verne.freenode.net changed the topic of #mlpack to: http://www.mlpack.org/ -- We don't respond instantly... but we will respond. Give it a few minutes. Or hours. -- Channel logs: http://www.mlpack.org/irc/
< ShikharJ>
Ah, I noticed that you had mentioned in your GSoC blog regarding implementing a stacking module for GANs and RBMs. Could you tell me what did you mean by a stacking module?
kaushik_ has quit [Quit: Connection closed for inactivity]
aashay has joined #mlpack
govg has joined #mlpack
kris__ has quit [Quit: Connection closed for inactivity]
luffy has joined #mlpack
luffy is now known as Guest60520
Guest60520 has quit [Ping timeout: 260 seconds]
luffy1996 has joined #mlpack
luffy1996 has quit [Ping timeout: 260 seconds]
luffy1996 has joined #mlpack
< luffy1996>
I am interested in working in deep reinforcement learning idea for GSOC 18.
< luffy1996>
I sincerely request to brief me how to proceed.
luffy1996_ has joined #mlpack
daivik has joined #mlpack
luffy1996___ has joined #mlpack
luffy1996_ has left #mlpack []
luffy1996___ has quit [Client Quit]
luffy1996_ has joined #mlpack
luffy1996 has quit [Ping timeout: 260 seconds]
luffy1996_ is now known as luffy1996
desai-aditya has joined #mlpack
< desai-aditya>
I am planning to write a blog or do a short series on video tutorials on how to use mlpack to achieve two things - 1) demonstrate how much I've learnt about it.. 2) help the people who want to get started (with videos , pretty much anyone can follow) . How's the idea??
< desai-aditya>
@rcurtin, @zoq : Do you think I should proceed with the idea? (that of making a series of videos or blogs)
alsc has joined #mlpack
alsc has quit [Client Quit]
desai-aditya has quit [Ping timeout: 260 seconds]
kaushik_ has joined #mlpack
sumedhghaisas2 has joined #mlpack
sumedhghaisas has quit [Ping timeout: 240 seconds]
< zoq>
desai-adity: That is a great idea, if you need any help let me know.
< zoq>
luffy1996: Hello, a good starting point is to go through the codebase, especially the rl part (src/mlpack/methods/reinforcement_learning/), checkout and run the tests: rl_components_test.cpp and q_learning_test.cpp, you can run each with: 'bin/mlpack_test -t RLComponentsTest' and 'bin/mlpack_test -t QLearningTest'. If you see something that you think could be improved or extend please feel free to open a PR.
< zoq>
luffy1996: Also if you like, you can work on a simple RL method e.g. (stochastic) Policy Gradients, but don't feel obligated.
sumedhghaisas2 has quit [Ping timeout: 260 seconds]
alsc has joined #mlpack
sumedhghaisas has joined #mlpack
vikash has joined #mlpack
alsc has quit [Ping timeout: 264 seconds]
alsc has joined #mlpack
vikash has quit [Ping timeout: 240 seconds]
< kaushik_>
zoq: hi can you please specify a starting task for the "string processing utilities" project
alsc has quit [Ping timeout: 260 seconds]
govg has quit [Ping timeout: 264 seconds]
alsc has joined #mlpack
desai-aditya has joined #mlpack
alsc has quit [Quit: alsc]
aneesh has joined #mlpack
< aneesh>
Hello all, I was wondering if mlpack can be run on windows with a docker image
< aneesh>
(I am aneesh, a CSE student, deep learning enthusiast, an aspirant for GSOC 18)
< rcurtin>
aneesh: if I understand correctly you mean that you plan to run a Linux Docker image from a Windows system
< rcurtin>
if this is the case, then yes, you should be able to build and run mlpack just fine
< aneesh>
Yes, I wanted to avoid installing all the dependencies for mlpack. Docker seems to be the way
< rcurtin>
I agree, I think that would be a good way to avoid installing dependencies
< aneesh>
Is there an existing docker image?
< rcurtin>
to be honest I think mlpack is a lot easier to develop for on Linux too
< rcurtin>
no, I'd just use an ubuntu image, install the dependencies, and then build mlpack
< rcurtin>
it is possible to build and use mlpack on Windows but it is a bit trickier
< aneesh>
hmmm, I prefer linux. It's just that my current "powerful machine" runs windows. I will partition it or just use it in a virtual box
< rcurtin>
yeah, either docker or virtualbox should work just fine for you
< aneesh>
I was also curious, what advantages would mlpack have over other libraries like tensorflow...
< aneesh>
w.r.t deep learning. tensorflow doesn't have more ML based algos like KNNs which are present in sklearn
< rcurtin>
flexibility, the ability to work natively in C++, a build system that isn't awful, more algorithms than just those that can be expressed by deep learning
< rcurtin>
really it depends on specifically what the user wants to do
< aneesh>
rcurtin: thanks for the info
aneeshj has joined #mlpack
aneesh has quit [Quit: Page closed]
krishnapaul has joined #mlpack
krishnapaul has quit [Quit: Page closed]
govg has joined #mlpack
sumedhghaisas2 has joined #mlpack
sumedhghaisas has quit [Ping timeout: 256 seconds]
ImQ009 has joined #mlpack
prakhar_code[m] has joined #mlpack
sumedhghaisas2 has quit [Read error: Connection reset by peer]
sumedhghaisas has joined #mlpack
desai-aditya has quit [Remote host closed the connection]
sumedhghaisas has quit [Read error: Connection reset by peer]
sumedhghaisas has joined #mlpack
ShikharJ has joined #mlpack
< ShikharJ>
rcurtin: Are you there?
< rcurtin>
ShikharJ: yes
< ShikharJ>
I had talked regarding this to Marcus earlier, but i thought I should talk to you regarding this as well. I had planned on proposing to implement Wasserstein GAN, DCGAN, SeqGAN and StackGAN, but Marcus suggested that I should try to focus on two out of them. How would you prioritize them?
desai-aditya has joined #mlpack
< ShikharJ>
rcurtin: ?
< rcurtin>
ShikharJ: I juggle a lot of things during a single day so I am not always able to respond immediately. Please be patient ...
< rcurtin>
I have no particular priority for any of them, personally
< rcurtin>
I am not sure I am the right person to ask
< rcurtin>
since I would not be mentoring the project
< ShikharJ>
Ah, cool, sorry for disturbing you. If I could ask, would Mikhail and Marcus be mentoring this project this year?
< rcurtin>
I would assume either of them would be mentoring it, yes, but I don't know their exact plans
< rcurtin>
keep in mind, we are likely to receive nearly 100 applications for lots of different projects. so it is hard to know which students will apply to what projects, and even then we do not know (and will not know) which projects will be selected
< rcurtin>
it may be that no GAN project is selected at all and they may choose to mentor other things
< rcurtin>
but at this time, none of that is known
< rcurtin>
that doesn't necessarily make a huge difference for you, but I thought that I should point it out so you are a little clearer on what this process looks like from our side :)
< ShikharJ>
This makes sense. Thanks.
ShikharJ has quit [Quit: Page closed]
robertohueso has joined #mlpack
pawan_sasanka has joined #mlpack
< pawan_sasanka>
Hello
pawan_sasanka is now known as pawsed
pawsed_ has joined #mlpack
pawsed has quit [Ping timeout: 248 seconds]
pawsed_ is now known as pawsed
< pawsed>
Oh so how do i get started if i want to contribute?