datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
ru_paraphraser
Dataset Card for ParaPhraser
Dataset Summary
ParaPhraser is a news headlines corpus annotated according to the following schema:
1: precise paraphrases
0: near paraphrases
-1: non-paraphrases
The Plus part is also available.
It contains clusters of news headline paraphrases labeled automatically by a fine-tuned paraphrase detection BERT model.In order to load it:
from datasets import load_dataset
corpus = load_dataset('merionum/ru_paraphraser', data_files='plus.jsonl')… See the full description on the dataset page: https://huggingface.co/datasets/merionum/ru_paraphraser.jailbreakbench-paraphrase-2025-08
JailbreakBench Paraphrase Dataset (2025-08)
This dataset contains 115 paraphrased prompts from the JailbreakBench dataset, created as part of research on semantic entropy-based jailbreak detection and its robustness to paraphrasing.
Dataset Description
Overview
This dataset was created to test the robustness of jailbreak detection methods (particularly semantic entropy) to paraphrased inputs. The paraphrases were generated from the original JailbreakBench… See the full description on the dataset page: https://huggingface.co/datasets/DhruvTre/jailbreakbench-paraphrase-2025-08.style-aware-paraphraser-author-bank-reddit
Style-Aware Paraphraser — Reddit Author Targets Bank
A bank of 12 000 anonymous Reddit authors, each represented by 16 exemplar
comments plus 5 Mistral-7B paraphrases of each. This is what feeds the
target-style side of our paraphraser: pick a row, pass reference_text
and paraphrase_reference_text to
rrivera1849/style-aware-paraphraser-mistral7b,
and the model will rewrite any machine text in that author's style.
Reddit usernames are not included; the bank carries only the… See the full description on the dataset page: https://huggingface.co/datasets/rrivera1849/style-aware-paraphraser-author-bank-reddit.chatgpt-paraphrases-kz
Kazakh Paraphrasing Dataset
This dataset is specifically designed for the paraphrasing task in the Kazakh language. It offers a unique resource for natural language processing applications, focusing on the development and evaluation of paraphrasing models.
Source and Translation Process
Originally sourced from humarin/chatgpt-paraphrases, this dataset has been translated using Google Translate.
Dataset Content and Structure
The dataset comprises 5.44 million… See the full description on the dataset page: https://huggingface.co/datasets/CCRss/chatgpt-paraphrases-kz.style-aware-paraphraser-outputs
Style-Aware Paraphraser Outputs
⚠️ Intended for evaluating machine-text detectors, not for training them.
Each row is a (human, machine, adversarial-paraphrase) triple drawn from a
specific evaluation slice of three domains. The author bank that produced
these outputs overlaps with the rows here, so training a detector against
this data would not generalize. Use it to score a detector, not to fit one.
Final outputs of the style-aware paraphraser from
Attacks on Machine-Text… See the full description on the dataset page: https://huggingface.co/datasets/rrivera1849/style-aware-paraphraser-outputs.nlp-paraphrases-27kcluebbers__Llama-3.1-8B-paraphrase-type-generation-apty-sigmoid-details
Dataset Card for Evaluation run of cluebbers/Llama-3.1-8B-paraphrase-type-generation-apty-sigmoid
Dataset automatically created during the evaluation run of model cluebbers/Llama-3.1-8B-paraphrase-type-generation-apty-sigmoid
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/cluebbers__Llama-3.1-8B-paraphrase-type-generation-apty-sigmoid-details.nlp-paraphrases-20k-prunedThe dataset contains 20703 records. The dataset was created by removing all dataset items from the original 27k dataset that had a BLEU score 0 or more than 0.3388.
en-wiki-paraphrased
Paraphrase-pairs from English Wikipedia
As used in paper More Room for Language: Investigating the Effect of Retrieval on Language Models.
We utilized the English Wikipedia as a clean and information-rich text corpus.
Due to the cost of paraphrasing, we selected only the top 10% most visited articles by page view count in the last year (about 400 million words).
The paraphrases were generated by a prompted instruction-tuned Mistral 7B language
model. We used the… See the full description on the dataset page: https://huggingface.co/datasets/ltg/en-wiki-paraphrased.ml-paraphrase-tr
Turkish ML Paraphrase Dataset
This dataset contains 60,000 Turkish sentence pairs designed for paraphrase detection and semantic similarity tasks. It is suitable for training and evaluating machine learning models, particularly in binary classification and fine-tuning scenarios.
