DeBERTa-small-ST-v1-toytest

SentenceTransformer based on microsoft/deberta-v3-small

This is a sentence-transformers model finetuned from microsoft/deberta-v3-small on the negation-triplets, vitaminc-pairs, scitail-pairs-qa, scitail-pairs-pos, xsum-pairs, sciq_pairs, qasc_pairs, openbookqa_pairs, msmarco_pairs, nq_pairs, trivia_pairs, gooaq_pairs and paws-pos datasets. 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: microsoft/deberta-v3-small <!-- at revision a36c739020e01763fe789b4b85e2df55d6180012 -->
• Maximum Sequence Length: 512 tokens
• Output Dimensionality: 768 tokens
• Similarity Function: Cosine Similarity
• Training Datasets:
  • negation-triplets
  • vitaminc-pairs
  • scitail-pairs-qa
  • scitail-pairs-pos
  • xsum-pairs
  • sciq_pairs
  • qasc_pairs
  • openbookqa_pairs
  • msmarco_pairs
  • nq_pairs
  • trivia_pairs
  • gooaq_pairs
  • paws-pos
• Language: en

<!-- - **License:** Unknown -->

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: DebertaV2Model 
  (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("bobox/DeBERTa-small-ST-v1-toytest")
# Run inference
sentences = [
    'who did ben assault in home and away',
    "List of Home and Away characters (2017) Ben and Maggie learn that Ziggy has been dating Brody in secret and they disapprove of the relationship. Ziggy leaves the house and Ben tells her not to come back. He apologises to her the next day, but does not accept her relationship with Brody, so Ziggy refuses to come home. Brody later breaks up with her. Ben begins making surf boards to sell at the pier. Ben finds Coco convulsing in the garden and he and Maggie learn she has bulimia. Ziggy later leaves home. Days later, Ben sees her with Brody, who is attempting to bring her home, and punches him in the face. Olivia Fraser Richards (Raechelle Banno) tells Sergeant Phillip McCarthy (Nicholas Cassim) and Ben is arrested. McCarthy and Kat Chapman (Pia Miller) learns he has a criminal record for assaulting his brother. Ben insults Kat, which leads to him being charged. Maggie secures a loan to get him out on bail. Maggie's mother, Diana (Sarah Chadwick) came to Summer Bay to visit the family and Diana told Ben and Maggie that she is the one who bailed Ben out of jail.",
    'Bad Things (Machine Gun Kelly and Camila Cabello song) "Bad Things" is a song by American rapper Machine Gun Kelly and Cuban-American singer Camila Cabello. The song was released on October 14, 2016 and was produced by The Futuristics. Its music video was directed by Hannah Lux Davis and premiered on December 1, 2016. The song features an interpolation of Fastball\'s 1999 single "Out of My Head". The single peaked at number four on the US Billboard Hot 100.',
]
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]

<!-- ### Direct Usage (Transformers) <details><summary>Click to see the direct usage in Transformers</summary> </details> --> <!-- ### Downstream Usage (Sentence Transformers) You can finetune this model on your own dataset. <details><summary>Click to expand</summary> </details> --> <!-- ### Out-of-Scope Use *List how the model may foreseeably be misused and address what users ought not to do with the model.* -->

Evaluation

Metrics

Semantic Similarity

• Dataset: sts-test
• Evaluated with <code>EmbeddingSimilarityEvaluator</code>

Triplet

• Dataset: NLI-v2
• Evaluated with <code>TripletEvaluator</code>

Binary Classification

• Dataset: VitaminC
• Evaluated with <code>BinaryClassificationEvaluator</code>

<!-- ## Bias, Risks and Limitations *What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.* --> <!-- ### Recommendations *What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.* -->

