Permalink
Find file Copy path
Fetching contributors…
Cannot retrieve contributors at this time
3210 lines (3209 sloc) 210 KB
Namespace(batch_size=50, data_name='SST-2', dropout=0.5, epochs=40, gpu=0, log_interval=30, lr=0.0001, model_mode='rand', save_prefix='sa-model')
Use gpu0
1614
53
Done! Tokenizing Time=4.58s, #Sentences=118038
Done! Tokenizing Time=0.78s, #Sentences=1745
SentimentNet(
(embedding): Embedding(17814 -> 300, float32)
(encoder): ConvolutionalEncoder(
(_convs): HybridConcurrent(
(0): HybridSequential(
(0): Conv1D(300 -> 100, kernel_size=(3,), stride=(1,))
(1): Activation(relu)
(2): HybridLambda(<lambda>)
)
(1): HybridSequential(
(0): Conv1D(300 -> 100, kernel_size=(4,), stride=(1,))
(1): Activation(relu)
(2): HybridLambda(<lambda>)
)
(2): HybridSequential(
(0): Conv1D(300 -> 100, kernel_size=(5,), stride=(1,))
(1): Activation(relu)
(2): HybridLambda(<lambda>)
)
)
)
(output): HybridSequential(
(0): Dropout(p = 0.5, axes=())
(1): Dense(None -> 2, linear)
)
)
[Epoch 0 Batch 30/2125] avg loss 0.013829, throughput 3.66685K wps
[Epoch 0 Batch 60/2125] avg loss 0.0137645, throughput 6.05735K wps
[Epoch 0 Batch 90/2125] avg loss 0.0137475, throughput 6.07005K wps
[Epoch 0 Batch 120/2125] avg loss 0.0136432, throughput 6.06206K wps
[Epoch 0 Batch 150/2125] avg loss 0.0136089, throughput 6.0656K wps
[Epoch 0 Batch 180/2125] avg loss 0.013504, throughput 6.0639K wps
[Epoch 0 Batch 210/2125] avg loss 0.0135624, throughput 6.06649K wps
[Epoch 0 Batch 240/2125] avg loss 0.0134624, throughput 6.0619K wps
[Epoch 0 Batch 270/2125] avg loss 0.0133424, throughput 6.06328K wps
[Epoch 0 Batch 300/2125] avg loss 0.0133814, throughput 6.06174K wps
[Epoch 0 Batch 330/2125] avg loss 0.0132914, throughput 6.05639K wps
[Epoch 0 Batch 360/2125] avg loss 0.0131282, throughput 6.06175K wps
[Epoch 0 Batch 390/2125] avg loss 0.0128756, throughput 6.06474K wps
[Epoch 0 Batch 420/2125] avg loss 0.0126722, throughput 6.05425K wps
[Epoch 0 Batch 450/2125] avg loss 0.0127548, throughput 6.0656K wps
[Epoch 0 Batch 480/2125] avg loss 0.0122959, throughput 6.05347K wps
[Epoch 0 Batch 510/2125] avg loss 0.012094, throughput 6.05032K wps
[Epoch 0 Batch 540/2125] avg loss 0.0116756, throughput 6.06943K wps
[Epoch 0 Batch 570/2125] avg loss 0.0115086, throughput 6.05083K wps
[Epoch 0 Batch 600/2125] avg loss 0.0112123, throughput 6.05777K wps
[Epoch 0 Batch 630/2125] avg loss 0.0107596, throughput 6.04489K wps
[Epoch 0 Batch 660/2125] avg loss 0.0103285, throughput 6.04414K wps
[Epoch 0 Batch 690/2125] avg loss 0.0100179, throughput 6.05513K wps
[Epoch 0 Batch 720/2125] avg loss 0.0101158, throughput 6.05751K wps
[Epoch 0 Batch 750/2125] avg loss 0.0092245, throughput 6.05337K wps
[Epoch 0 Batch 780/2125] avg loss 0.00903469, throughput 6.05519K wps
[Epoch 0 Batch 810/2125] avg loss 0.00905648, throughput 6.04936K wps
[Epoch 0 Batch 840/2125] avg loss 0.00860034, throughput 6.0433K wps
[Epoch 0 Batch 870/2125] avg loss 0.00784226, throughput 6.05035K wps
[Epoch 0 Batch 900/2125] avg loss 0.00819203, throughput 6.05377K wps
[Epoch 0 Batch 930/2125] avg loss 0.00793297, throughput 6.04908K wps
[Epoch 0 Batch 960/2125] avg loss 0.00782591, throughput 6.04933K wps
[Epoch 0 Batch 990/2125] avg loss 0.007448, throughput 6.043K wps
[Epoch 0 Batch 1020/2125] avg loss 0.00726903, throughput 6.04568K wps
[Epoch 0 Batch 1050/2125] avg loss 0.00739385, throughput 6.04146K wps
[Epoch 0 Batch 1080/2125] avg loss 0.0070646, throughput 6.05312K wps
[Epoch 0 Batch 1110/2125] avg loss 0.00699586, throughput 6.04355K wps
[Epoch 0 Batch 1140/2125] avg loss 0.00710965, throughput 6.04414K wps
[Epoch 0 Batch 1170/2125] avg loss 0.00695069, throughput 6.04976K wps
[Epoch 0 Batch 1200/2125] avg loss 0.00663305, throughput 6.03645K wps
[Epoch 0 Batch 1230/2125] avg loss 0.00641377, throughput 6.03661K wps
[Epoch 0 Batch 1260/2125] avg loss 0.00665456, throughput 6.03939K wps
[Epoch 0 Batch 1290/2125] avg loss 0.00656593, throughput 6.0487K wps
[Epoch 0 Batch 1320/2125] avg loss 0.00642296, throughput 6.0419K wps
[Epoch 0 Batch 1350/2125] avg loss 0.00607219, throughput 6.05294K wps
[Epoch 0 Batch 1380/2125] avg loss 0.00640395, throughput 6.05378K wps
[Epoch 0 Batch 1410/2125] avg loss 0.00625747, throughput 6.05652K wps
[Epoch 0 Batch 1440/2125] avg loss 0.00618365, throughput 6.05028K wps
[Epoch 0 Batch 1470/2125] avg loss 0.00638179, throughput 6.0515K wps
[Epoch 0 Batch 1500/2125] avg loss 0.00608974, throughput 6.05329K wps
[Epoch 0 Batch 1530/2125] avg loss 0.00608176, throughput 6.04791K wps
[Epoch 0 Batch 1560/2125] avg loss 0.00657479, throughput 6.04796K wps
[Epoch 0 Batch 1590/2125] avg loss 0.00601842, throughput 6.0459K wps
[Epoch 0 Batch 1620/2125] avg loss 0.00622085, throughput 6.04544K wps
[Epoch 0 Batch 1650/2125] avg loss 0.00621068, throughput 6.02408K wps
[Epoch 0 Batch 1680/2125] avg loss 0.00582171, throughput 6.04315K wps
[Epoch 0 Batch 1710/2125] avg loss 0.00585082, throughput 6.04712K wps
[Epoch 0 Batch 1740/2125] avg loss 0.00559648, throughput 6.04692K wps
[Epoch 0 Batch 1770/2125] avg loss 0.00583111, throughput 6.05233K wps
[Epoch 0 Batch 1800/2125] avg loss 0.00604723, throughput 6.0428K wps
[Epoch 0 Batch 1830/2125] avg loss 0.00555733, throughput 6.0459K wps
[Epoch 0 Batch 1860/2125] avg loss 0.00556889, throughput 6.04087K wps
[Epoch 0 Batch 1890/2125] avg loss 0.00571736, throughput 6.04176K wps
[Epoch 0 Batch 1920/2125] avg loss 0.00494089, throughput 6.03961K wps
[Epoch 0 Batch 1950/2125] avg loss 0.00565151, throughput 6.03557K wps
[Epoch 0 Batch 1980/2125] avg loss 0.00596137, throughput 6.03493K wps
[Epoch 0 Batch 2010/2125] avg loss 0.00559709, throughput 6.03625K wps
[Epoch 0 Batch 2040/2125] avg loss 0.00550116, throughput 6.04436K wps
[Epoch 0 Batch 2070/2125] avg loss 0.00499994, throughput 6.03316K wps
[Epoch 0 Batch 2100/2125] avg loss 0.00560363, throughput 6.04246K wps
Begin Testing...
[Batch 30/237] elapsed 0.30 s
[Batch 60/237] elapsed 0.28 s
[Batch 90/237] elapsed 0.28 s
[Batch 120/237] elapsed 0.28 s
[Batch 150/237] elapsed 0.28 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 0] train avg loss 0.00859733, test acc 0.8961, test avg loss 0.26706, throughput 5.95804K wps
Observed Improvement.
Begin Testing...
[Batch 30/35] elapsed 0.27 s
[Epoch 1 Batch 30/2125] avg loss 0.00508697, throughput 6.17923K wps
[Epoch 1 Batch 60/2125] avg loss 0.00468721, throughput 6.05207K wps
[Epoch 1 Batch 90/2125] avg loss 0.00464745, throughput 6.03797K wps
[Epoch 1 Batch 120/2125] avg loss 0.00480526, throughput 6.03109K wps
[Epoch 1 Batch 150/2125] avg loss 0.00461163, throughput 6.02916K wps
[Epoch 1 Batch 180/2125] avg loss 0.0041649, throughput 6.02806K wps
[Epoch 1 Batch 210/2125] avg loss 0.00453032, throughput 6.03678K wps
[Epoch 1 Batch 240/2125] avg loss 0.00432474, throughput 6.03632K wps
[Epoch 1 Batch 270/2125] avg loss 0.00481225, throughput 6.02714K wps
[Epoch 1 Batch 300/2125] avg loss 0.00433578, throughput 6.02622K wps
[Epoch 1 Batch 330/2125] avg loss 0.00467245, throughput 6.02281K wps
[Epoch 1 Batch 360/2125] avg loss 0.00403423, throughput 6.03134K wps
[Epoch 1 Batch 390/2125] avg loss 0.00492794, throughput 6.03659K wps
[Epoch 1 Batch 420/2125] avg loss 0.00407941, throughput 6.03136K wps
[Epoch 1 Batch 450/2125] avg loss 0.00462471, throughput 6.02857K wps
[Epoch 1 Batch 480/2125] avg loss 0.00447644, throughput 6.02654K wps
[Epoch 1 Batch 510/2125] avg loss 0.00415036, throughput 6.02889K wps
[Epoch 1 Batch 540/2125] avg loss 0.00407898, throughput 6.03537K wps
[Epoch 1 Batch 570/2125] avg loss 0.00465972, throughput 6.03405K wps
[Epoch 1 Batch 600/2125] avg loss 0.00454082, throughput 6.02579K wps
[Epoch 1 Batch 630/2125] avg loss 0.00457226, throughput 6.02746K wps
[Epoch 1 Batch 660/2125] avg loss 0.00469247, throughput 6.03245K wps
[Epoch 1 Batch 690/2125] avg loss 0.00457549, throughput 6.03122K wps
[Epoch 1 Batch 720/2125] avg loss 0.00415805, throughput 6.02865K wps
[Epoch 1 Batch 750/2125] avg loss 0.00415541, throughput 6.02474K wps
[Epoch 1 Batch 780/2125] avg loss 0.00460983, throughput 6.02889K wps
[Epoch 1 Batch 810/2125] avg loss 0.00462681, throughput 6.02913K wps
[Epoch 1 Batch 840/2125] avg loss 0.00438685, throughput 6.02683K wps
[Epoch 1 Batch 870/2125] avg loss 0.00384389, throughput 6.0347K wps
[Epoch 1 Batch 900/2125] avg loss 0.00411738, throughput 6.02614K wps
[Epoch 1 Batch 930/2125] avg loss 0.00386545, throughput 6.02518K wps
[Epoch 1 Batch 960/2125] avg loss 0.00410042, throughput 6.02562K wps
[Epoch 1 Batch 990/2125] avg loss 0.00418177, throughput 6.02193K wps
[Epoch 1 Batch 1020/2125] avg loss 0.00429555, throughput 6.03404K wps
[Epoch 1 Batch 1050/2125] avg loss 0.00447776, throughput 6.02487K wps
[Epoch 1 Batch 1080/2125] avg loss 0.00435861, throughput 6.02774K wps
[Epoch 1 Batch 1110/2125] avg loss 0.00436445, throughput 6.04158K wps
[Epoch 1 Batch 1140/2125] avg loss 0.00429003, throughput 6.03309K wps
[Epoch 1 Batch 1170/2125] avg loss 0.00429183, throughput 6.03113K wps
[Epoch 1 Batch 1200/2125] avg loss 0.00402015, throughput 6.0287K wps
[Epoch 1 Batch 1230/2125] avg loss 0.00370733, throughput 6.03442K wps
[Epoch 1 Batch 1260/2125] avg loss 0.00392001, throughput 6.02985K wps
[Epoch 1 Batch 1290/2125] avg loss 0.00398331, throughput 6.03933K wps
[Epoch 1 Batch 1320/2125] avg loss 0.00419159, throughput 6.03513K wps
[Epoch 1 Batch 1350/2125] avg loss 0.00449445, throughput 6.03314K wps
[Epoch 1 Batch 1380/2125] avg loss 0.003994, throughput 6.03455K wps
[Epoch 1 Batch 1410/2125] avg loss 0.00406692, throughput 6.03737K wps
[Epoch 1 Batch 1440/2125] avg loss 0.00441621, throughput 6.0322K wps
[Epoch 1 Batch 1470/2125] avg loss 0.00412627, throughput 6.02868K wps
[Epoch 1 Batch 1500/2125] avg loss 0.00417826, throughput 6.03309K wps
[Epoch 1 Batch 1530/2125] avg loss 0.00436325, throughput 6.02982K wps
[Epoch 1 Batch 1560/2125] avg loss 0.00416676, throughput 6.04032K wps
[Epoch 1 Batch 1590/2125] avg loss 0.00379787, throughput 6.02174K wps
[Epoch 1 Batch 1620/2125] avg loss 0.00464045, throughput 6.01529K wps
[Epoch 1 Batch 1650/2125] avg loss 0.00402388, throughput 6.02384K wps
[Epoch 1 Batch 1680/2125] avg loss 0.00433767, throughput 6.02607K wps
[Epoch 1 Batch 1710/2125] avg loss 0.00412046, throughput 6.03575K wps
[Epoch 1 Batch 1740/2125] avg loss 0.00447082, throughput 6.01659K wps
[Epoch 1 Batch 1770/2125] avg loss 0.00414897, throughput 6.02051K wps
[Epoch 1 Batch 1800/2125] avg loss 0.00421806, throughput 6.02261K wps
[Epoch 1 Batch 1830/2125] avg loss 0.00426355, throughput 6.0174K wps
[Epoch 1 Batch 1860/2125] avg loss 0.00376214, throughput 6.03443K wps
[Epoch 1 Batch 1890/2125] avg loss 0.00424035, throughput 6.02313K wps
[Epoch 1 Batch 1920/2125] avg loss 0.00426306, throughput 6.02931K wps
[Epoch 1 Batch 1950/2125] avg loss 0.00479304, throughput 6.03408K wps
[Epoch 1 Batch 1980/2125] avg loss 0.00444748, throughput 6.0338K wps
[Epoch 1 Batch 2010/2125] avg loss 0.00410632, throughput 6.03082K wps
[Epoch 1 Batch 2040/2125] avg loss 0.0041873, throughput 6.03016K wps
[Epoch 1 Batch 2070/2125] avg loss 0.00392489, throughput 6.0252K wps
[Epoch 1 Batch 2100/2125] avg loss 0.00462059, throughput 6.02999K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 1] train avg loss 0.00431313, test acc 0.9144, test avg loss 0.231972, throughput 6.0321K wps
Observed Improvement.
Begin Testing...
[Batch 30/35] elapsed 0.27 s
[Epoch 2 Batch 30/2125] avg loss 0.00314362, throughput 6.15894K wps
[Epoch 2 Batch 60/2125] avg loss 0.00308142, throughput 6.01439K wps
[Epoch 2 Batch 90/2125] avg loss 0.00338772, throughput 6.02925K wps
[Epoch 2 Batch 120/2125] avg loss 0.0030482, throughput 6.02136K wps
[Epoch 2 Batch 150/2125] avg loss 0.00291294, throughput 6.02453K wps
[Epoch 2 Batch 180/2125] avg loss 0.00331701, throughput 6.02185K wps
[Epoch 2 Batch 210/2125] avg loss 0.00347754, throughput 6.01208K wps
[Epoch 2 Batch 240/2125] avg loss 0.00360653, throughput 6.02371K wps
[Epoch 2 Batch 270/2125] avg loss 0.0027905, throughput 6.02059K wps
[Epoch 2 Batch 300/2125] avg loss 0.0034358, throughput 6.02302K wps
[Epoch 2 Batch 330/2125] avg loss 0.00278267, throughput 6.02687K wps
[Epoch 2 Batch 360/2125] avg loss 0.00293847, throughput 6.0222K wps
[Epoch 2 Batch 390/2125] avg loss 0.00357792, throughput 6.0258K wps
[Epoch 2 Batch 420/2125] avg loss 0.00331124, throughput 6.03765K wps
[Epoch 2 Batch 450/2125] avg loss 0.00319685, throughput 6.01817K wps
[Epoch 2 Batch 480/2125] avg loss 0.00349386, throughput 6.02044K wps
[Epoch 2 Batch 510/2125] avg loss 0.00345725, throughput 6.02068K wps
[Epoch 2 Batch 540/2125] avg loss 0.00350414, throughput 6.02721K wps
[Epoch 2 Batch 570/2125] avg loss 0.00337767, throughput 6.03502K wps
[Epoch 2 Batch 600/2125] avg loss 0.00302796, throughput 6.02627K wps
[Epoch 2 Batch 630/2125] avg loss 0.00304125, throughput 6.00929K wps
[Epoch 2 Batch 660/2125] avg loss 0.00319373, throughput 6.01309K wps
[Epoch 2 Batch 690/2125] avg loss 0.00352117, throughput 6.01663K wps
[Epoch 2 Batch 720/2125] avg loss 0.00310565, throughput 6.01505K wps
[Epoch 2 Batch 750/2125] avg loss 0.00322476, throughput 6.01698K wps
[Epoch 2 Batch 780/2125] avg loss 0.00295135, throughput 6.02647K wps
[Epoch 2 Batch 810/2125] avg loss 0.00325777, throughput 6.02007K wps
[Epoch 2 Batch 840/2125] avg loss 0.00311891, throughput 6.02495K wps
[Epoch 2 Batch 870/2125] avg loss 0.00376137, throughput 6.03409K wps
[Epoch 2 Batch 900/2125] avg loss 0.00306664, throughput 6.03132K wps
[Epoch 2 Batch 930/2125] avg loss 0.00353107, throughput 6.02501K wps
[Epoch 2 Batch 960/2125] avg loss 0.00280812, throughput 6.02909K wps
[Epoch 2 Batch 990/2125] avg loss 0.00356564, throughput 6.019K wps
[Epoch 2 Batch 1020/2125] avg loss 0.00306218, throughput 6.02777K wps
[Epoch 2 Batch 1050/2125] avg loss 0.00300588, throughput 6.03341K wps
[Epoch 2 Batch 1080/2125] avg loss 0.00336497, throughput 6.0355K wps
[Epoch 2 Batch 1110/2125] avg loss 0.00341811, throughput 6.03052K wps
[Epoch 2 Batch 1140/2125] avg loss 0.00343795, throughput 6.02125K wps
[Epoch 2 Batch 1170/2125] avg loss 0.00314894, throughput 6.02437K wps
[Epoch 2 Batch 1200/2125] avg loss 0.00374758, throughput 6.02336K wps
[Epoch 2 Batch 1230/2125] avg loss 0.00320125, throughput 6.02539K wps
[Epoch 2 Batch 1260/2125] avg loss 0.00337495, throughput 6.02558K wps
[Epoch 2 Batch 1290/2125] avg loss 0.00301344, throughput 6.03116K wps
[Epoch 2 Batch 1320/2125] avg loss 0.00325075, throughput 6.02569K wps
[Epoch 2 Batch 1350/2125] avg loss 0.00367398, throughput 6.03365K wps
[Epoch 2 Batch 1380/2125] avg loss 0.00292623, throughput 6.02599K wps
[Epoch 2 Batch 1410/2125] avg loss 0.00388576, throughput 6.02173K wps
[Epoch 2 Batch 1440/2125] avg loss 0.00347624, throughput 6.02922K wps
[Epoch 2 Batch 1470/2125] avg loss 0.00346544, throughput 6.03134K wps
[Epoch 2 Batch 1500/2125] avg loss 0.00333196, throughput 6.02616K wps
[Epoch 2 Batch 1530/2125] avg loss 0.00331523, throughput 6.02593K wps
[Epoch 2 Batch 1560/2125] avg loss 0.00378902, throughput 6.01847K wps
[Epoch 2 Batch 1590/2125] avg loss 0.00316463, throughput 6.0314K wps
[Epoch 2 Batch 1620/2125] avg loss 0.00300465, throughput 6.0332K wps
[Epoch 2 Batch 1650/2125] avg loss 0.00359532, throughput 6.0331K wps
[Epoch 2 Batch 1680/2125] avg loss 0.00374746, throughput 6.01796K wps
[Epoch 2 Batch 1710/2125] avg loss 0.00347461, throughput 6.02801K wps
[Epoch 2 Batch 1740/2125] avg loss 0.00337847, throughput 6.03183K wps
[Epoch 2 Batch 1770/2125] avg loss 0.00280696, throughput 6.02255K wps
[Epoch 2 Batch 1800/2125] avg loss 0.00348706, throughput 6.01956K wps
[Epoch 2 Batch 1830/2125] avg loss 0.00289692, throughput 6.02517K wps
[Epoch 2 Batch 1860/2125] avg loss 0.00327325, throughput 6.02422K wps
[Epoch 2 Batch 1890/2125] avg loss 0.00307346, throughput 6.02271K wps
[Epoch 2 Batch 1920/2125] avg loss 0.00338595, throughput 6.02511K wps
[Epoch 2 Batch 1950/2125] avg loss 0.00318515, throughput 6.01937K wps
[Epoch 2 Batch 1980/2125] avg loss 0.00305889, throughput 6.01414K wps
[Epoch 2 Batch 2010/2125] avg loss 0.0031303, throughput 6.01831K wps
[Epoch 2 Batch 2040/2125] avg loss 0.00321362, throughput 6.02186K wps
[Epoch 2 Batch 2070/2125] avg loss 0.00333849, throughput 6.01876K wps
[Epoch 2 Batch 2100/2125] avg loss 0.00354186, throughput 6.01233K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 2] train avg loss 0.00327839, test acc 0.9203, test avg loss 0.225931, throughput 6.02604K wps
Observed Improvement.
Begin Testing...
[Batch 30/35] elapsed 0.27 s
[Epoch 3 Batch 30/2125] avg loss 0.00238329, throughput 6.15746K wps
[Epoch 3 Batch 60/2125] avg loss 0.00273408, throughput 6.01289K wps
[Epoch 3 Batch 90/2125] avg loss 0.00277831, throughput 6.02173K wps
[Epoch 3 Batch 120/2125] avg loss 0.00259528, throughput 6.03131K wps
[Epoch 3 Batch 150/2125] avg loss 0.00260867, throughput 6.0261K wps
[Epoch 3 Batch 180/2125] avg loss 0.00261858, throughput 6.02391K wps
[Epoch 3 Batch 210/2125] avg loss 0.00245517, throughput 6.02365K wps
[Epoch 3 Batch 240/2125] avg loss 0.0022151, throughput 6.02424K wps
[Epoch 3 Batch 270/2125] avg loss 0.00227305, throughput 6.01542K wps
[Epoch 3 Batch 300/2125] avg loss 0.00239416, throughput 6.0152K wps
[Epoch 3 Batch 330/2125] avg loss 0.00261257, throughput 6.01598K wps
[Epoch 3 Batch 360/2125] avg loss 0.00296452, throughput 6.02392K wps
[Epoch 3 Batch 390/2125] avg loss 0.00230194, throughput 6.02513K wps
[Epoch 3 Batch 420/2125] avg loss 0.00247226, throughput 6.02798K wps
[Epoch 3 Batch 450/2125] avg loss 0.00285218, throughput 6.0303K wps
[Epoch 3 Batch 480/2125] avg loss 0.00290747, throughput 6.02527K wps
[Epoch 3 Batch 510/2125] avg loss 0.00300131, throughput 6.02043K wps
[Epoch 3 Batch 540/2125] avg loss 0.00299441, throughput 6.02515K wps
[Epoch 3 Batch 570/2125] avg loss 0.0028417, throughput 6.02152K wps
[Epoch 3 Batch 600/2125] avg loss 0.00291954, throughput 6.02418K wps
[Epoch 3 Batch 630/2125] avg loss 0.00253257, throughput 6.02197K wps
[Epoch 3 Batch 660/2125] avg loss 0.00266961, throughput 6.01394K wps
[Epoch 3 Batch 690/2125] avg loss 0.00231881, throughput 6.02464K wps
[Epoch 3 Batch 720/2125] avg loss 0.00288925, throughput 6.01866K wps
[Epoch 3 Batch 750/2125] avg loss 0.00252009, throughput 6.02882K wps
[Epoch 3 Batch 780/2125] avg loss 0.0024299, throughput 6.0224K wps
[Epoch 3 Batch 810/2125] avg loss 0.00304578, throughput 6.02104K wps
[Epoch 3 Batch 840/2125] avg loss 0.00287125, throughput 6.02843K wps
[Epoch 3 Batch 870/2125] avg loss 0.0025081, throughput 6.01218K wps
[Epoch 3 Batch 900/2125] avg loss 0.00263124, throughput 6.01111K wps
[Epoch 3 Batch 930/2125] avg loss 0.00263107, throughput 6.01456K wps
[Epoch 3 Batch 960/2125] avg loss 0.0024143, throughput 5.99362K wps
[Epoch 3 Batch 990/2125] avg loss 0.00246575, throughput 6.00334K wps
[Epoch 3 Batch 1020/2125] avg loss 0.00254204, throughput 6.02419K wps
[Epoch 3 Batch 1050/2125] avg loss 0.00278385, throughput 6.02448K wps
[Epoch 3 Batch 1080/2125] avg loss 0.00245835, throughput 6.02401K wps
[Epoch 3 Batch 1110/2125] avg loss 0.00263226, throughput 6.0216K wps
[Epoch 3 Batch 1140/2125] avg loss 0.003071, throughput 6.0168K wps
[Epoch 3 Batch 1170/2125] avg loss 0.0025007, throughput 6.01351K wps
[Epoch 3 Batch 1200/2125] avg loss 0.00264362, throughput 6.02229K wps
[Epoch 3 Batch 1230/2125] avg loss 0.00281793, throughput 6.01972K wps
[Epoch 3 Batch 1260/2125] avg loss 0.00264957, throughput 6.00981K wps
[Epoch 3 Batch 1290/2125] avg loss 0.00262275, throughput 6.0259K wps
[Epoch 3 Batch 1320/2125] avg loss 0.00252379, throughput 6.01957K wps
[Epoch 3 Batch 1350/2125] avg loss 0.00276722, throughput 6.02388K wps
[Epoch 3 Batch 1380/2125] avg loss 0.00235232, throughput 6.0223K wps
[Epoch 3 Batch 1410/2125] avg loss 0.00261438, throughput 6.02505K wps
[Epoch 3 Batch 1440/2125] avg loss 0.00258915, throughput 6.02491K wps
[Epoch 3 Batch 1470/2125] avg loss 0.00248594, throughput 6.00957K wps
[Epoch 3 Batch 1500/2125] avg loss 0.00281933, throughput 6.02309K wps
[Epoch 3 Batch 1530/2125] avg loss 0.00239254, throughput 6.0187K wps
[Epoch 3 Batch 1560/2125] avg loss 0.00268616, throughput 6.02759K wps
[Epoch 3 Batch 1590/2125] avg loss 0.00277859, throughput 6.02846K wps
[Epoch 3 Batch 1620/2125] avg loss 0.00271962, throughput 6.01869K wps
[Epoch 3 Batch 1650/2125] avg loss 0.0028008, throughput 6.0284K wps
[Epoch 3 Batch 1680/2125] avg loss 0.00304454, throughput 6.0219K wps
[Epoch 3 Batch 1710/2125] avg loss 0.00307218, throughput 6.02654K wps
[Epoch 3 Batch 1740/2125] avg loss 0.00298205, throughput 6.03345K wps
[Epoch 3 Batch 1770/2125] avg loss 0.00310105, throughput 6.0234K wps
[Epoch 3 Batch 1800/2125] avg loss 0.00298861, throughput 6.02282K wps
[Epoch 3 Batch 1830/2125] avg loss 0.0028988, throughput 6.01236K wps
[Epoch 3 Batch 1860/2125] avg loss 0.00256832, throughput 6.02101K wps
[Epoch 3 Batch 1890/2125] avg loss 0.00279202, throughput 6.01562K wps
[Epoch 3 Batch 1920/2125] avg loss 0.00312812, throughput 6.02395K wps
[Epoch 3 Batch 1950/2125] avg loss 0.00327366, throughput 6.02067K wps
[Epoch 3 Batch 1980/2125] avg loss 0.00259131, throughput 6.01293K wps
[Epoch 3 Batch 2010/2125] avg loss 0.00298956, throughput 6.01963K wps
[Epoch 3 Batch 2040/2125] avg loss 0.00244052, throughput 6.01786K wps
[Epoch 3 Batch 2070/2125] avg loss 0.00292767, throughput 6.01744K wps
[Epoch 3 Batch 2100/2125] avg loss 0.00241609, throughput 6.03382K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 3] train avg loss 0.00268678, test acc 0.9232, test avg loss 0.232664, throughput 6.02297K wps
Observed Improvement.
