statistics

baseline with momentum:


run:
Iteration: 1000
- For training data
loss function (cross entropy): 16.253588520282676
classification accuracy: 75.0
- For dev
Classification accuracy: 77.60000000000001
- For test
Classification accuracy: 76.7
Iteration: 2000
- For training data
loss function (cross entropy): 8.723658394718838
classification accuracy: 90.0
- For dev
Classification accuracy: 82.8
- For test
Classification accuracy: 82.39999999999999
Iteration: 3000
- For training data
loss function (cross entropy): 42.395302339453224
classification accuracy: 60.0
- For dev
Classification accuracy: 83.36
- For test
Classification accuracy: 82.28
Iteration: 4000
- For training data
loss function (cross entropy): 16.881328336219905
classification accuracy: 85.0
- For dev
Classification accuracy: 85.65
- For test
Classification accuracy: 84.92
Iteration: 5000
- For training data
loss function (cross entropy): 28.52292460972988
classification accuracy: 80.0
- For dev
Classification accuracy: 85.64
- For test
Classification accuracy: 84.47
Iteration: 6000
- For training data
loss function (cross entropy): 35.785970642088806
classification accuracy: 80.0
- For dev
Classification accuracy: 86.68
- For test
Classification accuracy: 86.0
Iteration: 7000
- For training data
loss function (cross entropy): 6.4970644810232265
classification accuracy: 85.0
- For dev
Classification accuracy: 87.07000000000001
- For test
Classification accuracy: 86.37
Iteration: 8000
- For training data
loss function (cross entropy): 1.4240411080412299
classification accuracy: 95.0
- For dev
Classification accuracy: 87.13
- For test
Classification accuracy: 86.71
Iteration: 9000
- For training data
loss function (cross entropy): 0.5715992384615941
classification accuracy: 100.0
- For dev
Classification accuracy: 87.51
- For test
Classification accuracy: 87.55
Iteration: 10000
- For training data
loss function (cross entropy): 3.0640075292723212
classification accuracy: 95.0
- For dev
Classification accuracy: 87.24
- For test
Classification accuracy: 86.58
Iteration: 11000
- For training data
loss function (cross entropy): 8.962425543846042
classification accuracy: 90.0
- For dev
Classification accuracy: 88.14
- For test
Classification accuracy: 87.25
Iteration: 12000
- For training data
loss function (cross entropy): 21.189376618003944
classification accuracy: 85.0
- For dev
Classification accuracy: 87.67
- For test
Classification accuracy: 87.22
Iteration: 13000
- For training data
loss function (cross entropy): 9.50920135101521
classification accuracy: 85.0
- For dev
Classification accuracy: 87.8
- For test
Classification accuracy: 86.97
Iteration: 14000
- For training data
loss function (cross entropy): 14.22118591751031
classification accuracy: 85.0
- For dev
Classification accuracy: 88.25
- For test
Classification accuracy: 87.5
Iteration: 15000
- For training data
loss function (cross entropy): 19.826557305110242
classification accuracy: 80.0
- For dev
Classification accuracy: 88.91
- For test
Classification accuracy: 88.41
Iteration: 16000
- For training data
loss function (cross entropy): 41.37576804319447
classification accuracy: 70.0
- For dev
Classification accuracy: 88.29
- For test
Classification accuracy: 87.7
Iteration: 17000
- For training data
loss function (cross entropy): 4.253381466687117
classification accuracy: 90.0
- For dev
Classification accuracy: 88.6
- For test
Classification accuracy: 88.37
Iteration: 18000
- For training data
loss function (cross entropy): 27.889588457408664
classification accuracy: 85.0
- For dev
Classification accuracy: 88.59
- For test
Classification accuracy: 87.81
Iteration: 19000
- For training data
loss function (cross entropy): 18.07955046606887
classification accuracy: 85.0
- For dev
Classification accuracy: 88.13
- For test
Classification accuracy: 87.86
Iteration: 20000
