MXNet

Apache MXNet (https://github.com/apache/incubator-mxnet) is a deep learning framework designed for both efficiency and flexibility. _It _allows you to mix symbolic and imperative programming to maximize _efficiency and productivity. At its core, MXNet contains a dynamic dependency scheduler that automatically parallelizes both symbolic and imperative operations on the fly. A graph optimization layer on top of that makes symbolic execution fast and memory efficient. MXNet is portable and lightweight, scaling effectively to multiple GPUs and multiple machines.

Here are some steps you can follow when getting started with MxNet.

1. Understanding the architecture of MXNet http://mxnet.io/architecture/index.html
2. Installing MXNet on different platforms http://mxnet.io/get_started/install.html
3. Checking some basic concepts about MXNet http://mxnet.io/tutorials/index.html
4. Going through common questions about MXNet e.g., how to train multiple GPUs http://mxnet.io/how_to/index.html
5. Exploring more about MXNet http://mxnet.io/api/python/index.html

Some Useful Links

Gluon: A more efficient interface for MXNet http://gluon.mxnet.io/

Some deep learning projects based on MXNet

1. Face detection: https://github.com/tornadomeet/mxnet-face
2. Object detection: https://github.com/zhreshold/mxnet-ssd
3. Machine Translation: https://github.com/awslabs/sockeye