SentenceTransformer based on bobox/DeBERTa-ST-AllLayers-v3-checkpoints-tmp
This is a sentence-transformers model finetuned from bobox/DeBERTa-ST-AllLayers-v3-checkpoints-tmp on the nli-pairs, vitaminc-pairs, qnli-contrastive, scitail-pairs-qa, scitail-pairs-pos, xsum-pairs, compression-pairs, compression-pairs2, compression-pairs3, sciq_pairs, qasc_pairs, qasc_facts_sym, openbookqa_pairs, msmarco_pairs, msmarco_pairs2, nq_pairs, nq_pairs2, trivia_pairs, quora_pairs, gooaq_pairs, gooaq_pairs2 and mrpc_pairs 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-ST-AllLayers-v3-checkpoints-tmp <!-- at revision 4859fef4e21d101c2d445bcd33db1e3308f35dc4 -->
- Maximum Sequence Length: 512 tokens
- Output Dimensionality: 768 tokens
- Similarity Function: Cosine Similarity
- Training Datasets:
- Language: en
<!-- - **License:** Unknown -->
Model Sources
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-ST-AllLayers-v3.1bis")
# Run inference
sentences = [
'what are light bulbs powered by?',
'an incandescent light bulb converts electricity into light by sending electricity through a filament. Incandescence is light from heat energy.. light bulbs are powered by heat',
'Viruses infect and live inside the cells of living organisms.. Life is a living organism.. viruses infect and live inside the cells of life.',
]
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
| Metric | Value |
|---|
| pearson_cosine | 0.8793 |
| spearman_cosine | 0.8948 |
| pearson_manhattan | 0.856 |
| spearman_manhattan | 0.8598 |
| pearson_euclidean | 0.8584 |
| spearman_euclidean | 0.8609 |
| pearson_dot | 0.7762 |
| spearman_dot | 0.7778 |
| pearson_max | 0.8793 |
| spearman_max | 0.8948 |
Binary Classification
| Metric | Value |
|---|
| cosine_accuracy | 0.7316 |
| cosine_accuracy_threshold | 0.7115 |
| cosine_f1 | 0.8214 |
| cosine_f1_threshold | 0.616 |
| cosine_precision | 0.7009 |
| cosine_recall | 0.9919 |
| cosine_ap | 0.8423 |
| dot_accuracy | 0.6974 |
| dot_accuracy_threshold | 6.443 |
| dot_f1 | 0.8048 |
| dot_f1_threshold | 6.3601 |
| dot_precision | 0.695 |
| dot_recall | 0.9556 |
| dot_ap | 0.7569 |
| manhattan_accuracy | 0.7342 |
| manhattan_accuracy_threshold | 51.9824 |
| manhattan_f1 | 0.8179 |
| manhattan_f1_threshold | 53.5271 |
| manhattan_precision | 0.734 |
| manhattan_recall | 0.9234 |
| manhattan_ap | 0.8236 |
| euclidean_accuracy | 0.7316 |
| euclidean_accuracy_threshold | 2.6294 |
| euclidean_f1 | 0.8169 |
| euclidean_f1_threshold | 2.7942 |
| euclidean_precision | 0.725 |
| euclidean_recall | 0.9355 |
| euclidean_ap | 0.8252 |
| max_accuracy | 0.7342 |
| max_accuracy_threshold | 51.9824 |
| max_f1 | 0.8214 |
| max_f1_threshold | 53.5271 |
| max_precision | 0.734 |
| max_recall | 0.9919 |
| max_ap | 0.8423 |
Binary Classification
| Metric | Value |
|---|
| cosine_accuracy | 0.5737 |
| cosine_accuracy_threshold | 0.7548 |
| cosine_f1 | 0.6717 |
| cosine_f1_threshold | 0.4373 |
| cosine_precision | 0.5175 |
| cosine_recall | 0.9568 |
| cosine_ap | 0.5583 |
| dot_accuracy | 0.5553 |
| dot_accuracy_threshold | 9.6254 |
| dot_f1 | 0.6704 |
| dot_f1_threshold | 6.3724 |
| dot_precision | 0.5114 |
| dot_recall | 0.973 |
| dot_ap | 0.5546 |
| manhattan_accuracy | 0.5737 |
| manhattan_accuracy_threshold | 57.3429 |
| manhattan_f1 | 0.6667 |
| manhattan_f1_threshold | 100.9737 |
| manhattan_precision | 0.5014 |
| manhattan_recall | 0.9946 |
| manhattan_ap | 0.5523 |
| euclidean_accuracy | 0.5789 |
| euclidean_accuracy_threshold | 2.8532 |
| euclidean_f1 | 0.6655 |
| euclidean_f1_threshold | 5.0141 |
| euclidean_precision | 0.5 |
| euclidean_recall | 0.9946 |
| euclidean_ap | 0.5536 |
| max_accuracy | 0.5789 |
| max_accuracy_threshold | 57.3429 |
| max_f1 | 0.6717 |
| max_f1_threshold | 100.9737 |
| max_precision | 0.5175 |
| max_recall | 0.9946 |
| max_ap | 0.5583 |
<!--
## 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
nli-pairs
- Dataset: nli-pairs at d482672
- Size: 10,000 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: 5 tokens</li><li>mean: 16.62 tokens</li><li>max: 62 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 9.46 tokens</li><li>max: 29 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>A person on a horse jumps over a broken down airplane.</code> | <code>A person is outdoors, on a horse.</code> |
