float64 parameters incorrectly mapped to SqlFloat (32-bit)

[Mandukhai-Alimaa]

When passing float64 values as query parameters, they are incorrectly mapped to SqlFloat instead of SqlDouble, causing precision loss and overflow/underflow issues.

https://github.com/databricks/databricks-sql-go/blob/main/parameters.go#L151