Home > Servers > PowerEdge Components > White Papers > Developing and Deploying Vision AI with Dell and NVIDIA Metropolis > Step 5: Run the Inference TensorRT DS-Triton
When you are within the container image, run the inference with the refence application:
$ cd /workspace/Deepstream_6.0_Triton/samples/configs/deepstream-app-trtis
$ deepstream-app -c source1_primary_retinanet_resnet18_retrained.txt