[platform] refactor: refine detection and platform-specific installs

[Wangmerlyn]

No description provided.

[gemini-code-assist]

Summary of Changes

Hello @Wangmerlyn, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request significantly refines the system's ability to detect and adapt to different computing platforms, specifically enhancing support for ROCm alongside existing CUDA and CPU environments. The changes aim to provide a more reliable and user-friendly experience by improving automatic detection, offering manual override options, and clarifying installation procedures across various hardware configurations.

Highlights

• Enhanced Platform Detection: The core platform detection logic has been refactored to more robustly identify CUDA, ROCm, and CPU environments, including improved checks using torch.cuda.is_available() and torch.version.hip.
• ROCm Support Integration: Explicit support for ROCm has been added, allowing for optional rocm-smi integration via a new rocm extra dependency in pyproject.toml.
• Documentation Updates: The README.md and docs/ files have been updated to reflect the new ROCm support and provide clear, platform-specific installation instructions for CUDA, ROCm, and CPU-only setups.
• Platform Detection Overrides and Caching: A new environment variable KEEP_GPU_PLATFORM allows users to manually override platform detection, and a caching mechanism has been implemented to optimize repeated calls to get_platform().
• Comprehensive Testing: A new test file tests/utilities/test_platform_manager.py has been introduced to thoroughly validate the updated platform detection logic, including environment variable overrides, caching behavior, and specific CUDA/ROCm detection scenarios.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request refactors the platform detection logic to better distinguish between CUDA, ROCm, and CPU environments, which is a great improvement. The introduction of caching and an environment variable override for the platform are also valuable additions. The accompanying documentation updates and new tests are thorough. My review includes a few suggestions for improvement: there's a contradiction between the docstring and implementation for the CUDA detection logic that should be resolved, I found a piece of unreachable code in the platform manager, and the names of a couple of tests could be clarified to better reflect what they are testing. Overall, this is a solid refactoring. Addressing these points will enhance the code's clarity and correctness.

[gemini-code-assist]

This return False statement is unreachable. The preceding try...except block for the pynvml check will always return a boolean value (True on success, False on exception), causing the function to exit before this line is reached. This line should be removed.