Why did we open-source our inference engine? Read the post

colbert-ir/colbertv2.0 (Score)

[](https://colab.research.google.com/github/stanford-futuredata/ColBERT/blob/main/docs/intro2new.ipynb)

Overview

Architecture
BERT
Parameters
110M
Tasks
Encode, Score
Outputs
Multi-Vec
Dimensions
Multi-Vec: 128
Max Sequence Length
512 tokens
License
mit
Languages
en

Benchmarks

AskUbuntuDupQuestions

technology reranking en

Duplicate question detection from AskUbuntu

Corpus: 6,743 Queries: 360
Quality
ndcg at 10 0.5989
map at 10 0.4425
mrr at 10 0.6798
Reference →

CMedQAv1-reranking

general reranking en

Quality
ndcg at 10 0.1846
map at 10 0.1367
mrr at 10 0.1955

CMedQAv2-reranking

general reranking en

Quality
ndcg at 10 0.1979
map at 10 0.1478
mrr at 10 0.2017

CQADupstackPhysicsRetrieval

scientific retrieval en

Duplicate question retrieval from StackExchange Physics

Corpus: 38,314 Queries: 1,039
Performance L4 b1 c16
Corpus 33.2K tok/s
Corpus p50 58.6ms
Query 2.9K tok/s
Query p50 58.1ms
Reference →

CosQA

technology retrieval en

Code search with natural language queries

Corpus: 6,267 Queries: 500
Performance L4 b1 c16
Corpus 14.7K tok/s
Corpus p50 56.1ms
Query 1.8K tok/s
Query p50 51.8ms
Reference →

FiQA2018

finance retrieval en

Financial opinion mining and question answering

Corpus: 57,599 Queries: 648
Performance L4 b1 c16
Corpus 38.0K tok/s
Corpus p50 59.5ms
Query 3.9K tok/s
Query p50 47.3ms
Reference →

LegalBenchConsumerContractsQA

legal retrieval en

Question answering on consumer contracts

Corpus: 153 Queries: 396
Performance L4 b1 c16
Corpus 77.2K tok/s
Corpus p50 96.9ms
Query 5.4K tok/s
Query p50 49.0ms
Reference →

MMarcoReranking

general reranking zh

Multilingual MARCO passage reranking (Chinese)

Quality
ndcg at 10 0.0869
map at 10 0.0659
mrr at 10 0.0675
Reference →

NFCorpus

medical retrieval en

Biomedical literature search from NutritionFacts.org

Corpus: 3,593 Queries: 323
Performance L4 b1 c16
Corpus 42.8K tok/s
Corpus p50 89.2ms
Query 1.4K tok/s
Query p50 57.1ms
Reference →

SCIDOCS

scientific retrieval en

Citation prediction, document classification, and recommendation for scientific papers

Corpus: 25,656 Queries: 1,000
Performance L4 b1 c16
Corpus 41.3K tok/s
Corpus p50 66.2ms
Query 3.7K tok/s
Query p50 48.2ms
Reference →

SciFact

scientific retrieval en

Scientific claim verification using research literature

Corpus: 5,183 Queries: 300
Performance L4 b1 c16
Corpus 53.2K tok/s
Corpus p50 72.8ms
Query 5.1K tok/s
Query p50 51.0ms
Reference →

StackOverflowQA

technology retrieval en

Programming question answering from Stack Overflow

Corpus: 19,931 Queries: 1,994
Performance L4 b1 c16
Corpus 40.5K tok/s
Corpus p50 74.0ms
Query 62.4K tok/s
Query p50 75.6ms
Reference →

T2Reranking

general reranking zh

Chinese passage ranking benchmark

Quality
ndcg at 10 0.6533
map at 10 0.4716
mrr at 10 0.6980
Reference →

Open source inference for agents

Open-source inference for the models behind your agents. Run it yourself, or let us run it for you.

Github 2.1K

Contact us

Tell us about your use case and we'll get back to you shortly.

Apply for an inference grant

Free capacity on our hosted cluster for selected projects. Tell us what you run and we reply by email.