SentenceTransformer based on sentence-transformers/LaBSE

This is a sentence-transformers model finetuned from sentence-transformers/LaBSE. 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/LaBSE
• Maximum Sequence Length: 256 tokens
• Output Dimensionality: 768 dimensions
• Similarity Function: Cosine Similarity

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': 256, 'do_lower_case': False, 'architecture': 'BertModel'})
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': True, 'pooling_mode_mean_tokens': False, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
  (2): Dense({'in_features': 768, 'out_features': 768, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
  (3): Normalize()
)

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 = [
    'އޭނާ ރީތި.',
    "She's pretty.",
    'Words of gratitude are being sent to the government and President Yameen for bringing two new generators to the village within five days. The people of Thonadhoo have shown the whole country that they have a people who love patience, unity and brotherhood. It is a beautiful example of unity. The burden and pain of the power outages is not easy for anyone to bear in such an era.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 768]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000,  0.9827, -0.0089],
#         [ 0.9827,  1.0000, -0.0044],
#         [-0.0089, -0.0044,  1.0000]])

Training Details

Training Dataset

Unnamed Dataset

• Size: 968,266 training samples
• Columns: sentence_0, sentence_1, and label
• Approximate statistics based on the first 1000 samples:

  	sentence_0	sentence_1	label
  type	string	string	float
  details	min: 3 tokensmean: 121.67 tokensmax: 256 tokens	min: 3 tokensmean: 64.68 tokensmax: 256 tokens	min: 0.0mean: 0.51max: 1.0

• Samples:

  sentence_0	sentence_1	label
  އިންތިހާބު ލަސްކުރަން ބްލެޓާ ބޭނުމެއްނުވޭ: ފީފާ	The Ponoru River is a tributary of the Horezu in Romania .	0.0
  ޖޯ އުފަންވީ 27 މާރޗް 1929 ގައި މެސެޗުސެޓްސްގެ ސޮމަރވިލް އަށް ކަމަށާއި ބޮޑުވީ މެސެޗުސެޓްސްގެ ކުއިންސީ ގައެވެ .	The National Inquiry Commission set up by the government of President Mohammed Vaheed Hassan Manik has said that the coup was not a coup and that the government was overthrown according to the rules of law.	0.0
  ސާބިތު ދަރަނީގެ މައްސަލައެއްގައި ޑރ. މުހައްމަދު ޖަމީލްގެ އަނބިކަނބަލުން ހައުލަތު ފަހީމް އާއި އެ އާއިލާގެ އިތުރު ހަތް މީހެއްގެ ޕާސްޕޯޓް ހިފަހައްޓައި ބޭންކް އެކައުންޓްތައް ފްރީޒްކުރުމަށް ސިވިލް ކޯޓުން މިއަދު އަމުރު ނެރެފި އެވެ.ވީބީ އައްޑޫ އެފްސީގެ މުއައްސިސެއް ކަމަށްވާ މުހަންމަދު ޝަވީދުގެ ވެސް ބައްޕަ މަރުހޫމް މޫސާ ފަހީމްގެ އަށް ވާރިސުންގެ ޕާސްޕޯޓާއި، ރާއްޖޭގެ ބޭންކްތަކުގައި ހުރި ހުރިހާ އެކައުންޓެއް ހިފަހައްޓަން ސިވިލް ކޯޓުން މިއަދު ހެނދުނު ނިންމީ، ތްރީއޭ ކޮމްޕެނީ ނުބަހާއިރު އެ ކުންފުނީގެ ހަތަރު ރިސޯޓެއްގެ ޖެނެރަލް މެނޭޖަރެއްގެ ގޮތުގައި ވަޒީފާ އަދާކުރި މުހަންމަދު އަނީސް ކޮށްފައިވާ ދައުވާއަކާ ގުޅިގެން ބޭއްވި ޝަރީއަތުގެ އަޑުއެހުމުގަ އެވެ. އެ އަޑުއެހުމަށް ވާރިސުންގެ ފަރާތުން ހާޒިރެއް ނުވެ އެވެ. ވާރިސުންގެ ވަކީލް ވިދާޅުވީ ސިވިލް ކޯޓުގެ ހުކުމް ހައި ކޯޓަށް އިސްތިއުނާފަށް ހުށަހަޅާފައިވާ ކަމަށެވެ.ސާބިތު ދަރަނީގެ ކޮންމެ މައްސަލައެއްގައި ވެސް ދަރަނި އަދާނުކުރާ ހާލަތެއްގައި، ހުކުމް ތަންފީޒުކުރުމަށް އެދި ހުށަހަޅެމުން ޕާސްޕޯޓް ހިފަހައްޓައި އެކައުންޓުތައް ފްރީޒްކުރުމަކީ ކޯޓުން އަމަލުކުރާ އާންމު އުސޫލެވ...	The Civil Court of Maldives has ordered the seizure of passports and freezing bank accounts belonging to Haulath Faheem, wife of former President Dr. Mohamed Jamil, as well as seven other members of his family in connection with a case of proven debt. This was decided by the court today after an action filed by Mohammad Aniis who served as General Manager at four resorts owned by Three A Company when it was not being divided into shares. The heir was not present at the court. The lawyer for the heirs said that he has appealed to the High Court against this decision. In any case of proven debt, it is a common practice in courts to hold passports and freeze accounts as part of an application for enforcement of judgment when there are no payments made by debtors. The family appealed the Civil Court’s order to pay them back, which was then reviewed by the Supreme Court. In addition to the three charges, Anies also brought another two cases against Musa Fahim’s heirs. The other accused are ...	1.0

• Loss: CosineSimilarityLoss with these parameters:

  {
      "loss_fct": "torch.nn.modules.loss.MSELoss"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• num_train_epochs: 1
• multi_dataset_batch_sampler: round_robin

All Hyperparameters

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: no
• prediction_loss_only: True
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• 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: 5e-05
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1
• num_train_epochs: 1
• max_steps: -1
• lr_scheduler_type: linear
• lr_scheduler_kwargs: {}
• warmup_ratio: 0.0
• 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: False
• 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_fused
• 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: None
• hub_always_push: False
• hub_revision: None
• gradient_checkpointing: False
• gradient_checkpointing_kwargs: None
• include_inputs_for_metrics: False
• include_for_metrics: []
• 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
• 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
• liger_kernel_config: None
• eval_use_gather_object: False
• average_tokens_across_devices: False
• prompts: None
• batch_sampler: batch_sampler
• multi_dataset_batch_sampler: round_robin
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 5.1.0
• Transformers: 4.55.2
• PyTorch: 2.8.0+cu128
• Accelerate: 1.9.0
• Datasets: 3.6.0
• Tokenizers: 0.21.4

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",
}