This page is not created by, affiliated with, or supported by Slack Technologies, Inc.
2017-09-11
Channels
- # aleph (4)
- # beginners (68)
- # boot (21)
- # chestnut (1)
- # cljs-dev (72)
- # clojure (64)
- # clojure-austin (9)
- # clojure-dusseldorf (16)
- # clojure-gamedev (2)
- # clojure-italy (32)
- # clojure-russia (80)
- # clojure-spec (9)
- # clojure-uk (20)
- # clojurescript (105)
- # cursive (5)
- # data-science (5)
- # datomic (23)
- # defnpodcast (3)
- # emacs (22)
- # fulcro (2)
- # graphql (63)
- # hoplon (7)
- # lein-figwheel (17)
- # lumo (63)
- # mount (2)
- # nrepl (4)
- # off-topic (66)
- # om (6)
- # onyx (3)
- # portkey (54)
- # re-frame (12)
- # reagent (12)
- # specter (42)
- # uncomplicate (1)
- # unrepl (38)
- # vim (9)
- # yada (3)
I would like to write neural network (DMN+) in Clojure or possible create with Cortex? What do you think, how to start? (I saw the code in py.. but I would like to improve.. and for me more natural the Clojure language) .. What do you think?
@sb, you could certainly start with Cortex, but at present it doesn’t have any recurrent layers (in master branch at least), so you won’t be able to implement a dynamic memory network (if I understand “DMN” correctly) without writing those yourself. If you wanted to implement the whole thing from scratch, I would probably look at Neanderthal for fast linear algebra.
Yes, I thought dynamic memory network. Thanks!! 👍 in this case I start implement from scratch.
I saw this presentation now. 15:37 - he created a neural network visualisation. I can’t find for this github repo. Could you help what did he use in this presentation for visualisation?
I think, I found it: https://github.com/nuroko/nurokit