Clojure from Scratch to GPU: A Simple Neural Network Training API

This series builds everything from matrix operations provided by Neanderthal [1].

If you have any questions about Clojure or numerical computing with Clojure/Java, or if you spot an error in the article, please discuss here.

I will definitively write a book. Please sign up for the newsletter at the bottom of the article if you'd like to be notified.

Can’t wait for you book to come out. For anyone looking to do some serious learning in this space I highly recommend this from dragon.rocks

> This series builds everything from matrix operations provided by Neanderthal

Do they perform convolutions as well?

Good idea for a cliffhanger: Do they perform convolutions as well? Follow the next episodes and watch the movie (the book) to find out!

drcode 7y ago ↗

This is like when my 4 year old daughter asked if she'll be able to have dessert and I used to say "maybe if you're lucky..."

She quickly figured out this always meant "yes".

That's so adorable

Well, I mean, you can write a convolution as a matrix operation. Except, perhaps that's not the most efficient way to do it.

That may be covered in the book.

... or not? Stay tuned :)

I'm on the edge of my seat :)

Clojure from Scratch to GPU: A Simple Neural Network Training API (dragan.rocks)