python-pytorchHow do I use nn.linear in Python Pytorch?

nn.linear is a module in Pytorch that allows you to create a linear layer in a neural network. To use it, you need to first import the module:

import torch.nn as nn

You can then create a linear layer with the following code:

linear = nn.Linear(in_features, out_features)

Where in_features is the number of input features, and out_features is the number of output features.

You can also specify the bias of the linear layer:

linear = nn.Linear(in_features, out_features, bias=True)

You can then use the linear layer in your neural network. For example, if you have a neural network with two linear layers, you can define it as:

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.linear1 = nn.Linear(3, 4)
        self.linear2 = nn.Linear(4, 1)

    def forward(self, x):
        x = self.linear1(x)
        x = self.linear2(x)
        return x

net = Net()

You can then pass an input x of shape (batch_size, 3) through the neural network:

x = torch.randn(2, 3)
out = net(x)
print(out)

This will output:

tensor([[-0.7137],
        [-0.8232]], grad_fn=<AddmmBackward>)

Code parts and Explanation

• import torch.nn as nn: imports the nn module from torch.nn
• nn.Linear(in_features, out_features): creates a linear layer with the given in_features and out_features
• nn.Linear(in_features, out_features, bias=True): creates a linear layer with the given in_features and out_features and sets the bias to True
• class Net(nn.Module):: defines a new neural network class Net that inherits from nn.Module
• self.linear1 = nn.Linear(3, 4): creates a linear layer with 3 input features and 4 output features
• self.linear2 = nn.Linear(4, 1): creates a linear layer with 4 input features and 1 output feature
• def forward(self, x):: defines the forward pass of the neural network
• x = self.linear1(x): passes the input x through the first linear layer
• x = self.linear2(x): passes the output of the first linear layer through the second linear layer
• net = Net(): creates an instance of the neural network
• x = torch.randn(2, 3): creates a random input x of shape (2, 3)
• out = net(x): passes the input x through the neural network
• print(out): prints the output of the neural network

Relevant Links

• PyTorch Documentation - nn.Linear

Edit this code on GitHub