README.md

---
tags:
  - sentence-transformers
  - feature-extraction
  - sentence-similarity
language:
  - es
  - en
inference: false
license: apache-2.0
---
<!-- TODO: add evaluation results here -->
<br><br>

<p align="center">
<img src="https://aeiljuispo.cloudimg.io/v7/https://cdn-uploads.huggingface.co/production/uploads/603763514de52ff951d89793/AFoybzd5lpBQXEBrQHuTt.png?w=200&h=200&f=face" alt="Finetuner logo: Finetuner helps you to create experiments in order to improve embeddings on search tasks. It accompanies you to deliver the last mile of performance-tuning for neural search applications." width="150px">
</p>


<p align="center">
<b>The text embedding set trained by <a href="https://jina.ai/"><b>Jina AI</b></a>.</b>
</p>

## Quick Start

The easiest way to starting using `jina-embeddings-v2-base-de` is to use Jina AI's [Embedding API](https://jina.ai/embeddings/).


## Intended Usage & Model Info

`jina-embeddings-v2-base-es` is a Spanish/English bilingual text **embedding model** supporting **8192 sequence length**.
It is based on a BERT architecture (JinaBERT) that supports the symmetric bidirectional variant of [ALiBi](https://arxiv.org/abs/2108.12409) to allow longer sequence length.
We have designed it for high performance in mono-lingual & cross-lingual applications and trained it specifically to support mixed Spanish-English input without bias. 
Additionally, we provide the following embedding models:

`jina-embeddings-v2-base-es` es un modelo (embedding) de texto bilingüe Inglés/Español que admite una longitud de secuencia de 8192.
Se basa en la arquitectura BERT (JinaBERT) que incorpora la variante bi-direccional simétrica de [ALiBi](https://arxiv.org/abs/2108.12409) para permitir una mayor longitud de secuencia.
Hemos diseñado este modelo para un alto rendimiento en aplicaciones monolingües y bilingües, y está entrenando específicamente para admitir entradas mixtas de español e inglés sin sesgo.
Adicionalmente, proporcionamos los siguientes modelos (embeddings):

- [`jina-embeddings-v2-small-en`](https://huggingface.co/jinaai/jina-embeddings-v2-small-en): 33 million parameters.
- [`jina-embeddings-v2-base-en`](https://huggingface.co/jinaai/jina-embeddings-v2-base-en): 137 million parameters.
- [`jina-embeddings-v2-base-zh`](): Chinese-English Bilingual embeddings (soon).
- [`jina-embeddings-v2-base-de`](): German-English Bilingual embeddings (soon).
- [`jina-embeddings-v2-base-es`](): Spanish-English Bilingual embeddings (soon) **(you are here)**.

## Data & Parameters

Jina Embeddings V2 [technical report](https://arxiv.org/abs/2310.19923)

## Usage

**<details><summary>Please apply mean pooling when integrating the model.</summary>**
<p>

### Why mean pooling?

`mean poooling` takes all token embeddings from model output and averaging them at sentence/paragraph level.
It has been proved to be the most effective way to produce high-quality sentence embeddings.
We offer an `encode` function to deal with this.

However, if you would like to do it without using the default `encode` function:

```python
import torch
import torch.nn.functional as F
from transformers import AutoTokenizer, AutoModel

def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0]
    input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
    return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)

sentences = ['How is the weather today?', 'What is the current weather like today?']

tokenizer = AutoTokenizer.from_pretrained('jinaai/jina-embeddings-v2-base-es')
model = AutoModel.from_pretrained('jinaai/jina-embeddings-v2-base-es', trust_remote_code=True)

encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

with torch.no_grad():
    model_output = model(**encoded_input)

embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
embeddings = F.normalize(embeddings, p=2, dim=1)
```

</p>
</details>

You can use Jina Embedding models directly from transformers package:
```python
!pip install transformers
from transformers import AutoModel
from numpy.linalg import norm

cos_sim = lambda a,b: (a @ b.T) / (norm(a)*norm(b))
model = AutoModel.from_pretrained('jinaai/jina-embeddings-v2-base-es', trust_remote_code=True) # trust_remote_code is needed to use the encode method
embeddings = model.encode(['How is the weather today?', '¿Qué tiempo hace hoy?'])
print(cos_sim(embeddings[0], embeddings[1]))
```

If you only want to handle shorter sequence, such as 2k, pass the `max_length` parameter to the `encode` function:

```python
embeddings = model.encode(
    ['Very long ... document'],
    max_length=2048
)
```

## Alternatives to Using Transformers Package

1. _Managed SaaS_: Get started with a free key on Jina AI's [Embedding API](https://jina.ai/embeddings/). 
2. _Private and high-performance deployment_: Get started by picking from our suite of models and deploy them on [AWS Sagemaker](https://aws.amazon.com/marketplace/seller-profile?id=seller-stch2ludm6vgy).

## Use Jina Embeddings for RAG

According to the latest blog post from [LLamaIndex](https://blog.llamaindex.ai/boosting-rag-picking-the-best-embedding-reranker-models-42d079022e83),

> In summary, to achieve the peak performance in both hit rate and MRR, the combination of OpenAI or JinaAI-Base embeddings with the CohereRerank/bge-reranker-large reranker stands out.

<img src="https://miro.medium.com/v2/resize:fit:4800/format:webp/1*ZP2RVejCZovF3FDCg-Bx3A.png" width="780px">


## Plans

1. Bilingual embedding models supporting more European & Asian languages, including French, Italian and Japanese.
2. Multimodal embedding models enable Multimodal RAG applications.
3. High-performt rerankers.

## Contact

Join our [Discord community](https://discord.jina.ai) and chat with other community members about ideas.

## Citation

If you find Jina Embeddings useful in your research, please cite the following paper:

```
@misc{günther2023jina,
      title={Jina Embeddings 2: 8192-Token General-Purpose Text Embeddings for Long Documents}, 
      author={Michael Günther and Jackmin Ong and Isabelle Mohr and Alaeddine Abdessalem and Tanguy Abel and Mohammad Kalim Akram and Susana Guzman and Georgios Mastrapas and Saba Sturua and Bo Wang and Maximilian Werk and Nan Wang and Han Xiao},
      year={2023},
      eprint={2310.19923},
      archivePrefix={arXiv},
      primaryClass={cs.CL}
}
```