Powered by
Better RAG example
# ❓|questions
w
wuodarr
02/19/2024, 1:45 PM
I'm wonedring whether someone could try to improve
https://github.com/promptfoo/promptfoo/tree/98524381203d00abf8eba0a9fd16fb4f79f32fd4/examples/rag-eval
? Currently, context is passed from static docs, but in reality, I'd like to test E2E and use my retriever to get relevant docs, then evaluate the output of LLM. Or is it the current example the right approach?
g
gingerhendrix.
02/20/2024, 12:15 AM
Have you seen the guide
https://www.promptfoo.dev/docs/guides/evaluate-rag
?
w
wuodarr
02/20/2024, 1:11 PM
Yes, but E2E example could be more specific (e.g. example test cases)
g
gustavisaksen
04/02/2024, 8:52 AM
Hi, I was looking for a better End-to-End example as well. Did you figure out how to implement this?
@wuodarr
i
ian
04/02/2024, 12:25 PM
I did update the example to show how to include context dynamically here:
. You can use a script to retrieve documents for the
context
var, for example. Are you running into any problems with that / What questions can I help answer? 🙂
g
gustavisaksen
04/02/2024, 12:27 PM
Thank you! This is exactly what i was looking for 🙂
Previous
Next