13:31 < shrit[m]> rcurtin: This is the first error I have noticed from the Windows build: https://dev.azure.com/mlpack/mlpack/_build/results?buildId=4481&view=logs&j=9bb75cf0-500a-52c4-dbba-9c305edd1534&t=a1596060-4428-5922-31ab-7ee9d07b5622&l=23220, I am not sure that this is related to the mlpack code but rather to the configurations of azure

13:47 < rcurtin> shrit[m]: this could be helpful: https://github.com/appveyor/ci/issues/742

13:47 < rcurtin> my guess is we'll need to change some build flags or something like this

14:14 < AakashkaushikGit> I will be trying to implement teacher forcing and i will open a PR for that so others don't do the same work and i will be helping in migrating the tests from boost to catch2 parallel and then i will implement the remaining activation functions that i have taken up, those are [RReLU](https://pytorch.org/docs/stable/generated/torch.nn.RReLU.html) and [Adaptive logsoftmax with

14:14 < AakashkaushikGit> loss](https://pytorch.org/docs/stable/generated/torch.nn.AdaptiveLogSoftmaxWithLoss.html#torch.nn.AdaptiveLogSoftmaxWithLoss)

14:14 < AakashkaushikGit> (edited) ... catch2 parallel and ... => ... catch2 parallely and ...

14:15 < AakashkaushikGit> I just wanted to inform this here so there aren't any duplicate PRs.

14:44 < rcurtin> AakashkaushikGit: sounds good; I saw your message from yesterday too---if you're looking for a different way to dive deep, maybe you can find some other issue on Github that makes enough sense to approach?

14:44 < rcurtin> I agree that the various refactoring issues are pretty surface-level, but they at least provide an entry point to become familiar with the structure of the codebase

14:45 < rcurtin> the next step would be to find some way to dig deeper into some portion of the code and understand it at an in-depth level (maybe teacher forcing is a good way towards this)

15:03 < AakashkaushikGit> > `rcurtin on Freenode` Aakash kaushik (Gitter): sounds good; I saw your message from yesterday too---if you're looking for a different way to dive deep, maybe you can find some other issue on Github that makes enough sense to approach?

15:03 < AakashkaushikGit> Yup it makes sense, i would try to find some good issues after i am done with teacher forcing.

15:04 < AakashkaushikGit> > `rcurtin on Freenode` I agree that the various refactoring issues are pretty surface-level, but they at least provide an entry point to become familiar with the structure of the codebase

15:04 < AakashkaushikGit> That's why i started out with those issues, now i feel comfortable and so thought about taking more insight in the codebase.

15:05 < AakashkaushikGit> > `rcurtin on Freenode` the next step would be to find some way to dig deeper into some portion of the code and understand it at an in-depth level (maybe teacher forcing is a good way towards this)

15:05 < AakashkaushikGit> I believe teacher forcing will be a good way as that would make me interact with the implementation of the sequential models and will make me understand how different units such as GRU, LSTM are implemented.

15:05 < AakashkaushikGit> Thanks for the advice @rcurtin, I will keep these points moving forward.

15:05 < rcurtin> sounds like a plan!

15:05 < rcurtin> :)

16:07 < AakashkaushikGit> Yes, 😄

21:02 < shrit[m]> rcurtin: I have tried their solution, It is not working, It there any possible way to allocate more resource for this build?

21:04 < rcurtin> I doubt it; do you know what file it is failing on?

21:10 < shrit[m]> Yes, we have at least three https://github.com/mlpack/mlpack/pull/2458/checks?check_run_id=1145958188

21:27 < rcurtin> msbuildArguments: /m /p:BuildInParallel=true

21:27 < rcurtin> that line is in .ci/windows.yaml... maybe remove those arguments? both of those, I believe, enable parallel builds

21:50 < abernauer[m]> Enjoyed my informational interview this morning. Went well and the professional I talked to enjoyed the conversation.