???
#1
by kalle07 - opened
no one know how to use?
is that an embedder model?
and is that only to embedd and and the extract-version is for receive?
The use instructions were linked in the readme (Model card).
https://www.liquid.ai/blog/introducing-liquid-nanos-frontier-grade-performance-on-everyday-devices
Your system prompt is the json template, and the data (document) gets fed into the user prompt. Out comes the formatted json filled with your document's data.
[EDIT] I'm sorry, that applies to the -EXTRACT- not the RAG. I don't know how to use the RAG yet.