DeBERTa-small-ST-v1-test-step3

SentenceTransformer based on bobox/DeBERTa-small-ST-v1-test-step2

This is a sentence-transformers model finetuned from bobox/DeBERTa-small-ST-v1-test-step2 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: bobox/DeBERTa-small-ST-v1-test-step2 <!-- at revision 227c804cec7dd9eaab6a3cd4f9df268d4b5a1ca2 -->
• 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-test-step3")
# Run inference
sentences = [
    'where was the first hudson bay company post',
    "Hudson's Bay Company Hudson's Bay Company's first inland trading post was established by Samuel Hearne in 1774 in Cumberland House, Saskatchewan.[38][39]",
    'Steven Ogg Steven Ogg is a Canadian actor.[1] He is best known for his roles as Trevor Philips in the 2013 video game Grand Theft Auto V and Simon in The Walking Dead.[2] He has also appeared in television series such as Better Call Saul, Law & Order, Person of Interest, Broad City, and Westworld.',
]
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>

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: 35,750 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: 4 tokens</li><li>mean: 21.52 tokens</li><li>max: 101 tokens</li></ul>	<ul><li>min: 5 tokens</li><li>mean: 14.34 tokens</li><li>max: 47 tokens</li></ul>	<ul><li>min: 5 tokens</li><li>mean: 14.62 tokens</li><li>max: 44 tokens</li></ul>

• Samples:

  anchor	entailment	negative
  <code>Brunette woman in a white bikini pouring a drink into a man's cup.</code>	<code>A woman is serving a beverage to a man.</code>	<code>A woman is not serving a beverage to a man.</code>
  <code>People under a tent at a reception.</code>	<code>People are outside.</code>	<code>People are inside.</code>
  <code>A man is smoking at sunset.</code>	<code>The time in the picture is probably somewhere between 3PM and 9PM.</code>	<code>The time in the picture is probably not somewhere between 3PM and 9PM.</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: 34,375 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: 18.14 tokens</li><li>max: 54 tokens</li></ul>	<ul><li>min: 6 tokens</li><li>mean: 37.45 tokens</li><li>max: 192 tokens</li></ul>

• Samples:

  claim	evidence
  <code>Dora and the Lost City of Gold grossed more than $ 58 million in North America , more than $ 40 million in other territories , and more than $ 98 million globally .</code>	<code>, Dora and the Lost City of Gold has grossed $ 58.4 million in the United States and Canada , and $ 40.4 million in other territories , for a worldwide total of $ 98.8 million , against a production budget of $ 49 million .</code>
  <code>After Juan Antonio and Vicky Cristina have sex , they go to bed .</code>	<code>Afterwards , Juan Antonio lets her know he thinks she 's beautiful , and they make love , and eventually go to bed .</code>
  <code>Tove Lo 's single `` Disco Tits '' reached number 55 in Sweden .</code>	<code>Its lead single , '' Disco Tits '' '' , peaked at number 55 in Sweden. ''</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>question</code>
• Approximate statistics based on the first 1000 samples:

  	sentence2	question
  type	string	string
  details	<ul><li>min: 7 tokens</li><li>mean: 16.29 tokens</li><li>max: 41 tokens</li></ul>	<ul><li>min: 8 tokens</li><li>mean: 15.21 tokens</li><li>max: 35 tokens</li></ul>

