Confusion about the layer object - l.output_gpu

[Lin-Mao]

From what I know so far, during the processing of inference, it uses l.forwark_gpu to compute each layer when enabling GPU. Then, the results for each layer computation will be stored to l.output_gpu used for later layer computing. But, it was all zero except in the YOLO layer when I print out the content of each l.output_gpu after computing. I have no idea what was happened. By the way, it can detect the image successfully.

Any response would be appreciated!

[Lin-Mao]

My bad. I got it. Do not need to reply!