Execution Time0.06s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-fedora27-gcc7 (sft-fedora-27-2.cern.ch) on 2019-11-14 01:14:54
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 var00.26644
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    0.08998     -0.8351 
   1 |      1.305      0.7056 
   2 |    -0.7269     -0.2413 
   3 |      3.149       1.914 
   4 |     0.6619      -0.363 
   5 |      1.146      -1.045 
   6 |      0.394      -1.848 
   7 |     -2.488      0.9749 
   8 |     -1.394      -2.638 
   9 |     0.5266      0.3246 

output BN 
output DL feature 0 mean 0.26644	output DL std 1.55483
output DL feature 1 mean -0.305126	output DL std 1.35958
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.1196     -0.4109 
   1 |     0.7043      0.7836 
   2 |    -0.6734     0.04947 
   3 |      1.954        1.72 
   4 |     0.2681    -0.04487 
   5 |     0.5962     -0.5734 
   6 |    0.08649      -1.196 
   7 |     -1.867      0.9924 
   8 |     -1.125      -1.809 
   9 |     0.1764      0.4882 

output BN feature 0 mean 1.66533e-17	output BN std 1.05407
output BN feature 1 mean 3.88578e-17	output BN std 1.05406
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.1402     -0.1157      0.1478      0.2014 
   1 |      1.636   0.0007072       2.728       3.046 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.3514      -1.243      -1.033     -0.1995 
   1 |    -0.1129      0.2607     -0.8183       -1.02 

 training batch 2 mu var00.266443
compute loss for weight  0.351414  0.351404 result 2.34714
 training batch 3 mu var00.26644
compute loss for weight  0.351394  0.351404 result 2.34714
 training batch 4 mu var00.266441
compute loss for weight  0.351409  0.351404 result 2.34714
 training batch 5 mu var00.26644
compute loss for weight  0.351399  0.351404 result 2.34714
   --dy = 0.140235 dy_ref = 0.140235
 training batch 6 mu var00.26644
compute loss for weight  -1.24266  -1.24267 result 2.34714
 training batch 7 mu var00.26644
compute loss for weight  -1.24268  -1.24267 result 2.34714
 training batch 8 mu var00.26644
compute loss for weight  -1.24266  -1.24267 result 2.34714
 training batch 9 mu var00.26644
compute loss for weight  -1.24267  -1.24267 result 2.34714
   --dy = -0.115693 dy_ref = -0.115693
 training batch 10 mu var00.266441
compute loss for weight  -1.03262  -1.03263 result 2.34714
 training batch 11 mu var00.26644
compute loss for weight  -1.03264  -1.03263 result 2.34714
 training batch 12 mu var00.26644
compute loss for weight  -1.03263  -1.03263 result 2.34714
 training batch 13 mu var00.26644
compute loss for weight  -1.03264  -1.03263 result 2.34714
   --dy = 0.147835 dy_ref = 0.147835
 training batch 14 mu var00.26644
compute loss for weight  -0.199464  -0.199474 result 2.34714
 training batch 15 mu var00.26644
compute loss for weight  -0.199484  -0.199474 result 2.34713
 training batch 16 mu var00.26644
compute loss for weight  -0.199469  -0.199474 result 2.34714
 training batch 17 mu var00.26644
compute loss for weight  -0.199479  -0.199474 result 2.34714
   --dy = 0.201362 dy_ref = 0.201362
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      4.831     -0.1363 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.26644
compute loss for weight  1.00001  1 result 2.34719
 training batch 19 mu var00.26644
compute loss for weight  0.99999  1 result 2.34709
 training batch 20 mu var00.26644
compute loss for weight  1.00001  1 result 2.34716
 training batch 21 mu var00.26644
compute loss for weight  0.999995  1 result 2.34711
   --dy = 4.83061 dy_ref = 4.83061
 training batch 22 mu var00.26644
compute loss for weight  1.00001  1 result 2.34714
 training batch 23 mu var00.26644
compute loss for weight  0.99999  1 result 2.34714
 training batch 24 mu var00.26644
compute loss for weight  1.00001  1 result 2.34714
 training batch 25 mu var00.26644
compute loss for weight  0.999995  1 result 2.34714
   --dy = -0.13634 dy_ref = -0.13634
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |   1.11e-16   1.214e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.26644
compute loss for weight  1e-05  0 result 2.34714
 training batch 27 mu var00.26644
compute loss for weight  -1e-05  0 result 2.34714
 training batch 28 mu var00.26644
compute loss for weight  5e-06  0 result 2.34714
 training batch 29 mu var00.26644
compute loss for weight  -5e-06  0 result 2.34714
   --dy = -7.40149e-12 dy_ref = 1.11022e-16
 training batch 30 mu var00.26644
compute loss for weight  1e-05  0 result 2.34714
 training batch 31 mu var00.26644
compute loss for weight  -1e-05  0 result 2.34714
 training batch 32 mu var00.26644
compute loss for weight  5e-06  0 result 2.34714
 training batch 33 mu var00.26644
compute loss for weight  -5e-06  0 result 2.34714
   --dy = -5.92119e-11 dy_ref = 1.21431e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |       3.05      0.8719 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.584     -0.1564 

 training batch 34 mu var00.26644
compute loss for weight  1.58369  1.58368 result 2.34717
 training batch 35 mu var00.26644
compute loss for weight  1.58367  1.58368 result 2.34711
 training batch 36 mu var00.26644
compute loss for weight  1.58369  1.58368 result 2.34715
 training batch 37 mu var00.26644
compute loss for weight  1.58368  1.58368 result 2.34712
   --dy = 3.05025 dy_ref = 3.05025
 training batch 38 mu var00.26644
compute loss for weight  -0.15636  -0.15637 result 2.34715
 training batch 39 mu var00.26644
compute loss for weight  -0.15638  -0.15637 result 2.34713
 training batch 40 mu var00.26644
compute loss for weight  -0.156365  -0.15637 result 2.34714
 training batch 41 mu var00.26644
compute loss for weight  -0.156375  -0.15637 result 2.34713
   --dy = 0.871905 dy_ref = 0.871905
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m1.60562e-09[NON-XML-CHAR-0x1B][39m