Scholar
Edouard Oyallon
Google Scholar ID: Y8XGVkYAAAAJ
CNRS, Sorbonne University
Citations & Impact
Citations
5,295
H-index
18
i10-index
22
Publications
20
Co-authors
22
list available
Contact
Publications
Resume (English only)
Academic Achievements
- - Publication: Decentralized Asynchronous Optimization with DADAO allows Decoupling and Acceleration, JMLR 2025
- - Preprints/Technical Reports: LO: Compute-Efficient Meta-Generalization of Learned Optimizers, Model Parallelism With Subnetwork Data Parallelism, etc.
- - Projects: SHARP project (PEPR partner, 2023), ADONIS project (funded by ANR, 2022, PI), VHS (collaborator, 2022), CoCa4AI (collaborator, 2022)
Research Experience
- - CNRS Researcher, MLIA team, Sorbonne University
- - Flatiron Institute (CCM)
- - Ecole Polytechnique (DepMap)
- - CentraleSupélec (Opis)
- - INRIA Lille (SequeL), Advisor: Michal Valko
- - Ecole Normale Supérieure (DATA), Advisor: Stéphane Mallat
Education
- Worked at Flatiron Institute (CCM), Ecole Polytechnique (DepMap), CentraleSupélec (Opis), INRIA Lille (SequeL) with Michal Valko, Ecole Normale Supérieure (DATA) with Stéphane Mallat, and ENS Cachan, campus de Ker Lann.
Background
- A CNRS researcher in the MLIA team at Sorbonne University. Research interests include the foundations of machine learning techniques, symmetries of deep neural networks, and developing algorithms for large-scale distributed and decentralized training.
Miscellany
- No specific personal interests mentioned; Contact: edouard.oyallon@cnrs.fr; Looking for highly motivated students/colleagues passionate about optimizing and accelerating the training of massive LLMs.
