About this Talk
Better models are released every day, should we only rely on better models and bigger RAG or is there a place for fine-tune?
In this talk I will showcase examples of fine-tune use for code completion and agentic use cases, specifically the case of use of fine-tune for heavy database with over 600 related tables, so that agent can generate correct queries faster and more accurately. I will explain how training data has been gathered and what results where achived
