01:00
Mathnerd314 has joined #mlpack
02:39
agobin has quit [Quit: Connection closed for inactivity]
03:08
travis-ci has joined #mlpack
03:08
<
travis-ci >
mlpack/mlpack#695 (master - fb75f38 : Ryan Curtin): The build passed.
03:08
travis-ci has left #mlpack []
03:20
travis-ci has joined #mlpack
03:20
<
travis-ci >
mlpack/mlpack#696 (master - e14c246 : Ryan Curtin): The build passed.
03:20
travis-ci has left #mlpack []
03:53
travis-ci has joined #mlpack
03:53
<
travis-ci >
mlpack/mlpack#697 (master - 9f72e89 : Ryan Curtin): The build passed.
03:53
travis-ci has left #mlpack []
03:56
Nilabhra has joined #mlpack
05:28
tsathoggua has joined #mlpack
06:05
tsathoggua has quit [Quit: Konversation terminated!]
06:34
Mathnerd314 has quit [Ping timeout: 264 seconds]
10:35
Nilabhra has quit [Read error: Connection reset by peer]
10:59
wasiq has quit [Ping timeout: 252 seconds]
11:41
wasiq has joined #mlpack
12:27
<
ranjan123 >
I am geting this build error : E: Some index files failed to download. They have been ignored, or old ones used instead.
12:30
mentekid has joined #mlpack
12:30
<
ranjan123 >
I not sure but I think the problem is from travis-ci.
12:30
<
ranjan123 >
please correct me if I am wrong
13:29
<
ranjan123 >
any reply ?
13:33
keonkim has joined #mlpack
13:41
<
zoq >
ranjan123: travis could connect to the ubuntu repository to get the necessary packages, so this isn't an mlpack problem. I've restarted the build, let's see what happens.
13:41
<
ranjan123 >
thanks zoq
13:43
<
ranjan123 >
now the ubuntu repository is reachable
13:46
<
ranjan123 >
no. ! zoq it is failing again
14:08
Mathnerd314 has joined #mlpack
14:15
ranjan123 has quit [Quit: Page closed]
14:47
ftuesca has joined #mlpack
15:01
wasiq has quit [Read error: Connection reset by peer]
15:49
Divyam has joined #mlpack
15:50
Divyam has quit [Client Quit]
16:07
mentekid has quit [Ping timeout: 244 seconds]
16:39
Nilabhra has joined #mlpack
17:08
agobin has joined #mlpack
17:45
mentekid has joined #mlpack
18:03
tsathoggua has joined #mlpack
18:26
mentekid has quit [Ping timeout: 244 seconds]
18:32
ank_95_ has joined #mlpack
18:44
ftuesca has quit [Quit: Leaving]
18:54
Nilabhra has quit [Remote host closed the connection]
19:02
virtualgod has quit [Quit: Connection closed for inactivity]
19:47
sumedhghaisas has joined #mlpack
19:53
tsathoggua has quit [Quit: Konversation terminated!]
20:15
travis-ci has joined #mlpack
20:15
<
travis-ci >
mlpack/mlpack#707 (master - 5bc514c : Marcus Edel): The build was broken.
20:15
travis-ci has left #mlpack []
20:35
ank_95_ has quit [Quit: Connection closed for inactivity]
21:27
sumedhghaisas has quit [Ping timeout: 252 seconds]
21:36
mentekid has joined #mlpack
21:46
zoq_ has joined #mlpack
21:47
zoq has quit [Read error: Connection reset by peer]
21:48
zoq_ is now known as zoq
21:49
mentekid has quit [Ping timeout: 240 seconds]
21:51
mentekid has joined #mlpack
22:37
mentekid has quit [Ping timeout: 276 seconds]
22:41
<
zoq >
It would be great to have sgd with momentum, so I'm curious, what's the best way to do this.
22:41
<
zoq >
I modified the minibatch sgd optimizer to get this done, not sure it would be a good idea to write a minibatch sgd momentum optimizer.
22:57
<
rcurtin >
hmm sumedh wrote some SVD updats rules with momentum that could probably be adapted
22:57
<
rcurtin >
but the update step is so similar to just regular SGD
22:58
<
rcurtin >
maybe you could just add momentum to the existing SGD class and set the default momentum to 0 for reverse compatibility?
23:00
<
zoq >
That's the question do I write another optimizer called e.g. sgd minibatch momentum or can I add momentum to the existing minibatch sgd class.
23:01
<
rcurtin >
I think adding to the existing one is fine
23:01
<
rcurtin >
no need to make a new class
23:02
<
rcurtin >
and we don't have lots of different momentum strategies at this point so it probably doesn't make sense (yet?) to use template parameters
23:04
<
zoq >
okay, good, standard momentum works for me.
23:59
tsathoggua has joined #mlpack