Token-Operations-Oriented Inference Optimization Techniques for Large Models
For practitioners deploying large model services, this paper offers a structured framework to optimize inference, but it is a survey/review without new experimental results.
This paper proposes a four-layer technical architecture for token-oriented inference optimization in large models, aiming to reduce costs and improve efficiency. It reviews key technologies and industry status across these layers, providing a practical path for transitioning large model services from callable to operable.
Large model inference optimization serves as a key foundation for supporting the scalable, low-cost, and highly stable operation of large model services. Centered on token-oriented inference optimization technology, this paper proposes for the first time a four-layer technical architecture consisting of Multi-model Fusion, Model Optimization, Compute-Model Fusion, and Compute-Network-Model Fusion. It systematically reviews the key technologies and current industry status across these four levels and analyzes the application value of related technologies in real-world business scenarios. This paper provides a practical technical path for reducing token production costs, improving token service efficiency, ensuring the stability of token supply, and driving the transition of large model services from being merely callable to being operable.