deeplearning4j: Keras model import for Inception does not work on Mac using 0.8.1-SNAPSHOT

Keras model import for Inception does not work on Mac

Using Mac 10.11.3 OS X, and DL4J 0.8.1-SNAPSHOT (built from scratch along with libnd4j from scratch also version 0.8.1-SNAPSHOT), I tried Keras model import of Inception as described in https://github.com/deeplearning4j/deeplearning4j/issues/3190 and in https://github.com/apache/tika/pull/165. The issue I’m having is that running Inception against an image produces no matches. For example, using Tika and our tika-dl proposed module:

LMC-053601:tika-gh mattmann$ java -Dorg.bytedeco.javacpp.nopointergc=true -cp ./tika-dl/target/tika-dl-1.15-SNAPSHOT-jar-with-dependencies.jar:tika-app/target/tika-app-1.15-SNAPSHOT.jar   org.apache.tika.cli.TikaCLI  --config=tika-dl/src/test/resources/org/apache/tika/dl/imagerec/dl4j-inception3-config.xml  https://images-na.ssl-images-amazon.com/images/G/01/img15/pet-products/small-tiles/23695_pets_vertical_store_dogs_small_tile_8._CB312176604_.jpg
WARN  JBIG2ImageReader not loaded. jbig2 files will be ignored
INFO  Cache exist at /Users/mattmann/git/tika-gh/.tmp-inception/inception-model-weights.h5. Not downloading it
WARN  File org/apache/tika/dl/imagerec/inceptionv3-model.json not found in local file system. Asking the classloader
INFO  Cache exist at /Users/mattmann/git/tika-gh/.tmp-inception/inceptionv3-model.json. Not downloading it
WARN  File org/apache/tika/dl/imagerec/imagenet_incpetionv3_class_index.json not found in local file system. Asking the classloader
INFO  Going to load Inception network...
INFO  Loaded [CpuBackend] backend
INFO  Number of threads used for NativeOps: 4
INFO  Reflections took 100 ms to scan 1 urls, producing 29 keys and 192 values 
INFO  Number of threads used for BLAS: 4
INFO  Backend used: [CPU]; OS: [Mac OS X]
INFO  Cores: [8]; Memory: [3.6GB];
INFO  Blas vendor: [OPENBLAS]
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libJ3DAudio.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:428)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:383)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:373)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:120)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1047)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:578)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:562)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libAppleScriptEngine.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:428)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:383)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:373)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:120)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1047)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:578)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:562)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libJ3D.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:428)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:383)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:373)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:120)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1047)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:578)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:562)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/usr/lib/java/libjdns_sd.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:428)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:383)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:373)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:120)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1047)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:578)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:562)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libmlib_jai.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:428)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:383)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:373)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:120)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1047)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:578)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:562)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libJ3DUtils.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:428)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:383)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:373)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:120)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1047)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:578)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:562)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
INFO  Reflections took 1825 ms to scan 12 urls, producing 3714 keys and 16629 values 
WARN  Layer "batchnormalization_1" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_2" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_3" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_4" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_5" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_9" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_7" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_10" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_6" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_8" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_11" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_12" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_16" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_14" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_17" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_13" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_15" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_18" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_19" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_23" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_21" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_24" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_20" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_22" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_25" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_26" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_28" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_29" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_27" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_30" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_35" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_36" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_32" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_37" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_33" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_38" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_31" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_34" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_39" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_40" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_45" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_46" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_42" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_47" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_43" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_48" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_41" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_44" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_49" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_50" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_55" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_56" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_52" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_57" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_53" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_58" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_51" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_54" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_59" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_60" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_65" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_66" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_62" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_67" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_63" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_68" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_61" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_64" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_69" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_70" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_73" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_74" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_71" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_75" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_72" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_76" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_81" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_78" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_82" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_79" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_80" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_83" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_84" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_77" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_85" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_90" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_87" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_91" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_88" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_89" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_92" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_93" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_86" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_94" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Model cannot be trained: output layer predictions is not an IOutputLayer (no loss function specified)
WARN  Model cannot be trained: output predictions is not an IOutputLayer (no loss function specified)
INFO  Reflections took 143 ms to scan 1 urls, producing 375 keys and 1457 values 
INFO  Loaded the Inception model. Time taken=3769ms
INFO  minConfidence = 0.015, topN=10
INFO  Recogniser = org.apache.tika.dl.imagerec.DL4JInceptionV3Net
INFO  Recogniser Available = true
WARN  NO objects
LMC-053601:tika-gh mattmann$

Oddly, if I use ND4J 0.8 release, it sporadically fails as mentioned in #3190 however, it does actually produce objects:

