metadata

tags:
  - sentence-transformers
  - sentence-similarity
  - feature-extraction
  - generated_from_trainer
  - dataset_size:334
  - loss:MatryoshkaLoss
  - loss:MultipleNegativesRankingLoss
base_model: Snowflake/snowflake-arctic-embed-l
widget:
  - source_sentence: >-
      QUESTION #1: What are the potential adverse effects associated with the
      use of peramivir?
    sentences:
      - |-
        although poultry-to-human and human-to-human trans -
        mission remains relatively low. Despite low transmissibility, 
        the reported fatality rate is high (approximately 60%).14
        Prevention
        The Centers for Disease Control and Prevention’s (CDC’s) 
        Advisory Committee on Immunization Practices (ACIP) 
        and the American Academy of Family Physicians (AAFP) 
        recommend annual influenza vaccination for all people six 
        months and older who do not have contraindications. 15,16 
        Vaccination efforts should target people at increased risk of 
        complicated or severe influenza (Table 117-19) and those who 
        care for or live with high-risk individuals, including health 
        care professionals. 15 Two previous FPM articles provided
      - >-
        increased sensitivity to pain.60 These cytokines are also

        associated with URTIs and may mediate mood changes

        associated with these infections.

        Anorexia

        Anorexia is a common behavioural response to URTIs,

        and this response has entered the folklore as advice to

        Figure 4: Fever is caused by cytokines released from macrophages and
        other

        immune cells

        The cytokines may act on vagal nerve endings or enter the brain to cause
        a

        resetting of the temperature control centre in the hypothalamus. The

        hypothalamus causes shivering and constriction of skin blood vessels and
        also

        initiates a sensation of chilliness that is perceived at the level of
        the cerebral

        cortex. IL=interleukin; TNF=tumour necrosis factor.

        Vagal

        nerves

        ShiveringMacrophages
      - |-
        older who have been 
        symptomatic for no 
        more than 48 hours
        Contraindicated in people 
        with serious hypersensitivity or 
        anaphylaxis to peramivir or any 
        component of the product
        Potential adverse effects include 
        diarrhea, nausea, vomiting, and 
        neutropenia
        Weigh risks and benefits during 
        pregnancy; no human data 
        available; no known risk of 
        embryo-fetal toxicity based on 
        animal data at 8 times the recom -
        mended human dose; possible 
        risk of embryo-fetal toxicity with 
        continuous intravenous infusion 
        based on limited animal data
        Baloxavir (Xofluza), 
        available as oral 
        tablets
        NA ($160) Adults and children 12 years 
        and older:  
        88 to 174 lb (40 to 79 kg):  
        single dose of 40 mg  
        ≥ 175 lb (80 kg):  single dose 
        of 80 mg
  - source_sentence: Why is Influenza A most responsible for causing pandemics?
    sentences:
      - |-
        on the first day of symptoms, medications containing ibu -
        profen and pseudoephedrine may reduce the severity of cold 
        symptoms.35 Antihistamine monotherapy is not effective 
        for relieving cough.6,23
        Ipratropium. Inhaled ipratropium is the only medication 
        that improves persistent cough related to URI in adults. 24,36  
        TABLE 1
        Differential Diagnosis for the Common Cold
        Diagnosis
        Symptom 
        onset Cough Sore throat Fever Rhinorrhea Aches Watery eyes Sneezing
        Nasal  
        congestion Headache
        Shortness  
        of breath
        Acute 
        bronchitis
        Gradual Prominent, per-
        sistent, dry or wet
        Common None or low 
        grade
        Uncommon Mild Common Uncommon Uncommon Common, mild Common
        Allergic 
        rhinitis
        Gradual Common, chronic Possible, especially 
        on awakening
        None Common,
      - >-
        Patient information:   Handouts on this topic are available 

        at https:// family doctor.org/preventing-the-flu and https:// 

        family doctor.org/flu-myths.

