README.md

metadata

license: apache-2.0
language:
  - en
inference: false

The text embedding suit trained by Jina AI, Finetuner team.

Intented Usage & Model Info

jina-embedding-l-en-v1 is a language model that has been trained using Jina AI's Linnaeus-Clean dataset. This dataset consists of 380 million pairs of sentences, which include both query-document pairs. These pairs were obtained from various domains and were carefully selected through a thorough cleaning process. The Linnaeus-Full dataset, from which the Linnaeus-Clean dataset is derived, originally contained 1.6 billion sentence pairs.

The model has a range of use cases, including information retrieval, semantic textual similarity, text reranking, and more.

With a size of 330 million parameters, the model enables single-gpu inference while delivering better performance than our small and base model. Additionally, we provide the following options:

• jina-embedding-s-en-v1: 35 million parameters.
• jina-embedding-b-en-v1: 110 million parameters.
• jina-embedding-l-en-v1: 330 million parameters (you are here).
• jina-embedding-1b-en-v1: 1.2 billion parameters, 10* bert-base size (soon).
• jina-embedding-6b-en-v1: 6 billion parameters 30* bert-base size(soon).

Data & Parameters

More info will be released together with the technique report.

Metrics

We compared the model against all-minilm-l6-v2/all-mpnet-base-v2 from sbert and text-embeddings-ada-002 from OpenAI:

Name	param	context
all-minilm-l6-v2	33m	128
all-mpnet-base-v2	110m	128
ada-embedding-002	Unknown/OpenAI API	8192
jina-embedding-s-en-v1	35m	512
jina-embedding-b-en-v1	110m	512
jina-embedding-l-en-v1	330m	512

Name	STS12	STS13	STS14	STS15	STS16	STS17	TRECOVID	Quora	SciFact
all-minilm-l6-v2	0.724	0.806	0.756	0.854	0.79	0.876	0.473	0.876	0.645
all-mpnet-base-v2	0.726	0.835	0.78	0.857	0.8	0.906	0.513	0.875	0.656
ada-embedding-002	0.698	0.833	0.761	0.861	0.86	0.903	0.685	0.876	0.726
jina-embedding-s-en-v1	0.736	0.78	0.745	0.84	0.79	0.868	0.484	0.856	0.606
jina-embedding-b-en-v1	0.74	0.792	0.752	0.851	0.801	0.88	0.505	0.871	0.64
jina-embedding-l-en-v1	0.739	0.844	0.778	0.863	0.829	0.896	0.526	0.882	0.652

For more tasks and metrics, please checkout MTEB benchmark.

Usage

!pip install finetuner
import finetuner

model = finetuner.build_model('jinaai/jina-embedding-l-en-v1')
embeddings = finetuner.encode(
    model=model,
    data=['how is the weather today', 'What is the current weather like today?']
)
print(finetuner.cos_sim(embeddings[0], embeddings[1]))

Fine-tuning

Please consider Finetuner.

Plans

1. The development of jina-embedding-s-en-v2 is currently underway with two main objectives: improving performance and increasing the maximum sequence length.
2. We are currently working on a bilingual embedding model that combines English and X language. The upcoming model will be called jina-embedding-s/b/l-de-v1.

Contact

Join our Discord community and chat with other community members about ideas.