8Opt/mmbert-sts-0003-0001

SentenceTransformer based on jhu-clsp/mmBERT-small

This is a sentence-transformers model finetuned from jhu-clsp/mmBERT-small on the vn-sts-0003 dataset. It maps sentences & paragraphs to a 384-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: jhu-clsp/mmBERT-small
• Maximum Sequence Length: 8192 tokens
• Output Dimensionality: 384 dimensions
• Similarity Function: Cosine Similarity
• Training Dataset:
  • vn-sts-0003

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': 8192, 'do_lower_case': False, 'architecture': 'ModernBertModel'})
  (1): Pooling({'word_embedding_dimension': 384, '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("8Opt/mmbert-sts-0003-0001")
# Run inference
sentences = [
    'Chi phí chăm sóc trẻ em tăng gần 6%',
    "Chăm sóc trẻ em 'tiêu tốn 15.000 bảng mỗi năm'",
    'Không có con chó nào đuổi theo quả bóng trong cỏ.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.9201, 0.8759],
#         [0.9201, 1.0000, 0.8777],
#         [0.8759, 0.8777, 1.0000]])

Evaluation

Metrics

Semantic Similarity

• Datasets: 8Opt-sts-dev-0001 and 8Opt-sts-test-0001
• Evaluated with EmbeddingSimilarityEvaluator

Metric	8Opt-sts-dev-0001	8Opt-sts-test-0001
pearson_cosine	0.6497	0.6495
spearman_cosine	0.6934	0.6941

Training Details

Training Dataset

vn-sts-0003

• Dataset: vn-sts-0003 at 5164e7d
• Size: 21,543 training samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 5 tokens mean: 17.83 tokens max: 79 tokens	min: 6 tokens mean: 17.71 tokens max: 87 tokens	min: 0.01 mean: 2.55 max: 5.0

• Samples:

  sentence1	sentence2	score
  Một tay đua xe đạp trần truồng	Một tay đua xe đạp đang mặc đồ màu đen.	3.315
  Thật không may, câu trả lời cho câu hỏi của bạn là chúng tôi không biết.	Nếu cuộc trò chuyện không phải về công việc, bạn sẽ biết câu trả lời cho câu hỏi của mình là gì.	0.4
  Tòa án Trung Quốc duy trì án tử hình cho những kẻ giết người Mekong	Tòa án duy trì án tử hình cho những kẻ giết người Mekong	4.4

• Loss: CoSENTLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "pairwise_cos_sim"
  }
  

Evaluation Dataset

vn-sts-0003

• Dataset: vn-sts-0003 at 5164e7d
• Size: 3,077 evaluation samples
• Columns: sentence1, sentence2, and score
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2	score
  type	string	string	float
  details	min: 5 tokens mean: 18.03 tokens max: 80 tokens	min: 6 tokens mean: 17.6 tokens max: 71 tokens	min: 0.04 mean: 2.44 max: 5.0

• Samples:

  sentence1	sentence2	score
  Có một vài thứ để giúp bánh của bạn tăng cân một chút.	John Cavan có một số phương pháp hợp lý và thích hợp để tăng trọng lượng thỏ trong câu trả lời của mình.	3.0
  Tháng 4 / tháng 5 / tháng 6 năm 1963 trở thành một nhà lãnh đạo của các cuộc biểu tình dân quyền ở Greensboro, Bắc Carolina.	Mùa xuân năm 1966 trở thành người đứng đầu chi nhánh Chicago của CLC Operation Breadbasket.	0.5
  một người đàn ông đi dọc theo một con đường.	một người đàn ông đi qua một sợi dây thừng chặt chẽ.	0.8

• Loss: CoSENTLoss with these parameters:

  {
      "scale": 20.0,
      "similarity_fct": "pairwise_cos_sim"
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: steps
• per_device_train_batch_size: 16
• per_device_eval_batch_size: 16
• 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: 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: 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
• 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}
• parallelism_config: 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
• project: huggingface
• trackio_space_id: trackio
• 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: no
• 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: True
• prompts: None
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional
• router_mapping: {}
• learning_rate_mapping: {}

Training Logs

Epoch	Step	Training Loss	Validation Loss	8Opt-sts-dev-0001_spearman_cosine	8Opt-sts-test-0001_spearman_cosine
0.0742	100	4.7338	4.7076	0.4797	-
0.1485	200	4.6967	4.6637	0.5562	-
0.2227	300	4.6581	4.6458	0.6063	-
0.2970	400	4.6176	4.6038	0.6419	-
0.3712	500	4.6214	4.6475	0.5919	-
0.4454	600	4.6049	4.5982	0.6413	-
0.5197	700	4.598	4.5892	0.6367	-
0.5939	800	4.6088	4.5742	0.6751	-
0.6682	900	4.581	4.5980	0.6726	-
0.7424	1000	4.5617	4.5716	0.6776	-
0.8166	1100	4.5584	4.5545	0.6890	-
0.8909	1200	4.5537	4.5487	0.6929	-
0.9651	1300	4.5175	4.5535	0.6934	-
-1	-1	-	-	-	0.6941

Framework Versions

• Python: 3.12.12
• Sentence Transformers: 5.1.1
• Transformers: 4.57.1
• PyTorch: 2.8.0+cu126
• Accelerate: 1.11.0
• Datasets: 4.0.0
• Tokenizers: 0.22.1

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",
}

CoSENTLoss

@article{10531646,
    author={Huang, Xiang and Peng, Hao and Zou, Dongcheng and Liu, Zhiwei and Li, Jianxin and Liu, Kay and Wu, Jia and Su, Jianlin and Yu, Philip S.},
    journal={IEEE/ACM Transactions on Audio, Speech, and Language Processing},
    title={CoSENT: Consistent Sentence Embedding via Similarity Ranking},
    year={2024},
    doi={10.1109/TASLP.2024.3402087}
}