This repository provides quick demo and visualization of how to deploy and use OLive in Jupyter notebook.
Install Docker.
build.sh
pip install docker
pip install jupyter
sh build.sh
pip install docker
pip install jupyter
Type the below in command line, and then choose onnx-pipeline.ipynb to use the notebook.
jupyter notebook
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