In this talk, we’ll cover creation of a multilayer perceptron model using gluon and MXNet’s new NumPy-compatible functions, a port of the classic NumPy with GPU accelerations and additional features for deep learning.
All the code snippets shown during the talk are available at https://github.com/haojin2/PyData-LA-Demo
Diving into deep learning requires understanding bulky new frameworks, which significantly increases the adoption curve for data scientists in industry. In this talk, we’ll cover creation of a multilayer perceptron model using gluon and MXNet’s new NumPy-compatible functions, a port of the classic NumPy with GPU accelerations and additional features for deep learning. These open source tools will give you a working foundation for building out more complicated models for real applications with faster performance and less hassle.