src/generator.py

import torch
import torch.nn as nn
import torch.nn.functional as F
from torch import optim
import numpy as np
import matplotlib.pyplot as plt
from torchvision import datasets
import torchvision.transforms as transforms


class Gerador(nn.Module):
    def __init__(self):
        super().__init__()
        self.input_dim = 100 + 10  # Ruído (100) + One-hot do número (10)

        self.dense0 = nn.Linear(self.input_dim, 32)
        self.dense1 = nn.Linear(32, 64)
        self.dense2 = nn.Linear(64, 128)
        self.dense3 = nn.Linear(128, 784)
        self.dropout = nn.Dropout(0.3)

    def forward(self, ruido, rotulo):
        X = torch.cat((ruido, rotulo), dim=1)  # Concatenando ruído e rótulo one-hot
        X = self.dropout(F.leaky_relu(self.dense0(X), 0.2))
        X = self.dropout(F.leaky_relu(self.dense1(X), 0.2))
        X = self.dropout(F.leaky_relu(self.dense2(X), 0.2))
        X = torch.tanh(self.dense3(X))
        X = X.view(X.shape[0], 28, 28)
        return X