Clara Deploy Python Client

An intuitive python 3 package to develop applications with NVIDIA Clara Deploy. Utilize the clients within the nvidia_clara package to manage jobs, pipelines, payloads, and models. Each client has an associated set of objects which are defined in seperate 'types' modules (also can be found in nvidia_clara). Look at the examples below to learn more on each client to get started!

Additional Resources to Learn More on Clara Deploy

• NVIDIA Clara Overview Homepage
• NVIDIA Clara Deploy SDK User Guide

Client Prerequisites

• Python 3.6 or higher
• Clara Deploy 0.7.0 or higher

Pypi Packages Needed

• Grpcio
• Protobuf
• Grpcio-Testing (For Running Pytests)
• Pytest (For Running Pytests)

Getting Started

Package Installation

Installing from source repository

$ git clone --recursive git@github.com:kubernetes-client/python.git
$ cd ./clara-platform-python-client
$ python3 -m pip install .

Installing directly from Pypi

$ python3 -m pip install nvidia-clara-client

Clara Client Guides

• Jobs : Learn to start and manage Clara jobs
• Pipelines : Learn to create and manage Clara pipelines
• Payloads : Learn to create, upload, download, and manage Clara payloads

Full Example(s) Running Pipeline

• Spleen Segmentation Pipeline

Running Pytests

Only for developing with source repository

$ pip3 install grpcio-testing
$ pip3 install pytest
$ export PYTHONPATH="${PYTHONPATH}:<INSERT PATH TO /clara-platform-python-client>"
$ pytest <INSERT PATH TO /clara-platform-python-client>