01:40 < rcurtin> hmmm, seems like mlpack-bot isn't doing its job... so I closed some PRs for inactivity

01:40 < rcurtin> but actually it looks like it's doing nothing at all as far as stale issues go so I guess I will debug it...

01:44 < rcurtin> oh wait, I feel a bit stupid, it was explicitly avoiding issues that are in projects. well, easy fix...

05:26 < jenkins-mlpack2> Project docker mlpack weekly build build #57: STILL UNSTABLE in 7 hr 39 min: http://ci.mlpack.org/job/docker%20mlpack%20weekly%20build/57/

07:56 < jenkins-mlpack2> Project docker mlpack nightly build build #391: STILL UNSTABLE in 3 hr 42 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/391/

07:56 < jenkins-mlpack2> Ryan Curtin: Try to get more output for why it's taking so long.

11:30 < abhijeet> Hi Everyone!

11:32 < zoq> abhijeet: Hello

11:33 < abhijeet> I am abhijeet and I'm a freshman at VIIT college of engineering

11:33 < abhijeet> I want to learn more about mlpack and how can i learn to contribute to mlpack

11:34 < abhijeet> as i only know C++ and a bit of python

11:34 < abhijeet> so in this current scenario what should i primarily do!!!

11:37 < zoq> abhijeet: mlpack has bindings to other languages like python, but the library is written in C++; about getting involved https://www.mlpack.org/community.html should answer some questions, https://www.mlpack.org/gsoc.html might be interesting as well.

11:38 < abhijeet> Ok! Thats something cool.

11:41 < abhijeet> is it necessary to learn linux

11:44 < zoq> Most members use linux or mac to build and test the code, but mlpack builds on windows as well. there is also an docker images that could be helpful.

11:44 < abhijeet> so primarily theres no use of windows inhere

11:48 < zoq> abhijeet: I'm sure someone uses mlpack on windows but it's not the most used environment. We do have an example project for VS (https://github.com/mlpack/mlpack/tree/master/doc/examples/sample-ml-app) and also have a build pipeline that tests everything on windows (https://ci.appveyor.com/project/mlpack/mlpack).

11:51 < abhijeet> Ohhkay

11:51 < abhijeet> how to learn to contribute in mlpack

11:52 < abhijeet> like how can onne solve a bug????

11:56 < zoq> abhijeet: So either you find an interesting issue on GitHub that you like to solve, another idea might be to just take a method you are already familiar with that mlpack implements and see if you can find something that could be improved.

11:56 < zoq> abhijeet: Going through some exsisting code is defently a good starting point.

11:56 < zoq> abhijeet: We are always open for new methods as well, so if you like to contribute something new, that's also a good starting point.

11:57 < abhijeet> ohkayyy

13:02 < KimSangYeon-DGU> Hi, Sumedh!

13:02 < KimSangYeon-DGU> sumedhghaisas: I'm ready :)

13:03 < sumedhghaisas> KimSangYeon-DGU: Hey Kim

13:03 < sumedhghaisas> How are things?

13:03 < KimSangYeon-DGU> I've organized our researches

13:04 < sumedhghaisas> Yeah I saw :)

13:04 < sumedhghaisas> I am currently opening all the things

13:04 < sumedhghaisas> just me just a minute

13:05 < sumedhghaisas> For some reason my net s super slow today

13:06 < sumedhghaisas> anyways... Did you get a chance to organize the research with objective results and conclusions?

13:07 < KimSangYeon-DGU> Yeah, you can see the paper in researches/NLL vs NLL with constraint

13:08 < KimSangYeon-DGU> I tested NLL and NLL with constraint again for better check.

13:08 < sumedhghaisas> okay let me try restrart my connection. I am trying to open that document for ges now

13:09 < KimSangYeon-DGU> Yeah

13:19 < sumedhghaisas> KimSangYeon-DGU: Hey

13:19 < KimSangYeon-DGU> Hi!

13:20 < sumedhghaisas> okay great... sorry for the network issues

13:20 < sumedhghaisas> okay so I went through the documents

13:20 < sumedhghaisas> have couple of comments about them

13:20 < KimSangYeon-DGU> Yeah

13:21 < sumedhghaisas> I would add more experiments and results in 'Validity of objective function' itself

13:21 < sumedhghaisas> I would add all the results of NLL with constraint in it

13:21 < sumedhghaisas> how alpha goes to zero

13:21 < KimSangYeon-DGU> Yeah, I'll reflect on it

13:22 < KimSangYeon-DGU> I'll do that

13:22 < sumedhghaisas> how constraint is not been able to satisfy sometimes

13:22 < sumedhghaisas> Also I don't think there is good reason to compare NLL with constraint to NLL without constraint

13:22 < KimSangYeon-DGU> Yeah, I'll write more details

13:22 < KimSangYeon-DGU> Ah

13:23 < KimSangYeon-DGU> Can you tell me the reason?

13:23 < sumedhghaisas> I would rather compare NLL with constraint with GMM

13:23 < KimSangYeon-DGU> Okay

13:23 < sumedhghaisas> ahh I mean NLL without constraint is not theoretically sound

13:23 < KimSangYeon-DGU> Ah~

13:23 < KimSangYeon-DGU> I agree

13:24 < KimSangYeon-DGU> as it is unnormalized

13:24 < sumedhghaisas> cause the optimizer has no idea about the constraint which is major part of the investigation

13:24 < sumedhghaisas> correct

13:24 < sumedhghaisas> showing results of that we won't be able to justify

13:24 < KimSangYeon-DGU> Right, I totally agree

13:24 < KimSangYeon-DGU> I think I missed

13:24 < sumedhghaisas> lets concentrate our efforts on adding as many details possible in 'Validity of objective function'

13:25 < KimSangYeon-DGU> Yeah

13:39 < sakshamB> ShikharJ: I am here

16:24 < ShikharJ> sakshamB: Toshal: I'll have to apologize for this time, I had an emergency in the morning, hence I couldn't connect. We can have this talk tomorrow if you guys want?

17:24 < sakshamB> ShikharJ: hope everthing is fine. We can have the discussion tomorrow. I will be starting to work on CGAN now.