| <code>Children smiling and waving at camera</code> | <code>There are children present</code> |
| <code>A boy is jumping on skateboard in the middle of a red bridge.</code> | <code>The boy does a skateboarding trick.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
vitaminc-pairs
- Dataset: vitaminc-pairs at be6febb
- Size: 11,500 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: 8 tokens</li><li>mean: 17.5 tokens</li><li>max: 56 tokens</li></ul> | <ul><li>min: 8 tokens</li><li>mean: 38.51 tokens</li><li>max: 381 tokens</li></ul> |
- Samples:
| claim | evidence |
|---|
| <code>Danny Trevathan made more than 100 tackles in the 2013 season .</code> | <code>In the 2013 season , Trevathan led the team with 125 tackles .</code> |
| <code>Katy Perry : Part of Me has grossed over $ 6.7 million internationally and more than $ 32 million globally .</code> | <code>Katy Perry : Part of Me has grossed $ 25,326,071 domestically and $ 6,882,884 internationally , totaling $ 32,208,955 worldwide .</code> |
| <code>Bark at the Moon became Platinum for selling more than 700,000 copies in the United States .</code> | <code>The album peaked at number 19 on the Billboard album chart and within several weeks of release was certified Platinum for over a million sales in the United States alone .</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
qnli-contrastive
- Dataset: qnli-contrastive at bcdcba7
- Size: 13,300 training samples
- Columns: <code>sentence1</code>, <code>sentence2</code>, and <code>label</code>
- Approximate statistics based on the first 1000 samples:
| sentence1 | sentence2 | label |
|---|
| type | string | string | int |
| details | <ul><li>min: 6 tokens</li><li>mean: 13.78 tokens</li><li>max: 34 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 35.59 tokens</li><li>max: 178 tokens</li></ul> | <ul><li>0: 100.00%</li></ul> |
- Samples:
| sentence1 | sentence2 | label |
|---|
| <code>How does the Center for Measuring University performance rank Washington University?</code> | <code>According to the Center for Measuring University Performance, it is considered to be one of the top 10 private research universities in the nation.</code> | <code>0</code> |
| <code>Besides tax shifting, what is another need?</code> | <code>Just as there is a need for tax shifting, there is also a need for subsidy shifting.</code> | <code>0</code> |
| <code>How large is the Irish diaspora that was caused by the Great Irish Famine?</code> | <code>However, since the Great Irish Famine, the population of Ireland has fallen to less than one tenth of the population of the British Isles.</code> | <code>0</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "OnlineContrastiveLoss",
"n_layers_per_step": -1,
"last_layer_weight": 0.75,
"prior_layers_weight": 1.25,
"kl_div_weight": 0.8,
"kl_temperature": 0.75
}
scitail-pairs-qa
- Dataset: scitail-pairs-qa at 0cc4353
- Size: 11,155 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: 15.95 tokens</li><li>max: 41 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 15.04 tokens</li><li>max: 41 tokens</li></ul> |
- Samples:
| sentence2 | sentence1 |
|---|
| <code>The first living cells may have evolved around 4 billion years ago.</code> | <code>The first living cells may have evolved around how long ago?</code> |
| <code>By 8 weeks, all major organs start developing.</code> | <code>By how many weeks do all major organs start developing?</code> |
| <code>Intrinsic muscles allow your fingers to also make precise movements for actions.</code> | <code>Which muscles allow your fingers to also make precise movements for actions?</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
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: 23.56 tokens</li><li>max: 62 tokens</li></ul> | <ul><li>min: 7 tokens</li><li>mean: 15.39 tokens</li><li>max: 35 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>Retina The retina uses nerve cells known as rods and cones (photoreceptors) to detect light and color.</code> | <code>The light-sensing cells in the retina are called rods and cones.</code> |
| <code>Breakwaters are structures that protect the coast like barrier islands.</code> | <code>The name of artificial barriers that people build to protect shorelines is breakwaters.</code> |
