mixedbread-ai/mxbai-colbert-large-v1 (Encode)
The crispy rerank family from Mixedbread.
Benchmarks
CQADupstackPhysicsRetrieval
Duplicate question retrieval from StackExchange Physics
Corpus: 38,314 Queries: 1,039
Performance L4 b1 c16
Corpus 32.3K tok/s
Corpus p50 65.1ms
Query 3.9K tok/s
Query p50 44.7ms
CosQA
Code search with natural language queries
Corpus: 6,267 Queries: 500
Performance L4 b1 c16
Corpus 16.3K tok/s
Corpus p50 51.4ms
Query 2.4K tok/s
Query p50 40.1ms
FiQA2018
Financial opinion mining and question answering
Corpus: 57,599 Queries: 648
Performance L4 b1 c16
Corpus 38.0K tok/s
Corpus p50 66.6ms
Query 4.5K tok/s
Query p50 41.8ms
LegalBenchConsumerContractsQA
Question answering on consumer contracts
Corpus: 153 Queries: 396
Performance L4 b1 c16
Corpus 79.7K tok/s
Corpus p50 98.6ms
Query 6.3K tok/s
Query p50 42.1ms
NFCorpus
Biomedical literature search from NutritionFacts.org
Corpus: 3,593 Queries: 323
Quality
ndcg at 10 0.3467
map at 10 0.1321
mrr at 10 0.5620
Performance L4 b1 c16
Corpus 46.5K tok/s
Corpus p50 95.7ms
Query 1.9K tok/s
Query p50 42.8ms
NanoFiQA2018Retrieval
Smaller subset of the FiQA financial QA dataset
Quality
ndcg at 10 0.4833
map at 10 0.4103
mrr at 10 0.5605
SCIDOCS
Citation prediction, document classification, and recommendation for scientific papers
Corpus: 25,656 Queries: 1,000
Performance L4 b1 c16
Corpus 40.1K tok/s
Corpus p50 75.6ms
Query 4.6K tok/s
Query p50 39.8ms
SciFact
Scientific claim verification using research literature
Corpus: 5,183 Queries: 300
Performance L4 b1 c16
Corpus 48.5K tok/s
Corpus p50 87.4ms
Query 6.3K tok/s
Query p50 41.8ms
StackOverflowQA
Programming question answering from Stack Overflow
Corpus: 19,931 Queries: 1,994
Performance L4 b1 c16
Corpus 49.7K tok/s
Corpus p50 74.1ms
Query 76.4K tok/s
Query p50 60.7ms