[None][feat] Support sparse mqa/gqa attention#12470

lfr-0531 merged 12 commits into

NVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

heyuhhh:user/yuhangh/add_sparse_attention_supportheyuhhh/TensorRT-LLM:user/yuhangh/add_sparse_attention_supportCopy head branch name to clipboard