Software

See the table below for our open-source software offerings.

For the most current listing of NREL-developed open-source software, see U.S. Department of Energy Office of Scientific and Technical Information's DOE CODE database.

Commercial Licensing

Explore more than 250 copyrighted software tools available for licensing via the Department of Energy's Lab Partnering Service.

BuildingsBench: A Benchmark for Universal Building Load Forecasting

Jun 1, 2023, 00:00 AM

BuildingsBench is a Python software framework that streamlines the training and comparison of generalized machine learning models for universal building load forecasting. The current approach involves crafting specialized forecasting models for individual buildings, which is time-consuming and costly. This software provides a platform for large-scale pretraining using the synthetic Buildings-900K dataset, representing the entire U.S. building stock, and subsequently benchmarking performance across diverse buildings, geographic regions, weather patterns, and more.

Title : BuildingsBench: A Benchmark for Universal Building Load Forecasting

Url : https://github.com/NREL/BuildingsBench

Software Id :

Categories :