Self-play with Execution Feedback: Improving Instruction-following Capabilities of Large Language Models
This addresses the challenge of enhancing complex instruction-following capabilities in LLMs for AI applications, representing a novel method rather than an incremental improvement.
The paper tackles the problem of automatically generating high-quality instruction-following training data for large language models without manual annotation, achieving significant improvements in models like Qwen2 and LLaMA3 across SFT, Offline DPO, and Online DPO training algorithms.
One core capability of large language models (LLMs) is to follow natural language instructions. However, the issue of automatically constructing high-quality training data to enhance the complex instruction-following abilities of LLMs without manual annotation remains unresolved. In this paper, we introduce AutoIF, the first scalable and reliable method for automatically generating instruction-following training data. AutoIF transforms the validation of instruction-following data quality into code verification, requiring LLMs to generate instructions, the corresponding code to check the correctness of the instruction responses, and unit test samples to verify the code's correctness. Then, execution feedback-based rejection sampling can generate data for Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF) training. AutoIF achieves significant improvements across three training algorithms, SFT, Offline DPO, and Online DPO, when applied to the top open-source LLMs, Qwen2 and LLaMA3, in self-alignment and strong-to-weak distillation settings. Our code is publicly available at https://github.com/QwenLM/AutoIF.