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-oursdatasets")
# Run inference
sentences = [
    '智能监管，实时报警提示不规范行为',
    '共享单车智能监管的主要功能是什么？',
    '如何进行手机号实名认证流程？',
]
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	0.1183
cosine_accuracy@3	0.2052
cosine_accuracy@5	0.2465
cosine_accuracy@10	0.3018
cosine_precision@1	0.1183
cosine_precision@3	0.0684
cosine_precision@5	0.0493
cosine_precision@10	0.0302
cosine_recall@1	0.1183
cosine_recall@3	0.2052
cosine_recall@5	0.2465
cosine_recall@10	0.3018
cosine_ndcg@10	0.2039
cosine_mrr@10	0.1734
cosine_map@100	0.1814

Information Retrieval

Dataset: dim_512
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1151
cosine_accuracy@3	0.1982
cosine_accuracy@5	0.2394
cosine_accuracy@10	0.3018
cosine_precision@1	0.1151
cosine_precision@3	0.0661
cosine_precision@5	0.0479
cosine_precision@10	0.0302
cosine_recall@1	0.1151
cosine_recall@3	0.1982
cosine_recall@5	0.2394
cosine_recall@10	0.3018
cosine_ndcg@10	0.2019
cosine_mrr@10	0.1708
cosine_map@100	0.1788

Information Retrieval

Dataset: dim_256
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1107
cosine_accuracy@3	0.1906
cosine_accuracy@5	0.2242
cosine_accuracy@10	0.2866
cosine_precision@1	0.1107
cosine_precision@3	0.0635
cosine_precision@5	0.0448
cosine_precision@10	0.0287
cosine_recall@1	0.1107
cosine_recall@3	0.1906
cosine_recall@5	0.2242
cosine_recall@10	0.2866
cosine_ndcg@10	0.1918
cosine_mrr@10	0.1624
cosine_map@100	0.1702

Information Retrieval

Dataset: dim_128
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.1031
cosine_accuracy@3	0.171
cosine_accuracy@5	0.2074
cosine_accuracy@10	0.2725
cosine_precision@1	0.1031
cosine_precision@3	0.057
cosine_precision@5	0.0415
cosine_precision@10	0.0273
cosine_recall@1	0.1031
cosine_recall@3	0.171
cosine_recall@5	0.2074
cosine_recall@10	0.2725
cosine_ndcg@10	0.1791
cosine_mrr@10	0.1504
cosine_map@100	0.158

Information Retrieval

Dataset: dim_64
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.0874
cosine_accuracy@3	0.1471
cosine_accuracy@5	0.1846
cosine_accuracy@10	0.2405
cosine_precision@1	0.0874
cosine_precision@3	0.049
cosine_precision@5	0.0369
cosine_precision@10	0.024
cosine_recall@1	0.0874
cosine_recall@3	0.1471
cosine_recall@5	0.1846
cosine_recall@10	0.2405
cosine_ndcg@10	0.1562
cosine_mrr@10	0.1302
cosine_map@100	0.1378

Training Details

Training Dataset

json

Dataset: json
Size: 16,572 training samples
Columns: positive and anchor
Approximate statistics based on the first 1000 samples:
positive anchor
type string string
details
min: 4 tokens
mean: 112.19 tokens
max: 512 tokens
min: 10 tokens
mean: 22.76 tokens
max: 52 tokens

Samples:

positive	anchor
`区块链去中心化及不可篡改等特性保证数据及交易的可靠性。企业数据共享交易在区块链上完成，由多个企业节点共识一致后才能确认交易并写入区块。`	`区块链技术如何保障VDBP平台的数据可靠性？`
`文本说明，通过AI行业垂直模型的深度学习和预测，为外贸企业发掘海外目标市场的需求匹配度和转化关键点。`	`垂直行业AI模型如何帮助外贸企业？`
`智慧管网系统由国塑科技与材料研究院基于"贵州智慧管网技术创新中心"项目研发。`	`根据文本，智慧管网系统的研发背景是什么？`

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
0.3089	10	10.581	-	-	-	-	-
0.6178	20	8.0326	-	-	-	-	-
0.9266	30	7.5337	-	-	-	-	-
0.9884	32	-	0.1739	0.1751	0.1630	0.1560	0.1411
1.2355	40	6.8775	-	-	-	-	-
1.5444	50	6.7937	-	-	-	-	-
1.8533	60	6.3144	-	-	-	-	-
1.9768	64	-	0.1762	0.1753	0.1648	0.1547	0.1377
2.1622	70	6.0413	-	-	-	-	-
2.4710	80	6.2465	-	-	-	-	-
2.7799	90	6.0414	-	-	-	-	-
2.9961	97	-	0.1805	0.1788	0.1709	0.1598	0.1397
3.0888	100	5.7624	-	-	-	-	-
3.3977	110	5.9827	-	-	-	-	-
3.7066	120	5.8302	-	-	-	-	-
3.9537	128	-	0.1814	0.1788	0.1702	0.1580	0.1378

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: 4 tokens mean: 112.19 tokens max: 512 tokens	min: 10 tokens mean: 22.76 tokens max: 52 tokens