121, in main net.params[pr][0].data[...] = batch_norm[1] # mean ValueError: could not broadcast input array from shape (16) into shape (1,16,1,1)

any ideas?