ChanServ changed the topic of #mlpack to: Due to ongoing spam on freenode, we've muted unregistered users. See http://www.mlpack.org/ircspam.txt for more information, or also you could join #mlpack-temp and chat there.
Dwarf23 has joined #mlpack
Dwarf23 has quit [Remote host closed the connection]
eNigmaFx24 has joined #mlpack
eNigmaFx24 has quit [Read error: Connection reset by peer]
Pilfers has joined #mlpack
Pilfers has quit [Remote host closed the connection]
witness has joined #mlpack
CrunchyChewie28 has joined #mlpack
CrunchyChewie28 has quit [Remote host closed the connection]
Lausefuchs22 has joined #mlpack
Lausefuchs22 has quit [Ping timeout: 256 seconds]
channels has joined #mlpack
channels has quit [Remote host closed the connection]
Shikhar has joined #mlpack
< Shikhar>
zoq: Ah, okay. Though could this potentiall also fix the SSRBM error that we were observing?
Shikhar has quit [Quit: Page closed]
dfgg29 has joined #mlpack
dfgg29 has quit [Remote host closed the connection]
channels has joined #mlpack
channels has quit [Remote host closed the connection]
nkuttler22 has joined #mlpack
nkuttler22 has quit [Remote host closed the connection]
< zoq>
Shikhar: Yes, that should solve the other issue as well.
CGML2 has joined #mlpack
CGML2 has quit [Remote host closed the connection]
Kirito has joined #mlpack
Kirito has quit [Remote host closed the connection]
sushichef has joined #mlpack
sushichef has quit [Remote host closed the connection]
pierte has joined #mlpack
pierte has quit [Remote host closed the connection]
preaction29 has joined #mlpack
Kinsmir27 has joined #mlpack
Kinsmir27 has quit [Remote host closed the connection]
preaction29 has quit [Remote host closed the connection]
cjlcarvalho has joined #mlpack
cjlcarvalho has quit [Read error: Connection reset by peer]
rs-afk has joined #mlpack
rs-afk has quit [Remote host closed the connection]
cjlcarvalho has joined #mlpack
cjlcarvalho has quit [Read error: Connection reset by peer]
witness has quit [Quit: Connection closed for inactivity]
ChickeNES has joined #mlpack
ChickeNES has quit [Remote host closed the connection]
cjlcarvalho has joined #mlpack
cjlcarvalho has quit [Client Quit]
crayfishx14 has joined #mlpack
crayfishx14 has quit [Remote host closed the connection]
ImQ009 has joined #mlpack
cjlcarvalho has joined #mlpack
cjlcarvalho has quit [Remote host closed the connection]
cjlcarvalho has joined #mlpack
caiojcarvalho has joined #mlpack
cjlcarvalho has quit [Ping timeout: 268 seconds]
ManyRaptors27 has joined #mlpack
ManyRaptors27 has quit [Remote host closed the connection]
caiojcarvalho has quit [Quit: Konversation terminated!]
cjlcarvalho has joined #mlpack
< zoq>
ShikharJ: I'll take a look at the conv PR later today, just wanted to make sure the test issues are resolved first.
Pyrrhic has joined #mlpack
Pyrrhic has quit [Remote host closed the connection]
cjlcarvalho has quit [Ping timeout: 260 seconds]
christel19 has joined #mlpack
christel19 has quit [Remote host closed the connection]
cjlcarvalho has joined #mlpack
cjlcarvalho has quit [Ping timeout: 256 seconds]
Sigals has joined #mlpack
vivekp has quit [Ping timeout: 272 seconds]
Sigals has quit [Remote host closed the connection]
Pugabyte0 has joined #mlpack
Pugabyte0 has quit [K-Lined]
ImQ009 has quit [Quit: Leaving]
Sharker has joined #mlpack
Sharker has quit [Remote host closed the connection]
me` has joined #mlpack
me` has quit [Excess Flood]
me` has joined #mlpack
me` has quit [Excess Flood]
< rcurtin>
disappointing, I see the spam is still going on
xa0 has quit [Excess Flood]
xa0 has joined #mlpack
xa0 has quit [Excess Flood]
< zoq>
not sure there is anything we can do
< rcurtin>
yeah, I think what we've done is the right thing
< rcurtin>
I see also the spam has descended into vulgar ASCII art ...
xa0 has joined #mlpack
xa0 has quit [Excess Flood]
< zoq>
interesting to see that the temp channel isn't affected
xa0 has joined #mlpack
xa0 has quit [Excess Flood]
xa0 has joined #mlpack
xa0 has quit [Excess Flood]
xa0 has joined #mlpack
xa0 has quit [Excess Flood]
xa0 has joined #mlpack
< rcurtin>
yeah, it seems like a lot of "not popular" channels don't have spam going on
< rcurtin>
so I guess the operators of the botnet must have just found a list of publicly advertised channels or something
< zoq>
Hopefully, at some point, they stop the attack.
