metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:178
  - loss:MultipleNegativesRankingLoss
base_model: BAAI/bge-reranker-base
widget:
  - source_sentence: >-
      What is the basis for the company's calculation of deferred tax assets and
      liabilities?
    sentences:
      - ' We could be subject to various penalties or restrictions on our ability to conduct our business, which could have a material and adverse impact on our business, operating results, and financial condition.'
      - ' The company''s calculation of deferred tax assets and liabilities is based on certain estimates and judgments and involves dealing with uncertainties in the application of complex tax laws.'
      - ' March 4, 2024.'
  - source_sentence: >-
      What are some situations that may result in excess or obsolete inventory
      or excess product purchase commitments?
    sentences:
      - ' 20%'
      - ' NVIDIA''s common stock is traded on the Nasdaq Global Select Market under the symbol NVDA.'
      - ' Changes in business and economic conditions, changes in market conditions, sudden and significant decreases in demand for products, inventory obsolescence due to changing technology and customer requirements, new product introductions, failure to estimate customer demand properly, ordering in advance of historical lead-times, government regulations, and changes in future demand or increase in demand for competitive products.'
  - source_sentence: >-
      What are the primary methods used by the company to protect its
      intellectual property?
    sentences:
      - ' The change resulted in an increase in operating income of $135 million and net income of $114 million after tax, or $0.05 per both basic and diluted share.'
      - ' Forfeitures are estimated based on historical experience and revised semi-annually if actual forfeitures differ from those estimates.'
      - ' The company relies primarily on a combination of patents, trademarks, trade secrets, employee and third-party nondisclosure agreements, and licensing arrangements to protect its intellectual property in the United States and internationally.'
  - source_sentence: >-
      What are the potential consequences of an unfavorable outcome in the
      litigation and regulatory proceedings mentioned in the text?
    sentences:
      - ' The new licensing requirements apply to exports of certain products, including A100, A800, H100, H800, L4, L40, L40S, and RTX 4090, exceeding certain performance thresholds to China, Country Groups D1, D4, and D5, and to parties headquartered in or with an ultimate parent headquartered in Country Group D5, including China.'
      - ' Adverse rulings could occur, including monetary damages or fines, an injunction stopping the company from manufacturing or selling certain products, engaging in certain business practices, or requiring other remedies, such as compulsory licensing of patents.'
      - ' The main components of the NVIDIA accelerated computing platform include GPUs, DPUs, interconnects, and fully optimized AI and high-performance computing software stacks.'
  - source_sentence: >-
      What are the potential risks associated with the company's acquisitions
      and strategic investments?
    sentences:
      - ' $7,280.'
      - ' The company could face significant consequences, including government enforcement actions, litigation, additional reporting requirements and/or oversight, bans on processing personal data, and orders to destroy or not use personal data, which could have a material adverse effect on its reputation, business, or financial condition.'
      - ' The potential risks include impairment of the company''s ability to grow its business, develop new products, or sell its products, as well as the possibility of regulatory conditions reducing the value of the acquisition, volatility in results, losses up to the value of the investment, and impairment losses due to the failure of the invested companies.'
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
  - dot_accuracy@1
  - dot_accuracy@3
  - dot_accuracy@5
  - dot_accuracy@10
  - dot_precision@1
  - dot_precision@3
  - dot_precision@5
  - dot_precision@10
  - dot_recall@1
  - dot_recall@3
  - dot_recall@5
  - dot_recall@10
  - dot_ndcg@10
  - dot_mrr@10
  - dot_map@100
model-index:
  - name: SentenceTransformer based on BAAI/bge-reranker-base
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: bge base en
          type: bge-base-en
        metrics:
          - type: cosine_accuracy@1
            value: 0.0056179775280898875
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.05056179775280899
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 0.0898876404494382
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 0.16853932584269662
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.0056179775280898875
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.016853932584269662
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.017977528089887642
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.016853932584269662
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.0056179775280898875
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.05056179775280899
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 0.0898876404494382
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 0.16853932584269662
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.07282854323415827
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.044208132691278754
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.07494028485986497
            name: Cosine Map@100
          - type: dot_accuracy@1
            value: 0.016853932584269662
            name: Dot Accuracy@1
          - type: dot_accuracy@3
            value: 0.033707865168539325
            name: Dot Accuracy@3
          - type: dot_accuracy@5
            value: 0.08426966292134831
            name: Dot Accuracy@5
          - type: dot_accuracy@10
            value: 0.1853932584269663
            name: Dot Accuracy@10
          - type: dot_precision@1
            value: 0.016853932584269662
            name: Dot Precision@1
          - type: dot_precision@3
            value: 0.011235955056179775
            name: Dot Precision@3
          - type: dot_precision@5
            value: 0.016853932584269662
            name: Dot Precision@5
          - type: dot_precision@10
            value: 0.018539325842696634
            name: Dot Precision@10
          - type: dot_recall@1
            value: 0.016853932584269662
            name: Dot Recall@1
          - type: dot_recall@3
            value: 0.033707865168539325
            name: Dot Recall@3
          - type: dot_recall@5
            value: 0.08426966292134831
            name: Dot Recall@5
          - type: dot_recall@10
            value: 0.1853932584269663
            name: Dot Recall@10
          - type: dot_ndcg@10
            value: 0.07894273048552719
            name: Dot Ndcg@10
          - type: dot_mrr@10
            value: 0.04782637774210808
            name: Dot Mrr@10
          - type: dot_map@100
            value: 0.07569131465859923
            name: Dot Map@100