• Samples:

  sentence2	question
  <code>When an atom gains or loses an electron it becames a(n) ion.</code>	<code>When an atom gains or loses an electron it becames an?</code>
  <code>The stratosphere is the layer above the troposphere.</code>	<code>What is the layer above the troposphere?</code>
  <code>Solar and wind are still expensive compared to fossil fuels.</code>	<code>Solar and wind are still expensive compared to what?</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: 7 tokens</li><li>mean: 23.52 tokens</li><li>max: 71 tokens</li></ul>	<ul><li>min: 7 tokens</li><li>mean: 15.54 tokens</li><li>max: 39 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>Ozone is a molecule formed of THREE oxygen atoms.</code>	<code>Three oxygen ions make up an ozone molecule.</code>
  <code>Animals live by eating the energy produced by plants (herbivores), or live by eating animals that eat plants (carnivores) Let's take the human being (a polyhagous animal) as an example and ponder upon it.</code>	<code>Herbivores is the term for animals that eat producers to get energy.</code>
  <code>Fertilization triggers Meiosis II, and then the sperm nucleus unites with the resulting egg nucleus.</code>	<code>When a sperm penetrates the egg, it triggers the egg to complete meiosis.</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: 38,500 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: 40 tokens</li><li>mean: 223.77 tokens</li><li>max: 443 tokens</li></ul>	<ul><li>min: 13 tokens</li><li>mean: 25.9 tokens</li><li>max: 45 tokens</li></ul>

• Samples:

