[TRTLLM-13321][feat] Extend rejection sampling to one-model speculative decoding (MTP/DRAFT_TARGET/PARD/DFLASH)#15775

zhaoyangwang-nvidia merged 8 commits into

NVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

zhaoyangwang-nvidia:rejection-one-model-speczhaoyangwang-nvidia/TensorRT-LLM:rejection-one-model-specCopy head branch name to clipboard