Metraipolis › archive › dc_3def89465262
COAL-SQL: Coverage-Guided Augmentation and Failure-Driven Learning for Text-to-SQL Post-Training — arXiv:2609.20842
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.20842
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-10T01:27:02.152Z
- SHA-256 of the archived text
448cea94c1912559706061a76594cb16556b65d513808fb09e8b2f63cb54c9ba- Stable URL of this page
https://metraipolis.com/documento/dc_3def89465262
How to cite
COAL-SQL: Coverage-Guided Augmentation and Failure-Driven Learning for Text-to-SQL Post-Training — arXiv:2609.20842 — arXiv (en), https://arxiv.org/abs/2609.20842 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_3def89465262 (Metraipolis archive dc_3def89465262, sha256 448cea94c1912559706061a76594cb16556b65d513808fb09e8b2f63cb54c9ba)
BibTeX
@misc{dc3def89465262,
title = {COAL-SQL: Coverage-Guided Augmentation and Failure-Driven Learning for Text-to-SQL Post-Training — arXiv:2609.20842},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.20842},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_3def89465262; sha256 448cea94c1912559706061a76594cb16556b65d513808fb09e8b2f63cb54c9ba},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.20842. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Text-to-SQL translates natural-language questions into executable SQL queries, but open-source large language models still require task-specific post-training for complex, real-world SQL generation. Effective post-training requires both training data that cover the capabilities demanded by the target task and a learning strategy that enables the model to acquire them. Existing datasets provide valuable supervision but incompletely cover SQL structures, while augmentation methods typically expand data without identifying structural gaps. Moreover, supervised fine-tuning (SFT) or reinforcement learning (RL) alone cannot dynamically address weaknesses exposed during training. We propose COAL-SQL, a unified framework combining Coverage-Guided Augmentation (CGA) and Failure-Driven Learning (FDL). CGA uses greedy selection to identify SQL structures missing from the original dataset and constructs complementary examples, improving structural coverage. FDL retains GRPO as the main optimization objective while supplying targeted supervision for unsolved examples. At the step level, it applies SFT to verified reasoning traces generated by a strong LLM for accumulated failures. At the epoch level, it retrieves structurally related examples based on accumulated failures to create targeted practice, helping the model acquire the corresponding SQL capabilities. With only 12,600 distinct post-training examples, COAL-SQL achieves 64.9% execution accuracy on the BIRD development set and outperforms baselines trained at comparable scale. The code is available at this https URL . Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-21 in cs.CL). Categorie arXiv: Computation and Language (cs.CL) ; Databases (cs.DB). Pagina del lavoro: https://arxiv.org/abs/2609.20842 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-10T01:00:30.249Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-09-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).