SentenceTransformer based on sentence-transformers/paraphrase-multilingual-mpnet-base-v2

This is a sentence-transformers model finetuned from sentence-transformers/paraphrase-multilingual-mpnet-base-v2 on the Omartificial-Intelligence-Space/arabic-n_li-triplet dataset. It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

Model Type: Sentence Transformer
Base model: sentence-transformers/paraphrase-multilingual-mpnet-base-v2
Maximum Sequence Length: 128 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity
Training Dataset:
- Omartificial-Intelligence-Space/arabic-n_li-triplet

Model Sources

Documentation: Sentence Transformers Documentation
Repository: Sentence Transformers on GitHub
Hugging Face: Sentence Transformers on Hugging Face

Full Model Architecture

SentenceTransformer(
  (0): Transformer({'max_seq_length': 128, 'do_lower_case': False}) with Transformer model: XLMRobertaModel 
  (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:

pip install -U sentence-transformers

Then you can load this model and run inference.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("Omartificial-Intelligence-Space/Arabic-Nli-Matryoshka")
# Run inference
sentences = [
    'يجلس شاب ذو شعر أشقر على الحائط يقرأ جريدة بينما تمر امرأة وفتاة شابة.',
    'ذكر شاب ينظر إلى جريدة بينما تمر إمرأتان بجانبه',
    'الشاب نائم بينما الأم تقود ابنتها إلى الحديقة',
]
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

Metrics

Semantic Similarity

Dataset: sts-test-768
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.8539
spearman_cosine	0.8616
pearson_manhattan	0.8497
spearman_manhattan	0.8513
pearson_euclidean	0.8516
spearman_euclidean	0.8541
pearson_dot	0.7281
spearman_dot	0.723
pearson_max	0.8539
spearman_max	0.8616

Semantic Similarity

Dataset: sts-test-512
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.8542
spearman_cosine	0.8609
pearson_manhattan	0.8487
spearman_manhattan	0.8512
pearson_euclidean	0.8505
spearman_euclidean	0.8539
pearson_dot	0.7076
spearman_dot	0.7029
pearson_max	0.8542
spearman_max	0.8609

Semantic Similarity

Dataset: sts-test-256
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.8486
spearman_cosine	0.8579
pearson_manhattan	0.8405
spearman_manhattan	0.8456
pearson_euclidean	0.8415
spearman_euclidean	0.8472
pearson_dot	0.6669
spearman_dot	0.6651
pearson_max	0.8486
spearman_max	0.8579

Semantic Similarity

Dataset: sts-test-128
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.839
spearman_cosine	0.8499
pearson_manhattan	0.8287
spearman_manhattan	0.8353
pearson_euclidean	0.8298
spearman_euclidean	0.8372
pearson_dot	0.5856
spearman_dot	0.5835
pearson_max	0.839
spearman_max	0.8499

Semantic Similarity

Dataset: sts-test-64
Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.8291
spearman_cosine	0.8429
pearson_manhattan	0.8101
spearman_manhattan	0.8221
pearson_euclidean	0.8129
spearman_euclidean	0.8255
pearson_dot	0.5067
spearman_dot	0.511
pearson_max	0.8291
spearman_max	0.8429

Training Details

Training Dataset

Omartificial-Intelligence-Space/arabic-n_li-triplet

Dataset: Omartificial-Intelligence-Space/arabic-n_li-triplet
Size: 557,850 training samples
Columns: anchor, positive, and negative

Approximate statistics based on the first 1000 samples:

	anchor	positive	negative
type	string	string	string
details	min: 5 tokens mean: 10.33 tokens max: 52 tokens	min: 5 tokens mean: 13.21 tokens max: 49 tokens	min: 5 tokens mean: 15.32 tokens max: 53 tokens

Samples:

anchor	positive	negative
`شخص على حصان يقفز فوق طائرة معطلة`	`شخص في الهواء الطلق، على حصان.`	`شخص في مطعم، يطلب عجة.`
`أطفال يبتسمون و يلوحون للكاميرا`	`هناك أطفال حاضرون`	`الاطفال يتجهمون`
`صبي يقفز على لوح التزلج في منتصف الجسر الأحمر.`	`الفتى يقوم بخدعة التزلج`	`الصبي يتزلج على الرصيف`

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "MultipleNegativesRankingLoss",
    "matryoshka_dims": [
        768,
        512,
        256,
        128,
        64
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

Evaluation Dataset

Omartificial-Intelligence-Space/arabic-n_li-triplet

Dataset: Omartificial-Intelligence-Space/arabic-n_li-triplet
Size: 6,584 evaluation samples
Columns: anchor, positive, and negative

Approximate statistics based on the first 1000 samples:

	anchor	positive	negative
type	string	string	string
details	min: 5 tokens mean: 21.86 tokens max: 105 tokens	min: 4 tokens mean: 10.22 tokens max: 49 tokens	min: 4 tokens mean: 11.2 tokens max: 33 tokens

