Subreport: Logs for com.simiacryptus.ref.lang.ReferenceCountingBase
Using Seed 2903102412465031168
In this apply, we use a network to learn this target input, given it's pre-evaluated output:
TrainingTester.java:332 executed in 0.01 seconds (0.000 gc):
return RefArrays.stream(RefUtil.addRef(input_target)).flatMap(RefArrays::stream).map(x -> {
try {
return x.prettyPrint();
} finally {
x.freeRef();
}
}).reduce((a, b) -> a + "\n" + b).orElse("");
Returns
[
[ [ -0.636 ], [ 0.092 ], [ 1.612 ], [ 0.392 ], [ 1.524 ], [ -1.572 ] ],
[ [ 1.048 ], [ 0.496 ], [ -0.712 ], [ -1.228 ], [ -1.476 ], [ 1.108 ] ],
[ [ -0.608 ], [ -0.128 ], [ -1.028 ], [ 0.7 ], [ 0.028 ], [ -0.068 ] ],
[ [ -0.804 ], [ 0.788 ], [ 1.64 ], [ 1.356 ], [ 0.08 ], [ 1.208 ] ],
[ [ 1.912 ], [ 1.764 ], [ 0.3 ], [ -0.852 ], [ 1.704 ], [ -1.492 ] ],
[ [ -0.408 ], [ -1.616 ], [ -0.556 ], [ -1.72 ], [ 1.556 ], [ 1.876 ] ],
[ [ -0.384 ], [ 1.552 ], [ -0.176 ], [ 0.636 ], [ -0.384 ], [ 0.048 ] ],
[ [ -1.688 ], [ 1.512 ], [ 1.032 ], [ 1.556 ], [ -1.516 ], [ -0.768 ] ]
]
[
[ [ 0.092 ], [ 1.552 ], [ 1.64 ], [ -1.476 ], [ 1.524 ], [ 0.636 ] ],
[ [ 1.764 ], [ 0.496 ], [ -0.176 ], [ 0.3 ], [ 1.512 ], [ -0.636 ] ],
[ [ 1.108 ], [ -0.384 ], [ 0.08 ], [ -1.492 ], [ -1.028 ], [ -0.408 ] ],
[ [ -0.384 ], [ -1.616 ], [ 0.048 ], [ -1.228 ], [ -0.852 ], [ 1.208 ] ],
[ [ -1.72 ], [ 1.912 ], [ -1.688 ], [ -1.516 ], [ -0.128 ], [ 1.556 ] ],
[ [ -0.768 ], [ 1.032 ], [ 1.612 ], [ 1.704 ], [ -1.572 ], [ 0.7 ] ],
[ [ -0.608 ], [ 1.876 ], [ -0.556 ], [ 1.048 ], [ 0.028 ], [ 1.356 ] ],
[ [ -0.712 ], [ 0.788 ], [ -0.804 ], [ -0.068 ], [ 0.392 ], [ 1.556 ] ]
]
[
[ [ 0.788 ], [ -1.72 ], [ -0.384 ], [ 1.356 ], [ 1.612 ], [ -0.128 ] ],
[ [ -0.636 ], [ -0.768 ], [ 0.3 ], [ 0.092 ], [ -0.384 ], [ 0.028 ] ],
[ [ -0.712 ], [ -1.028 ], [ 1.912 ], [ -0.852 ], [ 0.08 ], [ 1.876 ] ],
[ [ 1.552 ], [ -1.688 ], [ 1.512 ], [ -0.408 ], [ 1.208 ], [ -1.492 ] ],
[ [ -0.804 ], [ -0.608 ], [ 1.556 ], [ 1.764 ], [ 0.392 ], [ 0.636 ] ],
[ [ 1.704 ], [ -1.476 ], [ 0.496 ], [ 1.64 ], [ -1.572 ], [ -0.176 ] ],
[ [ 1.108 ], [ 1.032 ], [ -0.068 ], [ 0.7 ], [ -0.556 ], [ 1.556 ] ],
[ [ 0.048 ], [ 1.524 ], [ -1.616 ], [ -1.516 ], [ 1.048 ], [ -1.228 ] ]
]
[
[ [ -0.128 ], [ 1.108 ], [ -0.608 ], [ 0.7 ], [ 0.3 ], [ -1.688 ] ],
[ [ -0.384 ], [ -0.636 ], [ -0.768 ], [ 0.092 ], [ -1.492 ], [ -0.068 ] ],
[ [ 0.08 ], [ 1.556 ], [ 1.552 ], [ -0.852 ], [ -1.476 ], [ 1.764 ] ],
[ [ 1.032 ], [ 0.636 ], [ 1.556 ], [ 1.64 ], [ -0.712 ], [ 0.028 ] ],
[ [ 1.912 ], [ 1.512 ], [ -1.572 ], [ 1.524 ], [ 1.048 ], [ -1.516 ] ],
[ [ -1.228 ], [ 0.788 ], [ -1.028 ], [ 0.048 ], [ 0.392 ], [ -0.384 ] ],
[ [ 0.496 ], [ 1.704 ], [ -1.616 ], [ -0.176 ], [ -0.408 ], [ 1.208 ] ],
[ [ -0.804 ], [ -1.72 ], [ 1.612 ], [ -0.556 ], [ 1.356 ], [ 1.876 ] ]
]
[
[ [ 1.524 ], [ -1.228 ], [ -0.804 ], [ 1.704 ], [ -0.176 ], [ 1.556 ] ],
[ [ 0.636 ], [ 0.788 ], [ 1.876 ], [ 0.092 ], [ 0.7 ], [ -0.408 ] ],
[ [ -0.712 ], [ -1.688 ], [ 0.028 ], [ -1.516 ], [ 1.032 ], [ -1.028 ] ],
[ [ -1.492 ], [ 1.048 ], [ -0.852 ], [ 1.612 ], [ 1.64 ], [ -0.636 ] ],
[ [ 1.764 ], [ 0.3 ], [ 1.108 ], [ -0.556 ], [ 0.496 ], [ 1.556 ] ],
[ [ -0.128 ], [ 1.356 ], [ -0.384 ], [ 0.08 ], [ -0.068 ], [ 1.512 ] ],
[ [ 1.208 ], [ 1.912 ], [ -1.72 ], [ -1.476 ], [ -0.768 ], [ -1.572 ] ],
[ [ 1.552 ], [ -0.608 ], [ -1.616 ], [ -0.384 ], [ 0.392 ], [ 0.048 ] ]
]
First, we train using basic gradient descent method apply weak line search conditions.
