python train.py simus/ep/cnn/xpvgg4_cif10.json

Input   :	 (32, 32, 3)
Layer 1 :	 (16, 16, 128)
Layer 2 :	 (8, 8, 256)
Layer 3 :	 (4, 4, 512)
Layer 4 :	 (1, 1, 512)

Mean log L2 = -4.754443645477295
Total dropped = 0
param norm = 33.0522575378418

Epoch 1 in 596.96 sec
Train accuracy :	0.299	(14971/50000)		Train loss :	1.874
Val accuracy   :	0.403	(4025/10000)		Val loss :	1.5884
Top-5 val acc  :	0.901	(9012/10000)


Mean log L2 = -4.61715841293335
Total dropped = 0
param norm = 34.496604919433594

Epoch 2 in 230.73 sec
Train accuracy :	0.483	(24145/50000)		Train loss :	1.4144
Val accuracy   :	0.56	(5600/10000)		Val loss :	1.2058
Top-5 val acc  :	0.954	(9542/10000)


Mean log L2 = -4.371346950531006
Total dropped = 0
param norm = 36.52159118652344

Epoch 3 in 230.41 sec
Train accuracy :	0.587	(29370/50000)		Train loss :	1.1615
Val accuracy   :	0.63	(6302/10000)		Val loss :	1.0478
Top-5 val acc  :	0.964	(9637/10000)


Mean log L2 = -3.9359657764434814
Total dropped = 0
param norm = 38.39393997192383

Epoch 4 in 230.39 sec
Train accuracy :	0.652	(32582/50000)		Train loss :	0.995
Val accuracy   :	0.623	(6233/10000)		Val loss :	1.0534
Top-5 val acc  :	0.957	(9569/10000)


Mean log L2 = -3.8817646503448486
Total dropped = 0
param norm = 40.049495697021484

Epoch 5 in 230.40 sec
Train accuracy :	0.692	(34585/50000)		Train loss :	0.8866
Val accuracy   :	0.702	(7023/10000)		Val loss :	0.8684
Top-5 val acc  :	0.974	(9735/10000)


Mean log L2 = -3.8019683361053467
Total dropped = 0
param norm = 41.542633056640625

Epoch 6 in 230.34 sec
Train accuracy :	0.719	(35950/50000)		Train loss :	0.8164
Val accuracy   :	0.728	(7283/10000)		Val loss :	0.7925
Top-5 val acc  :	0.978	(9775/10000)


Mean log L2 = -3.3474748134613037
Total dropped = 0
param norm = 42.67722702026367

Epoch 7 in 230.38 sec
Train accuracy :	0.74	(36993/50000)		Train loss :	0.7644
Val accuracy   :	0.735	(7351/10000)		Val loss :	0.7681
Top-5 val acc  :	0.982	(9817/10000)


Mean log L2 = -3.461397886276245
Total dropped = 0
param norm = 43.72172164916992

Epoch 8 in 230.30 sec
Train accuracy :	0.754	(37693/50000)		Train loss :	0.718
Val accuracy   :	0.745	(7449/10000)		Val loss :	0.7318
Top-5 val acc  :	0.984	(9836/10000)


Mean log L2 = -3.615551710128784
Total dropped = 0
param norm = 44.482810974121094

Epoch 9 in 230.26 sec
Train accuracy :	0.764	(38191/50000)		Train loss :	0.6879
Val accuracy   :	0.76	(7600/10000)		Val loss :	0.6894
Top-5 val acc  :	0.985	(9847/10000)


Mean log L2 = -3.437678098678589
Total dropped = 0
param norm = 45.31754684448242

Epoch 10 in 230.74 sec
Train accuracy :	0.773	(38642/50000)		Train loss :	0.667
Val accuracy   :	0.767	(7673/10000)		Val loss :	0.68
Top-5 val acc  :	0.985	(9847/10000)


Mean log L2 = -3.437929630279541
Total dropped = 0
param norm = 46.10272216796875

