Customizing ML Deployment with Triton Inference Server Python Backend

In this video we explore how you can bring custom packages and dependencies to Triton via the Python Backend. This is especially handy when you have custom pre/post processing as part of larger model inference workflow. Video Resources - Sample Code: https://github.com/RamVegiraju/triton-inference-server-examples/tree/master/conda-python-backend - Python Backend: https://github.com/triton-inference-server/python_backend Timestamps 0:00 Introduction 1:30 What is a Model Server 6:35 Why Triton? 8:34 When Python Backend 10:01 Hands-On #sagemaker #nvidia #tritoninferenceserver #modelserving #machinelearning #transformers #huggingface