Documentation

Beginners guide

• Installation
• Quick start

Layers

This part of the documentation helps to understand how to combine layers into networks. It covers basics and some of the more advanced topics. These topics are relevant only for the deep learning models (a.k.a models with constructible architectures).

• Layers
  • Basics
  • Mutlilayer Perceptron (MLP)
  • Convolutional Neural Networks (CNN)
  • Parallel connections
  • Create custom layers
  • Parameter Initialization Methods
  • Debug layer connections
• Syntax sugar
  • Subnetworks
  • Layer definitions
  • Replicate layers and networks
• Surgery
  • Slice networks
  • Find layers by name
• Mixing NeuPy with Tensorflow
  • Network’s inputs and outputs
  • Propagate inputs for training
  • Access variables
  • Putting everything together

Training algorithms

These documentation articles help to understand what types of algorithms NeuPy supports and how to train neural networks. More advanced topics cover some special cases that might be useful for experienced users.

• Training Algorithms
  • Basics
  • Algorithms with constructible architecture
  • Train network on GPU
  • Signals and Early stopping
  • Scikit-learn compatibility
• Transfer Learning

Extra features

• Storage for Neural Networks
  • Save and load layer parameters
  • Save and load algorithms
  • Save and load networks with fixed architectures
• Visualizations
  • Training progress
  • Network structure
  • Saliency Map
  • Hinton diagram

Search

Install NeuPy

pip install neupy

Learn more about NeuPy reading tutorials and documentation.

Issues and feature requests

If you find a bug or want to suggest a new feature feel free to create an issue on Github