It seems that the python-function "darknet.load_network()" returns an integer instead of a tuple, which causes basically all python files in src-python to fail. Example from running darknet_images.py:
0 : compute_capability = 750, cudnn_half = 1, GPU: Tesla T4
Allocating workspace to transfer between CPU and GPU: 50.0 MiB
Traceback (most recent call last):
File "/home/bgd/GIT/darknet/src-python/darknet_images.py", line 442, in
main()
File "/home/bgd/GIT/darknet/src-python/darknet_images.py", line 430, in main
network, class_names, class_colors = darknet.load_network(
Does anyone have a fix for this?
print out from darknet version:
Darknet V3 "Jazz" v3.0-194-g3b4dc0e7
CUDA runtime version 12030 (v12.3), driver version 12030 (v12.3)
cuDNN version 12060 (v9.5.1), use of half-size floats is ENABLED
=> 0: Tesla T4 [#7.5], 14.6 GiB
OpenCV v4.2.0, Ubuntu 20.04, kvm
It seems that the python-function "darknet.load_network()" returns an integer instead of a tuple, which causes basically all python files in src-python to fail. Example from running darknet_images.py:
0 : compute_capability = 750, cudnn_half = 1, GPU: Tesla T4 Allocating workspace to transfer between CPU and GPU: 50.0 MiB Traceback (most recent call last): File "/home/bgd/GIT/darknet/src-python/darknet_images.py", line 442, in
main()