Jaswanth160/bge-base-financial-matryoshka

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: en
• 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("Jaswanth160/bge-base-financial-matryoshka")
# Run inference
sentences = [
    'The par call date for the 7% Notes due 2029 is August 15, 2025, allowing for redemption at par from this date onward.',
    'What is the earliest date on which the 7% Notes due 2029 can be redeemed at par?',
    'What are some of the initiatives managed by Visa for supporting underrepresented communities?',
]
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	0.6943
cosine_accuracy@3	0.8314
cosine_accuracy@5	0.8729
cosine_accuracy@10	0.9071
cosine_precision@1	0.6943
cosine_precision@3	0.2771
cosine_precision@5	0.1746
cosine_precision@10	0.0907
cosine_recall@1	0.6943
cosine_recall@3	0.8314
cosine_recall@5	0.8729
cosine_recall@10	0.9071
cosine_ndcg@10	0.8042
cosine_mrr@10	0.7709
cosine_map@100	0.7746

Information Retrieval

• Dataset: dim_512
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6986
cosine_accuracy@3	0.8371
cosine_accuracy@5	0.87
cosine_accuracy@10	0.9114
cosine_precision@1	0.6986
cosine_precision@3	0.279
cosine_precision@5	0.174
cosine_precision@10	0.0911
cosine_recall@1	0.6986
cosine_recall@3	0.8371
cosine_recall@5	0.87
cosine_recall@10	0.9114
cosine_ndcg@10	0.8076
cosine_mrr@10	0.7741
cosine_map@100	0.7777

Information Retrieval

• Dataset: dim_256
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.7
cosine_accuracy@3	0.83
cosine_accuracy@5	0.86
cosine_accuracy@10	0.9071
cosine_precision@1	0.7
cosine_precision@3	0.2767
cosine_precision@5	0.172
cosine_precision@10	0.0907
cosine_recall@1	0.7
cosine_recall@3	0.83
cosine_recall@5	0.86
cosine_recall@10	0.9071
cosine_ndcg@10	0.8048
cosine_mrr@10	0.772
cosine_map@100	0.7755

Information Retrieval

• Dataset: dim_128
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.67
cosine_accuracy@3	0.8186
cosine_accuracy@5	0.8571
cosine_accuracy@10	0.8971
cosine_precision@1	0.67
cosine_precision@3	0.2729
cosine_precision@5	0.1714
cosine_precision@10	0.0897
cosine_recall@1	0.67
cosine_recall@3	0.8186
cosine_recall@5	0.8571
cosine_recall@10	0.8971
cosine_ndcg@10	0.7868
cosine_mrr@10	0.7511
cosine_map@100	0.7552

Information Retrieval

• Dataset: dim_64
• Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.65
cosine_accuracy@3	0.7914
cosine_accuracy@5	0.8386
cosine_accuracy@10	0.8786
cosine_precision@1	0.65
cosine_precision@3	0.2638
cosine_precision@5	0.1677
cosine_precision@10	0.0879
cosine_recall@1	0.65
cosine_recall@3	0.7914
cosine_recall@5	0.8386
cosine_recall@10	0.8786
cosine_ndcg@10	0.7646
cosine_mrr@10	0.7278
cosine_map@100	0.7326

Training Details

Training Dataset

Unnamed Dataset

• Size: 6,300 training samples
• Columns: positive and anchor
• Approximate statistics based on the first 1000 samples:

  	positive	anchor
  type	string	string
  details	min: 6 tokens mean: 47.11 tokens max: 439 tokens	min: 7 tokens mean: 20.36 tokens max: 51 tokens

• Samples:

  positive	anchor
  For some of our medical membership, we share risk with providers under capitation contracts where physicians and hospitals accept varying levels of financial risk for a defined set of membership, primarily HMO membership.	What is the primary type of membership for which risk is shared with providers under capitation contracts?
  Revenue for Comcast's Theme Parks segment is primarily derived from guest spending at the theme parks, including ticket sales and in-park spending on food, beverages, and merchandise.	What is the primary revenue source for Comcast's Theme Parks segment?
  In August 2022, the Board of Directors authorized a program to repurchase up to $10.0 billion of the Company’s common stock, referred to as the "Share Repurchase Program". In February 2023, the Board of Directors authorized an additional $10.0 billion in repurchases under the Share Repurchase Program, bringing the aggregate total authorized to $20.0 billion.	What was the total authorization amount for the Share Repurchase Program of the Company as of February 2023?

• 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
• fp16: True
• tf32: False
• load_best_model_at_end: True
• optim: adamw_torch_fused
• 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: True
• 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_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: 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	Training Loss	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
0.8122	10	1.5811	-	-	-	-	-
0.9746	12	-	0.7341	0.7568	0.7632	0.7056	0.7660
1.6244	20	0.6854	-	-	-	-	-
1.9492	24	-	0.7516	0.7705	0.7722	0.7263	0.7702
2.4365	30	0.4874	-	-	-	-	-
2.9239	36	-	0.755	0.7747	0.7756	0.7321	0.7739
3.2487	40	0.3876	-	-	-	-	-
3.8985	48	-	0.7552	0.7755	0.7777	0.7326	0.7746

• 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: 0.33.0
• 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}
}