deep learning toolbox: how to change precision to double?

Hello,

It seems that the Deep Learning Toolbox uses the single precision and that it's hard coded.

Is there a way to change the precision to doube?

Thx,

D

Answer 1

[Bhaskar R]

MATLAB is by default double precision floating point(double) i.e. 64-bit.

This answer can help you

[John Smith]

Yes, but DLT is not. In the code there're hardcoded casts to single precision.