Climate Science Reranker
C
Climate Science Reranker
由 nicolauduran45 开发
这是一个基于交叉编码器的气候科学文本重排序模型,专门用于气候科学领域的语义搜索和文本相关性排序。
下载量 26
发布时间 : 5/12/2025
模型简介
该模型计算文本对的分数,可用于气候科学领域的文本重排序和语义搜索任务,基于MiniLM-L6-v2架构微调而来。
模型特点
气候科学领域优化
专门针对气候科学领域的文本进行了微调,能够更好地理解该领域的专业术语和概念。
高性能重排序
在气候科学评估数据集上取得了0.7068的NDCG@10分数,表现优异。
高效推理
基于MiniLM架构,在保持高性能的同时具有较高的推理效率。
模型能力
文本相关性评分
语义搜索重排序
气候科学领域文本理解
使用案例
学术研究
气候科学文献检索
用于气候科学领域的文献检索系统,提高搜索结果的相关性。
在气候科学评估数据集上NDCG@10达到0.7068
研究论文推荐
根据用户查询推荐最相关的气候科学研究论文。
信息检索
气候政策文档检索
帮助政策制定者快速找到与特定气候议题相关的政策文档。
🚀 气候科学重排器
这是一个基于 Cross Encoder 的模型,使用 sentence-transformers 库从 cross-encoder/ms-marco-MiniLM-L6-v2 微调而来。它可以计算文本对的得分,可用于文本重排和语义搜索。
✨ 主要特性
- 基于 Cross Encoder 架构,能够有效计算文本对之间的相关性得分。
- 从预训练模型 cross-encoder/ms-marco-MiniLM-L6-v2 微调而来,具有良好的泛化能力。
- 适用于文本重排和语义搜索任务,可帮助提高搜索结果的准确性。
📦 安装指南
首先,你需要安装 Sentence Transformers 库:
pip install -U sentence-transformers
💻 使用示例
基础用法
from sentence_transformers import CrossEncoder
# 从 Hugging Face Hub 下载模型
model = CrossEncoder("cross_encoder_model_id")
# 获取文本对的得分
pairs = [
["The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.", 'Currently there is renewed interest in harnessing the vast tidal resource to combat the twin challenges of climate change and energy security. However, within the UK no tidal barrage proposals have passed the development stage, this is due to a combination of high cost and environmental concerns. This paper demonstrates how a framework, such as the North West Hydro Resource Model can be applied to tidal barrages, with the Mersey barrage as a case study. The model materialised in order to provide developers with a tool to successfully identify the capacity of hydropower schemes in a specific location. A key feature of the resource model is the understanding that there is no single barrier to the utilisation of small hydropower but several obstacles, which together impede development. Thus, this paper contributes in part to a fully holistic treatment of tidal barrages, recognising that apart from energy generation, other environmental, societal and economic opportunities arise and must be fully investigated for robust decision-making. This study demonstrates how considering the societal needs of the people and the necessity for compensatory habitats, for example, an organic architectural design has developed, which aims to enhance rather than detract from the Mersey.'],
["The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.", 'Rainbows contribute to human wellbeing by providing an inspiring connection to nature. Because the rainbow is an atmospheric optical phenomenon that results from the refraction of sunlight by rainwater droplets, changes in precipitation and cloud cover due to anthropogenic climate forcing will alter rainbow distribution. Yet, we lack a basic understanding of the current spatial distribution of rainbows and how climate change might alter this pattern. To assess how climate change might affect rainbow viewing opportunities, we developed a global database of crowd-sourced photographed rainbows, trained an empirical model of rainbow occurrence, and applied this model to present-day climate and three future climate scenarios. Results suggest that the average terrestrial location on Earth currently has 117 ± 71 days per year with conditions suitable for rainbows. By 2100, climate change is likely to generate a 4.0–4.9 % net increase in mean global annual rainbow-days (i.e., days with at least one rainbow), with the greatest change under the highest emission scenario. Around 21–34 % of land areas will lose rainbow-days and 66–79 % will gain rainbow-days, with rainbow gain hotspots mainly in high-latitude and high-elevation regions with smaller human populations. Our research demonstrates that alterations to non-tangible environmental attributes due to climate change could be significant and are worthy of consideration and mitigation.'],
["The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.", 'The ascendancy of dinosaurs to become dominant components of terrestrial ecosystems was a pivotal event in the history of life, yet the drivers of their early evolution and biodiversity are poorly understood.1Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. The first 50 Myr of dinosaur evolution: macroevolutionary pattern and morphological disparity.Biol. Lett. 2008; 4: 733-736https://doi.org/10.1098/rsbl.2008.0441Crossref PubMed Scopus (105) Google Scholar,2Irmis R.B. Evaluating hypotheses for the early diversification of dinosaurs.Earth Environ. Sci. Trans. R. Soc. Edinb. 2010; 101: 397-426https://doi.org/10.1017/S1755691011020068Crossref Scopus (94) Google Scholar,3Benton M.J. Forth J. Langer M.C. Models for the rise of the dinosaurs.Curr. Biol. 2014; 24: R87-R95https://doi.org/10.1016/j.cub.2013.11.063Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar During their early diversification in the Late Triassic, dinosaurs were initially rare and geographically restricted, only attaining wider distributions and greater abundance following the end-Triassic mass extinction event.4Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. Superiority, competition, and opportunism in the evolutionary radiation of dinosaurs.Science. 2008; 321: 1485-1488https://doi.org/10.1126/science.1161833Crossref PubMed Scopus (334) Google Scholar,5Langer M.C. Ezcurra M.D. Bittencourt J.S. Novas F.E. The origin and early evolution of dinosaurs.Biol. Rev. Camb. Philos. Soc. 2010; 85: 55-110https://doi.org/10.1111/j.1469-185X.2009.00094.xCrossref PubMed Scopus (212) Google Scholar,6Langer M.C. Godoy P.L. So volcanoes created the dinosaurs? a quantitative characterization of the early evolution of terrestrial pan-aves.Front. Earth Sci. 2022; 10https://doi.org/10.3389/feart.2022.899562Crossref PubMed Scopus (3) Google Scholar This pattern is consistent with an opportunistic expansion model, initiated by the extinction of co-occurring groups such as aetosaurs, rauisuchians, and therapsids.4Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. Superiority, competition, and opportunism in the evolutionary radiation of dinosaurs.Science. 