TrainingTester.java:480 executed in 1.17 seconds (0.000 gc):
IterativeTrainer iterativeTrainer = new IterativeTrainer(trainable.addRef());
try {
iterativeTrainer.setLineSearchFactory(label -> new ArmijoWolfeSearch());
iterativeTrainer.setOrientation(new GradientDescent());
iterativeTrainer.setMonitor(TrainingTester.getMonitor(history));
iterativeTrainer.setTimeout(30, TimeUnit.SECONDS);
iterativeTrainer.setMaxIterations(250);
iterativeTrainer.setTerminateThreshold(0);
return iterativeTrainer.run();
} finally {
iterativeTrainer.freeRef();
}
Reset training subject: 1873427727227
Reset training subject: 1873477910228
Constructing line search parameters: GD
th(0)=2.3288808;dx=-0.15525872
New Minimum: 2.3288808 > 2.006396813386603
WOLFE (weak): th(2.154434690031884)=2.006396813386603; dx=-0.1441088942567351 evalInputDelta=0.32248398661339683
New Minimum: 2.006396813386603 > 1.7079343981423065
END: th(4.308869380063768)=1.7079343981423065; dx=-0.1329590685134702 evalInputDelta=0.6209464018576933
Fitness changed from 2.3288808 to 1.7079343981423065
Iteration 1 complete. Error: 1.7079343981423065 Total: 0.1758; Orientation: 0.0048; Line Search: 0.0881
th(0)=1.7079343981423065;dx=-0.11386229320948713
New Minimum: 1.7079343981423065 > 0.8144697502656415
END: th(9.283177667225559)=0.8144697502656415; dx=-0.07862882996077382 evalInputDelta=0.893464647876665
Fitness changed from 1.7079343981423065 to 0.8144697502656415
Iteration 2 complete. Error: 0.8144697502656415 Total: 0.0618; Orientation: 0.0026; Line Search: 0.0424
th(0)=0.8144697502656415;dx=-0.05429798335104277
New Minimum: 0.8144697502656415 > 0.09049663891840456
END: th(20.000000000000004)=0.09049663891840456; dx=-0.018099327783680917 evalInputDelta=0.723973111347237
Fitness changed from 0.8144697502656415 to 0.09049663891840456
Iteration 3 complete. Error: 0.09049663891840456 Total: 0.0506; Orientation: 0.0023; Line Search: 0.0362
th(0)=0.09049663891840456;dx=-0.006033109261226971
New Minimum: 0.09049663891840456 > 0.017225925154783935
WOLF (strong): th(43.088693800637685)=0.017225925154783935; dx=0.0026321839928663752 evalInputDelta=0.07327071376362063
New Minimum: 0.017225925154783935 > 0.00718926107179931
END: th(21.544346900318843)=0.00718926107179931; dx=-0.0017004626341802977 evalInputDelta=0.08330737784660526
Fitness changed from 0.09049663891840456 to 0.00718926107179931
Iteration 4 complete. Error: 0.00718926107179931 Total: 0.0720; Orientation: 0.0025; Line Search: 0.0570
th(0)=0.00718926107179931;dx=-4.792840714532874E-4
New Minimum: 0.00718926107179931 > 0.002152635628583541
WOLF (strong): th(46.4158883361278)=0.002152635628583541; dx=2.6226245994206217E-4 evalInputDelta=0.005036625443215769
New Minimum: 0.002152635628583541 > 3.6850572553024703E-4
END: th(23.2079441680639)=3.6850572553024703E-4; dx=-1.0851080575561264E-4 evalInputDelta=0.006820755346269063
Fitness changed from 0.00718926107179931 to 3.6850572553024703E-4
Iteration 5 complete. Error: 3.6850572553024703E-4 Total: 0.0625; Orientation: 0.0016; Line Search: 0.0496
th(0)=3.6850572553024703E-4;dx=-2.4567048368683137E-5
New Minimum: 3.6850572553024703E-4 > 1.637803224578878E-4
WOLF (strong): th(50.000000000000014)=1.637803224578878E-4; dx=1.6378032245788767E-5 evalInputDelta=2.0472540307235924E-4
New Minimum: 1.637803224578878E-4 > 1.0236270153617892E-5
END: th(25.000000000000007)=1.0236270153617892E-5; dx=-4.094508061447174E-6 evalInputDelta=3.582694553766291E-4
Fitness changed from 3.6850572553024703E-4 to 1.0236270153617892E-5
Iteration 6 complete. Error: 1.0236270153617892E-5 Total: 0.0470; Orientation: 0.0013; Line Search: 0.0364
th(0)=1.0236270153617892E-5;dx=-6.82418010241193E-7
New Minimum: 1.0236270153617892E-5 > 6.475475713128259E-6
WOLF (strong): th(53.860867250797114)=6.475475713128259E-6; dx=5.427695183972756E-7 evalInputDelta=3.760794440489633E-6
New Minimum: 6.475475713128259E-6 > 1.0716507870696338E-7
END: th(26.930433625398557)=1.0716507870696338E-7; dx=-6.982424592195598E-8 evalInputDelta=1.012910507491093E-5
Fitness changed from 1.0236270153617892E-5 to 1.0716507870696338E-7
