[None][feat] Optimize GDN of Qwen3-Next/3.5; adds BF16 TRTLLM MoE#12557

rosenrodt wants to merge 9 commits intoNVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

rosenrodt:qwen3next-3_5-pyt-perfrosenrodt/TensorRT-LLM:qwen3next-3_5-pyt-perfCopy head branch name to clipboard