This is a compact 1 billion parameter reranker model based on Llama architecture from NVIDIA's Nemotron family. It is designed to score and reorder candidate results for vision-language retrieval tasks. The model is provided as open weights on Hugging Face and is intended for integration into retrieval-augmented generation or search pipelines that involve both images and text.
In the Foundation models & chat space, Llama Nemotron Rerank Vl 1b takes a focused approach. It focuses on improving retrieval quality by reranking candidate results in vision-language search systems. Llama Nemotron Rerank Vl 1b is an open-source project aimed at developers building multimodal retrieval systems. The project is open source (Apache-2.0). It runs on the web, the command line, and API.
It is developed by NVIDIA (United States), and the product first shipped in 2023. The project is developed in the open on GitHub with 86.8k stars and 3k commits in the last 90 days. Among its 3 catalogued features are reranking, vision-language retrieval, and cross-modal scoring.
Latest indexed changes and source events
nvidia/llama-nemotron-rerank-vl-1b-v2 verified by the PulseGate indexer
