README.md

---
tags:
- sentence-transformers
- sentence-similarity
- feature-extraction
- generated_from_trainer
- loss:MatryoshkaLoss
- loss:MultipleNegativesRankingLoss
- mteb
base_model: aubmindlab/bert-base-arabertv02
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
- pearson_cosine
- spearman_cosine
- pearson_manhattan
- spearman_manhattan
- pearson_euclidean
- spearman_euclidean
- pearson_dot
- spearman_dot
- pearson_max
- spearman_max
model-index:
- name: omarelshehy/Arabic-STS-Matryoshka-V2
  results:
  - dataset:
      config: ar-ar
      name: MTEB STS17 (ar-ar)
      revision: faeb762787bd10488a50c8b5be4a3b82e411949c
      split: test
      type: mteb/sts17-crosslingual-sts
    metrics:
    - type: pearson
      value: 85.1977
    - type: spearman
      value: 86.0559
    - type: cosine_pearson
      value: 85.1977
    - type: cosine_spearman
      value: 86.0559
    - type: manhattan_pearson
      value: 83.01950000000001
    - type: manhattan_spearman
      value: 85.28620000000001
    - type: euclidean_pearson
      value: 83.1524
    - type: euclidean_spearman
      value: 85.3787
    - type: main_score
      value: 86.0559
    task:
      type: STS
  - dataset:
      config: en-ar
      name: MTEB STS17 (en-ar)
      revision: faeb762787bd10488a50c8b5be4a3b82e411949c
      split: test
      type: mteb/sts17-crosslingual-sts
    metrics:
    - type: pearson
      value: 16.234
    - type: spearman
      value: 13.337499999999999
    - type: cosine_pearson
      value: 16.234
    - type: cosine_spearman
      value: 13.337499999999999
    - type: manhattan_pearson
      value: 11.103200000000001
    - type: manhattan_spearman
      value: 8.8513
    - type: euclidean_pearson
      value: 10.7335
    - type: euclidean_spearman
      value: 7.857
    - type: main_score
      value: 13.337499999999999
    task:
      type: STS
  - dataset:
      config: ar
      name: MTEB STS22 (ar)
      revision: de9d86b3b84231dc21f76c7b7af1f28e2f57f6e3
      split: test
      type: mteb/sts22-crosslingual-sts
    metrics:
    - type: pearson
      value: 49.8116
    - type: spearman
      value: 58.7217
    - type: cosine_pearson
      value: 49.8116
    - type: cosine_spearman
      value: 58.7217
    - type: manhattan_pearson
      value: 55.281499999999994
    - type: manhattan_spearman
      value: 58.658
    - type: euclidean_pearson
      value: 54.600300000000004
    - type: euclidean_spearman
      value: 58.59029999999999
    - type: main_score
      value: 58.7217
    task:
      type: STS
---

# SentenceTransformer based on aubmindlab/bert-base-arabertv02

🚀 🚀 This is **Arabic only** [sentence-transformers](https://www.SBERT.net) model finetuned from [aubmindlab/bert-base-arabertv02](https://huggingface.co./aubmindlab/bert-base-arabertv02). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for **semantic textual similarity**, **semantic search**, **clustering**, and more.

# Matryoshka Embeddings 🪆

This model supports Matryoshka embeddings, allowing you to truncate embeddings into smaller sizes to optimize performance and memory usage, based on your task requirements. Available truncation sizes include: **768**, **512**, **256**, **128**, and **64**

You can select the appropriate embedding size for your use case, ensuring flexibility in resource management.

## Model Details

### Model Description
- **Model Type:** Sentence Transformer
- **Base model:** [aubmindlab/bert-base-arabertv02](https://huggingface.co./aubmindlab/bert-base-arabertv02) <!-- at revision 016fb9d6768f522a59c6e0d2d5d5d43a4e1bff60 -->
- **Maximum Sequence Length:** 512 tokens
- **Output Dimensionality:** 768 tokens
- **Similarity Function:** Cosine Similarity

### Full Model Architecture

```
SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': 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("omarelshehy/Arabic-STS-Matryoshka-V2")
# Run inference
sentences = [
    'أحب قراءة الكتب في أوقات فراغي.',
    'أستمتع بقراءة القصص في المساء قبل النوم.',
    'القراءة تعزز معرفتي وتفتح أمامي آفاق جديدة.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

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

# 📊 Evaluation (Performance vs Embedding size)

I evaluated this model on the MTEB STS17 for arabic for different Embedding sizes 🪆

The results are plotted below:

![Plot](https://huggingface.co./omarelshehy/Arabic-STS-Matryoshka-V2/resolve/main/sts_matryoshka_v2_eval.png)

as seen from the plot, only very small degradation of performance happens across smaller matryoshka embedding sizes. 

## Citation

### BibTeX

#### Sentence Transformers
```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",
}
```

#### MatryoshkaLoss
```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}
}
```

#### MultipleNegativesRankingLoss
```bibtex
@misc{henderson2017efficient,
    title={Efficient Natural Language Response Suggestion for Smart Reply},
    author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
    year={2017},
    eprint={1705.00652},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}
```