markdtw/awesome-architecture-search

Covers reinforcement learning, evolutionary algorithms, and gradient-based approaches (DARTS) for automated neural architecture design, alongside hyperparameter optimization techniques like Bayesian optimization and Hyperband. Organizes landmark papers and implementations across multiple search paradigms, enabling researchers to compare methodologies from controller-based NAS to differentiable architecture search and evolutionary strategies. Bridges the gap between foundational AutoML work and contemporary efficient search methods targeting deployment constraints like mobile inference.

1,190 stars. No commits in the last 6 months.