Run:ai Model Streamer is a Python package available for installation in PyPi. Installing the streamer is done in your Python project using pip or by adding runai-model-streamer to your requirements.txt file.
The Run:ai Streamer comes with precompiled C++ code compatible to run on any Linux machine. Installing the streamer with the pip command is supported for x86_64 processors.
To install the streamer, run:
pip install runai-model-streamerNote: To install a specific version of the package, use
runai-model-streamer==0.3.1. Visit our PyPi or GitHub repository for the list of available versions.
To stream models from object storage, run the following command as well:
pip install runai-model-streamer[s3]Warning: Make sure you install the S3 dependency in the same version of your runai-model-streamer by running
pip install runai-model-streamer[s3]==0.3.1.