| <code>The speed of the wave is equal to the wavelength times the frequency.</code> | <code>The product of a wave's wavelength and its frequency is its speed.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
xsum-pairs
- Dataset: xsum-pairs at 788ddaf
- Size: 7,000 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: 7 tokens</li><li>mean: 190.44 tokens</li><li>max: 512 tokens</li></ul> | <ul><li>min: 8 tokens</li><li>mean: 25.62 tokens</li><li>max: 75 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>The Surrey force said a 51-year-old woman from East Molesey voluntarily spoke to officers.<br>The incident at Hampton Court was filmed by cycling instructor David Williams, 47, who posted it on YouTube.<br>Officers said its traffic and process unit would consider whether to take further action.<br>The woman was given a referral notice of "driving whilst not in proper control of motor vehicle".<br>The police force said it would consider whether the woman should face penalties, including being required to attend a driver improvement course or receive a fixed penalty notice.<br>The alleged incident happened at the junction of Creek Road and the A309 Hampton Court Way in Molesey.</code> | <code>A woman who was filmed allegedly eating a bowl of cereal while driving at a busy junction in south-west London has been interviewed by police.</code> |
| <code>Paul Hegarty was sacked last month, with the Gable Endies picking up one point from three games since his departure.<br>Petrie, 46, leaves his post as as assistant manager at Junior Super League side Broughty Athletic.<br>Born in Dundee, he played for Forfar, Dunfermline and Ross County and had spells in Australia and Singapore.<br>Petrie has previously served as assistant manager at Forfar and Arbroath.<br>"He is highly motivated and enthusiastic, comes with a wealth of playing and coaching experience and most importantly understands the lower leagues of Scottish football," said chairman John Crawford.<br>"I would like to thank John Holt for his support in stepping in and looking after the team since the departure of Paul Hegarty and wish him every success for the future."</code> | <code>Stewart Petrie is the new manager at struggling Montrose, taking over with the club at the foot of League Two.</code> |
| <code>The 28-year-old former Bournemouth, Portsmouth and Leeds United centre-back has signed a three-year contract with the Addicks.<br>Pearce made 33 appearances last season as the Latics won the League One title and becomes Charlton's seventh signing of the summer transfer window.<br>"He's an absolute warrior," boss Russell Slade told the club website.<br>Find all the latest football transfers on our dedicated page.</code> | <code>League One club Charlton Athletic have signed defender Jason Pearce from Wigan for an undisclosed fee.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "MultipleNegativesSymmetricRankingLoss",
"n_layers_per_step": -1,
"last_layer_weight": 0.75,
"prior_layers_weight": 1.25,
"kl_div_weight": 0.8,
"kl_temperature": 0.75
}
compression-pairs
- Dataset: compression-pairs at 605bc91
- Size: 4,014 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: 11 tokens</li><li>mean: 31.5 tokens</li><li>max: 125 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 10.17 tokens</li><li>max: 28 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>one who survived one of the city's worst mass shootings but left behind a trail of death and violence in British Columbia and Alberta going back almost a decade.</code> | <code>Victim had survived mass shooting</code> |
| <code>``It's an awesome feeling,'' Vick said of getting the bankruptcy plan approved, which includes how his creditors will be repaid more than $20 million.</code> | <code>Michael Vick's bankruptcy plan approved</code> |
| <code>Kevin Youkilis put on pinstripes and played at Yankee Stadium for the first time as a member of the New York Yankees on Monday afternoon.</code> | <code>Youkilis downplays putting on pinstripes at Yankee Stadium</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "MultipleNegativesSymmetricRankingLoss",
"n_layers_per_step": -1,
"last_layer_weight": 0.75,
"prior_layers_weight": 1.25,
"kl_div_weight": 0.8,
"kl_temperature": 0.75
}
compression-pairs2
- Dataset: compression-pairs2 at 605bc91