  document	summary
  <code>William Heath, 12, from Earls Barton, has so far disguised himself as Ali G, Hercule Poirot and Groucho Marx as part of his Movember campaign.<br>He is raising money in memory of his grandfather, who died from cancer in February.<br>William came up with the idea as he cannot grow a moustache himself.<br>His granddad, Robin Wickham, died from pancreatic, testicular and prostate cancer, and William is now asking people to donate to The Movember Foundation, via his own "Mo Space" page.<br>William has already dressed as seven famous moustachioed men, including Biggles, Keith Lemon and Magnum PI.<br>His costumes were sourced from a combination of a local fancy dress shop and items around the home.<br>"My granddad and I were very close," he said. "He used to come on holiday with us and was always having fun."<br>His mum Clare said William had wanted to do something that would help, but was also fun.<br>She said: "We needed to fill 30 days. When you try to think of 30 people with moustaches it is harder than you think.<br>"The overriding thing is that it has been fun, but it has serious implications."<br>William said his favourite character so far had been Ali G, but he had struggled to create a successful Ned Flanders outfit.<br>He said: "I tried to paint my face yellow, but it didn't really work."</code>	<code>A schoolboy has decided to raise money for a men's health charity by dressing up as a different moustachioed famous face every day for a month.</code>
  <code>Media playback is not supported on this device<br>The Daily Mail claimed a £10m offer was made to owner Owen Oyston despite him stating the club was not for sale at a recent public meeting with fans.<br>Oyston was heckled throughout over the way the club is being run.<br>"The owners of the club have received no approach whatsoever, from any investor or third party," said a statement.<br>The Blackpool Supporters Trust launched a £16m bid to buy the club July 2015 but Oyston, whose son Karl is chairman, eventually ended those takeover talks.<br>The Seasiders, who will play in the fourth tier for the first time since 2000-01 this season, have also announced the signing of midfielder Danny Pugh.<br>The 33-year-old former Manchester United trainee joins on a free transfer after his release by Bury and has signed a one-year deal, with further 12-month option.<br>"He brings a wealth of experience to the team and will be a strong, vocal presence in the middle of the park," said boss Gary Bowyer.</code>	<code>League Two club Blackpool have denied claims they were subject of a new takeover offer.</code>
  <code>The Cochrane Collaboration carried out a systematic review of eight exercise trials involving more than 300 patients living at home or in care.<br>Exercise did little for patients' moods, the research concluded.<br>But it did help them carry out daily activities such as rising from a chair, and boosted their cognitive skills.<br>Whether these benefits improve quality of life is still unclear, but the study authors say the findings are reason for optimism.<br>Dementia affects some 800,000 people in the UK. And the number of people with the condition is steadily increasing because people are living longer.<br>It is estimated that by 2021, the number of people with dementia in the UK will have increased to around one million.<br>With no cure, ways to improve the lives of those living with the condition are vital.<br>Researcher Dorothy Forbes, of the University of Alberta, and colleagues who carried out the Cochrane review, said: "Clearly, further research is needed to be able to develop best practice guidelines to enable healthcare providers to advise people with dementia living at home or in institutions.<br>"We also need to understand what level and intensity of exercise is beneficial for someone with dementia."<br>Dr Laura Phipps of Alzheimer's Research UK said: "We do know that exercise is an important part of keeping healthy, and though we can't say that exercise will prevent dementia, evidence does suggest it can help reduce the risk of the condition as part of a healthy lifestyle."</code>	<code>People with dementia who exercise improve their thinking abilities and everyday life, a body of medical research concludes.</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: 16.92 tokens</li><li>max: 54 tokens</li></ul>	<ul><li>min: 2 tokens</li><li>mean: 85.56 tokens</li><li>max: 512 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>The bright color of poison dart frogs serves what purpose?</code>	<code>Poison dart frogs have toxins in their skin. Their bright colors warn potential predators not to take a bite!.</code>
  <code>What term describes motion that repeats itself at regular time intervals, such as exhibited by a vibrating guitar string?</code>	<code>When you pluck a guitar string, the resulting sound has a steady tone and lasts a long time. Each successive vibration of the string takes the same time as the previous one. We define periodic motion to be a motion that repeats itself at regular time intervals, such as exhibited by the guitar string or by an object on a spring moving up and down. The time to complete one oscillation remains constant and is called the period T . Its units are usually seconds, but may be any convenient unit of time. The word period refers to the time for some event whether repetitive or not; but we shall be primarily interested in periodic motion, which is by definition repetitive. A concept closely related to period is the frequency of an event. For example, if you get a paycheck twice a month, the frequency of payment is two per month and the period between checks is half a month. Frequency f is defined to be the number of events per unit time. For periodic motion, frequency is the number of oscillations per unit time. The relationship between frequency and period is.</code>
