Sean Sheng
Sean currently serves as the Head of Engineering at BentoML. He has led the team to successfully release multiple open-source projects, including BentoML and OpenLLM, aimed to help facilitate AI application development. Additionally, Sean has also led the launch of the AI deployment platform BentoCloud, designed for deploying and scaling AI applications in production. Prior to his role at BentoML, he led engineering teams at LinkedIn, where he supported the service infrastructure powering all of LinkedIn's backend services.
Sessions
Open source large language models (LLMs) are now inching towards matching the proficiency of proprietary models, such as GPT-4. In addition, operating your own LLMs can unveil advantages in aspects like data privacy, model customizability, and cost efficiency. However, running your own LLMs and realizing these benefits in a production environment is not easy - it necessitates a precise set of optimization and a robust infrastructure. Come to this talk to learn about the problems you might face when using your own large language models, and find out how OpenLLM can help you solve them.