---
license: cc-by-4.0
language:
- fra
task_categories:
- summarization
---

# Description
Dataframe containing 949 French books in `txt` format.  
More precisely :
- the `titre` column contains the title of the book
- the `auteur` column contains the author's name and dates of birth and death (if you want to filter the texts to keep only those from a given century to the present day; for example, Cligès and Erec et Enide are in very early French)
- the `résumé_wiki` column contains a summary of the book from French Wikipedia
- the `résumé_wiki_en` column contains a summary of the book from English Wikipedia
- the `résumé_autre` column contains a summary of the book from a site other than Wikipedia (the license for this column is very uncertain, which is why we put this dataset behind a gate)
- the `texte` column contains texts
- the `nb_words` column contains an estimate of the number of words per text (a simple `.split(" ")`; again if you want to filter)

All available summaries (`resumé_wiki` + `résumé_wiki_en` + `résumé_autre`) represent 1513 lines.

Assuming an average word length of 1.3 to 1.5 tokens, we have :
- Estimated number of texts with at least 2048 tokens: between 940 and 945
- Estimated number of texts with at least 4096 tokens: between 871 and 879
- Estimated number of texts with at least 8192 tokens: between 833 and 844
- Estimated number of texts with at least 16384 tokens: between 729 and 750
- Estimated number of texts with at least 32768 tokens: between 591 and 634
- Estimated number of texts with at least 65536 tokens: between 409 and 444
- Estimated number of texts with at least 131072 tokens: between 202 and 230
- Estimated number of texts with at least 262144 tokens: between 53 and 81