[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-72161":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":23,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":27,"readmeContent":28,"aiSummary":29,"trendingCount":16,"starSnapshotCount":16,"syncStatus":30,"lastSyncTime":31,"discoverSource":32},72161,"flow_matching","facebookresearch\u002Fflow_matching","facebookresearch","A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.","http:\u002F\u002Ffacebookresearch.github.io\u002Fflow_matching",null,"Python",4528,355,34,13,0,17,43,111,51,29.65,"Other",false,"main",true,[],"2026-06-12 02:02:59","\u003Cdiv align=\"center\">\n\n# Flow Matching\n\n[![arXiv](assets\u002FarXiv-2412.06264-red.svg)](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.06264)\n[![CI](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fflow_matching\u002Factions\u002Fworkflows\u002Fci.yaml\u002Fbadge.svg)](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fflow_matching\u002Factions\u002Fworkflows\u002Fci.yaml)\n[![Coverage](https:\u002F\u002Fgithub.com\u002Ffacebookresearch\u002Fflow_matching\u002Fraw\u002Frefs\u002Fheads\u002Fgh-pages\u002Fcoverage\u002Fcoverage-badge.svg)](https:\u002F\u002Fstunning-potato-4k4z71e.pages.github.io\u002Fcoverage\u002F)\n[![License: CC BY-NC 4.0](assets\u002FLicense-CC_BY--NC_4.0-lightgrey.svg)](https:\u002F\u002Fcreativecommons.org\u002Flicenses\u002Fby-nc\u002F4.0\u002F)\n[![PyPI](https:\u002F\u002Fimg.shields.io\u002Fpypi\u002Fv\u002Fflow-matching)](https:\u002F\u002Fpypi.org\u002Fproject\u002Fflow-matching\u002F)\n\n\n\u003C\u002Fdiv>\n\n`flow_matching` is a PyTorch library for Flow Matching algorithms, featuring continuous and discrete implementations. It includes examples for both text and image modalities. This repository is part of [Flow Matching Guide and Codebase](https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.06264).\n\n\n![](.\u002Fassets\u002Fteaser.png)\n\n## Installation\n\nThis repository requires Python 3.9 and Pytorch 2.1 or greater. To install the latest version run:\n```\npip install flow_matching\n```\n\n## Repository structure\n\nThe core and example folders are structured in the following way:\n```bash\n.\n├── flow_matching                  # Core library\n│   ├── loss                       # Loss functions\n│   │   └── ...\n│   ├── path                       # Path and schedulers\n│   │   ├── ...\n│   │   └── scheduler              # Schedulers and transformations\n│   │       └── ...\n│   ├── solver                     # Solvers for continuous and discrete flows\n│   │   └── ...\n│   └── utils\n│       └── ...\n└── examples                       # Synthetic, image, and text examples\n    ├── ...\n    ├── image\n    │       └── ...\n    └── text \n            └── ...\n```\n\n## Development\n\nTo create a conda environment with all required dependencies, run:\n```\nconda env create -f environment.yml\nconda activate flow_matching\n```\n\nInstall pre-commit hook. This will ensure that all linting is done on each commit\n```\npre-commit install\n```\n\nInstall the `flow_matching` package in an editable mode:\n```\npip install -e .\n```\n\n## FAQ\n\n#### I want to train a Flow Matching model, where can I find the training code?\n\nWe provide [training examples](examples). Under this folder, you can find synthetic data for [continuous](examples\u002F2d_flow_matching.ipynb), [discrete](examples\u002F2d_discrete_flow_matching.ipynb), and [Riemannian](examples\u002F2d_riemannian_flow_matching_flat_torus.ipynb) Flow Matching. We also provide full training [examples](examples\u002Fimage) (continuous and discrete) on CIFAR10 and face-blurred ImageNet, and a scalable discrete Flow Matching example for [text modeling](examples\u002Ftext).\n\n#### Do you release pre-trained models?\n\nIn this version, we don't release pre-trained models. All models under [examples](examples) can be trained from scratch by a single running command. \n\n#### How to contribute to this codebase?\nPlease follow the [contribution guide](CONTRIBUTING.md).\n\n## License\n\nThe code in this repository is CC BY-NC licensed. See the [LICENSE](LICENSE) for details.\n\n## Citation\n\nIf you found this repository useful, please cite the following.\n\n```\n@misc{lipman2024flowmatchingguidecode,\n      title={Flow Matching Guide and Code}, \n      author={Yaron Lipman and Marton Havasi and Peter Holderrieth and Neta Shaul and Matt Le and Brian Karrer and Ricky T. Q. Chen and David Lopez-Paz and Heli Ben-Hamu and Itai Gat},\n      year={2024},\n      eprint={2412.06264},\n      archivePrefix={arXiv},\n      primaryClass={cs.LG},\n      url={https:\u002F\u002Farxiv.org\u002Fabs\u002F2412.06264}, \n}\n```\n","`flow_matching` 是一个基于 PyTorch 的库，用于实现连续和离散的流匹配算法，并提供了文本和图像模态的实际示例。其核心功能包括多种损失函数、路径调度器以及针对连续和离散流的求解器，支持用户在不同应用场景下灵活选择。该项目特别适合需要进行数据生成或转换的研究人员和开发者使用，如图像处理、自然语言处理等领域中的任务。通过简洁的安装步骤与详细的文档说明，即使是初学者也能快速上手并根据自身需求定制模型。",2,"2026-06-11 03:40:37","high_star"]