09:50 < jenkins-mlpack> Project docker mlpack nightly build build #325: SUCCESS in 2 hr 36 min: http://masterblaster.mlpack.org/job/docker%20mlpack%20nightly%20build/325/

10:03 < ShikharJ> zoq: Is there a way of loading only the first n elements from a dataset? I'm currently debugging Kris' PR on mnist_first250_training_4s_and_9s.arm dataset, but the issue is that I have to wait for it to process 500 images and then return the error.

10:03 < ShikharJ> zoq: This is taking up a lot of time.

12:31 < ShikharJ> zoq: Also, how can I run gdb with the executable?

12:44 < manish7294> ShikharJ: Try this: gdb bin/"executable"

12:45 < manish7294> Then: run - input -l label ........

12:45 < manish7294> -i input.csv ------ the usual binding commands

12:52 < manish7294> ShikharJ: Regarding the dataset issue, I guess the best will be to use a smaller dataset, if it works.

17:12 < zoq> ShikharJ: Not directly, but one way would be to create another dataset that only contains the first sample, arma::save should be helpful here.

17:12 < zoq> ShikharJ: If you use gdb, you have to write a simple exectuable that links against libmlpack, debugging a test case is difficult.

17:29 < haritha1313> zoq: Would you be able to take a look into PR #1355 once again? I am still stuck at using template parameter for PARAM_MODEL_IN, I had replied on github to your suggestion of it being similar to PCA. Any direction would be really helpful.

17:49 < rcurtin> haritha1313: I can answer this one, maybe zoq can also, but I think I have a good idea---let me write it down there

17:53 < haritha1313> rcurtin: Thanks for the quick response :). I

17:54 < haritha1313> I'll try what you have suggested, thanks for helping out.

17:58 < rcurtin> sure, I hope it is helpful

17:58 < manish7294> zoq: Travis-CI people do help. They have increase my job timeout to 90 minutes without any question :)

17:58 < rcurtin> manish7294: I didn't have a chance to look into it too much, but I think you need to define 'programName' in lmnn_main.cpp so that the python bindings compile correctly

17:59 < rcurtin> however, you may have already figured that out---I glanced at the travis build several hours ago and haven't looked since then :)

17:59 < rcurtin> (alternately just comment out the python binding declaration in CMakeLists.txt for now...)

18:00 < manish7294> rcurtin: Right, I have noticed that. Currently I am getting just fine results with L-BFGS but I do think there may be some bug in my evaluate or gradient, if you get some time please look at those

18:01 < rcurtin> sure, I will try to do that later this afternoon, but I am still in this long all-day meeting

18:01 < rcurtin> do you have any information on exactly what is failing? or do you want me to just review the implementation itself and see if I see any bugs?

18:02 < manish7294> rcurtin: Yeah, that must me hectic. I remembered you mentioning that :)

18:03 < manish7294> you can just review evaluate and gradient for now as whole implementation is quite rough as of now.

18:03 < rcurtin> we're reviewing papers from ICLR, which is nice, but there is a lot of other stuff I needed to do today too and when this is over I still have to do it... :)

18:03 < rcurtin> ok, sounds good, I will take a look in a little while

18:04 < manish7294> fine, whenever you get time. No need to hurry!

18:22 < travis-ci> mlpack/mlpack#4942 (master - f025772 : Marcus Edel): The build has errored.

18:22 < travis-ci> Change view : https://github.com/mlpack/mlpack/compare/095f784a3074...f02577278a6b

18:22 < travis-ci> Build details : https://travis-ci.org/mlpack/mlpack/builds/381774002

21:34 < zoq> manish7294: yeah, great people

23:20 < travis-ci> mlpack/mlpack#4944 (master - 0853aff : Ryan Curtin): The build passed.

23:20 < travis-ci> Change view : https://github.com/mlpack/mlpack/compare/f02577278a6b...0853affad63d

23:20 < travis-ci> Build details : https://travis-ci.org/mlpack/mlpack/builds/381894821