Iteration 7 complete. Error: 1.0716507870696338E-7 Total: 0.0520; Orientation: 0.0011; Line Search: 0.0397
th(0)=1.0716507870696338E-7;dx=-7.144338580464226E-9
New Minimum: 1.0716507870696338E-7 > 9.348516801275299E-8
WOLF (strong): th(58.01986042015976)=9.348516801275299E-8; dx=6.672778993965915E-9 evalInputDelta=1.3679910694210395E-8
New Minimum: 9.348516801275299E-8 > 1.167192251848466E-10
END: th(29.00993021007988)=1.167192251848466E-10; dx=-2.3577979324926484E-10 evalInputDelta=1.0704835948177854E-7
Fitness changed from 1.0716507870696338E-7 to 1.167192251848466E-10
Iteration 8 complete. Error: 1.167192251848466E-10 Total: 0.0453; Orientation: 0.0010; Line Search: 0.0354
Low gradient: 2.789494878824798E-6
th(0)=1.167192251848466E-10;dx=-7.781281678989772E-12
Armijo: th(62.50000000000003)=1.3698297955741747E-10; dx=8.429721818911793E-12 evalInputDelta=-2.0263754372570858E-11
New Minimum: 1.167192251848466E-10 > 2.026375437367321E-13
WOLF (strong): th(31.250000000000014)=2.026375437367321E-13; dx=3.242200699683393E-13 evalInputDelta=1.1651658764110988E-10
END: th(10.416666666666671)=4.9736259342740154E-11; dx=-5.079447762668352E-12 evalInputDelta=6.698296584210645E-11
Fitness changed from 1.167192251848466E-10 to 2.026375437367321E-13
Iteration 9 complete. Error: 2.026375437367321E-13 Total: 0.0569; Orientation: 0.0014; Line Search: 0.0468
Low gradient: 1.1622895328810635E-7
th(0)=2.026375437367321E-13;dx=-1.3509169582448809E-14
New Minimum: 2.026375437367321E-13 > 1.2861391487330364E-14
END: th(22.442028021165466)=1.2861391487330364E-14; dx=-3.403397505127383E-15 evalInputDelta=1.8977615224940173E-13
Fitness changed from 2.026375437367321E-13 to 1.2861391487330364E-14
Iteration 10 complete. Error: 1.2861391487330364E-14 Total: 0.0369; Orientation: 0.0012; Line Search: 0.0261
Low gradient: 2.9281839067165122E-8
th(0)=1.2861391487330364E-14;dx=-8.574260991553577E-16
New Minimum: 1.2861391487330364E-14 > 4.811849990787794E-15
WOLF (strong): th(48.34988368346647)=4.811849990787794E-15; dx=5.244556395172631E-16 evalInputDelta=8.04954149654257E-15
New Minimum: 4.811849990787794E-15 > 4.848930728151569E-16
END: th(24.174941841733236)=4.848930728151569E-16; dx=-1.6648522976155268E-16 evalInputDelta=1.2376498414515207E-14
Fitness changed from 1.2861391487330364E-14 to 4.848930728151569E-16
Iteration 11 complete. Error: 4.848930728151569E-16 Total: 0.0482; Orientation: 0.0019; Line Search: 0.0376
Low gradient: 5.685613850266636E-9
th(0)=4.848930728151569E-16;dx=-3.232620485434379E-17
New Minimum: 4.848930728151569E-16 > 2.627439512942158E-16
WOLF (strong): th(52.083333333333364)=2.627439512942158E-16; dx=2.3795678590557222E-17 evalInputDelta=2.2214912152094116E-16
New Minimum: 2.627439512942158E-16 > 8.441666646464659E-18
END: th(26.041666666666682)=8.441666646464659E-18; dx=-4.2652631440955496E-18 evalInputDelta=4.764514061686923E-16
Fitness changed from 4.848930728151569E-16 to 8.441666646464659E-18
Iteration 12 complete. Error: 8.441666646464659E-18 Total: 0.0432; Orientation: 0.0012; Line Search: 0.0335
Low gradient: 7.501851614308145E-10
th(0)=8.441666646464659E-18;dx=-5.627777764309772E-19
New Minimum: 8.441666646464659E-18 > 6.391980119491586E-18
WOLF (strong): th(56.105070052913675)=6.391980119491586E-18; dx=4.897117763636407E-19 evalInputDelta=2.0496865269730722E-18
New Minimum: 6.391980119491586E-18 > 3.557336487665235E-20
END: th(28.052535026456837)=3.557336487665235E-20; dx=-3.6532998038712706E-20 evalInputDelta=8.406093281588007E-18
Fitness changed from 8.441666646464659E-18 to 3.557336487665235E-20
Iteration 13 complete. Error: 3.557336487665235E-20 Total: 0.0540; Orientation: 0.0011; Line Search: 0.0431
Zero gradient: 4.8698641238164855E-11
th(0)=3.557336487665235E-20;dx=-2.3715576584434905E-21
Armijo: th(60.4373546043331)=3.661814161351914E-20; dx=2.40613153119974E-21 evalInputDelta=-1.0447767368667872E-21
New Minimum: 3.557336487665235E-20 > 1.8901019402396014E-24
WOLF (strong): th(30.21867730216655)=1.8901019402396014E-24; dx=1.7286768976747066E-23 evalInputDelta=3.557147477471211E-20
END: th(10.072892434055516)=1.569534906217893E-20; dx=-1.5752761719848317E-21 evalInputDelta=1.987801581447342E-20
