This repo is for the ANZ LLM Bootcamp Series.
This series of notebooks have been developed and tested on ML Runtime 13.2 GA Edition
For optimal performance, we recommend running on g5.4xlarge instances on AWS.
On Databricks, you can select the Single Node instance type and select the runtime and infrastructure above. If you run into capacity issues, we have supplemented the notebooks with CPU editions. In this case, feel free to use i3.4xlarge or m series. Please note that this will not have optimal performance.
- We have a great catalog of LLM related talks at the Data and AI Summit link here
- For a set of great examples on fine-tuning these LLMs, we recommend looking at the Databricks ML examples repo