Cutting Carbon Footprint in AI Training by Optimization

Researchers from the University of Michigan have created an open-source optimization framework called Zeus that addresses the energy consumption issue in deep learning models. As the trend of using larger models with more parameters grows, the demand for energy to train these models is also increasing. Zeus seeks to solve this issue by identifying the […]

The post Cutting Carbon Footprint in AI Training by Optimization appeared first on MarkTechPost.

ai shorts ai training applications artificial intelligence carbon carbon footprint deep learning demand editors pick energy framework issue machine learning optimization researchers staff tech news technology training trend university university of michigan