SentenceTransformer based on BAAI/bge-reranker-base

This is a sentence-transformers model finetuned from BAAI/bge-reranker-base on the train 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-reranker-base
Maximum Sequence Length: 512 tokens
Output Dimensionality: 768 tokens
Similarity Function: Cosine Similarity
Training Dataset:
- train

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': False}) with Transformer model: XLMRobertaModel 
  (1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)

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("rezarahim/bge-finetuned-reranker")
# Run inference
sentences = [
    "What are the potential risks associated with the company's acquisitions and strategic investments?",
    " The potential risks include impairment of the company's ability to grow its business, develop new products, or sell its products, as well as the possibility of regulatory conditions reducing the value of the acquisition, volatility in results, losses up to the value of the investment, and impairment losses due to the failure of the invested companies.",
    ' $7,280.',
]
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: bge-base-en
Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.0056
cosine_accuracy@3	0.0506
cosine_accuracy@5	0.0899
cosine_accuracy@10	0.1685
cosine_precision@1	0.0056
cosine_precision@3	0.0169
cosine_precision@5	0.018
cosine_precision@10	0.0169
cosine_recall@1	0.0056
cosine_recall@3	0.0506
cosine_recall@5	0.0899
cosine_recall@10	0.1685
cosine_ndcg@10	0.0728
cosine_mrr@10	0.0442
cosine_map@100	0.0749
dot_accuracy@1	0.0169
dot_accuracy@3	0.0337
dot_accuracy@5	0.0843
dot_accuracy@10	0.1854
dot_precision@1	0.0169
dot_precision@3	0.0112
dot_precision@5	0.0169
dot_precision@10	0.0185
dot_recall@1	0.0169
dot_recall@3	0.0337
dot_recall@5	0.0843
dot_recall@10	0.1854
dot_ndcg@10	0.0789
dot_mrr@10	0.0478
dot_map@100	0.0757

Training Details

Training Dataset

train

Dataset: train
Size: 178 training samples
Columns: anchor and positive
Approximate statistics based on the first 178 samples:
anchor positive
type string string
details
min: 12 tokens
mean: 23.55 tokens
max: 50 tokens

min: 3 tokens
mean: 42.22 tokens
max: 135 tokens

	anchor	positive
type	string	string
details	min: 12 tokens mean: 23.55 tokens max: 50 tokens	min: 3 tokens mean: 42.22 tokens max: 135 tokens

Samples:

anchor	positive
`What is the publication date of the NVIDIA Corporation Annual Report 2024?`	`February 21st, 2024`
`What is the filing date of the 10-K report for NVIDIA Corporation in 2004?`	`The filing dates of the 10-K reports for NVIDIA Corporation in 2004 are May 20th, March 29th, and April 25th.`
`What is the purpose of the section of the filing that requires the registrant to indicate whether it has submitted electronically every Interactive Data File required to be submitted during the preceding 12 months?`	`The purpose of this section is to comply with Rule 405 of Regulation S-T, which requires the registrant to submit electronic files for certain financial information.`

Loss: MultipleNegativesRankingLoss with these parameters:

{
    "scale": 20.0,
    "similarity_fct": "cos_sim"
}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: epoch
per_device_train_batch_size: 4
per_device_eval_batch_size: 16
gradient_accumulation_steps: 16
learning_rate: 2e-05
num_train_epochs: 25
lr_scheduler_type: cosine
warmup_ratio: 0.1
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: 4
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: 25
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: False
fp16_opt_level: O1
half_precision_backend: auto
bf16_full_eval: False
fp16_full_eval: False
tf32: None
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
eval_on_start: False
use_liger_kernel: False
eval_use_gather_object: False
batch_sampler: no_duplicates
multi_dataset_batch_sampler: proportional

Training Logs

Epoch	Step	Training Loss	bge-base-en_dot_map@100
0	0	-	0.0362
0.7111	2	-	0.0369
1.7778	5	-	0.0539
2.8444	8	-	0.0393
3.5556	10	2.0824	-
3.9111	11	-	0.0559
4.9778	14	-	0.0632
5.6889	16	-	0.08
6.7556	19	-	0.0692
7.1111	20	1.2812	-
7.8222	22	-	0.0627
8.8889	25	-	0.0623
9.9556	28	-	0.0692
10.6667	30	1.0855	0.0884
11.7333	33	-	0.0754
12.8	36	-	0.0607
13.8667	39	-	0.0725
14.2222	40	0.8978	-
14.9333	42	-	0.0747
16.0	45	-	0.0766
16.7111	47	-	0.0756
17.7778	50	0.8563	0.0757

The bold row denotes the saved checkpoint.

Framework Versions

Python: 3.10.12
Sentence Transformers: 3.1.1
Transformers: 4.45.2
PyTorch: 2.5.1+cu121
Accelerate: 1.2.1
Datasets: 3.1.0
Tokenizers: 0.20.3

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",
}

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}
}