datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
nan-nli
Dataset Card for [Dataset Name]
Dataset Summary
[More Information Needed]
Supported Tasks and Leaderboards
Natural Language Inference
Text Classification
Languages
en
Dataset Structure
Data Instances
Data Fields
premise:
hypothesis:
label:
Data Splits
Evaluation: 258 samples
Dataset Creation
Curation Rationale
Extracting samples corresponding to different linguistics constructions of… See the full description on the dataset page: https://huggingface.co/datasets/joey234/nan-nli.all-nli-tr
Dataset Card for AllNLITR
This dataset is a formatted version of NLI-TR datasets, sharing the same licenses. The format is intended to be in line with AllNLI by Sentence Transformers for ease of training.
Despite originally being intended for Natural Language Inference (NLI), this dataset can be used for training/finetuning an embedding model for semantic textual similarity.
Dataset Subsets
pair-class subset
Columns: "premise", "hypothesis", "label"… See the full description on the dataset page: https://huggingface.co/datasets/emrecan/all-nli-tr.nli-veridicality-transitivity@inproceedings{yanaka-etal-2021-exploring,
title = "Exploring Transitivity in Neural {NLI} Models through Veridicality",
author = "Yanaka, Hitomi and
Mineshima, Koji and
Inui, Kentaro",
booktitle = "Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume",
year = "2021",
pages = "920--934",
}
help-nlihttps://github.com/verypluming/HELP
@InProceedings{yanaka-EtAl:2019:starsem,
author = {Yanaka, Hitomi and Mineshima, Koji and Bekki, Daisuke and Inui, Kentaro and Sekine, Satoshi and Abzianidze, Lasha and Bos, Johan},
title = {HELP: A Dataset for Identifying Shortcomings of Neural Models in Monotonicity Reasoning},
booktitle = {Proceedings of the Eighth Joint Conference on Lexical and Computational Semantics (*SEM2019)},
year = {2019},
}
Arabic-NLi-Triplet
Arabic NLI Triplet
Dataset Summary
The Arabic Version of SNLI and MultiNLI datasets. (Triplet Subset)
Originally used for Natural Language Inference (NLI),
Dataset may be used for training/finetuning an embedding model for semantic textual similarity.
Triplet Subset
Columns: "anchor", "positive", "negative"
Column types: str, str, str
Examples:
{
"anchor": "شخص على حصان يقفز فوق طائرة معطلة",
"positive": "شخص في الهواء الطلق، على حصان.",
"negative":… See the full description on the dataset page: https://huggingface.co/datasets/Omartificial-Intelligence-Space/Arabic-NLi-Triplet.vimedaqa-nli-g1tomi-nlitomi dataset (theory of mind question answering) recasted as natural language inference
https://colab.research.google.com/drive/1J_RqDSw9iPxJSBvCJu-VRbjXnrEjKVvr?usp=sharing
@article{sileo2023tasksource,
title={tasksource: Structured Dataset Preprocessing Annotations for Frictionless Extreme Multi-Task Learning and Evaluation},
author={Sileo, Damien},
url= {https://arxiv.org/abs/2301.05948},
journal={arXiv preprint arXiv:2301.05948},
year={2023}
}… See the full description on the dataset page: https://huggingface.co/datasets/tasksource/tomi-nli.Arabic-NLi-Pair-Score
Arabic NLI Pair-Score
Dataset Summary
The Arabic Version of SNLI and MultiNLI datasets. (Pair-Score Subset)
Originally used for Natural Language Inference (NLI),
Dataset may be used for training/finetuning an embedding model for semantic textual similarity.
Pair-Class Subset
Columns: "sentence1", "sentence2", "score"
Column types: str, str, float
Arabic Examples:
{
"sentence1": "شخص على حصان يقفز فوق طائرة معطلة",
"sentence2": "شخص يقوم… See the full description on the dataset page: https://huggingface.co/datasets/Omartificial-Intelligence-Space/Arabic-NLi-Pair-Score.Arabic-NLi-Pair-Class
Arabic NLI Pair-Class
Dataset Summary
The Arabic Version of SNLI and MultiNLI datasets. (Pair-Class Subset)
Originally used for Natural Language Inference (NLI),
Dataset may be used for training/finetuning an embedding model for semantic textual similarity.
Pair-Class Subset
Columns: "premise", "hypothesis", "label"
Column types: str, str, class with {"0": "entailment", "1": "neutral", "2": "contradiction"}
Arabic Examples:
{
"premise": "شخص… See the full description on the dataset page: https://huggingface.co/datasets/Omartificial-Intelligence-Space/Arabic-NLi-Pair-Class.Arabic-NLi-Pair
Arabic-NLI-PAir
Dataset Summary
The Arabic Version of SNLI and MultiNLI datasets. (Pair Subset)
Originally used for Natural Language Inference (NLI),
Dataset may be used for training/finetuning an embedding model for semantic textual similarity.
