Andre HolznerinTowards Data ScienceDifferentiable convolution of probability distributions with TensorflowConvolution operations in Tensorflow are designed for tensors but can also be used to convolute differentiable functionsMay 6, 2018May 6, 2018
Andre HolznerLearning discontinuous functions with PyTorchIn this article we look at an example how PyTorch can be used to learn a discontinuous function. We do this by using a combination of…Dec 26, 20171Dec 26, 20171
Andre HolznerLSTM cells in PyTorchThis is an annotated illustration of the LSTM cell in PyTorch (admittedly inspired by the diagrams in Christopher Olah’s excellent blog…Oct 8, 20171Oct 8, 20171
Andre HolznerLearning a function with a variable number of inputs with PyTorchFascinated by the idea that building dynamic neural networks can be constructed easily with PyTorch, I decided to give it a try.Jul 9, 20172Jul 9, 20172