{"id":26415614,"url":"https://github.com/hypergan/hypergan","last_synced_at":"2025-05-16T05:03:08.203Z","repository":{"id":41432213,"uuid":"61916488","full_name":"HyperGAN/HyperGAN","owner":"HyperGAN","description":"Composable GAN framework with api and user interface","archived":false,"fork":false,"pushed_at":"2023-01-21T17:54:27.000Z","size":39062,"stargazers_count":1190,"open_issues_count":20,"forks_count":167,"subscribers_count":41,"default_branch":"master","last_synced_at":"2024-04-13T21:53:23.911Z","etag":null,"topics":["artificial-intelligence","computer-vision","gan","generative-adversarial-network","hypergan","machine-learning","machine-learning-api","online-learning","python","pytorch","sponsors","unsupervised-learning"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/HyperGAN.png","metadata":{"files":{"readme":"README.md","changelog":"changelog.md","contributing":"CONTRIBUTING.md","funding":".github/FUNDING.yml","license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":"SECURITY.md","support":null},"funding":{"github":["hypergan"]}},"created_at":"2016-06-24T22:49:56.000Z","updated_at":"2024-03-28T11:11:27.000Z","dependencies_parsed_at":"2023-02-12T11:48:26.133Z","dependency_job_id":null,"html_url":"https://github.com/HyperGAN/HyperGAN","commit_stats":null,"previous_names":["255bits/hypergan"],"tags_count":16,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HyperGAN%2FHyperGAN","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HyperGAN%2FHyperGAN/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HyperGAN%2FHyperGAN/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/HyperGAN%2FHyperGAN/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/HyperGAN","download_url":"https://codeload.github.com/HyperGAN/HyperGAN/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":247856544,"owners_count":21007621,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["artificial-intelligence","computer-vision","gan","generative-adversarial-network","hypergan","machine-learning","machine-learning-api","online-learning","python","pytorch","sponsors","unsupervised-learning"],"created_at":"2025-03-18T00:53:55.869Z","updated_at":"2025-04-08T14:13:13.189Z","avatar_url":"https://github.com/HyperGAN.png","language":"Python","readme":"# README\n\n## HyperGAN 1.0\n\n[![docs](https://img.shields.io/badge/gitbook-docs-yellowgreen)](https://hypergan.gitbook.io/hypergan/) [![Discord](https://img.shields.io/badge/discord-join%20chat-brightgreen.svg)](https://discord.gg/t4WWBPF) [![Twitter](https://img.shields.io/badge/twitter-follow-blue.svg)](https://twitter.com/hypergan)\n\nA composable GAN built for developers, researchers, and artists.\n\nHyperGAN is in pre-release and open beta.\n\n![Colorizer 0.9 1](https://s3.amazonaws.com/hypergan-apidocs/0.9.0-images/colorizer-2.gif)\n\n_Logos generated with_ [_examples/colorizer_](./examples/colorizer.py)\n\nSee more on the [hypergan youtube](https://www.youtube.com/channel/UCU33XvBbMnS8002_NB7JSvA)\n\n## Table of contents\n\n* [About](#about)\n* [Documentation](https://hypergan.gitbook.io/hypergan/)\n* [Changelog](./changelog.md)\n* [Quick start](#quick-start)\n  * [Requirements](#requirements)\n  * [Install](#install)\n  * [Train](#train)\n* [API](#api)\n  * [Using a trained hypergan model](#using-a-trained-hypergan-model)\n  * [Training a gan](#training-a-gan)\n  * [Examples](#examples)\n  * [Tutorials](#tutorials)\n* [The pip package hypergan](#the-pip-package-hypergan)\n  * [Training](#training)\n  * [Sampling](#sampling)\n  * [Additional Arguments](#additional-arguments)\n  * [Running on CPU](#running-on-cpu)\n  * [Troubleshooting](#troubleshooting)\n  * [Development Mode](#development-mode)\n* [Datasets](#datasets)\n  * [Creating a Dataset](#creating-a-dataset)\n  * [Downloadable Datasets](#downloadable-datasets)\n  * [Cleaning up data](#cleaning-up-data)\n* [Features](#features)\n* [Showcase](#showcase)\n* [Sponsors](#sponsors)\n* [Contributing](./#contributing.md)\n* [Versioning](#Versioning)\n* [Citation](#citation)\n\n## About\n\nHyperGAN builds generative adversarial networks in pytorch and makes them easy to train and share.\n\nFor a general introduction to GANs see [http://blog.aylien.com/introduction-generative-adversarial-networks-code-tensorflow/](http://blog.aylien.com/introduction-generative-adversarial-networks-code-tensorflow/)\n\nJoin the community [discord](https://discord.gg/t4WWBPF).\n\n## Documentation\n\n* [Gitbook documentation](https://hypergan.gitbook.io/)\n\n## Changelog\n\nSee the full changelog here: [Changelog.md](changelog.md)\n\n## Quick start\n\n### Requirements\n\nOS: Windows, OSX, Linux\n\nFor training:\n\nGPU: Nvidia, GTX 1080+ recommended\n\n### Install\n\n1. Install HyperGAN\n  For users: `pip3 install hypergan`\n\n  For developers: Download this repo and run `python3 setup.py develop`\n\n2. Test it out\n  * `hypergan train preset:celeba -s 128x128x3`\n\n3. Join the community\n  * Once you've made something cool, be sure to share it on the Discord \\([https://discord.gg/t4WWBPF](https://discord.gg/t4WWBPF)\\).