Intuitive library to help with training neural networks in PyTorch

ignis is a high-level library that helps you write compact but full-featured training loops with metrics, early stops, and model checkpoints for deep learning library PyTorch.

With ignis you can solve Supervised Learning and Deep Reinforcement Learning problems.

You can extend ignis according to your own needs. You can implement custom functionalities by extending simple abstract classes.

Deep Reinforcement Learning algorithms

1. Deep Q Network (DQN)
2. Deep Deterministic Policy Gradients (DDPG)

Installation

1. Install PyTorch. You can find it here: PyTorch
2. pip install ignis

Examples

You can find examples in examples/ directory

You can also run examples: python examples/iris.py

You might want to export PYTHONPATH=/path/to/this/directory

Contribute

1. Implement new functionalities
2. Improve code design
3. Improve comments and readme
4. Tests