  <code>What do stored fats provide our body with for later use?</code>	<code>Fats are one type of lipid. Stored fat gives your body energy to use for later. It’s like having money in a savings account: it’s there in case you need it. Stored fat also cushions and protects internal organs. In addition, it insulates the body. It helps keep you warm in cold weather.</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: 4 tokens</li><li>mean: 11.4 tokens</li><li>max: 22 tokens</li></ul>	<ul><li>min: 16 tokens</li><li>mean: 34.47 tokens</li><li>max: 67 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>What do geologists use to determine the age of rock layers?</code>	<code>radioactive dating is used to determine the age of fossils. Geologists use fossils to determine the age of the rock layer that it was found in.. Geologists use radioactive dating to determine the age of rock layers.</code>
  <code>Air pollution can cause a decrease in what?</code>	<code>air pollution can cause the pH of soil to decrease. Research has shown that soil pH less than 6.0 can decrease crop yields.. Air pollution can cause a decrease in crop yields. </code>
  <code>What organism uses their sense of smell to find a mate?</code>	<code>Most salamanders use their sense of smell to find a mate.. Salamanders are a type of amphibian.. Some amphibians use their sense of smell to find a mate.</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: 30,250 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: 8.65 tokens</li><li>max: 31 tokens</li></ul>	<ul><li>min: 17 tokens</li><li>mean: 76.38 tokens</li><li>max: 207 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>propria definition</code>	<code>The lamina propria is a constituent of the moist linings known as mucous membranes or mucosa, which line various tubes in the body.The lamina propria is a thin layer of loose connective tissue which lies beneath the epithelium and together with the epithelium constitutes the mucosa.As its Latin name indicates it is a characteristic component of the mucosa, the mucosa's own special layer.he lamina propria is a constituent of the moist linings known as mucous membranes or mucosa, which line various tubes in the body.</code>
  <code>what family is the common bullfrog in</code>	<code>The American bullfrog (Rana catesbeiana), often simply known as the bullfrog in Canada and the United States, is an aquatic frog, a member of the family Ranidae, or “true frogs”. This frog has an olive green back and sides blotched with brownish markings and a whitish belly spotted with yellow or grey.</code>
  <code>why does well pump cycle on and off</code>	<code>1 When a well pump turns on every time a water-using fixture’s valve is opened, or if it turns on and off rapidly while the fixture is in use, that’s short cycling. It is typically caused by a water pressure tank that has lost it built-in cushion of pressurized air.</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: 30,250 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: 11.81 tokens</li><li>max: 22 tokens</li></ul>	<ul><li>min: 16 tokens</li><li>mean: 135.28 tokens</li><li>max: 512 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>who established the republic of china in 1912 ce</code>	<code>Republic of China (1912–1949) The Republic of China was a sovereign state in East Asia, that occupied the territories of modern China and for part of its history Mongolia and Taiwan. It was founded in 1912, after the Qing dynasty, the last imperial dynasty, was overthrown in the Xinhai Revolution. The Republic's first president, Sun Yat-sen, served only briefly before handing over the position to Yuan Shikai, former leader of the Beiyang Army. His party, then led by Song Jiaoren, won the parliamentary election held in December 1912. Song was assassinated shortly after, and the Beiyang Army led by Yuan Shikai maintained full control of the government in Beijing. Between late 1915 and early 1916, Yuan tried to reinstate the monarchy, before resigning after popular unrest. After Yuan's death in 1916, members of cliques in the former Beiyang Army claimed their autonomy and clashed with each other. During this period, the authority of the republican government was weakened by a restoration of the Qing government.</code>
  <code>what is the message in john lennon's imagine</code>	<code>Imagine (John Lennon song) "Imagine" is a song written and performed by English musician John Lennon. The best-selling single of his solo career, its lyrics encourage the listener to imagine a world at peace without the barriers of borders or the divisions of religion and nationality, and to consider the possibility that the whole of humanity would live unattached to material possessions.</code>
  <code>who is the original singer for these boots are made for walking</code>	<code>These Boots Are Made for Walkin' "These Boots Are Made for Walkin'" is a hit song written by Lee Hazlewood and recorded by Nancy Sinatra. It charted January 22, 1966,[3] and reached No. 1 in the United States Billboard Hot 100 and in the UK Singles Chart.[2]</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: 12,024 training samples
• Columns: <code>query</code> and <code>answer</code>
• Approximate statistics based on the first 1000 samples:

