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/
Suryo has joined #mlpack
< Suryo>
rcurtin, zoq: another detail that is included in the ideas page is: "Note mlpack has already an optimization infrastructure for handling unconstrained problems, However, there is no infrastructure to deal with constrained problems."
< Suryo>
So would you like me to propose a means of dealing with constraints in standard ways (such as accepting only feasible solutions, or using penalty functions, etc.)?
< ShikharJ>
rcurtin zoq: I think I need to apologize for my absence lately, I was busy with some final interviews, but most of them are over now. I'll get back to reviewing the PRs now.
pd09041999 has quit [Ping timeout: 252 seconds]
Toshal_ has joined #mlpack
pd09041999 has joined #mlpack
Toshal has quit [Ping timeout: 255 seconds]
pd09041999 has quit [Max SendQ exceeded]
pd09041999 has joined #mlpack
pd09041999 has quit [Max SendQ exceeded]
pd09041999 has joined #mlpack
< rcurtin>
ShikharJ: no worries! there are no expectations for an open source project :)
< rcurtin>
hopefully the interviews went well :)
< ShikharJ>
rcurtin: Oh yeah, I have an intern offer from Amazon India and Ola Cabs, and waiting on the results for the rest.
< rcurtin>
nice! congratulations!
< ShikharJ>
Thanks
Toshal_ has quit [Read error: Connection reset by peer]
< Hemal>
what is the command to run a specific test ?
< Hemal>
Also, if i want to search a specific string in the log of mlpack irc, how can I do that ?
< zoq>
mlpack_test -t YourTestSuite if you like to run all tests inside a test suite or mlpack_test -t YourTestSuite/YOURTESTCASE to run a single test
< zoq>
ctr + f?
< Hemal>
thanks for the test commands !
< Hemal>
the ctrl + f would search in a single day (for example Logs for the day 2018-12-16 (starts at 0:00 UTC) are shown below.) I want to search in the log for many days. I am clicking on each date and trying to find that string, that becomes somewhat tedious
< rcurtin>
Hemal: you're right... hmm...
< rcurtin>
ah, ok, here we go:
< rcurtin>
you can go to mlpack.org/irc/logs-all.html then download the entire log
< rcurtin>
and then ctrl+f through that
< rcurtin>
(you can also get to that page through the "list of all logs" link at the bottom of the calendar)
< rcurtin>
honestly I'm surprised that the irc log has been so useful for so many years without any real changes
< rcurtin>
it's a mess of bash scripts I threw together one weekend... :)
< rcurtin>
I was quite surprised to find at that time that there weren't any real or simple IRC logging solutions
< Hemal>
yes, indeed the logs are very useful! Thank you for your reply :)
< Hemal>
Maybe we could add a string search functionality to the logs so that the developers won't have to download. Just a suggestion.
< ShikharJ>
zoq: Thanks.
< rcurtin>
Hemal: I guess it would not be impossible to add a page that just has all the logs displayed on it
< rcurtin>
the problem is finding the time to do it :(
< rcurtin>
any additional development to that repo will be... tedious, because it's all pretty brittle
< sreenik>
rcurtin: I am working on the CLI for ann. I think we had agreed upon a simple text format for the model and related parameters. However, do you suggest to store the allowed layers and parameters (and their corresponding formats) in a json for easier updation or have them hard-coded?
< rcurtin>
it may be pretty hard to contribute to (or even make it work). I put it together once a long time ago so that it would work, not so that it would be well documented and easy to work with...
< rcurtin>
sreenik: I saw your updates to the CLI for ANN issue; I still need to comment
< rcurtin>
I think I'm ambivalent about JSON or hard-coded, but probably hard-coded is easier to do---since with JSON we also have to figure out how to distribute the JSON file, and then where to load it from when the ann program is called, and so forth
< rcurtin>
I'll probably develop an opinion on it but I don't have a strong one at the moment. still need to think about advantages/disadvantages of each approach
< rcurtin>
but anyway, I hope to find the time to respond to the issue soon
< sreenik>
Okay, but now I see that my comments are unnecessarily long. I hope to summarise everything in the last comment. You may consider only going through that when you have free time :)
< rcurtin>
hehe, there is nothing wrong with being verbose :)
< rcurtin>
or, well, the only problem I have with being verbose is that a lot of times people don't actually read the whole email I sent to them because it is too long :)
< rcurtin>
but personally I don't mind it at all, and prefer the extra context
pd09041999 has quit [Ping timeout: 250 seconds]
< sreenik>
Haha, we've all been victims of "oh I didn't read the entire email, it was too long"
kumaraditya1999 has joined #mlpack
zhuq has quit [Ping timeout: 245 seconds]
Hemal has quit [Ping timeout: 250 seconds]
Hemal1 has joined #mlpack
Toshal- has joined #mlpack
kumaraditya1999 has quit [Ping timeout: 256 seconds]
pd09041999 has joined #mlpack
Soonmok has quit [Quit: Connection closed for inactivity]
Piyush has joined #mlpack
umang has joined #mlpack
zhuq has joined #mlpack
Toshal has quit [Quit: WeeChat 2.4]
Toshal- is now known as Toshal
Piyush has left #mlpack []
umang has quit [Ping timeout: 256 seconds]
manish7294 has joined #mlpack
< manish7294>
Hemal: A google search may also come in handy for log search. Just a simple search consisting of strings like mlpack + irc + "string to search" may do the work. Atleast that's what I used to do :)
Hemal1 has quit [Read error: Connection reset by peer]
pd09041999 has quit [Ping timeout: 252 seconds]
ayesdie has joined #mlpack
manish7294 has quit [Quit: Page closed]
pd09041999 has joined #mlpack
Toshal has quit [Ping timeout: 272 seconds]
Toshal has joined #mlpack
Toshal has quit [Ping timeout: 268 seconds]
shubhangi has joined #mlpack
< shubhangi>
Hello Everyone!
< shubhangi>
If I update cmake do I have to run 'make ' command again?
MichalOp has joined #mlpack
< rcurtin>
shubhangi: do you mean updating the version of cmake, like from 3.10 to 3.12 or something
< rcurtin>
?
< rcurtin>
or do you mean that you updated the CMake configuration for mlpack or something like this
MichalOp has quit [Client Quit]
< shubhangi>
Yes updating version of cmake from 3.10 to 3.12
< rcurtin>
ah, in this case no need to run 'make'
< rcurtin>
CMake should still produce the same configuration and build rules between versions :)