05:38 < say4n> Hmm, I think these sort of spammy PRs might be bigger problem at projects that have a very low barrier of entry

05:40 < say4n> But, yes, at the same time DO's program does indeed help with getting a larger population of people get involved with OSS year over year. :)

08:37 < AakashkaushikGi4> Hi, this is a question: just like PyTorch and Tensorflow have model creation that allow them to create models by sub classing or sequentially by adding layers does mlpack also have something similar, i know we can add layers sequentially in mlpack but the sub classing method allows for great flexibility and even allows things like model stacking for ensemble of models. And adding two model outputs from somewhere in

08:37 < AakashkaushikGi4> between and all that kind of things.

08:45 < zoq> AakashkaushikGi4: There is the Concat layer.

08:52 < PulkitGitter[m]> Hey guys, I had a Question : I wanted to know if mlpack is planning to implement GCNs and GNN?

08:54 < zoq> PulkitGitter[m]: There are no plans yet, if you like to work on this please feel free :)

09:08 < AakashkaushikGi4> @zoq that is correct but suppose there is a complete model and i want to redefine the forward function for the layers can i do that ?

09:20 < zoq> AakashkaushikGi4: You can always inherit from a class/layer reimplement/adjust the method and use the layer the same way.

09:33 < AakashkaushikGi4> @zoq, Oh okay thanks, I didn't knew that i could kind of edit layers as a user without editing and recompiling the library.

11:45 < kartikdutt18Gitt> Hey @zoq, If I want to merge a PR, I can just simply click on the "Merge pull request" button, right?

11:45 < zoq> kartikdutt18Gitt: Yes :)

11:46 < kartikdutt18Gitt> Great, Thanks.

11:55 < yashwants19[m]> Hi @zoq do you have any idea about this issue: Error while cancelling workflow_id 2372494: Resource not accessible by integration.

11:56 < yashwants19[m]> I am facing this issue while testing for auto cancelling previous builds for github actions.

11:57 < yashwants19[m]> https://github.com/mlpack/mlpack/runs/1193145575?check_suite_focus=true#step:3:27

11:57 < yashwants19[m]> Here is the link.

11:59 < zoq> yashwants19[m]: https://github.com/actions/first-interaction/issues/10#issuecomment-670968624

12:00 < yashwants19[m]> Let me try this.

12:01 < zoq> I think once it's merged it should work with 'pull_request' as well.

12:02 < yashwants19[m]> Yeah this make sense.

12:02 < yashwants19[m]> Thank you for the help.

12:03 < yashwants19[m]> I think so with 'pull_request_target' it is not triggering the cancel job.

12:03 < zoq> Probably not.

12:11 < yashwants19[m]> I have revert this back to `pull_request`, Lets wait for merge.

14:55 < rcurtin> PulkitGitter[m]: if you want to look into GNNs, the first question is "how do we represent a graph in mlpack?"

14:56 < rcurtin> there are lots of possibilities, but we would need to come up with something there and all agree that the abstraction is good before moving forward

17:30 < mlpack-meeting> Hello everyone, video meeting in about 30 minutes - https://zoom.us/j/3820896170