Скачать или смотреть LLM evaluation datasets: test cases and synthetic data

How to design evaluation datasets for your LLM application? In this video, we discuss where to get the data, how to use synthetic data effectively and what to include in the test scenarios.

00:15 Where do you get evaluation data from?
01:33 Using synthetic data to generate inputs
02:00 Using synthetic data for RAG
02:26 Using synthetic data to diversify inputs
02:43 Structuring test cases: happy path
03:24 Structuring test cases: edge cases
04:15 Structuring test cases: adversarial inputs
04:43 What do top companies do
05:13 What do you need to design evaluation datasets

LINKS
Blog on designing test datasets https://www.evidentlyai.com/llm-guide...
Segment https://segment.com/blog/llm-as-judge
Wix https://www.wix.engineering/post/cust...

Sign up for LLM evaluation course: https://www.evidentlyai.com/llm-evalu...
Instructor: Elena Samuylova, CEO Evidently AI.
Full playlist:    • LLM evaluation course