Nemotron ColEmbed V2 π€ 8B    π€ 4B    π€ 3B   Model Overview Description The nvidia/nemotron colembed vl 4b v2 is a state of the art late interaction embedding model that ranks No. 3 in the ViDoRe V3: a comprehensive evaluation of retrieval for enterprise use case benchmark, (as of Jan 26, 2026) with a score of 61.42 on 8 public tasks. The model was fine tuned for query document retrieval. Users can input queries , which are text, or documents which are page images, to the model. The model outputs ColBERT style multi vector numerical representations for input queries and documents. β¨ Key Improvements: βοΈ Advanced Model Merging: Utilizes post training model merging to combine the strengths of multiple fine tuned checkpoints. This delivers the accuracy stability of an ensemble without any additional inference latency. π Enhanced Synthetic Data: We significantly enriched our training mixture with diverse multilingual synthetic data, improving semantic alignment across languages and complex document types. This model is for non commercial/research use only. See the Nemotron ColEmbed v2 paper for more details about its architecture, training and results. License/Teβ¦
We use cookies for essential functionality and analytics. You can accept or reject analytics cookies.Cookie policy