📦 Dataset Structure
Each entry consists of:
sentence1: First sentence
sentence2: Second sentence
label:
1 → paraphrase (semantically similar)
0 → not paraphrase (semantically unrelated)… See the full description on the dataset page: https://huggingface.co/datasets/dogukanvzr/ml-paraphrase-tr.Armenian-Paraphrase-PC
Armenian Paraphrase Detection Corpus
This data is orinally from https://github.com/ivannikov-lab/arpa-paraphrase-corpus
BibTeX Citation
If you use this dataset, please cite following paper:
@misc{malajyan2020arpa,
title={ARPA: Armenian Paraphrase Detection Corpus and Models},
author={Arthur Malajyan and Karen Avetisyan and Tsolak Ghukasyan},
year={2020},
eprint={2009.12615},
archivePrefix={arXiv},
primaryClass={cs.CL}
}… See the full description on the dataset page: https://huggingface.co/datasets/asparius/Armenian-Paraphrase-PC.sentence-paraphrases
Sentence Paraphrases Dataset
This dataset is a curated collection of sentence-length paraphrases derived from two primary sources:
humarin/chatgpt-paraphrases
xwjzds/paraphrase_collections.
Dataset Details
Dataset Description
The dataset is structured to provide pairs of sentences from an original text and its paraphrase(s). For each entry:
The "text" field contains the least readable paraphrase.
The "paraphrase" field contains the most readable paraphrase.… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/sentence-paraphrases.paraphrases
Датасет перефразировок коротких фраз (читчат+поэзия)
В датасете содержатся правильные и некорректные перефразировки коротких диалоговых реплик (проект диалоговой системы)
и фрагментов стихов (проект генеративной поэзии).
Датасет представляет из себя список сэмплов-кортежей. Каждый сэмпл состоит из двух списков:
paraphrases - примеры правильных перефразировокdistractors - примеры неправильных перефразировок
Датасет используется для создания моделей детектора перефразировок… See the full description on the dataset page: https://huggingface.co/datasets/inkoziev/paraphrases.DPO-arxiv_paraphraseSO-Python_basics_QA-filtered-2023-T5_paraphrased-tanh_scoresentence-paraphrases-2wikipedia-paragraphs-direct-paraphrases
Wikipedia Paragraphs Direct Paraphrases
Paraphrases of Wikipedia paragraphs using AI large language models.
Paragraphs from agentlans/wikipedia-paragraphs-complete sample_k10000 and sample_k50000 splits
Paraphrased using Qwen/Qwen3.5-9B and a distilled Qwen/Qwen3-4B-Instruct-2507 with the following prompt:
Rewrite the following paragraph entirely in your own words while preserving every fact, detail, meaning, nuance, and level of specificity. Do not add, remove, reinterpret… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/wikipedia-paragraphs-direct-paraphrases.nlp-paraphrases-europarl-llamacontextual_faithfulness_openbookqa_qwen_paraphrase_variantsHebrew-Paraphrase-DatasetHebrew Paraphrase Dataset
This repository contains a high-quality paraphrase dataset in Hebrew, consisting of 9785 instances.
The dataset includes both paragraph-level (75%) and sentence-level (25%) paraphrases generated with the help of a large language model.
Among these, 300 instances have been manually validated as gold standard examples.
What Is a Paraphrase?
A paraphrase is a restatement of a text using different words and structures while preserving the original meaning.
Dataset… See the full description on the dataset page: https://huggingface.co/datasets/HebArabNlpProject/Hebrew-Paraphrase-Dataset.MetaMath_Paraphrase_Answerauggemma4b_paraphrased_eagle_cotparaphrased_qwen1.5b_dragon_numswikipedia-paragraph-paraphrasesparaphraseTraincontextual_faithfulness_openbookqa_llama_paraphrase_variantsqwen7b_paraphrased_cat_numsqwen1.5b_paraphrased_dragon_cotllama8b_paraphrased_wolf_cotPolish-paraphrases-12K-synthetic12K sentence paraphrases in Polish.
First, lot of various sentences were generated using ChatGPT 5.2 and Gemini 3 Thinking, then all of them were paraphrased by GPT-4o.
Sentences range from short to medium to long, and are not only random general sentences but also cover tech topics and topics related to universities and student life.
Please mind the license terms for content generated by the AI models mentioned when using this dataset. ChatGPT itself belongs to OpenAI, Gemini to Google.
12… See the full description on the dataset page: https://huggingface.co/datasets/Wojtekb30/Polish-paraphrases-12K-synthetic.