DQN.py

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

class DQN(nn.Module):
    def __init__(self, h, w):
        super(DQN, self).__init__()
        self.conv1 = nn.Conv2d(3, 16, kernel_size=5, stride=2)
        self.bn1 = nn.BatchNorm2d(16)
        self.conv2 = nn.Conv2d(16, 32, kernel_size=5, stride=2)
        self.bn2 = nn.BatchNorm2d(32)
        self.conv3 = nn.Conv2d(32, 32, kernel_size=5, stride=2)
        self.bn3 = nn.BatchNorm2d(32)

        def conv2d_size_out(size, kernel_size=5, stride=2):
            return (size - kernel_size) // stride + 1
        convw = w
        convh = h
        for i in range(3):
            convw = conv2d_size_out(convw)
            convh = conv2d_size_out(convh)
        
        self.head = nn.Linear(convw * convh * 32, 2)
        
            
    
    # x.size: (N, input_channels, H, W)
    # output.size: (N, 2)
    # DQN is used to calculate Q(s_t)
    def forward(self, x):
        x = F.relu(self.bn1(self.conv1(x)))
        x = F.relu(self.bn2(self.conv2(x)))
        x = F.relu(self.bn3(self.conv3(x)))

        return self.head(x.view(x.size(0), -1))