Execution Time0.19s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-centos7-gcc62-opt-master (olhswep09.cern.ch) on 2019-11-13 23:14:21
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 var0-0.119009
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.1321      0.3312 
   1 |     -1.055     -0.2269 
   2 |      2.181      0.5238 
   3 |     0.3897      0.2728 
   4 |     0.1691      0.3995 
   5 |    -0.2464      0.6976 
   6 |      -2.83     0.08814 
   7 |      1.719     -0.6874 
   8 |     -1.298      0.2918 
   9 |   -0.08675    0.001172 

output BN 
output DL feature 0 mean -0.119009	output DL std 1.43918
output DL feature 1 mean 0.169171	output DL std 0.399352
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  -0.009618      0.4275 
   1 |    -0.6856      -1.045 
   2 |      1.684      0.9357 
   3 |     0.3726      0.2734 
   4 |      0.211      0.6079 
   5 |   -0.09331       1.394 
   6 |     -1.985     -0.2138 
   7 |      1.346       -2.26 
   8 |    -0.8636      0.3236 
   9 |    0.02362     -0.4433 

output BN feature 0 mean -2.01228e-17	output BN std 1.05406
output BN feature 1 mean 4.44089e-17	output BN std 1.05373
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.3106      0.5717     -0.3949      0.1171 
   1 |    -0.4017       1.099      -1.643       2.383 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.5442      0.5048      0.7546      -1.363 
   1 |   -0.02258     -0.3527      0.2554    -0.02414 

 training batch 2 mu var0-0.119006
compute loss for weight  -0.544141  -0.544151 result 3.32447
 training batch 3 mu var0-0.119009
compute loss for weight  -0.544161  -0.544151 result 3.32447
 training batch 4 mu var0-0.119009
compute loss for weight  -0.544146  -0.544151 result 3.32447
 training batch 5 mu var0-0.119009
compute loss for weight  -0.544156  -0.544151 result 3.32447
   --dy = -0.310552 dy_ref = -0.310552
 training batch 6 mu var0-0.11901
compute loss for weight  0.504763  0.504753 result 3.32448
 training batch 7 mu var0-0.119009
compute loss for weight  0.504743  0.504753 result 3.32446
 training batch 8 mu var0-0.119009
compute loss for weight  0.504758  0.504753 result 3.32447
 training batch 9 mu var0-0.119009
compute loss for weight  0.504748  0.504753 result 3.32447
   --dy = 0.571725 dy_ref = 0.571725
 training batch 10 mu var0-0.119009
compute loss for weight  0.754584  0.754574 result 3.32447
 training batch 11 mu var0-0.119009
compute loss for weight  0.754564  0.754574 result 3.32447
 training batch 12 mu var0-0.119009
compute loss for weight  0.754579  0.754574 result 3.32447
 training batch 13 mu var0-0.119009
compute loss for weight  0.754569  0.754574 result 3.32447
   --dy = -0.394921 dy_ref = -0.394921
 training batch 14 mu var0-0.119009
compute loss for weight  -1.36276  -1.36277 result 3.32447
 training batch 15 mu var0-0.119009
compute loss for weight  -1.36278  -1.36277 result 3.32447
 training batch 16 mu var0-0.119009
compute loss for weight  -1.36277  -1.36277 result 3.32447
 training batch 17 mu var0-0.119009
compute loss for weight  -1.36278  -1.36277 result 3.32447
   --dy = 0.117083 dy_ref = 0.117083
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.2389        6.41 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.119009
compute loss for weight  1.00001  1 result 3.32447
 training batch 19 mu var0-0.119009
compute loss for weight  0.99999  1 result 3.32447
 training batch 20 mu var0-0.119009
compute loss for weight  1.00001  1 result 3.32447
 training batch 21 mu var0-0.119009
compute loss for weight  0.999995  1 result 3.32447
   --dy = 0.238923 dy_ref = 0.238923
 training batch 22 mu var0-0.119009
compute loss for weight  1.00001  1 result 3.32453
 training batch 23 mu var0-0.119009
compute loss for weight  0.99999  1 result 3.32441
 training batch 24 mu var0-0.119009
compute loss for weight  1.00001  1 result 3.3245
 training batch 25 mu var0-0.119009
compute loss for weight  0.999995  1 result 3.32444
   --dy = 6.41002 dy_ref = 6.41002
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -1.388e-17   2.776e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.119009
compute loss for weight  1e-05  0 result 3.32447
 training batch 27 mu var0-0.119009
compute loss for weight  -1e-05  0 result 3.32447
 training batch 28 mu var0-0.119009
compute loss for weight  5e-06  0 result 3.32447
 training batch 29 mu var0-0.119009
compute loss for weight  -5e-06  0 result 3.32447
   --dy = 1.4803e-11 dy_ref = -1.38778e-17
 training batch 30 mu var0-0.119009
compute loss for weight  1e-05  0 result 3.32447
 training batch 31 mu var0-0.119009
compute loss for weight  -1e-05  0 result 3.32447
 training batch 32 mu var0-0.119009
compute loss for weight  5e-06  0 result 3.32447
 training batch 33 mu var0-0.119009
compute loss for weight  -5e-06  0 result 3.32447
   --dy = -1.4803e-11 dy_ref = 2.77556e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.7901      -3.595 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.3024      -1.783 

 training batch 34 mu var0-0.119009
compute loss for weight  0.302411  0.302401 result 3.32448
 training batch 35 mu var0-0.119009
compute loss for weight  0.302391  0.302401 result 3.32446
 training batch 36 mu var0-0.119009
compute loss for weight  0.302406  0.302401 result 3.32447
 training batch 37 mu var0-0.119009
compute loss for weight  0.302396  0.302401 result 3.32447
   --dy = 0.790086 dy_ref = 0.790086
 training batch 38 mu var0-0.119009
compute loss for weight  -1.78302  -1.78303 result 3.32443
 training batch 39 mu var0-0.119009
compute loss for weight  -1.78304  -1.78303 result 3.32451
 training batch 40 mu var0-0.119009
compute loss for weight  -1.78303  -1.78303 result 3.32445
 training batch 41 mu var0-0.119009
compute loss for weight  -1.78304  -1.78303 result 3.32449
   --dy = -3.59501 dy_ref = -3.59501
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m1.54702e-10[NON-XML-CHAR-0x1B][39m