Execution Time0.05s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu18-gcc7-opt-exp-pyroot (sft-ubuntu-1804-3) on 2019-11-14 03:54:10
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 var00.0261082
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    0.07457      0.5687 
   1 |     0.6631     -0.6823 
   2 |     -1.943     -0.4934 
   3 |     -1.268      -1.842 
   4 |     -0.378      0.1216 
   5 |    -0.1397      0.5222 
   6 |      2.638       3.135 
   7 |    -0.9415       -1.93 
   8 |      1.616       2.547 
   9 |   -0.06064     -0.2084 

output BN 
output DL feature 0 mean 0.0261082	output DL std 1.35192
output DL feature 1 mean 0.173773	output DL std 1.64967
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    0.03778      0.2523 
   1 |     0.4967      -0.547 
   2 |     -1.535     -0.4263 
   3 |     -1.009      -1.288 
   4 |    -0.3151    -0.03333 
   5 |    -0.1293      0.2226 
   6 |      2.037       1.892 
   7 |    -0.7544      -1.344 
   8 |       1.24       1.516 
   9 |   -0.06764     -0.2442 

output BN feature 0 mean -7.07767e-17	output BN std 1.05406
output BN feature 1 mean 3.60822e-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 |     0.1098     -0.1601      0.3598     0.07146 
   1 |    -0.4493      0.3725     -0.8174      -1.082 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.4197     -0.1273     -0.4586       1.381 
   1 |    -0.1099     -0.6491      0.5673       1.572 

 training batch 2 mu var00.026111
compute loss for weight  0.419739  0.419729 result 2.284
 training batch 3 mu var00.0261082
compute loss for weight  0.419719  0.419729 result 2.284
 training batch 4 mu var00.0261089
compute loss for weight  0.419734  0.419729 result 2.284
 training batch 5 mu var00.0261082
compute loss for weight  0.419724  0.419729 result 2.284
   --dy = 0.109838 dy_ref = 0.109838
 training batch 6 mu var00.0261077
compute loss for weight  -0.127264  -0.127274 result 2.284
 training batch 7 mu var00.0261082
compute loss for weight  -0.127284  -0.127274 result 2.284
 training batch 8 mu var00.026108
compute loss for weight  -0.127269  -0.127274 result 2.284
 training batch 9 mu var00.0261082
compute loss for weight  -0.127279  -0.127274 result 2.284
   --dy = -0.160057 dy_ref = -0.160057
 training batch 10 mu var00.0261085
compute loss for weight  -0.458593  -0.458603 result 2.284
 training batch 11 mu var00.0261082
compute loss for weight  -0.458613  -0.458603 result 2.284
 training batch 12 mu var00.0261083
compute loss for weight  -0.458598  -0.458603 result 2.284
 training batch 13 mu var00.0261082
compute loss for weight  -0.458608  -0.458603 result 2.284
   --dy = 0.35983 dy_ref = 0.35983
 training batch 14 mu var00.0261081
compute loss for weight  1.38145  1.38144 result 2.284
 training batch 15 mu var00.0261082
compute loss for weight  1.38143  1.38144 result 2.284
 training batch 16 mu var00.0261081
compute loss for weight  1.38145  1.38144 result 2.284
 training batch 17 mu var00.0261082
compute loss for weight  1.38144  1.38144 result 2.284
   --dy = 0.071462 dy_ref = 0.071462
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      2.876       1.692 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.0261082
compute loss for weight  1.00001  1 result 2.28403
 training batch 19 mu var00.0261082
compute loss for weight  0.99999  1 result 2.28397
 training batch 20 mu var00.0261082
compute loss for weight  1.00001  1 result 2.28401
 training batch 21 mu var00.0261082
compute loss for weight  0.999995  1 result 2.28398
   --dy = 2.87623 dy_ref = 2.87623
 training batch 22 mu var00.0261082
compute loss for weight  1.00001  1 result 2.28402
 training batch 23 mu var00.0261082
compute loss for weight  0.99999  1 result 2.28398
 training batch 24 mu var00.0261082
compute loss for weight  1.00001  1 result 2.28401
 training batch 25 mu var00.0261082
compute loss for weight  0.999995  1 result 2.28399
   --dy = 1.69177 dy_ref = 1.69177
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  2.082e-17   -7.98e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.0261082
compute loss for weight  1e-05  0 result 2.284
 training batch 27 mu var00.0261082
compute loss for weight  -1e-05  0 result 2.284
 training batch 28 mu var00.0261082
compute loss for weight  5e-06  0 result 2.284
 training batch 29 mu var00.0261082
compute loss for weight  -5e-06  0 result 2.284
   --dy = 0 dy_ref = 2.08167e-17
 training batch 30 mu var00.0261082
compute loss for weight  1e-05  0 result 2.284
 training batch 31 mu var00.0261082
compute loss for weight  -1e-05  0 result 2.284
 training batch 32 mu var00.0261082
compute loss for weight  5e-06  0 result 2.284
 training batch 33 mu var00.0261082
compute loss for weight  -5e-06  0 result 2.284
   --dy = 1.18424e-10 dy_ref = -7.97973e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -2.954      -2.836 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.9737     -0.5965 

 training batch 34 mu var00.0261082
compute loss for weight  -0.973686  -0.973696 result 2.28397
 training batch 35 mu var00.0261082
compute loss for weight  -0.973706  -0.973696 result 2.28403
 training batch 36 mu var00.0261082
compute loss for weight  -0.973691  -0.973696 result 2.28398
 training batch 37 mu var00.0261082
compute loss for weight  -0.973701  -0.973696 result 2.28401
   --dy = -2.95393 dy_ref = -2.95393
 training batch 38 mu var00.0261082
compute loss for weight  -0.596491  -0.596501 result 2.28397
 training batch 39 mu var00.0261082
compute loss for weight  -0.596511  -0.596501 result 2.28403
 training batch 40 mu var00.0261082
compute loss for weight  -0.596496  -0.596501 result 2.28398
 training batch 41 mu var00.0261082
compute loss for weight  -0.596506  -0.596501 result 2.28401
   --dy = -2.83616 dy_ref = -2.83616
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m1.7237e-09[NON-XML-CHAR-0x1B][39m