Samples:

anchor	positive	negative
`امرأتان يتعانقان بينما يحملان حزمة`	`إمرأتان يحملان حزمة`	`الرجال يتشاجرون خارج مطعم`
`طفلين صغيرين يرتديان قميصاً أزرق، أحدهما يرتدي الرقم 9 والآخر يرتدي الرقم 2 يقفان على خطوات خشبية في الحمام ويغسلان أيديهما في المغسلة.`	`طفلين يرتديان قميصاً مرقماً يغسلون أيديهم`	`طفلين يرتديان سترة يذهبان إلى المدرسة`
`رجل يبيع الدونات لعميل خلال معرض عالمي أقيم في مدينة أنجليس`	`رجل يبيع الدونات لعميل`	`امرأة تشرب قهوتها في مقهى صغير`

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "MultipleNegativesRankingLoss",
    "matryoshka_dims": [
        768,
        512,
        256,
        128,
        64
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 256
per_device_eval_batch_size: 256
num_train_epochs: 1
warmup_ratio: 0.1
fp16: True
batch_sampler: no_duplicates

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
prediction_loss_only: True
per_device_train_batch_size: 256
per_device_eval_batch_size: 256
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_steps: None
learning_rate: 5e-05
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1.0
num_train_epochs: 1
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.1
warmup_steps: 0
log_level: passive
log_level_replica: warning
log_on_each_node: True
logging_nan_inf_filter: True
save_safetensors: True
save_on_each_node: False
save_only_model: False
no_cuda: False
use_cpu: False
use_mps_device: False
seed: 42
data_seed: None
jit_mode_eval: False
use_ipex: False
bf16: False
fp16: True
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: None
local_rank: 0
ddp_backend: None
tpu_num_cores: None
tpu_metrics_debug: False
debug: []
dataloader_drop_last: False
dataloader_num_workers: 0
dataloader_prefetch_factor: None
past_index: -1
disable_tqdm: False
remove_unused_columns: True
label_names: None
load_best_model_at_end: False
ignore_data_skip: False
fsdp: []
fsdp_min_num_params: 0
fsdp_config: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
fsdp_transformer_layer_cls_to_wrap: None
accelerator_config: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'gradient_accumulation_kwargs': None}
deepspeed: None
label_smoothing_factor: 0.0
optim: adamw_torch
optim_args: None
adafactor: False
group_by_length: False
length_column_name: length
ddp_find_unused_parameters: None
ddp_bucket_cap_mb: None
ddp_broadcast_buffers: False
dataloader_pin_memory: True
dataloader_persistent_workers: False
skip_memory_metrics: True
use_legacy_prediction_loop: False
push_to_hub: False
resume_from_checkpoint: None
hub_model_id: None
hub_strategy: every_save
hub_private_repo: False
hub_always_push: False
gradient_checkpointing: False
gradient_checkpointing_kwargs: None
include_inputs_for_metrics: False
eval_do_concat_batches: True
fp16_backend: auto
push_to_hub_model_id: None
push_to_hub_organization: None
mp_parameters:
auto_find_batch_size: False
full_determinism: False
torchdynamo: None
ray_scope: last
ddp_timeout: 1800
torch_compile: False
torch_compile_backend: None
torch_compile_mode: None
dispatch_batches: None
split_batches: None
include_tokens_per_second: False
include_num_input_tokens_seen: False
neftune_noise_alpha: None
optim_target_modules: None
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	sts-test-128_spearman_cosine	sts-test-256_spearman_cosine	sts-test-512_spearman_cosine	sts-test-64_spearman_cosine	sts-test-768_spearman_cosine
0.2294	500	10.1279	-	-	-	-	-
0.4587	1000	8.0384	-	-	-	-	-
0.6881	1500	7.3484	-	-	-	-	-
0.9174	2000	4.2216	-	-	-	-	-
1.0	2180	-	0.8499	0.8579	0.8609	0.8429	0.8616

Framework Versions

Python: 3.9.18
Sentence Transformers: 3.0.1
Transformers: 4.40.0
PyTorch: 2.2.2+cu121
Accelerate: 0.26.1
Datasets: 2.19.0
Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

@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

@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

@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}
}

Acknowledgments

The author would like to thank Prince Sultan University for their invaluable support in this project. Their contributions and resources have been instrumental in the development and fine-tuning of these models.

## Citation

If you use the Arabic Matryoshka Embeddings Model, please cite it as follows:

@misc{nacar2024enhancingsemanticsimilarityunderstanding,
      title={Enhancing Semantic Similarity Understanding in Arabic NLP with Nested Embedding Learning}, 
      author={Omer Nacar and Anis Koubaa},
      year={2024},
      eprint={2407.21139},
      archivePrefix={arXiv},
      primaryClass={cs.CL},
      url={https://arxiv.org/abs/2407.21139}, 
}