Fitness changed from 3.557336487665235E-20 to 1.8901019402396014E-24
Iteration 14 complete. Error: 1.8901019402396014E-24 Total: 0.0619; Orientation: 0.0015; Line Search: 0.0506
Zero gradient: 3.549743596599245E-13
th(0)=1.8901019402396014E-24;dx=-1.2600679601597344E-25
New Minimum: 1.8901019402396014E-24 > 1.4462900305522458E-25
END: th(21.701388888888903)=1.4462900305522458E-25; dx=-3.4856120403152416E-26 evalInputDelta=1.7454729371843767E-24
Fitness changed from 1.8901019402396014E-24 to 1.4462900305522458E-25
Iteration 15 complete. Error: 1.4462900305522458E-25 Total: 0.0367; Orientation: 0.0014; Line Search: 0.0248
Zero gradient: 9.819334772282169E-14
th(0)=1.4462900305522458E-25;dx=-9.641933537014973E-27
New Minimum: 1.4462900305522458E-25 > 4.510816251498466E-26
WOLF (strong): th(46.75422504409473)=4.510816251498466E-26; dx=5.384730805626678E-27 evalInputDelta=9.952084054023992E-26
New Minimum: 4.510816251498466E-26 > 7.046511672515906E-27
END: th(23.377112522047366)=7.046511672515906E-27; dx=-2.1282538966358186E-27 evalInputDelta=1.3758249138270867E-25
Fitness changed from 1.4462900305522458E-25 to 7.046511672515906E-27
Iteration 16 complete. Error: 7.046511672515906E-27 Total: 0.0507; Orientation: 0.0013; Line Search: 0.0396
Zero gradient: 2.1674119240107396E-14
th(0)=7.046511672515906E-27;dx=-4.697674448343937E-28
New Minimum: 7.046511672515906E-27 > 3.2474724218531625E-27
WOLF (strong): th(50.364462170277584)=3.2474724218531625E-27; dx=3.1891019135402064E-28 evalInputDelta=3.799039250662744E-27
New Minimum: 3.2474724218531625E-27 > 1.8173810018500065E-28
END: th(25.182231085138792)=1.8173810018500065E-28; dx=-7.544256191329526E-29 evalInputDelta=6.864773572330906E-27
Fitness changed from 7.046511672515906E-27 to 1.8173810018500065E-28
Iteration 17 complete. Error: 1.8173810018500065E-28 Total: 0.0473; Orientation: 0.0012; Line Search: 0.0379
Zero gradient: 3.4807863114053283E-15
th(0)=1.8173810018500065E-28;dx=-1.211587334566671E-29
New Minimum: 1.8173810018500065E-28 > 1.188388845010657E-28
WOLF (strong): th(54.253472222222264)=1.188388845010657E-28; dx=9.797320326925668E-30 evalInputDelta=6.289921568393495E-29
New Minimum: 1.188388845010657E-28 > 1.6380708252493977E-30
END: th(27.126736111111132)=1.6380708252493977E-30; dx=-1.1493566862036308E-30 evalInputDelta=1.8010002935975125E-28
Fitness changed from 1.8173810018500065E-28 to 1.6380708252493977E-30
Iteration 18 complete. Error: 1.6380708252493977E-30 Total: 0.0415; Orientation: 0.0012; Line Search: 0.0320
Zero gradient: 3.3046137699176463E-16
th(0)=1.6380708252493977E-30;dx=-1.0920472168329317E-31
New Minimum: 1.6380708252493977E-30 > 1.4991695477769495E-30
WOLF (strong): th(58.44278130511842)=1.4991695477769495E-30; dx=1.0438026717260316E-31 evalInputDelta=1.3890127747244819E-31
New Minimum: 1.4991695477769495E-30 > 1.2919780043597706E-34
END: th(29.22139065255921)=1.2919780043597706E-34; dx=-2.359264181874364E-34 evalInputDelta=1.6379416274489617E-30
Fitness changed from 1.6380708252493977E-30 to 1.2919780043597706E-34
Iteration 19 complete. Error: 1.2919780043597706E-34 Total: 0.0422; Orientation: 0.0010; Line Search: 0.0336
Zero gradient: 2.934823111489312E-18
th(0)=1.2919780043597706E-34;dx=-8.613186695731804E-36
Armijo: th(62.955577712846996)=1.2919780043597706E-34; dx=8.613186695731804E-36 evalInputDelta=0.0
New Minimum: 1.2919780043597706E-34 > 0.0
END: th(31.477788856423498)=0.0; dx=0.0 evalInputDelta=1.2919780043597706E-34
Fitness changed from 1.2919780043597706E-34 to 0.0
Iteration 20 complete. Error: 0.0 Total: 0.0414; Orientation: 0.0010; Line Search: 0.0281
Zero gradient: 0.0
th(0)=0.0;dx=0.0 (ERROR: Starting derivative negative)
Fitness changed from 0.0 to 0.0
Static Iteration Total: 0.0274; Orientation: 0.0010; Line Search: 0.0188
Iteration 21 failed. Error: 0.0
Previous Error: 0.0 -> 0.0
Optimization terminated 21
Final threshold in iteration 21: 0.0 (> 0.0) after 1.156s (< 30.000s)
Returns
0.0
First, we use a conjugate gradient descent method, which converges the fastest for purely linear functions.
