Float16 support #72
Description
I was thinking of implementing support for float16 (<f2 and >f2) using the half crate behind a feature flag. Would that be an appreciated feature?
{{ message }}
I was thinking of implementing support for float16 (<f2 and >f2) using the half crate behind a feature flag. Would that be an appreciated feature?