bricks is our content library that provides off-the-shelf natural language modules like language translation, sentiment classification, keyword extraction, etc. which you can use in your project directly or in refinery to enrich your text data.
The library is available at bricks.kern.ai and its purpose is to make it easier for you to build better products using its distinguished modular automations stacked together into one collection. This is not a library in the sense that you can pip install it, but have to copy-paste the code from the online platform.
It consists of modules that can be categorised into three sections - classifiers, extractors and generators that are accessible from the left panel below the “home” icon.
bricks - i.e. classifiers, extractors, or generators - are sub-categorised into “Python function”, “Active learner”, or “Premium”.
The “Python function” badge is associated with the modules which can be executed in the bricks environment as well as in refinery without requiring any external changes in the code.
“Active learner” badge is associated with modules that are active transfer learning models. They can only be used in your project in refinery and not in bricks directly, since they require a dataset with some manual labelling.
“Premium” badge is associated with the modules that require an external API key/token. You can contact us at [email protected] if you require an API key. Due to confidentiality, we have only chosen the examples for this documentation which are marked with “Python function”.
In bricks, you can also see at the end of the modules page that you can add your own module.
If you come up with an idea for either a classifier, extractor, or generator, you can open a new issue on GitHub and if we like the idea, we will implement the module for you, or you can implement the module by yourself by following the steps shown in this video.
In order to access the bricks modules into refinery, the proper way is to use the bricks in refinery integration.
bricks comes with the following features.
Every brick is open-source and modular, such that you can customize the code itself immediately. Start with a great baseline, and adapt it to your needs.
Integration to refinery
Every brick is designed to be integrated into refinery. You can search within refinery for a brick, and then add it to your project within a few clicks.
We aim to add new bricks every week. If you have a great idea for a brick, please let us know. We are happy to help you implement it.
Just as refinery, bricks is open-source. You can find the code on GitHub. We are also happy to accept contributions.
Great, you're now set up with an API client and have made your first request to the API. Here are a few links that might be handy as you venture further into the Protocol API: