bourdoiscatie's picture
Update README.md
2caaa37
metadata
language:
  - fr
license: cc-by-sa-4.0
size_categories:
  - 100K<n<1M
task_categories:
  - text-generation
tags:
  - DFP
  - french prompts
annotations_creators:
  - found
language_creators:
  - found
multilinguality:
  - monolingual
source_datasets:
  - orange_sum

orange_sum_fr_prompt_text_generation_from_an_article

Summary

orange_sum_fr_prompt_text_generation_from_an_article is a subset of the Dataset of French Prompts (DFP).
It contains 539,400 rows that can be used for a text generation task.
The original data (without prompts) comes from the dataset orange_sum by Eddine et al.
A list of prompts (see below) was then applied in order to build the input and target columns and thus obtain the same format as the xP3 dataset by Muennighoff et al.

Prompts used

List

24 prompts were created for this dataset. The logic applied consists in proposing prompts in the indicative tense, in the form of tutoiement and in the form of vouvoiement.

'"'+document+'"\n Continuer le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Continue le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Continuez le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Poursuivre le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Poursuis le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Poursuivez le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Prolonger le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Prolonge le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Prolongez le texte sur 1000 caractères maximum :',  
'"'+document+'"\n Rédiger la suite du texte : ',  
'"'+document+'"\n Rédige la suite du texte : ',  
'"'+document+'"\n Rédigez la suite du texte : ',  
'"'+document+'"\n Imaginer la suite du texte : ',  
'"'+document+'"\n Imagine la suite du texte : ',  
'"'+document+'"\n Imaginez la suite du texte : ',  
'"'+document+'"\n Ecrire la suite du texte : ',  
'"'+document+'"\n Ecris la suite du texte : ',  
'"'+document+'"\n Ecriver la suite du texte : ',  
'"'+document+'"\n Développer la suite du texte : ',  
'"'+document+'"\n Développe la suite du texte : ',  
'"'+document+'"\n Développez la suite du texte : ',  
'"'+document+'"\nGénérer la suite du texte : ',  
'"'+document+'"\nGénère la suite du texte : ',  
'"'+document+'"\n Générez la suite du texte : ',  

Features used in the prompts

In the prompt list above, text and targets have been constructed from:

orange_sum = load_dataset('orange_sum','abstract')
if len(orange_sum['train'][i]['text']) > 1000:
        document = orange_sum['train'][i]['text'][:1000]
targets = orange_sum['train'][i]['summary'][1000:]

Splits

  • train with 472,944 samples
  • valid with 33,096 samples
  • test with 33,360 samples

How to use?

from datasets import load_dataset
dataset = load_dataset("CATIE-AQ/orange_sum_fr_prompt_text_generation_from_an_article")

Citation

Original data

@article{eddine2020barthez, title={BARThez: a Skilled Pretrained French Sequence-to-Sequence Model}, author={Eddine, Moussa Kamal and Tixier, Antoine J-P and Vazirgiannis, Michalis}, journal={arXiv preprint arXiv:2010.12321}, year={2020} }

This Dataset

@misc {centre_aquitain_des_technologies_de_l'information_et_electroniques_2023,
author = { {Centre Aquitain des Technologies de l'Information et Electroniques} },
title = { DFP (Revision 1d24c09) },
year = 2023,
url = { https://huggingface.co/datasets/CATIE-AQ/DFP },
doi = { 10.57967/hf/1200 },
publisher = { Hugging Face }
}

License

CC-BY-SA-4.0