torch中的summary

最新推荐文章于 2024-06-30 00:00:42 发布

step step

最新推荐文章于 2024-06-30 00:00:42 发布

阅读量1.5k

点赞数 1

分类专栏： torch python深度学习文章标签： pytorch

本文链接：https://blog.csdn.net/weixin_42618420/article/details/129734997

版权

python深度学习同时被 2 个专栏收录

6 篇文章 0 订阅

订阅专栏

torch

1 篇文章 0 订阅

订阅专栏

torch中的summary

通过summary可以查看模型每层的输出形状，每层的参数数量，总的参数数量，以及模型大小等信息。

定义一个简单的模型LeNet

import torch.nn as nn
import torch.nn.functional as F
from torchsummary import summary


class LeNet(nn.Module):
    def __init__(self):
        super(LeNet, self).__init__()
        self.conv1 = nn.Conv2d(3, 16, 5)
        self.pool1 = nn.MaxPool2d(2, 2)
        self.conv2 = nn.Conv2d(16, 32, 5)
        self.pool2 = nn.MaxPool2d(2, 2)
        self.fc1 = nn.Linear(32*5*5, 120)
        self.fc2 = nn.Linear(120, 84)
        self.fc3 = nn.Linear(84, 10)

    def forward(self, x):
        x = F.relu(self.conv1(x))    # input(3, 32, 32) output(16, 28, 28)
        x = self.pool1(x)            # output(16, 14, 14)
        x = F.relu(self.conv2(x))    # output(32, 10, 10)
        x = self.pool2(x)            # output(32, 5, 5)
        x = x.view(-1, 32*5*5)       # output(32*5*5)
        x = F.relu(self.fc1(x))      # output(120)
        x = F.relu(self.fc2(x))      # output(84)
        x = self.fc3(x)              # output(10)
        return x

查看参数等信息

if __name__ == '__main__':
    net = LeNet()
    summary(model=net, input_size=(3,32,32),batch_size=2, device="cpu")

####### 输出

----------------------------------------------------------------
        Layer (type)               Output Shape         Param #
================================================================
            Conv2d-1            [2, 16, 28, 28]           1,216
         MaxPool2d-2            [2, 16, 14, 14]               0
            Conv2d-3            [2, 32, 10, 10]          12,832
         MaxPool2d-4              [2, 32, 5, 5]               0
            Linear-5                   [2, 120]          96,120
            Linear-6                    [2, 84]          10,164
            Linear-7                    [2, 10]             850
================================================================
Total params: 121,182
Trainable params: 121,182
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 0.02
Forward/backward pass size (MB): 0.30
Params size (MB): 0.46
Estimated Total Size (MB): 0.79