🕷️ Crawler Inspector

URL Lookup

Direct Parameter Lookup

Raw Queries and Responses

1. Shard Calculation

Query:

Response:

Calculated Shard: 59 (from laksa163)

2. Crawled Status Check

Query:

curl -X POST \
  'http://laksa059.int.ahrefs:8124/' \
  -H 'Content-Type: text/plain' \
  -H 'X-ClickHouse-Database: crawler3' \
  -H 'Authorization: Basic YXBpOg==' \
  -d 'SELECT getAhrefsURLFromUnparsed(src_unparsed) AS found_url, ifNull(toUnixTimestamp(download_stamp), 0) AS crawl_time, ifNull(toUnixTimestamp(props_url_first_seen), 0) AS first_indexed_time, download_http_code AS http_code, src_unparsed AS src_unparsed, src_root_hash AS src_root_hash, history_drop_reason AS history_drop_reason, meta_title AS meta_title, meta_descriptions AS meta_descriptions, attrs_boilerpipe_text AS attrs_boilerpipe_text, attrs_markdown AS attrs_markdown, attrs_readable_markdown AS attrs_readable_markdown, meta_canonical AS meta_canonical FROM crawler3.page_info_local FINAL PREWHERE (src_root_hash, src_unparsed) IN ((getAhrefsRootHashFromUnparsed(getAhrefsUnparsedNoserviceFromURL(\'https://pypi.org/project/torch/\')), getAhrefsUnparsedNoserviceFromURL(\'https://pypi.org/project/torch/\'))) FORMAT JSONEachRow'

Response:

{"found_url":"https:\/\/pypi.org\/project\/torch\/","crawl_time":1775037277,"first_indexed_time":1485192561,"http_code":200,"src_unparsed":"org,pypi!\/project\/torch\/ s443","src_root_hash":"7813724874982801459","history_drop_reason":null,"meta_title":"torch · PyPI","meta_descriptions":["Tensors and Dynamic neural networks in Python with strong GPU acceleration"],"attrs_boilerpipe_text":"PyTorch is a Python package that provides two high-level features:\nTensor computation (like NumPy) with strong GPU acceleration\nDeep neural networks built on a tape-based autograd system\nYou can reuse your favorite Python packages such as NumPy, SciPy, and Cython to extend PyTorch when needed.\nOur trunk health (Continuous Integration signals) can be found at\nhud.pytorch.org\n.\nMore About PyTorch\nA GPU-Ready Tensor Library\nDynamic Neural Networks: Tape-Based Autograd\nPython First\nImperative Experiences\nFast and Lean\nExtensions Without Pain\nInstallation\nBinaries\nNVIDIA Jetson Platforms\nFrom Source\nPrerequisites\nNVIDIA CUDA Support\nAMD ROCm Support\nIntel GPU Support\nGet the PyTorch Source\nInstall Dependencies\nInstall PyTorch\nAdjust Build Options (Optional)\nDocker Image\nUsing pre-built images\nBuilding the image yourself\nBuilding the Documentation\nBuilding a PDF\nPrevious Versions\nGetting Started\nResources\nCommunication\nReleases and Contributing\nThe Team\nLicense\nMore About PyTorch\nLearn the basics of PyTorch\nAt a granular level, PyTorch is a library that consists of the following components:\nComponent\nDescription\ntorch\nA Tensor library like NumPy, with strong GPU support\ntorch.autograd\nA tape-based automatic differentiation library that supports all differentiable Tensor operations in torch\ntorch.jit\nA compilation stack (TorchScript) to create serializable and optimizable models from PyTorch code\ntorch.nn\nA neural networks library deeply integrated with autograd designed for maximum flexibility\ntorch.multiprocessing\nPython multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and Hogwild training\ntorch.utils\nDataLoader and other utility functions for convenience\nUsually, PyTorch is used either as:\nA replacement for NumPy to use the power of GPUs.\nA deep learning research platform that provides maximum flexibility and speed.\nElaborating Further:\nA GPU-Ready Tensor Library\nIf you use NumPy, then you have used Tensors (a.k.a. ndarray).\nPyTorch provides Tensors that can live either on the CPU or the GPU and accelerates the\ncomputation by a huge amount.\nWe provide a wide variety of tensor routines to accelerate and fit your scientific computation needs\nsuch as slicing, indexing, mathematical operations, linear algebra, reductions.\nAnd they are fast!\nDynamic Neural Networks: Tape-Based Autograd\nPyTorch has a unique way of building neural networks: using and replaying a tape recorder.\nMost frameworks such as TensorFlow, Theano, Caffe, and CNTK have a static view of the world.\nOne has to build a neural network and reuse the same structure again and again.\nChanging the way the network behaves means that one has to start from scratch.\nWith PyTorch, we use a technique called reverse-mode auto-differentiation, which allows you to\nchange the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes\nfrom several research papers on this topic, as well as current and past work such as\ntorch-autograd\n,\nautograd\n,\nChainer\n, etc.\nWhile this technique is not unique to PyTorch, it's one of the fastest implementations of it to date.\nYou get the best of speed and flexibility for your crazy research.\nPython First\nPyTorch is not a Python binding into a monolithic C++ framework.\nIt is built to be deeply integrated into Python.\nYou can use it naturally like you would use\nNumPy\n\/\nSciPy\n\/\nscikit-learn\netc.\nYou can write your new neural network layers in Python itself, using your favorite libraries\nand use packages such as\nCython\nand\nNumba\n.\nOur goal is to not reinvent the wheel where appropriate.\nImperative Experiences\nPyTorch is designed to be intuitive, linear in thought, and easy to use.\nWhen you execute a line of code, it gets executed. There isn't an asynchronous view of the world.\nWhen you drop into a debugger or receive error messages and stack traces, understanding them is straightforward.\nThe stack trace points to exactly where your code was defined.\nWe hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines.\nFast and Lean\nPyTorch has minimal framework overhead. We integrate acceleration libraries\nsuch as\nIntel MKL\nand NVIDIA (\ncuDNN\n,\nNCCL\n) to maximize speed.\nAt the core, its CPU and GPU Tensor and neural network backends\nare mature and have been tested for years.\nHence, PyTorch is quite fast — whether you run small or large neural networks.\nThe memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives.\nWe've written custom memory allocators for the GPU to make sure that\nyour deep learning models are maximally memory efficient.\nThis enables you to train bigger deep learning models than before.\nExtensions Without Pain\nWriting new neural network modules, or interfacing with PyTorch's Tensor API was designed to be straightforward\nand with minimal abstractions.\nYou can write new neural network layers in Python using the torch API\nor your favorite NumPy-based libraries such as SciPy\n.\nIf you want to write your layers in C\/C++, we provide a convenient extension API that is efficient and with minimal boilerplate.\nNo wrapper code needs to be written. You can see\na tutorial here\nand\nan example here\n.\nInstallation\nBinaries\nCommands to install binaries via Conda or pip wheels are on our website:\nhttps:\/\/pytorch.org\/get-started\/locally\/\nNVIDIA Jetson Platforms\nPython wheels for NVIDIA's Jetson Nano, Jetson TX1\/TX2, Jetson Xavier NX\/AGX, and Jetson AGX Orin are provided\nhere\nand the L4T container is published\nhere\nThey require JetPack 4.2 and above, and\n@dusty-nv\nand\n@ptrblck\nare maintaining them.\nFrom Source\nPrerequisites\nIf you are installing from source, you will need:\nPython 3.10 or later\nA compiler that fully supports C++17, such as clang or gcc (gcc 9.4.0 or newer is required, on Linux)\nVisual Studio or Visual Studio Build Tool (Windows only)\n* PyTorch CI uses Visual C++ BuildTools, which come with Visual Studio Enterprise,\nProfessional, or Community Editions. You can also install the build tools from\nhttps:\/\/visualstudio.microsoft.com\/visual-cpp-build-tools\/\n. The build tools\ndo not\ncome with Visual Studio Code by default.\nAn example of environment setup is shown below:\nLinux:\n$\nsource\n<CONDA_INSTALL_DIR>\/bin\/activate\n$\nconda\ncreate\n-y\n-n\n<CONDA_NAME>\n$\nconda\nactivate\n<CONDA_NAME>\nWindows:\n$\nsource\n<CONDA_INSTALL_DIR>\n\\S\ncripts\n\\a\nctivate.bat\n$\nconda\ncreate\n-y\n-n\n<CONDA_NAME>\n$\nconda\nactivate\n<CONDA_NAME>\n$\ncall\n\"C:\\Program Files\\Microsoft Visual Studio\\<VERSION>\\Community\\VC\\Auxiliary\\Build\\vcvarsall.bat\"\nx64\nA conda environment is not required.  You can also do a PyTorch build in a\nstandard virtual environment, e.g., created with tools like\nuv\n, provided\nyour system has installed all the necessary dependencies unavailable as pip\npackages (e.g., CUDA, MKL.)\nNVIDIA CUDA Support\nIf you want to compile with CUDA support,\nselect a supported version of CUDA from our support matrix\n, then install the following:\nNVIDIA CUDA\nNVIDIA cuDNN\nv8.5 or above\nCompiler\ncompatible with CUDA\nNote: You could refer to the\ncuDNN Support Matrix\nfor cuDNN versions with the various supported CUDA, CUDA driver, and NVIDIA hardware.\nIf you want to disable CUDA support, export the environment variable\nUSE_CUDA=0\n.\nOther potentially useful environment variables may be found in\nsetup.py\n.  If\nCUDA is installed in a non-standard location, set PATH so that the nvcc you\nwant to use can be found (e.g.,\nexport PATH=\/usr\/local\/cuda-12.8\/bin:$PATH\n).\nIf you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are\navailable here\nAMD ROCm Support\nIf you want to compile with ROCm support, install\nAMD ROCm\n4.0 and above installation\nROCm is currently supported only for Linux systems.\nBy default the build system expects ROCm to be installed in\n\/opt\/rocm\n. If ROCm is installed in a different directory, the\nROCM_PATH\nenvironment variable must be set to the ROCm installation directory. The build system automatically detects the AMD GPU architecture. Optionally, the AMD GPU architecture can be explicitly set with the\nPYTORCH_ROCM_ARCH\nenvironment variable\nAMD GPU architecture\nIf you want to disable ROCm support, export the environment variable\nUSE_ROCM=0\n.\nOther potentially useful environment variables may be found in\nsetup.py\n.\nIntel GPU Support\nIf you want to compile with Intel GPU support, follow these\nPyTorch Prerequisites for Intel GPUs\ninstructions.\nIntel GPU is supported for Linux and Windows.\nIf you want to disable Intel GPU support, export the environment variable\nUSE_XPU=0\n.\nOther potentially useful environment variables may be found in\nsetup.py\n.\nGet the PyTorch Source\ngit\nclone\nhttps:\/\/github.com\/pytorch\/pytorch\ncd\npytorch\n# if you are updating an existing checkout\ngit\nsubmodule\nsync\ngit\nsubmodule\nupdate\n--init\n--recursive\nInstall Dependencies\nCommon\n# Run this command from the PyTorch directory after cloning the source code using the “Get the PyTorch Source“ section above\npip\ninstall\n--group\ndev\nOn Linux\npip\ninstall\nmkl-static\nmkl-include\n# CUDA only: Add LAPACK support for the GPU if needed\n# magma installation: run with active conda environment. specify CUDA version to install\n.ci\/docker\/common\/install_magma_conda.sh\n12\n.4\n# (optional) If using torch.compile with inductor\/triton, install the matching version of triton\n# Run from the pytorch directory after cloning\n# For Intel GPU support, please explicitly `export USE_XPU=1` before running command.\nmake\ntriton\nOn MacOS\n# Add this package on intel x86 processor machines only\npip\ninstall\nmkl-static\nmkl-include\n# Add these packages if torch.distributed is needed\nconda\ninstall\npkg-config\nlibuv\nOn Windows\npip\ninstall\nmkl-static\nmkl-include\n# Add these packages if torch.distributed is needed.\n# Distributed package support on Windows is a prototype feature and is subject to changes.\nconda\ninstall\n-c\nconda-forge\nlibuv\n=\n1\n.51\nInstall PyTorch\nOn Linux\nIf you're compiling for AMD ROCm then first run this command:\n# Only run this if you're compiling for ROCm\npython\ntools\/amd_build\/build_amd.py\nInstall PyTorch\n# the CMake prefix for conda environment\nexport\nCMAKE_PREFIX_PATH\n=\n\"\n${\nCONDA_PREFIX\n:-\n'$(dirname $(which conda))\/..\/'\n}\n:\n${\nCMAKE_PREFIX_PATH\n}\n\"\npython\n-m\npip\ninstall\n--no-build-isolation\n-v\n-e\n.\n# the CMake prefix for non-conda environment, e.g. Python venv\n# call following after activating the venv\nexport\nCMAKE_PREFIX_PATH\n=\n\"\n${\nVIRTUAL_ENV\n}\n:\n${\nCMAKE_PREFIX_PATH\n}\n\"\nOn macOS\npython\n-m\npip\ninstall\n--no-build-isolation\n-v\n-e\n.\nOn Windows\nIf you want to build legacy python code, please refer to\nBuilding on legacy code and CUDA\nCPU-only builds\nIn this mode PyTorch computations will run on your CPU, not your GPU.\npython -m pip install --no-build-isolation -v -e .\nNote on OpenMP: The desired OpenMP implementation is Intel OpenMP (iomp). In order to link against iomp, you'll need to manually download the library and set up the building environment by tweaking\nCMAKE_INCLUDE_PATH\nand\nLIB\n. The instruction\nhere\nis an example for setting up both MKL and Intel OpenMP. Without these configurations for CMake, Microsoft Visual C OpenMP runtime (vcomp) will be used.\nCUDA based build\nIn this mode PyTorch computations will leverage your GPU via CUDA for faster number crunching\nNVTX\nis needed to build PyTorch with CUDA.\nNVTX is a part of CUDA distributive, where it is called \"Nsight Compute\". To install it onto an already installed CUDA run CUDA installation once again and check the corresponding checkbox.\nMake sure that CUDA with Nsight Compute is installed after Visual Studio.\nCurrently, VS 2017 \/ 2019, and Ninja are supported as the generator of CMake. If\nninja.exe\nis detected in\nPATH\n, then Ninja will be used as the default generator, otherwise, it will use VS 2017 \/ 2019.\nIf Ninja is selected as the generator, the latest MSVC will get selected as the underlying toolchain.\nAdditional libraries such as\nMagma\n,\noneDNN, a.k.a. MKLDNN or DNNL\n, and\nSccache\nare often needed. Please refer to the\ninstallation-helper\nto install them.\nYou can refer to the\nbuild_pytorch.bat\nscript for some other environment variables configurations\ncmd\n\n:: Set the environment variables after you have downloaded and unzipped the mkl package,\n:: else CMake would throw an error as `Could NOT find OpenMP`.\nset CMAKE_INCLUDE_PATH={Your directory}\\mkl\\include\nset LIB={Your directory}\\mkl\\lib;%LIB%\n\n:: Read the content in the previous section carefully before you proceed.\n:: [Optional] If you want to override the underlying toolset used by Ninja and Visual Studio with CUDA, please run the following script block.\n:: \"Visual Studio 2019 Developer Command Prompt\" will be run automatically.\n:: Make sure you have CMake >= 3.12 before you do this when you use the Visual Studio generator.\nset CMAKE_GENERATOR_TOOLSET_VERSION=14.27\nset DISTUTILS_USE_SDK=1\nfor \/f \"usebackq tokens=*\" %i in (`\"%ProgramFiles(x86)%\\Microsoft Visual Studio\\Installer\\vswhere.exe\" -version [15^,17^) -products * -latest -property installationPath`) do call \"%i\\VC\\Auxiliary\\Build\\vcvarsall.bat\" x64 -vcvars_ver=%CMAKE_GENERATOR_TOOLSET_VERSION%\n\n:: [Optional] If you want to override the CUDA host compiler\nset CUDAHOSTCXX=C:\\Program Files (x86)\\Microsoft Visual Studio\\2019\\Community\\VC\\Tools\\MSVC\\14.27.29110\\bin\\HostX64\\x64\\cl.exe\n\npython -m pip install --no-build-isolation -v -e .\nIntel GPU builds\nIn this mode PyTorch with Intel GPU support will be built.\nPlease make sure\nthe common prerequisites\nas well as\nthe prerequisites for Intel GPU\nare properly installed and the environment variables are configured prior to starting the build. For build tool support,\nVisual Studio 2022\nis required.\nThen PyTorch can be built with the command:\n:: CMD Commands:\n:: Set the CMAKE_PREFIX_PATH to help find corresponding packages\n:: %CONDA_PREFIX% only works after `conda activate custom_env`\n\nif defined CMAKE_PREFIX_PATH (\n    set \"CMAKE_PREFIX_PATH=%CONDA_PREFIX%\\Library;%CMAKE_PREFIX_PATH%\"\n) else (\n    set \"CMAKE_PREFIX_PATH=%CONDA_PREFIX%\\Library\"\n)\n\npython -m pip install --no-build-isolation -v -e .\nAdjust Build Options (Optional)\nYou can adjust the configuration of cmake variables optionally (without building first), by doing\nthe following. For example, adjusting the pre-detected directories for CuDNN or BLAS can be done\nwith such a step.\nOn Linux\nexport\nCMAKE_PREFIX_PATH\n=\n\"\n${\nCONDA_PREFIX\n:-\n'$(dirname $(which conda))\/..\/'\n}\n:\n${\nCMAKE_PREFIX_PATH\n}\n\"\nCMAKE_ONLY\n=\n1\npython\nsetup.py\nbuild\nccmake\nbuild\n# or cmake-gui build\nOn macOS\nexport\nCMAKE_PREFIX_PATH\n=\n\"\n${\nCONDA_PREFIX\n:-\n'$(dirname $(which conda))\/..\/'\n}\n:\n${\nCMAKE_PREFIX_PATH\n}\n\"\nMACOSX_DEPLOYMENT_TARGET\n=\n11\n.0\nCMAKE_ONLY\n=\n1\npython\nsetup.py\nbuild\nccmake\nbuild\n# or cmake-gui build\nDocker Image\nUsing pre-built images\nYou can also pull a pre-built docker image from Docker Hub and run with docker v19.03+\ndocker\nrun\n--gpus\nall\n--rm\n-ti\n--ipc\n=\nhost\npytorch\/pytorch:latest\nPlease note that PyTorch uses shared memory to share data between processes, so if torch multiprocessing is used (e.g.\nfor multithreaded data loaders) the default shared memory segment size that container runs with is not enough, and you\nshould increase shared memory size either with\n--ipc=host\nor\n--shm-size\ncommand line options to\nnvidia-docker run\n.\nBuilding the image yourself\nNOTE:\nMust be built with a docker version > 18.06\nThe\nDockerfile\nis supplied to build images with CUDA 11.1 support and cuDNN v8.\nYou can pass\nPYTHON_VERSION=x.y\nmake variable to specify which Python version is to be used by Miniconda, or leave it\nunset to use the default.\nmake\n-f\ndocker.Makefile\n# images are tagged as docker.io\/${your_docker_username}\/pytorch\nYou can also pass the\nCMAKE_VARS=\"...\"\nenvironment variable to specify additional CMake variables to be passed to CMake during the build.\nSee\nsetup.py\nfor the list of available variables.\nmake\n-f\ndocker.Makefile\nBuilding the Documentation\nTo build documentation in various formats, you will need\nSphinx\nand the pytorch_sphinx_theme2.\nBefore you build the documentation locally, ensure\ntorch\nis\ninstalled in your environment. For small fixes, you can install the\nnightly version as described in\nGetting Started\n.\nFor more complex fixes, such as adding a new module and docstrings for\nthe new module, you might need to install torch\nfrom source\n.\nSee\nDocstring Guidelines\nfor docstring conventions.\ncd\ndocs\/\npip\ninstall\n-r\nrequirements.txt\nmake\nhtml\nmake\nserve\nRun\nmake\nto get a list of all available output formats.\nIf you get a katex error run\nnpm install katex\n.  If it persists, try\nnpm install -g katex\n[!NOTE]\nIf you installed\nnodejs\nwith a different package manager (e.g.,\nconda\n) then\nnpm\nwill probably install a version of\nkatex\nthat is not\ncompatible with your version of\nnodejs\nand doc builds will fail.\nA combination of versions that is known to work is\nnode@6.13.1\nand\nkatex@0.13.18\n. To install the latter with\nnpm\nyou can run\nnpm install -g katex@0.13.18\n[!NOTE]\nIf you see a numpy incompatibility error, run:\npip install 'numpy<2'\nWhen you make changes to the dependencies run by CI, edit the\n.ci\/docker\/requirements-docs.txt\nfile.\nBuilding a PDF\nTo compile a PDF of all PyTorch documentation, ensure you have\ntexlive\nand LaTeX installed. On macOS, you can install them using:\nbrew install --cask mactex\nTo create the PDF:\nRun:\nmake latexpdf\nThis will generate the necessary files in the\nbuild\/latex\ndirectory.\nNavigate to this directory and execute:\nmake LATEXOPTS=\"-interaction=nonstopmode\"\nThis will produce a\npytorch.pdf\nwith the desired content. Run this\ncommand one more time so that it generates the correct table\nof contents and index.\n[!NOTE]\nTo view the Table of Contents, switch to the\nTable of Contents\nview in your PDF viewer.\nPrevious Versions\nInstallation instructions and binaries for previous PyTorch versions may be found\non\nour website\n.\nGetting Started\nThree pointers to get you started:\nTutorials: get you started with understanding and using PyTorch\nExamples: easy to understand PyTorch code across all domains\nThe API Reference\nGlossary\nResources\nPyTorch.org\nPyTorch Tutorials\nPyTorch Examples\nPyTorch Models\nIntro to Deep Learning with PyTorch from Udacity\nIntro to Machine Learning with PyTorch from Udacity\nDeep Neural Networks with PyTorch from Coursera\nPyTorch Twitter\nPyTorch Blog\nPyTorch YouTube\nCommunication\nForums: Discuss implementations, research, etc.\nhttps:\/\/discuss.pytorch.org\nGitHub Issues: Bug reports, feature requests, install issues, RFCs, thoughts, etc.\nSlack: The\nPyTorch Slack\nhosts a primary audience of moderate to experienced PyTorch users and developers for general chat, online discussions, collaboration, etc. If you are a beginner looking for help, the primary medium is\nPyTorch Forums\n. If you need a slack invite, please fill this form:\nhttps:\/\/goo.gl\/forms\/PP1AGvNHpSaJP8to1\nNewsletter: No-noise, a one-way email newsletter with important announcements about PyTorch. You can sign-up here:\nhttps:\/\/eepurl.com\/cbG0rv\nFacebook Page: Important announcements about PyTorch.\nhttps:\/\/www.facebook.com\/pytorch\nFor brand guidelines, please visit our website at\npytorch.org\nReleases and Contributing\nTypically, PyTorch has three minor releases a year. Please let us know if you encounter a bug by\nfiling an issue\n.\nWe appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion.\nIf you plan to contribute new features, utility functions, or extensions to the core, please first open an issue and discuss the feature with us.\nSending a PR without discussion might end up resulting in a rejected PR because we might be taking the core in a different direction than you might be aware of.\nTo learn more about making a contribution to PyTorch, please see our\nContribution page\n. For more information about PyTorch releases, see\nRelease page\n.\nThe Team\nPyTorch is a community-driven project with several skillful engineers and researchers contributing to it.\nPyTorch is currently maintained by\nSoumith Chintala\n,\nGregory Chanan\n,\nDmytro Dzhulgakov\n,\nEdward Yang\n,\nAlban Desmaison\n,\nPiotr Bialecki\nand\nNikita Shulga\nwith major contributions coming from hundreds of talented individuals in various forms and means.\nA non-exhaustive but growing list needs to mention:\nTrevor Killeen\n,\nSasank Chilamkurthy\n,\nSergey Zagoruyko\n,\nAdam Lerer\n,\nFrancisco Massa\n,\nAlykhan Tejani\n,\nLuca Antiga\n,\nAlban Desmaison\n,\nAndreas Koepf\n,\nJames Bradbury\n,\nZeming Lin\n,\nYuandong Tian\n,\nGuillaume Lample\n,\nMarat Dukhan\n,\nNatalia Gimelshein\n,\nChristian Sarofeen\n,\nMartin Raison\n,\nEdward Yang\n,\nZachary Devito\n.\nNote: This project is unrelated to\nhughperkins\/pytorch\nwith the same name. Hugh is a valuable contributor to the Torch community and has helped with many things Torch and PyTorch.\nLicense\nPyTorch has a BSD-style license, as found in the\nLICENSE\nfile.","attrs_markdown":"[Skip to main content](https:\/\/pypi.org\/project\/torch\/#content)\n\nSwitch to mobile version\n\nWarning  Some features may not work without JavaScript. Please try enabling it if you encounter problems.\n\n[![PyPI](https:\/\/pypi.org\/static\/images\/logo-small.8998e9d1.svg)](https:\/\/pypi.org\/)\n\n- [Help](https:\/\/pypi.org\/help\/)\n- [Docs](https:\/\/docs.pypi.org\/)\n- [Sponsors](https:\/\/pypi.org\/sponsors\/)\n- [Log in](https:\/\/pypi.org\/account\/login\/?next=https%3A%2F%2Fpypi.org%2Fproject%2Ftorch%2F)\n- [Register](https:\/\/pypi.org\/account\/register\/)\n\nMenu\n\n- [Help](https:\/\/pypi.org\/help\/)\n- [Docs](https:\/\/docs.pypi.org\/)\n- [Sponsors](https:\/\/pypi.org\/sponsors\/)\n- [Log in](https:\/\/pypi.org\/account\/login\/?next=https%3A%2F%2Fpypi.org%2Fproject%2Ftorch%2F)\n- [Register](https:\/\/pypi.org\/account\/register\/)\n\n# torch 2.11.0\npip install torch Copy PIP instructions\n\n[Latest version](https:\/\/pypi.org\/project\/torch\/)\n\nReleased: Mar 23, 2026\n\nTensors and Dynamic neural networks in Python with strong GPU acceleration\n\n### Navigation\n- [Project description](https:\/\/pypi.org\/project\/torch\/#description)\n- [Release history](https:\/\/pypi.org\/project\/torch\/#history)\n- [Download files](https:\/\/pypi.org\/project\/torch\/#files)\n\n### Verified details\n*These details have been [verified by PyPI](https:\/\/docs.pypi.org\/project_metadata\/#verified-details)*\n\n###### Owner\n- [Meta Platforms](https:\/\/pypi.org\/org\/meta-platforms\/)\n###### Maintainers\n[![Avatar for atalman from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/4a1e9de14640155ddeb5938095a666ac957155a3\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f34363564656263366230376638366139613236613461323633346535396166383f73697a653d3530) atalman](https:\/\/pypi.org\/user\/atalman\/)     [![Avatar for malfet from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/d8937f01c4383b85866102a5e806397f40b3a557\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f35336331663335393833326462633337366337613732626365323539616536643f73697a653d3530) malfet](https:\/\/pypi.org\/user\/malfet\/)     [![Avatar for seemethere from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/0bde43daede4ad73d5e20183d999d5a5aeb0c0e0\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f62343230313862343337386564393666393330356133663334386462326435623f73697a653d3530) seemethere](https:\/\/pypi.org\/user\/seemethere\/)     [![Avatar for soumith from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/73d38504f107197034b9b146995199672719c24e\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f65386363373633626131316639323936353765323863623033373162373563383f73697a653d3530) soumith](https:\/\/pypi.org\/user\/soumith\/)\n\n### Unverified details\n*These details have **not** been verified by PyPI*\n\n###### Project links\n- [Documentation](https:\/\/pytorch.org\/docs)\n- [Forum](https:\/\/discuss.pytorch.org\/)\n- [Homepage](https:\/\/pytorch.org\/)\n- [Issue Tracker](https:\/\/github.com\/pytorch\/pytorch\/issues)\n- [Repository](https:\/\/github.com\/pytorch\/pytorch)\n\n###### Meta\n- **License:** BSD-3-Clause\n- **Author:** [PyTorch Team](mailto:packages@pytorch.org)\n- Tags  pytorch ,   machine learning\n- **Requires:** Python \\>=3.10\n- **Provides-Extra:** `optree` , `opt-einsum` , `pyyaml`\n\n###### Classifiers\n- **Development Status**\n  - [5 - Production\/Stable](https:\/\/pypi.org\/search\/?c=Development+Status+%3A%3A+5+-+Production%2FStable)\n- **Intended Audience**\n  - [Developers](https:\/\/pypi.org\/search\/?c=Intended+Audience+%3A%3A+Developers)\n  - [Education](https:\/\/pypi.org\/search\/?c=Intended+Audience+%3A%3A+Education)\n  - [Science\/Research](https:\/\/pypi.org\/search\/?c=Intended+Audience+%3A%3A+Science%2FResearch)\n- **Programming Language**\n  - [C++](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+C%2B%2B)\n  - [Python :: 3 :: Only](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3+%3A%3A+Only)\n  - [Python :: 3.10](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.10)\n  - [Python :: 3.11](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.11)\n  - [Python :: 3.12](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.12)\n  - [Python :: 3.13](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.13)\n  - [Python :: 3.14](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.14)\n- **Topic**\n  - [Scientific\/Engineering](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Scientific%2FEngineering)\n  - [Scientific\/Engineering :: Artificial Intelligence](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Artificial+Intelligence)\n  - [Scientific\/Engineering :: Mathematics](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Mathematics)\n  - [Software Development](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Software+Development)\n  - [Software Development :: Libraries](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries)\n  - [Software Development :: Libraries :: Python Modules](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries+%3A%3A+Python+Modules)\n\n[Report project as malware](https:\/\/pypi.org\/project\/torch\/submit-malware-report\/)\n\n- [Project description](https:\/\/pypi.org\/project\/torch\/#description)\n- [Project details](https:\/\/pypi.org\/project\/torch\/#data)\n- [Release history](https:\/\/pypi.org\/project\/torch\/#history)\n- [Download files](https:\/\/pypi.org\/project\/torch\/#files)\n\n## Project description\n![PyTorch Logo](https:\/\/pypi-camo.freetls.fastly.net\/ed3839e1c11e779b508097969affa63d0968692b\/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f7079746f7263682d6c6f676f2d6461726b2e706e67)\n***\nPyTorch is a Python package that provides two high-level features:\n\n- Tensor computation (like NumPy) with strong GPU acceleration\n- Deep neural networks built on a tape-based autograd system\n\nYou can reuse your favorite Python packages such as NumPy, SciPy, and Cython to extend PyTorch when needed.\n\nOur trunk health (Continuous Integration signals) can be found at [hud.pytorch.org](https:\/\/hud.pytorch.org\/ci\/pytorch\/pytorch\/main).\n\n- [More About PyTorch](https:\/\/pypi.org\/project\/torch\/#more-about-pytorch)\n  - [A GPU-Ready Tensor Library](https:\/\/pypi.org\/project\/torch\/#a-gpu-ready-tensor-library)\n  - [Dynamic Neural Networks: Tape-Based Autograd](https:\/\/pypi.org\/project\/torch\/#dynamic-neural-networks-tape-based-autograd)\n  - [Python First](https:\/\/pypi.org\/project\/torch\/#python-first)\n  - [Imperative Experiences](https:\/\/pypi.org\/project\/torch\/#imperative-experiences)\n  - [Fast and Lean](https:\/\/pypi.org\/project\/torch\/#fast-and-lean)\n  - [Extensions Without Pain](https:\/\/pypi.org\/project\/torch\/#extensions-without-pain)\n- [Installation](https:\/\/pypi.org\/project\/torch\/#installation)\n  - [Binaries](https:\/\/pypi.org\/project\/torch\/#binaries)\n    - [NVIDIA Jetson Platforms](https:\/\/pypi.org\/project\/torch\/#nvidia-jetson-platforms)\n  - [From Source](https:\/\/pypi.org\/project\/torch\/#from-source)\n    - [Prerequisites](https:\/\/pypi.org\/project\/torch\/#prerequisites)\n      - [NVIDIA CUDA Support](https:\/\/pypi.org\/project\/torch\/#nvidia-cuda-support)\n      - [AMD ROCm Support](https:\/\/pypi.org\/project\/torch\/#amd-rocm-support)\n      - [Intel GPU Support](https:\/\/pypi.org\/project\/torch\/#intel-gpu-support)\n    - [Get the PyTorch Source](https:\/\/pypi.org\/project\/torch\/#get-the-pytorch-source)\n    - [Install Dependencies](https:\/\/pypi.org\/project\/torch\/#install-dependencies)\n    - [Install PyTorch](https:\/\/pypi.org\/project\/torch\/#install-pytorch)\n      - [Adjust Build Options (Optional)](https:\/\/pypi.org\/project\/torch\/#adjust-build-options-optional)\n  - [Docker Image](https:\/\/pypi.org\/project\/torch\/#docker-image)\n    - [Using pre-built images](https:\/\/pypi.org\/project\/torch\/#using-pre-built-images)\n    - [Building the image yourself](https:\/\/pypi.org\/project\/torch\/#building-the-image-yourself)\n  - [Building the Documentation](https:\/\/pypi.org\/project\/torch\/#building-the-documentation)\n    - [Building a PDF](https:\/\/pypi.org\/project\/torch\/#building-a-pdf)\n  - [Previous Versions](https:\/\/pypi.org\/project\/torch\/#previous-versions)\n- [Getting Started](https:\/\/pypi.org\/project\/torch\/#getting-started)\n- [Resources](https:\/\/pypi.org\/project\/torch\/#resources)\n- [Communication](https:\/\/pypi.org\/project\/torch\/#communication)\n- [Releases and Contributing](https:\/\/pypi.org\/project\/torch\/#releases-and-contributing)\n- [The Team](https:\/\/pypi.org\/project\/torch\/#the-team)\n- [License](https:\/\/pypi.org\/project\/torch\/#license)\n## More About PyTorch\n[Learn the basics of PyTorch](https:\/\/pytorch.org\/tutorials\/beginner\/basics\/intro.html)\n\nAt a granular level, PyTorch is a library that consists of the following components:\n\n| Component | Description |\n|---|---|\n| [**torch**](https:\/\/pytorch.org\/docs\/stable\/torch.html) | A Tensor library like NumPy, with strong GPU support |\n| [**torch.autograd**](https:\/\/pytorch.org\/docs\/stable\/autograd.html) | A tape-based automatic differentiation library that supports all differentiable Tensor operations in torch |\n| [**torch.jit**](https:\/\/pytorch.org\/docs\/stable\/jit.html) | A compilation stack (TorchScript) to create serializable and optimizable models from PyTorch code |\n| [**torch.nn**](https:\/\/pytorch.org\/docs\/stable\/nn.html) | A neural networks library deeply integrated with autograd designed for maximum flexibility |\n| [**torch.multiprocessing**](https:\/\/pytorch.org\/docs\/stable\/multiprocessing.html) | Python multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and Hogwild training |\n| [**torch.utils**](https:\/\/pytorch.org\/docs\/stable\/data.html) | DataLoader and other utility functions for convenience |\n\nUsually, PyTorch is used either as:\n\n- A replacement for NumPy to use the power of GPUs.