ViRanker: A BGE-M3 & Blockwise Parallel Transformer Cross-Encoder for Vietnamese Reranking
This work addresses the problem of improving retrieval systems for Vietnamese, an underrepresented language, though it is incremental as it builds on existing methods like BGE-M3 and Blockwise Parallel Transformer.
The paper tackles the lack of competitive reranking models for Vietnamese, a low-resource language, by introducing ViRanker, which achieves strong early-rank accuracy on the MMARCO-VI benchmark, surpassing multilingual baselines and competing closely with PhoRanker.
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.