Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Repo info
Activity
    matrixbot
    @matrixbot
    jonpsy vigsterkr, cool just @ me when you're free, and we could go from there :)
    matrixbot
    @matrixbot
    jonpsy vigsterkr hi! , you up? I'll be around for the next few hours, if I miss it, you may type it anyway I'll have a look when I'm back.
    matrixbot
    @matrixbot
    braceletboy on Freenode Hi everyone, I have a question and hope someone can clarify this. How is ParameterProperties::SETTING different from ParameterProperties::HYPER ? https://usercontent.irccloud-cdn.com/file/bcvEFqw0/image.png
    braceletboy on Freenode I looked at some of the code where ParameterProperties::SETTING is used. It seemed like it's the same as ParameterProperties::HYPER
    braceletboy on Freenode Any reason for having a ParameterProperties::SETTING ?
    matrixbot
    @matrixbot

    jonpsy SETTING is for how you'd train your model. For ex: in tensorflow you can set batch_size which has nothing to do with the model itself.

    Whereas HYPER , relates to the hyper parameters governing the training. For ex: the C in SVM

    matrixbot
    @matrixbot
    braceletboy on Freenode Okay. It's kind of like a more generally applicable (compared to HYPER) parameter. Thank you.
    braceletboy on Freenode While HYPER is specific to the model we are dealing with.
    matrixbot
    @matrixbot
    @a_essam:matrix.org was kicked by @appservice-irc:matrix.org ("Idle for 30+ days").
    matrixbot
    @matrixbot
    @gf712:matrix.org was kicked by @appservice-irc:matrix.org ("Idle for 30+ days").
    matrixbot
    @matrixbot
    Amaan36 on Freenode Hello
    matrixbot
    @matrixbot
    HeikoS on Freenode Viktor Gal (Gitter): ping
    HeikoS on Freenode vigsterkr: ping
    Aryan71
    @freenode_Aryan71:matrix.org
    [m]
    hey can anyone guide me about upcoming projects of this organisation
    fiach
    @freenode_fiach:matrix.org
    [m]
    Hi guys! I'm a final year in Computer Science and I came across this community in the hope of completing the Google Summer of Code this summer. I have experience in developing machine learning algorithms from scratch and a huge passion for applied mathematics and computational sciences. If I can assist in this project in anyway please do let me
    know!
    Nim
    @freenode_Nim:matrix.org
    [m]
    Hey everyone ^-^
    vigsterkr
    @vigsterkr:matrix.org
    [m]
    m ertmost
    Aditya Nikhil
    @AdityaNikhil
    Hi everyone..!
    arya29
    @freenode_arya29:matrix.org
    [m]
    hey!! when i close this tab and again open the freenode i again have join this channel what should i do
    siddhant_jain
    @siddhant_jain:matrix.org
    [m]
    Try using element.io
    arya29
    @freenode_arya29:matrix.org
    [m]
    got it . thanks:)
    arya29
    @arya2910
    image.png
    i am getting this error
    how to solve it??
    siddhant_jain
    @siddhant_jain:matrix.org
    [m]
    I am assuming you have installed shogun on your device.
    if so have you added it to your env variable
    arya29
    @arya2910
    i have installed the miniconda.
    its actually little messy to understand. i am new here so m trying to learn things precisely.
    I will be thankful if i get a little guidance from u all
    Parth66
    @freenode_Parth66:matrix.org
    [m]
    Hello, I am Parth. I wanted to confirm if Shogun has applied for GSoC'21.
    Debo
    @freenode_Debo:matrix.org
    [m]
    Hello I am cs prefinal year student want to perticipate in gsoc 2021,can anyone guide me how start contribute in shogun
    arya29
    @arya2910
    read the wiki
    world-wake
    @world-wake

    Hi. I'm trying to compile Shogun for Windows and I'm getting this error

    D:\python-packages\shogun\build\src\interfaces\python\shogunPYTHON_wrap.cxx(419444,5): error C2440: 'initializing': cannot convert from 'int (__cdecl *)(PyObject *,FILE *,int)' to 'Py_ssize_t'

    that's currently the only error
    world-wake
    @world-wake
    hey- I'll be checking this gitter regularly so if anybody comes up with a solution to that, I'm all ears.
    Rei 머호
    @rei_wu:matrix.org
    [m]

    techprog:matrix.org

    jonpsy
    @jonpsy:matrix.org
    [m]
    @HeikoS mic testing 1..2..3
    HeikoS
    @freenode_HeikoS:matrix.org
    [m]
    jonpsy: hello!
    HeikoS
    @freenode_HeikoS:matrix.org
    [m]
    not much going on here of late :)
    jonpsy
    @jonpsy:matrix.org
    [m]
    Yeah, rooms been a graveyard for months
    jonpsy
    @jonpsy:matrix.org
    [m]

    HeikoS looks like shogun isn't participating in GSoc :(. No worries though, I have applied for mlpack and tensorflow :).

    Would you like to review the tensorflow proposal? :D

    1 reply
    Yuhui Liu
    @LiuYuHui
    @jonpsy:matrix.org Hope your proposal can be accepted :D
    glames
    @freenode_glames:matrix.org
    [m]
    hello. I'm very new to using shogun, I am not sure if this is the right platform to ask doubts regarding how to use a particular function but I have been trying to use BinaryLables on a numpy 1Darray, and I keep getting wrong type of argument error. I initialized a dummy numpy 1Darray and used BinaryLabels, and it worked, and I'm really confused as
    to why it isn't working on my dataset
    glames
    @freenode_glames:matrix.org
    [m]
    Could anyone please help
    jonpsy
    @jonpsy:matrix.org
    [m]
    @HeikoS if you're there. I wanted to let you know that I got selected for GSoc 2021 under mlpack ! Thank you for all the guidance to everyone, @gf712. I know you probably wont be able to read this, but I said it anyway :)
    Yuhui Liu
    @LiuYuHui
    @jonpsy:matrix.org Congratulations!
    1 reply
    anadon
    @freenode_anadon:matrix.org
    [m]
    Hello all, I have a chunk of code good for computing Hidden Markov Models which may or may not be useful to Shogun. If someone could take a look and tell me if it would be a good candidate for an addition, it would be much appreciated. https://gitlab.com/anadon/chmm/-/blob/master/include/hmm.hpp
    It is header-only, has its requirements and environment fully defined in guix, aims to be as constexpr as possible, builds with meson, and is intended to both serve as a usable and educational implementation.
    It does not handle training and doesn't try to handle higher order HMMs since those can technically be encoded into the base case, which if encoded that way allows for fewer pointer dereferences and should be faster. Should be. I'm still looking for feedback and ideas for improvement.