Anthill Inside 2018

On the current state of academic research, practice and development regarding Deep Learning and Artificial Intelligence.

Make your own DL framework

Submitted by Nithish Divakar (@nithishdivakar) on Monday, 14 May 2018

videocam_off

Technical level

Intermediate

Section

Full talk

Status

Submitted

Vote on this proposal

Login to vote

Total votes:  +12

Abstract

We have all used all the high end frameworks that works really well. How about writing a small strip down version of one. In this session, I’ll walk you through how to write a small Deep Learning Framework in Pure python and numpy which has auto grad and optimizers and easy to create models. Also, the framework will be extendible so that you can easily play around with.

Outline

  • Computational graph and how it is usefull in moder day deep learning.
  • Bare basics Computations needed for Deep Learning
  • Implenting the forward pass
  • Implementing the auto grad
  • The complete tiny frame

Requirements

Basic Knowledge in using Python and numpy.
Basic knowledge in any of the modern day deep learning frameworks.

Speaker bio

Nithish Divakar is a Computer Vision Research Engineer at Cogknit Semantics. He Is a masters graduate from IISc bangalore and has been working in deep learning for past 4 years. He has prior published research work in GANs and Image captioning. See everythingproject.in for more.

Comments

  • 1
    Charlotte Treesa George (@charlotte) 10 months ago

    go ahead!

  • 1
    Akash J123 (@akashjp) 10 months ago

    Great initiative Nitesh.

  • 1
    Arya Sasi 10 months ago

    all d bst da

  • 1
    Subair t (@subairt) 10 months ago

    ഇജ്ജ് പോളിക്ക് മുത്തെ

  • 1
    REMEEZ K.RASACK (@remeez0802) 10 months ago

    all the best man…

  • 1
    Zainab Bawa (@zainabbawa) Reviewer 9 months ago

    Nitish, this looks like a workshop proposal. Submit slides detailing outline and a preview video by 7 June.

  • 0
    rinlexx (@rinlexx) 10 months ago

    – .- -.- . .. - .... .- .–. .–. . -.

Login with Twitter or Google to leave a comment