        Influenza is an acute viral respiratory infection that causes
        significant morbidity and mortality worldwide. Three types of influ-

        enza cause disease in humans. Influenza A is the type most responsible
        for causing pandemics because of its high susceptibility 

        to antigenic variation. Influenza is highly contagious, and the hallmark
        of infection is abrupt onset of fever, cough, chills or 

        sweats, myalgias, and malaise. For most patients in the outpatient
        setting, the diagnosis is made clinically, and laboratory con-
      - |-
        www.aafp.org/fpm/2017/0900/p6.html
         22.  Centers for Disease Control and Prevention. Influenza (flu):  immuno -
        genicity, efficacy, and effectiveness of influenza vaccines. Updated 
        August 23, 2018. Accessed January 22, 2019. https:// www.cdc.gov/flu/
        professionals/acip/2018-2019/background/immunogenicity.htm
         23.  DiazGranados CA, Dunning AJ, Kimmel M, et al. Efficacy of high-dose 
        versus standard-dose influenza vaccine in older adults. N Engl J Med. 
        2014; 371(7): 635-645.
         24.  DiazGranados CA, Robertson CA, Talbot HK, et al. Prevention of serious 
        events in adults 65 years of age or older:  a comparison between high-
        dose and standard-dose inactivated influenza vaccines. Vaccine. 2015;  
        33(38): 4988-4993.
  - source_sentence: >-
      How does the negative likelihood ratio for digital immunoassays compare
      between adults and children for Influenza A?
    sentences:
      - |-
        17.  Erlikh IV, Abraham S, Kondamudi VK. Management of influenza. Am 
        Fam Physician . 2010;  82(9): 1087-1095. Accessed September  5, 2019. 
        https:// www.aafp.org/afp/2010/1101/p1087.html
         18.  Centers for Disease Control and Prevention. Influenza (flu):  for clini -
        cians:  antiviral medication. Updated Decemebr 27, 2018. Accessed 
        February 24, 2019. https:// www.cdc.gov/flu/professionals/antivirals/
        summary-clinicians.htm
         19.  Centers for Disease Control and Prevention. Influenza (flu):  guide for  
        considering influenza testing. Updated March 4, 2019. Accessed Octo -
        ber 5, 2019. https:// www.cdc.gov/flu/professionals/diagnosis/consider-
        influenza-testing.htm
      - |-
        TABLE 3
        Accuracy of Point-of-Care Tests for Influenza
        Test
        Positive  
        likelihood  
        ratio
        Negative 
        likelihood 
        ratio
        Low prevalence (5%) High prevalence (33%)
        Positive 
        predictive 
        value (%)
        Negative 
        predictive 
        value (%)
        Positive 
        predictive 
        value (%)
        Negative 
        predictive 
        value (%)
        Influenza A
        Adults       
        Commercially available rapid influenza tests 85 0.58 82 3 98 22
        Digital immunoassays 23 0.25 55 1 92 11
        Rapid nucleic acid amplification tests 44 0.13 70 1 96 6
        Children       
        Commercially available rapid influenza tests 76 0.39 80 2 97 16
        Digital immunoassays 46 0.13 71 1 96 6
        Rapid nucleic acid amplification tests 90 0.10 83 0 98 5
        Influenza B
        Adults       
        Commercially available rapid influenza tests 332 0.67 95 3 99 25
      - |-
        recommended dosages. 28 However, extended treatment 
        courses may be indicated in critically ill patients. 18 Support-
        ive treatment and management of complications, including 
        potential secondary bacterial pneumonia, are paramount. 
        Corticosteroids are not recommended unless the patient 
        has another approved indication for their use.18,28 Treatment 
        resistance should be considered in patients who take anti -
        virals and develop lower respiratory tract disease, although 
        this is less likely than natural disease progression and more 
        common in immunosuppressed patients.18
        Pregnancy is an independent risk factor for complicated 
        influenza. The risk of maternal death increases with each
  - source_sentence: >-
      What is the role of ipratropium in the treatment of the common cold
      according to the context?
    sentences:
      - >-
        sistent, dry or wet