- For training data
loss function (cross entropy): 1.9292750860433454
classification accuracy: 95.0
- For dev
Classification accuracy: 88.81
- For test
Classification accuracy: 88.37
Iteration: 21000
- For training data
loss function (cross entropy): 18.670185606202114
classification accuracy: 80.0
- For dev
Classification accuracy: 88.81
- For test
Classification accuracy: 88.19
Iteration: 22000
- For training data
loss function (cross entropy): 16.40703095128173
classification accuracy: 75.0
- For dev
Classification accuracy: 88.74
- For test
Classification accuracy: 88.44999999999999
Iteration: 23000
- For training data
loss function (cross entropy): 9.767569146248006
classification accuracy: 90.0
- For dev
Classification accuracy: 89.02
- For test
Classification accuracy: 88.57000000000001
Iteration: 24000
- For training data
loss function (cross entropy): 11.968768641694226
classification accuracy: 85.0
- For dev
Classification accuracy: 87.96000000000001
- For test
Classification accuracy: 87.41
Iteration: 25000
- For training data
loss function (cross entropy): 30.836978008212373
classification accuracy: 95.0
- For dev
Classification accuracy: 88.14
- For test
Classification accuracy: 87.86
Iteration: 26000
- For training data
loss function (cross entropy): 7.24090613136463
classification accuracy: 95.0
- For dev
Classification accuracy: 88.61
- For test
Classification accuracy: 88.42
Iteration: 27000
- For training data
loss function (cross entropy): 3.5655777702877356
classification accuracy: 95.0
- For dev
Classification accuracy: 87.96000000000001
- For test
Classification accuracy: 87.64
Iteration: 28000
- For training data
loss function (cross entropy): 20.066780733570184
classification accuracy: 90.0
- For dev
Classification accuracy: 88.48
- For test
Classification accuracy: 87.58
Iteration: 29000
- For training data
loss function (cross entropy): 10.1302422538356
classification accuracy: 85.0
- For dev
Classification accuracy: 88.81
- For test
Classification accuracy: 87.76
Iteration: 30000
- For training data
loss function (cross entropy): 50.20108276565142
classification accuracy: 75.0
- For dev
Classification accuracy: 88.62
- For test
Classification accuracy: 87.92999999999999
BUILD SUCCESSFUL (total time: 7 minutes 48 seconds)







our ...



run:
Iteration: 1000
- For training data
loss function (cross entropy): 27.395779572983074
classification accuracy: 60.0
- For dev
Classification accuracy: 66.10000000000001
- For test
Classification accuracy: 65.71000000000001
Iteration: 2000
- For training data
loss function (cross entropy): 18.128060489270517
classification accuracy: 70.0
- For dev
Classification accuracy: 75.82
- For test
Classification accuracy: 75.21
Iteration: 3000
- For training data
loss function (cross entropy): 15.981851460711729
classification accuracy: 70.0
- For dev
Classification accuracy: 80.95
- For test
Classification accuracy: 80.21000000000001
Iteration: 4000
- For training data
loss function (cross entropy): 16.763053002558213
classification accuracy: 80.0
- For dev
Classification accuracy: 83.77
- For test
Classification accuracy: 83.23
Iteration: 5000
- For training data
loss function (cross entropy): 10.50691985191854
classification accuracy: 85.0
- For dev
Classification accuracy: 85.44
- For test
Classification accuracy: 84.95
Iteration: 6000
- For training data
loss function (cross entropy): 1.6216220296494208
classification accuracy: 95.0
- For dev
Classification accuracy: 85.36
- For test
Classification accuracy: 84.82
Iteration: 7000
- For training data
loss function (cross entropy): 11.255429328373042
classification accuracy: 85.0
- For dev
Classification accuracy: 86.98
- For test
Classification accuracy: 86.42999999999999
Iteration: 8000
- For training data
loss function (cross entropy): 15.178644454108895
classification accuracy: 80.0
- For dev