\n- A deep learning research platform that provides maximum flexibility and speed.\n\nElaborating Further:\n\n### A GPU-Ready Tensor Library\nIf you use NumPy, then you have used Tensors (a.k.a. ndarray).\n\n![Tensor illustration](https:\/\/pypi-camo.freetls.fastly.net\/e5f621cdea28d4de40577ac3346794f746d96b86\/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f74656e736f725f696c6c757374726174696f6e2e706e67)\n\nPyTorch provides Tensors that can live either on the CPU or the GPU and accelerates the computation by a huge amount.\n\nWe provide a wide variety of tensor routines to accelerate and fit your scientific computation needs such as slicing, indexing, mathematical operations, linear algebra, reductions. And they are fast\\!\n\n### Dynamic Neural Networks: Tape-Based Autograd\nPyTorch has a unique way of building neural networks: using and replaying a tape recorder.\n\nMost frameworks such as TensorFlow, Theano, Caffe, and CNTK have a static view of the world. One has to build a neural network and reuse the same structure again and again. Changing the way the network behaves means that one has to start from scratch.\n\nWith PyTorch, we use a technique called reverse-mode auto-differentiation, which allows you to change the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes from several research papers on this topic, as well as current and past work such as [torch-autograd](https:\/\/github.com\/twitter\/torch-autograd), [autograd](https:\/\/github.com\/HIPS\/autograd), [Chainer](https:\/\/chainer.org\/), etc.\n\nWhile this technique is not unique to PyTorch, it's one of the fastest implementations of it to date. You get the best of speed and flexibility for your crazy research.\n\n![Dynamic graph](https:\/\/pypi-camo.freetls.fastly.net\/360dc3e00cca046f73204c87f117feefe993580f\/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f64796e616d69635f67726170682e676966)\n\n### Python First\nPyTorch is not a Python binding into a monolithic C++ framework. It is built to be deeply integrated into Python. You can use it naturally like you would use [NumPy](https:\/\/www.numpy.org\/) \/ [SciPy](https:\/\/www.scipy.org\/) \/ [scikit-learn](https:\/\/scikit-learn.org\/) etc. You can write your new neural network layers in Python itself, using your favorite libraries and use packages such as [Cython](https:\/\/cython.org\/) and [Numba](http:\/\/numba.pydata.org\/). Our goal is to not reinvent the wheel where appropriate.\n\n### Imperative Experiences\nPyTorch is designed to be intuitive, linear in thought, and easy to use. When you execute a line of code, it gets executed. There isn't an asynchronous view of the world. When you drop into a debugger or receive error messages and stack traces, understanding them is straightforward. The stack trace points to exactly where your code was defined. We hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines.\n\n### Fast and Lean\nPyTorch has minimal framework overhead. We integrate acceleration libraries such as [Intel MKL](https:\/\/software.intel.com\/mkl) and NVIDIA ([cuDNN](https:\/\/developer.nvidia.com\/cudnn), [NCCL](https:\/\/developer.nvidia.com\/nccl)) to maximize speed. At the core, its CPU and GPU Tensor and neural network backends are mature and have been tested for years.\n\nHence, PyTorch is quite fast — whether you run small or large neural networks.\n\nThe memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives. We've written custom memory allocators for the GPU to make sure that your deep learning models are maximally memory efficient. This enables you to train bigger deep learning models than before.\n\n### Extensions Without Pain\nWriting new neural network modules, or interfacing with PyTorch's Tensor API was designed to be straightforward and with minimal abstractions.\n\nYou can write new neural network layers in Python using the torch API [or your favorite NumPy-based libraries such as SciPy](https:\/\/pytorch.org\/tutorials\/advanced\/numpy_extensions_tutorial.html).\n\nIf you want to write your layers in C\/C++, we provide a convenient extension API that is efficient and with minimal boilerplate. No wrapper code needs to be written. You can see [a tutorial here](https:\/\/pytorch.org\/tutorials\/advanced\/cpp_extension.html) and [an example here](https:\/\/github.com\/pytorch\/extension-cpp).\n\n## Installation\n### Binaries\nCommands to install binaries via Conda or pip wheels are on our website: <https:\/\/pytorch.org\/get-started\/locally\/>\n\n#### NVIDIA Jetson Platforms\nPython wheels for NVIDIA's Jetson Nano, Jetson TX1\/TX2, Jetson Xavier NX\/AGX, and Jetson AGX Orin are provided [here](https:\/\/forums.developer.nvidia.com\/t\/pytorch-for-jetson-version-1-10-now-available\/72048) and the L4T container is published [here](https:\/\/catalog.ngc.nvidia.com\/orgs\/nvidia\/containers\/l4t-pytorch)\n\nThey require JetPack 4.2 and above, and [@dusty-nv](https:\/\/github.com\/dusty-nv) and [@ptrblck](https:\/\/github.com\/ptrblck) are maintaining them.\n\n### From Source\n#### Prerequisites\nIf you are installing from source, you will need:\n\n- Python 3.10 or later\n- A compiler that fully supports C++17, such as clang or gcc (gcc 9.4.0 or newer is required, on Linux)\n- Visual Studio or Visual Studio Build Tool (Windows only)\n\n\\* PyTorch CI uses Visual C++ BuildTools, which come with Visual Studio Enterprise, Professional, or Community Editions. You can also install the build tools from <https:\/\/visualstudio.microsoft.com\/visual-cpp-build-tools\/>. The build tools *do not* come with Visual Studio Code by default.\n\nAn example of environment setup is shown below:\n\n- Linux:\n```\n$ source <CONDA_INSTALL_DIR>\/bin\/activate\n$ conda create -y -n <CONDA_NAME>\n$ conda activate <CONDA_NAME>\n```\n- Windows:\n```\n$ source <CONDA_INSTALL_DIR>\\Scripts\\activate.bat\n$ conda create -y -n <CONDA_NAME>\n$ conda activate <CONDA_NAME>\n$ call \"C:\\Program Files\\Microsoft Visual Studio\\<VERSION>\\Community\\VC\\Auxiliary\\Build\\vcvarsall.bat\" x64\n```\nA conda environment is not required. You can also do a PyTorch build in a standard virtual environment, e.g., created with tools like `uv`, provided your system has installed all the necessary dependencies unavailable as pip packages (e.g., CUDA, MKL.)\n\n##### NVIDIA CUDA Support\nIf you want to compile with CUDA support, [select a supported version of CUDA from our support matrix](https:\/\/pytorch.org\/get-started\/locally\/), then install the following:\n\n- [NVIDIA CUDA](https:\/\/developer.nvidia.com\/cuda-downloads)\n- [NVIDIA cuDNN](https:\/\/developer.nvidia.com\/cudnn) v8.5 or above\n- [Compiler](https:\/\/gist.github.com\/ax3l\/9489132) compatible with CUDA\n\nNote: You could refer to the [cuDNN Support Matrix](https:\/\/docs.nvidia.com\/deeplearning\/cudnn\/backend\/latest\/reference\/support-matrix.html) for cuDNN versions with the various supported CUDA, CUDA driver, and NVIDIA hardware.\n\nIf you want to disable CUDA support, export the environment variable `USE_CUDA=0`. Other potentially useful environment variables may be found in `setup.py`. If CUDA is installed in a non-standard location, set PATH so that the nvcc you want to use can be found (e.g., `export PATH=\/usr\/local\/cuda-12.8\/bin:$PATH`).\n\nIf you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are [available here](https:\/\/devtalk.nvidia.com\/default\/topic\/1049071\/jetson-nano\/pytorch-for-jetson-nano\/)\n\n##### AMD ROCm Support\nIf you want to compile with ROCm support, install\n\n- [AMD ROCm](https:\/\/rocm.docs.amd.com\/en\/latest\/deploy\/linux\/quick_start.html) 4.0 and above installation\n- ROCm is currently supported only for Linux systems.\n\nBy default the build system expects ROCm to be installed in `\/opt\/rocm`. If ROCm is installed in a different directory, the `ROCM_PATH` environment variable must be set to the ROCm installation directory. The build system automatically detects the AMD GPU architecture. Optionally, the AMD GPU architecture can be explicitly set with the `PYTORCH_ROCM_ARCH` environment variable [AMD GPU architecture](https:\/\/rocm.docs.amd.com\/projects\/install-on-linux\/en\/latest\/reference\/system-requirements.html#supported-gpus)\n\nIf you want to disable ROCm support, export the environment variable `USE_ROCM=0`. Other potentially useful environment variables may be found in `setup.py`.\n\n##### Intel GPU Support\nIf you want to compile with Intel GPU support, follow these\n\n- [PyTorch Prerequisites for Intel GPUs](https:\/\/www.intel.com\/content\/www\/us\/en\/developer\/articles\/tool\/pytorch-prerequisites-for-intel-gpu.html) instructions.\n- Intel GPU is supported for Linux and Windows.\n\nIf you want to disable Intel GPU support, export the environment variable `USE_XPU=0`. Other potentially useful environment variables may be found in `setup.py`.\n\n#### Get the PyTorch Source\n```\ngit clone https:\/\/github.com\/pytorch\/pytorch\ncd pytorch\n# if you are updating an existing checkout\ngit submodule sync\ngit submodule update --init --recursive\n```\n#### Install Dependencies\n**Common**\n```\n# Run this command from the PyTorch directory after cloning the source code using the “Get the PyTorch Source“ section above\npip install --group dev\n```\n**On Linux**\n```\npip install mkl-static mkl-include\n# CUDA only: Add LAPACK support for the GPU if needed\n# magma installation: run with active conda environment. specify CUDA version to install\n.ci\/docker\/common\/install_magma_conda.sh 12.4\n\n# (optional) If using torch.compile with inductor\/triton, install the matching version of triton\n# Run from the pytorch directory after cloning\n# For Intel GPU support, please explicitly `export USE_XPU=1` before running command.\nmake triton\n```\n**On MacOS**\n```\n# Add this package on intel x86 processor machines only\npip install mkl-static mkl-include\n# Add these packages if torch.distributed is needed\nconda install pkg-config libuv\n```\n**On Windows**\n```\npip install mkl-static mkl-include\n# Add these packages if torch.distributed is needed.\n# Distributed package support on Windows is a prototype feature and is subject to changes.\nconda install -c conda-forge libuv=1.51\n```\n#### Install PyTorch\n**On Linux**\n\nIf you're compiling for AMD ROCm then first run this command:\n```\n# Only run this if you're compiling for ROCm\npython tools\/amd_build\/build_amd.py\n```\nInstall PyTorch\n```\n# the CMake prefix for conda environment\nexport CMAKE_PREFIX_PATH=\"${CONDA_PREFIX:-'$(dirname $(which conda))\/..\/'}:${CMAKE_PREFIX_PATH}\"\npython -m pip install --no-build-isolation -v -e .\n\n# the CMake prefix for non-conda environment, e.g. Python venv\n# call following after activating the venv\nexport CMAKE_PREFIX_PATH=\"${VIRTUAL_ENV}:${CMAKE_PREFIX_PATH}\"\n```\n**On macOS**\n```\npython -m pip install --no-build-isolation -v -e .\n```\n**On Windows**\n\nIf you want to build legacy python code, please refer to [Building on legacy code and CUDA](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/CONTRIBUTING.md#building-on-legacy-code-and-cuda)\n\n**CPU-only builds**\n\nIn this mode PyTorch computations will run on your CPU, not your GPU.\n```\npython -m pip install --no-build-isolation -v -e .\n```\nNote on OpenMP: The desired OpenMP implementation is Intel OpenMP (iomp). In order to link against iomp, you'll need to manually download the library and set up the building environment by tweaking `CMAKE_INCLUDE_PATH` and `LIB`. The instruction [here](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/docs\/source\/notes\/windows.rst#building-from-source) is an example for setting up both MKL and Intel OpenMP. Without these configurations for CMake, Microsoft Visual C OpenMP runtime (vcomp) will be used.\n\n**CUDA based build**\n\nIn this mode PyTorch computations will leverage your GPU via CUDA for faster number crunching\n\n[NVTX](https:\/\/docs.nvidia.com\/gameworks\/content\/gameworkslibrary\/nvtx\/nvidia_tools_extension_library_nvtx.htm) is needed to build PyTorch with CUDA. NVTX is a part of CUDA distributive, where it is called \"Nsight Compute\". To install it onto an already installed CUDA run CUDA installation once again and check the corresponding checkbox. Make sure that CUDA with Nsight Compute is installed after Visual Studio.\n\nCurrently, VS 2017 \/ 2019, and Ninja are supported as the generator of CMake. If `ninja.exe` is detected in `PATH`, then Ninja will be used as the default generator, otherwise, it will use VS 2017 \/ 2019.   \n If Ninja is selected as the generator, the latest MSVC will get selected as the underlying toolchain.\n\nAdditional libraries such as [Magma](https:\/\/developer.nvidia.com\/magma), [oneDNN, a.k.a. MKLDNN or DNNL](https:\/\/github.com\/oneapi-src\/oneDNN), and [Sccache](https:\/\/github.com\/mozilla\/sccache) are often needed. Please refer to the [installation-helper](https:\/\/github.com\/pytorch\/pytorch\/tree\/main\/.ci\/pytorch\/win-test-helpers\/installation-helpers) to install them.\n\nYou can refer to the [build\\_pytorch.bat](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/.ci\/pytorch\/win-test-helpers\/build_pytorch.bat) script for some other environment variables configurations\n```\ncmd\n\n:: Set the environment variables after you have downloaded and unzipped the mkl package,\n:: else CMake would throw an error as `Could NOT find OpenMP`.\nset CMAKE_INCLUDE_PATH={Your directory}\\mkl\\include\nset LIB={Your directory}\\mkl\\lib;%LIB%\n\n:: Read the content in the previous section carefully before you proceed.\n:: [Optional] If you want to override the underlying toolset used by Ninja and Visual Studio with CUDA, please run the following script block.\n:: \"Visual Studio 2019 Developer Command Prompt\" will be run automatically.\n:: Make sure you have CMake >= 3.12 before you do this when you use the Visual Studio generator.\nset CMAKE_GENERATOR_TOOLSET_VERSION=14.27\nset DISTUTILS_USE_SDK=1\nfor \/f \"usebackq tokens=*\" %i in (`\"%ProgramFiles(x86)%\\Microsoft Visual Studio\\Installer\\vswhere.exe\" -version [15^,17^) -products * -latest -property installationPath`) do call \"%i\\VC\\Auxiliary\\Build\\vcvarsall.bat\" x64 -vcvars_ver=%CMAKE_GENERATOR_TOOLSET_VERSION%\n\n:: [Optional] If you want to override the CUDA host compiler\nset CUDAHOSTCXX=C:\\Program Files (x86)\\Microsoft Visual Studio\\2019\\Community\\VC\\Tools\\MSVC\\14.27.29110\\bin\\HostX64\\x64\\cl.exe\n\npython -m pip install --no-build-isolation -v -e .\n```\n**Intel GPU builds**\n\nIn this mode PyTorch with Intel GPU support will be built.\n\nPlease make sure [the common prerequisites](https:\/\/pypi.org\/project\/torch\/#prerequisites) as well as [the prerequisites for Intel GPU](https:\/\/pypi.org\/project\/torch\/#intel-gpu-support) are properly installed and the environment variables are configured prior to starting the build. For build tool support, `Visual Studio 2022` is required.\n\nThen PyTorch can be built with the command:\n```\n:: CMD Commands:\n:: Set the CMAKE_PREFIX_PATH to help find corresponding packages\n:: %CONDA_PREFIX% only works after `conda activate custom_env`\n\nif defined CMAKE_PREFIX_PATH (\n    set \"CMAKE_PREFIX_PATH=%CONDA_PREFIX%\\Library;%CMAKE_PREFIX_PATH%\"\n) else (\n    set \"CMAKE_PREFIX_PATH=%CONDA_PREFIX%\\Library\"\n)\n\npython -m pip install --no-build-isolation -v -e .\n```\n##### Adjust Build Options (Optional)\nYou can adjust the configuration of cmake variables optionally (without building first), by doing the following. For example, adjusting the pre-detected directories for CuDNN or BLAS can be done with such a step.\n\nOn Linux\n```\nexport CMAKE_PREFIX_PATH=\"${CONDA_PREFIX:-'$(dirname $(which conda))\/..\/'}:${CMAKE_PREFIX_PATH}\"\nCMAKE_ONLY=1 python setup.py build\nccmake build  # or cmake-gui build\n```\nOn macOS\n```\nexport CMAKE_PREFIX_PATH=\"${CONDA_PREFIX:-'$(dirname $(which conda))\/..\/'}:${CMAKE_PREFIX_PATH}\"\nMACOSX_DEPLOYMENT_TARGET=11.0 CMAKE_ONLY=1 python setup.py build\nccmake build  # or cmake-gui build\n```\n### Docker Image\n#### Using pre-built images\nYou can also pull a pre-built docker image from Docker Hub and run with docker v19.03+\n```\ndocker run --gpus all --rm -ti --ipc=host pytorch\/pytorch:latest\n```\nPlease note that PyTorch uses shared memory to share data between processes, so if torch multiprocessing is used (e.g. for multithreaded data loaders) the default shared memory segment size that container runs with is not enough, and you should increase shared memory size either with `--ipc=host` or `--shm-size` command line options to `nvidia-docker run`.\n\n#### Building the image yourself\n**NOTE:** Must be built with a docker version \\> 18.06\n\nThe `Dockerfile` is supplied to build images with CUDA 11.1 support and cuDNN v8. You can pass `PYTHON_VERSION=x.y` make variable to specify which Python version is to be used by Miniconda, or leave it unset to use the default.\n```\nmake -f docker.Makefile\n# images are tagged as docker.io\/${your_docker_username}\/pytorch\n```\nYou can also pass the `CMAKE_VARS=\"...\"` environment variable to specify additional CMake variables to be passed to CMake during the build. See [setup.py](https:\/\/pypi.org\/project\/torch\/setup.py) for the list of available variables.\n```\nmake -f docker.Makefile\n```\n### Building the Documentation\nTo build documentation in various formats, you will need [Sphinx](http:\/\/www.sphinx-doc.org\/) and the pytorch\\_sphinx\\_theme2.\n\nBefore you build the documentation locally, ensure `torch` is installed in your environment. For small fixes, you can install the nightly version as described in [Getting Started](https:\/\/pytorch.org\/get-started\/locally\/).\n\nFor more complex fixes, such as adding a new module and docstrings for the new module, you might need to install torch [from source](https:\/\/pypi.org\/project\/torch\/#from-source). See [Docstring Guidelines](https:\/\/github.com\/pytorch\/pytorch\/wiki\/Docstring-Guidelines) for docstring conventions.\n```\ncd docs\/\npip install -r requirements.txt\nmake html\nmake serve\n```\nRun `make` to get a list of all available output formats.\n\nIf you get a katex error run `npm install katex`. If it persists, try `npm install -g katex`\n\n> \\[!NOTE\\] If you installed `nodejs` with a different package manager (e.g., `conda`) then `npm` will probably install a version of `katex` that is not compatible with your version of `nodejs` and doc builds will fail. A combination of versions that is known to work is `node@6.13.1` and `katex@0.13.18`. To install the latter with `npm` you can run `npm install -g katex@0.13.18`\n\n> \\[!NOTE\\] If you see a numpy incompatibility error, run:\n> ```\n> pip install 'numpy<2'\n> ```\n\nWhen you make changes to the dependencies run by CI, edit the `.ci\/docker\/requirements-docs.txt` file.\n\n#### Building a PDF\nTo compile a PDF of all PyTorch documentation, ensure you have `texlive` and LaTeX installed. On macOS, you can install them using:\n```\nbrew install --cask mactex\n```\nTo create the PDF:\n\n1. Run:\n   ```\n   make latexpdf\n   ```\n   This will generate the necessary files in the `build\/latex` directory.\n2. Navigate to this directory and execute:\n   ```\n   make LATEXOPTS=\"-interaction=nonstopmode\"\n   ```\n   This will produce a `pytorch.pdf` with the desired content. Run this command one more time so that it generates the correct table of contents and index.\n\n> \\[!NOTE\\] To view the Table of Contents, switch to the **Table of Contents** view in your PDF viewer.\n### Previous Versions\nInstallation instructions and binaries for previous PyTorch versions may be found on [our website](https:\/\/pytorch.org\/get-started\/previous-versions).\n\n## Getting Started\nThree pointers to get you started:\n\n- [Tutorials: get you started with understanding and using PyTorch](https:\/\/pytorch.org\/tutorials\/)\n- [Examples: easy to understand PyTorch code across all domains](https:\/\/github.com\/pytorch\/examples)\n- [The API Reference](https:\/\/pytorch.org\/docs\/)\n- [Glossary](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/GLOSSARY.md)\n## Resources\n- [PyTorch.org](https:\/\/pytorch.org\/)\n- [PyTorch Tutorials](https:\/\/pytorch.org\/tutorials\/)\n- [PyTorch Examples](https:\/\/github.com\/pytorch\/examples)\n- [PyTorch Models](https:\/\/pytorch.org\/hub\/)\n- [Intro to Deep Learning with PyTorch from Udacity](https:\/\/www.udacity.com\/course\/deep-learning-pytorch--ud188)\n- [Intro to Machine Learning with PyTorch from Udacity](https:\/\/www.udacity.com\/course\/intro-to-machine-learning-nanodegree--nd229)\n- [Deep Neural Networks with PyTorch from Coursera](https:\/\/www.coursera.org\/learn\/deep-neural-networks-with-pytorch)\n- [PyTorch Twitter](https:\/\/twitter.com\/PyTorch)\n- [PyTorch Blog](https:\/\/pytorch.org\/blog\/)\n- [PyTorch YouTube](https:\/\/www.youtube.com\/channel\/UCWXI5YeOsh03QvJ59PMaXFw)\n## Communication\n- Forums: Discuss implementations, research, etc. <https:\/\/discuss.pytorch.org>\n- GitHub Issues: Bug reports, feature requests, install issues, RFCs, thoughts, etc.\n- Slack: The [PyTorch Slack](https:\/\/pytorch.slack.com\/) hosts a primary audience of moderate to experienced PyTorch users and developers for general chat, online discussions, collaboration, etc. If you are a beginner looking for help, the primary medium is [PyTorch Forums](https:\/\/discuss.pytorch.org\/). If you need a slack invite, please fill this form: <https:\/\/goo.gl\/forms\/PP1AGvNHpSaJP8to1>\n- Newsletter: No-noise, a one-way email newsletter with important announcements about PyTorch. You can sign-up here: <https:\/\/eepurl.com\/cbG0rv>\n- Facebook Page: Important announcements about PyTorch. <https:\/\/www.facebook.com\/pytorch>\n- For brand guidelines, please visit our website at [pytorch.org](https:\/\/pytorch.org\/)\n## Releases and Contributing\nTypically, PyTorch has three minor releases a year. Please let us know if you encounter a bug by [filing an issue](https:\/\/github.com\/pytorch\/pytorch\/issues).\n\nWe appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion.\n\nIf you plan to contribute new features, utility functions, or extensions to the core, please first open an issue and discuss the feature with us. Sending a PR without discussion might end up resulting in a rejected PR because we might be taking the core in a different direction than you might be aware of.\n\nTo learn more about making a contribution to PyTorch, please see our [Contribution page](https:\/\/pypi.org\/project\/torch\/CONTRIBUTING.md). For more information about PyTorch releases, see [Release page](https:\/\/pypi.org\/project\/torch\/RELEASE.md).\n\n## The Team\nPyTorch is a community-driven project with several skillful engineers and researchers contributing to it.\n\nPyTorch is currently maintained by [Soumith Chintala](http:\/\/soumith.ch\/), [Gregory Chanan](https:\/\/github.com\/gchanan), [Dmytro Dzhulgakov](https:\/\/github.com\/dzhulgakov), [Edward Yang](https:\/\/github.com\/ezyang), [Alban Desmaison](https:\/\/github.com\/albanD), [Piotr Bialecki](https:\/\/github.com\/ptrblck) and [Nikita Shulga](https:\/\/github.com\/malfet) with major contributions coming from hundreds of talented individuals in various forms and means. A non-exhaustive but growing list needs to mention: [Trevor Killeen](https:\/\/github.com\/killeent), [Sasank Chilamkurthy](https:\/\/github.com\/chsasank), [Sergey Zagoruyko](https:\/\/github.com\/szagoruyko), [Adam Lerer](https:\/\/github.com\/adamlerer), [Francisco Massa](https:\/\/github.com\/fmassa), [Alykhan Tejani](https:\/\/github.com\/alykhantejani), [Luca Antiga](https:\/\/github.com\/lantiga), [Alban Desmaison](https:\/\/github.com\/albanD), [Andreas Koepf](https:\/\/github.com\/andreaskoepf), [James Bradbury](https:\/\/github.com\/jekbradbury), [Zeming Lin](https:\/\/github.com\/ebetica), [Yuandong Tian](https:\/\/github.com\/yuandong-tian), [Guillaume Lample](https:\/\/github.com\/glample), [Marat Dukhan](https:\/\/github.com\/Maratyszcza), [Natalia Gimelshein](https:\/\/github.com\/ngimel), [Christian Sarofeen](https:\/\/github.com\/csarofeen), [Martin Raison](https:\/\/github.com\/martinraison), [Edward Yang](https:\/\/github.com\/ezyang), [Zachary Devito](https:\/\/github.com\/zdevito).\n\nNote: This project is unrelated to [hughperkins\/pytorch](https:\/\/github.com\/hughperkins\/pytorch) with the same name. Hugh is a valuable contributor to the Torch community and has helped with many things Torch and PyTorch.\n\n## License\nPyTorch has a BSD-style license, as found in the [LICENSE](https:\/\/pypi.org\/project\/torch\/LICENSE) file.\n\n## Project details\n### Verified details\n*These details have been [verified by PyPI](https:\/\/docs.pypi.org\/project_metadata\/#verified-details)*\n\n###### Owner\n- [Meta Platforms](https:\/\/pypi.org\/org\/meta-platforms\/)\n###### Maintainers\n[![Avatar for atalman from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/4a1e9de14640155ddeb5938095a666ac957155a3\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f34363564656263366230376638366139613236613461323633346535396166383f73697a653d3530) atalman](https:\/\/pypi.org\/user\/atalman\/)     [![Avatar for malfet from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/d8937f01c4383b85866102a5e806397f40b3a557\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f35336331663335393833326462633337366337613732626365323539616536643f73697a653d3530) malfet](https:\/\/pypi.org\/user\/malfet\/)     [![Avatar for seemethere from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/0bde43daede4ad73d5e20183d999d5a5aeb0c0e0\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f62343230313862343337386564393666393330356133663334386462326435623f73697a653d3530) seemethere](https:\/\/pypi.org\/user\/seemethere\/)     [![Avatar for soumith from gravatar.com](https:\/\/pypi-camo.freetls.fastly.net\/73d38504f107197034b9b146995199672719c24e\/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f65386363373633626131316639323936353765323863623033373162373563383f73697a653d3530) soumith](https:\/\/pypi.org\/user\/soumith\/)\n\n### Unverified details\n*These details have **not** been verified by PyPI*\n\n###### Project links\n- [Documentation](https:\/\/pytorch.org\/docs)\n- [Forum](https:\/\/discuss.pytorch.org\/)\n- [Homepage](https:\/\/pytorch.org\/)\n- [Issue Tracker](https:\/\/github.com\/pytorch\/pytorch\/issues)\n- [Repository](https:\/\/github.com\/pytorch\/pytorch)\n\n###### Meta\n- **License:** BSD-3-Clause\n- **Author:** [PyTorch Team](mailto:packages@pytorch.org)\n- Tags  pytorch ,   machine learning\n- **Requires:** Python \\>=3.10\n- **Provides-Extra:** `optree` , `opt-einsum` , `pyyaml`\n\n###### Classifiers\n- **Development Status**\n  - [5 - Production\/Stable](https:\/\/pypi.org\/search\/?c=Development+Status+%3A%3A+5+-+Production%2FStable)\n- **Intended Audience**\n  - [Developers](https:\/\/pypi.org\/search\/?c=Intended+Audience+%3A%3A+Developers)\n  - [Education](https:\/\/pypi.org\/search\/?c=Intended+Audience+%3A%3A+Education)\n  - [Science\/Research](https:\/\/pypi.org\/search\/?c=Intended+Audience+%3A%3A+Science%2FResearch)\n- **Programming Language**\n  - [C++](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+C%2B%2B)\n  - [Python :: 3 :: Only](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3+%3A%3A+Only)\n  - [Python :: 3.10](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.10)\n  - [Python :: 3.11](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.11)\n  - [Python :: 3.12](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.12)\n  - [Python :: 3.13](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.13)\n  - [Python :: 3.14](https:\/\/pypi.org\/search\/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.14)\n- **Topic**\n  - [Scientific\/Engineering](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Scientific%2FEngineering)\n  - [Scientific\/Engineering :: Artificial Intelligence](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Artificial+Intelligence)\n  - [Scientific\/Engineering :: Mathematics](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Mathematics)\n  - [Software Development](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Software+Development)\n  - [Software Development :: Libraries](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries)\n  - [Software Development :: Libraries :: Python Modules](https:\/\/pypi.org\/search\/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries+%3A%3A+Python+Modules)\n\n## Release history  [Release notifications](https:\/\/pypi.org\/help\/#project-release-notifications) \\| [RSS feed](https:\/\/pypi.org\/rss\/project\/torch\/releases.xml)\nThis version\n\n![](https:\/\/pypi.org\/static\/images\/blue-cube.572a5bfb.svg)\n\n[2\\.11.0 Mar 23, 2026](https:\/\/pypi.org\/project\/torch\/2.11.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.10.0 Jan 21, 2026](https:\/\/pypi.org\/project\/torch\/2.10.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.9.1 Nov 12, 2025](https:\/\/pypi.org\/project\/torch\/2.9.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.9.0 Oct 15, 2025](https:\/\/pypi.org\/project\/torch\/2.9.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.8.0 Aug 6, 2025](https:\/\/pypi.org\/project\/torch\/2.8.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.7.1 Jun 4, 2025](https:\/\/pypi.org\/project\/torch\/2.7.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.7.0 Apr 23, 2025](https:\/\/pypi.org\/project\/torch\/2.7.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.6.0 Jan 29, 2025](https:\/\/pypi.org\/project\/torch\/2.6.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.5.1 Oct 29, 2024](https:\/\/pypi.org\/project\/torch\/2.5.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.5.0 Oct 17, 2024](https:\/\/pypi.org\/project\/torch\/2.5.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.4.1 Sep 4, 2024](https:\/\/pypi.org\/project\/torch\/2.4.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.4.0 Jul 24, 2024](https:\/\/pypi.org\/project\/torch\/2.4.