Execution Time0.50s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-mac1015-clang110 (macphsft19.dyndns.cern.ch) on 2019-11-13 01:12:45

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 var0-0.220308
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.3357     -0.4369 
   1 |     -0.452       0.684 
   2 |    -0.1158     -0.4363 
   3 |    -0.4508       1.975 
   4 |    -0.3532   -0.001448 
   5 |    -0.8413     -0.2627 
   6 |      0.522     -0.4584 
   7 |   -0.01617     -0.6164 
   8 |     -0.155      -1.769 
   9 |  -0.005106       0.356 

output BN 
output DL feature 0 mean -0.220308	output DL std 0.361726
output DL feature 1 mean -0.096557	output DL std 0.975988
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.3361     -0.3676 
   1 |    -0.6749       0.843 
   2 |     0.3044     -0.3669 
   3 |    -0.6715       2.238 
   4 |    -0.3871      0.1027 
   5 |     -1.809     -0.1794 
   6 |      2.162     -0.3908 
   7 |     0.5946     -0.5614 
   8 |     0.1902      -1.806 
   9 |     0.6268      0.4888 

output BN feature 0 mean -7.77156e-17	output BN std 1.05365
output BN feature 1 mean -1.66533e-17	output BN std 1.05403
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |      0.898      0.6236       1.528       0.809 
   1 |     0.5706      -0.026      0.5212    -0.09055 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.4109   -0.002812      0.1042      0.2626 
   1 |   -0.04688     -0.4045     -0.7407     -0.4348 

 training batch 2 mu var0-0.220306
compute loss for weight  -0.410908  -0.410918 result 0.94098
 training batch 3 mu var0-0.220308
compute loss for weight  -0.410928  -0.410918 result 0.940962
 training batch 4 mu var0-0.220308
compute loss for weight  -0.410913  -0.410918 result 0.940975
 training batch 5 mu var0-0.220308
compute loss for weight  -0.410923  -0.410918 result 0.940966
   --dy = 0.89802 dy_ref = 0.89802
 training batch 6 mu var0-0.220309
compute loss for weight  -0.00280151  -0.00281151 result 0.940977
 training batch 7 mu var0-0.220308
compute loss for weight  -0.00282151  -0.00281151 result 0.940965
 training batch 8 mu var0-0.220309
compute loss for weight  -0.00280651  -0.00281151 result 0.940974
 training batch 9 mu var0-0.220308
compute loss for weight  -0.00281651  -0.00281151 result 0.940968
   --dy = 0.623567 dy_ref = 0.623567
 training batch 10 mu var0-0.220308
compute loss for weight  0.104179  0.104169 result 0.940986
 training batch 11 mu var0-0.220308
compute loss for weight  0.104159  0.104169 result 0.940956
 training batch 12 mu var0-0.220308
compute loss for weight  0.104174  0.104169 result 0.940979
 training batch 13 mu var0-0.220308
compute loss for weight  0.104164  0.104169 result 0.940963
   --dy = 1.52815 dy_ref = 1.52815
 training batch 14 mu var0-0.220309
compute loss for weight  0.262573  0.262563 result 0.940979
 training batch 15 mu var0-0.220308
compute loss for weight  0.262553  0.262563 result 0.940963
 training batch 16 mu var0-0.220309
compute loss for weight  0.262568  0.262563 result 0.940975
 training batch 17 mu var0-0.220308
compute loss for weight  0.262558  0.262563 result 0.940967
   --dy = 0.808994 dy_ref = 0.808994
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.9792      0.9027 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.220308
compute loss for weight  1.00001  1 result 0.940981
 training batch 19 mu var0-0.220308
compute loss for weight  0.99999  1 result 0.940961
 training batch 20 mu var0-0.220308
compute loss for weight  1.00001  1 result 0.940976
 training batch 21 mu var0-0.220308
compute loss for weight  0.999995  1 result 0.940966
   --dy = 0.979249 dy_ref = 0.979249
 training batch 22 mu var0-0.220308
compute loss for weight  1.00001  1 result 0.94098
 training batch 23 mu var0-0.220308
compute loss for weight  0.99999  1 result 0.940962
 training batch 24 mu var0-0.220308
compute loss for weight  1.00001  1 result 0.940975
 training batch 25 mu var0-0.220308
compute loss for weight  0.999995  1 result 0.940966
   --dy = 0.902693 dy_ref = 0.902693
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -6.072e-17   1.353e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.220308
compute loss for weight  1e-05  0 result 0.940971
 training batch 27 mu var0-0.220308
compute loss for weight  -1e-05  0 result 0.940971
 training batch 28 mu var0-0.220308
compute loss for weight  5e-06  0 result 0.940971
 training batch 29 mu var0-0.220308
compute loss for weight  -5e-06  0 result 0.940971
   --dy = 0 dy_ref = -6.07153e-17
 training batch 30 mu var0-0.220308
compute loss for weight  1e-05  0 result 0.940971
 training batch 31 mu var0-0.220308
compute loss for weight  -1e-05  0 result 0.940971
 training batch 32 mu var0-0.220308
compute loss for weight  5e-06  0 result 0.940971
 training batch 33 mu var0-0.220308
compute loss for weight  -5e-06  0 result 0.940971
   --dy = 0 dy_ref = 1.35308e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.585      -1.541 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.6178     -0.5858 

 training batch 34 mu var0-0.220308
compute loss for weight  0.617836  0.617826 result 0.940987
 training batch 35 mu var0-0.220308
compute loss for weight  0.617816  0.617826 result 0.940955
 training batch 36 mu var0-0.220308
compute loss for weight  0.617831  0.617826 result 0.940979
 training batch 37 mu var0-0.220308
compute loss for weight  0.617821  0.617826 result 0.940963
   --dy = 1.58499 dy_ref = 1.58499
 training batch 38 mu var0-0.220308
compute loss for weight  -0.585777  -0.585787 result 0.940955
 training batch 39 mu var0-0.220308
compute loss for weight  -0.585797  -0.585787 result 0.940986
 training batch 40 mu var0-0.220308
compute loss for weight  -0.585782  -0.585787 result 0.940963
 training batch 41 mu var0-0.220308
compute loss for weight  -0.585792  -0.585787 result 0.940979
   --dy = -1.54099 dy_ref = -1.54099
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m3.18695e-11[NON-XML-CHAR-0x1B][39m