**This is a repo for training CGAN model based on jittor. You may use it as a practice on how to get start with jittor. **
Introduction
This repo provides a digital image dataset, MNIST, and you can train a Conditional GAN model that maps random noise and category labels to digital images, and use it to generate any input numbers you like.
A sample code framework is provided in ./warm_up_competition, offering data download, model definition, training steps and other functions.
An example generated by Conditional GAN is as the following (with phone number as input):
Installing
To run the model, please follow the instructions below:
# Clone the repo to local
git clone https://github.com/Jittor/gan-jittor.git
cd gan-jittor
# Install the pakages required
sudo python3 -m pip install -r requirements.txt
cd warm_up_competition
# Run the CGAN model !
python3 CGAN.py
Structure
The descriptions about each directory are as the following:
–CGAN_jittor (root)
–assets: contains the graphs used for model training
–data: contains the scripts for downloading specific datasets
–model: contains all the models that might be used for training
–warm_up_competition: contains the main code CGAN.py; results will also be put into this directory
Other Usage
Besides running the Conditional GAN model, there are also several models provided in directory “models”, and they can all be trained by yourself! Explore them now! The introduction to this part can be find in “./models/README.md”
CGAN_jittor
**This is a repo for training CGAN model based on jittor. You may use it as a practice on how to get start with jittor. **
Introduction
This repo provides a digital image dataset, MNIST, and you can train a Conditional GAN model that maps random noise and category labels to digital images, and use it to generate any input numbers you like.
A sample code framework is provided in ./warm_up_competition, offering data download, model definition, training steps and other functions.
An example generated by Conditional GAN is as the following (with phone number as input):
Installing
To run the model, please follow the instructions below:
Structure
The descriptions about each directory are as the following:
–CGAN_jittor (root)
Other Usage
Besides running the Conditional GAN model, there are also several models provided in directory “models”, and they can all be trained by yourself! Explore them now! The introduction to this part can be find in “./models/README.md”