Classification accuracy: 87.17
- For test
Classification accuracy: 86.65
Iteration: 9000
- For training data
loss function (cross entropy): 2.8683699506574953
classification accuracy: 95.0
- For dev
Classification accuracy: 87.56
- For test
Classification accuracy: 87.48
Iteration: 10000
- For training data
loss function (cross entropy): 10.504669960170279
classification accuracy: 85.0
- For dev
Classification accuracy: 88.03999999999999
- For test
Classification accuracy: 87.83
Iteration: 11000
- For training data
loss function (cross entropy): 20.406281795172788
classification accuracy: 80.0
- For dev
Classification accuracy: 88.71
- For test
Classification accuracy: 88.06
Iteration: 12000
- For training data
loss function (cross entropy): 4.988916228584073
classification accuracy: 90.0
- For dev
Classification accuracy: 88.53999999999999
- For test
Classification accuracy: 88.19
Iteration: 13000
- For training data
loss function (cross entropy): 10.030566930717486
classification accuracy: 80.0
- For dev
Classification accuracy: 89.03
- For test
Classification accuracy: 88.42
Iteration: 14000
- For training data
loss function (cross entropy): 0.3638699273325693
classification accuracy: 100.0
- For dev
Classification accuracy: 89.13
- For test
Classification accuracy: 88.69
Iteration: 15000
- For training data
loss function (cross entropy): 8.96439323946428
classification accuracy: 90.0
- For dev
Classification accuracy: 89.4
- For test
Classification accuracy: 88.75
Iteration: 16000
- For training data
loss function (cross entropy): 16.496873438423226
classification accuracy: 90.0
- For dev
Classification accuracy: 89.37
- For test
Classification accuracy: 89.23
Iteration: 17000
- For training data
loss function (cross entropy): 22.443768788187292
classification accuracy: 85.0
- For dev
Classification accuracy: 89.38000000000001
- For test
Classification accuracy: 88.66000000000001
Iteration: 18000
- For training data
loss function (cross entropy): 13.671469471757966
classification accuracy: 75.0
- For dev
Classification accuracy: 89.84
- For test
Classification accuracy: 89.37
Iteration: 19000
- For training data
loss function (cross entropy): 12.69506493423111
classification accuracy: 85.0
- For dev
Classification accuracy: 90.2
- For test
Classification accuracy: 89.48
Iteration: 20000
- For training data
loss function (cross entropy): 0.9515018764827561
classification accuracy: 100.0
- For dev
Classification accuracy: 89.96
- For test
Classification accuracy: 89.38000000000001
Iteration: 21000
- For training data
loss function (cross entropy): 17.94899758710934
classification accuracy: 85.0
- For dev
Classification accuracy: 90.25
- For test
Classification accuracy: 89.66
Iteration: 22000
- For training data
loss function (cross entropy): 31.071050991784894
classification accuracy: 70.0
- For dev
Classification accuracy: 90.34
- For test
Classification accuracy: 89.85
Iteration: 23000
- For training data
loss function (cross entropy): 10.965924957882935
classification accuracy: 90.0
- For dev
Classification accuracy: 90.29
- For test
Classification accuracy: 89.89
Iteration: 24000
- For training data
loss function (cross entropy): 7.446215904467733
classification accuracy: 90.0
- For dev
Classification accuracy: 90.38000000000001
- For test
Classification accuracy: 90.14
Iteration: 25000
- For training data
loss function (cross entropy): 16.756611552208888
classification accuracy: 75.0
- For dev
Classification accuracy: 90.29
- For test
Classification accuracy: 90.02
Iteration: 26000
- For training data
loss function (cross entropy): 5.974265672102913
classification accuracy: 95.0
- For dev
Classification accuracy: 90.92
- For test
Classification accuracy: 90.19
Iteration: 27000
- For training data
loss function (cross entropy): 2.4667593858419665
classification accuracy: 95.0
- For dev
Classification accuracy: 90.75