Pair Subset
Columns: "anchor", "positive"
Column types: str, str
Examples:
{
"anchor": "كيف أكون جيولوجياً جيداً؟",
"positive": "ماذا علي أن أفعل لأكون جيولوجياً عظيماً؟"
}
Disclaimer
Please note… See the full description on the dataset page: https://huggingface.co/datasets/Omartificial-Intelligence-Space/Arabic-NLi-Pair.temporal-nli@inproceedings{thukral-etal-2021-probing,
title = "Probing Language Models for Understanding of Temporal Expressions",
author = "Thukral, Shivin and
Kukreja, Kunal and
Kavouras, Christian",
booktitle = "Proceedings of the Fourth BlackboxNLP Workshop on Analyzing and Interpreting Neural Networks for NLP",
month = nov,
year = "2021",
address = "Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url =… See the full description on the dataset page: https://huggingface.co/datasets/tasksource/temporal-nli.squadid-nli
Dataset Card for SQuAD-ID-NLI
Dataset Summary
The SQuAD-ID-NLI dataset is derived from the SQuAD-ID question answering dataset, utilizing named entity recognition (NER), chunking tags, Regex, and embedding similarity techniques to determine its contradiction sets.
Collected through this process, the dataset comprises various columns beyond premise, hypothesis, and label, including properties aligned with NER and chunking tags.
This dataset is designed to facilitate… See the full description on the dataset page: https://huggingface.co/datasets/muhammadravi251001/squadid-nli.figurative-nli @inproceedings{chakrabarty-etal-2021-figurative,
title = "Figurative Language in Recognizing Textual Entailment",
author = "Chakrabarty, Tuhin and
Ghosh, Debanjan and
Poliak, Adam and
Muresan, Smaranda",
booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
month = aug,
year = "2021",
address = "Online",
publisher =… See the full description on the dataset page: https://huggingface.co/datasets/tasksource/figurative-nli.SDOH-NLI
Dataset Card for Dataset Name
This dataset card aims to be a base template for new datasets. It has been generated using this raw template.
Dataset Details
Dataset Description
Curated by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by [optional]: [More Information Needed]
Language(s) (NLP): [More Information Needed]
License: [More Information Needed]
Dataset Sources [optional]
Repository: [More… See the full description on the dataset page: https://huggingface.co/datasets/davanstrien/SDOH-NLI.idkmrc-nli
Dataset Card for IDK-MRC-NLI
Dataset Summary
The IDKMRC-NLI dataset is derived from the IDK-MRC question answering dataset, utilizing named entity recognition (NER), chunking tags, Regex, and embedding similarity techniques to determine its contradiction sets.
Collected through this process, the dataset comprises various columns beyond premise, hypothesis, and label, including properties aligned with NER and chunking tags.
This dataset is designed to facilitate Natural… See the full description on the dataset page: https://huggingface.co/datasets/muhammadravi251001/idkmrc-nli.TRIAGEsyntactic-augmentation-nlihttps://github.com/Aatlantise/syntactic-augmentation-nli/tree/master/datasets
@inproceedings{min-etal-2020-syntactic,
title = "Syntactic Data Augmentation Increases Robustness to Inference Heuristics",
author = "Min, Junghyun and
McCoy, R. Thomas and
Das, Dipanjan and
Pitler, Emily and
Linzen, Tal",
booktitle = "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
month = jul,
year = "2020",
address =… See the full description on the dataset page: https://huggingface.co/datasets/tasksource/syntactic-augmentation-nli.nli_tr_for_simcse
NLI-TR for Supervised SimCSE
This dataset is a modified version of NLI-TR dataset. Its intended use is to train Supervised SimCSE models for sentence-embeddings. Steps followed to produce this dataset are listed below:
Merge train split of snli_tr and multinli_tr subsets.
Find every premise that has an entailment hypothesis and a contradiction hypothesis.
Write found triplets into sent0 (premise), sent1 (entailment hypothesis), hard_neg (contradiction hypothesis) format.
See this… See the full description on the dataset page: https://huggingface.co/datasets/emrecan/nli_tr_for_simcse.newsph-nli-fil-pairclassificationref: https://huggingface.co/datasets/jcblaise/newsph_nli
contradictory-watson-nli
🔬 Contradictory, My Dear Watson — Multilingual NLI Solution
Kaggle Competition | Model: mDeBERTa-v3-base-mnli-xnli
📋 Problem
Given a premise and hypothesis in one of 15 languages, predict the relationship:
0 = Entailment (hypothesis follows from premise)
1 = Neutral (hypothesis is possible but not certain)
2 = Contradiction (hypothesis contradicts premise)
🏗️ Approach
Model
mDeBERTa-v3-base-mnli-xnli — 279M params, already fine-tuned on MultiNLI… See the full description on the dataset page: https://huggingface.co/datasets/Rock2346/contradictory-watson-nli.attempto-nliNatural language inference using attempto controlled english
Paper to come
@inproceedings{fuchs2012first,
title={First-order reasoning for attempto controlled english},
author={Fuchs, Norbert E},
booktitle={Controlled Natural Language: Second International Workshop, CNL 2010, Marettimo Island, Italy, September 13-15, 2010. Revised Papers 2},
pages={73--94},
year={2012},
organization={Springer}
}
SynCSE-scratch-NLI
Dataset Card for Dataset Name
Dataset Summary
The SynCSE-scratch-NLI is a Natural Language Inference dataset generated by GPT-3.5-Turbo. You can use it to learn better sentence representation with contrastive learning. More details can be found in paper and code
Supported Tasks and Leaderboards
Natural Language Inference
Contrastive Learning of Sentence Embeddings
Languages
English
Dataset Structure
Data Instances
[More… See the full description on the dataset page: https://huggingface.co/datasets/hkust-nlp/SynCSE-scratch-NLI.tydiqaid-nli
Dataset Card for TyDI-QA-ID-NLI
Dataset Summary
The TyDI-QA-ID-NLI dataset is derived from the TyDI-QA-ID question answering dataset, utilizing named entity recognition (NER), chunking tags, Regex, and embedding similarity techniques to determine its contradiction sets.
