For mobile device inference
I don't know the answer, but even if it works, 180B is big enough to be not reliable/doable on model devices.
· Sign up or log in to comment