A Jittor implementation of Conditional GAN (CGAN) on MNIST dataset.
Project Overview
This project implements a Conditional GAN (CGAN) model using Jittor deep learning framework. The generator takes random noise and a specified class label to generate a 32x32 MNIST-like digit image.
Requirements
Python 3.7+
Jittor >= 1.3.8.5
numpy
Pillow
Usage
1. Install dependencies
pip install jittor numpy Pillow
2. Train the CGAN model
python CGAN.py
3. Generate specific digit sequence images
After training, the model will automatically generate images for the sequence “28131342805930” and save it as result.png.
Notes
The dataset (MNIST) will be automatically downloaded by Jittor’s built-in dataset loader.
If the dataset or model files are too large, consider providing a Baidu Cloud download link.
License
This project is open-sourced for educational purposes.
关于
A Jittor implementation of Conditional GAN (CGAN) on MNIST dataset.
CGAN_jittor
A Jittor implementation of Conditional GAN (CGAN) on MNIST dataset.
Project Overview
This project implements a Conditional GAN (CGAN) model using Jittor deep learning framework. The generator takes random noise and a specified class label to generate a 32x32 MNIST-like digit image.
Requirements
Usage
1. Install dependencies
2. Train the CGAN model
3. Generate specific digit sequence images
After training, the model will automatically generate images for the sequence “28131342805930” and save it as result.png.
Notes
The dataset (MNIST) will be automatically downloaded by Jittor’s built-in dataset loader.
If the dataset or model files are too large, consider providing a Baidu Cloud download link.
License
This project is open-sourced for educational purposes.