[Misc.] Add More Backends Support

[uniartisan]

This pull request aims to enhance the project's compatibility by adding preliminary support for multiple backends. Triton, a powerful framework, is well - known for its extensive support across a wide range of hardware backends, including AMD, NVIDIA, and Intel. Additionally, it has shown great potential in supporting certain Chinese - made hardware, such as GPUs from companies like MetaX, Moore Threads, and iluvatar. This PR takes a step forward in leveraging Triton's capabilities to integrate these diverse backends into our project.

Changes Made

Backend Detection: Implemented a mechanism to detect the availability of different backends at runtime. This includes checking for the presence of hardware from AMD, NVIDIA, and Intel, as well as the newly supported Chinese - made hardware from MetaX, Moore Threads, and iluvatar.

Testing

• Conducted a series of tests on various hardware platforms, including AMD, NVIDIA, and Intel GPUs.
• All existing test cases were re - run to ensure that the changes did not break any existing functionality.

By adding support for these multiple backends, we believe that FLA project will become more accessible and versatile, catering to a broader range of users and hardware configurations.