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/
< abernauer[m]>
rcurtin: Just wanted to reiterate I am available to help with code review and bug fixing for the R bindings once GSoC starts ramping up.
< rcurtin>
abernauer[m]: yes, like I said before, you are more than happy to review any PRs that you'd like
asadoll[m] has joined #mlpack
ImQ009 has joined #mlpack
TaapasAgrawalGit has joined #mlpack
robotcatorGitter has joined #mlpack
JoelJosephGitter has joined #mlpack
siddhant2001Gitt has joined #mlpack
maa1388[m] has joined #mlpack
himanshu_pathak[ has joined #mlpack
kartikdutt18[m] has joined #mlpack
eddelbuettel[m] has joined #mlpack
robertohueso has joined #mlpack
SriramSKGitter[m has joined #mlpack
RyanBirminghamGi has joined #mlpack
Cadair has joined #mlpack
ImQ009 has quit [Ping timeout: 272 seconds]
favre49 has joined #mlpack
ImQ009 has joined #mlpack
favre49 has quit [Ping timeout: 260 seconds]
favre49 has joined #mlpack
favre49 has quit [Ping timeout: 256 seconds]
favre49 has joined #mlpack
favre49 has quit [Ping timeout: 260 seconds]
favre49 has joined #mlpack
favre49 has quit [Ping timeout: 256 seconds]
favre49 has joined #mlpack
< jeffin143[m]>
Hi @walragatver:matrix.org Ryan Birmingham (Gitter) : are you both there ??
< RyanBirminghamGi>
I'm here!
< jeffin143[m]>
So today I don't have anything on my list
< jeffin143[m]>
Everything is clear :) and I am still figuring out cmake things
< RyanBirminghamGi>
How is that going? :)
< walragatver[m]>
jeffin143: I am also here.
< jeffin143[m]>
Cmake is really vast, and there are.my ways to do a single tasl
< jeffin143[m]>
Task*
< jeffin143[m]>
And when you come down to 3rd task 1st task way is not compatible with 3rd task or may be I did something wrong :-p
< jeffin143[m]>
Are many*
< jeffin143[m]>
I am taking refrence from ensmallen since I am planning , mlboard would be header only and hence we can take some bits from ensmallen
< RyanBirminghamGi>
that makes sense. cmake is kind of between paradigms right now, and that's not helped by every blog having a different opinion on what "good modern cmake" is
< walragatver[m]>
birm: jeffin143: Are we going to use modern cmake?
< jeffin143[m]>
Same version as in mlpack
< walragatver[m]>
jeffin143: If I am not wrong. We are looking to upgrade the cmake version in mlpack.
< walragatver[m]>
right?
< walragatver[m]>
or is it updated?
< RyanBirminghamGi>
That's an eventually kind of thing
< RyanBirminghamGi>
I don't think mlboard will be intricate enough for it to be a problem either way
< RyanBirminghamGi>
As long as it works cross-platform, we should be good!
< walragatver[m]>
<RyanBirminghamGi "As long as it works cross-platfo"> Okay sounds cool
< walragatver[m]>
I don't mind anything as long as the build is working
< jeffin143[m]>
There are lot of hardship in being updating that,it would be a separate gsoc project
< jeffin143[m]>
Since there are some complexity , i.am not sure what but I saw an issue
< jeffin143[m]>
So I wanted to be aligned with mlpack
< jeffin143[m]>
And i am sure mlboard would have to complex cmake so that we couldn't update later
< jeffin143[m]>
Sorry guys could we postpone the meeting a little bit
< jeffin143[m]>
I don't have electricity and hence don't have wifi
< jeffin143[m]>
> <@jeffin143:matrix.org> Sorry guys could we postpone the meeting a little bit
< jeffin143[m]>
> I don't have electricity and hence don't have wifi
< jeffin143[m]>
Tmrw may be sorry
< walragatver[m]>
birm: jeffin143: I don't have too much today on my agenda.
< walragatver[m]>
jeffin143: I just have one doubt. Feel free to answer whenever you can. Have you asked ryan about making repo in mlpack? And what's your plan about the repo?
< RyanBirminghamGi>
My schedule has inverted a bit; I'm free tomorrow 15:00 UTC on
< walragatver[m]>
birm: sounds good to me.
< walragatver[m]>
Let's meet tommorow
< walragatver[m]>
jeffin143: birm: Till then take care.
< RyanBirminghamGi>
Same time as today?
< walragatver[m]>
birm: Fine with me.
< RyanBirminghamGi>
ok!
< jeffin143[m]>
Hi , are you both still there , I am back
< walragatver[m]>
jeffin143: I am here
< jeffin143[m]>
Yes I did post a msg on orc
< jeffin143[m]>
Irc
< jeffin143[m]>
But ryan didn't answer
< jeffin143[m]>
May be he was busy
< jeffin143[m]>
I will ask
< rcurtin>
sorry, I'm here now, we can make a repository, that's no problem
< rcurtin>
but I mean then there are permission issues; maybe it is easier to work in a personal repository to start with?
< rcurtin>
otherwise everything will need PR approvals, etc., etc. (maybe that's desired, I dunno)
< jeffin143[m]>
Yes I don't wanna commit directly
< jeffin143[m]>
I will anyways make a pr
< jeffin143[m]>
Even against my own repo
< jeffin143[m]>
So that @walragatver:matrix.org can review
< RyanBirminghamGi>
We can always make a blank repo within mlpack and copy over, or otherwise transfer ownership.
< RyanBirminghamGi>
So I guess it's a matter of preference.
