/SEM2

This is a python 3 port of the SEM model with and updated to a more current version of tensor flow (2+)

Primary LanguageJupyter NotebookMIT LicenseMIT

SEM2

This the current working build of the Structured Event Memory (SEM) model, originally presented in Franklin et al., 2019 [pdf]. For the simulations that ocompany the paper, please see our earlier github repository https://github.com/ProjectSEM/SEM, which is maintained for archival purposes.

Tutorials and Demonstrations

The tutorials run in Google Colab and don't require local installation. These will download and install the SEM2 package

Installation Requirements

This package has been tested in Python v3.7.9 and tensorflow 2.3.1

To install, use the command:

pip install git+https://github.com/nicktfranklin/SEM2

SEM can then be used in python as a package. To import SEM:

>>import sem

The demonstrations require addtional plotting libraries (found in the requirements.txt file). These can be installed via

pip install -r requirements.txt