Stop Sequence

Define a stop sequence in an LLM API call

What is the EOS token and how does the runtime use it?

Flashcard: what is a stop sequence in an LLM API call and what is it used for?

In a production prompt for generating a single SQL query, what's the right way to make sure the model returns only the SQL without trailing commentary?

In a few-shot prompt that extracts structured fields from invoices, why is a distinctive stop sequence (like a custom END_INVOICE marker) better than a paragraph break or the natural end of response?