Python-Pytorch v1.10.1: Tensors and Dynamic neural networks in Python with strong GPU acceleration

icon
Latest Release: v1.10.1

This release is meant to fix the following issues (regressions / silent correctness):

  • torch.nn.cross_entropy silently incorrect in PyTorch 1.10 on CUDA on non-contiguous inputs #67167
  • channels_last significantly degrades accuracy #67239
  • Potential strict aliasing rule violation in bitwise_binary_op (on ARM/NEON) #66119
  • torch.get_autocast_cpu_dtype() returns a new dtype #65786
  • Conv2d grad bias gets wrong value for bfloat16 case #68048

The release tracker should contain all relevant pull requests related to this release as well as links to related issues

Source code(tar.gz)
Source code(zip)
pytorch-v1.10.1.tar.gz(105.54 MB)

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 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

On Linux

# Add LAPACK support for the GPU if needed
conda install -c pytorch magma-cuda102  # or [ magma-cuda101 | magma-cuda100 | magma-cuda92 ] depending on your cuda version

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

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

At least Visual Studio 2017 Update 3 (version 15.3.3 with the toolset 14.11) and NVTX are needed.

If the version of Visual Studio 2017 is higher than 15.4.5, installing of "VC++ 2017 version 15.4 v14.11 toolset" is strongly recommended.
If the version of Visual Studio 2017 is lesser than 15.3.3, please update Visual Studio 2017 to the latest version along with installing "VC++ 2017 version 15.4 v14.11 toolset".
There is no guarantee of the correct building with VC++ 2017 toolsets, others than version 15.4 v14.11.
"VC++ 2017 version 15.4 v14.11 toolset" might be installed onto already installed Visual Studio 2017 by running its installation once again and checking the corresponding checkbox under "Individual components"/"Compilers, build tools, and runtimes".

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. Be sure that CUDA with Nsight Compute is installed after Visual Studio 2017.

Currently VS 2017, VS 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.
If Ninja is selected as the generator, the latest MSVC which is newer than VS 2015 (14.0) will get selected as the underlying toolchain if you have Python > 3.5, otherwise VS 2015 will be selected so you'll have to activate the environment. If you use CMake <= 3.14.2 and has VS 2019 installed, then even if you specify VS 2017 as the generator, VS 2019 will get selected as the generator.

CUDA and MSVC have strong version dependencies, so even if you use VS 2017 / 2019, you will get build errors like nvcc fatal : Host compiler targets unsupported OS. For this kind of problem, please install the corresponding VS toolchain in the table below and then you can either specify the toolset during activation (recommended) or set CUDAHOSTCXX to override the cuda host compiler (not recommended if there are big version differences).

CUDA version Newest supported VS version
9.2 Visual Studio 2017 Update 5 (15.5) (_MSC_VER <= 1912)
10.0 Visual Studio 2017 (15.X) (_MSC_VER < 1920)
10.1 Visual Studio 2019 (16.X) (_MSC_VER < 1930)
cmd

:: [Optional] If you want to build with VS 2019 generator, please change the value in the next line to `Visual Studio 16 2019`.
:: 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 15 2017

:: 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 2017 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=14.11
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\2017\Enterprise\VC\Tools\MSVC\14.11.25503\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.

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

  • 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, one-way email newsletter with important announcements about pytorch. You can sign-up here: https://eepurl.com/cbG0rv
  • for brand guidelines, please visit our website at pytorch.org

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.

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 in the Torch community and has helped with many things Torch and PyTorch.

License

PyTorch is BSD-style licensed, as found in the LICENSE file.

Comments

  • DOC: Initialize numpy compatibility note
    DOC: Initialize numpy compatibility note

    Jan 23, 2022

    Fixes #48628, #46829.

    This will need to be revised when #59790 is merged in.

    @rgommers

    open source ciflow/default 
    Reply
  • torch.distributions.categorical.Categorical does not work with 0 batch size
    torch.distributions.categorical.Categorical does not work with 0 batch size

    Jan 23, 2022

    🐛 Describe the bug

    Initializing a Categorical with logits that have a batch dimension of 0 raises an error. I would expect this to work and result in probabilities/entropies/... with batch dimension 0.

    Python 3.8.10 (default, Nov 26 2021, 20:14:08)
    [GCC 9.3.0] on linux
    Type "help", "copyright", "credits" or "license" for more information.
    >>> from torch.distributions.categorical import Categorical
    >>> import torch
    >>> Categorical(logits=torch.zeros((2, 5)))
    Categorical(logits: torch.Size([2, 5]))
    >>> Categorical(logits=torch.zeros((0, 5)))
    Traceback (most recent call last):
      File "<stdin>", line 1, in <module>
      File "/home/clemens/.cache/pypoetry/virtualenvs/incubator-Qe8CM38i-py3.8/lib/python3.8/site-packages/torch/distributions/categorical.py", line 64, in __init__
        super(Categorical, self).__init__(batch_shape, validate_args=validate_args)
      File "/home/clemens/.cache/pypoetry/virtualenvs/incubator-Qe8CM38i-py3.8/lib/python3.8/site-packages/torch/distributions/distribution.py", line 53, in __init__
        valid = constraint.check(value)
      File "/home/clemens/.cache/pypoetry/virtualenvs/incubator-Qe8CM38i-py3.8/lib/python3.8/site-packages/torch/distributions/constraints.py", line 206, in check
        result = result.reshape(result.shape[:result.dim() - self.reinterpreted_batch_ndims] + (-1,))
    RuntimeError: cannot reshape tensor of 0 elements into shape [0, -1] because the unspecified dimension size -1 can be any value and is ambiguous
    

    Versions

    Collecting environment information...
    PyTorch version: 1.10.1+cu102
    Is debug build: False
    CUDA used to build PyTorch: 10.2
    ROCM used to build PyTorch: N/A
    
    OS: Ubuntu 20.04.3 LTS (x86_64)
    GCC version: (Ubuntu 9.3.0-17ubuntu1~20.04) 9.3.0
    Clang version: 10.0.0-4ubuntu1
    CMake version: Could not collect
    Libc version: glibc-2.31
    
    Python version: 3.8.10 (default, Nov 26 2021, 20:14:08)  [GCC 9.3.0] (64-bit runtime)
    Python platform: Linux-5.4.0-92-generic-x86_64-with-glibc2.29
    Is CUDA available: True
    CUDA runtime version: Could not collect
    GPU models and configuration:
    GPU 0: NVIDIA GeForce RTX 2080 Ti
    GPU 1: NVIDIA GeForce RTX 2080 Ti
    GPU 2: NVIDIA GeForce RTX 2080 Ti
    
    Nvidia driver version: 465.19.01
    cuDNN version: /usr/lib/x86_64-linux-gnu/libcudnn.so.7.6.5
    HIP runtime version: N/A
    MIOpen runtime version: N/A
    
    Versions of relevant libraries:
    [pip3] msgpack-numpy==0.4.7.1
    [pip3] mypy==0.910
    [pip3] mypy-extensions==0.4.3
    [pip3] numpy==1.21.5
    [pip3] torch==1.10.1
    [pip3] torch-scatter==2.0.9
    [conda] msgpack-numpy             0.4.7.1                  pypi_0    pypi
    [conda] mypy                      0.910                    pypi_0    pypi
    [conda] mypy-extensions           0.4.3                    pypi_0    pypi
    [conda] numpy                     1.21.2                   pypi_0    pypi
    [conda] torch                     1.10.0                   pypi_0    pypi
    [conda] torch-scatter             2.0.9                    pypi_0    pypi
    
    Reply
  • Use of Python round() inbuilt function inside Functional.CenterCrop() NonDeterministic
    Use of Python round() inbuilt function inside Functional.CenterCrop() NonDeterministic

    Jan 23, 2022

    🐛 Describe the bug

    Barring the padding process, CenterCrop is implemented with these lines: crop_top = int(round((image_height - crop_height) / 2.0)) crop_left = int(round((image_width - crop_width) / 2.0)) return crop(img, crop_top, crop_left, crop_height, crop_width)

    Note that round(11/2)=6 and round(9/2)=4, i.e., python round() selects floor() sometimes, and sometimes ceil(). This causes crop_top and crop_left coordinates to be non-deterministic (depending on input sizes).

    Remedy: Just remove the round() since you already have int() outside (change it to floor() to avoid further confusion)

    Versions

    Collecting environment information... PyTorch version: 1.9.1+cu111 Is debug build: False CUDA used to build PyTorch: 11.1 ROCM used to build PyTorch: N/A

    OS: Red Hat Enterprise Linux release 8.5 (Ootpa) (x86_64) GCC version: Could not collect Clang version: Could not collect CMake version: version 3.20.2 Libc version: glibc-2.3.4

    Python version: 3.6.8 (default, Sep 9 2021, 07:49:02) [GCC 8.5.0 20210514 (Red Hat 8.5.0-3)] (64-bit runtime) Python platform: Linux-4.18.0-348.2.1.el8_5.x86_64-x86_64-with-redhat-8.5-Ootpa Is CUDA available: True CUDA runtime version: Could not collect GPU models and configuration: GPU 0: Quadro RTX 6000 Nvidia driver version: 470.57.02 cuDNN version: Could not collect HIP runtime version: N/A MIOpen runtime version: N/A

    Versions of relevant libraries: [pip3] numpy==1.19.5 [pip3] torch==1.9.1+cu111 [pip3] torchvision==0.10.1+cu111 [conda] Could not collect

    Reply
  • Kill the test_torch.py mixin and creates test_scatter_gather_ops
    Kill the test_torch.py mixin and creates test_scatter_gather_ops

    Jan 23, 2022

    Per title.