\n\n### Create a new model\n\n```bash\n  hypergan new mymodel\n```\n\nThis will create a mymodel.json based off the default configuration. You can change configuration templates with the `-c` flag.\n\n### List configuration templates\n\n```bash\n  hypergan new mymodel -l\n```\n\nSee all configuration templates with `--list-templates` or `-l`.\n\n### Train\n\n```bash\n  hypergan train folder/ -s 32x32x3 -c mymodel --resize\n```\n\n## API\n\n```python\nimport hypergan as hg\n```\n\nNote this API is currently under work in 1.0. If you are reading this before 1.0 is released check the examples.\n\nSee the [gitbook documentation](https://hypergan.gitbook.io/) for more details.\n\n### Using a trained hypergan model\n\n```python\nmy_gan = hg.GAN('model.hypergan')\nbatch_sample = my_gan.sample()\n```\n\n### Training a gan\n\n```python\ngan = hg.GAN(\"default.json\", inputs=hg.inputs.ImageLoader(...))\ntrainable_gan = hg.TrainableGAN(gan)\nfor step in trainable_gan.train():\n    print(\"I'm on step \", step)\n```\n\n### Examples\n\nSee the examples [https://github.com/hypergan/HyperGAN/tree/master/examples](https://github.com/hypergan/HyperGAN/tree/master/examples)\n\n### Tutorials\n\nSee the tutorials [https://hypergan.gitbook.io/hypergan/tutorials](https://hypergan.gitbook.io/hypergan/tutorials)\n\n## The pip package hypergan\n\n```bash\npip install hypergan\n```\n\n### Training\n\n```bash\n  # Train a 32x32 gan with batch size 32 on a folder of pngs\n  hypergan train [folder] -s 32x32x3 -b 32 --config [name]\n```\n\n### Sampling\n\n```bash\n  hypergan sample [folder] -s 32x32x3 -b 32 --config [name] --sampler batch_walk --save_samples\n```\n\nBy default hypergan will not save training samples to disk. To change this, use `--save_samples`.\n\n### Additional Arguments\n\nTo see a detailed list, run\n\n```bash\n  hypergan -h\n```\n\n### Running on CPU\n\nYou can switch the backend with:\n\n```bash\n  hypergan [...] -B cpu\n```\n\nDon't train on CPU! It's too slow.\n\n### Troubleshooting\n\nMake sure that your cuda, nvidia drivers, pillow, pytorch, and pytorch vision are the latest version.\n\nCheck the discord for help.\n\n### Development mode\n\nIf you wish to modify hypergan\n\n```bash\ngit clone https://github.com/hypergan/hypergan\ncd hypergan\npython3 setup.py develop\n```\n\nMake sure to `pip3 uninstall hypergan` to avoid version conflicts.\n\n## Datasets\n\nTo build a new network you need a dataset.\n\n### Creating a Dataset\n\nDatasets in HyperGAN are meant to be simple to create. Just use a folder of images. Nested folders work too.\n\n### Cleaning up data\n\nHyperGAN is built to be resilient to all types of unclean data. By default images are resized then cropped if necessary.\n\nSee `--nocrop`, `--random_crop` and `--resize` for additional image scaling options.\n\n## Features\n\nA list of features in the 1.0 release:\n\n* API\n* CLI\n* Viewer - an electron app to explore and create models\n* Cross platform - Windows, OSX, Linux\n* Inference - Add AI content generation to your project\n* Training - Train custom models using accelerated parallel training backends\n* Sharing - Share built models with each other. Use them in python projects as hypergan models, or in any project as onxx models\n* Customizable - Define custom architectures in the json, or replace any component with your own pytorch creation\n* Data - Built to work on unclean data and multiple data types\n* Unsupervised learning\n* Unsupervised alignment - Align one distribution to another or discover new novel distributions.\n* Transfer learning\n* Online learning\n\n## Showcase\n\n### 1.0 models are still training\n\nSubmit your showcase with a pull request!\n\nFor more, see the \\#showcase room in [![Discord](https://img.shields.io/badge/discord-join%20chat-brightgreen.svg)](https://discord.gg/t4WWBPF)\n\n## Sponsors\n\nWe are now accepting financial sponsors. Sponsor to (optionally) be listed here.\n\nhttps://github.com/sponsors/hypergan\n\n## Contributing\n\nContributions are welcome and appreciated! We have many open issues in the _Issues_ tab. Join the discord.\n\nSee [how to contribute.](./)\n\n## Versioning\n\nHyperGAN uses semantic versioning. [http://semver.org/](http://semver.org/)\n\nTLDR: _x.y.z_\n\n* _x_ is incremented on stable public releases.\n* _y_ is incremented on API breaking changes.  This includes configuration file changes and graph construction changes.\n* _z_ is incremented on non-API breaking changes.  _z_ changes will be able to reload a saved graph.\n\n## Citation\n\n```text\n  HyperGAN Community\n  HyperGAN, (2016-2020+), \n  GitHub repository, \n  https://github.com/HyperGAN/HyperGAN\n```\n\nHyperGAN comes with no warranty or support.\n\n","funding_links":["https://github.com/sponsors/hypergan"],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhypergan%2Fhypergan","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fhypergan%2Fhypergan","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fhypergan%2Fhypergan/lists"}