From the Cloud to the Edge: The Future of Language Models - tinyML Talk

Explore the future of language models in this one-hour tinyML TALK featuring Mahesh Yadav from Google. Delve into the shift from Large Language Models (LLMs) to Small Language Models (SLMs) and edge computing, addressing critical issues such as high costs, latency, and security concerns. Discover how LLMs may evolve to function like operating systems, with SLMs driving applications directly on devices. Learn about techniques enabling this transition, including distillation and pruning for training, and performance optimization for inference. Participate in a hands-on lab to experience running an SLM on the edge. Benefit from Mahesh Yadav's 20 years of experience in AI product development across Meta, Microsoft, and AWS, gaining insights into the entire AI stack from chips to LLMs and understanding how GenAI companies deliver value to customers.