E5 Base vs E5 Base v2
MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages) Embeddings assign examples to categories, group similar examples, estimate semantic similarity between pairs, classify relationships between pairs, rank relevant documents for queries, find translated sentence pairs, rerank documents according to instructions, assign multiple labels to each example, reorder candidate documents by relevance on MTEB Multilingual v2 · mean over tasks in 1037 languages including 'Auhelawa, Abau, Achi, Achinese, Achuar-Shiwiar, Adzera, Afrikaans, Agarabi; tasks receive equal weight
E5 Base—
E5 Base v246
| Benchmark | E5 Base | E5 Base v2 |
|---|---|---|
| MTEB multilingual v2 · Mean over tasks · multilingual (1037 languages) | — | 46 |
Specs
| Spec | E5 Base | E5 Base v2 |
|---|---|---|
| Price | — | !$0.005 |
| Context | 1K | 1K |
| Weights | Open | Open |
| Released | Dec 2022 | Feb 2024 |
| Dimensions | 768 | 768 |
| Max input | 1K tokens | 1K tokens |
| Languages | en | en |
| Memory (4-bit) | 0.1 GB | 0.1 GB |
| Licence | Apache-2.0 | MIT |
| Commercial use | Allowed | Allowed |
What breaks when you switch
E5 Base → E5 Base v2
Re-embed your whole corpus›
Vectors from different models can't be mixed.
E5 Base v2 → E5 Base
Re-embed your whole corpus›
Vectors from different models can't be mixed.