1. Test Modules
  2. Network Diagram
  3. Differential Validation
    1. Feedback Validation
    2. Learning Validation
    3. Total Accuracy
    4. Frozen and Alive Status
  4. Results

Subreport: Logs for com.simiacryptus.ref.lang.ReferenceCountingBase

Test Modules

Network Diagram

This is a network apply the following layout:

LayerTests.java:203 executed in 0.07 seconds (0.000 gc):

    return Graphviz.fromGraph((Graph) TestUtil.toGraph(((DAGNetwork) layer).addRef())).height(400).width(600)
        .render(Format.PNG).toImage();
Logging 
executing command [/bin/sh, -c, dot -Tsvg /tmp/GraphvizJava/DotEngine1476882440843522288/dotfile.dot -ooutfile.svg]

Returns

Result

Using Seed 7159579659087751168

Differential Validation

SingleDerivativeTester.java:101 executed in 0.00 seconds (0.000 gc):

        log.info(RefString.format("Inputs: %s", prettyPrint(inputPrototype)));
        log.info(RefString.format("Inputs Statistics: %s", printStats(inputPrototype)));
        log.info(RefString.format("Output: %s", outputPrototype.prettyPrint()));
        assert outputPrototype != null;
        log.info(RefString.format("Outputs Statistics: %s", outputPrototype.getScalarStatistics()));
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype)));
Logging 
Inputs: [
[ [ 0.08, -1.72, -0.712 ], [ 0.496, -0.852, 1.356 ], [ 0.048, -0.804, 1.032 ] ],
[ [ 0.7, -1.028, 1.048 ], [ -0.608, 1.912, 1.512 ], [ 1.524, 1.108, 0.3 ] ],
[ [ -0.128, -0.384, -1.616 ], [ 1.764, -1.688, 1.556 ], [ 1.208, 0.028, 0.636 ] ]
]
Inputs Statistics: {meanExponent=-0.18240580381395474, negative=10, min=-1.72, max=1.912, mean=0.2506666666666667, count=27, sum=6.768000000000001, positive=17, stdDev=1.0872753578113061, zeros=0}
Output: [
[ [ 1.176576, -0.49487999999999993 ], [ 0.5904639999999999, 1.755984 ], [ 0.380544, 0.742752 ] ],
[ [ 0.8380799999999999, 1.8861599999999998 ], [ -1.6576, 0.13884800000000003 ], [ 0.04383999999999998, 2.986976 ] ],
[ [ 0.376832, -1.387712 ], [ 1.70208, 4.065856 ], [ 0.5007359999999998, 2.5783519999999998 ] ]
]
Outputs Statistics: {meanExponent=-0.07891120377813803, negative=3, min=-1.6576, max=4.065856, mean=0.9013271111111111, count=18, sum=16.223888, positive=15, stdDev=1.4039688704668984, zeros=0}

Feedback Validation

We validate the agreement between the implemented derivative of the inputs apply finite difference estimations:

SingleDerivativeTester.java:117 executed in 0.53 seconds (0.000 gc):

        return testFeedback(
            statistics,
            component.addRef(),
            RefUtil.addRef(inputPrototype),
            outputPrototype.addRef());
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype),
      component.addRef()));
Logging 
Feedback for input 0
Inputs Values: [
[ [ 0.08, -1.72, -0.712 ], [ 0.496, -0.852, 1.356 ], [ 0.048, -0.804, 1.032 ] ],
[ [ 0.7, -1.028, 1.048 ], [ -0.608, 1.912, 1.512 ], [ 1.524, 1.108, 0.3 ] ],
[ [ -0.128, -0.384, -1.616 ], [ 1.764, -1.688, 1.556 ], [ 1.208, 0.028, 0.636 ] ]
]
Value Statistics: {meanExponent=-0.18240580381395474, negative=10, min=-1.72, max=1.912, mean=0.2506666666666667, count=27, sum=6.768000000000001, positive=17, stdDev=1.0872753578113061, zeros=0}
Implemented Feedback: [ [ 0.496, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.496, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.496, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.496, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.496, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.496, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.496, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.496, ... ], ... ]
Implemented Statistics: {meanExponent=-0.4031196944645328, negative=18, min=-0.608, max=1.764, mean=0.042666666666666644, count=486, sum=20.73599999999999, positive=36, stdDev=0.27694952073805257, zeros=432}
Measured Feedback: [ [ 0.49600000000094013, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.4959999999998299, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.496000000000385, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.4959999999998299, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.49600000000094013, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.4959999999987197, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.4959999999998299, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.4959999999998299, ... ], ... ]
Measured Statistics: {meanExponent=-0.4031196944643072, negative=18, min=-0.6079999999999974, max=1.7640000000002098, mean=0.042666666666674416, count=486, sum=20.736000000003767, positive=36, stdDev=0.2769495207380164, zeros=432}
Feedback Error: [ [ 9.401368572525826E-13, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, -1.7008616737257398E-13, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 3.850253449400043E-13, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, -1.7008616737257398E-13, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 9.401368572525826E-13, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, -1.2803091919977305E-12, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.7008616737257398E-13, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.7008616737257398E-13, ... ], ... ]
Error Statistics: {meanExponent=-12.56554715879918, negative=21, min=-2.1404406025382627E-12, max=2.3004514959623634E-12, mean=7.747946089007655E-15, count=486, sum=3.76550179925772E-12, positive=33, stdDev=2.4278970089060296E-13, zeros=432}

