Execution Time0.04s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu14-gcc48 (ec-ubuntu-14-04-x86-64-2) on 2019-11-14 00:56:17
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 var0-0.413823
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.5741     -0.5839 
   1 |    -0.7632     0.05719 
   2 |     -0.192      -0.904 
   3 |     -1.608      -1.173 
   4 |    -0.8071     -0.8301 
   5 |     -1.648      -1.457 
   6 |     0.3912      0.2947 
   7 |      1.132       1.249 
   8 |    0.09913    -0.05525 
   9 |    -0.1684    -0.09238 

output BN 
output DL feature 0 mean -0.413823	output DL std 0.860427
output DL feature 1 mean -0.349446	output DL std 0.801858
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.1964     -0.3082 
   1 |     -0.428      0.5345 
   2 |     0.2718      -0.729 
   3 |     -1.462      -1.083 
   4 |    -0.4818     -0.6318 
   5 |     -1.512      -1.455 
   6 |     0.9862      0.8467 
   7 |      1.894       2.101 
   8 |     0.6284      0.3867 
   9 |     0.3006      0.3379 

output BN feature 0 mean -6.10623e-17	output BN std 1.05401
output BN feature 1 mean 1.66533e-17	output BN std 1.054
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |   -0.01329      0.1019     -0.2568      0.0231 
   1 |     0.2796     -0.7449        0.49     -0.3554 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.5388      0.5764      0.2883      0.3574 
   1 |    -0.1023      0.6801     -0.2259      0.3142 

 training batch 2 mu var0-0.41382
compute loss for weight  -0.538765  -0.538775 result 3.23735
 training batch 3 mu var0-0.413823
compute loss for weight  -0.538785  -0.538775 result 3.23735
 training batch 4 mu var0-0.413822
compute loss for weight  -0.53877  -0.538775 result 3.23735
 training batch 5 mu var0-0.413823
compute loss for weight  -0.53878  -0.538775 result 3.23735
   --dy = -0.0132918 dy_ref = -0.0132918
 training batch 6 mu var0-0.413824
compute loss for weight  0.576403  0.576393 result 3.23736
 training batch 7 mu var0-0.413823
compute loss for weight  0.576383  0.576393 result 3.23735
 training batch 8 mu var0-0.413823
compute loss for weight  0.576398  0.576393 result 3.23735
 training batch 9 mu var0-0.413823
compute loss for weight  0.576388  0.576393 result 3.23735
   --dy = 0.101899 dy_ref = 0.101899
 training batch 10 mu var0-0.413823
compute loss for weight  0.288346  0.288336 result 3.23735
 training batch 11 mu var0-0.413823
compute loss for weight  0.288326  0.288336 result 3.23736
 training batch 12 mu var0-0.413823
compute loss for weight  0.288341  0.288336 result 3.23735
 training batch 13 mu var0-0.413823
compute loss for weight  0.288331  0.288336 result 3.23736
   --dy = -0.256807 dy_ref = -0.256807
 training batch 14 mu var0-0.413823
compute loss for weight  0.35743  0.35742 result 3.23735
 training batch 15 mu var0-0.413823
compute loss for weight  0.35741  0.35742 result 3.23735
 training batch 16 mu var0-0.413823
compute loss for weight  0.357425  0.35742 result 3.23735
 training batch 17 mu var0-0.413823
compute loss for weight  0.357415  0.35742 result 3.23735
   --dy = 0.0231044 dy_ref = 0.0231044
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.7071       5.768 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.413823
compute loss for weight  1.00001  1 result 3.23736
 training batch 19 mu var0-0.413823
compute loss for weight  0.99999  1 result 3.23735
 training batch 20 mu var0-0.413823
compute loss for weight  1.00001  1 result 3.23736
 training batch 21 mu var0-0.413823
compute loss for weight  0.999995  1 result 3.23735
   --dy = 0.707077 dy_ref = 0.707077
 training batch 22 mu var0-0.413823
compute loss for weight  1.00001  1 result 3.23741
 training batch 23 mu var0-0.413823
compute loss for weight  0.99999  1 result 3.2373
 training batch 24 mu var0-0.413823
compute loss for weight  1.00001  1 result 3.23738
 training batch 25 mu var0-0.413823
compute loss for weight  0.999995  1 result 3.23733
   --dy = 5.76763 dy_ref = 5.76763
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -3.469e-17    2.22e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.413823
compute loss for weight  1e-05  0 result 3.23735
 training batch 27 mu var0-0.413823
compute loss for weight  -1e-05  0 result 3.23735
 training batch 28 mu var0-0.413823
compute loss for weight  5e-06  0 result 3.23735
 training batch 29 mu var0-0.413823
compute loss for weight  -5e-06  0 result 3.23735
   --dy = 1.4803e-11 dy_ref = -3.46945e-17
 training batch 30 mu var0-0.413823
compute loss for weight  1e-05  0 result 3.23735
 training batch 31 mu var0-0.413823
compute loss for weight  -1e-05  0 result 3.23735
 training batch 32 mu var0-0.413823
compute loss for weight  5e-06  0 result 3.23735
 training batch 33 mu var0-0.413823
compute loss for weight  -5e-06  0 result 3.23735
   --dy = -5.92119e-11 dy_ref = 2.22045e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -3.283      -3.593 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.2154      -1.605 

 training batch 34 mu var0-0.413823
compute loss for weight  -0.215397  -0.215407 result 3.23732
 training batch 35 mu var0-0.413823
compute loss for weight  -0.215417  -0.215407 result 3.23739
 training batch 36 mu var0-0.413823
compute loss for weight  -0.215402  -0.215407 result 3.23734
 training batch 37 mu var0-0.413823
compute loss for weight  -0.215412  -0.215407 result 3.23737
   --dy = -3.28252 dy_ref = -3.28252
 training batch 38 mu var0-0.413823
compute loss for weight  -1.60533  -1.60534 result 3.23732
 training batch 39 mu var0-0.413823
compute loss for weight  -1.60535  -1.60534 result 3.23739
 training batch 40 mu var0-0.413823
compute loss for weight  -1.60533  -1.60534 result 3.23734
 training batch 41 mu var0-0.413823
compute loss for weight  -1.60534  -1.60534 result 3.23737
   --dy = -3.59278 dy_ref = -3.59278
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m1.81976e-08[NON-XML-CHAR-0x1B][39m