ChanServ changed the topic of #mlpack to: "mlpack: a fast, flexible machine learning library :: We don't always respond instantly, but we will respond; please be patient :: Logs at http://www.mlpack.org/irc/
< AakashkaushikGit> Thanks for the update.
< civdex> No problem
< civdex> let me know if you have any advice
< civdex> i opened up an issue on github in the event anyone has any feedback https://github.com/mlpack/mlpack/issues/2623
< rcurtin> civdex: thanks for opening the issue---I saw it, I'll try and take a look in the morning or sometime tomorrow
< civdex> thank you very much
< civdex> cheers
< rcurtin> we had some issues some time back with the convolution layer and some invalid memory accesses... but... I thought we had it solved!
< rcurtin> anyway, getting late here, so at least for me it will wait until tomorrow :)
< civdex> maybe just a beginners mistake on my end of some kind
< civdex> adios!
< civdex> :)
d1 is now known as DrGibbby
civdex has quit [Remote host closed the connection]
ImQ009 has joined #mlpack
civdex has joined #mlpack
< civdex> Hey guys, so I'm trying to rebuild mlpack from source using CMake, and while not in red, I'm getting a message that says "
< zoq> civdex: Looks like the message got cut off.
civdex has quit [Remote host closed the connection]
< rcurtin> zoq: oops, missed your earlier message... yeah, masterblaster had 256GB :)
< rcurtin> but... it also had 72 cores; only 16 in mlpack.org now
civdex has joined #mlpack
< civdex> Any idea why "using namespace ens" would be giving an error "name must be a namespace name"
< rcurtin> civdex: maybe a `#include <ensmallen.hpp>` is needed?
< civdex> I got that line in there and there was no error associated with. Checked my properties and made sure that there was an include directory leading to where that .hpp file would be found
< rcurtin> make sure also that you didn't #include <ensmallen.hpp> from inside another namespace
< civdex> okay thanks ill check
ImQ009 has quit [Quit: Leaving]
< zoq> rcurtin: I guess even if we count all current build nodes cores we are < 72.
< rcurtin> if dealgood comes back online we'll be closer
< rcurtin> but I am still waiting on a response...
< rcurtin> if COVID wasn't going on I would just walk over there and bang on doors until someone helped me out but I can't do that now :-D
< zoq> I guess RAM we are pretty close to 256GB if we count every node.
< rcurtin> :)
< zoq> Right, probably not the best idea.
< rcurtin> at the same time, I don't want to overload mlpack.org 24/7 like we did to masterblaster, because then website latency will get slow (and my screen session on that system will get slow too :))
< zoq> The current jenkins setup looks good to me, don't think there are any jobs that have to wait > 3 hours.
< zoq> Sure, we removed the build matrix.
< rcurtin> it might be nice to bring back, e.g., the nightly build matrix (but maybe run it weekly?)
< rcurtin> it would require a lot of setup though, since it depended on a docker registry running on masterblaster
< rcurtin> (and that's now gone)
< abernauer[m]> My application got passed on for that role with the startup, but on a positive my career advisor is connecting me with a professional who is an Associate Director of Analytics at a Media Group.
< rcurtin> sorry to hear that! hang in there, there are always more opportunities :)
< abernauer[m]> Yeah I am using it as a learning experience. Job market is a lot tougher than other periods in which I was actively applying.
< rcurtin> yeah, I have to say, with COVID you are pretty unlucky :( but, keep at it, I am sure something will come through eventually
< rcurtin> I finished my undergraduate degree in Spring 2008... things went south over the summer; a friend who had taken a job at Cisco was laid off before he even started
< rcurtin> I was lucky, my plan was to go back to grad school, so I was mostly unaffected by the crisis
< abernauer[m]> Yeah I remember you mentioning that in some initial emails.
civdex has quit [Remote host closed the connection]
< rcurtin> shrit[m]: ok, I think I debugged the failure... I just didn't recurse correctly for loading a variant
< rcurtin> let me double-check my patch and then I'll post it
< shrit[m]> OK
< shrit[m]> I think the two other tests are passing, (this is the case on my machine, I did not check the build farm)
< rcurtin> awesome, we are getting closer :)
< rcurtin> I accidentally closed my cereal PR tab so now I have to reload all the comments :-O
< rcurtin> looks like gitdub forgot all the emails it sent to the mlpack-git list over the past month and decided to send them again... my "mark as read" key just got a lot of exercise