[None][fix] Fix Qwen3 w4a8 model execution failure and add unit test (re-open of #14527) - #17196

[None][fix] Fix Qwen3 w4a8 model execution failure and add unit test (re-open of #14527)#17196

yuanjingx87 merged 5 commits into

NVIDIA:mainNVIDIA/TensorRT-LLM:mainfrom

rosenrodt:mixed-w4a8-quant-fixrosenrodt/TensorRT-LLM:mixed-w4a8-quant-fixCopy head branch name to clipboard