04:34 < ShikharJ> zoq: Ah, but I think this wouldn't be applicable to transposed convolution case, as the non-padded input is always passed to the `Backward` convolution rule in the `Forward` routine.

05:37 < ShikharJ> zoq: I've updated the PR, please take a look whenever free. Also, I'd like to work up on Layer Normalization and completing the PR on Atrous Convolutions.

05:38 < ShikharJ> The task on Layer Normalization was planned for the later, but I'd like to finish it now, so that I can focus on big tasks at once, without worrying about the smaller dependencies.

05:42 < ShikharJ> This would also complete the convolutional toolbox of mlpack (maybe depth-wise separable convolutions would be remaining), but I don't think they're used that often (I could only find a couple of papers).

05:47 < ShikharJ> They're mainly used in the Xception model if I'm not wrong.

06:41 < travis-ci> ShikharJ/mlpack#128 (master - acce0c0 : Ryan Curtin): The build has errored.

06:41 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/d390433adc55...acce0c0000b8

06:41 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/377146717

10:53 < zoq> ShikharJ: Agreed, depth-wise separable convolutions isn't used that much, if you like to shift the work, that's fine with me; will take a look at the PR later today.

11:19 < ShikharJ> zoq: Thanks. I'll push both the PRs over the weekend.

11:22 < zoq> ShikharJ: Sounds good; I think I have already mentioned that working over the weekend isn't required :)

11:31 < ShikharJ> zoq: Yeah, just wanted to get a head start. Already the work on Transposed Convolution layer took a bit longer than I expected it would take. So taking no risks.

12:30 < travis-ci> ShikharJ/mlpack#129 (Deconv - e708df9 : Shikhar Jaiswal): The build has errored.

12:30 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/df31c1ac1645...e708df9bdcca

12:30 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/377239952

13:38 < manish7294> rcurtin: Are you there?

13:40 < manish7294> rcurtin: I want to discuss the constraint generation algo.

13:48 < rcurtin> manish7294: yeah, I am here, but unfortunately today a lot of people from out of town are coming to my office

13:48 < rcurtin> so I'll be in a meeting in about 10 minutes... and it's likely to go all day :(

13:49 < rcurtin> maybe an email to the list is a better way to discuss in this case?

13:49 < manish7294> rcurtin: It's fine. Just ping whenever you are free :)

13:49 < manish7294> sure :)

13:49 < rcurtin> sure, I will do that

13:49 < rcurtin> or if you send an email I'll respond when I can :)

13:50 < manish7294> Yup! I will do that

13:50 < rcurtin> sorry to not be available on the spot... I am not looking forward to a day of talking to lots of people inside the company

13:50 < rcurtin> (I'd rather be getting work done! :))

13:50 < manish7294> It's cool :)

15:50 < Atharva> zoq: I am in Berlin till 12th night

15:50 < Atharva> Will you be free tomorrow or day after tomorrow?

16:29 < robertohueso> Jenkins tests seem to be down

16:44 < zoq> Atharva: Unfortunately, I'm out of town :( ... do you like me to recommend something?

16:46 < zoq> robertohues: It's up, but the working queue is filled up with the matrix build.

16:51 < robertohueso> zoq: thanks! :)

16:55 < zoq> robertohues: The style, static check is queued, but it might take some hours until we can see the results.

17:37 < rcurtin> robertohueso: zoq: I could kill the monthly build, it is not high priority

17:37 < rcurtin> let me know if you'd like me to. I mostly just wanted to see if any tests were still randomly failing

17:46 < robertohueso> rcurtin: Don't worry I can wait or fix any mistakes later :)

18:04 < rcurtin> ok, sounds good. the monthly matrix build may take up to a handful pf days

18:04 < rcurtin> *of days

18:05 < rcurtin> the build server machine is probably very warm right now

18:05 < rcurtin> load average currently 74.23... :)

18:08 < zoq> hopefully everything passes

18:08 < ShikharJ> zoq: Hopefully the work for Layer Normalization would also be finished by tomorrow.

18:09 < rcurtin> ha

18:09 < rcurtin> "Build Queue (1,652)"

18:39 < zoq> ShikharJ: Great :)

18:41 < zoq> rcurtin: I wonder if it's a good idea to run the PR jobs simultaneously.

18:41 < rcurtin> yeah, I wonder if there is a job priority plugin for jenkins

18:42 < rcurtin> I'm writing the HISTORY.md file for mlpack 3.0.1 now, but once that is done I'll look into that (unless you'd like to)

18:42 < zoq> sure, let's see if there is something useful

18:46 < zoq> Priority Sorter Plugin

18:56 < zoq> okay, let's see if that works, everything inside the PR view has the highest priority

19:03 < rcurtin> I might need to restart the monthly matrix build, but let's watch for a little longer and see if any PR jobs get done

19:03 < rcurtin> or maybe someone needs to push a PR commit so that the job gets put to the front of the list

19:04 < zoq> let's see what happens if an executor becomes free

19:12 < travis-ci> ShikharJ/mlpack#130 (LayerNorm - 4d237fc : Shikhar Jaiswal): The build has errored.

19:12 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/commit/4d237fccd111

19:12 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/377400445

19:54 < travis-ci> ShikharJ/mlpack#131 (LayerNorm - 32f41ee : Shikhar Jaiswal): The build has errored.

19:54 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/4d237fccd111...32f41eecd6cb

19:54 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/377401664

21:51 < travis-ci> mlpack/mlpack#4866 (mlpack-3.0.1 - 81ad1d8 : Ryan Curtin): The build passed.

21:51 < travis-ci> Change view : https://github.com/mlpack/mlpack/compare/mlpack-3.0.1

21:51 < travis-ci> Build details : https://travis-ci.org/mlpack/mlpack/builds/377442408