akhooli/sbert_ar_nli_500k_p100

BERT base trained on 500k Arabic NLI triplets

This is a sentence-transformers model finetuned from 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, paraphrase mining, text classification, clustering, and more.

Model Details

Model Description

• Model Type: Sentence Transformer
• Base model: aubmindlab/bert-base-arabertv02
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 tokens
• Similarity Function: Cosine Similarity
• Language: ar
• License: apache-2.0

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': 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:

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("sentence_transformers_model_id")
# Run inference
sentences = [
    'في أي مدينة تقع الحديقة الوطنية الجليدية',
    'الحديقة الجليدية الوطنية هي حديقة وطنية تقع في ولاية مونتانا الأمريكية ، على الحدود الكندية للولايات المتحدة مع المقاطعات الكندية في ألبرتا وكولومبيا البريطانية. حرائق الغابات الكبيرة غير شائعة في المنتزه. ومع ذلك ، في عام 2003 تم حرق أكثر من 13٪ من المتنزه. حديقة جلاسير الوطنية تقع على حدود متنزه ووترتون ليكس الوطني في كندا - يُعرف المنتزهان باسم منتزه واترتون-جلاسير الدولي للسلام وتم تصنيفهما كأول منتزه سلام دولي في العالم في عام 1932.',
    'تصوير: ايرين كونويل - رويترز. 1 بواسطة Alex Dobuzinskis. (2 رويترز) - قال مسؤولون إن حريقًا هائلًا في منتزه مونتانا الجليدي الوطني اندلع لليوم الرابع من خلال الأخشاب الثقيلة يوم الجمعة خلال ذروة موسم الزائرين ، بينما اجتاح حريق آخر في شمال كاليفورنيا الجبال فوق منطقة نبيذ وادي نابا.',
]
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]

Training Details

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• learning_rate: 2e-05
• 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
• eval_strategy: steps
• prediction_loss_only: True
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 32
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 1
• eval_accumulation_steps: None
• torch_empty_cache_steps: None
• learning_rate: 2e-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
• restore_callback_states_from_checkpoint: 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, 'non_blocking': False, '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_eval_metrics: False
• eval_on_start: False
• use_liger_kernel: False
• eval_use_gather_object: False
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	loss
0.016	250	4.087	-
0.032	500	1.9943	-
0.048	750	1.4472	-
0.064	1000	1.2324	-
0.08	1250	1.0402	-
0.096	1500	1.0357	-
0.112	1750	0.8857	-
0.128	2000	0.8617	-
0.144	2250	0.8101	-
0.16	2500	0.8452	-
0.176	2750	0.7949	-
0.192	3000	0.7706	-
0.208	3250	0.7518	-
0.224	3500	0.7217	-
0.24	3750	0.7225	-
0.256	4000	0.6761	-
0.272	4250	0.6492	-
0.288	4500	0.6379	-
0.304	4750	0.6225	-
0.32	5000	0.5899	0.5937
0.336	5250	0.6406	-
0.352	5500	0.6109	-
0.368	5750	0.5964	-
0.384	6000	0.5325	-
0.4	6250	0.5633	-
0.416	6500	0.5652	-
0.432	6750	0.6109	-
0.448	7000	0.527	-
0.464	7250	0.5215	-
0.48	7500	0.5508	-
0.496	7750	0.5832	-
0.512	8000	0.5817	-
0.528	8250	0.5617	-
0.544	8500	0.4963	-
0.56	8750	0.5168	-
0.576	9000	0.5251	-
0.592	9250	0.5439	-
0.608	9500	0.4962	-
0.624	9750	0.5638	-
0.64	10000	0.4764	0.4306
0.656	10250	0.531	-
0.672	10500	0.4901	-
0.688	10750	0.5076	-
0.704	11000	0.4384	-
0.72	11250	0.4971	-
0.736	11500	0.4457	-
0.752	11750	0.4603	-
0.768	12000	0.4854	-
0.784	12250	0.4702	-
0.8	12500	0.5154	-
0.816	12750	0.4619	-
0.832	13000	0.4829	-
0.848	13250	0.5101	-
0.864	13500	0.4641	-
0.88	13750	0.4797	-
0.896	14000	0.4632	-
0.912	14250	0.4578	-
0.928	14500	0.4552	-
0.944	14750	0.4636	-
0.96	15000	0.4764	0.4142
0.976	15250	0.5066	-
0.992	15500	0.4567	-

Framework Versions

• Python: 3.10.14
• Sentence Transformers: 3.1.1
• Transformers: 4.45.1
• PyTorch: 2.4.0
• Accelerate: 0.34.2
• Datasets: 3.0.1
• Tokenizers: 0.20.0

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",
}

Matryoshka2dLoss

@misc{li20242d,
    title={2D Matryoshka Sentence Embeddings},
    author={Xianming Li and Zongxi Li and Jing Li and Haoran Xie and Qing Li},
    year={2024},
    eprint={2402.14776},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

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}
}