IlhamEbdesk/bge-base-financial-matryoshka_test_my

BGE base Financial Matryoshka

This is a sentence-transformers model finetuned from BAAI/bge-base-en-v1.5. 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: BAAI/bge-base-en-v1.5
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 tokens
• Similarity Function: Cosine Similarity
• Language: my
• 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': True}) with Transformer model: 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): 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("IlhamEbdesk/bge-base-financial-matryoshka_test_my")
# Run inference
sentences = [
    'Penyelaras kempen iklan adalah individu yang menyelaraskan semua aspek kempen iklan, termasuk jadual, pelaksanaan, dan laporan prestasi.',
    'Apakah itu penyelaras kempen iklan?',
    'Apakah itu pembuat roti?',
]
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]

Evaluation

Metrics

Information Retrieval

• Dataset: dim_768
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.8226
cosine_accuracy@3	0.9769
cosine_accuracy@5	0.9871
cosine_accuracy@10	0.9974
cosine_precision@1	0.8226
cosine_precision@3	0.3256
cosine_precision@5	0.1974
cosine_precision@10	0.0997
cosine_recall@1	0.8226
cosine_recall@3	0.9769
cosine_recall@5	0.9871
cosine_recall@10	0.9974
cosine_ndcg@10	0.9255
cosine_mrr@10	0.901
cosine_map@100	0.9011

Information Retrieval

• Dataset: dim_512
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.8046
cosine_accuracy@3	0.9743
cosine_accuracy@5	0.9871
cosine_accuracy@10	0.9923
cosine_precision@1	0.8046
cosine_precision@3	0.3248
cosine_precision@5	0.1974
cosine_precision@10	0.0992
cosine_recall@1	0.8046
cosine_recall@3	0.9743
cosine_recall@5	0.9871
cosine_recall@10	0.9923
cosine_ndcg@10	0.9159
cosine_mrr@10	0.8896
cosine_map@100	0.89

Information Retrieval

• Dataset: dim_256
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.7892
cosine_accuracy@3	0.9666
cosine_accuracy@5	0.9743
cosine_accuracy@10	0.9871
cosine_precision@1	0.7892
cosine_precision@3	0.3222
cosine_precision@5	0.1949
cosine_precision@10	0.0987
cosine_recall@1	0.7892
cosine_recall@3	0.9666
cosine_recall@5	0.9743
cosine_recall@10	0.9871
cosine_ndcg@10	0.9046
cosine_mrr@10	0.8764
cosine_map@100	0.8771

Information Retrieval

• Dataset: dim_128
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.7481
cosine_accuracy@3	0.9409
cosine_accuracy@5	0.9537
cosine_accuracy@10	0.9692
cosine_precision@1	0.7481
cosine_precision@3	0.3136
cosine_precision@5	0.1907
cosine_precision@10	0.0969
cosine_recall@1	0.7481
cosine_recall@3	0.9409
cosine_recall@5	0.9537
cosine_recall@10	0.9692
cosine_ndcg@10	0.8765
cosine_mrr@10	0.845
cosine_map@100	0.8461

Information Retrieval

• Dataset: dim_64
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.7224
cosine_accuracy@3	0.8972
cosine_accuracy@5	0.9254
cosine_accuracy@10	0.9434
cosine_precision@1	0.7224
cosine_precision@3	0.2991
cosine_precision@5	0.1851
cosine_precision@10	0.0943
cosine_recall@1	0.7224
cosine_recall@3	0.8972
cosine_recall@5	0.9254
cosine_recall@10	0.9434
cosine_ndcg@10	0.8455
cosine_mrr@10	0.8127
cosine_map@100	0.8146

Training Details

Training Dataset

Unnamed Dataset

• Size: 389 training samples
• Columns: positive and anchor
• Approximate statistics based on the first 1000 samples:

  	positive	anchor
  type	string	string
  details	min: 27 tokens mean: 61.59 tokens max: 139 tokens	min: 8 tokens mean: 15.26 tokens max: 24 tokens

• Samples:

  positive	anchor
  Dokter adalah profesional medis yang mendiagnosis dan merawat penyakit serta cedera pasien.	Apa itu dokter?
  Pereka sistem akuakultur adalah individu yang merancang dan membangunkan sistem untuk membiakkan ikan secara berkesan, termasuk reka bentuk kolam, sistem aliran air, dan pemantauan kualiti air.	Apakah itu pereka sistem akuakultur?
  Ahli sejarah seni adalah individu yang mengkaji perkembangan seni sepanjang sejarah dan konteks sosial, politik, dan budaya yang mempengaruhi penciptaannya. Mereka bekerja di muzium, galeri, dan institusi akademik, menganalisis karya seni	Apakah itu ahli sejarah seni?

• Loss: MatryoshkaLoss with these parameters:

  {
      "loss": "MultipleNegativesRankingLoss",
      "matryoshka_dims": [
          768,
          512,
          256,
          128,
          64
      ],
      "matryoshka_weights": [
          1,
          1,
          1,
          1,
          1
      ],
      "n_dims_per_step": -1
  }
  

Training Hyperparameters

Non-Default Hyperparameters

• eval_strategy: epoch
• per_device_train_batch_size: 32
• per_device_eval_batch_size: 16
• gradient_accumulation_steps: 16
• learning_rate: 2e-05
• num_train_epochs: 4
• lr_scheduler_type: cosine
• warmup_ratio: 0.1
• tf32: False
• 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: 32
• per_device_eval_batch_size: 16
• per_gpu_train_batch_size: None
• per_gpu_eval_batch_size: None
• gradient_accumulation_steps: 16
• eval_accumulation_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: 4
• max_steps: -1
• lr_scheduler_type: cosine
• 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: False
• fp16_opt_level: O1
• half_precision_backend: auto
• bf16_full_eval: False
• fp16_full_eval: False
• tf32: False
• 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}
• 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
• batch_sampler: no_duplicates
• multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	dim_128_cosine_map@100	dim_256_cosine_map@100	dim_512_cosine_map@100	dim_64_cosine_map@100	dim_768_cosine_map@100
1.0	1	0.6375	0.7065	0.7339	0.5984	0.7483
2.0	3	0.8282	0.8712	0.8821	0.7994	0.8929
2.4615	4	0.8461	0.8771	0.89	0.8146	0.9011

• The bold row denotes the saved checkpoint.

Framework Versions

• Python: 3.10.12
• Sentence Transformers: 3.0.1
• Transformers: 4.41.2
• PyTorch: 2.1.2+cu121
• Accelerate: 0.32.1
• Datasets: 2.19.1
• Tokenizers: 0.19.1

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}
}