Data Brew by Databricks LLMs: Internals, Hallucinations, and Applications | Data Brew | Episode 33
Jul 21, 2023
Databricks experts discuss best practices for prompt engineering with large language models, risks and mitigation strategies for hallucinations, criteria for choosing and fine-tuning language models, and strategies for keeping up with LLMs in the rapidly evolving field.
Chapters
Transcript
Episode notes
1 2 3 4 5 6
Introduction
00:00 • 2min
Background in AI, ML and Prompt Engineering
01:32 • 13min
Intrinsic and Extrinsic Hallucinations in Language Models
14:37 • 2min
Risks of Hallucinations in Language Models and Mitigation Strategies
17:05 • 4min
Choosing and Fine-tuning Language Models
21:35 • 14min
Challenges and Strategies for Keeping Up with Large Language Models
35:34 • 3min