DARTS: Differentiable Architecture Search

E2187847 UNEXPLORED

DARTS: Differentiable Architecture Search is a neural architecture search method that formulates the search over network structures as a differentiable optimization problem, enabling efficient gradient-based discovery of high-performing deep learning architectures.

Try in SPARQL Jump to: Surface forms Referenced by

All labels observed (1)

Label Occurrences
DARTS: Differentiable Architecture Search canonical 1

Referenced by (1)

Full triples — surface form annotated when it differs from this entity's canonical label.

DARTS citationTitle DARTS: Differentiable Architecture Search