notebook

OLive Notebook Demo

This repository provides quick demo and visualization of how to deploy and use OLive in Jupyter notebook.

Prerequisites

Install Docker.

Windows

build.sh
pip install docker
pip install jupyter

Linux

sh build.sh
pip install docker
pip install jupyter

Start Notebook

Type the below in command line, and then choose onnx-pipeline.ipynb to use the notebook.

jupyter notebook

Convert model to ONNX and Performance test tool

This command is used to convert model from major model frameworks to onnx and then performance test that onnx model.

IMPORTANT: Any path in the notebook must be under the current directory (/cmd-tool).

Supported frameworks are - tensorflow, pytorch, cntk, coreml, keras and scikit-learn.

Use Jupyter Notebook and see onnx-pipeline.ipynb