        Common None or low 

        grade

        Uncommon Mild Common Uncommon Uncommon Common, mild Common

        Allergic 

        rhinitis

        Gradual Common, chronic Possible, especially 

        on awakening

        None Common, 

        prominent

        None Common Prominent Common Uncommon Uncommon

        Bacterial 

        sinusitis

        Gradual Common Common Common Common Common Uncommon Uncommon Common
        Common Uncommon

        Common 

        cold

        Gradual Common, dry Common None or low 

        grade

        Common Mild Common Common Common Common, mild Uncommon

        Influenza Abrupt Common, dry 

        hacking

        Common Characteristic;    

        high and rises 

        rapidly

        Common Early, 

        prominent

        Uncommon Uncommon Possible Prominent Uncommon

        Pertussis Gradual Prominent, parox-

        ysmal, whoop-like

        Uncommon None or low 

        grade
      - |-
        common cold are inhibited by intranasal administration
        of ipratropium.25 The nasal discharge also consists of a
        protein-rich plasma exudate derived from subepithelial
        capillaries,28 which may explain why anticholinergics
        only partly inhibit nasal discharge associated with
        URTIs.27
        The colour of nasal discharge and sputum is often
        used as a clinical marker to determine whether or not to
        prescribe antibiotics but there is no evidence from the
        literature that supports this concept,29 since colour
        changes in nasal discharge or sputum reﬂect the severity
        of the inﬂammatory response30 rather than the nature of
        the infection. Much of the literature relates to colour
        changes in sputum and the lower airways but the same
      - |-
        release by leukocytic pyrogen (interleukin-1). A mechanism for the
        increased degradation of muscle proteins during fever. N Engl J
        Med1983; 308: 553–58.
        64 Kotler DP. Cachexia. Ann Intern Med2000; 133: 622–34. 
        65 Ferreira SH. Prostaglandins, pain, and inﬂammation. Agents
        Actions Suppl1986; 19: 91–98.
  - source_sentence: >-
      QUESTION #1: How might changes in posture from sitting to supine affect
      sinus pain according to the context?
    sentences:
      - |-
        gas absorption from the sinus and “vacuum maxillary
        sinusitis”.37 However, sinuses with patent ostia may also
        be painful, indicating that the generation of
        inﬂammatory mediators within the sinus may be
        sufﬁcient to trigger the sensation of pain either by direct
        stimulation of pain nerve ﬁbres or via distension of blood
        vessels that are also served by sensory nerves.36 Changes
        in posture from sitting to supine cause an increase in
        sinus pain that may be related to dilation of the blood
        vessels draining the sinus caused by an increase in
        venous pressure. Pressure changes in the sinus may also
        cause pain by stimulation of branches of the trigeminal
        nerve that course in and around the sinuses.37
        Watery eyes
      - |-
        American Indians and Alaska Natives
        Children younger than 5 years (particularly those younger 
        than 2 years)
        Institutionalized adults (e.g., residents of nursing homes or 
        chronic care facilities)
        Pregnant and postpartum women (up to 2 weeks postpartum, 
        including pregnancy loss)
        Adapted with permission from Erlikh IV, Abraham S, Kondamudi VK. 
        Management of influenza. Am Fam Physician. 2010; 82(9): 1088, with 
        additional information from references 18 and 19.
      - |-
        sary Antibiotics
        Step Examples
        Explain why 
        antibiotics will 
        not help
        “The common cold is caused by a virus, so antibiot -
        ics won’t help.”
        “Antibiotics can’t fight viruses like colds. Taking them 
        won’t do any good this time and may hurt their 
        chances of fighting bacterial infections you might 
        get in the future.”
        Suggest treat-
        ments that might 
        help
        “You can try honey for your cough, ibuprofen or 
        acetaminophen for your muscle aches, and nasal or 
        oral decongestants with or without an antihistamine 
        for your congestion.”
        Manage expec-
        tations for length 
        of illness
        “Cold viruses can make you feel lousy. Most people 
        start to feel better after about a week, but some -
        times the cough can last even longer, especially if 
        you smoke.”
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
model-index:
  - name: SentenceTransformer based on Snowflake/snowflake-arctic-embed-l
    results:
      - task:
          type: information-retrieval
          name: Information Retrieval
        dataset:
          name: Unknown
          type: unknown
        metrics:
          - type: cosine_accuracy@1
            value: 0.75
            name: Cosine Accuracy@1
          - type: cosine_accuracy@3
            value: 0.9166666666666666
            name: Cosine Accuracy@3
          - type: cosine_accuracy@5
            value: 1
            name: Cosine Accuracy@5
          - type: cosine_accuracy@10
            value: 1
            name: Cosine Accuracy@10
          - type: cosine_precision@1
            value: 0.75
            name: Cosine Precision@1
          - type: cosine_precision@3
            value: 0.3055555555555555
            name: Cosine Precision@3
          - type: cosine_precision@5
            value: 0.19999999999999998
            name: Cosine Precision@5
          - type: cosine_precision@10
            value: 0.09999999999999999
            name: Cosine Precision@10
          - type: cosine_recall@1
            value: 0.75
            name: Cosine Recall@1
          - type: cosine_recall@3
            value: 0.9166666666666666
            name: Cosine Recall@3
          - type: cosine_recall@5
            value: 1
            name: Cosine Recall@5
          - type: cosine_recall@10
            value: 1
            name: Cosine Recall@10
          - type: cosine_ndcg@10
            value: 0.8864909792836682
            name: Cosine Ndcg@10
          - type: cosine_mrr@10
            value: 0.8486111111111113
            name: Cosine Mrr@10
          - type: cosine_map@100
            value: 0.8486111111111111
            name: Cosine Map@100

