Execution Time0.89s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-mac1014-clang100-dbg (macitois21.cern.ch) on 2019-11-15 20:01:40

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 var00.0771013
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.2016    0.002651 
   1 |     0.9135      0.5483 
   2 |     0.3865      0.5306 
   3 |      2.882       2.552 
   4 |     0.4213      0.5691 
   5 |     0.4879      0.7519 
   6 |      -2.14      -1.331 
   7 |   -0.01063     -0.7725 
   8 |     -2.349      -1.727 
   9 |     0.3814      0.3407 

output BN 
output DL feature 0 mean 0.0771013	output DL std 1.48798
output DL feature 1 mean 0.146475	output DL std 1.21284
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.1974      -0.125 
   1 |     0.5925      0.3492 
   2 |     0.2192      0.3338 
   3 |      1.987       2.091 
   4 |     0.2438      0.3673 
   5 |      0.291      0.5262 
   6 |     -1.571      -1.284 
   7 |   -0.06215     -0.7986 
   8 |     -1.718      -1.628 
   9 |     0.2156      0.1688 

output BN feature 0 mean -8.60423e-17	output BN std 1.05407
output BN feature 1 mean -2.498e-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 |  -0.002407    0.008194   -0.002131  -0.0008851 
   1 |     0.1604       1.099      0.7826       1.643 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.2032     -0.2502     -0.7059      -1.283 
   1 |    0.06727      -0.571     -0.4353     -0.9147 

 training batch 2 mu var00.0771041
compute loss for weight  0.203161  0.203151 result 0.979328
 training batch 3 mu var00.0771013
compute loss for weight  0.203141  0.203151 result 0.979328
 training batch 4 mu var00.077102
compute loss for weight  0.203156  0.203151 result 0.979328
 training batch 5 mu var00.0771013
compute loss for weight  0.203146  0.203151 result 0.979328
   --dy = -0.00240704 dy_ref = -0.00240704
 training batch 6 mu var00.0771008
compute loss for weight  -0.250231  -0.250241 result 0.979328
 training batch 7 mu var00.0771013
compute loss for weight  -0.250251  -0.250241 result 0.979328
 training batch 8 mu var00.0771011
compute loss for weight  -0.250236  -0.250241 result 0.979328
 training batch 9 mu var00.0771013
compute loss for weight  -0.250246  -0.250241 result 0.979328
   --dy = 0.00819444 dy_ref = 0.00819444
 training batch 10 mu var00.0771016
compute loss for weight  -0.705904  -0.705914 result 0.979328
 training batch 11 mu var00.0771013
compute loss for weight  -0.705924  -0.705914 result 0.979328
 training batch 12 mu var00.0771015
compute loss for weight  -0.705909  -0.705914 result 0.979328
 training batch 13 mu var00.0771013
compute loss for weight  -0.705919  -0.705914 result 0.979328
   --dy = -0.0021311 dy_ref = -0.0021311
 training batch 14 mu var00.0771013
compute loss for weight  -1.28331  -1.28332 result 0.979328
 training batch 15 mu var00.0771013
compute loss for weight  -1.28333  -1.28332 result 0.979328
 training batch 16 mu var00.0771013
compute loss for weight  -1.28332  -1.28332 result 0.979328
 training batch 17 mu var00.0771013
compute loss for weight  -1.28333  -1.28332 result 0.979328
   --dy = -0.000885085 dy_ref = -0.000885085
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      2.006    -0.04696 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.0771013
compute loss for weight  1.00001  1 result 0.979348
 training batch 19 mu var00.0771013
compute loss for weight  0.99999  1 result 0.979308
 training batch 20 mu var00.0771013
compute loss for weight  1.00001  1 result 0.979338
 training batch 21 mu var00.0771013
compute loss for weight  0.999995  1 result 0.979318
   --dy = 2.00562 dy_ref = 2.00562
 training batch 22 mu var00.0771013
compute loss for weight  1.00001  1 result 0.979327
 training batch 23 mu var00.0771013
compute loss for weight  0.99999  1 result 0.979328
 training batch 24 mu var00.0771013
compute loss for weight  1.00001  1 result 0.979327
 training batch 25 mu var00.0771013
compute loss for weight  0.999995  1 result 0.979328
   --dy = -0.0469619 dy_ref = -0.0469619
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -1.874e-16   2.385e-18 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.0771013
compute loss for weight  1e-05  0 result 0.979328
 training batch 27 mu var00.0771013
compute loss for weight  -1e-05  0 result 0.979328
 training batch 28 mu var00.0771013
compute loss for weight  5e-06  0 result 0.979328
 training batch 29 mu var00.0771013
compute loss for weight  -5e-06  0 result 0.979328
   --dy = 4.62593e-11 dy_ref = -1.8735e-16
 training batch 30 mu var00.0771013
compute loss for weight  1e-05  0 result 0.979328
 training batch 31 mu var00.0771013
compute loss for weight  -1e-05  0 result 0.979328
 training batch 32 mu var00.0771013
compute loss for weight  5e-06  0 result 0.979328
 training batch 33 mu var00.0771013
compute loss for weight  -5e-06  0 result 0.979328
   --dy = 1.4803e-11 dy_ref = 2.38524e-18
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.979      -1.897 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.013     0.02476 

 training batch 34 mu var00.0771013
compute loss for weight  -1.01338  -1.01339 result 0.979308
 training batch 35 mu var00.0771013
compute loss for weight  -1.0134  -1.01339 result 0.979347
 training batch 36 mu var00.0771013
compute loss for weight  -1.01338  -1.01339 result 0.979318
 training batch 37 mu var00.0771013
compute loss for weight  -1.01339  -1.01339 result 0.979338
   --dy = -1.97912 dy_ref = -1.97912
 training batch 38 mu var00.0771013
compute loss for weight  0.02477  0.02476 result 0.979309
 training batch 39 mu var00.0771013
compute loss for weight  0.02475  0.02476 result 0.979347
 training batch 40 mu var00.0771013
compute loss for weight  0.024765  0.02476 result 0.979318
 training batch 41 mu var00.0771013
compute loss for weight  0.024755  0.02476 result 0.979337
   --dy = -1.89669 dy_ref = -1.89669
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m4.15787e-08[NON-XML-CHAR-0x1B][39m