Nikita Sivukhin 4091efb5b1 accept K parameter as float if there is no loss in the precision after rounding to the integer ...

- it cah be hard in some cliens (JS for example) to distinguish between INTEGERs and FLOATs - so we always bind FLOATs instead
- this create unnecessary error: "vector index(search): third parameter (k) must be a non-negative integer"

2024-09-29 15:09:47 +04:00

25 KiB

Raw Permalink Blame History

View Raw