SentenceTransformer based on Snowflake/snowflake-arctic-embed-l

This is a sentence-transformers model finetuned from Snowflake/snowflake-arctic-embed-l. It maps sentences & paragraphs to a 1024-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: Snowflake/snowflake-arctic-embed-l
Maximum Sequence Length: 512 tokens
Output Dimensionality: 1024 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': 512, 'do_lower_case': False}) with Transformer model: BertModel 
  (1): Pooling({'word_embedding_dimension': 1024, '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.

from sentence_transformers import SentenceTransformer

# Download from the 🤗 Hub
model = SentenceTransformer("Gonalb/flucold-ft-v0")
# Run inference
sentences = [
    'QUESTION #1: How might changes in posture from sitting to supine affect sinus pain according to the context?',
    'gas absorption from the sinus and “vacuum maxillary\nsinusitis”.37 However, sinuses with patent ostia may also\nbe painful, indicating that the generation of\ninﬂammatory mediators within the sinus may be\nsufﬁcient to trigger the sensation of pain either by direct\nstimulation of pain nerve ﬁbres or via distension of blood\nvessels that are also served by sensory nerves.36 Changes\nin posture from sitting to supine cause an increase in\nsinus pain that may be related to dilation of the blood\nvessels draining the sinus caused by an increase in\nvenous pressure. Pressure changes in the sinus may also\ncause pain by stimulation of branches of the trigeminal\nnerve that course in and around the sinuses.37\nWatery eyes',
    'American Indians and Alaska Natives\nChildren younger than 5 years (particularly those younger \nthan 2 years)\nInstitutionalized adults (e.g., residents of nursing homes or \nchronic care facilities)\nPregnant and postpartum women (up to 2 weeks postpartum, \nincluding pregnancy loss)\nAdapted with permission from Erlikh IV, Abraham S, Kondamudi VK. \nManagement of influenza. Am Fam Physician. 2010; 82(9): 1088, with \nadditional information from references 18 and 19.',
]
embeddings = model.encode(sentences)
print(embeddings.shape)
# [3, 1024]

# Get the similarity scores for the embeddings
similarities = model.similarity(embeddings, embeddings)
print(similarities.shape)
# [3, 3]

Evaluation

Metrics

Information Retrieval

Evaluated with InformationRetrievalEvaluator

Metric	Value
cosine_accuracy@1	0.75
cosine_accuracy@3	0.9167
cosine_accuracy@5	1.0
cosine_accuracy@10	1.0
cosine_precision@1	0.75
cosine_precision@3	0.3056
cosine_precision@5	0.2
cosine_precision@10	0.1
cosine_recall@1	0.75
cosine_recall@3	0.9167
cosine_recall@5	1.0
cosine_recall@10	1.0
cosine_ndcg@10	0.8865
cosine_mrr@10	0.8486
cosine_map@100	0.8486