< rcurtin>
that or the spammers get bored and move on... :)
cjlcarvalho has joined #mlpack
msm17 has joined #mlpack
msm17 has quit [Remote host closed the connection]
cjlcarvalho has quit [Ping timeout: 240 seconds]
cjlcarvalho has joined #mlpack
< akhandait>
zoq: ShikharJ: Sorry it took so long to open the PR. I obviously underestimated the time it would take me to get it done completely.
< rcurtin>
akhandait: I think I underestimate the time things will take me every day, you are not alone in that respect :)
< akhandait>
Oh, glad to know I am not alone. :)
< zoq>
akhandait: No problem at all, huge thanks for looking into the code.
< akhandait>
zoq: Oh it was a lot of fun actually, I got to understand things a lot better by working on this.
< akhandait>
This also helped reduce the time to train on celebA by a huge margin, so I am glad I did this.
< akhandait>
zoq: There was a thing I wanted to discuss with you, about support for conditional VAEs.
< zoq>
Oh that's great to hear :)
< zoq>
sure
< akhandait>
I know that officially GSoC ends tomorrow, but there are a couple things I wish to do that remain
< akhandait>
I think, I will open two WIP PRs tomorrow and continue working on them thereafter, I have already wrote some code I wish to add
< zoq>
More than happy to work with everyone after GSoC has ended; this is an Open Source project after all.
< zoq>
So, yeah happy to take a look at the PR's and extenting the work sounds like a great idea to me.
< akhandait>
So, about conditional VAEs, in that we have to concatenate labels to the output of the Reparametrization layer(input of decoder) in the middle of a forward pass through the network, I think currently we can't do that.
< akhandait>
Is there a way?
< akhandait>
I tried doing a partial forward pass and then another forward pass with labels concatenated but the gradients can't flow then.
< akhandait>
If there isn't a way to do this currently, I had an idea.
< zoq>
Since it could be in the middle of the network, and there are layers that don't need the extra information, I think there is no way to use the subview layer, so currently there is no way.
< akhandait>
zoq: So, will it be a good idea to add a new Concatenate layer which concatenates a constant matrix to the input.
andries21 has joined #mlpack
< zoq>
hm, that means we can't use e.g. the linear layer before the Reparametrization layer right?
andries21 has quit [Remote host closed the connection]
< zoq>
since the layer would process the input
< akhandait>
zoq: I think it won't affect this, we can still use that
< akhandait>
zoq: The lyer will just concatenate a matrix to the input of the layer, it won't affect the rest of the network
< zoq>
ah I see, that should work
< zoq>
and once again there is no need to change the existing codebase
< akhandait>
zoq: Yeah, I was just wondering what's the cost of adding a new layer to the module?
< zoq>
With module you mean the network codebase? build time should increase slightly but that is for almost every addition.
caiojcarvalho has joined #mlpack
< akhandait>
akhandait: Oh okay then, this layer should be fairly easy to do, but I have learnt never to underestimate tasks that seem easy at first. :p
< zoq>
haha perhaps not tomorrow :)
cjlcarvalho has quit [Ping timeout: 272 seconds]
< akhandait>
zoq: Haha yeah, still I will try to do it ASAP.
caiojcarvalho has quit [Quit: Konversation terminated!]
cjlcarvalho has joined #mlpack
deetwelve13 has joined #mlpack
deetwelve13 has quit [Remote host closed the connection]
thurin20 has joined #mlpack
thurin20 has quit [Remote host closed the connection]
badpixel15 has joined #mlpack
badpixel15 has quit [Killed (Unit193 (Spam is not permitted on freenode.))]
cjlcarvalho has quit [Quit: Konversation terminated!]