Execution Time0.83s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-mac1013-clang100 (macitois22.cern.ch) on 2019-11-13 01:56:46
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 var0-1.06667
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.998      -1.591 
   1 |    -0.4059     -0.2889 
   2 |     -1.428      -1.172 
   3 |    -0.2502     -0.4909 
   4 |      -1.81      -1.516 
   5 |     -4.069      -3.309 
   6 |     0.4374      0.1876 
   7 |      1.243       1.396 
   8 |     -2.629      -1.957 
   9 |      0.243      0.1389 

output BN 
output DL feature 0 mean -1.06667	output DL std 1.61312
output DL feature 1 mean -0.86013	output DL std 1.32961
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.6088     -0.5792 
   1 |     0.4317      0.4529 
   2 |     -0.236      -0.247 
   3 |     0.5335      0.2927 
   4 |    -0.4857     -0.5203 
   5 |     -1.962      -1.941 
   6 |     0.9828      0.8306 
   7 |      1.509       1.789 
   8 |     -1.021     -0.8697 
   9 |     0.8558       0.792 

output BN feature 0 mean -1.11022e-17	output BN std 1.05407
output BN feature 1 mean 3.33067e-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 |    0.01913     0.09877     0.04403     0.01354 
   1 |      4.304      -3.786       3.727     -0.3605 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |     -1.211      0.5234     -0.7089      0.2203 
   1 |    -0.8894      0.6246     -0.5451       0.152 

 training batch 2 mu var0-1.06667
compute loss for weight  -1.21125  -1.21126 result 4.32325
 training batch 3 mu var0-1.06667
compute loss for weight  -1.21127  -1.21126 result 4.32325
 training batch 4 mu var0-1.06667
compute loss for weight  -1.21126  -1.21126 result 4.32325
 training batch 5 mu var0-1.06667
compute loss for weight  -1.21127  -1.21126 result 4.32325
   --dy = 0.0191318 dy_ref = 0.0191318
 training batch 6 mu var0-1.06667
compute loss for weight  0.523408  0.523398 result 4.32325
 training batch 7 mu var0-1.06667
compute loss for weight  0.523388  0.523398 result 4.32325
 training batch 8 mu var0-1.06667
compute loss for weight  0.523403  0.523398 result 4.32325
 training batch 9 mu var0-1.06667
compute loss for weight  0.523393  0.523398 result 4.32325
   --dy = 0.0987706 dy_ref = 0.0987706
 training batch 10 mu var0-1.06667
compute loss for weight  -0.708859  -0.708869 result 4.32325
 training batch 11 mu var0-1.06667
compute loss for weight  -0.708879  -0.708869 result 4.32325
 training batch 12 mu var0-1.06667
compute loss for weight  -0.708864  -0.708869 result 4.32325
 training batch 13 mu var0-1.06667
compute loss for weight  -0.708874  -0.708869 result 4.32325
   --dy = 0.0440338 dy_ref = 0.0440338
 training batch 14 mu var0-1.06667
compute loss for weight  0.220347  0.220337 result 4.32325
 training batch 15 mu var0-1.06667
compute loss for weight  0.220327  0.220337 result 4.32325
 training batch 16 mu var0-1.06667
compute loss for weight  0.220342  0.220337 result 4.32325
 training batch 17 mu var0-1.06667
compute loss for weight  0.220332  0.220337 result 4.32325
   --dy = 0.013538 dy_ref = 0.013538
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      6.824       1.822 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-1.06667
compute loss for weight  1.00001  1 result 4.32332
 training batch 19 mu var0-1.06667
compute loss for weight  0.99999  1 result 4.32318
 training batch 20 mu var0-1.06667
compute loss for weight  1.00001  1 result 4.32329
 training batch 21 mu var0-1.06667
compute loss for weight  0.999995  1 result 4.32322
   --dy = 6.82447 dy_ref = 6.82447
 training batch 22 mu var0-1.06667
compute loss for weight  1.00001  1 result 4.32327
 training batch 23 mu var0-1.06667
compute loss for weight  0.99999  1 result 4.32323
 training batch 24 mu var0-1.06667
compute loss for weight  1.00001  1 result 4.32326
 training batch 25 mu var0-1.06667
compute loss for weight  0.999995  1 result 4.32324
   --dy = 1.82204 dy_ref = 1.82204
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  6.661e-16   1.388e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-1.06667
compute loss for weight  1e-05  0 result 4.32325
 training batch 27 mu var0-1.06667
compute loss for weight  -1e-05  0 result 4.32325
 training batch 28 mu var0-1.06667
compute loss for weight  5e-06  0 result 4.32325
 training batch 29 mu var0-1.06667
compute loss for weight  -5e-06  0 result 4.32325
   --dy = 1.18424e-10 dy_ref = 6.66134e-16
 training batch 30 mu var0-1.06667
compute loss for weight  1e-05  0 result 4.32325
 training batch 31 mu var0-1.06667
compute loss for weight  -1e-05  0 result 4.32325
 training batch 32 mu var0-1.06667
compute loss for weight  5e-06  0 result 4.32325
 training batch 33 mu var0-1.06667
compute loss for weight  -5e-06  0 result 4.32325
   --dy = 1.18424e-10 dy_ref = 1.38778e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      4.157       4.134 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.642      0.4408 

 training batch 34 mu var0-1.06667
compute loss for weight  1.64183  1.64182 result 4.32329
 training batch 35 mu var0-1.06667
compute loss for weight  1.64181  1.64182 result 4.32321
 training batch 36 mu var0-1.06667
compute loss for weight  1.64183  1.64182 result 4.32327
 training batch 37 mu var0-1.06667
compute loss for weight  1.64182  1.64182 result 4.32323
   --dy = 4.15663 dy_ref = 4.15663
 training batch 38 mu var0-1.06667
compute loss for weight  0.440771  0.440761 result 4.32329
 training batch 39 mu var0-1.06667
compute loss for weight  0.440751  0.440761 result 4.32321
 training batch 40 mu var0-1.06667
compute loss for weight  0.440766  0.440761 result 4.32327
 training batch 41 mu var0-1.06667
compute loss for weight  0.440756  0.440761 result 4.32323
   --dy = 4.13384 dy_ref = 4.13384
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m1.83772e-08[NON-XML-CHAR-0x1B][39m