*Data Synthesis* is a promising way to train a small model with very little labeled data. One approach for data synthesis is to leverage the rich knowledge from large language models to synthesize pseudo training examples for small models, making it possible to achieve both data and compute efficiency at the same time. However, a key challenge in data synthesis is that the synthesized dataset often suffers from a large distributional discrepancy from the *real task* data distribution. Thus, in this paper, we propose *Synthesis Step by Step* (**S3**), a data synthesis framework that shrinks this distribution gap by iteratively extrapolating the errors made by a small model trained on the synthesized dataset on a small real-world validation dataset using a large language model. Extensive experiments on multiple NLP tasks show that our approach improves the performance of a small model by reducing the gap between the synthetic dataset and the real data, resulting in significant improvement compared to several baselines: 9.48% improvement compared to ZeroGen and 2.73% compared to GoldGen, and at most 15.17% improvement compared to the small model trained on human-annotated data.
Self-Error-Instruct: Generalizing from Errors for LLMs Mathematical Reasoning
The Hong Kong Polytechnic University(香港理工大学);Research Centre for Data Science & Artificial Intelligence(数据科学与人工智能研究中心);The Chinese University of Hong Kong(香港中文大学)
Iterative Learning of Computable Phenotypes for Treatment Resistant Hypertension using Large Language Models
Federal University of ABC(亚巴西联邦大学);Department of Pathology and Laboratory Medicine, University of Pennsylvania(宾夕法尼亚大学病理学与实验室医学系);Computational Health Informatics Program, Boston Children’s Hospital, Harvard Medical School(哈佛医学院波士顿儿童医院计算健康信息学项目)
Not All Documents Are What You Need for Extracting Instruction Tuning Data
Department of Computer Science, Beijing Institute of Technology, Beijing, China(北京理工大学计算机科学系);Sensetime Research, Shenzhen, China(深睿时代研究);Shanghai Artificial Intelligence Laboratory, Shanghai, China(上海人工智能实验室);University of Arizona, USA(亚利桑那大学);Meta, USA(Meta)
Mining or Synthesis? Rethinking Exploration Efficiency in Iterative Alignment of Mathematical Reasoning
Institute of Computing and Intelligence, Harbin Institute of Technology, Shenzhen, China(哈尔滨工业大学深圳研究院);Huawei Large Model Data Technology Lab(华为大模型数据技术实验室);Huawei Multimodal Model Lab(华为多模态模型实验室);Department of Statistics and Data Science, Tsinghua University, Beijing, China(清华大学统计与数据科学系)
BRIDGE: Bootstrapping Text to Control Time-Series Generation via Multi-Agent Iterative Optimization and Diffusion Modeling
Microsoft Research(微软研究院);The University of Manchester, UK(英国曼彻斯特大学);Nanjing University, China(南京大学);Imperial College London, Imperial Global Singapore, Singapore(伦敦帝国学院,新加坡帝国全球分校,新加坡);The University of Tokyo, Japan(日本东京大学)
DialogueAgents: A Hybrid Agent-Based Speech Synthesis Framework for Multi-Party Dialogue
State Key Laboratory of Networking and Switching Technology, Beijing University of Posts and Telecommunications(网络与交换技术国家重点实验室,北京邮电大学);Department of Computer Science and Engineering, Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州)计算机科学与工程系);Hong Kong University of Science and Technology(香港科技大学)