2008; 321: 1485-1488https://doi.org/10.1126/science.1161833Crossref PubMed Scopus (334) Google Scholar,7Tucker M.E. Benton M.J. Triassic environments, climates and reptile evolution.Palaeogeogr. Palaeoclimatol. Palaeoecol. 1982; 40: 361-379https://doi.org/10.1016/0031-0182(82)90034-7Crossref Scopus (89) Google Scholar,8Benton M.J. Dinosaur success in the triassic: a noncompetitive ecological model.Q. Rev. Biol. 1983; 58: 29-55Crossref Scopus (170) Google Scholar However, this pattern could instead be a response to changes in global climatic distributions through the Triassic to Jurassic transition, especially given the increasing evidence that climate played a key role in constraining Triassic dinosaur distributions.7Tucker M.E. Benton M.J. Triassic environments, climates and reptile evolution.Palaeogeogr. Palaeoclimatol. Palaeoecol. 1982; 40: 361-379https://doi.org/10.1016/0031-0182(82)90034-7Crossref Scopus (89) Google Scholar,9Whiteside J.H. Lindström S. Irmis R.B. Glasspool I.J. Schaller M.F. Dunlavey M. Nesbitt S.J. Smith N.D. Turner A.H. Extreme ecosystem instability suppressed tropical dinosaur dominance for 30 million years.Proc. Natl. Acad. Sci. USA. 2015; 112: 7909-7913https://doi.org/10.1073/pnas.1505252112Crossref PubMed Scopus (61) Google Scholar,10Bernardi M. Gianolla P. Petti F.M. Mietto P. Benton M.J. Dinosaur diversification linked with the Carnian pluvial episode.Nat. Commun. 2018; 9: 1499https://doi.org/10.1038/s41467-018-03996-1Crossref PubMed Scopus (87) Google Scholar,11Lovelace D.M. Hartman S.A. Mathewson P.D. Linzmeier B.J. Porter W.P. Modeling Dragons: using linked mechanistic physiological and microclimate models to explore environmental, physiological, and morphological constraints on the early evolution of dinosaurs.PLoS One. 2020; 15e0223872https://doi.org/10.1371/journal.pone.0223872Crossref Scopus (8) Google Scholar,12Mancuso A.C. Benavente C.A. Irmis R.B. Mundil R. Evidence for the Carnian pluvial episode in Gondwana: new multiproxy climate records and their bearing on early dinosaur diversification.Gondwana Res. 2020; 86: 104-125https://doi.org/10.1016/j.gr.2020.05.009Crossref Scopus (35) Google Scholar,13Mancuso A.C. Irmis R.B. Pedernera T.E. Gaetano L.C. Benavente C.A. Breeden III B.T. Paleoenvironmental and biotic changes in the late triassic of Argentina: testing hypotheses of abiotic forcing at the basin scale.Front. Earth Sci. 2022; 10https://doi.org/10.3389/feart.2022.883788Crossref PubMed Scopus (4) Google Scholar,14Kent D.V. Clemmensen L.B. Northward dispersal of dinosaurs from Gondwana to Greenland at the mid-Norian (215–212 Ma, Late Triassic) dip in atmospheric pCO2.Proc. Natl. Acad. Sci. USA. 2021; 118e2020778118https://doi.org/10.1073/pnas.2020778118Crossref Scopus (16) Google Scholar,15Griffin C.T. Wynd B.M. Munyikwa D. Broderick T.J. Zondo M. Tolan S. Langer M.C. Nesbitt S.J. Taruvinga H.R. Africa\'s oldest dinosaurs reveal early suppression of dinosaur distribution.Nature. 2022; 609: 313-319https://doi.org/10.1038/s41586-022-05133-xCrossref PubMed Scopus (4) Google Scholar,16Olsen P. Sha J. Fang Y. Chang C. Whiteside J.H. Kinney S. Sues H.-D. Kent D. Schaller M. Vajda V. Arctic ice and the ecological rise of the dinosaurs.Sci. Adv. 2022; 8eabo6342https://doi.org/10.1126/sciadv.abo6342Crossref Scopus (5) Google Scholar Here, we test this hypothesis and elucidate how climate influenced early dinosaur distribution by quantitatively examining changes in dinosaur and tetrapod "climatic niche space" across the Triassic-Jurassic boundary. Statistical analyses show that Late Triassic sauropodomorph dinosaurs occupied a more restricted climatic niche space than other tetrapods and dinosaurs, being excluded from the hottest, low-latitude climate zones. A subsequent, earliest Jurassic expansion of sauropodomorph geographic distribution is linked to the expansion of their preferred climatic conditions. Evolutionary model-fitting analyses provide evidence for an important evolutionary shift from cooler to warmer climatic niches during the origin of Sauropoda. These results are consistent with the hypothesis that global abundance of sauropodomorph dinosaurs was facilitated by climatic change and provide support for the key role of climate in the ascendancy of dinosaurs.'],
["The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.", 'The development of technologies to slow climate change has been identified as a global imperative. Nonetheless, such ‘green’ technologies can potentially have negative impacts on biodiversity. We explored how climate change and the mining of lithium for green technologies influence surface water availability, primary productivity and the abundance of three threatened and economically important flamingo species in the ‘Lithium Triangle’ of the Chilean Andes. We combined climate and primary productivity data with remotely sensed measures of surface water levels and a 30-year dataset on flamingo abundance using structural equation modelling. We found that, regionally, flamingo abundance fluctuated dramatically from year-to-year in response to variation in surface water levels and primary productivity but did not exhibit any temporal trends. Locally, in the Salar de Atacama—where lithium mining is focused—we found that mining was negatively correlated with the abundance of two of the three flamingo species. These results suggest continued increases in lithium mining and declines in surface water could soon have dramatic effects on flamingo abundance across their range. Efforts to slow the expansion of mining and the impacts of climate change are, therefore, urgently needed to benefit local biodiversity and the local human economy that depends on it.'],
