Execution Time0.31s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: PR-4624-i686-ubuntu18-gcc7-opt (sft-ubuntu-1804-i386-2) on 2019-11-14 21:02:05
Repository revision: dfb00e27dc8e66d7a36bd3789ae6ae978a535876

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 var0-0.713571
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.411      0.8588 
   1 |    -0.1265     -0.6566 
   2 |    -0.8922       2.211 
   3 |      0.501       2.268 
   4 |      -1.13       1.447 
   5 |     -2.673       2.222 
   6 |   -0.02029     -0.8811 
   7 |     0.6588      -2.192 
   8 |     -2.307     -0.4427 
   9 |     0.2636      0.1903 

output BN 
output DL feature 0 mean -0.713571	output DL std 1.16461
output DL feature 1 mean 0.502318	output DL std 1.54625
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -0.631       0.243 
   1 |     0.5314       -0.79 
   2 |    -0.1617       1.165 
   3 |      1.099       1.203 
   4 |    -0.3766      0.6437 
   5 |     -1.773       1.172 
   6 |     0.6275     -0.9431 
   7 |      1.242      -1.837 
   8 |     -1.442     -0.6442 
   9 |     0.8844     -0.2127 

output BN feature 0 mean -7.77156e-17	output BN std 1.05405
output BN feature 1 mean 1.08247e-16	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 |    0.01952      0.1939      0.0179      0.1841 
   1 |    -0.1661       1.109     -0.1858      0.8781 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.8306      0.2144     -0.6268    -0.07561 
   1 |    -0.2399      -1.294      0.6556     -0.7809 

 training batch 2 mu var0-0.713568
compute loss for weight  -0.830594  -0.830604 result 1.55061
 training batch 3 mu var0-0.713571
compute loss for weight  -0.830614  -0.830604 result 1.55061
 training batch 4 mu var0-0.71357
compute loss for weight  -0.830599  -0.830604 result 1.55061
 training batch 5 mu var0-0.713571
compute loss for weight  -0.830609  -0.830604 result 1.55061
   --dy = 0.0195174 dy_ref = 0.0195174
 training batch 6 mu var0-0.713571
compute loss for weight  0.214448  0.214438 result 1.55061
 training batch 7 mu var0-0.713571
compute loss for weight  0.214428  0.214438 result 1.5506
 training batch 8 mu var0-0.713571
compute loss for weight  0.214443  0.214438 result 1.55061
 training batch 9 mu var0-0.713571
compute loss for weight  0.214433  0.214438 result 1.5506
   --dy = 0.193941 dy_ref = 0.193941
 training batch 10 mu var0-0.71357
compute loss for weight  -0.626813  -0.626823 result 1.55061
 training batch 11 mu var0-0.713571
compute loss for weight  -0.626833  -0.626823 result 1.55061
 training batch 12 mu var0-0.71357
compute loss for weight  -0.626818  -0.626823 result 1.55061
 training batch 13 mu var0-0.713571
compute loss for weight  -0.626828  -0.626823 result 1.55061
   --dy = 0.0178959 dy_ref = 0.0178959
 training batch 14 mu var0-0.713571
compute loss for weight  -0.0755974  -0.0756074 result 1.55061
 training batch 15 mu var0-0.713571
compute loss for weight  -0.0756174  -0.0756074 result 1.5506
 training batch 16 mu var0-0.713571
compute loss for weight  -0.0756024  -0.0756074 result 1.55061
 training batch 17 mu var0-0.713571
compute loss for weight  -0.0756124  -0.0756074 result 1.5506
   --dy = 0.184108 dy_ref = 0.184108
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      2.925      0.1763 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.713571
compute loss for weight  1.00001  1 result 1.55063
 training batch 19 mu var0-0.713571
compute loss for weight  0.99999  1 result 1.55058
 training batch 20 mu var0-0.713571
compute loss for weight  1.00001  1 result 1.55062
 training batch 21 mu var0-0.713571
compute loss for weight  0.999995  1 result 1.55059
   --dy = 2.92496 dy_ref = 2.92496
 training batch 22 mu var0-0.713571
compute loss for weight  1.00001  1 result 1.55061
 training batch 23 mu var0-0.713571
compute loss for weight  0.99999  1 result 1.5506
 training batch 24 mu var0-0.713571
compute loss for weight  1.00001  1 result 1.55061
 training batch 25 mu var0-0.713571
compute loss for weight  0.999995  1 result 1.5506
   --dy = 0.176251 dy_ref = 0.176251
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |          0   1.388e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.713571
compute loss for weight  1e-05  0 result 1.55061
 training batch 27 mu var0-0.713571
compute loss for weight  -1e-05  0 result 1.55061
 training batch 28 mu var0-0.713571
compute loss for weight  5e-06  0 result 1.55061
 training batch 29 mu var0-0.713571
compute loss for weight  -5e-06  0 result 1.55061
   --dy = 6.29126e-11 dy_ref = 0
 training batch 30 mu var0-0.713571
compute loss for weight  1e-05  0 result 1.55061
 training batch 31 mu var0-0.713571
compute loss for weight  -1e-05  0 result 1.55061
 training batch 32 mu var0-0.713571
compute loss for weight  5e-06  0 result 1.55061
 training batch 33 mu var0-0.713571
compute loss for weight  -5e-06  0 result 1.55061
   --dy = 2.96059e-11 dy_ref = 1.38778e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      2.476      -1.211 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.181     -0.1456 

 training batch 34 mu var0-0.713571
compute loss for weight  1.18139  1.18138 result 1.55063
 training batch 35 mu var0-0.713571
compute loss for weight  1.18137  1.18138 result 1.55058
 training batch 36 mu var0-0.713571
compute loss for weight  1.18139  1.18138 result 1.55062
 training batch 37 mu var0-0.713571
compute loss for weight  1.18138  1.18138 result 1.55059
   --dy = 2.47588 dy_ref = 2.47588
 training batch 38 mu var0-0.713571
compute loss for weight  -0.145571  -0.145581 result 1.55059
 training batch 39 mu var0-0.713571
compute loss for weight  -0.145591  -0.145581 result 1.55062
 training batch 40 mu var0-0.713571
compute loss for weight  -0.145576  -0.145581 result 1.5506
 training batch 41 mu var0-0.713571
compute loss for weight  -0.145586  -0.145581 result 1.55061
   --dy = -1.21067 dy_ref = -1.21067
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m3.44115e-09[NON-XML-CHAR-0x1B][39m