A base-sized reranker model trained to evaluate and reorder candidate passages for relevance. It supports many languages and is designed for use in retrieval-augmented generation and search systems. The model uses the sentence-transformers library and is hosted publicly on Hugging Face.
In the Foundation models & chat space, Gte Multilingual Reranker Base takes a focused approach. It focuses on re-ranking retrieved documents to improve relevance in multilingual search and RAG pipelines. It is built as an open-source project for developers. Gte Multilingual Reranker Base is open source under the Open Source license. Gte Multilingual Reranker Base is available on API.
Alibaba-NLP builds and maintains Gte Multilingual Reranker Base, and the product first shipped in 2024.
Latest indexed changes and source events
Alibaba-NLP/gte-multilingual-reranker-base verified by the PulseGate indexer
Other apps tracked under the same category.
