Execution Time1.01s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu18-clang91-opt (sft-ubuntu-1804-3) on 2019-11-14 00:51:22
Repository revision: 32b17abcda23e44b64218a42d0ca69cb30cda7e0

Test Timing: Passed
Processors1

Show Command Line
Display graphs:

Test output
Testing Backpropagation:
DEEP NEURAL NETWORK:   Depth = 3  Input = ( 1, 10, 4 )  Batch size = 10  Loss function = R
	Layer 0	 DENSE Layer: 	 ( Input =     4 , Width =     2 ) 	Output = (  1 ,    10 ,     2 ) 	 Activation Function = Identity
	Layer 1	 BATCH NORM Layer: 	 ( Input =     2 ) 
	Layer 2	 DENSE Layer: 	 ( Input =     2 , Width =     1 ) 	Output = (  1 ,    10 ,     1 ) 	 Activation Function = Identity
input 

10x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.9989     -0.4348      0.7818    -0.03005 
   1 |     0.8243    -0.05672     -0.9009     -0.0747 
   2 |   0.007912     -0.4108       1.391     -0.9851 
   3 |   -0.04894      -1.443      -1.061      -1.388 
   4 |     0.7674      -0.736      0.5797     -0.3821 
   5 |      2.061      -1.235       1.165     -0.4542 
   6 |    -0.1348     -0.4996     -0.1824       1.844 
   7 |    -0.2428       1.997    0.004806     -0.4222 
   8 |      1.541     0.09474       1.525       1.217 
   9 |    -0.1363     -0.1992     -0.2938     -0.1184 

 training batch 1 mu var0-0.11029
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.4192       1.311 
   1 |      1.901      -1.235 
   2 |     -2.172       1.174 
   3 |    -0.2277      -2.737 
   4 |    -0.7152      0.6731 
   5 |    -0.6579       1.715 
   6 |    -0.3954       1.376 
   7 |        1.8     -0.4965 
   8 |    -0.2395       3.583 
   9 |    0.02345     -0.5507 

output BN 
output DL feature 0 mean -0.11029	output DL std 1.19414
output DL feature 1 mean 0.481089	output DL std 1.77861
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.2727      0.4918 
   1 |      1.775      -1.017 
   2 |      -1.82      0.4104 
   3 |    -0.1036      -1.907 
   4 |    -0.5339      0.1138 
   5 |    -0.4833       0.731 
   6 |    -0.2517      0.5303 
   7 |      1.686     -0.5794 
   8 |    -0.1141       1.838 
   9 |      0.118     -0.6115 

output BN feature 0 mean 4.16334e-17	output BN std 1.05405
output BN feature 1 mean 1.11022e-17	output BN std 1.05407
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.1833     -0.1204     -0.2499      -0.269 
   1 |    -0.2213     -0.1835     -0.2746     -0.3382 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |      1.009        1.03      -1.253      0.0145 
   1 |     0.1768    -0.04206       1.462      0.8922 

 training batch 2 mu var0-0.110287
compute loss for weight  1.00941  1.0094 result 0.814778
 training batch 3 mu var0-0.11029
compute loss for weight  1.00939  1.0094 result 0.814782
 training batch 4 mu var0-0.110289
compute loss for weight  1.0094  1.0094 result 0.814779
 training batch 5 mu var0-0.11029
compute loss for weight  1.00939  1.0094 result 0.814781
   --dy = -0.183333 dy_ref = -0.183333
 training batch 6 mu var0-0.11029
compute loss for weight  1.03005  1.03004 result 0.814779
 training batch 7 mu var0-0.11029
compute loss for weight  1.03003  1.03004 result 0.814781
 training batch 8 mu var0-0.11029
compute loss for weight  1.03005  1.03004 result 0.814779
 training batch 9 mu var0-0.11029
compute loss for weight  1.03004  1.03004 result 0.81478
   --dy = -0.120386 dy_ref = -0.120386
 training batch 10 mu var0-0.110289
compute loss for weight  -1.25253  -1.25254 result 0.814777
 training batch 11 mu var0-0.11029
compute loss for weight  -1.25255  -1.25254 result 0.814782
 training batch 12 mu var0-0.11029
compute loss for weight  -1.25254  -1.25254 result 0.814779
 training batch 13 mu var0-0.11029
compute loss for weight  -1.25255  -1.25254 result 0.814781
   --dy = -0.249927 dy_ref = -0.249927
 training batch 14 mu var0-0.11029
compute loss for weight  0.0145109  0.0145009 result 0.814777
 training batch 15 mu var0-0.11029
compute loss for weight  0.0144909  0.0145009 result 0.814783
 training batch 16 mu var0-0.11029
compute loss for weight  0.0145059  0.0145009 result 0.814779
 training batch 17 mu var0-0.11029
compute loss for weight  0.0144959  0.0145009 result 0.814781
   --dy = -0.268981 dy_ref = -0.268981
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.094      0.5354 

