Tensors and Dynamic neural networks in Python with strong GPU acceleration

Overview

PyTorch Logo


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.

System 3.6 3.7 3.8
Linux CPU Build Status Build Status
Linux GPU Build Status Build Status
Windows CPU / GPU Build Status
Linux (ppc64le) CPU Build Status
Linux (ppc64le) GPU Build Status
Linux (aarch64) CPU Build Status Build Status Build Status

See also the ci.pytorch.org HUD.

More About 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).

Tensor illustration

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, math 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.

Dynamic graph

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 (TH, THC, THNN, THCUNN) 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 from binaries via Conda or pip wheels are on our website: https://pytorch.org

NVIDIA Jetson Platforms

Python wheels for NVIDIA's Jetson Nano, Jetson TX2, and Jetson AGX Xavier are available via the following URLs:

They require JetPack 4.2 and above, and @dusty-nv maintains them

From Source

If you are installing from source, you will need Python 3.6.2 or later and a C++14 compiler. Also, we highly recommend installing an Anaconda environment. You will get a high-quality BLAS library (MKL) and you get controlled dependency versions regardless of your Linux distro.

Once you have Anaconda installed, here are the instructions.

If you want to compile with CUDA support, install

If you want to disable CUDA support, export environment variable USE_CUDA=0. Other potentially useful environment variables may be found in setup.py.

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

Install Dependencies

Common

conda install numpy ninja pyyaml mkl mkl-include setuptools cmake cffi typing_extensions future six requests dataclasses

On Linux

# Add LAPACK support for the GPU if needed
conda install -c pytorch magma-cuda110  # or the magma-cuda* that matches your CUDA version from https://anaconda.org/pytorch/repo

On MacOS

# Add these packages if torch.distributed is needed
conda install pkg-config libuv

On Windows

# 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=1.39

Get the PyTorch Source

git clone --recursive https://github.com/pytorch/pytorch
cd pytorch
# if you are updating an existing checkout
git submodule sync
git submodule update --init --recursive

Install PyTorch

On Linux

export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname $(which conda))/../"}
python setup.py install

Note that if you are using Anaconda, you may experience an error caused by the linker:

build/temp.linux-x86_64-3.7/torch/csrc/stub.o: file not recognized: file format not recognized
collect2: error: ld returned 1 exit status
error: command 'g++' failed with exit status 1

This is caused by ld from Conda environment shadowing the system ld. You should use a newer version of Python that fixes this issue. The recommended Python version is 3.6.10+, 3.7.6+ and 3.8.1+.

On macOS

export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname $(which conda))/../"}
MACOSX_DEPLOYMENT_TARGET=10.9 CC=clang CXX=clang++ python setup.py install

Each CUDA version only supports one particular XCode version. The following combinations have been reported to work with PyTorch.

CUDA version XCode version
10.0 XCode 9.4
10.1 XCode 10.1

On Windows

Build with CPU

It's fairly easy to build with CPU. Visual Studio 2019 version 16.7.6 (MSVC toolchain version 14.27) or higher is recommended.

Build with CUDA

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 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.

CUDA, MSVC, and PyTorch versions are interdependent; please install matching versions from this table:

CUDA version Newest supported VS version PyTorch version
9.2 Visual Studio 2017 Update 5 (15.5) (_MSC_VER <= 1912) 0.4.1 ~ 1.5.1
10.1 Visual Studio 2019 (16.X) (_MSC_VER < 1930) 1.3.0 ~ 1.7.0
10.2 Visual Studio 2019 (16.X) (_MSC_VER < 1930) 1.5.0 ~ 1.7.0
11.0 Visual Studio 2019 (16.X) (_MSC_VER < 1930) 1.7.0

Note: There's a compilation issue in several Visual Studio 2019 versions since 16.7.1, so please make sure your Visual Studio 2019 version is not in 16.7.1 ~ 16.7.5

