An Jittor implementation of Conditional GAN by Holder, targeting at generating & recognizing handwriting numbers.
TOC
CGAN.py is the core training code of the Conditional GAN.
result.png is an example image of handwriting numbers.
Usage
Make sure that Jittor has been installed in your Python environment:
python3 -m pip install jittor
Then, create a new directory for images generated when training the model:
mkdir images
Then, run the Python code:
python3 CGAN.py
The images generated when training will be saved in images folder. The final image, the final generator & discriminator will be saved in the project directory.
CGAN Jittor Holder ver.
An Jittor implementation of Conditional GAN by Holder, targeting at generating & recognizing handwriting numbers.
TOC
CGAN.pyis the core training code of the Conditional GAN.result.pngis an example image of handwriting numbers.Usage
Make sure that Jittor has been installed in your Python environment:
Then, create a new directory for images generated when training the model:
Then, run the Python code:
The images generated when training will be saved in
imagesfolder. The final image, the final generator & discriminator will be saved in the project directory.The model we trained
The model we trained has been uploaded to URL: https://cloud.tsinghua.edu.cn/d/6852e44d9b7b409eaa9d/.
The model is saved in form of Pickle File (
.pkl).