Begin Testing...
[Batch 30/35] elapsed 0.27 s
[Epoch 4 Batch 30/2125] avg loss 0.00197623, throughput 6.15294K wps
[Epoch 4 Batch 60/2125] avg loss 0.00184827, throughput 6.02155K wps
[Epoch 4 Batch 90/2125] avg loss 0.00214668, throughput 6.02516K wps
[Epoch 4 Batch 120/2125] avg loss 0.00228548, throughput 6.02233K wps
[Epoch 4 Batch 150/2125] avg loss 0.0024132, throughput 6.01614K wps
[Epoch 4 Batch 180/2125] avg loss 0.00186548, throughput 6.00777K wps
[Epoch 4 Batch 210/2125] avg loss 0.00200755, throughput 6.02317K wps
[Epoch 4 Batch 240/2125] avg loss 0.00230009, throughput 6.02821K wps
[Epoch 4 Batch 270/2125] avg loss 0.00198461, throughput 6.01693K wps
[Epoch 4 Batch 300/2125] avg loss 0.00195834, throughput 6.02209K wps
[Epoch 4 Batch 330/2125] avg loss 0.00218797, throughput 6.01K wps
[Epoch 4 Batch 360/2125] avg loss 0.00217512, throughput 6.0091K wps
[Epoch 4 Batch 390/2125] avg loss 0.00214169, throughput 6.0075K wps
[Epoch 4 Batch 420/2125] avg loss 0.00221047, throughput 6.01875K wps
[Epoch 4 Batch 450/2125] avg loss 0.00218145, throughput 6.02316K wps
[Epoch 4 Batch 480/2125] avg loss 0.00217336, throughput 6.0177K wps
[Epoch 4 Batch 510/2125] avg loss 0.00209655, throughput 6.02092K wps
[Epoch 4 Batch 540/2125] avg loss 0.00201126, throughput 6.02621K wps
[Epoch 4 Batch 570/2125] avg loss 0.00232639, throughput 6.02245K wps
[Epoch 4 Batch 600/2125] avg loss 0.00267902, throughput 6.02982K wps
[Epoch 4 Batch 630/2125] avg loss 0.00213633, throughput 6.02685K wps
[Epoch 4 Batch 660/2125] avg loss 0.00217018, throughput 6.03212K wps
[Epoch 4 Batch 690/2125] avg loss 0.00235743, throughput 6.02445K wps
[Epoch 4 Batch 720/2125] avg loss 0.00207556, throughput 6.03138K wps
[Epoch 4 Batch 750/2125] avg loss 0.0018993, throughput 6.02012K wps
[Epoch 4 Batch 780/2125] avg loss 0.00191106, throughput 6.02186K wps
[Epoch 4 Batch 810/2125] avg loss 0.00242145, throughput 6.02486K wps
[Epoch 4 Batch 840/2125] avg loss 0.00253679, throughput 6.02464K wps
[Epoch 4 Batch 870/2125] avg loss 0.00212724, throughput 6.01567K wps
[Epoch 4 Batch 900/2125] avg loss 0.00219012, throughput 6.02053K wps
[Epoch 4 Batch 930/2125] avg loss 0.00206217, throughput 6.02703K wps
[Epoch 4 Batch 960/2125] avg loss 0.00228351, throughput 6.02261K wps
[Epoch 4 Batch 990/2125] avg loss 0.00239689, throughput 6.01803K wps
[Epoch 4 Batch 1020/2125] avg loss 0.0020606, throughput 6.02067K wps
[Epoch 4 Batch 1050/2125] avg loss 0.00272871, throughput 6.01987K wps
[Epoch 4 Batch 1080/2125] avg loss 0.0025229, throughput 6.01812K wps
[Epoch 4 Batch 1110/2125] avg loss 0.00245847, throughput 6.01565K wps
[Epoch 4 Batch 1140/2125] avg loss 0.002153, throughput 6.02758K wps
[Epoch 4 Batch 1170/2125] avg loss 0.00245732, throughput 6.02199K wps
[Epoch 4 Batch 1200/2125] avg loss 0.00222894, throughput 6.02602K wps
[Epoch 4 Batch 1230/2125] avg loss 0.00260402, throughput 6.02724K wps
[Epoch 4 Batch 1260/2125] avg loss 0.00209676, throughput 6.02284K wps
[Epoch 4 Batch 1290/2125] avg loss 0.00243215, throughput 6.02223K wps
[Epoch 4 Batch 1320/2125] avg loss 0.00271236, throughput 6.02232K wps
[Epoch 4 Batch 1350/2125] avg loss 0.00213127, throughput 6.021K wps
[Epoch 4 Batch 1380/2125] avg loss 0.00228534, throughput 6.01215K wps
[Epoch 4 Batch 1410/2125] avg loss 0.00228203, throughput 6.01883K wps
[Epoch 4 Batch 1440/2125] avg loss 0.0023127, throughput 6.02022K wps
[Epoch 4 Batch 1470/2125] avg loss 0.00235965, throughput 6.02326K wps
[Epoch 4 Batch 1500/2125] avg loss 0.00224759, throughput 6.02781K wps
[Epoch 4 Batch 1530/2125] avg loss 0.00195852, throughput 6.01533K wps
[Epoch 4 Batch 1560/2125] avg loss 0.0023748, throughput 6.01792K wps
[Epoch 4 Batch 1590/2125] avg loss 0.00251534, throughput 6.02512K wps
[Epoch 4 Batch 1620/2125] avg loss 0.00224816, throughput 6.02711K wps
[Epoch 4 Batch 1650/2125] avg loss 0.0023237, throughput 6.01569K wps
[Epoch 4 Batch 1680/2125] avg loss 0.00218529, throughput 6.02487K wps
[Epoch 4 Batch 1710/2125] avg loss 0.00220012, throughput 6.0172K wps
[Epoch 4 Batch 1740/2125] avg loss 0.00250381, throughput 6.01555K wps
[Epoch 4 Batch 1770/2125] avg loss 0.00199925, throughput 6.0251K wps
[Epoch 4 Batch 1800/2125] avg loss 0.00247794, throughput 6.02896K wps
[Epoch 4 Batch 1830/2125] avg loss 0.00239459, throughput 6.01883K wps
[Epoch 4 Batch 1860/2125] avg loss 0.00246685, throughput 6.01585K wps
[Epoch 4 Batch 1890/2125] avg loss 0.00311176, throughput 6.02159K wps
[Epoch 4 Batch 1920/2125] avg loss 0.00229637, throughput 6.02174K wps
[Epoch 4 Batch 1950/2125] avg loss 0.002782, throughput 6.03186K wps
[Epoch 4 Batch 1980/2125] avg loss 0.00273546, throughput 6.02169K wps
[Epoch 4 Batch 2010/2125] avg loss 0.00265926, throughput 6.0199K wps
[Epoch 4 Batch 2040/2125] avg loss 0.00235093, throughput 6.017K wps
[Epoch 4 Batch 2070/2125] avg loss 0.00252596, throughput 6.02465K wps
[Epoch 4 Batch 2100/2125] avg loss 0.00238402, throughput 6.02324K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 4] train avg loss 0.00228829, test acc 0.9270, test avg loss 0.237708, throughput 6.0231K wps
Observed Improvement.
Begin Testing...
[Batch 30/35] elapsed 0.27 s
[Epoch 5 Batch 30/2125] avg loss 0.0015214, throughput 6.15398K wps
[Epoch 5 Batch 60/2125] avg loss 0.00163773, throughput 6.01606K wps
[Epoch 5 Batch 90/2125] avg loss 0.00170908, throughput 6.02076K wps
[Epoch 5 Batch 120/2125] avg loss 0.00177498, throughput 6.02317K wps
[Epoch 5 Batch 150/2125] avg loss 0.00192649, throughput 6.02125K wps
[Epoch 5 Batch 180/2125] avg loss 0.00174042, throughput 6.03462K wps
[Epoch 5 Batch 210/2125] avg loss 0.00190934, throughput 6.0257K wps
[Epoch 5 Batch 240/2125] avg loss 0.00157959, throughput 6.03177K wps
[Epoch 5 Batch 270/2125] avg loss 0.00174691, throughput 6.0255K wps
[Epoch 5 Batch 300/2125] avg loss 0.00192083, throughput 6.01453K wps
[Epoch 5 Batch 330/2125] avg loss 0.00187042, throughput 6.01651K wps
[Epoch 5 Batch 360/2125] avg loss 0.00164148, throughput 6.00793K wps
[Epoch 5 Batch 390/2125] avg loss 0.00182407, throughput 6.00613K wps
[Epoch 5 Batch 420/2125] avg loss 0.00194486, throughput 6.01612K wps
[Epoch 5 Batch 450/2125] avg loss 0.00215397, throughput 6.01802K wps
[Epoch 5 Batch 480/2125] avg loss 0.00184424, throughput 6.01337K wps
[Epoch 5 Batch 510/2125] avg loss 0.00144629, throughput 6.02035K wps
[Epoch 5 Batch 540/2125] avg loss 0.001615, throughput 6.02126K wps
[Epoch 5 Batch 570/2125] avg loss 0.00181364, throughput 6.02948K wps
[Epoch 5 Batch 600/2125] avg loss 0.00198801, throughput 6.02037K wps
[Epoch 5 Batch 630/2125] avg loss 0.00196187, throughput 6.02832K wps
[Epoch 5 Batch 660/2125] avg loss 0.00191681, throughput 6.01369K wps
[Epoch 5 Batch 690/2125] avg loss 0.00214142, throughput 6.02031K wps
[Epoch 5 Batch 720/2125] avg loss 0.00194812, throughput 6.01665K wps
[Epoch 5 Batch 750/2125] avg loss 0.0021861, throughput 6.01853K wps
[Epoch 5 Batch 780/2125] avg loss 0.00187838, throughput 6.02035K wps
[Epoch 5 Batch 810/2125] avg loss 0.00209187, throughput 6.01887K wps
[Epoch 5 Batch 840/2125] avg loss 0.00197526, throughput 6.01715K wps
[Epoch 5 Batch 870/2125] avg loss 0.00183353, throughput 6.01982K wps
[Epoch 5 Batch 900/2125] avg loss 0.00233926, throughput 6.01652K wps
[Epoch 5 Batch 930/2125] avg loss 0.00173277, throughput 6.02425K wps
[Epoch 5 Batch 960/2125] avg loss 0.00182582, throughput 6.00898K wps
[Epoch 5 Batch 990/2125] avg loss 0.00221033, throughput 6.01745K wps
[Epoch 5 Batch 1020/2125] avg loss 0.00230544, throughput 6.01811K wps
[Epoch 5 Batch 1050/2125] avg loss 0.00218735, throughput 6.012K wps
[Epoch 5 Batch 1080/2125] avg loss 0.00219576, throughput 6.00381K wps
[Epoch 5 Batch 1110/2125] avg loss 0.00165669, throughput 6.00991K wps
[Epoch 5 Batch 1140/2125] avg loss 0.0020985, throughput 6.02275K wps
[Epoch 5 Batch 1170/2125] avg loss 0.00253055, throughput 6.0203K wps
[Epoch 5 Batch 1200/2125] avg loss 0.00203626, throughput 6.00957K wps
[Epoch 5 Batch 1230/2125] avg loss 0.0022801, throughput 6.02109K wps
[Epoch 5 Batch 1260/2125] avg loss 0.00193983, throughput 6.00579K wps
[Epoch 5 Batch 1290/2125] avg loss 0.00193076, throughput 6.00873K wps
[Epoch 5 Batch 1320/2125] avg loss 0.00193145, throughput 6.02161K wps
[Epoch 5 Batch 1350/2125] avg loss 0.00186283, throughput 6.01171K wps
[Epoch 5 Batch 1380/2125] avg loss 0.0020561, throughput 6.00527K wps
[Epoch 5 Batch 1410/2125] avg loss 0.00232482, throughput 6.01825K wps
[Epoch 5 Batch 1440/2125] avg loss 0.00240091, throughput 6.01856K wps
[Epoch 5 Batch 1470/2125] avg loss 0.00187529, throughput 6.01792K wps
[Epoch 5 Batch 1500/2125] avg loss 0.0019871, throughput 6.00666K wps
[Epoch 5 Batch 1530/2125] avg loss 0.00193169, throughput 6.0077K wps
[Epoch 5 Batch 1560/2125] avg loss 0.00191817, throughput 6.01608K wps
[Epoch 5 Batch 1590/2125] avg loss 0.00220137, throughput 6.01543K wps
[Epoch 5 Batch 1620/2125] avg loss 0.00177599, throughput 6.01436K wps
[Epoch 5 Batch 1650/2125] avg loss 0.00183143, throughput 6.01666K wps
[Epoch 5 Batch 1680/2125] avg loss 0.00223401, throughput 6.01525K wps
[Epoch 5 Batch 1710/2125] avg loss 0.00177578, throughput 6.01818K wps
[Epoch 5 Batch 1740/2125] avg loss 0.00220512, throughput 6.00759K wps
[Epoch 5 Batch 1770/2125] avg loss 0.0020589, throughput 6.02069K wps
[Epoch 5 Batch 1800/2125] avg loss 0.00187366, throughput 6.01438K wps
[Epoch 5 Batch 1830/2125] avg loss 0.00173914, throughput 6.00908K wps
[Epoch 5 Batch 1860/2125] avg loss 0.00225509, throughput 6.01486K wps
[Epoch 5 Batch 1890/2125] avg loss 0.00228213, throughput 6.01906K wps
[Epoch 5 Batch 1920/2125] avg loss 0.00204147, throughput 6.02079K wps
[Epoch 5 Batch 1950/2125] avg loss 0.00240567, throughput 6.01158K wps
[Epoch 5 Batch 1980/2125] avg loss 0.00263055, throughput 6.00124K wps
[Epoch 5 Batch 2010/2125] avg loss 0.00229786, throughput 5.999K wps
[Epoch 5 Batch 2040/2125] avg loss 0.00230623, throughput 6.01608K wps
[Epoch 5 Batch 2070/2125] avg loss 0.00209841, throughput 6.01176K wps
[Epoch 5 Batch 2100/2125] avg loss 0.00162761, throughput 6.01739K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 5] train avg loss 0.00198567, test acc 0.9257, test avg loss 0.251908, throughput 6.01807K wps
[Epoch 6 Batch 30/2125] avg loss 0.00143401, throughput 6.15987K wps
[Epoch 6 Batch 60/2125] avg loss 0.00162062, throughput 6.02889K wps
[Epoch 6 Batch 90/2125] avg loss 0.00159454, throughput 6.01238K wps
[Epoch 6 Batch 120/2125] avg loss 0.00160784, throughput 6.01926K wps
[Epoch 6 Batch 150/2125] avg loss 0.00163463, throughput 6.0244K wps
[Epoch 6 Batch 180/2125] avg loss 0.00158036, throughput 6.02308K wps
[Epoch 6 Batch 210/2125] avg loss 0.0014567, throughput 6.02443K wps
[Epoch 6 Batch 240/2125] avg loss 0.0013044, throughput 6.00074K wps
[Epoch 6 Batch 270/2125] avg loss 0.00179579, throughput 6.0256K wps
[Epoch 6 Batch 300/2125] avg loss 0.00144325, throughput 6.00409K wps
[Epoch 6 Batch 330/2125] avg loss 0.001594, throughput 6.03011K wps
[Epoch 6 Batch 360/2125] avg loss 0.00140858, throughput 6.02536K wps
[Epoch 6 Batch 390/2125] avg loss 0.00168392, throughput 6.02124K wps
[Epoch 6 Batch 420/2125] avg loss 0.00205177, throughput 6.02735K wps
[Epoch 6 Batch 450/2125] avg loss 0.00180042, throughput 6.00918K wps
[Epoch 6 Batch 480/2125] avg loss 0.00171754, throughput 6.02905K wps
[Epoch 6 Batch 510/2125] avg loss 0.00152203, throughput 6.01732K wps
[Epoch 6 Batch 540/2125] avg loss 0.00158528, throughput 6.01437K wps
[Epoch 6 Batch 570/2125] avg loss 0.00149926, throughput 6.02519K wps
[Epoch 6 Batch 600/2125] avg loss 0.00158698, throughput 6.0241K wps
[Epoch 6 Batch 630/2125] avg loss 0.00176024, throughput 6.00785K wps
[Epoch 6 Batch 660/2125] avg loss 0.00182589, throughput 6.02473K wps
[Epoch 6 Batch 690/2125] avg loss 0.00187512, throughput 6.02701K wps
[Epoch 6 Batch 720/2125] avg loss 0.00193131, throughput 6.02549K wps
[Epoch 6 Batch 750/2125] avg loss 0.00149832, throughput 6.01486K wps
[Epoch 6 Batch 780/2125] avg loss 0.00193317, throughput 6.02798K wps
[Epoch 6 Batch 810/2125] avg loss 0.00153479, throughput 6.01887K wps
[Epoch 6 Batch 840/2125] avg loss 0.00142075, throughput 6.02631K wps
[Epoch 6 Batch 870/2125] avg loss 0.00200754, throughput 6.02921K wps
[Epoch 6 Batch 900/2125] avg loss 0.00207879, throughput 6.02225K wps
[Epoch 6 Batch 930/2125] avg loss 0.001911, throughput 6.01305K wps
[Epoch 6 Batch 960/2125] avg loss 0.00172938, throughput 6.02176K wps
[Epoch 6 Batch 990/2125] avg loss 0.00156606, throughput 6.03392K wps
[Epoch 6 Batch 1020/2125] avg loss 0.00179803, throughput 6.02219K wps
[Epoch 6 Batch 1050/2125] avg loss 0.00180095, throughput 6.02793K wps
[Epoch 6 Batch 1080/2125] avg loss 0.00148448, throughput 6.0177K wps
[Epoch 6 Batch 1110/2125] avg loss 0.0017663, throughput 6.02777K wps
[Epoch 6 Batch 1140/2125] avg loss 0.00142393, throughput 6.02614K wps
[Epoch 6 Batch 1170/2125] avg loss 0.00154842, throughput 6.01913K wps
[Epoch 6 Batch 1200/2125] avg loss 0.00176048, throughput 6.02981K wps
[Epoch 6 Batch 1230/2125] avg loss 0.00163366, throughput 5.99733K wps
[Epoch 6 Batch 1260/2125] avg loss 0.00170489, throughput 6.00375K wps
[Epoch 6 Batch 1290/2125] avg loss 0.00182386, throughput 6.00806K wps
[Epoch 6 Batch 1320/2125] avg loss 0.00227901, throughput 6.00041K wps
[Epoch 6 Batch 1350/2125] avg loss 0.00181659, throughput 6.01915K wps
[Epoch 6 Batch 1380/2125] avg loss 0.00144348, throughput 6.01553K wps
[Epoch 6 Batch 1410/2125] avg loss 0.00182094, throughput 6.02111K wps
[Epoch 6 Batch 1440/2125] avg loss 0.00207621, throughput 6.0164K wps
[Epoch 6 Batch 1470/2125] avg loss 0.00174451, throughput 6.0301K wps
[Epoch 6 Batch 1500/2125] avg loss 0.00198384, throughput 6.02397K wps
[Epoch 6 Batch 1530/2125] avg loss 0.00214062, throughput 6.02576K wps
[Epoch 6 Batch 1560/2125] avg loss 0.00154858, throughput 6.02461K wps
[Epoch 6 Batch 1590/2125] avg loss 0.00186165, throughput 6.02458K wps
[Epoch 6 Batch 1620/2125] avg loss 0.00193522, throughput 6.01978K wps
[Epoch 6 Batch 1650/2125] avg loss 0.00186315, throughput 6.00939K wps
[Epoch 6 Batch 1680/2125] avg loss 0.00166569, throughput 6.01436K wps
[Epoch 6 Batch 1710/2125] avg loss 0.00217489, throughput 6.01442K wps
[Epoch 6 Batch 1740/2125] avg loss 0.00169276, throughput 6.02607K wps
[Epoch 6 Batch 1770/2125] avg loss 0.00172876, throughput 6.02117K wps
[Epoch 6 Batch 1800/2125] avg loss 0.00235071, throughput 6.02515K wps
[Epoch 6 Batch 1830/2125] avg loss 0.00247296, throughput 6.01836K wps
[Epoch 6 Batch 1860/2125] avg loss 0.00172835, throughput 6.0113K wps
[Epoch 6 Batch 1890/2125] avg loss 0.00157904, throughput 6.01913K wps
[Epoch 6 Batch 1920/2125] avg loss 0.00271029, throughput 6.02813K wps
[Epoch 6 Batch 1950/2125] avg loss 0.00191968, throughput 6.02262K wps
[Epoch 6 Batch 1980/2125] avg loss 0.00198379, throughput 6.01343K wps
[Epoch 6 Batch 2010/2125] avg loss 0.00194769, throughput 6.01069K wps
[Epoch 6 Batch 2040/2125] avg loss 0.00173524, throughput 6.02112K wps
[Epoch 6 Batch 2070/2125] avg loss 0.00212147, throughput 6.02804K wps
[Epoch 6 Batch 2100/2125] avg loss 0.0023538, throughput 6.02476K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 6] train avg loss 0.00177826, test acc 0.9263, test avg loss 0.265486, throughput 6.02212K wps
[Epoch 7 Batch 30/2125] avg loss 0.00138391, throughput 6.15913K wps
[Epoch 7 Batch 60/2125] avg loss 0.00144632, throughput 6.01288K wps
[Epoch 7 Batch 90/2125] avg loss 0.0012511, throughput 6.0161K wps
[Epoch 7 Batch 120/2125] avg loss 0.00136493, throughput 6.00793K wps
[Epoch 7 Batch 150/2125] avg loss 0.00122343, throughput 6.02266K wps
[Epoch 7 Batch 180/2125] avg loss 0.00136592, throughput 6.01762K wps
[Epoch 7 Batch 210/2125] avg loss 0.00130141, throughput 6.01188K wps
[Epoch 7 Batch 240/2125] avg loss 0.00106083, throughput 6.0227K wps
[Epoch 7 Batch 270/2125] avg loss 0.0014105, throughput 6.01135K wps
[Epoch 7 Batch 300/2125] avg loss 0.00162967, throughput 6.01143K wps
[Epoch 7 Batch 330/2125] avg loss 0.00142019, throughput 6.00941K wps
[Epoch 7 Batch 360/2125] avg loss 0.00141422, throughput 6.00225K wps
[Epoch 7 Batch 390/2125] avg loss 0.00124699, throughput 5.99908K wps
[Epoch 7 Batch 420/2125] avg loss 0.00150662, throughput 6.01749K wps
[Epoch 7 Batch 450/2125] avg loss 0.00134305, throughput 6.01555K wps
[Epoch 7 Batch 480/2125] avg loss 0.00115115, throughput 6.0168K wps
[Epoch 7 Batch 510/2125] avg loss 0.00162686, throughput 6.01759K wps
[Epoch 7 Batch 540/2125] avg loss 0.00199012, throughput 6.01895K wps
[Epoch 7 Batch 570/2125] avg loss 0.0016047, throughput 6.01207K wps
[Epoch 7 Batch 600/2125] avg loss 0.00134874, throughput 6.01205K wps
[Epoch 7 Batch 630/2125] avg loss 0.00132161, throughput 6.00142K wps
[Epoch 7 Batch 660/2125] avg loss 0.00137177, throughput 6.00224K wps
[Epoch 7 Batch 690/2125] avg loss 0.00145827, throughput 6.00869K wps
[Epoch 7 Batch 720/2125] avg loss 0.00136913, throughput 6.01565K wps
[Epoch 7 Batch 750/2125] avg loss 0.00167497, throughput 6.01713K wps
[Epoch 7 Batch 780/2125] avg loss 0.00121769, throughput 6.01469K wps
[Epoch 7 Batch 810/2125] avg loss 0.00181145, throughput 6.01506K wps
[Epoch 7 Batch 840/2125] avg loss 0.00140501, throughput 6.012K wps
[Epoch 7 Batch 870/2125] avg loss 0.00154237, throughput 6.02024K wps
[Epoch 7 Batch 900/2125] avg loss 0.00134491, throughput 6.02847K wps
[Epoch 7 Batch 930/2125] avg loss 0.00125617, throughput 6.01883K wps
[Epoch 7 Batch 960/2125] avg loss 0.00185526, throughput 6.0139K wps
[Epoch 7 Batch 990/2125] avg loss 0.00157221, throughput 6.0126K wps
[Epoch 7 Batch 1020/2125] avg loss 0.00154186, throughput 6.01933K wps
[Epoch 7 Batch 1050/2125] avg loss 0.00162592, throughput 6.00922K wps
[Epoch 7 Batch 1080/2125] avg loss 0.00162722, throughput 6.01486K wps
[Epoch 7 Batch 1110/2125] avg loss 0.00180182, throughput 6.01074K wps
[Epoch 7 Batch 1140/2125] avg loss 0.00204016, throughput 6.00733K wps
[Epoch 7 Batch 1170/2125] avg loss 0.00191146, throughput 6.01985K wps
[Epoch 7 Batch 1200/2125] avg loss 0.0014543, throughput 6.01607K wps
[Epoch 7 Batch 1230/2125] avg loss 0.0015092, throughput 6.0146K wps
[Epoch 7 Batch 1260/2125] avg loss 0.00167479, throughput 6.01437K wps
[Epoch 7 Batch 1290/2125] avg loss 0.00117421, throughput 6.00134K wps
[Epoch 7 Batch 1320/2125] avg loss 0.00180587, throughput 6.02072K wps
[Epoch 7 Batch 1350/2125] avg loss 0.00142445, throughput 6.01448K wps
[Epoch 7 Batch 1380/2125] avg loss 0.00184895, throughput 6.00705K wps
[Epoch 7 Batch 1410/2125] avg loss 0.0014999, throughput 6.01452K wps
[Epoch 7 Batch 1440/2125] avg loss 0.00146374, throughput 6.02128K wps
[Epoch 7 Batch 1470/2125] avg loss 0.00164917, throughput 6.01768K wps
[Epoch 7 Batch 1500/2125] avg loss 0.00121465, throughput 6.01936K wps
[Epoch 7 Batch 1530/2125] avg loss 0.00188805, throughput 6.02653K wps
[Epoch 7 Batch 1560/2125] avg loss 0.00155888, throughput 6.02329K wps
[Epoch 7 Batch 1590/2125] avg loss 0.00185553, throughput 6.02579K wps
[Epoch 7 Batch 1620/2125] avg loss 0.00172886, throughput 6.02417K wps
[Epoch 7 Batch 1650/2125] avg loss 0.00221982, throughput 6.01969K wps
[Epoch 7 Batch 1680/2125] avg loss 0.00187133, throughput 6.01379K wps
[Epoch 7 Batch 1710/2125] avg loss 0.00164991, throughput 6.01124K wps
[Epoch 7 Batch 1740/2125] avg loss 0.00199331, throughput 6.02625K wps
[Epoch 7 Batch 1770/2125] avg loss 0.00141228, throughput 6.02251K wps
[Epoch 7 Batch 1800/2125] avg loss 0.00175644, throughput 6.00732K wps
[Epoch 7 Batch 1830/2125] avg loss 0.00205022, throughput 6.01073K wps
[Epoch 7 Batch 1860/2125] avg loss 0.00194963, throughput 6.02183K wps
[Epoch 7 Batch 1890/2125] avg loss 0.00150715, throughput 6.01553K wps
