04:30 < ShikharJaiswalGi> @khimraj Might be worthwhile to lookup existing PRs on GANs and see if you can help with those.

07:05 < himanshu_pathak[> Hey Everyone I want to implement PACGAN in mlpack should I start working on it from now or it will be too early to by the When I read the paper I like it. The idea of mitigating mode collapse. Right now working on my pr of neural Turing machine after debugging from last two days I realised that I have not added ResetCell() function in FFN it's so silly Mistake. I wasted my 3 days because of this

07:05 < himanshu_pathak[> Sigh..

07:38 < Param-29Gitter[m> Will we be having video chat today it has been 2 weeks :)?

08:02 < jenkins-mlpack2> Project docker mlpack nightly build build #630: STILL FAILING in 2 hr 48 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/630/

12:29 < rcurtin> Param-29Gitter[m: ah, thanks, I need to send that email

12:48 < rcurtin> PParam-29Gitter[m: it'll be tomorrow though, at 1800 UTC; that seems to be a much better time for people

12:49 < Saksham[m]> How do we signup for the video chat?

12:50 < rcurtin> Saksham[m]: here's the email I sent to the mlpack list: http://knife.lugatgt.org/pipermail/mlpack/2020-March/004202.html

12:53 < Saksham[m]> Thankssss

12:53 < Saksham[m]> Thanks

13:35 < zoq> himanshu_pathak[: If you feel comfortable with mlpack and the paper, please feel free, but please add a comprehensive test suite.

13:43 < himanshu_pathak[> Sure zoq I will take care of tests also.

14:30 < AbishaiEbenezerG> hey mlpack! quick question- considering that i'm relatively new to neural networks , how must i start coding my neural networks? Should i start with tensorflow, or could i directly try jumping into mlpack?

14:31 < AbishaiEbenezerG> also, considering that i'm trying to work my way to the mlpack codebase, what would be advisable to do?

14:41 < zoq> AbishaiEbenezerG: The internal design of tf and mlpack is different so if I like to work on mlpack I would look into the mlpack codebase.

14:42 < AbishaiEbenezerG> hey thanks!

14:43 < AbishaiEbenezerG> although, would experience with a framework like pytorch or tensorflow be helpful?

14:44 < zoq> AbishaiEbenezerG: absolutely

14:44 < AbishaiEbenezerG> got it ! thanks alot!

14:51 < AbhiSaphire> Hey, I have been looking on GSOC ideas for automatic binding of new languages. It seems like the documentation and reference links provided are not working. I wanted to work on implementing automatic binding for MATLAB, since I have coding experience in MATLAB. Can you please guide me how do I start. Thanks

15:21 < favre49> Sorry for pushing to the main website repo by mistake

15:23 < favre49> completely zoned out there :)

15:23 < zoq> AbishaiEbenezerG: Link updated.

15:25 < zoq> AbhiSaphire: Also, there are some informations about the bindings project in the mail archive, and there are two PR's open for go and java.

15:27 < AbishaiEbenezerG> @zoq sorry , i didn't catch you ...

15:27 < AbishaiEbenezerG> link..?

15:28 < rcurtin> favre49: no worries :)

15:29 < zoq> AbishaiEbenezerG: The link to the bindings documentation on the GSoC ideas page.

15:30 < zoq> Ahh, wrong name, sorry.

15:32 < AbishaiEbenezerG> for AbhiSaphire ig...

15:36 < favre49> Also, is GOVERNANCE.md meant to say that it is a draft, or it's revision date?

15:38 < rcurtin> favre49: oh, it shouldn't say (DRAFT), nice catch :)

15:38 < rcurtin> want me to open a PR to fix that?

15:38 < rcurtin> also I see I botched the markdown in "Code of Conduct"...

15:38 < favre49> I'm opening one to fix the license copyright year, I'll fix it

15:38 < rcurtin> cool, thank you so much :)

15:39 < favre49> The markdown in code of conduct looks fine though, where is the issue?

15:44 < rcurtin> I have the (here)[https://...] for the link backwards, it should be [here](https://...)

15:44 < favre49> Oh completely missed that

15:44 < rcurtin> apparently I did too :-D

15:44 < favre49> I think after joining mlpack I've become exceedingly pedantic about these things

15:46 < rcurtin> :)

15:46 < rcurtin> I don't think that's necessarily a bad thing :)

15:46 < rcurtin> but then, I am also very pedantic...

16:51 < Param-29Gitter[m> Hey @rcurtin i had an idea to speed up decision trees using openmp...can open a PR once i am done working on it?

16:54 < Param-29Gitter[m> Also should i used the gini_gain version that #2169 has or the original one?

17:02 < rcurtin> Param-29Gitter[m: you don't need my permission to work on something; if you make decision trees faster, that sounds great, but I have no idea how long it might be until I am able to review it

17:08 < SriramSKGitter[m> Speaking of pedantic fixes to markdown, I noticed shogun's github page has a small banner (?) showing 'powered by Numfocus' along with build status and code coverage banners.

17:08 < SriramSKGitter[m> I think that'd look cool on the mlpack page :)

17:10 < Param-29Gitter[m> Ok. I'll open a PR. once i am done.

17:13 < SriramSKGitter[m> Was talking about [this](https://pasteboard.co/IXyRjCb.png) if it wasn't clear.

17:15 < rcurtin> SriramSKGitter[m: NumFOCUS helped us set up the Github 'sponsor' button, so we have that plus the second paragraph mentions NumFOCUS in the readme

17:16 < rcurtin> now, I wouldn't be opposed to removing the NumFOCUS image from the readme and adding the kind of small banner you suggested to the list of banners we already have (e.g., right to the right of the 'license' one?)

17:16 < rcurtin> if you wanted to make that change I don't see any issue with it

17:18 < SriramSKGitter[m> Oh yeah, it definitely isn't necessary just a cosmetic improvement. I liked the aesthetic and wanted to see what you guys thought :)

17:18 < SriramSKGitter[m> Powered by NumFOCUS has a nice ring to it

17:20 < rcurtin> sure, why not give it a shot and open a PR, then we can see it? :)

17:21 < SriramSKGitter[m> Sounds good :)

17:40 < SriramSKGitter[m> #2256 let me know what you think :)