import os import torch import torch.nn as nn import numpy as np import utils.general as utils import utils.cgan as cgan_utils from dense_CGAN import Generator, Discriminator import torchvision if __name__ == '__main__': epochs = 200 batch_size = 100 latent_dim = 100 dataloader = utils.get_dataloader(batch_size, pad=False) device = utils.get_device() step_per_epoch = np.ceil(dataloader.dataset.__len__() / batch_size) sample_dir = './samples' checkpoint_dir = './checkpoints' utils.makedirs(sample_dir, checkpoint_dir) G = Generator(latent_dim=latent_dim).to(device) D = Discriminator().to(device) g_optim = utils.get_optim(G, 0.0005) d_optim = utils.get_optim(D, 0.0005) g_log = [] d_log = []
""" import os import torch import torch.nn as nn import numpy as np import utils.general as utils from GAN import Generator, Discriminator import torchvision if __name__ == '__main__': epochs = 200 batch_size = 100 latent_dim = 100 dataloader = utils.get_dataloader(batch_size) device = utils.get_device() step_per_epoch = np.ceil(dataloader.dataset.__len__() / batch_size) sample_dir = './samples' checkpoint_dir = './checkpoints' utils.makedirs(sample_dir, checkpoint_dir) G = Generator(latent_dim = latent_dim).to(device) D = Discriminator().to(device) g_optim = utils.get_optim(G, 0.0002) d_optim = utils.get_optim(D, 0.0002) g_log = [] d_log = []