Ask Your Question

Revision history [back]

threshold function: why 8-bit or 32-bit floating point only?

For the threshold function, is there a specific reason to only support these two data types? Thanks!