- Size: 7,960 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: 11 tokens</li><li>mean: 31.9 tokens</li><li>max: 329 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 10.05 tokens</li><li>max: 22 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>Industrial Parks in Asia are seeing very strong demand as Foreign Companies take advantage of the many investment incentives on offer.</code> | <code>Industrial Parks in Asia see strong demand</code> |
| <code>Western Australia has the highest loan delinquency rates in the country with two percent of its home loan payments delayed by more than a month compared to the national average of 1.54 percent says Fitch Ratings.</code> | <code>Western Australia has highest delinquency rate in the country</code> |
| <code>Governor Chet Culver today issued a disaster emergency proclamation for Des Moines County in response to severe thunderstorms that hit the area on May 12.</code> | <code>Governor Culver issues disaster proclamation for Des Moines County</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": 3,
"last_layer_weight": 0.15,
"prior_layers_weight": 2.5,
"kl_div_weight": 0.75,
"kl_temperature": 0.5
}
compression-pairs3
- Dataset: compression-pairs3 at 605bc91
- Size: 7,960 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: 32.24 tokens</li><li>max: 166 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 10.08 tokens</li><li>max: 22 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>WHILE London 2012 might seem a bit farfetched for a 17-year-old, water polo player Tuesday Birmingham is hoping her undefeated streak makes selectors sit up and take notice.</code> | <code>Birmingham hoping selectors sit up and take notice</code> |
| <code>Police are searching for a man accused of trying to kidnap a girl from her North Charlotte bus stop.</code> | <code>Police search for man accused of trying to kidnap girl at bus stop</code> |
| <code>NASCAR has ``called a mandatory meeting for all Sprint Cup drivers and team owners for Tuesday morning'' at its Research and Development Center in Concord, North Carolina, according to David Newton of ESPN.com.</code> | <code>NASCAR calls mandatory meeting for Sprint Cup drivers, owners</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "MultipleNegativesRankingLoss",
"n_layers_per_step": 3,
"last_layer_weight": 0.05,
"prior_layers_weight": 10,
"kl_div_weight": 5,
"kl_temperature": 0.2
}
sciq_pairs
- Dataset: sciq_pairs at 2c94ad3
- Size: 10,750 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: 6 tokens</li><li>mean: 16.74 tokens</li><li>max: 43 tokens</li></ul> | <ul><li>min: 2 tokens</li><li>mean: 79.87 tokens</li><li>max: 512 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>What do secondary spermatocytes form when completing meiosis?</code> | <code>Spermatogonia lining the seminiferous tubule undergo mitosis to form primary spermatocytes , which are also diploid. The primary spermatocytes undergo the first meiotic division to form secondary spermatocytes , which are haploid. Spermatocytes make up the next layer of cells inside the seminiferous tubule. Finally, the secondary spermatocytes complete meiosis to form spermatids . Spermatids make up a third layer of cells in the tubule.</code> |
| <code>Unlike ammonia, oxygen cannot be liquefied at room temperature because its what is below room temperature?</code> | <code>Check Your Learning Ammonia can be liquefied by compression at room temperature; oxygen cannot be liquefied under these conditions. Why do the two gases exhibit different behavior? Answer: The critical temperature of ammonia is 405.5 K, which is higher than room temperature. The critical temperature of oxygen is below room temperature; thus oxygen cannot be liquefied at room temperature.</code> |
| <code>The distance between two consecutive z discs or z lines is called what?</code> | <code>When a sarcomere shortens, some regions shorten whereas others stay the same length. A sarcomere is defined as the distance between two consecutive Z discs or Z lines; when a muscle contracts, the distance between the Z discs is reduced. The H zone—the central region of the A zone—contains only thick filaments and is shortened during contraction. The I band contains only thin filaments and also shortens. The A band does not shorten—it remains the same length—but A bands of different sarcomeres move closer together during contraction, eventually disappearing. Thin filaments are pulled by the thick filaments toward the center of the sarcomere until the Z discs approach the thick filaments. The zone of overlap, in which thin filaments and thick filaments occupy the same area, increases as the thin filaments move inward.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