weights for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |          1           1 

 training batch 18 mu var0-0.11029
compute loss for weight  1.00001  1 result 0.814791
 training batch 19 mu var0-0.11029
compute loss for weight  0.99999  1 result 0.814769
 training batch 20 mu var0-0.11029
compute loss for weight  1.00001  1 result 0.814785
 training batch 21 mu var0-0.11029
compute loss for weight  0.999995  1 result 0.814774
   --dy = 1.09419 dy_ref = 1.09419
 training batch 22 mu var0-0.11029
compute loss for weight  1.00001  1 result 0.814785
 training batch 23 mu var0-0.11029
compute loss for weight  0.99999  1 result 0.814775
 training batch 24 mu var0-0.11029
compute loss for weight  1.00001  1 result 0.814783
 training batch 25 mu var0-0.11029
compute loss for weight  0.999995  1 result 0.814777
   --dy = 0.535372 dy_ref = 0.535372
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  6.245e-17  -5.551e-17 

weights for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |          0           0 

 training batch 26 mu var0-0.11029
compute loss for weight  1e-05  0 result 0.81478
 training batch 27 mu var0-0.11029
compute loss for weight  -1e-05  0 result 0.81478
 training batch 28 mu var0-0.11029
compute loss for weight  5e-06  0 result 0.81478
 training batch 29 mu var0-0.11029
compute loss for weight  -5e-06  0 result 0.81478
   --dy = 1.4803e-11 dy_ref = 6.245e-17
 training batch 30 mu var0-0.11029
compute loss for weight  1e-05  0 result 0.81478
 training batch 31 mu var0-0.11029
compute loss for weight  -1e-05  0 result 0.81478
 training batch 32 mu var0-0.11029
compute loss for weight  5e-06  0 result 0.81478
 training batch 33 mu var0-0.11029
compute loss for weight  -5e-06  0 result 0.81478
   --dy = -2.96059e-11 dy_ref = -5.55112e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.659      -1.356 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.6597     -0.3947 

 training batch 34 mu var0-0.11029
compute loss for weight  0.659755  0.659745 result 0.814796
 training batch 35 mu var0-0.11029
compute loss for weight  0.659735  0.659745 result 0.814763
 training batch 36 mu var0-0.11029
compute loss for weight  0.65975  0.659745 result 0.814788
 training batch 37 mu var0-0.11029
compute loss for weight  0.65974  0.659745 result 0.814772
   --dy = 1.6585 dy_ref = 1.6585
 training batch 38 mu var0-0.11029
compute loss for weight  -0.394739  -0.394749 result 0.814766
 training batch 39 mu var0-0.11029
compute loss for weight  -0.394759  -0.394749 result 0.814793
 training batch 40 mu var0-0.11029
compute loss for weight  -0.394744  -0.394749 result 0.814773
 training batch 41 mu var0-0.11029
compute loss for weight  -0.394754  -0.394749 result 0.814787
   --dy = -1.35623 dy_ref = -1.35623
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m2.53188e-10[NON-XML-CHAR-0x1B][39m