07:45 < AlexNguyenGitter> Hi, I have a question, if I want to test a file individually how can I do that without rebuilding the whole source. For example: i modify the linear regression test file, and just want to run that file.

07:48 < RishabhGarg108Gi> Hey @rxng8 , if you will do the make command after doing any changes then it will only build the changed files. After that, you can test it whatever way you like.

07:49 < AlexNguyenGitter> Oh nice! Thanks

13:07 < EmmanuelLykosGit> Hello, I'm Manos(Emmanuel) and I'm new to mlpack wishing to contribute in the future. May I have a good luck!!!

13:07 < zoq> EmmanuelLykosGit: Hello there, if you have any questions please don't hestiate to ask.

13:08 < zoq> EmmanuelLykosGit: Also we have a get involved page - https://www.mlpack.org/community.html that could help you get started.

13:09 < turska79Gitter[m> I got some out of heap errors as well with Visual studio 2019, but they went away by switching build tool architecture to 64-bit, default was 32

13:14 < turska79Gitter[m> But my question is how to apply QLearning to Snake game. Or which layers to use and what sizes. Now i give it vision of 24 values, info from every 45 degree andgle about distance to wall, food, body. plus its location x,y and direction vector. Learning process doesnt seem to be getting anywhere. I used cartpole code as a starting point for mlpack related things.

16:27 < abernauer[m]> Writing VBA code for this EY assessment for the first time interesting experience.

16:31 < zoq> abernauer[m]: I hope it's a good experience :)

16:35 < zoq> Hello everyone, we have a casual video meetup in about 25 minutes with no particular agenda, feel free to join.

16:36 < RishabhGarg108Gi> On which platform ?

16:37 < zoq> Right, for more information see the Video Meetup section on the community page - https://www.mlpack.org/community.html

16:37 < RishabhGarg108Gi> okay.

17:34 < zoq> turska79Gitter[m: Interesting, so maybe we should set it to 64-bit as default.

17:36 < zoq> turska79Gitter[m: Did you modify the network for the Cartpole example? The network used on that is very small, so maybe not the best one to solve the env.

17:37 < zoq> turska79Gitter[m: The network used in https://github.com/mlpack/examples/blob/master/reinforcement_learning_gym/bipedal_walker_sac.ipynb might be better.

17:37 < zoq> turska79Gitter[m: But there are a bunch of settings you could tune, that could also help.

17:40 < rcurtin> what, weird, why did the meeting end?

17:41 < zoq> no idea

17:41 < rcurtin> did I need to be logged in?

17:41 < zoq> I guess we can rejoin

17:41 < rcurtin> yeah

17:41 < AakashkaushikGit> the timer ended

17:41 < AakashkaushikGit> there was a timer above

17:41 < RishabhGarg108Gi> it said that the host didn't arrive so the meeting ended

18:12 < RishabhGarg108Gi> Hey, I want to make a PR. Can someone guide me how to do it.

18:26 < zoq> RishabhGarg108Gi: https://github.com/firstcontributions/first-contributions should be helpful, there is nothing special about opening a PR in mlpack.

18:26 < abernauer[m]> Is the meeting still going?

18:27 < zoq> abernauer[m]: It's usually for one hour, so no.

18:29 < abernauer[m]> ok thanks for the info. Got news from the recruiter that I should know if I have a job offer by the end of next week on a positive note.

18:30 < RishabhGarg108Gi> I mean i am actually new to open source, so I just wanted to know that is their any specific practice that we follow here for creating a PR or it is just I work on a separate branch and then just push it to my fork and create a PR from there.

19:18 < AakashkaushikGit> Hey @RishabhGarg108 basically you can create a PR from the branch you are working in.

19:19 < AakashkaushikGit> so need need to push those changed to the master in your forked repository of mlpack if that was what you were referring to.

20:47 < pradkrish> yeah, I agree. People doing machine learning are spoilt with DL frameworks like tensorflow that naturally uses GPU for high computer operations. Being able to do that will add a lot of value to mlpack and armadillo as well.

20:49 < pradkrish> sorry that was a wrong window ... faceinpalm

20:57 < zoq> pradkrish: Nah, you are right adding GPU support will add value.