Benchmarking On-Device Machine Learning on Apple Silicon with MLX

TL;DR

This paper evaluates MLX, a framework optimized for on-device machine learning on Apple Silicon, benchmarking transformer model inference latency and demonstrating its efficiency compared to traditional frameworks and GPU performance.

Contribution

It introduces MLX-transformers, enabling seamless execution of transformer models on Apple Silicon without checkpoint conversion, and provides a comprehensive performance evaluation against NVIDIA GPUs.

Findings

MLX achieves lower inference latency on Apple Silicon compared to Pytorch implementations.

Transformer models run efficiently on Apple Silicon, approaching GPU performance levels.

MLX facilitates easier deployment of transformer models directly from Hugging Face.

Abstract

The recent widespread adoption of Large Language Models (LLMs) and machine learning in general has sparked research interest in exploring the possibilities of deploying these models on smaller devices such as laptops and mobile phones. This creates a need for frameworks and approaches that are capable of taking advantage of on-device hardware. The MLX framework was created to address this need. It is a framework optimized for machine learning (ML) computations on Apple silicon devices, facilitating easier research, experimentation, and prototyping. This paper presents a performance evaluation of MLX, focusing on inference latency of transformer models. We compare the performance of different transformer architecture implementations in MLX with their Pytorch counterparts. For this research we create a framework called MLX-transformers which includes different transformer implementations in MLX and downloads the model checkpoints in pytorch and converts it to the MLX format. By leveraging the advanced architecture and capabilities of Apple Silicon, MLX-Transformers enables seamless execution of transformer models directly sourced from Hugging Face, eliminating the need for checkpoint conversion often required when porting models between frameworks. Our study benchmarks different transformer models on two Apple Silicon macbook devices against an NVIDIA CUDA GPU. Specifically, we compare the inference latency performance of models with the same parameter sizes and checkpoints. We evaluate the performance of BERT, RoBERTa, and XLM-RoBERTa models, with the intention of extending future work to include models of different modalities, thus providing a more comprehensive assessment of MLX's capabilities. The results highlight MLX's potential in enabling efficient and more accessible on-device ML applications within Apple's ecosystem.