Defining the structure of the confined but happy neural network

neural-network.py

import torch.nn as nn
import torch.nn.functional as F

class Network(nn.Module):
  def __init__(self, input_size, lin1_size, lin2_size):
    super().__init__() # initializes Network from the parent class Module.
    # linear layers. 
    self.lin1 = nn.Linear(input_size, lin1_size) # Creates random weights and biases.
    self.lin2 = nn.Linear(lin1_size, lin2_size) # Use lin1.weight & lin1.bias to inspect.
  
  def forward(self, x):
    x = self.lin1(x) # affine function: x = x @ weights + bias.
    x = F.relu(x) # ReLU activation.
    x = self.lin2(x) # 2nd affine function
    return x

# Creating a Network object
net = Network(input_size=3, lin1_size=7, lin2_size=2)