Training Details

Training Dataset

Unnamed Dataset

Size: 334 training samples
Columns: sentence_0 and sentence_1
Approximate statistics based on the first 334 samples:
sentence_0 sentence_1
type string string
details
min: 12 tokens
mean: 24.85 tokens
max: 61 tokens

min: 61 tokens
mean: 159.74 tokens
max: 248 tokens

	sentence_0	sentence_1
type	string	string
details	min: 12 tokens mean: 24.85 tokens max: 61 tokens	min: 61 tokens mean: 159.74 tokens max: 248 tokens

Samples:

sentence_0	sentence_1
`QUESTION #1: What is the source website from which the document was downloaded?`	`Downloaded from the American Family Physician website at www.aafp.org/afp. Copyright © 2019 American Academy of Family Physicians. For the private, noncom - mercial use of one individual user of the website. All other rights reserved. Contact copyrights@aafp.org for copyright questions and/or permission requests.`
`QUESTION #2: Who should be contacted for copyright questions and/or permission requests regarding the document?`	`Downloaded from the American Family Physician website at www.aafp.org/afp. Copyright © 2019 American Academy of Family Physicians. For the private, noncom - mercial use of one individual user of the website. All other rights reserved. Contact copyrights@aafp.org for copyright questions and/or permission requests.`
`QUESTION #1: Why is early diagnosis essential for antiviral therapy and public-health measures in the community?`	syndrome (SARS) 3 because early diagnosis is essential for any antiviral therapy and for the initiation of public- health measures in the community (eg, isolation of infected cases). Here, I discuss the mechanisms that generate symptoms associated with URTIs, especially common cold and ﬂu, but will not review virology in any detail except as regards relevance to symptoms. Is it a cold or ﬂu? The clinical expression of URTIs is variable and is partly inﬂuenced by the nature of the infecting virus but to a greater extent is modulated by the age, physiological state, and immunological experience of the host. 4 Depending on these factors, URTIs may occur without symptoms, may kill, or most commonly

Loss: MatryoshkaLoss with these parameters:

{
    "loss": "MultipleNegativesRankingLoss",
    "matryoshka_dims": [
        768,
        512,
        256,
        128,
        64
    ],
    "matryoshka_weights": [
        1,
        1,
        1,
        1,
        1
    ],
    "n_dims_per_step": -1
}

Training Hyperparameters

Non-Default Hyperparameters

eval_strategy: steps
per_device_train_batch_size: 10
per_device_eval_batch_size: 10
num_train_epochs: 10
multi_dataset_batch_sampler: round_robin

All Hyperparameters

Click to expand

overwrite_output_dir: False
do_predict: False
eval_strategy: steps
prediction_loss_only: True
per_device_train_batch_size: 10
per_device_eval_batch_size: 10
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: 10
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
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
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
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
average_tokens_across_devices: False
prompts: None
batch_sampler: batch_sampler
multi_dataset_batch_sampler: round_robin

Training Logs

Epoch	Step	cosine_ndcg@10
1.0	34	0.9108
1.4706	50	0.9098
2.0	68	0.8834
2.9412	100	0.9051
3.0	102	0.9066
4.0	136	0.9205
4.4118	150	0.9019
5.0	170	0.9156
5.8824	200	0.9247
6.0	204	0.9238
7.0	238	0.9019
7.3529	250	0.8856
8.0	272	0.8856
8.8235	300	0.8879
9.0	306	0.8879
10.0	340	0.8865

Framework Versions

Python: 3.11.11
Sentence Transformers: 3.4.1
Transformers: 4.48.3
PyTorch: 2.5.1+cu124
Accelerate: 1.3.0
Datasets: 3.3.2
Tokenizers: 0.21.0

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",
}

MatryoshkaLoss

@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

@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}
}