Epoch 11 in 230.42 sec
Train accuracy :	0.784	(39198/50000)		Train loss :	0.6348
Val accuracy   :	0.769	(7695/10000)		Val loss :	0.6695
Top-5 val acc  :	0.985	(9854/10000)


Mean log L2 = -3.353205442428589
Total dropped = 0
param norm = 46.893131256103516

Epoch 12 in 230.30 sec
Train accuracy :	0.789	(39456/50000)		Train loss :	0.6211
Val accuracy   :	0.769	(7687/10000)		Val loss :	0.7026
Top-5 val acc  :	0.985	(9845/10000)


Mean log L2 = -3.1555869579315186
Total dropped = 0
param norm = 47.381858825683594

Epoch 13 in 230.42 sec
Train accuracy :	0.797	(39844/50000)		Train loss :	0.5935
Val accuracy   :	0.786	(7860/10000)		Val loss :	0.6235
Top-5 val acc  :	0.988	(9879/10000)


Mean log L2 = -3.230078935623169
Total dropped = 0
param norm = 47.799251556396484

Epoch 14 in 230.34 sec
Train accuracy :	0.807	(40335/50000)		Train loss :	0.5724
Val accuracy   :	0.799	(7988/10000)		Val loss :	0.594
Top-5 val acc  :	0.987	(9873/10000)


Mean log L2 = -3.1174156665802
Total dropped = 0
param norm = 48.22359085083008

Epoch 15 in 230.35 sec
Train accuracy :	0.813	(40659/50000)		Train loss :	0.5565
Val accuracy   :	0.789	(7887/10000)		Val loss :	0.6222
Top-5 val acc  :	0.986	(9862/10000)


Mean log L2 = -3.074605703353882
Total dropped = 0
param norm = 48.769195556640625

Epoch 16 in 230.25 sec
Train accuracy :	0.816	(40781/50000)		Train loss :	0.5445
Val accuracy   :	0.795	(7951/10000)		Val loss :	0.6013
Top-5 val acc  :	0.989	(9885/10000)


Mean log L2 = -2.871687889099121
Total dropped = 0
param norm = 49.210140228271484

Epoch 17 in 230.38 sec
Train accuracy :	0.818	(40916/50000)		Train loss :	0.5362
Val accuracy   :	0.779	(7792/10000)		Val loss :	0.6425
Top-5 val acc  :	0.985	(9848/10000)


Mean log L2 = -3.367900848388672
Total dropped = 0
param norm = 49.607051849365234

Epoch 18 in 230.45 sec
Train accuracy :	0.826	(41287/50000)		Train loss :	0.5158
Val accuracy   :	0.794	(7944/10000)		Val loss :	0.6038
Top-5 val acc  :	0.989	(9886/10000)


Mean log L2 = -3.1582114696502686
Total dropped = 0
param norm = 50.047393798828125

Epoch 19 in 230.32 sec
Train accuracy :	0.827	(41357/50000)		Train loss :	0.5112
Val accuracy   :	0.808	(8077/10000)		Val loss :	0.5733
Top-5 val acc  :	0.989	(9889/10000)


Mean log L2 = -2.7811455726623535
Total dropped = 0
param norm = 50.2242317199707

Epoch 20 in 230.18 sec
Train accuracy :	0.835	(41741/50000)		Train loss :	0.4857
Val accuracy   :	0.789	(7893/10000)		Val loss :	0.615
Top-5 val acc  :	0.987	(9869/10000)


Mean log L2 = -3.3323090076446533
Total dropped = 0
param norm = 50.527809143066406

Epoch 21 in 230.33 sec
Train accuracy :	0.836	(41783/50000)		Train loss :	0.4831
Val accuracy   :	0.799	(7989/10000)		Val loss :	0.5995
Top-5 val acc  :	0.989	(9887/10000)


Mean log L2 = -3.256892442703247
Total dropped = 0
param norm = 50.612144470214844

Epoch 22 in 230.13 sec
Train accuracy :	0.844	(42182/50000)		Train loss :	0.4636
Val accuracy   :	0.814	(8138/10000)		Val loss :	0.5477
Top-5 val acc  :	0.988	(9879/10000)


