3 answers

How to learn pytorch and fastai

Updated Jaipur, Rajasthan, India

Currently I do most of my projects in keras and tensorflow. But now I want to switch to pytorch as it seems to be more flexible. #engineering #computer-science # #student

3 answers

Dinesh’s Answer


Deep learning continues to be the hottest thing in data science. Deep learning frameworks are changing rapidly. It is often taught in a bottom-up manner, requiring that you first get familiar with linear algebra, calculus, and mathematical optimization before eventually learning the neural network techniques. This can take years, and most of the background theory will not help you to get good results, fast. Instead, a top-down approach can be used where first you learn how to get results with deep learning models on real-world problems and later learn more about how the methods work.

Dinesh recommends the following next steps:

  • https://towardsdatascience.com/simplifying-deep-learning-with-fast-ai-37aa0d321f5e
  • https://www.pluralsight.com/courses/getting-started-nlp-deep-learning-pytorch-fastai
  • https://github.com/fastai/fastai

Dhairya’s Answer

if you're interested in fastai, go through the fastai course which available for free https://course.fast.ai/. The course introduces pytorch and their fastai wrapper library.

The docs for the fastai library can be found here: https://docs.fast.ai/

Jason’s Answer

Updated Boxborough, Massachusetts

Hi Pawan,

I'm not an expert in this field. However, I did find a few free resources online.

  1. Deep Learning with Pytorch: A 60 minute Blitz
  2. Tutorials available on pytorch.org
  3. A free Intro to Deep Learning with PyTorch from Udacity