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/
< GopiManoharTatir>
Heyy, I am planning to create a Trading Env for mlpack under GSoC. I have a doubt, should I use Gym based env here? Can I create gym-based env? It can be totally independent from gym, right?
almost_strange[m has joined #mlpack
Ayush has joined #mlpack
Ayush has quit [Quit: Connection closed]
ImQ009 has joined #mlpack
AyushRaj8271Gitt has joined #mlpack
< AyushRaj8271Gitt>
Hello everyone! My name is Ayush Raj . I am a btech 1st yr student from India. I started learning about the mlpack C++ library and trying to use it but statring to face some issues.
rcurtin[m] has quit [Quit: Idle for 30+ days]
< AyushRaj8271Gitt>
Hey, I am planning to make a simple program to understanding how ml pack works but I its looks like vs code can't recognize the mlpack c++ library. I am using a pre-build ml pack for Windows. Can any one help me with this issue?
< RishabhGargGitte>
A quick question, let's say I want to use `mlpackMain` twice in a unit test. Then to call it second time, do I have to again specify all the input parameters or the `IO` object retains them ?
< zoq>
AyushRaj8271Gitt: Do you get any error message?
< zoq>
ShahAnwaarKhalid: I'll answer on the issue.
< ABHINAVANAND[m]>
<ABHINAVANAND[m] "Hi, While I was going through po"> zoq did you go through the improved implementation as mentioned in the last 4 pages of the document.
< RishabhGargGitte>
Thanks @NippunSharma. In my case, actually `std::move` was used inside the bindings that was causing the error. Its resolved now :)
< NippunSharmaGitt>
Yes, that can cause some issues
hasan1101 has joined #mlpack
hasan1101 has quit [Client Quit]
hasan1101 has joined #mlpack
hasan1101 has quit [Client Quit]
< OleksandrNikolsk>
Hey,
< OleksandrNikolsk>
Is it actually currently possible to call an ensmallen-optimizer using the mlpack python bindings?
< OleksandrNikolsk>
Can you pass an optimizer from ensmallen while calling the ffn Train method?
< rcurtin[m]1>
there's no FFN binding to Python, so you couldn't do it from there, but if you are using C++, the FFN.Train() method does support any ensmallen optimizer
< OleksandrNikolsk>
ok, was just thinking a little about possible API for a GSOC proposal.
< OleksandrNikolsk>
I guess if there is no bindings for ffn, it should not be part of a proposal, which is not dealing directly with ffn.
< rcurtin[m]1>
yeah, that makes sense---but, you could discuss how what you are implementing could eventually fit into an FFN binding, once it was created (I do expect us to make one at some point, it's definitely useful and relevant)
< OleksandrNikolsk>
Yeah, would be nice to have some easy going neural network programming in mlpack python
< OleksandrNikolsk>
Took me a while to code a 3-layer one in cpp, but perhaps because I have no excercise
< OleksandrNikolsk>
Btw, is it ok to suubmit a proposal around the second half of the submission period?
< OleksandrNikolsk>
I guess posting a draft as early as possible is helpful to have time for feedback
< rcurtin[m]1>
yes, of course, it is completely okay to do that; the proposal only needs to be submitted by the deadline :)
< OleksandrNikolsk>
23:59 😉
ImQ009 has quit [Quit: Leaving]
< OleksandrNikolsk>
`as_scalar(P.col(i) * v.t() )` did not work either
< rcurtin[m]1>
I'm very surprised that gives a compiler error... can you paste what the error is? maybe the output type is wrong? you could also use `arma::dot()`
< OleksandrNikolsk>
Sorry, both worked, I transposed the second vector and got an n x n matrix, my fault
< OleksandrNikolsk>
as_scalar is necessary, because the result of multiplication is 1 x 1 mattrix