3 papers
cs.LG2025
Purifying Shampoo: Investigating Shampoo's Heuristics by Decomposing its Preconditioner
Runa Eschenhagen, Aaron Defazio, Tsung-Hsien Lee +2
The recent success of Shampoo in the AlgoPerf contest has sparked renewed interest in Kronecker-factorization-based optimization algorithms for training neural networks. Despite it…
stat.ML2025
Spectral-factorized Positive-definite Curvature Learning for NN Training
Wu Lin, Felix Dangel, Runa Eschenhagen +3
Many training methods, such as Adam(W) and Shampoo, learn a positive-definite curvature matrix and apply an inverse root before preconditioning. Recently, non-diagonal training met…
cs.LG2024
Can We Remove the Square-Root in Adaptive Gradient Methods? A Second-Order Perspective
Wu Lin, Felix Dangel, Runa Eschenhagen +3
Adaptive gradient optimizers like Adam(W) are the default training algorithms for many deep learning architectures, such as transformers. Their diagonal preconditioner is based on…