Returns

    {
      "absoluteTol" : {
        "count" : 486,
        "sum" : 2.6075155923344084E-11,
        "min" : 0.0,
        "max" : 2.3004514959623634E-12,
        "sumOfSquare" : 2.867733859019325E-23,
        "standardDeviation" : 2.3691405558838235E-13,
        "average" : 5.3652584204411694E-14
      },
      "relativeTol" : {
        "count" : 54,
        "sum" : 8.207728913728923E-11,
        "min" : 2.0999284182877183E-15,
        "max" : 1.4377821849558047E-11,
        "sumOfSquare" : 7.106355002328198E-22,
        "standardDeviation" : 3.2938836166428785E-12,
        "average" : 1.5199497988386894E-12
      }
    }

Learning Validation

We validate the agreement between the implemented derivative of the internal weights apply finite difference estimations:

SingleDerivativeTester.java:133 executed in 0.18 seconds (0.000 gc):

        return testLearning(
            statistics,
            component.addRef(),
            RefUtil.addRef(inputPrototype),
            outputPrototype.addRef());
      },
      outputPrototype.addRef(),
      RefUtil.addRef(inputPrototype),
      component.addRef()));
Logging 
Learning Gradient for weight setByCoord 0
Weights: [ 0.496, -0.608, -0.128, 1.764, 0.08, 0.7, -0.608, -0.128, 0.0 ]
Implemented Gradient: [ [ 0.08, 0.7, -0.128, 0.496, -0.608, 1.764, 0.048, 1.524, ... ], [ -1.72, -1.028, -0.384, -0.852, 1.912, -1.688, -0.804, 1.108, ... ], [ -0.712, 1.048, -1.616, 1.356, 1.512, 1.556, 1.032, 0.3, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ] ]
Implemented Statistics: {meanExponent=-0.1824058038139548, negative=20, min=-1.72, max=1.912, mean=0.08355555555555556, count=162, sum=13.536000000000001, positive=34, stdDev=0.6387636205694271, zeros=108}
Measured Gradient: [ [ 0.08000000000008001, 0.700000000000145, -0.12799999999979494, 0.4959999999998299, -0.6079999999997199, 1.7639999999996547, 0.04800000000027005, 1.5239999999994147, ... ], [ -1.72000000000061, -1.0279999999995848, -0.38399999999993994, -0.8520000000000749, 1.9120000000016901, -1.6879999999996897, -0.8039999999998049, 1.1079999999998036, ... ], [ -0.71200000000049, 1.04800000000016, -1.6159999999998398, 1.3560000000001349, 1.5119999999990696, 1.556000000000335, 1.0319999999996998, 0.2999999999999531, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ] ]
Measured Statistics: {meanExponent=-0.1824058038139651, negative=20, min=-1.72000000000061, max=1.9120000000016901, mean=0.0835555555556577, count=162, sum=13.536000000016548, positive=34, stdDev=0.6387636205695167, zeros=108}
Gradient Error: [ [ 8.000544671205034E-14, 1.4499512701604544E-13, 2.0505819264826641E-13, -1.7008616737257398E-13, 2.80109269112927E-13, -3.452793606584237E-13, 2.700478729522615E-13, -5.853095785823825E-13, ... ], [ -6.09956529729061E-13, 4.1522341120980855E-13, 6.006306563222097E-14, -7.494005416219807E-14, 1.6902035326893383E-12, 3.1019631308026874E-13, 1.9517720772910252E-13, -1.965094753586527E-13, ... ], [ -4.900524430695441E-13, 1.5987211554602254E-13, 1.603162047558726E-13, 1.34781075189494E-13, -9.303668946358812E-13, 3.348432642269472E-13, -3.0020430585864233E-13, -4.6906922790412864E-14, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ], [ 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ... ] ]
Error Statistics: {meanExponent=-12.469784754941802, negative=18, min=-1.4152880256634859E-12, max=2.9953817204386723E-12, mean=1.0215041261422923E-13, count=162, sum=1.6548366843505136E-11, positive=36, stdDev=5.240956084524837E-13, zeros=108}

