Convex Distillation: Efficient Compression of Deep Networks via Convex Optimization

Prateek Varshney, Mert Pilanci

Deploying large and complex deep neural networks on resource-constrained edge devices poses significant challenges due to their computational demands and the complexities of non-co…