{"id":21354664,"url":"https://github.com/iveevi/zhetapi","last_synced_at":"2025-09-10T14:33:20.126Z","repository":{"id":41340931,"uuid":"220569420","full_name":"iveevi/zhetapi","owner":"iveevi","description":"A C++ ML and numerical analysis API, with an accompanying scripting language.","archived":false,"fork":false,"pushed_at":"2023-03-20T19:56:00.000Z","size":42754,"stargazers_count":47,"open_issues_count":0,"forks_count":2,"subscribers_count":3,"default_branch":"focused","last_synced_at":"2024-01-25T23:53:50.663Z","etag":null,"topics":["machine-learning","numerical-analysis","numerical-methods","scripting-language","symbolic-computation"],"latest_commit_sha":null,"homepage":"https://zhetapi.readthedocs.io/en/latest/","language":"C++","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/iveevi.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null}},"created_at":"2019-11-09T00:33:49.000Z","updated_at":"2024-01-25T23:53:52.718Z","dependencies_parsed_at":"2024-01-26T00:04:01.843Z","dependency_job_id":null,"html_url":"https://github.com/iveevi/zhetapi","commit_stats":null,"previous_names":["iveevi/zhetapi"],"tags_count":4,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iveevi%2Fzhetapi","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iveevi%2Fzhetapi/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iveevi%2Fzhetapi/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/iveevi%2Fzhetapi/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/iveevi","download_url":"https://codeload.github.com/iveevi/zhetapi/tar.gz/refs/heads/focused","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":225839605,"owners_count":17532307,"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":["machine-learning","numerical-analysis","numerical-methods","scripting-language","symbolic-computation"],"created_at":"2024-11-22T04:14:11.405Z","updated_at":"2024-11-22T04:14:12.005Z","avatar_url":"https://github.com/iveevi.png","language":"C++","funding_links":[],"categories":[],"sub_categories":[],"readme":"![Zhetapi Logo](zhetapi_logo.svg)\n\n![build badge](https://github.com/vedavamadathil/zhetapi/actions/workflows/cmake.yml/badge.svg)\n\nZhetapi is a modern C++ machine learning and numerical analysis library with an\nemphasis on intuitive usage.\n\n# Linear Algebra\n\nZhetapi provides a basic linear algebra interface using C++ 20 concepts. The\n`Field` structure enforces a contraint on types to ensure they behave like\nalgebraic fields. As a result these structures support basic arithmetic along\nwith other useful methods.\n\nFor now, the following template classes are provided: `Tensor`, `Matrix`,\n`Vector`.\n\n# Auto Differentiation\n\nThe auto differentiation facilities in Zhetapi belong in the `zhetapi::autograd`\nnamespace. All operations that depend on autodiff use `float`s as the underlying\ntype; in particular `Constant` is a `Tensor \u003cfloat\u003e` and is the basis of all\nnumerical values in this module.\n\nTo provide a seamless, operator based interface into the autodiff facilities,\ntwo notable classes are provided, `Variable` and `Function`. As one can expect,\n`Variable`s can store arbitrary `Constant` values, and `Function`s are\ncompositions of `Variables` under varying operations. For example:\n\n```cpp\nFunction f = x + y;\nFunction g = x * y;\n\n// f and g are now functions of *two* variables\nConstant a = f(1, 2);\nConstant b = g(1, 2);\n\n// Composition of functions is done likewise\nFunction h = f(x, g(x, y)); // NOTE: h is still a function of two variables\n```\n\n## Symbolic Differentiation\n\n## Backward Pass\n\nCurrently, only backward mode is enabled for autograd.\n\n# Building\n\nZhetapi is primarily a header-only library, but for now there are some examples\nthat one can play around with in the `experimental` directory.\n\nThis project is developed using C++ 20. Additional dependenies include PNG\n(`libpng-dev` on Ubuntu systems), OpenMP (Optional) and CUDA (Optional).\n\nGenerate the build configuration using CMake as follows:\n\n```\n$ cd zhetapi\n$ mkdir build \u0026\u0026 cd build\n$ cmake -DZHETAPI_ENABLE_CUDA=\u003cON|OFF\u003e # ON by default\n```\n\nAnd build the targets as one would usually do (e.g. `make` or `ninja`).\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fiveevi%2Fzhetapi","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fiveevi%2Fzhetapi","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fiveevi%2Fzhetapi/lists"}