This repository contains deep baselines and three recent state-of-the-art models for the task of energy disaggregation implemented using NILMTK's Rapid Experimentation API and Pytorch lightning.
The repository facilitates also the experiments management and model evaluation through the implementation of recent best practices in DL discipline. It performs automatic management of experiments with MLflow and allows automatic hyperparameters optimisation with Optuna.
Deep-NILMtk was originally developped to enable PyTorch community from interacting with NILMtk API and benifiting from its features. However, it was further developed to include more recent models and sofisticated evaluation practices from DL.
It impelements a variety of features such as CV and hyper-param optimisation with minimal coding to facilitate prototyping and speed up the reseach process.
pip install .
Check the documentation for more details.
For any enquiries, please contact the main authors.