Training Details

Training Datasets

negation-triplets

• Dataset: negation-triplets
• Size: 26,000 training samples
• Columns: <code>anchor</code>, <code>entailment</code>, and <code>negative</code>
• Approximate statistics based on the first 1000 samples:

  	anchor	entailment	negative
  type	string	string	string
  details	<ul><li>min: 5 tokens</li><li>mean: 22.32 tokens</li><li>max: 124 tokens</li></ul>	<ul><li>min: 4 tokens</li><li>mean: 14.05 tokens</li><li>max: 42 tokens</li></ul>	<ul><li>min: 4 tokens</li><li>mean: 14.36 tokens</li><li>max: 42 tokens</li></ul>

• Samples:

  anchor	entailment	negative
  <code>BraÅ ov is part of the Transylvania area .</code>	<code>Like many other cities in Transylvania , BraÅ ov is also home for a significant ethnic Hungarian minority .</code>	<code>Like many other cities in Transylvania, BraÅ ov is also home for a significant ethnic Romanian majority.</code>
  <code>If some of the principles of supersymmetry are correct , it is possible to recreate these superparticles with particle accelerators . This attempt could prove or disprove the ideas of supersymmetry .</code>	<code>It is possible to have more than one kind of supersymmetry transformation .</code>	<code>It is impossible to find even one kind of supersymmetry transformation.</code>
  <code>A group of people running a bicycle race past a red building.</code>	<code>A bunch of people run past a building</code>	<code>A few people stay still near a building</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

vitaminc-pairs

• Dataset: vitaminc-pairs at be6febb
• Size: 24,000 training samples
• Columns: <code>claim</code> and <code>evidence</code>
• Approximate statistics based on the first 1000 samples:

  	claim	evidence
  type	string	string
  details	<ul><li>min: 7 tokens</li><li>mean: 17.43 tokens</li><li>max: 67 tokens</li></ul>	<ul><li>min: 8 tokens</li><li>mean: 37.41 tokens</li><li>max: 366 tokens</li></ul>

• Samples:

  claim	evidence
  <code>By March 2016 , Baby was above the 8th most viewed YouTube video .</code>	<code>On March 5 , 2014 , Baby '' became the second video , after Gangnam Style '' , to receive 1 billion views on YouTube , and is the ninth most viewed video on the site , with over 1.33 billion views as of March 2016 .</code>
  <code>The movie Think Like A Man had a rating of less than 50 % on Metacritic .</code>	<code>Early reviews for the film were mixed , the film currently holds a 47 % on Metacritic , indicating `` mixed or average reviews '' .</code>
  <code>Animal consumption at the Huanan Seafood Market is suspected to be where the severe acute respiratory syndrome coronavirus 2 originated .</code>	<code>Animals sold for food are suspected to be the reservoir because many of first identified infected individuals were workers at the Huanan Seafood Market .</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

scitail-pairs-qa

• Dataset: scitail-pairs-qa at 0cc4353
• Size: 14,237 training samples
• Columns: <code>sentence2</code> and <code>sentence1</code>
• Approximate statistics based on the first 1000 samples:

  	sentence2	sentence1
  type	string	string
  details	<ul><li>min: 7 tokens</li><li>mean: 16.12 tokens</li><li>max: 41 tokens</li></ul>	<ul><li>min: 7 tokens</li><li>mean: 15.23 tokens</li><li>max: 41 tokens</li></ul>

• Samples:

  sentence2	sentence1
  <code>Instruments that measure the angle of the slope of a volcano are called tilt meters.</code>	<code>Instruments that measure the angle of the slope of a volcano are called what?</code>
  <code>Ultrasound, a diagnostic technology, uses high-frequency vibrations transmitted into any tissue in contact with the transducer.</code>	<code>What diagnostic technology uses high-frequency vibrations transmitted into any tissue in contact with the transducer?</code>
  <code>Many species of birds in new england fly south for the winter months to find an environment with more food.</code>	<code>Which of the following best explains why many species of birds in New England fly south for the winter months?</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

