Notes: MXNet
title
MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems
notes
- Imperative:
the user specifies exactly “how” computation needs to be performed
e.g. Matlab, numpy
- Declarative:
the user specification focuses on “what” to be done
e.g. Caffe, CXXNet
- Comparison of Frameworks: