Bump the python-packages group across 1 directory with 2 updates #249

dependabot · 2024-10-28T05:08:32Z

Updates the requirements on torch and numpy to permit the latest version.
Updates torch to 2.5.0

Release notes

PyTorch 2.5.0 Release, SDPA CuDNN backend, Flex Attention

PyTorch 2.5 Release Notes

Highlights

Backwards Incompatible Change

Deprecations

New Features

Improvements

Bug fixes

Performance

Documentation

Developers

Security

Highlights

We are excited to announce the release of PyTorch® 2.5! This release features a new CuDNN backend for SDPA, enabling speedups by default for users of SDPA on H100s or newer GPUs. As well, regional compilation of torch.compile offers a way to reduce the cold start up time for torch.compile by allowing users to compile a repeated nn.Module (e.g. a transformer layer in LLM) without recompilations. Finally, TorchInductor CPP backend offers solid performance speedup with numerous enhancements like FP16 support, CPP wrapper, AOT-Inductor mode, and max-autotune mode. This release is composed of 4095 commits from 504 contributors since PyTorch 2.4. We want to sincerely thank our dedicated community for your contributions. As always, we encourage you to try these out and report any issues as we improve 2.5. More information about how to get started with the PyTorch 2-series can be found at our Getting Started page. As well, please check out our new ecosystem projects releases with TorchRec and TorchFix.

Beta Prototype

CuDNN backend for SDPA FlexAttention

torch.compile regional compilation without recompilations Compiled Autograd

TorchDynamo added support for exception handling & MutableMapping types Flight Recorder

TorchInductor CPU backend optimization Max-autotune Support on CPU with GEMM Template

TorchInductor on Windows

FP16 support on CPU path for both eager mode and TorchInductor CPP backend

Autoload Device Extension

Enhanced Intel GPU support

*To see a full list of public feature submissions click here.

BETA FEATURES

[Beta] CuDNN backend for SDPA

The cuDNN "Fused Flash Attention" backend was landed for torch.nn.functional.scaled_dot_product_attention. On NVIDIA H100 GPUs this can provide up to 75% speed-up over FlashAttentionV2. This speedup is enabled by default for all users of SDPA on H100 or newer GPUs.

[Beta] torch.compile regional compilation without recompilations

Regional compilation without recompilations, via torch._dynamo.config.inline_inbuilt_nn_modules which default to True in 2.5+. This option allows users to compile a repeated nn.Module (e.g. a transformer layer in LLM) without recompilations. Compared to compiling the full model, this option can result in smaller compilation latencies with 1%-5% performance degradation compared to full model compilation.

See the tutorial for more information.

[Beta] TorchInductor CPU backend optimization

This feature advances Inductor’s CPU backend optimization, including CPP backend code generation and FX fusions with customized CPU kernels. The Inductor CPU backend supports vectorization of common data types and all Inductor IR operations, along with the static and symbolic shapes. It is compatible with 10000 both Linux and Windows OS and supports the default Python wrapper, the CPP wrapper, and AOT-Inductor mode.

Additionally, it extends the max-autotune mode of the GEMM template (prototyped in 2.5), offering further performance gains. The backend supports various FX fusions, lowering to customized kernels such as oneDNN for Linear/Conv operations and SDPA. The Inductor CPU backend consistently achieves performance speedups across three benchmark suites—TorchBench, Hugging Face, and timms—outperforming eager mode in 97.5% of the 193 models tested.

PROTOTYPE FEATURES

[Prototype] FlexAttention

We've introduced a flexible API that enables implementing various attention mechanisms such as Sliding Window, Causal Mask, and PrefixLM with just a few lines of idiomatic PyTorch code. This API leverages torch.compile to generate a fused FlashAttention kernel, which eliminates extra memory allocation and achieves performance comparable to handwritten implementations. Additionally, we automatically generate the backwards pass using PyTorch's autograd machinery. Furthermore, our API can take advantage of sparsity in the attention mask, resulting in significant improvements over standard attention implementations.

