Execution Time0.06s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-fedora27-gcc7 (sft-fedora-27-2.cern.ch) on 2019-11-13 01:16:06
Repository revision: 30660dce2d9e89e4852dbf83dbd8b2cfcc137eff

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 var00.852365
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.707      -1.136 
   1 |    -0.8426     -0.8514 
   2 |       2.79      0.5964 
   3 |     0.5531      0.1288 
   4 |      1.817     -0.8476 
   5 |      3.433      -2.292 
   6 |     -1.083      -1.551 
   7 |     -1.315       1.737 
   8 |      1.672      -2.207 
   9 |    -0.2076     0.07623 

output BN 
output DL feature 0 mean 0.852365	output DL std 1.67564
output DL feature 1 mean -0.634746	output DL std 1.2752
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.5376     -0.4146 
   1 |     -1.066     -0.1791 
   2 |      1.219       1.018 
   3 |    -0.1883      0.6312 
   4 |     0.6067     -0.1759 
   5 |      1.623       -1.37 
   6 |     -1.217     -0.7576 
   7 |     -1.363       1.961 
   8 |     0.5154        -1.3 
   9 |    -0.6668      0.5877 

output BN feature 0 mean -6.66134e-17	output BN std 1.05407
output BN feature 1 mean 0	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 |    -0.2638      0.1832    -0.02617     -0.4047 
   1 |     0.1409     -0.1923      0.4334     -0.4753 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.3201     -0.7574       1.329     -0.6374 
   1 |    -0.9669      0.5953     0.08446     -0.7422 

 training batch 2 mu var00.852368
compute loss for weight  0.320092  0.320082 result 1.44806
 training batch 3 mu var00.852365
compute loss for weight  0.320072  0.320082 result 1.44807
 training batch 4 mu var00.852366
compute loss for weight  0.320087  0.320082 result 1.44806
 training batch 5 mu var00.852365
compute loss for weight  0.320077  0.320082 result 1.44806
   --dy = -0.263766 dy_ref = -0.263766
 training batch 6 mu var00.852365
compute loss for weight  -0.757356  -0.757366 result 1.44806
 training batch 7 mu var00.852365
compute loss for weight  -0.757376  -0.757366 result 1.44806
 training batch 8 mu var00.852365
compute loss for weight  -0.757361  -0.757366 result 1.44806
 training batch 9 mu var00.852365
compute loss for weight  -0.757371  -0.757366 result 1.44806
   --dy = 0.183187 dy_ref = 0.183187
 training batch 10 mu var00.852366
compute loss for weight  1.32867  1.32866 result 1.44806
 training batch 11 mu var00.852365
compute loss for weight  1.32865  1.32866 result 1.44806
 training batch 12 mu var00.852365
compute loss for weight  1.32867  1.32866 result 1.44806
 training batch 13 mu var00.852365
compute loss for weight  1.32866  1.32866 result 1.44806
   --dy = -0.0261696 dy_ref = -0.0261696
 training batch 14 mu var00.852365
compute loss for weight  -0.637363  -0.637373 result 1.44806
 training batch 15 mu var00.852365
compute loss for weight  -0.637383  -0.637373 result 1.44807
 training batch 16 mu var00.852365
compute loss for weight  -0.637368  -0.637373 result 1.44806
 training batch 17 mu var00.852365
compute loss for weight  -0.637378  -0.637373 result 1.44807
   --dy = -0.404682 dy_ref = -0.404682
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.1054       3.002 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.852365
compute loss for weight  1.00001  1 result 1.44806
 training batch 19 mu var00.852365
compute loss for weight  0.99999  1 result 1.44806
 training batch 20 mu var00.852365
compute loss for weight  1.00001  1 result 1.44806
 training batch 21 mu var00.852365
compute loss for weight  0.999995  1 result 1.44806
   --dy = -0.105394 dy_ref = -0.105394
 training batch 22 mu var00.852365
compute loss for weight  1.00001  1 result 1.44809
 training batch 23 mu var00.852365
compute loss for weight  0.99999  1 result 1.44803
 training batch 24 mu var00.852365
compute loss for weight  1.00001  1 result 1.44808
 training batch 25 mu var00.852365
compute loss for weight  0.999995  1 result 1.44805
   --dy = 3.00152 dy_ref = 3.00152
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 | -2.776e-17  -8.327e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.852365
compute loss for weight  1e-05  0 result 1.44806
 training batch 27 mu var00.852365
compute loss for weight  -1e-05  0 result 1.44806
 training batch 28 mu var00.852365
compute loss for weight  5e-06  0 result 1.44806
 training batch 29 mu var00.852365
compute loss for weight  -5e-06  0 result 1.44806
   --dy = -5.92119e-11 dy_ref = -2.77556e-17
 training batch 30 mu var00.852365
compute loss for weight  1e-05  0 result 1.44806
 training batch 31 mu var00.852365
compute loss for weight  -1e-05  0 result 1.44806
 training batch 32 mu var00.852365
compute loss for weight  5e-06  0 result 1.44806
 training batch 33 mu var00.852365
compute loss for weight  -5e-06  0 result 1.44806
   --dy = -3.33067e-11 dy_ref = -8.32667e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.2649      -2.294 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.3979      -1.308 

 training batch 34 mu var00.852365
compute loss for weight  -0.397922  -0.397932 result 1.44807
 training batch 35 mu var00.852365
compute loss for weight  -0.397942  -0.397932 result 1.44806
 training batch 36 mu var00.852365
compute loss for weight  -0.397927  -0.397932 result 1.44806
 training batch 37 mu var00.852365
compute loss for weight  -0.397937  -0.397932 result 1.44806
   --dy = 0.264854 dy_ref = 0.264854
 training batch 38 mu var00.852365
compute loss for weight  -1.30839  -1.3084 result 1.44804
 training batch 39 mu var00.852365
compute loss for weight  -1.30841  -1.3084 result 1.44809
 training batch 40 mu var00.852365
compute loss for weight  -1.3084  -1.3084 result 1.44805
 training batch 41 mu var00.852365
compute loss for weight  -1.30841  -1.3084 result 1.44807
   --dy = -2.29403 dy_ref = -2.29403
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m1.03349e-09[NON-XML-CHAR-0x1B][39m