GDOPT - General Dynamic Optimizer: Adaptive Mesh Refinement for Collocation-Based Dynamic Optimization

Implementation of back-propagation through different convex optimization algorithms - gradient descent, gradient descent with momentum, Nesterov's accelerated momentum, RMSProp and ADAM - for a deep neural network

mao ni new

Gerencia de Desarrollo Organizacional, Procesos y Tecnología

raphyy

GDOPT - General Dynamic Optimizer: Adaptive Mesh Refinement for Collocation-Based Dynamic Optimization. This repository contains additional developments regarding training of Neural ODEs.

No description available