Add new SentenceTransformer model.

0d00981 verified 8 months ago

20.3 kB

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - dense
  - generated_from_trainer
  - dataset_size:1000
  - loss:TripletLoss
base_model: sentence-transformers/all-MiniLM-L6-v2
widget:
  - source_sentence: What are the key points in passage fam_402.10_30?
    sentences:
      - >-
        ( v ) A dependent applying under [ paragraph ( s)(2 ) ( iii)](/current /
        title-8 / section-214.2#p-214.2(s)(2)(iii ) ) or [ ( iv)](/current /
        title-8 / section-214.2#p-214.2(s)(2)(iv ) ) of this section must also
        submit a certified statement from the post - secondary educational
        institution confirming that he or she is pursuing studies on a full -
        time basis .
      - |-
        ( b )   ( U ) The criteria for 
         qualifying as an H-1B physician are found in subparagraph 3 below .
      - ( ii ) * What are the requirements for participation ? *
  - source_sentence: What are the key points in passage 8cfr_214.3_93?
    sentences:
      - >-
        ( vii ) Whether the student has been certified for practical training ,
        and the beginning and end dates of certification .
      - ( D ) Similarity of jobs and working conditions ;
      - ( ii ) * What are the requirements for participation ? *
  - source_sentence: Explain the significance of passage fam_402_62.
    sentences:
      - >-
        ( * i * ) Has competency in oral and written English which shall be
        demonstrated by the passage of the English language proficiency test
        given by the Educational Commission for Foreign Medical Graduates ; or
      - |-
        Derivative beneficiaries are entitled to apply for visas to 
         follow and/or join principals who are maintaining status in the United States , 
         even when the principal was never issued a visa in the classification being 
         sought by the dependent .   Take , for instance , a world - class soccer player , who 
         changes their status from F-1 to O-1 .   The spouse and/or children are entitled 
         to apply for nonimmigrant O-3 visas .   Typical documentation for establishing 
         entitlement to visas in such an instance might include marriage and birth 
         certificates for the spouse and dependent(s ) , a copy of the principal 
         beneficiary 's approval notice , and any Form I-797 , Notice of Action notices 
         relating to the dependents ' own change of status filings .   Another example 
         would be a foreign national who entered the United States on a B-1 visa and 
         subsequently changed status to F-1 .   The spouse and/or child of the F-1 would 
         be entitled to seek F-2 visas .   In such cases , the dependent would need to 
         present a properly endorsed Form I-20 , Certificate of Eligibility for 
         Nonimmigrant ( F-1 ) Student Status - for Academic and Language Students , as 
         evidence that the principal is enrolled , or will be enrolled within 60 days , in 
         a full course of study or is in approved practical training .
      - >-
        ( 1 ) Meaning of term * Designated Official . * As used in [ § §
        214.2(f)](/current / title-8 / section-214.2#p-214.2(f ) ) and [ (
        m)](/current / title-8 / section-214.2#p-214.2(m ) ) , [ 214.3](/current
        / title-8 / section-214.3 ) and [ 214.4](/current / title-8 /
        section-214.4 ) , a * Designated Official , Designated School Official (
        DSO ) , * or * Principal Designated School Official ( PDSO ) , * means a
        regularly employed member of the school administration whose office is
        located at the school and whose compensation does not come from
        commissions for recruitment of foreign students . An individual whose
        principal obligation to the school is to recruit foreign students for
        compensation does not qualify as a designated official . The PDSO and
        any other DSO must be named by the president , owner , or head of a
        school or school system . The PDSO and DSO may not delegate this
        designation to any other person .
  - source_sentence: What is the main topic of passage fam_402.9_141?
    sentences:
      - |-
        ( 1 )   The title of the position to which the applicant 
         is destined , its place in the firm 's organizational structure , the duties 
         of the position , the degree to which the applicant will have ultimate control 
         and responsibility for the firm 's overall operations or a major component 
         thereof , the number and skill levels of the employees the applicant will 
         supervise , the level of pay , and whether the applicant possesses qualifying 
         executive or supervisory experience ;
      - >-
        describes methods of oversight and supervision . The Form I-983 must
        explain how the training is directly related to the student 's
        qualifying STEM degree .
      - >-
        ( A ) A nurse who is granted H-1C classification based on passage of the
        CGFNS examination must , upon admission to the United States , be able
        to obtain temporary licensure or other temporary authorization to
        practice as a registered nurse from the State Board of Nursing in the
        state of intended employment .
  - source_sentence: Explain the significance of passage uscis_pm_volume_2_part_f_chapter_7_1.
    sentences:
      - ( C ) A common formal code of doctrine and discipline ;
      - >-
        ( * i * ) Has competency in oral and written English which shall be
        demonstrated by the passage of the English language proficiency test
        given by the Educational Commission for Foreign Medical Graduates ; or
      - Chapter 7 - Absences From the United States | USCIS
pipeline_tag: sentence-similarity
library_name: sentence-transformers

