PalmBench: A Comprehensive Benchmark of Compressed Large Language Models on Mobile Platforms

TL;DR

This paper introduces PalmBench, a comprehensive benchmarking framework for evaluating compressed large language models on mobile devices, focusing on resource efficiency, performance, and safety aspects across various hardware configurations.

Contribution

We present a lightweight, automated benchmarking framework and provide extensive evaluations of quantized LLMs on multiple mobile platforms, highlighting their efficiency and safety trade-offs.

Findings

Quantization affects memory, speed, and power consumption.

Energy efficiency varies across mobile devices.

Compressed models show increased hallucinations and toxicity.

Abstract

Deploying large language models (LLMs) locally on mobile devices is advantageous in scenarios where transmitting data to remote cloud servers is either undesirable due to privacy concerns or impractical due to network connection. Recent advancements (MLC, 2023a; Gerganov, 2023) have facilitated the local deployment of LLMs. However, local deployment also presents challenges, particularly in balancing quality (generative performance), latency, and throughput within the hardware constraints of mobile devices. In this paper, we introduce our lightweight, all-in-one automated benchmarking framework that allows users to evaluate LLMs on mobile devices. We provide a comprehensive benchmark of various popular LLMs with different quantization configurations (both weights and activations) across multiple mobile platforms with varying hardware capabilities. Unlike traditional benchmarks that assess full-scale models on high-end GPU clusters, we focus on evaluating resource efficiency (memory and power consumption) and harmful output for compressed models on mobile devices. Our key observations include i) differences in energy efficiency and throughput across mobile platforms; ii) the impact of quantization on memory usage, GPU execution time, and power consumption; and iii) accuracy and performance degradation of quantized models compared to their non-quantized counterparts; and iv) the frequency of hallucinations and toxic content generated by compressed LLMs on mobile devices.