Add new SentenceTransformer model

5619cce verified over 1 year ago

30 kB

language:
  - en
license: apache-2.0
tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:6300
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
base_model: BAAI/bge-base-en-v1.5
widget:
  - source_sentence: >-
      What year do the patent families related to DARZALEX expire in the United
      States?
    sentences:
      - >-
        Amortization for owned content predominantly monetized on an individual
        basis and accrued costs associated with participations and residuals
        payments are recorded using the individual film forecast computation
        method, which recognizes the costs in the same ratio as the associated
        ultimate revenue.
      - The two patent families both expire in the United States in 2029.
      - >-
        For the year ended December 31, 2022, net cash used in investing
        activities of $371.9 million was primarily from the purchase of $247.3
        million marketable securities, net of sale and maturities, $62.2 million
        net cash used to acquire GreenCom, SolarLeadFactory and ClipperCreek,
        $46.4 million used in purchases of test and assembly equipment to expand
        our supply capacity, related facility improvements and information
        technology enhancements, including capitalized costs related to
        internal-use software and $16.0 million used to invest in private
        companies.
  - source_sentence: >-
      What legal claims does Fortis Advisors LLC allege against Ethicon Inc. in
      the lawsuit related to the acquisition of Auris Health Inc.?
    sentences:
      - >-
        Payments include a single lump-sum per treatment, referred to as bundled
        rates, or in other cases separate payments for dialysis treatments and
        pharmaceuticals, referred to as FFS rates.
      - >-
        In October 2020, Fortis Advisors LLC filed a complaint against Ethicon
        Inc. and others in Delaware's Court of Chancery. The lawsuit alleges
        breach of contract and fraud related to Ethicon's acquisition of Auris
        Health Inc. in 2019. The case underwent a partial dismissal in December
        2021, and as of January 2024, the trial's decision is pending.
      - >-
        On September 5, 2023, ICE acquired 100% of Black Knight for aggregate
        transaction consideration of approximately $11.8 billion, or $76 per
        share of Black Knight common stock, with cash comprising 90% of the
        value of the aggregate transaction consideration. The aggregate cash
        component of the transaction consideration was $10.5 billion. ICE issued
        10.9 million shares of its common stock to Black Knight stockholders,
        which was based on the market price of the common stock and the average
        of the volume weighted averages of the trading prices of the common
        stock on each of the ten consecutive trading days ending three trading
        days prior to the closing of the merger.
  - source_sentence: >-
      What caused the increase in net cash provided by operating activities
      between 2022 and 2023?
    sentences:
      - >-
        Net cash provided by operating activities was $712.2 million and $223.7
        million for the year ended December 31, 2023 and 2022, respectively. The
        increase was primarily driven by timing of payments to vendors and
        timing of the receipt of payments from our customers, as well as an
        increase in interest income.
      - >-
        Joanne D. Smith held the position of Vice President - Marketing at Delta
        from November 2005 to February 2007.
      - >-
        Experienced management team with a proven track in the gaming and resort
        industry. Mr. Robert G. Goldstein, our Chairman and Chief Executive
        Officer, has been an integral part of our executive team from the
        beginning, joining our founder and previous Chairman and Chief Executive
        Officer, Mr. Sheldon G. Adelson, before The Venetian Resort Las Vegas
        was constructed. Mr. Goldstein is one of the most respected and
        experienced executives in our industry today.
  - source_sentence: >-
      What does the company believe adds significant value to its business
      regarding intellectual property?
    sentences:
      - >-
        In 2022, the net interest expense on pre-acquisition-related debt was
        $59 million and additional adjustments included costs of $30 million
        associated with the May and June 2022 extinguishment of four series of
        senior notes.
      - >-
        Fluctuations in foreign currency exchange rates decreased our
        consolidated net operating revenues by 4%.
      - >-
        We believe that, to varying degrees, our trademarks, trade names,
        copyrights, proprietary processes, trade secrets, trade dress, domain
        names and similar intellectual property add significant value to our
        business
  - source_sentence: >-
      What does it mean for financial statements to be incorporated by
      reference?
    sentences:
      - >-
        The consolidated financial statements are incorporated by reference in
        the Annual Report on Form 10-K, indicating they are treated as part of
        the document for legal and reporting purposes.
      - >-
        The Consolidated Financial Statements, together with the Notes thereto
        and the report thereon dated February 16, 2024, of
        PricewaterhouseCoopers LLP, the Firm’s independent registered public
        accounting firm (PCAOB ID 238), appear on pages 163–309.
      - >-
        The Goldman Sachs Group, Inc. manages and reports its activities in
        three business segments: Global Banking & Markets, Asset & Wealth
        Samantha Management and Platform Solutions.
pipeline_tag: sentence-similarity
library_name: sentence-transformers
metrics:
  - cosine_accuracy@1
  - cosine_accuracy@3
  - cosine_accuracy@5
  - cosine_accuracy@10
  - cosine_precision@1