scitail-pairs-pos

• Dataset: scitail-pairs-pos at 0cc4353
• Size: 8,600 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 8 tokens</li><li>mean: 23.36 tokens</li><li>max: 74 tokens</li></ul>	<ul><li>min: 7 tokens</li><li>mean: 15.79 tokens</li><li>max: 41 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>Anyway, what makes it possible for insects to walk on water is called Surface tension .</code>	<code>Surface tension is responsible for the fact that small insects can walk on water.</code>
  <code>Elastic potential energy is the potential energy of an elastic object (for example a bow or a catapult) that is deformed under tension or compression (or stressed in formal terminology).</code>	<code>The term elastic potential energy is used to describe potential energy due to an object’s shape.</code>
  <code>But, season or not, tornadoes can occur at any time of the year, if the weather conditions are right.</code>	<code>Tornadoes can occur in any.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

xsum-pairs

• Dataset: xsum-pairs
• Size: 24,000 training samples
• Columns: <code>document</code> and <code>summary</code>
• Approximate statistics based on the first 1000 samples:

  	document	summary
  type	string	string
  details	<ul><li>min: 45 tokens</li><li>mean: 255.53 tokens</li><li>max: 487 tokens</li></ul>	<ul><li>min: 8 tokens</li><li>mean: 25.67 tokens</li><li>max: 42 tokens</li></ul>

• Samples:

  document	summary
  <code>Jean Galligan, 82, from Dumfries, died when her car caught fire after it was involved in a collision with a Vauxhall Zafira on 14 May.<br>Police Scotland said Mrs Galligan was driving a red Daihatsu which was burnt out as a result of the accident on the A76 at Holywood.<br>Neither the driver nor the front seat passenger in the Zafira were injured.</code>	<code>A woman killed in a road accident near Dumfries has been named by police.</code>
  <code>Police officers carried out arrests on Thursday in connection with alleged sex offences against females which occurred between 2008 and 2015.<br>Six men were charged with the rape of a girl under 16 as well as other sexual offences, while a seventh man was charged with conspiracy to rape.<br>Six men - all from Oxford - will appear before Oxford magistrates.<br>They are: Shabir Dogar, 22; Shabaz Khan, 23; Shohab Dogar, 23; Yasin Hamid, 20; Usman Iddris, 22; and Joseph Suraina, 22.<br>Waqas Hussain, 24, of no fixed abode, will appear at Oxford Magistrates' Court on 4 April.<br>Mr Hussain has also been charged with the attempted sexual assault of a girl under 13, as have Shabir Dogar and Shohab Dogar.<br>The raids were part of what the police are calling Operation Nautical.<br>A further 10 men were also arrested on Wednesday as part of the same operation.</code>	<code>Seven men have been charged in connection with a major child sexual exploitation investigation in Oxford.</code>
  <code>In February 1957, 11-year-old Moira Anderson left her grandmother's house in Coatbridge to go to the shops but never returned.<br>Bus driver and convicted paedophile Alexander Gartshore, who died in 2006, is suspected of her murder.<br>Police are now looking at an area of Monkland Canal in an attempt to find her remains.<br>Moira Anderson was last seen on 23 February 1957 when she left on an errand during a heavy snowstorm, and boarded a Baxter's bus that was driven by Gartshore.<br>Later that year, he was jailed for raping a 17-year-old babysitter.<br>In 1999, convicted child abuser James Gallogley named his former friend Gartshore as Moira's murderer.<br>Gartshore's own daughter Sandra Brown was convinced he was the killer and campaigned to have him charged.<br>In 2014 prosecutors took the unusual step of announcing that Gartshore would have faced prosecution for the schoolgirl's murder if he were still alive.<br>In 1957 a witness reported seeing a tall man carrying a large, heavy sack towards the canal the morning after Moira disappeared but the possible sighting was not followed up.<br>Four years ago a grave in Old Monkland Cemetery in Coatbridge was exhumed as part of the search but no evidence was found that Moira was buried there.</code>	<code>A new search to find a schoolgirl who disappeared 60 years in North Lanarkshire has begun.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