["The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.", 'Rivers can abruptly shift pathways in rare events called avulsions, which cause devastating floods. The controls on avulsion locations are poorly understood as a result of sparse data on such features. We analyzed nearly 50 years of satellite imagery and documented 113 avulsions across the globe that indicate three distinct controls on avulsion location. Avulsions on fans coincide with valley-confinement change, whereas avulsions on deltas are primarily clustered within the backwater zone, indicating a control by spatial flow deceleration or acceleration during floods. However, 38% of avulsions on deltas occurred upstream of backwater effects. These events occurred in steep, sediment-rich rivers in tropical and desert environments. Our results indicate that avulsion location on deltas is set by the upstream extent of flood-driven erosion, which is typically limited to the backwater zone but can extend far upstream in steep, sediment-laden rivers. Our findings elucidate how avulsion hazards might respond to land use and climate change.'],
]
scores = model.predict(pairs)
print(scores.shape)
# (5,)
高级用法
# 根据与单个文本的相似度对不同文本进行排序
ranks = model.rank(
"The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.",
[
'Currently there is renewed interest in harnessing the vast tidal resource to combat the twin challenges of climate change and energy security. However, within the UK no tidal barrage proposals have passed the development stage, this is due to a combination of high cost and environmental concerns. This paper demonstrates how a framework, such as the North West Hydro Resource Model can be applied to tidal barrages, with the Mersey barrage as a case study. The model materialised in order to provide developers with a tool to successfully identify the capacity of hydropower schemes in a specific location. A key feature of the resource model is the understanding that there is no single barrier to the utilisation of small hydropower but several obstacles, which together impede development. Thus, this paper contributes in part to a fully holistic treatment of tidal barrages, recognising that apart from energy generation, other environmental, societal and economic opportunities arise and must be fully investigated for robust decision-making. This study demonstrates how considering the societal needs of the people and the necessity for compensatory habitats, for example, an organic architectural design has developed, which aims to enhance rather than detract from the Mersey.',
'Rainbows contribute to human wellbeing by providing an inspiring connection to nature. Because the rainbow is an atmospheric optical phenomenon that results from the refraction of sunlight by rainwater droplets, changes in precipitation and cloud cover due to anthropogenic climate forcing will alter rainbow distribution. Yet, we lack a basic understanding of the current spatial distribution of rainbows and how climate change might alter this pattern. To assess how climate change might affect rainbow viewing opportunities, we developed a global database of crowd-sourced photographed rainbows, trained an empirical model of rainbow occurrence, and applied this model to present-day climate and three future climate scenarios. Results suggest that the average terrestrial location on Earth currently has 117 ± 71 days per year with conditions suitable for rainbows. By 2100, climate change is likely to generate a 4.0–4.9 % net increase in mean global annual rainbow-days (i.e., days with at least one rainbow), with the greatest change under the highest emission scenario. Around 21–34 % of land areas will lose rainbow-days and 66–79 % will gain rainbow-days, with rainbow gain hotspots mainly in high-latitude and high-elevation regions with smaller human populations. Our research demonstrates that alterations to non-tangible environmental attributes due to climate change could be significant and are worthy of consideration and mitigation.',
'The ascendancy of dinosaurs to become dominant components of terrestrial ecosystems was a pivotal event in the history of life, yet the drivers of their early evolution and biodiversity are poorly understood.1Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. The first 50 Myr of dinosaur evolution: macroevolutionary pattern and morphological disparity.Biol. Lett. 2008; 4: 733-736https://doi.org/10.1098/rsbl.2008.0441Crossref PubMed Scopus (105) Google Scholar,2Irmis R.B. Evaluating hypotheses for the early diversification of dinosaurs.Earth Environ. Sci. Trans. R. Soc. Edinb. 2010; 101: 397-426https://doi.org/10.1017/S1755691011020068Crossref Scopus (94) Google Scholar,3Benton M.J. Forth J. Langer M.C. Models for the rise of the dinosaurs.Curr. Biol. 2014; 24: R87-R95https://doi.org/10.1016/j.cub.2013.11.063Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar During their early diversification in the Late Triassic, dinosaurs were initially rare and geographically restricted, only attaining wider distributions and greater abundance following the end-Triassic mass extinction event.4Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. Superiority, competition, and opportunism in the evolutionary radiation of dinosaurs.Science. 2008; 321: 1485-1488https://doi.org/10.1126/science.1161833Crossref PubMed Scopus (334) Google Scholar,5Langer M.C. Ezcurra M.D. Bittencourt J.S. Novas F.E. The origin and early evolution of dinosaurs.Biol. Rev. Camb. Philos. Soc. 2010; 85: 55-110https://doi.org/10.1111/j.1469-185X.2009.00094.xCrossref PubMed Scopus (212) Google Scholar,6Langer M.C. Godoy P.L. So volcanoes created the dinosaurs? a quantitative characterization of the early evolution of terrestrial pan-aves.Front. Earth Sci. 2022; 10https://doi.org/10.3389/feart.2022.899562Crossref PubMed Scopus (3) Google Scholar This pattern is consistent with an opportunistic expansion model, initiated by the extinction of co-occurring groups such as aetosaurs, rauisuchians, and therapsids.4Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. Superiority, competition, and opportunism in the evolutionary radiation of dinosaurs.Science. 2008; 321: 1485-1488https://doi.org/10.1126/science.1161833Crossref PubMed Scopus (334) Google Scholar,7Tucker M.E. Benton M.J. Triassic environments, climates and reptile evolution.Palaeogeogr. Palaeoclimatol. Palaeoecol. 