how to run age test.py

Hello, How can I run the age/gender detection test.py file without using GPU? I get the following error regarding GPU value :

RuntimeError: simple_bind error. Arguments: data: (1, 3, 112, 112) [14:32:59] C:\Jenkins\workspace\mxnet-tag\mxnet\src\storage\storage.cc:143: Compile with USE_CUDA=1 to enable GPU usage

[hopeai]

Hi, You can change set line 46 in face_model.py to elif args.gpu==-1: and pass -1 for --gpu while running the code. So, mxnet will run on CPU. python test.py --gpu -1

[JAYM629]

Hi @ImOmid , thanks for the reply. There’s something I couldn’t understand in the solution you provided. In face_model.py file the lines 45-46 are:

45    self.args = args
46   ctx = mx.gpu(args.gpu)

How do I add this line “elif args.gpu==-1:” ? I would appreciate it if you could explain more specifically what to change. Based on my understanding I tried the below 2 ways but both failed. way 1:

self.args = args
if args.gpu==-1:
  ctx = mx.gpu(args.gpu)

resulted error:

  File "C:\Users\malabdou\Documents\Python\PyProj\GenderAge\X1insightface-master\deploy\helper.py", line 156, in detect_first_stage
    output = net.predict(input_buf)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\mxnet\model.py", line 699, in predict
    self._init_predictor(data_shapes, type_dict)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\mxnet\model.py", line 622, in _init_predictor
    self.ctx[0], grad_req='null', type_dict=type_dict, **shapes)
  File "C:\Users\malabdou\Anaconda3\envs\py5cv-33\lib\site-packages\mxnet\symbol\symbol.py", line 1529, in simple_bind
    raise RuntimeError(error_msg)
RuntimeError: simple_bind error. Arguments:
conv2_weight: (16, 10, 3, 3)
conv2_bias: (16,)
conv4_2_bias: (4,)
prelu1_gamma: (10,)
conv3_bias: (32,)
data: (1, 3, 68, 68)
conv4_1_bias: (2,)
prelu3_gamma: (32,)
conv4_2_weight: (4, 32, 1, 1)
conv3_weight: (32, 16, 3, 3)
conv1_weight: (10, 3, 3, 3)
prelu2_gamma: (16,)
conv1_bias: (10,)
conv4_1_weight: (2, 32, 1, 1)
[15:59:03] C:\Jenkins\workspace\mxnet-tag\mxnet\include\mxnet/base.h:299: Please compile with CUDA enabled for cuda features

2nd Way:

    self.args = args
    if args.gpu==-1: 
      ctx = mx.cpu()
    else:
      ctx = mx.gpu(args.gpu)

resulted error:

Traceback (most recent call last):
  File "test.py", line 23, in <module>
    gender, age = model.get_ga(img)
  File "C:\Users\malabdou\Documents\Python\PyProj\GenderAge\X1insightface-master\deploy\face_model.py", line 107, in get_ga
    self.ga_model.forward(db, is_train=False)
AttributeError: 'NoneType' object has no attribute 'forward'

Have you solve this problem? I also meet this problem