For more information and examples, please refer to the official blog post and Attention Gym.

... (truncated)

Changelog

Sourced from torch's changelog.

Releasing PyTorch

Release Compatibility Matrix

Release Cadence

General Overview

Frequently Asked Questions

Cutting a release branch preparations

Cutting release branches

pytorch/pytorch

pytorch/builder / PyTorch domain libraries

Making release branch specific changes for PyTorch

Making release branch specific changes for domain libraries

Running Launch Execution team Core XFN sync

Drafting RCs (https://github.com/pytorch/pytorch/blob/main/Release Candidates) for PyTorch and domain libraries

Release Candidate Storage

Release Candidate health validation

Cherry Picking Fixes

How to do Cherry Picking

Cherry Picking Reverts

Preparing and Creating Final Release candidate

Promoting RCs to Stable

Additional Steps to prepare for release day

Modify release matrix

Open Google Colab issue

Patch Releases

Patch Release Criteria

Patch Release Process

Patch Release Process Description

Triage

Issue Tracker for Patch releases

Building a release schedule / cherry picking

Building Binaries / Promotion to Stable

Hardware / Software Support in Binary Build Matrix

Python

Accelerator Software

Special support cases

Operating Systems

Submitting Tutorials

Special Topics

Updating submodules for a release

Triton dependency for the release

Release Compatibility Matrix

Following is the Release Compatibility Matrix for PyTorch releases:

... (truncated)

Commits

32f585d [Release only] use triton 3.1.x from pypi (#137895)
417a076 [split build] move periodic split builds into own concurrency group (#135510)...
119e734 [RELEASE-ONLY CHANGES] Fix dependency on filesystem on Linux (#137242)
783a6a4 [MPS] Add regression test for fft.fftfreq (#137215)
5375201 [MPS] Add missing dispatch to rshift.Tensor (#137212)
1de132e [MPS] Fix 5D+ reductions over negative dimentions (#137211)
0b1b609 [NCCL] Don't override waitUntilInitialized's setting of `comm->initialized_...
0b45af9 Fix addmm silent correctness on aarch64 (#137208)
1a0b166 [ONNX] Add assertion nodes to ignoring list (#137214)
3a541ef Clarify that libtorch API is C++17 compatible (#137206)
Additional commits viewable in compare view

Updates numpy to 2.1.2

Release notes

Sourced from numpy's releases.

2.1.2 (Oct 5, 2024)

NumPy 2.1.2 Release Notes

NumPy 2.1.2 is a maintenance release that fixes bugs and regressions discovered after the 2.1.1 release.

The Python versions supported by this release are 3.10-3.13.

Contributors

A total of 11 people contributed to this release. People with a "+" by their names contributed a patch for the first time.

Charles Harris

Chris Sidebottom

Ishan Koradia +

João Eiras +

Katie Rust +

Marten van Kerkwijk

Matti Picus

Nathan Goldbaum

Peter Hawkins

Pieter Eendebak

Slava Gorloff +

Pull requests merged

A total of 14 pull requests were merged for this release.

#27333: MAINT: prepare 2.1.x for further development

#27400: BUG: apply critical sections around populating the dispatch cache

#27406: BUG: Stub out get_build_msvc_version if distutils.msvccompiler...

#27416: BUILD: fix missing include for std::ptrdiff_t for C++23 language...

#27433: BLD: pin setuptools to avoid breaking numpy.distutils

#27437: BUG: Allow unsigned shift argument for np.roll

#27439: BUG: Disable SVE VQSort

#27471: BUG: rfftn axis bug

#27479: BUG: Fix extra decref of PyArray_UInt8DType.

#27480: CI: use PyPI not scientific-python-nightly-wheels for CI doc...

