Tensor Parallel Attention (AMD)
===============================

Tensor Parallel Attention layer for AMD GPUs.

See ``python/triton_dist/layers/amd/tp_attn.py`` for implementation details.