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/
divyansh997 has quit [Ping timeout: 250 seconds]
travis-ci has joined #mlpack
< travis-ci> mlpack/mlpack#6591 (master - 2635297 : Marcus Edel): The build passed.
travis-ci has left #mlpack []
robb9 has joined #mlpack
< robb9> What is the default training optimizer for an FFN?
< robb9> thank you. I have another question. Ever since I started using a constant initializer, all of my networks seems to have the same MSE regardless of amount of layers, neurons, etc... Is there any reasoning for this?
< robb9> it's the same training data, but I would assume that the predictions would change based on the amount of neurons and layers
< robb9> nevermind, I think it was because linear layers technically do not stack
jeffin143 has quit [Ping timeout: 246 seconds]
robb9 has quit [Quit: Page closed]
Toshal has joined #mlpack
Toshal has quit [Read error: Connection reset by peer]
V has joined #mlpack
V is now known as Guest32828
rajiv_ has joined #mlpack
jeffin143 has joined #mlpack
< rajiv_> zoq: I have 2 PRs open but I'm finding it difficult to work on them due to semester exams... Working on them will help me work on the GSoC project more smoothly... So can I work on them in the Community Bonding period? Can I add those PRs to the timeline of community bonding?
< jeffin143> Rcurtin : have shared my proposal on the dashboard, if you are anytime free, please provide with feedback
< jeffin143> Thanks
< jeffin143> And if not free , no issues with it , i can understand. Thanks
Guest32828 has quit [Ping timeout: 256 seconds]
rajiv_ has quit [Quit: Page closed]
favre49 has joined #mlpack
< favre49> zoq: I've uploaded the draft proposal, looking forward to your feedback :)
favre49 has quit [Client Quit]
rf_sust2018 has joined #mlpack
rf_sust2018 has quit [Ping timeout: 245 seconds]
rf_sust2018 has joined #mlpack
rf_sust2018 has quit [Read error: Connection reset by peer]
rf_sust2018 has joined #mlpack
mulx10 has joined #mlpack
< mulx10> Hemal: sorry for the delay. I guess you can have a look at
mulx10 has quit [Ping timeout: 256 seconds]
rf_sust2018 has quit [Quit: Leaving.]
< jenkins-mlpack2> Project docker mlpack nightly build build #274: UNSTABLE in 3 hr 36 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/274/
dendrite has joined #mlpack
< dendrite> i want to implement String Processing Utilities
< dendrite> Should i put my proposal here to get reviewed by others
jeffin143 has quit [Ping timeout: 250 seconds]
rf_sust2018 has joined #mlpack
gauravcr7rm has joined #mlpack
< gauravcr7rm> Zoq: I have uploaded my proposal on dashboard. Looking forward for feedback from your side. Thanks in advance
mulx10 has joined #mlpack
gauravcr7rm has quit [Ping timeout: 256 seconds]
mulx10 has quit [Ping timeout: 256 seconds]
dendrite has quit [Quit: Page closed]
halwai has joined #mlpack
rajiv_ has joined #mlpack
jeffin143 has joined #mlpack
Suryo has joined #mlpack
halwai has quit [Ping timeout: 256 seconds]
< Suryo> zoq, rcurtin: I have shared a draft of my proposal, too! It would be wonderful if you could tell me what to change or improve.
< Suryo> zoq: I'll be working on changes with respect to PR#86. My final proposal will also mention the changes I've made and the state at which the PR is after the updates. I hope that's okay.
< rajiv_> zoq: Is there a way to import pretrained weights into FFN?
rajiv_ has quit [Client Quit]
< zoq> Suryo: Sure, will take a look at the proposal this week.
< Suryo> Thanks again, zoq!
Suryo has quit [Quit: Page closed]
Suryo has joined #mlpack
< Suryo> I've also included my resume at the end, I hope that's okay. I didn't see any other place to input the resume, so...