#27481: MAINT: Check for SVE support on demand

#27484: BUG: initialize the promotion state to be weak

#27501: MAINT: Bump pypa/cibuildwheel from 2.20.0 to 2.21.2

#27506: BUG: avoid segfault on bad arguments in ndarray.__array_function__

Checksums

MD5
4aae28b7919b126485c1aaccee37a6ba  numpy-2.1.2-cp310-cp310-macosx_10_9_x86_64.whl
172614423a82ef73d8752ad8a59cbafc  numpy-2.1.2-cp310-cp310-macosx_11_0_arm64.whl

... (truncated)

Changelog

Sourced from numpy's changelog.

This is a walkthrough of the NumPy 2.1.0 release on Linux, modified for building with GitHub Actions and cibuildwheels and uploading to the anaconda.org staging repository for NumPy <https://anaconda.org/multibuild-wheels-staging/numpy>_. The commands can be copied into the command line, but be sure to replace 2.1.0 by the correct version. This should be read together with the :ref:general release guide <prepare_release>.

Facility preparation

Before beginning to make a release, use the requirements/*_requirements.txt files to ensure that you have the needed software. Most software can be installed with pip, but some will require apt-get, dnf, or whatever your system uses for software. You will also need a GitHub personal access token (PAT) to push the documentation. There are a few ways to streamline things:

Git can be set up to use a keyring to store your GitHub personal access token. Search online for the details.

You can use the keyring app to store the PyPI password for twine. See the online twine documentation for details.

Prior to release

Add/drop Python versions

When adding or dropping Python versions, three files need to be edited:

.github/workflows/wheels.yml # for github cibuildwheel

.travis.yml # for cibuildwheel aarch64 builds

setup.py # for classifier and minimum version check.

Make these changes in an ordinary PR against main and backport if necessary. Using the BLD: prefix (build label) for the commit summary will cause the wheel builds to be run so that the changes will be tested, We currently release wheels for new Python versions after the first Python rc once manylinux and cibuildwheel support it. For Python 3.11 we were able to release within a week of the rc1 announcement.

Backport pull requests

Changes that have been marked for this release must be backported to the maintenance/2.1.x branch.

Update 2.1.0 milestones

... (truncated)

Commits

f5afe3d Merge pull request #27507 from charris/prepare-2.1.2
6b9ef48 REL: Prepare for the NumPy 2.1.2 release [wheel build]
6d85a24 MAINT: Pin setuptools for Python < 3.12 [wheel build]
7f327d8 Merge pull request #27506 from charris/backport-27503
09e9cd9 BUG: avoid segfault on bad arguments in ndarray.array_function
06b4793 Merge pull request #27501 from charris/backport-27495
364efb5 MAINT: Bump pypa/cibuildwheel from 2.21.1 to 2.21.2
df0e261 Merge pull request #27484 from ngoldbaum/use-weak-default
fa61bc6 Merge pull request #27481 from charris/backport-27472
d6d06ae Merge pull request #27480 from charris/backport-27475
Additional commits viewable in compare view

Dependabot will resolve any conflicts with this PR as long as you don't alter it yourself. You can also trigger a rebase manually by commenting @dependabot rebase.

Dependabot commands and options

You can trigger Dependabot actions by commenting on this PR:

@dependabot rebase will rebase this PR
@dependabot recreate will recreate this PR, overwriting any edits that have been made to it
@dependabot merge will merge this PR after your CI passes on it
@dependabot squash and merge will squash and merge this PR after your CI passes on it
@dependabot cancel merge will cancel a previously requested merge and block automerging
@dependabot reopen will reopen this PR if it is closed
@dependabot close will close this PR and stop Dependabot recreating it. You can achieve the same result by closing it manually
@dependabot show <dependency name> ignore conditions will show all of the ignore conditions of the specified dependency
@dependabot ignore <dependency name> major version will close this group update PR and stop Dependabot creating any more for the specific dependency's major version (unless you unignore this specific dependency's major version or upgrade to it yourself)
@dependabot ignore <dependency name> minor version will close this group update PR and stop Dependabot creating any more for the specific dependency's minor version (unless you unignore this specific dependency's minor version or upgrade to it yourself)
@dependabot ignore <dependency name> will close this group update PR and stop Dependabot creating any more for the specific dependency (unless you unignore this specific dependency or upgrade to it yourself)
@dependabot unignore <dependency name> will remove all of the ignore conditions of the specified dependency
@dependabot unignore <dependency name> <ignore condition> will remove the ignore condition of the specified dependency and ignore conditions