1982; 40: 361-379https://doi.org/10.1016/0031-0182(82)90034-7Crossref Scopus (89) Google Scholar,8Benton M.J. Dinosaur success in the triassic: a noncompetitive ecological model.Q. Rev. Biol. 1983; 58: 29-55Crossref Scopus (170) Google Scholar However, this pattern could instead be a response to changes in global climatic distributions through the Triassic to Jurassic transition, especially given the increasing evidence that climate played a key role in constraining Triassic dinosaur distributions.7Tucker M.E. Benton M.J. Triassic environments, climates and reptile evolution.Palaeogeogr. Palaeoclimatol. Palaeoecol. 1982; 40: 361-379https://doi.org/10.1016/0031-0182(82)90034-7Crossref Scopus (89) Google Scholar,9Whiteside J.H. Lindström S. Irmis R.B. Glasspool I.J. Schaller M.F. Dunlavey M. Nesbitt S.J. Smith N.D. Turner A.H. Extreme ecosystem instability suppressed tropical dinosaur dominance for 30 million years.Proc. Natl. Acad. Sci. USA. 2015; 112: 7909-7913https://doi.org/10.1073/pnas.1505252112Crossref PubMed Scopus (61) Google Scholar,10Bernardi M. Gianolla P. Petti F.M. Mietto P. Benton M.J. Dinosaur diversification linked with the Carnian pluvial episode.Nat. Commun. 2018; 9: 1499https://doi.org/10.1038/s41467-018-03996-1Crossref PubMed Scopus (87) Google Scholar,11Lovelace D.M. Hartman S.A. Mathewson P.D. Linzmeier B.J. Porter W.P. Modeling Dragons: using linked mechanistic physiological and microclimate models to explore environmental, physiological, and morphological constraints on the early evolution of dinosaurs.PLoS One. 2020; 15e0223872https://doi.org/10.1371/journal.pone.0223872Crossref Scopus (8) Google Scholar,12Mancuso A.C. Benavente C.A. Irmis R.B. Mundil R. Evidence for the Carnian pluvial episode in Gondwana: new multiproxy climate records and their bearing on early dinosaur diversification.Gondwana Res. 2020; 86: 104-125https://doi.org/10.1016/j.gr.2020.05.009Crossref Scopus (35) Google Scholar,13Mancuso A.C. Irmis R.B. Pedernera T.E. Gaetano L.C. Benavente C.A. Breeden III B.T. Paleoenvironmental and biotic changes in the late triassic of Argentina: testing hypotheses of abiotic forcing at the basin scale.Front. Earth Sci. 2022; 10https://doi.org/10.3389/feart.2022.883788Crossref PubMed Scopus (4) Google Scholar,14Kent D.V. Clemmensen L.B. Northward dispersal of dinosaurs from Gondwana to Greenland at the mid-Norian (215–212 Ma, Late Triassic) dip in atmospheric pCO2.Proc. Natl. Acad. Sci. USA. 2021; 118e2020778118https://doi.org/10.1073/pnas.2020778118Crossref Scopus (16) Google Scholar,15Griffin C.T. Wynd B.M. Munyikwa D. Broderick T.J. Zondo M. Tolan S. Langer M.C. Nesbitt S.J. Taruvinga H.R. Africa\'s oldest dinosaurs reveal early suppression of dinosaur distribution.Nature. 2022; 609: 313-319https://doi.org/10.1038/s41586-022-05133-xCrossref PubMed Scopus (4) Google Scholar,16Olsen P. Sha J. Fang Y. Chang C. Whiteside J.H. Kinney S. Sues H.-D. Kent D. Schaller M. Vajda V. Arctic ice and the ecological rise of the dinosaurs.Sci. Adv. 2022; 8eabo6342https://doi.org/10.1126/sciadv.abo6342Crossref Scopus (5) Google Scholar Here, we test this hypothesis and elucidate how climate influenced early dinosaur distribution by quantitatively examining changes in dinosaur and tetrapod "climatic niche space" across the Triassic-Jurassic boundary. Statistical analyses show that Late Triassic sauropodomorph dinosaurs occupied a more restricted climatic niche space than other tetrapods and dinosaurs, being excluded from the hottest, low-latitude climate zones. A subsequent, earliest Jurassic expansion of sauropodomorph geographic distribution is linked to the expansion of their preferred climatic conditions. Evolutionary model-fitting analyses provide evidence for an important evolutionary shift from cooler to warmer climatic niches during the origin of Sauropoda. These results are consistent with the hypothesis that global abundance of sauropodomorph dinosaurs was facilitated by climatic change and provide support for the key role of climate in the ascendancy of dinosaurs.',
'The development of technologies to slow climate change has been identified as a global imperative. Nonetheless, such ‘green’ technologies can potentially have negative impacts on biodiversity. We explored how climate change and the mining of lithium for green technologies influence surface water availability, primary productivity and the abundance of three threatened and economically important flamingo species in the ‘Lithium Triangle’ of the Chilean Andes. We combined climate and primary productivity data with remotely sensed measures of surface water levels and a 30-year dataset on flamingo abundance using structural equation modelling. We found that, regionally, flamingo abundance fluctuated dramatically from year-to-year in response to variation in surface water levels and primary productivity but did not exhibit any temporal trends. Locally, in the Salar de Atacama—where lithium mining is focused—we found that mining was negatively correlated with the abundance of two of the three flamingo species. These results suggest continued increases in lithium mining and declines in surface water could soon have dramatic effects on flamingo abundance across their range. Efforts to slow the expansion of mining and the impacts of climate change are, therefore, urgently needed to benefit local biodiversity and the local human economy that depends on it.',
'Rivers can abruptly shift pathways in rare events called avulsions, which cause devastating floods. The controls on avulsion locations are poorly understood as a result of sparse data on such features. We analyzed nearly 50 years of satellite imagery and documented 113 avulsions across the globe that indicate three distinct controls on avulsion location. Avulsions on fans coincide with valley-confinement change, whereas avulsions on deltas are primarily clustered within the backwater zone, indicating a control by spatial flow deceleration or acceleration during floods. However, 38% of avulsions on deltas occurred upstream of backwater effects. These events occurred in steep, sediment-rich rivers in tropical and desert environments. Our results indicate that avulsion location on deltas is set by the upstream extent of flood-driven erosion, which is typically limited to the backwater zone but can extend far upstream in steep, sediment-laden rivers. Our findings elucidate how avulsion hazards might respond to land use and climate change.',
]
)
# [{'corpus_id': ..., 'score': ...}, {'corpus_id': ..., 'score': ...}, ...]
📚 详细文档
模型详情
模型描述
属性 | 详情 |
---|---|
模型类型 | Cross Encoder |
基础模型 | cross-encoder/ms-marco-MiniLM-L6-v2 |
最大序列长度 | 512 个词元 |
输出标签数量 | 1 个标签 |
语言 | 英文 |
许可证 | apache-2.0 |
模型来源
- 文档:Sentence Transformers 文档
- 文档:Cross Encoder 文档
- 仓库:GitHub 上的 Sentence Transformers
- Hugging Face:Hugging Face 上的 Cross Encoders
评估
指标
Cross Encoder 重排
- 数据集:
climate-science-eval
- 使用
CrossEncoderRerankingEvaluator
进行评估,参数如下:
{
"at_k": 10,
"always_rerank_positives": true
}
指标 | 值 |
---|---|
map | 0.6629 (+0.4483) |
mrr@10 | 0.6554 (+0.4475) |
ndcg@10 | 0.7068 (+0.4669) |
训练详情
训练数据集
未命名数据集
- 大小:263,476 个训练样本
- 列:
query
、answer
和label
- 基于前 1000 个样本的近似统计信息:
| | 查询 | 答案 | 标签 |
|------|------|------|------|
| 类型 | 字符串 | 字符串 | 整数 |
| 详情 |
- 最小:55 个字符
- 平均:178.19 个字符
- 最大:593 个字符
- 最小:13 个字符
- 平均:1510.36 个字符
- 最大:29945 个字符
- 0:~74.40%
- 1:~25.60%
- 样本:
| 查询 | 答案 | 标签 |
|------|------|------|
|
The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.
|Currently there is renewed interest in harnessing the vast tidal resource to combat the twin challenges of climate change and energy security. However, within the UK no tidal barrage proposals have passed the development stage, this is due to a combination of high cost and environmental concerns. This paper demonstrates how a framework, such as the North West Hydro Resource Model can be applied to tidal barrages, with the Mersey barrage as a case study. The model materialised in order to provide developers with a tool to successfully identify the capacity of hydropower schemes in a specific location. A key feature of the resource model is the understanding that there is no single barrier to the utilisation of small hydropower but several obstacles, which together impede development. Thus, this paper contributes in part to a fully holistic treatment of tidal barrages, recognising that apart from energy generation, other environmental, societal and economic opportunities arise and must b...
|1
| |The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.
|Rainbows contribute to human wellbeing by providing an inspiring connection to nature. Because the rainbow is an atmospheric optical phenomenon that results from the refraction of sunlight by rainwater droplets, changes in precipitation and cloud cover due to anthropogenic climate forcing will alter rainbow distribution. Yet, we lack a basic understanding of the current spatial distribution of rainbows and how climate change might alter this pattern. To assess how climate change might affect rainbow viewing opportunities, we developed a global database of crowd-sourced photographed rainbows, trained an empirical model of rainbow occurrence, and applied this model to present-day climate and three future climate scenarios. Results suggest that the average terrestrial location on Earth currently has 117 ± 71 days per year with conditions suitable for rainbows. By 2100, climate change is likely to generate a 4.0–4.9 % net increase in mean global annual rainbow-days (i.e., days with at leas...
|0
| |The researchers say that with the right design a Mersey barrage has the potential to become a globally identifiable piece of architectural infrastructure - a 'hydropower landmark' boosting tourism to the region.
|The ascendancy of dinosaurs to become dominant components of terrestrial ecosystems was a pivotal event in the history of life, yet the drivers of their early evolution and biodiversity are poorly understood.1Brusatte S.L. Benton M.J. Ruta M. Lloyd G.T. The first 50 Myr of dinosaur evolution: macroevolutionary pattern and morphological disparity.Biol. Lett. 2008; 4: 733-736https://doi.org/10.1098/rsbl.2008.0441Crossref PubMed Scopus (105) Google Scholar,2Irmis R.B. Evaluating hypotheses for the early diversification of dinosaurs.Earth Environ. Sci. Trans. R. Soc. Edinb. 2010; 101: 397-426https://doi.org/10.1017/S1755691011020068Crossref Scopus (94) Google Scholar,3Benton M.J. Forth J. Langer M.C. Models for the rise of the dinosaurs.Curr. Biol. 2014; 24: R87-R95https://doi.org/10.1016/j.cub.2013.11.063Abstract Full Text Full Text PDF PubMed Scopus (93) Google Scholar During their early diversification in the Late Triassic, dinosaurs were initially rare and geographically restricted, on...
