Generative Adversarial Networks Projects

Starting the training

To train the network for a specified number of epochs, perform the following steps:

1. Start by loading the dataset for both domains, as follows:

imagesA, imagesB = load_images(data_dir=data_dir)

We have already defined the load_images function in the Loading the data subsection in the Training a CycleGAN section.

2. Next, create a for loop, which should run for the number of times specified by the number of epochs, as follows:

for epoch in range(epochs):    print("Epoch:{}".format(epoch))

3. Create two lists to store the losses for all mini-batches, as follows:

dis_losses = []gen_losses = []

4. Calculate the number of mini-batches inside the epochs loop, as follows:

num_batches = int(min(imagesA.shape[0], imagesB.shape[0