mlpack IRC logs, 2018-05-15

Logs for the day 2018-05-15 (starts at 0:00 UTC) are shown below.

>
May 2018
Sun
Mon
Tue
Wed
Thu
Fri
Sat
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
--- Log opened Tue May 15 00:00:20 2018
02:40 < rcurtin> ok, I changed the configuration so it should only send one email and one IRC notification on failing tests
02:40 < rcurtin> let's see if that actually works...
02:40 < rcurtin> or maybe we will get 1000 more messages :)
04:26 < jenkins-mlpack> Project docker mlpack nightly build build #318: STILL UNSTABLE in 2 hr 14 min: http://masterblaster.mlpack.org/job/docker%20mlpack%20nightly%20build/318/
06:17 -!- ImQ009 [~ImQ009@unaffiliated/imq009] has joined #mlpack
06:26 -!- ImQ009_ [~ImQ009@unaffiliated/imq009] has joined #mlpack
06:30 -!- ImQ009 [~ImQ009@unaffiliated/imq009] has quit [Ping timeout: 240 seconds]
06:53 -!- travis-ci [~travis-ci@ec2-54-158-40-27.compute-1.amazonaws.com] has joined #mlpack
06:53 < travis-ci> ShikharJ/mlpack#144 (AtrousConv - 1d0ce3f : Shikhar Jaiswal): The build has errored.
06:53 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/84e6c520df2c...1d0ce3ffb29a
06:53 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/379061871
06:53 -!- travis-ci [~travis-ci@ec2-54-158-40-27.compute-1.amazonaws.com] has left #mlpack []
09:08 -!- sumedhghaisas2 [~yaaic@85.255.234.10] has joined #mlpack
09:08 -!- sumedhghaisas [~yaaic@host-92-8-33-72.as43234.net] has quit [Ping timeout: 248 seconds]
09:38 -!- sumedhghaisas2 [~yaaic@85.255.234.10] has quit [Read error: Connection reset by peer]
09:39 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
09:45 < jenkins-mlpack> Project docker mlpack nightly build build #319: STILL UNSTABLE in 2 hr 31 min: http://masterblaster.mlpack.org/job/docker%20mlpack%20nightly%20build/319/
10:04 -!- P [2f0f3558@gateway/web/freenode/ip.47.15.53.88] has joined #mlpack
10:04 -!- P is now known as Guest26708
10:05 -!- Guest26708 [2f0f3558@gateway/web/freenode/ip.47.15.53.88] has quit [Client Quit]
11:20 -!- sumedhghaisas2 [~yaaic@85.255.234.60] has joined #mlpack
11:21 -!- sumedhghaisas2 [~yaaic@85.255.234.60] has quit [Read error: Connection reset by peer]
11:21 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
11:21 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
11:22 -!- travis-ci [~travis-ci@ec2-54-158-40-27.compute-1.amazonaws.com] has joined #mlpack
11:22 < travis-ci> ShikharJ/mlpack#145 (LayerNorm - 11cbce4 : Shikhar Jaiswal): The build has errored.
11:22 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/2f18c223f02a...11cbce4e1d7d
11:22 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/379150087
11:22 -!- travis-ci [~travis-ci@ec2-54-158-40-27.compute-1.amazonaws.com] has left #mlpack []
11:28 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
11:32 -!- sumedhghaisas [~yaaic@85.255.234.60] has joined #mlpack
11:44 -!- sumedhghaisas [~yaaic@85.255.234.60] has quit [Read error: Connection reset by peer]
11:44 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
11:46 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
11:49 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Read error: Connection reset by peer]
11:49 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
11:51 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
11:58 < ShikharJ> zoq: Are you there?
12:06 < zoq> ShikharJ: Currently in a meeting, but I answer as soon as possible.
12:09 < ShikharJ> Sure, just wanted to give an update regarding future work. Since the PRs on Layer Norm and Dilated Convolutions are complete now (I don't think any major changes would be required), I'll start debugging the GAN PR and see if I can get that merged.
