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/
< rcurtin>
shrit[m]: I kept upgrading to woody, and now I'm trying to upgrade to sarge, but now the apt archives are so large they are making my machine swap :-O
< rcurtin>
it took 40 minutes to get to the yes/no prompt when I did `apt-get dist-upgrade` :)
petris has quit [Ping timeout: 265 seconds]
petris has joined #mlpack
ImQ009 has joined #mlpack
mlozhnikov[m] has left #mlpack []
ShikharJaiswalGi has joined #mlpack
HimanshuPathakGi has joined #mlpack
RyanBirminghamGi has joined #mlpack
jeffinsamGitter[ has joined #mlpack
LolitaNazarov[m] has joined #mlpack
jacob-earleGitte has joined #mlpack
birm[m] has joined #mlpack
Param-29Gitter[m has joined #mlpack
Eddie-XiaoGitter has joined #mlpack
< Eddie-XiaoGitter>
@rcurtin Hi Ryan, when will you add classification to LSTM and regressor to RandomForest on these two most popular models?
travis-ci has joined #mlpack
< travis-ci>
UtR491/ensmallen#36 (moead - d2c8a2d : Utkarsh Rai): The build was broken.
< shrit[m]>
rcurtin: Just a question why you are trying to upgrade up to sarge if the machine is too old. I think woody is fine right?
< rcurtin>
well because I wanted to see how new of a system I could run :)
< rcurtin>
but I'm pretty sure this adventure will end with me reinstalling slink and then only stepping up to potato
< rcurtin>
even woody is pretty slow
< shrit[m]>
I have woody on my machine and it eat 12 MB of memory out of nothing
< rcurtin>
uh-oh, I only have 8MB :-D
< shrit[m]>
I can not imagine how much memory you have consumed with woody
< rcurtin>
actually I type free -m and it says only 5, I think because 3MB is claimed for buffers, etc.
< shrit[m]>
good, so it mange well the memory when there is nothing, but notice that half of the memory is consumed already without running anything, even not an SSH server.
< rcurtin>
I'll at least end up with an amusing benchmark for how long it takes to boot and how long it takes to ssh in
< rcurtin>
Eddie-XiaoGitter: that would be really awesome to add regression for RandomForest and DecisionTree, but at least from my side I don't have time to do it
< rcurtin>
if you like, you could open an issue for it and we could see if anyone is interested in contributing it
< rcurtin>
for classification with an LSTM, this should already be possible---you just use a softmax layer as the last layer, with number of nodes equal to the number of classes, then take the maximum prediction as the predicted class
< rcurtin>
possibly that API could be made simpler---but again, please do feel free to open an issue on Github, that's probably the best way to help make it happen :-D
say4n has left #mlpack []
< abernauer[m]>
Had my phone screen was pretty low pressure. I like the company alot, but a little concerned about the responsibilites.
< rcurtin>
glad to hear it went well---hopefully further discussions can make it a bit clearer what the position is, what the responsibilities would be, etc.
< rcurtin>
when I interview I try to just gather as much information as I possibly can about who I will be working with, what the trajectory of the group inside the company is, what working there is like, etc. etc.
< rcurtin>
in some sense I am interviewing them just as much as they are interviewing me :-D
< rcurtin>
shrit[m]: I got myself into dependency hell trying to go from woody to sarge :(
< rcurtin>
I think I'll start over and stop at potato :-D
< abernauer[m]>
That sounds like a solid approach. The only bad vibe I got from the phone call was she seemed slightly concerned I was over qualified or not interested in the position.
< rcurtin>
that's better than the opposite situation of being underqualified :)
< abernauer[m]>
The company does Software for VR, AR and 3D Ecommerce in the retail space.
< rcurtin>
it sounds like a cool opportunity for sure!
< shrit[m]>
rcurtin: Agreed, I used to use debian testing, and then I stopped, I was unable to resolve the dependency hell that I was facing each time I need to install a new software.
< rcurtin>
:)
< shrit[m]>
I think it will be fast to re-install potato. Hope that it will be not so slow
< rcurtin>
yeah, I should be able to have it done by tomorrow
< shrit[m]>
Imagine, I spent the entire week deubging a strange issue in my thesis code, It started as a mutex issue, then I undserstood that it was a copying issue, in which a vector of object is being copied, and it should not be, finally I found that I forget to add `&` to a private data member only once out of 10.
< rcurtin>
ha! yeah, I have definitely spent way too long digging after single-character fixes too :(