qasc_pairs
- Dataset: qasc_pairs at a34ba20
- Size: 7,889 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: 11.41 tokens</li><li>max: 29 tokens</li></ul> | <ul><li>min: 16 tokens</li><li>mean: 34.81 tokens</li><li>max: 75 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>What is a type of the basic units of the structure and function of living things?</code> | <code>Cells are the basic units of the structure and function of living things.. Osteogenic cell types are indicated.. Osteogenic is a type of the basic units of the structure and function of living things.</code> |
| <code>bacteria can cause people to need what?</code> | <code>bacteria can cause people to become ill. Serious illness needing hospitalization is uncommon.. bacteria can cause people to need hospitalization.</code> |
| <code>What type of energy generated by the bulbs is wasted?</code> | <code>some light bulbs convert electricity into light and heat energy. Since the purpose of a light bulb is to generate light, the heat is wasted energy.. The heat energy generated by the bulbs gets wasted.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
qasc_facts_sym
- Dataset: qasc_facts_sym at a34ba20
- Size: 7,889 training samples
- Columns: <code>combinedfact</code> and <code>facts</code>
- Approximate statistics based on the first 1000 samples:
| combinedfact | facts |
|---|
| type | string | string |
| details | <ul><li>min: 5 tokens</li><li>mean: 11.82 tokens</li><li>max: 26 tokens</li></ul> | <ul><li>min: 12 tokens</li><li>mean: 25.18 tokens</li><li>max: 45 tokens</li></ul> |
- Samples:
| combinedfact | facts |
|---|
| <code>weathering means that rocks are eroded down again.</code> | <code>weathering means breaking down rocks from larger whole into smaller pieces by weather. Erosion breaks the rock down again, and the cycle continues..</code> |
| <code>cutting down trees has a negative impact on the health of humankind</code> | <code>cutting down trees has a negative impact on an ecosystem. When ecosystems are healthy, humankind is healthy..</code> |
| <code>Fossils are formed with sand and mud cover the remains over time</code> | <code>fossils are formed when layers of sediment cover the remains of organisms over time. Sand and mud are examples of sediments..</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "MultipleNegativesSymmetricRankingLoss",
"n_layers_per_step": -1,
"last_layer_weight": 0.75,
"prior_layers_weight": 1.25,
"kl_div_weight": 0.8,
"kl_temperature": 0.75
}
openbookqa_pairs
- Dataset: openbookqa_pairs
- Size: 4,505 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: 13.81 tokens</li><li>max: 78 tokens</li></ul> | <ul><li>min: 4 tokens</li><li>mean: 11.49 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>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
msmarco_pairs
- Dataset: msmarco_pairs at 28ff31e
- Size: 6,875 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: 8.68 tokens</li><li>max: 32 tokens</li></ul> | <ul><li>min: 17 tokens</li><li>mean: 77.4 tokens</li><li>max: 232 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>benefits of reishi mushroom</code> | <code>Reishi mushroom has been used to help enhance the immune system, reduce stress, improve sleep, and lessen fatigue. People also take reishi mushroom for health conditions such as: High blood pressure</code> |
| <code>shoulder replacement how long driving</code> | <code>Shoulder Replacement Surger. After undergoing shoulder replacement surgery, it is important to have realistic expectations about the types of activities you may do. Minimal stretching of your new shoulder is recommended throughout your lifetime to maintain your full range of motion.You may give you permission to drive within four weeks after the surgery. If your surgery was on the right side, driving permission may not be. given until a month or six weeks following the surgery.inimal stretching of your new shoulder is recommended throughout your lifetime to maintain your full range of motion. You may give you permission to drive within four weeks after the surgery. If your surgery was on the right side, driving permission may not be. given until a month or six weeks following the surgery.</code> |
