The latest version of Streamlit is 1.8.0. Streamlit is an open-source Python library that makes it simple to build custom machine learning and data science web apps.
The following are the major changes in the new release:
- Dataframe performance improvements
Performance enhancements for dataframes – The time required to parse dataframes has been drastically reduced in this update. A 150MB dataframe now goes from taking 3s to process on the frontend to 125ms, which is primarily the time spent by arrow converting the byte array into a table.
- ‘st.slider’ handles timezones better by removing backend timezone conversions
This update removes any backend timezone conversation and uses the UTC timezone in the frontend (with moments) instead of the browser timezone. As a result, DateTime instances are always displayed with their exact date and time.
- Design enhancements to our header
This update includes several design changes to the app header, such as styled top buttons, different hover states for buttons, a semi-opaque background for the top header, and the background removal behind the running man animation.
Streamlit recently agreed to be acquired by Snowflake for joining forces and pushing the boundaries of data science and data application development.