[Epoch 7 Batch 1920/2125] avg loss 0.00189169, throughput 6.01508K wps
[Epoch 7 Batch 1950/2125] avg loss 0.00182158, throughput 6.01283K wps
[Epoch 7 Batch 1980/2125] avg loss 0.00152759, throughput 6.01923K wps
[Epoch 7 Batch 2010/2125] avg loss 0.00187992, throughput 6.0235K wps
[Epoch 7 Batch 2040/2125] avg loss 0.00187159, throughput 6.02249K wps
[Epoch 7 Batch 2070/2125] avg loss 0.00186872, throughput 6.00769K wps
[Epoch 7 Batch 2100/2125] avg loss 0.00186226, throughput 6.0197K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 7] train avg loss 0.00158719, test acc 0.9267, test avg loss 0.280398, throughput 6.01729K wps
[Epoch 8 Batch 30/2125] avg loss 0.00118313, throughput 6.16944K wps
[Epoch 8 Batch 60/2125] avg loss 0.00103229, throughput 6.01234K wps
[Epoch 8 Batch 90/2125] avg loss 0.00115928, throughput 6.02321K wps
[Epoch 8 Batch 120/2125] avg loss 0.00128744, throughput 6.01534K wps
[Epoch 8 Batch 150/2125] avg loss 0.0011415, throughput 6.02214K wps
[Epoch 8 Batch 180/2125] avg loss 0.00119047, throughput 6.01959K wps
[Epoch 8 Batch 210/2125] avg loss 0.00121781, throughput 5.99558K wps
[Epoch 8 Batch 240/2125] avg loss 0.00126624, throughput 6.01533K wps
[Epoch 8 Batch 270/2125] avg loss 0.00157652, throughput 6.01508K wps
[Epoch 8 Batch 300/2125] avg loss 0.00140678, throughput 6.01988K wps
[Epoch 8 Batch 330/2125] avg loss 0.00126451, throughput 6.01458K wps
[Epoch 8 Batch 360/2125] avg loss 0.00116266, throughput 6.02513K wps
[Epoch 8 Batch 390/2125] avg loss 0.00133267, throughput 6.01997K wps
[Epoch 8 Batch 420/2125] avg loss 0.00134687, throughput 6.02202K wps
[Epoch 8 Batch 450/2125] avg loss 0.0014634, throughput 6.01535K wps
[Epoch 8 Batch 480/2125] avg loss 0.00118032, throughput 6.02309K wps
[Epoch 8 Batch 510/2125] avg loss 0.00146095, throughput 6.02544K wps
[Epoch 8 Batch 540/2125] avg loss 0.0014587, throughput 6.01473K wps
[Epoch 8 Batch 570/2125] avg loss 0.00146558, throughput 6.0277K wps
[Epoch 8 Batch 600/2125] avg loss 0.00133723, throughput 6.02472K wps
[Epoch 8 Batch 630/2125] avg loss 0.00110727, throughput 6.01801K wps
[Epoch 8 Batch 660/2125] avg loss 0.00137143, throughput 6.01331K wps
[Epoch 8 Batch 690/2125] avg loss 0.00160191, throughput 6.02431K wps
[Epoch 8 Batch 720/2125] avg loss 0.00117946, throughput 6.0219K wps
[Epoch 8 Batch 750/2125] avg loss 0.00165824, throughput 6.01452K wps
[Epoch 8 Batch 780/2125] avg loss 0.00117383, throughput 6.02555K wps
[Epoch 8 Batch 810/2125] avg loss 0.00139891, throughput 5.98497K wps
[Epoch 8 Batch 840/2125] avg loss 0.001541, throughput 5.99543K wps
[Epoch 8 Batch 870/2125] avg loss 0.00154169, throughput 6.01287K wps
[Epoch 8 Batch 900/2125] avg loss 0.00131683, throughput 6.01173K wps
[Epoch 8 Batch 930/2125] avg loss 0.00123108, throughput 6.00593K wps
[Epoch 8 Batch 960/2125] avg loss 0.00153428, throughput 6.01741K wps
[Epoch 8 Batch 990/2125] avg loss 0.00167105, throughput 6.01922K wps
[Epoch 8 Batch 1020/2125] avg loss 0.00150442, throughput 6.01537K wps
[Epoch 8 Batch 1050/2125] avg loss 0.00149242, throughput 6.01913K wps
[Epoch 8 Batch 1080/2125] avg loss 0.00128549, throughput 6.01056K wps
[Epoch 8 Batch 1110/2125] avg loss 0.00138098, throughput 6.02725K wps
[Epoch 8 Batch 1140/2125] avg loss 0.00150342, throughput 6.01501K wps
[Epoch 8 Batch 1170/2125] avg loss 0.00155106, throughput 6.02463K wps
[Epoch 8 Batch 1200/2125] avg loss 0.00146795, throughput 6.02137K wps
[Epoch 8 Batch 1230/2125] avg loss 0.00128118, throughput 6.01237K wps
[Epoch 8 Batch 1260/2125] avg loss 0.00130548, throughput 6.0313K wps
[Epoch 8 Batch 1290/2125] avg loss 0.00163696, throughput 6.02677K wps
[Epoch 8 Batch 1320/2125] avg loss 0.0014831, throughput 6.003K wps
[Epoch 8 Batch 1350/2125] avg loss 0.001374, throughput 6.00561K wps
[Epoch 8 Batch 1380/2125] avg loss 0.00177027, throughput 6.03038K wps
[Epoch 8 Batch 1410/2125] avg loss 0.00144212, throughput 6.01974K wps
[Epoch 8 Batch 1440/2125] avg loss 0.00142428, throughput 6.02418K wps
[Epoch 8 Batch 1470/2125] avg loss 0.00155867, throughput 6.02216K wps
[Epoch 8 Batch 1500/2125] avg loss 0.00151979, throughput 6.00798K wps
[Epoch 8 Batch 1530/2125] avg loss 0.00181596, throughput 6.01994K wps
[Epoch 8 Batch 1560/2125] avg loss 0.00159266, throughput 6.02719K wps
[Epoch 8 Batch 1590/2125] avg loss 0.0013084, throughput 6.01571K wps
[Epoch 8 Batch 1620/2125] avg loss 0.00139701, throughput 6.01546K wps
[Epoch 8 Batch 1650/2125] avg loss 0.00152573, throughput 6.02627K wps
[Epoch 8 Batch 1680/2125] avg loss 0.00122975, throughput 6.02299K wps
[Epoch 8 Batch 1710/2125] avg loss 0.00143349, throughput 6.02038K wps
[Epoch 8 Batch 1740/2125] avg loss 0.00165565, throughput 6.02008K wps
[Epoch 8 Batch 1770/2125] avg loss 0.00182221, throughput 5.97152K wps
[Epoch 8 Batch 1800/2125] avg loss 0.001575, throughput 6.00129K wps
[Epoch 8 Batch 1830/2125] avg loss 0.00151679, throughput 6.01527K wps
[Epoch 8 Batch 1860/2125] avg loss 0.00156741, throughput 6.00701K wps
[Epoch 8 Batch 1890/2125] avg loss 0.00131145, throughput 6.01097K wps
[Epoch 8 Batch 1920/2125] avg loss 0.00144851, throughput 6.01112K wps
[Epoch 8 Batch 1950/2125] avg loss 0.00142256, throughput 6.01013K wps
[Epoch 8 Batch 1980/2125] avg loss 0.00177985, throughput 6.01671K wps
[Epoch 8 Batch 2010/2125] avg loss 0.00153835, throughput 6.01914K wps
[Epoch 8 Batch 2040/2125] avg loss 0.00185871, throughput 6.00884K wps
[Epoch 8 Batch 2070/2125] avg loss 0.00185945, throughput 6.01032K wps
[Epoch 8 Batch 2100/2125] avg loss 0.00130488, throughput 6.02301K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 8] train avg loss 0.00143642, test acc 0.9258, test avg loss 0.300093, throughput 6.01822K wps
[Epoch 9 Batch 30/2125] avg loss 0.000969259, throughput 6.15332K wps
[Epoch 9 Batch 60/2125] avg loss 0.0011693, throughput 6.01684K wps
[Epoch 9 Batch 90/2125] avg loss 0.000880694, throughput 6.02025K wps
[Epoch 9 Batch 120/2125] avg loss 0.00105629, throughput 6.01855K wps
[Epoch 9 Batch 150/2125] avg loss 0.00110347, throughput 6.02174K wps
[Epoch 9 Batch 180/2125] avg loss 0.000885239, throughput 6.01683K wps
[Epoch 9 Batch 210/2125] avg loss 0.000924654, throughput 6.01687K wps
[Epoch 9 Batch 240/2125] avg loss 0.00114026, throughput 6.01165K wps
[Epoch 9 Batch 270/2125] avg loss 0.00136323, throughput 6.01408K wps
[Epoch 9 Batch 300/2125] avg loss 0.00122665, throughput 6.01633K wps
[Epoch 9 Batch 330/2125] avg loss 0.00101408, throughput 6.03399K wps
[Epoch 9 Batch 360/2125] avg loss 0.00131196, throughput 6.00678K wps
[Epoch 9 Batch 390/2125] avg loss 0.00118139, throughput 6.00527K wps
[Epoch 9 Batch 420/2125] avg loss 0.00134069, throughput 6.02231K wps
[Epoch 9 Batch 450/2125] avg loss 0.00128085, throughput 6.03223K wps
[Epoch 9 Batch 480/2125] avg loss 0.00120814, throughput 6.02938K wps
[Epoch 9 Batch 510/2125] avg loss 0.0012161, throughput 6.01559K wps
[Epoch 9 Batch 540/2125] avg loss 0.00128244, throughput 6.01672K wps
[Epoch 9 Batch 570/2125] avg loss 0.00112483, throughput 6.01173K wps
[Epoch 9 Batch 600/2125] avg loss 0.00135648, throughput 6.00779K wps
[Epoch 9 Batch 630/2125] avg loss 0.0012007, throughput 6.01362K wps
[Epoch 9 Batch 660/2125] avg loss 0.00101158, throughput 6.02409K wps
[Epoch 9 Batch 690/2125] avg loss 0.00128408, throughput 6.02907K wps
[Epoch 9 Batch 720/2125] avg loss 0.00140932, throughput 6.03003K wps
[Epoch 9 Batch 750/2125] avg loss 0.00132323, throughput 6.02063K wps
[Epoch 9 Batch 780/2125] avg loss 0.0011177, throughput 6.0165K wps
[Epoch 9 Batch 810/2125] avg loss 0.00130684, throughput 6.01982K wps
[Epoch 9 Batch 840/2125] avg loss 0.00104136, throughput 6.02558K wps
[Epoch 9 Batch 870/2125] avg loss 0.00129281, throughput 6.02152K wps
[Epoch 9 Batch 900/2125] avg loss 0.00117668, throughput 6.02065K wps
[Epoch 9 Batch 930/2125] avg loss 0.00117492, throughput 6.01698K wps
[Epoch 9 Batch 960/2125] avg loss 0.00137053, throughput 6.02799K wps
[Epoch 9 Batch 990/2125] avg loss 0.00119182, throughput 6.02296K wps
[Epoch 9 Batch 1020/2125] avg loss 0.00122679, throughput 6.02139K wps
[Epoch 9 Batch 1050/2125] avg loss 0.00123281, throughput 6.01148K wps
[Epoch 9 Batch 1080/2125] avg loss 0.0015326, throughput 6.025K wps
[Epoch 9 Batch 1110/2125] avg loss 0.00127969, throughput 6.02339K wps
[Epoch 9 Batch 1140/2125] avg loss 0.00132585, throughput 6.01758K wps
[Epoch 9 Batch 1170/2125] avg loss 0.00117789, throughput 6.02279K wps
[Epoch 9 Batch 1200/2125] avg loss 0.00168645, throughput 6.01745K wps
[Epoch 9 Batch 1230/2125] avg loss 0.00130571, throughput 6.01714K wps
[Epoch 9 Batch 1260/2125] avg loss 0.00154435, throughput 6.02337K wps
[Epoch 9 Batch 1290/2125] avg loss 0.00148727, throughput 6.02328K wps
[Epoch 9 Batch 1320/2125] avg loss 0.00125417, throughput 6.02381K wps
[Epoch 9 Batch 1350/2125] avg loss 0.00137377, throughput 6.02153K wps
[Epoch 9 Batch 1380/2125] avg loss 0.00138344, throughput 6.02026K wps
[Epoch 9 Batch 1410/2125] avg loss 0.00131722, throughput 6.02709K wps
[Epoch 9 Batch 1440/2125] avg loss 0.00144407, throughput 6.01192K wps
[Epoch 9 Batch 1470/2125] avg loss 0.00163246, throughput 6.01234K wps
[Epoch 9 Batch 1500/2125] avg loss 0.0015646, throughput 6.02548K wps
[Epoch 9 Batch 1530/2125] avg loss 0.00142445, throughput 6.02376K wps
[Epoch 9 Batch 1560/2125] avg loss 0.00138671, throughput 6.02692K wps
[Epoch 9 Batch 1590/2125] avg loss 0.00139743, throughput 6.03028K wps
[Epoch 9 Batch 1620/2125] avg loss 0.00121396, throughput 6.01977K wps
[Epoch 9 Batch 1650/2125] avg loss 0.00140288, throughput 6.02442K wps
[Epoch 9 Batch 1680/2125] avg loss 0.00156761, throughput 6.02991K wps
[Epoch 9 Batch 1710/2125] avg loss 0.00149254, throughput 6.02511K wps
[Epoch 9 Batch 1740/2125] avg loss 0.00174833, throughput 6.03226K wps
[Epoch 9 Batch 1770/2125] avg loss 0.0015996, throughput 6.0168K wps
[Epoch 9 Batch 1800/2125] avg loss 0.00155534, throughput 6.01864K wps
[Epoch 9 Batch 1830/2125] avg loss 0.0015644, throughput 6.02337K wps
[Epoch 9 Batch 1860/2125] avg loss 0.00185167, throughput 6.02149K wps
[Epoch 9 Batch 1890/2125] avg loss 0.00162232, throughput 6.02119K wps
[Epoch 9 Batch 1920/2125] avg loss 0.00140311, throughput 6.01599K wps
[Epoch 9 Batch 1950/2125] avg loss 0.00166109, throughput 6.01052K wps
[Epoch 9 Batch 1980/2125] avg loss 0.00163539, throughput 6.02679K wps
[Epoch 9 Batch 2010/2125] avg loss 0.00126912, throughput 6.01843K wps
[Epoch 9 Batch 2040/2125] avg loss 0.00153095, throughput 6.01568K wps
[Epoch 9 Batch 2070/2125] avg loss 0.00154654, throughput 6.01734K wps
[Epoch 9 Batch 2100/2125] avg loss 0.00118063, throughput 6.02042K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 9] train avg loss 0.00132753, test acc 0.9264, test avg loss 0.310232, throughput 6.02234K wps
[Epoch 10 Batch 30/2125] avg loss 0.000945357, throughput 6.16222K wps
[Epoch 10 Batch 60/2125] avg loss 0.0010266, throughput 6.01541K wps
[Epoch 10 Batch 90/2125] avg loss 0.00133443, throughput 6.01228K wps
[Epoch 10 Batch 120/2125] avg loss 0.00116895, throughput 6.02132K wps
[Epoch 10 Batch 150/2125] avg loss 0.00085818, throughput 6.01387K wps
[Epoch 10 Batch 180/2125] avg loss 0.00104651, throughput 6.01404K wps
[Epoch 10 Batch 210/2125] avg loss 0.00107097, throughput 6.01559K wps
[Epoch 10 Batch 240/2125] avg loss 0.0010347, throughput 6.02622K wps
[Epoch 10 Batch 270/2125] avg loss 0.00101927, throughput 6.01865K wps
[Epoch 10 Batch 300/2125] avg loss 0.00126273, throughput 6.03022K wps
[Epoch 10 Batch 330/2125] avg loss 0.00100666, throughput 6.01894K wps
[Epoch 10 Batch 360/2125] avg loss 0.00123826, throughput 6.01596K wps
[Epoch 10 Batch 390/2125] avg loss 0.00127604, throughput 6.01302K wps
[Epoch 10 Batch 420/2125] avg loss 0.00122692, throughput 6.01977K wps
[Epoch 10 Batch 450/2125] avg loss 0.00149378, throughput 6.0172K wps
[Epoch 10 Batch 480/2125] avg loss 0.00123578, throughput 6.01546K wps
[Epoch 10 Batch 510/2125] avg loss 0.00112775, throughput 6.01487K wps
[Epoch 10 Batch 540/2125] avg loss 0.00139981, throughput 6.02091K wps
[Epoch 10 Batch 570/2125] avg loss 0.000953831, throughput 6.01519K wps
[Epoch 10 Batch 600/2125] avg loss 0.000996425, throughput 6.01146K wps
[Epoch 10 Batch 630/2125] avg loss 0.00124566, throughput 6.01722K wps
[Epoch 10 Batch 660/2125] avg loss 0.00122947, throughput 6.0256K wps
[Epoch 10 Batch 690/2125] avg loss 0.00107161, throughput 6.01062K wps
[Epoch 10 Batch 720/2125] avg loss 0.00131284, throughput 6.01694K wps
[Epoch 10 Batch 750/2125] avg loss 0.00108788, throughput 6.01941K wps
[Epoch 10 Batch 780/2125] avg loss 0.00105862, throughput 6.00563K wps
[Epoch 10 Batch 810/2125] avg loss 0.00119511, throughput 6.00536K wps
[Epoch 10 Batch 840/2125] avg loss 0.00123623, throughput 6.01732K wps
[Epoch 10 Batch 870/2125] avg loss 0.00116228, throughput 6.02465K wps
[Epoch 10 Batch 900/2125] avg loss 0.0012188, throughput 6.02179K wps
[Epoch 10 Batch 930/2125] avg loss 0.00108208, throughput 6.01295K wps
[Epoch 10 Batch 960/2125] avg loss 0.00138802, throughput 6.01354K wps
[Epoch 10 Batch 990/2125] avg loss 0.00129621, throughput 6.00209K wps
[Epoch 10 Batch 1020/2125] avg loss 0.00112338, throughput 6.01444K wps
[Epoch 10 Batch 1050/2125] avg loss 0.00117441, throughput 6.01454K wps
[Epoch 10 Batch 1080/2125] avg loss 0.00123106, throughput 6.02048K wps
[Epoch 10 Batch 1110/2125] avg loss 0.00106179, throughput 6.01585K wps
[Epoch 10 Batch 1140/2125] avg loss 0.00103955, throughput 6.01094K wps
[Epoch 10 Batch 1170/2125] avg loss 0.00118204, throughput 6.02165K wps
[Epoch 10 Batch 1200/2125] avg loss 0.00113691, throughput 6.02374K wps
[Epoch 10 Batch 1230/2125] avg loss 0.00105751, throughput 6.01713K wps
[Epoch 10 Batch 1260/2125] avg loss 0.00113869, throughput 6.01997K wps
[Epoch 10 Batch 1290/2125] avg loss 0.000979387, throughput 6.01012K wps
[Epoch 10 Batch 1320/2125] avg loss 0.00143551, throughput 6.02835K wps
[Epoch 10 Batch 1350/2125] avg loss 0.00125474, throughput 6.01088K wps
[Epoch 10 Batch 1380/2125] avg loss 0.00122672, throughput 6.02501K wps
[Epoch 10 Batch 1410/2125] avg loss 0.00116667, throughput 6.01443K wps
[Epoch 10 Batch 1440/2125] avg loss 0.0011223, throughput 6.01753K wps
[Epoch 10 Batch 1470/2125] avg loss 0.00123099, throughput 6.02383K wps
[Epoch 10 Batch 1500/2125] avg loss 0.00131237, throughput 6.02562K wps
[Epoch 10 Batch 1530/2125] avg loss 0.00119264, throughput 6.01518K wps
[Epoch 10 Batch 1560/2125] avg loss 0.00136504, throughput 6.00845K wps
[Epoch 10 Batch 1590/2125] avg loss 0.00132757, throughput 6.01974K wps
[Epoch 10 Batch 1620/2125] avg loss 0.0015589, throughput 6.0236K wps
[Epoch 10 Batch 1650/2125] avg loss 0.00130379, throughput 6.01339K wps
[Epoch 10 Batch 1680/2125] avg loss 0.00168383, throughput 6.00995K wps
[Epoch 10 Batch 1710/2125] avg loss 0.00138744, throughput 6.00376K wps
[Epoch 10 Batch 1740/2125] avg loss 0.00133728, throughput 6.02386K wps
[Epoch 10 Batch 1770/2125] avg loss 0.00122172, throughput 6.01825K wps
[Epoch 10 Batch 1800/2125] avg loss 0.00146909, throughput 6.00923K wps
[Epoch 10 Batch 1830/2125] avg loss 0.00133232, throughput 6.01002K wps
[Epoch 10 Batch 1860/2125] avg loss 0.00129973, throughput 5.99266K wps
[Epoch 10 Batch 1890/2125] avg loss 0.0016672, throughput 5.99162K wps
[Epoch 10 Batch 1920/2125] avg loss 0.00116836, throughput 6.02214K wps
[Epoch 10 Batch 1950/2125] avg loss 0.00120944, throughput 6.02604K wps
[Epoch 10 Batch 1980/2125] avg loss 0.00121911, throughput 6.01723K wps
[Epoch 10 Batch 2010/2125] avg loss 0.00170782, throughput 6.01955K wps
[Epoch 10 Batch 2040/2125] avg loss 0.0016061, throughput 6.02669K wps
[Epoch 10 Batch 2070/2125] avg loss 0.00134302, throughput 6.01626K wps
[Epoch 10 Batch 2100/2125] avg loss 0.00112688, throughput 6.01026K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 10] train avg loss 0.00122541, test acc 0.9266, test avg loss 0.329704, throughput 6.01833K wps
[Epoch 11 Batch 30/2125] avg loss 0.000871779, throughput 6.15935K wps
[Epoch 11 Batch 60/2125] avg loss 0.00114995, throughput 6.0271K wps
[Epoch 11 Batch 90/2125] avg loss 0.00100125, throughput 6.02005K wps
[Epoch 11 Batch 120/2125] avg loss 0.000825593, throughput 6.01108K wps
[Epoch 11 Batch 150/2125] avg loss 0.000918682, throughput 6.01376K wps
[Epoch 11 Batch 180/2125] avg loss 0.00104146, throughput 6.01439K wps
[Epoch 11 Batch 210/2125] avg loss 0.000834111, throughput 6.01118K wps
[Epoch 11 Batch 240/2125] avg loss 0.000824702, throughput 6.01038K wps
[Epoch 11 Batch 270/2125] avg loss 0.000887079, throughput 6.00336K wps
[Epoch 11 Batch 300/2125] avg loss 0.000816447, throughput 6.00762K wps
[Epoch 11 Batch 330/2125] avg loss 0.00100532, throughput 6.01002K wps
[Epoch 11 Batch 360/2125] avg loss 0.00104165, throughput 6.01865K wps
[Epoch 11 Batch 390/2125] avg loss 0.000861591, throughput 6.01766K wps
[Epoch 11 Batch 420/2125] avg loss 0.000796927, throughput 6.0099K wps
[Epoch 11 Batch 450/2125] avg loss 0.00101545, throughput 6.02204K wps
[Epoch 11 Batch 480/2125] avg loss 0.00120154, throughput 6.01031K wps
[Epoch 11 Batch 510/2125] avg loss 0.0011003, throughput 6.01709K wps
[Epoch 11 Batch 540/2125] avg loss 0.00100812, throughput 6.01502K wps
[Epoch 11 Batch 570/2125] avg loss 0.00118204, throughput 6.01504K wps
[Epoch 11 Batch 600/2125] avg loss 0.00106542, throughput 6.0174K wps
[Epoch 11 Batch 630/2125] avg loss 0.00119238, throughput 6.02205K wps
[Epoch 11 Batch 660/2125] avg loss 0.000968815, throughput 6.01697K wps
[Epoch 11 Batch 690/2125] avg loss 0.000972758, throughput 6.02032K wps
[Epoch 11 Batch 720/2125] avg loss 0.000951895, throughput 6.01912K wps
[Epoch 11 Batch 750/2125] avg loss 0.00124134, throughput 6.0149K wps
[Epoch 11 Batch 780/2125] avg loss 0.0010466, throughput 6.00883K wps
[Epoch 11 Batch 810/2125] avg loss 0.0010532, throughput 6.01635K wps
[Epoch 11 Batch 840/2125] avg loss 0.00135194, throughput 6.01854K wps
[Epoch 11 Batch 870/2125] avg loss 0.00107797, throughput 6.0118K wps
[Epoch 11 Batch 900/2125] avg loss 0.00122478, throughput 6.02094K wps
[Epoch 11 Batch 930/2125] avg loss 0.00119027, throughput 6.01395K wps
[Epoch 11 Batch 960/2125] avg loss 0.00128709, throughput 6.00514K wps
[Epoch 11 Batch 990/2125] avg loss 0.00106464, throughput 6.02722K wps
[Epoch 11 Batch 1020/2125] avg loss 0.00127232, throughput 6.02677K wps
[Epoch 11 Batch 1050/2125] avg loss 0.00113073, throughput 6.01069K wps
[Epoch 11 Batch 1080/2125] avg loss 0.00140409, throughput 6.00576K wps
[Epoch 11 Batch 1110/2125] avg loss 0.00146768, throughput 6.01187K wps
[Epoch 11 Batch 1140/2125] avg loss 0.00115781, throughput 6.00807K wps
[Epoch 11 Batch 1170/2125] avg loss 0.00124002, throughput 6.00731K wps
[Epoch 11 Batch 1200/2125] avg loss 0.00137068, throughput 6.01601K wps
[Epoch 11 Batch 1230/2125] avg loss 0.00112048, throughput 6.01075K wps
[Epoch 11 Batch 1260/2125] avg loss 0.0010931, throughput 6.00297K wps
[Epoch 11 Batch 1290/2125] avg loss 0.000907075, throughput 6.00519K wps
[Epoch 11 Batch 1320/2125] avg loss 0.00152328, throughput 6.01479K wps
[Epoch 11 Batch 1350/2125] avg loss 0.00110495, throughput 6.00907K wps
[Epoch 11 Batch 1380/2125] avg loss 0.00150946, throughput 6.00716K wps
[Epoch 11 Batch 1410/2125] avg loss 0.00107715, throughput 6.01002K wps
[Epoch 11 Batch 1440/2125] avg loss 0.00125197, throughput 6.01223K wps
[Epoch 11 Batch 1470/2125] avg loss 0.00148732, throughput 6.0135K wps
[Epoch 11 Batch 1500/2125] avg loss 0.0010248, throughput 6.00459K wps
[Epoch 11 Batch 1530/2125] avg loss 0.00111837, throughput 6.00362K wps
[Epoch 11 Batch 1560/2125] avg loss 0.00130822, throughput 6.00158K wps
[Epoch 11 Batch 1590/2125] avg loss 0.00123279, throughput 6.01285K wps
