[NPU] Add group norm support on NPU

[orangeH25]

Summary

This PR introduces a functional GroupNorm operator for Ascend NPU.

Key improvements:

• Fixes the runtime error grid should be less than 65536! and ub overflow that occurs when the original GPU-oriented liger-kernel GroupNorm implementation is executed on NPU.
• Adjusts the kernel launch and tiling strategy to comply with Ascend NPU execution constraints.
• Resolves numerical accuracy issues with PyTorch reference outputs.

While the current implementation is still slower than the HuggingFace implementation in end-to-end benchmarks, it provides a stable and functional GroupNorm path for Ascend NPU.

This PR mainly focuses on correctness and NPU compatibility. Further kernel-level optimizations will be explored in follow-up work.

Testing Done

• Hardware Type: Atlas 800I A2
• run make test to ensure correctness
• run make checkstyle to ensure code style
• run make test-convergence to ensure convergence

[orangeH25]

Hi @Tcc0403 , please take a look. Thanks!