w00t! the change that you've made seems correct Tim. I see they've done the same change in dl4j tests https://github.com/deeplearning4j/deeplearning4j/commit/015fd10bddbc59416595f706af6be1d0d23f573f
On Wed, May 8, 2019 at 9:48 PM Chris Mattmann <mattm...@apache.org> wrote: > Yayy Tim > > > > > > > > From: Tim Allison <talli...@apache.org> > Reply-To: "dev@tika.apache.org" <dev@tika.apache.org> > Date: Wednesday, May 8, 2019 at 8:53 AM > To: "dev@tika.apache.org" <dev@tika.apache.org> > Subject: Re: [EXTERNAL] Re: DL4JVGG16NetTest failures > > > > I think so. Works locally now. Actually works locally. :D > > > > On Wed, May 8, 2019 at 11:50 AM Chris Mattmann <mattm...@apache.org> > wrote: > > > > Great work ☺ So it’s fixed? > > > > > > > > > > > > > > > > From: Tim Allison <talli...@apache.org> > > Reply-To: "dev@tika.apache.org" <dev@tika.apache.org> > > Date: Wednesday, May 8, 2019 at 8:43 AM > > To: "dev@tika.apache.org" <dev@tika.apache.org> > > Cc: Thamme Gowda N <t...@isi.edu>, Thejan Wijesinghe < > > thejan.k.wijesin...@gmail.com> > > Subject: Re: [EXTERNAL] Re: DL4JVGG16NetTest failures > > > > > > > > As is my habit, I should have kicked the tires a bit more before > > > > asking for help. > > > > > > > > I _think_ I fixed it. I was able to reproduce it locally (darned > > > > proxy!!!). If anyone w knowledge of dl4j would be willing to confirm > > > > that my commit/fix is sane, I'd appreciate it. Thank you!!! > > > > > > > > Cheers, > > > > > > > > Tim > > > > > > > > On Wed, May 8, 2019 at 11:32 AM Chris Mattmann <mattm...@apache.org> > > wrote: > > > > > > > > Thejan, Thamme any ideas? > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > From: Tim Allison <talli...@apache.org> > > > > Reply-To: "dev@tika.apache.org" <dev@tika.apache.org> > > > > Date: Wednesday, May 8, 2019 at 7:50 AM > > > > To: "dev@tika.apache.org" <dev@tika.apache.org> > > > > Subject: [EXTERNAL] Re: DL4JVGG16NetTest failures > > > > > > > > > > > > > > > > Any recommendations? > > > > > > > > > > > > > > > > java.lang.IllegalStateException: Number of indices (got 2) must be > > > > > > > > same as array rank (1) - indices [0, 0] > > > > > > > > at > > org.nd4j.base.Preconditions.throwStateEx(Preconditions.java:641) > > > > > > > > at org.nd4j.base.Preconditions.checkState(Preconditions.java:376) > > > > > > > > at org.nd4j.linalg.api.shape.Shape.getOffset(Shape.java:884) > > > > > > > > at org.nd4j.linalg.api.shape.Shape.getDouble(Shape.java:680) > > > > > > > > at > > org.nd4j.linalg.api.ndarray.BaseNDArray.getDouble(BaseNDArray.java:1906) > > > > > > > > at > > org.nd4j.linalg.api.ndarray.BaseNDArray.getInt(BaseNDArray.java:1851) > > > > > > > > at > > org.apache.tika.dl.imagerec.DL4JVGG16Net.predict(DL4JVGG16Net.java:147) > > > > > > > > at > > org.apache.tika.dl.imagerec.DL4JVGG16Net.recognise(DL4JVGG16Net.java:128) > > > > > > > > at > > > org.apache.tika.parser.recognition.ObjectRecognitionParser.parse(ObjectRecognitionParser.java:118) > > > > > > > > > > > > > > > > On Wed, May 8, 2019 at 10:11 AM Tim Allison <talli...@apache.org> wrote: > > > > > > > > > > > > > > > > Yay! I am able to reproduce this on our vm... Onward to debugging... > > > > > > > > > > > > > > > > On Wed, May 8, 2019 at 9:57 AM Tim Allison <talli...@apache.org> wrote: > > > > > > > > > > > > > > > > > > All, > > > > > > > > > Apologies for the broken builds...I'm not able to reproduce this > > > > > > > > > test failure on my mac or Windows machine. I'm testing the build on > > > > > > > > > our regression vm now. > > > > > > > > > If anyone has any idea why this is failing on our build vms (since > > > > > > > > > we upgraded to dl4j-beta3), please let me know. > > > > > > > > > Thank you. > > > > > > > > > > > > > > > > > > Best, > > > > > > > > > > > > > > > > > > Tim > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > https://builds.apache.org/job/tika-branch-1x/189/org.apache.tika$tika-dl/testReport/junit/org.apache.tika.dl.imagerec/DL4JVGG16NetTest/recognise/ > > > > > > > > > > > > > > > > > > > > > > > >
