Helix: Serving Large Language Models over Heterogeneous GPUs and Network via Max-Flow

Devs

Helix: Serving Large Language Models over Heterogeneous GPUs and Network via Max-Flow | Read Paper on Bytez