Pytorch-可视化小结

2020-11-05

1. pytorch可视化

1.1. torchviz

安装
- 安装graphviz并配置在系统的环境变量(系统->高级设置->环境变量->…./bin
  - dot -version在命令行窗口中验证是否安装成功
- python环境中安装 pip install graphviz
- 验证出现如下界面表示安装成功
- python环境中安装- pip install torchviz

使用

import torch
from torch import nn
from torchviz import make_dot, make_dot_from_trace

model = nn.Sequential()
model.add_module('W0', nn.Linear(8, 16))
model.add_module('tanh', nn.Tanh())
model.add_module('W1', nn.Linear(16, 1))

x = torch.randn(1, 8)

vis_graph = make_dot(model(x), params=dict(model.named_parameters()))
vis_graph.view()  # 会在当前目录下保存一个“Digraph.gv.pdf”文件，并在默认浏览器中打开

# with torch.onnx.set_training(model, False): # 版本兼容性问题这里的onnx会报错
#     trace, _ = torch.jit.get_trace_graph(model, args=(x,))
# make_dot_from_trace(trace)

1.2. torchsummary

安装
1
pip install torchsummary

使用

1 2	from torchsummary import summary summary(your_model, input_size=(channels, H, W))

代码实现

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.conv2_drop = nn.Dropout2d()
        self.fc1 = nn.Linear(320, 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, 320)
        x = F.relu(self.fc1(x))
        x = F.dropout(x, training=self.training)
        x = self.fc2(x)
        return F.log_softmax(x, dim=1)

device = torch.device("cuda" if torch.cuda.is_available() else "cpu") # PyTorch v0.4.0
model = Net().to(device)

summary(model, (1, 28, 28))

结果输出

----------------------------------------------------------------
        Layer (type)               Output Shape         Param #
================================================================
            Conv2d-1           [-1, 10, 24, 24]             260
            Conv2d-2             [-1, 20, 8, 8]           5,020
         Dropout2d-3             [-1, 20, 8, 8]               0
            Linear-4                   [-1, 50]          16,050
            Linear-5                   [-1, 10]             510
================================================================
Total params: 21,840
Trainable params: 21,840
Non-trainable params: 0
----------------------------------------------------------------
Input size (MB): 0.00
Forward/backward pass size (MB): 0.06
Params size (MB): 0.08
Estimated Total Size (MB): 0.15
----------------------------------------------------------------

1.3. netron

使用方法:
- 1.直接将保存的模型在https://netron.app/网站中打开即可
- 2.
  1
  pip install netron

第二种方法的使用

'''
可视化netron:
pip install netron
'''
import torch
import torch.nn as nn
import torch.nn.functional as F
import torch.onnx

import netron


class model(nn.Module):
    def __init__(self):
        super(model, self).__init__()
        self.block1 = nn.Sequential(
            nn.Conv2d(64, 64, 3, padding=1, bias=False),
            nn.BatchNorm2d(64),
            nn.ReLU(inplace=True),
            nn.Conv2d(64, 32, 1, bias=False),
            nn.BatchNorm2d(32),
            nn.ReLU(inplace=True),
            nn.Conv2d(32, 64, 3, padding=1, bias=False),
            nn.BatchNorm2d(64)
        )

        self.conv1 = nn.Conv2d(3, 64, 3, padding=1, bias=False)  # 64*3*3*3(feature_map/filter/weight   )
        self.output = nn.Sequential(
            nn.Conv2d(64, 1, 3, padding=1, bias=True),
            nn.Sigmoid()
        )

    def forward(self, x):
        x = self.conv1(x)
        identity = x
        x = F.relu(self.block1(x) + identity)
        x = self.output(x)
        return x


d = torch.rand(1, 3, 416, 416)
m = model()
o = m(d)

onnx_path = "onnx_model_name.onnx"
# 保存导出模型模型
torch.onnx.export(m, d, onnx_path)
# 打开模型显示数据
netron.start(onnx_path)

运行之后结果显示

1604562037021

本文作者： SparkParis
本文链接： https://sparkparis.github.io/2020/11/05/Pytorch1-%E5%8F%AF%E8%A7%86%E5%8C%96/
版权声明： 本作品采用知识共享署名-非商业性使用-相同方式共享 4.0 国际许可协议进行许可。转载请注明出处！