LMC-053601:tika-gh mattmann$ java -Xmx400m -cp ./tika-dl/target/tika-dl-1.15-SNAPSHOT-jar-with-dependencies.jar:tika-app/target/tika-app-1.15-SNAPSHOT.jar   org.apache.tika.cli.TikaCLI  --config=tika-dl/src/test/resources/org/apache/tika/dl/imagerec/dl4j-inception3-config.xml  https://images-na.ssl-images-amazon.com/images/G/01/img15/pet-products/small-tiles/23695_pets_vertical_store_dogs_small_tile_8._CB312176604_.jpg
WARN  JBIG2ImageReader not loaded. jbig2 files will be ignored
INFO  Cache doesn't exist. Going to make a copy
INFO  This might take a while! GET https://raw.githubusercontent.com/USCDataScience/dl4j-kerasimport-examples/98ec48b56a5b8fb7d54a2994ce9cb23bfefac821/dl4j-import-example/data/inception-model-weights.h5
WARN  File org/apache/tika/dl/imagerec/inceptionv3-model.json not found in local file system. Asking the classloader
INFO  Cache doesn't exist. Going to make a copy
INFO  This might take a while! GET jar:file:/Users/mattmann/git/tika-gh/tika-dl/target/tika-dl-1.15-SNAPSHOT-jar-with-dependencies.jar!/org/apache/tika/dl/imagerec/inceptionv3-model.json
WARN  File org/apache/tika/dl/imagerec/imagenet_incpetionv3_class_index.json not found in local file system. Asking the classloader
INFO  Going to load Inception network...
INFO  Loaded [CpuBackend] backend
INFO  Number of threads used for NativeOps: 4
INFO  Reflections took 94 ms to scan 1 urls, producing 29 keys and 189 values 
INFO  Number of threads used for BLAS: 4
INFO  Backend used: [CPU]; OS: [Mac OS X]
INFO  Cores: [8]; Memory: [0.3GB];
INFO  Blas vendor: [OPENBLAS]
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libJ3DAudio.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:431)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:386)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:376)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:123)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1019)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:559)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:543)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libAppleScriptEngine.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:431)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:386)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:376)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:123)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1019)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:559)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:543)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libJ3D.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:431)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:386)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:376)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:123)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1019)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:559)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:543)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/usr/lib/java/libjdns_sd.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:431)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:386)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:376)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:123)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1019)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:559)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:543)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libmlib_jai.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:431)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:386)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:376)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:123)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1019)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:559)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:543)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
WARN  could not create Vfs.Dir from url. ignoring the exception and continuing
org.reflections.ReflectionsException: could not create Vfs.Dir from url, no matching UrlType was found [file:/System/Library/Java/Extensions/libJ3DUtils.jnilib]
either use fromURL(final URL url, final List<UrlType> urlTypes) or use the static setDefaultURLTypes(final List<UrlType> urlTypes) or addDefaultURLTypes(UrlType urlType) with your specialized UrlType.
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:109)
	at org.reflections.vfs.Vfs.fromURL(Vfs.java:91)
	at org.reflections.Reflections.scan(Reflections.java:237)
	at org.reflections.Reflections.scan(Reflections.java:204)
	at org.reflections.Reflections.<init>(Reflections.java:129)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.registerSubtypes(NeuralNetConfiguration.java:431)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.configureMapper(NeuralNetConfiguration.java:386)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.initMapper(NeuralNetConfiguration.java:376)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration.<clinit>(NeuralNetConfiguration.java:123)
	at org.deeplearning4j.nn.conf.NeuralNetConfiguration$Builder.build(NeuralNetConfiguration.java:1019)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:559)
	at org.deeplearning4j.nn.conf.ComputationGraphConfiguration$GraphBuilder.addLayer(ComputationGraphConfiguration.java:543)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraphConfiguration(KerasModel.java:428)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:482)
	at org.deeplearning4j.nn.modelimport.keras.KerasModel.getComputationGraph(KerasModel.java:471)
	at org.deeplearning4j.nn.modelimport.keras.KerasModelImport.importKerasModelAndWeights(KerasModelImport.java:178)
	at org.apache.tika.dl.imagerec.DL4JInceptionV3Net.initialize(DL4JInceptionV3Net.java:265)
	at org.apache.tika.parser.recognition.ObjectRecognitionParser.initialize(ObjectRecognitionParser.java:100)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOne(TikaConfig.java:603)
	at org.apache.tika.config.TikaConfig$XmlLoader.loadOverall(TikaConfig.java:525)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:185)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:162)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:137)
	at org.apache.tika.config.TikaConfig.<init>(TikaConfig.java:133)
	at org.apache.tika.cli.TikaCLI.configure(TikaCLI.java:675)
	at org.apache.tika.cli.TikaCLI.process(TikaCLI.java:407)
	at org.apache.tika.cli.TikaCLI.main(TikaCLI.java:143)
INFO  Reflections took 1501 ms to scan 12 urls, producing 3710 keys and 16614 values 
WARN  Layer "batchnormalization_1" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_2" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_3" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_4" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_5" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_9" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_7" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_10" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_6" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_8" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_11" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_12" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_16" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_14" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_17" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_13" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_15" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_18" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_19" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_23" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_21" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_24" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_20" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_22" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_25" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_26" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_28" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_29" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_27" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_30" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_35" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_36" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_32" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_37" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_33" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_38" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_31" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_34" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_39" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_40" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_45" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_46" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_42" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_47" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_43" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_48" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_41" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_44" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_49" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_50" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_55" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_56" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_52" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_57" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_53" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_58" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_51" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_54" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_59" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_60" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_65" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_66" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_62" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_67" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_63" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_68" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_61" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_64" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_69" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_70" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_73" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_74" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_71" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_75" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_72" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_76" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_81" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_78" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_82" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_79" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_80" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_83" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_84" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_77" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_85" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_90" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_87" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_91" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_88" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_89" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_92" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_93" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_86" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Layer "batchnormalization_94" momentum has been set but will not be applied unless the updater is set to NESTEROVS.
WARN  Model cannot be trained: output layer predictions is not an IOutputLayer (no loss function specified)
WARN  Model cannot be trained: output predictions is not an IOutputLayer (no loss function specified)
INFO  Reflections took 107 ms to scan 1 urls, producing 371 keys and 1443 values 
INFO  Loaded the Inception model. Time taken=4944ms
INFO  minConfidence = 0.015, topN=10
INFO  Recogniser = org.apache.tika.dl.imagerec.DL4JInceptionV3Net
INFO  Recogniser Available = true
<?xml version="1.0" encoding="UTF-8"?><html xmlns="http://www.w3.org/1999/xhtml">
<head>
<meta name="org.apache.tika.parser.recognition.object.rec.impl" content="org.apache.tika.dl.imagerec.DL4JInceptionV3Net"/>
<meta name="X-Parsed-By" content="org.apache.tika.parser.CompositeParser"/>
<meta name="X-Parsed-By" content="org.apache.tika.parser.recognition.ObjectRecognitionParser"/>
<meta name="resourceName" content="23695_pets_vertical_store_dogs_small_tile_8._CB312176604_.jpg"/>
<meta name="Content-Length" content="56996"/>
<meta name="OBJECT" content="beagle (0.60227)"/>
<meta name="OBJECT" content="tennis_ball (0.11380)"/>
<meta name="OBJECT" content="Walker_hound (0.05603)"/>
<meta name="OBJECT" content="Labrador_retriever (0.02694)"/>
<meta name="OBJECT" content="bluetick (0.02557)"/>
<meta name="OBJECT" content="English_foxhound (0.02299)"/>
<meta name="Content-Type" content="image/jpeg"/>
<title/>
</head>
<body><ol id="objects">	<li id="162"> beagle [en](confidence = 0.602268 )</li>
	<li id="852"> tennis_ball [en](confidence = 0.113798 )</li>
	<li id="166"> Walker_hound [en](confidence = 0.056029 )</li>
	<li id="208"> Labrador_retriever [en](confidence = 0.026943 )</li>
	<li id="164"> bluetick [en](confidence = 0.025571 )</li>
	<li id="167"> English_foxhound [en](confidence = 0.022993 )</li>
</ol>
</body></html>LMC-053601:tika-gh mattmann$

I’d be happy to help fix it but don’t really know the code base well.

About this issue

Original URL
State: closed
Created 7 years ago
Comments: 22 (9 by maintainers)

Most upvoted comments

perhaps the issue is that in the tika-dl pom.xml, we have:

  <dependencies>
    <dependency>
      <groupId>org.apache.tika</groupId>
      <artifactId>tika-parsers</artifactId>
      <version>${project.version}</version>
      <scope>provided</scope>
    </dependency>
    <dependency>
      <groupId>junit</groupId>
      <artifactId>junit</artifactId>
      <scope>test</scope>
    </dependency>
    <dependency>
      <groupId>org.deeplearning4j</groupId>
      <artifactId>deeplearning4j-modelimport</artifactId>
      <version>${dl4j.version}</version>
    </dependency>
    <dependency>
      <groupId>org.datavec</groupId>
      <artifactId>datavec-data-image</artifactId>
      <version>${dl4j.version}</version>
    </dependency>
    <dependency>
      <groupId>org.nd4j</groupId>
      <artifactId>nd4j-native-platform</artifactId>
      <version>${dl4j.version}</version>
    </dependency>
  </dependencies>

assuming they should all be the same version. Is that not the case in this patch release? If not, what should I set them to?

chrismattmann on May 7, 2017

I’ve released deeplearning4j-modelimport and deeplearning4j-keras version 0.8.0-1 based on the current master and https://github.com/deeplearning4j/deeplearning4j/commit/344cb1e44714bccfcfdc5b2c91032ad45bc8eb23, but it basically only includes the fix I made over 0.8.0: http://repo1.maven.org/maven2/org/deeplearning4j/deeplearning4j-modelimport/0.8.0-1/ http://repo1.maven.org/maven2/org/deeplearning4j/deeplearning4j-keras/0.8.0-1/ @chrismattmann Let us know if that works! Thanks

saudet on May 7, 2017

Hmm let me get back to you there. I wouldn’t want to hold up your release…were you guys using anything that was relying on 0.8.1? The only thing I could think of might be the hdf5 fix? What do you think @saudet ?

Unfortunately this release is looking pretty big but we might be able to cut it after workspaces and our new ETL is in place.

@raver119 how long do you think we have on this before it’s beaten in to shape? It seems like we’re close.

agibsonccc on May 7, 2017