TrainingTester.java:452 executed in 0.32 seconds (0.000 gc):
IterativeTrainer iterativeTrainer = new IterativeTrainer(trainable.addRef());
try {
iterativeTrainer.setLineSearchFactory(label -> new QuadraticSearch());
iterativeTrainer.setOrientation(new GradientDescent());
iterativeTrainer.setMonitor(TrainingTester.getMonitor(history));
iterativeTrainer.setTimeout(30, TimeUnit.SECONDS);
iterativeTrainer.setMaxIterations(250);
iterativeTrainer.setTerminateThreshold(0);
return iterativeTrainer.run();
} finally {
iterativeTrainer.freeRef();
}
Reset training subject: 1874591210357
Reset training subject: 1874599571161
Constructing line search parameters: GD
F(0.0) = LineSearchPoint{point=PointSample{avg=2.3288808}, derivative=-0.15525872}
New Minimum: 2.3288808 > 2.328880799984474
F(1.0E-10) = LineSearchPoint{point=PointSample{avg=2.328880799984474}, derivative=-0.15525871999948246}, evalInputDelta = -1.552580286556804E-11
New Minimum: 2.328880799984474 > 2.328880799891319
F(7.000000000000001E-10) = LineSearchPoint{point=PointSample{avg=2.328880799891319}, derivative=-0.1552587199963773}, evalInputDelta = -1.0868106414818612E-10
New Minimum: 2.328880799891319 > 2.328880799239232
F(4.900000000000001E-9) = LineSearchPoint{point=PointSample{avg=2.328880799239232}, derivative=-0.15525871997464108}, evalInputDelta = -7.607678931265127E-10
New Minimum: 2.328880799239232 > 2.3288807946746255
F(3.430000000000001E-8) = LineSearchPoint{point=PointSample{avg=2.3288807946746255}, derivative=-0.1552587198224875}, evalInputDelta = -5.325374363707169E-9
New Minimum: 2.3288807946746255 > 2.3288807627223815
F(2.4010000000000004E-7) = LineSearchPoint{point=PointSample{avg=2.3288807627223815}, derivative=-0.1552587187574127}, evalInputDelta = -3.7277618325504136E-8
New Minimum: 2.3288807627223815 > 2.3288805390566765
F(1.6807000000000003E-6) = LineSearchPoint{point=PointSample{avg=2.3288805390566765}, derivative=-0.15525871130188898}, evalInputDelta = -2.6094332339354764E-7
New Minimum: 2.3288805390566765 > 2.3288789733970434
F(1.1764900000000001E-5) = LineSearchPoint{point=PointSample{avg=2.3288789733970434}, derivative=-0.15525865911322284}, evalInputDelta = -1.8266029564451003E-6
New Minimum: 2.3288789733970434 > 2.328868013794345
F(8.235430000000001E-5) = LineSearchPoint{point=PointSample{avg=2.328868013794345}, derivative=-0.15525829379255984}, evalInputDelta = -1.2786205654702343E-5
New Minimum: 2.328868013794345 > 2.328791297297519
F(5.764801000000001E-4) = LineSearchPoint{point=PointSample{avg=2.328791297297519}, derivative=-0.15525573654791897}, evalInputDelta = -8.950270248098136E-5
New Minimum: 2.328791297297519 > 2.328254317200548
F(0.004035360700000001) = LineSearchPoint{point=PointSample{avg=2.328254317200548}, derivative=-0.15523783583543266}, evalInputDelta = -6.264827994519884E-4
New Minimum: 2.328254317200548 > 2.324497190181713
F(0.028247524900000005) = LineSearchPoint{point=PointSample{avg=2.324497190181713}, derivative=-0.1551125308480286}, evalInputDelta = -0.004383609818286782
New Minimum: 2.324497190181713 > 2.298282250387911
F(0.19773267430000002) = LineSearchPoint{point=PointSample{avg=2.298282250387911}, derivative=-0.15423539593620017}, evalInputDelta = -0.030598549612089077
New Minimum: 2.298282250387911 > 2.1189401893953974
F(1.3841287201) = LineSearchPoint{point=PointSample{avg=2.1189401893953974}, derivative=-0.14809545155340118}, evalInputDelta = -0.20994061060460245
New Minimum: 2.1189401893953974 > 1.0675091230889762
F(9.688901040700001) = LineSearchPoint{point=PointSample{avg=1.0675091230889762}, derivative=-0.10511584087380833}, evalInputDelta = -1.2613716769110237
F(67.8223072849) = LineSearchPoint{point=PointSample{avg=3.7016963303613424}, derivative=0.19574143388334167}, evalInputDelta = 1.3728155303613425
F(5.217100560376924) = LineSearchPoint{point=PointSample{avg=1.589311333327564}, derivative=-0.12825870816281987}, evalInputDelta = -0.739569466672436
New Minimum: 1.0675091230889762 > 0.10999184789760559
F(36.51970392263847) = LineSearchPoint{point=PointSample{avg=0.10999184789760559}, derivative=0.03374136286026093}, evalInputDelta = -2.2188889521023945
0.10999184789760559 <= 2.3288808
New Minimum: 0.10999184789760559 > 1.7233863119025113E-32
F(30.0) = LineSearchPoint{point=PointSample{avg=1.7233863119025113E-32}, derivative=-2.5642451125425693E-18}, evalInputDelta = -2.3288808
Left bracket at 30.0
F(33.259851961319235) = LineSearchPoint{point=PointSample{avg=0.02749796197440139}, derivative=0.016870681430130464}, evalInputDelta = -2.3013828380255985
Right bracket at 33.259851961319235
Converged to left
Fitness changed from 2.3288808 to 1.7233863119025113E-32
Iteration 1 complete. Error: 1.7233863119025113E-32 Total: 0.2713; Orientation: 0.0009; Line Search: 0.2358
Zero gradient: 3.389578451570354E-17
F(0.0) = LineSearchPoint{point=PointSample{avg=1.7233863119025113E-32}, derivative=-1.1489242079350077E-33}
New Minimum: 1.7233863119025113E-32 > 0.0
F(30.0) = LineSearchPoint{point=PointSample{avg=0.0}, derivative=0.0}, evalInputDelta = -1.7233863119025113E-32
0.0 <= 1.7233863119025113E-32
Converged to right
Fitness changed from 1.7233863119025113E-32 to 0.0
Iteration 2 complete. Error: 0.0 Total: 0.0310; Orientation: 0.0011; Line Search: 0.0212
Zero gradient: 0.0
F(0.0) = LineSearchPoint{point=PointSample{avg=0.0}, derivative=0.0}
Fitness changed from 0.0 to 0.0
Static Iteration Total: 0.0206; Orientation: 0.0021; Line Search: 0.0086
Iteration 3 failed. Error: 0.0
Previous Error: 0.0 -> 0.0
Optimization terminated 3
Final threshold in iteration 3: 0.0 (> 0.0) after 0.323s (< 30.000s)
Returns
0.0
Next, we apply the same optimization using L-BFGS, which is nearly ideal for purely second-order or quadratic functions.
