2018-11-12 22:39
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/
02:16
vivekp has joined #mlpack
03:12
KimSangYeon-DGU has joined #mlpack
03:36
xiaohong has joined #mlpack
04:06
xiaohong has quit [Ping timeout: 245 seconds]
06:32
xiaohong has joined #mlpack
06:40
xiaohong has quit [Remote host closed the connection]
07:11
eku43 has joined #mlpack
07:14
eku43 has quit [Remote host closed the connection]
07:21
ImQ009 has joined #mlpack
08:02
xiaohong has joined #mlpack
08:08
xiaohong has quit [Remote host closed the connection]
15:16
<
jenkins-mlpack2 >
Ryan Curtin: Add callback documentation to docs.
15:19
<
jenkins-mlpack2 >
Ryan Curtin: Move callbacks.md to the right directory.
15:19
<
jenkins-mlpack2 >
Yippee, build fixed!
15:19
vivekp has quit [Read error: Connection reset by peer]
15:19
<
jenkins-mlpack2 >
Ryan Curtin: Extract callbacks.md from the tar.
15:21
vivekp has joined #mlpack
15:33
travis-ci has joined #mlpack
15:33
<
travis-ci >
mlpack/ensmallen#453 (2.10.0 - e921973 : Ryan Curtin): The build passed.
15:33
travis-ci has left #mlpack []
16:45
vivekp has quit [Ping timeout: 245 seconds]
17:07
drymaiker has joined #mlpack
17:08
drymaiker has left #mlpack []
17:31
vivekp has joined #mlpack
17:39
travis-ci has joined #mlpack
17:39
<
travis-ci >
coatless/ensmallen#15 (patch-2 - 73644c6 : James J Balamuta): The build passed.
17:39
travis-ci has left #mlpack []
19:03
vivekp has quit [Ping timeout: 258 seconds]
19:28
MLJens has joined #mlpack
19:31
<
MLJens >
Hey, hope I'm right here for a question about convolutional MLPack model.
19:32
<
MLJens >
I tried to copy the MNIST example and just use other data and a different Inputsize.
19:32
<
MLJens >
I created an image of all letters of all on my computer installed fonts.
19:32
<
MLJens >
Stuff like "Windows Dings" or "Symbols" I skipped.
19:33
<
MLJens >
The resolution is 100x100 of each input image.
19:33
<
MLJens >
I dimensioned my network the following way:
19:33
<
MLJens >
100x100x1 --conv (6 filters of size 11x11. stride = 1) -> 90x90x6
19:33
<
MLJens >
90x90x6 --------------- Leaky ReLU ---------------------> 90x90x6
19:34
<
MLJens >
90x90x6 --- max pooling (over 3x3 fields. stride = 3) --> 30x30x6
19:34
<
MLJens >
30x30x6 --- conv (16 filters of size 9x9. stride = 1) --> 22x22x16
19:34
<
MLJens >
22x22x16 ------------- Leaky ReLU ---------------------> 22x22x16
19:34
<
MLJens >
22x22x16 - max pooling (over 2x2 fields. stride = 2) --> 11x11x16
19:35
<
MLJens >
11x11x16 ----------------- Dense ----------------------> 62 //[0-9] && [a-z] && [A-Z]
19:35
<
MLJens >
Training my network leads to an access violation
19:35
<
MLJens >
The variables trainX and trainY seem to be ok.
19:35
<
MLJens >
The have dimension: 10000 x 392 and 1 x 392
19:36
<
MLJens >
I'm not really skilled in using MLPack and hope, you can help me a little bit.
19:36
<
MLJens >
If it's helpfull, I can open a ticket at stackoverflow to show you my code....
19:38
<
MLJens >
The error occures in "model.train" routine.
20:13
ImQ009 has quit [Quit: Leaving]