clojure-cortex

getting error aswell. Is it possible to use only CPU and avoid the CUDA library? On Thursday, 15

I think a lot of the team has moved to supporting the Clojure library for MXNet. On Tue, Oct 9, 2018

I haven't done the Titantic one, but there is a cats and dogs in the example folder of the cortex

Hi, As the name suggests, infinite-class-balanced-dataset is an abstraction of an infinite dataset.

You'll find that clojure libraries mimic Unix command line tools, they stick to the "do one

Hi Nico, I'm new to ML and Cortex, I followed the example in the article. Is is possible to find

Indeed. NIPS is now the most heavily attended AI conference in the world. Somebody actually did a

Nice work sticking with it! If batch-norm helped that much then I would say potentially the inputs

I think that's that same ami that I have used before. I just took the defaults for the p2 and it

Hi Harold, Thanks for pointing out where the learning rate and optimiser can be set. I've also

I was sure you were behind that :-). Yes for sure, I intend to refactor cortex to remove the old (pre

Oh no! Bobby, I'll definitely be happy to help you get involved in contributing. Feel free to

Hi Rickesh, The last call in your `train-xor` function is to `execute/run`. `execute/run` returns

Glad to hear it. You are welcome. -Harold On Saturday, November 4, 2017 at 9:52:59 PM UTC-6, Tapojit

Hi Shawn, Sure, you can train networks on any input data you like. We have built models for genomic

Thank you for your explanation. I have worked it out. On Tuesday, October 17, 2017 at 11:34:32 PM UTC

Hey TongKe - Check out the unit tests for the tensors. Specifically, follow the from the tests: https

Hi Peter, Thanks for your interest. Experiment is a library of higher level functions that simplify

Reminder - NIPS registration opens today and it's a good bet that it will sell out quickly https:

Thank you! I suggest that we start with "simple RNN" rather than a LSTM (see diagram in

The performance is back. I checked out the xor example and did some regression on it. It looks fine

Hi, That error indicates that the port that the classification example is trying to use to display

"Machine learning" is a broad subject. "Deep learning" ie neural network is the

Boris, Thanks! I've linked this from the README, here: https://github.com/thinktopic/cortex/blob/

John, Thanks for your interest and for sending this over. I've added a cortex issue with some

Thanks folks! I'm really enjoying playing with Cortex, and I'm finding the MNIST example in

Nicolas* (sorry) :) On Wednesday, June 14, 2017 at 1:12:24 PM UTC-6, Harold wrote: Hi Nicholas,

Hey Harold, Boris: I replied on the issue. Agreed it would be good to have a network that takes

On Friday, May 19, 2017 at 1:44:43 PM UTC-6, John Newman wrote: Congrats on the 1.0!!! Not quite

Because 0.9.5 had some subtle bugs associated with multi-stream batch uploads. My apologies, this