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

:: [Optional] If you want to build with the VS 2017 generator for old CUDA and PyTorch, please change the value in the next line to `Visual Studio 15 2017`.
:: Note: This value is useless if Ninja is detected. However, you can force that by using `set USE_NINJA=OFF`.
set CMAKE_GENERATOR=Visual Studio 16 2019

:: 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^,16^) -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 setup.py install
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))/../"}
python setup.py build --cmake-only
ccmake build  # or cmake-gui build

On macOS

export CMAKE_PREFIX_PATH=${CONDA_PREFIX:-"$(dirname $(which conda))/../"}
MACOSX_DEPLOYMENT_TARGET=10.9 CC=clang CXX=clang++ python setup.py build --cmake-only
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 support and cuDNN v7. 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

Building the Documentation

To build documentation in various formats, you will need Sphinx and the readthedocs theme.

cd docs/
pip install -r requirements.txt

You can then build the documentation by running make <format> from the docs/ folder. 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

Previous Versions

Installation instructions and binaries for previous PyTorch versions may be found on Our Website.

Getting Started

Three-pointers to get you started:

Resources

Communication

Releases and Contributing

PyTorch has a 90-day release cycle (major releases). 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.

The Team

PyTorch is a community-driven project with several skillful engineers and researchers contributing to it.

PyTorch is currently maintained by Adam Paszke, Sam Gross, Soumith Chintala and Gregory Chanan 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.

Issues
  • add frozen_pyyaml as a builtin library to torch::deploy

    add frozen_pyyaml as a builtin library to torch::deploy

    Summary: add frozen_pyyaml as a builtin library to torch::deploy

    Test Plan: unittests pass

    buck test mode/dev-nosan caffe2/torch/csrc/deploy/... -- --regex ".TestPyYAML."

    predictor also works

    buck run mode/dev-nosan mode/inplace hpc/torchrec/models/examples:sparsenn_packager -- package_path=/tmp/sparsenn_example_zyan.zip

    buck run mode/dev-nosan mode/inplace hpc/predictor:server -- --package_path /tmp/sparsenn_example_zyan.zip --n_gpu 1

    Differential Revision: D31852201

    fb-exported cla signed ciflow/default 
    opened by zyan0 4
  • [LTC] Add support for non-structured in-place operator variants

    [LTC] Add support for non-structured in-place operator variants

    Summary: This commit teached the LTC code-gen engine about in-place operator variants and generated zero_ for validation.

    Test Plan: lazy_tensor_core/test/cpp/build/test_ptltc --gtest_filter=AtenLtcTsTensorTest.TestZeroInPlace

    Fixes #65576.

    cla signed ciflow/default 
    opened by alanwaketan 4
  • [SR] Factor operator() implementation into separate function

    [SR] Factor operator() implementation into separate function

    Summary: Using explicit template instantiations in D31659973 was a bad idea. The problem is that the lvalue instantiation was for a const vector of IValue, meaning that if you tried to pass SR a non-const vector of arguments, the linker would fail to find the symbol.

    The reason we didn't catch this in D31659973 was because predictor always passes a const reference anyways. But we should fix this to prevent unexpected problems in the future.

    Test Plan: buck test caffe2/benchmarks/static_runtime/...

    Differential Revision: D31873406

    oncall: jit fb-exported cla signed ciflow/default 
    opened by mikeiovine 3
  • [ONNX] Fix new_full and full_like for Python 3.9

    [ONNX] Fix new_full and full_like for Python 3.9

    Previously new_full would fail with errors like: TypeError: only integer tensors of a single element can be converted to an index

    And full_like would trigger warnings like: DeprecationWarning: an integer is required (got type float). Implicit conversion to integers using __int__ is deprecated, and may be removed in a future version of Python.

