TRT_Triton Hands_On (2022/04/21 AI Developer Meetup)

Set up environments

docker run --gpus '"device=0"' -it --rm -p 8887:8887 -v $(pwd):/hands_on nvcr.io/nvidia/pytorch:22.03-py3

cd /hands_on

jupyter notebook --ip 0.0.0.0 --port 8887

Additional Resources

• NGC

• DLI Deploying a Model for Inference at Production Scale

• TRT Quick Start

• TRT Documentation

• TF-TRT

• Torch-TRT

• Triton Server

• Triton Client

• Triton Model Analyzer

• GTC On-Demand - To watch more deep dive sessions