noace/bge-m3-banking-73pct

SentenceTransformer based on BAAI/bge-m3

This is a sentence-transformers model finetuned from BAAI/bge-m3. It maps sentences & paragraphs to a 1024-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: BAAI/bge-m3
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 1024 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': 512, 'do_lower_case': False, 'architecture': 'XLMRobertaModel'})
  (1): Pooling({'word_embedding_dimension': 1024, '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): 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 = [
    'Quy đổi 1 lượt golf thành 1 đêm nghỉ dưỡng tiêu chuẩn cho 2 người.',
    'Mỗi lượt golf trong tài khoản tương đương với 01 đêm phòng tiêu chuẩn dành cho 02 khách.',
    'Giao dịch ở siêu thị bằng thẻ được hoàn lại giá trị',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[ 1.0000,  0.8742, -0.0163],
#         [ 0.8742,  1.0000, -0.0454],
#         [-0.0163, -0.0454,  1.0000]])

Evaluation

Metrics

Semantic Similarity

• Dataset: banking-val
• Evaluated with EmbeddingSimilarityEvaluator

Metric	Value
pearson_cosine	0.5023
spearman_cosine	0.5016

Training Details

Training Dataset

Unnamed Dataset

• Size: 1,109 training samples
• Columns: sentence1 and sentence2
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	min: 7 tokens mean: 17.54 tokens max: 31 tokens	min: 5 tokens mean: 19.86 tokens max: 34 tokens

• Samples:

  sentence1	sentence2
  Hạn mức chuyển tiền qua internet banking	Giới hạn giao dịch trên mobile banking mỗi ngày
  Lãi suất tiền gửi Tương lai kỳ hạn 1 năm là 3,70%/năm.	Sản phẩm Tiền gửi Tương lai 12 tháng có lãi suất 3,70%.
  Chi tiêu khác ngoài siêu thị và di chuyển được hoàn 0,5%.	Các giao dịch chi tiêu thông thường khác áp dụng tỷ lệ hoàn tiền là 0,5%.

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          1024,
          768,
          512,
          256,
          128
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• learning_rate: 5e-06
• num_train_epochs: 6
• warmup_ratio: 0.1
• fp16: True
• load_best_model_at_end: True
• batch_sampler: no_duplicates

All Hyperparameters

Click to expand

• overwrite_output_dir: False
• do_predict: False
• eval_strategy: epoch
• prediction_loss_only: True
• per_device_train_batch_size: 8
• per_device_eval_batch_size: 8
• 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-06
• weight_decay: 0.0
• adam_beta1: 0.9
• adam_beta2: 0.999
• adam_epsilon: 1e-08
• max_grad_norm: 1.0
• num_train_epochs: 6
• 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: True
• 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	banking-val_spearman_cosine
0.1429	10	0.9599	-
0.2857	20	1.241	-
0.4286	30	0.9945	-
0.5714	40	1.0378	-
0.7143	50	0.8056	-
0.8571	60	0.4594	-
1.0	70	0.5231	0.4477
1.1429	80	0.1718	-
1.2857	90	0.2573	-
1.4286	100	0.4365	-
1.5714	110	0.4087	-
1.7143	120	0.1634	-
1.8571	130	0.2878	-
2.0	140	0.2623	0.4541
2.1429	150	0.3152	-
2.2857	160	0.1694	-
2.4286	170	0.4442	-
2.5714	180	0.0521	-
2.7143	190	0.398	-
2.8571	200	0.2821	-
3.0	210	0.0689	0.4757
3.1429	220	0.0884	-
3.2857	230	0.1287	-
3.4286	240	0.114	-
3.5714	250	0.0855	-
3.7143	260	0.1124	-
3.8571	270	0.341	-
4.0	280	0.1434	0.4852
4.1429	290	0.0775	-
4.2857	300	0.221	-
4.4286	310	0.1457	-
4.5714	320	0.1224	-
4.7143	330	0.0609	-
4.8571	340	0.0364	-
5.0	350	0.0739	0.4978
5.1429	360	0.1004	-
5.2857	370	0.1364	-
5.4286	380	0.0411	-
5.5714	390	0.3235	-
5.7143	400	0.0252	-
5.8571	410	0.1668	-
6.0	420	0.0637	0.5016

• The bold row denotes the saved checkpoint.

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: 4.4.2
• 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",
}

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}
}