Suryo has quit [Client Quit]
abhi_ has joined #mlpack
abhi_ has quit [Client Quit]
rf_sust2018 has quit [Quit: Leaving.]
< zoq> Suryo: That works.
rf_sust2018 has joined #mlpack
rf_sust2018 has quit [Quit: Leaving.]
Lorcan has joined #mlpack
Lorcan has quit [Client Quit]
Lorcan has joined #mlpack
jeffin143 has quit [Ping timeout: 250 seconds]
sayan__ has joined #mlpack
Abdulrahman has joined #mlpack
Abdulrahman has quit [Client Quit]
< ShikharJ> rcurtin: I didn't find any issues with the links, so it seems all good.
mavis_ has joined #mlpack
mavis_ has quit [Ping timeout: 256 seconds]
< rcurtin> ShikharJ: awesome, thanks for checking :)
AndroUser has joined #mlpack
Hemal has joined #mlpack
< AndroUser> Could some one check my pr1817, appveyor is failing
< AndroUser> Couldn't understand what exactly the issues is
< AndroUser> Building fine in my local environment , and also travis
< AndroUser> Bt building failed in appveyor
< rcurtin> AndroUser: the AppVeyor logs are really long but if you search for "error " (with the space included) it should bring you right to the line
< AndroUser> Thanks would do so*
< Hemal> zoq, rcurtin : I have shared my draft proposal for Google Summer of Code. It would be great if you could go through it and suggest improvements. Thank you !
< AndroUser> U are great recurtin*
< AndroUser> I tried almost for 5 year
< AndroUser> To read line by line, couldn't get that thing
< AndroUser> Thanks*
< AndroUser> 5 hours .. :) typo mistake
< Hemal> mulx10: thanks for the link.
Hemal has quit [Quit: Leaving.]
robb10 has joined #mlpack
abhinav__ has joined #mlpack
abhinav__ has quit [Client Quit]
rf_sust2018 has joined #mlpack
robb10 has quit [Ping timeout: 256 seconds]
rf_sust20181 has joined #mlpack
rajiv_ has joined #mlpack
rf_sust2018 has quit [Ping timeout: 250 seconds]
nagyation has joined #mlpack
< rajiv_> rcurtin, zoq, Shikharj: I have submitted my proposal draft "Implementation of object detection algorithms" in the official website... It would be great if you could review and provide suggestions to improve it :)
rajiv_ has quit [Client Quit]
< ShikharJ> rcurtin: Hmm, did you get my previous message (I sent it using "/msg format")?
AndroUser has quit [Read error: Connection reset by peer]
Lorcan has quit [Remote host closed the connection]
vivekp has joined #mlpack
nagyation has quit [Quit: nagyation]
nagyation has joined #mlpack
Lorcan has joined #mlpack
< rcurtin> ShikharJ: I did, just responded. but I was out for a while, so I didn't see it
< rcurtin> AndroUser: glad you got it figured out and that you did not spend five years working on it :)
< rcurtin> but I will say, I think it might take five years to read the full appveyor log...
saksham189 has joined #mlpack
sayan__ has quit [Ping timeout: 256 seconds]
KimSangYeon-DGU_ has joined #mlpack
Soonmok has joined #mlpack
KimSangYeon-DGU has quit [Ping timeout: 256 seconds]
KimSangYeon-DGU_ has quit [Quit: Page closed]
ryuk has joined #mlpack
KimSangYeon-DGU has joined #mlpack
< ryuk> what are the prerequisites needed for this program
nagyation has quit [Quit: nagyation]
nagyation has joined #mlpack
sayan__ has joined #mlpack
nagyation has quit [Quit: nagyation]
sayan__ has quit [Ping timeout: 256 seconds]
nagyation has joined #mlpack
nagyation has quit [Quit: nagyation]
robb9 has joined #mlpack
sreenik has joined #mlpack
jeffin143 has joined #mlpack
mulx10 has joined #mlpack
< mulx10> Hello, I needed help regarding Concat Layer. Is it possible to concat along channels from two conv layer?
