verne.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
Cyrinika has quit [Remote host closed the connection]
Cyrinika has joined #mlpack
mayank has joined #mlpack
mayank has quit [Quit: Page closed]
vivekp has quit [Ping timeout: 240 seconds]
vivekp has joined #mlpack
vivekp has quit [Ping timeout: 255 seconds]
vivekp has joined #mlpack
vivekp has quit [Ping timeout: 255 seconds]
vivekp has joined #mlpack
vivekp has quit [Ping timeout: 258 seconds]
vivekp has joined #mlpack
raghav9_97 has joined #mlpack
raghav9_97 has quit [Client Quit]
kaushik_ has joined #mlpack
< rcurtin>
kaushik_: ok, in this case I will probably go ahead and finish that PR then
kaushik_ has quit [Quit: Connection closed for inactivity]
kaushik_ has joined #mlpack
< rcurtin>
zoq: oops, I pushed to the wrong benchmarks branch, let me redo that to make the correct logistic regression PR
< rcurtin>
not sure why I have another PR open on the master branch anyway, that was a stupid idea...
< zoq>
rcurtin: If you like we can use the sweep PR, if that's easier for you; looks like it's ready anyway (we could change the colors later).
< rcurtin>
nah, I am splitting it
< rcurtin>
how can I add new datasets to the datasets/ repo? should I open a PR on bitbucket? or is it easier for you to do it?
< zoq>
If you don't have an bitbucket account, it's probably easier if I just push the datasets.
< zoq>
I was thinking to host the datasets on mlpack not sure it's a good idea, but sometimes bitbucket is super slow.
< rcurtin>
hosting on mlpack.org is fine, we don't have bandwidth issues (unless suddenly the benchmarks repo becomes super popular)
< rcurtin>
the terms of the hosting are that we pay for 5Mbps 95th percentile bandwidth
< rcurtin>
so basically if we took all the (discretely sampled) bandwidth usage over a month, the 95th percentile has to be at or below 5Mbps
< rcurtin>
I think we have come nowhere close to this in the two years it's been hosted there
< zoq>
Sounds like that would be possible; so maybe compress the datasets, write a simple script for the download part in the fashion of the libraries install scripts.
< zoq>
Also, do you like to clean #108, looks like it includes the sweep commits? Or we can just manually merge the PR.
< zoq>
If no one beats me with the download script; I'll open a PR tomorrow, should be straightforward.
< rcurtin>
blah you are right
< rcurtin>
I didn't rebase it correctly
< rcurtin>
hang on...
< rcurtin>
I only did half the job, which was removing the logistic regression changes from the sweep PR
< rcurtin>
if you want to handle the download script, you will probably get to it before me; I think you have write access to mlpack.org:/var/www/www.mlpack.org?