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.

python

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 tokens
mean: 121.67 tokens
max: 256 tokens
min: 3 tokens
mean: 64.68 tokens
max: 256 tokens
min: 0.0
mean: 0.51
max: 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:
json
1{ 2 "loss_fct": "torch.nn.modules.loss.MSELoss" 3}

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

Click to expand

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

Epoch	Step	Training Loss
0.0661	500	0.0528
0.1322	1000	0.0298
0.1983	1500	0.0261
0.2644	2000	0.0242
0.3305	2500	0.0235
0.3966	3000	0.0223
0.4627	3500	0.0207
0.5288	4000	0.0208
0.5948	4500	0.0196
0.6609	5000	0.0192
0.7270	5500	0.019
0.7931	6000	0.0181
0.8592	6500	0.0181
0.9253	7000	0.0175
0.9914	7500	0.0178

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

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",
}

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

Devion333/labse-dhivehi-finetuned

SentenceTransformer based on sentence-transformers/LaBSE