DrNAS: Dirichlet Neural Architecture Search

Ruochen Wang
Ruochen Wang
Xiaocheng Tang
Xiaocheng Tang
Cited by: 0|Bibtex|Views28|Links

Abstract:

This paper proposes a novel differentiable architecture search method by formulating it into a distribution learning problem. We treat the continuously relaxed architecture mixing weight as random variables, modeled by Dirichlet distribution. With recently developed pathwise derivatives, the Dirichlet parameters can be easily optimized ...More

Code:

Data:

Your rating :
0

 

Tags
Comments