[Feature] Enhance hadamard transform

> Status:
> 1. support `deterministic_hadamard_matrix` and `deterministic_hadamard_matrix`
> 3. support inference with transformers with a triton mxfp4 kernel
> 4. support rtn and autoroud tuning (iters>0)
>
> Todo:
> 1. align the implementation of [SpinQuant: LLM Quantization with Learned Rotations](https://arxiv.org/pdf/2405.16406), especially R2/R3, and implement structured fused
> 2. support online transform for vllm inference
> 3. support nvfp4
> 4. verify multi-cards

_Originally posted by @lkk12014402 in [#1323](https://github.com/intel/auto-round/issues/1323#issuecomment-4039450678)_