# Serve a Tensorflow CV model on an Amazon SageMaker Real-time Endpoint running on a GPU with Triton Server from NVIDIA.

The Jupyter notebook `triton-cv-sme-tensorflow-backend.ipynb` shows you how to deploy a Tensorflow CV model using the NVIDIA Triton Inference Server on an Amazon SageMaker real-time endpoint. 

## Steps to run the notebook
1. Launch the notebook in SageMaker studio. This notebook was tested with the Data Science Python3 kernel. 
1. The notebook is self-contained with no local dependencies. Further documentation on the workings of the code can be found within the notebook itself.