Comparing Elixir and Python when working with Simple Neural Networks

Speaker:
Adolfo Neto


Speaker:
Lucas C. Tavano


Abstract:

With a wide range of libraries focused on the machine learning market, such as TensorFlow, NumPy, Pandas, Keras, and others, Python has made a name for itself as one of the main programming languages. In February 2021, José Valim and Sean Moriarity published the first version of Numerical Elixir (Nx), a library for tensor operations written in Elixir. Nx aims to allow the language to be a good choice for GPU-intensive operations.

This talk aims to compare Python and Elixir when training convolutional neural networks using MNIST and CIFAR-10 datasets as examples, analyzing development experience, and performance difference.

OBJECTIVES

  • Teach about the new Nx library for Elixir
  • Compare Nx and Keras by the use of resources and time to train similar neural networks
  • Talk about the experience of developing in both languages and how different this experience was.

AUDIENCE

  • Numerical Elixir (Nx) interested people
  • Python AI developers
  • Data scientists

Level:

Tags:
Innovation,Nx,Performance