Synthetic Data (Almost) from Scratch: Generalized Instruction Tuning for Language Models

TL;DR

This paper presents GLAN, a scalable method for instruction tuning of large language models using a human knowledge taxonomy to generate synthetic training data, improving performance across diverse tasks without task-specific data.

Contribution

Introduces a novel generalized instruction tuning framework that leverages a structured human knowledge taxonomy to generate synthetic data for large language models.

Findings

GLAN improves performance on mathematical reasoning, coding, and academic exams.

It enables easy customization for new fields or skills.

Demonstrates effectiveness without task-specific training data.

Abstract

We introduce Generalized Instruction Tuning (called GLAN), a general and scalable method for instruction tuning of Large Language Models (LLMs). Unlike prior work that relies on seed examples or existing datasets to construct instruction tuning data, GLAN exclusively utilizes a pre-curated taxonomy of human knowledge and capabilities as input and generates large-scale synthetic instruction data across all disciplines. Specifically, inspired by the systematic structure in human education system, we build the taxonomy by decomposing human knowledge and capabilities to various fields, sub-fields and ultimately, distinct disciplines semi-automatically, facilitated by LLMs. Subsequently, we generate a comprehensive list of subjects for every discipline and proceed to design a syllabus tailored to each subject, again utilizing LLMs. With the fine-grained key concepts detailed in every class session of the syllabus, we are able to generate diverse instructions with a broad coverage across the entire spectrum of human knowledge and skills. Extensive experiments on large language models (e.g., Mistral) demonstrate that GLAN excels in multiple dimensions from mathematical reasoning, coding, academic exams, logical reasoning to general instruction following without using task-specific training data of these tasks. In addition, GLAN allows for easy customization and new fields or skills can be added by simply incorporating a new node into our taxonomy.