- For test
Classification accuracy: 90.48
Iteration: 28000
- For training data
loss function (cross entropy): 16.474680891945972
classification accuracy: 90.0
- For dev
Classification accuracy: 90.94
- For test
Classification accuracy: 90.34
Iteration: 29000
- For training data
loss function (cross entropy): 6.3080540397678515
classification accuracy: 95.0
- For dev
Classification accuracy: 90.96
- For test
Classification accuracy: 90.5
Iteration: 30000
- For training data
loss function (cross entropy): 25.263500579665873
classification accuracy: 80.0
- For dev
Classification accuracy: 90.96
- For test
Classification accuracy: 90.0
BUILD SUCCESSFUL (total time: 8 minutes 14 seconds)






----baseline part 2



run:
Iteration: 1000
- For training data
loss function (cross entropy): 12.902256420036167
classification accuracy: 80.0
- For dev
Classification accuracy: 79.73
- For test
Classification accuracy: 78.56
Iteration: 2000
- For training data
loss function (cross entropy): 14.79685073978909
classification accuracy: 85.0
- For dev
Classification accuracy: 83.48
- For test
Classification accuracy: 82.34
Iteration: 3000
- For training data
loss function (cross entropy): 12.148209758059933
classification accuracy: 85.0
- For dev
Classification accuracy: 85.11
- For test
Classification accuracy: 84.67
Iteration: 4000
- For training data
loss function (cross entropy): 8.730418295157376
classification accuracy: 90.0
- For dev
Classification accuracy: 85.79
- For test
Classification accuracy: 84.89
Iteration: 5000
- For training data
loss function (cross entropy): 9.866252475944286
classification accuracy: 85.0
- For dev
Classification accuracy: 86.67
- For test
Classification accuracy: 85.65
Iteration: 6000
- For training data
loss function (cross entropy): 32.48734450464802
classification accuracy: 75.0
- For dev
Classification accuracy: 86.29
- For test
Classification accuracy: 85.39999999999999
Iteration: 7000
- For training data
loss function (cross entropy): 7.3956642868050215
classification accuracy: 80.0
- For dev
Classification accuracy: 87.61
- For test
Classification accuracy: 86.11
Iteration: 8000
- For training data
loss function (cross entropy): 26.30566769362767
classification accuracy: 85.0
- For dev
Classification accuracy: 86.74
- For test
Classification accuracy: 85.65
Iteration: 9000
- For training data
loss function (cross entropy): 8.39078910523949
classification accuracy: 95.0
- For dev
Classification accuracy: 87.35000000000001
- For test
Classification accuracy: 86.4
Iteration: 10000
- For training data
loss function (cross entropy): 17.72089472304587
classification accuracy: 85.0
- For dev
Classification accuracy: 87.28
- For test
Classification accuracy: 86.50999999999999
Iteration: 11000
- For training data
loss function (cross entropy): 65.15840277451308
classification accuracy: 65.0
- For dev
Classification accuracy: 87.75
- For test
Classification accuracy: 87.21
Iteration: 12000
- For training data
loss function (cross entropy): 54.93250871250317
classification accuracy: 75.0
- For dev
Classification accuracy: 88.39
- For test
Classification accuracy: 87.33999999999999
Iteration: 13000
- For training data
loss function (cross entropy): 16.323934710590642
classification accuracy: 80.0
- For dev
Classification accuracy: 88.36
- For test
Classification accuracy: 87.06
Iteration: 14000
- For training data
loss function (cross entropy): 6.507589592640616
classification accuracy: 85.0
- For dev
Classification accuracy: 88.36
- For test
Classification accuracy: 87.69
Iteration: 15000
- For training data
loss function (cross entropy): 26.1535649791688
classification accuracy: 85.0
- For dev
Classification accuracy: 88.57000000000001
- For test
Classification accuracy: 87.76
Iteration: 16000
- For training data
loss function (cross entropy): 11.113900628824316
classification accuracy: 90.0
