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 tokens
Similarity Function: Cosine Similarity
Training Dataset:
- json

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.

python

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("ybWw/bge-base-financial-matryoshka")
# Run inference
sentences = [
    'A 1% increase in medical cost PMPM trend factors led to an increase in medical costs payable by $1,128 million for the most recent two months as of December 31, 2023.',
    'What financial impact would a 1% increase in medical cost PMPM trend factors have on medical costs payable for the most recent two months as of December 31, 2023?',
    'How does Kroger present its financial performance in its reporting?',
]
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	0.6843
cosine_accuracy@3	0.8186
cosine_accuracy@5	0.8571
cosine_accuracy@10	0.9
cosine_precision@1	0.6843
cosine_precision@3	0.2729
cosine_precision@5	0.1714
cosine_precision@10	0.09
cosine_recall@1	0.6843
cosine_recall@3	0.8186
cosine_recall@5	0.8571
cosine_recall@10	0.9
cosine_ndcg@10	0.7945
cosine_mrr@10	0.7606
cosine_map@100	0.7643

Information Retrieval

Dataset: dim_512
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6771
cosine_accuracy@3	0.8129
cosine_accuracy@5	0.8571
cosine_accuracy@10	0.9
cosine_precision@1	0.6771
cosine_precision@3	0.271
cosine_precision@5	0.1714
cosine_precision@10	0.09
cosine_recall@1	0.6771
cosine_recall@3	0.8129
cosine_recall@5	0.8571
cosine_recall@10	0.9
cosine_ndcg@10	0.7898
cosine_mrr@10	0.7544
cosine_map@100	0.7579

Information Retrieval

Dataset: dim_256
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6829
cosine_accuracy@3	0.8129
cosine_accuracy@5	0.8557
cosine_accuracy@10	0.89
cosine_precision@1	0.6829
cosine_precision@3	0.271
cosine_precision@5	0.1711
cosine_precision@10	0.089
cosine_recall@1	0.6829
cosine_recall@3	0.8129
cosine_recall@5	0.8557
cosine_recall@10	0.89
cosine_ndcg@10	0.789
cosine_mrr@10	0.7562
cosine_map@100	0.7601

Information Retrieval

Dataset: dim_128
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6657
cosine_accuracy@3	0.8
cosine_accuracy@5	0.8314
cosine_accuracy@10	0.88
cosine_precision@1	0.6657
cosine_precision@3	0.2667
cosine_precision@5	0.1663
cosine_precision@10	0.088
cosine_recall@1	0.6657
cosine_recall@3	0.8
cosine_recall@5	0.8314
cosine_recall@10	0.88
cosine_ndcg@10	0.7745
cosine_mrr@10	0.7406
cosine_map@100	0.7444

Information Retrieval

Dataset: dim_64
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.6357
cosine_accuracy@3	0.7686
cosine_accuracy@5	0.8086
cosine_accuracy@10	0.8514
cosine_precision@1	0.6357
cosine_precision@3	0.2562
cosine_precision@5	0.1617
cosine_precision@10	0.0851
cosine_recall@1	0.6357
cosine_recall@3	0.7686
cosine_recall@5	0.8086
cosine_recall@10	0.8514
cosine_ndcg@10	0.7447
cosine_mrr@10	0.7103
cosine_map@100	0.7149

Training Details

Training Dataset

json

Dataset: json
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: 46.17 tokens
max: 272 tokens
min: 8 tokens
mean: 20.72 tokens
max: 42 tokens

Samples:

positive	anchor
`For fiscal 2023, the net cash provided by operating activities was $7,111 million.`	`What was the net cash provided by operating activities in fiscal 2023?`
`Penalties for impermissible use or disclosure of PHI were increased by the HITECH Act by imposing tiered penalties of more than $50,000 per violation and up to $1.5 million per year for identical violations.`	`What are the consequences of impermissible use or disclosure of PHI according to the HITECH Act?`
`Global Day of Joy is Hasbro’s annual, company-wide day of service and has become a cherished tradition. Global Day of Joy takes place every December, and employees from each Hasbro office participate in service projects to benefit a variety of organizations.`	`What is the purpose of the Global Day of Joy at Hasbro?`

Loss: MatryoshkaLoss with these parameters:
json
1{ 2 "loss": "MultipleNegativesRankingLoss", 3 "matryoshka_dims": [ 4 768, 5 512, 6 256, 7 128, 8 64 9 ], 10 "matryoshka_weights": [ 11 1, 12 1, 13 1, 14 1, 15 1 16 ], 17 "n_dims_per_step": -1 18}

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
bf16: True
tf32: True
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: True
fp16: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: True
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_768_cosine_map@100	dim_512_cosine_map@100	dim_256_cosine_map@100	dim_128_cosine_map@100	dim_64_cosine_map@100
0.8122	10	1.5383	-	-	-	-	-
0.9746	12	-	0.7575	0.7597	0.7428	0.7303	0.6889
1.6244	20	0.6278	-	-	-	-	-
1.9492	24	-	0.7623	0.7580	0.7582	0.7417	0.7100
2.4365	30	0.4388	-	-	-	-	-
2.9239	36	-	0.7649	0.7576	0.7571	0.7465	0.7142
3.2487	40	0.3729	-	-	-	-	-
3.8985	48	-	0.7643	0.7579	0.7601	0.7444	0.7149

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.8.20
Sentence Transformers: 3.2.1
Transformers: 4.41.2
PyTorch: 2.1.2+cu121
Accelerate: 1.0.1
Datasets: 2.19.1
Tokenizers: 0.19.1

Citation

BibTeX

Sentence Transformers

bibtex

@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

bibtex

@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

bibtex

@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}
}

	positive	anchor
type	string	string
details	min: 6 tokens mean: 46.17 tokens max: 272 tokens	min: 8 tokens mean: 20.72 tokens max: 42 tokens