You can deploy this model and start earning money today!

ChristianBernhard/
bge-base-financial-matryoshka-2

Name: bge-base-financial-matryoshka-2
Availability: InStock

Views

No views yet

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.

python

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("ChristianBernhard/bge-base-financial-matryoshka-2")
# Run inference
sentences = [
    'What was the total long-term debt, including the current portion, for AbbVie as of December 31, 2023?',
    "The following table summarizes AbbVie's estimated material contractual obligations as of December 31, 2023, which includes long-term debt, both current and long-term portions, totaling $59,245 million.",
    'In the first quarter of 2023, we began production shipments of Enphase branded EV chargers at our existing contract manufacturing facility in Mexico.',
]
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	0.7329	0.73	0.7314	0.7157	0.6843
cosine_accuracy@3	0.8429	0.85	0.8386	0.8286	0.7971
cosine_accuracy@5	0.8857	0.8886	0.8771	0.8657	0.8514
cosine_accuracy@10	0.9171	0.9186	0.92	0.9143	0.8957
cosine_precision@1	0.7329	0.73	0.7314	0.7157	0.6843
cosine_precision@3	0.281	0.2833	0.2795	0.2762	0.2657
cosine_precision@5	0.1771	0.1777	0.1754	0.1731	0.1703
cosine_precision@10	0.0917	0.0919	0.092	0.0914	0.0896
cosine_recall@1	0.7329	0.73	0.7314	0.7157	0.6843
cosine_recall@3	0.8429	0.85	0.8386	0.8286	0.7971
cosine_recall@5	0.8857	0.8886	0.8771	0.8657	0.8514
cosine_recall@10	0.9171	0.9186	0.92	0.9143	0.8957
cosine_ndcg@10	0.8263	0.8269	0.8245	0.8143	0.7898
cosine_mrr@10	0.797	0.7971	0.794	0.7825	0.7559
cosine_map@100	0.8009	0.8009	0.7977	0.786	0.7602

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: 2 tokens
mean: 20.58 tokens
max: 42 tokens
min: 7 tokens
mean: 46.12 tokens
max: 512 tokens

Samples:

anchor	positive
`How many new stores did the company open in fiscal 2023?`	`We had net new store openings of 197 for fiscal 2023.`
`What can users do with Windows 365?`	`From another internet-connected device, including Android or macOS devices, users can run Windows 365, just like a virtual machine.`
`What is the effective date for the fifteen percent minimum tax based on adjusted financial statement income as specified in the Inflation Reduction Act of 2022?`	`On August 16, 2022, the U.S. government enacted the Inflation Reduction Act of 2022 that includes a fifteen percent minimum tax based on adjusted financial statement income, which is effective for NIKE beginning June 1, 2023.`

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: 10
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: 10
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
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	1.7456	-	-	-	-	-
0.9746	12	-	0.7938	0.7946	0.7889	0.7703	0.7223
1.6244	20	0.7675	-	-	-	-	-
1.9492	24	-	0.8159	0.8171	0.8085	0.7989	0.7705
2.4365	30	0.4625	-	-	-	-	-
2.9239	36	-	0.8221	0.8223	0.8196	0.8059	0.7776
3.2487	40	0.3201	-	-	-	-	-
3.9797	49	-	0.8232	0.8257	0.8221	0.8135	0.7865
4.0609	50	0.2489	-	-	-	-	-
4.8731	60	0.196	-	-	-	-	-
4.9543	61	-	0.8253	0.8255	0.8219	0.816	0.7856
5.6853	70	0.1628	-	-	-	-	-
5.9289	73	-	0.8259	0.8252	0.8219	0.8155	0.7864
6.4975	80	0.1605	-	-	-	-	-
6.9848	86	-	0.8264	0.8250	0.8240	0.8139	0.7887
7.3096	90	0.1357	-	-	-	-	-
7.9594	98	-	0.8266	0.8256	0.8240	0.8147	0.7904
8.1218	100	0.1374	-	-	-	-	-
8.9340	110	0.1272	0.8267	0.8254	0.8249	0.8143	0.7882
9.7462	120	0.1253	0.8263	0.8269	0.8245	0.8143	0.7898

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.3.1
Transformers: 4.41.2
PyTorch: 2.1.2+cu121
Accelerate: 1.2.0
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}
}

	anchor	positive
type	string	string
details	min: 2 tokens mean: 20.58 tokens max: 42 tokens	min: 7 tokens mean: 46.12 tokens max: 512 tokens