FasterPy: An LLM-based Code Execution Efficiency Optimization Framework

FasterPy is a newly proposed framework that leverages Large Language Models (LLMs) to optimize Python code execution efficiency. It integrates Retrieval-Augmented Generation (RAG) with a knowledge base of performance-improving code pairs and employs Low-Rank Adaptation (LoRA) for enhanced optimization performance. Experimental results on the Performance Improving Code Edits (PIE) benchmark indicate that FasterPy surpasses existing optimization models, providing a cost-effective solution for practitioners seeking to automate code performance improvements.