The Flower team is excited to announce the release of Flower 1.41 stable, packed with new updates!
Flower is a friendly framework for collaborative AI and data science. It makes novel approaches such as federated learning, federated evaluation, federated analytics, and fleet learning accessible to a wide audience of researchers and engineers.
Thanks to our contributors
We would like to give our special thanks to all the contributors who made the new version of Flower possible (in git shortlog order):
Charles Beauville, Chong Shen Ng, Daniel Nata Nugraha, Heng Pan, Javier, Taner Topal
What's new?
-
Use GitHub, Notion, Slack, Attio, and Google connectors through MCP (#8328, #8339, #8343, #8362, #8382)
GitHub, Notion, Slack, and Attio connectors now integrate through MCP, with tools discovered dynamically from each connected account. New Google Drive, Gmail, and Google Calendar connectors use Google's remote MCP servers with OAuth/PKCE and token refresh.
-
Store and reuse compacted AgentApp contexts (experimental) (#8350, #8363, #8364, #8367, #8370, #8371, #8376, #8377, #8380)
AgentApps can now filter run-series events and read or store compacted contexts through AgentSession.events. Compacted contexts are persisted per run series and exposed through authenticated Runtime API endpoints with newest-first paginated reads.
-
Lay the groundwork for faster AgentApp responses (#8375, #8388)
-
General improvements (#8341, #8345, #8349, #8356, #8357, #8358, #8359, #8360, #8361, #8365, #8366, #8372, #8373, #8374, #8378, #8384, #8387, #8389, #8394)
As always, many parts of the Flower framework and quality infrastructure were improved and updated.