    Also annotates test_torch.py with additional cleanup tasks and adds empty sample inputs to elementwise unary and binary OpInfos.

    module: scatter & gather ops cla signed ciflow/default 
    Reply
  • Back out
    Back out "Move upgraders from python to cpp"

    Jan 24, 2022

    Summary: Original commit changeset: 713c54fbbb2b

    Original Phabricator Diff: D33402543 (https://github.com/pytorch/pytorch/commit/c9bd1c60ed9f854b56c27c0eb51dcda2962d287e)

    Test Plan: sandcastle

    Reviewed By: cccclai

    Differential Revision: D33735253

    oncall: jit fb-exported cla signed ciflow/default 
    Reply
  • Exit once there's an environment error
    Exit once there's an environment error

    Jan 24, 2022

    Narrow the scope of #69730. Once there's an error, stop the script. Since it's a random error, it most likely has something with the environment.

    Let's see the stat.

    open source cla signed ciflow/default 
    Reply
  • Implement value mappings for vulkanInsertPrePackedOps(). Test vulkanInsertPrePackedOps().
    Implement value mappings for vulkanInsertPrePackedOps(). Test vulkanInsertPrePackedOps().

    Jan 24, 2022

    Description:

    1. Same as title.
    2. Refactor test_graph_rewrite_passes.py for increased modularity. Add tests for vulkanInsertPrePackedOps() to this file.

    Test plan:

    1. pip3 install cmake pyyaml typing_extensions numpy expecttest hypothesis monkeytype
    2. python3 test/test_jit.py TestGraphRewritePasses
    oncall: jit cla signed ciflow/default 
    Reply
  • torch.onnx.export creates model with TopK node that may be invalid
    torch.onnx.export creates model with TopK node that may be invalid

    Jan 24, 2022

    🐛 Describe the bug

    The ONNX model exported from torchvision.models.detection.ssdlite320_mobilenet_v3_large may fail due to an invalid TopK node. This appears to be dependent on the input. Possibly the TopK export needs something to limit the 'K' value based on the input size as the ONNX spec requires that to be less than or equal to the number of inputs.

    The below repro will convert the model and run with ONNX. If zeros are used as input onnx model execution will fail.

    import numpy as np
    import onnxruntime as ort
    import torch
    import torchvision
    import urllib
    
    from torch.onnx import TrainingMode
    from torchvision import transforms
    from PIL import Image
    
    ssdlite = torchvision.models.detection.ssdlite320_mobilenet_v3_large(pretrained=True)
    ssdlite = ssdlite.eval()
    
    try:
        urllib.URLopener().retrieve("https://github.com/pytorch/hub/raw/master/images/dog.jpg", "dog.jpg")
    except:
        urllib.request.urlretrieve("https://github.com/pytorch/hub/raw/master/images/dog.jpg", "dog.jpg")
    
    input_image = Image.open("dog.jpg")
    preprocess = transforms.Compose([
        transforms.Resize(352),
        transforms.CenterCrop(320),
        transforms.ToTensor(),
        transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225]),
    ])
    input_tensor = preprocess(input_image)
    input_batch = input_tensor.unsqueeze(0)  # create a mini-batch as expected by the model
    
    with torch.no_grad():
        torch.onnx.export(ssdlite, input_batch, "ssdlite320_mobilenet_v3_large.onnx",
                          input_names=['image'], output_names=['boxes', 'scores', 'labels'],
                          opset_version=12,
                          do_constant_folding=False,
                          training=TrainingMode.EVAL,
                          export_params=True,
                          keep_initializers_as_inputs=False
                          )
    
    print("Running model with pytorch")
    torch_results = ssdlite(input_batch)
    print(torch_results)
    
    s = ort.InferenceSession("ssdlite320_mobilenet_v3_large.onnx")
    print("Running ONNX model with dog.jpg")
    np_input = input_batch.detach().cpu().numpy()
    ort_results = s.run(None, {'image': np_input})
    print(ort_results)
    
    print("Running ONNX model with zeros")
    np_zeros_input = np.zeros((1, 3, 320, 320), np.float32)
    zeros_results = s.run(None, {'image': np_zeros_input})
    print(zeros_results)
    

    Error from the second last line in the python

    onnxruntime.capi.onnxruntime_pybind11_state.Fail: [ONNXRuntimeError] : 1 : FAIL : Non-zero status code returned while running TopK node. Name:'TopK_2118' Status Message: k argument [300] should not be greater than specified axis dim value [183]
    

    FWIW there are some other warnings during export. As the model produces the same results as torch when dog.jpg is used as input I assume they can be ignored. Including here for completeness.

