Clojurians Log v2
Clojure programming
Channels
Apps
numerical-computing
qqq 2017-03-25T06:06:10.998904Z
active channel here 🙂
qqq 2017-03-25T06:06:14.999031Z
so my choices right now are:
qqq 2017-03-25T06:06:53.000457Z
(1) use Hy + tensorflow (2) try to get Tensorflow 1.0's compleetly undocumented java api to work (3) use Clojure + Cortex (4) use Clojure + neathenaderal (only provides blas) (5) use osmething else
qqq 2017-03-25T06:06:56.000583Z
suggestions ?
2017-03-25T08:53:40.387290Z
@qqq A few comments regarding option 4, although you didn't write what you are trying to do:
2017-03-25T08:54:24.389034Z
1) Neanderthal also supports LAPACK from the upcoming version 0.9.0 that is going to be released in a few days.
2017-03-25T08:55:04.390457Z
2) Neanderthal also supports straightforward communication with GPUs (OpenCL based, though)
2017-03-25T08:55:38.391846Z
3) I am going to create ClojureCUDA soon (a couple of months I guess)
2017-03-25T08:56:21.393439Z
4) In one of the next versions, Neanderthal will also have CUDA engine and easy transfer to CUDA
2017-03-25T08:56:50.394570Z
5) Then, cuDNN is a (relatively) straightforward library away...
qqq 2017-03-25T17:32:06.109066Z
@blueberry: sorry, my XY problem is: I am doing deep learning in Hy/Tensorflow
qqq 2017-03-25T17:32:20.110122Z
I'm sick and tired of Hy being sorta clojure but not really clojure and want my entire stack in clojure.
qqq 2017-03-25T17:32:42.111873Z
I'm willing to give up numpy and it's entire ecosystem -- I use mostly tensorflow, not so much numpy