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.3.1 Jun 5, 2024](https:\/\/pypi.org\/project\/torch\/2.3.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.3.0 Apr 24, 2024](https:\/\/pypi.org\/project\/torch\/2.3.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.2.2 Mar 27, 2024](https:\/\/pypi.org\/project\/torch\/2.2.2\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.2.1 Feb 22, 2024](https:\/\/pypi.org\/project\/torch\/2.2.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.2.0 Jan 30, 2024](https:\/\/pypi.org\/project\/torch\/2.2.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.1.2 Dec 14, 2023](https:\/\/pypi.org\/project\/torch\/2.1.2\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.1.1 Nov 15, 2023](https:\/\/pypi.org\/project\/torch\/2.1.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.1.0 Oct 4, 2023](https:\/\/pypi.org\/project\/torch\/2.1.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.0.1 May 8, 2023](https:\/\/pypi.org\/project\/torch\/2.0.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[2\\.0.0 Mar 15, 2023](https:\/\/pypi.org\/project\/torch\/2.0.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.13.1 Dec 15, 2022](https:\/\/pypi.org\/project\/torch\/1.13.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.13.0 Oct 28, 2022](https:\/\/pypi.org\/project\/torch\/1.13.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.12.1 Aug 5, 2022](https:\/\/pypi.org\/project\/torch\/1.12.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.12.0 Jun 28, 2022](https:\/\/pypi.org\/project\/torch\/1.12.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.11.0 Mar 10, 2022](https:\/\/pypi.org\/project\/torch\/1.11.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.10.2 Jan 27, 2022](https:\/\/pypi.org\/project\/torch\/1.10.2\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.10.1 Dec 15, 2021](https:\/\/pypi.org\/project\/torch\/1.10.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.10.0 Oct 21, 2021](https:\/\/pypi.org\/project\/torch\/1.10.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.9.1 Sep 20, 2021](https:\/\/pypi.org\/project\/torch\/1.9.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.9.0 Jun 15, 2021](https:\/\/pypi.org\/project\/torch\/1.9.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.8.1 Mar 25, 2021](https:\/\/pypi.org\/project\/torch\/1.8.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.8.0 Mar 4, 2021](https:\/\/pypi.org\/project\/torch\/1.8.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.7.1 Dec 10, 2020](https:\/\/pypi.org\/project\/torch\/1.7.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.7.0 Oct 27, 2020](https:\/\/pypi.org\/project\/torch\/1.7.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.6.0 Jul 28, 2020](https:\/\/pypi.org\/project\/torch\/1.6.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.5.1 Jun 18, 2020](https:\/\/pypi.org\/project\/torch\/1.5.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.5.0 Apr 21, 2020](https:\/\/pypi.org\/project\/torch\/1.5.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.4.0 Jan 15, 2020](https:\/\/pypi.org\/project\/torch\/1.4.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.3.1 Nov 7, 2019](https:\/\/pypi.org\/project\/torch\/1.3.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.3.0 Oct 10, 2019](https:\/\/pypi.org\/project\/torch\/1.3.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.2.0 Aug 8, 2019](https:\/\/pypi.org\/project\/torch\/1.2.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.1.0 Apr 30, 2019](https:\/\/pypi.org\/project\/torch\/1.1.0\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.0.1 Feb 5, 2019](https:\/\/pypi.org\/project\/torch\/1.0.1\/)\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n[1\\.0.0 Dec 7, 2018](https:\/\/pypi.org\/project\/torch\/1.0.0\/)\n\n## Download files\nDownload the file for your platform. If you're not sure which to choose, learn more about [installing packages](https:\/\/packaging.python.org\/tutorials\/installing-packages\/ \"External link\").\n\n### Source Distributions\nNo source distribution files available for this release.See tutorial on [generating distribution archives](https:\/\/packaging.python.org\/tutorials\/packaging-projects\/#generating-distribution-archives \"External link\").\n\n### Built Distributions\nFilter files by name, interpreter, ABI, and platform.\n\nIf you're not sure about the file name format, learn more about [wheel file names](https:\/\/packaging.python.org\/en\/latest\/specifications\/binary-distribution-format\/ \"External link\").\n\nThe dropdown lists show the available interpreters, ABIs, and platforms.\n\nEnable javascript to be able to filter the list of wheel files.\n\nCopy a direct link to the current filters <https:\/\/pypi.org\/project\/torch\/#files> Copy\n\nShowing 28 of 28 files.\n\nFile name\n\nInterpreter\n\nABI\n\nPlatform\n\n[torch-2.11.0-cp314-cp314t-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/cf\/bf\/c8d12a2c86dbfd7f40fb2f56fbf5a505ccf2d9ce131eb559dfc7c51e1a04\/torch-2.11.0-cp314-cp314t-win_amd64.whl) (114.8 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314t-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14t``Windows x86-64`\n\n[torch-2.11.0-cp314-cp314t-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/b1\/e7\/0b6665f533aa9e337662dc190425abc0af1fe3234088f4454c52393ded61\/torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14t``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp314-cp314t-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/fd\/66\/54a56a4a6ceaffb567231994a9745821d3af922a854ed33b0b3a278e0a99\/torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14t``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp314-cp314t-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/bf\/46\/4419098ed6d801750f26567b478fc185c3432e11e2cad712bc6b4c2ab0d0\/torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl) (81.0 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14t``macOS 11.0+ ARM64`\n\n[torch-2.11.0-cp314-cp314-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/78\/88\/d4a4cda8362f8a30d1ed428564878c3cafb0d87971fbd3947d4c84552095\/torch-2.11.0-cp314-cp314-win_amd64.whl) (114.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14``Windows x86-64`\n\n[torch-2.11.0-cp314-cp314-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/47\/e8\/b98ca2d39b2e0e4730c0ee52537e488e7008025bc77ca89552ff91021f7c\/torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl) (530.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp314-cp314-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/c7\/86\/7cd7c66cb9cec6be330fff36db5bd0eef386d80c031b581ec81be1d4b26c\/torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp314-cp314-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/26\/0d\/8603382f61abd0db35841148ddc1ffd607bf3100b11c6e1dab6d2fc44e72\/torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl) (80.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.14``macOS 11.0+ ARM64`\n\n[torch-2.11.0-cp313-cp313t-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/48\/6b\/30d1459fa7e4b67e9e3fe1685ca1d8bb4ce7c62ef436c3a615963c6c866c\/torch-2.11.0-cp313-cp313t-win_amd64.whl) (114.8 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313t-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13t``Windows x86-64`\n\n[torch-2.11.0-cp313-cp313t-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/07\/f4\/1b666b6d61d3394cca306ea543ed03a64aad0a201b6cd159f1d41010aeb1\/torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13t``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp313-cp313t-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/6d\/6c\/56bfb37073e7136e6dd86bfc6af7339946dd684e0ecf2155ac0eee687ae1\/torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13t``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp313-cp313t-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/db\/38\/8ac78069621b8c2b4979c2f96dc8409ef5e9c4189f6aac629189a78677ca\/torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl) (81.0 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13t``macOS 11.0+ ARM64`\n\n[torch-2.11.0-cp313-cp313-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/66\/82\/3e3fcdd388fbe54e29fd3f991f36846ff4ac90b0d0181e9c8f7236565f82\/torch-2.11.0-cp313-cp313-win_amd64.whl) (114.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13``Windows x86-64`\n\n[torch-2.11.0-cp313-cp313-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/3d\/e1\/b73f7c575a4b8f87a5928f50a1e35416b5e27295d8be9397d5293e7e8d4c\/torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl) (530.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp313-cp313-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/32\/d1\/8ed2173589cbfe744ed54e5a73efc107c0085ba5777ee93a5f4c1ab90553\/torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp313-cp313-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/87\/89\/5ea6722763acee56b045435fb84258db7375c48165ec8be7880ab2b281c5\/torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl) (80.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.13``macOS 11.0+ ARM64`\n\n[torch-2.11.0-cp312-cp312-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/1c\/ff\/6756f1c7ee302f6d202120e0f4f05b432b839908f9071157302cedfc5232\/torch-2.11.0-cp312-cp312-win_amd64.whl) (114.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp312-cp312-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.12``Windows x86-64`\n\n[torch-2.11.0-cp312-cp312-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/1a\/c9\/82638ef24d7877510f83baf821f5619a61b45568ce21c0a87a91576510aa\/torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl) (530.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.12``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp312-cp312-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/13\/16\/42e5915ebe4868caa6bac83a8ed59db57f12e9a61b7d749d584776ed53d5\/torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.12``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp312-cp312-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/6f\/8b\/69e3008d78e5cee2b30183340cc425081b78afc5eff3d080daab0adda9aa\/torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl) (80.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.12``macOS 11.0+ ARM64`\n\n[torch-2.11.0-cp311-cp311-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/d1\/bd\/9912d30b68845256aabbb4a40aeefeef3c3b20db5211ccda653544ada4b6\/torch-2.11.0-cp311-cp311-win_amd64.whl) (114.5 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp311-cp311-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.11``Windows x86-64`\n\n[torch-2.11.0-cp311-cp311-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/8c\/8b\/d7be22fbec9ffee6cff31a39f8750d4b3a65d349a286cf4aec74c2375662\/torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.11``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp311-cp311-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/84\/03\/acea680005f098f79fd70c1d9d5ccc0cb4296ec2af539a0450108232fc0c\/torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.11``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp311-cp311-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/ae\/0d\/98b410492609e34a155fa8b121b55c7dca229f39636851c3a9ec20edea21\/torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl) (80.5 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.11``macOS 11.0+ ARM64`\n\n[torch-2.11.0-cp310-cp310-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/35\/40\/2d532e8c0e23705be9d1debce5bc37b68d59a39bda7584c26fe9668076fe\/torch-2.11.0-cp310-cp310-win_amd64.whl) (114.5 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp310-cp310-win_amd64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.10``Windows x86-64`\n\n[torch-2.11.0-cp310-cp310-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/f9\/1e\/18a9b10b4bd34f12d4e561c52b0ae7158707b8193c6cfc0aad2b48167090\/torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.10``manylinux: glibc 2.28+ x86-64`\n\n[torch-2.11.0-cp310-cp310-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/a4\/f0\/98ae802fa8c09d3149b0c8690741f3f5753c90e779bd28c9613257295945\/torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.10``manylinux: glibc 2.28+ ARM64`\n\n[torch-2.11.0-cp310-cp310-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/ac\/f2\/c1690994afe461aae2d0cac62251e6802a703dec0a6c549c02ecd0de92a9\/torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl) (80.5 MB [view details](https:\/\/pypi.org\/project\/torch\/#torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl))\n\nUploaded Mar 23, 2026 `CPython 3.10``macOS 11.0+ ARM64`\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314t-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314t-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/cf\/bf\/c8d12a2c86dbfd7f40fb2f56fbf5a505ccf2d9ce131eb559dfc7c51e1a04\/torch-2.11.0-cp314-cp314t-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.8 MB\n- Tags: CPython 3.14t, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `b2a43985ff5ef6ddd923bbcf99943e5f58059805787c5c9a2622bf05ca2965b0` | Copy |\n| MD5 | `5e5a354ee5a4478788038514491c9b53` | Copy |\n| BLAKE2b-256 | `cfbfc8d12a2c86dbfd7f40fb2f56fbf5a505ccf2d9ce131eb559dfc7c51e1a04` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314t-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/b1\/e7\/0b6665f533aa9e337662dc190425abc0af1fe3234088f4454c52393ded61\/torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.6 MB\n- Tags: CPython 3.14t, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `563ed3d25542d7e7bbc5b235ccfacfeb97fb470c7fee257eae599adb8005c8a2` | Copy |\n| MD5 | `1fd06fd8ab11531a465edbd203708a92` | Copy |\n| BLAKE2b-256 | `b1e70b6665f533aa9e337662dc190425abc0af1fe3234088f4454c52393ded61` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314t-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/fd\/66\/54a56a4a6ceaffb567231994a9745821d3af922a854ed33b0b3a278e0a99\/torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.14t, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `ab9a8482f475f9ba20e12db84b0e55e2f58784bdca43a854a6ccd3fd4b9f75e6` | Copy |\n| MD5 | `1042d08d26381bf18f22fb2eeff9acf4` | Copy |\n| BLAKE2b-256 | `fd6654a56a4a6ceaffb567231994a9745821d3af922a854ed33b0b3a278e0a99` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314t-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/bf\/46\/4419098ed6d801750f26567b478fc185c3432e11e2cad712bc6b4c2ab0d0\/torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 81.0 MB\n- Tags: CPython 3.14t, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `8245477871c3700d4370352ffec94b103cfcb737229445cf9946cddb7b2ca7cd` | Copy |\n| MD5 | `b173521d650ee78e110f4e52d758ffea` | Copy |\n| BLAKE2b-256 | `bf464419098ed6d801750f26567b478fc185c3432e11e2cad712bc6b4c2ab0d0` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/78\/88\/d4a4cda8362f8a30d1ed428564878c3cafb0d87971fbd3947d4c84552095\/torch-2.11.0-cp314-cp314-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.6 MB\n- Tags: CPython 3.14, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `2b4e811728bd0cc58fb2b0948fe939a1ee2bf1422f6025be2fca4c7bd9d79718` | Copy |\n| MD5 | `78e8aec77b35ce2adba124b9fe686764` | Copy |\n| BLAKE2b-256 | `7888d4a4cda8362f8a30d1ed428564878c3cafb0d87971fbd3947d4c84552095` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/47\/e8\/b98ca2d39b2e0e4730c0ee52537e488e7008025bc77ca89552ff91021f7c\/torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.7 MB\n- Tags: CPython 3.14, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `4dc8b3809469b6c30b411bb8c4cad3828efd26236153d9beb6a3ec500f211a60` | Copy |\n| MD5 | `73e8a53e185b61d9e87ee0516a65fbc0` | Copy |\n| BLAKE2b-256 | `47e8b98ca2d39b2e0e4730c0ee52537e488e7008025bc77ca89552ff91021f7c` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/c7\/86\/7cd7c66cb9cec6be330fff36db5bd0eef386d80c031b581ec81be1d4b26c\/torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.14, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `2bb3cc54bd0dea126b0060bb1ec9de0f9c7f7342d93d436646516b0330cd5be7` | Copy |\n| MD5 | `808fab6c041012c9003e03b0fee4c96e` | Copy |\n| BLAKE2b-256 | `c7867cd7c66cb9cec6be330fff36db5bd0eef386d80c031b581ec81be1d4b26c` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp314-cp314-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/26\/0d\/8603382f61abd0db35841148ddc1ffd607bf3100b11c6e1dab6d2fc44e72\/torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 80.6 MB\n- Tags: CPython 3.14, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `01018087326984a33b64e04c8cb5c2795f9120e0d775ada1f6638840227b04d7` | Copy |\n| MD5 | `dbbef6f8b3c61fe3244f2f4cdc0f11a1` | Copy |\n| BLAKE2b-256 | `260d8603382f61abd0db35841148ddc1ffd607bf3100b11c6e1dab6d2fc44e72` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313t-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313t-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/48\/6b\/30d1459fa7e4b67e9e3fe1685ca1d8bb4ce7c62ef436c3a615963c6c866c\/torch-2.11.0-cp313-cp313t-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.8 MB\n- Tags: CPython 3.13t, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `a97b94bbf62992949b4730c6cd2cc9aee7b335921ee8dc207d930f2ed09ae2db` | Copy |\n| MD5 | `0f7535cb00b28ab4a04c8160a22fbbad` | Copy |\n| BLAKE2b-256 | `486b30d1459fa7e4b67e9e3fe1685ca1d8bb4ce7c62ef436c3a615963c6c866c` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313t-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/07\/f4\/1b666b6d61d3394cca306ea543ed03a64aad0a201b6cd159f1d41010aeb1\/torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.6 MB\n- Tags: CPython 3.13t, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `98bb213c3084cfe176302949bdc360074b18a9da7ab59ef2edc9d9f742504778` | Copy |\n| MD5 | `c46ed32625c87b1e75b502b11ebf720d` | Copy |\n| BLAKE2b-256 | `07f41b666b6d61d3394cca306ea543ed03a64aad0a201b6cd159f1d41010aeb1` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313t-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/6d\/6c\/56bfb37073e7136e6dd86bfc6af7339946dd684e0ecf2155ac0eee687ae1\/torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.13t, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `2658f34ce7e2dabf4ec73b45e2ca68aedad7a5be87ea756ad656eaf32bf1e1ea` | Copy |\n| MD5 | `af3c69887fbfc4fbc68b9602a8fcb308` | Copy |\n| BLAKE2b-256 | `6d6c56bfb37073e7136e6dd86bfc6af7339946dd684e0ecf2155ac0eee687ae1` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313t-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/db\/38\/8ac78069621b8c2b4979c2f96dc8409ef5e9c4189f6aac629189a78677ca\/torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 81.0 MB\n- Tags: CPython 3.13t, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `8b394322f49af4362d4f80e424bcaca7efcd049619af03a4cf4501520bdf0fb4` | Copy |\n| MD5 | `380d859cfad5f8262ba0b834da35bcf4` | Copy |\n| BLAKE2b-256 | `db388ac78069621b8c2b4979c2f96dc8409ef5e9c4189f6aac629189a78677ca` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/66\/82\/3e3fcdd388fbe54e29fd3f991f36846ff4ac90b0d0181e9c8f7236565f82\/torch-2.11.0-cp313-cp313-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.6 MB\n- Tags: CPython 3.13, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `4dda3b3f52d121063a731ddb835f010dc137b920d7fec2778e52f60d8e4bf0cd` | Copy |\n| MD5 | `0fe7969556fd22b5a190345511e1ae68` | Copy |\n| BLAKE2b-256 | `66823e3fcdd388fbe54e29fd3f991f36846ff4ac90b0d0181e9c8f7236565f82` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/3d\/e1\/b73f7c575a4b8f87a5928f50a1e35416b5e27295d8be9397d5293e7e8d4c\/torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.7 MB\n- Tags: CPython 3.13, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `cc89b9b173d9adfab59fd227f0ab5e5516d9a52b658ae41d64e59d2e55a418db` | Copy |\n| MD5 | `8f55c6619d670cc6ee89fbfde6cb4f5c` | Copy |\n| BLAKE2b-256 | `3de1b73f7c575a4b8f87a5928f50a1e35416b5e27295d8be9397d5293e7e8d4c` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/32\/d1\/8ed2173589cbfe744ed54e5a73efc107c0085ba5777ee93a5f4c1ab90553\/torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.13, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `63a68fa59de8f87acc7e85a5478bb2dddbb3392b7593ec3e78827c793c4b73fd` | Copy |\n| MD5 | `a15b86b7cda0a0a1613b306676ec374d` | Copy |\n| BLAKE2b-256 | `32d18ed2173589cbfe744ed54e5a73efc107c0085ba5777ee93a5f4c1ab90553` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp313-cp313-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/87\/89\/5ea6722763acee56b045435fb84258db7375c48165ec8be7880ab2b281c5\/torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 80.6 MB\n- Tags: CPython 3.13, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `1e6debd97ccd3205bbb37eb806a9d8219e1139d15419982c09e23ef7d4369d18` | Copy |\n| MD5 | `82d7d812c6109516d7cad4ca8002395e` | Copy |\n| BLAKE2b-256 | `87895ea6722763acee56b045435fb84258db7375c48165ec8be7880ab2b281c5` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp312-cp312-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp312-cp312-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/1c\/ff\/6756f1c7ee302f6d202120e0f4f05b432b839908f9071157302cedfc5232\/torch-2.11.0-cp312-cp312-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.6 MB\n- Tags: CPython 3.12, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `fbf39280699d1b869f55eac536deceaa1b60bd6788ba74f399cc67e60a5fab10` | Copy |\n| MD5 | `1799208f1391b3d694aee962cb9e5a57` | Copy |\n| BLAKE2b-256 | `1cff6756f1c7ee302f6d202120e0f4f05b432b839908f9071157302cedfc5232` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp312-cp312-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/1a\/c9\/82638ef24d7877510f83baf821f5619a61b45568ce21c0a87a91576510aa\/torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.7 MB\n- Tags: CPython 3.12, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `0f68f4ac6d95d12e896c3b7a912b5871619542ec54d3649cf48cc1edd4dd2756` | Copy |\n| MD5 | `2b481e9bac55320ccdcef091e3a4ac52` | Copy |\n| BLAKE2b-256 | `1ac982638ef24d7877510f83baf821f5619a61b45568ce21c0a87a91576510aa` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp312-cp312-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/13\/16\/42e5915ebe4868caa6bac83a8ed59db57f12e9a61b7d749d584776ed53d5\/torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.12, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `f99924682ef0aa6a4ab3b1b76f40dc6e273fca09f367d15a524266db100a723f` | Copy |\n| MD5 | `d44ddbd0e2d0166ef511d5d3ab19c64a` | Copy |\n| BLAKE2b-256 | `131642e5915ebe4868caa6bac83a8ed59db57f12e9a61b7d749d584776ed53d5` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp312-cp312-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/6f\/8b\/69e3008d78e5cee2b30183340cc425081b78afc5eff3d080daab0adda9aa\/torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 80.6 MB\n- Tags: CPython 3.12, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `4b5866312ee6e52ea625cd211dcb97d6a2cdc1131a5f15cc0d87eec948f6dd34` | Copy |\n| MD5 | `0a02d3ea665d48edc1fd7f2d70a76e08` | Copy |\n| BLAKE2b-256 | `6f8b69e3008d78e5cee2b30183340cc425081b78afc5eff3d080daab0adda9aa` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp311-cp311-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp311-cp311-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/d1\/bd\/9912d30b68845256aabbb4a40aeefeef3c3b20db5211ccda653544ada4b6\/torch-2.11.0-cp311-cp311-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.5 MB\n- Tags: CPython 3.11, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `73e24aaf8f36ab90d95cd1761208b2eb70841c2a9ca1a3f9061b39fc5331b708` | Copy |\n| MD5 | `66da62379541d52ab250ea3ae052dfe6` | Copy |\n| BLAKE2b-256 | `d1bd9912d30b68845256aabbb4a40aeefeef3c3b20db5211ccda653544ada4b6` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp311-cp311-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/8c\/8b\/d7be22fbec9ffee6cff31a39f8750d4b3a65d349a286cf4aec74c2375662\/torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.6 MB\n- Tags: CPython 3.11, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `7aa2f9bbc6d4595ba72138026b2074be1233186150e9292865e04b7a63b8c67a` | Copy |\n| MD5 | `aa9c5f374a0a6834cb940b42a29978ae` | Copy |\n| BLAKE2b-256 | `8c8bd7be22fbec9ffee6cff31a39f8750d4b3a65d349a286cf4aec74c2375662` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp311-cp311-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/84\/03\/acea680005f098f79fd70c1d9d5ccc0cb4296ec2af539a0450108232fc0c\/torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.11, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `d91aac77f24082809d2c5a93f52a5f085032740a1ebc9252a7b052ef5a4fddc6` | Copy |\n| MD5 | `f471c3392876a56dcf1f57a200b649de` | Copy |\n| BLAKE2b-256 | `8403acea680005f098f79fd70c1d9d5ccc0cb4296ec2af539a0450108232fc0c` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp311-cp311-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/ae\/0d\/98b410492609e34a155fa8b121b55c7dca229f39636851c3a9ec20edea21\/torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 80.5 MB\n- Tags: CPython 3.11, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `7b6a60d48062809f58595509c524b88e6ddec3ebe25833d6462eeab81e5f2ce4` | Copy |\n| MD5 | `544775e2610b48889b875137a15f0f79` | Copy |\n| BLAKE2b-256 | `ae0d98b410492609e34a155fa8b121b55c7dca229f39636851c3a9ec20edea21` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp310-cp310-win_amd64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp310-cp310-win\\_amd64.whl](https:\/\/files.pythonhosted.org\/packages\/35\/40\/2d532e8c0e23705be9d1debce5bc37b68d59a39bda7584c26fe9668076fe\/torch-2.11.0-cp310-cp310-win_amd64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 114.5 MB\n- Tags: CPython 3.10, Windows x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `b3c712ae6fb8e7a949051a953fc412fe0a6940337336c3b6f905e905dac5157f` | Copy |\n| MD5 | `92a04a9fc443c7ff23226e10a3af704f` | Copy |\n| BLAKE2b-256 | `35402d532e8c0e23705be9d1debce5bc37b68d59a39bda7584c26fe9668076fe` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp310-cp310-manylinux\\_2\\_28\\_x86\\_64.whl](https:\/\/files.pythonhosted.org\/packages\/f9\/1e\/18a9b10b4bd34f12d4e561c52b0ae7158707b8193c6cfc0aad2b48167090\/torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 530.6 MB\n- Tags: CPython 3.10, manylinux: glibc 2.28+ x86-64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `1b32ceda909818a03b112006709b02be1877240c31750a8d9c6b7bf5f2d8a6e5` | Copy |\n| MD5 | `1552a6d321733cf7a046d92afbb6ac67` | Copy |\n| BLAKE2b-256 | `f91e18a9b10b4bd34f12d4e561c52b0ae7158707b8193c6cfc0aad2b48167090` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp310-cp310-manylinux\\_2\\_28\\_aarch64.whl](https:\/\/files.pythonhosted.org\/packages\/a4\/f0\/98ae802fa8c09d3149b0c8690741f3f5753c90e779bd28c9613257295945\/torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 419.7 MB\n- Tags: CPython 3.10, manylinux: glibc 2.28+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `4cf8687f4aec3900f748d553483ef40e0ac38411c3c48d0a86a438f6d7a99b18` | Copy |\n| MD5 | `893da691a4d1d289f24de063a7835f57` | Copy |\n| BLAKE2b-256 | `a4f098ae802fa8c09d3149b0c8690741f3f5753c90e779bd28c9613257295945` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n## File details\nDetails for the file `torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl`.\n\n### File metadata\n- Download URL: [torch-2.11.0-cp310-cp310-macosx\\_11\\_0\\_arm64.whl](https:\/\/files.pythonhosted.org\/packages\/ac\/f2\/c1690994afe461aae2d0cac62251e6802a703dec0a6c549c02ecd0de92a9\/torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl)\n- Upload date:\n  Mar 23, 2026\n- Size: 80.5 MB\n- Tags: CPython 3.10, macOS 11.0+ ARM64\n- Uploaded using Trusted Publishing? No\n- Uploaded via: twine\/6.2.0 CPython\/3.11.6\n### File hashes\n| Algorithm | Hash digest |   |\n|---|---|---|\n| SHA256 | `2c0d7fcfbc0c4e8bb5ebc3907cbc0c6a0da1b8f82b1fc6e14e914fa0b9baf74e` | Copy |\n| MD5 | `a052f0ae699d2599a2d0f43a083f9275` | Copy |\n| BLAKE2b-256 | `acf2c1690994afe461aae2d0cac62251e6802a703dec0a6c549c02ecd0de92a9` | Copy |\n\n[See more details on using hashes here.](https:\/\/pip.pypa.io\/en\/stable\/topics\/secure-installs\/#hash-checking-mode \"External link\")\n\n![](https:\/\/pypi.org\/static\/images\/white-cube.2351a86c.svg)\n\n## Help\n- [Installing packages](https:\/\/packaging.python.org\/tutorials\/installing-packages\/ \"External link\")\n- [Uploading packages](https:\/\/packaging.python.org\/tutorials\/packaging-projects\/ \"External link\")\n- [User guide](https:\/\/packaging.python.org\/ \"External link\")\n- [Project name retention](https:\/\/www.python.org\/dev\/peps\/pep-0541\/ \"External link\")\n- [FAQs](https:\/\/pypi.org\/help\/)\n\n## About PyPI\n- [PyPI Blog](https:\/\/blog.pypi.org\/ \"External link\")\n- [Infrastructure dashboard](https:\/\/dtdg.co\/pypi \"External link\")\n- [Statistics](https:\/\/pypi.org\/stats\/)\n- [Logos & trademarks](https:\/\/pypi.org\/trademarks\/)\n- [Our sponsors](https:\/\/pypi.org\/sponsors\/)\n\n## Contributing to PyPI\n- [Bugs and feedback](https:\/\/pypi.org\/help\/#feedback)\n- [Contribute on GitHub](https:\/\/github.com\/pypi\/warehouse \"External link\")\n- [Translate PyPI](https:\/\/hosted.weblate.org\/projects\/pypa\/warehouse\/ \"External link\")\n- [Sponsor PyPI](https:\/\/pypi.org\/sponsors\/)\n- [Development credits](https:\/\/github.com\/pypi\/warehouse\/graphs\/contributors \"External link\")\n\n## Using PyPI\n- [Terms of Service](https:\/\/policies.python.org\/pypi.org\/Terms-of-Service\/ \"External link\")\n- [Report security issue](https:\/\/pypi.org\/security\/)\n- [Code of conduct](https:\/\/policies.python.org\/python.org\/code-of-conduct\/ \"External link\")\n- [Privacy Notice](https:\/\/policies.python.org\/pypi.org\/Privacy-Notice\/ \"External link\")\n- [Acceptable Use Policy](https:\/\/policies.python.org\/pypi.org\/Acceptable-Use-Policy\/ \"External link\")\n***\nStatus: [all systems operational](https:\/\/status.python.org\/ \"External link\")\n\nDeveloped and maintained by the Python community, for the Python community.   \n [Donate today\\!](https:\/\/donate.pypi.org\/)\n\n\"PyPI\", \"Python Package Index\", and the blocks logos are registered [trademarks](https:\/\/pypi.org\/trademarks\/) of the [Python Software Foundation](https:\/\/www.python.org\/psf-landing).