News for hackers. Stuff that matters.
Re:Flare: Clojure Dynamic Neural Net Library by jgh on Friday December 08, 2017 @11:17AM (#15879714)
Very cool. As someone who wants to begin learning how to use ML and who enjoys clojure as a hobby language...this is great. Gives me an excuse to use Clojure more :) I hope you decide to continue working on it!
Re:Flare: Clojure Dynamic Neural Net Library by aria on Friday December 08, 2017 @10:04PM (#15884038)
I'll try! I've got a full time management job so this was a fun way to code again. The library is pretty small and I think easy to follow if you want to learn
Re:Flare: Clojure Dynamic Neural Net Library by thom on Saturday December 09, 2017 @10:30AM (#15886179)
This made me chuckle. If anybody remembers the dim and distant past of various singularitarian mailing lists on the internet, you might remember Eliezer Yudkowsky's theoretical 'annotative' language Flare[1]. It was one of the early steps on the (what was then quite straight in some people's minds) line to AGI.

As for this, possibly worthwhile for the authors to look into Dragan Djuric's work, he's done some really fast matrix stuff on Clojure[2].

1: http://yudkowsky.net/obsolete/plan.html#tech_comp_flare

2: https://github.com/uncomplicate/neanderthal

Re:Flare: Clojure Dynamic Neural Net Library by thom on Saturday December 09, 2017 @10:31AM (#15886188)
And I read the article and see Neantherthal's being used (stupidly thought it wasn't given a comment downthread).
Re:Flare: Clojure Dynamic Neural Net Library by nickbauman on Friday December 08, 2017 @01:47PM (#15880874)
This is for Dynamic Neural Nets, which do not need back propagation. The benchmarks are amazing.
Re:Flare: Clojure Dynamic Neural Net Library by stochastic_monk on Friday December 08, 2017 @01:59PM (#15880985)
What is the optimization algorithm for these if not backpropagation? This page does mention the use of hyperparameters for optimization and compares itself to an adaptive stochastic gradient descent method, which makes me think that it is using backpropagation.
Re:Flare: Clojure Dynamic Neural Net Library by aria on Friday December 08, 2017 @09:53PM (#15883993)
Definitely doing the standard back-prop algorithm. Check out the code, it's pretty straightforward and the whole lib is about 2,500 lines
Re:Flare: Clojure Dynamic Neural Net Library by nickbauman on Friday December 08, 2017 @03:01PM (#15881523)
Ber ber ber, I think you're right. I was assuming a Cascade-Correlation DNN.
Re:Flare: Clojure Dynamic Neural Net Library by aria on Friday December 08, 2017 @09:52PM (#15883988)
Author here, happy to answer any questions!
Re:Flare: Clojure Dynamic Neural Net Library by ghosthamlet on Friday December 08, 2017 @10:24PM (#15884109)
Thinks for your great lib in both great Clojure and ML. Do you know whether this Clojure lib can do Dynamic net: https://github.com/thinktopic/cortex? you didn't compare it in your blog.
Re:Flare: Clojure Dynamic Neural Net Library by aria on Friday December 08, 2017 @10:52PM (#15884192)
Yes this does dynamic nets, the model is the same as PyTorch, you build a graph for each input(s). I think Cortex is in the Keras-style of model (only layer abstractions and not "dynamic" in the sense we talk about neural nets) and isn't dynamic, but I could be missing something. The examples seem to all be in that mold: https://github.com/thinktopic/cortex/blob/master/examples/ca...
Re:Flare: Clojure Dynamic Neural Net Library by visarga on Friday December 08, 2017 @02:36PM (#15881322)
Needs more GPU.
Re:Flare: Clojure Dynamic Neural Net Library by aria on Friday December 08, 2017 @09:52PM (#15883986)
Author here. This is coming up. Will do OpenCL first and then CUDA.
Re:Flare: Clojure Dynamic Neural Net Library by nickbauman on Friday December 08, 2017 @02:59PM (#15881509)
True, I think it's on the "wish list", no?