This is a repository of implementations of deep learning papers, along with mini projects and experiments. Most are written using a mix of TFLearn and Tensorflow.
Basic descriptions of the projects. They are all commented with the relevant information, so if you're curious about anything, check there. If it's not there, feel free to ask me!
DeXpression- Implementation of the DeXpression paper: https://arxiv.org/abs/1509.05371
GRU Cart- A network that utilizes gated recurrent units to solve OpenAI's cartpole environment.
Attention Mech- A function made in Tensorflow that replicates the attention mechanism described in http://www.cs.cmu.edu/~./hovy/papers/16HLT-hierarchical-attention-networks.pdf
seq2seq- A basic sequence2sequence function using LSTM cells without attention that's based off of the Tensorflow tutorial with some minor changes and work arounds for issues with Deepcopy.
Q-Learn Func- A basic function for experimenting with Q-Learning and the involved hyperparameters.