---
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- dataset_size:2280319
- loss:MatryoshkaLoss
- loss:MultipleNegativesRankingLoss
- Arabic
- EuroBert
- Semantic
base_model: EuroBERT/EuroBERT-2.1B
widget:
- source_sentence: امرأة شقراء تطل على مشهد (سياتل سبيس نيدل)
  sentences:
  - رجل يستمتع بمناظر جسر البوابة الذهبية
  - فتاة بالخارج تلعب في الثلج
  - شخص ما يأخذ في نظرة إبرة الفضاء.
- source_sentence: سوق الشرق الأوسط
  sentences:
  - مسرح أمريكي
  - متجر في الشرق الأوسط
  - البالغون صغار
- source_sentence: رجلين يتنافسان في ملابس فنون الدفاع عن النفس
  sentences:
  - هناك العديد من الناس الحاضرين.
  - الكلب الأبيض على الشاطئ
  - هناك شخص واحد فقط موجود.
- source_sentence: مجموعة من الناس تمشي بجانب شاحنة.
  sentences:
  - الناس يقفون
  - بعض الناس بالخارج
  - بعض الرجال يقودون على الطريق
- source_sentence: لاعبة كرة ناعمة ترمي الكرة إلى زميلتها في الفريق
  sentences:
  - شخصان يلعبان كرة البيسبول
  - الرجل ينظف
  - لاعبين لكرة البيسبول يجلسان على مقعد
pipeline_tag: feature-extraction
library_name: sentence-transformers
metrics:
- pearson_cosine
- spearman_cosine
model-index:
- name: SentenceTransformer based on EuroBERT/EuroBERT-2.1B
  results:
  - task:
      type: semantic-similarity
      name: Semantic Similarity
    dataset:
      name: sts dev 2304
      type: sts-dev-2304
    metrics:
    - type: pearson_cosine
      value: 0.7268081506530346
      name: Pearson Cosine
    - type: spearman_cosine
      value: 0.7297515677640904
      name: Spearman Cosine
  - task:
      type: semantic-similarity
      name: Semantic Similarity
    dataset:
      name: sts dev 1152
      type: sts-dev-1152
    metrics:
    - type: pearson_cosine
      value: 0.7267206088516627
      name: Pearson Cosine
    - type: spearman_cosine
      value: 0.72990680019899
      name: Spearman Cosine
  - task:
      type: semantic-similarity
      name: Semantic Similarity
    dataset:
      name: sts dev 960
      type: sts-dev-960
    metrics:
    - type: pearson_cosine
      value: 0.7262504495546784
      name: Pearson Cosine
    - type: spearman_cosine
      value: 0.7297095834107545
      name: Spearman Cosine
  - task:
      type: semantic-similarity
      name: Semantic Similarity
    dataset:
      name: sts dev 580
      type: sts-dev-580
    metrics:
    - type: pearson_cosine
      value: 0.7246438630418187
      name: Pearson Cosine
    - type: spearman_cosine
      value: 0.7285747858670478
      name: Spearman Cosine
license: mit
language:
- ar
---

# Ara-EuroBERT: Arabic Semantic Text Embeddings at Scale

<img src="https://i.ibb.co/d4svDscP/Clear-Familiar-situations-that-you-already-have-best-practices-for-4.png" width="170" align="left"/>

**Ara-EuroBERT-2.1B** is a [sentence-transformers](https://www.SBERT.net) model fine-tuned from [EuroBERT/EuroBERT-2.1B](https://huggingface.co/EuroBERT/EuroBERT-2.1B), specifically optimized for **Arabic semantic embeddings**.

This model maps sentences and paragraphs to a **2304-dimensional dense vector space** and supports **up to 8,192 tokens** in a single input sequence.

This model supports **Matryoshka (nested) embeddings** with the following dimensions:
- Full dimension: 2304
- Reduced dimensions:  1151, 960 , 580

You can select the embedding dimension based on your specific requirements, trading off between performance and computational efficiency.

