Execution Time0.05s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu19-gcc8 (root-ubuntu1904-1) on 2019-11-13 01:37:57
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 var00.418307
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.5984       0.547 
   1 |      1.282      0.6565 
   2 |    -0.7976       1.149 
   3 |     0.6897       2.171 
   4 |     0.5153      0.9608 
   5 |      1.509       1.754 
   6 |     0.3103       -2.27 
   7 |    -0.7208    -0.01343 
   8 |     0.7328     -0.9544 
   9 |    0.06404      0.1591 

output BN 
output DL feature 0 mean 0.418307	output DL std 0.750075
output DL feature 1 mean 0.416042	output DL std 1.29642
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      0.253      0.1065 
   1 |      1.213      0.1955 
   2 |     -1.709      0.5962 
   3 |     0.3814       1.427 
   4 |     0.1362      0.4429 
   5 |      1.533       1.088 
   6 |    -0.1518      -2.184 
   7 |     -1.601     -0.3492 
   8 |     0.4419      -1.114 
   9 |    -0.4978     -0.2089 

output BN feature 0 mean 1.22125e-16	output BN std 1.05399
output BN feature 1 mean -9.99201e-17	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 |   -0.05067      0.4063     0.04485       1.317 
   1 |     -0.462      0.3593      0.0745     -0.2906 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     0.9429      -0.219     -0.5564      0.1228 
   1 |     0.5223     -0.2099     -0.1329      -1.263 

 training batch 2 mu var00.41831
compute loss for weight  0.942877  0.942867 result 3.75202
 training batch 3 mu var00.418307
compute loss for weight  0.942857  0.942867 result 3.75202
 training batch 4 mu var00.418308
compute loss for weight  0.942872  0.942867 result 3.75202
 training batch 5 mu var00.418307
compute loss for weight  0.942862  0.942867 result 3.75202
   --dy = -0.0506733 dy_ref = -0.0506733
 training batch 6 mu var00.418307
compute loss for weight  -0.218995  -0.219005 result 3.75202
 training batch 7 mu var00.418307
compute loss for weight  -0.219015  -0.219005 result 3.75201
 training batch 8 mu var00.418307
compute loss for weight  -0.219  -0.219005 result 3.75202
 training batch 9 mu var00.418307
compute loss for weight  -0.21901  -0.219005 result 3.75202
   --dy = 0.406297 dy_ref = 0.406297
 training batch 10 mu var00.418307
compute loss for weight  -0.556425  -0.556435 result 3.75202
 training batch 11 mu var00.418307
compute loss for weight  -0.556445  -0.556435 result 3.75202
 training batch 12 mu var00.418307
compute loss for weight  -0.55643  -0.556435 result 3.75202
 training batch 13 mu var00.418307
compute loss for weight  -0.55644  -0.556435 result 3.75202
   --dy = 0.0448453 dy_ref = 0.0448453
 training batch 14 mu var00.418307
compute loss for weight  0.122787  0.122777 result 3.75203
 training batch 15 mu var00.418307
compute loss for weight  0.122767  0.122777 result 3.752
 training batch 16 mu var00.418307
compute loss for weight  0.122782  0.122777 result 3.75202
 training batch 17 mu var00.418307
compute loss for weight  0.122772  0.122777 result 3.75201
   --dy = 1.31701 dy_ref = 1.31701
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |   -0.06892       7.573 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.418307
compute loss for weight  1.00001  1 result 3.75202
 training batch 19 mu var00.418307
compute loss for weight  0.99999  1 result 3.75202
 training batch 20 mu var00.418307
compute loss for weight  1.00001  1 result 3.75202
 training batch 21 mu var00.418307
compute loss for weight  0.999995  1 result 3.75202
   --dy = -0.068917 dy_ref = -0.068917
 training batch 22 mu var00.418307
compute loss for weight  1.00001  1 result 3.75209
 training batch 23 mu var00.418307
compute loss for weight  0.99999  1 result 3.75194
 training batch 24 mu var00.418307
compute loss for weight  1.00001  1 result 3.75206
 training batch 25 mu var00.418307
compute loss for weight  0.999995  1 result 3.75198
   --dy = 7.57295 dy_ref = 7.57295
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  5.378e-17  -3.886e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.418307
compute loss for weight  1e-05  0 result 3.75202
 training batch 27 mu var00.418307
compute loss for weight  -1e-05  0 result 3.75202
 training batch 28 mu var00.418307
compute loss for weight  5e-06  0 result 3.75202
 training batch 29 mu var00.418307
compute loss for weight  -5e-06  0 result 3.75202
   --dy = 5.92119e-11 dy_ref = 5.37764e-17
 training batch 30 mu var00.418307
compute loss for weight  1e-05  0 result 3.75202
 training batch 31 mu var00.418307
compute loss for weight  -1e-05  0 result 3.75202
 training batch 32 mu var00.418307
compute loss for weight  5e-06  0 result 3.75202
 training batch 33 mu var00.418307
compute loss for weight  -5e-06  0 result 3.75202
   --dy = 7.40149e-11 dy_ref = -3.88578e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.2532      -3.837 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.2722      -1.974 

 training batch 34 mu var00.418307
compute loss for weight  0.272175  0.272165 result 3.75202
 training batch 35 mu var00.418307
compute loss for weight  0.272155  0.272165 result 3.75202
 training batch 36 mu var00.418307
compute loss for weight  0.27217  0.272165 result 3.75202
 training batch 37 mu var00.418307
compute loss for weight  0.27216  0.272165 result 3.75202
   --dy = -0.253218 dy_ref = -0.253218
 training batch 38 mu var00.418307
compute loss for weight  -1.97363  -1.97364 result 3.75198
 training batch 39 mu var00.418307
compute loss for weight  -1.97365  -1.97364 result 3.75206
 training batch 40 mu var00.418307
compute loss for weight  -1.97363  -1.97364 result 3.752
 training batch 41 mu var00.418307
compute loss for weight  -1.97364  -1.97364 result 3.75204
   --dy = -3.83705 dy_ref = -3.83705
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m3.2938e-09[NON-XML-CHAR-0x1B][39m