Fix logging issue for unsupported torch.compile devices

[rghosh08]

Fixes #137285

Description

This PR addresses the issue of insufficient logging when torch.compile is not supported on devices with CUDA capability below 7.0. Instead of exiting the program with a simple message, the code now logs detailed information about the device capabilities and reasons for the lack of support for torch.compile. It also ensures that the program continues to execute, showcasing the TORCH_LOGS logging system even when the function is not compiled.

Changes made:

• Added detailed logging to explain why torch.compile is not supported on devices with lower CUDA capabilities.
• Introduced a fallback mechanism where the function runs uncompiled if the device doesn't support compilation.
• Consolidated the device selection logic for cleaner code.
• Ensured the program runs and demonstrates TORCH_LOGS regardless of device capabilities.
• Refactored code to adhere to PyTorch standards and PEP 8 linting guidelines.

---

Checklist

• The issue that is being fixed is referred to in the description (see above "Fixes #137285").
• Only one issue is addressed in this pull request.
• Labels from the issue that this PR is fixing are added to this pull request.
• No unnecessary issues are included in this pull request.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/tutorials/3077

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[facebook-github-bot]

Hi @rghosh08!

Thank you for your pull request and welcome to our community.

Action Required

In order to merge any pull request (code, docs, etc.), we require contributors to sign our Contributor License Agreement, and we don't seem to have one on file for you.

Process

In order for us to review and merge your suggested changes, please sign at https://code.facebook.com/cla. If you are contributing on behalf of someone else (eg your employer), the individual CLA may not be sufficient and your employer may need to sign the corporate CLA.

Once the CLA is signed, our tooling will perform checks and validations. Afterwards, the pull request will be tagged with CLA signed. The tagging process may take up to 1 hour after signing. Please give it that time before contacting us about it.

If you have received this in error or have any questions, please contact us at [email protected]. Thanks!

[facebook-github-bot]

Thank you for signing our Contributor License Agreement. We can now accept your code for this (and any) Meta Open Source project. Thanks!

[Tharusha-Lekamge]

Small suggestion - Link the PR to the issue. Currently the issue has no linked PRs

[svekars]

I don't understand why everything is being deleted here.

[mlazos]

@rghosh08 I don't think this PR makes sense, as logging is explicitly meant for torch.compile at the moment. If you run without torch.compile nothing of note is printed so it is kind of a useless tutorial. When I ran this on my machine (not compiling the function) and get:

INFO:__main__:CUDA Device Capability: (8, 0)
INFO:__main__:Device supports torch.compile.
INFO:__main__:Running the function without compilation...
===================Dynamo Tracing=========================
===================Traced Graph=========================
===================Fusion Decisions=========================
===================Output Code=========================
============================================

I also don't understand why so much of the tutorial aspect of the script is getting removed. Can you explain further?