sciq_pairs

• Dataset: sciq_pairs at 2c94ad3
• Size: 11,095 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 7 tokens</li><li>mean: 16.67 tokens</li><li>max: 60 tokens</li></ul>	<ul><li>min: 2 tokens</li><li>mean: 82.57 tokens</li><li>max: 512 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>The punnett square shows the possible what, and their most likely ratios?</code>	<code>If the parents had four offspring, their most likely genotypes would be one BB, two Bb, and one bb. But the genotype ratios of their actual offspring may differ. That's because which gametes happen to unite is a matter of chance, like a coin toss. The Punnett square just shows the possible genotypes and their most likely ratios.</code>
  <code>Which hormones work together to control the level of glucose in the blood?</code>	<code>The pancreas is located near the stomach. Its hormones include insulin and glucagon. These two hormones work together to control the level of glucose in the blood. Insulin causes excess blood glucose to be taken up by the liver, which stores the glucose as glycogen. Glucagon stimulates the liver to break down glycogen into glucose and release it back into the blood. The pancreas also secretes digestive enzymes into the digestive tract.</code>
  <code>What is the“packet” of energy called that the nucleus emits during gamma decay?</code>	<code>Gamma rays are produced during gamma decay of an excited nucleus. During gamma decay, the nucleus emits a “packet” of energy called a gamma particle.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

qasc_pairs

• Dataset: qasc_pairs at a34ba20
• Size: 7,727 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 5 tokens</li><li>mean: 11.36 tokens</li><li>max: 22 tokens</li></ul>	<ul><li>min: 14 tokens</li><li>mean: 34.66 tokens</li><li>max: 66 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>Orbiting spacecraft reentering the Earth's atmosphere</code>	<code>friction causes an object to lose energy. Dear Ashlee, The heat in the reentry phase is due to friction between the spacecraft and the air.. Spacecraft that are reentering the atmosphere lose energy</code>
  <code>The chance of you developing cancer depends most on your</code>	<code>Cancer genes can be inherited.. Genes are inherited from parents.. Developing cancer can depend on your parents</code>
  <code>Why does a snake look for shelter in the winter?</code>	<code>shelter is used for protection by animals against weather. Many snakes seek shelter from the winter weather by holding up in dens.. Snakes use shelter to protect themselves in the winter</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

openbookqa_pairs

• Dataset: openbookqa_pairs
• Size: 4,522 training samples
• Columns: <code>question</code> and <code>fact</code>
• Approximate statistics based on the first 1000 samples:

  	question	fact
  type	string	string
  details	<ul><li>min: 3 tokens</li><li>mean: 13.8 tokens</li><li>max: 78 tokens</li></ul>	<ul><li>min: 4 tokens</li><li>mean: 11.5 tokens</li><li>max: 30 tokens</li></ul>

• Samples:

  question	fact
  <code>What is animal competition?</code>	<code>if two animals eat the same prey then those animals compete for that pey</code>
  <code>If you wanted to make a metal bed frame, where would you start?</code>	<code>alloys are made of two or more metals</code>
  <code>Places lacking warmth have few what</code>	<code>cold environments contain few organisms</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

msmarco_pairs

• Dataset: msmarco_pairs at 28ff31e
• Size: 22,000 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 4 tokens</li><li>mean: 8.75 tokens</li><li>max: 38 tokens</li></ul>	<ul><li>min: 14 tokens</li><li>mean: 76.85 tokens</li><li>max: 201 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>what is hydrolysis in digestion</code>	<code>Digestion and Hydrolysis The digestion process relies upon hydrolysis to render the biochemical reactions that break down food. The digestive tract secretes enzymes, such as proteases, carbohydrases, nucleases and lipases that, along with water, catalyze the hydrolysis that releases various nutrients.</code>
  <code>is cartier a good watch</code>	<code>CARTIER WATCHES. AuthenticWatches.com is one of the largest Internet Dealers for authentic Cartier watches. Cartier watches have no equal with respect to elegance and luxury. The name Cartier is synonymous with exquisite luxury and quality. Founded in 1847 by Louis-Fran ois Cartier, Cartier has led the industry in jewelry and watches alike. Cartier watches boast a large variety of design and functionality, yet maintain the utmost quality and sophistication in every series.</code>
  <code>what vitamin is a precursor for a neurotransmitter</code>	<code>Tryptophan is an essential amino acid which is the precursor of serotonin. Serotonin is a brain neurotransmitter, platelet clotting factor and neurohormone found in organs throughout the body. Metabolism of tryptophan to serotonin requires nutrients such as vitamin B6, niacin and glutathione.ower doses, as little as 1000 to 2000 mg, have been found to be effective clinically, as well as experimentally in animals. The minimum daily requirement for adults of tyrosine and its precursor, phenylalanine, is 16 mg/kg a day or about 1000 mg total. Hence, 6 g is at least six times the minimum daily requirement.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

