

Frontier Local AI - Cohere Labs in Conversation with Tim Dettmers
Frontier Local AI
Talk by Tim Dettmers, Assistant Professor, Carnegie Mellon University and Research Scientist, Allen Institute for Artificial Intelligence,
Hosted by Alejandro Salamanca, Open Science ML Engineer, Cohere Labs
As smaller models become more and more capable and with rising AI costs, it is becoming increasingly attractive to run models locally on your own hardware. In this talk, I will present our co-designed methods at the intersection of autonomous research, model compression, and agent harnesses that make it possible to deploy capable 300B agents on personal hardware, while harnesses remove weaknesses and also increase their strengths. The combination of these factors leads to frontier autonomous research agents that you can run on your laptop.
This talk is part of Cohere Labs in Conversation, a limited series of talks in which Cohere Labs scientists and engineers host external researchers for technical talks and Q&A discussions on subjects related to our current explorations at Cohere Labs. We look forward to sharing these talks with you, giving you a glimpse into the problems we're exploring, and learning together from some of the greatest minds in the field.