Fully generalist synthetic dataset and SOTA small reasoners
AI & ML interests
Open Science LLMs
Recent Activity
The annotations and the golden subset for the HellaSwag validation set.
The first generation of models pretrained on Common Corpus.
Tools for de-toxifying public domain data, especially multilingual and historical text data and data with OCR errors.
PleIAs collection of models for the data processing of challenging document and data sources.
New generation of small reasoning models for RAG, search, and source summarization.
Models, tools and other research artifacts used for the training of the first series of Pleias models.
Largest multilingual pretraining data.
A large collection of multimodal financial documents in open data.
A multilingual dataset of public domain books and newspapers.
Fully generalist synthetic dataset and SOTA small reasoners
New generation of small reasoning models for RAG, search, and source summarization.
The annotations and the golden subset for the HellaSwag validation set.
Models, tools and other research artifacts used for the training of the first series of Pleias models.
The first generation of models pretrained on Common Corpus.
Largest multilingual pretraining data.
Tools for de-toxifying public domain data, especially multilingual and historical text data and data with OCR errors.
A large collection of multimodal financial documents in open data.
PleIAs collection of models for the data processing of challenging document and data sources.
A multilingual dataset of public domain books and newspapers.