Code for visualizing the loss landscape of neural nets

POT : Python Optimal Transport

KErnel OPerationS, on CPUs and GPUs, with autodiff and without memory overflows

Riemannian Adaptive Optimization Methods with pytorch optim

Keras Generative Adversarial Networks

Fast CUDA implementation of (differentiable) soft dynamic time warping for PyTorch

Python implementation of soft-DTW.

An implementation of SoftDTW for PyTorch.

Pytorch Custom CUDA kernel for searchsorted

Code for Optimal Transport for structured data with application on graphs

python script for running long/parallel simulations (either by direct execution or using oarsub)