Tooka-SBERT: Lightweight Sentence Embedding models for Persian

ACL ARR 2025 July Submission1346 Authors

Abstract:

We introduce Tooka-SBERT, a family of Persian sentence embedding models designed to enhance semantic understanding for Persian. The models are released in two sizes—Small (123M parameters) and Large (353M parameters)—both built upon the TookaBERT backbone. Tooka-SBERT is pretrained on the Targoman News corpus and fine-tuned using high-quality synthetic Persian sentence pair datasets to improve semantic alignment. We evaluate Tooka-SBERT on PTEB, a Persian adaptation of the MTEB benchmark, where the Large model achieves an average score of 70.54\% and the Small model 69.49\%, outperforming some strong multilingual baselines.
Tooka-SBERT provides a compact and high-performing open-source solution for Persian sentence representation, with efficient inference suitable for both GPU and CPU environments.

Paper Type: Long

Research Area: Multilingualism and Cross-Lingual NLP

Research Area Keywords: Multilingualism and Cross-Lingual NLP, NLP Applications, Semantics: Lexical and Sentence-Level, Information Retrieval and Text Mining, Resources and Evaluation

Contribution Types: Publicly available software and/or pre-trained models

Languages Studied: Persian

Submission Number: 1346