Uvdr2026/wgan-cifar10-airplanes

WGAN-GP – CIFAR-10 Airplanes

This repository contains a Wasserstein GAN with Gradient Penalty (WGAN-GP)
trained on CIFAR-10 airplane images.

Files

• wgan_generator.keras – Trained generator (use for inference)
• wgan_critic.keras – Trained critic
• training_config.json – Training hyperparameters

Usage

import tensorflow as tf
from tensorflow import keras

generator = keras.models.load_model("wgan_generator.keras")

LATENT_DIM = 128
noise = tf.random.normal([9, LATENT_DIM])
images = (generator(noise, training=False) + 1) / 2