12:11 < ShikharJ> Most of the smaller changes are now done (except for Weight Clipping and Gradient Penalty methods, but that'll have to wait for WGAN work to commence). So we can start off with the big tasks now.
12:14 -!- travis-ci [~travis-ci@ec2-54-205-2-111.compute-1.amazonaws.com] has joined #mlpack
12:14 < travis-ci> ShikharJ/mlpack#146 (LayerNorm - 1ea8586 : Shikhar Jaiswal): The build has errored.
12:14 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/11cbce4e1d7d...1ea8586a5215
12:14 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/379167604
12:14 -!- travis-ci [~travis-ci@ec2-54-205-2-111.compute-1.amazonaws.com] has left #mlpack []
12:18 < zoq> That sounds like a really good plan to me, don't think there is much work left for the conv operations if any; so I guess this could be merged by the end of the week.
12:18 < zoq> Getting stable results for the existing GAN code is tricky, but I'm sure we can figure out what went wrong.
12:19 < ShikharJ> One issue was with the Naive Convolution Rule itself. So maybe we can expect different results now.
12:21 < ShikharJ> Also, now that I have understood a lot about the existing codebase, I'm hoping this would be easier now :)
12:29 < zoq> ShikharJ: Right, that was definitely an issue, perhaps that solves all the issues :)
12:57 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
12:57 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
14:08 < ShikharJ> zoq: I was wrong, I found another bug in the Batch Normalization layer (I'm guessing it didn't affect the outputs much). I'll recheck the tests again against tensorflow's output.
14:09 < zoq> ShikharJ: Oh, okay.
14:10 < ShikharJ> This would also need to be corrected in Layer Normalization.
14:14 < zoq> right, hopefully you don't have to refactor the complete layer
14:21 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has joined #mlpack
14:25 < manish7294> rcurtin: Are you available for a short talk?
14:28 < zoq> hm, looks like the job prioritization doesn't work as I thought, perhaps it's because the matrix build is a single job ... not sure.
14:28 < zoq> I guess an easy solution is to use one of the benchmark systems in such a situation.
14:28 < zoq> If we agree on that, I'll go ahead and install the necessary packages.
14:29 < manish7294> rcurtin: I assume you are not here. Well, I am leaving some of my findings here. So, that you can have a look when you are back.
14:30 < manish7294> rcurtin: The same SDP which was giving bad_alloc() for primal dual work seamlessly with LRSDP, though it didn't converge when I initialized impostors part.
14:31 < manish7294> rcurtin: Yesterday while calculating the memory for matrix, we missed the SparseA matrices vector, which has constraint times objective function size.
14:31 < ShikharJ> zoq: No major issues, just had to add a pair of braces. It didn't affect the tests at all, probably the reason why it remained undetected.
14:33 < manish7294> rcurtin: Currently, I have created a pull request w.r.t inequality constraints, though some tests are failing. will look into that. All that remains in this part is to adapt primal dual with inequality constraints.
14:34 < manish7294> rcurtin: The main problem I am facing is, how we are going to initialize Eijl (slack) part of the matrix from yesterday's repersentation.
14:36 < manish7294> rcurtin: As we take impostors and target neighbors constraints independently (leaving us with 2 * k * n constraints), how will we be able to initialize sparseA matrix, as it depends on the combination of triplets ijl
14:37 < manish7294> rcurtin: It will be very helpful, if you could look into this. Thanks!
14:40 < rcurtin> manish7294: in a meeting now, available later, and I will answer then :(
14:40 < manish7294> rcurtin: It's fine. Just leaving the message here :)
15:06 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has quit [Quit: Page closed]
15:15 < ShikharJ> zoq: Can I also squash Kris' commits into one (there are more than 45, and everytime I do a rebase, some or the other conflict crops up)?
15:23 -!- travis-ci [~travis-ci@ec2-54-157-202-111.compute-1.amazonaws.com] has joined #mlpack
15:23 < travis-ci> ShikharJ/mlpack#147 (LayerNorm - 7ad07a8 : Shikhar Jaiswal): The build has errored.