    C:\Users\s\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\torchvision\ops\boxes.py:156: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
      boxes_x = torch.min(boxes_x, torch.tensor(width, dtype=boxes.dtype, device=boxes.device))
    
    C:\Users\s\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\torchvision\ops\boxes.py:158: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
      boxes_y = torch.min(boxes_y, torch.tensor(height, dtype=boxes.dtype, device=boxes.device))
    
    C:\Users\s\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\torchvision\models\detection\transform.py:291: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
      torch.tensor(s, dtype=torch.float32, device=boxes.device)
    
    C:\Users\s\AppData\Local\Packages\PythonSoftwareFoundation.Python.3.9_qbz5n2kfra8p0\LocalCache\local-packages\Python39\site-packages\torchvision\models\detection\transform.py:292: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).
      / torch.tensor(s_orig, dtype=torch.float32, device=boxes.device)
    
    WARNING: ComputeShapeFromReshape(), shape_ratio overflows, skip shape inference.
    
    WARNING: The shape inference of prim::Constant type is missing, so it may result in wrong shape inference for the exported graph. Please consider adding it in symbolic function.
    
    

    Versions

    Collecting environment information... PyTorch version: 1.11.0.dev20220114+cpu Is debug build: False CUDA used to build PyTorch: Could not collect ROCM used to build PyTorch: N/A

    OS: Microsoft Windows 11 Enterprise GCC version: Could not collect Clang version: Could not collect CMake version: version 3.22.1 Libc version: N/A

    Python version: 3.9.10 (tags/v3.9.10:f2f3f53, Jan 17 2022, 15:14:21) [MSC v.1929 64 bit (AMD64)] (64-bit runtime) Python platform: Windows-10-10.0.22000-SP0 Is CUDA available: False CUDA runtime version: 11.5.119 GPU models and configuration: GPU 0: Quadro P620 Nvidia driver version: 496.49 cuDNN version: C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v11.5\bin\cudnn_ops_train64_8.dll HIP runtime version: N/A MIOpen runtime version: N/A

    Versions of relevant libraries: [pip3] numpy==1.21.5 [pip3] torch==1.11.0.dev20220114+cpu [pip3] torchaudio==0.11.0.dev20220119+cpu [pip3] torchvision==0.12.0.dev20220119+cpu [conda] Could not collect

    Reply
  • Fix RNN modules with inputs shapes containing-0 in CUDA
    Fix RNN modules with inputs shapes containing-0 in CUDA

    Jan 24, 2022

    We found a discrepancy between cpu & CUDA when using RNN modules where input shapes containing 0s would cause an invalid configuration argument error in CUDA (kernel grid size is 0), while returning a valid tensor in CPU cases.

    A reproducer:

    import torch
    
    x = torch.zeros((5, 0, 3)).cuda()
    gru = torch.nn.GRU(input_size=3, hidden_size=4).to("cuda")
    gru(x)
    

    Run with CUDA_LAUNCH_BLOCKING=1 set.

    cc @ngimel @albanD

    open source cla signed ciflow/default 
    Reply
  • Libtorch dlls delayed loading
    Libtorch dlls delayed loading

    Jan 24, 2022

    🐛 Describe the bug

    Pytorch Release ver. 1.10.0 Microsoft Visual Studio Community 2019 The c++ application we are developing has to load the pytorch dlls in a delayed way so that the end user is not forced to install them if he does not need them. The problem is that we can’t link the pytorch dlls in a delayed way (/DELAYLOAD). There are two different problems:

    1. linker error If I add the c10.dll or torch_cpu.dll in the project properties → linker → input → “Delay Loaded Dlls” I have the following linker errors:
    • ''' fatal error LNK1194: cannot delay-load ‘c10.dll’ due to import of data symbol '"__declspec (dllimport) struct std :: atomic <struct c10 :: impl :: DeviceGuardImplInterface const *> * c10 :: impl :: device_guard_impl_registry "(_imp? device_guard_impl_registry @ impl @ c10 @@ 3PAU? $ atomic @ PEBUDeviceGuardImplInterface @ impl @ c10 @@@ std @@ A) '; link without /DELAYLOAD:c10.dll '''
    • ''' fatal error LNK1194: cannot delay-load ‘torch_cpu.dll’ due to import of data symbol ‘"__declspec(dllimport) struct torch::enumtype::kValid const torch::kValid" ([email protected]@@[email protected]@B)’; link without /DELAYLOAD:torch_cpu.dll '''
    1. anomaly in runtime If I add the dll torch_cuda_cu.dll in the project properties → linker → input → “Delay Loaded Dlls” it happens that the CPU is always used and the GPU is ignored. The function torch::cuda::is_available() returns always false.

    Thanks for your help

    Versions

    Pytorch Release ver. 1.10.0 Microsoft Visual Studio Community 2019

    Reply