[Epoch 11 Batch 1620/2125] avg loss 0.00142636, throughput 6.01107K wps
[Epoch 11 Batch 1650/2125] avg loss 0.00121303, throughput 6.00793K wps
[Epoch 11 Batch 1680/2125] avg loss 0.0013154, throughput 6.0087K wps
[Epoch 11 Batch 1710/2125] avg loss 0.00114358, throughput 6.00854K wps
[Epoch 11 Batch 1740/2125] avg loss 0.00105089, throughput 6.00459K wps
[Epoch 11 Batch 1770/2125] avg loss 0.00150311, throughput 5.99859K wps
[Epoch 11 Batch 1800/2125] avg loss 0.00127863, throughput 6.01231K wps
[Epoch 11 Batch 1830/2125] avg loss 0.00117567, throughput 6.00747K wps
[Epoch 11 Batch 1860/2125] avg loss 0.000966579, throughput 6.02073K wps
[Epoch 11 Batch 1890/2125] avg loss 0.00117558, throughput 6.01669K wps
[Epoch 11 Batch 1920/2125] avg loss 0.00107088, throughput 6.01453K wps
[Epoch 11 Batch 1950/2125] avg loss 0.001303, throughput 6.01275K wps
[Epoch 11 Batch 1980/2125] avg loss 0.0015025, throughput 6.00978K wps
[Epoch 11 Batch 2010/2125] avg loss 0.00133716, throughput 6.01023K wps
[Epoch 11 Batch 2040/2125] avg loss 0.00109812, throughput 6.01888K wps
[Epoch 11 Batch 2070/2125] avg loss 0.00136779, throughput 6.028K wps
[Epoch 11 Batch 2100/2125] avg loss 0.00138631, throughput 6.01259K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 11] train avg loss 0.00114561, test acc 0.9256, test avg loss 0.342964, throughput 6.015K wps
[Epoch 12 Batch 30/2125] avg loss 0.000971687, throughput 6.15241K wps
[Epoch 12 Batch 60/2125] avg loss 0.000842094, throughput 6.02106K wps
[Epoch 12 Batch 90/2125] avg loss 0.000750695, throughput 6.0101K wps
[Epoch 12 Batch 120/2125] avg loss 0.00111169, throughput 6.01065K wps
[Epoch 12 Batch 150/2125] avg loss 0.000913355, throughput 6.00637K wps
[Epoch 12 Batch 180/2125] avg loss 0.000687627, throughput 6.0153K wps
[Epoch 12 Batch 210/2125] avg loss 0.000840865, throughput 6.03082K wps
[Epoch 12 Batch 240/2125] avg loss 0.00119181, throughput 6.01475K wps
[Epoch 12 Batch 270/2125] avg loss 0.000950309, throughput 6.01129K wps
[Epoch 12 Batch 300/2125] avg loss 0.00101251, throughput 6.0088K wps
[Epoch 12 Batch 330/2125] avg loss 0.000963666, throughput 6.01012K wps
[Epoch 12 Batch 360/2125] avg loss 0.00106978, throughput 6.00924K wps
[Epoch 12 Batch 390/2125] avg loss 0.00113868, throughput 6.01384K wps
[Epoch 12 Batch 420/2125] avg loss 0.000935335, throughput 6.00727K wps
[Epoch 12 Batch 450/2125] avg loss 0.00104527, throughput 6.00933K wps
[Epoch 12 Batch 480/2125] avg loss 0.00100805, throughput 6.00805K wps
[Epoch 12 Batch 510/2125] avg loss 0.00112906, throughput 5.99436K wps
[Epoch 12 Batch 540/2125] avg loss 0.000769744, throughput 6.00029K wps
[Epoch 12 Batch 570/2125] avg loss 0.000796751, throughput 6.00923K wps
[Epoch 12 Batch 600/2125] avg loss 0.00119287, throughput 6.02134K wps
[Epoch 12 Batch 630/2125] avg loss 0.00101279, throughput 6.00473K wps
[Epoch 12 Batch 660/2125] avg loss 0.000949899, throughput 6.02122K wps
[Epoch 12 Batch 690/2125] avg loss 0.00100545, throughput 6.01783K wps
[Epoch 12 Batch 720/2125] avg loss 0.00113364, throughput 6.01722K wps
[Epoch 12 Batch 750/2125] avg loss 0.000905199, throughput 6.01623K wps
[Epoch 12 Batch 780/2125] avg loss 0.000921404, throughput 6.01311K wps
[Epoch 12 Batch 810/2125] avg loss 0.00099655, throughput 6.01014K wps
[Epoch 12 Batch 840/2125] avg loss 0.00102455, throughput 6.00781K wps
[Epoch 12 Batch 870/2125] avg loss 0.000898202, throughput 6.0017K wps
[Epoch 12 Batch 900/2125] avg loss 0.00108687, throughput 6.01031K wps
[Epoch 12 Batch 930/2125] avg loss 0.000882231, throughput 6.00863K wps
[Epoch 12 Batch 960/2125] avg loss 0.00104166, throughput 6.00511K wps
[Epoch 12 Batch 990/2125] avg loss 0.00100018, throughput 6.00866K wps
[Epoch 12 Batch 1020/2125] avg loss 0.000999968, throughput 6.01874K wps
[Epoch 12 Batch 1050/2125] avg loss 0.000888523, throughput 6.01518K wps
[Epoch 12 Batch 1080/2125] avg loss 0.00125621, throughput 6.00922K wps
[Epoch 12 Batch 1110/2125] avg loss 0.00134194, throughput 6.0091K wps
[Epoch 12 Batch 1140/2125] avg loss 0.00118755, throughput 6.00617K wps
[Epoch 12 Batch 1170/2125] avg loss 0.00104344, throughput 6.01272K wps
[Epoch 12 Batch 1200/2125] avg loss 0.00112199, throughput 6.02178K wps
[Epoch 12 Batch 1230/2125] avg loss 0.000855964, throughput 6.01754K wps
[Epoch 12 Batch 1260/2125] avg loss 0.00128413, throughput 6.01355K wps
[Epoch 12 Batch 1290/2125] avg loss 0.0012698, throughput 6.01435K wps
[Epoch 12 Batch 1320/2125] avg loss 0.000872924, throughput 6.01078K wps
[Epoch 12 Batch 1350/2125] avg loss 0.00108328, throughput 6.01304K wps
[Epoch 12 Batch 1380/2125] avg loss 0.00135713, throughput 6.01387K wps
[Epoch 12 Batch 1410/2125] avg loss 0.00100481, throughput 6.02246K wps
[Epoch 12 Batch 1440/2125] avg loss 0.00131962, throughput 6.02601K wps
[Epoch 12 Batch 1470/2125] avg loss 0.00121021, throughput 6.02346K wps
[Epoch 12 Batch 1500/2125] avg loss 0.00104488, throughput 6.01577K wps
[Epoch 12 Batch 1530/2125] avg loss 0.00113479, throughput 6.02202K wps
[Epoch 12 Batch 1560/2125] avg loss 0.00100491, throughput 6.01512K wps
[Epoch 12 Batch 1590/2125] avg loss 0.000907024, throughput 6.02263K wps
[Epoch 12 Batch 1620/2125] avg loss 0.00117419, throughput 6.01755K wps
[Epoch 12 Batch 1650/2125] avg loss 0.00135167, throughput 6.00877K wps
[Epoch 12 Batch 1680/2125] avg loss 0.00105655, throughput 6.01256K wps
[Epoch 12 Batch 1710/2125] avg loss 0.00120385, throughput 6.01598K wps
[Epoch 12 Batch 1740/2125] avg loss 0.00105561, throughput 6.01749K wps
[Epoch 12 Batch 1770/2125] avg loss 0.00143946, throughput 6.013K wps
[Epoch 12 Batch 1800/2125] avg loss 0.00126841, throughput 6.00747K wps
[Epoch 12 Batch 1830/2125] avg loss 0.000978508, throughput 6.02678K wps
[Epoch 12 Batch 1860/2125] avg loss 0.0010328, throughput 6.02295K wps
[Epoch 12 Batch 1890/2125] avg loss 0.000854251, throughput 6.02233K wps
[Epoch 12 Batch 1920/2125] avg loss 0.00104609, throughput 6.02348K wps
[Epoch 12 Batch 1950/2125] avg loss 0.00107899, throughput 6.02671K wps
[Epoch 12 Batch 1980/2125] avg loss 0.00126193, throughput 6.02171K wps
[Epoch 12 Batch 2010/2125] avg loss 0.00106937, throughput 6.01417K wps
[Epoch 12 Batch 2040/2125] avg loss 0.00136814, throughput 6.01866K wps
[Epoch 12 Batch 2070/2125] avg loss 0.00149425, throughput 6.01354K wps
[Epoch 12 Batch 2100/2125] avg loss 0.00150886, throughput 6.0081K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 12] train avg loss 0.00106802, test acc 0.9254, test avg loss 0.353518, throughput 6.01597K wps
[Epoch 13 Batch 30/2125] avg loss 0.00101892, throughput 6.14629K wps
[Epoch 13 Batch 60/2125] avg loss 0.000967282, throughput 6.02034K wps
[Epoch 13 Batch 90/2125] avg loss 0.000779802, throughput 6.02476K wps
[Epoch 13 Batch 120/2125] avg loss 0.000895655, throughput 6.01894K wps
[Epoch 13 Batch 150/2125] avg loss 0.000998215, throughput 6.00395K wps
[Epoch 13 Batch 180/2125] avg loss 0.000807327, throughput 6.01838K wps
[Epoch 13 Batch 210/2125] avg loss 0.000930469, throughput 6.01801K wps
[Epoch 13 Batch 240/2125] avg loss 0.000854809, throughput 6.01415K wps
[Epoch 13 Batch 270/2125] avg loss 0.000743591, throughput 6.0223K wps
[Epoch 13 Batch 300/2125] avg loss 0.00106553, throughput 6.01352K wps
[Epoch 13 Batch 330/2125] avg loss 0.000815823, throughput 6.01629K wps
[Epoch 13 Batch 360/2125] avg loss 0.000907055, throughput 6.01114K wps
[Epoch 13 Batch 390/2125] avg loss 0.0012251, throughput 6.02238K wps
[Epoch 13 Batch 420/2125] avg loss 0.00086331, throughput 6.01906K wps
[Epoch 13 Batch 450/2125] avg loss 0.000909162, throughput 6.01343K wps
[Epoch 13 Batch 480/2125] avg loss 0.000979442, throughput 6.01894K wps
[Epoch 13 Batch 510/2125] avg loss 0.00108855, throughput 6.02376K wps
[Epoch 13 Batch 540/2125] avg loss 0.00109598, throughput 6.02002K wps
[Epoch 13 Batch 570/2125] avg loss 0.000933722, throughput 6.02238K wps
[Epoch 13 Batch 600/2125] avg loss 0.000669939, throughput 6.01648K wps
[Epoch 13 Batch 630/2125] avg loss 0.000876336, throughput 6.02043K wps
[Epoch 13 Batch 660/2125] avg loss 0.000766274, throughput 6.01609K wps
[Epoch 13 Batch 690/2125] avg loss 0.000855382, throughput 5.93764K wps
[Epoch 13 Batch 720/2125] avg loss 0.00107411, throughput 5.9686K wps
[Epoch 13 Batch 750/2125] avg loss 0.00106039, throughput 6.00045K wps
[Epoch 13 Batch 780/2125] avg loss 0.000749605, throughput 6.01881K wps
[Epoch 13 Batch 810/2125] avg loss 0.00108954, throughput 6.01454K wps
[Epoch 13 Batch 840/2125] avg loss 0.00112498, throughput 6.01226K wps
[Epoch 13 Batch 870/2125] avg loss 0.00119517, throughput 6.0117K wps
[Epoch 13 Batch 900/2125] avg loss 0.00104399, throughput 6.0036K wps
[Epoch 13 Batch 930/2125] avg loss 0.00107735, throughput 6.00786K wps
[Epoch 13 Batch 960/2125] avg loss 0.00094741, throughput 6.01341K wps
[Epoch 13 Batch 990/2125] avg loss 0.000900494, throughput 6.01453K wps
[Epoch 13 Batch 1020/2125] avg loss 0.000921927, throughput 6.01232K wps
[Epoch 13 Batch 1050/2125] avg loss 0.00107684, throughput 6.01579K wps
[Epoch 13 Batch 1080/2125] avg loss 0.00123712, throughput 6.00172K wps
[Epoch 13 Batch 1110/2125] avg loss 0.00099662, throughput 6.00782K wps
[Epoch 13 Batch 1140/2125] avg loss 0.00100538, throughput 6.01296K wps
[Epoch 13 Batch 1170/2125] avg loss 0.000749233, throughput 6.01274K wps
[Epoch 13 Batch 1200/2125] avg loss 0.00079767, throughput 6.00491K wps
[Epoch 13 Batch 1230/2125] avg loss 0.000952016, throughput 6.0145K wps
[Epoch 13 Batch 1260/2125] avg loss 0.000866607, throughput 6.02461K wps
[Epoch 13 Batch 1290/2125] avg loss 0.000767177, throughput 6.00339K wps
[Epoch 13 Batch 1320/2125] avg loss 0.000894903, throughput 6.00807K wps
[Epoch 13 Batch 1350/2125] avg loss 0.000948048, throughput 5.99838K wps
[Epoch 13 Batch 1380/2125] avg loss 0.00140652, throughput 6.00402K wps
[Epoch 13 Batch 1410/2125] avg loss 0.000967229, throughput 6.00511K wps
[Epoch 13 Batch 1440/2125] avg loss 0.00129539, throughput 6.00673K wps
[Epoch 13 Batch 1470/2125] avg loss 0.0012502, throughput 6.00912K wps
[Epoch 13 Batch 1500/2125] avg loss 0.00111007, throughput 6.00721K wps
[Epoch 13 Batch 1530/2125] avg loss 0.00124584, throughput 6.00417K wps
[Epoch 13 Batch 1560/2125] avg loss 0.000991881, throughput 6.01537K wps
[Epoch 13 Batch 1590/2125] avg loss 0.000873129, throughput 6.00922K wps
[Epoch 13 Batch 1620/2125] avg loss 0.00106931, throughput 6.0116K wps
[Epoch 13 Batch 1650/2125] avg loss 0.00117206, throughput 6.01019K wps
[Epoch 13 Batch 1680/2125] avg loss 0.00105947, throughput 6.01143K wps
[Epoch 13 Batch 1710/2125] avg loss 0.00098258, throughput 6.00807K wps
[Epoch 13 Batch 1740/2125] avg loss 0.00123489, throughput 5.99525K wps
[Epoch 13 Batch 1770/2125] avg loss 0.00115244, throughput 6.00892K wps
[Epoch 13 Batch 1800/2125] avg loss 0.00110693, throughput 6.00711K wps
[Epoch 13 Batch 1830/2125] avg loss 0.000994606, throughput 6.01649K wps
[Epoch 13 Batch 1860/2125] avg loss 0.00124693, throughput 6.00857K wps
[Epoch 13 Batch 1890/2125] avg loss 0.0011056, throughput 6.01963K wps
[Epoch 13 Batch 1920/2125] avg loss 0.000877281, throughput 6.03002K wps
[Epoch 13 Batch 1950/2125] avg loss 0.00123282, throughput 6.01767K wps
[Epoch 13 Batch 1980/2125] avg loss 0.000896427, throughput 6.00998K wps
[Epoch 13 Batch 2010/2125] avg loss 0.0013074, throughput 6.01344K wps
[Epoch 13 Batch 2040/2125] avg loss 0.00101372, throughput 6.0004K wps
[Epoch 13 Batch 2070/2125] avg loss 0.000870462, throughput 5.99952K wps
[Epoch 13 Batch 2100/2125] avg loss 0.00109607, throughput 6.00974K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 13] train avg loss 0.00100241, test acc 0.9263, test avg loss 0.368209, throughput 6.0126K wps
[Epoch 14 Batch 30/2125] avg loss 0.00074417, throughput 6.14467K wps
[Epoch 14 Batch 60/2125] avg loss 0.000845636, throughput 6.00666K wps
[Epoch 14 Batch 90/2125] avg loss 0.000850947, throughput 6.00226K wps
[Epoch 14 Batch 120/2125] avg loss 0.00087104, throughput 6.00726K wps
[Epoch 14 Batch 150/2125] avg loss 0.000718609, throughput 6.01114K wps
[Epoch 14 Batch 180/2125] avg loss 0.000697669, throughput 6.0042K wps
[Epoch 14 Batch 210/2125] avg loss 0.000852047, throughput 6.01636K wps
[Epoch 14 Batch 240/2125] avg loss 0.000685869, throughput 6.00879K wps
[Epoch 14 Batch 270/2125] avg loss 0.00103066, throughput 6.02196K wps
[Epoch 14 Batch 300/2125] avg loss 0.00102494, throughput 6.00824K wps
[Epoch 14 Batch 330/2125] avg loss 0.000887455, throughput 6.01462K wps
[Epoch 14 Batch 360/2125] avg loss 0.000649308, throughput 6.00821K wps
[Epoch 14 Batch 390/2125] avg loss 0.000754829, throughput 6.01091K wps
[Epoch 14 Batch 420/2125] avg loss 0.000796099, throughput 6.00929K wps
[Epoch 14 Batch 450/2125] avg loss 0.000866744, throughput 6.00949K wps
[Epoch 14 Batch 480/2125] avg loss 0.000761737, throughput 6.01441K wps
[Epoch 14 Batch 510/2125] avg loss 0.000855048, throughput 6.00894K wps
[Epoch 14 Batch 540/2125] avg loss 0.000884833, throughput 6.00753K wps
[Epoch 14 Batch 570/2125] avg loss 0.000831817, throughput 6.01118K wps
[Epoch 14 Batch 600/2125] avg loss 0.000997149, throughput 6.00756K wps
[Epoch 14 Batch 630/2125] avg loss 0.000844022, throughput 6.01244K wps
[Epoch 14 Batch 660/2125] avg loss 0.00102909, throughput 6.01585K wps
[Epoch 14 Batch 690/2125] avg loss 0.00103769, throughput 6.00036K wps
[Epoch 14 Batch 720/2125] avg loss 0.000854342, throughput 6.01908K wps
[Epoch 14 Batch 750/2125] avg loss 0.000842531, throughput 6.01924K wps
[Epoch 14 Batch 780/2125] avg loss 0.00108849, throughput 6.00772K wps
[Epoch 14 Batch 810/2125] avg loss 0.00113799, throughput 6.00283K wps
[Epoch 14 Batch 840/2125] avg loss 0.000822118, throughput 6.00927K wps
[Epoch 14 Batch 870/2125] avg loss 0.000903704, throughput 6.01076K wps
[Epoch 14 Batch 900/2125] avg loss 0.000885693, throughput 6.01114K wps
[Epoch 14 Batch 930/2125] avg loss 0.000819587, throughput 6.00417K wps
[Epoch 14 Batch 960/2125] avg loss 0.000722968, throughput 6.01305K wps
[Epoch 14 Batch 990/2125] avg loss 0.000926265, throughput 6.00449K wps
[Epoch 14 Batch 1020/2125] avg loss 0.000976518, throughput 6.01587K wps
[Epoch 14 Batch 1050/2125] avg loss 0.000884543, throughput 6.00866K wps
[Epoch 14 Batch 1080/2125] avg loss 0.000904935, throughput 6.01262K wps
[Epoch 14 Batch 1110/2125] avg loss 0.000990453, throughput 6.00365K wps
[Epoch 14 Batch 1140/2125] avg loss 0.000946054, throughput 6.02183K wps
[Epoch 14 Batch 1170/2125] avg loss 0.00108744, throughput 6.00711K wps
[Epoch 14 Batch 1200/2125] avg loss 0.000699143, throughput 6.0145K wps
[Epoch 14 Batch 1230/2125] avg loss 0.00101952, throughput 6.01561K wps
[Epoch 14 Batch 1260/2125] avg loss 0.000832426, throughput 6.0099K wps
[Epoch 14 Batch 1290/2125] avg loss 0.000850756, throughput 6.00296K wps
[Epoch 14 Batch 1320/2125] avg loss 0.00118597, throughput 6.01112K wps
[Epoch 14 Batch 1350/2125] avg loss 0.00080498, throughput 6.01188K wps
[Epoch 14 Batch 1380/2125] avg loss 0.000772171, throughput 6.02084K wps
[Epoch 14 Batch 1410/2125] avg loss 0.000883625, throughput 6.00734K wps
[Epoch 14 Batch 1440/2125] avg loss 0.000854712, throughput 6.01156K wps
[Epoch 14 Batch 1470/2125] avg loss 0.000993395, throughput 6.01344K wps
[Epoch 14 Batch 1500/2125] avg loss 0.000936918, throughput 6.00962K wps
[Epoch 14 Batch 1530/2125] avg loss 0.00110218, throughput 6.00934K wps
[Epoch 14 Batch 1560/2125] avg loss 0.00117019, throughput 6.0172K wps
[Epoch 14 Batch 1590/2125] avg loss 0.00101987, throughput 6.01455K wps
[Epoch 14 Batch 1620/2125] avg loss 0.00137667, throughput 6.0096K wps
[Epoch 14 Batch 1650/2125] avg loss 0.00103904, throughput 6.01349K wps
[Epoch 14 Batch 1680/2125] avg loss 0.00108417, throughput 6.00741K wps
[Epoch 14 Batch 1710/2125] avg loss 0.00103327, throughput 6.01531K wps
[Epoch 14 Batch 1740/2125] avg loss 0.00115898, throughput 6.00869K wps
[Epoch 14 Batch 1770/2125] avg loss 0.00110905, throughput 6.00697K wps
[Epoch 14 Batch 1800/2125] avg loss 0.000948657, throughput 6.00057K wps
[Epoch 14 Batch 1830/2125] avg loss 0.00112475, throughput 6.0057K wps
[Epoch 14 Batch 1860/2125] avg loss 0.00101103, throughput 6.01498K wps
[Epoch 14 Batch 1890/2125] avg loss 0.00104754, throughput 6.01724K wps
[Epoch 14 Batch 1920/2125] avg loss 0.0012768, throughput 6.01751K wps
[Epoch 14 Batch 1950/2125] avg loss 0.0010171, throughput 6.0065K wps
[Epoch 14 Batch 1980/2125] avg loss 0.00100872, throughput 6.00326K wps
[Epoch 14 Batch 2010/2125] avg loss 0.000968363, throughput 6.00315K wps
[Epoch 14 Batch 2040/2125] avg loss 0.0011414, throughput 6.0157K wps
[Epoch 14 Batch 2070/2125] avg loss 0.00130195, throughput 6.01118K wps
[Epoch 14 Batch 2100/2125] avg loss 0.00142889, throughput 6.02069K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 14] train avg loss 0.000951867, test acc 0.9255, test avg loss 0.377104, throughput 6.0126K wps
[Epoch 15 Batch 30/2125] avg loss 0.000905626, throughput 6.1505K wps
[Epoch 15 Batch 60/2125] avg loss 0.000715662, throughput 6.00838K wps
[Epoch 15 Batch 90/2125] avg loss 0.000662512, throughput 6.00789K wps
[Epoch 15 Batch 120/2125] avg loss 0.000853108, throughput 6.00784K wps
[Epoch 15 Batch 150/2125] avg loss 0.000773665, throughput 6.00796K wps
[Epoch 15 Batch 180/2125] avg loss 0.000779711, throughput 6.01364K wps
[Epoch 15 Batch 210/2125] avg loss 0.000746945, throughput 6.00711K wps
[Epoch 15 Batch 240/2125] avg loss 0.00073244, throughput 6.00511K wps
[Epoch 15 Batch 270/2125] avg loss 0.000735796, throughput 6.01381K wps
[Epoch 15 Batch 300/2125] avg loss 0.000968134, throughput 6.00796K wps
[Epoch 15 Batch 330/2125] avg loss 0.000728643, throughput 6.01805K wps
[Epoch 15 Batch 360/2125] avg loss 0.00078673, throughput 6.01615K wps
[Epoch 15 Batch 390/2125] avg loss 0.000782574, throughput 6.0104K wps
[Epoch 15 Batch 420/2125] avg loss 0.000766435, throughput 6.00408K wps
[Epoch 15 Batch 450/2125] avg loss 0.000686568, throughput 6.00607K wps
[Epoch 15 Batch 480/2125] avg loss 0.000948556, throughput 6.00697K wps
[Epoch 15 Batch 510/2125] avg loss 0.000825495, throughput 6.01544K wps
[Epoch 15 Batch 540/2125] avg loss 0.000991408, throughput 6.00805K wps
[Epoch 15 Batch 570/2125] avg loss 0.000745565, throughput 6.00513K wps
[Epoch 15 Batch 600/2125] avg loss 0.000948933, throughput 6.00325K wps
[Epoch 15 Batch 630/2125] avg loss 0.000811216, throughput 6.00514K wps
[Epoch 15 Batch 660/2125] avg loss 0.000917463, throughput 6.01605K wps
[Epoch 15 Batch 690/2125] avg loss 0.000757205, throughput 6.01779K wps
[Epoch 15 Batch 720/2125] avg loss 0.000899646, throughput 6.01467K wps
[Epoch 15 Batch 750/2125] avg loss 0.000676076, throughput 6.02406K wps
[Epoch 15 Batch 780/2125] avg loss 0.000887567, throughput 6.02245K wps
[Epoch 15 Batch 810/2125] avg loss 0.000711167, throughput 6.0201K wps
[Epoch 15 Batch 840/2125] avg loss 0.000801454, throughput 6.01267K wps
[Epoch 15 Batch 870/2125] avg loss 0.001151, throughput 6.01172K wps
[Epoch 15 Batch 900/2125] avg loss 0.00081163, throughput 6.01199K wps
[Epoch 15 Batch 930/2125] avg loss 0.000720714, throughput 6.00944K wps
[Epoch 15 Batch 960/2125] avg loss 0.000927055, throughput 6.00904K wps
[Epoch 15 Batch 990/2125] avg loss 0.00118453, throughput 6.01488K wps
[Epoch 15 Batch 1020/2125] avg loss 0.000664371, throughput 6.02225K wps
[Epoch 15 Batch 1050/2125] avg loss 0.00105412, throughput 6.01292K wps
[Epoch 15 Batch 1080/2125] avg loss 0.000862353, throughput 6.01039K wps
[Epoch 15 Batch 1110/2125] avg loss 0.000811533, throughput 6.00748K wps
[Epoch 15 Batch 1140/2125] avg loss 0.00096826, throughput 6.0193K wps
[Epoch 15 Batch 1170/2125] avg loss 0.000800381, throughput 6.0212K wps
[Epoch 15 Batch 1200/2125] avg loss 0.00105327, throughput 6.01544K wps
