Project Limit Request: onnxruntime-gpu - 100 GiB

[sanaa-hamel-microsoft]

Project URL

https://pypi.org/project/onnxruntime-gpu

Does this project already exist?

• Yes

New limit

100 GiB

Update issue title

• I have updated the title.

Which indexes

PyPI

About the project

ONNX Runtime is a performance-focused scoring engine for Open Neural Network Exchange (ONNX) models. For more information on ONNX Runtime, please see aka.ms/onnxruntime or the Github project. This project hosts the CUDA accelerated version of the project.

The project is widely used by nVidia, AMD, Intel, Qualcomm, and Microsoft. We have been publishing for at least 8 years to PyPi. We will continue to publish going into the future. We understand our releases are quite large (~2.5G), and will work towards reducing their size in the future.

How large is each release?

~3GiB per release, with ~250MiB per wheel in a release.

6x ABIs ([3.11, 3.14], and 3.13t & 3.14t on linux)
2x Platforms (Windows, ManyLinux 2_28)
1x Variant(s) per platform (x86_64)

How frequently do you make a release?

Roughly once a month.

Code of Conduct

• I agree to follow the PSF Code of Conduct

[github-actions]

Thanks for filing.

There are currently 18 other open limit request issues ahead of this one.

PyPI support is one paid staffer, plus volunteers, plus whatever part-time hours other PSF staff have to give. Often that's zero. If you'd like to help fund more staffing, the PSF takes donations at https://www.python.org/psf/donations/. Either way, someone will get to your request eventually.