  - cosine_precision@3
  - cosine_precision@5
  - cosine_precision@10
  - cosine_recall@1
  - cosine_recall@3
  - cosine_recall@5
  - cosine_recall@10
  - cosine_ndcg@10
  - cosine_mrr@10
  - cosine_map@100
model-index:
  - name: BGE base Financial Matryoshka
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 768
          type: dim_768
        metrics:
          - type: cosine_accuracy@1
            value: 0.7
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.8285714285714286
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.8728571428571429
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9071428571428571
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.7
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.2761904761904762
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.17457142857142854
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09071428571428569
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.7
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.8285714285714286
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.8728571428571429
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9071428571428571
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.8045805359515339
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.7714971655328795
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.775178941729297
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 512
          type: dim_512
        metrics:
          - type: cosine_accuracy@1
            value: 0.7014285714285714
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.83
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.8671428571428571
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.9042857142857142
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.7014285714285714
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.27666666666666667
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.1734285714285714
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09042857142857141
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.7014285714285714
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.83
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.8671428571428571
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.9042857142857142
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.8036464537429646
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.771175736961451
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.7751075563277001
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 256
          type: dim_256
        metrics:
          - type: cosine_accuracy@1
            value: 0.6928571428571428
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.8185714285714286
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.8628571428571429
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.8971428571428571
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.6928571428571428
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.27285714285714285
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.17257142857142854
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.0897142857142857
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.6928571428571428
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.8185714285714286
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.8628571428571429
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.8971428571428571
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.7963364154792727
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.7638741496598634
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.7683107318753077
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 128
          type: dim_128
        metrics:
          - type: cosine_accuracy@1
            value: 0.6771428571428572
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.8142857142857143
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.8514285714285714
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.8885714285714286
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.6771428571428572
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.2714285714285714
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.17028571428571426
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.08885714285714284
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.6771428571428572
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.8142857142857143
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.8514285714285714
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.8885714285714286
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.786332288682679
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.7531507936507934
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.7576033800206036
            name: Cosine Map@100
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: dim 64
          type: dim_64
        metrics:
          - type: cosine_accuracy@1
            value: 0.6571428571428571
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.7814285714285715
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.8171428571428572
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.86
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.6571428571428571
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.2604761904761905
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.16342857142857142
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.08599999999999998
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.6571428571428571
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.7814285714285715
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.8171428571428572
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.86
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.7602042820067257
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.7281371882086165
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.7334805218687248
            name: Cosine Map@100