[Epoch 15 Batch 1230/2125] avg loss 0.000912642, throughput 6.00275K wps
[Epoch 15 Batch 1260/2125] avg loss 0.0011147, throughput 6.00343K wps
[Epoch 15 Batch 1290/2125] avg loss 0.000937972, throughput 6.01045K wps
[Epoch 15 Batch 1320/2125] avg loss 0.000921672, throughput 6.00546K wps
[Epoch 15 Batch 1350/2125] avg loss 0.000863945, throughput 6.01536K wps
[Epoch 15 Batch 1380/2125] avg loss 0.00103239, throughput 6.01673K wps
[Epoch 15 Batch 1410/2125] avg loss 0.00081878, throughput 6.01032K wps
[Epoch 15 Batch 1440/2125] avg loss 0.000868317, throughput 6.01128K wps
[Epoch 15 Batch 1470/2125] avg loss 0.0008904, throughput 6.00563K wps
[Epoch 15 Batch 1500/2125] avg loss 0.00110283, throughput 6.01708K wps
[Epoch 15 Batch 1530/2125] avg loss 0.000885455, throughput 6.01626K wps
[Epoch 15 Batch 1560/2125] avg loss 0.000960313, throughput 6.00723K wps
[Epoch 15 Batch 1590/2125] avg loss 0.000912937, throughput 6.01221K wps
[Epoch 15 Batch 1620/2125] avg loss 0.000995622, throughput 6.01777K wps
[Epoch 15 Batch 1650/2125] avg loss 0.00101872, throughput 6.0218K wps
[Epoch 15 Batch 1680/2125] avg loss 0.00100788, throughput 6.01854K wps
[Epoch 15 Batch 1710/2125] avg loss 0.0011567, throughput 6.00451K wps
[Epoch 15 Batch 1740/2125] avg loss 0.000748502, throughput 6.00192K wps
[Epoch 15 Batch 1770/2125] avg loss 0.000687576, throughput 5.99583K wps
[Epoch 15 Batch 1800/2125] avg loss 0.000721836, throughput 6.01283K wps
[Epoch 15 Batch 1830/2125] avg loss 0.000907815, throughput 6.0128K wps
[Epoch 15 Batch 1860/2125] avg loss 0.00132678, throughput 6.01558K wps
[Epoch 15 Batch 1890/2125] avg loss 0.000914684, throughput 6.03105K wps
[Epoch 15 Batch 1920/2125] avg loss 0.00126829, throughput 6.02222K wps
[Epoch 15 Batch 1950/2125] avg loss 0.00118509, throughput 6.01763K wps
[Epoch 15 Batch 1980/2125] avg loss 0.000830785, throughput 6.01843K wps
[Epoch 15 Batch 2010/2125] avg loss 0.00117355, throughput 6.01535K wps
[Epoch 15 Batch 2040/2125] avg loss 0.000722919, throughput 6.00968K wps
[Epoch 15 Batch 2070/2125] avg loss 0.00108623, throughput 6.02637K wps
[Epoch 15 Batch 2100/2125] avg loss 0.00111851, throughput 6.01485K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 15] train avg loss 0.000895013, test acc 0.9264, test avg loss 0.393888, throughput 6.01444K wps
[Epoch 16 Batch 30/2125] avg loss 0.000700255, throughput 6.15397K wps
[Epoch 16 Batch 60/2125] avg loss 0.000672233, throughput 6.02799K wps
[Epoch 16 Batch 90/2125] avg loss 0.000680336, throughput 6.0094K wps
[Epoch 16 Batch 120/2125] avg loss 0.000632582, throughput 6.01402K wps
[Epoch 16 Batch 150/2125] avg loss 0.000789214, throughput 6.00674K wps
[Epoch 16 Batch 180/2125] avg loss 0.000743008, throughput 6.01359K wps
[Epoch 16 Batch 210/2125] avg loss 0.000728884, throughput 6.01535K wps
[Epoch 16 Batch 240/2125] avg loss 0.000747101, throughput 6.0237K wps
[Epoch 16 Batch 270/2125] avg loss 0.00105457, throughput 6.01778K wps
[Epoch 16 Batch 300/2125] avg loss 0.000580589, throughput 6.01212K wps
[Epoch 16 Batch 330/2125] avg loss 0.000874428, throughput 6.01911K wps
[Epoch 16 Batch 360/2125] avg loss 0.000729789, throughput 6.01066K wps
[Epoch 16 Batch 390/2125] avg loss 0.000815021, throughput 6.01345K wps
[Epoch 16 Batch 420/2125] avg loss 0.000483204, throughput 6.02216K wps
[Epoch 16 Batch 450/2125] avg loss 0.000723333, throughput 6.0155K wps
[Epoch 16 Batch 480/2125] avg loss 0.00066828, throughput 6.01085K wps
[Epoch 16 Batch 510/2125] avg loss 0.000622571, throughput 6.01187K wps
[Epoch 16 Batch 540/2125] avg loss 0.000504421, throughput 6.01158K wps
[Epoch 16 Batch 570/2125] avg loss 0.000644185, throughput 6.01506K wps
[Epoch 16 Batch 600/2125] avg loss 0.000833485, throughput 6.01912K wps
[Epoch 16 Batch 630/2125] avg loss 0.000545795, throughput 6.01979K wps
[Epoch 16 Batch 660/2125] avg loss 0.000581739, throughput 6.02424K wps
[Epoch 16 Batch 690/2125] avg loss 0.000767658, throughput 6.01488K wps
[Epoch 16 Batch 720/2125] avg loss 0.000680114, throughput 6.01152K wps
[Epoch 16 Batch 750/2125] avg loss 0.000601711, throughput 6.01716K wps
[Epoch 16 Batch 780/2125] avg loss 0.000852451, throughput 6.01768K wps
[Epoch 16 Batch 810/2125] avg loss 0.001029, throughput 6.01853K wps
[Epoch 16 Batch 840/2125] avg loss 0.000986434, throughput 6.01852K wps
[Epoch 16 Batch 870/2125] avg loss 0.000742192, throughput 6.02273K wps
[Epoch 16 Batch 900/2125] avg loss 0.00062693, throughput 6.02528K wps
[Epoch 16 Batch 930/2125] avg loss 0.000870318, throughput 6.02061K wps
[Epoch 16 Batch 960/2125] avg loss 0.000657596, throughput 6.0166K wps
[Epoch 16 Batch 990/2125] avg loss 0.0011207, throughput 6.01419K wps
[Epoch 16 Batch 1020/2125] avg loss 0.000825926, throughput 6.0216K wps
[Epoch 16 Batch 1050/2125] avg loss 0.00123872, throughput 6.02049K wps
[Epoch 16 Batch 1080/2125] avg loss 0.000914023, throughput 6.02283K wps
[Epoch 16 Batch 1110/2125] avg loss 0.000897164, throughput 6.01838K wps
[Epoch 16 Batch 1140/2125] avg loss 0.000921938, throughput 6.02153K wps
[Epoch 16 Batch 1170/2125] avg loss 0.000749688, throughput 6.00959K wps
[Epoch 16 Batch 1200/2125] avg loss 0.000724331, throughput 5.99917K wps
[Epoch 16 Batch 1230/2125] avg loss 0.000920361, throughput 6.0166K wps
[Epoch 16 Batch 1260/2125] avg loss 0.00063115, throughput 6.02252K wps
[Epoch 16 Batch 1290/2125] avg loss 0.000692472, throughput 6.02165K wps
[Epoch 16 Batch 1320/2125] avg loss 0.000669155, throughput 6.02406K wps
[Epoch 16 Batch 1350/2125] avg loss 0.00105881, throughput 6.02328K wps
[Epoch 16 Batch 1380/2125] avg loss 0.000962101, throughput 6.00837K wps
[Epoch 16 Batch 1410/2125] avg loss 0.00100853, throughput 6.02056K wps
[Epoch 16 Batch 1440/2125] avg loss 0.000862195, throughput 6.02384K wps
[Epoch 16 Batch 1470/2125] avg loss 0.000809816, throughput 6.0143K wps
[Epoch 16 Batch 1500/2125] avg loss 0.000939134, throughput 6.01596K wps
[Epoch 16 Batch 1530/2125] avg loss 0.000751404, throughput 6.01773K wps
[Epoch 16 Batch 1560/2125] avg loss 0.000932248, throughput 6.02147K wps
[Epoch 16 Batch 1590/2125] avg loss 0.000848558, throughput 6.02491K wps
[Epoch 16 Batch 1620/2125] avg loss 0.00102344, throughput 6.00831K wps
[Epoch 16 Batch 1650/2125] avg loss 0.00100238, throughput 6.00883K wps
[Epoch 16 Batch 1680/2125] avg loss 0.00111599, throughput 6.00308K wps
[Epoch 16 Batch 1710/2125] avg loss 0.000782975, throughput 6.00867K wps
[Epoch 16 Batch 1740/2125] avg loss 0.000841482, throughput 6.0108K wps
[Epoch 16 Batch 1770/2125] avg loss 0.00112291, throughput 6.00629K wps
[Epoch 16 Batch 1800/2125] avg loss 0.000762345, throughput 6.00753K wps
[Epoch 16 Batch 1830/2125] avg loss 0.000717917, throughput 6.00958K wps
[Epoch 16 Batch 1860/2125] avg loss 0.0012141, throughput 6.00274K wps
[Epoch 16 Batch 1890/2125] avg loss 0.00111475, throughput 6.00688K wps
[Epoch 16 Batch 1920/2125] avg loss 0.00101266, throughput 6.01046K wps
[Epoch 16 Batch 1950/2125] avg loss 0.00123361, throughput 6.00462K wps
[Epoch 16 Batch 1980/2125] avg loss 0.00132155, throughput 6.00685K wps
[Epoch 16 Batch 2010/2125] avg loss 0.000856864, throughput 6.00917K wps
[Epoch 16 Batch 2040/2125] avg loss 0.000840745, throughput 6.01409K wps
[Epoch 16 Batch 2070/2125] avg loss 0.000905719, throughput 6.01139K wps
[Epoch 16 Batch 2100/2125] avg loss 0.00114506, throughput 6.02391K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 16] train avg loss 0.000837895, test acc 0.9250, test avg loss 0.406746, throughput 6.01718K wps
[Epoch 17 Batch 30/2125] avg loss 0.000599889, throughput 6.15582K wps
[Epoch 17 Batch 60/2125] avg loss 0.00066103, throughput 6.01735K wps
[Epoch 17 Batch 90/2125] avg loss 0.000677178, throughput 6.01189K wps
[Epoch 17 Batch 120/2125] avg loss 0.000661652, throughput 6.0114K wps
[Epoch 17 Batch 150/2125] avg loss 0.000590205, throughput 5.99599K wps
[Epoch 17 Batch 180/2125] avg loss 0.000819459, throughput 6.0178K wps
[Epoch 17 Batch 210/2125] avg loss 0.00059392, throughput 6.02073K wps
[Epoch 17 Batch 240/2125] avg loss 0.000689386, throughput 6.00765K wps
[Epoch 17 Batch 270/2125] avg loss 0.000843374, throughput 6.0135K wps
[Epoch 17 Batch 300/2125] avg loss 0.000791896, throughput 6.00888K wps
[Epoch 17 Batch 330/2125] avg loss 0.000718978, throughput 6.02029K wps
[Epoch 17 Batch 360/2125] avg loss 0.000656822, throughput 6.0037K wps
[Epoch 17 Batch 390/2125] avg loss 0.000545736, throughput 6.00961K wps
[Epoch 17 Batch 420/2125] avg loss 0.0006776, throughput 6.02178K wps
[Epoch 17 Batch 450/2125] avg loss 0.000540761, throughput 6.02535K wps
[Epoch 17 Batch 480/2125] avg loss 0.000861532, throughput 6.02923K wps
[Epoch 17 Batch 510/2125] avg loss 0.000661893, throughput 6.02548K wps
[Epoch 17 Batch 540/2125] avg loss 0.000698913, throughput 6.02562K wps
[Epoch 17 Batch 570/2125] avg loss 0.000623042, throughput 6.01652K wps
[Epoch 17 Batch 600/2125] avg loss 0.000695064, throughput 6.02272K wps
[Epoch 17 Batch 630/2125] avg loss 0.000715071, throughput 6.02393K wps
[Epoch 17 Batch 660/2125] avg loss 0.000676741, throughput 6.02173K wps
[Epoch 17 Batch 690/2125] avg loss 0.000670723, throughput 6.01435K wps
[Epoch 17 Batch 720/2125] avg loss 0.000751837, throughput 6.01127K wps
[Epoch 17 Batch 750/2125] avg loss 0.000808519, throughput 6.00993K wps
[Epoch 17 Batch 780/2125] avg loss 0.000486168, throughput 6.02306K wps
[Epoch 17 Batch 810/2125] avg loss 0.000673599, throughput 6.01232K wps
[Epoch 17 Batch 840/2125] avg loss 0.000771459, throughput 6.01617K wps
[Epoch 17 Batch 870/2125] avg loss 0.000835609, throughput 6.02484K wps
[Epoch 17 Batch 900/2125] avg loss 0.000810488, throughput 6.02551K wps
[Epoch 17 Batch 930/2125] avg loss 0.000863506, throughput 6.03122K wps
[Epoch 17 Batch 960/2125] avg loss 0.000710065, throughput 6.02089K wps
[Epoch 17 Batch 990/2125] avg loss 0.000953335, throughput 6.01679K wps
[Epoch 17 Batch 1020/2125] avg loss 0.000806199, throughput 6.02037K wps
[Epoch 17 Batch 1050/2125] avg loss 0.00108213, throughput 6.0145K wps
[Epoch 17 Batch 1080/2125] avg loss 0.000871811, throughput 6.02432K wps
[Epoch 17 Batch 1110/2125] avg loss 0.000873889, throughput 6.01705K wps
[Epoch 17 Batch 1140/2125] avg loss 0.000768565, throughput 6.01609K wps
[Epoch 17 Batch 1170/2125] avg loss 0.000885983, throughput 6.00557K wps
[Epoch 17 Batch 1200/2125] avg loss 0.000807054, throughput 6.01822K wps
[Epoch 17 Batch 1230/2125] avg loss 0.000898227, throughput 6.01414K wps
[Epoch 17 Batch 1260/2125] avg loss 0.00088569, throughput 6.01532K wps
[Epoch 17 Batch 1290/2125] avg loss 0.000704245, throughput 6.02655K wps
[Epoch 17 Batch 1320/2125] avg loss 0.000771858, throughput 6.00473K wps
[Epoch 17 Batch 1350/2125] avg loss 0.000681807, throughput 6.01941K wps
[Epoch 17 Batch 1380/2125] avg loss 0.000723327, throughput 6.01846K wps
[Epoch 17 Batch 1410/2125] avg loss 0.000716552, throughput 6.01447K wps
[Epoch 17 Batch 1440/2125] avg loss 0.000762629, throughput 6.01509K wps
[Epoch 17 Batch 1470/2125] avg loss 0.000801244, throughput 6.00774K wps
[Epoch 17 Batch 1500/2125] avg loss 0.00113637, throughput 6.02062K wps
[Epoch 17 Batch 1530/2125] avg loss 0.00108486, throughput 6.01832K wps
[Epoch 17 Batch 1560/2125] avg loss 0.000910646, throughput 6.02458K wps
[Epoch 17 Batch 1590/2125] avg loss 0.000897684, throughput 6.01883K wps
[Epoch 17 Batch 1620/2125] avg loss 0.00107565, throughput 6.02328K wps
[Epoch 17 Batch 1650/2125] avg loss 0.00107981, throughput 6.02655K wps
[Epoch 17 Batch 1680/2125] avg loss 0.00103463, throughput 6.02595K wps
[Epoch 17 Batch 1710/2125] avg loss 0.000919373, throughput 6.01432K wps
[Epoch 17 Batch 1740/2125] avg loss 0.00105089, throughput 6.01287K wps
[Epoch 17 Batch 1770/2125] avg loss 0.000642832, throughput 6.01156K wps
[Epoch 17 Batch 1800/2125] avg loss 0.0011586, throughput 6.00842K wps
[Epoch 17 Batch 1830/2125] avg loss 0.000863036, throughput 6.01587K wps
[Epoch 17 Batch 1860/2125] avg loss 0.0010085, throughput 6.00903K wps
[Epoch 17 Batch 1890/2125] avg loss 0.00112593, throughput 6.01638K wps
[Epoch 17 Batch 1920/2125] avg loss 0.000829175, throughput 6.01706K wps
[Epoch 17 Batch 1950/2125] avg loss 0.00111297, throughput 6.01999K wps
[Epoch 17 Batch 1980/2125] avg loss 0.000909638, throughput 6.01022K wps
[Epoch 17 Batch 2010/2125] avg loss 0.000737868, throughput 6.01643K wps
[Epoch 17 Batch 2040/2125] avg loss 0.000948975, throughput 6.00767K wps
[Epoch 17 Batch 2070/2125] avg loss 0.00093076, throughput 6.02259K wps
[Epoch 17 Batch 2100/2125] avg loss 0.000851536, throughput 6.00522K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 17] train avg loss 0.000812684, test acc 0.9237, test avg loss 0.415757, throughput 6.01873K wps
[Epoch 18 Batch 30/2125] avg loss 0.000707391, throughput 6.15916K wps
[Epoch 18 Batch 60/2125] avg loss 0.000706771, throughput 6.03008K wps
[Epoch 18 Batch 90/2125] avg loss 0.0005944, throughput 6.01371K wps
[Epoch 18 Batch 120/2125] avg loss 0.000720784, throughput 6.0142K wps
[Epoch 18 Batch 150/2125] avg loss 0.000533055, throughput 6.01695K wps
[Epoch 18 Batch 180/2125] avg loss 0.000442523, throughput 6.01173K wps
[Epoch 18 Batch 210/2125] avg loss 0.000570719, throughput 6.01993K wps
[Epoch 18 Batch 240/2125] avg loss 0.000612429, throughput 6.0218K wps
[Epoch 18 Batch 270/2125] avg loss 0.000773289, throughput 6.01589K wps
[Epoch 18 Batch 300/2125] avg loss 0.000754451, throughput 6.01799K wps
[Epoch 18 Batch 330/2125] avg loss 0.000698454, throughput 6.01952K wps
[Epoch 18 Batch 360/2125] avg loss 0.000734355, throughput 6.01157K wps
[Epoch 18 Batch 390/2125] avg loss 0.000749488, throughput 6.01074K wps
[Epoch 18 Batch 420/2125] avg loss 0.000575669, throughput 6.01598K wps
[Epoch 18 Batch 450/2125] avg loss 0.000817338, throughput 6.01495K wps
[Epoch 18 Batch 480/2125] avg loss 0.000635876, throughput 6.02218K wps
[Epoch 18 Batch 510/2125] avg loss 0.000931148, throughput 6.02134K wps
[Epoch 18 Batch 540/2125] avg loss 0.000678365, throughput 6.0072K wps
[Epoch 18 Batch 570/2125] avg loss 0.000662703, throughput 6.01786K wps
[Epoch 18 Batch 600/2125] avg loss 0.000539033, throughput 5.99132K wps
[Epoch 18 Batch 630/2125] avg loss 0.00065606, throughput 6.00704K wps
[Epoch 18 Batch 660/2125] avg loss 0.000587945, throughput 6.0255K wps
[Epoch 18 Batch 690/2125] avg loss 0.000871621, throughput 6.01338K wps
[Epoch 18 Batch 720/2125] avg loss 0.000781989, throughput 6.00563K wps
[Epoch 18 Batch 750/2125] avg loss 0.000714171, throughput 6.00734K wps
[Epoch 18 Batch 780/2125] avg loss 0.000782531, throughput 6.0007K wps
[Epoch 18 Batch 810/2125] avg loss 0.000842307, throughput 6.00755K wps
[Epoch 18 Batch 840/2125] avg loss 0.000733922, throughput 6.0061K wps
[Epoch 18 Batch 870/2125] avg loss 0.000789025, throughput 6.0113K wps
[Epoch 18 Batch 900/2125] avg loss 0.000925252, throughput 6.00777K wps
[Epoch 18 Batch 930/2125] avg loss 0.000689079, throughput 6.01723K wps
[Epoch 18 Batch 960/2125] avg loss 0.000761132, throughput 6.01908K wps
[Epoch 18 Batch 990/2125] avg loss 0.000846551, throughput 6.00863K wps
[Epoch 18 Batch 1020/2125] avg loss 0.000654831, throughput 6.00737K wps
[Epoch 18 Batch 1050/2125] avg loss 0.000722014, throughput 6.00542K wps
[Epoch 18 Batch 1080/2125] avg loss 0.000734986, throughput 6.01244K wps
[Epoch 18 Batch 1110/2125] avg loss 0.000854089, throughput 6.01811K wps
[Epoch 18 Batch 1140/2125] avg loss 0.000787235, throughput 6.0096K wps
[Epoch 18 Batch 1170/2125] avg loss 0.000860193, throughput 6.01254K wps
[Epoch 18 Batch 1200/2125] avg loss 0.000804146, throughput 6.01343K wps
[Epoch 18 Batch 1230/2125] avg loss 0.000852169, throughput 6.02125K wps
[Epoch 18 Batch 1260/2125] avg loss 0.000768605, throughput 6.01401K wps
[Epoch 18 Batch 1290/2125] avg loss 0.000789258, throughput 6.00669K wps
[Epoch 18 Batch 1320/2125] avg loss 0.000742646, throughput 6.00837K wps
[Epoch 18 Batch 1350/2125] avg loss 0.000952375, throughput 6.00751K wps
[Epoch 18 Batch 1380/2125] avg loss 0.00101653, throughput 6.01594K wps
[Epoch 18 Batch 1410/2125] avg loss 0.000742798, throughput 6.01971K wps
[Epoch 18 Batch 1440/2125] avg loss 0.000915005, throughput 6.01803K wps
[Epoch 18 Batch 1470/2125] avg loss 0.000846176, throughput 6.01084K wps
[Epoch 18 Batch 1500/2125] avg loss 0.000893334, throughput 6.0133K wps
[Epoch 18 Batch 1530/2125] avg loss 0.000596528, throughput 6.01351K wps
[Epoch 18 Batch 1560/2125] avg loss 0.000783842, throughput 6.00564K wps
[Epoch 18 Batch 1590/2125] avg loss 0.00103868, throughput 6.00694K wps
[Epoch 18 Batch 1620/2125] avg loss 0.000765857, throughput 6.00625K wps
[Epoch 18 Batch 1650/2125] avg loss 0.00108066, throughput 6.01947K wps
[Epoch 18 Batch 1680/2125] avg loss 0.000610276, throughput 6.01457K wps
[Epoch 18 Batch 1710/2125] avg loss 0.000897808, throughput 6.01234K wps
[Epoch 18 Batch 1740/2125] avg loss 0.00104791, throughput 6.01668K wps
[Epoch 18 Batch 1770/2125] avg loss 0.0010708, throughput 6.02384K wps
[Epoch 18 Batch 1800/2125] avg loss 0.000880172, throughput 6.01869K wps
[Epoch 18 Batch 1830/2125] avg loss 0.000834756, throughput 6.0264K wps
[Epoch 18 Batch 1860/2125] avg loss 0.000888168, throughput 6.01927K wps
[Epoch 18 Batch 1890/2125] avg loss 0.000754886, throughput 6.01023K wps
[Epoch 18 Batch 1920/2125] avg loss 0.00118919, throughput 6.01299K wps
[Epoch 18 Batch 1950/2125] avg loss 0.000737579, throughput 6.00699K wps
[Epoch 18 Batch 1980/2125] avg loss 0.000841736, throughput 6.00817K wps
[Epoch 18 Batch 2010/2125] avg loss 0.00090521, throughput 6.00362K wps
[Epoch 18 Batch 2040/2125] avg loss 0.00106706, throughput 6.01836K wps
[Epoch 18 Batch 2070/2125] avg loss 0.000768554, throughput 6.02136K wps
[Epoch 18 Batch 2100/2125] avg loss 0.000899584, throughput 6.00915K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 18] train avg loss 0.000786137, test acc 0.9249, test avg loss 0.421145, throughput 6.01541K wps
[Epoch 19 Batch 30/2125] avg loss 0.000439053, throughput 6.14901K wps
[Epoch 19 Batch 60/2125] avg loss 0.000656189, throughput 5.99926K wps
[Epoch 19 Batch 90/2125] avg loss 0.000570596, throughput 6.01807K wps
[Epoch 19 Batch 120/2125] avg loss 0.000401377, throughput 6.01257K wps
[Epoch 19 Batch 150/2125] avg loss 0.000601915, throughput 6.00283K wps
[Epoch 19 Batch 180/2125] avg loss 0.000502293, throughput 6.01107K wps
[Epoch 19 Batch 210/2125] avg loss 0.000802866, throughput 6.00659K wps
[Epoch 19 Batch 240/2125] avg loss 0.00078467, throughput 6.01079K wps
[Epoch 19 Batch 270/2125] avg loss 0.000543126, throughput 6.01631K wps
[Epoch 19 Batch 300/2125] avg loss 0.000618054, throughput 6.01487K wps
[Epoch 19 Batch 330/2125] avg loss 0.000568777, throughput 6.0143K wps
[Epoch 19 Batch 360/2125] avg loss 0.000415589, throughput 6.01977K wps
[Epoch 19 Batch 390/2125] avg loss 0.000698215, throughput 6.00698K wps
[Epoch 19 Batch 420/2125] avg loss 0.000602706, throughput 6.00802K wps
[Epoch 19 Batch 450/2125] avg loss 0.00051907, throughput 6.01179K wps
[Epoch 19 Batch 480/2125] avg loss 0.000801319, throughput 6.01188K wps
[Epoch 19 Batch 510/2125] avg loss 0.000652575, throughput 6.01918K wps
[Epoch 19 Batch 540/2125] avg loss 0.000610735, throughput 6.01363K wps
[Epoch 19 Batch 570/2125] avg loss 0.000831128, throughput 6.022K wps
[Epoch 19 Batch 600/2125] avg loss 0.000704277, throughput 6.01813K wps
[Epoch 19 Batch 630/2125] avg loss 0.000638594, throughput 6.01636K wps
[Epoch 19 Batch 660/2125] avg loss 0.000834625, throughput 6.0078K wps
[Epoch 19 Batch 690/2125] avg loss 0.000952288, throughput 6.02567K wps
[Epoch 19 Batch 720/2125] avg loss 0.000668521, throughput 6.02011K wps
[Epoch 19 Batch 750/2125] avg loss 0.000712958, throughput 6.02089K wps