|0
| - 损失函数:
BinaryCrossEntropyLoss
,参数如下:
{
"activation_fn": "torch.nn.modules.linear.Identity",
"pos_weight": 6
}
训练超参数
非默认超参数
eval_strategy
:stepsper_device_train_batch_size
:16per_device_eval_batch_size
:16learning_rate
:2e-05warmup_ratio
:0.1fp16
:Truedataloader_num_workers
:4load_best_model_at_end
:True
所有超参数
点击展开
overwrite_output_dir
:Falsedo_predict
:Falseeval_strategy
:stepsprediction_loss_only
:Trueper_device_train_batch_size
:16per_device_eval_batch_size
:16per_gpu_train_batch_size
:Noneper_gpu_eval_batch_size
:Nonegradient_accumulation_steps
:1eval_accumulation_steps
:Nonetorch_empty_cache_steps
:Nonelearning_rate
:2e-05weight_decay
:0.0adam_beta1
:0.9adam_beta2
:0.999adam_epsilon
:1e-08max_grad_norm
:1.0num_train_epochs
:3max_steps
:-1lr_scheduler_type
:linearlr_scheduler_kwargs
:{}warmup_ratio
:0.1warmup_steps
:0log_level
:passivelog_level_replica
:warninglog_on_each_node
:Truelogging_nan_inf_filter
:Truesave_safetensors
:Truesave_on_each_node
:Falsesave_only_model
:Falserestore_callback_states_from_checkpoint
:Falseno_cuda
:Falseuse_cpu
:Falseuse_mps_device
:Falseseed
:42data_seed
:Nonejit_mode_eval
:Falseuse_ipex
:Falsebf16
:Falsefp16
:Truefp16_opt_level
:O1half_precision_backend
:autobf16_full_eval
:Falsefp16_full_eval
:Falsetf32
:Nonelocal_rank
:0ddp_backend
:Nonetpu_num_cores
:Nonetpu_metrics_debug
:Falsedebug
:[]dataloader_drop_last
:Falsedataloader_num_workers
:4dataloader_prefetch_factor
:Nonepast_index
:-1disable_tqdm
:Falseremove_unused_columns
:Truelabel_names
:Noneload_best_model_at_end
:Trueignore_data_skip
:Falsefsdp
:[]fsdp_min_num_params
:0fsdp_config
:{'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}tp_size
:0fsdp_transformer_layer_cls_to_wrap
:Noneaccelerator_config
:{'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}deepspeed
:Nonelabel_smoothing_factor
:0.0optim
:adamw_torchoptim_args
:Noneadafactor
:Falsegroup_by_length
:Falselength_column_name
:lengthddp_find_unused_parameters
:Noneddp_bucket_cap_mb
:Noneddp_broadcast_buffers
:Falsedataloader_pin_memory
:Truedataloader_persistent_workers
:Falseskip_memory_metrics
:Trueuse_legacy_prediction_loop
:Falsepush_to_hub
:Falseresume_from_checkpoint
:Nonehub_model_id
:Nonehub_strategy
:every_savehub_private_repo
:Nonehub_always_push
:Falsegradient_checkpointing
:Falsegradient_checkpointing_kwargs
:Noneinclude_inputs_for_metrics
:Falseinclude_for_metrics
:[]eval_do_concat_batches
:Truefp16_backend
:autopush_to_hub_model_id
:Nonepush_to_hub_organization
:Nonemp_parameters
:auto_find_batch_size
:Falsefull_determinism
:Falsetorchdynamo
:Noneray_scope
:lastddp_timeout
:1800torch_compile
:Falsetorch_compile_backend
:Nonetorch_compile_mode
:Noneinclude_tokens_per_second
:Falseinclude_num_input_tokens_seen
:Falseneftune_noise_alpha
:Noneoptim_target_modules
:Nonebatch_eval_metrics
:Falseeval_on_start
:Falseuse_liger_kernel
:Falseeval_use_gather_object
:Falseaverage_tokens_across_devices
:Falseprompts
:Nonebatch_sampler
:batch_samplermulti_dataset_batch_sampler
:proportional
训练日志
点击展开
轮次 | 步数 | 训练损失 | climate-science-eval_ndcg@10 |
---|---|---|---|
0.0001 | 1 | 6.4826 | - |
0.0061 | 100 | 6.3516 | - |
0.0121 | 200 | 5.1792 | - |
0.0182 | 300 | 2.9628 | - |
0.0243 | 400 | 1.8946 | - |
0.0304 | 500 | 1.3992 | - |
0.0364 | 600 | 1.4469 | - |
0.0425 | 700 | 1.1841 | - |
0.0486 | 800 | 0.9967 | - |
0.0547 | 900 | 0.9914 | - |
0.0607 | 1000 | 0.7138 | 0.6113 (+0.3713) |
0.0668 | 1100 | 0.6944 | - |
0.0729 | 1200 | 0.7374 | - |
0.0789 | 1300 | 0.7249 | - |
0.0850 | 1400 | 0.8826 | - |
0.0911 | 1500 | 0.6886 | - |
0.0972 | 1600 | 0.8185 | - |
0.1032 | 1700 | 0.6946 | - |
0.1093 | 1800 | 0.7231 | - |
0.1154 | 1900 | 0.668 | - |
0.1214 | 2000 | 0.6434 | 0.6325 (+0.3926) |
0.1275 | 2100 | 0.7417 | - |
0.1336 | 2200 | 0.6777 | - |
0.1397 | 2300 | 0.779 | - |
