Bug: Python `int` and `float` objects are downcast to 32-bit types before sending.

Pyhon `int` and `float` objects are currently converted to 32-bit C `int` and C `float` types instead of the 64-bit `int64_t` and `double` types in C.
This is due to an error in the Cython binding code which uses the wrong datatypes.
A fix is imminent which will transmit these numbers correctly.