The code hasn't been modified. When testing the pictures or video, sometimes the results can be trained and some will be displayed：
InvalidArgumentError (see above for traceback): input and filter must have the same depth: 26 vs 512
[[Node: yolov3/darknet53_body/Conv_43/Conv2D = _MklConv2D[T=DT_FLOAT, _kernel="MklOp", data_format="NHWC", dilations=[1, 1, 1, 1], padding
="VALID", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device="/job:localhost/replica:0/task:0/device:CPU:0"](yolov3/darknet53_body/Pad_4, yolov3/
darknet53_body/Conv_43/weights/read, DMT/_432, yolov3/darknet53_body/Conv_43/weights/read:1)]]
Shape is not in the right shape，Video and picture are both, 50% of the test results can be run, and the other half will report the above errors
The code hasn't been modified. When testing the pictures or video, sometimes the results can be trained and some will be displayed： InvalidArgumentError (see above for traceback): input and filter must have the same depth: 26 vs 512 [[Node: yolov3/darknet53_body/Conv_43/Conv2D = _MklConv2D[T=DT_FLOAT, _kernel="MklOp", data_format="NHWC", dilations=[1, 1, 1, 1], padding ="VALID", strides=[1, 2, 2, 1], use_cudnn_on_gpu=true, _device="/job:localhost/replica:0/task:0/device:CPU:0"](yolov3/darknet53_body/Pad_4, yolov3/ darknet53_body/Conv_43/weights/read, DMT/_432, yolov3/darknet53_body/Conv_43/weights/read:1)]]
Shape is not in the right shape，Video and picture are both, 50% of the test results can be run, and the other half will report the above errors