conditional_gan

Introduction

This is a simple conditional GAN network reproduced using the Jittor framework.Using the Touge online evaluation platform, you can get a score of 0.9689

Build & Environment Requisitions

You need to make sure that the JITTOR environment is configured locally. Besides, numpy and matplotlib are required.

Getting Started

then run the command below in shell:

python CGAN.py

Demo

After 100 epochs of training, the generator can produce the following results.

Modify the string variable number in CGAN.py.

For eaxmple:

number = "13788951377748" 

A simulation generated by the generator is saved in result.png.