\n\n© 2026 [Python Software Foundation](https:\/\/www.python.org\/psf-landing\/ \"External link\")   \n [Site map](https:\/\/pypi.org\/sitemap\/)\n\nSwitch to desktop version\n\nSupported by\n\n[![](https:\/\/pypi-camo.freetls.fastly.net\/ed7074cadad1a06f56bc520ad9bd3e00d0704c5b\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f6177732d77686974652d6c6f676f2d7443615473387a432e706e67) AWS Cloud computing and Security Sponsor](https:\/\/aws.amazon.com\/)   [![](https:\/\/pypi-camo.freetls.fastly.net\/8855f7c063a3bdb5b0ce8d91bfc50cf851cc5c51\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f64617461646f672d77686974652d6c6f676f2d6668644c4e666c6f2e706e67) Datadog Monitoring](https:\/\/www.datadoghq.com\/)   [![](https:\/\/pypi-camo.freetls.fastly.net\/60f709d24f3e4d469f9adc77c65e2f5291a3d165\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f6465706f742d77686974652d6c6f676f2d7038506f476831302e706e67) Depot Continuous Integration](https:\/\/depot.dev\/)   [![](https:\/\/pypi-camo.freetls.fastly.net\/df6fe8829cbff2d7f668d98571df1fd011f36192\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f666173746c792d77686974652d6c6f676f2d65684d3077735f6f2e706e67) Fastly CDN](https:\/\/www.fastly.com\/)   [![](https:\/\/pypi-camo.freetls.fastly.net\/420cc8cf360bac879e24c923b2f50ba7d1314fb0\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f676f6f676c652d77686974652d6c6f676f2d616734424e3774332e706e67) Google Download Analytics](https:\/\/careers.google.com\/)   [![](https:\/\/pypi-camo.freetls.fastly.net\/d01053c02f3a626b73ffcb06b96367fdbbf9e230\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f70696e67646f6d2d77686974652d6c6f676f2d67355831547546362e706e67) Pingdom Monitoring](https:\/\/www.pingdom.com\/)   [![](https:\/\/pypi-camo.freetls.fastly.net\/67af7117035e2345bacb5a82e9aa8b5b3e70701d\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f73656e7472792d77686974652d6c6f676f2d4a2d6b64742d706e2e706e67) Sentry Error logging](https:\/\/sentry.io\/for\/python\/?utm_source=pypi&utm_medium=paid-community&utm_campaign=python-na-evergreen&utm_content=static-ad-pypi-sponsor-learnmore)   [![](https:\/\/pypi-camo.freetls.fastly.net\/b611884ff90435a0575dbab7d9b0d3e60f136466\/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f737461747573706167652d77686974652d6c6f676f2d5467476c6a4a2d502e706e67) StatusPage Status page](https:\/\/statuspage.io\/)","attrs_readable_markdown":"![PyTorch Logo](https:\/\/pypi-camo.freetls.fastly.net\/ed3839e1c11e779b508097969affa63d0968692b\/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f7079746f7263682d6c6f676f2d6461726b2e706e67)\n***\nPyTorch is a Python package that provides two high-level features:\n\n- Tensor computation (like NumPy) with strong GPU acceleration\n- Deep neural networks built on a tape-based autograd system\n\nYou can reuse your favorite Python packages such as NumPy, SciPy, and Cython to extend PyTorch when needed.\n\nOur trunk health (Continuous Integration signals) can be found at [hud.pytorch.org](https:\/\/hud.pytorch.org\/ci\/pytorch\/pytorch\/main).\n\n- [More About PyTorch](https:\/\/pypi.org\/project\/torch\/#more-about-pytorch)\n  - [A GPU-Ready Tensor Library](https:\/\/pypi.org\/project\/torch\/#a-gpu-ready-tensor-library)\n  - [Dynamic Neural Networks: Tape-Based Autograd](https:\/\/pypi.org\/project\/torch\/#dynamic-neural-networks-tape-based-autograd)\n  - [Python First](https:\/\/pypi.org\/project\/torch\/#python-first)\n  - [Imperative Experiences](https:\/\/pypi.org\/project\/torch\/#imperative-experiences)\n  - [Fast and Lean](https:\/\/pypi.org\/project\/torch\/#fast-and-lean)\n  - [Extensions Without Pain](https:\/\/pypi.org\/project\/torch\/#extensions-without-pain)\n- [Installation](https:\/\/pypi.org\/project\/torch\/#installation)\n  - [Binaries](https:\/\/pypi.org\/project\/torch\/#binaries)\n    - [NVIDIA Jetson Platforms](https:\/\/pypi.org\/project\/torch\/#nvidia-jetson-platforms)\n  - [From Source](https:\/\/pypi.org\/project\/torch\/#from-source)\n    - [Prerequisites](https:\/\/pypi.org\/project\/torch\/#prerequisites)\n      - [NVIDIA CUDA Support](https:\/\/pypi.org\/project\/torch\/#nvidia-cuda-support)\n      - [AMD ROCm Support](https:\/\/pypi.org\/project\/torch\/#amd-rocm-support)\n      - [Intel GPU Support](https:\/\/pypi.org\/project\/torch\/#intel-gpu-support)\n    - [Get the PyTorch Source](https:\/\/pypi.org\/project\/torch\/#get-the-pytorch-source)\n    - [Install Dependencies](https:\/\/pypi.org\/project\/torch\/#install-dependencies)\n    - [Install PyTorch](https:\/\/pypi.org\/project\/torch\/#install-pytorch)\n      - [Adjust Build Options (Optional)](https:\/\/pypi.org\/project\/torch\/#adjust-build-options-optional)\n  - [Docker Image](https:\/\/pypi.org\/project\/torch\/#docker-image)\n    - [Using pre-built images](https:\/\/pypi.org\/project\/torch\/#using-pre-built-images)\n    - [Building the image yourself](https:\/\/pypi.org\/project\/torch\/#building-the-image-yourself)\n  - [Building the Documentation](https:\/\/pypi.org\/project\/torch\/#building-the-documentation)\n    - [Building a PDF](https:\/\/pypi.org\/project\/torch\/#building-a-pdf)\n  - [Previous Versions](https:\/\/pypi.org\/project\/torch\/#previous-versions)\n- [Getting Started](https:\/\/pypi.org\/project\/torch\/#getting-started)\n- [Resources](https:\/\/pypi.org\/project\/torch\/#resources)\n- [Communication](https:\/\/pypi.org\/project\/torch\/#communication)\n- [Releases and Contributing](https:\/\/pypi.org\/project\/torch\/#releases-and-contributing)\n- [The Team](https:\/\/pypi.org\/project\/torch\/#the-team)\n- [License](https:\/\/pypi.org\/project\/torch\/#license)\n## More About PyTorch\n[Learn the basics of PyTorch](https:\/\/pytorch.org\/tutorials\/beginner\/basics\/intro.html)\n\nAt a granular level, PyTorch is a library that consists of the following components:\n\n| Component | Description |\n|---|---|\n| [**torch**](https:\/\/pytorch.org\/docs\/stable\/torch.html) | A Tensor library like NumPy, with strong GPU support |\n| [**torch.autograd**](https:\/\/pytorch.org\/docs\/stable\/autograd.html) | A tape-based automatic differentiation library that supports all differentiable Tensor operations in torch |\n| [**torch.jit**](https:\/\/pytorch.org\/docs\/stable\/jit.html) | A compilation stack (TorchScript) to create serializable and optimizable models from PyTorch code |\n| [**torch.nn**](https:\/\/pytorch.org\/docs\/stable\/nn.html) | A neural networks library deeply integrated with autograd designed for maximum flexibility |\n| [**torch.multiprocessing**](https:\/\/pytorch.org\/docs\/stable\/multiprocessing.html) | Python multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and Hogwild training |\n| [**torch.utils**](https:\/\/pytorch.org\/docs\/stable\/data.html) | DataLoader and other utility functions for convenience |\n\nUsually, PyTorch is used either as:\n\n- A replacement for NumPy to use the power of GPUs.\n- A deep learning research platform that provides maximum flexibility and speed.\n\nElaborating Further:\n\n### A GPU-Ready Tensor Library\nIf you use NumPy, then you have used Tensors (a.k.a. ndarray).\n\n![Tensor illustration](https:\/\/pypi-camo.freetls.fastly.net\/e5f621cdea28d4de40577ac3346794f746d96b86\/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f74656e736f725f696c6c757374726174696f6e2e706e67)\n\nPyTorch provides Tensors that can live either on the CPU or the GPU and accelerates the computation by a huge amount.\n\nWe provide a wide variety of tensor routines to accelerate and fit your scientific computation needs such as slicing, indexing, mathematical operations, linear algebra, reductions. And they are fast\\!\n\n### Dynamic Neural Networks: Tape-Based Autograd\nPyTorch has a unique way of building neural networks: using and replaying a tape recorder.\n\nMost frameworks such as TensorFlow, Theano, Caffe, and CNTK have a static view of the world. One has to build a neural network and reuse the same structure again and again. Changing the way the network behaves means that one has to start from scratch.\n\nWith PyTorch, we use a technique called reverse-mode auto-differentiation, which allows you to change the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes from several research papers on this topic, as well as current and past work such as [torch-autograd](https:\/\/github.com\/twitter\/torch-autograd), [autograd](https:\/\/github.com\/HIPS\/autograd), [Chainer](https:\/\/chainer.org\/), etc.\n\nWhile this technique is not unique to PyTorch, it's one of the fastest implementations of it to date. You get the best of speed and flexibility for your crazy research.\n\n![Dynamic graph](https:\/\/pypi-camo.freetls.fastly.net\/360dc3e00cca046f73204c87f117feefe993580f\/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f64796e616d69635f67726170682e676966)\n\n### Python First\nPyTorch is not a Python binding into a monolithic C++ framework. It is built to be deeply integrated into Python. You can use it naturally like you would use [NumPy](https:\/\/www.numpy.org\/) \/ [SciPy](https:\/\/www.scipy.org\/) \/ [scikit-learn](https:\/\/scikit-learn.org\/) etc. You can write your new neural network layers in Python itself, using your favorite libraries and use packages such as [Cython](https:\/\/cython.org\/) and [Numba](http:\/\/numba.pydata.org\/). Our goal is to not reinvent the wheel where appropriate.\n\n### Imperative Experiences\nPyTorch is designed to be intuitive, linear in thought, and easy to use. When you execute a line of code, it gets executed. There isn't an asynchronous view of the world. When you drop into a debugger or receive error messages and stack traces, understanding them is straightforward. The stack trace points to exactly where your code was defined. We hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines.\n\n### Fast and Lean\nPyTorch has minimal framework overhead. We integrate acceleration libraries such as [Intel MKL](https:\/\/software.intel.com\/mkl) and NVIDIA ([cuDNN](https:\/\/developer.nvidia.com\/cudnn), [NCCL](https:\/\/developer.nvidia.com\/nccl)) to maximize speed. At the core, its CPU and GPU Tensor and neural network backends are mature and have been tested for years.\n\nHence, PyTorch is quite fast — whether you run small or large neural networks.\n\nThe memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives. We've written custom memory allocators for the GPU to make sure that your deep learning models are maximally memory efficient. This enables you to train bigger deep learning models than before.\n\n### Extensions Without Pain\nWriting new neural network modules, or interfacing with PyTorch's Tensor API was designed to be straightforward and with minimal abstractions.\n\nYou can write new neural network layers in Python using the torch API [or your favorite NumPy-based libraries such as SciPy](https:\/\/pytorch.org\/tutorials\/advanced\/numpy_extensions_tutorial.html).\n\nIf you want to write your layers in C\/C++, we provide a convenient extension API that is efficient and with minimal boilerplate. No wrapper code needs to be written. You can see [a tutorial here](https:\/\/pytorch.org\/tutorials\/advanced\/cpp_extension.html) and [an example here](https:\/\/github.com\/pytorch\/extension-cpp).\n\n## Installation\n### Binaries\nCommands to install binaries via Conda or pip wheels are on our website: <https:\/\/pytorch.org\/get-started\/locally\/>\n\n#### NVIDIA Jetson Platforms\nPython wheels for NVIDIA's Jetson Nano, Jetson TX1\/TX2, Jetson Xavier NX\/AGX, and Jetson AGX Orin are provided [here](https:\/\/forums.developer.nvidia.com\/t\/pytorch-for-jetson-version-1-10-now-available\/72048) and the L4T container is published [here](https:\/\/catalog.ngc.nvidia.com\/orgs\/nvidia\/containers\/l4t-pytorch)\n\nThey require JetPack 4.2 and above, and [@dusty-nv](https:\/\/github.com\/dusty-nv) and [@ptrblck](https:\/\/github.com\/ptrblck) are maintaining them.\n\n### From Source\n#### Prerequisites\nIf you are installing from source, you will need:\n\n- Python 3.10 or later\n- A compiler that fully supports C++17, such as clang or gcc (gcc 9.4.0 or newer is required, on Linux)\n- Visual Studio or Visual Studio Build Tool (Windows only)\n\n\\* PyTorch CI uses Visual C++ BuildTools, which come with Visual Studio Enterprise, Professional, or Community Editions. You can also install the build tools from <https:\/\/visualstudio.microsoft.com\/visual-cpp-build-tools\/>. The build tools *do not* come with Visual Studio Code by default.\n\nAn example of environment setup is shown below:\n\n- Linux:\n```\n$ source <CONDA_INSTALL_DIR>\/bin\/activate\n$ conda create -y -n <CONDA_NAME>\n$ conda activate <CONDA_NAME>\n```\n- Windows:\n```\n$ source <CONDA_INSTALL_DIR>\\Scripts\\activate.bat\n$ conda create -y -n <CONDA_NAME>\n$ conda activate <CONDA_NAME>\n$ call \"C:\\Program Files\\Microsoft Visual Studio\\<VERSION>\\Community\\VC\\Auxiliary\\Build\\vcvarsall.bat\" x64\n```\nA conda environment is not required. You can also do a PyTorch build in a standard virtual environment, e.g., created with tools like `uv`, provided your system has installed all the necessary dependencies unavailable as pip packages (e.g., CUDA, MKL.)\n\n##### NVIDIA CUDA Support\nIf you want to compile with CUDA support, [select a supported version of CUDA from our support matrix](https:\/\/pytorch.org\/get-started\/locally\/), then install the following:\n\n- [NVIDIA CUDA](https:\/\/developer.nvidia.com\/cuda-downloads)\n- [NVIDIA cuDNN](https:\/\/developer.nvidia.com\/cudnn) v8.5 or above\n- [Compiler](https:\/\/gist.github.com\/ax3l\/9489132) compatible with CUDA\n\nNote: You could refer to the [cuDNN Support Matrix](https:\/\/docs.nvidia.com\/deeplearning\/cudnn\/backend\/latest\/reference\/support-matrix.html) for cuDNN versions with the various supported CUDA, CUDA driver, and NVIDIA hardware.\n\nIf you want to disable CUDA support, export the environment variable `USE_CUDA=0`. Other potentially useful environment variables may be found in `setup.py`. If CUDA is installed in a non-standard location, set PATH so that the nvcc you want to use can be found (e.g., `export PATH=\/usr\/local\/cuda-12.8\/bin:$PATH`).\n\nIf you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are [available here](https:\/\/devtalk.nvidia.com\/default\/topic\/1049071\/jetson-nano\/pytorch-for-jetson-nano\/)\n\n##### AMD ROCm Support\nIf you want to compile with ROCm support, install\n\n- [AMD ROCm](https:\/\/rocm.docs.amd.com\/en\/latest\/deploy\/linux\/quick_start.html) 4.0 and above installation\n- ROCm is currently supported only for Linux systems.\n\nBy default the build system expects ROCm to be installed in `\/opt\/rocm`. If ROCm is installed in a different directory, the `ROCM_PATH` environment variable must be set to the ROCm installation directory. The build system automatically detects the AMD GPU architecture. Optionally, the AMD GPU architecture can be explicitly set with the `PYTORCH_ROCM_ARCH` environment variable [AMD GPU architecture](https:\/\/rocm.docs.amd.com\/projects\/install-on-linux\/en\/latest\/reference\/system-requirements.html#supported-gpus)\n\nIf you want to disable ROCm support, export the environment variable `USE_ROCM=0`. Other potentially useful environment variables may be found in `setup.py`.\n\n##### Intel GPU Support\nIf you want to compile with Intel GPU support, follow these\n\n- [PyTorch Prerequisites for Intel GPUs](https:\/\/www.intel.com\/content\/www\/us\/en\/developer\/articles\/tool\/pytorch-prerequisites-for-intel-gpu.html) instructions.\n- Intel GPU is supported for Linux and Windows.\n\nIf you want to disable Intel GPU support, export the environment variable `USE_XPU=0`. Other potentially useful environment variables may be found in `setup.py`.\n\n#### Get the PyTorch Source\n```\ngit clone https:\/\/github.com\/pytorch\/pytorch\ncd pytorch\n# if you are updating an existing checkout\ngit submodule sync\ngit submodule update --init --recursive\n```\n#### Install Dependencies\n**Common**\n```\n# Run this command from the PyTorch directory after cloning the source code using the “Get the PyTorch Source“ section above\npip install --group dev\n```\n**On Linux**\n```\npip install mkl-static mkl-include\n# CUDA only: Add LAPACK support for the GPU if needed\n# magma installation: run with active conda environment. specify CUDA version to install\n.ci\/docker\/common\/install_magma_conda.sh 12.4\n\n# (optional) If using torch.compile with inductor\/triton, install the matching version of triton\n# Run from the pytorch directory after cloning\n# For Intel GPU support, please explicitly `export USE_XPU=1` before running command.\nmake triton\n```\n**On MacOS**\n```\n# Add this package on intel x86 processor machines only\npip install mkl-static mkl-include\n# Add these packages if torch.distributed is needed\nconda install pkg-config libuv\n```\n**On Windows**\n```\npip install mkl-static mkl-include\n# Add these packages if torch.distributed is needed.\n# Distributed package support on Windows is a prototype feature and is subject to changes.\nconda install -c conda-forge libuv=1.51\n```\n#### Install PyTorch\n**On Linux**\n\nIf you're compiling for AMD ROCm then first run this command:\n```\n# Only run this if you're compiling for ROCm\npython tools\/amd_build\/build_amd.py\n```\nInstall PyTorch\n```\n# the CMake prefix for conda environment\nexport CMAKE_PREFIX_PATH=\"${CONDA_PREFIX:-'$(dirname $(which conda))\/..\/'}:${CMAKE_PREFIX_PATH}\"\npython -m pip install --no-build-isolation -v -e .\n\n# the CMake prefix for non-conda environment, e.g. Python venv\n# call following after activating the venv\nexport CMAKE_PREFIX_PATH=\"${VIRTUAL_ENV}:${CMAKE_PREFIX_PATH}\"\n```\n**On macOS**\n```\npython -m pip install --no-build-isolation -v -e .\n```\n**On Windows**\n\nIf you want to build legacy python code, please refer to [Building on legacy code and CUDA](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/CONTRIBUTING.md#building-on-legacy-code-and-cuda)\n\n**CPU-only builds**\n\nIn this mode PyTorch computations will run on your CPU, not your GPU.\n```\npython -m pip install --no-build-isolation -v -e .\n```\nNote on OpenMP: The desired OpenMP implementation is Intel OpenMP (iomp). In order to link against iomp, you'll need to manually download the library and set up the building environment by tweaking `CMAKE_INCLUDE_PATH` and `LIB`. The instruction [here](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/docs\/source\/notes\/windows.rst#building-from-source) is an example for setting up both MKL and Intel OpenMP. Without these configurations for CMake, Microsoft Visual C OpenMP runtime (vcomp) will be used.\n\n**CUDA based build**\n\nIn this mode PyTorch computations will leverage your GPU via CUDA for faster number crunching\n\n[NVTX](https:\/\/docs.nvidia.com\/gameworks\/content\/gameworkslibrary\/nvtx\/nvidia_tools_extension_library_nvtx.htm) is needed to build PyTorch with CUDA. NVTX is a part of CUDA distributive, where it is called \"Nsight Compute\". To install it onto an already installed CUDA run CUDA installation once again and check the corresponding checkbox. Make sure that CUDA with Nsight Compute is installed after Visual Studio.\n\nCurrently, VS 2017 \/ 2019, and Ninja are supported as the generator of CMake. If `ninja.exe` is detected in `PATH`, then Ninja will be used as the default generator, otherwise, it will use VS 2017 \/ 2019.   \n If Ninja is selected as the generator, the latest MSVC will get selected as the underlying toolchain.\n\nAdditional libraries such as [Magma](https:\/\/developer.nvidia.com\/magma), [oneDNN, a.k.a. MKLDNN or DNNL](https:\/\/github.com\/oneapi-src\/oneDNN), and [Sccache](https:\/\/github.com\/mozilla\/sccache) are often needed. Please refer to the [installation-helper](https:\/\/github.com\/pytorch\/pytorch\/tree\/main\/.ci\/pytorch\/win-test-helpers\/installation-helpers) to install them.\n\nYou can refer to the [build\\_pytorch.bat](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/.ci\/pytorch\/win-test-helpers\/build_pytorch.bat) script for some other environment variables configurations\n```\ncmd\n\n:: Set the environment variables after you have downloaded and unzipped the mkl package,\n:: else CMake would throw an error as `Could NOT find OpenMP`.\nset CMAKE_INCLUDE_PATH={Your directory}\\mkl\\include\nset LIB={Your directory}\\mkl\\lib;%LIB%\n\n:: Read the content in the previous section carefully before you proceed.\n:: [Optional] If you want to override the underlying toolset used by Ninja and Visual Studio with CUDA, please run the following script block.\n:: \"Visual Studio 2019 Developer Command Prompt\" will be run automatically.\n:: Make sure you have CMake >= 3.12 before you do this when you use the Visual Studio generator.\nset CMAKE_GENERATOR_TOOLSET_VERSION=14.27\nset DISTUTILS_USE_SDK=1\nfor \/f \"usebackq tokens=*\" %i in (`\"%ProgramFiles(x86)%\\Microsoft Visual Studio\\Installer\\vswhere.exe\" -version [15^,17^) -products * -latest -property installationPath`) do call \"%i\\VC\\Auxiliary\\Build\\vcvarsall.bat\" x64 -vcvars_ver=%CMAKE_GENERATOR_TOOLSET_VERSION%\n\n:: [Optional] If you want to override the CUDA host compiler\nset CUDAHOSTCXX=C:\\Program Files (x86)\\Microsoft Visual Studio\\2019\\Community\\VC\\Tools\\MSVC\\14.27.29110\\bin\\HostX64\\x64\\cl.exe\n\npython -m pip install --no-build-isolation -v -e .\n```\n**Intel GPU builds**\n\nIn this mode PyTorch with Intel GPU support will be built.\n\nPlease make sure [the common prerequisites](https:\/\/pypi.org\/project\/torch\/#prerequisites) as well as [the prerequisites for Intel GPU](https:\/\/pypi.org\/project\/torch\/#intel-gpu-support) are properly installed and the environment variables are configured prior to starting the build. For build tool support, `Visual Studio 2022` is required.\n\nThen PyTorch can be built with the command:\n```\n:: CMD Commands:\n:: Set the CMAKE_PREFIX_PATH to help find corresponding packages\n:: %CONDA_PREFIX% only works after `conda activate custom_env`\n\nif defined CMAKE_PREFIX_PATH (\n    set \"CMAKE_PREFIX_PATH=%CONDA_PREFIX%\\Library;%CMAKE_PREFIX_PATH%\"\n) else (\n    set \"CMAKE_PREFIX_PATH=%CONDA_PREFIX%\\Library\"\n)\n\npython -m pip install --no-build-isolation -v -e .\n```\n##### Adjust Build Options (Optional)\nYou can adjust the configuration of cmake variables optionally (without building first), by doing the following. For example, adjusting the pre-detected directories for CuDNN or BLAS can be done with such a step.\n\nOn Linux\n```\nexport CMAKE_PREFIX_PATH=\"${CONDA_PREFIX:-'$(dirname $(which conda))\/..\/'}:${CMAKE_PREFIX_PATH}\"\nCMAKE_ONLY=1 python setup.py build\nccmake build  # or cmake-gui build\n```\nOn macOS\n```\nexport CMAKE_PREFIX_PATH=\"${CONDA_PREFIX:-'$(dirname $(which conda))\/..\/'}:${CMAKE_PREFIX_PATH}\"\nMACOSX_DEPLOYMENT_TARGET=11.0 CMAKE_ONLY=1 python setup.py build\nccmake build  # or cmake-gui build\n```\n### Docker Image\n#### Using pre-built images\nYou can also pull a pre-built docker image from Docker Hub and run with docker v19.03+\n```\ndocker run --gpus all --rm -ti --ipc=host pytorch\/pytorch:latest\n```\nPlease note that PyTorch uses shared memory to share data between processes, so if torch multiprocessing is used (e.g. for multithreaded data loaders) the default shared memory segment size that container runs with is not enough, and you should increase shared memory size either with `--ipc=host` or `--shm-size` command line options to `nvidia-docker run`.\n\n#### Building the image yourself\n**NOTE:** Must be built with a docker version \\> 18.06\n\nThe `Dockerfile` is supplied to build images with CUDA 11.1 support and cuDNN v8. You can pass `PYTHON_VERSION=x.y` make variable to specify which Python version is to be used by Miniconda, or leave it unset to use the default.\n```\nmake -f docker.Makefile\n# images are tagged as docker.io\/${your_docker_username}\/pytorch\n```\nYou can also pass the `CMAKE_VARS=\"...\"` environment variable to specify additional CMake variables to be passed to CMake during the build. See [setup.py](https:\/\/pypi.org\/project\/torch\/setup.py) for the list of available variables.\n```\nmake -f docker.Makefile\n```\n### Building the Documentation\nTo build documentation in various formats, you will need [Sphinx](http:\/\/www.sphinx-doc.org\/) and the pytorch\\_sphinx\\_theme2.\n\nBefore you build the documentation locally, ensure `torch` is installed in your environment. For small fixes, you can install the nightly version as described in [Getting Started](https:\/\/pytorch.org\/get-started\/locally\/).\n\nFor more complex fixes, such as adding a new module and docstrings for the new module, you might need to install torch [from source](https:\/\/pypi.org\/project\/torch\/#from-source). See [Docstring Guidelines](https:\/\/github.com\/pytorch\/pytorch\/wiki\/Docstring-Guidelines) for docstring conventions.\n```\ncd docs\/\npip install -r requirements.txt\nmake html\nmake serve\n```\nRun `make` to get a list of all available output formats.\n\nIf you get a katex error run `npm install katex`. If it persists, try `npm install -g katex`\n\n> \\[!NOTE\\] If you installed `nodejs` with a different package manager (e.g., `conda`) then `npm` will probably install a version of `katex` that is not compatible with your version of `nodejs` and doc builds will fail. A combination of versions that is known to work is `node@6.13.1` and `katex@0.13.18`. To install the latter with `npm` you can run `npm install -g katex@0.13.18`\n\n> \\[!NOTE\\] If you see a numpy incompatibility error, run:\n> ```\n> pip install 'numpy<2'\n> ```\n\nWhen you make changes to the dependencies run by CI, edit the `.ci\/docker\/requirements-docs.txt` file.\n\n#### Building a PDF\nTo compile a PDF of all PyTorch documentation, ensure you have `texlive` and LaTeX installed. On macOS, you can install them using:\n```\nbrew install --cask mactex\n```\nTo create the PDF:\n\n1. Run:\n   ```\n   make latexpdf\n   ```\n   This will generate the necessary files in the `build\/latex` directory.\n2. Navigate to this directory and execute:\n   ```\n   make LATEXOPTS=\"-interaction=nonstopmode\"\n   ```\n   This will produce a `pytorch.pdf` with the desired content. Run this command one more time so that it generates the correct table of contents and index.\n\n> \\[!NOTE\\] To view the Table of Contents, switch to the **Table of Contents** view in your PDF viewer.\n### Previous Versions\nInstallation instructions and binaries for previous PyTorch versions may be found on [our website](https:\/\/pytorch.org\/get-started\/previous-versions).\n\n## Getting Started\nThree pointers to get you started:\n\n- [Tutorials: get you started with understanding and using PyTorch](https:\/\/pytorch.org\/tutorials\/)\n- [Examples: easy to understand PyTorch code across all domains](https:\/\/github.com\/pytorch\/examples)\n- [The API Reference](https:\/\/pytorch.org\/docs\/)\n- [Glossary](https:\/\/github.com\/pytorch\/pytorch\/blob\/main\/GLOSSARY.md)\n## Resources\n- [PyTorch.org](https:\/\/pytorch.org\/)\n- [PyTorch Tutorials](https:\/\/pytorch.org\/tutorials\/)\n- [PyTorch Examples](https:\/\/github.com\/pytorch\/examples)\n- [PyTorch Models](https:\/\/pytorch.org\/hub\/)\n- [Intro to Deep Learning with PyTorch from Udacity](https:\/\/www.udacity.com\/course\/deep-learning-pytorch--ud188)\n- [Intro to Machine Learning with PyTorch from Udacity](https:\/\/www.udacity.com\/course\/intro-to-machine-learning-nanodegree--nd229)\n- [Deep Neural Networks with PyTorch from Coursera](https:\/\/www.coursera.org\/learn\/deep-neural-networks-with-pytorch)\n- [PyTorch Twitter](https:\/\/twitter.com\/PyTorch)\n- [PyTorch Blog](https:\/\/pytorch.org\/blog\/)\n- [PyTorch YouTube](https:\/\/www.youtube.com\/channel\/UCWXI5YeOsh03QvJ59PMaXFw)\n## Communication\n- Forums: Discuss implementations, research, etc. [https:\/\/discuss.pytorch.org](https:\/\/discuss.pytorch.org\/)\n- GitHub Issues: Bug reports, feature requests, install issues, RFCs, thoughts, etc.\n- Slack: The [PyTorch Slack](https:\/\/pytorch.slack.com\/) hosts a primary audience of moderate to experienced PyTorch users and developers for general chat, online discussions, collaboration, etc. If you are a beginner looking for help, the primary medium is [PyTorch Forums](https:\/\/discuss.pytorch.org\/). If you need a slack invite, please fill this form: <https:\/\/goo.gl\/forms\/PP1AGvNHpSaJP8to1>\n- Newsletter: No-noise, a one-way email newsletter with important announcements about PyTorch. You can sign-up here: <https:\/\/eepurl.com\/cbG0rv>\n- Facebook Page: Important announcements about PyTorch. <https:\/\/www.facebook.com\/pytorch>\n- For brand guidelines, please visit our website at [pytorch.org](https:\/\/pytorch.org\/)\n## Releases and Contributing\nTypically, PyTorch has three minor releases a year. Please let us know if you encounter a bug by [filing an issue](https:\/\/github.com\/pytorch\/pytorch\/issues).\n\nWe appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion.\n\nIf you plan to contribute new features, utility functions, or extensions to the core, please first open an issue and discuss the feature with us. Sending a PR without discussion might end up resulting in a rejected PR because we might be taking the core in a different direction than you might be aware of.\n\nTo learn more about making a contribution to PyTorch, please see our [Contribution page](https:\/\/pypi.org\/project\/torch\/CONTRIBUTING.md). For more information about PyTorch releases, see [Release page](https:\/\/pypi.org\/project\/torch\/RELEASE.md).\n\n## The Team\nPyTorch is a community-driven project with several skillful engineers and researchers contributing to it.\n\nPyTorch is currently maintained by [Soumith Chintala](http:\/\/soumith.ch\/), [Gregory Chanan](https:\/\/github.com\/gchanan), [Dmytro Dzhulgakov](https:\/\/github.com\/dzhulgakov), [Edward Yang](https:\/\/github.com\/ezyang), [Alban Desmaison](https:\/\/github.com\/albanD), [Piotr Bialecki](https:\/\/github.com\/ptrblck) and [Nikita Shulga](https:\/\/github.com\/malfet) with major contributions coming from hundreds of talented individuals in various forms and means. A non-exhaustive but growing list needs to mention: [Trevor Killeen](https:\/\/github.com\/killeent), [Sasank Chilamkurthy](https:\/\/github.com\/chsasank), [Sergey Zagoruyko](https:\/\/github.com\/szagoruyko), [Adam Lerer](https:\/\/github.com\/adamlerer), [Francisco Massa](https:\/\/github.com\/fmassa), [Alykhan Tejani](https:\/\/github.com\/alykhantejani), [Luca Antiga](https:\/\/github.com\/lantiga), [Alban Desmaison](https:\/\/github.com\/albanD), [Andreas Koepf](https:\/\/github.com\/andreaskoepf), [James Bradbury](https:\/\/github.com\/jekbradbury), [Zeming Lin](https:\/\/github.com\/ebetica), [Yuandong Tian](https:\/\/github.com\/yuandong-tian), [Guillaume Lample](https:\/\/github.com\/glample), [Marat Dukhan](https:\/\/github.com\/Maratyszcza), [Natalia Gimelshein](https:\/\/github.com\/ngimel), [Christian Sarofeen](https:\/\/github.com\/csarofeen), [Martin Raison](https:\/\/github.com\/martinraison), [Edward Yang](https:\/\/github.com\/ezyang), [Zachary Devito](https:\/\/github.com\/zdevito).\n\nNote: This project is unrelated to [hughperkins\/pytorch](https:\/\/github.com\/hughperkins\/pytorch) with the same name. Hugh is a valuable contributor to the Torch community and has helped with many things Torch and PyTorch.\n\n## License\nPyTorch has a BSD-style license, as found in the [LICENSE](https:\/\/pypi.org\/project\/torch\/LICENSE) file.","meta_canonical":null}

