说明与安装

这个包可以计算出一个网络模型的参数量和运算量，甚至给出每一层的运算量，比如：

example

安装方法为：

1	pip install torchstat

示例

import torch
import torch.nn as nn
import torch.nn.functional as F
from torchstat import stat


class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(3, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.conv2_drop = nn.Dropout2d()
        self.fc1 = nn.Linear(56180, 50)
        self.fc2 = nn.Linear(50, 10)

    def forward(self, x):
        x = F.relu(F.max_pool2d(self.conv1(x), 2))
        x = F.relu(F.max_pool2d(self.conv2_drop(self.conv2(x)), 2))
        x = x.view(-1, 56180)
        x = F.relu(self.fc1(x))
        x = F.dropout(x, training=self.training)
        x = self.fc2(x)
        return F.log_softmax(x, dim=1)


if __name__ == '__main__':
    model = Net()
    stat(model, (3, 224, 224))

然后在命令行输入

1	torchstat -f example.py -m Net

如果要更改输入图像的尺寸，只改example.py里的（3，224，224）没有起作用，于是我使用了-s选项：

1	torchstat -f example.py -m Net -s '3x32x32'

这里选项内容是用字符串表示的，‘x’就是字母x。

另一个示例

from torchstat import stat
import torchvision.models as models

model = models.resnet18()
stat(model, (3, 224, 224))

torchstat的Github