spragunr / overcoming-catastrophic

Implementation of "Overcoming catastrophic forgetting in neural networks" in Tensorflow

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

This is an implementation of "Overcoming catastrophic forgetting in neural networks" (https://arxiv.org/abs/1612.00796) for supervised learning in TensorFlow.

model.py defines a simple fully-connected network and methods to compute the diagonal of the Fisher information matrix.

experiment.ipynb trains and tests a single network on three MNIST classification tasks sequentially (i.e., once the network begins training on a given task, it is never exposed to previous task training data again).

About

Implementation of "Overcoming catastrophic forgetting in neural networks" in Tensorflow


Languages

Language:Jupyter Notebook 78.3%Language:Python 21.5%Language:Shell 0.2%