Execution Time0.19s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-centos7-gcc62-opt-no-rt-cxxmodules (olsnba08.cern.ch) on 2019-11-14 01:02:24
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 var00.891684
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.652      -1.079 
   1 |     0.8183     -0.9233 
   2 |      1.671     -0.1367 
   3 |    -0.1626     -0.2315 
   4 |      1.364     -0.8975 
   5 |       3.36      -2.307 
   6 |     -3.041       0.396 
   7 |      1.854       0.273 
   8 |      1.799      -1.441 
   9 |    -0.3971      0.1151 

output BN 
output DL feature 0 mean 0.891684	output DL std 1.7478
output DL feature 1 mean -0.623232	output DL std 0.859954
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.4583     -0.5587 
   1 |   -0.04427     -0.3678 
   2 |     0.4701      0.5963 
   3 |    -0.6358      0.4802 
   4 |     0.2846     -0.3362 
   5 |      1.489      -2.064 
   6 |     -2.372       1.249 
   7 |     0.5806       1.099 
   8 |     0.5469      -1.002 
   9 |    -0.7772       0.905 

output BN feature 0 mean -1.11022e-17	output BN std 1.05407
output BN feature 1 mean -7.77156e-17	output BN std 1.05401
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.5395      0.5472     -0.1071     -0.3438 
   1 |      3.115     -0.5466       2.438      -1.228 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |      1.549      0.8515       0.559       -1.25 
   1 |     -1.078     0.03689     0.02373      0.1483 

 training batch 2 mu var00.891686
compute loss for weight  1.54899  1.54898 result 0.830446
 training batch 3 mu var00.891684
compute loss for weight  1.54897  1.54898 result 0.830457
 training batch 4 mu var00.891684
compute loss for weight  1.54898  1.54898 result 0.830448
 training batch 5 mu var00.891684
compute loss for weight  1.54897  1.54898 result 0.830454
   --dy = -0.539514 dy_ref = -0.539514
 training batch 6 mu var00.891683
compute loss for weight  0.851467  0.851457 result 0.830457
 training batch 7 mu var00.891684
compute loss for weight  0.851447  0.851457 result 0.830446
 training batch 8 mu var00.891683
compute loss for weight  0.851462  0.851457 result 0.830454
 training batch 9 mu var00.891684
compute loss for weight  0.851452  0.851457 result 0.830448
   --dy = 0.547211 dy_ref = 0.547211
 training batch 10 mu var00.891684
compute loss for weight  0.558968  0.558958 result 0.83045
 training batch 11 mu var00.891684
compute loss for weight  0.558948  0.558958 result 0.830452
 training batch 12 mu var00.891684
compute loss for weight  0.558963  0.558958 result 0.830451
 training batch 13 mu var00.891684
compute loss for weight  0.558953  0.558958 result 0.830452
   --dy = -0.107105 dy_ref = -0.107105
 training batch 14 mu var00.891684
compute loss for weight  -1.24971  -1.24972 result 0.830448
 training batch 15 mu var00.891684
compute loss for weight  -1.24973  -1.24972 result 0.830455
 training batch 16 mu var00.891684
compute loss for weight  -1.24972  -1.24972 result 0.830449
 training batch 17 mu var00.891684
compute loss for weight  -1.24973  -1.24972 result 0.830453
   --dy = -0.343835 dy_ref = -0.343835
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.724    -0.06321 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.891684
compute loss for weight  1.00001  1 result 0.830468
 training batch 19 mu var00.891684
compute loss for weight  0.99999  1 result 0.830434
 training batch 20 mu var00.891684
compute loss for weight  1.00001  1 result 0.83046
 training batch 21 mu var00.891684
compute loss for weight  0.999995  1 result 0.830443
   --dy = 1.72412 dy_ref = 1.72412
 training batch 22 mu var00.891684
compute loss for weight  1.00001  1 result 0.830451
 training batch 23 mu var00.891684
compute loss for weight  0.99999  1 result 0.830452
 training batch 24 mu var00.891684
compute loss for weight  1.00001  1 result 0.830451
 training batch 25 mu var00.891684
compute loss for weight  0.999995  1 result 0.830451
   --dy = -0.0632148 dy_ref = -0.0632148
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  -1.11e-16  -5.551e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.891684
compute loss for weight  1e-05  0 result 0.830451
 training batch 27 mu var00.891684
compute loss for weight  -1e-05  0 result 0.830451
 training batch 28 mu var00.891684
compute loss for weight  5e-06  0 result 0.830451
 training batch 29 mu var00.891684
compute loss for weight  -5e-06  0 result 0.830451
   --dy = 5.55112e-12 dy_ref = -1.11022e-16
 training batch 30 mu var00.891684
compute loss for weight  1e-05  0 result 0.830451
 training batch 31 mu var00.891684
compute loss for weight  -1e-05  0 result 0.830451
 training batch 32 mu var00.891684
compute loss for weight  5e-06  0 result 0.830451
 training batch 33 mu var00.891684
compute loss for weight  -5e-06  0 result 0.830451
   --dy = 0 dy_ref = -5.55112e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      -1.35     0.07365 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.277     -0.8584 

 training batch 34 mu var00.891684
compute loss for weight  -1.27686  -1.27687 result 0.830438
 training batch 35 mu var00.891684
compute loss for weight  -1.27688  -1.27687 result 0.830465
 training batch 36 mu var00.891684
compute loss for weight  -1.27686  -1.27687 result 0.830444
 training batch 37 mu var00.891684
compute loss for weight  -1.27687  -1.27687 result 0.830458
   --dy = -1.35027 dy_ref = -1.35027
 training batch 38 mu var00.891684
compute loss for weight  -0.858349  -0.858359 result 0.830452
 training batch 39 mu var00.891684
compute loss for weight  -0.858369  -0.858359 result 0.83045
 training batch 40 mu var00.891684
compute loss for weight  -0.858354  -0.858359 result 0.830452
 training batch 41 mu var00.891684
compute loss for weight  -0.858364  -0.858359 result 0.830451
   --dy = 0.0736461 dy_ref = 0.0736461
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m4.45694e-10[NON-XML-CHAR-0x1B][39m