Mean log L2 = -2.9336516857147217
Total dropped = 0
param norm = 50.851409912109375

Epoch 23 in 230.37 sec
Train accuracy :	0.846	(42286/50000)		Train loss :	0.4623
Val accuracy   :	0.795	(7950/10000)		Val loss :	0.6069
Top-5 val acc  :	0.989	(9886/10000)


Mean log L2 = -3.283986806869507
Total dropped = 0
param norm = 51.000282287597656

Epoch 24 in 230.18 sec
Train accuracy :	0.85	(42512/50000)		Train loss :	0.4482
Val accuracy   :	0.809	(8095/10000)		Val loss :	0.5785
Top-5 val acc  :	0.989	(9888/10000)


Mean log L2 = -3.2005937099456787
Total dropped = 0
param norm = 51.234500885009766

Epoch 25 in 230.37 sec
Train accuracy :	0.85	(42523/50000)		Train loss :	0.4463
Val accuracy   :	0.816	(8159/10000)		Val loss :	0.5514
Top-5 val acc  :	0.989	(9887/10000)


Mean log L2 = -3.193877935409546
Total dropped = 0
param norm = 51.36049270629883

Epoch 26 in 230.25 sec
Train accuracy :	0.856	(42778/50000)		Train loss :	0.4265
Val accuracy   :	0.82	(8199/10000)		Val loss :	0.5287
Top-5 val acc  :	0.99	(9897/10000)


Mean log L2 = -3.425182819366455
Total dropped = 0
param norm = 51.4461669921875

Epoch 27 in 230.46 sec
Train accuracy :	0.862	(43095/50000)		Train loss :	0.4131
Val accuracy   :	0.812	(8118/10000)		Val loss :	0.5621
Top-5 val acc  :	0.99	(9895/10000)


Mean log L2 = -3.3118813037872314
Total dropped = 0
param norm = 51.573974609375

Epoch 28 in 230.35 sec
Train accuracy :	0.865	(43240/50000)		Train loss :	0.4074
Val accuracy   :	0.789	(7886/10000)		Val loss :	0.6395
Top-5 val acc  :	0.988	(9881/10000)


Mean log L2 = -3.024928092956543
Total dropped = 0
param norm = 51.5788688659668

Epoch 29 in 230.37 sec
Train accuracy :	0.863	(43173/50000)		Train loss :	0.4025
Val accuracy   :	0.812	(8124/10000)		Val loss :	0.5615
Top-5 val acc  :	0.99	(9903/10000)


Mean log L2 = -2.887359380722046
Total dropped = 0
param norm = 51.587791442871094

Epoch 30 in 230.37 sec
Train accuracy :	0.869	(43437/50000)		Train loss :	0.3944
Val accuracy   :	0.834	(8342/10000)		Val loss :	0.4896
Top-5 val acc  :	0.992	(9915/10000)


Mean log L2 = -3.12064790725708
Total dropped = 0
param norm = 51.609947204589844

Epoch 31 in 230.36 sec
Train accuracy :	0.872	(43608/50000)		Train loss :	0.3831
Val accuracy   :	0.823	(8232/10000)		Val loss :	0.528
Top-5 val acc  :	0.989	(9894/10000)


Mean log L2 = -3.125314474105835
Total dropped = 0
param norm = 51.59976577758789

Epoch 32 in 230.19 sec
Train accuracy :	0.878	(43906/50000)		Train loss :	0.3683
Val accuracy   :	0.835	(8348/10000)		Val loss :	0.4973
Top-5 val acc  :	0.992	(9923/10000)


Mean log L2 = -3.3872292041778564
Total dropped = 0
param norm = 51.556739807128906

Epoch 33 in 230.60 sec
Train accuracy :	0.881	(44056/50000)		Train loss :	0.3594
Val accuracy   :	0.828	(8281/10000)		Val loss :	0.5157
Top-5 val acc  :	0.99	(9901/10000)


Mean log L2 = -3.0777509212493896
Total dropped = 0
param norm = 51.472450256347656

