Execution Time1.40s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu18-clang91-dbg (sft-ubuntu-1804-3) on 2019-11-14 10:13:30
Repository revision: 039d38b9f83f4fc0b537c106b114d9ef73dbfd81

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 var0-0.192229
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.3797       1.021 
   1 |    0.06767     -0.6263 
   2 |    -0.3587     -0.5274 
   3 |  -0.008739      -3.519 
   4 |    -0.3544      0.1061 
   5 |    -0.7441      0.9874 
   6 |   -0.04742       3.071 
   7 |     0.3866     -0.7689 
   8 |    -0.5307       4.157 
   9 |    0.04725     -0.5374 

output BN 
output DL feature 0 mean -0.192229	output DL std 0.33713
output DL feature 1 mean 0.336272	output DL std 2.14841
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.5859       0.336 
   1 |     0.8122     -0.4723 
   2 |    -0.5204     -0.4237 
   3 |     0.5734      -1.892 
   4 |    -0.5069     -0.1129 
   5 |     -1.725      0.3195 
   6 |     0.4525       1.342 
   7 |      1.809     -0.5422 
   8 |     -1.058       1.874 
   9 |     0.7484     -0.4286 

output BN feature 0 mean 2.22045e-17	output BN std 1.05358
output BN feature 1 mean 3.33067e-17	output BN std 1.05408
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.2372       1.193      0.7152         2.7 
   1 |     -0.243      0.3146     -0.2323       0.179 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.1382      0.1754     -0.2119   -0.009267 
   1 |     0.3889     0.03918      0.8999       1.793 

 training batch 2 mu var0-0.192226
compute loss for weight  -0.138232  -0.138242 result 1.5594
 training batch 3 mu var0-0.192229
compute loss for weight  -0.138252  -0.138242 result 1.5594
 training batch 4 mu var0-0.192228
compute loss for weight  -0.138237  -0.138242 result 1.5594
 training batch 5 mu var0-0.192229
compute loss for weight  -0.138247  -0.138242 result 1.5594
   --dy = 0.237188 dy_ref = 0.237188
 training batch 6 mu var0-0.192229
compute loss for weight  0.175364  0.175354 result 1.55941
 training batch 7 mu var0-0.192229
compute loss for weight  0.175344  0.175354 result 1.55939
 training batch 8 mu var0-0.192229
compute loss for weight  0.175359  0.175354 result 1.55941
 training batch 9 mu var0-0.192229
compute loss for weight  0.175349  0.175354 result 1.55939
   --dy = 1.19273 dy_ref = 1.19273
 training batch 10 mu var0-0.192228
compute loss for weight  -0.211859  -0.211869 result 1.55941
 training batch 11 mu var0-0.192229
compute loss for weight  -0.211879  -0.211869 result 1.55939
 training batch 12 mu var0-0.192229
compute loss for weight  -0.211864  -0.211869 result 1.5594
 training batch 13 mu var0-0.192229
compute loss for weight  -0.211874  -0.211869 result 1.5594
   --dy = 0.715167 dy_ref = 0.715167
 training batch 14 mu var0-0.192229
compute loss for weight  -0.00925707  -0.00926707 result 1.55943
 training batch 15 mu var0-0.192229
compute loss for weight  -0.00927707  -0.00926707 result 1.55937
 training batch 16 mu var0-0.192229
compute loss for weight  -0.00926207  -0.00926707 result 1.55941
 training batch 17 mu var0-0.192229
compute loss for weight  -0.00927207  -0.00926707 result 1.55939
   --dy = 2.70017 dy_ref = 2.70017
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.1891       3.308 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.192229
compute loss for weight  1.00001  1 result 1.5594
 training batch 19 mu var0-0.192229
compute loss for weight  0.99999  1 result 1.5594
 training batch 20 mu var0-0.192229
compute loss for weight  1.00001  1 result 1.5594
 training batch 21 mu var0-0.192229
compute loss for weight  0.999995  1 result 1.5594
   --dy = -0.189103 dy_ref = -0.189103
 training batch 22 mu var0-0.192229
compute loss for weight  1.00001  1 result 1.55943
 training batch 23 mu var0-0.192229
compute loss for weight  0.99999  1 result 1.55937
 training batch 24 mu var0-0.192229
compute loss for weight  1.00001  1 result 1.55942
 training batch 25 mu var0-0.192229
compute loss for weight  0.999995  1 result 1.55938
   --dy = 3.3079 dy_ref = 3.3079
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -4.163e-17   2.776e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.192229
compute loss for weight  1e-05  0 result 1.5594
 training batch 27 mu var0-0.192229
compute loss for weight  -1e-05  0 result 1.5594
 training batch 28 mu var0-0.192229
compute loss for weight  5e-06  0 result 1.5594
 training batch 29 mu var0-0.192229
compute loss for weight  -5e-06  0 result 1.5594
   --dy = 5.55112e-11 dy_ref = -4.16334e-17
 training batch 30 mu var0-0.192229
compute loss for weight  1e-05  0 result 1.5594
 training batch 31 mu var0-0.192229
compute loss for weight  -1e-05  0 result 1.5594
 training batch 32 mu var0-0.192229
compute loss for weight  5e-06  0 result 1.5594
 training batch 33 mu var0-0.192229
compute loss for weight  -5e-06  0 result 1.5594
   --dy = 0 dy_ref = 2.77556e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.4661       2.392 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.4057       1.383 

 training batch 34 mu var0-0.192229
compute loss for weight  0.405683  0.405673 result 1.5594
 training batch 35 mu var0-0.192229
compute loss for weight  0.405663  0.405673 result 1.5594
 training batch 36 mu var0-0.192229
compute loss for weight  0.405678  0.405673 result 1.5594
 training batch 37 mu var0-0.192229
compute loss for weight  0.405668  0.405673 result 1.5594
   --dy = -0.466146 dy_ref = -0.466146
 training batch 38 mu var0-0.192229
compute loss for weight  1.38311  1.3831 result 1.55942
 training batch 39 mu var0-0.192229
compute loss for weight  1.38309  1.3831 result 1.55938
 training batch 40 mu var0-0.192229
compute loss for weight  1.3831  1.3831 result 1.55941
 training batch 41 mu var0-0.192229
compute loss for weight  1.38309  1.3831 result 1.55939
   --dy = 2.39166 dy_ref = 2.39166
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m2.89042e-10[NON-XML-CHAR-0x1B][39m