0.1457 | 2400 | 0.6876 | - |
0.1518 | 2500 | 0.6619 | - |
0.1579 | 2600 | 0.6626 | - |
0.1640 | 2700 | 0.7394 | - |
0.1700 | 2800 | 0.6654 | - |
0.1761 | 2900 | 0.6026 | - |
0.1822 | 3000 | 0.6838 | 0.6417 (+0.4018) |
0.1882 | 3100 | 0.6423 | - |
0.1943 | 3200 | 0.6559 | - |
0.2004 | 3300 | 0.6097 | - |
0.2065 | 3400 | 0.6564 | - |
0.2125 | 3500 | 0.6912 | - |
0.2186 | 3600 | 0.6183 | - |
0.2247 | 3700 | 0.5585 | - |
0.2308 | 3800 | 0.6748 | - |
0.2368 | 3900 | 0.6165 | - |
0.2429 | 4000 | 0.6358 | 0.6529 (+0.4130) |
0.2490 | 4100 | 0.6473 | - |
0.2550 | 4200 | 0.6766 | - |
0.2611 | 4300 | 0.6603 | - |
0.2672 | 4400 | 0.5778 | - |
0.2733 | 4500 | 0.6732 | - |
0.2793 | 4600 | 0.605 | - |
0.2854 | 4700 | 0.6943 | - |
0.2915 | 4800 | 0.5776 | - |
0.2975 | 4900 | 0.706 | - |
0.3036 | 5000 | 0.5758 | 0.6559 (+0.4160) |
0.3097 | 5100 | 0.6596 | - |
0.3158 | 5200 | 0.6466 | - |
0.3218 | 5300 | 0.6116 | - |
0.3279 | 5400 | 0.5654 | - |
0.3340 | 5500 | 0.643 | - |
0.3401 | 5600 | 0.7281 | - |
0.3461 | 5700 | 0.6295 | - |
0.3522 | 5800 | 0.6555 | - |
0.3583 | 5900 | 0.6671 | - |
0.3643 | 6000 | 0.6647 | 0.6537 (+0.4138) |
0.3704 | 6100 | 0.5458 | - |
0.3765 | 6200 | 0.6279 | - |
0.3826 | 6300 | 0.6575 | - |
0.3886 | 6400 | 0.6206 | - |
0.3947 | 6500 | 0.5802 | - |
0.4008 | 6600 | 0.7117 | - |
0.4068 | 6700 | 0.589 | - |
0.4129 | 6800 | 0.6245 | - |
0.4190 | 6900 | 0.5346 | - |
0.4251 | 7000 | 0.7323 | 0.6559 (+0.4160) |
0.4311 | 7100 | 0.5407 | - |
0.4372 | 7200 | 0.53 | - |
0.4433 | 7300 | 0.5586 | - |
0.4494 | 7400 | 0.6219 | - |
0.4554 | 7500 | 0.6396 | - |
0.4615 | 7600 | 0.54 | - |
0.4676 | 7700 | 0.6284 | - |
0.4736 | 7800 | 0.6021 | - |
0.4797 | 7900 | 0.6326 | - |
0.4858 | 8000 | 0.6375 | 0.6691 (+0.4291) |
0.4919 | 8100 | 0.5402 | - |
0.4979 | 8200 | 0.582 | - |
0.5040 | 8300 | 0.5382 | - |
0.5101 | 8400 | 0.581 | - |
0.5162 | 8500 | 0.6062 | - |
0.5222 | 8600 | 0.5804 | - |
0.5283 | 8700 | 0.6233 | - |
0.5344 | 8800 | 0.5813 | - |
0.5404 | 8900 | 0.5619 | - |
0.5465 | 9000 | 0.5328 | 0.6694 (+0.4295) |
0.5526 | 9100 | 0.5371 | - |
0.5587 | 9200 | 0.6534 | - |
0.5647 | 9300 | 0.5395 | - |
0.5708 | 9400 | 0.577 | - |
0.5769 | 9500 | 0.5936 | - |
0.5829 | 9600 | 0.5947 | - |
0.5890 | 9700 | 0.5806 | - |
0.5951 | 9800 | 0.6236 | - |
0.6012 | 9900 | 0.6087 | - |
0.6072 | 10000 | 0.5466 | 0.6712 (+0.4313) |
0.6133 | 10100 | 0.6824 | - |
0.6194 | 10200 | 0.5657 | - |
0.6255 | 10300 | 0.5772 | - |
0.6315 | 10400 | 0.6068 | - |
0.6376 | 10500 | 0.4815 | - |
0.6437 | 10600 | 0.527 | - |
0.6497 | 10700 | 0.6041 | - |
0.6558 | 10800 | 0.5542 | - |
0.6619 | 10900 | 0.5846 | - |
0.6680 | 11000 | 0.5559 | 0.6683 (+0.4284) |
0.6740 | 11100 | 0.6235 | - |
0.6801 | 11200 | 0.581 | - |
0.6862 | 11300 | 0.5931 | - |
0.6923 | 11400 | 0.532 | - |
0.6983 | 11500 | 0.5832 | - |
0.7044 | 11600 | 0.4815 | - |
0.7105 | 11700 | 0.7507 | - |
0.7165 | 11800 | 0.555 | - |
0.7226 | 11900 | 0.585 | - |
0.7287 | 12000 | 0.6486 | 0.6711 (+0.4311) |
0.7348 | 12100 | 0.6077 | - |
0.7408 | 12200 | 0.5116 | - |
0.7469 | 12300 | 0.6163 | - |
0.7530 | 12400 | 0.6205 | - |
0.7590 | 12500 | 0.5086 | - |
0.7651 | 12600 | 0.5544 | - |
0.7712 | 12700 | 0.4743 | - |
0.7773 | 12800 | 0.5854 | - |
0.7833 | 12900 | 0.5681 | - |
0.7894 | 13000 | 0.6179 | 0.6760 (+0.4360) |
0.7955 | 13100 | 0.5958 | - |
0.8016 | 13200 | 0.5162 | - |
0.8076 | 13300 | 0.609 | - |
0.8137 | 13400 | 0.4877 | - |
0.8198 | 13500 | 0.6157 | - |
0.8258 | 13600 | 0.5638 | - |
0.8319 | 13700 | 0.5049 | - |
0.8380 | 13800 | 0.7226 | - |
0.8441 | 13900 | 0.515 | - |
0.8501 | 14000 | 0.5564 | 0.6822 (+0.4423) |
0.8562 | 14100 | 0.5618 | - |
0.8623 | 14200 | 0.5448 | - |
0.8684 | 14300 | 0.5693 | - |
0.8744 | 14400 | 0.6417 | - |
0.8805 | 14500 | 0.5609 | - |
0.8866 | 14600 | 0.6033 | - |
0.8926 | 14700 | 0.6355 | - |
0.8987 | 14800 | 0.5322 | - |