Epoch 34 in 230.29 sec
Train accuracy :	0.88	(44004/50000)		Train loss :	0.3565
Val accuracy   :	0.837	(8367/10000)		Val loss :	0.4937
Top-5 val acc  :	0.993	(9925/10000)


Mean log L2 = -3.170912742614746
Total dropped = 0
param norm = 51.4102783203125

Epoch 35 in 230.23 sec
Train accuracy :	0.886	(44305/50000)		Train loss :	0.3435
Val accuracy   :	0.827	(8269/10000)		Val loss :	0.5053
Top-5 val acc  :	0.99	(9901/10000)


Mean log L2 = -3.3517251014709473
Total dropped = 0
param norm = 51.327816009521484

Epoch 36 in 230.20 sec
Train accuracy :	0.891	(44554/50000)		Train loss :	0.3288
Val accuracy   :	0.835	(8354/10000)		Val loss :	0.5007
Top-5 val acc  :	0.99	(9899/10000)


Mean log L2 = -3.4436745643615723
Total dropped = 0
param norm = 51.24701690673828

Epoch 37 in 230.41 sec
Train accuracy :	0.893	(44630/50000)		Train loss :	0.3251
Val accuracy   :	0.841	(8405/10000)		Val loss :	0.4847
Top-5 val acc  :	0.991	(9912/10000)


Mean log L2 = -3.246882200241089
Total dropped = 0
param norm = 51.178802490234375

Epoch 38 in 230.47 sec
Train accuracy :	0.894	(44710/50000)		Train loss :	0.3204
Val accuracy   :	0.846	(8455/10000)		Val loss :	0.4752
Top-5 val acc  :	0.992	(9918/10000)


Mean log L2 = -3.3895480632781982
Total dropped = 0
param norm = 51.00271987915039

Epoch 39 in 230.54 sec
Train accuracy :	0.9	(44976/50000)		Train loss :	0.3068
Val accuracy   :	0.838	(8377/10000)		Val loss :	0.5109
Top-5 val acc  :	0.99	(9902/10000)


Mean log L2 = -3.3426787853240967
Total dropped = 0
param norm = 50.895267486572266

Epoch 40 in 230.37 sec
Train accuracy :	0.902	(45094/50000)		Train loss :	0.2999
Val accuracy   :	0.836	(8363/10000)		Val loss :	0.5023
Top-5 val acc  :	0.99	(9902/10000)


Mean log L2 = -3.323812246322632
Total dropped = 0
param norm = 50.757110595703125

Epoch 41 in 230.50 sec
Train accuracy :	0.906	(45286/50000)		Train loss :	0.2887
Val accuracy   :	0.841	(8410/10000)		Val loss :	0.5038
Top-5 val acc  :	0.991	(9913/10000)


Mean log L2 = -3.4337220191955566
Total dropped = 0
param norm = 50.59711456298828

Epoch 42 in 230.14 sec
Train accuracy :	0.908	(45421/50000)		Train loss :	0.2805
Val accuracy   :	0.845	(8447/10000)		Val loss :	0.4771
Top-5 val acc  :	0.991	(9908/10000)


Mean log L2 = -3.3813302516937256
Total dropped = 0
param norm = 50.44116973876953

Epoch 43 in 230.19 sec
Train accuracy :	0.91	(45522/50000)		Train loss :	0.2766
Val accuracy   :	0.846	(8462/10000)		Val loss :	0.4695
Top-5 val acc  :	0.992	(9915/10000)


Mean log L2 = -3.4886136054992676
Total dropped = 0
param norm = 50.326053619384766

Epoch 44 in 230.36 sec
Train accuracy :	0.912	(45591/50000)		Train loss :	0.2699
Val accuracy   :	0.846	(8461/10000)		Val loss :	0.4757
Top-5 val acc  :	0.992	(9917/10000)


Mean log L2 = -3.419174909591675
Total dropped = 0
param norm = 50.188819885253906