SentenceTransformer based on sentence-transformers/all-MiniLM-L6-v2

This is a sentence-transformers model finetuned from sentence-transformers/all-MiniLM-L6-v2. It maps sentences & paragraphs to a 384-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: sentence-transformers/all-MiniLM-L6-v2
Maximum Sequence Length: 256 tokens
Output Dimensionality: 384 dimensions
Similarity Function: Cosine Similarity

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': 256, 'do_lower_case': False, 'architecture': 'BertModel'})
  (1): Pooling({'word_embedding_dimension': 384, '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})
  (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("vayishu/visa-minilm")
# Run inference
sentences = [
    'Explain the significance of passage uscis_pm_volume_2_part_f_chapter_7_1.',
    'Chapter 7 - Absences From the United States | USCIS',
    '( * i * ) Has competency in oral and written English which shall be demonstrated by the passage of the English language proficiency test given by the Educational Commission for Foreign Medical Graduates ; or',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 384]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities)
# tensor([[1.0000, 0.3954, 0.4014],
#         [0.3954, 1.0000, 0.1409],
#         [0.4014, 0.1409, 1.0000]])

Training Details

Training Dataset

Unnamed Dataset

Size: 1,000 training samples
Columns: sentence_0, sentence_1, and sentence_2

Approximate statistics based on the first 1000 samples:

	sentence_0	sentence_1	sentence_2
type	string	string	string
details	min: 14 tokens mean: 20.62 tokens max: 30 tokens	min: 5 tokens mean: 74.9 tokens max: 256 tokens	min: 14 tokens mean: 48.16 tokens max: 143 tokens

Samples:

sentence_0	sentence_1	sentence_2
`Explain the significance of passage 8cfr_214.1_85.`	`# # # # § 214.1 Requirements for admission , extension , and maintenance of status .`	`( * 5 * ) Evidence of the alien 's original scientific , scholarly , or business - related contributions of major significance in the field ;`
`Can you summarize the content of passage 8cfr_214.2_1843?`	`( C ) A common formal code of doctrine and discipline ;`	`The Office of the Federal Register publishes documents on behalf of Federal agencies but does not have any authority over their programs . We recommend you directly contact the agency associated with the content in question .`
`What is the main topic of passage uscis_pm_volume_2_part_f_chapter_5_85?`	`If the Form I-765 for the STEM OPT extension is denied and the student 's post - completion OPT EAD is expired , OPT employment authorization ends on the date of the decision and the student 's F-1 status ends 60 days after the date of denial . If the Form I-765 for the STEM OPT extension is denied and the student 's post - completion OPT EAD is unexpired , the student will remain employment authorized until the expiration date of the EAD .`	`( A ) A nurse who is granted H-1C classification based on passage of the CGFNS examination must , upon admission to the United States , be able to obtain temporary licensure or other temporary authorization to practice as a registered nurse from the State Board of Nursing in the state of intended employment .`

Loss: TripletLoss with these parameters:

{
    "distance_metric": "TripletDistanceMetric.EUCLIDEAN",
    "triplet_margin": 5
}

Training Hyperparameters

Non-Default Hyperparameters

per_device_train_batch_size: 32
per_device_eval_batch_size: 32
num_train_epochs: 1
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: no
prediction_loss_only: True
per_device_train_batch_size: 32
per_device_eval_batch_size: 32
per_gpu_train_batch_size: None
per_gpu_eval_batch_size: None
gradient_accumulation_steps: 1
eval_accumulation_steps: None
torch_empty_cache_steps: None
learning_rate: 5e-05
weight_decay: 0.0
adam_beta1: 0.9
adam_beta2: 0.999
adam_epsilon: 1e-08
max_grad_norm: 1
num_train_epochs: 1
max_steps: -1
lr_scheduler_type: linear
lr_scheduler_kwargs: {}
warmup_ratio: 0.0
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: False
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
hub_revision: None
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
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
liger_kernel_config: None
eval_use_gather_object: False
average_tokens_across_devices: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin
router_mapping: {}
learning_rate_mapping: {}

Framework Versions

Python: 3.12.11
Sentence Transformers: 5.1.0
Transformers: 4.55.2
PyTorch: 2.8.0+cu128
Accelerate: 1.10.0
Datasets: 4.0.0
Tokenizers: 0.21.4

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",
}

TripletLoss

@misc{hermans2017defense,
    title={In Defense of the Triplet Loss for Person Re-Identification},
    author={Alexander Hermans and Lucas Beyer and Bastian Leibe},
    year={2017},
    eprint={1703.07737},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}