[None][fix] Enable kv cache reuse for VLMs with special tokens#7685

chang-l wants to merge 2 commits intoNVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

chang-l:kvcache_reuse_special_tokenschang-l/TensorRT-LLM:kvcache_reuse_special_tokensCopy head branch name to clipboard