nq_pairs

• Dataset: nq_pairs at f9e894e
• Size: 22,000 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 10 tokens</li><li>mean: 11.92 tokens</li><li>max: 27 tokens</li></ul>	<ul><li>min: 15 tokens</li><li>mean: 132.65 tokens</li><li>max: 512 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>friends episode with the turkey on the head</code>	<code>The One with All the Thanksgivings "The One with All the Thanksgivings" (also known as "The One with the Thanksgiving Flashbacks"[2]) is the eighth episode of the fifth season of Friends. It first aired on the NBC network in the United States on November 19, 1998. In the episode, the main characters spend Thanksgiving at Monica's (Courteney Cox) apartment and begin telling stories about their worst Thanksgivings: Chandler (Matthew Perry) learning of his parents' divorce, Phoebe (Lisa Kudrow) losing arms in past lives and Joey (Matt LeBlanc) having his head stuck in a turkey. Rachel (Jennifer Aniston) reveals Monica's worst Thanksgiving—accidentally cutting off Chandler's toe after he called her "fat" in their first encounter. When Monica begs Chandler to forgive her, he accidentally reveals that he loves her.</code>
  <code>who played the first buford pusser in walking tall</code>	<code>Walking Tall (1973 film) Buford Pusser (Joe Don Baker), at his wife Pauline's (Elizabeth Hartman) behest, retires from the professional wrestling ring and moves back to Tennessee to start a logging business with his father, Carl Pusser (Noah Beery, Jr.).</code>
  <code>when did the us let go of the philippines</code>	<code>History of the Philippines (1898–1946) The history of the Philippines from 1898 to 1946 covers the period of American rule in the Philippines and began with the outbreak of the Spanish–American War in April 1898, when the Philippines was still part of the Spanish East Indies, and concluded when the United States formally recognised the independence of the Republic of the Philippines on July 4, 1946.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

trivia_pairs

• Dataset: trivia_pairs at a7c36e3
• Size: 20,000 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 8 tokens</li><li>mean: 18.89 tokens</li><li>max: 60 tokens</li></ul>	<ul><li>min: 15 tokens</li><li>mean: 456.48 tokens</li><li>max: 512 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>With the symbol Wb what is the unit of magnetic flux?</code>	<code>schoolphysics ::Welcome:: HOME > AGE 16 - 19 > ELECTRICITY AND MAGNETISM > ELECTROMAGNETISM > FLUX AND FLUX DENSITY Flux and flux density To understand the meaning of magnetic flux (Φ) and magnetic flux density (B) think first about an ordinary bar magnet. Around the magnet there is a magnetic field and this gives a �flow of magnetic energy� around the magnet. It is this flow of energy that we call magnetic flux (Φ). We think of magnetic flux as flowing from the north pole of a magnet round to its south pole as shown by the arrows on the lines in the diagram. Looking at the diagram you should see that there is as much flux flowing �from the north pole� as there is �flowing into the south pole�. Magnetic flux is given the symbol Φ and is measured in units called Webers (Wb). However the amount of magnetic flux flowing through a given area will change from one point to another around the magnet and you can understand this by thinking about a loop of wire placed in the field at two different points (A and B). You can see that in position B there are a smaller number of magnetic field lines passing through the loop than there is when it is in position A. We call the amount of flux passing through a unit area at right angles to the magnetic field lines the flux density (B) at that point. Flux density is measured in Tesla (T) where 1 T = 1 Wbm-2 So: Flux (Φ) = Flux density (B) x area through which flux passes (A)    Φ = BA If we now use more than one loop of wire, in others words a coil of N turns as shown in position C the flux flowing through the N turns is simply N times that flowing through the single loop. The quantity NΦ is called the the flux linkage for the coil at that point. Therefore:</code>
  <code>The informal term for a gangster, especially belonging to the Mafia is?</code>	<code>goodfella - definition of goodfella in English
  <code>What did the band S Club 7 change their name to when Paul Cattermole left in June 2002?</code>	<code>BBC News

