DatabricksRM KeyError 'score'

[kvnlng]

Sending a list of vectors to DatabricksRM instance returns:

KeyError: 'score'
File , line 1
----> 1 retriever(qv)

239 # Sorting results by score in descending order

--> 240 sorted_docs = sorted(items, key=lambda x: x["score"], reverse=True)[: self.k]
242 # Returning the prediction
243 return Prediction(
244 docs=[doc[self.text_column_name] for doc in sorted_docs],
245 doc_ids=[self._extract_doc_ids(doc) for doc in sorted_docs],
246 extra_columns=[self._get_extra_columns(item) for item in sorted_docs],
247 )

Note:
This is a STORAGE_OPTIMIZED index being tested under private beta.

Happy to provide more details

[chenmoneygithub]

@dbczumar Any clues?