--- title: cross-encoder/ms-marco-MiniLM-L-6-v2 description: This model was trained on the MS Marco Passage Ranking task.. BERT, 23M parameters. canonical_url: https://superlinked.com/models/cross-encoder-ms-marco-minilm-l-6-v2 last_updated: 2026-10-14 --- # cross-encoder/ms-marco-MiniLM-L-6-v2 This model was trained on the MS Marco Passage Ranking task. Source: [cross-encoder/ms-marco-MiniLM-L-6-v2 on HuggingFace](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L-6-v2) Base model: [cross-encoder/ms-marco-MiniLM-L12-v2](https://huggingface.co/cross-encoder/ms-marco-MiniLM-L12-v2) ## Overview | Field | Value | |-------|-------| | Architecture | BERT | | Parameters | 23M | | Tasks | Score | | Outputs | Score | | Max sequence length | 512 tokens | | License | apache-2.0 | | Inputs | text | | Languages | en | ## Benchmarks ### AskUbuntuDupQuestions Domain: technology · Task: reranking · Language: en Duplicate question detection from AskUbuntu Corpus: 6,743 · Queries: 360 **Quality:** ndcg at 10: 0.6027 · map at 10: 0.4439 · mrr at 10: 0.6776 **Performance (L4 b1 c16):** Query 948 tok/s · Query p50 362.8ms [Reference](https://github.com/taolei87/askubuntu) ### CMedQAv1Reranking Domain: medical · Task: reranking · Language: zh Chinese medical question answering reranking (v1) Corpus: 100,000 · Queries: 2,000 **Quality:** map at 10: 0.0835 · mrr at 10: 0.1371 [Reference](https://github.com/zhangsheng93/cMedQA) ### CMedQAv2Reranking Domain: medical · Task: reranking · Language: zh Chinese medical question answering reranking (v2) Corpus: 108,000 · Queries: 4,000 **Quality:** map at 10: 0.0926 · mrr at 10: 0.1425 [Reference](https://github.com/zhangsheng93/cMedQA2) ### CQADupstackPhysicsRetrieval Domain: scientific · Task: retrieval · Language: en Duplicate question retrieval from StackExchange Physics Corpus: 38,314 · Queries: 1,039 **Quality:** map at 10: 0.3669 · mrr at 10: 0.4301 · ndcg at 10: 0.4251 [Reference](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) ### CQADupstackPhysicsRetrieval (candidates: gte-multilingual-base, k=50) Domain: scientific · Task: retrieval · Language: en Duplicate question retrieval from StackExchange Physics **Performance (L4 b1 c16):** Query 44.3K tok/s · Query p50 44.6ms [Reference](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) ### CosQA Domain: technology · Task: retrieval · Language: en Code search with natural language queries Corpus: 6,267 · Queries: 500 **Quality:** map at 10: 0.1937 · mrr at 10: 0.2108 · ndcg at 10: 0.2618 [Reference](https://arxiv.org/abs/2105.13239) ### CosQA (candidates: gte-multilingual-base, k=50) Domain: technology · Task: retrieval · Language: en Code search with natural language queries **Performance (L4 b1 c16):** Query 20.5K tok/s · Query p50 43.6ms [Reference](https://arxiv.org/abs/2105.13239) ### FiQA2018 Domain: finance · Task: retrieval · Language: en Financial opinion mining and question answering Corpus: 57,599 · Queries: 648 **Quality:** map at 10: 0.3160 · mrr at 10: 0.4782 · ndcg at 10: 0.3977 [Reference](https://sites.google.com/view/fiqa/) ### FiQA2018 (candidates: gte-multilingual-base, k=50) Domain: finance · Task: retrieval · Language: en Financial opinion mining and question answering **Performance (L4 b1 c16):** Query 51.1K tok/s · Query p50 43.4ms [Reference](https://sites.google.com/view/fiqa/) ### LegalBenchConsumerContractsQA Domain: legal · Task: retrieval · Language: en Question answering on consumer contracts Corpus: 153 · Queries: 396 **Quality:** map at 10: 0.6822 · mrr at 10: 0.6859 · ndcg at 10: 0.7409 [Reference](https://huggingface.co/datasets/nguha/legalbench) ### LegalBenchConsumerContractsQA (candidates: gte-multilingual-base, k=50) Domain: legal · Task: retrieval · Language: en Question answering on consumer contracts **Performance (L4 b1 c16):** Query 91.7K tok/s · Query p50 45.6ms [Reference](https://huggingface.co/datasets/nguha/legalbench) ### MMarcoReranking Domain: general · Task: reranking · Language: zh Multilingual MARCO passage reranking (Chinese) **Quality:** map at 10: 0.0543 · mrr at 10: 0.0544 [Reference](https://arxiv.org/abs/2304.03679) ### NFCorpus Domain: medical · Task: retrieval · Language: en Biomedical literature search from NutritionFacts.org Corpus: 3,593 · Queries: 323 **Quality:** map at 10: 0.2606 · mrr at 10: 0.5808 · ndcg at 10: 0.3591 [Reference](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) ### NFCorpus (candidates: gte-multilingual-base, k=50) Domain: medical · Task: retrieval · Language: en Biomedical literature search from NutritionFacts.org **Performance (L4 b1 c16):** Query 70.8K tok/s · Query p50 45.9ms [Reference](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) ### NanoFiQA2018Retrieval Domain: finance · Task: retrieval · Language: en Smaller subset of the FiQA financial QA dataset **Performance (L4 b1 c16):** Query 7.5K tok/s · Query p50 388.1ms [Reference](https://sites.google.com/view/fiqa/) ### SCIDOCS Domain: scientific · Task: retrieval · Language: en Citation prediction, document classification, and recommendation for scientific papers Corpus: 25,656 · Queries: 1,000 **Quality:** map at 10: 0.1065 · mrr at 10: 0.3257 · ndcg at 10: 0.1833 [Reference](https://allenai.org/data/scidocs) ### SCIDOCS (candidates: gte-multilingual-base, k=50) Domain: scientific · Task: retrieval · Language: en Citation prediction, document classification, and recommendation for scientific papers **Performance (L4 b1 c16):** Query 53.7K tok/s · Query p50 42.5ms [Reference](https://allenai.org/data/scidocs) ### SciFact Domain: scientific · Task: retrieval · Language: en Scientific claim verification using research literature Corpus: 5,183 · Queries: 300 **Quality:** map at 10: 0.6569 · mrr at 10: 0.6764 · ndcg at 10: 0.7071 [Reference](https://github.com/allenai/scifact) ### SciFact (candidates: gte-multilingual-base, k=50) Domain: scientific · Task: retrieval · Language: en Scientific claim verification using research literature **Performance (L4 b1 c16):** Query 67.4K tok/s · Query p50 42.1ms [Reference](https://github.com/allenai/scifact) ### StackOverflowQA Domain: technology · Task: retrieval · Language: en Programming question answering from Stack Overflow Corpus: 19,931 · Queries: 1,994 **Quality:** map at 10: 0.5244 · mrr at 10: 0.5349 · ndcg at 10: 0.5815 [Reference](https://arxiv.org/abs/2407.02883) ### StackOverflowQA (candidates: gte-multilingual-base, k=50) Domain: technology · Task: retrieval · Language: en Programming question answering from Stack Overflow **Performance (L4 b1 c16):** Query 98.6K tok/s · Query p50 47.2ms [Reference](https://arxiv.org/abs/2407.02883) ### T2Reranking Domain: general · Task: reranking · Language: zh Chinese passage ranking benchmark **Quality:** map at 10: 0.4714 · mrr at 10: 0.7102 [Reference](https://arxiv.org/abs/2304.03679)