- For dev
Classification accuracy: 88.09
- For test
Classification accuracy: 87.53999999999999
Iteration: 17000
- For training data
loss function (cross entropy): 37.71369886584154
classification accuracy: 85.0
- For dev
Classification accuracy: 86.74
- For test
Classification accuracy: 85.47
Iteration: 18000
- For training data
loss function (cross entropy): 3.9924888734661845
classification accuracy: 95.0
- For dev
Classification accuracy: 87.41
- For test
Classification accuracy: 86.72999999999999
Iteration: 19000
- For training data
loss function (cross entropy): 12.91074685193738
classification accuracy: 85.0
- For dev
Classification accuracy: 88.28
- For test
Classification accuracy: 87.55
Iteration: 20000
- For training data
loss function (cross entropy): 26.31955305612145
classification accuracy: 80.0
- For dev
Classification accuracy: 88.4
- For test
Classification accuracy: 87.45
Iteration: 21000
- For training data
loss function (cross entropy): 10.947068698778509
classification accuracy: 90.0
- For dev
Classification accuracy: 88.02
- For test
Classification accuracy: 87.14
Iteration: 22000
- For training data
loss function (cross entropy): 8.874312440902179
classification accuracy: 90.0
- For dev
Classification accuracy: 88.46000000000001
- For test
Classification accuracy: 87.53
Iteration: 23000
- For training data
loss function (cross entropy): 0.2890543683260612
classification accuracy: 100.0
- For dev
Classification accuracy: 87.63
- For test
Classification accuracy: 86.99
Iteration: 24000
- For training data
loss function (cross entropy): 28.191826456527988
classification accuracy: 95.0
- For dev
Classification accuracy: 87.49
- For test
Classification accuracy: 86.68
Iteration: 25000
- For training data
loss function (cross entropy): 74.59036115293726
classification accuracy: 80.0
- For dev
Classification accuracy: 88.14
- For test
Classification accuracy: 87.14
Iteration: 26000
- For training data
loss function (cross entropy): 27.258641864097527
classification accuracy: 90.0
- For dev
Classification accuracy: 88.52
- For test
Classification accuracy: 87.58
Iteration: 27000
- For training data
loss function (cross entropy): 13.275842742933428
classification accuracy: 85.0
- For dev
Classification accuracy: 88.14999999999999
- For test
Classification accuracy: 87.46000000000001
Iteration: 28000
- For training data
loss function (cross entropy): 11.475335655606752
classification accuracy: 90.0
- For dev
Classification accuracy: 87.87
- For test
Classification accuracy: 87.21
Iteration: 29000
- For training data
loss function (cross entropy): 25.227614069947258
classification accuracy: 90.0
- For dev
Classification accuracy: 88.51
- For test
Classification accuracy: 87.77000000000001
Iteration: 30000
- For training data
loss function (cross entropy): 17.59826019360478
classification accuracy: 85.0
- For dev
Classification accuracy: 87.85
- For test
Classification accuracy: 87.38
BUILD SUCCESSFUL (total time: 7 minutes 21 seconds)





















-----------baseline -----------------


run:
Iteration: 1000
- For training data
loss function (cross entropy): 20.688032776678163
classification accuracy: 80.0
- For dev
Classification accuracy: 75.64999999999999
- For test
Classification accuracy: 74.17
Iteration: 2000
- For training data
loss function (cross entropy): 29.67877407215493
classification accuracy: 75.0
- For dev
Classification accuracy: 82.92
- For test
Classification accuracy: 82.25
Iteration: 3000
- For training data
loss function (cross entropy): 15.586824424622126
classification accuracy: 75.0
- For dev
Classification accuracy: 83.37
- For test
Classification accuracy: 82.78999999999999
Iteration: 4000
- For training data
loss function (cross entropy): 7.526171255441129
classification accuracy: 80.0
- For dev
Classification accuracy: 85.31
- For test
Classification accuracy: 84.77
