Glossary

L’Attention de Requête Groupée est une stratégie d’optimisation des modèles de langage, permettant un calcul conjoint et rapide des requêtes, pour une efficacité maximale.

The Nvidia H100 is a top-tier, AI-optimized computing device tailored for large-scale data centers. Its remarkable performance, scalability, and security ensure optimum HPC results.

Hallucination in AI is when large language models like chatbots or vision tools generate illogical, inaccurate outputs not based on training, often due to faulty decoding.