TrainingTester.java:509 executed in 6.79 seconds (0.112 gc):
IterativeTrainer iterativeTrainer = new IterativeTrainer(trainable.addRef());
try {
iterativeTrainer.setLineSearchFactory(label -> new ArmijoWolfeSearch());
iterativeTrainer.setOrientation(new LBFGS());
iterativeTrainer.setMonitor(TrainingTester.getMonitor(history));
iterativeTrainer.setTimeout(30, TimeUnit.SECONDS);
iterativeTrainer.setIterationsPerSample(100);
iterativeTrainer.setMaxIterations(250);
iterativeTrainer.setTerminateThreshold(0);
return iterativeTrainer.run();
} finally {
iterativeTrainer.freeRef();
}
Reset training subject: 1874919464382
Reset training subject: 1874926327847
Adding measurement 30f65f2c to history. Total: 0
LBFGS Accumulation History: 1 points
Constructing line search parameters: GD
Non-optimal measurement 2.3288808 < 2.3288808. Total: 1
th(0)=2.3288808;dx=-0.15525872
Adding measurement 7b609d3c to history. Total: 1
New Minimum: 2.3288808 > 2.006396813386603
WOLFE (weak): th(2.154434690031884)=2.006396813386603; dx=-0.1441088942567351 evalInputDelta=0.32248398661339683
Adding measurement 6f1b683d to history. Total: 2
New Minimum: 2.006396813386603 > 1.7079343981423065
END: th(4.308869380063768)=1.7079343981423065; dx=-0.1329590685134702 evalInputDelta=0.6209464018576933
Fitness changed from 2.3288808 to 1.7079343981423065
Iteration 1 complete. Error: 1.7079343981423065 Total: 0.0623; Orientation: 0.0045; Line Search: 0.0379
Non-optimal measurement 1.7079343981423065 < 1.7079343981423065. Total: 3
LBFGS Accumulation History: 3 points
Non-optimal measurement 1.7079343981423065 < 1.7079343981423065. Total: 3
th(0)=1.7079343981423065;dx=-0.11386229320948713
Adding measurement 60e83735 to history. Total: 3
New Minimum: 1.7079343981423065 > 0.8144697502656415
END: th(9.283177667225559)=0.8144697502656415; dx=-0.07862882996077382 evalInputDelta=0.893464647876665
Fitness changed from 1.7079343981423065 to 0.8144697502656415
Iteration 2 complete. Error: 0.8144697502656415 Total: 0.0331; Orientation: 0.0025; Line Search: 0.0227
Non-optimal measurement 0.8144697502656415 < 0.8144697502656415. Total: 4
Rejected: LBFGS Orientation magnitude: 6.991e+00, gradient 2.330e-01, dot -1.000; [f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 0.8144697502656415, 1.7079343981423065, 2.006396813386603, 2.3288808
LBFGS Accumulation History: 3 points
Removed measurement 60e83735 to history. Total: 3
Adding measurement 552e7c8d to history. Total: 3
th(0)=0.8144697502656415;dx=-0.05429798335104277
Adding measurement a9bbac7 to history. Total: 4
New Minimum: 0.8144697502656415 > 0.09049663891840456
END: th(20.000000000000004)=0.09049663891840456; dx=-0.018099327783680917 evalInputDelta=0.723973111347237
Fitness changed from 0.8144697502656415 to 0.09049663891840456
Iteration 3 complete. Error: 0.09049663891840456 Total: 0.0893; Orientation: 0.0621; Line Search: 0.0179
Non-optimal measurement 0.09049663891840456 < 0.09049663891840456. Total: 5
Rejected: LBFGS Orientation magnitude: 2.330e+00, gradient 7.767e-02, dot -1.000; [8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00]
Orientation rejected. Popping history element from 0.09049663891840456, 0.8144697502656415, 1.7079343981423065, 2.006396813386603, 2.3288808
Rejected: LBFGS Orientation magnitude: 2.330e+00, gradient 7.767e-02, dot -1.000; [8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 0.09049663891840456, 0.8144697502656415, 1.7079343981423065, 2.006396813386603
LBFGS Accumulation History: 3 points
Removed measurement a9bbac7 to history. Total: 4
Removed measurement 552e7c8d to history. Total: 3
Adding measurement 17409657 to history. Total: 3
th(0)=0.09049663891840456;dx=-0.006033109261226971
Adding measurement 6ab6c7 to history. Total: 4
New Minimum: 0.09049663891840456 > 0.017225925154783935
WOLF (strong): th(43.088693800637685)=0.017225925154783935; dx=0.0026321839928663757 evalInputDelta=0.07327071376362063
Adding measurement 107c64c2 to history. Total: 5
New Minimum: 0.017225925154783935 > 0.00718926107179931
END: th(21.544346900318843)=0.00718926107179931; dx=-0.0017004626341802977 evalInputDelta=0.08330737784660526
Fitness changed from 0.09049663891840456 to 0.00718926107179931
Iteration 4 complete. Error: 0.00718926107179931 Total: 0.2339; Orientation: 0.1935; Line Search: 0.0334
Non-optimal measurement 0.00718926107179931 < 0.00718926107179931. Total: 6
Rejected: LBFGS Orientation magnitude: 6.568e-01, gradient 2.189e-02, dot -1.000; [8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00]
Orientation rejected. Popping history element from 0.00718926107179931, 0.017225925154783935, 0.09049663891840456, 1.7079343981423065, 2.006396813386603, 2.3288808
Rejected: LBFGS Orientation magnitude: 6.568e-01, gradient 2.189e-02, dot -1.000; [8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00]
Orientation rejected. Popping history element from 0.00718926107179931, 0.017225925154783935, 0.09049663891840456, 1.7079343981423065, 2.006396813386603
Rejected: LBFGS Orientation magnitude: 6.568e-01, gradient 2.189e-02, dot -1.000; [f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00]
Orientation rejected. Popping history element from 0.00718926107179931, 0.017225925154783935, 0.09049663891840456, 1.7079343981423065
LBFGS Accumulation History: 3 points
Removed measurement 107c64c2 to history. Total: 5
Removed measurement 6ab6c7 to history. Total: 4
Removed measurement 17409657 to history. Total: 3
Adding measurement 390aae93 to history. Total: 3
th(0)=0.00718926107179931;dx=-4.792840714532874E-4
Adding measurement 388539b0 to history. Total: 4
New Minimum: 0.00718926107179931 > 0.002152635628583541
WOLF (strong): th(46.4158883361278)=0.002152635628583541; dx=2.622624599420621E-4 evalInputDelta=0.005036625443215769
Adding measurement 2f29b79d to history. Total: 5
New Minimum: 0.002152635628583541 > 3.6850572553024703E-4
END: th(23.2079441680639)=3.6850572553024703E-4; dx=-1.0851080575561264E-4 evalInputDelta=0.006820755346269063
Fitness changed from 0.00718926107179931 to 3.6850572553024703E-4
Iteration 5 complete. Error: 3.6850572553024703E-4 Total: 0.3611; Orientation: 0.3073; Line Search: 0.0468
Non-optimal measurement 3.6850572553024703E-4 < 3.6850572553024703E-4. Total: 6
Rejected: LBFGS Orientation magnitude: 1.487e-01, gradient 4.957e-03, dot -1.000; [8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 3.6850572553024703E-4, 0.002152635628583541, 0.00718926107179931, 1.7079343981423065, 2.006396813386603, 2.3288808
Rejected: LBFGS Orientation magnitude: 1.487e-01, gradient 4.957e-03, dot -1.000; [8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 3.6850572553024703E-4, 0.002152635628583541, 0.00718926107179931, 1.7079343981423065, 2.006396813386603
Rejected: LBFGS Orientation magnitude: 1.487e-01, gra
...skipping 27826 bytes...
