Installation
Kern AI refinery is available as open-source, so you can install it on your local machine. The open-source version can be used with only a single user and therefore doesn't contain any user management.
If you enjoy working with it, please consider giving us a GitHub star ⭐.
You can install the application by either cloning our repository or by installing it with pip. Due to the containerization of all services, one prerequisite is that you need to have docker installed on your local machine.
From repository (recommended)
As the open-source version is hosted on GitHub, just clone the repository.
git clone https://github.com/code-kern-ai/refinery.git
After that, if you want to start the application you can use the start scripts we developed for all operating systems.
cd refinery
./start
To stop refinery, you can use the stop scripts that shut down the containers in a controlled way.
./stop
With pip
Refinery is also registered on pypi, which means you can install it using pip. Do not get confused with the displayed version number on pypi as we implemented a separate updating service that lets up update refinery without pushing a new version to pypi.
pip install kern-refinery
Starting the service then pulls all the required containers and connects them, while stopping it will remove everything safely. Since it was installed with pip, these commands are operating system independent.
refinery start
Accessing the UI
After starting the app, you can visit localhost:4455
access the UI. You now just have to register a user and after that will be redirected to the home screen.
Only register a single user! While in theory you could log out of the registered user and create a new one, we strongly advise against it as the open-source version is not built to handle multiple users and might break projects when doing so. If you need multiple users, please look at our managed version.
From here, we really recommend trying the quick start of this documentation, which gives you a good overview of refinery without taking too much of your time.