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

← Catalog

intfloat/multilingual-e5-small

Open comparison →

Primitive: /encode · Encode · BERT

Multilingual E5 Text Embeddings: A Technical Report. Liang Wang, Nan Yang, Xiaolong Huang, Linjun Yang, Rangan Majumder, Furu Wei, arXiv 2024

MultilingualDense

Overview

Hardware: — drives latency, throughput & cost

Size118M params
Tasks /encode
Licensemit
Languagesmultilingual, af, am, ar, as, az, be, bg, bn, br, bs, ca, cs, cy, da, de, el, en, eo, es, et, eu, fa, fi, fr, fy, ga, gd, gl, gu, ha, he, hi, hr, hu, hy, id, is, it, ja, jv, ka, kk, km, kn, ko, ku, ky, la, lo, lt, lv, mg, mk, ml, mn, mr, ms, my, ne, nl, no, om, or, pa, pl, ps, pt, ro, ru, sa, sd, si, sk, sl, so, sq, sr, su, sv, sw, ta, te, th, tl, tr, ug, uk, ur, uz, vi, xh, yi, zh
Latency28 ms
Throughput85.0K tok/s
Cost$0.0026 /1M tok

Cost is approximate — computed from list GPU prices; your actual price depends on the provider you deploy SIE with.

Embedding

Output typesDense
Dimensionsdense: 384
Max sequence length512
Inputstext

Benchmarks

CQADupstackPhysicsRetrieval

scientific retrieval en

Duplicate question retrieval from StackExchange Physics

Corpus: 38,314 Queries: 1,039
Quality
ndcg at 10 0.3843
map at 10 0.3361
mrr at 10 0.3894
Performance L4 b1 c16
Corpus 66.3K tok/s
Corpus p50 26.3ms
Query 5.6K tok/s
Query p50 25.0ms
Reference →

CosQA

technology retrieval en

Code search with natural language queries

Corpus: 6,267 Queries: 500
Quality
ndcg at 10 0.2870
map at 10 0.2222
mrr at 10 0.2448
Performance L4 b1 c16
Corpus 28.5K tok/s
Corpus p50 26.7ms
Query 2.5K tok/s
Query p50 26.1ms
Reference →

FiQA2018

finance retrieval en

Financial opinion mining and question answering

Corpus: 57,599 Queries: 648
Quality
ndcg at 10 0.3328
map at 10 0.2642
mrr at 10 0.4086
Performance L4 b1 c16
Corpus 78.7K tok/s
Corpus p50 27.5ms
Query 5.3K tok/s
Query p50 26.4ms
Reference →

LegalBenchConsumerContractsQA

legal retrieval en

Question answering on consumer contracts

Corpus: 153 Queries: 396
Quality
ndcg at 10 0.6698
map at 10 0.5987
mrr at 10 0.5986
Performance L4 b1 c16
Corpus 201.5K tok/s
Corpus p50 29.2ms
Query 8.6K tok/s
Query p50 24.2ms
Reference →

NFCorpus

medical retrieval en

Biomedical literature search from NutritionFacts.org

Corpus: 3,593 Queries: 323
Quality
ndcg at 10 0.3101
map at 10 0.1156
mrr at 10 0.5054
Performance L4 b1 c16
Corpus 118.9K tok/s
Corpus p50 29.6ms
Query 2.6K tok/s
Query p50 25.4ms
Reference →

SCIDOCS

scientific retrieval en

Citation prediction, document classification, and recommendation for scientific papers

Corpus: 25,656 Queries: 1,000
Quality
ndcg at 10 0.1390
map at 10 0.0787
mrr at 10 0.2561
Performance L4 b1 c16
Corpus 91.2K tok/s
Corpus p50 27.3ms
Query 5.8K tok/s
Query p50 24.6ms
Reference →

SciFact

scientific retrieval en

Scientific claim verification using research literature

Corpus: 5,183 Queries: 300
Quality
ndcg at 10 0.6770
map at 10 0.6294
mrr at 10 0.6451
Performance L4 b1 c16
Corpus 122.9K tok/s
Corpus p50 28.8ms
Query 8.6K tok/s
Query p50 25.5ms
Reference →

StackOverflowQA

technology retrieval en

Programming question answering from Stack Overflow

Corpus: 19,931 Queries: 1,994
Quality
ndcg at 10 0.8194
map at 10 0.7966
mrr at 10 0.7966
Performance L4 b1 c16
Corpus 78.8K tok/s
Corpus p50 29.6ms
Query 133.1K tok/s
Query p50 27.2ms
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 2.8K

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.