< sreenik> zoq: Do you remember I was having an issue with creating an FFN model where I wanted to declare the init type and loss type separately and not together as in `FFN<MeanSquaredError<>, GlorotInitialization> model`. So you had suggested that I take a look at the kernel_pca_main.cpp file (now this has only the kernelType unknown, but I have two both the init type and loss type unknown).
< sreenik> If something like model.setLossType(MeanSquaredError<>) and model.setInitType(GlorotInitialization) existed, my problem would have been solved. Can you suggest a workaround? :)
Lorcan has quit [Remote host closed the connection]
< robb9> Is nvBlas support for armadillo tested? It loads onto my GPU and everything but then does a Segmentation Fault
< robb9> it works with OpenBlas and Lapack fine though, with no segfault
< robb9> was just hoping for some gpu acceleration
ryuk has quit [Ping timeout: 256 seconds]
< robb9> ah, nevermind, it was a problem with the nvblas config file. You have to specify a CPU fallback
robb9 has quit [Quit: Page closed]
rf_sust20181 has quit [Quit: Leaving.]
nagyation has joined #mlpack
mulx10 has quit [Ping timeout: 256 seconds]
nsquare has joined #mlpack
nsquare has quit [Client Quit]
namelessstar has joined #mlpack
Lorcan has joined #mlpack
nagyation has quit [Quit: nagyation]
nagyation has joined #mlpack
Lorcan has quit [Ping timeout: 246 seconds]
nagyation has quit [Quit: nagyation]
nagyation has joined #mlpack
nagyation has quit [Quit: nagyation]
nagyation has joined #mlpack
nagyation has quit [Quit: nagyation]
nagyation has joined #mlpack
nagyation has quit [Remote host closed the connection]
nagyation has joined #mlpack
< zoq> sreenik: And I think you don't know the type at compile time?
nagyation has quit [Quit: nagyation]
namelessstar has quit [Quit: Page closed]
< sreenik> zoq: Yep. I don't. The CLI which I am working on gets the init type and loss type as strings. So depending on that the specific type needs to be selected
< zoq> sreenik: I'l lsketch something up later today.
< sreenik> zoq: I have been trying to find a solution for quite some time now, I think I have been close but couldn't just hit the right block. Thanks, looking forward to your suggestion :)
Lorcan has joined #mlpack
jeffin143 has quit [Read error: Connection reset by peer]
vnr has joined #mlpack
vnr has quit [Quit: Page closed]
Soonmok has quit [Quit: Connection closed for inactivity]
impatidar has joined #mlpack
impatidar has quit [Client Quit]
sreenik has quit [Quit: Page closed]
arun_ has joined #mlpack
arun_ has quit [Client Quit]
saksham189 has quit [Quit: Page closed]
Lorcan has quit [Remote host closed the connection]
jeffin143 has joined #mlpack
jeffin143 has quit [Read error: Connection reset by peer]
jeffin143 has joined #mlpack
jeffin143 has quit [Ping timeout: 246 seconds]
< zoq> sreenik: https://gist.github.com/zoq/254ff987c4c767a3afc484fd3406262c this is what I had in mind.
< zoq> sreenik: We could implement something like: model.setInitType(GlorotInitialization) this as well, but perhaps what I proposed works?
< rcurtin> you can actually use multiple functions to "build up" the template type of the FFN
< rcurtin> I'll add a comment to the gist
< rcurtin> I think that idea is done in a few different mlpack programs, maybe CF? can't remember
< zoq> ahh, nice that reduces the number of lines
< zoq> mulx10: Unfortunately that's not implemented right now.
sunanda has quit [Quit: Connection closed for inactivity]
nagyation has joined #mlpack
robb9 has joined #mlpack
< robb9> how would I go about using k-means to classify 1d time-series data? Or should my second dimension be the time?
< robb9> or should the second dimension be a constant?
robb9 has quit [Ping timeout: 256 seconds]
nagyation has quit [Quit: nagyation]