20191129
beta3运行自带的dl4j-example/org.deeplearning4j.example/catcharecognition/MultiDigitNumberRecognition.java是正常的,但是beta5运行同样的文件就有问题。
20191122
上述文件在beta6被放到了dl4j-example/org.deeplearning4j.examples/convolution/catcharecognition/中。
20200210
beta6运行上述文件,没有配置虚拟机参数,4个epoch计算完以后,除了报以下错误以外,其余运行正常。程序停止前的混淆矩阵分布不记得了,没空重新测试。
Exception in thread "main" org.nd4j.linalg.exception.ND4JIllegalStateException: Axis array [1] contains values above array rank (rank=1) at org.nd4j.linalg.api.shape.Shape.normalizeAxis(Shape.java:3394) at org.nd4j.linalg.api.ops.BaseOp.defineDimensions(BaseOp.java:344) at org.nd4j.linalg.api.ops.BaseIndexAccumulation.<init>(BaseIndexAccumulation.java:95) at org.nd4j.linalg.api.ops.impl.indexaccum.IMax.<init>(IMax.java:47) at org.nd4j.linalg.factory.Nd4j.argMax(Nd4j.java:578) at org.deeplearning4j.examples.convolution.captcharecognition.MultiDigitNumberRecognition.modelPredict(MultiDigitNumberRecognition.java:177) at org.deeplearning4j.examples.convolution.captcharecognition.MultiDigitNumberRecognition.main(MultiDigitNumberRecognition.java:103)
20200527
beta7运行MultiDigitNumberRecognition.java后和20200210beta6的情况一样。
程序停止前的混淆矩阵似乎有问题。
0 1 2 3 4 5 6 7 8 9
-------------------------------
0 0 0 0 0 0 0 0 0 22 | 0 = 0
0 0 0 0 0 0 0 0 0 0 | 1 = 1
0 0 0 0 0 0 0 0 0 13 | 2 = 2
0 0 0 0 0 0 0 0 0 7 | 3 = 3
0 0 0 0 0 0 0 0 0 16 | 4 = 4
0 0 0 0 0 0 0 0 0 6 | 5 = 5
0 0 0 0 0 0 0 0 0 14 | 6 = 6
0 0 0 0 0 0 0 0 0 0 | 7 = 7
0 0 0 0 0 0 0 0 0 15 | 8 = 8
0 0 0 0 0 0 0 0 0 15 | 9 = 9
暂时就讨论到这里。