000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.8173810018500065E-28, 3.2474724218531625E-27, 7.046511672515906E-27, 1.7079343981423065, 2.006396813386603
Rejected: LBFGS Orientation magnitude: 1.044e-13, gradient 3.481e-15, dot -1.000; [f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.8173810018500065E-28, 3.2474724218531625E-27, 7.046511672515906E-27, 1.7079343981423065
LBFGS Accumulation History: 3 points
Removed measurement 760cf8d9 to history. Total: 5
Removed measurement 5dbe67bd to history. Total: 4
Removed measurement 447d33d9 to history. Total: 3
Adding measurement 19dd035b to history. Total: 3
th(0)=1.8173810018500065E-28;dx=-1.211587334566671E-29
Adding measurement 2bed5ba1 to history. Total: 4
New Minimum: 1.8173810018500065E-28 > 1.188388845010657E-28
WOLF (strong): th(54.253472222222264)=1.188388845010657E-28; dx=9.797320326925668E-30 evalInputDelta=6.289921568393495E-29
Adding measurement 877e600 to history. Total: 5
New Minimum: 1.188388845010657E-28 > 1.6380708252493977E-30
END: th(27.126736111111132)=1.6380708252493977E-30; dx=-1.1493566862036308E-30 evalInputDelta=1.8010002935975125E-28
Fitness changed from 1.8173810018500065E-28 to 1.6380708252493977E-30
Iteration 18 complete. Error: 1.6380708252493977E-30 Total: 0.3777; Orientation: 0.3370; Line Search: 0.0345
Non-optimal measurement 1.6380708252493977E-30 < 1.6380708252493977E-30. Total: 6
Rejected: LBFGS Orientation magnitude: 9.914e-15, gradient 3.305e-16, dot -1.000; [8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.6380708252493977E-30, 1.188388845010657E-28, 1.8173810018500065E-28, 1.7079343981423065, 2.006396813386603, 2.3288808
Rejected: LBFGS Orientation magnitude: 9.914e-15, gradient 3.305e-16, dot -1.000; [77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.6380708252493977E-30, 1.188388845010657E-28, 1.8173810018500065E-28, 1.7079343981423065, 2.006396813386603
Rejected: LBFGS Orientation magnitude: 9.914e-15, gradient 3.305e-16, dot -1.000; [8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.6380708252493977E-30, 1.188388845010657E-28, 1.8173810018500065E-28, 1.7079343981423065
LBFGS Accumulation History: 3 points
Removed measurement 877e600 to history. Total: 5
Removed measurement 2bed5ba1 to history. Total: 4
Removed measurement 19dd035b to history. Total: 3
Adding measurement 4c05979f to history. Total: 3
th(0)=1.6380708252493977E-30;dx=-1.0920472168329317E-31
Adding measurement 73fd4086 to history. Total: 4
New Minimum: 1.6380708252493977E-30 > 1.4991695477769495E-30
WOLF (strong): th(58.44278130511842)=1.4991695477769495E-30; dx=1.0438026717260316E-31 evalInputDelta=1.3890127747244819E-31
Adding measurement 2aca99d7 to history. Total: 5
New Minimum: 1.4991695477769495E-30 > 1.2919780043597706E-34
END: th(29.22139065255921)=1.2919780043597706E-34; dx=-2.359264181874364E-34 evalInputDelta=1.6379416274489617E-30
Fitness changed from 1.6380708252493977E-30 to 1.2919780043597706E-34
Iteration 19 complete. Error: 1.2919780043597706E-34 Total: 0.5183; Orientation: 0.4798; Line Search: 0.0300
Non-optimal measurement 1.2919780043597706E-34 < 1.2919780043597706E-34. Total: 6
Rejected: LBFGS Orientation magnitude: 8.804e-17, gradient 2.935e-18, dot -1.000; [8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 0.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.2919780043597706E-34, 1.4991695477769495E-30, 1.6380708252493977E-30, 1.7079343981423065, 2.006396813386603, 2.3288808
Rejected: LBFGS Orientation magnitude: 8.804e-17, gradient 2.935e-18, dot -1.000; [e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 0.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.2919780043597706E-34, 1.4991695477769495E-30, 1.6380708252493977E-30, 1.7079343981423065, 2.006396813386603
Rejected: LBFGS Orientation magnitude: 8.804e-17, gradient 2.935e-18, dot -1.000; [e5966327-753c-4a97-bd24-87ca74a8aaef = 1.000/1.000e+00, 8614c49c-445e-4053-b1a7-51401eb745b7 = 1.000/1.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 0.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 1.000/1.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 1.000/1.000e+00]
Orientation rejected. Popping history element from 1.2919780043597706E-34, 1.4991695477769495E-30, 1.6380708252493977E-30, 1.7079343981423065
LBFGS Accumulation History: 3 points
Removed measurement 2aca99d7 to history. Total: 5
Removed measurement 73fd4086 to history. Total: 4
Removed measurement 4c05979f to history. Total: 3
Adding measurement 659f8dce to history. Total: 3
th(0)=1.2919780043597706E-34;dx=-8.613186695731804E-36
Non-optimal measurement 1.2919780043597706E-34 < 1.2919780043597706E-34. Total: 4
Armijo: th(62.955577712846996)=1.2919780043597706E-34; dx=8.613186695731805E-36 evalInputDelta=0.0