[Epoch 19 Batch 780/2125] avg loss 0.000537426, throughput 6.01525K wps
[Epoch 19 Batch 810/2125] avg loss 0.000832736, throughput 6.01899K wps
[Epoch 19 Batch 840/2125] avg loss 0.000615841, throughput 6.01804K wps
[Epoch 19 Batch 870/2125] avg loss 0.000685709, throughput 6.01315K wps
[Epoch 19 Batch 900/2125] avg loss 0.000927133, throughput 6.01885K wps
[Epoch 19 Batch 930/2125] avg loss 0.000581872, throughput 6.01836K wps
[Epoch 19 Batch 960/2125] avg loss 0.000751955, throughput 6.01648K wps
[Epoch 19 Batch 990/2125] avg loss 0.000644666, throughput 6.01903K wps
[Epoch 19 Batch 1020/2125] avg loss 0.000811615, throughput 6.01461K wps
[Epoch 19 Batch 1050/2125] avg loss 0.000540838, throughput 6.0083K wps
[Epoch 19 Batch 1080/2125] avg loss 0.000845235, throughput 6.01042K wps
[Epoch 19 Batch 1110/2125] avg loss 0.00060721, throughput 6.01613K wps
[Epoch 19 Batch 1140/2125] avg loss 0.000851434, throughput 6.00761K wps
[Epoch 19 Batch 1170/2125] avg loss 0.000727925, throughput 6.00417K wps
[Epoch 19 Batch 1200/2125] avg loss 0.000682514, throughput 5.99449K wps
[Epoch 19 Batch 1230/2125] avg loss 0.000863878, throughput 6.01775K wps
[Epoch 19 Batch 1260/2125] avg loss 0.000610734, throughput 6.01701K wps
[Epoch 19 Batch 1290/2125] avg loss 0.000835102, throughput 6.00802K wps
[Epoch 19 Batch 1320/2125] avg loss 0.00062132, throughput 6.00247K wps
[Epoch 19 Batch 1350/2125] avg loss 0.000612866, throughput 6.01159K wps
[Epoch 19 Batch 1380/2125] avg loss 0.000735835, throughput 6.01647K wps
[Epoch 19 Batch 1410/2125] avg loss 0.000650769, throughput 6.01466K wps
[Epoch 19 Batch 1440/2125] avg loss 0.000712138, throughput 6.01636K wps
[Epoch 19 Batch 1470/2125] avg loss 0.000965356, throughput 6.01377K wps
[Epoch 19 Batch 1500/2125] avg loss 0.000763949, throughput 6.022K wps
[Epoch 19 Batch 1530/2125] avg loss 0.000892489, throughput 6.01478K wps
[Epoch 19 Batch 1560/2125] avg loss 0.000857499, throughput 6.02171K wps
[Epoch 19 Batch 1590/2125] avg loss 0.000782075, throughput 6.03051K wps
[Epoch 19 Batch 1620/2125] avg loss 0.00081909, throughput 6.01991K wps
[Epoch 19 Batch 1650/2125] avg loss 0.000720364, throughput 6.01178K wps
[Epoch 19 Batch 1680/2125] avg loss 0.000984517, throughput 6.00366K wps
[Epoch 19 Batch 1710/2125] avg loss 0.000737386, throughput 6.01372K wps
[Epoch 19 Batch 1740/2125] avg loss 0.000721006, throughput 6.01693K wps
[Epoch 19 Batch 1770/2125] avg loss 0.000748822, throughput 6.00991K wps
[Epoch 19 Batch 1800/2125] avg loss 0.000785671, throughput 6.01781K wps
[Epoch 19 Batch 1830/2125] avg loss 0.000791185, throughput 5.99999K wps
[Epoch 19 Batch 1860/2125] avg loss 0.00099042, throughput 6.01181K wps
[Epoch 19 Batch 1890/2125] avg loss 0.00114218, throughput 6.01412K wps
[Epoch 19 Batch 1920/2125] avg loss 0.000750673, throughput 6.02058K wps
[Epoch 19 Batch 1950/2125] avg loss 0.000862198, throughput 6.02022K wps
[Epoch 19 Batch 1980/2125] avg loss 0.000804845, throughput 6.03001K wps
[Epoch 19 Batch 2010/2125] avg loss 0.000882999, throughput 6.01403K wps
[Epoch 19 Batch 2040/2125] avg loss 0.000821997, throughput 6.00787K wps
[Epoch 19 Batch 2070/2125] avg loss 0.000874674, throughput 6.01158K wps
[Epoch 19 Batch 2100/2125] avg loss 0.00105837, throughput 6.01524K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 19] train avg loss 0.000732303, test acc 0.9249, test avg loss 0.425258, throughput 6.01599K wps
[Epoch 20 Batch 30/2125] avg loss 0.000560184, throughput 6.14813K wps
[Epoch 20 Batch 60/2125] avg loss 0.000668134, throughput 6.01882K wps
[Epoch 20 Batch 90/2125] avg loss 0.000624524, throughput 6.01941K wps
[Epoch 20 Batch 120/2125] avg loss 0.000444279, throughput 6.0086K wps
[Epoch 20 Batch 150/2125] avg loss 0.000570434, throughput 6.01373K wps
[Epoch 20 Batch 180/2125] avg loss 0.000487019, throughput 6.01443K wps
[Epoch 20 Batch 210/2125] avg loss 0.000640316, throughput 6.01122K wps
[Epoch 20 Batch 240/2125] avg loss 0.000721325, throughput 6.02264K wps
[Epoch 20 Batch 270/2125] avg loss 0.000526939, throughput 6.01916K wps
[Epoch 20 Batch 300/2125] avg loss 0.000350804, throughput 6.00831K wps
[Epoch 20 Batch 330/2125] avg loss 0.000574666, throughput 6.01856K wps
[Epoch 20 Batch 360/2125] avg loss 0.000524954, throughput 6.02043K wps
[Epoch 20 Batch 390/2125] avg loss 0.000744241, throughput 6.01469K wps
[Epoch 20 Batch 420/2125] avg loss 0.000564602, throughput 6.01277K wps
[Epoch 20 Batch 450/2125] avg loss 0.000571813, throughput 6.00613K wps
[Epoch 20 Batch 480/2125] avg loss 0.000638296, throughput 6.02308K wps
[Epoch 20 Batch 510/2125] avg loss 0.000626649, throughput 6.02438K wps
[Epoch 20 Batch 540/2125] avg loss 0.000713338, throughput 6.01147K wps
[Epoch 20 Batch 570/2125] avg loss 0.000510447, throughput 6.01771K wps
[Epoch 20 Batch 600/2125] avg loss 0.000524962, throughput 6.01332K wps
[Epoch 20 Batch 630/2125] avg loss 0.000839797, throughput 6.00433K wps
[Epoch 20 Batch 660/2125] avg loss 0.000641997, throughput 6.01524K wps
[Epoch 20 Batch 690/2125] avg loss 0.000671449, throughput 6.01047K wps
[Epoch 20 Batch 720/2125] avg loss 0.000669292, throughput 6.01561K wps
[Epoch 20 Batch 750/2125] avg loss 0.000974328, throughput 6.00979K wps
[Epoch 20 Batch 780/2125] avg loss 0.00057325, throughput 6.01357K wps
[Epoch 20 Batch 810/2125] avg loss 0.000537563, throughput 6.0204K wps
[Epoch 20 Batch 840/2125] avg loss 0.000407437, throughput 6.02402K wps
[Epoch 20 Batch 870/2125] avg loss 0.00089206, throughput 6.01842K wps
[Epoch 20 Batch 900/2125] avg loss 0.000633254, throughput 6.01551K wps
[Epoch 20 Batch 930/2125] avg loss 0.000629708, throughput 6.01676K wps
[Epoch 20 Batch 960/2125] avg loss 0.000748553, throughput 6.00918K wps
[Epoch 20 Batch 990/2125] avg loss 0.000686213, throughput 6.00545K wps
[Epoch 20 Batch 1020/2125] avg loss 0.000578891, throughput 6.01366K wps
[Epoch 20 Batch 1050/2125] avg loss 0.000543255, throughput 6.0151K wps
[Epoch 20 Batch 1080/2125] avg loss 0.000553692, throughput 6.01683K wps
[Epoch 20 Batch 1110/2125] avg loss 0.000554613, throughput 6.02331K wps
[Epoch 20 Batch 1140/2125] avg loss 0.000818497, throughput 6.0145K wps
[Epoch 20 Batch 1170/2125] avg loss 0.000590303, throughput 6.01724K wps
[Epoch 20 Batch 1200/2125] avg loss 0.000806386, throughput 6.02102K wps
[Epoch 20 Batch 1230/2125] avg loss 0.000752377, throughput 6.0225K wps
[Epoch 20 Batch 1260/2125] avg loss 0.00101708, throughput 6.01636K wps
[Epoch 20 Batch 1290/2125] avg loss 0.000660204, throughput 6.01754K wps
[Epoch 20 Batch 1320/2125] avg loss 0.000730564, throughput 6.02132K wps
[Epoch 20 Batch 1350/2125] avg loss 0.000867173, throughput 6.01539K wps
[Epoch 20 Batch 1380/2125] avg loss 0.00077384, throughput 6.01314K wps
[Epoch 20 Batch 1410/2125] avg loss 0.000881282, throughput 6.01565K wps
[Epoch 20 Batch 1440/2125] avg loss 0.000714256, throughput 6.01566K wps
[Epoch 20 Batch 1470/2125] avg loss 0.000566462, throughput 6.01278K wps
[Epoch 20 Batch 1500/2125] avg loss 0.000906905, throughput 6.01752K wps
[Epoch 20 Batch 1530/2125] avg loss 0.000622677, throughput 6.02037K wps
[Epoch 20 Batch 1560/2125] avg loss 0.0010323, throughput 6.01569K wps
[Epoch 20 Batch 1590/2125] avg loss 0.000647772, throughput 6.0048K wps
[Epoch 20 Batch 1620/2125] avg loss 0.000664852, throughput 6.00696K wps
[Epoch 20 Batch 1650/2125] avg loss 0.000726782, throughput 6.00244K wps
[Epoch 20 Batch 1680/2125] avg loss 0.000873103, throughput 6.02135K wps
[Epoch 20 Batch 1710/2125] avg loss 0.000807888, throughput 6.01107K wps
[Epoch 20 Batch 1740/2125] avg loss 0.000901429, throughput 6.01365K wps
[Epoch 20 Batch 1770/2125] avg loss 0.000755701, throughput 6.01325K wps
[Epoch 20 Batch 1800/2125] avg loss 0.000615265, throughput 6.0176K wps
[Epoch 20 Batch 1830/2125] avg loss 0.000845683, throughput 6.00965K wps
[Epoch 20 Batch 1860/2125] avg loss 0.00106253, throughput 6.01747K wps
[Epoch 20 Batch 1890/2125] avg loss 0.000678386, throughput 6.01485K wps
[Epoch 20 Batch 1920/2125] avg loss 0.000942274, throughput 6.01203K wps
[Epoch 20 Batch 1950/2125] avg loss 0.00104466, throughput 6.0076K wps
[Epoch 20 Batch 1980/2125] avg loss 0.000753422, throughput 6.0095K wps
[Epoch 20 Batch 2010/2125] avg loss 0.000770432, throughput 6.00156K wps
[Epoch 20 Batch 2040/2125] avg loss 0.000893668, throughput 6.01521K wps
[Epoch 20 Batch 2070/2125] avg loss 0.000749015, throughput 6.01072K wps
[Epoch 20 Batch 2100/2125] avg loss 0.000596037, throughput 6.00889K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 20] train avg loss 0.000697211, test acc 0.9261, test avg loss 0.442336, throughput 6.01641K wps
[Epoch 21 Batch 30/2125] avg loss 0.00039619, throughput 6.14179K wps
[Epoch 21 Batch 60/2125] avg loss 0.00048563, throughput 6.0184K wps
[Epoch 21 Batch 90/2125] avg loss 0.000725033, throughput 6.0134K wps
[Epoch 21 Batch 120/2125] avg loss 0.000606385, throughput 6.00383K wps
[Epoch 21 Batch 150/2125] avg loss 0.000854238, throughput 6.01011K wps
[Epoch 21 Batch 180/2125] avg loss 0.000826935, throughput 6.00768K wps
[Epoch 21 Batch 210/2125] avg loss 0.000468602, throughput 6.01734K wps
[Epoch 21 Batch 240/2125] avg loss 0.000551663, throughput 6.01436K wps
[Epoch 21 Batch 270/2125] avg loss 0.000479449, throughput 6.00294K wps
[Epoch 21 Batch 300/2125] avg loss 0.000404127, throughput 6.00379K wps
[Epoch 21 Batch 330/2125] avg loss 0.000520148, throughput 6.00216K wps
[Epoch 21 Batch 360/2125] avg loss 0.000708235, throughput 6.01284K wps
[Epoch 21 Batch 390/2125] avg loss 0.00056685, throughput 6.00951K wps
[Epoch 21 Batch 420/2125] avg loss 0.000391314, throughput 6.00696K wps
[Epoch 21 Batch 450/2125] avg loss 0.000548245, throughput 6.01318K wps
[Epoch 21 Batch 480/2125] avg loss 0.000491156, throughput 6.01043K wps
[Epoch 21 Batch 510/2125] avg loss 0.000652402, throughput 6.00956K wps
[Epoch 21 Batch 540/2125] avg loss 0.000504269, throughput 6.01022K wps
[Epoch 21 Batch 570/2125] avg loss 0.000493398, throughput 6.00955K wps
[Epoch 21 Batch 600/2125] avg loss 0.000546528, throughput 6.0101K wps
[Epoch 21 Batch 630/2125] avg loss 0.00060661, throughput 6.01465K wps
[Epoch 21 Batch 660/2125] avg loss 0.000935998, throughput 6.01135K wps
[Epoch 21 Batch 690/2125] avg loss 0.000658706, throughput 5.99809K wps
[Epoch 21 Batch 720/2125] avg loss 0.000502668, throughput 6.01145K wps
[Epoch 21 Batch 750/2125] avg loss 0.000706524, throughput 6.00761K wps
[Epoch 21 Batch 780/2125] avg loss 0.00050079, throughput 6.01567K wps
[Epoch 21 Batch 810/2125] avg loss 0.000478498, throughput 6.02146K wps
[Epoch 21 Batch 840/2125] avg loss 0.000669822, throughput 6.01647K wps
[Epoch 21 Batch 870/2125] avg loss 0.000713259, throughput 6.01581K wps
[Epoch 21 Batch 900/2125] avg loss 0.000584195, throughput 6.01519K wps
[Epoch 21 Batch 930/2125] avg loss 0.000572599, throughput 6.01828K wps
[Epoch 21 Batch 960/2125] avg loss 0.00079627, throughput 6.00279K wps
[Epoch 21 Batch 990/2125] avg loss 0.000618401, throughput 6.01097K wps
[Epoch 21 Batch 1020/2125] avg loss 0.000605419, throughput 6.01338K wps
[Epoch 21 Batch 1050/2125] avg loss 0.000506593, throughput 6.00346K wps
[Epoch 21 Batch 1080/2125] avg loss 0.000834857, throughput 6.01373K wps
[Epoch 21 Batch 1110/2125] avg loss 0.000550195, throughput 6.01359K wps
[Epoch 21 Batch 1140/2125] avg loss 0.000519227, throughput 6.01978K wps
[Epoch 21 Batch 1170/2125] avg loss 0.000716951, throughput 6.0119K wps
[Epoch 21 Batch 1200/2125] avg loss 0.000619133, throughput 6.02245K wps
[Epoch 21 Batch 1230/2125] avg loss 0.000662163, throughput 6.01596K wps
[Epoch 21 Batch 1260/2125] avg loss 0.000560288, throughput 6.02519K wps
[Epoch 21 Batch 1290/2125] avg loss 0.000862792, throughput 6.01407K wps
[Epoch 21 Batch 1320/2125] avg loss 0.000795563, throughput 5.9953K wps
[Epoch 21 Batch 1350/2125] avg loss 0.000976533, throughput 6.01219K wps
[Epoch 21 Batch 1380/2125] avg loss 0.000690544, throughput 6.02095K wps
[Epoch 21 Batch 1410/2125] avg loss 0.000833229, throughput 6.03059K wps
[Epoch 21 Batch 1440/2125] avg loss 0.000588659, throughput 6.02915K wps
[Epoch 21 Batch 1470/2125] avg loss 0.000554693, throughput 6.00545K wps
[Epoch 21 Batch 1500/2125] avg loss 0.000843535, throughput 6.01707K wps
[Epoch 21 Batch 1530/2125] avg loss 0.000556222, throughput 6.01407K wps
[Epoch 21 Batch 1560/2125] avg loss 0.000780572, throughput 6.01692K wps
[Epoch 21 Batch 1590/2125] avg loss 0.000690894, throughput 6.02087K wps
[Epoch 21 Batch 1620/2125] avg loss 0.000848758, throughput 6.02637K wps
[Epoch 21 Batch 1650/2125] avg loss 0.000767833, throughput 6.01782K wps
[Epoch 21 Batch 1680/2125] avg loss 0.00073271, throughput 6.00553K wps
[Epoch 21 Batch 1710/2125] avg loss 0.000622312, throughput 6.02057K wps
[Epoch 21 Batch 1740/2125] avg loss 0.000958657, throughput 6.02261K wps
[Epoch 21 Batch 1770/2125] avg loss 0.000579405, throughput 6.02198K wps
[Epoch 21 Batch 1800/2125] avg loss 0.000620973, throughput 6.03092K wps
[Epoch 21 Batch 1830/2125] avg loss 0.000867364, throughput 6.01295K wps
[Epoch 21 Batch 1860/2125] avg loss 0.000559711, throughput 6.00581K wps
[Epoch 21 Batch 1890/2125] avg loss 0.000602671, throughput 6.01235K wps
[Epoch 21 Batch 1920/2125] avg loss 0.000502658, throughput 6.01962K wps
[Epoch 21 Batch 1950/2125] avg loss 0.00089296, throughput 6.02184K wps
[Epoch 21 Batch 1980/2125] avg loss 0.000893674, throughput 6.02594K wps
[Epoch 21 Batch 2010/2125] avg loss 0.000905217, throughput 6.01176K wps
[Epoch 21 Batch 2040/2125] avg loss 0.00103153, throughput 6.00578K wps
[Epoch 21 Batch 2070/2125] avg loss 0.00084201, throughput 6.00968K wps
[Epoch 21 Batch 2100/2125] avg loss 0.000629403, throughput 6.00254K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 21] train avg loss 0.000663326, test acc 0.9250, test avg loss 0.459595, throughput 6.01528K wps
[Epoch 22 Batch 30/2125] avg loss 0.000290195, throughput 6.14975K wps
[Epoch 22 Batch 60/2125] avg loss 0.00029681, throughput 6.00314K wps
[Epoch 22 Batch 90/2125] avg loss 0.000615914, throughput 6.01149K wps
[Epoch 22 Batch 120/2125] avg loss 0.000495977, throughput 6.01655K wps
[Epoch 22 Batch 150/2125] avg loss 0.000507743, throughput 6.02092K wps
[Epoch 22 Batch 180/2125] avg loss 0.000485192, throughput 6.01376K wps
[Epoch 22 Batch 210/2125] avg loss 0.000537761, throughput 6.01522K wps
[Epoch 22 Batch 240/2125] avg loss 0.00062583, throughput 6.01959K wps
[Epoch 22 Batch 270/2125] avg loss 0.000453176, throughput 6.00807K wps
[Epoch 22 Batch 300/2125] avg loss 0.000506248, throughput 6.01291K wps
[Epoch 22 Batch 330/2125] avg loss 0.000457156, throughput 6.0123K wps
[Epoch 22 Batch 360/2125] avg loss 0.000444959, throughput 6.01358K wps
[Epoch 22 Batch 390/2125] avg loss 0.00054904, throughput 6.02094K wps
[Epoch 22 Batch 420/2125] avg loss 0.00061308, throughput 6.02036K wps
[Epoch 22 Batch 450/2125] avg loss 0.000657776, throughput 6.01287K wps
[Epoch 22 Batch 480/2125] avg loss 0.00059497, throughput 6.01297K wps
[Epoch 22 Batch 510/2125] avg loss 0.000633118, throughput 6.02423K wps
[Epoch 22 Batch 540/2125] avg loss 0.00054359, throughput 6.01101K wps
[Epoch 22 Batch 570/2125] avg loss 0.000522441, throughput 6.00847K wps
[Epoch 22 Batch 600/2125] avg loss 0.000712409, throughput 6.00993K wps
[Epoch 22 Batch 630/2125] avg loss 0.000900664, throughput 6.01963K wps
[Epoch 22 Batch 660/2125] avg loss 0.000571884, throughput 6.0114K wps
[Epoch 22 Batch 690/2125] avg loss 0.000627811, throughput 6.02125K wps
[Epoch 22 Batch 720/2125] avg loss 0.000560262, throughput 6.01167K wps
[Epoch 22 Batch 750/2125] avg loss 0.000539228, throughput 6.01553K wps
[Epoch 22 Batch 780/2125] avg loss 0.000709209, throughput 6.00754K wps
[Epoch 22 Batch 810/2125] avg loss 0.000461627, throughput 6.02223K wps
[Epoch 22 Batch 840/2125] avg loss 0.000499415, throughput 6.01186K wps
[Epoch 22 Batch 870/2125] avg loss 0.00060505, throughput 6.0258K wps
[Epoch 22 Batch 900/2125] avg loss 0.000496182, throughput 6.00706K wps
[Epoch 22 Batch 930/2125] avg loss 0.000632283, throughput 6.01429K wps
[Epoch 22 Batch 960/2125] avg loss 0.000733675, throughput 6.01963K wps
[Epoch 22 Batch 990/2125] avg loss 0.000653397, throughput 6.01478K wps
[Epoch 22 Batch 1020/2125] avg loss 0.000586935, throughput 6.01233K wps
[Epoch 22 Batch 1050/2125] avg loss 0.000676327, throughput 6.01226K wps
[Epoch 22 Batch 1080/2125] avg loss 0.000463647, throughput 6.01634K wps
[Epoch 22 Batch 1110/2125] avg loss 0.00081571, throughput 6.01493K wps
[Epoch 22 Batch 1140/2125] avg loss 0.000716583, throughput 6.01794K wps
[Epoch 22 Batch 1170/2125] avg loss 0.00053909, throughput 6.02223K wps
[Epoch 22 Batch 1200/2125] avg loss 0.000599394, throughput 6.01607K wps
[Epoch 22 Batch 1230/2125] avg loss 0.000667581, throughput 6.02318K wps
[Epoch 22 Batch 1260/2125] avg loss 0.000731853, throughput 6.0154K wps
[Epoch 22 Batch 1290/2125] avg loss 0.000648828, throughput 6.00108K wps
[Epoch 22 Batch 1320/2125] avg loss 0.000543689, throughput 6.00997K wps
[Epoch 22 Batch 1350/2125] avg loss 0.000624873, throughput 6.02363K wps
[Epoch 22 Batch 1380/2125] avg loss 0.00064303, throughput 6.0107K wps
[Epoch 22 Batch 1410/2125] avg loss 0.000633293, throughput 6.00774K wps
[Epoch 22 Batch 1440/2125] avg loss 0.000825683, throughput 6.01679K wps
[Epoch 22 Batch 1470/2125] avg loss 0.000648122, throughput 6.01358K wps
[Epoch 22 Batch 1500/2125] avg loss 0.000781286, throughput 6.00134K wps
[Epoch 22 Batch 1530/2125] avg loss 0.000544011, throughput 6.00759K wps
[Epoch 22 Batch 1560/2125] avg loss 0.0006343, throughput 6.00615K wps
[Epoch 22 Batch 1590/2125] avg loss 0.000805804, throughput 6.01177K wps
[Epoch 22 Batch 1620/2125] avg loss 0.000607564, throughput 6.01358K wps
[Epoch 22 Batch 1650/2125] avg loss 0.000705795, throughput 6.01341K wps
[Epoch 22 Batch 1680/2125] avg loss 0.000752438, throughput 6.01276K wps
[Epoch 22 Batch 1710/2125] avg loss 0.000893468, throughput 6.01366K wps
[Epoch 22 Batch 1740/2125] avg loss 0.000702251, throughput 6.00929K wps
[Epoch 22 Batch 1770/2125] avg loss 0.000836152, throughput 6.00391K wps
[Epoch 22 Batch 1800/2125] avg loss 0.000634814, throughput 6.00176K wps
[Epoch 22 Batch 1830/2125] avg loss 0.000804042, throughput 6.01501K wps
[Epoch 22 Batch 1860/2125] avg loss 0.000723603, throughput 6.01739K wps
[Epoch 22 Batch 1890/2125] avg loss 0.000763126, throughput 6.01637K wps
[Epoch 22 Batch 1920/2125] avg loss 0.000894262, throughput 6.00687K wps
[Epoch 22 Batch 1950/2125] avg loss 0.00085442, throughput 6.01715K wps
[Epoch 22 Batch 1980/2125] avg loss 0.000963072, throughput 6.00629K wps
[Epoch 22 Batch 2010/2125] avg loss 0.000667212, throughput 6.01218K wps
[Epoch 22 Batch 2040/2125] avg loss 0.000400209, throughput 5.99755K wps
[Epoch 22 Batch 2070/2125] avg loss 0.000883759, throughput 6.00884K wps
[Epoch 22 Batch 2100/2125] avg loss 0.000712644, throughput 6.01801K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 22] train avg loss 0.000638586, test acc 0.9248, test avg loss 0.462367, throughput 6.01513K wps
[Epoch 23 Batch 30/2125] avg loss 0.000407457, throughput 6.15218K wps
[Epoch 23 Batch 60/2125] avg loss 0.00045725, throughput 6.02667K wps
[Epoch 23 Batch 90/2125] avg loss 0.000521373, throughput 6.00934K wps
[Epoch 23 Batch 120/2125] avg loss 0.000614969, throughput 6.00638K wps
[Epoch 23 Batch 150/2125] avg loss 0.000646022, throughput 6.00587K wps
[Epoch 23 Batch 180/2125] avg loss 0.00049718, throughput 4.93103K wps
[Epoch 23 Batch 210/2125] avg loss 0.000400283, throughput 6.01089K wps
[Epoch 23 Batch 240/2125] avg loss 0.000726196, throughput 6.01232K wps
[Epoch 23 Batch 270/2125] avg loss 0.000601271, throughput 6.00933K wps
[Epoch 23 Batch 300/2125] avg loss 0.000428004, throughput 6.0059K wps