| <code>cabinet has how many people in it trump</code> | <code>Should all of President-elect Donald Trumpâs Cabinet nominees eventually be confirmed, he will start his administration with one of the most heavily business-oriented Cabinets in U.S. history. Five of the 15 people Trump has nominated to be Cabinet secretaries have spent all or nearly all their careers in the business world, with no significant public office or senior military service on their résumés.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
msmarco_pairs2
- Dataset: msmarco_pairs2 at 28ff31e
- Size: 4,500 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: 24 tokens</li></ul> | <ul><li>min: 19 tokens</li><li>mean: 76.9 tokens</li><li>max: 229 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>when is grant of probate not required</code> | <code>We are frequently asked whether a grant of probate is always required after somebodyâs death. Very small estates under £15,000 do not require a grant while those assets held in joint names (such as bank accounts or property) pass to the survivor on the Deceased persons death, meaning that no grant is required.he grant is a legal document which states the Deceasedâs full name, as well as the Executors entrusted with the administration of the estate. The grant can then be presented to various financial institutions to collect the Deceasedâs assets.</code> |
| <code>what does speculation mean</code> | <code>Wiktionary(3.00 / 1 vote)Rate this definition: 1 speculation(Noun) The process of thinking or meditating on a subject. 2 speculation(Noun) A judgment or conclusion reached by speculating. 3 speculation(Noun) An investment involving higher than normal risk in order to obtain a higher than normal return.</code> |
| <code>what county is mint hill nc</code> | <code>8. 21 Acres Mint Hill, Mecklenburg County, North Carolina $2,310,000. 21 Acres Mint Hill, Mecklenburg County, North Carolina $2,310,000. Great piece of property just waiting for the right buyer. Flat and open and conveniently located at 485 and Idlewild Rd....a rare find indeed!</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": 3,
"last_layer_weight": 0.15,
"prior_layers_weight": 2.5,
"kl_div_weight": 0.75,
"kl_temperature": 0.5
}
nq_pairs
- Dataset: nq_pairs at f9e894e
- Size: 3,667 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: 11.85 tokens</li><li>max: 22 tokens</li></ul> | <ul><li>min: 19 tokens</li><li>mean: 132.66 tokens</li><li>max: 512 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>where are the powers of congress listed in the constitution</code> | <code>Powers of the United States Congress Article I of the Constitution sets forth most of the powers of Congress, which include numerous explicit powers enumerated in Section 8. Constitutional amendments have granted Congress additional powers. Congress also has implied powers derived from the Necessary and Proper Clause of the Constitution.</code> |
| <code>who is the man in we are never ever getting back together</code> | <code>We Are Never Ever Getting Back Together After writing Speak Now (2010) entirely solo, Swift opted to collaborate with different songwriters and producers for Red. Thus, she called Max Martin and Shellback, two songwriters and producers whose work she admired, to discuss a possible collaboration. The trio conceived the concept for "We Are Never Ever Getting Back Together" shortly after a friend of Swift's ex-boyfriend walked into the recording studio and spoke of rumors he heard that Swift and her former flame were reuniting. After the friend left, Martin and Shellback asked Swift to elaborate on the details of the relationship, which she described as "break up, get back together, break up, get back together, just, ugh, the worst". When Martin suggested that they write about the incident. Swift began playing the guitar and singing, "We are never ever......", and the song flowed rapidly afterwards. She described the process as one of the most humorous experiences she had while recording, and said the musical partners matched her expectations. An audio clip of her sarcastically speaking about breakups can be heard before the final chorus.[2] The song is reportedly about Swift's ex, Jake Gyllenhaal, as the two had broken up in January 2011 but had been seen on a date a few days later.[3] After the release of the music video, more clues linking the song to Gyllenhaal emerged,[3] with the actor looking like Gyllenhaal,[4] the actor in the video giving her a scarf as Gyllenhaal had reportedly done for Swift and a bracelet Swift wears in the video that is speculated to look similar to that of which Gyllenhaal was rumored to have given Swift for her birthday.[3]</code> |