3. Robots.txt Check

Query:

Response:

4. Spam/Ban Check

Query:

Response:

5. Seen Status Check

ℹ️ Skipped - page is already crawled

📄

INDEXABLE

✅

CRAWLED

5 days ago

🤖

ROBOTS ALLOWED

Page Info Filters

Filter	Status	Condition	Details
HTTP status	PASS	`download_http_code = 200`	HTTP 200
Age cutoff	PASS	`download_stamp > now() - 6 MONTH`	0.2 months ago
History drop	PASS	`isNull(history_drop_reason)`	No drop reason
Spam/ban	PASS	`fh_dont_index != 1 AND ml_spam_score = 0`	ml_spam_score=0
Canonical	PASS	`meta_canonical IS NULL OR = '' OR = src_unparsed`	Not set

Page Details

Property	Value
URL	https://pypi.org/project/torch/
Last Crawled	2026-04-01 09:54:37 (5 days ago)
First Indexed	2017-01-23 17:29:21 (9 years ago)
HTTP Status Code	200
Meta Title	torch · PyPI
Meta Description	Tensors and Dynamic neural networks in Python with strong GPU acceleration
Meta Canonical	null
Boilerpipe Text	PyTorch is a Python package that provides two high-level features: Tensor computation (like NumPy) with strong GPU acceleration Deep neural networks built on a tape-based autograd system You can reuse your favorite Python packages such as NumPy, SciPy, and Cython to extend PyTorch when needed. Our trunk health (Continuous Integration signals) can be found at hud.pytorch.org . More About PyTorch A GPU-Ready Tensor Library Dynamic Neural Networks: Tape-Based Autograd Python First Imperative Experiences Fast and Lean Extensions Without Pain Installation Binaries NVIDIA Jetson Platforms From Source Prerequisites NVIDIA CUDA Support AMD ROCm Support Intel GPU Support Get the PyTorch Source Install Dependencies Install PyTorch Adjust Build Options (Optional) Docker Image Using pre-built images Building the image yourself Building the Documentation Building a PDF Previous Versions Getting Started Resources Communication Releases and Contributing The Team License More About PyTorch Learn the basics of PyTorch At a granular level, PyTorch is a library that consists of the following components: Component Description torch A Tensor library like NumPy, with strong GPU support torch.autograd A tape-based automatic differentiation library that supports all differentiable Tensor operations in torch torch.jit A compilation stack (TorchScript) to create serializable and optimizable models from PyTorch code torch.nn A neural networks library deeply integrated with autograd designed for maximum flexibility torch.multiprocessing Python multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and Hogwild training torch.utils DataLoader and other utility functions for convenience Usually, PyTorch is used either as: A replacement for NumPy to use the power of GPUs. A deep learning research platform that provides maximum flexibility and speed. Elaborating Further: A GPU-Ready Tensor Library If you use NumPy, then you have used Tensors (a.k.a. ndarray). PyTorch provides Tensors that can live either on the CPU or the GPU and accelerates the computation by a huge amount. We provide a wide variety of tensor routines to accelerate and fit your scientific computation needs such as slicing, indexing, mathematical operations, linear algebra, reductions. And they are fast! Dynamic Neural Networks: Tape-Based Autograd PyTorch has a unique way of building neural networks: using and replaying a tape recorder. Most frameworks such as TensorFlow, Theano, Caffe, and CNTK have a static view of the world. One has to build a neural network and reuse the same structure again and again. Changing the way the network behaves means that one has to start from scratch. With PyTorch, we use a technique called reverse-mode auto-differentiation, which allows you to change the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes from several research papers on this topic, as well as current and past work such as torch-autograd , autograd , Chainer , etc. While this technique is not unique to PyTorch, it's one of the fastest implementations of it to date. You get the best of speed and flexibility for your crazy research. Python First PyTorch is not a Python binding into a monolithic C++ framework. It is built to be deeply integrated into Python. You can use it naturally like you would use NumPy / SciPy / scikit-learn etc. You can write your new neural network layers in Python itself, using your favorite libraries and use packages such as Cython and Numba . Our goal is to not reinvent the wheel where appropriate. Imperative Experiences PyTorch is designed to be intuitive, linear in thought, and easy to use. When you execute a line of code, it gets executed. There isn't an asynchronous view of the world. When you drop into a debugger or receive error messages and stack traces, understanding them is straightforward. The stack trace points to exactly where your code was defined. We hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines. Fast and Lean PyTorch has minimal framework overhead. We integrate acceleration libraries such as Intel MKL and NVIDIA ( cuDNN , NCCL ) to maximize speed. At the core, its CPU and GPU Tensor and neural network backends are mature and have been tested for years. Hence, PyTorch is quite fast — whether you run small or large neural networks. The memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives. We've written custom memory allocators for the GPU to make sure that your deep learning models are maximally memory efficient. This enables you to train bigger deep learning models than before. Extensions Without Pain Writing new neural network modules, or interfacing with PyTorch's Tensor API was designed to be straightforward and with minimal abstractions. You can write new neural network layers in Python using the torch API or your favorite NumPy-based libraries such as SciPy . If you want to write your layers in C/C++, we provide a convenient extension API that is efficient and with minimal boilerplate. No wrapper code needs to be written. You can see a tutorial here and an example here . Installation Binaries Commands to install binaries via Conda or pip wheels are on our website: https://pytorch.org/get-started/locally/ NVIDIA Jetson Platforms Python wheels for NVIDIA's Jetson Nano, Jetson TX1/TX2, Jetson Xavier NX/AGX, and Jetson AGX Orin are provided here and the L4T container is published here They require JetPack 4.2 and above, and @dusty-nv and @ptrblck are maintaining them. From Source Prerequisites If you are installing from source, you will need: Python 3.10 or later A compiler that fully supports C++17, such as clang or gcc (gcc 9.4.0 or newer is required, on Linux) Visual Studio or Visual Studio Build Tool (Windows only) * PyTorch CI uses Visual C++ BuildTools, which come with Visual Studio Enterprise, Professional, or Community Editions. You can also install the build tools from https://visualstudio.microsoft.com/visual-cpp-build-tools/ . The build tools do not come with Visual Studio Code by default. An example of environment setup is shown below: Linux: $ source <CONDA_INSTALL_DIR>/bin/activate $ conda create -y -n <CONDA_NAME> $ conda activate <CONDA_NAME> Windows: $ source <CONDA_INSTALL_DIR> \S cripts \a ctivate.bat $ conda create -y -n <CONDA_NAME> $ conda activate <CONDA_NAME> $ call "C:\Program Files\Microsoft Visual Studio\<VERSION>\Community\VC\Auxiliary\Build\vcvarsall.bat" x64 A conda environment is not required. You can also do a PyTorch build in a standard virtual environment, e.g., created with tools like uv , provided your system has installed all the necessary dependencies unavailable as pip packages (e.g., CUDA, MKL.) NVIDIA CUDA Support If you want to compile with CUDA support, select a supported version of CUDA from our support matrix , then install the following: NVIDIA CUDA NVIDIA cuDNN v8.5 or above Compiler compatible with CUDA Note: You could refer to the cuDNN Support Matrix for cuDNN versions with the various supported CUDA, CUDA driver, and NVIDIA hardware. If you want to disable CUDA support, export the environment variable USE_CUDA=0 . Other potentially useful environment variables may be found in setup.py . If CUDA is installed in a non-standard location, set PATH so that the nvcc you want to use can be found (e.g., export PATH=/usr/local/cuda-12.8/bin:$PATH ). If you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are available here AMD ROCm Support If you want to compile with ROCm support, install AMD ROCm 4.0 and above installation ROCm is currently supported only for Linux systems. By default the build system expects ROCm to be installed in /opt/rocm . If ROCm is installed in a different directory, the ROCM_PATH environment variable must be set to the ROCm installation directory. The build system automatically detects the AMD GPU architecture. Optionally, the AMD GPU architecture can be explicitly set with the PYTORCH_ROCM_ARCH environment variable AMD GPU architecture If you want to disable ROCm support, export the environment variable USE_ROCM=0 . Other potentially useful environment variables may be found in setup.py . Intel GPU Support If you want to compile with Intel GPU support, follow these PyTorch Prerequisites for Intel GPUs instructions. Intel GPU is supported for Linux and Windows. If you want to disable Intel GPU support, export the environment variable USE_XPU=0 . Other potentially useful environment variables may be found in setup.py . Get the PyTorch Source git clone https://github.com/pytorch/pytorch cd pytorch # if you are updating an existing checkout git submodule sync git submodule update --init --recursive Install Dependencies Common # Run this command from the PyTorch directory after cloning the source code using the “Get the PyTorch Source“ section above pip install --group dev On Linux pip install mkl-static mkl-include # CUDA only: Add LAPACK support for the GPU if needed # magma installation: run with active conda environment. specify CUDA version to install .ci/docker/common/install_magma_conda.sh 12 .4 # (optional) If using torch.compile with inductor/triton, install the matching version of triton # Run from the pytorch directory after cloning # For Intel GPU support, please explicitly `export USE_XPU=1` before running command. make triton On MacOS # Add this package on intel x86 processor machines only pip install mkl-static mkl-include # Add these packages if torch.distributed is needed conda install pkg-config libuv On Windows pip install mkl-static mkl-include # Add these packages if torch.distributed is needed. # Distributed package support on Windows is a prototype feature and is subject to changes. conda install -c conda-forge libuv = 1 .51 Install PyTorch On Linux If you're compiling for AMD ROCm then first run this command: # Only run this if you're compiling for ROCm python tools/amd_build/build_amd.py Install PyTorch # the CMake prefix for conda environment export CMAKE_PREFIX_PATH = " ${ CONDA_PREFIX :- '$(dirname $(which conda))/../' } : ${ CMAKE_PREFIX_PATH } " python -m pip install --no-build-isolation -v -e . # the CMake prefix for non-conda environment, e.g. Python venv # call following after activating the venv export CMAKE_PREFIX_PATH = " ${ VIRTUAL_ENV } : ${ CMAKE_PREFIX_PATH } " On macOS python -m pip install --no-build-isolation -v -e . On Windows If you want to build legacy python code, please refer to Building on legacy code and CUDA CPU-only builds In this mode PyTorch computations will run on your CPU, not your GPU. python -m pip install --no-build-isolation -v -e . Note on OpenMP: The desired OpenMP implementation is Intel OpenMP (iomp). In order to link against iomp, you'll need to manually download the library and set up the building environment by tweaking CMAKE_INCLUDE_PATH and LIB . The instruction here is an example for setting up both MKL and Intel OpenMP. Without these configurations for CMake, Microsoft Visual C OpenMP runtime (vcomp) will be used. CUDA based build In this mode PyTorch computations will leverage your GPU via CUDA for faster number crunching NVTX is needed to build PyTorch with CUDA. NVTX is a part of CUDA distributive, where it is called "Nsight Compute". To install it onto an already installed CUDA run CUDA installation once again and check the corresponding checkbox. Make sure that CUDA with Nsight Compute is installed after Visual Studio. Currently, VS 2017 / 2019, and Ninja are supported as the generator of CMake. If ninja.exe is detected in PATH , then Ninja will be used as the default generator, otherwise, it will use VS 2017 / 2019. If Ninja is selected as the generator, the latest MSVC will get selected as the underlying toolchain. Additional libraries such as Magma , oneDNN, a.k.a. MKLDNN or DNNL , and Sccache are often needed. Please refer to the installation-helper to install them. You can refer to the build_pytorch.bat script for some other environment variables configurations cmd :: Set the environment variables after you have downloaded and unzipped the mkl package, :: else CMake would throw an error as `Could NOT find OpenMP`. set CMAKE_INCLUDE_PATH={Your directory}\mkl\include set LIB={Your directory}\mkl\lib;%LIB% :: Read the content in the previous section carefully before you proceed. :: [Optional] If you want to override the underlying toolset used by Ninja and Visual Studio with CUDA, please run the following script block. :: "Visual Studio 2019 Developer Command Prompt" will be run automatically. :: Make sure you have CMake >= 3.12 before you do this when you use the Visual Studio generator. set CMAKE_GENERATOR_TOOLSET_VERSION=14.27 set DISTUTILS_USE_SDK=1 for /f "usebackq tokens=" %i in (`"%ProgramFiles(x86)%\Microsoft Visual Studio\Installer\vswhere.exe" -version [15^,17^) -products -latest -property installationPath`) do call "%i\VC\Auxiliary\Build\vcvarsall.bat" x64 -vcvars_ver=%CMAKE_GENERATOR_TOOLSET_VERSION% :: [Optional] If you want to override the CUDA host compiler set CUDAHOSTCXX=C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\HostX64\x64\cl.exe python -m pip install --no-build-isolation -v -e . Intel GPU builds In this mode PyTorch with Intel GPU support will be built. Please make sure the common prerequisites as well as the prerequisites for Intel GPU are properly installed and the environment variables are configured prior to starting the build. For build tool support, Visual Studio 2022 is required. Then PyTorch can be built with the command: :: CMD Commands: :: Set the CMAKE_PREFIX_PATH to help find corresponding packages :: %CONDA_PREFIX% only works after `conda activate custom_env` if defined CMAKE_PREFIX_PATH ( set "CMAKE_PREFIX_PATH=%CONDA_PREFIX%\Library;%CMAKE_PREFIX_PATH%" ) else ( set "CMAKE_PREFIX_PATH=%CONDA_PREFIX%\Library" ) python -m pip install --no-build-isolation -v -e . Adjust Build Options (Optional) You can adjust the configuration of cmake variables optionally (without building first), by doing the following. For example, adjusting the pre-detected directories for CuDNN or BLAS can be done with such a step. On Linux export CMAKE_PREFIX_PATH = " ${ CONDA_PREFIX :- '$(dirname $(which conda))/../' } : ${ CMAKE_PREFIX_PATH } " CMAKE_ONLY = 1 python setup.py build ccmake build # or cmake-gui build On macOS export CMAKE_PREFIX_PATH = " ${ CONDA_PREFIX :- '$(dirname $(which conda))/../' } : ${ CMAKE_PREFIX_PATH } " MACOSX_DEPLOYMENT_TARGET = 11 .0 CMAKE_ONLY = 1 python setup.py build ccmake build # or cmake-gui build Docker Image Using pre-built images You can also pull a pre-built docker image from Docker Hub and run with docker v19.03+ docker run --gpus all --rm -ti --ipc = host pytorch/pytorch:latest Please note that PyTorch uses shared memory to share data between processes, so if torch multiprocessing is used (e.g. for multithreaded data loaders) the default shared memory segment size that container runs with is not enough, and you should increase shared memory size either with --ipc=host or --shm-size command line options to nvidia-docker run . Building the image yourself NOTE: Must be built with a docker version > 18.06 The Dockerfile is supplied to build images with CUDA 11.1 support and cuDNN v8. You can pass PYTHON_VERSION=x.y make variable to specify which Python version is to be used by Miniconda, or leave it unset to use the default. make -f docker.Makefile # images are tagged as docker.io/${your_docker_username}/pytorch You can also pass the CMAKE_VARS="..." environment variable to specify additional CMake variables to be passed to CMake during the build. See setup.py for the list of available variables. make -f docker.Makefile Building the Documentation To build documentation in various formats, you will need Sphinx and the pytorch_sphinx_theme2. Before you build the documentation locally, ensure torch is installed in your environment. For small fixes, you can install the nightly version as described in Getting Started . For more complex fixes, such as adding a new module and docstrings for the new module, you might need to install torch from source . See Docstring Guidelines for docstring conventions. cd docs/ pip install -r requirements.txt make html make serve Run make to get a list of all available output formats. If you get a katex error run npm install katex . If it persists, try npm install -g katex [!NOTE] If you installed nodejs with a different package manager (e.g., conda ) then npm will probably install a version of katex that is not compatible with your version of nodejs and doc builds will fail. A combination of versions that is known to work is node@6.13.1 and katex@0.13.18 . To install the latter with npm you can run npm install -g katex@0.13.18 [!NOTE] If you see a numpy incompatibility error, run: pip install 'numpy<2' When you make changes to the dependencies run by CI, edit the .ci/docker/requirements-docs.txt file. Building a PDF To compile a PDF of all PyTorch documentation, ensure you have texlive and LaTeX installed. On macOS, you can install them using: brew install --cask mactex To create the PDF: Run: make latexpdf This will generate the necessary files in the build/latex directory. Navigate to this directory and execute: make LATEXOPTS="-interaction=nonstopmode" This will produce a pytorch.pdf with the desired content. Run this command one more time so that it generates the correct table of contents and index. [!NOTE] To view the Table of Contents, switch to the Table of Contents view in your PDF viewer. Previous Versions Installation instructions and binaries for previous PyTorch versions may be found on our website . Getting Started Three pointers to get you started: Tutorials: get you started with understanding and using PyTorch Examples: easy to understand PyTorch code across all domains The API Reference Glossary Resources PyTorch.org PyTorch Tutorials PyTorch Examples PyTorch Models Intro to Deep Learning with PyTorch from Udacity Intro to Machine Learning with PyTorch from Udacity Deep Neural Networks with PyTorch from Coursera PyTorch Twitter PyTorch Blog PyTorch YouTube Communication Forums: Discuss implementations, research, etc. https://discuss.pytorch.org GitHub Issues: Bug reports, feature requests, install issues, RFCs, thoughts, etc. Slack: The PyTorch Slack hosts a primary audience of moderate to experienced PyTorch users and developers for general chat, online discussions, collaboration, etc. If you are a beginner looking for help, the primary medium is PyTorch Forums . If you need a slack invite, please fill this form: https://goo.gl/forms/PP1AGvNHpSaJP8to1 Newsletter: No-noise, a one-way email newsletter with important announcements about PyTorch. You can sign-up here: https://eepurl.com/cbG0rv Facebook Page: Important announcements about PyTorch. https://www.facebook.com/pytorch For brand guidelines, please visit our website at pytorch.org Releases and Contributing Typically, PyTorch has three minor releases a year. Please let us know if you encounter a bug by filing an issue . We appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion. If you plan to contribute new features, utility functions, or extensions to the core, please first open an issue and discuss the feature with us. Sending a PR without discussion might end up resulting in a rejected PR because we might be taking the core in a different direction than you might be aware of. To learn more about making a contribution to PyTorch, please see our Contribution page . For more information about PyTorch releases, see Release page . The Team PyTorch is a community-driven project with several skillful engineers and researchers contributing to it. PyTorch is currently maintained by Soumith Chintala , Gregory Chanan , Dmytro Dzhulgakov , Edward Yang , Alban Desmaison , Piotr Bialecki and Nikita Shulga with major contributions coming from hundreds of talented individuals in various forms and means. A non-exhaustive but growing list needs to mention: Trevor Killeen , Sasank Chilamkurthy , Sergey Zagoruyko , Adam Lerer , Francisco Massa , Alykhan Tejani , Luca Antiga , Alban Desmaison , Andreas Koepf , James Bradbury , Zeming Lin , Yuandong Tian , Guillaume Lample , Marat Dukhan , Natalia Gimelshein , Christian Sarofeen , Martin Raison , Edward Yang , Zachary Devito . Note: This project is unrelated to hughperkins/pytorch with the same name. Hugh is a valuable contributor to the Torch community and has helped with many things Torch and PyTorch. License PyTorch has a BSD-style license, as found in the LICENSE file.
Markdown	[Skip to main content](https://pypi.org/project/torch/#content) Switch to mobile version Warning Some features may not work without JavaScript. Please try enabling it if you encounter problems. [![PyPI](https://pypi.org/static/images/logo-small.8998e9d1.svg)](https://pypi.org/) - [Help](https://pypi.org/help/) - [Docs](https://docs.pypi.org/) - [Sponsors](https://pypi.org/sponsors/) - [Log in](https://pypi.org/account/login/?next=https%3A%2F%2Fpypi.org%2Fproject%2Ftorch%2F) - [Register](https://pypi.org/account/register/) Menu - [Help](https://pypi.org/help/) - [Docs](https://docs.pypi.org/) - [Sponsors](https://pypi.org/sponsors/) - [Log in](https://pypi.org/account/login/?next=https%3A%2F%2Fpypi.org%2Fproject%2Ftorch%2F) - [Register](https://pypi.org/account/register/) # torch 2.11.0 pip install torch Copy PIP instructions [Latest version](https://pypi.org/project/torch/) Released: Mar 23, 2026 Tensors and Dynamic neural networks in Python with strong GPU acceleration ### Navigation - [Project description](https://pypi.org/project/torch/#description) - [Release history](https://pypi.org/project/torch/#history) - [Download files](https://pypi.org/project/torch/#files) ### Verified details These details have been [verified by PyPI](https://docs.pypi.org/project_metadata/#verified-details) ###### Owner - [Meta Platforms](https://pypi.org/org/meta-platforms/) ###### Maintainers [![Avatar for atalman from gravatar.com](https://pypi-camo.freetls.fastly.net/4a1e9de14640155ddeb5938095a666ac957155a3/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f34363564656263366230376638366139613236613461323633346535396166383f73697a653d3530) atalman](https://pypi.org/user/atalman/) [![Avatar for malfet from gravatar.com](https://pypi-camo.freetls.fastly.net/d8937f01c4383b85866102a5e806397f40b3a557/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f35336331663335393833326462633337366337613732626365323539616536643f73697a653d3530) malfet](https://pypi.org/user/malfet/) [![Avatar for seemethere from gravatar.com](https://pypi-camo.freetls.fastly.net/0bde43daede4ad73d5e20183d999d5a5aeb0c0e0/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f62343230313862343337386564393666393330356133663334386462326435623f73697a653d3530) seemethere](https://pypi.org/user/seemethere/) [![Avatar for soumith from gravatar.com](https://pypi-camo.freetls.fastly.net/73d38504f107197034b9b146995199672719c24e/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f65386363373633626131316639323936353765323863623033373162373563383f73697a653d3530) soumith](https://pypi.org/user/soumith/) ### Unverified details These details have not* been verified by PyPI* ###### Project links - [Documentation](https://pytorch.org/docs) - [Forum](https://discuss.pytorch.org/) - [Homepage](https://pytorch.org/) - [Issue Tracker](https://github.com/pytorch/pytorch/issues) - [Repository](https://github.com/pytorch/pytorch) ###### Meta - License: BSD-3-Clause - Author: [PyTorch Team](mailto:packages@pytorch.org) - Tags pytorch , machine learning - Requires: Python \>=3.10 - Provides-Extra: `optree` , `opt-einsum` , `pyyaml` ###### Classifiers - Development Status - [5 - Production/Stable](https://pypi.org/search/?c=Development+Status+%3A%3A+5+-+Production%2FStable) - Intended Audience - [Developers](https://pypi.org/search/?c=Intended+Audience+%3A%3A+Developers) - [Education](https://pypi.org/search/?c=Intended+Audience+%3A%3A+Education) - [Science/Research](https://pypi.org/search/?c=Intended+Audience+%3A%3A+Science%2FResearch) - Programming Language - [C++](https://pypi.org/search/?c=Programming+Language+%3A%3A+C%2B%2B) - [Python :: 3 :: Only](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3+%3A%3A+Only) - [Python :: 3.10](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.10) - [Python :: 3.11](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.11) - [Python :: 3.12](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.12) - [Python :: 3.13](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.13) - [Python :: 3.14](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.14) - Topic - [Scientific/Engineering](https://pypi.org/search/?c=Topic+%3A%3A+Scientific%2FEngineering) - [Scientific/Engineering :: Artificial Intelligence](https://pypi.org/search/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Artificial+Intelligence) - [Scientific/Engineering :: Mathematics](https://pypi.org/search/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Mathematics) - [Software Development](https://pypi.org/search/?c=Topic+%3A%3A+Software+Development) - [Software Development :: Libraries](https://pypi.org/search/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries) - [Software Development :: Libraries :: Python Modules](https://pypi.org/search/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries+%3A%3A+Python+Modules) [Report project as malware](https://pypi.org/project/torch/submit-malware-report/) - [Project description](https://pypi.org/project/torch/#description) - [Project details](https://pypi.org/project/torch/#data) - [Release history](https://pypi.org/project/torch/#history) - [Download files](https://pypi.org/project/torch/#files) ## Project description ![PyTorch Logo](https://pypi-camo.freetls.fastly.net/ed3839e1c11e779b508097969affa63d0968692b/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f7079746f7263682d6c6f676f2d6461726b2e706e67) * PyTorch is a Python package that provides two high-level features: - Tensor computation (like NumPy) with strong GPU acceleration - Deep neural networks built on a tape-based autograd system You can reuse your favorite Python packages such as NumPy, SciPy, and Cython to extend PyTorch when needed. Our trunk health (Continuous Integration signals) can be found at [hud.pytorch.org](https://hud.pytorch.org/ci/pytorch/pytorch/main). - [More About PyTorch](https://pypi.org/project/torch/#more-about-pytorch) - [A GPU-Ready Tensor Library](https://pypi.org/project/torch/#a-gpu-ready-tensor-library) - [Dynamic Neural Networks: Tape-Based Autograd](https://pypi.org/project/torch/#dynamic-neural-networks-tape-based-autograd) - [Python First](https://pypi.org/project/torch/#python-first) - [Imperative Experiences](https://pypi.org/project/torch/#imperative-experiences) - [Fast and Lean](https://pypi.org/project/torch/#fast-and-lean) - [Extensions Without Pain](https://pypi.org/project/torch/#extensions-without-pain) - [Installation](https://pypi.org/project/torch/#installation) - [Binaries](https://pypi.org/project/torch/#binaries) - [NVIDIA Jetson Platforms](https://pypi.org/project/torch/#nvidia-jetson-platforms) - [From Source](https://pypi.org/project/torch/#from-source) - [Prerequisites](https://pypi.org/project/torch/#prerequisites) - [NVIDIA CUDA Support](https://pypi.org/project/torch/#nvidia-cuda-support) - [AMD ROCm Support](https://pypi.org/project/torch/#amd-rocm-support) - [Intel GPU Support](https://pypi.org/project/torch/#intel-gpu-support) - [Get the PyTorch Source](https://pypi.org/project/torch/#get-the-pytorch-source) - [Install Dependencies](https://pypi.org/project/torch/#install-dependencies) - [Install PyTorch](https://pypi.org/project/torch/#install-pytorch) - [Adjust Build Options (Optional)](https://pypi.org/project/torch/#adjust-build-options-optional) - [Docker Image](https://pypi.org/project/torch/#docker-image) - [Using pre-built images](https://pypi.org/project/torch/#using-pre-built-images) - [Building the image yourself](https://pypi.org/project/torch/#building-the-image-yourself) - [Building the Documentation](https://pypi.org/project/torch/#building-the-documentation) - [Building a PDF](https://pypi.org/project/torch/#building-a-pdf) - [Previous Versions](https://pypi.org/project/torch/#previous-versions) - [Getting Started](https://pypi.org/project/torch/#getting-started) - [Resources](https://pypi.org/project/torch/#resources) - [Communication](https://pypi.org/project/torch/#communication) - [Releases and Contributing](https://pypi.org/project/torch/#releases-and-contributing) - [The Team](https://pypi.org/project/torch/#the-team) - [License](https://pypi.org/project/torch/#license) ## More About PyTorch [Learn the basics of PyTorch](https://pytorch.org/tutorials/beginner/basics/intro.html) At a granular level, PyTorch is a library that consists of the following components: \| Component \| Description \| \|---\|---\| \| [torch](https://pytorch.org/docs/stable/torch.html) \| A Tensor library like NumPy, with strong GPU support \| \| [torch.autograd](https://pytorch.org/docs/stable/autograd.html) \| A tape-based automatic differentiation library that supports all differentiable Tensor operations in torch \| \| [torch.jit](https://pytorch.org/docs/stable/jit.html) \| A compilation stack (TorchScript) to create serializable and optimizable models from PyTorch code \| \| [torch.nn](https://pytorch.org/docs/stable/nn.html) \| A neural networks library deeply integrated with autograd designed for maximum flexibility \| \| [torch.multiprocessing](https://pytorch.org/docs/stable/multiprocessing.html) \| Python multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and Hogwild training \| \| [torch.utils*](https://pytorch.org/docs/stable/data.html) \| DataLoader and other utility functions for convenience \| Usually, PyTorch is used either as: - A replacement for NumPy to use the power of GPUs. - A deep learning research platform that provides maximum flexibility and speed. Elaborating Further: ### A GPU-Ready Tensor Library If you use NumPy, then you have used Tensors (a.k.a. ndarray). ![Tensor illustration](https://pypi-camo.freetls.fastly.net/e5f621cdea28d4de40577ac3346794f746d96b86/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f74656e736f725f696c6c757374726174696f6e2e706e67) PyTorch provides Tensors that can live either on the CPU or the GPU and accelerates the computation by a huge amount. We provide a wide variety of tensor routines to accelerate and fit your scientific computation needs such as slicing, indexing, mathematical operations, linear algebra, reductions. And they are fast\! ### Dynamic Neural Networks: Tape-Based Autograd PyTorch has a unique way of building neural networks: using and replaying a tape recorder. Most frameworks such as TensorFlow, Theano, Caffe, and CNTK have a static view of the world. One has to build a neural network and reuse the same structure again and again. Changing the way the network behaves means that one has to start from scratch. With PyTorch, we use a technique called reverse-mode auto-differentiation, which allows you to change the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes from several research papers on this topic, as well as current and past work such as [torch-autograd](https://github.com/twitter/torch-autograd), [autograd](https://github.com/HIPS/autograd), [Chainer](https://chainer.org/), etc. While this technique is not unique to PyTorch, it's one of the fastest implementations of it to date. You get the best of speed and flexibility for your crazy research. ![Dynamic graph](https://pypi-camo.freetls.fastly.net/360dc3e00cca046f73204c87f117feefe993580f/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f64796e616d69635f67726170682e676966) ### Python First PyTorch is not a Python binding into a monolithic C++ framework. It is built to be deeply integrated into Python. You can use it naturally like you would use [NumPy](https://www.numpy.org/) / [SciPy](https://www.scipy.org/) / [scikit-learn](https://scikit-learn.org/) etc. You can write your new neural network layers in Python itself, using your favorite libraries and use packages such as [Cython](https://cython.org/) and [Numba](http://numba.pydata.org/). Our goal is to not reinvent the wheel where appropriate. ### Imperative Experiences PyTorch is designed to be intuitive, linear in thought, and easy to use. When you execute a line of code, it gets executed. There isn't an asynchronous view of the world. When you drop into a debugger or receive error messages and stack traces, understanding them is straightforward. The stack trace points to exactly where your code was defined. We hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines. ### Fast and Lean PyTorch has minimal framework overhead. We integrate acceleration libraries such as [Intel MKL](https://software.intel.com/mkl) and NVIDIA ([cuDNN](https://developer.nvidia.com/cudnn), [NCCL](https://developer.nvidia.com/nccl)) to maximize speed. At the core, its CPU and GPU Tensor and neural network backends are mature and have been tested for years. Hence, PyTorch is quite fast — whether you run small or large neural networks. The memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives. We've written custom memory allocators for the GPU to make sure that your deep learning models are maximally memory efficient. This enables you to train bigger deep learning models than before. ### Extensions Without Pain Writing new neural network modules, or interfacing with PyTorch's Tensor API was designed to be straightforward and with minimal abstractions. You can write new neural network layers in Python using the torch API [or your favorite NumPy-based libraries such as SciPy](https://pytorch.org/tutorials/advanced/numpy_extensions_tutorial.html). If you want to write your layers in C/C++, we provide a convenient extension API that is efficient and with minimal boilerplate. No wrapper code needs to be written. You can see [a tutorial here](https://pytorch.org/tutorials/advanced/cpp_extension.html) and [an example here](https://github.com/pytorch/extension-cpp). ## Installation ### Binaries Commands to install binaries via Conda or pip wheels are on our website: <https://pytorch.org/get-started/locally/> #### NVIDIA Jetson Platforms Python wheels for NVIDIA's Jetson Nano, Jetson TX1/TX2, Jetson Xavier NX/AGX, and Jetson AGX Orin are provided [here](https://forums.developer.nvidia.com/t/pytorch-for-jetson-version-1-10-now-available/72048) and the L4T container is published [here](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/l4t-pytorch) They require JetPack 4.2 and above, and [@dusty-nv](https://github.com/dusty-nv) and [@ptrblck](https://github.com/ptrblck) are maintaining them. ### From Source #### Prerequisites If you are installing from source, you will need: - Python 3.10 or later - A compiler that fully supports C++17, such as clang or gcc (gcc 9.4.0 or newer is required, on Linux) - Visual Studio or Visual Studio Build Tool (Windows only) \ PyTorch CI uses Visual C++ BuildTools, which come with Visual Studio Enterprise, Professional, or Community Editions. You can also install the build tools from <https://visualstudio.microsoft.com/visual-cpp-build-tools/>. The build tools do not come with Visual Studio Code by default. An example of environment setup is shown below: - Linux: ``` $ source <CONDA_INSTALL_DIR>/bin/activate $ conda create -y -n <CONDA_NAME> $ conda activate <CONDA_NAME> ``` - Windows: ``` $ source <CONDA_INSTALL_DIR>\Scripts\activate.bat $ conda create -y -n <CONDA_NAME> $ conda activate <CONDA_NAME> $ call "C:\Program Files\Microsoft Visual Studio\<VERSION>\Community\VC\Auxiliary\Build\vcvarsall.bat" x64 ``` A conda environment is not required. You can also do a PyTorch build in a standard virtual environment, e.g., created with tools like `uv`, provided your system has installed all the necessary dependencies unavailable as pip packages (e.g., CUDA, MKL.) ##### NVIDIA CUDA Support If you want to compile with CUDA support, [select a supported version of CUDA from our support matrix](https://pytorch.org/get-started/locally/), then install the following: - [NVIDIA CUDA](https://developer.nvidia.com/cuda-downloads) - [NVIDIA cuDNN](https://developer.nvidia.com/cudnn) v8.5 or above - [Compiler](https://gist.github.com/ax3l/9489132) compatible with CUDA Note: You could refer to the [cuDNN Support Matrix](https://docs.nvidia.com/deeplearning/cudnn/backend/latest/reference/support-matrix.html) for cuDNN versions with the various supported CUDA, CUDA driver, and NVIDIA hardware. If you want to disable CUDA support, export the environment variable `USE_CUDA=0`. Other potentially useful environment variables may be found in `setup.py`. If CUDA is installed in a non-standard location, set PATH so that the nvcc you want to use can be found (e.g., `export PATH=/usr/local/cuda-12.8/bin:$PATH`). If you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are [available here](https://devtalk.nvidia.com/default/topic/1049071/jetson-nano/pytorch-for-jetson-nano/) ##### AMD ROCm Support If you want to compile with ROCm support, install - [AMD ROCm](https://rocm.docs.amd.com/en/latest/deploy/linux/quick_start.html) 4.0 and above installation - ROCm is currently supported only for Linux systems. By default the build system expects ROCm to be installed in `/opt/rocm`. If ROCm is installed in a different directory, the `ROCM_PATH` environment variable must be set to the ROCm installation directory. The build system automatically detects the AMD GPU architecture. Optionally, the AMD GPU architecture can be explicitly set with the `PYTORCH_ROCM_ARCH` environment variable [AMD GPU architecture](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/system-requirements.html#supported-gpus) If you want to disable ROCm support, export the environment variable `USE_ROCM=0`. Other potentially useful environment variables may be found in `setup.py`. ##### Intel GPU Support If you want to compile with Intel GPU support, follow these - [PyTorch Prerequisites for Intel GPUs](https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpu.html) instructions. - Intel GPU is supported for Linux and Windows. If you want to disable Intel GPU support, export the environment variable `USE_XPU=0`. Other potentially useful environment variables may be found in `setup.py`. #### Get the PyTorch Source ``` git clone https://github.com/pytorch/pytorch cd pytorch # if you are updating an existing checkout git submodule sync git submodule update --init --recursive ``` #### Install Dependencies Common ``` # Run this command from the PyTorch directory after cloning the source code using the “Get the PyTorch Source“ section above pip install --group dev ``` On Linux ``` pip install mkl-static mkl-include # CUDA only: Add LAPACK support for the GPU if needed # magma installation: run with active conda environment. specify CUDA version to install .ci/docker/common/install_magma_conda.sh 12.4 # (optional) If using torch.compile with inductor/triton, install the matching version of triton # Run from the pytorch directory after cloning # For Intel GPU support, please explicitly `export USE_XPU=1` before running command. make triton ``` On MacOS ``` # Add this package on intel x86 processor machines only pip install mkl-static mkl-include # Add these packages if torch.distributed is needed conda install pkg-config libuv ``` On Windows ``` pip