15:23 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/1ea8586a5215...7ad07a8dd82c
15:23 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/379245550
15:23 -!- travis-ci [~travis-ci@ec2-54-157-202-111.compute-1.amazonaws.com] has left #mlpack []
15:40 [Users #mlpack]
15:40 [ anirudhm] [ ImQ009_ ] [ killer_bee[m]] [ prakhar_code[m]] [ sumedhghaisas2] [ xa0]
15:40 [ govg ] [ jenkins-mlpack] [ lozhnikov ] [ rcurtin ] [ vivekp ] [ zoq]
15:40 [ gtank ] [ K4k ] [ petris ] [ ShikharJ ] [ wiking ]
15:40 -!- Irssi: #mlpack: Total of 17 nicks [0 ops, 0 halfops, 0 voices, 17 normal]
15:52 < rcurtin> zoq: sure, using the benchmark system for the PR jobs is just fine with me
15:52 < rcurtin> even multiple benchmark systems would be fine
15:57 -!- travis-ci [~travis-ci@ec2-54-157-202-111.compute-1.amazonaws.com] has joined #mlpack
15:57 < travis-ci> ShikharJ/mlpack#148 (master - d77f040 : Marcus Edel): The build has errored.
15:57 < travis-ci> Change view : https://github.com/ShikharJ/mlpack/compare/acce0c0000b8...d77f04097759
15:57 < travis-ci> Build details : https://travis-ci.org/ShikharJ/mlpack/builds/379262064
15:57 -!- travis-ci [~travis-ci@ec2-54-157-202-111.compute-1.amazonaws.com] has left #mlpack []
16:04 < rcurtin> manish7294: have you taken a look at the original LMNN implementation and how they handle this?
16:04 < rcurtin> I could see that you could initialize the slack variables to whatever is necessary to make the current matrix M a feasible solution
16:05 < rcurtin> then you could run the SDP holding the slack variables constant, then you could update the slack variables, etc.
16:05 < rcurtin> but you should probably take a look at the existing LMNN or BoostMetric implementation to get an idea
16:07 < zoq> ShikharJ: Sure, fine with me.
16:45 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has joined #mlpack
16:48 < manish7294> rcurtin: The original implementation calculates a N * N * k slack matrix at each gradient step, which they uses to calculate loss part. That's the largest size matrix they use.
16:49 < manish7294> rcutin: you can refer to https://it.mathworks.com/matlabcentral/mlc-downloads/downloads/submissions/56937/versions/28/previews/FSLib_v6.0_2018/lib/drtoolbox/techniques/lmnn.m/index.html?access_key=
16:55 < manish7294> rcurtin: And if we do something like that we will end up having a sparse matrix as a combination of diag matrix of N * N * k + a dense matrix of order d.
16:59 < rcurtin> manish7294: thanks for the link, I will study it soon when I have a chance
17:00 < manish7294> rcurtin: Sure! Till then I will updating primal dual part too :)
17:06 -!- haritha1313 [0e8bf0fb@gateway/web/freenode/ip.14.139.240.251] has joined #mlpack
17:08 < haritha1313> zoq: Thought it would be better to keep you updated on progress. I am currently refactoring CF class wrt PR #1355. Working on policy classes for various factorizers as of now.
17:09 < haritha1313> I guess it should be done by tomorrow, and once its settled I can start working on NCF and introduce it as a new DecompositionPolicy.
17:10 < haritha1313> As you know, a PR has been made for multiplymerge layer, and as discussed, embedding layer can be implemented by an alias in lookup layer. Should I make that addition in the existing PR itself?
