Run yolov5 and submit them to the following error: the File "D:/yolov5 - master/detect. Py", line 53, detect in
_=model (img. Half () if half the else img) if device. The type! The else='CPU' None # run once
The File "C: \ Users \ tenbig \ Anaconda3 \ envs \ yolov5 \ lib \ site - packages, torch, nn \ modules \ module py", line 550, in __call__
Input result=self. Forward (* and * * kwargs)
The File "D: \ yolov5 - master \ models \ yolo py", line 100, in the forward
Return the self. Forward_once (x, profile) # single - scale inference, "train"
The File "D: \ yolov5 - master \ models \ yolo py", line 123, in forward_once
X=m (x) # run
The File "C: \ Users \ tenbig \ Anaconda3 \ envs \ yolov5 \ lib \ site - packages, torch, nn \ modules \ module py", line 550, in __call__
Input result=self. Forward (* and * * kwargs)
The File "D: \ yolov5 - master \ models \ yolo py", line 27, the in the forward
[I] x=self. M [I] [I] (x) # conv
The File "C: \ Users \ tenbig \ Anaconda3 \ envs \ yolov5 \ lib \ site - packages, torch, nn \ modules \ module py", line 594, in __getattr__
Type (self). __name__, name))
AttributeError: 'Detect object has no attribute' m '

Give advice or comments please the great god!

CodePudding user response:

Same, now also in solution

CodePudding user response:

Seems to be pytorch version is wrong, should upgrade to pytorch1.6 version can,

CodePudding user response:

My version is pytorch 1.6 python3.7 cuda - 10.2 or not

CodePudding user response:

I tried like direct download zip there is a problem, use the git clone https://github.com/ultralytics/yolov5.git can run, you can have a try

CodePudding user response:

Meet the same problem, bosses solved? I have to run the project https://github.com/mikel-brostrom/Yolov5_DeepSort_Pytorch/issues/10, direct download yolov5 can run,

CodePudding user response:

Upstairs, I also run YoloV5 + Deepsort engineering https://github.com/mikel-brostrom/Yolov5_DeepSort_Pytorch encounter this problem, a separate run YoloV5 doesn't come along, who can discuss thank you have any solution,

CodePudding user response:

Convenient to add a QQ communication under? 2455538070

CodePudding user response:

My version is pytorch 1.6 python3.8 cuda 10.2 appear the same mistake

CodePudding user response:

Weight file has a problem, I have successfully solved this problem, is simple

CodePudding user response:

The building Lord solved excuse me?