install mkl-static mkl-include # Add these packages if torch.distributed is needed. # Distributed package support on Windows is a prototype feature and is subject to changes. conda install -c conda-forge libuv=1.51 ``` #### Install PyTorch On Linux If you're compiling for AMD ROCm then first run this command: ``` # Only run this if you're compiling for ROCm python tools/amd_build/build_amd.py ``` Install PyTorch ``` # the CMake prefix for conda environment export CMAKE_PREFIX_PATH="${CONDA_PREFIX:-'$(dirname $(which conda))/../'}:${CMAKE_PREFIX_PATH}" python -m pip install --no-build-isolation -v -e . # the CMake prefix for non-conda environment, e.g. Python venv # call following after activating the venv export CMAKE_PREFIX_PATH="${VIRTUAL_ENV}:${CMAKE_PREFIX_PATH}" ``` On macOS ``` python -m pip install --no-build-isolation -v -e . ``` On Windows If you want to build legacy python code, please refer to [Building on legacy code and CUDA](https://github.com/pytorch/pytorch/blob/main/CONTRIBUTING.md#building-on-legacy-code-and-cuda) CPU-only builds In this mode PyTorch computations will run on your CPU, not your GPU. ``` python -m pip install --no-build-isolation -v -e . ``` Note on OpenMP: The desired OpenMP implementation is Intel OpenMP (iomp). In order to link against iomp, you'll need to manually download the library and set up the building environment by tweaking `CMAKE_INCLUDE_PATH` and `LIB`. The instruction [here](https://github.com/pytorch/pytorch/blob/main/docs/source/notes/windows.rst#building-from-source) is an example for setting up both MKL and Intel OpenMP. Without these configurations for CMake, Microsoft Visual C OpenMP runtime (vcomp) will be used. CUDA based build In this mode PyTorch computations will leverage your GPU via CUDA for faster number crunching [NVTX](https://docs.nvidia.com/gameworks/content/gameworkslibrary/nvtx/nvidia_tools_extension_library_nvtx.htm) is needed to build PyTorch with CUDA. NVTX is a part of CUDA distributive, where it is called "Nsight Compute". To install it onto an already installed CUDA run CUDA installation once again and check the corresponding checkbox. Make sure that CUDA with Nsight Compute is installed after Visual Studio. Currently, VS 2017 / 2019, and Ninja are supported as the generator of CMake. If `ninja.exe` is detected in `PATH`, then Ninja will be used as the default generator, otherwise, it will use VS 2017 / 2019. If Ninja is selected as the generator, the latest MSVC will get selected as the underlying toolchain. Additional libraries such as [Magma](https://developer.nvidia.com/magma), [oneDNN, a.k.a. MKLDNN or DNNL](https://github.com/oneapi-src/oneDNN), and [Sccache](https://github.com/mozilla/sccache) are often needed. Please refer to the [installation-helper](https://github.com/pytorch/pytorch/tree/main/.ci/pytorch/win-test-helpers/installation-helpers) to install them. You can refer to the [build\_pytorch.bat](https://github.com/pytorch/pytorch/blob/main/.ci/pytorch/win-test-helpers/build_pytorch.bat) script for some other environment variables configurations ``` cmd :: Set the environment variables after you have downloaded and unzipped the mkl package, :: else CMake would throw an error as `Could NOT find OpenMP`. set CMAKE_INCLUDE_PATH={Your directory}\mkl\include set LIB={Your directory}\mkl\lib;%LIB% :: Read the content in the previous section carefully before you proceed. :: [Optional] If you want to override the underlying toolset used by Ninja and Visual Studio with CUDA, please run the following script block. :: "Visual Studio 2019 Developer Command Prompt" will be run automatically. :: Make sure you have CMake >= 3.12 before you do this when you use the Visual Studio generator. set CMAKE_GENERATOR_TOOLSET_VERSION=14.27 set DISTUTILS_USE_SDK=1 for /f "usebackq tokens=" %i in (`"%ProgramFiles(x86)%\Microsoft Visual Studio\Installer\vswhere.exe" -version [15^,17^) -products -latest -property installationPath`) do call "%i\VC\Auxiliary\Build\vcvarsall.bat" x64 -vcvars_ver=%CMAKE_GENERATOR_TOOLSET_VERSION% :: [Optional] If you want to override the CUDA host compiler set CUDAHOSTCXX=C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\HostX64\x64\cl.exe python -m pip install --no-build-isolation -v -e . ``` Intel GPU builds In this mode PyTorch with Intel GPU support will be built. Please make sure [the common prerequisites](https://pypi.org/project/torch/#prerequisites) as well as [the prerequisites for Intel GPU](https://pypi.org/project/torch/#intel-gpu-support) are properly installed and the environment variables are configured prior to starting the build. For build tool support, `Visual Studio 2022` is required. Then PyTorch can be built with the command: ``` :: CMD Commands: :: Set the CMAKE_PREFIX_PATH to help find corresponding packages :: %CONDA_PREFIX% only works after `conda activate custom_env` if defined CMAKE_PREFIX_PATH ( set "CMAKE_PREFIX_PATH=%CONDA_PREFIX%\Library;%CMAKE_PREFIX_PATH%" ) else ( set "CMAKE_PREFIX_PATH=%CONDA_PREFIX%\Library" ) python -m pip install --no-build-isolation -v -e . ``` ##### Adjust Build Options (Optional) You can adjust the configuration of cmake variables optionally (without building first), by doing the following. For example, adjusting the pre-detected directories for CuDNN or BLAS can be done with such a step. On Linux ``` export CMAKE_PREFIX_PATH="${CONDA_PREFIX:-'$(dirname $(which conda))/../'}:${CMAKE_PREFIX_PATH}" CMAKE_ONLY=1 python setup.py build ccmake build # or cmake-gui build ``` On macOS ``` export CMAKE_PREFIX_PATH="${CONDA_PREFIX:-'$(dirname $(which conda))/../'}:${CMAKE_PREFIX_PATH}" MACOSX_DEPLOYMENT_TARGET=11.0 CMAKE_ONLY=1 python setup.py build ccmake build # or cmake-gui build ``` ### Docker Image #### Using pre-built images You can also pull a pre-built docker image from Docker Hub and run with docker v19.03+ ``` docker run --gpus all --rm -ti --ipc=host pytorch/pytorch:latest ``` Please note that PyTorch uses shared memory to share data between processes, so if torch multiprocessing is used (e.g. for multithreaded data loaders) the default shared memory segment size that container runs with is not enough, and you should increase shared memory size either with `--ipc=host` or `--shm-size` command line options to `nvidia-docker run`. #### Building the image yourself NOTE: Must be built with a docker version \> 18.06 The `Dockerfile` is supplied to build images with CUDA 11.1 support and cuDNN v8. You can pass `PYTHON_VERSION=x.y` make variable to specify which Python version is to be used by Miniconda, or leave it unset to use the default. ``` make -f docker.Makefile # images are tagged as docker.io/${your_docker_username}/pytorch ``` You can also pass the `CMAKE_VARS="..."` environment variable to specify additional CMake variables to be passed to CMake during the build. See [setup.py](https://pypi.org/project/torch/setup.py) for the list of available variables. ``` make -f docker.Makefile ``` ### Building the Documentation To build documentation in various formats, you will need [Sphinx](http://www.sphinx-doc.org/) and the pytorch\_sphinx\_theme2. Before you build the documentation locally, ensure `torch` is installed in your environment. For small fixes, you can install the nightly version as described in [Getting Started](https://pytorch.org/get-started/locally/). For more complex fixes, such as adding a new module and docstrings for the new module, you might need to install torch [from source](https://pypi.org/project/torch/#from-source). See [Docstring Guidelines](https://github.com/pytorch/pytorch/wiki/Docstring-Guidelines) for docstring conventions. ``` cd docs/ pip install -r requirements.txt make html make serve ``` Run `make` to get a list of all available output formats. If you get a katex error run `npm install katex`. If it persists, try `npm install -g katex` > \[!NOTE\] If you installed `nodejs` with a different package manager (e.g., `conda`) then `npm` will probably install a version of `katex` that is not compatible with your version of `nodejs` and doc builds will fail. A combination of versions that is known to work is `node@6.13.1` and `katex@0.13.18`. To install the latter with `npm` you can run `npm install -g katex@0.13.18` > \[!NOTE\] If you see a numpy incompatibility error, run: > ``` > pip install 'numpy<2' > ``` When you make changes to the dependencies run by CI, edit the `.ci/docker/requirements-docs.txt` file. #### Building a PDF To compile a PDF of all PyTorch documentation, ensure you have `texlive` and LaTeX installed. On macOS, you can install them using: ``` brew install --cask mactex ``` To create the PDF: 1. Run: ``` make latexpdf ``` This will generate the necessary files in the `build/latex` directory. 2. Navigate to this directory and execute: ``` make LATEXOPTS="-interaction=nonstopmode" ``` This will produce a `pytorch.pdf` with the desired content. Run this command one more time so that it generates the correct table of contents and index. > \[!NOTE\] To view the Table of Contents, switch to the Table of Contents view in your PDF viewer. ### Previous Versions Installation instructions and binaries for previous PyTorch versions may be found on [our website](https://pytorch.org/get-started/previous-versions). ## Getting Started Three pointers to get you started: - [Tutorials: get you started with understanding and using PyTorch](https://pytorch.org/tutorials/) - [Examples: easy to understand PyTorch code across all domains](https://github.com/pytorch/examples) - [The API Reference](https://pytorch.org/docs/) - [Glossary](https://github.com/pytorch/pytorch/blob/main/GLOSSARY.md) ## Resources - [PyTorch.org](https://pytorch.org/) - [PyTorch Tutorials](https://pytorch.org/tutorials/) - [PyTorch Examples](https://github.com/pytorch/examples) - [PyTorch Models](https://pytorch.org/hub/) - [Intro to Deep Learning with PyTorch from Udacity](https://www.udacity.com/course/deep-learning-pytorch--ud188) - [Intro to Machine Learning with PyTorch from Udacity](https://www.udacity.com/course/intro-to-machine-learning-nanodegree--nd229) - [Deep Neural Networks with PyTorch from Coursera](https://www.coursera.org/learn/deep-neural-networks-with-pytorch) - [PyTorch Twitter](https://twitter.com/PyTorch) - [PyTorch Blog](https://pytorch.org/blog/) - [PyTorch YouTube](https://www.youtube.com/channel/UCWXI5YeOsh03QvJ59PMaXFw) ## Communication - Forums: Discuss implementations, research, etc. <https://discuss.pytorch.org> - GitHub Issues: Bug reports, feature requests, install issues, RFCs, thoughts, etc. - Slack: The [PyTorch Slack](https://pytorch.slack.com/) hosts a primary audience of moderate to experienced PyTorch users and developers for general chat, online discussions, collaboration, etc. If you are a beginner looking for help, the primary medium is [PyTorch Forums](https://discuss.pytorch.org/). If you need a slack invite, please fill this form: <https://goo.gl/forms/PP1AGvNHpSaJP8to1> - Newsletter: No-noise, a one-way email newsletter with important announcements about PyTorch. You can sign-up here: <https://eepurl.com/cbG0rv> - Facebook Page: Important announcements about PyTorch. <https://www.facebook.com/pytorch> - For brand guidelines, please visit our website at [pytorch.org](https://pytorch.org/) ## Releases and Contributing Typically, PyTorch has three minor releases a year. Please let us know if you encounter a bug by [filing an issue](https://github.com/pytorch/pytorch/issues). We appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion. If you plan to contribute new features, utility functions, or extensions to the core, please first open an issue and discuss the feature with us. Sending a PR without discussion might end up resulting in a rejected PR because we might be taking the core in a different direction than you might be aware of. To learn more about making a contribution to PyTorch, please see our [Contribution page](https://pypi.org/project/torch/CONTRIBUTING.md). For more information about PyTorch releases, see [Release page](https://pypi.org/project/torch/RELEASE.md). ## The Team PyTorch is a community-driven project with several skillful engineers and researchers contributing to it. PyTorch is currently maintained by [Soumith Chintala](http://soumith.ch/), [Gregory Chanan](https://github.com/gchanan), [Dmytro Dzhulgakov](https://github.com/dzhulgakov), [Edward Yang](https://github.com/ezyang), [Alban Desmaison](https://github.com/albanD), [Piotr Bialecki](https://github.com/ptrblck) and [Nikita Shulga](https://github.com/malfet) with major contributions coming from hundreds of talented individuals in various forms and means. A non-exhaustive but growing list needs to mention: [Trevor Killeen](https://github.com/killeent), [Sasank Chilamkurthy](https://github.com/chsasank), [Sergey Zagoruyko](https://github.com/szagoruyko), [Adam Lerer](https://github.com/adamlerer), [Francisco Massa](https://github.com/fmassa), [Alykhan Tejani](https://github.com/alykhantejani), [Luca Antiga](https://github.com/lantiga), [Alban Desmaison](https://github.com/albanD), [Andreas Koepf](https://github.com/andreaskoepf), [James Bradbury](https://github.com/jekbradbury), [Zeming Lin](https://github.com/ebetica), [Yuandong Tian](https://github.com/yuandong-tian), [Guillaume Lample](https://github.com/glample), [Marat Dukhan](https://github.com/Maratyszcza), [Natalia Gimelshein](https://github.com/ngimel), [Christian Sarofeen](https://github.com/csarofeen), [Martin Raison](https://github.com/martinraison), [Edward Yang](https://github.com/ezyang), [Zachary Devito](https://github.com/zdevito). Note: This project is unrelated to [hughperkins/pytorch](https://github.com/hughperkins/pytorch) with the same name. Hugh is a valuable contributor to the Torch community and has helped with many things Torch and PyTorch. ## License PyTorch has a BSD-style license, as found in the [LICENSE](https://pypi.org/project/torch/LICENSE) file. ## Project details ### Verified details These details have been [verified by PyPI](https://docs.pypi.org/project_metadata/#verified-details) ###### Owner - [Meta Platforms](https://pypi.org/org/meta-platforms/) ###### Maintainers [![Avatar for atalman from gravatar.com](https://pypi-camo.freetls.fastly.net/4a1e9de14640155ddeb5938095a666ac957155a3/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f34363564656263366230376638366139613236613461323633346535396166383f73697a653d3530) atalman](https://pypi.org/user/atalman/) [![Avatar for malfet from gravatar.com](https://pypi-camo.freetls.fastly.net/d8937f01c4383b85866102a5e806397f40b3a557/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f35336331663335393833326462633337366337613732626365323539616536643f73697a653d3530) malfet](https://pypi.org/user/malfet/) [![Avatar for seemethere from gravatar.com](https://pypi-camo.freetls.fastly.net/0bde43daede4ad73d5e20183d999d5a5aeb0c0e0/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f62343230313862343337386564393666393330356133663334386462326435623f73697a653d3530) seemethere](https://pypi.org/user/seemethere/) [![Avatar for soumith from gravatar.com](https://pypi-camo.freetls.fastly.net/73d38504f107197034b9b146995199672719c24e/68747470733a2f2f7365637572652e67726176617461722e636f6d2f6176617461722f65386363373633626131316639323936353765323863623033373162373563383f73697a653d3530) soumith](https://pypi.org/user/soumith/) ### Unverified details These details have not* been verified by PyPI* ###### Project links - [Documentation](https://pytorch.org/docs) - [Forum](https://discuss.pytorch.org/) - [Homepage](https://pytorch.org/) - [Issue Tracker](https://github.com/pytorch/pytorch/issues) - [Repository](https://github.com/pytorch/pytorch) ###### Meta - License: BSD-3-Clause - Author: [PyTorch Team](mailto:packages@pytorch.org) - Tags pytorch , machine learning - Requires: Python \>=3.10 - Provides-Extra: `optree` , `opt-einsum` , `pyyaml` ###### Classifiers - Development Status - [5 - Production/Stable](https://pypi.org/search/?c=Development+Status+%3A%3A+5+-+Production%2FStable) - Intended Audience - [Developers](https://pypi.org/search/?c=Intended+Audience+%3A%3A+Developers) - [Education](https://pypi.org/search/?c=Intended+Audience+%3A%3A+Education) - [Science/Research](https://pypi.org/search/?c=Intended+Audience+%3A%3A+Science%2FResearch) - Programming Language - [C++](https://pypi.org/search/?c=Programming+Language+%3A%3A+C%2B%2B) - [Python :: 3 :: Only](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3+%3A%3A+Only) - [Python :: 3.10](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.10) - [Python :: 3.11](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.11) - [Python :: 3.12](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.12) - [Python :: 3.13](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.13) - [Python :: 3.14](https://pypi.org/search/?c=Programming+Language+%3A%3A+Python+%3A%3A+3.14) - Topic - [Scientific/Engineering](https://pypi.org/search/?c=Topic+%3A%3A+Scientific%2FEngineering) - [Scientific/Engineering :: Artificial Intelligence](https://pypi.org/search/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Artificial+Intelligence) - [Scientific/Engineering :: Mathematics](https://pypi.org/search/?c=Topic+%3A%3A+Scientific%2FEngineering+%3A%3A+Mathematics) - [Software Development](https://pypi.org/search/?c=Topic+%3A%3A+Software+Development) - [Software Development :: Libraries](https://pypi.org/search/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries) - [Software Development :: Libraries :: Python Modules](https://pypi.org/search/?c=Topic+%3A%3A+Software+Development+%3A%3A+Libraries+%3A%3A+Python+Modules) ## Release history [Release notifications](https://pypi.org/help/#project-release-notifications) \\| [RSS feed](https://pypi.org/rss/project/torch/releases.xml) This version ![](https://pypi.org/static/images/blue-cube.572a5bfb.svg) [2\.11.0 Mar 23, 2026](https://pypi.org/project/torch/2.11.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.10.0 Jan 21, 2026](https://pypi.org/project/torch/2.10.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.9.1 Nov 12, 2025](https://pypi.org/project/torch/2.9.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.9.0 Oct 15, 2025](https://pypi.org/project/torch/2.9.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.8.0 Aug 6, 2025](https://pypi.org/project/torch/2.8.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.7.1 Jun 4, 2025](https://pypi.org/project/torch/2.7.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.7.0 Apr 23, 2025](https://pypi.org/project/torch/2.7.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.6.0 Jan 29, 2025](https://pypi.org/project/torch/2.6.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.5.1 Oct 29, 2024](https://pypi.org/project/torch/2.5.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.5.0 Oct 17, 2024](https://pypi.org/project/torch/2.5.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.4.1 Sep 4, 2024](https://pypi.org/project/torch/2.4.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.4.0 Jul 24, 2024](https://pypi.org/project/torch/2.4.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.3.1 Jun 5, 2024](https://pypi.org/project/torch/2.3.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.3.0 Apr 24, 2024](https://pypi.org/project/torch/2.3.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.2.2 Mar 27, 2024](https://pypi.org/project/torch/2.2.2/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.2.1 Feb 22, 2024](https://pypi.org/project/torch/2.2.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.2.0 Jan 30, 2024](https://pypi.org/project/torch/2.2.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.1.2 Dec 14, 2023](https://pypi.org/project/torch/2.1.2/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.1.1 Nov 15, 2023](https://pypi.org/project/torch/2.1.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.1.0 Oct 4, 2023](https://pypi.org/project/torch/2.1.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.0.1 May 8, 2023](https://pypi.org/project/torch/2.0.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [2\.0.0 Mar 15, 2023](https://pypi.org/project/torch/2.0.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.13.1 Dec 15, 2022](https://pypi.org/project/torch/1.13.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.13.0 Oct 28, 2022](https://pypi.org/project/torch/1.13.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.12.1 Aug 5, 2022](https://pypi.org/project/torch/1.12.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.12.0 Jun 28, 2022](https://pypi.org/project/torch/1.12.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.11.0 Mar 10, 2022](https://pypi.org/project/torch/1.11.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.10.2 Jan 27, 2022](https://pypi.org/project/torch/1.10.2/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.10.1 Dec 15, 2021](https://pypi.org/project/torch/1.10.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.10.0 Oct 21, 2021](https://pypi.org/project/torch/1.10.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.9.1 Sep 20, 2021](https://pypi.org/project/torch/1.9.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.9.0 Jun 15, 2021](https://pypi.org/project/torch/1.9.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.8.1 Mar 25, 2021](https://pypi.org/project/torch/1.8.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.8.0 Mar 4, 2021](https://pypi.org/project/torch/1.8.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.7.1 Dec 10, 2020](https://pypi.org/project/torch/1.7.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.7.0 Oct 27, 2020](https://pypi.org/project/torch/1.7.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.6.0 Jul 28, 2020](https://pypi.org/project/torch/1.6.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.5.1 Jun 18, 2020](https://pypi.org/project/torch/1.5.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.5.0 Apr 21, 2020](https://pypi.org/project/torch/1.5.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.4.0 Jan 15, 2020](https://pypi.org/project/torch/1.4.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.3.1 Nov 7, 2019](https://pypi.org/project/torch/1.3.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.3.0 Oct 10, 2019](https://pypi.org/project/torch/1.3.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.2.0 Aug 8, 2019](https://pypi.org/project/torch/1.2.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.1.0 Apr 30, 2019](https://pypi.org/project/torch/1.1.0/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.0.1 Feb 5, 2019](https://pypi.org/project/torch/1.0.1/) ![](https://pypi.org/static/images/white-cube.2351a86c.svg) [1\.0.0 Dec 7, 2018](https://pypi.org/project/torch/1.0.0/) ## Download files Download the file for your platform. If you're not sure which to choose, learn more about [installing packages](https://packaging.python.org/tutorials/installing-packages/ "External link"). ### Source Distributions No source distribution files available for this release.See tutorial on [generating distribution archives](https://packaging.python.org/tutorials/packaging-projects/#generating-distribution-archives "External link"). ### Built Distributions Filter files by name, interpreter, ABI, and platform. If you're not sure about the file name format, learn more about [wheel file names](https://packaging.python.org/en/latest/specifications/binary-distribution-format/ "External link"). The dropdown lists show the available interpreters, ABIs, and platforms. Enable javascript to be able to filter the list of wheel files. Copy a direct link to the current filters <https://pypi.org/project/torch/#files> Copy Showing 28 of 28 files. File name Interpreter ABI Platform [torch-2.11.0-cp314-cp314t-win\_amd64.whl](https://files.pythonhosted.org/packages/cf/bf/c8d12a2c86dbfd7f40fb2f56fbf5a505ccf2d9ce131eb559dfc7c51e1a04/torch-2.11.0-cp314-cp314t-win_amd64.whl) (114.8 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314t-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.14t``Windows x86-64` [torch-2.11.0-cp314-cp314t-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/b1/e7/0b6665f533aa9e337662dc190425abc0af1fe3234088f4454c52393ded61/torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.14t``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp314-cp314t-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/fd/66/54a56a4a6ceaffb567231994a9745821d3af922a854ed33b0b3a278e0a99/torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.14t``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp314-cp314t-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/bf/46/4419098ed6d801750f26567b478fc185c3432e11e2cad712bc6b4c2ab0d0/torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl) (81.0 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.14t``macOS 11.0+ ARM64` [torch-2.11.0-cp314-cp314-win\_amd64.whl](https://files.pythonhosted.org/packages/78/88/d4a4cda8362f8a30d1ed428564878c3cafb0d87971fbd3947d4c84552095/torch-2.11.0-cp314-cp314-win_amd64.whl) (114.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.14``Windows x86-64` [torch-2.11.0-cp314-cp314-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/47/e8/b98ca2d39b2e0e4730c0ee52537e488e7008025bc77ca89552ff91021f7c/torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl) (530.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.14``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp314-cp314-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/c7/86/7cd7c66cb9cec6be330fff36db5bd0eef386d80c031b581ec81be1d4b26c/torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.14``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp314-cp314-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/26/0d/8603382f61abd0db35841148ddc1ffd607bf3100b11c6e1dab6d2fc44e72/torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl) (80.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.14``macOS 11.0+ ARM64` [torch-2.11.0-cp313-cp313t-win\_amd64.whl](https://files.pythonhosted.org/packages/48/6b/30d1459fa7e4b67e9e3fe1685ca1d8bb4ce7c62ef436c3a615963c6c866c/torch-2.11.0-cp313-cp313t-win_amd64.whl) (114.8 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313t-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.13t``Windows x86-64` [torch-2.11.0-cp313-cp313t-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/07/f4/1b666b6d61d3394cca306ea543ed03a64aad0a201b6cd159f1d41010aeb1/torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.13t``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp313-cp313t-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/6d/6c/56bfb37073e7136e6dd86bfc6af7339946dd684e0ecf2155ac0eee687ae1/torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.13t``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp313-cp313t-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/db/38/8ac78069621b8c2b4979c2f96dc8409ef5e9c4189f6aac629189a78677ca/torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl) (81.0 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.13t``macOS 11.0+ ARM64` [torch-2.11.0-cp313-cp313-win\_amd64.whl](https://files.pythonhosted.org/packages/66/82/3e3fcdd388fbe54e29fd3f991f36846ff4ac90b0d0181e9c8f7236565f82/torch-2.11.0-cp313-cp313-win_amd64.whl) (114.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.13``Windows x86-64` [torch-2.11.0-cp313-cp313-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/3d/e1/b73f7c575a4b8f87a5928f50a1e35416b5e27295d8be9397d5293e7e8d4c/torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl) (530.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.13``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp313-cp313-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/32/d1/8ed2173589cbfe744ed54e5a73efc107c0085ba5777ee93a5f4c1ab90553/torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.13``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp313-cp313-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/87/89/5ea6722763acee56b045435fb84258db7375c48165ec8be7880ab2b281c5/torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl) (80.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.13``macOS 11.0+ ARM64` [torch-2.11.0-cp312-cp312-win\_amd64.whl](https://files.pythonhosted.org/packages/1c/ff/6756f1c7ee302f6d202120e0f4f05b432b839908f9071157302cedfc5232/torch-2.11.0-cp312-cp312-win_amd64.whl) (114.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp312-cp312-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.12``Windows x86-64` [torch-2.11.0-cp312-cp312-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/1a/c9/82638ef24d7877510f83baf821f5619a61b45568ce21c0a87a91576510aa/torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl) (530.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.12``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp312-cp312-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/13/16/42e5915ebe4868caa6bac83a8ed59db57f12e9a61b7d749d584776ed53d5/torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.12``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp312-cp312-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/6f/8b/69e3008d78e5cee2b30183340cc425081b78afc5eff3d080daab0adda9aa/torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl) (80.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.12``macOS 11.0+ ARM64` [torch-2.11.0-cp311-cp311-win\_amd64.whl](https://files.pythonhosted.org/packages/d1/bd/9912d30b68845256aabbb4a40aeefeef3c3b20db5211ccda653544ada4b6/torch-2.11.0-cp311-cp311-win_amd64.whl) (114.5 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp311-cp311-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.11``Windows x86-64` [torch-2.11.0-cp311-cp311-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/8c/8b/d7be22fbec9ffee6cff31a39f8750d4b3a65d349a286cf4aec74c2375662/torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.11``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp311-cp311-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/84/03/acea680005f098f79fd70c1d9d5ccc0cb4296ec2af539a0450108232fc0c/torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.11``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp311-cp311-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/ae/0d/98b410492609e34a155fa8b121b55c7dca229f39636851c3a9ec20edea21/torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl) (80.5 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.11``macOS 11.0+ ARM64` [torch-2.11.0-cp310-cp310-win\_amd64.whl](https://files.pythonhosted.org/packages/35/40/2d532e8c0e23705be9d1debce5bc37b68d59a39bda7584c26fe9668076fe/torch-2.11.0-cp310-cp310-win_amd64.whl) (114.5 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp310-cp310-win_amd64.whl)) Uploaded Mar 23, 2026 `CPython 3.10``Windows x86-64` [torch-2.11.0-cp310-cp310-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/f9/1e/18a9b10b4bd34f12d4e561c52b0ae7158707b8193c6cfc0aad2b48167090/torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl) (530.6 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl)) Uploaded Mar 23, 2026 `CPython 3.10``manylinux: glibc 2.28+ x86-64` [torch-2.11.0-cp310-cp310-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/a4/f0/98ae802fa8c09d3149b0c8690741f3f5753c90e779bd28c9613257295945/torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl) (419.7 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl)) Uploaded Mar 23, 2026 `CPython 3.10``manylinux: glibc 2.28+ ARM64` [torch-2.11.0-cp310-cp310-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/ac/f2/c1690994afe461aae2d0cac62251e6802a703dec0a6c549c02ecd0de92a9/torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl) (80.5 MB [view details](https://pypi.org/project/torch/#torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl)) Uploaded Mar 23, 2026 `CPython 3.10``macOS 11.0+ ARM64` ## File details Details for the file `torch-2.11.0-cp314-cp314t-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314t-win\_amd64.whl](https://files.pythonhosted.org/packages/cf/bf/c8d12a2c86dbfd7f40fb2f56fbf5a505ccf2d9ce131eb559dfc7c51e1a04/torch-2.11.0-cp314-cp314t-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.8 MB - Tags: CPython 3.14t, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `b2a43985ff5ef6ddd923bbcf99943e5f58059805787c5c9a2622bf05ca2965b0` \| Copy \| \| MD5 \| `5e5a354ee5a4478788038514491c9b53` \| Copy \| \| BLAKE2b-256 \| `cfbfc8d12a2c86dbfd7f40fb2f56fbf5a505ccf2d9ce131eb559dfc7c51e1a04` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314t-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/b1/e7/0b6665f533aa9e337662dc190425abc0af1fe3234088f4454c52393ded61/torch-2.11.0-cp314-cp314t-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.6 MB - Tags: CPython 3.14t, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `563ed3d25542d7e7bbc5b235ccfacfeb97fb470c7fee257eae599adb8005c8a2` \| Copy \| \| MD5 \| `1fd06fd8ab11531a465edbd203708a92` \| Copy \| \| BLAKE2b-256 \| `b1e70b6665f533aa9e337662dc190425abc0af1fe3234088f4454c52393ded61` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314t-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/fd/66/54a56a4a6ceaffb567231994a9745821d3af922a854ed33b0b3a278e0a99/torch-2.11.0-cp314-cp314t-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.14t, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `ab9a8482f475f9ba20e12db84b0e55e2f58784bdca43a854a6ccd3fd4b9f75e6` \| Copy \| \| MD5 \| `1042d08d26381bf18f22fb2eeff9acf4` \| Copy \| \| BLAKE2b-256 \| `fd6654a56a4a6ceaffb567231994a9745821d3af922a854ed33b0b3a278e0a99` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314t-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/bf/46/4419098ed6d801750f26567b478fc185c3432e11e2cad712bc6b4c2ab0d0/torch-2.11.0-cp314-cp314t-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 81.0 MB - Tags: CPython 3.14t, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `8245477871c3700d4370352ffec94b103cfcb737229445cf9946cddb7b2ca7cd` \| Copy \| \| MD5 \| `b173521d650ee78e110f4e52d758ffea` \| Copy \| \| BLAKE2b-256 \| `bf464419098ed6d801750f26567b478fc185c3432e11e2cad712bc6b4c2ab0d0` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314-win\_amd64.whl](https://files.pythonhosted.org/packages/78/88/d4a4cda8362f8a30d1ed428564878c3cafb0d87971fbd3947d4c84552095/torch-2.11.0-cp314-cp314-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.6 MB - Tags: CPython 3.14, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `2b4e811728bd0cc58fb2b0948fe939a1ee2bf1422f6025be2fca4c7bd9d79718` \| Copy \| \| MD5 \| `78e8aec77b35ce2adba124b9fe686764` \| Copy \| \| BLAKE2b-256 \| `7888d4a4cda8362f8a30d1ed428564878c3cafb0d87971fbd3947d4c84552095` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/47/e8/b98ca2d39b2e0e4730c0ee52537e488e7008025bc77ca89552ff91021f7c/torch-2.11.0-cp314-cp314-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.7 MB - Tags: CPython 3.14, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `4dc8b3809469b6c30b411bb8c4cad3828efd26236153d9beb6a3ec500f211a60` \| Copy \| \| MD5 \| `73e8a53e185b61d9e87ee0516a65fbc0` \| Copy \| \| BLAKE2b-256 \| `47e8b98ca2d39b2e0e4730c0ee52537e488e7008025bc77ca89552ff91021f7c` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/c7/86/7cd7c66cb9cec6be330fff36db5bd0eef386d80c031b581ec81be1d4b26c/torch-2.11.0-cp314-cp314-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.14, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `2bb3cc54bd0dea126b0060bb1ec9de0f9c7f7342d93d436646516b0330cd5be7` \| Copy \| \| MD5 \| `808fab6c041012c9003e03b0fee4c96e` \| Copy \| \| BLAKE2b-256 \| `c7867cd7c66cb9cec6be330fff36db5bd0eef386d80c031b581ec81be1d4b26c` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp314-cp314-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/26/0d/8603382f61abd0db35841148ddc1ffd607bf3100b11c6e1dab6d2fc44e72/torch-2.11.0-cp314-cp314-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 80.6 MB - Tags: CPython 3.14, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `01018087326984a33b64e04c8cb5c2795f9120e0d775ada1f6638840227b04d7` \| Copy \| \| MD5 \| `dbbef6f8b3c61fe3244f2f4cdc0f11a1` \| Copy \| \| BLAKE2b-256 \| `260d8603382f61abd0db35841148ddc1ffd607bf3100b11c6e1dab6d2fc44e72` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313t-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313t-win\_amd64.whl](https://files.pythonhosted.org/packages/48/6b/30d1459fa7e4b67e9e3fe1685ca1d8bb4ce7c62ef436c3a615963c6c866c/torch-2.11.0-cp313-cp313t-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.8 MB - Tags: CPython 3.13t, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `a97b94bbf62992949b4730c6cd2cc9aee7b335921ee8dc207d930f2ed09ae2db` \| Copy \| \| MD5 \| `0f7535cb00b28ab4a04c8160a22fbbad` \| Copy \| \| BLAKE2b-256 \| `486b30d1459fa7e4b67e9e3fe1685ca1d8bb4ce7c62ef436c3a615963c6c866c` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313t-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/07/f4/1b666b6d61d3394cca306ea543ed03a64aad0a201b6cd159f1d41010aeb1/torch-2.11.0-cp313-cp313t-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.6 MB - Tags: CPython 3.13t, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `98bb213c3084cfe176302949bdc360074b18a9da7ab59ef2edc9d9f742504778` \| Copy \| \| MD5 \| `c46ed32625c87b1e75b502b11ebf720d` \| Copy \| \| BLAKE2b-256 \| `07f41b666b6d61d3394cca306ea543ed03a64aad0a201b6cd159f1d41010aeb1` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313t-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/6d/6c/56bfb37073e7136e6dd86bfc6af7339946dd684e0ecf2155ac0eee687ae1/torch-2.11.0-cp313-cp313t-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.13t, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `2658f34ce7e2dabf4ec73b45e2ca68aedad7a5be87ea756ad656eaf32bf1e1ea` \| Copy \| \| MD5 \| `af3c69887fbfc4fbc68b9602a8fcb308` \| Copy \| \| BLAKE2b-256 \| `6d6c56bfb37073e7136e6dd86bfc6af7339946dd684e0ecf2155ac0eee687ae1` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313t-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/db/38/8ac78069621b8c2b4979c2f96dc8409ef5e9c4189f6aac629189a78677ca/torch-2.11.0-cp313-cp313t-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 81.0 MB - Tags: CPython 3.13t, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `8b394322f49af4362d4f80e424bcaca7efcd049619af03a4cf4501520bdf0fb4` \| Copy \| \| MD5 \| `380d859cfad5f8262ba0b834da35bcf4` \| Copy \| \| BLAKE2b-256 \| `db388ac78069621b8c2b4979c2f96dc8409ef5e9c4189f6aac629189a78677ca` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313-win\_amd64.whl](https://files.pythonhosted.org/packages/66/82/3e3fcdd388fbe54e29fd3f991f36846ff4ac90b0d0181e9c8f7236565f82/torch-2.11.0-cp313-cp313-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.6 MB - Tags: CPython 3.13, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `4dda3b3f52d121063a731ddb835f010dc137b920d7fec2778e52f60d8e4bf0cd` \| Copy \| \| MD5 \| `0fe7969556fd22b5a190345511e1ae68` \| Copy \| \| BLAKE2b-256 \| `66823e3fcdd388fbe54e29fd3f991f36846ff4ac90b0d0181e9c8f7236565f82` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/3d/e1/b73f7c575a4b8f87a5928f50a1e35416b5e27295d8be9397d5293e7e8d4c/torch-2.11.0-cp313-cp313-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.7 MB - Tags: CPython 3.13, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `cc89b9b173d9adfab59fd227f0ab5e5516d9a52b658ae41d64e59d2e55a418db` \| Copy \| \| MD5 \| `8f55c6619d670cc6ee89fbfde6cb4f5c` \| Copy \| \| BLAKE2b-256 \| `3de1b73f7c575a4b8f87a5928f50a1e35416b5e27295d8be9397d5293e7e8d4c` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/32/d1/8ed2173589cbfe744ed54e5a73efc107c0085ba5777ee93a5f4c1ab90553/torch-2.11.0-cp313-cp313-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.13, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `63a68fa59de8f87acc7e85a5478bb2dddbb3392b7593ec3e78827c793c4b73fd` \| Copy \| \| MD5 \| `a15b86b7cda0a0a1613b306676ec374d` \| Copy \| \| BLAKE2b-256 \| `32d18ed2173589cbfe744ed54e5a73efc107c0085ba5777ee93a5f4c1ab90553` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp313-cp313-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/87/89/5ea6722763acee56b045435fb84258db7375c48165ec8be7880ab2b281c5/torch-2.11.0-cp313-cp313-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 80.6 MB - Tags: