Fine-Tuning Open Source Models for Agents

Open source LLMs and multi-modal foundation models offer cost-effective and flexible options for developing domain-specific generative AI applications. However, building AI agents – e.g., application copilots or domain-specific assistants that leverage tools – requires models optimized for retrieval, chat, planning, and function calling. This talk will provide an overview of the open source foundation model ecosystem and best practices for fine-tuning models for AI agent use cases.