These datasets are used to evaluate models on French performance using: https://github.com/EleutherAI/lm-evaluation-harness (from CroissantLLM paper)
Manuel Faysse
manu
AI & ML interests
NLP, Privacy, multi-modal DL
Recent Activity
upvoted an article 6 days ago
BidirLM: Turning Generative LLMs into the Best Open-Source Omnimodal Encoders new activity 15 days ago
manu/bge-fr-en:Size of dataset new activity about 1 month ago
vidore/vidore_v3_finance_en:Upload multimodal_elements_enriched.jsonOrganizations