  	query	answer
  type	string	string
  details	<ul><li>min: 7 tokens</li><li>mean: 17.49 tokens</li><li>max: 61 tokens</li></ul>	<ul><li>min: 12 tokens</li><li>mean: 200.72 tokens</li><li>max: 482 tokens</li></ul>

• Samples:

  query	answer
  <code>What is the name of 'Bob the Builder's' cement mixer?</code>	<code>BOB the BUILDER DIZZY the CEMENT MIXER BOB the BUILDER DIZZY THE CEMENT MIXER Large Friction Action Dizzy the Cement Mixer and clicky noise turn by hand Mixer from the Large Bob The Builder Friction toy Collection  Measures 5" x  4" High  in EXCELLENT Played with Condition �8.00 .. US $15.60  .. EUROS 12.00 Ref # 0845</code>
  <code>Which cartoon character, who cooks at the Krusty Krab, lives in a pineapple under the sea?</code>	<code>SpongeBob SquarePants from SpongeBob SquarePants
  <code>Chinese Democracy was the long awaited 2008 album release of which band?</code>	<code>Guns N Roses 'Chinese Democracy' Set For Wal-Mart Release

• 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: 30,250 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: 11.37 tokens</li><li>max: 24 tokens</li></ul>	<ul><li>min: 13 tokens</li><li>mean: 56.7 tokens</li><li>max: 154 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>do namibians need a visa for london?</code>	<code>The Embassy is closed every Friday. Namibia tourist visa is not required for citizens of United Kingdom for a stay up to 90 days. Sounds good!</code>
  <code>what is the meaning of personal finance?</code>	<code>Personal finance is a term that covers managing your money as well as saving and investing. ... It often refers to the entire industry that provides financial services to individuals and households and advises them about financial and investment opportunities.</code>
  <code>what are trim tabs for?</code>	<code>Trim tabs are either flat plates or vertical blades fitted either side of the boat and attached to the transom. They are used to trim or level the boat, both fore and aft and side to side. Trim tabs are controlled up and downwards by either a hydraulic, or electric ram system on the transom.</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: 25.63 tokens</li><li>max: 68 tokens</li></ul>	<ul><li>min: 10 tokens</li><li>mean: 25.55 tokens</li><li>max: 68 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>Thompson was born in Hendon , North London , Dena Holmes in 1960 and worked for a building society bank .</code>	<code>Thompson was born Dena Holmes in 1960 in Hendon , North London . She worked for a building society .</code>
  <code>One of his first cousins was Elizabeth , aka Lady Elizabeth Hervey , aka Bess Foster , Duchess of Devonshire . His younger brother married Lord Bishop Foster .</code>	<code>One of his first cousins was Elizabeth , alias Lady Elizabeth Hervey , alias Bess Foster , Duchess of Devonshire , his younger brother married Lord Bishop Foster .</code>
  <code>At the executive level , EEAA represents the central arm of the Ministry .</code>	<code>At executive level , EEAA represents the central arm of the ministry .</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: 11 tokens</li><li>mean: 13.64 tokens</li><li>max: 19 tokens</li></ul>	<ul><li>min: 9 tokens</li><li>mean: 13.23 tokens</li><li>max: 21 tokens</li></ul>	<ul><li>min: 10 tokens</li><li>mean: 13.52 tokens</li><li>max: 22 tokens</li></ul>

• Samples:

  anchor	entailment	negative
  <code>Set of toy animals sitting in front of a red wooden wagon.</code>	<code>Several toy animals - a bull, giraffe, deer and parakeet.</code>	<code>Several toy animals - a cow, lion, wolf and canary.</code>
  <code>A bathroom with a toilette with it's seat down.</code>	<code>A bathroom with a sink and a toilet</code>	<code>A bathroom without a sink or a toilet</code>
  <code>A striped plane flying up into the sky as the sun shines behind it.</code>	<code>An airplane is ascending into the white sky</code>	<code>An airplane is descending into the black sky</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}
  

scitail-pairs-qa

• Dataset: scitail-pairs-qa at 0cc4353
• Size: 128 evaluation samples
• Columns: <code>sentence2</code> and <code>question</code>
• Approximate statistics based on the first 1000 samples:

  	sentence2	question
  type	string	string
  details	<ul><li>min: 7 tokens</li><li>mean: 15.31 tokens</li><li>max: 33 tokens</li></ul>	<ul><li>min: 7 tokens</li><li>mean: 14.91 tokens</li><li>max: 26 tokens</li></ul>

• Samples:

  sentence2	question
  <code>It takes earth one week to rotate on its axis seven times.</code>	<code>How long does it take for Earth to rotate on its axis seven times?</code>
  <code>Both hurricanes and tornadoes always have high winds.</code>	<code>Both hurricanes and tornadoes always</code>
  <code>Seeds of a pine cone are easily carried by the wind and dispersed because seeds have wings.</code>	<code>Why are seeds of a pine cone easily carried by the wind and dispersed?</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: 63 tokens</li><li>mean: 214.16 tokens</li><li>max: 341 tokens</li></ul>	<ul><li>min: 15 tokens</li><li>mean: 25.86 tokens</li><li>max: 43 tokens</li></ul>

