ViRanker: A BGE-M3 & Blockwise Parallel Transformer Cross-Encoder for Vietnamese Reranking

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dang, Phuong-Nam, Nguyen, Kieu-Linh, Pham, Thanh-Hieu
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918139101446144
author Dang, Phuong-Nam
Nguyen, Kieu-Linh
Pham, Thanh-Hieu
author_facet Dang, Phuong-Nam
Nguyen, Kieu-Linh
Pham, Thanh-Hieu
contents This paper presents ViRanker, a cross-encoder reranking model tailored to the Vietnamese language. Built on the BGE-M3 encoder and enhanced with the Blockwise Parallel Transformer, ViRanker addresses the lack of competitive rerankers for Vietnamese, a low-resource language with complex syntax and diacritics. The model was trained on an 8 GB curated corpus and fine-tuned with hybrid hard-negative sampling to strengthen robustness. Evaluated on the MMARCO-VI benchmark, ViRanker achieves strong early-rank accuracy, surpassing multilingual baselines and competing closely with PhoRanker. By releasing the model openly on Hugging Face, we aim to support reproducibility and encourage wider adoption in real-world retrieval systems. Beyond Vietnamese, this study illustrates how careful architectural adaptation and data curation can advance reranking in other underrepresented languages.
format Preprint
id arxiv_https___arxiv_org_abs_2509_09131
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ViRanker: A BGE-M3 & Blockwise Parallel Transformer Cross-Encoder for Vietnamese Reranking
Dang, Phuong-Nam
Nguyen, Kieu-Linh
Pham, Thanh-Hieu
Computation and Language
Artificial Intelligence
This paper presents ViRanker, a cross-encoder reranking model tailored to the Vietnamese language. Built on the BGE-M3 encoder and enhanced with the Blockwise Parallel Transformer, ViRanker addresses the lack of competitive rerankers for Vietnamese, a low-resource language with complex syntax and diacritics. The model was trained on an 8 GB curated corpus and fine-tuned with hybrid hard-negative sampling to strengthen robustness. Evaluated on the MMARCO-VI benchmark, ViRanker achieves strong early-rank accuracy, surpassing multilingual baselines and competing closely with PhoRanker. By releasing the model openly on Hugging Face, we aim to support reproducibility and encourage wider adoption in real-world retrieval systems. Beyond Vietnamese, this study illustrates how careful architectural adaptation and data curation can advance reranking in other underrepresented languages.
title ViRanker: A BGE-M3 & Blockwise Parallel Transformer Cross-Encoder for Vietnamese Reranking
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2509.09131