Children's English Reading Story Generation via Supervised Fine-Tuning of Compact LLMs with Controllable Difficulty and Safety

TL;DR

This paper demonstrates that fine-tuning compact 8B LLMs with a curriculum-based approach can produce children's reading stories that are easier, safe, and controllable, outperforming larger models in difficulty metrics.

Contribution

It introduces a fine-tuning method focused on controllability for small LLMs to generate suitable children's stories, reducing operational costs and safety concerns.

Findings

Fine-tuned 8B LLMs outperform zero-shot GPT-4o and Llama 3.3 70B in difficulty metrics.

The method ensures safety with minimal issues.

Controllable story generation is feasible with compact models.

Abstract

Large Language Models (LLMs) are widely applied in educational practices, such as for generating children's stories. However, the generated stories are often too difficult for children to read, and the operational cost of LLMs hinders their widespread adoption in educational settings. We used an existing expert-designed children's reading curriculum and its corresponding generated stories from GPT-4o and Llama 3.3 70B to design different experiments for fine-tuning three 8B-parameter LLMs, which then generated new English reading stories that were subjected to quantitative and qualitative evaluation. Our method prioritizes controllability over scale, enabling educators to target reading levels and error patterns with a compact, affordable model. Our evaluation results show that with appropriate fine-tuning designs, children's English reading stories generated by 8B LLMs perform better on difficulty-related metrics than those from zero-shot GPT-4o and Llama 3.3 70B, with almost no discernible safety issues. Such fine-tuned LLMs could be more broadly used by teachers, parents, and children in classrooms and at home to generate engaging English reading stories with children's interests, controllable difficulty and safety.