OpenSTEF 4 Launch: A New Way of Building Energy...
summary
OpenSTEF 4 offers a modular Lego-block API that separates preprocessing, models, metalearning and backtesting so users swap pieces without reengineering; it adds a metalearning stack, a Kronos foundational model option (first supported open-source foundational TS model), a median forecaster for slow signals, strict multi-horizon handling, and an integrated backtesting workflow that reproduces production training/forecasting. Documentation generation runs on AWS with Amazon Bedrock docs pipeline, producing 12 pages in ~10 minutes and opening PRs for human review.
AI-written summary · about
chapters (5)
- 0:00Introduction to OpenSTEF 4
- 3:11Key improvements in version 4
- 8:46New automated documentation
- 14:44Upgrading from version 3 to 4
- 22:11Q&A session
related talks
25mThe Value of Open Source Short-term Energy Forecasting Demo...
11mOpenSTEF: Forecasting The Load On An Energy Grid
30mHow Open Source Helps a Grid Operator with the Challenges of...
32mAdopting Open-Source Tools for Time Series Forecasting
28mBuilding a Sustainable Future through Open Source
40mInstructLab: Applying Open Source Methods to Building and Trai...
43mMore Renewable Energy Into the Power Grid with Open Source
42mHow Open Source...