CGAN_jittor

Introduction

This project is a Jittor implementation of Conditional GAN (CGAN). We use the MNIST digit image dataset to train a Conditional GAN model that maps random noise and class labels to digit images, and generate specified digit sequences.

Program Running

First, you need to install the Jittor framework. Please refer to the website https://cg.cs.tsinghua.edu.cn/jittor/download/.

Then, you only need to enter the following command:

python CGAN.py

After training for a period of time, the image ‘result.png’ will be generated.