    Co-Authored-By: Bowen Bao [email protected]

    open source cla signed ciflow/default 
    opened by garymm 2
  • Make fb::sigrid_hash_compute_multipler_shift return a std::tuple<int64_t, int64_t>

    Make fb::sigrid_hash_compute_multipler_shift return a std::tuple

    Summary: Makes sigrid_hash_compute_multipler_shift return a tuple instead of a tensor and modifies functions that depends on it.

    Test Plan:

    buck test //caffe2/benchmarks/static_runtime/fb:test_fb_operators
    

    Benchmarks: local - P463676888 local_ro - P463676984

    Differential Revision: D31867186

    oncall: jit fb-exported cla signed ciflow/default 
    opened by pilleye 3
  • [ONNX] Support opset 15

    [ONNX] Support opset 15

    null

    open source cla signed ciflow/default 
    opened by garymm 2
  • Replace lazy_tensors::Span with std::vector in ir_util

    Replace lazy_tensors::Span with std::vector in ir_util

    Summary: Replace lazy_tensors::Span with std::vector as a preparation to land ir_util in master. If we use ArrayRef, we need to deal with the const qualifier problem and add nasty const_cast here and there.

    cla signed ciflow/default 
    opened by desertfire 4
  • Request) draw_keypoints function

    Request) draw_keypoints function

    There is no draw_keypoints function in torchvision.utils. Could you make this function in future version.

    torchvision.utils functions are as follows

    .. currentmodule:: torchvision.utils

    .. autofunction:: make_grid

    .. autofunction:: save_image

    .. autofunction:: draw_bounding_boxes

    .. autofunction:: draw_segmentation_masks

    opened by epsilon-deltta 0
  • [quant][fx] add pass to duplicate dequant nodes with multi use

    [quant][fx] add pass to duplicate dequant nodes with multi use

    Stack from ghstack:

    • -> #67118

    Summary: Fixes a bug in the reference pattern support for nn.Linear when the same quantized input is shared across multiple Linear nodes.

    This PR adds a pass to duplicate the dequant nodes for each use so that for a case like

    x -> quant -> dequant -> linear1 - quant1
                         |
                       linear2 - quant2
    

    We duplicate the dequant nodes

    x -> quant -> dequant1 -> linear1 - quant1
                |
              dequant2-> linear2 - quant2
    

    So that we can match each pattern in the loweing step

    We also add a pass to remove the extra/duplicate dequant nodes that may be leftover from the above pass if we don't lower them based on pattern match

    Test Plan: python test/test_quantization.py test_ref_pattern_multi_use

    Reviewers:

    Subscribers:

    Tasks:

    Tags:

    Differential Revision: D31873511

    cla signed ciflow/default fx 
    opened by supriyar 3
  • Add FSDP tests to verify forward overlap and memory usage

    Add FSDP tests to verify forward overlap and memory usage

    Stack from ghstack:

    • -> #67117
    • #66904

    Add FSDP tests to verify forward overlap and memory usage

    Differential Revision: D31845629

    cc @pietern @mrshenli @pritamdamania87 @zhaojuanmao @satgera @rohan-varma @gqchen @aazzolini @osalpekar @jiayisuse @SciPioneer @H-Huang

    oncall: distributed cla signed ciflow/default 
    opened by zhaojuanmao 2
Releases(v1.10.0)
The Incredible PyTorch: a curated list of tutorials, papers, projects, communities and more relating to PyTorch.

This is a curated list of tutorials, projects, libraries, videos, papers, books and anything related to the incredible PyTorch. Feel free to make a pu

Ritchie Ng 8.5k Oct 22, 2021
A comprehensive list of published machine learning applications to cosmology

ml-in-cosmology This github attempts to maintain a comprehensive list of published machine learning applications to cosmology, organized by subject ma

George Stein 211 Oct 15, 2021
A curated list of neural network pruning resources.

A curated list of neural network pruning and related resources. Inspired by awesome-deep-vision, awesome-adversarial-machine-learning, awesome-deep-learning-papers and Awesome-NAS.

