Episode Notes: https://thedataexchange.media/bestpr...
Waleed Kadous, Chief Scientist at Anyscale, ia a leading expert on for best practices on building applications leveraging large language models.
*Sections*
Open Source LLMs: when and how to use them 00:01:06
Code Llama vs. GitHub Copilot 00:09:39
Deploying open source LLMs 00:13:06
Fine tuning LLMs 00:21:03
Using GPT to create fine tuning datasets 00:24:53
Retrieval augmented generation 00:32:15
RAG at scale and the role of LLMs in RAG 00:37:27
Evaluating RAG and experimenting with different RAG configurations and settings 00:41:01
Reimagining "AutoML" in the age of LLMs 00:45:56
Mixture of experts 00:49:29
AMD and other hardware options for LLM inference 00:54:30
Supply of open source LLMs 00:56:54