Iteration: 5000
- For training data
loss function (cross entropy): 10.026711773641924
classification accuracy: 85.0
- For dev
Classification accuracy: 85.28999999999999
- For test
Classification accuracy: 84.28
Iteration: 6000
- For training data
loss function (cross entropy): 21.689228963289224
classification accuracy: 80.0
- For dev
Classification accuracy: 85.33
- For test
Classification accuracy: 84.55
Iteration: 7000
- For training data
loss function (cross entropy): 9.191393669098312
classification accuracy: 85.0
- For dev
Classification accuracy: 84.76
- For test
Classification accuracy: 84.83000000000001
Iteration: 8000
- For training data
loss function (cross entropy): 19.372275059759694
classification accuracy: 85.0
- For dev
Classification accuracy: 85.68
- For test
Classification accuracy: 85.66
Iteration: 9000
- For training data
loss function (cross entropy): 24.7160599936386
classification accuracy: 75.0
- For dev
Classification accuracy: 85.25
- For test
Classification accuracy: 85.33
Iteration: 10000
- For training data
loss function (cross entropy): 10.476185690343943
classification accuracy: 80.0
- For dev
Classification accuracy: 85.72999999999999
- For test
Classification accuracy: 85.61
Iteration: 11000
- For training data
loss function (cross entropy): 2.5054501138906775
classification accuracy: 95.0
- For dev
Classification accuracy: 86.24000000000001
- For test
Classification accuracy: 86.11999999999999
Iteration: 12000
- For training data
loss function (cross entropy): 12.11245984966836
classification accuracy: 90.0
- For dev
Classification accuracy: 87.2
- For test
Classification accuracy: 86.71
Iteration: 13000
- For training data
loss function (cross entropy): 28.091383309933224
classification accuracy: 80.0
- For dev
Classification accuracy: 87.22999999999999
- For test
Classification accuracy: 86.15
Iteration: 14000
- For training data
loss function (cross entropy): 25.62558027656708
classification accuracy: 80.0
- For dev
Classification accuracy: 86.47
- For test
Classification accuracy: 85.17
Iteration: 15000
- For training data
loss function (cross entropy): 12.825298328329591
classification accuracy: 85.0
- For dev
Classification accuracy: 86.77
- For test
Classification accuracy: 86.4
Iteration: 16000
- For training data
loss function (cross entropy): 23.71586203742414
classification accuracy: 90.0
- For dev
Classification accuracy: 87.61
- For test
Classification accuracy: 87.06
Iteration: 17000
- For training data
loss function (cross entropy): 1.5596315127792824
classification accuracy: 95.0
- For dev
Classification accuracy: 85.99
- For test
Classification accuracy: 85.61
Iteration: 18000
- For training data
loss function (cross entropy): 10.9912470832097
classification accuracy: 85.0
- For dev
Classification accuracy: 86.28
- For test
Classification accuracy: 84.85000000000001
Iteration: 19000
- For training data
loss function (cross entropy): 7.431006022959574
classification accuracy: 90.0
- For dev
Classification accuracy: 87.32
- For test
Classification accuracy: 86.50999999999999
Iteration: 20000
- For training data
loss function (cross entropy): 1.4422320193826335
classification accuracy: 95.0
- For dev
Classification accuracy: 87.6
- For test
Classification accuracy: 87.12
Iteration: 21000
- For training data
loss function (cross entropy): 29.772624863569153
classification accuracy: 85.0
- For dev
Classification accuracy: 87.2
- For test
Classification accuracy: 86.83999999999999
Iteration: 22000
- For training data
loss function (cross entropy): 26.18918640553418
classification accuracy: 75.0
- For dev
Classification accuracy: 87.06
- For test
Classification accuracy: 86.1
Iteration: 23000
- For training data
loss function (cross entropy): 0.3259996316320848
classification accuracy: 100.0
- For dev
Classification accuracy: 88.24
- For test
Classification accuracy: 87.29