Updates the requirements on [torch](https://github.com/pytorch/pytorch) and [numpy](https://github.com/numpy/numpy) to permit the latest version. Updates `torch` to 2.5.0 - [Release notes](https://github.com/pytorch/pytorch/releases) - [Changelog](https://github.com/pytorch/pytorch/blob/main/RELEASE.md) - [Commits](pytorch/pytorch@v2.0.1...v2.5.0) Updates `numpy` to 2.1.2 - [Release notes](https://github.com/numpy/numpy/releases) - [Changelog](https://github.com/numpy/numpy/blob/main/doc/RELEASE_WALKTHROUGH.rst) - [Commits](numpy/numpy@v0.2.0...v2.1.2) --- updated-dependencies: - dependency-name: torch dependency-type: direct:production dependency-group: python-packages - dependency-name: numpy dependency-type: direct:production dependency-group: python-packages ... Signed-off-by: dependabot[bot] <support@github.com>

IsaevIlya · 2024-10-29T12:03:22Z

PyTorch 2.5.0 contains bug in cpuinfo pytorch/pytorch#138333
We need to wait until 2.5.1 will become available

dependabot · 2024-10-29T12:03:24Z

This pull request was built based on a group rule. Closing it will not ignore any of these versions in future pull requests.

To ignore these dependencies, configure ignore rules in dependabot.yml

dependabot bot added dependencies Pull requests that update a dependency file python Pull requests that update Python code labels Oct 28, 2024

dependabot bot had a problem deploying to integration-tests October 28, 2024 05:08 Failure

dependabot bot force-pushed the dependabot/pip/s3torchconnector/python-packages-dddb8c5026 branch from 5a5c82a to 1e0bcd0 Compare October 29, 2024 11:51

dependabot bot had a problem deploying to integration-tests October 29, 2024 11:52 Error

IsaevIlya closed this Oct 29, 2024

dependabot bot deleted the dependabot/pip/s3torchconnector/python-packages-dddb8c5026 branch October 29, 2024 12:03

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Bump the python-packages group across 1 directory with 2 updates #249

Bump the python-packages group across 1 directory with 2 updates #249

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Beta	Prototype
CuDNN backend for SDPA	FlexAttention
torch.compile regional compilation without recompilations	Compiled Autograd
TorchDynamo added support for exception handling & MutableMapping types	Flight Recorder
TorchInductor CPU backend optimization	Max-autotune Support on CPU with GEMM Template
	TorchInductor on Windows
	FP16 support on CPU path for both eager mode and TorchInductor CPP backend
	Autoload Device Extension
	Enhanced Intel GPU support

Bump the python-packages group across 1 directory with 2 updates #249

Bump the python-packages group across 1 directory with 2 updates #249

Uh oh!

Conversation

Uh oh!

PyTorch 2.5.0 Release, SDPA CuDNN backend, Flex Attention

PyTorch 2.5 Release Notes

Highlights

BETA FEATURES

[Beta] CuDNN backend for SDPA

[Beta] torch.compile regional compilation without recompilations

[Beta] TorchInductor CPU backend optimization

PROTOTYPE FEATURES

[Prototype] FlexAttention

Releasing PyTorch

Release Compatibility Matrix

2.1.2 (Oct 5, 2024)

NumPy 2.1.2 Release Notes

Contributors

Pull requests merged

Checksums

MD5

Facility preparation

Prior to release

Add/drop Python versions

Backport pull requests

Update 2.1.0 milestones

Uh oh!

Uh oh!

Uh oh!

Uh oh!