[Epoch 23 Batch 330/2125] avg loss 0.000599645, throughput 6.00511K wps
[Epoch 23 Batch 360/2125] avg loss 0.000368194, throughput 6.01324K wps
[Epoch 23 Batch 390/2125] avg loss 0.000608887, throughput 6.02167K wps
[Epoch 23 Batch 420/2125] avg loss 0.000466015, throughput 6.02405K wps
[Epoch 23 Batch 450/2125] avg loss 0.000612515, throughput 6.00395K wps
[Epoch 23 Batch 480/2125] avg loss 0.000581691, throughput 5.97265K wps
[Epoch 23 Batch 510/2125] avg loss 0.000544012, throughput 6.03746K wps
[Epoch 23 Batch 540/2125] avg loss 0.000623947, throughput 6.00772K wps
[Epoch 23 Batch 570/2125] avg loss 0.000441507, throughput 6.00923K wps
[Epoch 23 Batch 600/2125] avg loss 0.000653524, throughput 6.00736K wps
[Epoch 23 Batch 630/2125] avg loss 0.000549624, throughput 6.01457K wps
[Epoch 23 Batch 660/2125] avg loss 0.000485816, throughput 6.01147K wps
[Epoch 23 Batch 690/2125] avg loss 0.000590541, throughput 6.01099K wps
[Epoch 23 Batch 720/2125] avg loss 0.000620451, throughput 6.02324K wps
[Epoch 23 Batch 750/2125] avg loss 0.000630684, throughput 6.0171K wps
[Epoch 23 Batch 780/2125] avg loss 0.000800198, throughput 6.00809K wps
[Epoch 23 Batch 810/2125] avg loss 0.000633709, throughput 6.02147K wps
[Epoch 23 Batch 840/2125] avg loss 0.000418382, throughput 6.02891K wps
[Epoch 23 Batch 870/2125] avg loss 0.000864962, throughput 6.02209K wps
[Epoch 23 Batch 900/2125] avg loss 0.00065591, throughput 6.01016K wps
[Epoch 23 Batch 930/2125] avg loss 0.000439099, throughput 6.02171K wps
[Epoch 23 Batch 960/2125] avg loss 0.000496203, throughput 6.02153K wps
[Epoch 23 Batch 990/2125] avg loss 0.000548647, throughput 6.01383K wps
[Epoch 23 Batch 1020/2125] avg loss 0.000513168, throughput 6.0097K wps
[Epoch 23 Batch 1050/2125] avg loss 0.00068441, throughput 6.0129K wps
[Epoch 23 Batch 1080/2125] avg loss 0.000581021, throughput 6.02435K wps
[Epoch 23 Batch 1110/2125] avg loss 0.00053705, throughput 6.02118K wps
[Epoch 23 Batch 1140/2125] avg loss 0.000720621, throughput 6.01649K wps
[Epoch 23 Batch 1170/2125] avg loss 0.000595768, throughput 6.01839K wps
[Epoch 23 Batch 1200/2125] avg loss 0.00054517, throughput 6.01606K wps
[Epoch 23 Batch 1230/2125] avg loss 0.00062558, throughput 6.02608K wps
[Epoch 23 Batch 1260/2125] avg loss 0.000525051, throughput 6.01842K wps
[Epoch 23 Batch 1290/2125] avg loss 0.000749353, throughput 6.00738K wps
[Epoch 23 Batch 1320/2125] avg loss 0.00069362, throughput 6.01182K wps
[Epoch 23 Batch 1350/2125] avg loss 0.000620591, throughput 6.00403K wps
[Epoch 23 Batch 1380/2125] avg loss 0.000684608, throughput 6.00943K wps
[Epoch 23 Batch 1410/2125] avg loss 0.000416293, throughput 6.00857K wps
[Epoch 23 Batch 1440/2125] avg loss 0.000687062, throughput 6.00962K wps
[Epoch 23 Batch 1470/2125] avg loss 0.000625315, throughput 6.00288K wps
[Epoch 23 Batch 1500/2125] avg loss 0.000510812, throughput 6.01494K wps
[Epoch 23 Batch 1530/2125] avg loss 0.000667301, throughput 6.00951K wps
[Epoch 23 Batch 1560/2125] avg loss 0.000746795, throughput 6.01697K wps
[Epoch 23 Batch 1590/2125] avg loss 0.000883162, throughput 6.00779K wps
[Epoch 23 Batch 1620/2125] avg loss 0.000564085, throughput 6.01505K wps
[Epoch 23 Batch 1650/2125] avg loss 0.00076655, throughput 6.0089K wps
[Epoch 23 Batch 1680/2125] avg loss 0.000573836, throughput 6.00774K wps
[Epoch 23 Batch 1710/2125] avg loss 0.000665816, throughput 6.00236K wps
[Epoch 23 Batch 1740/2125] avg loss 0.000363151, throughput 6.01136K wps
[Epoch 23 Batch 1770/2125] avg loss 0.000932397, throughput 6.01156K wps
[Epoch 23 Batch 1800/2125] avg loss 0.000454997, throughput 6.02231K wps
[Epoch 23 Batch 1830/2125] avg loss 0.000506721, throughput 6.02319K wps
[Epoch 23 Batch 1860/2125] avg loss 0.000664384, throughput 6.02701K wps
[Epoch 23 Batch 1890/2125] avg loss 0.000953529, throughput 6.01383K wps
[Epoch 23 Batch 1920/2125] avg loss 0.000783683, throughput 6.02703K wps
[Epoch 23 Batch 1950/2125] avg loss 0.000550741, throughput 6.02508K wps
[Epoch 23 Batch 1980/2125] avg loss 0.000506892, throughput 6.01359K wps
[Epoch 23 Batch 2010/2125] avg loss 0.00087812, throughput 6.01169K wps
[Epoch 23 Batch 2040/2125] avg loss 0.000813206, throughput 6.01093K wps
[Epoch 23 Batch 2070/2125] avg loss 0.000964453, throughput 6.01K wps
[Epoch 23 Batch 2100/2125] avg loss 0.00100281, throughput 6.02474K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 23] train avg loss 0.000615001, test acc 0.9248, test avg loss 0.469554, throughput 5.99736K wps
[Epoch 24 Batch 30/2125] avg loss 0.000419331, throughput 6.14221K wps
[Epoch 24 Batch 60/2125] avg loss 0.000515763, throughput 6.01093K wps
[Epoch 24 Batch 90/2125] avg loss 0.000582948, throughput 6.01968K wps
[Epoch 24 Batch 120/2125] avg loss 0.000336269, throughput 6.01816K wps
[Epoch 24 Batch 150/2125] avg loss 0.000477618, throughput 6.02558K wps
[Epoch 24 Batch 180/2125] avg loss 0.000477769, throughput 6.02449K wps
[Epoch 24 Batch 210/2125] avg loss 0.000451511, throughput 6.01122K wps
[Epoch 24 Batch 240/2125] avg loss 0.000540809, throughput 6.01691K wps
[Epoch 24 Batch 270/2125] avg loss 0.00052557, throughput 6.01702K wps
[Epoch 24 Batch 300/2125] avg loss 0.000531388, throughput 6.01451K wps
[Epoch 24 Batch 330/2125] avg loss 0.000427567, throughput 6.00446K wps
[Epoch 24 Batch 360/2125] avg loss 0.000522747, throughput 6.00946K wps
[Epoch 24 Batch 390/2125] avg loss 0.000528415, throughput 6.01477K wps
[Epoch 24 Batch 420/2125] avg loss 0.00063799, throughput 6.018K wps
[Epoch 24 Batch 450/2125] avg loss 0.000469251, throughput 6.02026K wps
[Epoch 24 Batch 480/2125] avg loss 0.000476049, throughput 6.01559K wps
[Epoch 24 Batch 510/2125] avg loss 0.000678732, throughput 6.01374K wps
[Epoch 24 Batch 540/2125] avg loss 0.000639008, throughput 6.01546K wps
[Epoch 24 Batch 570/2125] avg loss 0.000601423, throughput 6.01019K wps
[Epoch 24 Batch 600/2125] avg loss 0.000603147, throughput 6.00425K wps
[Epoch 24 Batch 630/2125] avg loss 0.000448751, throughput 6.00356K wps
[Epoch 24 Batch 660/2125] avg loss 0.000612507, throughput 6.01228K wps
[Epoch 24 Batch 690/2125] avg loss 0.000824992, throughput 6.0092K wps
[Epoch 24 Batch 720/2125] avg loss 0.000649829, throughput 6.00863K wps
[Epoch 24 Batch 750/2125] avg loss 0.000627064, throughput 5.99614K wps
[Epoch 24 Batch 780/2125] avg loss 0.000582563, throughput 6.01068K wps
[Epoch 24 Batch 810/2125] avg loss 0.000518798, throughput 6.01177K wps
[Epoch 24 Batch 840/2125] avg loss 0.000590077, throughput 6.01053K wps
[Epoch 24 Batch 870/2125] avg loss 0.000576128, throughput 6.00993K wps
[Epoch 24 Batch 900/2125] avg loss 0.000565188, throughput 6.02095K wps
[Epoch 24 Batch 930/2125] avg loss 0.000576594, throughput 6.01392K wps
[Epoch 24 Batch 960/2125] avg loss 0.000502363, throughput 6.01613K wps
[Epoch 24 Batch 990/2125] avg loss 0.000773562, throughput 6.01621K wps
[Epoch 24 Batch 1020/2125] avg loss 0.000407554, throughput 6.01306K wps
[Epoch 24 Batch 1050/2125] avg loss 0.000585698, throughput 6.01548K wps
[Epoch 24 Batch 1080/2125] avg loss 0.00073649, throughput 6.01667K wps
[Epoch 24 Batch 1110/2125] avg loss 0.000636003, throughput 6.01197K wps
[Epoch 24 Batch 1140/2125] avg loss 0.000650988, throughput 6.01907K wps
[Epoch 24 Batch 1170/2125] avg loss 0.000644103, throughput 6.01215K wps
[Epoch 24 Batch 1200/2125] avg loss 0.000671624, throughput 6.01286K wps
[Epoch 24 Batch 1230/2125] avg loss 0.000628573, throughput 6.00665K wps
[Epoch 24 Batch 1260/2125] avg loss 0.000575092, throughput 6.0117K wps
[Epoch 24 Batch 1290/2125] avg loss 0.000513482, throughput 6.01247K wps
[Epoch 24 Batch 1320/2125] avg loss 0.000734791, throughput 6.02152K wps
[Epoch 24 Batch 1350/2125] avg loss 0.000750718, throughput 6.00972K wps
[Epoch 24 Batch 1380/2125] avg loss 0.000586817, throughput 6.01878K wps
[Epoch 24 Batch 1410/2125] avg loss 0.00062986, throughput 6.02086K wps
[Epoch 24 Batch 1440/2125] avg loss 0.000632759, throughput 6.01446K wps
[Epoch 24 Batch 1470/2125] avg loss 0.000715695, throughput 6.0186K wps
[Epoch 24 Batch 1500/2125] avg loss 0.000546277, throughput 6.01871K wps
[Epoch 24 Batch 1530/2125] avg loss 0.000622273, throughput 6.01683K wps
[Epoch 24 Batch 1560/2125] avg loss 0.000403001, throughput 6.01827K wps
[Epoch 24 Batch 1590/2125] avg loss 0.000980262, throughput 6.01641K wps
[Epoch 24 Batch 1620/2125] avg loss 0.000610188, throughput 6.01951K wps
[Epoch 24 Batch 1650/2125] avg loss 0.000834244, throughput 6.01739K wps
[Epoch 24 Batch 1680/2125] avg loss 0.000598953, throughput 6.01429K wps
[Epoch 24 Batch 1710/2125] avg loss 0.000440807, throughput 6.01757K wps
[Epoch 24 Batch 1740/2125] avg loss 0.00060386, throughput 6.01445K wps
[Epoch 24 Batch 1770/2125] avg loss 0.000717238, throughput 6.00765K wps
[Epoch 24 Batch 1800/2125] avg loss 0.0008942, throughput 6.01098K wps
[Epoch 24 Batch 1830/2125] avg loss 0.000560582, throughput 6.01369K wps
[Epoch 24 Batch 1860/2125] avg loss 0.000548625, throughput 6.00154K wps
[Epoch 24 Batch 1890/2125] avg loss 0.000589853, throughput 6.00807K wps
[Epoch 24 Batch 1920/2125] avg loss 0.000699839, throughput 6.00189K wps
[Epoch 24 Batch 1950/2125] avg loss 0.000823463, throughput 6.01389K wps
[Epoch 24 Batch 1980/2125] avg loss 0.000515272, throughput 6.01315K wps
[Epoch 24 Batch 2010/2125] avg loss 0.000778411, throughput 6.00275K wps
[Epoch 24 Batch 2040/2125] avg loss 0.000679197, throughput 6.02181K wps
[Epoch 24 Batch 2070/2125] avg loss 0.000447823, throughput 6.00428K wps
[Epoch 24 Batch 2100/2125] avg loss 0.000800966, throughput 6.00506K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 24] train avg loss 0.000602932, test acc 0.9241, test avg loss 0.479316, throughput 6.01508K wps
[Epoch 25 Batch 30/2125] avg loss 0.000271705, throughput 6.14206K wps
[Epoch 25 Batch 60/2125] avg loss 0.000387683, throughput 6.01369K wps
[Epoch 25 Batch 90/2125] avg loss 0.000510608, throughput 6.00557K wps
[Epoch 25 Batch 120/2125] avg loss 0.000390805, throughput 6.01755K wps
[Epoch 25 Batch 150/2125] avg loss 0.000544998, throughput 6.01429K wps
[Epoch 25 Batch 180/2125] avg loss 0.000416377, throughput 6.00039K wps
[Epoch 25 Batch 210/2125] avg loss 0.000434481, throughput 6.01391K wps
[Epoch 25 Batch 240/2125] avg loss 0.000465505, throughput 6.00619K wps
[Epoch 25 Batch 270/2125] avg loss 0.00073504, throughput 6.01001K wps
[Epoch 25 Batch 300/2125] avg loss 0.000634894, throughput 6.02153K wps
[Epoch 25 Batch 330/2125] avg loss 0.000390469, throughput 6.00871K wps
[Epoch 25 Batch 360/2125] avg loss 0.000474315, throughput 6.02239K wps
[Epoch 25 Batch 390/2125] avg loss 0.000617041, throughput 6.01037K wps
[Epoch 25 Batch 420/2125] avg loss 0.000587929, throughput 6.01346K wps
[Epoch 25 Batch 450/2125] avg loss 0.000373358, throughput 6.02185K wps
[Epoch 25 Batch 480/2125] avg loss 0.000541302, throughput 6.01791K wps
[Epoch 25 Batch 510/2125] avg loss 0.000693112, throughput 6.01708K wps
[Epoch 25 Batch 540/2125] avg loss 0.000511134, throughput 6.01074K wps
[Epoch 25 Batch 570/2125] avg loss 0.000767926, throughput 6.00051K wps
[Epoch 25 Batch 600/2125] avg loss 0.000406837, throughput 6.01315K wps
[Epoch 25 Batch 630/2125] avg loss 0.000571921, throughput 6.01105K wps
[Epoch 25 Batch 660/2125] avg loss 0.000605171, throughput 6.02046K wps
[Epoch 25 Batch 690/2125] avg loss 0.000541737, throughput 6.01998K wps
[Epoch 25 Batch 720/2125] avg loss 0.000449484, throughput 6.0083K wps
[Epoch 25 Batch 750/2125] avg loss 0.000515273, throughput 6.0222K wps
[Epoch 25 Batch 780/2125] avg loss 0.000743272, throughput 6.02382K wps
[Epoch 25 Batch 810/2125] avg loss 0.000490807, throughput 6.01516K wps
[Epoch 25 Batch 840/2125] avg loss 0.000491697, throughput 6.00669K wps
[Epoch 25 Batch 870/2125] avg loss 0.000546006, throughput 6.02151K wps
[Epoch 25 Batch 900/2125] avg loss 0.000580528, throughput 6.02159K wps
[Epoch 25 Batch 930/2125] avg loss 0.000799338, throughput 6.02171K wps
[Epoch 25 Batch 960/2125] avg loss 0.000630431, throughput 6.01141K wps
[Epoch 25 Batch 990/2125] avg loss 0.000364279, throughput 6.01689K wps
[Epoch 25 Batch 1020/2125] avg loss 0.000429925, throughput 6.01476K wps
[Epoch 25 Batch 1050/2125] avg loss 0.000715964, throughput 6.01155K wps
[Epoch 25 Batch 1080/2125] avg loss 0.000757848, throughput 6.01084K wps
[Epoch 25 Batch 1110/2125] avg loss 0.000546666, throughput 6.00046K wps
[Epoch 25 Batch 1140/2125] avg loss 0.000485318, throughput 6.0128K wps
[Epoch 25 Batch 1170/2125] avg loss 0.000495412, throughput 6.01067K wps
[Epoch 25 Batch 1200/2125] avg loss 0.000554646, throughput 6.01328K wps
[Epoch 25 Batch 1230/2125] avg loss 0.00055897, throughput 6.01944K wps
[Epoch 25 Batch 1260/2125] avg loss 0.00040022, throughput 6.01783K wps
[Epoch 25 Batch 1290/2125] avg loss 0.000692616, throughput 6.01083K wps
[Epoch 25 Batch 1320/2125] avg loss 0.000306764, throughput 6.01327K wps
[Epoch 25 Batch 1350/2125] avg loss 0.000806729, throughput 6.0139K wps
[Epoch 25 Batch 1380/2125] avg loss 0.000642317, throughput 6.00579K wps
[Epoch 25 Batch 1410/2125] avg loss 0.000697555, throughput 6.02204K wps
[Epoch 25 Batch 1440/2125] avg loss 0.000487706, throughput 6.01689K wps
[Epoch 25 Batch 1470/2125] avg loss 0.0007974, throughput 6.01679K wps
[Epoch 25 Batch 1500/2125] avg loss 0.000580874, throughput 5.99511K wps
[Epoch 25 Batch 1530/2125] avg loss 0.000609445, throughput 5.99689K wps
[Epoch 25 Batch 1560/2125] avg loss 0.000367798, throughput 6.0127K wps
[Epoch 25 Batch 1590/2125] avg loss 0.000918323, throughput 6.00985K wps
[Epoch 25 Batch 1620/2125] avg loss 0.000450097, throughput 6.01396K wps
[Epoch 25 Batch 1650/2125] avg loss 0.000670622, throughput 6.00591K wps
[Epoch 25 Batch 1680/2125] avg loss 0.000550048, throughput 6.00985K wps
[Epoch 25 Batch 1710/2125] avg loss 0.000629211, throughput 6.0144K wps
[Epoch 25 Batch 1740/2125] avg loss 0.000682177, throughput 6.01351K wps
[Epoch 25 Batch 1770/2125] avg loss 0.00042311, throughput 6.02667K wps
[Epoch 25 Batch 1800/2125] avg loss 0.00069979, throughput 6.01044K wps
[Epoch 25 Batch 1830/2125] avg loss 0.000867429, throughput 6.02411K wps
[Epoch 25 Batch 1860/2125] avg loss 0.000789773, throughput 6.01204K wps
[Epoch 25 Batch 1890/2125] avg loss 0.000737013, throughput 6.01368K wps
[Epoch 25 Batch 1920/2125] avg loss 0.000677094, throughput 6.0158K wps
[Epoch 25 Batch 1950/2125] avg loss 0.000720097, throughput 6.01047K wps
[Epoch 25 Batch 1980/2125] avg loss 0.000426482, throughput 6.00858K wps
[Epoch 25 Batch 2010/2125] avg loss 0.000683929, throughput 6.01327K wps
[Epoch 25 Batch 2040/2125] avg loss 0.00054306, throughput 6.01358K wps
[Epoch 25 Batch 2070/2125] avg loss 0.000824048, throughput 6.02177K wps
[Epoch 25 Batch 2100/2125] avg loss 0.000669657, throughput 6.01459K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 25] train avg loss 0.000577146, test acc 0.9236, test avg loss 0.494024, throughput 6.01515K wps
[Epoch 26 Batch 30/2125] avg loss 0.000425627, throughput 6.15206K wps
[Epoch 26 Batch 60/2125] avg loss 0.000291608, throughput 6.01709K wps
[Epoch 26 Batch 90/2125] avg loss 0.000461954, throughput 6.01188K wps
[Epoch 26 Batch 120/2125] avg loss 0.000426445, throughput 6.01343K wps
[Epoch 26 Batch 150/2125] avg loss 0.000473252, throughput 6.00181K wps
[Epoch 26 Batch 180/2125] avg loss 0.000399594, throughput 6.02278K wps
[Epoch 26 Batch 210/2125] avg loss 0.000476684, throughput 6.01302K wps
[Epoch 26 Batch 240/2125] avg loss 0.000438476, throughput 6.01803K wps
[Epoch 26 Batch 270/2125] avg loss 0.00059276, throughput 6.02536K wps
[Epoch 26 Batch 300/2125] avg loss 0.000360117, throughput 6.01018K wps
[Epoch 26 Batch 330/2125] avg loss 0.00037346, throughput 6.00765K wps
[Epoch 26 Batch 360/2125] avg loss 0.000283371, throughput 6.01316K wps
[Epoch 26 Batch 390/2125] avg loss 0.00035877, throughput 6.00908K wps
[Epoch 26 Batch 420/2125] avg loss 0.000320211, throughput 6.01153K wps
[Epoch 26 Batch 450/2125] avg loss 0.00043382, throughput 6.00688K wps
[Epoch 26 Batch 480/2125] avg loss 0.000664841, throughput 6.01129K wps
[Epoch 26 Batch 510/2125] avg loss 0.000409762, throughput 6.00952K wps
[Epoch 26 Batch 540/2125] avg loss 0.000290099, throughput 6.00507K wps
[Epoch 26 Batch 570/2125] avg loss 0.00041799, throughput 6.01309K wps
[Epoch 26 Batch 600/2125] avg loss 0.000701869, throughput 6.01202K wps
[Epoch 26 Batch 630/2125] avg loss 0.000471855, throughput 5.99946K wps
[Epoch 26 Batch 660/2125] avg loss 0.000704069, throughput 6.00752K wps
[Epoch 26 Batch 690/2125] avg loss 0.000478916, throughput 6.01364K wps
[Epoch 26 Batch 720/2125] avg loss 0.000523917, throughput 6.0184K wps
[Epoch 26 Batch 750/2125] avg loss 0.000574208, throughput 6.01505K wps
[Epoch 26 Batch 780/2125] avg loss 0.000258236, throughput 6.00405K wps
[Epoch 26 Batch 810/2125] avg loss 0.000661136, throughput 6.01438K wps
[Epoch 26 Batch 840/2125] avg loss 0.000440092, throughput 6.01601K wps
[Epoch 26 Batch 870/2125] avg loss 0.000649803, throughput 6.01531K wps
[Epoch 26 Batch 900/2125] avg loss 0.000565071, throughput 6.01449K wps
[Epoch 26 Batch 930/2125] avg loss 0.000544992, throughput 6.01059K wps
[Epoch 26 Batch 960/2125] avg loss 0.00045454, throughput 6.00178K wps
[Epoch 26 Batch 990/2125] avg loss 0.000541862, throughput 6.0079K wps
[Epoch 26 Batch 1020/2125] avg loss 0.000487193, throughput 6.0094K wps
[Epoch 26 Batch 1050/2125] avg loss 0.000390949, throughput 6.01077K wps
[Epoch 26 Batch 1080/2125] avg loss 0.000636319, throughput 6.02216K wps
[Epoch 26 Batch 1110/2125] avg loss 0.000646407, throughput 6.02327K wps
[Epoch 26 Batch 1140/2125] avg loss 0.00077826, throughput 6.02184K wps
[Epoch 26 Batch 1170/2125] avg loss 0.000359855, throughput 5.99922K wps
[Epoch 26 Batch 1200/2125] avg loss 0.000486319, throughput 6.00737K wps
[Epoch 26 Batch 1230/2125] avg loss 0.000436942, throughput 6.00914K wps
[Epoch 26 Batch 1260/2125] avg loss 0.000551878, throughput 6.01797K wps
[Epoch 26 Batch 1290/2125] avg loss 0.000389224, throughput 6.00898K wps
[Epoch 26 Batch 1320/2125] avg loss 0.000499443, throughput 6.01202K wps
[Epoch 26 Batch 1350/2125] avg loss 0.000602729, throughput 6.01276K wps
[Epoch 26 Batch 1380/2125] avg loss 0.00089074, throughput 6.01477K wps
[Epoch 26 Batch 1410/2125] avg loss 0.000492661, throughput 6.01338K wps
[Epoch 26 Batch 1440/2125] avg loss 0.000516658, throughput 6.00909K wps
[Epoch 26 Batch 1470/2125] avg loss 0.000631157, throughput 6.01936K wps
[Epoch 26 Batch 1500/2125] avg loss 0.000676466, throughput 6.01925K wps
[Epoch 26 Batch 1530/2125] avg loss 0.000647959, throughput 6.00268K wps
[Epoch 26 Batch 1560/2125] avg loss 0.000646636, throughput 5.99608K wps
[Epoch 26 Batch 1590/2125] avg loss 0.000707123, throughput 6.00691K wps
[Epoch 26 Batch 1620/2125] avg loss 0.000666422, throughput 6.01301K wps
[Epoch 26 Batch 1650/2125] avg loss 0.000400134, throughput 6.01431K wps
[Epoch 26 Batch 1680/2125] avg loss 0.000374508, throughput 6.01535K wps
[Epoch 26 Batch 1710/2125] avg loss 0.00070017, throughput 6.00943K wps
[Epoch 26 Batch 1740/2125] avg loss 0.000531205, throughput 6.01704K wps
[Epoch 26 Batch 1770/2125] avg loss 0.000880355, throughput 6.01737K wps
[Epoch 26 Batch 1800/2125] avg loss 0.000699727, throughput 6.00671K wps
[Epoch 26 Batch 1830/2125] avg loss 0.00049437, throughput 6.01202K wps
[Epoch 26 Batch 1860/2125] avg loss 0.000829629, throughput 6.01102K wps
[Epoch 26 Batch 1890/2125] avg loss 0.000611524, throughput 6.01407K wps
[Epoch 26 Batch 1920/2125] avg loss 0.000937884, throughput 6.00623K wps
[Epoch 26 Batch 1950/2125] avg loss 0.000551543, throughput 6.00406K wps
[Epoch 26 Batch 1980/2125] avg loss 0.000793608, throughput 6.01682K wps
