Revolutionizing Mixed Precision Quantization: Towards Training-free Automatic Proxy Discovery via Large Language Models

TL;DR

This paper introduces TAP, a training-free, LLM-driven framework that automatically discovers proxies for mixed-precision quantization, eliminating the need for manual design or costly optimization, and achieves state-of-the-art results.

Contribution

Proposes a novel LLM-based, training-free proxy discovery framework for MPQ that automates proxy design using evolutionary search and preference optimization.

Findings

Achieves state-of-the-art performance on benchmarks.

Eliminates manual proxy design and training for MPQ.

Demonstrates effectiveness of LLM-driven proxy discovery.

Abstract

Mixed-Precision Quantization (MPQ) liberates Deep Neural Networks (DNNs) from the Out-Of-Memory (OOM) bottleneck and has garnered increasing research attention. However, conventional methods either rely on costly differentiable optimization search, which is neither efficient nor flexible, or learn a quantized DNN from a proxy (e.g., HAWQ) manually designed by human experts, which is labor-intensive and requires extensive expert knowledge. Can we design a proxy without involving any human experts or training? In this paper, we provide an affirmative answer by proposing a novel Large Language Model (LLM)-driven Training-free Automatic Proxy (dubbed TAP) discovery framework. It reforms the design paradigm of MPQ by utilizing LLMs and evolutionary search strategies to automatically find superior TAP tailored for MPQ. In addition, to bridge the gap between black-box LLMs and the challenging MPQ task, we introduce a lightweight Direct Preference Optimization (DPO)-based strategy controller that dynamically reweights the selection probabilities of the three prompt templates for evolutionary search strategies according to fitness signals, without fine-tuning the LLM. This forms a task-aware feedback loop that improves proxy generation across evolutions. Extensive experiments on mainstream benchmarks demonstrate that TAP achieves state-of-the-art performance. Finally, we believe that our TAP will significantly contribute to the MPQ community by providing a new perspective on LLM-driven design algorithms.