Epoch 45 in 230.14 sec
Train accuracy :	0.916	(45776/50000)		Train loss :	0.2622
Val accuracy   :	0.835	(8351/10000)		Val loss :	0.5223
Top-5 val acc  :	0.992	(9915/10000)


Mean log L2 = -3.5869851112365723
Total dropped = 0
param norm = 49.97089767456055

Epoch 46 in 230.31 sec
Train accuracy :	0.919	(45955/50000)		Train loss :	0.2507
Val accuracy   :	0.86	(8597/10000)		Val loss :	0.4482
Top-5 val acc  :	0.992	(9922/10000)


Mean log L2 = -3.515195608139038
Total dropped = 0
param norm = 49.75188064575195

Epoch 47 in 230.39 sec
Train accuracy :	0.922	(46109/50000)		Train loss :	0.2392
Val accuracy   :	0.854	(8541/10000)		Val loss :	0.4731
Top-5 val acc  :	0.993	(9931/10000)


Mean log L2 = -3.244732618331909
Total dropped = 0
param norm = 49.46633529663086

Epoch 48 in 230.27 sec
Train accuracy :	0.926	(46291/50000)		Train loss :	0.2287
Val accuracy   :	0.856	(8563/10000)		Val loss :	0.445
Top-5 val acc  :	0.994	(9941/10000)


Mean log L2 = -3.4304234981536865
Total dropped = 0
param norm = 49.23505783081055

Epoch 49 in 230.42 sec
Train accuracy :	0.928	(46421/50000)		Train loss :	0.2235
Val accuracy   :	0.856	(8560/10000)		Val loss :	0.4378
Top-5 val acc  :	0.994	(9939/10000)


Mean log L2 = -3.345200777053833
Total dropped = 0
param norm = 49.0129280090332

Epoch 50 in 230.59 sec
Train accuracy :	0.931	(46567/50000)		Train loss :	0.2167
Val accuracy   :	0.853	(8526/10000)		Val loss :	0.465
Top-5 val acc  :	0.992	(9921/10000)


Mean log L2 = -3.5493741035461426
Total dropped = 0
param norm = 48.80921173095703

Epoch 51 in 230.33 sec
Train accuracy :	0.934	(46712/50000)		Train loss :	0.208
Val accuracy   :	0.851	(8509/10000)		Val loss :	0.4592
Top-5 val acc  :	0.993	(9932/10000)


Mean log L2 = -3.5869882106781006
Total dropped = 0
param norm = 48.53509521484375

Epoch 52 in 230.33 sec
Train accuracy :	0.937	(46848/50000)		Train loss :	0.2004
Val accuracy   :	0.858	(8576/10000)		Val loss :	0.4417
Top-5 val acc  :	0.993	(9930/10000)


Mean log L2 = -3.677743673324585
Total dropped = 0
param norm = 48.32620620727539

Epoch 53 in 230.30 sec
Train accuracy :	0.939	(46950/50000)		Train loss :	0.1931
Val accuracy   :	0.857	(8570/10000)		Val loss :	0.455
Top-5 val acc  :	0.993	(9932/10000)


Mean log L2 = -3.7251365184783936
Total dropped = 0
param norm = 48.07477569580078

Epoch 54 in 230.20 sec
Train accuracy :	0.941	(47065/50000)		Train loss :	0.1859
Val accuracy   :	0.854	(8537/10000)		Val loss :	0.4693
Top-5 val acc  :	0.994	(9935/10000)


Mean log L2 = -3.691392660140991
Total dropped = 0
param norm = 47.76181411743164

Epoch 55 in 230.39 sec
Train accuracy :	0.945	(47274/50000)		Train loss :	0.1744
Val accuracy   :	0.845	(8452/10000)		Val loss :	0.4843
Top-5 val acc  :	0.992	(9918/10000)


Mean log L2 = -3.6250054836273193
Total dropped = 0
param norm = 47.487762451171875

Epoch 56 in 230.27 sec
Train accuracy :	0.947	(47369/50000)		Train loss :	0.1678
Val accuracy   :	0.859	(8594/10000)		Val loss :	0.4379
Top-5 val acc  :	0.993	(9927/10000)


