cross-encoder/ms-marco-MiniLM-L-6-v2
This model was trained on the MS Marco Passage Ranking task.
Overview
Benchmarks
AskUbuntuDupQuestions
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 827 tok/s
Query p50 411.2ms
CMedQAv1Reranking
Chinese medical question answering reranking (v1)
Corpus: 100,000 Queries: 2,000
Quality
map at 10 0.0835
mrr at 10 0.1371
CMedQAv2Reranking
Chinese medical question answering reranking (v2)
Corpus: 108,000 Queries: 4,000
Quality
map at 10 0.0926
mrr at 10 0.1425
CQADupstackPhysicsRetrieval?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 44.3K tok/s
Query p50 44.6ms
CosQA?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 20.5K tok/s
Query p50 43.6ms
FiQA2018?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 51.1K tok/s
Query p50 43.4ms
LegalBenchConsumerContractsQA?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 91.7K tok/s
Query p50 45.6ms
MMarcoReranking
Multilingual MARCO passage reranking (Chinese)
Quality
map at 10 0.0543
mrr at 10 0.0544
Performance L4 b1 c16
NFCorpus?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 70.8K tok/s
Query p50 45.9ms
NanoFiQA2018Retrieval
Smaller subset of the FiQA financial QA dataset
Performance L4 b1 c16
Query 7.5K tok/s
Query p50 388.1ms
SCIDOCS?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 53.7K tok/s
Query p50 42.5ms
SciFact?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 67.4K tok/s
Query p50 42.1ms
StackOverflowQA?candidates_model=Alibaba-NLP
Performance L4 b1 c16
Query 98.6K tok/s
Query p50 47.2ms