Skip to content

Install tensorflow-macos dependency conditionally #5124

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Oct 19, 2022
Merged

Install tensorflow-macos dependency conditionally #5124

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 6 additions & 2 deletions setup.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -132,7 +132,8 @@
"py7zr",
"rarfile>=4.0",
"s3fs>=2021.11.1", # aligned with fsspec[http]>=2021.11.1
"tensorflow>=2.3,!=2.6.0,!=2.6.1",
"tensorflow>=2.3,!=2.6.0,!=2.6.1; sys_platform != 'darwin' or platform_machine != 'arm64'",
"tensorflow-macos; sys_platform == 'darwin' and platform_machine == 'arm64'",
Copy link
Collaborator

@severo severo Oct 17, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

BTW I think and platform_machine == 'arm64'" is not needed, as I was able to install tensorflow-macos on my MacOS with an Intel chip

$ sysctl -a | grep Intel
machdep.cpu.vendor: GenuineIntel
machdep.cpu.brand_string: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz

See https://github.com/huggingface/datasets-server/blob/9b346a79c5e3f51719fb2717683c4e5f86291201/services/worker/pyproject.toml#L34-L35

But maybe it's better to limit the exception cases and only use the fork for arm64.

Copy link
Member Author

@albertvillanova albertvillanova Oct 17, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@severo I'm not an expert on this subject. But it appears they recommend tensorflow-macos also for Intel x86-64: https://developer.apple.com/metal/tensorflow-plugin/

So you think I can safely remove and platform_machine == 'arm64'? Or better keeping to reduce the scope?

Copy link
Collaborator

@severo severo Oct 17, 2022

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I really don't know, I just wanted to share my recent experience with the same problem, and use the same solution if you opt for a reduced scope.

"torch",
"torchaudio<0.12.0",
"soundfile",
Expand Down Expand Up @@ -172,7 +173,10 @@
"audio": AUDIO_REQUIRE,
"vision": VISION_REQURE,
"apache-beam": ["apache-beam>=2.26.0"],
"tensorflow": ["tensorflow>=2.2.0,!=2.6.0,!=2.6.1"],
"tensorflow": [
"tensorflow>=2.2.0,!=2.6.0,!=2.6.1; sys_platform != 'darwin' or platform_machine != 'arm64'",
"tensorflow-macos; sys_platform == 'darwin' and platform_machine == 'arm64'",
],
"tensorflow_gpu": ["tensorflow-gpu>=2.2.0,!=2.6.0,!=2.6.1"],
"torch": ["torch"],
"s3": [
Expand Down