LeGo-Code: Can Modular Curriculum Learning Advance Complex Code Generation? Insights from Text-to-SQL

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chafik, Salmane, Ezzini, Saad, Berrada, Ismail
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913047604363264
author Chafik, Salmane
Ezzini, Saad
Berrada, Ismail
author_facet Chafik, Salmane
Ezzini, Saad
Berrada, Ismail
contents Recently, code-oriented large language models (LLMs) have demonstrated strong capabilities in translating natural language into executable code. Text-to-SQL is a significant application of this ability, enabling non-technical users to interact with relational databases using natural language. However, state-of-the-art models continue to struggle with highly complex logic, particularly deeply nested statements involving multiple joins and conditions, as well as with real-world database schemas that are noisy or poorly structured. In this paper, we investigate whether curriculum learning can improve the performance of code-based LLMs on Text-to-SQL tasks. Employing benchmarks including Spider and BIRD, we fine-tune models under different curriculum strategies. Our experiments show that naive curriculum, simply ordering training samples by complexity in a single epoch, fails to surpass standard fine-tuning due to catastrophic forgetting. To overcome this, we propose a Modular Adapter Composition (MAC) strategy. By sequentially training tier-specific adapters on incremental complexity levels (Easy to Extra-Hard), we create a scaffolded learning environment that improves performance on complex queries. Our approach not only produces measurable performance gains on the Spider and BIRD benchmarks but also provides a flexible, "Lego-like" architecture, allowing models to be composed and deployed based on specific schema difficulty requirements. These findings demonstrate that structured, modular learning is a superior alternative to monolithic fine-tuning for mastering the syntax and logic of complex code generation.
format Preprint
id arxiv_https___arxiv_org_abs_2604_18254
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle LeGo-Code: Can Modular Curriculum Learning Advance Complex Code Generation? Insights from Text-to-SQL
Chafik, Salmane
Ezzini, Saad
Berrada, Ismail
Artificial Intelligence
Databases
Software Engineering
Recently, code-oriented large language models (LLMs) have demonstrated strong capabilities in translating natural language into executable code. Text-to-SQL is a significant application of this ability, enabling non-technical users to interact with relational databases using natural language. However, state-of-the-art models continue to struggle with highly complex logic, particularly deeply nested statements involving multiple joins and conditions, as well as with real-world database schemas that are noisy or poorly structured. In this paper, we investigate whether curriculum learning can improve the performance of code-based LLMs on Text-to-SQL tasks. Employing benchmarks including Spider and BIRD, we fine-tune models under different curriculum strategies. Our experiments show that naive curriculum, simply ordering training samples by complexity in a single epoch, fails to surpass standard fine-tuning due to catastrophic forgetting. To overcome this, we propose a Modular Adapter Composition (MAC) strategy. By sequentially training tier-specific adapters on incremental complexity levels (Easy to Extra-Hard), we create a scaffolded learning environment that improves performance on complex queries. Our approach not only produces measurable performance gains on the Spider and BIRD benchmarks but also provides a flexible, "Lego-like" architecture, allowing models to be composed and deployed based on specific schema difficulty requirements. These findings demonstrate that structured, modular learning is a superior alternative to monolithic fine-tuning for mastering the syntax and logic of complex code generation.
title LeGo-Code: Can Modular Curriculum Learning Advance Complex Code Generation? Insights from Text-to-SQL
topic Artificial Intelligence
Databases
Software Engineering
url https://arxiv.org/abs/2604.18254