pytorch-CycleGAN-and-pix2pix README

The project's README from github.com/junyanz/pytorch-CycleGAN-and-pix2pix. As one markdown file: README.md.




CycleGAN and pix2pix in PyTorch

Udpate in 2025: we recently updated the code to support Python 3.11 and PyTorch 2.4. It also supports DDP for single-machine multiple-GPU training. (Please use torchrun --nproc_per_node=4 train.py ...)

New: Please check out img2img-turbo repo that includes both pix2pix-turbo and CycleGAN-Turbo. Our new one-step image-to-image translation methods can support both paired and unpaired training and produce better results by leveraging the pre-trained StableDiffusion-Turbo model. The inference time for 512x512 image is 0.29 sec on A6000 and 0.11 sec on A100.

Please check out contrastive-unpaired-translation (CUT), our new unpaired image-to-image translation model that enables fast and memory-efficient training.

We provide PyTorch implementations for both unpaired and paired image-to-image translation.

The code was written by Jun-Yan Zhu and Taesung Park, and supported by Tongzhou Wang.

This PyTorch implementation produces results comparable to or better than our original Torch software. If you would like to reproduce the same results as in the papers, check out the original CycleGAN Torch and pix2pix Torch code in Lua/Torch.

Note: The current software works well with PyTorch 2.4+. Check out the older branch that supports PyTorch 0.1-0.3.

You may find useful information in training/test tips and frequently asked questions. To implement custom models and datasets, check out our templates. To help users better understand and adapt our codebase, we provide an overview of the code structure of this repository.

CycleGAN: Project | Paper | Torch | Tensorflow Core Tutorial | PyTorch Colab

Pix2pix: Project | Paper | Torch | Tensorflow Core Tutorial | PyTorch Colab

EdgesCats Demo | pix2pix-tensorflow | by Christopher Hesse

If you use this code for your research, please cite:

Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks.
Jun-Yan Zhu*, Taesung Park*, Phillip Isola, Alexei A. Efros. In ICCV 2017. (* equal contributions) [Bibtex]

Image-to-Image Translation with Conditional Adversarial Networks.
Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, Alexei A. Efros. In CVPR 2017. [Bibtex]

Talks and Course

pix2pix slides: keynote | pdf, CycleGAN slides: pptx | pdf

CycleGAN course assignment code and handout designed by Prof. Roger Grosse for CSC321 "Intro to Neural Networks and Machine Learning" at University of Toronto. Please contact the instructor if you would like to adopt it in your course.

Colab Notebook

TensorFlow Core CycleGAN Tutorial: Google Colab | Code

TensorFlow Core pix2pix Tutorial: Google Colab | Code

PyTorch Colab notebook: CycleGAN and pix2pix

ZeroCostDL4Mic Colab notebook: CycleGAN and pix2pix

Other implementations

CycleGAN

[Tensorflow] (by Harry Yang), [Tensorflow] (by Archit Rathore), [Tensorflow] (by Van Huy), [Tensorflow] (by Xiaowei Hu), [Tensorflow2] (by Zhenliang He), [TensorLayer1.0] (by luoxier), [TensorLayer2.0] (by zsdonghao), [Chainer] (by Yanghua Jin), [Minimal PyTorch] (by yunjey), [Mxnet] (by Ldpe2G), [lasagne/Keras] (by tjwei), [Keras] (by Simon Karlsson), [OneFlow] (by Ldpe2G)

pix2pix

[Tensorflow] (by Christopher Hesse), [Tensorflow] (by Eyyüb Sariu), [Tensorflow (face2face)] (by Dat Tran), [Tensorflow (film)] (by Arthur Juliani), [Tensorflow (zi2zi)] (by Yuchen Tian), [Chainer] (by mattya), [tf/torch/keras/lasagne] (by tjwei), [Pytorch] (by taey16)

Prerequisites

Getting Started

Installation

git clone https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix
cd pytorch-CycleGAN-and-pix2pix

conda env create -f environment.yml

and then activate the environment by

conda activate pytorch-img2img

CycleGAN train/test

bash ./datasets/download_cyclegan_dataset.sh maps

#!./scripts/train_cyclegan.sh
python train.py --dataroot ./datasets/maps --name maps_cyclegan --model cycle_gan --use_wandb

To see more intermediate results, check out ./checkpoints/maps_cyclegan/web/index.html.

#!./scripts/test_cyclegan.sh
python test.py --dataroot ./datasets/maps --name maps_cyclegan --model cycle_gan

pix2pix train/test

bash ./datasets/download_pix2pix_dataset.sh facades

#!./scripts/train_pix2pix.sh
python train.py --dataroot ./datasets/facades --name facades_pix2pix --model pix2pix --direction BtoA  --use_wandb

To see more intermediate results, check out ./checkpoints/facades_pix2pix/web/index.html.

