LLMs have taken the world by storm over the past few months. But there's often a common perception that building an LLM model yourself is prohibitively challenging and expensive. In this talk I will show you how you can use state of the art Open Source foundation models and fine tune those models simply and cheaply to build something bespoke on your data. At Databricks we believe in selecting the right model for the right use-case, and with Dolly we demonstrated how fine-tuning on a bespoke dataset can provide instruction-following capability.