Home

Awesome

Replication of the basic model (No full context embeddings) found in the paper "Matching Networks for One Shot Learning" (http://arxiv.org/pdf/1606.04080.pdf). The currect version only works with the Omniglot dataset (included). Additionally, the current version doesn't utilize a seperate validation set.

To run, simply clone the repo, cd into the directory, and enter "python3 match_net.py"

The only prerequites are:

-Python 3

-Tensorflow r.9

-numpy

This code has only been tested on Ubuntu 16.4 and Mac OS X, but should probably work on Windows as well.