osushkov Goto Github PK

characterrnn

Character-level language modelling using a Recurrent Neural Network.

cuda_rnn

CUDA implementation of a Recurrent Neural Network.

cudafun

Playing with CUDA

cudann

CUDA implementation of a feed-forward neural network.

ddpg

Deep Deterministic Policy Gradient implementation

ddpg_fd

DDPG From Demonstrations

deephandwriting

Handwritten digit recognition using deep neural nets

dqn

Deep Q-network for playing connect-four

dqn_enhancements

DQN implementation with a bunch of enhancements to the core algorithm.

driving_rnn

Training an RNN to drive around a track using Reinforcement Learning.

experimental

experimental

explicitnn

A simple implementation of an explicit neural network with back propagation.

fractals

Messing around with C++, SDL, and Cairo to render some fractals.

greekcoursebuilder

gridrl

Grid game reinforcement learning

gym_dqn_extension

Apply extended DQN to some OpenAI gym tasks.

gym_simple

Simple learning task using the OpenAI gym.

handwriting

Handwriting recognition using a neural net.

logisticregression

Just testing out some logistic regression

mcts

Monte-Carlo Tree Search (MCTS) basic implementation.

mnist_draw

Neural Network trained to draw digit images.

osushkov.github.io

parallel_cuda

Trying out some constructs for running CUDA in parallel

phd_src

A backup of my old PhD related code

policy_gradient

Using a policy gradient algorithm to learn to play connect-four.

python_cpp

Trying to interface Python and C++ to get the good TensorFlow Python APIs in Python and the fast C++ simulation together.

rbm

Simple Restricted Boltzmann Machine implementation. Just to have a play around.

regressionlearning

rnn_optimised

Optimised cuda implementation of an RNN.

rnn_rl

Reinforcement Learning of a POMDP task using a Recurrent Neural Network.