06:02 < rcurtin> oops, looks like docker is broken on masterblaster

06:02 < rcurtin> I'll see if I can fix it tomorrow

06:47 < Kunvar_thaman> Hi all! I'm Kunvar Thaman - an Electrical and electrical engineering sophomore and I'm working on the 'Essential Deep Learning modules' project

06:49 < Kunvar_thaman> So far I'm going through NN code, built MLPack and am exploring how to implement GANs first of all

08:36 < niteya> Hello , I am sorry if this is a bit theoretical but what is the formula of SGD with momentum ? The mlpack implementation follows the MIT book http://www.deeplearningbook.org/contents/optimization.html (Pg 293) but other online material - https://towardsdatascience.com/stochastic-gradient-descent-with-momentum-a84097641a5d , http://ruder.io/optimizing-gradient-descent/index.html#momentum . Many more resources show either of these two

08:36 < niteya> methods .

08:37 < niteya> shows subtracting velocity as well*

08:39 < jenkins-mlpack2> Project docker mlpack nightly build build #211: FAILURE in 3 hr 25 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/211/

11:05 < zoq> rcurtin: Can we restart jenkins once more?

12:13 < zoq> Kunvar_thaman: Hello and welcome, sounds like a good start, let us know if we should clarify anything.

12:17 < zoq> niteya: http://ruder.io/optimizing-gradient-descent/index.html#momentum describes the same idea, "It does this by adding a fraction γ of the update vector of the past time step to the current update vector:"

12:17 < zoq> niteya: We call the "update vector of the past time step" velocity, 'momentum * velocity' and momentum is the fraction term.

12:17 < zoq> niteya: Let me know if anything I said was helpful.

13:58 < sumedhghaisas> zoq: Hey Marcus. Hows it going? I was thinking of mentoring a project fully focused on models repo. Trying to implement many NN applications. Would you be interested in co-mentoring this?

14:01 < sumedhghaisas> Currently there is only basic VAE on MNIST implemented by Atharva. Would be cool to have simple GANs modelling MNIST as well.

14:02 < zoq> sumedhghais: I'm in, there might be room for RL methods as well.

14:02 < sumedhghaisas> Great.

14:03 < sumedhghaisas> I will try to create an entry right now on projects page

14:03 < sumedhghaisas> and try to write down model ideas

14:04 < sumedhghaisas> but I am not sure which GAN implementations are already in so many you can update the list

14:04 < sumedhghaisas> and also RL

14:04 < sumedhghaisas> how many implementations do you think are reasonable?

14:39 < Abhishek_Chakrav> Hi, I'm Abhishek. I would like to work with mlpack. I have already built mlpack and will start resolving issues soon. I'll post any queries I have here soon.

14:56 < zoq> sumedhghais: I think that depends on the model itself, something like an inception network might take longer, so I guess the idea should reflect that.

14:57 < zoq> Abhishek_Chakrav: Hello and welcome, sounds good :)

14:57 < zoq> sumedhghais: I can add some RL ideas if you like.

14:59 < sreenik> Hello, I am new to this community and want to contribute to it. While trying to fix a bug, I realised that the documentation is not quite complete. I think I can discover faster if I work on the documentation for a while. However, are the python bindings adapted to python 3?

15:23 < zoq> sreenik: Hello, the python bindings should work with python3, travis does build against python and python3: https://travis-ci.org/mlpack/mlpack

15:43 < sreenik> Okay, that's nice

17:06 < rcurtin> zoq: restarted

17:54 < jenkins-mlpack2> Project docker mlpack nightly build build #212: ABORTED in 17 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/212/

18:10 < jenkins-mlpack2> Project docker mlpack nightly build build #213: ABORTED in 15 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/213/

18:12 < jenkins-mlpack2> Project docker mlpack nightly build build #214: ABORTED in 1 min 38 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/214/

18:17 < jenkins-mlpack2> Project docker mlpack nightly build build #215: STILL FAILING in 1 min 4 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/215/

18:19 < jenkins-mlpack2> Project docker mlpack nightly build build #216: STILL FAILING in 22 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/216/

18:21 < jenkins-mlpack2> Project docker mlpack nightly build build #217: STILL FAILING in 50 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/217/

18:24 < jenkins-mlpack2> Project docker mlpack nightly build build #218: ABORTED in 2 min 32 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/218/

18:25 < jenkins-mlpack2> Project docker mlpack nightly build build #219: STILL FAILING in 1 min 2 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/219/

18:51 < jenkins-mlpack2> Project docker mlpack nightly build build #220: STILL FAILING in 1 min 4 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/220/

18:54 < jenkins-mlpack2> Project docker mlpack nightly build build #221: STILL FAILING in 1 min 5 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/221/

19:09 < jenkins-mlpack2> Project docker mlpack nightly build build #222: STILL FAILING in 1 min 11 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/222/

19:32 < trancenoid> Hello everyone, I am Shivansh Beohar, I have set up mlpack on my windows pc and tried the sample app please suggest a starting point to get my hands dirty on mlpack.

19:36 < jenkins-mlpack2> Project docker mlpack nightly build build #223: ABORTED in 23 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/223/

19:46 < jenkins-mlpack2> Project docker mlpack nightly build build #224: ABORTED in 10 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/224/

19:54 < jenkins-mlpack2> Project docker mlpack nightly build build #225: ABORTED in 7 min 49 sec: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/225/

20:22 < zoq> rcurtin: Looks like we have the same problem as described here: https://groups.google.com/forum/#!topic/jenkinsci-users/--aquDzIA0Q

20:22 < zoq> rcurtin: Not sure it's feasible to use java 1.8?