Image
OmniGen
OmniGen: one unified model for text-to-image, editing, subject-driven and identity-preserving generation
★ 4.3k starsMIT
View guide →MMagic — OpenMMLab's generative-AI toolbox: diffusion, text-to-image, restoration and a model zoo.
mmagic is an open-source image project by open-mmlab (open-mmlab/mmagic on GitHub). It has 7,472 GitHub stars and 1,097 forks, is released under the Apache-2.0 license, and was last updated on GitHub on 6 August 2024.
Pick your task (generation, restoration, inpainting and so on) and the algorithm/config you will start from in the model zoo.
Clone open-mmlab/mmagic from GitHub (or use the install path in the run steps below) and skim its README for the parts relevant to your goal.
Clone the repo and install it with the OpenMMLab toolchain (mim/pip) per the official documentation for your PyTorch and CUDA versions.
Run a documented demo config — for example a text-to-image or super-resolution example — on a sample input before training anything.
Text-to-image, diffusion, image/video restoration and enhancement, inpainting, matting, colorisation and video interpolation in one toolbox.
A curated zoo of algorithms (33+ accelerated by PyTorch 2.0) driven by OpenMMLab-style configs for reproducible experiments.
Stable Diffusion and Disco Diffusion generation, DreamBooth and DreamBooth LoRA fine-tuning, and ControlNet-based controllability.
xFormers-based acceleration and optimisation strategies to improve training and inference efficiency.
Refactored DataSample and DataPreprocessor unify data formats across tasks during training and inference.
Multi-dataset evaluation supporting generation metrics (for example FID) and reconstruction metrics (for example SSIM), with TensorBoard and wandb visualisation.
Get the source from the official GitHub repository for mmagic.
git clone https://github.com/open-mmlab/mmagic.git
cd mmagicClone the repository as the first step of the documented OpenMMLab installation.
git clone https://github.com/open-mmlab/mmagic.git
cd mmagicInstall MMagic with the OpenMMLab toolchain for your environment, then run a documented demo config from the docs.
# Install per https://mmagic.readthedocs.io/en/latest/ (mim/pip for your PyTorch/CUDA), then run a demo — see READMEopen-mmlab (OpenMMLab), through the open-mmlab/mmagic repository, with documentation at mmagic.readthedocs.io.
Yes — the project is open source under the Apache-2.0 license as recorded for this listing. See the license terms for any conditions on redistribution.
No API keys are involved — models run locally on your GPUs; MMagic is free under Apache-2.0, and you supply the compute and datasets.
Single-model libraries do one pipeline well. MMagic is a research toolbox spanning many generative tasks and algorithms with shared configs, data handling and evaluation — broader and more configurable, with a steeper OpenMMLab learning curve.
Clone the repository (or use the documented installer), follow the install and configure steps on this page, and run the smallest real task first — run a documented demo config — for example a text-to-image or super-resolution example — on a sample input before training anything.
OmniGen: one unified model for text-to-image, editing, subject-driven and identity-preserving generation
Kandinsky 2.2 — multilingual text-to-image latent diffusion with ControlNet, inpainting and image fusion
C# generative AI SDK — text, image, video and audio via ONNX
Diffusion model(SD,Flux,Wan,Qwen Image,Z-Image,...) inference in pure C/C++
Tencent's HunyuanImage-3.0 — a powerful native multimodal model for image generation.
AI LinkedIn content studio — text, image and carousel posts on a calendar
The most-starred other Image projects in the LibreAI catalogue — the realistic shortlist if mmagic is not the right fit.
Stable Diffusion web UI
The node-graph powerhouse for precise, professional AI image and video generation.
High performance self-hosted photo and video management solution
Focus on prompting and generating
🆙 Upscayl - #1 Free and Open Source AI Image Upscaler for Linux, MacOS and Windows
Free and open — grab the source, star the repo, and start creating.