Execution Time0.10s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu19-gcc9 (root-ubuntu-1910-1) on 2019-11-14 00:48:35
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 var00.248455
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.5992      0.6422 
   1 |   -0.06074       1.224 
   2 |     0.3308      0.3826 
   3 |     -1.282       1.991 
   4 |     0.2337      0.9102 
   5 |     0.8457       1.955 
   6 |    -0.1658      -1.704 
   7 |     0.7358     -0.1518 
   8 |      1.541     -0.4789 
   9 |    -0.2927      0.1559 

output BN 
output DL feature 0 mean 0.248455	output DL std 0.767312
output DL feature 1 mean 0.492632	output DL std 1.12677
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.4817      0.1399 
   1 |    -0.4247       0.684 
   2 |     0.1131     -0.1029 
   3 |     -2.103       1.402 
   4 |   -0.02029      0.3907 
   5 |     0.8204       1.368 
   6 |     -0.569      -2.055 
   7 |     0.6694     -0.6028 
   8 |      1.776     -0.9088 
   9 |    -0.7433      -0.315 

output BN feature 0 mean 5.55112e-17	output BN std 1.05399
output BN feature 1 mean -3.88578e-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 |     -1.191      0.9523     -0.1968       1.864 
   1 |     -1.382    -0.06134      -1.221      0.4966 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.4503      0.4449      0.4425      0.1073 
   1 |     0.8946     -0.1666     -0.4508     -0.9484 

 training batch 2 mu var00.248458
compute loss for weight  0.450319  0.450309 result 3.16579
 training batch 3 mu var00.248455
compute loss for weight  0.450299  0.450309 result 3.16582
 training batch 4 mu var00.248456
compute loss for weight  0.450314  0.450309 result 3.1658
 training batch 5 mu var00.248455
compute loss for weight  0.450304  0.450309 result 3.16581
   --dy = -1.19073 dy_ref = -1.19073
 training batch 6 mu var00.248454
compute loss for weight  0.444863  0.444853 result 3.16582
 training batch 7 mu var00.248455
compute loss for weight  0.444843  0.444853 result 3.1658
 training batch 8 mu var00.248455
compute loss for weight  0.444858  0.444853 result 3.16581
 training batch 9 mu var00.248455
compute loss for weight  0.444848  0.444853 result 3.1658
   --dy = 0.952297 dy_ref = 0.952297
 training batch 10 mu var00.248455
compute loss for weight  0.442537  0.442527 result 3.1658
 training batch 11 mu var00.248455
compute loss for weight  0.442517  0.442527 result 3.16581
 training batch 12 mu var00.248455
compute loss for weight  0.442532  0.442527 result 3.16581
 training batch 13 mu var00.248455
compute loss for weight  0.442522  0.442527 result 3.16581
   --dy = -0.196846 dy_ref = -0.196846
 training batch 14 mu var00.248455
compute loss for weight  0.107308  0.107298 result 3.16582
 training batch 15 mu var00.248455
compute loss for weight  0.107288  0.107298 result 3.16579
 training batch 16 mu var00.248455
compute loss for weight  0.107303  0.107298 result 3.16582
 training batch 17 mu var00.248455
compute loss for weight  0.107293  0.107298 result 3.1658
   --dy = 1.86391 dy_ref = 1.86391
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.685       4.647 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.248455
compute loss for weight  1.00001  1 result 3.16582
 training batch 19 mu var00.248455
compute loss for weight  0.99999  1 result 3.16579
 training batch 20 mu var00.248455
compute loss for weight  1.00001  1 result 3.16581
 training batch 21 mu var00.248455
compute loss for weight  0.999995  1 result 3.1658
   --dy = 1.68479 dy_ref = 1.68479
 training batch 22 mu var00.248455
compute loss for weight  1.00001  1 result 3.16585
 training batch 23 mu var00.248455
compute loss for weight  0.99999  1 result 3.16576
 training batch 24 mu var00.248455
compute loss for weight  1.00001  1 result 3.16583
 training batch 25 mu var00.248455
compute loss for weight  0.999995  1 result 3.16578
   --dy = 4.64682 dy_ref = 4.64682
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |   1.11e-16  -3.851e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.248455
compute loss for weight  1e-05  0 result 3.16581
 training batch 27 mu var00.248455
compute loss for weight  -1e-05  0 result 3.16581
 training batch 28 mu var00.248455
compute loss for weight  5e-06  0 result 3.16581
 training batch 29 mu var00.248455
compute loss for weight  -5e-06  0 result 3.16581
   --dy = -5.18104e-11 dy_ref = 1.11022e-16
 training batch 30 mu var00.248455
compute loss for weight  1e-05  0 result 3.16581
 training batch 31 mu var00.248455
compute loss for weight  -1e-05  0 result 3.16581
 training batch 32 mu var00.248455
compute loss for weight  5e-06  0 result 3.16581
 training batch 33 mu var00.248455
compute loss for weight  -5e-06  0 result 3.16581
   --dy = 1.18424e-10 dy_ref = -3.85109e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      2.258      -3.255 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.7462      -1.428 

 training batch 34 mu var00.248455
compute loss for weight  0.746172  0.746162 result 3.16583
 training batch 35 mu var00.248455
compute loss for weight  0.746152  0.746162 result 3.16578
 training batch 36 mu var00.248455
compute loss for weight  0.746167  0.746162 result 3.16582
 training batch 37 mu var00.248455
compute loss for weight  0.746157  0.746162 result 3.1658
   --dy = 2.25795 dy_ref = 2.25795
 training batch 38 mu var00.248455
compute loss for weight  -1.42752  -1.42753 result 3.16577
 training batch 39 mu var00.248455
compute loss for weight  -1.42754  -1.42753 result 3.16584
 training batch 40 mu var00.248455
compute loss for weight  -1.42753  -1.42753 result 3.16579
 training batch 41 mu var00.248455
compute loss for weight  -1.42754  -1.42753 result 3.16582
   --dy = -3.25514 dy_ref = -3.25514
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m2.39146e-10[NON-XML-CHAR-0x1B][39m