<div align="center"> <img src="https://user-images.githubusercontent.com/12726765/114528756-de55af80-9c7b-11eb-94d7-d3224ada1585.png" width="400"/> <div> </div> <div align="center"> <b><font size="5">OpenMMLab website</font></b> <sup> <a href="https://openmmlab.com"> <i><font size="4">HOT</font></i> </a> </sup>      <b><font size="5">OpenMMLab platform</font></b> <sup> <a href="https://platform.openmmlab.com"> <i><font size="4">TRY IT OUT</font></i> </a> </sup> </div> <div> </div> </div>

English | 简体中文

What's New

MMGeneration has been merged in MMEditing. And we have supported new generation tasks and models. We highlight the following new features:

🌟 Text2Image
- ✅ GLIDE
- ✅ Disco-Diffusion
- ✅ Stable-Diffusion
🌟 3D-aware Generation
- ✅ EG3D

Introduction

MMGeneration is a powerful toolkit for generative models, especially for GANs now. It is based on PyTorch and MMCV. The master branch works with PyTorch 1.5+.

Major Features

High-quality Training Performance: We currently support training on Unconditional GANs, Internal GANs, and Image Translation Models. Support for conditional models will come soon.
Powerful Application Toolkit: A plentiful toolkit containing multiple applications in GANs is provided to users. GAN interpolation, GAN projection, and GAN manipulations are integrated into our framework. It's time to play with your GANs! (Tutorial for applications)
Efficient Distributed Training for Generative Models: For the highly dynamic training in generative models, we adopt a new way to train dynamic models with MMDDP. (Tutorial for DDP)
New Modular Design for Flexible Combination: A new design for complex loss modules is proposed for customizing the links between modules, which can achieve flexible combination among different modules. (Tutorial for new modular design)

<table> <thead> <tr> <td> <div align="center"> <b> Training Visualization</b> <br/> <img src="https://user-images.githubusercontent.com/12726765/114509105-b6f4e780-9c67-11eb-8644-110b3cb01314.gif" width="200"/> </div></td> <td> <div align="center"> <b> GAN Interpolation</b> <br/> <img src="https://user-images.githubusercontent.com/12726765/114679300-9fd4f900-9d3e-11eb-8f37-c36a018c02f7.gif" width="200"/> </div></td> <td> <div align="center"> <b> GAN Projector</b> <br/> <img src="https://user-images.githubusercontent.com/12726765/114524392-c11ee200-9c77-11eb-8b6d-37bc637f5626.gif" width="200"/> </div></td> <td> <div align="center"> <b> GAN Manipulation</b> <br/> <img src="https://user-images.githubusercontent.com/12726765/114523716-20302700-9c77-11eb-804e-327ae1ca0c5b.gif" width="200"/> </div></td> </tr> </thead> </table>

Highlight

Positional Encoding as Spatial Inductive Bias in GANs (CVPR2021) has been released in MMGeneration. [Config], [Project Page]
Conditional GANs have been supported in our toolkit. More methods and pre-trained weights will come soon.
Mixed-precision training (FP16) for StyleGAN2 has been supported. Please check the comparison between different implementations.

Changelog

v0.7.3 was released on 14/04/2023. Please refer to changelog.md for details and release history.

Installation

MMGeneration depends on PyTorch and MMCV. Below are quick steps for installation.

Step 1. Install PyTorch following official instructions, e.g.

pip3 install torch torchvision

Step 2. Install MMCV with MIM.

pip3 install openmim
mim install mmcv-full

Step 3. Install MMGeneration from source.

git clone https://github.com/open-mmlab/mmgeneration.git
cd mmgeneration
pip3 install -e .

Please refer to get_started.md for more detailed instruction.

Getting Started

Please see get_started.md for the basic usage of MMGeneration. docs/en/quick_run.md can offer full guidance for quick run. For other details and tutorials, please go to our documentation.

ModelZoo

These methods have been carefully studied and supported in our frameworks:

<details open> <summary>Unconditional GANs (click to collapse)</summary>

✅ DCGAN (ICLR'2016)
✅ WGAN-GP (NIPS'2017)
✅ LSGAN (ICCV'2017)
✅ GGAN (arXiv'2017)
✅ PGGAN (ICLR'2018)
✅ StyleGANV1 (CVPR'2019)
✅ StyleGANV2 (CVPR'2020)
✅ StyleGANV3 (NeurIPS'2021)
✅ Positional Encoding in GANs (CVPR'2021)

</details> <details open> <summary>Conditional GANs (click to collapse)</summary>

✅ SNGAN (ICLR'2018)
✅ Projection GAN (ICLR'2018)
✅ SAGAN (ICML'2019)
✅ BIGGAN/BIGGAN-DEEP (ICLR'2019)

</details> <details open> <summary>Tricks for GANs (click to collapse)</summary>

✅ ADA (NeurIPS'2020)

</details> <details open> <summary>Image2Image Translation (click to collapse)</summary>

✅ Pix2Pix (CVPR'2017)
✅ CycleGAN (ICCV'2017)

</details> <details open> <summary>Internal Learning (click to collapse)</summary>

✅ SinGAN (ICCV'2019)

</details> <details open> <summary>Denoising Diffusion Probabilistic Models (click to collapse)</summary>

✅ Improved DDPM (arXiv'2021)

</details>

Related-Applications

✅ MMGEN-FaceStylor

Contributing

We appreciate all contributions to improve MMGeneration. Please refer to CONTRIBUTING.md in MMCV for more details about the contributing guideline.

Citation

If you find this project useful in your research, please consider cite:

@misc{2021mmgeneration,
    title={{MMGeneration}: OpenMMLab Generative Model Toolbox and Benchmark},
    author={MMGeneration Contributors},
    howpublished = {\url{https://github.com/open-mmlab/mmgeneration}},
    year={2021}
}

License

This project is released under the Apache 2.0 license. Some operations in MMGeneration are with other licenses instead of Apache2.0. Please refer to LICENSES.md for the careful check, if you are using our code for commercial matters.

Projects in OpenMMLab

MMCV: OpenMMLab foundational library for computer vision.
MIM: MIM installs OpenMMLab packages.
MMClassification: OpenMMLab image classification toolbox and benchmark.
MMDetection: OpenMMLab detection toolbox and benchmark.
MMDetection3D: OpenMMLab's next-generation platform for general 3D object detection.
MMRotate: OpenMMLab rotated object detection toolbox and benchmark.
MMSegmentation: OpenMMLab semantic segmentation toolbox and benchmark.
[

Mmgeneration

Install / Use

README