torch.hub.load(‘pytorch/vision‘, ‘deeplabv3_resnet101‘, pretrained=True)运行方法

1. 首先下载 vision-master.zip

 torch_1.7.1_gpu) E:\PythonPRO\PytorchPro\G5>pip install vision-master.zip

实现上安装到 C:\Users\Administrator.cache\torch\hub (每个人安装的地方不一样，通过运行程序可以看出在什么地方)

2. 下面主要用到2个文件

D:\ProgramData\Anaconda3\envs\pytorch_gpu\Lib\site-packages\torch\hub.py
C:\Users\Administrator\.cache\torch\hub\pytorch_vision_master\hubconf.py

在hubconf.py中，并按着下面注释掉不要的：

# Optional list of dependencies required by the package
dependencies = ['torch']

# classification
from torchvision.models.alexnet import alexnet
from torchvision.models.densenet import densenet121, densenet169, densenet201, densenet161
from torchvision.models.inception import inception_v3
from torchvision.models.resnet import resnet18, resnet34, resnet50, resnet101, resnet152,\
    resnext50_32x4d, resnext101_32x8d, wide_resnet50_2, wide_resnet101_2
from torchvision.models.squeezenet import squeezenet1_0, squeezenet1_1
from torchvision.models.vgg import vgg11, vgg13, vgg16, vgg19, vgg11_bn, vgg13_bn, vgg16_bn, vgg19_bn
from torchvision.models.googlenet import googlenet
from torchvision.models.shufflenetv2 import shufflenet_v2_x0_5, shufflenet_v2_x1_0
# from torchvision.models.mobilenetv2 import mobilenet_v2
# from torchvision.models.mobilenetv2 import mobilenetv2
# from torchvision.models.mobilenetv3 import mobilenet_v3_large, mobilenet_v3_small
# from torchvision.models.mnasnet import mnasnet0_5, mnasnet0_75, mnasnet1_0, \
#     mnasnet1_3

# segmentation
# from torchvision.models.segmentation import fcn_resnet50, fcn_resnet101, \
#     deeplabv3_resnet50, deeplabv3_resnet101, deeplabv3_mobilenet_v3_large, lraspp_mobilenet_v3_large

from torchvision.models.segmentation import fcn_resnet50, fcn_resnet101, \
   deeplabv3_resnet50, deeplabv3_resnet101

3. 根据网址：
   https://download.pytorch.org/models/resnet101-5d3b4d8f.pth 下载后放到：
   C:\Users\Administrator.cache\torch\hub\checkpoints

4. https://download.pytorch.org/models/deeplabv3_resnet101_coco-586e9e4e.pth

下载后放到
C:\Users\Administrator.cache\torch\hub\checkpoints
再次运行即可成功。

另外 1_seg_Demo.py文件内容也需要改
将

model = torch.hub.load('pytorch/vision', 'deeplabv3plus_resnet101', pretrained=True) # pretrained=True 设为预训练模式

改为：

model = torch.hub.load('pytorch/vision', 'deeplabv3_resnet101', pretrained=True)  # pretrained=True 设为预训练模式