Iteration: 24000
- For training data
loss function (cross entropy): 34.81648304788943
classification accuracy: 80.0
- For dev
Classification accuracy: 87.46000000000001
- For test
Classification accuracy: 86.61
Iteration: 25000
- For training data
loss function (cross entropy): 15.135875309613596
classification accuracy: 95.0
- For dev
Classification accuracy: 87.44
- For test
Classification accuracy: 87.4
Iteration: 26000
- For training data
loss function (cross entropy): 7.3022177120915295
classification accuracy: 90.0
- For dev
Classification accuracy: 87.13
- For test
Classification accuracy: 87.22999999999999
Iteration: 27000
- For training data
loss function (cross entropy): 12.51571363681879
classification accuracy: 90.0
- For dev
Classification accuracy: 88.34
- For test
Classification accuracy: 87.72
Iteration: 28000
- For training data
loss function (cross entropy): 5.60592400089843
classification accuracy: 85.0
- For dev
Classification accuracy: 88.4
- For test
Classification accuracy: 87.83999999999999
Iteration: 29000
- For training data
loss function (cross entropy): 12.46117488691281
classification accuracy: 90.0
- For dev
Classification accuracy: 88.57000000000001
- For test
Classification accuracy: 87.7
Iteration: 30000
- For training data
loss function (cross entropy): 27.163382563833057
classification accuracy: 75.0
- For dev
Classification accuracy: 87.94999999999999
- For test
Classification accuracy: 87.16000000000001
BUILD SUCCESSFUL (total time: 7 minutes 58 seconds)




----our 


run:
Iteration: 1000
- For training data
loss function (cross entropy): 22.08602157828443
classification accuracy: 75.0
- For dev
Classification accuracy: 64.23
- For test
Classification accuracy: 62.970000000000006
Iteration: 2000
- For training data
loss function (cross entropy): 14.9946307650128
classification accuracy: 70.0
- For dev
Classification accuracy: 76.83
- For test
Classification accuracy: 75.25
Iteration: 3000
- For training data
loss function (cross entropy): 10.084852839918174
classification accuracy: 80.0
- For dev
Classification accuracy: 80.03
- For test
Classification accuracy: 78.78
Iteration: 4000
- For training data
loss function (cross entropy): 22.338626632396586
classification accuracy: 85.0
- For dev
Classification accuracy: 83.37
- For test
Classification accuracy: 82.02000000000001
Iteration: 5000
- For training data
loss function (cross entropy): 6.8827655601747075
classification accuracy: 95.0
- For dev
Classification accuracy: 84.39
- For test
Classification accuracy: 83.34
Iteration: 6000
- For training data
loss function (cross entropy): 4.187136260098993
classification accuracy: 85.0
- For dev
Classification accuracy: 85.72
- For test
Classification accuracy: 84.63000000000001
Iteration: 7000
- For training data
loss function (cross entropy): 9.524923875103026
classification accuracy: 80.0
- For dev
Classification accuracy: 86.41
- For test
Classification accuracy: 85.61999999999999
Iteration: 8000
- For training data
loss function (cross entropy): 13.150365359076762
classification accuracy: 90.0
- For dev
Classification accuracy: 87.18
- For test
Classification accuracy: 86.72
Iteration: 9000
- For training data
loss function (cross entropy): 4.6576102521873635
classification accuracy: 95.0
- For dev
Classification accuracy: 87.64999999999999
- For test
Classification accuracy: 87.05000000000001
Iteration: 10000
- For training data
loss function (cross entropy): 9.659942714283023
classification accuracy: 95.0
- For dev
Classification accuracy: 88.5
- For test
Classification accuracy: 87.94999999999999
Iteration: 11000
- For training data
loss function (cross entropy): 12.107443378152997
classification accuracy: 90.0
- For dev
Classification accuracy: 88.31
- For test
Classification accuracy: 87.78
Iteration: 12000
- For training data