Mean log L2 = -3.769522190093994
Total dropped = 0
param norm = 47.22748565673828

Epoch 57 in 230.42 sec
Train accuracy :	0.95	(47491/50000)		Train loss :	0.1601
Val accuracy   :	0.86	(8597/10000)		Val loss :	0.4644
Top-5 val acc  :	0.993	(9931/10000)


Mean log L2 = -3.5286366939544678
Total dropped = 0
param norm = 46.9337158203125

Epoch 58 in 230.28 sec
Train accuracy :	0.954	(47709/50000)		Train loss :	0.1493
Val accuracy   :	0.869	(8688/10000)		Val loss :	0.4315
Top-5 val acc  :	0.994	(9940/10000)


Mean log L2 = -3.6493728160858154
Total dropped = 0
param norm = 46.659812927246094

Epoch 59 in 230.41 sec
Train accuracy :	0.956	(47802/50000)		Train loss :	0.1452
Val accuracy   :	0.856	(8564/10000)		Val loss :	0.4682
Top-5 val acc  :	0.993	(9927/10000)


Mean log L2 = -3.451794385910034
Total dropped = 0
param norm = 46.3692626953125

Epoch 60 in 230.24 sec
Train accuracy :	0.959	(47934/50000)		Train loss :	0.1396
Val accuracy   :	0.863	(8627/10000)		Val loss :	0.4736
Top-5 val acc  :	0.993	(9926/10000)


Mean log L2 = -3.5502402782440186
Total dropped = 0
param norm = 46.08327865600586

Epoch 61 in 230.57 sec
Train accuracy :	0.962	(48088/50000)		Train loss :	0.1298
Val accuracy   :	0.872	(8722/10000)		Val loss :	0.4371
Top-5 val acc  :	0.993	(9933/10000)


Mean log L2 = -3.445631742477417
Total dropped = 0
param norm = 45.81013107299805

Epoch 62 in 230.56 sec
Train accuracy :	0.963	(48151/50000)		Train loss :	0.124
Val accuracy   :	0.862	(8624/10000)		Val loss :	0.4793
Top-5 val acc  :	0.993	(9925/10000)


Mean log L2 = -3.6677708625793457
Total dropped = 0
param norm = 45.549625396728516

Epoch 63 in 230.47 sec
Train accuracy :	0.964	(48195/50000)		Train loss :	0.1195
Val accuracy   :	0.868	(8677/10000)		Val loss :	0.4633
Top-5 val acc  :	0.993	(9934/10000)


Mean log L2 = -3.8268704414367676
Total dropped = 0
param norm = 45.29446792602539

Epoch 64 in 230.72 sec
Train accuracy :	0.965	(48246/50000)		Train loss :	0.1192
Val accuracy   :	0.868	(8681/10000)		Val loss :	0.4537
Top-5 val acc  :	0.993	(9929/10000)


Mean log L2 = -3.5689148902893066
Total dropped = 0
param norm = 45.04667663574219

Epoch 65 in 230.58 sec
Train accuracy :	0.969	(48470/50000)		Train loss :	0.1058
Val accuracy   :	0.871	(8712/10000)		Val loss :	0.4522
Top-5 val acc  :	0.994	(9943/10000)


Mean log L2 = -3.7777938842773438
Total dropped = 0
param norm = 44.77680206298828

Epoch 66 in 230.19 sec
Train accuracy :	0.973	(48672/50000)		Train loss :	0.0946
Val accuracy   :	0.871	(8710/10000)		Val loss :	0.463
Top-5 val acc  :	0.994	(9937/10000)


Mean log L2 = -3.6637864112854004
Total dropped = 0
param norm = 44.51341247558594

Epoch 67 in 230.38 sec
Train accuracy :	0.973	(48632/50000)		Train loss :	0.0944
Val accuracy   :	0.874	(8739/10000)		Val loss :	0.447
Top-5 val acc  :	0.994	(9943/10000)


Mean log L2 = -3.6260292530059814
Total dropped = 0
param norm = 44.26926040649414

