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/
jeffin143 has joined #mlpack
< jeffin143> lozhnikov : could you help me with travis build issue , cmake properties are profile and debug off* , for some reason my local computer doesn't throw that error* after building
jeffin143 has quit [Ping timeout: 276 seconds]
favre49 has joined #mlpack
< favre49> This seems to be the same issue with the travis build affecting #1941 and #1945
favre49 has quit [Remote host closed the connection]
jeffin143 has joined #mlpack
jeffin143 has quit [Client Quit]
favre49 has joined #mlpack
< favre49> zoq I'll be going out of station from 4th to 10th of July. I'll still be working, but my output might be less and I may be less responsive for that period. Hope thats fine
favre49 has quit [Remote host closed the connection]
xiaohong has joined #mlpack
xiaohong has quit [Ping timeout: 260 seconds]
< jenkins-mlpack2> Project docker mlpack nightly build build #373: STILL UNSTABLE in 3 hr 23 min: http://ci.mlpack.org/job/docker%20mlpack%20nightly%20build/373/
< zoq> favre49: Sure, have fun!
< zoq> jeffin143 favre49: Looks like some issue with the pytest package.
< lozhnikov> jeffin143: I am not sure it's your fault. Maybe Travis settings have been changed. Try to merge mlpack master branch into your branch.
< zoq> lozhnikov: Same issue with the latest master: https://github.com/mlpack/mlpack/pull/1945
< zoq> lozhnikov: So I agree it's not Jeffin's fault.
< lozhnikov> Yes, I see. According to the build history it started 2-3 days ago. Now pytest throws some errors. And actually pytest-5.0.0 was released on the 29th of June. Possibly it's not a coincidence.
jeffin143 has joined #mlpack
< jeffin143> zoq lozhnikov : ok, probably I should have just asked you before hand, tried debugging it , for a full day :-p
xiaohong has joined #mlpack
KimSangYeon-DGU has joined #mlpack
xiaohong has quit [Ping timeout: 260 seconds]
< Toshal> ShikharJ: Sorry, I was just busy in my last week
< Toshal> Now I am highly free from today.
< Toshal> I will start working dedicatedelly now onwards.
< Toshal> Yesterday I got placed in my college's first company. And now I am quite free to work.
< Toshal> I Write blog soon.
< Toshal> Sorry for the inconveinence.
jeffin143 has quit [Ping timeout: 252 seconds]
< Toshal> ShikharJ: I am here
< ShikharJ> Toshal: sakshamB: Okay, let's make this one quick.
< sakshamB> ShikharJ: alright
< Toshal> Yes
< ShikharJ> First of congrats to you both, sorry I wasn't that active this month, but I think things should be a lot better this phase.
< Toshal> ShikharJ: Thanks
< ShikharJ> Toshal: Oh, okay, you could have let me know that your college's placement season was going on, anyways, do you have a set task to do this week?
< Toshal> I was thinking to work on FID this week.
< ShikharJ> sakshamB: As I understand, you were working on spectral norm? What's the progress on that.
< Toshal> Let me know your thoughts on it.
< Toshal> If you have any other plans please let me know.
< ShikharJ> sakshamB: Also, I'm not sure what the check Inception score to-do item means on your PR, could you elaborate what you have in mind regarding that?
< sakshamB> ShikharJ: yes I am trying to finish my work on regularizers and starting on spectralNorm. I am currently trying to understand the paper and looking at other implementations. Have written the base code for the layer.
< ShikharJ> Toshal: OKay, FID sounds good. Go ahead. I'll review the remaining PRs soon hopefully.
< sakshamB> ShikharJ: train a GAN model and compute the inception score for generated images. Compare Inception score for different GAN models.
< ShikharJ> sakshamB: Do you wish to implement the test in that PR itself?
< ShikharJ> Ah I remember you mentioned about the total runtime being out of scope.
< sakshamB> ShikharJ: yes we discussed this before
< ShikharJ> In that case, do you think you could push a commented out script of the test?
< ShikharJ> So that if users wish to do that, they can do it without hassle?
< sakshamB> sakshamB: yes I can do that. Or maybe we can add that to the models repository.
< ShikharJ> sakshamB: Yeah, just add it to the PR for now. Moving to the models repository is something we wish to do in the future.
< sakshamB> ShikharJ: alright, I will finish that.
< ShikharJ> Toshal: sakshamB: Okay, you guys have set tasks to do in this week. All the best. If you need help, please leave your messages in the IRC or on the PR. I'll log off for now, have a good week.
< Toshal> ShikharJ: Okay Thanks.
< sakshamB> ShikharJ: you too. thanks :)
KimSangYeon-DGU has quit [Remote host closed the connection]
favre49 has joined #mlpack
< favre49> zoq I was thinking for the NEAT config feature, I could make a NEATConfig class, which would be serializable itself, and that could be passed to the constructor for NEAT, as well as the genomes. I would keep the old constructors as well, too, and they would create a NEATConfig instance themselves. The model by itself would be serializable as well,
< favre49> to save and load progress.
< favre49> Does this make sense, or do you have a different idea?
< zoq> favre49: I was thinking if we would serialize the NEAT class, do we need a seperate config class?
< favre49> If we serialize the NEAT class, we would also be serializing the current population and related statistics, which we would need to save, but not for a config file, is why I thought we should put it in a different file
< zoq> Right, what we could do is to check if the model files contains genome settings, etc. That way we could probably get the same format as with an extra class.
< favre49> Ah right, that makes sense. I'll try that (y)
favre49 has quit [Remote host closed the connection]
KimSangYeon has quit [Remote host closed the connection]