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/
< chopper_inbound[>
How can I get/set the `weight` of the linear layer (`linear.hpp`)? There is such method for `bias` and for concatenated form of `weight` and `bias` but not just `weight`. I can get the `weight` from `weights` but that would only increase the lines of code or make the expression complex. Is this something intentional? I require to initialize the `weight` of that particular layer with different initialization rule than
< chopper_inbound[>
that of the network's. Is there any other way to do this?
< walragatver[m]>
jeffin143 : Let's have a meet day after tomorrow? Let me know what you think about it.
< walragatver[m]>
Also, I hope you have seen the mail I have sent you two days ago. If you have any doubts feel free to ask about it.
< walragatver[m]>
chopper_inbound : There is no direct way to access the `weight` of a layer directly
< walragatver[m]>
I think you need to use the extra lines of code for that
< walragatver[m]>
Regarding change in intialization rule. I am sure whether it is possible currently or not.
< walragatver[m]>
But I think it can be done or atleast we can expose that feature inorder to make it possible.
< walragatver[m]>
It would be great if you could open an issue for it. I think zoq would be knowing about it in detail.
< chopper_inbound[>
@walragatver:matrix.org: Right. Let me open an issue for that.
< jeffin143[m]>
walragatver: oops , I have no clue why did that mail was not In my inbox and was in social section ,
< KhizirSiddiquiGi>
nevermind, I did a clean build and it works as expected, thanks. :thumbsup:
< yashwants19[m]>
Happy to hear, clean build worked. :)
< KhizirSiddiquiGi>
thanks :)
favre49 has joined #mlpack
favre49 has quit [Ping timeout: 245 seconds]
ImQ009 has quit [Quit: Leaving]
< HimanshuPathakGi>
Hey @saksham189 When you get time time take a look at my RBFN pr now it is giving right prediction and tests are passing now. I have to add multiple non-linear activation functions I will open a separate PR for that