Epoch 68 in 230.39 sec
Train accuracy :	0.976	(48784/50000)		Train loss :	0.0869
Val accuracy   :	0.876	(8760/10000)		Val loss :	0.4571
Top-5 val acc  :	0.995	(9946/10000)


Mean log L2 = -3.711087465286255
Total dropped = 0
param norm = 44.02077865600586

Epoch 69 in 230.35 sec
Train accuracy :	0.976	(48819/50000)		Train loss :	0.0836
Val accuracy   :	0.875	(8754/10000)		Val loss :	0.4475
Top-5 val acc  :	0.994	(9942/10000)


Mean log L2 = -3.737269639968872
Total dropped = 0
param norm = 43.78630828857422

Epoch 70 in 230.38 sec
Train accuracy :	0.978	(48915/50000)		Train loss :	0.0789
Val accuracy   :	0.877	(8775/10000)		Val loss :	0.4415
Top-5 val acc  :	0.995	(9952/10000)


Mean log L2 = -3.5619118213653564
Total dropped = 0
param norm = 43.551055908203125

Epoch 71 in 230.27 sec
Train accuracy :	0.98	(49022/50000)		Train loss :	0.0705
Val accuracy   :	0.871	(8711/10000)		Val loss :	0.4787
Top-5 val acc  :	0.994	(9936/10000)


Mean log L2 = -3.518988847732544
Total dropped = 0
param norm = 43.3398323059082

Epoch 72 in 230.35 sec
Train accuracy :	0.982	(49078/50000)		Train loss :	0.0675
Val accuracy   :	0.878	(8781/10000)		Val loss :	0.4651
Top-5 val acc  :	0.993	(9931/10000)


Mean log L2 = -3.729463577270508
Total dropped = 0
param norm = 43.13300323486328

Epoch 73 in 230.27 sec
Train accuracy :	0.983	(49125/50000)		Train loss :	0.064
Val accuracy   :	0.88	(8801/10000)		Val loss :	0.4702
Top-5 val acc  :	0.995	(9948/10000)


Mean log L2 = -3.7140214443206787
Total dropped = 0
param norm = 42.950538635253906

Epoch 74 in 230.30 sec
Train accuracy :	0.984	(49188/50000)		Train loss :	0.0601
Val accuracy   :	0.882	(8820/10000)		Val loss :	0.4672
Top-5 val acc  :	0.994	(9943/10000)


Mean log L2 = -3.594597339630127
Total dropped = 0
param norm = 42.76486587524414

Epoch 75 in 230.26 sec
Train accuracy :	0.986	(49277/50000)		Train loss :	0.0561
Val accuracy   :	0.873	(8732/10000)		Val loss :	0.4728
Top-5 val acc  :	0.994	(9944/10000)


Mean log L2 = -3.7250077724456787
Total dropped = 0
param norm = 42.60726547241211

Epoch 76 in 230.39 sec
Train accuracy :	0.986	(49310/50000)		Train loss :	0.0551
Val accuracy   :	0.88	(8798/10000)		Val loss :	0.4735
Top-5 val acc  :	0.994	(9938/10000)


Mean log L2 = -3.8431472778320312
Total dropped = 0
param norm = 42.45656967163086

Epoch 77 in 230.36 sec
Train accuracy :	0.988	(49392/50000)		Train loss :	0.0483
Val accuracy   :	0.881	(8807/10000)		Val loss :	0.481
Top-5 val acc  :	0.994	(9942/10000)


Mean log L2 = -3.591344118118286
Total dropped = 0
param norm = 42.32920837402344

Epoch 78 in 230.28 sec
Train accuracy :	0.988	(49402/50000)		Train loss :	0.0474
Val accuracy   :	0.883	(8827/10000)		Val loss :	0.4889
Top-5 val acc  :	0.995	(9947/10000)


Mean log L2 = -3.582199811935425
Total dropped = 0
param norm = 42.20594024658203

