End to End Neural Art

This is an implementation of blog: http://dmlc.ml/mxnet/2016/06/20/end-to-end-neural-style.html

We will release a Multi-GPU training code soon.

How to use

1. First use download.sh to download pre-trained model and sample inputs

2. Then prepare training dataset according to the blog

3. Modify boost_train.py

Pretrained Model

Weight https://github.com/dmlc/web-data/raw/master/mxnet/art/model.zip Inference boost_inference.py