Yang He 1.3k Oct 15, 2021
PyTorch implementation of neural style transfer algorithm

neural-style-pt This is a PyTorch implementation of the paper A Neural Algorithm of Artistic Style by Leon A. Gatys, Alexander S. Ecker, and Matthias

null 594 Oct 23, 2021
Dense matching library based on PyTorch

Dense Matching A general dense matching library based on PyTorch. For any questions, issues or recommendations, please contact Prune at [email protected]

Prune Truong 191 Oct 21, 2021
Geometric Deep Learning Extension Library for PyTorch

Documentation | Paper | Colab Notebooks | External Resources | OGB Examples PyTorch Geometric (PyG) is a geometric deep learning extension library for

Matthias Fey 12.8k Oct 22, 2021
General purpose GPU compute framework for cross vendor graphics cards (AMD, Qualcomm, NVIDIA & friends)

General purpose GPU compute framework for cross vendor graphics cards (AMD, Qualcomm, NVIDIA & friends). Blazing fast, mobile-enabled, asynchronous and optimized for advanced GPU data processing usecases. Backed by the Linux Foundation.

The Kompute Project 581 Oct 19, 2021
Annotated notes and summaries of the TensorFlow white paper, along with SVG figures and links to documentation

TensorFlow White Paper Notes Features Notes broken down section by section, as well as subsection by subsection Relevant links to documentation, resou

Sam Abrahams 438 Sep 10, 2021
A curated list of neural rendering resources.

Awesome-of-Neural-Rendering A curated list of neural rendering and related resources. Please feel free to pull requests or open an issue to add papers

Zhiwei ZHANG 25 Oct 4, 2021
DRIFT is a tool for Diachronic Analysis of Scientific Literature.

About DRIFT is a tool for Diachronic Analysis of Scientific Literature. The application offers user-friendly and customizable utilities for two modes:

Rajaswa Patil 88 Oct 13, 2021
Tensors and Dynamic neural networks in Python with strong GPU acceleration

PyTorch is a Python package that provides two high-level features: Tensor computation (like NumPy) with strong GPU acceleration Deep neural networks b

null 51.6k Oct 24, 2021
Tensors and Dynamic neural networks in Python with strong GPU acceleration

PyTorch is a Python package that provides two high-level features: Tensor computation (like NumPy) with strong GPU acceleration Deep neural networks b

null 46.1k Feb 13, 2021
All course materials for the Zero to Mastery Deep Learning with TensorFlow course.

All course materials for the Zero to Mastery Deep Learning with TensorFlow course.

Daniel Bourke 1.7k Oct 24, 2021
A curated list of resources for Image and Video Deblurring

A curated list of resources for Image and Video Deblurring

Subeesh Vasu 1.1k Oct 23, 2021
Must-read Papers on Physics-Informed Neural Networks.

PINNpapers Contributed by IDRL lab. Introduction Physics-Informed Neural Network (PINN) has achieved great success in scientific computing since 2017.

IDRL 30 Oct 15, 2021
PyTorch implementation of DeepDream algorithm

neural-dream This is a PyTorch implementation of DeepDream. The code is based on neural-style-pt. Here we DeepDream a photograph of the Golden Gate Br

null 96 Oct 15, 2021
An unofficial styleguide and best practices summary for PyTorch

A PyTorch Tools, best practices & Styleguide This is not an official style guide for PyTorch. This document summarizes best practices from more than a

IgorSusmelj 1.2k Oct 22, 2021
Official Implementation of Neural Splines

Neural Splines: Fitting 3D Surfaces with Inifinitely-Wide Neural Networks This repository contains the official implementation of the CVPR 2021 (Oral)

Francis Williams 21 Oct 15, 2021
Channel Pruning for Accelerating Very Deep Neural Networks (ICCV'17)

Channel Pruning for Accelerating Very Deep Neural Networks (ICCV'17)

Yihui He 1k Oct 12, 2021