Automated Program Repair using quantized language models and parameter-efficient fine-tuning

doi:10.1016/j.infsof.2026.108181

Automated Program Repair using quantized language models and parameter-efficient fine-tuning

2026-05-05

0

PRE

AI

DOI：10.1016/j.infsof.2026.108181

Original

Original request for help

Save

Abstract

En 中文

Large language models have demonstrated remarkable capabilities in Automated Program Repair (APR), outperforming traditional template- and rule-based approaches. However, their extensive memory and computational requirements pose significant challenges for local deployment on consumer-grade GPUs, which are essential for maintaining data privacy and avoiding dependency on cloud-based API services.

Keywords:

Automated Program Repair

Large language models

Quantization

Parameter-efficient fine-tuning

Local deployment

Automated Program Repair using quantized language models and parameter-efficient fine-tuning

Abstract

Journal

Organization