Collaborative Language Modeling
Thomas discusses the innovative approach to training large language models by gathering researchers from various fields, akin to large-scale physics collaborations. With over 1,000 participants contributing, they recently completed an extensive 800GB data set and are set to begin a four-month training process. This initiative not only enhances research capabilities but also expands the resources available for future projects.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Big Science and Embodied Learning at Hugging Face 🤗 with Thomas Wolf - #564
Related Questions