datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
roots_fr_hal_archives_ouvertesROOTS Subset: roots_fr_hal_archives_ouvertes
Licensing
Each row of the dataset corresponds to a specific document in the HAL Open Archive and is individually licensed: the corresponding identifier (HAL Id) is in the meta column.
BigScience processing steps
Filters applied to: fr
dedup_document
dedup_template_soft
filter_remove_empty_docs
remove_references_fr
filter_small_docs_bytes_1024
bok_words_700krx_report