BGE base Financial Matryoshka

This is a sentence-transformers model finetuned from BAAI/bge-base-en-v1.5 on the json dataset. 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 dimensions
Similarity Function: Cosine Similarity
Training Dataset:
- json
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("Fe2x/bge-base-financial-matryoshka")
# Run inference
sentences = [
    'What does it mean for financial statements to be incorporated by reference?',
    'The consolidated financial statements are incorporated by reference in the Annual Report on Form 10-K, indicating they are treated as part of the document for legal and reporting purposes.',
    'The Consolidated Financial Statements, together with the Notes thereto and the report thereon dated February 16, 2024, of PricewaterhouseCoopers LLP, the Firm’s independent registered public accounting firm (PCAOB ID 238), appear on pages 163–309.',
]
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

Datasets: dim_768, dim_512, dim_256, dim_128 and dim_64
Evaluated with InformationRetrievalEvaluator

Metric	dim_768	dim_512	dim_256	dim_128	dim_64
cosine_accuracy@1	0.7	0.7014	0.6929	0.6771	0.6571
cosine_accuracy@3	0.8286	0.83	0.8186	0.8143	0.7814
cosine_accuracy@5	0.8729	0.8671	0.8629	0.8514	0.8171
cosine_accuracy@10	0.9071	0.9043	0.8971	0.8886	0.86
cosine_precision@1	0.7	0.7014	0.6929	0.6771	0.6571
cosine_precision@3	0.2762	0.2767	0.2729	0.2714	0.2605
cosine_precision@5	0.1746	0.1734	0.1726	0.1703	0.1634
cosine_precision@10	0.0907	0.0904	0.0897	0.0889	0.086
cosine_recall@1	0.7	0.7014	0.6929	0.6771	0.6571
cosine_recall@3	0.8286	0.83	0.8186	0.8143	0.7814
cosine_recall@5	0.8729	0.8671	0.8629	0.8514	0.8171
cosine_recall@10	0.9071	0.9043	0.8971	0.8886	0.86
cosine_ndcg@10	0.8046	0.8036	0.7963	0.7863	0.7602
cosine_mrr@10	0.7715	0.7712	0.7639	0.7532	0.7281
cosine_map@100	0.7752	0.7751	0.7683	0.7576	0.7335

Training Details

Training Dataset

json

Dataset: json
Size: 6,300 training samples
Columns: anchor and positive
Approximate statistics based on the first 1000 samples:
anchor positive
type string string
details
min: 7 tokens
mean: 20.44 tokens
max: 45 tokens

min: 8 tokens
mean: 45.16 tokens
max: 512 tokens

	anchor	positive
type	string	string
details	min: 7 tokens mean: 20.44 tokens max: 45 tokens	min: 8 tokens mean: 45.16 tokens max: 512 tokens

Samples:

anchor	positive
`What was the amount of cash generated from operations by the company in fiscal year 2023?`	`Highlights during fiscal year 2023 include the following: We generated $18,085 million of cash from operations.`
`How much were unrealized losses on U.S. government and agency securities for those held for 12 months or greater as of June 30, 2023?`	`U.S. government and agency securities`
`How is the impairment of assets assessed for projects still under development?`	`For assets under development, assets are grouped and assessed for impairment by estimating the undiscounted cash flows, which include remaining construction costs, over the asset's remaining useful life. If cash flows do not exceed the carrying amount, impairment based on fair value versus carrying value is considered.`

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
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: 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: None
hub_always_push: False
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
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
eval_on_start: False
use_liger_kernel: False
eval_use_gather_object: False
average_tokens_across_devices: False
prompts: None
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	dim_768_cosine_ndcg@10	dim_512_cosine_ndcg@10	dim_256_cosine_ndcg@10	dim_128_cosine_ndcg@10	dim_64_cosine_ndcg@10
0.8122	10	1.5872	-	-	-	-	-
1.0	13	-	0.7879	0.7860	0.7782	0.7698	0.7320
1.5685	20	0.6329	-	-	-	-	-
2.0	26	-	0.7988	0.7969	0.7923	0.7826	0.7520
2.3249	30	0.4465	-	-	-	-	-
3.0	39	-	0.8046	0.8026	0.7959	0.7855	0.7596
3.0812	40	0.349	-	-	-	-	-
3.731	48	-	0.8046	0.8036	0.7963	0.7863	0.7602

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.9.20
Sentence Transformers: 3.3.1
Transformers: 4.47.1
PyTorch: 2.1.2+cu121
Accelerate: 1.2.1
Datasets: 2.19.1
Tokenizers: 0.21.0

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}
}