Execution Time0.41s

Test: TMVA-DNN-BatchNormalization (Passed)
Build: master-x86_64-ubuntu18-gcc7 (sft-ubuntu-1804-3) on 2019-11-14 03:09:45
Repository revision: 32b17abcda23e44b64218a42d0ca69cb30cda7e0

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 var00.81459
output DL 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.494    -0.01089 
   1 |     0.2089      0.8801 
   2 |      1.015     -0.1952 
   3 |     0.6837       1.004 
   4 |      1.483      0.1219 
   5 |      3.152      0.3603 
   6 |     0.3231      -1.238 
   7 |     -1.968      0.6498 
   8 |      1.823     -0.7642 
   9 |   -0.06918      0.1061 

output BN 
output DL feature 0 mean 0.81459	output DL std 1.3571
output DL feature 1 mean 0.0913567	output DL std 0.701843
output of BN 

10x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.5279     -0.1536 
   1 |    -0.4705       1.184 
   2 |     0.1554     -0.4303 
   3 |    -0.1017       1.371 
   4 |     0.5192      0.0459 
   5 |      1.815      0.4039 
   6 |    -0.3817      -1.997 
   7 |     -2.161      0.8386 
   8 |     0.7834      -1.285 
   9 |    -0.6864     0.02209 

output BN feature 0 mean 2.22045e-17	output BN std 1.05406
output BN feature 1 mean -5.20417e-18	output BN std 1.05397
Testing weight gradients   for    layer 0
weight gradient for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |    -0.2046      0.2425      0.6439       1.448 
   1 |     -2.206       2.439      0.2974       3.687 

weights for layer 0

2x4 matrix is as follows

     |      0    |      1    |      2    |      3    |
---------------------------------------------------------
   0 |      0.728     -0.8905      0.4873     0.03535 
   1 |     0.4767       0.254     -0.5055     -0.6179 

 training batch 2 mu var00.814592
compute loss for weight  0.727989  0.727979 result 4.50354
 training batch 3 mu var00.81459
compute loss for weight  0.727969  0.727979 result 4.50355
 training batch 4 mu var00.81459
compute loss for weight  0.727984  0.727979 result 4.50354
 training batch 5 mu var00.81459
compute loss for weight  0.727974  0.727979 result 4.50354
   --dy = -0.204564 dy_ref = -0.204564
 training batch 6 mu var00.814589
compute loss for weight  -0.890477  -0.890487 result 4.50355
 training batch 7 mu var00.81459
compute loss for weight  -0.890497  -0.890487 result 4.50354
 training batch 8 mu var00.814589
compute loss for weight  -0.890482  -0.890487 result 4.50354
 training batch 9 mu var00.81459
compute loss for weight  -0.890492  -0.890487 result 4.50354
   --dy = 0.242543 dy_ref = 0.242543
 training batch 10 mu var00.81459
compute loss for weight  0.487352  0.487342 result 4.50355
 training batch 11 mu var00.81459
compute loss for weight  0.487332  0.487342 result 4.50354
 training batch 12 mu var00.81459
compute loss for weight  0.487347  0.487342 result 4.50355
 training batch 13 mu var00.81459
compute loss for weight  0.487337  0.487342 result 4.50354
   --dy = 0.643865 dy_ref = 0.643865
 training batch 14 mu var00.81459
compute loss for weight  0.0353636  0.0353536 result 4.50356
 training batch 15 mu var00.81459
compute loss for weight  0.0353436  0.0353536 result 4.50353
 training batch 16 mu var00.81459
compute loss for weight  0.0353586  0.0353536 result 4.50355
 training batch 17 mu var00.81459
compute loss for weight  0.0353486  0.0353536 result 4.50354
   --dy = 1.44841 dy_ref = 1.44841
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      1.481       7.526 

weights for layer 1

1x2 matrix is as follows

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

 training batch 18 mu var00.81459
compute loss for weight  1.00001  1 result 4.50356
 training batch 19 mu var00.81459
compute loss for weight  0.99999  1 result 4.50353
 training batch 20 mu var00.81459
compute loss for weight  1.00001  1 result 4.50355
 training batch 21 mu var00.81459
compute loss for weight  0.999995  1 result 4.50354
   --dy = 1.48084 dy_ref = 1.48084
 training batch 22 mu var00.81459
compute loss for weight  1.00001  1 result 4.50362
 training batch 23 mu var00.81459
compute loss for weight  0.99999  1 result 4.50347
 training batch 24 mu var00.81459
compute loss for weight  1.00001  1 result 4.50358
 training batch 25 mu var00.81459
compute loss for weight  0.999995  1 result 4.50351
   --dy = 7.52625 dy_ref = 7.52625
Testing weight gradients   for    layer 1
weight gradient for layer 1

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |  5.551e-17  -1.943e-16 

weights for layer 1

1x2 matrix is as follows

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

 training batch 26 mu var00.81459
compute loss for weight  1e-05  0 result 4.50354
 training batch 27 mu var00.81459
compute loss for weight  -1e-05  0 result 4.50354
 training batch 28 mu var00.81459
compute loss for weight  5e-06  0 result 4.50354
 training batch 29 mu var00.81459
compute loss for weight  -5e-06  0 result 4.50354
   --dy = -8.88178e-11 dy_ref = 5.55112e-17
 training batch 30 mu var00.81459
compute loss for weight  1e-05  0 result 4.50354
 training batch 31 mu var00.81459
compute loss for weight  -1e-05  0 result 4.50354
 training batch 32 mu var00.81459
compute loss for weight  5e-06  0 result 4.50354
 training batch 33 mu var00.81459
compute loss for weight  -5e-06  0 result 4.50354
   --dy = 0 dy_ref = -1.94289e-16
Testing weight gradients   for    layer 2
weight gradient for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |      2.171      -4.029 

weights for layer 2

1x2 matrix is as follows

     |      0    |      1    |
-------------------------------
   0 |     0.6822      -1.868 

 training batch 34 mu var00.81459
compute loss for weight  0.682181  0.682171 result 4.50357
 training batch 35 mu var00.81459
compute loss for weight  0.682161  0.682171 result 4.50352
 training batch 36 mu var00.81459
compute loss for weight  0.682176  0.682171 result 4.50355
 training batch 37 mu var00.81459
compute loss for weight  0.682166  0.682171 result 4.50353
   --dy = 2.17077 dy_ref = 2.17077
 training batch 38 mu var00.81459
compute loss for weight  -1.86783  -1.86784 result 4.5035
 training batch 39 mu var00.81459
compute loss for weight  -1.86785  -1.86784 result 4.50358
 training batch 40 mu var00.81459
compute loss for weight  -1.86783  -1.86784 result 4.50352
 training batch 41 mu var00.81459
compute loss for weight  -1.86784  -1.86784 result 4.50356
   --dy = -4.02939 dy_ref = -4.02939
Testing weight gradients:      maximum relative error: [NON-XML-CHAR-0x1B][32m6.30639e-10[NON-XML-CHAR-0x1B][39m