| <code>which agreement settled a boundary dispute between british canada and the united states</code> | <code>Alaska boundary dispute Finally, in 1903, the Hay-Herbert Treaty between the United States and Britain entrusted the decision to an arbitration by a mixed tribunal of six members: three Americans (Elihu Root, Secretary of War; Henry Cabot Lodge, senator from Massachusetts; and George Turner, ex-senator from Washington), two Canadians (Sir Louis A. Jette, Lieutenant Governor of Quebec; and Allen B. Aylesworth, K.C., from Toronto), and one Briton (Baron Alverstone). All sides respected Root, but he was a member of the U.S. Cabinet. Canadians ridiculed the choice of the obscure ex-Senator Turner and, especially, Lodge, a leading historian and diplomatic specialist whom they saw as an unobjective.[8]</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": -1,
"last_layer_weight": 1.75,
"prior_layers_weight": 0.5,
"kl_div_weight": 1.25,
"kl_temperature": 0.9
}
nq_pairs2
- Dataset: nq_pairs2 at f9e894e
- Size: 3,001 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: 11.8 tokens</li><li>max: 23 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 132.85 tokens</li><li>max: 512 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>what do i need to carry a gun in indiana</code> | <code>Gun laws in Indiana Indiana requires a license for carrying a handgun generally, although there are several exceptions. Manner of carry, whether open or concealed, is not explicitly specified in the code, thus the license is required for concealed, open, or transport within vehicle, unless covered by one of the exceptions. A license is not required if carrying on property "owned, rented, leased, or otherwise legally controlled" by the carrier, or carried on another person's legally controlled property if permission has been obtained. A license is not required for carrying at a "gun show, firearm expo, gun owner's club or convention, hunting club, shooting club, or training course", or places where carrier is receiving "firearm related services." A license is also not required at a "shooting range", at a location where one is participating in a "firearms instructional course", or during "legal hunting activity." One may transport a handgun in a vehicle without a license if the handgun is "unloaded", "not readily accessible", and "secured in a case." A violation of is a class A misdemeanor. [5]</code> |
| <code>who plays kristen and susan on days of our lives</code> | <code>Susan Banks Susan Banks is a fictional character on NBC's daytime drama Days of Our Lives. She was played by Eileen Davidson from November 4, 1996 to April 8, 1998, and again in 2014 and 2017. Susan is the eccentric mother of Elvis "EJ" DiMera, and once acted as Kristen Blake's doppelganger. In November 2011, it was announced that Brynn Thayer would take over the role of Susan, since Davidson was committed to The Young and the Restless. Thayer made her brief one-off appearance as Susan on December 7, 2011.[1]</code> |
| <code>what is the revelation in the hero journey</code> | <code>Hero's journey This is the point of realization in which a greater understanding is achieved. Armed with this new knowledge and perception, the hero is resolved and ready for the more difficult part of the adventure.</code> |
- Loss: <code>AdaptiveLayerLoss</code> with these parameters:
{
"loss": "GISTEmbedLoss",
"n_layers_per_step": 3,
"last_layer_weight": 0.15,
"prior_layers_weight": 2.5,
"kl_div_weight": 0.75,
"kl_temperature": 0.5
}
trivia_pairs
- Dataset: trivia_pairs at a7c36e3
- Size: 9,700 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: 16.55 tokens</li><li>max: 43 tokens</li></ul> | <ul><li>min: 27 tokens</li><li>mean: 446.79 tokens</li><li>max: 512 tokens</li></ul> |
