{"id":14958080,"url":"https://github.com/pytorch/fbgemm","last_synced_at":"2025-05-13T00:11:29.210Z","repository":{"id":37253653,"uuid":"150154628","full_name":"pytorch/FBGEMM","owner":"pytorch","description":"FB (Facebook) + GEMM (General Matrix-Matrix Multiplication) - https://code.fb.com/ml-applications/fbgemm/","archived":false,"fork":false,"pushed_at":"2025-05-12T22:45:27.000Z","size":28248,"stargazers_count":1320,"open_issues_count":490,"forks_count":572,"subscribers_count":65,"default_branch":"main","last_synced_at":"2025-05-13T00:11:20.930Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":"","language":"C++","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"other","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/pytorch.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null}},"created_at":"2018-09-24T19:07:42.000Z","updated_at":"2025-05-12T22:03:21.000Z","dependencies_parsed_at":"2023-12-15T21:09:05.490Z","dependency_job_id":"589d2e1f-8a0e-46b5-94da-73b87777d6cb","html_url":"https://github.com/pytorch/FBGEMM","commit_stats":{"total_commits":2843,"total_committers":342,"mean_commits":8.312865497076023,"dds":0.884277172001407,"last_synced_commit":"8f6d96df19bcc5a3bb64400e1069469b07447978"},"previous_names":[],"tags_count":47,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pytorch%2FFBGEMM","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pytorch%2FFBGEMM/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pytorch%2FFBGEMM/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/pytorch%2FFBGEMM/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/pytorch","download_url":"https://codeload.github.com/pytorch/FBGEMM/tar.gz/refs/heads/main","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":253843219,"owners_count":21972874,"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":[],"created_at":"2024-09-24T13:16:12.029Z","updated_at":"2025-05-13T00:11:29.182Z","avatar_url":"https://github.com/pytorch.png","language":"C++","funding_links":[],"categories":[],"sub_categories":[],"readme":"# The FBGEMM Project\n\nThe FBGEMM Project is a repository of highly-optimized kernels used across\ndeep learning applications.\n\nThe codebase is organized and published as three related packages: FBGEMM,\nFBGEMM-GPU, and FBGEMM-GenAI. Each package has its own set of features and\ndocumentation.\n\n### Project Overview\n\n* **FBGEMM**: A low-precision, high-performance matrix multiplication and\n  convolution library for server-side inference.  The documentation below\n  provides an overview of FBGEMM, including its features, documentation, and\n  community resources.\n\n* **FBGEMM_GPU**: A collection of PyTorch GPU operator libraries built on top of\n  FBGEMM for training and inference, with focus on recommendation systems\n  applications.  Please see [the documentation](fbgemm_gpu/README.md) for more\n  information.\n\n* **FBGEMM_GPU GenAI**: A collection of PyTorch GPU operator libraries that are\n  designed for generative AI applications, such as FP8 row-wise quantization and\n  collective communications. Please see [the documentation](fbgemm_gpu/experimental/gen_ai/README.md)\n  for more information.\n\n\n## FBGEMM\n\n[![FBGEMM CI](https://github.com/pytorch/FBGEMM/actions/workflows/fbgemm_ci.yml/badge.svg)](https://github.com/pytorch/FBGEMM/actions/workflows/fbgemm_ci.yml)\n\nFBGEMM (Facebook GEneral Matrix Multiplication) is a low-precision,\nhigh-performance matrix-matrix multiplications and convolution library for\nserver-side inference.\n\nThe library provides efficient low-precision general matrix multiplication for\nsmall batch sizes and support for accuracy-loss minimizing techniques such as\nrow-wise quantization and outlier-aware quantization. FBGEMM also exploits\nfusion opportunities in order to overcome the unique challenges of matrix\nmultiplication at lower precision with bandwidth-bound operations.\n\nFBGEMM is used as a backend of PyTorch quantized operators for x86 machines:\n\n  * PyTorch: https://github.com/pytorch/pytorch/tree/master/aten/src/ATen/native/quantized/cpu\n\nSee the full [Documentation](https://pytorch.org/FBGEMM) for more information\non building, installing, and developing with FBGEMM, as well as the most\nup-to-date support matrix and API documentation for this library.\n\n### What's New?\n\n* [New Features and Recent Improvements](https://github.com/pytorch/FBGEMM/wiki/Recent-feature-additions-and-improvements-in-FBGEMM) (January, 2020)\n\n### Citation\n\nFor a high-level overview, design philosophy and brief descriptions of various\nparts of FBGEMM please see [our blog post](https://code.fb.com/ml-applications/fbgemm).\n\nFor those looking for the appropriate article to cite regarding FBGEMM, we\nrecommend citing our [paper](https://arxiv.org/pdf/2101.05615.pdf):\n\n```\n@article{fbgemm,\n  title={FBGEMM: Enabling High-Performance Low-Precision Deep Learning Inference},\n  author={Khudia, Daya and Huang, Jianyu and Basu, Protonu and Deng, Summer and Liu, Haixin and Park, Jongsoo and Smelyanskiy, Mikhail},\n  journal={arXiv preprint arXiv:2101.05615},\n  year={2021}\n}\n```\n\n\n## Join the FBGEMM community\n\nFor questions, support, news updates, or feature requests, please feel free to:\n\n* File a ticket in [GitHub Issues](https://github.com/pytorch/FBGEMM/issues)\n* Post a discussion in [GitHub Discussions](https://github.com/pytorch/FBGEMM/discussions)\n* Reach out to us on the `#fbgemm` channel in [PyTorch Slack](https://bit.ly/ptslack)\n\nFor contributions, please see the [`CONTRIBUTING`](./CONTRIBUTING.md) file for\nways to help out.\n\n\n## License\n\nFBGEMM is BSD licensed, as found in the [`LICENSE`](LICENSE) file.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpytorch%2Ffbgemm","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fpytorch%2Ffbgemm","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fpytorch%2Ffbgemm/lists"}