Collected through this process, the dataset comprises various columns beyond premise, hypothesis, and label, including properties aligned with NER and chunking tags.
This dataset is designed to… See the full description on the dataset page: https://huggingface.co/datasets/muhammadravi251001/tydiqaid-nli.breaking_nli_fr
Dataset Card for Breaking_NLI-FR
Dataset Details
Dataset Description
This repository contains a machine-translated, manually verified French version of the Breaking_NLI dataset, originally written in English. Breaking_NLI is a dataset meant for testing methods trained to solve the natural language inference task, requiring some lexical and world knowledge.
Curated by: [More Information Needed]
Funded by [optional]: [More Information Needed]
Shared by… See the full description on the dataset page: https://huggingface.co/datasets/maximoss/breaking_nli_fr.multilingual-nli-datasetUse it like this.
from datasets import load_dataset, Dataset, DatasetDict
import pandas as pd
data_files = {"train": "multilingual_nli_train_df.csv",
"validation": "multilingual_nli_validation_df.csv",
"test": "multilingual_nli_test_df.csv"}
dataset = load_dataset("muhammadravi251001/multilingual-nli-dataset", data_files=data_files)
selected_columns = ["premise", "hypothesis", "label"]
# selected_columns = dataset.column_names['train'] # Uncomment this line to… See the full description on the dataset page: https://huggingface.co/datasets/muhammadravi251001/multilingual-nli-dataset.idk_mrc_nli_chunkingYou can download this Dataset just like this:
data_files = {"train": "data_nli_train_chunking_df.csv",
"validation": "data_nli_val_chunking_df.csv",
"test": "data_nli_test_chunking_df.csv"}
dataset = load_dataset("muhammadravi251001/idk_mrc_nli_chunking", data_files=data_files)
This is some modification from IDK-MRC dataset to IDK-MRC-NLI dataset. By convert QAS dataset to NLI dataset. You can find the original IDK-MRC in this link:… See the full description on the dataset page: https://huggingface.co/datasets/muhammadravi251001/idk_mrc_nli_chunking.SynCSE-partial-NLI
Dataset Card for Dataset Name
Dataset Summary
The SynCSE-scratch-NLI is a Natural Language Inference dataset generated by GPT-3.5-Turbo. You can use it to learn better sentence representation with contrastive learning. More details can be found in paper and code
Supported Tasks and Leaderboards
Natural Language Inference
Contrastive Learning of Sentence Embeddings
Languages
English
Dataset Structure
Data Instances
[More… See the full description on the dataset page: https://huggingface.co/datasets/hkust-nlp/SynCSE-partial-NLI.SDOH-NLISDOH-NLI is a natural language inference dataset containing ~30k premise-hypothesis pairs with binary entailment labels in the domain of social and behavioral determinants of health.
@misc{lelkes2023sdohnli,
title={SDOH-NLI: a Dataset for Inferring Social Determinants of Health from Clinical Notes},
author={Adam D. Lelkes and Eric Loreaux and Tal Schuster and Ming-Jun Chen and Alvin Rajkomar},
year={2023},
eprint={2310.18431},
archivePrefix={arXiv}… See the full description on the dataset page: https://huggingface.co/datasets/tasksource/SDOH-NLI.nli-high-quality
NLI High-Quality Balanced Dataset
A combined, filtered, and class-balanced natural language inference (NLI)
dataset built from MNLI, SNLI, FEVER-NLI, and ANLI, intended for fine-tuning
NLI models for use in zero-shot text classification via the entailment trick
(hypothesis = "This example is about {label}.").
The goal of this dataset was quality and generalization over raw volume:
rather than concatenating the four source datasets as-is, several filtering
stages were applied to… See the full description on the dataset page: https://huggingface.co/datasets/Pankaj8922/nli-high-quality.MultiCSR_NLIThis is an unofficial synthetic dataset generated using the methodology provided by MultiCSR (https://arxiv.org/pdf/2310.10962). We utilize the premise component of NLI to synthesize both positive and hard negative samples.