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

gooaq_pairs

• Dataset: gooaq_pairs at b089f72
• Size: 20,000 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 8 tokens</li><li>mean: 11.52 tokens</li><li>max: 19 tokens</li></ul>	<ul><li>min: 16 tokens</li><li>mean: 57.68 tokens</li><li>max: 121 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>gyan is called in english?</code>	<code>Gyan (Sanskrit), a Sanskrit word that roughly translates to 'knowledge' in English.</code>
  <code>are mud baths good for dogs?</code>	<code>Mud has many benefits for your dog. It can soothe irritations by removing dead irritated skin. It can soothe hot spots. The mud applied to your dogs coat during the bath can help moisturize the skin and remove dandruff.</code>
  <code>how many calories do you burn doing interval running?</code>	<code>Cost in Calories exerciser who performs sprint intervals at a speed of 12 mph for 20 minutes expends roughly 608 calories. To demonstrate how body weight affects calorie expenditure, a 250-lb. person burns approximately 845 calories using the identical sprint speed and workout duration.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

paws-pos

• Dataset: paws-pos at 161ece9
• Size: 21,829 training samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 9 tokens</li><li>mean: 25.38 tokens</li><li>max: 51 tokens</li></ul>	<ul><li>min: 9 tokens</li><li>mean: 25.41 tokens</li><li>max: 50 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>After some protests from the girl 's father , the man of Douji and Hime was killed before Suzu and her father 's corpse were consumed by the Orochi .</code>	<code>After some protesting from the girl 's father , the man was killed by Douji and Hime before Suzu and her father 's corpse were consumed by the Orochi .</code>
  <code>162 . Fighter Escadrille was a unit of the Łódź Army at the start of the Second World War . The unit was attached to the Polish Air Force .</code>	<code>At the beginning of the Second World War , the Fighter Escadrille was a unit of the Łódź army , which was attached to the Polish Air Force .</code>
  <code>The first music video for the album filmed and edited for the song 'Trust You ' was made by Pierre Bouvier-Patron of Friends Studio , London .</code>	<code>The first music video for the album , filmed and edited for the song 'Trust You ' , was made by Pierre Bouvier-Patron of Friends Studio , London .</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

Evaluation Datasets

vitaminc-pairs

• Dataset: vitaminc-pairs at be6febb
• Size: 108 evaluation samples
• Columns: <code>claim</code> and <code>evidence</code>
• Approximate statistics based on the first 1000 samples:

  	claim	evidence
  type	string	string
  details	<ul><li>min: 9 tokens</li><li>mean: 21.36 tokens</li><li>max: 41 tokens</li></ul>	<ul><li>min: 11 tokens</li><li>mean: 36.11 tokens</li><li>max: 79 tokens</li></ul>