#!./scripts/test_pix2pix.sh
python test.py --dataroot ./datasets/facades --name facades_pix2pix --model pix2pix --direction BtoA

Apply a pre-trained model (CycleGAN)

bash ./scripts/download_cyclegan_model.sh horse2zebra

bash ./datasets/download_cyclegan_dataset.sh horse2zebra

python test.py --dataroot datasets/horse2zebra/testA --name horse2zebra_pretrained --model test --no_dropout

Apply a pre-trained model (pix2pix)

Download a pre-trained model with ./scripts/download_pix2pix_model.sh.

bash ./scripts/download_pix2pix_model.sh facades_label2photo

bash ./datasets/download_pix2pix_dataset.sh facades

python test.py --dataroot ./datasets/facades/ --direction BtoA --model pix2pix --name facades_label2photo_pretrained

Multi-GPU training

To train a model on multiple GPUs, please use torchrun --nproc_per_node=4 train.py ... instead of python train.py .... We also need to use synchronized batchnorm by setting --norm sync_batch (or --norm sync_instance for instance normgalization). The --norm batch is not compatible with DDP.

Docker

We provide the pre-built Docker image and Dockerfile that can run this code repo. See docker.

Datasets

Download pix2pix/CycleGAN datasets and create your own datasets.

Training/Test Tips

Best practice for training and testing your models.

Frequently Asked Questions

Before you post a new question, please first look at the above Q & A and existing GitHub issues.

Custom Model and Dataset

If you plan to implement custom models and dataset for your new applications, we provide a dataset template and a model template as a starting point.

Code structure

To help users better understand and use our code, we briefly overview the functionality and implementation of each package and each module.

Pull Request

You are always welcome to contribute to this repository by sending a pull request. Please run flake8 --ignore E501 . and pytest scripts/test_before_push.py -v before you commit the code. Please also update the code structure overview accordingly if you add or remove files.

Citation

If you use this code for your research, please cite our papers.

@inproceedings{CycleGAN2017,
  title={Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks},
  author={Zhu, Jun-Yan and Park, Taesung and Isola, Phillip and Efros, Alexei A},
  booktitle={Computer Vision (ICCV), 2017 IEEE International Conference on},
  year={2017}
}


@inproceedings{isola2017image,
  title={Image-to-Image Translation with Conditional Adversarial Networks},
  author={Isola, Phillip and Zhu, Jun-Yan and Zhou, Tinghui and Efros, Alexei A},
  booktitle={Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on},
  year={2017}
}

Other Languages

Spanish

img2img-turbo
contrastive-unpaired-translation (CUT)
CycleGAN-Torch | pix2pix-Torch | pix2pixHD| BicycleGAN | vid2vid | SPADE/GauGAN
iGAN | GAN Dissection | GAN Paint

Cat Paper Collection

If you love cats, and love reading cool graphics, vision, and learning papers, please check out the Cat Paper Collection.

Acknowledgments

Our code is inspired by pytorch-DCGAN.

pytorch-CycleGAN-and-pix2pix (junyanz/pytorch-CycleGAN-and-pix2pix)

Searching Database…

25.3K ·GitHub
Pytorch-CycleGAN-and-pix2pix
What is this?Open-source PyTorch implementations of two classic image-to-image translation models: CycleGAN and pix2pix. They turn images from one domain into another, like horses into zebras or building label maps into photos.
CycleGAN and pix2pix are two different models - when do I use which?CycleGAN learns the translation from two unpaired image collections, so no one-to-one matches are needed; pix2pix trains on aligned input-output pairs from your dataset.
How do I get it running?Clone the repo, install the dependencies, download a dataset with the bundled script and run the train and test scripts - test results are saved to an HTML page under ./results/. A pre-built Docker image and Colab notebooks are also available.
What do I need installed first?Python 3 and PyTorch 2.4+; a Conda environment file is provided to set up the dependencies, and there are pre-built Docker and Repl options if you prefer.
Will it run on Windows?The prerequisites only list Linux or macOS, so Windows is not officially supported.
Do I need an NVIDIA GPU?No - the prerequisites list CPU as an option, though an NVIDIA GPU with CUDA CuDNN is what you want for sensible training times.
Can I try it without training anything myself?Yes - pre-trained models are downloaded with the bundled scripts, e.g. horse2zebra for CycleGAN or label2photo on Facades for pix2pix. The checkpoint lands in ./checkpoints/horse2zebra_pretrained/latest_net_G.pth and results are written to an HTML page in ./results/. For the Facades model you must pass --direction BtoA, since its A-to-B direction is photos to labels.
How do I train on my own images?Download a dataset with the bundled scripts or prepare your own, then run the training script. Intermediate results are written to an HTML page under ./checkpoints, W&B logging is available with the --use_wandb flag, and a model template plus a dataset template are provided if you want a custom architecture or data loader.
Does training support multiple GPUs?Yes - since the 2025 update you launch DDP on a single machine with torchrun --nproc_per_node=4 train.py, and you must switch to --norm sync_batch (or sync_instance) because plain batch norm is incompatible with DDP.
Is there a faster alternative from the same authors?Yes - img2img-turbo does one-step image-to-image translation built on StableDiffusion-Turbo and supports both paired and unpaired training: 0.29 sec per 512x512 image on an A6000 and 0.11 sec on an A100. The authors also point to CUT for fast, memory-efficient unpaired training.
What are similar projects?CycleGAN, pix2pix, Bringing-Old-Photos-Back-to-Life, PyTorch-GAN, DragGAN, stylegan2, lama, GFPGAN
Which topics is it in?image-to-image-translation, image-manipulation, generative-adversarial-network, gan, computer-graphics, dataset, neural-network, image-generation, image-processing, nvidia, computer-vision, cuda
Creator
junyanz
Size
8.3 MB
Stars
25.3K
Updated
CommitAug 06
Age
9y
Commits · 90 Days