V1.0.0

V1.0.0

• feed forward neural network: 1 input -> x hidden -> 1 output layer
• xavier initialisation
• cuda/cpu training
• simple and scalable
• easy to use
• easy to extend