deepface ResourceExhaustedError: 分配内存失败 [Op:AddV2]

Question

我是深度学习的新手。 我正在尝试在我的本地机器中使用deepface库。我使用

pip install deepface

来安装库，尝试了python 3.7.13、3.8.13和3.9.13，它们都是使用conda虚拟环境创建的。

然而，当运行下面的代码片段时，我在本地机器上运行时遇到了同样的错误。我需要 GPU 来运行库吗？如果是，我该如何设置？因为从在线指南/文章中，都没有提到需要安装/设置GPU。

我的本地电脑上有一个 GeForce MX450。

代码

import cv2
from deepface import DeepFace
import numpy as np

def analyse_face():
    imagepath = "happy_face_woman.png"
    image = cv2.imread(imagepath)
    face_analysis = DeepFace.analyze(image)
    print(face_analysis)

print(analyse_face())

错误：

ResourceExhaustedError                    Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_14196\3829791526.py in <module>
     12     print(face_analysis)
     13 
---> 14 analyse_face()

~\AppData\Local\Temp\ipykernel_14196\3829791526.py in analyse_face()
      9     imagepath = "happy_face_woman.png"
     10     image = cv2.imread(imagepath)
---> 11     face_analysis = DeepFace.analyze(image)
     12     print(face_analysis)
     13 

c:\Users\user_name\anaconda3\envs\deepFacepy37\lib\site-packages\deepface\DeepFace.py in analyze(img_path, actions, models, enforce_detection, detector_backend, prog_bar)
    352 
    353         if 'age' in actions and 'age' not in built_models:
--> 354                 models['age'] = build_model('Age')
    355 
    356         if 'gender' in actions and 'gender' not in built_models:

c:\Users\user_name\anaconda3\envs\deepFacepy37\lib\site-packages\deepface\DeepFace.py in build_model(model_name)
     61                 model = models.get(model_name)
     62                 if model:
---> 63                         model = model()
...
-> 1922         seed=self.make_legacy_seed())
   1923 
   1924   def truncated_normal(self, shape, mean=0., stddev=1., dtype=None):

ResourceExhaustedError: failed to allocate memory [Op:AddV2]

不同的错误输出

ResourceExhaustedError                    Traceback (most recent call last)
~\AppData\Local\Temp\ipykernel_14196\3829791526.py in <module>
     12     print(face_analysis)
     13 
---> 14 analyse_face()

~\AppData\Local\Temp\ipykernel_14196\3829791526.py in analyse_face()
      9     imagepath = "happy_face_woman.png"
     10     image = cv2.imread(imagepath)
---> 11     face_analysis = DeepFace.analyze(image)
     12     print(face_analysis)
     13 

c:\Users\user_name\anaconda3\envs\deepFacepy37\lib\site-packages\deepface\DeepFace.py in analyze(img_path, actions, models, enforce_detection, detector_backend, prog_bar)
    352 
    353         if 'age' in actions and 'age' not in built_models:
--> 354                 models['age'] = build_model('Age')
    355 
    356         if 'gender' in actions and 'gender' not in built_models:

c:\Users\user_name\anaconda3\envs\deepFacepy37\lib\site-packages\deepface\DeepFace.py in build_model(model_name)
     61                 model = models.get(model_name)
     62                 if model:
---> 63                         model = model()
...
-> 1922         seed=self.make_legacy_seed())
   1923 
   1924   def truncated_normal(self, shape, mean=0., stddev=1., dtype=None):

ResourceExhaustedError: OOM when allocating tensor with shape[7,7,512,4096] and type float on /job:localhost/replica:0/task:0/device:GPU:0 by allocator GPU_0_bfc [Op:RandomUniform]

附加信息 我已经运行了命令来检查我的 GPU 使用情况，详细信息如下：

!nvidia-smi

Answer 1

为什么不禁用 GPU？

import os
os.environ["CUDA_VISIBLE_DEVICES"]=""