fix: standalone nvidia-modelopt[torch] command breaks Docker build

[andrewwhitecdw]

Problem

fix: standalone nvidia-modelopt[torch] command breaks Docker build

Fix

Replace:

    uv pip install onnxruntime-gpu onnx onnxscript pycuda cuda-python tensorrt-cu12 tensorrt-lean-cu12 tensorrt-dispatch-cu12 && nvidia-modelopt[torch] &&\

with:

    uv pip install onnxruntime-gpu onnx onnxscript pycuda cuda-python tensorrt-cu12 tensorrt-lean-cu12 tensorrt-dispatch-cu12 &&\

Files changed

• docker/dockerfile_cpp