It is suitable for:
- Semantic textual similarity
- Semantic search
- Paraphrase mining
- Text classification
- Clustering
- And other Arabic NLP tasks

You can find more information on the base model at [EuroBERT/EuroBERT-2.1B](https://huggingface.co/EuroBERT/EuroBERT-2.1B)

<br clear="left"/>

---

## 🧠 Model Details & Benchmark Performance


![image/png](https://cdn-uploads.huggingface.co/production/uploads/628f7a71dd993507cfcbe587/qcT6MrSY1RY_RX9lSJuQl.png)


### 🔬 Benchmark Highlights

- **STS17 Benchmark**: AraEuroBERT-2.1B achieves a score of **79**, significantly outperforming the standard EuroBERT-2.1B (**12**).
- **STS22.v2 Benchmark**: Scores **55**, competitive with much smaller and more efficient models.

## 🧪 Metrics

### Semantic Similarity

* Datasets: `sts-dev-2304`, `sts-dev-1152`, `sts-dev-960`, `sts-dev-580`
* Evaluated with [`EmbeddingSimilarityEvaluator`](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.EmbeddingSimilarityEvaluator)

| Metric              | sts-dev-2304 | sts-dev-1152 | sts-dev-960 | sts-dev-580 |
|---------------------|--------------|--------------|-------------|-------------|
| Pearson (cosine)    | 0.7268       | 0.7267       | 0.7263      | 0.7246      |
| **Spearman (cosine)** | **0.7298**   | **0.7299**   | **0.7297**  | **0.7286**  |


## 📦 Model Description

- **Model Type:** Sentence Transformer
- **Base Model:** [EuroBERT/EuroBERT-2.1B](https://huggingface.co/EuroBERT/EuroBERT-2.1B)
- **Maximum Sequence Length:** 8192 tokens
- **Output Dimensionality:** 2304 dimensions (supports nested dimensions: 2304, 1152, 960, 580)
- **Similarity Function:** Cosine Similarity
- **Language:** Arabic


## 🧬 Full Model Architecture

```python
SentenceTransformer(
  (0): Transformer({'max_seq_length': 8192, 'do_lower_case': False}) with Transformer model: EuroBertModel 
  (1): Pooling({
        'word_embedding_dimension': 2304,
        'pooling_mode_cls_token': False,
        'pooling_mode_mean_tokens': True,
        'pooling_mode_max_tokens': False,
        'include_prompt': True
  })
)
```

## Usage

### Direct Usage (Sentence Transformers)

First install the Sentence Transformers library:

```bash
pip install -U sentence-transformers
```

Then you can load this model and run inference:

```python
from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("Omartificial-Intelligence-Space/AraEuroBert-2.1B")

# Run inference
sentences = [
    'لاعبة كرة ناعمة ترمي الكرة إلى زميلتها في الفريق',
    'شخصان يلعبان كرة البيسبول',
    'لاعبين لكرة البيسبول يجلسان على مقعد',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1152]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]
```

## Citation

If you use this model in your research, please cite the following works:

```bibtex
@misc{boizard2025eurobertscalingmultilingualencoders,
      title={EuroBERT: Scaling Multilingual Encoders for European Languages}, 
      author={Nicolas Boizard and Hippolyte Gisserot-Boukhlef and Duarte M. Alves and André Martins and Ayoub Hammal and Caio Corro and Céline Hudelot and Emmanuel Malherbe and Etienne Malaboeuf and Fanny Jourdan and Gabriel Hautreux and João Alves and Kevin El-Haddad and Manuel Faysse and Maxime Peyrard and Nuno M. Guerreiro and Patrick Fernandes and Ricardo Rei and Pierre Colombo},
      year={2025},
      eprint={2503.05500},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2503.05500}, 
}
```

```bibtex
@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "https://arxiv.org/abs/1908.10084",
}
```

```bibtex
@misc{kusupati2024matryoshka,
    title={Matryoshka Representation Learning},
    author={Aditya Kusupati and Gantavya Bhatt and Aniket Rege and Matthew Wallingford and Aditya Sinha and Vivek Ramanujan and William Howard-Snyder and Kaifeng Chen and Sham Kakade and Prateek Jain and Ali Farhadi},
    year={2024},
    eprint={2205.13147},
    archivePrefix={arXiv},
    primaryClass={cs.LG}
}
```