pi-tau

vae

Pytorch implementation of a Variational Autoencoder trained on CIFAR-10. The encoder and decoder modules are modelled using a resnet-style U-Net architecture with residual blocks.

transformer

The Transformer model implemented from scratch using PyTorch. The model uses weight sharing between the embedding layers and the pre-softmax linear layer. Training on the Multi30k machine translation task is shown.

pi-tau.github.io

pixelcnn

PixelCNN implemented with PyTorch <3

playing-with-RL-models

Single-file implementations of reinforcement learning algorithms

deep-conv-nets

Modern deep convolutional neural networks implemented with PyTorch <3. ResNet, DenseNet and InceptionNet are trained on the CIFAR-10 dataset and the results are compared.

dl-from-scratch

entanglement-control

Supplementary code for the Master thesis "Developing Artificial Intelligence Agents to Manipulate Quantum Entanglement"

fun-with-algorithms

realnvp

PyTorch implementation of the RealNVP model

machine-translation

NMT with RNNs implemented with PyTorch <3

made

PyTorch implementation of the MADE model

max-ent

random-walk-in-AI

rubiks-cube

test