loss function (cross entropy): 24.34838937152884
classification accuracy: 70.0
- For dev
Classification accuracy: 88.78
- For test
Classification accuracy: 88.4
Iteration: 13000
- For training data
loss function (cross entropy): 9.48833640987718
classification accuracy: 90.0
- For dev
Classification accuracy: 89.05
- For test
Classification accuracy: 88.71
Iteration: 14000
- For training data
loss function (cross entropy): 4.807193206341512
classification accuracy: 95.0
- For dev
Classification accuracy: 89.18
- For test
Classification accuracy: 89.1
Iteration: 15000
- For training data
loss function (cross entropy): 10.054251935132367
classification accuracy: 95.0
- For dev
Classification accuracy: 89.4
- For test
Classification accuracy: 89.16
Iteration: 16000
- For training data
loss function (cross entropy): 24.50675072061591
classification accuracy: 85.0
- For dev
Classification accuracy: 89.60000000000001
- For test
Classification accuracy: 89.25
Iteration: 17000
- For training data
loss function (cross entropy): 22.941410014116062
classification accuracy: 85.0
- For dev
Classification accuracy: 89.62
- For test
Classification accuracy: 89.42999999999999
Iteration: 18000
- For training data
loss function (cross entropy): 3.1173602427817784
classification accuracy: 95.0
- For dev
Classification accuracy: 90.13
- For test
Classification accuracy: 89.56
Iteration: 19000
- For training data
loss function (cross entropy): 25.67684171098444
classification accuracy: 90.0
- For dev
Classification accuracy: 89.92
- For test
Classification accuracy: 89.7
Iteration: 20000
- For training data
loss function (cross entropy): 25.56728459235009
classification accuracy: 80.0
- For dev
Classification accuracy: 90.38000000000001
- For test
Classification accuracy: 90.16999999999999
Iteration: 21000
- For training data
loss function (cross entropy): 14.322819533335927
classification accuracy: 85.0
- For dev
Classification accuracy: 90.44
- For test
Classification accuracy: 90.18
Iteration: 22000
- For training data
loss function (cross entropy): 4.065657125703203
classification accuracy: 90.0
- For dev
Classification accuracy: 90.27
- For test
Classification accuracy: 90.24
Iteration: 23000
- For training data
loss function (cross entropy): 4.106227985089191
classification accuracy: 95.0
- For dev
Classification accuracy: 90.55
- For test
Classification accuracy: 90.42999999999999
Iteration: 24000
- For training data
loss function (cross entropy): 18.262258910432408
classification accuracy: 90.0
- For dev
Classification accuracy: 90.60000000000001
- For test
Classification accuracy: 90.5
Iteration: 25000
- For training data
loss function (cross entropy): 38.56961916585929
classification accuracy: 90.0
- For dev
Classification accuracy: 90.53999999999999
- For test
Classification accuracy: 90.14
Iteration: 26000
- For training data
loss function (cross entropy): 24.44727628392236
classification accuracy: 80.0
- For dev
Classification accuracy: 90.84
- For test
Classification accuracy: 90.57
Iteration: 27000
- For training data
loss function (cross entropy): 11.622123987412271
classification accuracy: 90.0
- For dev
Classification accuracy: 91.03
- For test
Classification accuracy: 90.45
Iteration: 28000
- For training data
loss function (cross entropy): 15.73396907559002
classification accuracy: 90.0
- For dev
Classification accuracy: 90.9
- For test
Classification accuracy: 90.73
Iteration: 29000
- For training data
loss function (cross entropy): 46.149505100285666
classification accuracy: 80.0
- For dev
Classification accuracy: 90.96
- For test
Classification accuracy: 90.71000000000001
Iteration: 30000
- For training data
loss function (cross entropy): 21.704330388382726
classification accuracy: 90.0
- For dev
Classification accuracy: 91.13
- For test
Classification accuracy: 90.75999999999999
BUILD SUCCESSFUL (total time: 7 minutes 53 seconds)