• Samples:

  claim	evidence
  <code>Dragon Con had over 5000 guests .</code>	<code>Among the more than 6000 guests and musical performers at the 2009 convention were such notables as Patrick Stewart , William Shatner , Leonard Nimoy , Terry Gilliam , Bruce Boxleitner , James Marsters , and Mary McDonnell .</code>
  <code>COVID-19 has reached more than 185 countries .</code>	<code>As of , more than cases of COVID-19 have been reported in more than 190 countries and 200 territories , resulting in more than deaths .</code>
  <code>In March , Italy had 3.6x times more cases of coronavirus than China .</code>	<code>As of 12 March , among nations with at least one million citizens , Italy has the world 's highest per capita rate of positive coronavirus cases at 206.1 cases per million people ( 3.6x times the rate of China ) and is the country with the second-highest number of positive cases as well as of deaths in the world , after China .</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

negation-triplets

• Dataset: negation-triplets
• Size: 64 evaluation samples
• Columns: <code>anchor</code>, <code>entailment</code>, and <code>negative</code>
• Approximate statistics based on the first 1000 samples:

  	anchor	entailment	negative
  type	string	string	string
  details	<ul><li>min: 10 tokens</li><li>mean: 13.88 tokens</li><li>max: 18 tokens</li></ul>	<ul><li>min: 10 tokens</li><li>mean: 13.31 tokens</li><li>max: 21 tokens</li></ul>	<ul><li>min: 10 tokens</li><li>mean: 13.64 tokens</li><li>max: 22 tokens</li></ul>

• Samples:

  anchor	entailment	negative
  <code>1 military jet fighter flying in formation alongside a 1 military propeller pilot. </code>	<code>The two planes are different in design, but flying in a similar flight pattern.</code>	<code>The two planes are identical in design, but flying in different flight patterns.</code>
  <code>A random plane in the sky flying alone</code>	<code>An airplane flying high in the blue sky. </code>	<code>A helicopter flying low in the cloudy sky. </code>
  <code>A picture of a white gas range with figurines above.</code>	<code>a white stove turned off with a digital clock</code>	<code>a black stove turned on with a digital clock</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

scitail-pairs-pos

• Dataset: scitail-pairs-pos at 0cc4353
• Size: 54 evaluation samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 9 tokens</li><li>mean: 20.81 tokens</li><li>max: 45 tokens</li></ul>	<ul><li>min: 10 tokens</li><li>mean: 15.48 tokens</li><li>max: 23 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>humans normally have 23 pairs of chromosomes.</code>	<code>Humans typically have 23 pairs pairs of chromosomes.</code>
  <code>A solution is a homogenous mixture of two or more substances that exist in a single phase.</code>	<code>Solution is the term for a homogeneous mixture of two or more substances.</code>
  <code>Upwelling The physical process in near-shore ocean systems of rising of nutrients and colder bottom waters to the surface because of constant wind patterns along the shoreline.</code>	<code>Upwelling is the term for when deep ocean water rises to the surface.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

xsum-pairs

• Dataset: xsum-pairs
• Size: 128 evaluation samples
• Columns: <code>document</code> and <code>summary</code>
• Approximate statistics based on the first 1000 samples:

  	document	summary
  type	string	string
  details	<ul><li>min: 74 tokens</li><li>mean: 242.33 tokens</li><li>max: 374 tokens</li></ul>	<ul><li>min: 12 tokens</li><li>mean: 25.18 tokens</li><li>max: 38 tokens</li></ul>

• Samples:

  document	summary
  <code>The region is already struggling to cope with a huge influx of migrants arriving from Tunisia.<br>Since January, at least 15,000 migrants have arrived, many of them landing on the tiny island of Lampedusa which is struggling to cope.<br>Thousands of people are living in basic camps on the island, leading to health concerns and rising local tensions.<br>"Until now the only migrants to arrive in Lampedusa were Tunisians," said Laura Boldrini a spokeswoman for the UN's refugee agency.<br>"This is the first boat coming from Libya with people fleeing the military escalation, the vendettas and the retaliation attacks," she said.<br>Overnight on Saturday, a boat carrying some 300 migrants was escorted by the Italian coastguard to Linosa, an even smaller island some 50 km (35 miles) north of Lampedusa.<br>The passengers were mostly Somalis, Eritreans and Ethiopians and included a woman who had just given birth - she and the baby were flown to Lampedusa for medical care.<br>Several other boats from Libya, each carrying hundreds more migrants, are expected to reach Italy within hours.<br>Officials on Lampedusa, which is less than 160km from the Tunisian coast, have moved thousands of migrants to reception centres on the mainland, but some 5,000 remain.<br>The island's mayor has said he is desperate for help to relieve pressure on the island's very limited resources. Local people have said they are afraid of an outbreak of disease in the camps.<br>The Italian government has appealed to the international community for help.</code>	<code>Boatloads of migrants fleeing fighting in Libya are beginning to arrive in southern Italy, say officials.</code>
  <code>The Belgium midfielder was one of a handful of players Mourinho had deemed to have underperformed this season.<br>Mourinho has also been under scrutiny, with the champions 15th in the league, 14 points behind leaders Leicester.<br>"I don't have a problem with him. We hope we can win a lot of trophies together," said Hazard, 24.<br>"Maybe not this season because it will be difficult, but next season and on."<br>The Blues moved above Norwich into 15th with a 1-0 win over the Canaries on Saturday - it was only their fourth victory in 13 league games this season.<br>Hazard, who was recently linked with a move to Real Madrid, has yet to score for the Blues in 18 appearances in all competitions in 2015-16.<br>He managed 20 last season en route to winning the Professional Footballers' Association and Football Writers' Association player of the year awards.<br>On his own form, he said: "I didn't start the season well. I tried to find out why, but I don't know.<br>"Sometimes you don't know. You have to keep going. I gave everything in training, on the pitch when I played.<br>"I hope I can get a lot of form and try to help the team win games."</code>	<code>Eden Hazard has denied having a strained relationship with Chelsea boss Jose Mourinho and suggested he wants to stay at the Premier League champions.</code>
  <code>Amadou Gallo Fall, the NBA's vice-president for Africa, told the BBC the centre would train boys and girls aged between 16 and 18.<br>He said the centre would be part of its global network of elite training academies.<br>Several Africans have played for top teams in the NBA league.<br>Mr Fall, who is originally from Senegal, said the pan-African academy would use its network to scout for players from around the continent.<br>He said the players would be given access to facilities and resources available to elite players including nutritionists, personal coaches and physiotherapists.<br>The centre will be in Thies, 60km (40 miles) east of the capital, Dakar.<br>Senegal's national teams - men and women - have traditionally been among the strongest in Africa.<br>Mr Fall said those who don't make it to the NBA would have other avenues, such as "other great leagues around the world, including the NBA development league or in US universities".<br>He said there were 14 African-born players on the NBA opening roster this year, including Senegal's Gorgui Dieng and Cameroon's Pascal Siakam.<br>He added the NBA had a long association with the continent, citing legendary players such as Hakeem Olajuwon, Manute Bol and Dikembe Mutombo.<br>"That generation has paved the way and they've inspired and ushered a significant number of other young players over the years, a lot of them from Senegal," he said.<br>The NBA launched three academy centres in China in October, one in India last month and is planning to open another global centre based in Australia.<br>The NBA held its first game in Africa in 1 August 2015 in the South African city of Johannesburg.</code>	<code>The US National Basketball Association (NBA) has announced it will open its first African training academy in Senegal early next year.</code>

• Loss: <code>CachedGISTEmbedLoss</code> with these parameters:

  {'guide': SentenceTransformer(
    (0): Transformer({'max_seq_length': 512, 'do_lower_case': False}) 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()
  ), 'temperature': 0.05}
  

sciq_pairs

• Dataset: sciq_pairs at 2c94ad3
• Size: 128 evaluation samples
• Columns: <code>sentence1</code> and <code>sentence2</code>
• Approximate statistics based on the first 1000 samples:

  	sentence1	sentence2
  type	string	string
  details	<ul><li>min: 8 tokens</li><li>mean: 16.24 tokens</li><li>max: 37 tokens</li></ul>	<ul><li>min: 2 tokens</li><li>mean: 71.48 tokens</li><li>max: 375 tokens</li></ul>

• Samples: | sentence1 | sentence2 |

The model card is truncated. Read the rest on Hugging Face.