[SYSTEMDS-1077] Max Pooling Null Pointer

• Type: Bug
• Status: Resolved
• Priority: Blocker
• Resolution: Fixed
• Affects Version/s: SystemML 0.11
• Fix Version/s: SystemML 0.11
• Epic Link: Deep Learning
• Assignee: Niketan Pansare
• Reporter: Mike Dusenberry
• Created: 31/Oct/16 20:22
• Updated: 01/Nov/16 21:30
• Resolved: 01/Nov/16 21:30

Description

Currently, if a matrix of all zeros is passed into `max_pool`, a null pointer exception will be raised. This is a possibility within the course of training a neural net that uses ReLU & Dropout layers.

script = """
N = 2
C = 3
Hin = 10
Win = 10
f = 2
stride = 2
X = matrix(0, rows=N, cols=C*Hin*Win)
out = max_pool(X, input_shape=[N,C,Hin,Win], pool_size=[f,f], stride=[stride,stride], padding=[0,0])
"""
script = dml(script).output("out")
out = ml.execute(script).get("out")
out.toDF().collect()

Caused by: org.apache.sysml.runtime.DMLRuntimeException: Error while executing multi-threaded MaxPooling_Forward
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN.runParallelConvTask(LibMatrixDNN.java:889)
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN.runConvTask(LibMatrixDNN.java:849)
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN.maxpooling(LibMatrixDNN.java:733)
	at org.apache.sysml.runtime.instructions.cp.ConvolutionCPInstruction.processInstruction(ConvolutionCPInstruction.java:192)
	at org.apache.sysml.runtime.controlprogram.ProgramBlock.executeSingleInstruction(ProgramBlock.java:305)
	... 17 more
Caused by: java.util.concurrent.ExecutionException: java.lang.NullPointerException
	at java.util.concurrent.FutureTask.report(FutureTask.java:122)
	at java.util.concurrent.FutureTask.get(FutureTask.java:192)
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN.runParallelConvTask(LibMatrixDNN.java:883)
	... 21 more
Caused by: java.lang.NullPointerException
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN.doPooling(LibMatrixDNN.java:756)
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN.access$000(LibMatrixDNN.java:41)
	at org.apache.sysml.runtime.matrix.data.LibMatrixDNN$ConvTask.call(LibMatrixDNN.java:912)
	at java.util.concurrent.FutureTask.run(FutureTask.java:266)
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1142)
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:617)
	... 1 more

Issue Links

• relates to: SYSTEMDS-1076 Sparse Max Pooling Unsupported