17:19 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Read error: Connection reset by peer]
17:21 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
17:31 -!- manish72942 [~yaaic@2405:205:1005:c8d3:6156:9454:32d0:8d24] has joined #mlpack
17:31 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has quit [Quit: Page closed]
17:32 -!- manish72942 [~yaaic@2405:205:1005:c8d3:6156:9454:32d0:8d24] has quit [Client Quit]
17:32 -!- manish7294 [~yaaic@2405:205:1005:c8d3:6156:9454:32d0:8d24] has joined #mlpack
17:32 -!- manish7294 [~yaaic@2405:205:1005:c8d3:6156:9454:32d0:8d24] has quit [Client Quit]
17:33 -!- manish7294 [~yaaic@2405:205:1005:c8d3:6156:9454:32d0:8d24] has joined #mlpack
17:34 -!- haritha1313 [0e8bf0fb@gateway/web/freenode/ip.14.139.240.251] has quit [Ping timeout: 260 seconds]
17:41 -!- manish7294 [~yaaic@2405:205:1005:c8d3:6156:9454:32d0:8d24] has quit [Quit: Yaaic - Yet another Android IRC client - http://www.yaaic.org]
17:41 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has joined #mlpack
17:45 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has quit [Client Quit]
17:46 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has joined #mlpack
17:47 < manish7294> "/home/travis/build/mlpack/mlpack/src/mlpack/tests/lrsdp_test.cpp(270): fatal error in "GaussianMatrixSensingSDP": difference{0.0514407%} between measurement{0.00023853556661135311} and b(i){0.00023841292538047398} exceeds 0.050000000000000003%"
17:47 < manish7294> Is this some kind of random error as two of travis job failed and one passed. Moreover I can't seem to reproduce it on my system.
17:52 -!- manish7294 [9d25071c@gateway/web/freenode/ip.157.37.7.28] has quit [Quit: Page closed]
18:08 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
18:08 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Read error: Connection reset by peer]
18:11 -!- sumedhghaisas [~yaaic@85.255.234.61] has joined #mlpack
18:13 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
18:15 -!- sumedhghaisas [~yaaic@85.255.234.61] has quit [Read error: Connection reset by peer]
18:15 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
18:46 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
18:46 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Read error: Connection reset by peer]
19:08 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
19:08 -!- sumedhghaisas [~yaaic@148.252.129.31] has joined #mlpack
19:19 -!- sumedhghaisas [~yaaic@148.252.129.31] has quit [Read error: Connection reset by peer]
19:19 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
19:22 -!- sumedhghaisas2 [~yaaic@148.252.129.31] has joined #mlpack
19:23 -!- sumedhghaisas2 [~yaaic@148.252.129.31] has quit [Read error: Connection reset by peer]
19:23 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has joined #mlpack
19:24 -!- sumedhghaisas [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
20:06 -!- ImQ009_ [~ImQ009@unaffiliated/imq009] has quit [Quit: Leaving]
20:22 < zoq> haritha1313: Sounds great, about the adjustments, let's open another PR.
20:24 < zoq> manish7294: Did you use mlpack::math::RandomSeed(time(NULL)); inside the test case?
20:24 < zoq> while(true); do bin/mlpack_test -t TestSuite/TestCase; sleep 1;
20:24 < zoq> might be helpful as well
20:33 -!- sumedhghaisas [~yaaic@148.252.129.13] has joined #mlpack
20:34 -!- sumedhghaisas2 [~yaaic@2a00:79e0:d:fd00:4a3:d217:b63c:d812] has quit [Ping timeout: 256 seconds]
20:55 -!- sumedhghaisas [~yaaic@148.252.129.13] has quit [Read error: Connection reset by peer]
20:55 -!- sumedhghaisas2 [~yaaic@host-92-8-33-72.as43234.net] has joined #mlpack
21:31 -!- vivekp [~vivek@unaffiliated/vivekp] has quit [Ping timeout: 240 seconds]
21:51 -!- vivekp [~vivek@unaffiliated/vivekp] has joined #mlpack
22:30 < zoq> okay, benchmark systems configured, they are not as fast as masterblaster but they are still fast enough; and since we have more than one we can build in parallel.
--- Log closed Wed May 16 00:00:21 2018