CPython 3.13, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `1e6debd97ccd3205bbb37eb806a9d8219e1139d15419982c09e23ef7d4369d18` \| Copy \| \| MD5 \| `82d7d812c6109516d7cad4ca8002395e` \| Copy \| \| BLAKE2b-256 \| `87895ea6722763acee56b045435fb84258db7375c48165ec8be7880ab2b281c5` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp312-cp312-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp312-cp312-win\_amd64.whl](https://files.pythonhosted.org/packages/1c/ff/6756f1c7ee302f6d202120e0f4f05b432b839908f9071157302cedfc5232/torch-2.11.0-cp312-cp312-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.6 MB - Tags: CPython 3.12, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `fbf39280699d1b869f55eac536deceaa1b60bd6788ba74f399cc67e60a5fab10` \| Copy \| \| MD5 \| `1799208f1391b3d694aee962cb9e5a57` \| Copy \| \| BLAKE2b-256 \| `1cff6756f1c7ee302f6d202120e0f4f05b432b839908f9071157302cedfc5232` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp312-cp312-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/1a/c9/82638ef24d7877510f83baf821f5619a61b45568ce21c0a87a91576510aa/torch-2.11.0-cp312-cp312-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.7 MB - Tags: CPython 3.12, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `0f68f4ac6d95d12e896c3b7a912b5871619542ec54d3649cf48cc1edd4dd2756` \| Copy \| \| MD5 \| `2b481e9bac55320ccdcef091e3a4ac52` \| Copy \| \| BLAKE2b-256 \| `1ac982638ef24d7877510f83baf821f5619a61b45568ce21c0a87a91576510aa` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp312-cp312-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/13/16/42e5915ebe4868caa6bac83a8ed59db57f12e9a61b7d749d584776ed53d5/torch-2.11.0-cp312-cp312-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.12, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `f99924682ef0aa6a4ab3b1b76f40dc6e273fca09f367d15a524266db100a723f` \| Copy \| \| MD5 \| `d44ddbd0e2d0166ef511d5d3ab19c64a` \| Copy \| \| BLAKE2b-256 \| `131642e5915ebe4868caa6bac83a8ed59db57f12e9a61b7d749d584776ed53d5` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp312-cp312-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/6f/8b/69e3008d78e5cee2b30183340cc425081b78afc5eff3d080daab0adda9aa/torch-2.11.0-cp312-cp312-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 80.6 MB - Tags: CPython 3.12, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `4b5866312ee6e52ea625cd211dcb97d6a2cdc1131a5f15cc0d87eec948f6dd34` \| Copy \| \| MD5 \| `0a02d3ea665d48edc1fd7f2d70a76e08` \| Copy \| \| BLAKE2b-256 \| `6f8b69e3008d78e5cee2b30183340cc425081b78afc5eff3d080daab0adda9aa` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp311-cp311-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp311-cp311-win\_amd64.whl](https://files.pythonhosted.org/packages/d1/bd/9912d30b68845256aabbb4a40aeefeef3c3b20db5211ccda653544ada4b6/torch-2.11.0-cp311-cp311-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.5 MB - Tags: CPython 3.11, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `73e24aaf8f36ab90d95cd1761208b2eb70841c2a9ca1a3f9061b39fc5331b708` \| Copy \| \| MD5 \| `66da62379541d52ab250ea3ae052dfe6` \| Copy \| \| BLAKE2b-256 \| `d1bd9912d30b68845256aabbb4a40aeefeef3c3b20db5211ccda653544ada4b6` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp311-cp311-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/8c/8b/d7be22fbec9ffee6cff31a39f8750d4b3a65d349a286cf4aec74c2375662/torch-2.11.0-cp311-cp311-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.6 MB - Tags: CPython 3.11, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `7aa2f9bbc6d4595ba72138026b2074be1233186150e9292865e04b7a63b8c67a` \| Copy \| \| MD5 \| `aa9c5f374a0a6834cb940b42a29978ae` \| Copy \| \| BLAKE2b-256 \| `8c8bd7be22fbec9ffee6cff31a39f8750d4b3a65d349a286cf4aec74c2375662` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp311-cp311-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/84/03/acea680005f098f79fd70c1d9d5ccc0cb4296ec2af539a0450108232fc0c/torch-2.11.0-cp311-cp311-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.11, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `d91aac77f24082809d2c5a93f52a5f085032740a1ebc9252a7b052ef5a4fddc6` \| Copy \| \| MD5 \| `f471c3392876a56dcf1f57a200b649de` \| Copy \| \| BLAKE2b-256 \| `8403acea680005f098f79fd70c1d9d5ccc0cb4296ec2af539a0450108232fc0c` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp311-cp311-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/ae/0d/98b410492609e34a155fa8b121b55c7dca229f39636851c3a9ec20edea21/torch-2.11.0-cp311-cp311-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 80.5 MB - Tags: CPython 3.11, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `7b6a60d48062809f58595509c524b88e6ddec3ebe25833d6462eeab81e5f2ce4` \| Copy \| \| MD5 \| `544775e2610b48889b875137a15f0f79` \| Copy \| \| BLAKE2b-256 \| `ae0d98b410492609e34a155fa8b121b55c7dca229f39636851c3a9ec20edea21` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp310-cp310-win_amd64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp310-cp310-win\_amd64.whl](https://files.pythonhosted.org/packages/35/40/2d532e8c0e23705be9d1debce5bc37b68d59a39bda7584c26fe9668076fe/torch-2.11.0-cp310-cp310-win_amd64.whl) - Upload date: Mar 23, 2026 - Size: 114.5 MB - Tags: CPython 3.10, Windows x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `b3c712ae6fb8e7a949051a953fc412fe0a6940337336c3b6f905e905dac5157f` \| Copy \| \| MD5 \| `92a04a9fc443c7ff23226e10a3af704f` \| Copy \| \| BLAKE2b-256 \| `35402d532e8c0e23705be9d1debce5bc37b68d59a39bda7584c26fe9668076fe` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp310-cp310-manylinux\_2\_28\_x86\_64.whl](https://files.pythonhosted.org/packages/f9/1e/18a9b10b4bd34f12d4e561c52b0ae7158707b8193c6cfc0aad2b48167090/torch-2.11.0-cp310-cp310-manylinux_2_28_x86_64.whl) - Upload date: Mar 23, 2026 - Size: 530.6 MB - Tags: CPython 3.10, manylinux: glibc 2.28+ x86-64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `1b32ceda909818a03b112006709b02be1877240c31750a8d9c6b7bf5f2d8a6e5` \| Copy \| \| MD5 \| `1552a6d321733cf7a046d92afbb6ac67` \| Copy \| \| BLAKE2b-256 \| `f91e18a9b10b4bd34f12d4e561c52b0ae7158707b8193c6cfc0aad2b48167090` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp310-cp310-manylinux\_2\_28\_aarch64.whl](https://files.pythonhosted.org/packages/a4/f0/98ae802fa8c09d3149b0c8690741f3f5753c90e779bd28c9613257295945/torch-2.11.0-cp310-cp310-manylinux_2_28_aarch64.whl) - Upload date: Mar 23, 2026 - Size: 419.7 MB - Tags: CPython 3.10, manylinux: glibc 2.28+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `4cf8687f4aec3900f748d553483ef40e0ac38411c3c48d0a86a438f6d7a99b18` \| Copy \| \| MD5 \| `893da691a4d1d289f24de063a7835f57` \| Copy \| \| BLAKE2b-256 \| `a4f098ae802fa8c09d3149b0c8690741f3f5753c90e779bd28c9613257295945` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ## File details Details for the file `torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl`. ### File metadata - Download URL: [torch-2.11.0-cp310-cp310-macosx\_11\_0\_arm64.whl](https://files.pythonhosted.org/packages/ac/f2/c1690994afe461aae2d0cac62251e6802a703dec0a6c549c02ecd0de92a9/torch-2.11.0-cp310-cp310-macosx_11_0_arm64.whl) - Upload date: Mar 23, 2026 - Size: 80.5 MB - Tags: CPython 3.10, macOS 11.0+ ARM64 - Uploaded using Trusted Publishing? No - Uploaded via: twine/6.2.0 CPython/3.11.6 ### File hashes \| Algorithm \| Hash digest \| \| \|---\|---\|---\| \| SHA256 \| `2c0d7fcfbc0c4e8bb5ebc3907cbc0c6a0da1b8f82b1fc6e14e914fa0b9baf74e` \| Copy \| \| MD5 \| `a052f0ae699d2599a2d0f43a083f9275` \| Copy \| \| BLAKE2b-256 \| `acf2c1690994afe461aae2d0cac62251e6802a703dec0a6c549c02ecd0de92a9` \| Copy \| [See more details on using hashes here.](https://pip.pypa.io/en/stable/topics/secure-installs/#hash-checking-mode "External link") ![](https://pypi.org/static/images/white-cube.2351a86c.svg) ## Help - [Installing packages](https://packaging.python.org/tutorials/installing-packages/ "External link") - [Uploading packages](https://packaging.python.org/tutorials/packaging-projects/ "External link") - [User guide](https://packaging.python.org/ "External link") - [Project name retention](https://www.python.org/dev/peps/pep-0541/ "External link") - [FAQs](https://pypi.org/help/) ## About PyPI - [PyPI Blog](https://blog.pypi.org/ "External link") - [Infrastructure dashboard](https://dtdg.co/pypi "External link") - [Statistics](https://pypi.org/stats/) - [Logos & trademarks](https://pypi.org/trademarks/) - [Our sponsors](https://pypi.org/sponsors/) ## Contributing to PyPI - [Bugs and feedback](https://pypi.org/help/#feedback) - [Contribute on GitHub](https://github.com/pypi/warehouse "External link") - [Translate PyPI](https://hosted.weblate.org/projects/pypa/warehouse/ "External link") - [Sponsor PyPI](https://pypi.org/sponsors/) - [Development credits](https://github.com/pypi/warehouse/graphs/contributors "External link") ## Using PyPI - [Terms of Service](https://policies.python.org/pypi.org/Terms-of-Service/ "External link") - [Report security issue](https://pypi.org/security/) - [Code of conduct](https://policies.python.org/python.org/code-of-conduct/ "External link") - [Privacy Notice](https://policies.python.org/pypi.org/Privacy-Notice/ "External link") - [Acceptable Use Policy](https://policies.python.org/pypi.org/Acceptable-Use-Policy/ "External link") *** Status: [all systems operational](https://status.python.org/ "External link") Developed and maintained by the Python community, for the Python community. [Donate today\!](https://donate.pypi.org/) "PyPI", "Python Package Index", and the blocks logos are registered [trademarks](https://pypi.org/trademarks/) of the [Python Software Foundation](https://www.python.org/psf-landing). © 2026 [Python Software Foundation](https://www.python.org/psf-landing/ "External link") [Site map](https://pypi.org/sitemap/) Switch to desktop version Supported by [![](https://pypi-camo.freetls.fastly.net/ed7074cadad1a06f56bc520ad9bd3e00d0704c5b/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f6177732d77686974652d6c6f676f2d7443615473387a432e706e67) AWS Cloud computing and Security Sponsor](https://aws.amazon.com/) [![](https://pypi-camo.freetls.fastly.net/8855f7c063a3bdb5b0ce8d91bfc50cf851cc5c51/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f64617461646f672d77686974652d6c6f676f2d6668644c4e666c6f2e706e67) Datadog Monitoring](https://www.datadoghq.com/) [![](https://pypi-camo.freetls.fastly.net/60f709d24f3e4d469f9adc77c65e2f5291a3d165/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f6465706f742d77686974652d6c6f676f2d7038506f476831302e706e67) Depot Continuous Integration](https://depot.dev/) [![](https://pypi-camo.freetls.fastly.net/df6fe8829cbff2d7f668d98571df1fd011f36192/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f666173746c792d77686974652d6c6f676f2d65684d3077735f6f2e706e67) Fastly CDN](https://www.fastly.com/) [![](https://pypi-camo.freetls.fastly.net/420cc8cf360bac879e24c923b2f50ba7d1314fb0/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f676f6f676c652d77686974652d6c6f676f2d616734424e3774332e706e67) Google Download Analytics](https://careers.google.com/) [![](https://pypi-camo.freetls.fastly.net/d01053c02f3a626b73ffcb06b96367fdbbf9e230/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f70696e67646f6d2d77686974652d6c6f676f2d67355831547546362e706e67) Pingdom Monitoring](https://www.pingdom.com/) [![](https://pypi-camo.freetls.fastly.net/67af7117035e2345bacb5a82e9aa8b5b3e70701d/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f73656e7472792d77686974652d6c6f676f2d4a2d6b64742d706e2e706e67) Sentry Error logging](https://sentry.io/for/python/?utm_source=pypi&utm_medium=paid-community&utm_campaign=python-na-evergreen&utm_content=static-ad-pypi-sponsor-learnmore) [![](https://pypi-camo.freetls.fastly.net/b611884ff90435a0575dbab7d9b0d3e60f136466/68747470733a2f2f73746f726167652e676f6f676c65617069732e636f6d2f707970692d6173736574732f73706f6e736f726c6f676f732f737461747573706167652d77686974652d6c6f676f2d5467476c6a4a2d502e706e67) StatusPage Status page](https://statuspage.io/)
Readable Markdown	![PyTorch Logo](https://pypi-camo.freetls.fastly.net/ed3839e1c11e779b508097969affa63d0968692b/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f7079746f7263682d6c6f676f2d6461726b2e706e67) * PyTorch is a Python package that provides two high-level features: - Tensor computation (like NumPy) with strong GPU acceleration - Deep neural networks built on a tape-based autograd system You can reuse your favorite Python packages such as NumPy, SciPy, and Cython to extend PyTorch when needed. Our trunk health (Continuous Integration signals) can be found at [hud.pytorch.org](https://hud.pytorch.org/ci/pytorch/pytorch/main). - [More About PyTorch](https://pypi.org/project/torch/#more-about-pytorch) - [A GPU-Ready Tensor Library](https://pypi.org/project/torch/#a-gpu-ready-tensor-library) - [Dynamic Neural Networks: Tape-Based Autograd](https://pypi.org/project/torch/#dynamic-neural-networks-tape-based-autograd) - [Python First](https://pypi.org/project/torch/#python-first) - [Imperative Experiences](https://pypi.org/project/torch/#imperative-experiences) - [Fast and Lean](https://pypi.org/project/torch/#fast-and-lean) - [Extensions Without Pain](https://pypi.org/project/torch/#extensions-without-pain) - [Installation](https://pypi.org/project/torch/#installation) - [Binaries](https://pypi.org/project/torch/#binaries) - [NVIDIA Jetson Platforms](https://pypi.org/project/torch/#nvidia-jetson-platforms) - [From Source](https://pypi.org/project/torch/#from-source) - [Prerequisites](https://pypi.org/project/torch/#prerequisites) - [NVIDIA CUDA Support](https://pypi.org/project/torch/#nvidia-cuda-support) - [AMD ROCm Support](https://pypi.org/project/torch/#amd-rocm-support) - [Intel GPU Support](https://pypi.org/project/torch/#intel-gpu-support) - [Get the PyTorch Source](https://pypi.org/project/torch/#get-the-pytorch-source) - [Install Dependencies](https://pypi.org/project/torch/#install-dependencies) - [Install PyTorch](https://pypi.org/project/torch/#install-pytorch) - [Adjust Build Options (Optional)](https://pypi.org/project/torch/#adjust-build-options-optional) - [Docker Image](https://pypi.org/project/torch/#docker-image) - [Using pre-built images](https://pypi.org/project/torch/#using-pre-built-images) - [Building the image yourself](https://pypi.org/project/torch/#building-the-image-yourself) - [Building the Documentation](https://pypi.org/project/torch/#building-the-documentation) - [Building a PDF](https://pypi.org/project/torch/#building-a-pdf) - [Previous Versions](https://pypi.org/project/torch/#previous-versions) - [Getting Started](https://pypi.org/project/torch/#getting-started) - [Resources](https://pypi.org/project/torch/#resources) - [Communication](https://pypi.org/project/torch/#communication) - [Releases and Contributing](https://pypi.org/project/torch/#releases-and-contributing) - [The Team](https://pypi.org/project/torch/#the-team) - [License](https://pypi.org/project/torch/#license) ## More About PyTorch [Learn the basics of PyTorch](https://pytorch.org/tutorials/beginner/basics/intro.html) At a granular level, PyTorch is a library that consists of the following components: \| Component \| Description \| \|---\|---\| \| [torch](https://pytorch.org/docs/stable/torch.html) \| A Tensor library like NumPy, with strong GPU support \| \| [torch.autograd](https://pytorch.org/docs/stable/autograd.html) \| A tape-based automatic differentiation library that supports all differentiable Tensor operations in torch \| \| [torch.jit](https://pytorch.org/docs/stable/jit.html) \| A compilation stack (TorchScript) to create serializable and optimizable models from PyTorch code \| \| [torch.nn](https://pytorch.org/docs/stable/nn.html) \| A neural networks library deeply integrated with autograd designed for maximum flexibility \| \| [torch.multiprocessing](https://pytorch.org/docs/stable/multiprocessing.html) \| Python multiprocessing, but with magical memory sharing of torch Tensors across processes. Useful for data loading and Hogwild training \| \| [torch.utils*](https://pytorch.org/docs/stable/data.html) \| DataLoader and other utility functions for convenience \| Usually, PyTorch is used either as: - A replacement for NumPy to use the power of GPUs. - A deep learning research platform that provides maximum flexibility and speed. Elaborating Further: ### A GPU-Ready Tensor Library If you use NumPy, then you have used Tensors (a.k.a. ndarray). ![Tensor illustration](https://pypi-camo.freetls.fastly.net/e5f621cdea28d4de40577ac3346794f746d96b86/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f74656e736f725f696c6c757374726174696f6e2e706e67) PyTorch provides Tensors that can live either on the CPU or the GPU and accelerates the computation by a huge amount. We provide a wide variety of tensor routines to accelerate and fit your scientific computation needs such as slicing, indexing, mathematical operations, linear algebra, reductions. And they are fast\! ### Dynamic Neural Networks: Tape-Based Autograd PyTorch has a unique way of building neural networks: using and replaying a tape recorder. Most frameworks such as TensorFlow, Theano, Caffe, and CNTK have a static view of the world. One has to build a neural network and reuse the same structure again and again. Changing the way the network behaves means that one has to start from scratch. With PyTorch, we use a technique called reverse-mode auto-differentiation, which allows you to change the way your network behaves arbitrarily with zero lag or overhead. Our inspiration comes from several research papers on this topic, as well as current and past work such as [torch-autograd](https://github.com/twitter/torch-autograd), [autograd](https://github.com/HIPS/autograd), [Chainer](https://chainer.org/), etc. While this technique is not unique to PyTorch, it's one of the fastest implementations of it to date. You get the best of speed and flexibility for your crazy research. ![Dynamic graph](https://pypi-camo.freetls.fastly.net/360dc3e00cca046f73204c87f117feefe993580f/68747470733a2f2f6769746875622e636f6d2f7079746f7263682f7079746f7263682f7261772f6d61696e2f646f63732f736f757263652f5f7374617469632f696d672f64796e616d69635f67726170682e676966) ### Python First PyTorch is not a Python binding into a monolithic C++ framework. It is built to be deeply integrated into Python. You can use it naturally like you would use [NumPy](https://www.numpy.org/) / [SciPy](https://www.scipy.org/) / [scikit-learn](https://scikit-learn.org/) etc. You can write your new neural network layers in Python itself, using your favorite libraries and use packages such as [Cython](https://cython.org/) and [Numba](http://numba.pydata.org/). Our goal is to not reinvent the wheel where appropriate. ### Imperative Experiences PyTorch is designed to be intuitive, linear in thought, and easy to use. When you execute a line of code, it gets executed. There isn't an asynchronous view of the world. When you drop into a debugger or receive error messages and stack traces, understanding them is straightforward. The stack trace points to exactly where your code was defined. We hope you never spend hours debugging your code because of bad stack traces or asynchronous and opaque execution engines. ### Fast and Lean PyTorch has minimal framework overhead. We integrate acceleration libraries such as [Intel MKL](https://software.intel.com/mkl) and NVIDIA ([cuDNN](https://developer.nvidia.com/cudnn), [NCCL](https://developer.nvidia.com/nccl)) to maximize speed. At the core, its CPU and GPU Tensor and neural network backends are mature and have been tested for years. Hence, PyTorch is quite fast — whether you run small or large neural networks. The memory usage in PyTorch is extremely efficient compared to Torch or some of the alternatives. We've written custom memory allocators for the GPU to make sure that your deep learning models are maximally memory efficient. This enables you to train bigger deep learning models than before. ### Extensions Without Pain Writing new neural network modules, or interfacing with PyTorch's Tensor API was designed to be straightforward and with minimal abstractions. You can write new neural network layers in Python using the torch API [or your favorite NumPy-based libraries such as SciPy](https://pytorch.org/tutorials/advanced/numpy_extensions_tutorial.html). If you want to write your layers in C/C++, we provide a convenient extension API that is efficient and with minimal boilerplate. No wrapper code needs to be written. You can see [a tutorial here](https://pytorch.org/tutorials/advanced/cpp_extension.html) and [an example here](https://github.com/pytorch/extension-cpp). ## Installation ### Binaries Commands to install binaries via Conda or pip wheels are on our website: <https://pytorch.org/get-started/locally/> #### NVIDIA Jetson Platforms Python wheels for NVIDIA's Jetson Nano, Jetson TX1/TX2, Jetson Xavier NX/AGX, and Jetson AGX Orin are provided [here](https://forums.developer.nvidia.com/t/pytorch-for-jetson-version-1-10-now-available/72048) and the L4T container is published [here](https://catalog.ngc.nvidia.com/orgs/nvidia/containers/l4t-pytorch) They require JetPack 4.2 and above, and [@dusty-nv](https://github.com/dusty-nv) and [@ptrblck](https://github.com/ptrblck) are maintaining them. ### From Source #### Prerequisites If you are installing from source, you will need: - Python 3.10 or later - A compiler that fully supports C++17, such as clang or gcc (gcc 9.4.0 or newer is required, on Linux) - Visual Studio or Visual Studio Build Tool (Windows only) \ PyTorch CI uses Visual C++ BuildTools, which come with Visual Studio Enterprise, Professional, or Community Editions. You can also install the build tools from <https://visualstudio.microsoft.com/visual-cpp-build-tools/>. The build tools do not come with Visual Studio Code by default. An example of environment setup is shown below: - Linux: ``` $ source <CONDA_INSTALL_DIR>/bin/activate $ conda create -y -n <CONDA_NAME> $ conda activate <CONDA_NAME> ``` - Windows: ``` $ source <CONDA_INSTALL_DIR>\Scripts\activate.bat $ conda create -y -n <CONDA_NAME> $ conda activate <CONDA_NAME> $ call "C:\Program Files\Microsoft Visual Studio\<VERSION>\Community\VC\Auxiliary\Build\vcvarsall.bat" x64 ``` A conda environment is not required. You can also do a PyTorch build in a standard virtual environment, e.g., created with tools like `uv`, provided your system has installed all the necessary dependencies unavailable as pip packages (e.g., CUDA, MKL.) ##### NVIDIA CUDA Support If you want to compile with CUDA support, [select a supported version of CUDA from our support matrix](https://pytorch.org/get-started/locally/), then install the following: - [NVIDIA CUDA](https://developer.nvidia.com/cuda-downloads) - [NVIDIA cuDNN](https://developer.nvidia.com/cudnn) v8.5 or above - [Compiler](https://gist.github.com/ax3l/9489132) compatible with CUDA Note: You could refer to the [cuDNN Support Matrix](https://docs.nvidia.com/deeplearning/cudnn/backend/latest/reference/support-matrix.html) for cuDNN versions with the various supported CUDA, CUDA driver, and NVIDIA hardware. If you want to disable CUDA support, export the environment variable `USE_CUDA=0`. Other potentially useful environment variables may be found in `setup.py`. If CUDA is installed in a non-standard location, set PATH so that the nvcc you want to use can be found (e.g., `export PATH=/usr/local/cuda-12.8/bin:$PATH`). If you are building for NVIDIA's Jetson platforms (Jetson Nano, TX1, TX2, AGX Xavier), Instructions to install PyTorch for Jetson Nano are [available here](https://devtalk.nvidia.com/default/topic/1049071/jetson-nano/pytorch-for-jetson-nano/) ##### AMD ROCm Support If you want to compile with ROCm support, install - [AMD ROCm](https://rocm.docs.amd.com/en/latest/deploy/linux/quick_start.html) 4.0 and above installation - ROCm is currently supported only for Linux systems. By default the build system expects ROCm to be installed in `/opt/rocm`. If ROCm is installed in a different directory, the `ROCM_PATH` environment variable must be set to the ROCm installation directory. The build system automatically detects the AMD GPU architecture. Optionally, the AMD GPU architecture can be explicitly set with the `PYTORCH_ROCM_ARCH` environment variable [AMD GPU architecture](https://rocm.docs.amd.com/projects/install-on-linux/en/latest/reference/system-requirements.html#supported-gpus) If you want to disable ROCm support, export the environment variable `USE_ROCM=0`. Other potentially useful environment variables may be found in `setup.py`. ##### Intel GPU Support If you want to compile with Intel GPU support, follow these - [PyTorch Prerequisites for Intel GPUs](https://www.intel.com/content/www/us/en/developer/articles/tool/pytorch-prerequisites-for-intel-gpu.html) instructions. - Intel GPU is supported for Linux and Windows. If you want to disable Intel GPU support, export the environment variable `USE_XPU=0`. Other potentially useful environment variables may be found in `setup.py`. #### Get the PyTorch Source ``` git clone https://github.com/pytorch/pytorch cd pytorch # if you are updating an existing checkout git submodule sync git submodule update --init --recursive ``` #### Install Dependencies Common ``` # Run this command from the PyTorch directory after cloning the source code using the “Get the PyTorch Source“ section above pip install --group dev ``` On Linux ``` pip install mkl-static mkl-include # CUDA only: Add LAPACK support for the GPU if needed # magma installation: run with active conda environment. specify CUDA version to install .ci/docker/common/install_magma_conda.sh 12.4 # (optional) If using torch.compile with inductor/triton, install the matching version of triton # Run from the pytorch directory after cloning # For Intel GPU support, please explicitly `export USE_XPU=1` before running command. make triton ``` On MacOS ``` # Add this package on intel x86 processor machines only pip install mkl-static mkl-include # Add these packages if torch.distributed is needed conda install pkg-config libuv ``` On Windows ``` pip install mkl-static mkl-include # Add these packages if torch.distributed is needed. # Distributed package support on Windows is a prototype feature and is subject to changes. conda install -c conda-forge libuv=1.51 ``` #### Install PyTorch On Linux If you're compiling for AMD ROCm then first run this command: ``` # Only run this if you're compiling for ROCm python tools/amd_build/build_amd.py ``` Install PyTorch ``` # the CMake prefix for conda environment export CMAKE_PREFIX_PATH="${CONDA_PREFIX:-'$(dirname $(which conda))/../'}:${CMAKE_PREFIX_PATH}" python -m pip install --no-build-isolation -v -e . # the CMake prefix for non-conda environment, e.g. Python venv # call following after activating the venv export CMAKE_PREFIX_PATH="${VIRTUAL_ENV}:${CMAKE_PREFIX_PATH}" ``` On macOS ``` python -m pip install --no-build-isolation -v -e . ``` On Windows If you want to build legacy python code, please refer to [Building on legacy code and CUDA](https://github.com/pytorch/pytorch/blob/main/CONTRIBUTING.md#building-on-legacy-code-and-cuda) CPU-only builds In this mode PyTorch computations will run on your CPU, not your GPU. ``` python -m pip install --no-build-isolation -v -e . ``` Note on OpenMP: The desired OpenMP implementation is Intel OpenMP (iomp). In order to link against iomp, you'll need to manually download the library and set up the building environment by tweaking `CMAKE_INCLUDE_PATH` and `LIB`. The instruction [here](https://github.com/pytorch/pytorch/blob/main/docs/source/notes/windows.rst#building-from-source) is an example for setting up both MKL and Intel OpenMP. Without these configurations for CMake, Microsoft Visual C OpenMP runtime (vcomp) will be used. CUDA based build In this mode PyTorch computations will leverage your GPU via CUDA for faster number crunching [NVTX](https://docs.nvidia.com/gameworks/content/gameworkslibrary/nvtx/nvidia_tools_extension_library_nvtx.htm) is needed to build PyTorch with CUDA. NVTX is a part of CUDA distributive, where it is called "Nsight Compute". To install it onto an already installed CUDA run CUDA installation once again and check the corresponding checkbox. Make sure that CUDA with Nsight Compute is installed after Visual Studio. Currently, VS 2017 / 2019, and Ninja are supported as the generator of CMake. If `ninja.exe` is detected in `PATH`, then Ninja will be used as the default generator, otherwise, it will use VS 2017 / 2019. If Ninja is selected as the generator, the latest MSVC will get selected as the underlying toolchain. Additional libraries such as [Magma](https://developer.nvidia.com/magma), [oneDNN, a.k.a. MKLDNN or DNNL](https://github.com/oneapi-src/oneDNN), and [Sccache](https://github.com/mozilla/sccache) are often needed. Please refer to the [installation-helper](https://github.com/pytorch/pytorch/tree/main/.ci/pytorch/win-test-helpers/installation-helpers) to install them. You can refer to the [build\_pytorch.bat](https://github.com/pytorch/pytorch/blob/main/.ci/pytorch/win-test-helpers/build_pytorch.bat) script for some other environment variables configurations ``` cmd :: Set the environment variables after you have downloaded and unzipped the mkl package, :: else CMake would throw an error as `Could NOT find OpenMP`. set CMAKE_INCLUDE_PATH={Your directory}\mkl\include set LIB={Your directory}\mkl\lib;%LIB% :: Read the content in the previous section carefully before you proceed. :: [Optional] If you want to override the underlying toolset used by Ninja and Visual Studio with CUDA, please run the following script block. :: "Visual Studio 2019 Developer Command Prompt" will be run automatically. :: Make sure you have CMake >= 3.12 before you do this when you use the Visual Studio generator. set CMAKE_GENERATOR_TOOLSET_VERSION=14.27 set DISTUTILS_USE_SDK=1 for /f "usebackq tokens=" %i in (`"%ProgramFiles(x86)%\Microsoft Visual Studio\Installer\vswhere.exe" -version [15^,17^) -products -latest -property installationPath`) do call "%i\VC\Auxiliary\Build\vcvarsall.bat" x64 -vcvars_ver=%CMAKE_GENERATOR_TOOLSET_VERSION% :: [Optional] If you want to override the CUDA host compiler set CUDAHOSTCXX=C:\Program Files (x86)\Microsoft Visual Studio\2019\Community\VC\Tools\MSVC\14.27.29110\bin\HostX64\x64\cl.exe python -m pip install --no-build-isolation -v -e . ``` Intel GPU builds In this mode PyTorch with Intel GPU support will be built. Please make sure [the common prerequisites](https://pypi.org/project/torch/#prerequisites) as well as [the prerequisites for Intel GPU](https://pypi.org/project/torch/#intel-gpu-support) are properly installed and the environment variables are configured prior to starting the build. For build tool support, `Visual Studio 2022` is required. Then PyTorch can be built with the command: ``` :: CMD Commands: :: Set the CMAKE_PREFIX_PATH to help find corresponding packages :: %CONDA_PREFIX% only works after `conda activate custom_env` if defined CMAKE_PREFIX_PATH ( set "CMAKE_PREFIX_PATH=%CONDA_PREFIX%\Library;%CMAKE_PREFIX_PATH%" ) else ( set "CMAKE_PREFIX_PATH=%CONDA_PREFIX%\Library" ) python -m pip install --no-build-isolation -v -e . ``` ##### Adjust Build Options (Optional) You can adjust the configuration of cmake variables optionally (without building first), by doing the following. For example, adjusting the pre-detected directories for CuDNN or BLAS can be done with such a step. On Linux ``` export CMAKE_PREFIX_PATH="${CONDA_PREFIX:-'$(dirname $(which conda))/../'}:${CMAKE_PREFIX_PATH}" CMAKE_ONLY=1 python setup.py build ccmake build # or cmake-gui build ``` On macOS ``` export CMAKE_PREFIX_PATH="${CONDA_PREFIX:-'$(dirname $(which conda))/../'}:${CMAKE_PREFIX_PATH}" MACOSX_DEPLOYMENT_TARGET=11.0 CMAKE_ONLY=1 python setup.py build ccmake build # or cmake-gui build ``` ### Docker Image #### Using pre-built images You can also pull a pre-built docker image from Docker Hub and run with docker v19.03+ ``` docker run --gpus all --rm -ti --ipc=host pytorch/pytorch:latest ``` Please note that PyTorch uses shared memory to share data between processes, so if torch multiprocessing is used (e.g. for multithreaded data loaders) the default shared memory segment size that container runs with is not enough, and you should increase shared memory size either with `--ipc=host` or `--shm-size` command line options to `nvidia-docker run`. #### Building the image yourself NOTE: Must be built with a docker version \> 18.06 The `Dockerfile` is supplied to build images with CUDA 11.1 support and cuDNN v8. You can pass `PYTHON_VERSION=x.y` make variable to specify which Python version is to be used by Miniconda, or leave it unset to use the default. ``` make -f docker.Makefile # images are tagged as docker.io/${your_docker_username}/pytorch ``` You can also pass the `CMAKE_VARS="..."` environment variable to specify additional CMake variables to be passed to CMake during the build. See [setup.py](https://pypi.org/project/torch/setup.py) for the list of available variables. ``` make -f docker.Makefile ``` ### Building the Documentation To build documentation in various formats, you will need [Sphinx](http://www.sphinx-doc.org/) and the pytorch\_sphinx\_theme2. Before you build the documentation locally, ensure `torch` is installed in your environment. For small fixes, you can install the nightly version as described in [Getting Started](https://pytorch.org/get-started/locally/). For more complex fixes, such as adding a new module and docstrings for the new module, you might need to install torch [from source](https://pypi.org/project/torch/#from-source). See [Docstring Guidelines](https://github.com/pytorch/pytorch/wiki/Docstring-Guidelines) for docstring conventions. ``` cd docs/ pip install -r requirements.txt make html make serve ``` Run `make` to get a list of all available output formats. If you get a katex error run `npm install katex`. If it persists, try `npm install -g katex` > \[!NOTE\] If you installed `nodejs` with a different package manager (e.g., `conda`) then `npm` will probably install a version of `katex` that is not compatible with your version of `nodejs` and doc builds will fail. A combination of versions that is known to work is `node@6.13.1` and `katex@0.13.18`. To install the latter with `npm` you can run `npm install -g katex@0.13.18` > \[!NOTE\] If you see a numpy incompatibility error, run: > ``` > pip install 'numpy<2' > ``` When you make changes to the dependencies run by CI, edit the `.ci/docker/requirements-docs.txt` file. #### Building a PDF To compile a PDF of all PyTorch documentation, ensure you have `texlive` and LaTeX installed. On macOS, you can install them using: ``` brew install --cask mactex ``` To create the PDF: 1. Run: ``` make latexpdf ``` This will generate the necessary files in the `build/latex` directory. 2. Navigate to this directory and execute: ``` make LATEXOPTS="-interaction=nonstopmode" ``` This will produce a `pytorch.pdf` with the desired content. Run this command one more time so that it generates the correct table of contents and index. > \[!NOTE\] To view the Table of Contents, switch to the Table of Contents view in your PDF viewer. ### Previous Versions Installation instructions and binaries for previous PyTorch versions may be found on [our website](https://pytorch.org/get-started/previous-versions). ## Getting Started Three pointers to get you started: - [Tutorials: get you started with understanding and using PyTorch](https://pytorch.org/tutorials/) - [Examples: easy to understand PyTorch code across all domains](https://github.com/pytorch/examples) - [The API Reference](https://pytorch.org/docs/) - [Glossary](https://github.com/pytorch/pytorch/blob/main/GLOSSARY.md) ## Resources - [PyTorch.org](https://pytorch.org/) - [PyTorch Tutorials](https://pytorch.org/tutorials/) - [PyTorch Examples](https://github.com/pytorch/examples) - [PyTorch Models](https://pytorch.org/hub/) - [Intro to Deep Learning with PyTorch from Udacity](https://www.udacity.com/course/deep-learning-pytorch--ud188) - [Intro to Machine Learning with PyTorch from Udacity](https://www.udacity.com/course/intro-to-machine-learning-nanodegree--nd229) - [Deep Neural Networks with PyTorch from Coursera](https://www.coursera.org/learn/deep-neural-networks-with-pytorch) - [PyTorch Twitter](https://twitter.com/PyTorch) - [PyTorch Blog](https://pytorch.org/blog/) - [PyTorch YouTube](https://www.youtube.com/channel/UCWXI5YeOsh03QvJ59PMaXFw) ## Communication - Forums: Discuss implementations, research, etc. [https://discuss.pytorch.org](https://discuss.pytorch.org/) - GitHub Issues: Bug reports, feature requests, install issues, RFCs, thoughts, etc. - Slack: The [PyTorch Slack](https://pytorch.slack.com/) hosts a primary audience of moderate to experienced PyTorch users and developers for general chat, online discussions, collaboration, etc. If you are a beginner looking for help, the primary medium is [PyTorch Forums](https://discuss.pytorch.org/). If you need a slack invite, please fill this form: <https://goo.gl/forms/PP1AGvNHpSaJP8to1> - Newsletter: No-noise, a one-way email newsletter with important announcements about PyTorch. You can sign-up here: <https://eepurl.com/cbG0rv> - Facebook Page: Important announcements about PyTorch. <https://www.facebook.com/pytorch> - For brand guidelines, please visit our website at [pytorch.org](https://pytorch.org/) ## Releases and Contributing Typically, PyTorch has three minor releases a year. Please let us know if you encounter a bug by [filing an issue](https://github.com/pytorch/pytorch/issues). We appreciate all contributions. If you are planning to contribute back bug-fixes, please do so without any further discussion. If you plan to contribute new features, utility functions, or extensions to the core, please first open an issue and discuss the feature with us. Sending a PR without discussion might end up resulting in a rejected PR because we might be taking the core in a different direction than you might be aware of. To learn more about making a contribution to PyTorch, please see our [Contribution page](https://pypi.org/project/torch/CONTRIBUTING.md). For more information about PyTorch releases, see [Release page](https://pypi.org/project/torch/RELEASE.md). ## The Team PyTorch is a community-driven project with several skillful engineers and researchers contributing to it. PyTorch is currently maintained by [Soumith Chintala](http://soumith.ch/), [Gregory Chanan](https://github.com/gchanan), [Dmytro Dzhulgakov](https://github.com/dzhulgakov), [Edward Yang](https://github.com/ezyang), [Alban Desmaison](https://github.com/albanD), [Piotr Bialecki](https://github.com/ptrblck) and [Nikita Shulga](https://github.com/malfet) with major contributions coming from hundreds of talented individuals in various forms and means. A non-exhaustive but growing list needs to mention: [Trevor Killeen](https://github.com/killeent), [Sasank Chilamkurthy](https://github.com/chsasank), [Sergey Zagoruyko](https://github.com/szagoruyko), [Adam Lerer](https://github.com/adamlerer), [Francisco Massa](https://github.com/fmassa), [Alykhan Tejani](https://github.com/alykhantejani), [Luca Antiga](https://github.com/lantiga), [Alban Desmaison](https://github.com/albanD), [Andreas Koepf](https://github.com/andreaskoepf), [James Bradbury](https://github.com/jekbradbury), [Zeming Lin](https://github.com/ebetica), [Yuandong Tian](https://github.com/yuandong-tian), [Guillaume Lample](https://github.com/glample), [Marat Dukhan](https://github.com/Maratyszcza), [Natalia Gimelshein](https://github.com/ngimel), [Christian Sarofeen](https://github.com/csarofeen), [Martin Raison](https://github.com/martinraison), [Edward Yang](https://github.com/ezyang), [Zachary Devito](https://github.com/zdevito). Note: This project is unrelated to [hughperkins/pytorch](https://github.com/hughperkins/pytorch) with the same name. Hugh is a valuable contributor to the Torch community and has helped with many things Torch and PyTorch. ## License PyTorch has a BSD-style license, as found in the [LICENSE](https://pypi.org/project/torch/LICENSE) file.
Shard	59 (laksa)
Root Hash	7813724874982801459
Unparsed URL	org,pypi!/project/torch/ s443