- Samples:
| sentence1 | sentence2 |
|---|
| <code>Most of the Three Tenors come from which country?</code> | <code>The Three Tenors - Pavarotti, Domingo, and Carreras The Three Tenors The Three Tenors Jose Carreras, Placido Domingo, and Luciano Pavarotti The Three Tenors - Luciano Pavarotti, Placido Domingo, and Jose Carreras. Photo By: Yoshikatsu Tsuno / Getty Images By Aaron Green Who Are the Three Tenors? The Three Tenors are made up of three of the world's most famous and beloved operatic tenors which include Jose Carreras, Placido Domingo, and Luciano Pavarotti. Jose Carreras (1946-) Born in Barcelona, Spain, Jose Carreras has been performing since he was 11 years old. Professionally, Carreras began his operatic career on December 19, 1970, when he sang the principal role of Gennaro alongside Montserrat Caballe in Donizetti's Lucrezia Borgia . Aside from performing, Carreras serves as president for the José Carreras International Leukaemia Foundation, which he founded after successfully overcoming his own battle with cancer. Placido Domingo (1941-) With over 100 operas and 147 roles under his belt, Placido Domingo is a seasoned operatic star. Born in Madrid, Spain, the celebrated tenor made his operatic debut as “Alfredo” in La Triviata at Monterrey, Mexico in 1961. Just as Carreras and Pavarotti, Domingo has performed in opera houses around the world. Now in his mid-70s and understanding the changes in his voice, Domingo sings baritone roles instead. In 1993, Domingo founded a young singer competition called Operalia. The competition is open to 18-32 year olds, and is hosted in a different city every year. Out of nearly 1,000 entrants, only the top 40 are selected for the competition. continue reading below our video Great Singers Gone too Soon Luciano Pavarotti (1935-2007) Born in Modena, Italy, Pavarotti had dreams of becoming a soccer goalkeeper, and it turned out he was quite good. However, his interest in music edged its way ahead after he won first place in the Llangollen International Singing Competition in Wales. Pavarotti went on to become one of the first opera stars to have nearly his entire performing career recorded musically and visually. He easily sold out shows and performed for millions of people in single performances. The Origin of the Three Tenors The idea for the Three Tenors came from Mario Dradi, an Italian manager and producer. Dradi's idea was to create a group of tenors for a concert and donate a portion of the proceeds to Jose Carreras's foundation after his successful treatment of leukemia. Jose Carreras, along with his two friends, Placido Domingo and Luciano Pavarotti , agreed to perform as the Three Tenors . Dradi's idea came to fruition on July 7, 1990, the day before the FIFA World Cup in Rome. The concert was watched by over 800 million viewers, and was so well received that when a recording of the concert was released, it became the biggest selling classical album in history. The album, Carreras - Domingo - Pavarotti: the Three Tenors in Concert , set a Guinness World Record. Because of the trio's instant success, they performed at the following three FIFA World Cups: Los Angeles in 1994, Paris in 1998, and Yokohama in 2002. The tremendous reception of the Three Tenors was due largely in part to their incredible voices, down-to-earth, likable personalities, and song selections. The trio would regularly perform classic and well-known operatic arias, as well as popular Broadway show tunes that even the most novice classical music listener could love and appreciate. Given the trio's enormous popularity, imitations of the Three Tenors quickly arose all over the world, including the Three Canadian Tenors, the Chinese Tenors, as well as the Three Mo' Tenors. The Three Tenors: Recommended YouTube Videos The Three Tenors sing "O Sole Mio" (1994) ( Watch on YouTube ) The Three Tenors sing "La donna e mobile" (1994) ( Watch on YouTube ) The Three Tenors sing "Singin' in the Rain" (1994) ( Watch on YouTube ) The Three Tenors sing "New York, New York" (1996) ( Watch on YouTube ) The Three Tenors sing "Ti Voglio Tanto Bene" (1998) ( Watch on YouTube ) The Three Tenors sing "Nessun Dorm</code> |
The model card is truncated. Read the rest on Hugging Face.