Skip to content

Latest commit

 

History

History
23 lines (17 loc) · 676 Bytes

README.md

File metadata and controls

23 lines (17 loc) · 676 Bytes

The Role of Embedding Complexity in Domain-invariant Representations

This repository is the official PyTorch implementation of the algorithm Multilayer Divergence Minimization (MDM) proposed in the following paper:

Ching-Yao Chuang, Antonio Torralba, Stefanie Jegelka. The Role of Embedding Complecity in Domain-invariant Representations link

Environment

  • Pytorch 1.0
  • Python 2.7

Dataset

Download the MNIST-M dataset from Google Drive and unzip it.

cd dataset
tar -zvxf mnist_m.tar.gz

Training

Simply run

python main.py