## Feature Request **Why do you want this feature?** TensorRT-LLM boasted some good performance gain https://github.com/NVIDIA/TensorRT-LLM