This is a package of python code for creating, training and testing machine learning models for timeseries reconstruction and prediction tasks. Specifically, this code uses pytorch and pytorch-lightning to implement recon./pred. models for neural field potential (e.g. ECoG, LFP) signals datasets.
Currently implemented models included in this package:
- LFADS
- MVAR
Not currently licensed as of version 0.0.1