# Dense Retrieval

A 7-billion-parameter multimodal embedding model specialized in visual document retrieval tasks, achieving outstanding performance on the Vidore-v2 benchmark

Nomic Embed Multimodal 3B is a cutting-edge multimodal embedding model focused on visual document retrieval tasks, supporting unified text-image encoding, achieving an outstanding performance of 58.8 NDCG@5 in the Vidore-v2 test.

DRAMA-base is a dense retrieval model based on a pruned large language model architecture, supporting multilingual text retrieval tasks.

MedCPT is a model capable of generating biomedical text embeddings, particularly suitable for semantic search (dense retrieval) tasks.

MedCPT is a model capable of generating biomedical text embeddings, particularly suitable for semantic search (dense retrieval) tasks.

DRAGON+ is a dense retrieval model based on the BERT architecture, employing an asymmetric dual-encoder architecture, suitable for text retrieval tasks.

DRAGON+ is a dense retrieval model based on the BERT architecture, with initial weights derived from RetroMAE and trained on enhanced data from the MS MARCO corpus.

T5-ANCE is a dense retrieval model trained on the MS MARCO passage dataset, optimized with the T5 architecture and ANCE training process, suitable for information retrieval tasks.

A dense retriever based on BERT-base architecture, trained on Wikipedia articles to emulate BM25 behavior

RankGen is a set of encoder models capable of mapping prefixes and generated content from pre-trained language models into a shared vector space to enhance generation quality and retrieval performance.

Scientific literature embedding model trained without SciDocs leakage data

TCT-ColBERT-V2 is a dense retrieval model based on the tightly-coupled teacher mechanism and in-batch negative sample knowledge distillation, designed for efficient text retrieval.

TCT-ColBERT-V2 is a dense retrieval model based on knowledge distillation, which improves retrieval efficiency and quality through tightly coupled teacher mechanisms and batch negative optimization training.

A fine-tuned version of the Contriever pre-trained model, optimized for dense information retrieval tasks and trained using contrastive learning methods