← Catalog
intfloat/e5-large-v2
Open comparison →
Primitive: /encode · Encode ·
BERT
Text Embeddings by Weakly-Supervised Contrastive Pre-training. Liang Wang, Nan Yang, Xiaolong Huang, Binxing Jiao, Linjun Yang, Daxin Jiang, Rangan Majumder, Furu Wei, arXiv 2022
Dense
View on Hugging Face →
Overview
Hardware:
— drives latency, throughput & cost
| Size | 335M params |
|---|
| Tasks | /encode |
|---|
| License | mit |
|---|
| Languages | en |
|---|
| Latency | 87 ms |
|---|
| Throughput | 33.2K tok/s |
|---|
| Cost | $0.0067 /1M tok |
|---|
Cost is approximate — computed from list GPU prices; your actual price depends on the provider you deploy SIE with.
Embedding
| Output types | Dense |
|---|
| Dimensions | dense: 1,024 |
|---|
| Max sequence length | 512 |
|---|
| Inputs | text |
|---|
Benchmarks
Duplicate question retrieval from StackExchange Physics
Corpus: 38,314 Queries: 1,039
Quality
map at 10 0.3397
mrr at 10 0.3843
ndcg at 10 0.3941
Performance L4 b1 c16
Corpus 26.5K tok/s
Corpus p50 74.3ms
Query 2.7K tok/s
Query p50 53.9ms
Reference →
Code search with natural language queries
Corpus: 6,267 Queries: 500
Quality
map at 10 0.2378
mrr at 10 0.2511
ndcg at 10 0.3097
Performance L4 b1 c16
Corpus 13.9K tok/s
Corpus p50 60.1ms
Query 1.4K tok/s
Query p50 57.6ms
Reference →
Financial opinion mining and question answering
Corpus: 57,599 Queries: 648
Quality
map at 10 0.2797
mrr at 10 0.4311
ndcg at 10 0.3513
Performance L4 b1 c16
Corpus 31.6K tok/s
Corpus p50 80.9ms
Query 2.8K tok/s
Query p50 55.5ms
Reference →
Question answering on consumer contracts
Corpus: 153 Queries: 396
Quality
map at 10 0.7228
mrr at 10 0.7241
ndcg at 10 0.7740
Performance L4 b1 c16
Corpus 57.2K tok/s
Corpus p50 140.9ms
Query 3.5K tok/s
Query p50 58.3ms
Reference →
Biomedical literature search from NutritionFacts.org
Corpus: 3,593 Queries: 323
Quality
ndcg at 10 0.3715
map at 10 0.1416
mrr at 10 0.5706
Performance L4 b1 c16
Corpus 42.8K tok/s
Corpus p50 112.1ms
Query 1.3K tok/s
Query p50 53.6ms
Reference →
Smaller subset of the FiQA financial QA dataset
Quality
ndcg at 10 0.4531
map at 10 0.3742
mrr at 10 0.5054
Performance L4 b1 c16
Corpus 27.3K tok/s
Corpus p50 86.6ms
Query 2.8K tok/s
Query p50 49.5ms
Reference →
Citation prediction, document classification, and recommendation for scientific papers
Corpus: 25,656 Queries: 1,000
Quality
map at 10 0.0797
mrr at 10 0.2531
ndcg at 10 0.1397
Performance L4 b1 c16
Corpus 33.2K tok/s
Corpus p50 86.0ms
Query 2.7K tok/s
Query p50 56.2ms
Reference →
Scientific claim verification using research literature
Corpus: 5,183 Queries: 300
Quality
map at 10 0.6302
mrr at 10 0.6422
ndcg at 10 0.6775
Performance L4 b1 c16
Corpus 39.4K tok/s
Corpus p50 108.1ms
Query 3.6K tok/s
Query p50 58.1ms
Reference →
Programming question answering from Stack Overflow
Corpus: 19,931 Queries: 1,994
Quality
map at 10 0.8838
mrr at 10 0.8837
ndcg at 10 0.8991
Performance L4 b1 c16
Corpus 39.2K tok/s
Corpus p50 95.0ms
Query 42.1K tok/s
Query p50 107.3ms
Reference →
