Replication package for Automated Extract Method Refactoring with Open-Source LLMs
收藏Figshare2025-07-25 更新2026-04-08 收录
下载链接：
https://figshare.com/articles/dataset/Replication_package_for_Automated_Extract_Method_Refactoring_with_Open-Source_LLMs/29645171/1
下载链接
链接失效反馈官方服务：
资源简介：
Comparison of LLMs for Automated Code RefactoringThis project compares various Large Language Models (LLMs) for the task of automated code refactoring. The framework allows you to test and evaluate multiple models on real-world Python code samples.<br>---Features- Supports 5 popular LLMs for code refactoring- Works on sample Python files in batch- Flexible model configuration via `model.yaml`- Designed for experimentation and evaluation in research or production- Uses CodeNet dataset<br><br>Specify one of the following in `CodeBase/model.yaml`:<br>- `Qwen/CodeQwen1.5-7B-Chat`- `deepseek-ai/deepseek-coder-6.7b-instruct`- `meta-llama/Llama-3.2-3B-Instruct`- `Qwen/Qwen2.5-Coder-7B-Instruct`- `microsoft/Phi-4-mini-instruct`<br>Specify one of the following in `CodeBase/prompts.yaml`:<br>- `few_shot`- `zero_shot`- `rci`<br>Dependencies:<br>- pip install -r requirements.txt<br>Data:<br>- data/Python_wrapped - data/Problem_descriptions - selected_files.txt<br>Run :<br>- inference.main
提供机构：
Chand, Sivajeet
创建时间：
2025-07-25