< rcurtin>
you guys decide and let me know what you'd like me to do :-D
< walragatver[m]>
birm: jeffin143: let's make a personal repo then
< jeffin143[m]>
Yes
< jeffin143[m]>
Ok I am alright
< walragatver[m]>
jeffin143: Just let us know about the repo so that we can review.
< jeffin143[m]>
Yes sure :)
< walragatver[m]>
Okay so that's from my side.
< walragatver[m]>
birm: Anything you would like to discuss?
< RyanBirminghamGi>
I don't have much to talk about myself now. Just let me know if you want me to look at anything in more detail :)
< RyanBirminghamGi>
jeffin143: Oh, have you decided how you would like to do updates? Maybe you have already said, and I forgot.
< walragatver[m]>
birm: jeffin143: Let's have integration testing mechanism in the repo. What do you think?
< walragatver[m]>
I am not sure how tough it is to set up.
< jeffin143[m]>
Yes mlpack blog
< jeffin143[m]>
Like last year
< RyanBirminghamGi>
ok!
< jeffin143[m]>
@walragatver:matrix.org: we need to plan it out
< RyanBirminghamGi>
walragatver what did you have in mind, generally speaking?
< jeffin143[m]>
About the integration testing
< jeffin143[m]>
So type of browser testing framework
< jeffin143[m]>
So that everything is rendered properly
< walragatver[m]>
birm: Basically after adding the personal repo to mlpack we will need to setup a integration testing mechanism. So I think it's better to use it from start. So I think a normal setup similar to ensmallen would be fine.
< walragatver[m]>
<jeffin143[m] "So type of browser testing frame"> Like travis ci or azure pipelines
< jeffin143[m]>
Oh I see
< RyanBirminghamGi>
Oh, you mean just some CI testing generally. I thought you meant a specific mlpack-mlboard integration test. my bad
< jeffin143[m]>
Then I guess it would be good to open it in mlpack
< jeffin143[m]>
So that everything is shared *
< walragatver[m]>
<RyanBirminghamGi "Oh, you mean just some CI testin"> Yes
< RyanBirminghamGi>
Sorry, there are too many testing terms/jargon, haha
< walragatver[m]>
Is it tough to set that up?
< RyanBirminghamGi>
Travis is pretty straightforward, usually. I don't know azure pipelines well.
< jeffin143[m]>
I have never done it for c++
< jeffin143[m]>
I will take a look
< jeffin143[m]>
I can borrow some things from ensmallen
< jeffin143[m]>
:)
< RyanBirminghamGi>
I can create a repo, but it's possible/likely rcurtin does some extra magic wrt mlpack bot that I'm unaware of.
< walragatver[m]>
birm: Yeah I see. I think we don't need a bot at this moment.
< walragatver[m]>
So normal simple setup would be fine
< walragatver[m]>
I will also look into it.
< RyanBirminghamGi>
What do you want to call the repo?
< walragatver[m]>
MLboard would suffice I suppose. What do you guys think?
< rcurtin>
we should probably match the capitalization of 'mlpack' with 'mlboard', but that's just a minor note
< rcurtin>
RyanBirminghamGi: if you can create the repo, awesome, if not, just let me know
< rcurtin>
I can set up mlpack-bot and mlpack-git list integration for it too if you like
< jeffin143[m]>
Ping pong
< RyanBirminghamGi>
I dunno if ping pong is a good name ;)
< jeffin143[m]>
Haha was checking internet
< jeffin143[m]>
mlboard
< jeffin143[m]>
Is good
< RyanBirminghamGi>
I'll use stock bsd 3 clause license for now
< rcurtin>
yeah, I figured weather would be a problem... it has been pouring here in Atlanta most of the day (Atlanta is a few hundred miles from Cape Canaveral)
< shrit[m]>
they can wait several hours
< shrit[m]>
rcurtin You can see it from home directly
< rcurtin>
nope, too far :(
< jjb[m]>
😢 No launch today.
< jeffin143[m]>
Driving to the launch location is allowed ?
< rcurtin>
I think you can drive over to it, but I have no idea what kind of crowd control they have going on there
< rcurtin>
I guess maybe I could see part of the launch from Atlanta? but it wouldn't be much, just a little twinkle in the sky... and I'd also have to figure out how to get above all the trees in my neighborhood :)
< zoq>
NASA told people to stay home due to COVID-19
< rcurtin>
heh, I think a lot of people around here are not heeding a lot of the warnings...
< rcurtin>
here's a picture from a few weeks ago in one of the popular bike paths in Atlanta:
< rcurtin>
yeah... "wow" is about all I could manage to say when I saw the picture too
< zoq>
I think you can find similar pictures here as well
< rcurtin>
yeah, I suppose there are rebellious people everywhere :)
< zoq>
was there a GitHub ui update, it looks different
< rcurtin>
if so, I think it's very subtle (at least based on what I'm seeing)
< jeffin143[m]>
Pr icon
< jeffin143[m]>
Changed
< jjb[m]>
zoq yes, there was a github ui update earlier today…
< zoq>
Good to know, I couldn't find any announcement from GitHub.
< abernauer[m]>
Yeah I didn't spot a single face mask in that photo.
< birm[m]>
the city has asked people to avoid the path, but has otherwise no enforcement
< zoq>
shrit[m]: I enjoyed reading your update post, thanks!
< zoq>
Ohh, I can switch to a dark mode as well, nice.
< shrit[m]>
zoq You are welcome
< shrit[m]>
rcurtin After some debugging, it seems that by commenting everything in mlpack_knn and leaving only main function, the binary size is 1.8 MB
< shrit[m]>
Adding typedef knn and using it in program option to recover the input model will go up tp 4.32 MB