Execution Time0.11s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-aarch64-centos7-gcc48 (techlab-arm64-moonshot-xgene-004) on 2019-11-15 00:49:21
Repository revision: 14de58de35eff907054671888ccc2de0f7f27e77

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 var0-0.642132
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     -1.283      0.2282 
   1 |    -0.5243     -0.6626 
   2 |     0.5868      0.6946 
   3 |      1.234     -0.1496 
   4 |     -0.749      0.2642 
   5 |      -2.25      0.3293 
   6 |     -2.056      0.3357 
   7 |      1.503     -0.6684 
   8 |     -3.131      0.3823 
   9 |     0.2485    -0.04477 

output BN 
output DL feature 0 mean -0.642132	output DL std 1.5507
output DL feature 1 mean 0.0708931	output DL std 0.451109
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |    -0.4358      0.3675 
   1 |    0.08006      -1.713 
   2 |     0.8353       1.457 
   3 |      1.276     -0.5152 
   4 |   -0.07262      0.4517 
   5 |     -1.093      0.6037 
   6 |    -0.9613      0.6185 
   7 |      1.458      -1.727 
   8 |     -1.692      0.7275 
   9 |     0.6054     -0.2702 

output BN feature 0 mean 1.11022e-17	output BN std 1.05407
output BN feature 1 mean 5.55112e-17	output BN std 1.0538
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    0.02451      0.1169     -0.1455      0.0521 
   1 |     -2.175       6.047      -8.288      -1.601 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.9608       0.404     -0.2314      -1.099 
   1 |    -0.2942     -0.3476      0.4788      0.1137 

 training batch 2 mu var0-0.642129
compute loss for weight  -0.960769  -0.960779 result 1.03512
 training batch 3 mu var0-0.642132
compute loss for weight  -0.960789  -0.960779 result 1.03512
 training batch 4 mu var0-0.642131
compute loss for weight  -0.960774  -0.960779 result 1.03512
 training batch 5 mu var0-0.642132
compute loss for weight  -0.960784  -0.960779 result 1.03512
   --dy = 0.0245109 dy_ref = 0.0245109
 training batch 6 mu var0-0.642132
compute loss for weight  0.404018  0.404008 result 1.03512
 training batch 7 mu var0-0.642132
compute loss for weight  0.403998  0.404008 result 1.03512
 training batch 8 mu var0-0.642132
compute loss for weight  0.404013  0.404008 result 1.03512
 training batch 9 mu var0-0.642132
compute loss for weight  0.404003  0.404008 result 1.03512
   --dy = 0.116852 dy_ref = 0.116852
 training batch 10 mu var0-0.642131
compute loss for weight  -0.231356  -0.231366 result 1.03511
 training batch 11 mu var0-0.642132
compute loss for weight  -0.231376  -0.231366 result 1.03512
 training batch 12 mu var0-0.642132
compute loss for weight  -0.231361  -0.231366 result 1.03512
 training batch 13 mu var0-0.642132
compute loss for weight  -0.231371  -0.231366 result 1.03512
   --dy = -0.14545 dy_ref = -0.14545
 training batch 14 mu var0-0.642132
compute loss for weight  -1.0986  -1.09861 result 1.03512
 training batch 15 mu var0-0.642132
compute loss for weight  -1.09862  -1.09861 result 1.03512
 training batch 16 mu var0-0.642132
compute loss for weight  -1.0986  -1.09861 result 1.03512
 training batch 17 mu var0-0.642132
compute loss for weight  -1.09861  -1.09861 result 1.03512
   --dy = 0.052099 dy_ref = 0.052099
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.634      0.4362 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var0-0.642132
compute loss for weight  1.00001  1 result 1.03513
 training batch 19 mu var0-0.642132
compute loss for weight  0.99999  1 result 1.0351
 training batch 20 mu var0-0.642132
compute loss for weight  1.00001  1 result 1.03512
 training batch 21 mu var0-0.642132
compute loss for weight  0.999995  1 result 1.03511
   --dy = 1.63402 dy_ref = 1.63402
 training batch 22 mu var0-0.642132
compute loss for weight  1.00001  1 result 1.03512
 training batch 23 mu var0-0.642132
compute loss for weight  0.99999  1 result 1.03511
 training batch 24 mu var0-0.642132
compute loss for weight  1.00001  1 result 1.03512
 training batch 25 mu var0-0.642132
compute loss for weight  0.999995  1 result 1.03511
   --dy = 0.436213 dy_ref = 0.436213
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  2.776e-17   3.469e-17 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var0-0.642132
compute loss for weight  1e-05  0 result 1.03512
 training batch 27 mu var0-0.642132
compute loss for weight  -1e-05  0 result 1.03512
 training batch 28 mu var0-0.642132
compute loss for weight  5e-06  0 result 1.03512
 training batch 29 mu var0-0.642132
compute loss for weight  -5e-06  0 result 1.03512
   --dy = -3.70074e-12 dy_ref = 2.77556e-17
 training batch 30 mu var0-0.642132
compute loss for weight  1e-05  0 result 1.03512
 training batch 31 mu var0-0.642132
compute loss for weight  -1e-05  0 result 1.03512
 training batch 32 mu var0-0.642132
compute loss for weight  5e-06  0 result 1.03512
 training batch 33 mu var0-0.642132
compute loss for weight  -5e-06  0 result 1.03512
   --dy = 0 dy_ref = 3.46945e-17
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.965      -1.431 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.8318     -0.3049 

 training batch 34 mu var0-0.642132
compute loss for weight  0.831782  0.831772 result 1.03514
 training batch 35 mu var0-0.642132
compute loss for weight  0.831762  0.831772 result 1.0351
 training batch 36 mu var0-0.642132
compute loss for weight  0.831777  0.831772 result 1.03513
 training batch 37 mu var0-0.642132
compute loss for weight  0.831767  0.831772 result 1.03511
   --dy = 1.9645 dy_ref = 1.9645
 training batch 38 mu var0-0.642132
compute loss for weight  -0.304885  -0.304895 result 1.0351
 training batch 39 mu var0-0.642132
compute loss for weight  -0.304905  -0.304895 result 1.03513
 training batch 40 mu var0-0.642132
compute loss for weight  -0.30489  -0.304895 result 1.03511
 training batch 41 mu var0-0.642132
compute loss for weight  -0.3049  -0.304895 result 1.03512
   --dy = -1.4307 dy_ref = -1.4307
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][33m2.59634e-09[NON-XML-CHAR-0x1B][39m