DConf Online '22 Q & A: Max Haughton
This is the Q & A session with Max Haughton that ran in the DConf Online 2022 Day One Q & A Livestream in parallel with the premiere of his Day One presentation video, 'Differentiable Programming in D'.
Viewers could ask questions in the video's live chat or in the livestream's live chat. Between questions, Mike Parker kept Max talking. This video is for those interested in more context on the talk and Max's thoughts related topics.
Max's talk is located here:
• DConf Online '22 - Differentiable Programm...
Feng Li's talk on Vlang (mentioned in the parting thoughts):
• DConf Online '22 Addendum: D-based Next Ge...
All of the links Max posted during the Q & A session
=========================================
The Enzyme Automatic Differentiation Framework
https://enzyme.mit.edu/
https://enzyme.mit.edu/explorer
The Differentiable Programming Manifesto (for Swift)
https://github.com/apple/swift/blob/m...
The Vectorflow D library
https://code.dlang.org/packages/vecto...
The grain D library
https://github.com/ShigekiKarita/grain
Book recommendation: The Elments of Statistical Learning
https://link.springer.com/book/10.100...
Book recommendation: Machine Learning: A Probabilistic Perspective
https://github.com/probml/pml-book
The tinygrad Python library
https://github.com/geohot/tinygrad
The LibNC C library
https://bellard.org/libnc/
Peter Norvig's response to Noam Chomsky on statistical learning
https://norvig.com/chomsky.html
The British Rail Class 802 Train
https://en.wikipedia.org/wiki/British...
=========================================
Learn more about DConf and the D Programming Language at:
https://dconf.org/
https://dlang.org/
#dlang #DConfOnline #dconf
TIMECODES:
00:00 Title & Introduction
01:12 What are some practical use cases for the D community?
02:31 Why wouldn't one just do the symbolic differentiation oneself?
04:47 Is this done in the front end of the intermediate code?
05:25 Where did Max compile the data for his travel speed example?
05:59 What are some interesting things Max didn't say in the talk?
07:34 Any suggestions for people with no background?
09:31 Any book recommendations?
10:19 Are there any plans for an easy-to-use, ML library?
11:52 About the tinygrad library
12:49 About LibNC
13:30 Does enzyme require a minimum version of LDC?
15:28 In a different world, would Max spend his time on this kind of work?
16:18 Max's parting thoughts, touching on Feng Li's talk about Vlang