Returns

    {
      "absoluteTol" : {
        "count" : 648,
        "sum" : 5.873558583946448E-11,
        "min" : 0.0,
        "max" : 2.9953817204386723E-12,
        "sumOfSquare" : 7.486530659281595E-23,
        "standardDeviation" : 3.2759277879926045E-13,
        "average" : 9.064133617201308E-14
      },
      "relativeTol" : {
        "count" : 108,
        "sum" : 1.4824126962855074E-10,
        "min" : 2.0999284182877183E-15,
        "max" : 2.5273000458915258E-11,
        "sumOfSquare" : 1.6030821196403657E-21,
        "standardDeviation" : 3.599904201829834E-12,
        "average" : 1.3726043484125068E-12
      }
    }

Total Accuracy

The overall agreement accuracy between the implemented derivative and the finite difference estimations:

SingleDerivativeTester.java:148 executed in 0.00 seconds (0.000 gc):

    //log.info(String.format("Component: %s\nInputs: %s\noutput=%s", component, Arrays.toStream(inputPrototype), outputPrototype));
    log.info(RefString.format("Finite-Difference Derivative Accuracy:"));
    log.info(RefString.format("absoluteTol: %s", statistics.absoluteTol));
    log.info(RefString.format("relativeTol: %s", statistics.relativeTol));
Logging 
Finite-Difference Derivative Accuracy:
absoluteTol: 9.0641e-14 +- 3.2759e-13 [0.0000e+00 - 2.9954e-12] (648#)
relativeTol: 1.3726e-12 +- 3.5999e-12 [2.0999e-15 - 2.5273e-11] (108#)

Frozen and Alive Status

SingleDerivativeTester.java:156 executed in 0.03 seconds (0.000 gc):

    testFrozen(component.addRef(), RefUtil.addRef(inputPrototype));
    testUnFrozen(component.addRef(), RefUtil.addRef(inputPrototype));

LayerTests.java:425 executed in 0.00 seconds (0.000 gc):

    throwException(exceptions.addRef());

Results

classdetailsresult
com.simiacryptus.mindseye.test.unit.SingleDerivativeTesterToleranceStatistics{absoluteTol=9.0641e-14 +- 3.2759e-13 [0.0000e+00 - 2.9954e-12] (648#), relativeTol=1.3726e-12 +- 3.5999e-12 [2.0999e-15 - 2.5273e-11] (108#)}OK
 
  {
    "result": "OK",
    "performance": {
      "execution_time": "1.104",
      "gc_time": "0.208"
    },
    "created_on": 1586746088763,
    "file_name": "derivativeTest",
    "report": {
      "simpleName": "BandLimit",
      "canonicalName": "com.simiacryptus.mindseye.layers.cudnn.conv.ConvolutionLayerTest.BandLimit",
      "link": "https://github.com/SimiaCryptus/mindseye-cudnn/tree/59d5b3318556370acb2d83ee6ec123ce0fc6974f/src/test/java/com/simiacryptus/mindseye/layers/cudnn/conv/ConvolutionLayerTest.java",
      "javaDoc": ""
    },
    "archive": "s3://code.simiacrypt.us/tests/com/simiacryptus/mindseye/layers/cudnn/conv/ConvolutionLayer/BandLimit/derivativeTest/202004134808",
    "id": "f27ea485-2689-4ee5-ba4f-6c0269615b95",
    "report_type": "Components",
    "display_name": "Derivative Validation",
    "target": {
      "simpleName": "ConvolutionLayer",
      "canonicalName": "com.simiacryptus.mindseye.layers.cudnn.conv.ConvolutionLayer",
      "link": "https://github.com/SimiaCryptus/mindseye-cudnn/tree/59d5b3318556370acb2d83ee6ec123ce0fc6974f/src/main/java/com/simiacryptus/mindseye/layers/cudnn/conv/ConvolutionLayer.java",
      "javaDoc": ""
    }
  }