OLMo: A Truly Open Language Model - Keynote

Explore the groundbreaking OLMo project in this 17-minute keynote address delivered by Hanna Hajishirzi, Senior Director of NLP Research at the Allen Institute for AI and Associate Professor at the University of Washington. Delve into the importance of open-source language models in the face of increasing commercialization and proprietary development. Learn about OLMo, a competitive and truly open language model designed to accelerate scientific study and innovation in the field of natural language processing. Discover how OLMo differs from previous efforts by providing not only model weights and inference code but also open training data and comprehensive training and evaluation code. Gain insights into the potential impact of this release on empowering the open research community and fostering new waves of innovation in language model development and applications.