Kairntech Sherpa building blocks for Streamlit apps
This package contains utilities for visualizing and building interactive Kairntech Sherpa-powered apps with Streamlit. It includes various building blocks you can use in your own Streamlit app, like visualizers for named entities, text classification, and more.
You can simply pip install sherpa-streamlit.
The package includes building blocks that call into Streamlit and set up all the required elements for you. You can either use the individual components directly and combine them with other elements in your app, or call the visualize function to embed the whole visualizer.
Put the following example code in a file.
# streamlit_app.py
import sherpa_streamlit
default_text = "Sundar Pichai is the CEO of Google."
sherpa_streamlit.visualize(default_text)
You can then run your app with streamlit run streamlit_app.py. The app should pop up in your web browser.
You will need to install flit (for building the package) and tox (for orchestrating testing and documentation building):
python3 -m pip install flit tox
Clone the repository:
git clone https://github.com/oterrier/sherpa_streamlit
You can run the full test suite against all supported versions of Python (3.8) with:
tox
You can build the HTML documentation with:
tox -e docs
The built documentation is available at `docs/_build/index.html.