Skip to content

Latest commit

 

History

History
23 lines (15 loc) · 640 Bytes

README.md

File metadata and controls

23 lines (15 loc) · 640 Bytes

Continual Learning for Deep Discriminative Models

Requirement: Tensorflow 1.4.0.

To run the Permuted MNIST experiment:

python run_permuted.py

To run the Split MNIST experiment:

python run_split.py

The printed results are matrices where each row i contains the test set accuracies on all previously observed tasks after seeing task i. The average accuracy will also be plotted and saved to results/. They should be similar to the following figures:

Permuted MNIST

Split MNIST