Execution Time0.05s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: PR-4623-x86_64-ubuntu16-gcc54-opt (sft-ubuntu-1604-4) on 2019-11-14 10:13:55
Repository revision: 23cd6774db1f8460248730c7b87c10dd00181367

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 var00.0428542
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.6387      0.9485 
   1 |     -1.591     -0.9862 
   2 |     0.7787       1.141 
   3 |     -4.298       -1.15 
   4 |    -0.2637      0.7072 
   5 |     -0.058       1.446 
   6 |      1.057      0.9549 
   7 |       1.49      -1.082 
   8 |      3.407       2.082 
   9 |    -0.7315     -0.2701 

output BN 
output DL feature 0 mean 0.0428542	output DL std 2.04262
output DL feature 1 mean 0.379141	output DL std 1.16163
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.3075      0.5166 
   1 |    -0.8431      -1.239 
   2 |     0.3797      0.6916 
   3 |      -2.24      -1.388 
   4 |    -0.1582      0.2977 
   5 |   -0.05205      0.9681 
   6 |     0.5231      0.5224 
   7 |      0.747      -1.326 
   8 |      1.736       1.546 
   9 |    -0.3996     -0.5891 

output BN feature 0 mean -2.77556e-17	output BN std 1.05408
output BN feature 1 mean 6.66134e-17	output BN std 1.05405
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.2281      0.4006     -0.2427    -0.03201 
   1 |    -0.9954     -0.1708      -2.083      -1.419 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.1158      0.9332       1.521      0.9677 
   1 |   -0.04402     -0.4447       1.041       0.497 

 training batch 2 mu var00.042857
compute loss for weight  -0.115835  -0.115845 result 0.705786
 training batch 3 mu var00.0428542
compute loss for weight  -0.115855  -0.115845 result 0.705791
 training batch 4 mu var00.0428549
compute loss for weight  -0.11584  -0.115845 result 0.705787
 training batch 5 mu var00.0428542
compute loss for weight  -0.11585  -0.115845 result 0.705789
   --dy = -0.228054 dy_ref = -0.228054
 training batch 6 mu var00.0428537
compute loss for weight  0.933201  0.933191 result 0.705792
 training batch 7 mu var00.0428542
compute loss for weight  0.933181  0.933191 result 0.705784
 training batch 8 mu var00.042854
compute loss for weight  0.933196  0.933191 result 0.70579
 training batch 9 mu var00.0428542
compute loss for weight  0.933186  0.933191 result 0.705786
   --dy = 0.400569 dy_ref = 0.400569
 training batch 10 mu var00.0428545
compute loss for weight  1.52111  1.5211 result 0.705786
 training batch 11 mu var00.0428542
compute loss for weight  1.52109  1.5211 result 0.705791
 training batch 12 mu var00.0428543
compute loss for weight  1.5211  1.5211 result 0.705787
 training batch 13 mu var00.0428542
compute loss for weight  1.52109  1.5211 result 0.70579
   --dy = -0.242729 dy_ref = -0.242729
 training batch 14 mu var00.0428541
compute loss for weight  0.967691  0.967681 result 0.705788
 training batch 15 mu var00.0428542
compute loss for weight  0.967671  0.967681 result 0.705789
 training batch 16 mu var00.0428541
compute loss for weight  0.967686  0.967681 result 0.705788
 training batch 17 mu var00.0428542
compute loss for weight  0.967676  0.967681 result 0.705788
   --dy = -0.0320052 dy_ref = -0.0320052
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.566     -0.1547 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.0428542
compute loss for weight  1.00001  1 result 0.705804
 training batch 19 mu var00.0428542
compute loss for weight  0.99999  1 result 0.705773
 training batch 20 mu var00.0428542
compute loss for weight  1.00001  1 result 0.705796
 training batch 21 mu var00.0428542
compute loss for weight  0.999995  1 result 0.70578
   --dy = 1.56629 dy_ref = 1.56629
 training batch 22 mu var00.0428542
compute loss for weight  1.00001  1 result 0.705787
 training batch 23 mu var00.0428542
compute loss for weight  0.99999  1 result 0.70579
 training batch 24 mu var00.0428542
compute loss for weight  1.00001  1 result 0.705788
 training batch 25 mu var00.0428542
compute loss for weight  0.999995  1 result 0.705789
   --dy = -0.154708 dy_ref = -0.154708
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -1.145e-16    7.98e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.0428542
compute loss for weight  1e-05  0 result 0.705788
 training batch 27 mu var00.0428542
compute loss for weight  -1e-05  0 result 0.705788
 training batch 28 mu var00.0428542
compute loss for weight  5e-06  0 result 0.705788
 training batch 29 mu var00.0428542
compute loss for weight  -5e-06  0 result 0.705788
   --dy = 2.96059e-11 dy_ref = -1.14492e-16
 training batch 30 mu var00.0428542
compute loss for weight  1e-05  0 result 0.705788
 training batch 31 mu var00.0428542
compute loss for weight  -1e-05  0 result 0.705788
 training batch 32 mu var00.0428542
compute loss for weight  5e-06  0 result 0.705788
 training batch 33 mu var00.0428542
compute loss for weight  -5e-06  0 result 0.705788
   --dy = -1.29526e-11 dy_ref = 7.97973e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.403      0.2491 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.116      -0.621 

 training batch 34 mu var00.0428542
compute loss for weight  1.11634  1.11633 result 0.705802
 training batch 35 mu var00.0428542
compute loss for weight  1.11632  1.11633 result 0.705774
 training batch 36 mu var00.0428542
compute loss for weight  1.11634  1.11633 result 0.705795
 training batch 37 mu var00.0428542
compute loss for weight  1.11633  1.11633 result 0.705781
   --dy = 1.40306 dy_ref = 1.40306
 training batch 38 mu var00.0428542
compute loss for weight  -0.621037  -0.621047 result 0.705791
 training batch 39 mu var00.0428542
compute loss for weight  -0.621057  -0.621047 result 0.705786
 training batch 40 mu var00.0428542
compute loss for weight  -0.621042  -0.621047 result 0.70579
 training batch 41 mu var00.0428542
compute loss for weight  -0.621052  -0.621047 result 0.705787
   --dy = 0.249109 dy_ref = 0.249109
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m8.38326e-10[NON-XML-CHAR-0x1B][39m