Adding measurement 579a4603 to history. Total: 4
New Minimum: 1.2919780043597706E-34 > 0.0
END: th(31.477788856423498)=0.0; dx=0.0 evalInputDelta=1.2919780043597706E-34
Fitness changed from 1.2919780043597706E-34 to 0.0
Iteration 20 complete. Error: 0.0 Total: 0.4365; Orientation: 0.4007; Line Search: 0.0294
Non-optimal measurement 0.0 < 0.0. Total: 5
Rejected: LBFGS Orientation magnitude: 0.000e+00, gradient 0.000e+00, dot NaN; [8614c49c-445e-4053-b1a7-51401eb745b7 = 0.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 0.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 0.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 0.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 0.000e+00]
Orientation rejected. Popping history element from 0.0, 1.2919780043597706E-34, 1.7079343981423065, 2.006396813386603, 2.3288808
Rejected: LBFGS Orientation magnitude: 0.000e+00, gradient 0.000e+00, dot NaN; [8614c49c-445e-4053-b1a7-51401eb745b7 = 0.000e+00, 8175d92f-5a21-4e35-a286-b71c676270b9 = 0.000e+00, e5966327-753c-4a97-bd24-87ca74a8aaef = 0.000e+00, f20ba27d-c267-436a-b83d-418c912ac79e = 0.000e+00, 77cc75fa-f5c4-4e41-8979-2e68e1ae7591 = 0.000e+00]
Orientation rejected. Popping history element from 0.0, 1.2919780043597706E-34, 1.7079343981423065, 2.006396813386603
LBFGS Accumulation History: 3 points
Removed measurement 579a4603 to history. Total: 4
Removed measurement 659f8dce to history. Total: 3
Adding measurement 108d77da to history. Total: 3
th(0)=0.0;dx=0.0 (ERROR: Starting derivative negative)
Non-optimal measurement 0.0 < 0.0. Total: 4
Fitness changed from 0.0 to 0.0
Static Iteration Total: 0.2131; Orientation: 0.1880; Line Search: 0.0186
Iteration 21 failed. Error: 0.0
Previous Error: 0.0 -> 0.0
Optimization terminated 21
Final threshold in iteration 21: 0.0 (> 0.0) after 6.794s (< 30.000s)
Returns
0.0
TrainingTester.java:432 executed in 0.17 seconds (0.000 gc):
return TestUtil.compare(title + " vs Iteration", runs);
Plotting range=[1.0, -33.88874488004565], [20.0, 0.23247118539751013]; valueStats=DoubleSummaryStatistics{count=39, sum=5.240938, min=0.000000, average=0.134383, max=1.707934}
Plotting 20 points for GD
Plotting 2 points for CjGD
Plotting 20 points for LBFGS
Returns
TrainingTester.java:435 executed in 0.02 seconds (0.000 gc):
return TestUtil.compareTime(title + " vs Time", runs);
Plotting range=[0.0, -33.88874488004565], [6.519, 0.23247118539751013]; valueStats=DoubleSummaryStatistics{count=39, sum=5.240938, min=0.000000, average=0.134383, max=1.707934}
Plotting 20 points for GD
Plotting 2 points for CjGD
Plotting 20 points for LBFGS
Returns
TrainingTester.java:255 executed in 0.00 seconds (0.000 gc):
return grid(inputLearning, modelLearning, completeLearning);
Returns
TrainingTester.java:258 executed in 0.00 seconds (0.000 gc):
return new ComponentResult(null == inputLearning ? null : inputLearning.value,
null == modelLearning ? null : modelLearning.value, null == completeLearning ? null : completeLearning.value);
Returns
{"input":{ "LBFGS": { "type": "Converged", "value": 0.0 }, "CjGD": { "type": "Converged", "value": 0.0 }, "GD": { "type": "Converged", "value": 0.0 } }, "model":null, "complete":null}
LayerTests.java:425 executed in 0.00 seconds (0.000 gc):
throwException(exceptions.addRef());
| details | result |
|---|---|
| {"input":{ "LBFGS": { "type": "Converged", "value": 0.0 }, "CjGD": { "type": "Converged", "value": 0.0 }, "GD": { "type": "Converged", "value": 0.0 } }, "model":null, "complete":null} | OK |
{
"result": "OK",
"performance": {
"execution_time": "9.258",
"gc_time": "0.459"
},
"created_on": 1586736505038,
"file_name": "trainingTest",
"report": {
"simpleName": "LL",
"canonicalName": "com.simiacryptus.mindseye.layers.cudnn.ImgTileSelectLayerTest.LL",
"link": "https://github.com/SimiaCryptus/mindseye-cudnn/tree/59d5b3318556370acb2d83ee6ec123ce0fc6974f/src/test/java/com/simiacryptus/mindseye/layers/cudnn/ImgTileSelectLayerTest.java",
"javaDoc": ""
},
"training_analysis": {
"input": {
"LBFGS": {
"type": "Converged",
"value": 0.0
},
"CjGD": {
"type": "Converged",
"value": 0.0
},
"GD": {
"type": "Converged",
"value": 0.0
}
}
},
"archive": "s3://code.simiacrypt.us/tests/com/simiacryptus/mindseye/layers/cudnn/ImgTileSelectLayer/LL/trainingTest/202004130825",
"id": "7e21d5d1-9743-4eca-9e77-2ce2a06fb3a3",
"report_type": "Components",
"display_name": "Comparative Training",
"target": {
"simpleName": "ImgTileSelectLayer",
"canonicalName": "com.simiacryptus.mindseye.layers.cudnn.ImgTileSelectLayer",
"link": "https://github.com/SimiaCryptus/mindseye-cudnn/tree/59d5b3318556370acb2d83ee6ec123ce0fc6974f/src/main/java/com/simiacryptus/mindseye/layers/cudnn/ImgTileSelectLayer.java",
"javaDoc": ""
}
}