Issue executing GenerateTxtCharCompGraphModel.java #1063
Description
Issue Description
The GenerateTxtCharCompGraphModel example seems to be broken. After getting around the missing input data ("java.io.IOException: Server returned HTTP response code: 403 for URL: https://s3.amazonaws.com/dl4j-distribution/pg100.txt", changed url to https://www.gutenberg.org/cache/epub/100/pg100.txt to fix), the ComputationGraphConfiguration appears to be misconfigured. Running the example results in the following error:
java.lang.reflect.InvocationTargetException
at sun.reflect.NativeMethodAccessorImpl.invoke0 (Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke (NativeMethodAccessorImpl.java:62)
at sun.reflect.DelegatingMethodAccessorImpl.invoke (DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke (Method.java:498)
at org.codehaus.mojo.exec.ExecJavaMojo$1.run (ExecJavaMojo.java:293)
at java.lang.Thread.run (Thread.java:750)
Caused by: java.lang.IllegalStateException: Sequence lengths do not match for RnnOutputLayer input and labels:Arrays should be rank 3 with shape [minibatch, size, sequenceLength] - mismatch on dimension 2 (sequence length) - input=[32, 400, 50] vs. label=[32, 77, 50]
at org.nd4j.common.base.Preconditions.throwStateEx (Preconditions.java:639)
at org.nd4j.common.base.Preconditions.checkState (Preconditions.java:337)
at org.deeplearning4j.nn.layers.recurrent.RnnOutputLayer.backpropGradient (RnnOutputLayer.java:59)
at org.deeplearning4j.nn.graph.vertex.impl.LayerVertex.doBackward (LayerVertex.java:148)
at org.deeplearning4j.nn.graph.ComputationGraph.calcBackpropGradients (ComputationGraph.java:2776)
at org.deeplearning4j.nn.graph.ComputationGraph.computeGradientAndScore (ComputationGraph.java:1385)
at org.deeplearning4j.nn.graph.ComputationGraph.computeGradientAndScore (ComputationGraph.java:1345)
at org.deeplearning4j.optimize.solvers.BaseOptimizer.gradientAndScore (BaseOptimizer.java:174)
at org.deeplearning4j.optimize.solvers.StochasticGradientDescent.optimize (StochasticGradientDescent.java:61)
at org.deeplearning4j.optimize.Solver.optimize (Solver.java:52)
at org.deeplearning4j.nn.graph.ComputationGraph.doTruncatedBPTT (ComputationGraph.java:3739)
at org.deeplearning4j.nn.graph.ComputationGraph.fitHelper (ComputationGraph.java:1160)
at org.deeplearning4j.nn.graph.ComputationGraph.fit (ComputationGraph.java:1119)
at org.deeplearning4j.nn.graph.ComputationGraph.fit (ComputationGraph.java:1106)
at org.deeplearning4j.nn.graph.ComputationGraph.fit (ComputationGraph.java:988)
at org.deeplearning4j.examples.advanced.modelling.charmodelling.generatetext.GenerateTxtCharCompGraphModel.main (GenerateTxtCharCompGraphModel.java:116)
at sun.reflect.NativeMethodAccessorImpl.invoke0 (Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke (NativeMethodAccessorImpl.java:62)
at sun.reflect.DelegatingMethodAccessorImpl.invoke (DelegatingMethodAccessorImpl.java:43)
at java.lang.reflect.Method.invoke (Method.java:498)
at org.codehaus.mojo.exec.ExecJavaMojo$1.run (ExecJavaMojo.java:293)
at java.lang.Thread.run (Thread.java:750)
Input size to "outputLayer" is correctly set to 2*lstmLayerSize, and the output sizes of both layers flowing to "outputLayer" are lstmLayerSize each. This appears to be the correct configuration. But the network seems to be expecting the input size (i.e. "iter.inputColumns()") at that point.
How can this be fixed?
Version Information
Please indicate relevant versions, including, if relevant:
- Deeplearning4j version - 1.0.0-M2
- platform information - MacOS x86