ó
    6ljX  ã                   ó   • S r g)uQ  FinanceBench benchmark: gold questions, answers, and evidence in Postgres.

The benchmark drives the playground corpus: every ingested document takes its
identity from a benchmark `doc_name` (see `load.doc_key_for`), and evaluation
reads gold answers from the `financebench` schema â€” which retrieval and the
answer agents never touch.
N)Ú__doc__© ó    ÚA/home/mande/repo/quber/src/quber/playground/benchmark/__init__.pyÚ<module>r      s
   ðòr   