Tex2Py-45K: A Parallel Corpus Dataset for Bidirectional Conversion Between LaTeX and Python
LaTeX and Python, integral to the scientific research workflow, each offers specialized capabilities. Bidirectional LaTeX-Python conversion models could augment the efficacy and precision of scientific research. In this paper, we release the Tex2Py-45K dataset1, containing 45,668 LaTeX-Python pairs. Based on the Tex2Py-45K dataset, we propose SMML format and Hybrid training mode, which trained effective LaTeX-Python bidirectional conversion models. At last, we analyze failure cases of the LaTeX-Python bidirectional conversion, providing direction for future research.
Authors
- Manni Duan
- Lei Li (ORCID: https://orcid.org/0000-0002-8891-1786)
- Yongheng Wang (ORCID: https://orcid.org/0000-0002-7675-2327)
Institutions
- Zhejiang Lab (CN)
Publication Details
- Published
- 2024-12-06
- DOI
- https://doi.org/10.1145/3709026.3709044
- Primary Topic
- Natural Language Processing Techniques
- Type
- article
- Field-Weighted Citation Impact
- 0.00