• Samples:

  document	summary
  <code>The rescue on Monday afternoon involved a challenging winch down to the casualty by a coastguard helicopter's paramedic.<br>The Inverness Airport-based helicopter could not be brought too close in case its downdraft blew the man and his fellow walkers off the ridge.<br>The walker was taken to hospital for treatment.<br>The alarm was raised at 14:25 on Monday after the man became unwell on Carn Mor Dearg.<br>He was airlifted to Torlundy and from there was taken to hospital by ambulance.<br>Scott Sharman, paramedic winchman, said: "It was an extremely steep ridge and we needed to make sure we kept at a safe distance because the downdraft could very easily have blown them over the ridge."</code>	<code>A walker had to be rescued from a 1,000m (3,500ft) ridge near Ben Nevis after he became ill.</code>
  <code>The 23-year-old, who did not make a senior appearances during a year with the Hornets, has signed a two-year deal with the O's.<br>Woods began his career in Cambridge United's youth set-up and subsequently joined Manchester United's academy.<br>He moved to Doncaster Rovers in 2009 and went on to make 82 appearances, before leaving the club last summer.<br>Woods, a former England Under-19 international, will provide competition for fellow new arrival Adam Legzdins.</code>	<code>League One side Leyton Orient have signed goalkeeper Gary Woods following his departure from Watford.</code>
  <code>The 32-year-old, the number one ranked Test bowler, has been out of action with a groin strain since the first Test against India in early November.<br>But in Cape Town this week he was able to complete two bowling spells at match intensity in the nets, run sprints and do agility tests and fielding drills.<br>The first of the four-match Test series begins in Durban on 26 December.<br>Proteas team manager, Dr Mohammed Moosajee said: "Dale has put a lot of work into his training and rehab since his return from India so we are very happy to have him back from injury.<br>"He was put through a thorough fitness test on Thursday by physiotherapist, Shane Jabaar, he came through the tests without any discomfort and will be available for selection for the first Test match against England."<br>Steyn has taken 402 wickets in 81 Tests at an average of 22.56, including 46 in 11 matches against England.<br>South Africa remain at the top of the Test rankings despite a 3-0 defeat in India this month.</code>	<code>South Africa paceman Dale Steyn will be fit for the Test series with England after passing a fitness test.</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: 16.72 tokens</li><li>max: 36 tokens</li></ul>	<ul><li>min: 2 tokens</li><li>mean: 73.28 tokens</li><li>max: 332 tokens</li></ul>

• Samples:

  sentence1	sentence2
  <code>The skull is a part of a vertebrate endoskeleton that encloses and protects what organ?</code>	<code>part of a vertebrate endoskeleton that encloses and protects the brain; also called the skull.</code>
  <code>The action of sunlight on the skin triggers the body to produce what vitamin?</code>	<code>Milk and other dairy foods are not the only sources of calcium. This important nutrient is also found in green leafy vegetables, broccoli, and intact salmon and canned sardines with their soft bones. Nuts, beans, seeds, and shellfish provide calcium in smaller quantities. Except for fatty fish like salmon and tuna, or fortified milk or cereal, vitamin D is not found naturally in many foods. The action of sunlight on the skin triggers the body to produce its own vitamin D (Figure 6.22), but many people, especially those of darker complexion and those living in northern latitudes where the sun’s rays are not as strong, are deficient in vitamin D. In cases of deficiency, a doctor can prescribe a vitamin D supplement.</code>
  <code>What phenomenon is essential in order for evolution to occur because it increases genetic variation and the potential for individuals to differ?</code>	<code>Mutations are essential for evolution to occur because they increase genetic variation and the potential for individuals to differ. The majority of mutations are neutral in their effects on the organisms in which they occur. Beneficial mutations may become more common through natural selection. Harmful mutations may cause genetic disorders or cancer.</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: 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: 4 tokens</li><li>mean: 11.34 tokens</li><li>max: 25 tokens</li></ul>	<ul><li>min: 18 tokens</li><li>mean: 35.58 tokens</li><li>max: 66 tokens</li></ul>

• Samples:

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