Execution Time0.16s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-centos7-gcc62-opt-no-rt-cxxmodules (olhswep09.cern.ch) on 2019-11-13 01:03:03
Repository revision: 30660dce2d9e89e4852dbf83dbd8b2cfcc137eff

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 var0-0.58917
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.8644     -0.4759 
   1 |     -0.914      0.1481 
   2 |   -0.04545     -0.6058 
   3 |     -1.955      -0.321 
   4 |     -1.124     -0.5314 
   5 |      -2.32      -1.007 
   6 |    -0.3716   3.469e-05 
   7 |      2.254      0.7301 
   8 |    -0.3234     -0.4589 
   9 |    -0.2281     0.01404 

output BN 
output DL feature 0 mean -0.58917	output DL std 1.24532
output DL feature 1 mean -0.250738	output DL std 0.486421
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -0.233     -0.4879 
   1 |    -0.2749       0.864 
   2 |     0.4602     -0.7693 
   3 |     -1.156     -0.1521 
   4 |    -0.4524      -0.608 
   5 |     -1.465      -1.638 
   6 |     0.1842      0.5433 
   7 |      2.407       2.125 
   8 |     0.2249     -0.4509 
   9 |     0.3056      0.5736 

output BN feature 0 mean -7.77156e-17	output BN std 1.05405
output BN feature 1 mean 9.99201e-17	output BN std 1.05385
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |      0.301    -0.08706      0.8779    -0.02818 
   1 |      4.169      -6.969        2.55      -1.344 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.6631       1.062      0.3351     0.07104 
   1 |   -0.09501      0.3685     -0.2799     0.06523 

 training batch 2 mu var0-0.589167
compute loss for weight  -0.663043  -0.663053 result 0.625714
 training batch 3 mu var0-0.58917
compute loss for weight  -0.663063  -0.663053 result 0.625708
 training batch 4 mu var0-0.589169
compute loss for weight  -0.663048  -0.663053 result 0.625713
 training batch 5 mu var0-0.58917
compute loss for weight  -0.663058  -0.663053 result 0.62571
   --dy = 0.301005 dy_ref = 0.301005
 training batch 6 mu var0-0.58917
compute loss for weight  1.06244  1.06243 result 0.62571
 training batch 7 mu var0-0.58917
compute loss for weight  1.06242  1.06243 result 0.625712
 training batch 8 mu var0-0.58917
compute loss for weight  1.06243  1.06243 result 0.625711
 training batch 9 mu var0-0.58917
compute loss for weight  1.06242  1.06243 result 0.625712
   --dy = -0.0870557 dy_ref = -0.0870557
 training batch 10 mu var0-0.589169
compute loss for weight  0.335109  0.335099 result 0.62572
 training batch 11 mu var0-0.58917
compute loss for weight  0.335089  0.335099 result 0.625702
 training batch 12 mu var0-0.58917
compute loss for weight  0.335104  0.335099 result 0.625715
 training batch 13 mu var0-0.58917
compute loss for weight  0.335094  0.335099 result 0.625707
   --dy = 0.87789 dy_ref = 0.87789
 training batch 14 mu var0-0.58917
compute loss for weight  0.0710536  0.0710436 result 0.625711
 training batch 15 mu var0-0.58917
compute loss for weight  0.0710336  0.0710436 result 0.625711
 training batch 16 mu var0-0.58917
compute loss for weight  0.0710486  0.0710436 result 0.625711
 training batch 17 mu var0-0.58917
compute loss for weight  0.0710386  0.0710436 result 0.625711
   --dy = -0.0281787 dy_ref = -0.0281787
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.472     -0.2205 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.58917
compute loss for weight  1.00001  1 result 0.625726
 training batch 19 mu var0-0.58917
compute loss for weight  0.99999  1 result 0.625696
 training batch 20 mu var0-0.58917
compute loss for weight  1.00001  1 result 0.625718
 training batch 21 mu var0-0.58917
compute loss for weight  0.999995  1 result 0.625704
   --dy = 1.4719 dy_ref = 1.4719
 training batch 22 mu var0-0.58917
compute loss for weight  1.00001  1 result 0.625709
 training batch 23 mu var0-0.58917
compute loss for weight  0.99999  1 result 0.625713
 training batch 24 mu var0-0.58917
compute loss for weight  1.00001  1 result 0.62571
 training batch 25 mu var0-0.58917
compute loss for weight  0.999995  1 result 0.625712
   --dy = -0.220474 dy_ref = -0.220474
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -1.318e-16   6.765e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.58917
compute loss for weight  1e-05  0 result 0.625711
 training batch 27 mu var0-0.58917
compute loss for weight  -1e-05  0 result 0.625711
 training batch 28 mu var0-0.58917
compute loss for weight  5e-06  0 result 0.625711
 training batch 29 mu var0-0.58917
compute loss for weight  -5e-06  0 result 0.625711
   --dy = -1.66533e-11 dy_ref = -1.31839e-16
 training batch 30 mu var0-0.58917
compute loss for weight  1e-05  0 result 0.625711
 training batch 31 mu var0-0.58917
compute loss for weight  -1e-05  0 result 0.625711
 training batch 32 mu var0-0.58917
compute loss for weight  5e-06  0 result 0.625711
 training batch 33 mu var0-0.58917
compute loss for weight  -5e-06  0 result 0.625711
   --dy = 0 dy_ref = 6.76542e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.215     -0.2794 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.212      0.7891 

 training batch 34 mu var0-0.58917
compute loss for weight  -1.21188  -1.21189 result 0.625699
 training batch 35 mu var0-0.58917
compute loss for weight  -1.2119  -1.21189 result 0.625723
 training batch 36 mu var0-0.58917
compute loss for weight  -1.21188  -1.21189 result 0.625705
 training batch 37 mu var0-0.58917
compute loss for weight  -1.21189  -1.21189 result 0.625717
   --dy = -1.21455 dy_ref = -1.21455
 training batch 38 mu var0-0.58917
compute loss for weight  0.789107  0.789097 result 0.625708
 training batch 39 mu var0-0.58917
compute loss for weight  0.789087  0.789097 result 0.625714
 training batch 40 mu var0-0.58917
compute loss for weight  0.789102  0.789097 result 0.62571
 training batch 41 mu var0-0.58917
compute loss for weight  0.789092  0.789097 result 0.625712
   --dy = -0.2794 dy_ref = -0.2794
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m9.68665e-10[NON-XML-CHAR-0x1B][39m