src/optimum/nvidia/utils/patching.py (1 line):
	- line 7: # TODO: Remove this once available natively in TensorRT-LLM - `optimize(network)` lowers predictive performance for Whisper.