0.9048 | 14900 | 0.519 | - |
0.9109 | 15000 | 0.5662 | 0.6764 (+0.4365) |
0.9169 | 15100 | 0.593 | - |
0.9230 | 15200 | 0.6004 | - |
0.9291 | 15300 | 0.5673 | - |
0.9351 | 15400 | 0.5142 | - |
0.9412 | 15500 | 0.5859 | - |
0.9473 | 15600 | 0.6421 | - |
0.9534 | 15700 | 0.4822 | - |
0.9594 | 15800 | 0.6082 | - |
0.9655 | 15900 | 0.5373 | - |
0.9716 | 16000 | 0.6102 | 0.6729 (+0.4330) |
0.9777 | 16100 | 0.5109 | - |
0.9837 | 16200 | 0.6156 | - |
0.9898 | 16300 | 0.6408 | - |
0.9959 | 16400 | 0.5031 | - |
1.0019 | 16500 | 0.4652 | - |
1.0080 | 16600 | 0.3893 | - |
1.0141 | 16700 | 0.6276 | - |
1.0202 | 16800 | 0.5526 | - |
1.0262 | 16900 | 0.551 | - |
1.0323 | 17000 | 0.5066 | 0.6832 (+0.4432) |
1.0384 | 17100 | 0.5074 | - |
1.0444 | 17200 | 0.48 | - |
1.0505 | 17300 | 0.6073 | - |
1.0566 | 17400 | 0.485 | - |
1.0627 | 17500 | 0.4927 | - |
1.0687 | 17600 | 0.597 | - |
1.0748 | 17700 | 0.4376 | - |
1.0809 | 17800 | 0.4935 | - |
1.0870 | 17900 | 0.5702 | - |
1.0930 | 18000 | 0.4482 | 0.6825 (+0.4426) |
1.0991 | 18100 | 0.5183 | - |
1.1052 | 18200 | 0.4593 | - |
1.1112 | 18300 | 0.4775 | - |
1.1173 | 18400 | 0.5831 | - |
1.1234 | 18500 | 0.4942 | - |
1.1295 | 18600 | 0.5684 | - |
1.1355 | 18700 | 0.5214 | - |
1.1416 | 18800 | 0.5292 | - |
1.1477 | 18900 | 0.5163 | - |
1.1538 | 19000 | 0.5305 | 0.6868 (+0.4469) |
1.1598 | 19100 | 0.4507 | - |
1.1659 | 19200 | 0.4699 | - |
1.1720 | 19300 | 0.4532 | - |
1.1780 | 19400 | 0.4853 | - |
1.1841 | 19500 | 0.5169 | - |
1.1902 | 19600 | 0.5927 | - |
1.1963 | 19700 | 0.5777 | - |
1.2023 | 19800 | 0.5041 | - |
1.2084 | 19900 | 0.5309 | - |
1.2145 | 20000 | 0.4426 | 0.6809 (+0.4410) |
1.2205 | 20100 | 0.54 | - |
1.2266 | 20200 | 0.5692 | - |
1.2327 | 20300 | 0.5004 | - |
1.2388 | 20400 | 0.5044 | - |
1.2448 | 20500 | 0.4574 | - |
1.2509 | 20600 | 0.6132 | - |
1.2570 | 20700 | 0.4477 | - |
1.2631 | 20800 | 0.4805 | - |
1.2691 | 20900 | 0.6127 | - |
1.2752 | 21000 | 0.4349 | 0.6914 (+0.4515) |
1.2813 | 21100 | 0.6595 | - |
1.2873 | 21200 | 0.5234 | - |
1.2934 | 21300 | 0.4525 | - |
1.2995 | 21400 | 0.3841 | - |
1.3056 | 21500 | 0.5215 | - |
1. |
Jina Embeddings V3
Jina Embeddings V3 是一个多语言句子嵌入模型,支持超过100种语言,专注于句子相似度和特征提取任务。
文本嵌入
Transformers 支持多种语言

J
jinaai
3.7M
911
Ms Marco MiniLM L6 V2
Apache-2.0
基于MS Marco段落排序任务训练的交叉编码器模型,用于信息检索中的查询-段落相关性评分
文本嵌入 英语
M
cross-encoder
2.5M
86
Opensearch Neural Sparse Encoding Doc V2 Distill
Apache-2.0
基于蒸馏技术的稀疏检索模型,专为OpenSearch优化,支持免推理文档编码,在搜索相关性和效率上优于V1版本
文本嵌入
Transformers 英语

O
opensearch-project
1.8M
7
Sapbert From PubMedBERT Fulltext
Apache-2.0
基于PubMedBERT的生物医学实体表征模型,通过自对齐预训练优化语义关系捕捉
文本嵌入 英语
S
cambridgeltl
1.7M
49
Gte Large
MIT
GTE-Large 是一个强大的句子转换器模型,专注于句子相似度和文本嵌入任务,在多个基准测试中表现出色。
文本嵌入 英语
G
thenlper
1.5M
278
Gte Base En V1.5
Apache-2.0
GTE-base-en-v1.5 是一个英文句子转换器模型,专注于句子相似度任务,在多个文本嵌入基准测试中表现优异。
文本嵌入
Transformers 支持多种语言

G
Alibaba-NLP
1.5M
63
Gte Multilingual Base
Apache-2.0
GTE Multilingual Base 是一个多语言的句子嵌入模型,支持超过50种语言,适用于句子相似度计算等任务。
文本嵌入
Transformers 支持多种语言

G
Alibaba-NLP
1.2M
246
Polybert
polyBERT是一个化学语言模型,旨在实现完全由机器驱动的超快聚合物信息学。它将PSMILES字符串映射为600维密集指纹,以数值形式表示聚合物化学结构。
文本嵌入
Transformers

P
kuelumbus
1.0M
5
Bert Base Turkish Cased Mean Nli Stsb Tr
Apache-2.0
基于土耳其语BERT的句子嵌入模型,专为语义相似度任务优化
文本嵌入
Transformers 其他

B
emrecan
1.0M
40
GIST Small Embedding V0
MIT
基于BAAI/bge-small-en-v1.5模型微调的文本嵌入模型,通过MEDI数据集与MTEB分类任务数据集训练,优化了检索任务的查询编码能力。
文本嵌入
Safetensors 英语
G
avsolatorio
945.68k
29
精选推荐AI模型
Llama 3 Typhoon V1.5x 8b Instruct
专为泰语设计的80亿参数指令模型,性能媲美GPT-3.5-turbo,优化了应用场景、检索增强生成、受限生成和推理任务
大型语言模型
Transformers 支持多种语言

L
scb10x
3,269
16
Cadet Tiny
Openrail
Cadet-Tiny是一个基于SODA数据集训练的超小型对话模型,专为边缘设备推理设计,体积仅为Cosmo-3B模型的2%左右。
对话系统
Transformers 英语

C
ToddGoldfarb
2,691
6
Roberta Base Chinese Extractive Qa
基于RoBERTa架构的中文抽取式问答模型,适用于从给定文本中提取答案的任务。
问答系统 中文
R
uer
2,694
98
智启未来,您的人工智能解决方案智库
简体中文