[TRTLLM-Gen Fmha] add optimized trtllm-gen decode kernels for high throughput + speculative decoding#2265

yzh119 merged 6 commits intoflashinfer-ai:mainflashinfer-ai/flashinfer:mainfrom

PerkzZheng:user/perkzz/trtllm-gen-groups-tokens-headsPerkzZheng/flashinfer:user/perkzz/trtllm-gen-groups-tokens-headsCopy head branch name to clipboard