Epoch 79 in 230.30 sec
Train accuracy :	0.989	(49457/50000)		Train loss :	0.0437
Val accuracy   :	0.88	(8797/10000)		Val loss :	0.5001
Top-5 val acc  :	0.994	(9942/10000)


Mean log L2 = -3.65021014213562
Total dropped = 0
param norm = 42.09934997558594

Epoch 80 in 230.33 sec
Train accuracy :	0.99	(49490/50000)		Train loss :	0.0423
Val accuracy   :	0.883	(8833/10000)		Val loss :	0.4944
Top-5 val acc  :	0.994	(9943/10000)


Mean log L2 = -3.6043126583099365
Total dropped = 0
param norm = 42.009586334228516

Epoch 81 in 230.31 sec
Train accuracy :	0.99	(49502/50000)		Train loss :	0.0404
Val accuracy   :	0.882	(8825/10000)		Val loss :	0.503
Top-5 val acc  :	0.995	(9947/10000)


Mean log L2 = -3.596050262451172
Total dropped = 0
param norm = 41.93499755859375

Epoch 82 in 230.42 sec
Train accuracy :	0.991	(49539/50000)		Train loss :	0.0374
Val accuracy   :	0.884	(8837/10000)		Val loss :	0.5029
Top-5 val acc  :	0.995	(9950/10000)


Mean log L2 = -3.4703500270843506
Total dropped = 0
param norm = 41.87162399291992

Epoch 83 in 230.38 sec
Train accuracy :	0.991	(49550/50000)		Train loss :	0.0343
Val accuracy   :	0.882	(8816/10000)		Val loss :	0.5122
Top-5 val acc  :	0.995	(9945/10000)


Mean log L2 = -3.6603798866271973
Total dropped = 0
param norm = 41.8201904296875

Epoch 84 in 230.32 sec
Train accuracy :	0.991	(49548/50000)		Train loss :	0.0366
Val accuracy   :	0.886	(8858/10000)		Val loss :	0.5047
Top-5 val acc  :	0.994	(9942/10000)


Mean log L2 = -3.4793331623077393
Total dropped = 0
param norm = 41.78134536743164

Epoch 85 in 230.49 sec
Train accuracy :	0.992	(49617/50000)		Train loss :	0.0322
Val accuracy   :	0.885	(8855/10000)		Val loss :	0.514
Top-5 val acc  :	0.994	(9943/10000)


Mean log L2 = -3.509418487548828
Total dropped = 0
param norm = 41.75230407714844

Epoch 86 in 230.29 sec
Train accuracy :	0.993	(49647/50000)		Train loss :	0.0303
Val accuracy   :	0.884	(8840/10000)		Val loss :	0.5229
Top-5 val acc  :	0.994	(9944/10000)


Mean log L2 = -3.639014482498169
Total dropped = 0
param norm = 41.732444763183594

Epoch 87 in 230.20 sec
Train accuracy :	0.993	(49637/50000)		Train loss :	0.0306
Val accuracy   :	0.884	(8838/10000)		Val loss :	0.5238
Top-5 val acc  :	0.994	(9943/10000)


Mean log L2 = -3.682337999343872
Total dropped = 0
param norm = 41.718753814697266

Epoch 88 in 230.20 sec
Train accuracy :	0.992	(49604/50000)		Train loss :	0.0333
Val accuracy   :	0.884	(8844/10000)		Val loss :	0.5235
Top-5 val acc  :	0.995	(9947/10000)


Mean log L2 = -3.620806932449341
Total dropped = 0
param norm = 41.71140670776367

Epoch 89 in 230.16 sec
Train accuracy :	0.993	(49674/50000)		Train loss :	0.0297
Val accuracy   :	0.885	(8854/10000)		Val loss :	0.5186
Top-5 val acc  :	0.995	(9948/10000)


Mean log L2 = -3.6419949531555176
Total dropped = 0
param norm = 41.70822525024414

Epoch 90 in 230.25 sec
Train accuracy :	0.993	(49663/50000)		Train loss :	0.03
Val accuracy   :	0.885	(8852/10000)		Val loss :	0.5192
Top-5 val acc  :	0.994	(9943/10000)


