通过专家模型互斗生成高质量代码训练数据，提升代码大模型性能。
WarriorCoder: Learning from Expert Battles to Augment Code Large Language Models
- 让多个顶尖代码大模型在竞技场中互相挑战，自动生成训练数据。
- 在相同规模下超越现有模型，无需依赖闭源大模型数据。
- 适合想提升代码生成能力的研究者和开发者使用。
尽管代码大语言模型（LLMs）取得了显著进展，但其强大能力主要依赖于高质量数据的微调，带来了数据收集与标注的挑战。当前方法常设计数据飞轮以收集复杂代码指令，使模型能处理更复杂的任务。然而，这些方法通常依赖现成数据集及少数专有大模型（如 Claude、GPT4 等）的数据增强，限制了数据多样性，易引入系统性偏差。本文提出 WarriorCoder，一种从专家对决中学习的新范式。我们构建一个竞技场，让领先的代码大模型相互挑战，由中立裁判评估。这一竞争机制从零开始生成新颖训练数据，融合所有参与者的优点。实验表明，WarriorCoder 在同等模型规模下达到当前最优性能，且不依赖专有大模型。
原文摘要 · Abstract (English)
Despite recent progress achieved by code large language models (LLMs), their remarkable abilities are largely dependent on fine-tuning on the high-quality data, posing challenges for data collection and annotation. To address this, current methods often design various data flywheels to collect complex code instructions, enabling models to handle more intricate tasks. However, these approaches typically rely on off-the-shelf datasets and data augmentation from a limited set of proprietary LLMs (e.g., Claude, GPT4, and so on), which restricts the diversity of the constructed data and makes it prone to systemic biases. In this paper, we propose WarriorCoder, a novel paradigm learns from expert battles to address these limitations. Specifically, we create an arena where leading expert code LLMs challenge each other, with evaluations conducted by impartial judges. This competitive framework generates novel training data from scratch, leveraging the strengths of all participants. Experimental results show that WarriorCoder achieves state-of-the-art performance compared to previous models of the same size, even without relying on proprietary LLMs.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。