[Epoch 26 Batch 2010/2125] avg loss 0.000735326, throughput 6.0129K wps
[Epoch 26 Batch 2040/2125] avg loss 0.000749898, throughput 6.01368K wps
[Epoch 26 Batch 2070/2125] avg loss 0.000802252, throughput 6.01672K wps
[Epoch 26 Batch 2100/2125] avg loss 0.000719998, throughput 6.01052K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 26] train avg loss 0.000553531, test acc 0.9240, test avg loss 0.491739, throughput 6.01372K wps
[Epoch 27 Batch 30/2125] avg loss 0.000461976, throughput 6.14939K wps
[Epoch 27 Batch 60/2125] avg loss 0.000453594, throughput 6.01692K wps
[Epoch 27 Batch 90/2125] avg loss 0.000361077, throughput 6.01317K wps
[Epoch 27 Batch 120/2125] avg loss 0.000310093, throughput 6.01136K wps
[Epoch 27 Batch 150/2125] avg loss 0.000369974, throughput 6.00323K wps
[Epoch 27 Batch 180/2125] avg loss 0.000321167, throughput 6.00864K wps
[Epoch 27 Batch 210/2125] avg loss 0.000491113, throughput 6.01113K wps
[Epoch 27 Batch 240/2125] avg loss 0.000411316, throughput 5.99902K wps
[Epoch 27 Batch 270/2125] avg loss 0.000298734, throughput 6.00131K wps
[Epoch 27 Batch 300/2125] avg loss 0.000344618, throughput 6.00589K wps
[Epoch 27 Batch 330/2125] avg loss 0.000491089, throughput 6.00293K wps
[Epoch 27 Batch 360/2125] avg loss 0.00046421, throughput 6.01314K wps
[Epoch 27 Batch 390/2125] avg loss 0.000520696, throughput 6.01032K wps
[Epoch 27 Batch 420/2125] avg loss 0.00066248, throughput 6.00601K wps
[Epoch 27 Batch 450/2125] avg loss 0.000419275, throughput 6.0079K wps
[Epoch 27 Batch 480/2125] avg loss 0.000394808, throughput 6.01274K wps
[Epoch 27 Batch 510/2125] avg loss 0.00041254, throughput 6.00857K wps
[Epoch 27 Batch 540/2125] avg loss 0.000345647, throughput 6.01324K wps
[Epoch 27 Batch 570/2125] avg loss 0.000537557, throughput 6.0158K wps
[Epoch 27 Batch 600/2125] avg loss 0.000577103, throughput 6.00652K wps
[Epoch 27 Batch 630/2125] avg loss 0.000608617, throughput 6.02208K wps
[Epoch 27 Batch 660/2125] avg loss 0.000458672, throughput 6.01417K wps
[Epoch 27 Batch 690/2125] avg loss 0.000699829, throughput 6.01188K wps
[Epoch 27 Batch 720/2125] avg loss 0.000423679, throughput 6.00653K wps
[Epoch 27 Batch 750/2125] avg loss 0.00032367, throughput 6.01709K wps
[Epoch 27 Batch 780/2125] avg loss 0.000772496, throughput 6.0149K wps
[Epoch 27 Batch 810/2125] avg loss 0.000505426, throughput 6.00223K wps
[Epoch 27 Batch 840/2125] avg loss 0.00053431, throughput 6.01329K wps
[Epoch 27 Batch 870/2125] avg loss 0.000584356, throughput 6.0213K wps
[Epoch 27 Batch 900/2125] avg loss 0.000596516, throughput 6.01038K wps
[Epoch 27 Batch 930/2125] avg loss 0.0004771, throughput 6.00667K wps
[Epoch 27 Batch 960/2125] avg loss 0.000469958, throughput 6.01663K wps
[Epoch 27 Batch 990/2125] avg loss 0.000463939, throughput 6.02029K wps
[Epoch 27 Batch 1020/2125] avg loss 0.000385905, throughput 6.01826K wps
[Epoch 27 Batch 1050/2125] avg loss 0.000487574, throughput 6.00783K wps
[Epoch 27 Batch 1080/2125] avg loss 0.000541038, throughput 6.01245K wps
[Epoch 27 Batch 1110/2125] avg loss 0.00055638, throughput 6.00166K wps
[Epoch 27 Batch 1140/2125] avg loss 0.000545533, throughput 6.01434K wps
[Epoch 27 Batch 1170/2125] avg loss 0.000481674, throughput 6.00992K wps
[Epoch 27 Batch 1200/2125] avg loss 0.000431579, throughput 6.00735K wps
[Epoch 27 Batch 1230/2125] avg loss 0.000607412, throughput 6.00403K wps
[Epoch 27 Batch 1260/2125] avg loss 0.000784415, throughput 6.01528K wps
[Epoch 27 Batch 1290/2125] avg loss 0.000554818, throughput 6.00607K wps
[Epoch 27 Batch 1320/2125] avg loss 0.000605115, throughput 6.01556K wps
[Epoch 27 Batch 1350/2125] avg loss 0.000576746, throughput 6.01629K wps
[Epoch 27 Batch 1380/2125] avg loss 0.000426442, throughput 6.02052K wps
[Epoch 27 Batch 1410/2125] avg loss 0.000702706, throughput 6.00994K wps
[Epoch 27 Batch 1440/2125] avg loss 0.000532238, throughput 6.0166K wps
[Epoch 27 Batch 1470/2125] avg loss 0.00052942, throughput 6.00647K wps
[Epoch 27 Batch 1500/2125] avg loss 0.000467343, throughput 6.00978K wps
[Epoch 27 Batch 1530/2125] avg loss 0.000382823, throughput 6.01018K wps
[Epoch 27 Batch 1560/2125] avg loss 0.000724156, throughput 6.00453K wps
[Epoch 27 Batch 1590/2125] avg loss 0.000645474, throughput 6.00818K wps
[Epoch 27 Batch 1620/2125] avg loss 0.000527545, throughput 6.01169K wps
[Epoch 27 Batch 1650/2125] avg loss 0.000536952, throughput 6.0137K wps
[Epoch 27 Batch 1680/2125] avg loss 0.000929757, throughput 6.00587K wps
[Epoch 27 Batch 1710/2125] avg loss 0.000657407, throughput 6.01382K wps
[Epoch 27 Batch 1740/2125] avg loss 0.000450992, throughput 6.01089K wps
[Epoch 27 Batch 1770/2125] avg loss 0.000485031, throughput 6.00935K wps
[Epoch 27 Batch 1800/2125] avg loss 0.000591052, throughput 6.01575K wps
[Epoch 27 Batch 1830/2125] avg loss 0.000645612, throughput 6.01191K wps
[Epoch 27 Batch 1860/2125] avg loss 0.000509999, throughput 5.99955K wps
[Epoch 27 Batch 1890/2125] avg loss 0.000669564, throughput 6.00786K wps
[Epoch 27 Batch 1920/2125] avg loss 0.000480801, throughput 6.00878K wps
[Epoch 27 Batch 1950/2125] avg loss 0.000676231, throughput 6.01112K wps
[Epoch 27 Batch 1980/2125] avg loss 0.000671441, throughput 6.0153K wps
[Epoch 27 Batch 2010/2125] avg loss 0.000743092, throughput 6.01026K wps
[Epoch 27 Batch 2040/2125] avg loss 0.000684335, throughput 6.00854K wps
[Epoch 27 Batch 2070/2125] avg loss 0.000747929, throughput 6.02837K wps
[Epoch 27 Batch 2100/2125] avg loss 0.00054055, throughput 6.01561K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 27] train avg loss 0.000530265, test acc 0.9243, test avg loss 0.49879, throughput 6.01299K wps
[Epoch 28 Batch 30/2125] avg loss 0.000366069, throughput 6.15494K wps
[Epoch 28 Batch 60/2125] avg loss 0.000407429, throughput 6.01397K wps
[Epoch 28 Batch 90/2125] avg loss 0.000385355, throughput 6.01191K wps
[Epoch 28 Batch 120/2125] avg loss 0.000430377, throughput 6.01111K wps
[Epoch 28 Batch 150/2125] avg loss 0.000522123, throughput 6.01848K wps
[Epoch 28 Batch 180/2125] avg loss 0.000456576, throughput 6.00973K wps
[Epoch 28 Batch 210/2125] avg loss 0.000229669, throughput 6.00001K wps
[Epoch 28 Batch 240/2125] avg loss 0.000304334, throughput 6.00778K wps
[Epoch 28 Batch 270/2125] avg loss 0.000256109, throughput 6.01274K wps
[Epoch 28 Batch 300/2125] avg loss 0.00042944, throughput 6.02029K wps
[Epoch 28 Batch 330/2125] avg loss 0.000332265, throughput 5.98126K wps
[Epoch 28 Batch 360/2125] avg loss 0.000634776, throughput 6.0024K wps
[Epoch 28 Batch 390/2125] avg loss 0.000564547, throughput 6.0095K wps
[Epoch 28 Batch 420/2125] avg loss 0.000404276, throughput 6.01687K wps
[Epoch 28 Batch 450/2125] avg loss 0.000564746, throughput 6.01813K wps
[Epoch 28 Batch 480/2125] avg loss 0.000383654, throughput 6.01947K wps
[Epoch 28 Batch 510/2125] avg loss 0.000430749, throughput 6.02304K wps
[Epoch 28 Batch 540/2125] avg loss 0.000345829, throughput 6.00666K wps
[Epoch 28 Batch 570/2125] avg loss 0.000428243, throughput 6.01174K wps
[Epoch 28 Batch 600/2125] avg loss 0.000445451, throughput 6.01642K wps
[Epoch 28 Batch 630/2125] avg loss 0.000449991, throughput 6.02275K wps
[Epoch 28 Batch 660/2125] avg loss 0.000455503, throughput 6.02637K wps
[Epoch 28 Batch 690/2125] avg loss 0.000543967, throughput 6.01363K wps
[Epoch 28 Batch 720/2125] avg loss 0.000264862, throughput 6.01773K wps
[Epoch 28 Batch 750/2125] avg loss 0.000485244, throughput 6.02486K wps
[Epoch 28 Batch 780/2125] avg loss 0.000601875, throughput 6.0129K wps
[Epoch 28 Batch 810/2125] avg loss 0.000441945, throughput 6.0103K wps
[Epoch 28 Batch 840/2125] avg loss 0.000604503, throughput 6.01736K wps
[Epoch 28 Batch 870/2125] avg loss 0.000395081, throughput 6.00834K wps
[Epoch 28 Batch 900/2125] avg loss 0.000566211, throughput 6.0101K wps
[Epoch 28 Batch 930/2125] avg loss 0.000521209, throughput 6.01352K wps
[Epoch 28 Batch 960/2125] avg loss 0.000311402, throughput 6.01322K wps
[Epoch 28 Batch 990/2125] avg loss 0.000551264, throughput 6.00245K wps
[Epoch 28 Batch 1020/2125] avg loss 0.00062246, throughput 6.02113K wps
[Epoch 28 Batch 1050/2125] avg loss 0.000528235, throughput 6.01123K wps
[Epoch 28 Batch 1080/2125] avg loss 0.000739136, throughput 6.00634K wps
[Epoch 28 Batch 1110/2125] avg loss 0.000500869, throughput 6.01744K wps
[Epoch 28 Batch 1140/2125] avg loss 0.000364663, throughput 6.01212K wps
[Epoch 28 Batch 1170/2125] avg loss 0.000688496, throughput 6.00459K wps
[Epoch 28 Batch 1200/2125] avg loss 0.000749937, throughput 6.00607K wps
[Epoch 28 Batch 1230/2125] avg loss 0.000552835, throughput 6.00584K wps
[Epoch 28 Batch 1260/2125] avg loss 0.000647316, throughput 6.00688K wps
[Epoch 28 Batch 1290/2125] avg loss 0.000482144, throughput 6.0135K wps
[Epoch 28 Batch 1320/2125] avg loss 0.000462863, throughput 6.01106K wps
[Epoch 28 Batch 1350/2125] avg loss 0.000423301, throughput 6.00844K wps
[Epoch 28 Batch 1380/2125] avg loss 0.000607143, throughput 6.01972K wps
[Epoch 28 Batch 1410/2125] avg loss 0.00041306, throughput 6.00383K wps
[Epoch 28 Batch 1440/2125] avg loss 0.000360716, throughput 6.0107K wps
[Epoch 28 Batch 1470/2125] avg loss 0.000550084, throughput 6.01116K wps
[Epoch 28 Batch 1500/2125] avg loss 0.000787316, throughput 6.0092K wps
[Epoch 28 Batch 1530/2125] avg loss 0.000457375, throughput 6.00812K wps
[Epoch 28 Batch 1560/2125] avg loss 0.00051971, throughput 6.01985K wps
[Epoch 28 Batch 1590/2125] avg loss 0.000727272, throughput 6.01354K wps
[Epoch 28 Batch 1620/2125] avg loss 0.00034806, throughput 6.00767K wps
[Epoch 28 Batch 1650/2125] avg loss 0.000450711, throughput 6.01317K wps
[Epoch 28 Batch 1680/2125] avg loss 0.000466072, throughput 6.01621K wps
[Epoch 28 Batch 1710/2125] avg loss 0.000469118, throughput 6.01494K wps
[Epoch 28 Batch 1740/2125] avg loss 0.000503739, throughput 6.01315K wps
[Epoch 28 Batch 1770/2125] avg loss 0.00069716, throughput 6.02171K wps
[Epoch 28 Batch 1800/2125] avg loss 0.000682285, throughput 6.01848K wps
[Epoch 28 Batch 1830/2125] avg loss 0.00070891, throughput 5.99775K wps
[Epoch 28 Batch 1860/2125] avg loss 0.000424039, throughput 6.00816K wps
[Epoch 28 Batch 1890/2125] avg loss 0.000606046, throughput 6.01371K wps
[Epoch 28 Batch 1920/2125] avg loss 0.000459375, throughput 6.01008K wps
[Epoch 28 Batch 1950/2125] avg loss 0.000899292, throughput 6.01136K wps
[Epoch 28 Batch 1980/2125] avg loss 0.000775539, throughput 6.01848K wps
[Epoch 28 Batch 2010/2125] avg loss 0.000818529, throughput 6.02261K wps
[Epoch 28 Batch 2040/2125] avg loss 0.000575843, throughput 6.00668K wps
[Epoch 28 Batch 2070/2125] avg loss 0.000641248, throughput 6.02572K wps
[Epoch 28 Batch 2100/2125] avg loss 0.000580702, throughput 6.01901K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 28] train avg loss 0.000515155, test acc 0.9227, test avg loss 0.509851, throughput 6.01458K wps
[Epoch 29 Batch 30/2125] avg loss 0.000702477, throughput 6.14417K wps
[Epoch 29 Batch 60/2125] avg loss 0.000377888, throughput 6.02186K wps
[Epoch 29 Batch 90/2125] avg loss 0.00031659, throughput 6.01923K wps
[Epoch 29 Batch 120/2125] avg loss 0.000389192, throughput 6.0181K wps
[Epoch 29 Batch 150/2125] avg loss 0.000189584, throughput 6.01443K wps
[Epoch 29 Batch 180/2125] avg loss 0.000416641, throughput 6.01067K wps
[Epoch 29 Batch 210/2125] avg loss 0.000421578, throughput 6.00761K wps
[Epoch 29 Batch 240/2125] avg loss 0.000290926, throughput 6.01689K wps
[Epoch 29 Batch 270/2125] avg loss 0.000424312, throughput 6.0094K wps
[Epoch 29 Batch 300/2125] avg loss 0.000432992, throughput 6.00784K wps
[Epoch 29 Batch 330/2125] avg loss 0.000297893, throughput 6.01148K wps
[Epoch 29 Batch 360/2125] avg loss 0.000493928, throughput 6.01492K wps
[Epoch 29 Batch 390/2125] avg loss 0.000576036, throughput 6.00937K wps
[Epoch 29 Batch 420/2125] avg loss 0.000410303, throughput 6.01211K wps
[Epoch 29 Batch 450/2125] avg loss 0.00064364, throughput 6.00531K wps
[Epoch 29 Batch 480/2125] avg loss 0.000449344, throughput 6.01388K wps
[Epoch 29 Batch 510/2125] avg loss 0.000381135, throughput 6.01134K wps
[Epoch 29 Batch 540/2125] avg loss 0.000283846, throughput 6.009K wps
[Epoch 29 Batch 570/2125] avg loss 0.000423784, throughput 6.00562K wps
[Epoch 29 Batch 600/2125] avg loss 0.000399777, throughput 6.01982K wps
[Epoch 29 Batch 630/2125] avg loss 0.000476612, throughput 6.01221K wps
[Epoch 29 Batch 660/2125] avg loss 0.000445986, throughput 6.01523K wps
[Epoch 29 Batch 690/2125] avg loss 0.000380113, throughput 6.0062K wps
[Epoch 29 Batch 720/2125] avg loss 0.000559469, throughput 6.00535K wps
[Epoch 29 Batch 750/2125] avg loss 0.000629033, throughput 5.98085K wps
[Epoch 29 Batch 780/2125] avg loss 0.000320712, throughput 6.00275K wps
[Epoch 29 Batch 810/2125] avg loss 0.00062407, throughput 6.01145K wps
[Epoch 29 Batch 840/2125] avg loss 0.000516259, throughput 6.00645K wps
[Epoch 29 Batch 870/2125] avg loss 0.000465215, throughput 6.01289K wps
[Epoch 29 Batch 900/2125] avg loss 0.000518875, throughput 6.00634K wps
[Epoch 29 Batch 930/2125] avg loss 0.0003085, throughput 6.00805K wps
[Epoch 29 Batch 960/2125] avg loss 0.00039209, throughput 6.00787K wps
[Epoch 29 Batch 990/2125] avg loss 0.00048387, throughput 6.01435K wps
[Epoch 29 Batch 1020/2125] avg loss 0.000421762, throughput 6.00299K wps
[Epoch 29 Batch 1050/2125] avg loss 0.000393854, throughput 6.01008K wps
[Epoch 29 Batch 1080/2125] avg loss 0.000634149, throughput 6.00082K wps
[Epoch 29 Batch 1110/2125] avg loss 0.000542649, throughput 6.00014K wps
[Epoch 29 Batch 1140/2125] avg loss 0.000715564, throughput 5.99746K wps
[Epoch 29 Batch 1170/2125] avg loss 0.000470743, throughput 6.01035K wps
[Epoch 29 Batch 1200/2125] avg loss 0.000401443, throughput 6.00233K wps
[Epoch 29 Batch 1230/2125] avg loss 0.000679325, throughput 6.00608K wps
[Epoch 29 Batch 1260/2125] avg loss 0.000634383, throughput 5.99922K wps
[Epoch 29 Batch 1290/2125] avg loss 0.000421802, throughput 6.01458K wps
[Epoch 29 Batch 1320/2125] avg loss 0.000619327, throughput 6.0238K wps
[Epoch 29 Batch 1350/2125] avg loss 0.000445965, throughput 6.02322K wps
[Epoch 29 Batch 1380/2125] avg loss 0.000490787, throughput 6.02307K wps
[Epoch 29 Batch 1410/2125] avg loss 0.000434538, throughput 6.00824K wps
[Epoch 29 Batch 1440/2125] avg loss 0.000356366, throughput 6.00884K wps
[Epoch 29 Batch 1470/2125] avg loss 0.000667965, throughput 6.01395K wps
[Epoch 29 Batch 1500/2125] avg loss 0.000431863, throughput 6.00954K wps
[Epoch 29 Batch 1530/2125] avg loss 0.000465705, throughput 6.00982K wps
[Epoch 29 Batch 1560/2125] avg loss 0.000629518, throughput 6.01106K wps
[Epoch 29 Batch 1590/2125] avg loss 0.000586818, throughput 6.01488K wps
[Epoch 29 Batch 1620/2125] avg loss 0.000514784, throughput 6.01723K wps
[Epoch 29 Batch 1650/2125] avg loss 0.000559293, throughput 6.00859K wps
[Epoch 29 Batch 1680/2125] avg loss 0.000673843, throughput 6.00788K wps
[Epoch 29 Batch 1710/2125] avg loss 0.000392863, throughput 6.00982K wps
[Epoch 29 Batch 1740/2125] avg loss 0.000719019, throughput 6.01456K wps
[Epoch 29 Batch 1770/2125] avg loss 0.000822342, throughput 6.02062K wps
[Epoch 29 Batch 1800/2125] avg loss 0.000580247, throughput 6.00909K wps
[Epoch 29 Batch 1830/2125] avg loss 0.000606566, throughput 6.00562K wps
[Epoch 29 Batch 1860/2125] avg loss 0.000536771, throughput 6.00524K wps
[Epoch 29 Batch 1890/2125] avg loss 0.000398209, throughput 6.00745K wps
[Epoch 29 Batch 1920/2125] avg loss 0.000664524, throughput 6.01052K wps
[Epoch 29 Batch 1950/2125] avg loss 0.000418618, throughput 6.01263K wps
[Epoch 29 Batch 1980/2125] avg loss 0.00068462, throughput 6.02005K wps
[Epoch 29 Batch 2010/2125] avg loss 0.000648657, throughput 6.01766K wps
[Epoch 29 Batch 2040/2125] avg loss 0.000528578, throughput 6.02147K wps
[Epoch 29 Batch 2070/2125] avg loss 0.000655935, throughput 6.0254K wps
[Epoch 29 Batch 2100/2125] avg loss 0.000483738, throughput 6.00875K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 29] train avg loss 0.000500141, test acc 0.9242, test avg loss 0.518326, throughput 6.0127K wps
[Epoch 30 Batch 30/2125] avg loss 0.000356415, throughput 6.15403K wps
[Epoch 30 Batch 60/2125] avg loss 0.000371083, throughput 6.01991K wps
[Epoch 30 Batch 90/2125] avg loss 0.000370818, throughput 6.01319K wps
[Epoch 30 Batch 120/2125] avg loss 0.000355733, throughput 6.01732K wps
[Epoch 30 Batch 150/2125] avg loss 0.000396997, throughput 6.0196K wps
[Epoch 30 Batch 180/2125] avg loss 0.000341529, throughput 6.01874K wps
[Epoch 30 Batch 210/2125] avg loss 0.000523252, throughput 6.01502K wps
[Epoch 30 Batch 240/2125] avg loss 0.000491438, throughput 6.00093K wps
[Epoch 30 Batch 270/2125] avg loss 0.000489321, throughput 6.01601K wps
[Epoch 30 Batch 300/2125] avg loss 0.000402919, throughput 6.01321K wps
[Epoch 30 Batch 330/2125] avg loss 0.000553189, throughput 6.0167K wps
[Epoch 30 Batch 360/2125] avg loss 0.000390117, throughput 6.01016K wps
[Epoch 30 Batch 390/2125] avg loss 0.000332792, throughput 6.00325K wps
[Epoch 30 Batch 420/2125] avg loss 0.000324521, throughput 6.01811K wps
[Epoch 30 Batch 450/2125] avg loss 0.000348055, throughput 6.01301K wps
[Epoch 30 Batch 480/2125] avg loss 0.000405639, throughput 6.00947K wps
[Epoch 30 Batch 510/2125] avg loss 0.000256765, throughput 6.01544K wps
[Epoch 30 Batch 540/2125] avg loss 0.000377747, throughput 6.00713K wps
[Epoch 30 Batch 570/2125] avg loss 0.000290538, throughput 6.00883K wps
[Epoch 30 Batch 600/2125] avg loss 0.000310053, throughput 6.01411K wps
[Epoch 30 Batch 630/2125] avg loss 0.000544997, throughput 6.00878K wps
[Epoch 30 Batch 660/2125] avg loss 0.000514896, throughput 6.01354K wps
[Epoch 30 Batch 690/2125] avg loss 0.00048612, throughput 6.00956K wps
[Epoch 30 Batch 720/2125] avg loss 0.000258263, throughput 6.0151K wps
[Epoch 30 Batch 750/2125] avg loss 0.000424863, throughput 6.00981K wps
[Epoch 30 Batch 780/2125] avg loss 0.000412065, throughput 6.01299K wps
[Epoch 30 Batch 810/2125] avg loss 0.000444386, throughput 6.0022K wps
[Epoch 30 Batch 840/2125] avg loss 0.000371668, throughput 6.00156K wps
[Epoch 30 Batch 870/2125] avg loss 0.000465691, throughput 6.00748K wps
[Epoch 30 Batch 900/2125] avg loss 0.000385243, throughput 6.01326K wps
[Epoch 30 Batch 930/2125] avg loss 0.000410661, throughput 5.99525K wps
[Epoch 30 Batch 960/2125] avg loss 0.000339601, throughput 6.01523K wps
[Epoch 30 Batch 990/2125] avg loss 0.000390288, throughput 6.0114K wps
[Epoch 30 Batch 1020/2125] avg loss 0.000681509, throughput 6.01697K wps
[Epoch 30 Batch 1050/2125] avg loss 0.000574416, throughput 6.01091K wps
[Epoch 30 Batch 1080/2125] avg loss 0.000532651, throughput 6.01398K wps
[Epoch 30 Batch 1110/2125] avg loss 0.000562277, throughput 6.00834K wps
[Epoch 30 Batch 1140/2125] avg loss 0.000379846, throughput 6.01463K wps
[Epoch 30 Batch 1170/2125] avg loss 0.000429224, throughput 6.01346K wps
[Epoch 30 Batch 1200/2125] avg loss 0.000646746, throughput 6.01071K wps
[Epoch 30 Batch 1230/2125] avg loss 0.000494272, throughput 6.0123K wps
[Epoch 30 Batch 1260/2125] avg loss 0.000354805, throughput 6.00156K wps
[Epoch 30 Batch 1290/2125] avg loss 0.000503618, throughput 5.99895K wps
[Epoch 30 Batch 1320/2125] avg loss 0.00066009, throughput 6.00717K wps
[Epoch 30 Batch 1350/2125] avg loss 0.000764344, throughput 6.00553K wps
[Epoch 30 Batch 1380/2125] avg loss 0.000448199, throughput 5.92127K wps
[Epoch 30 Batch 1410/2125] avg loss 0.000375576, throughput 5.97365K wps
[Epoch 30 Batch 1440/2125] avg loss 0.000540009, throughput 6.00971K wps
[Epoch 30 Batch 1470/2125] avg loss 0.000624595, throughput 6.01058K wps
[Epoch 30 Batch 1500/2125] avg loss 0.000475863, throughput 6.01328K wps
[Epoch 30 Batch 1530/2125] avg loss 0.000588111, throughput 6.01036K wps
[Epoch 30 Batch 1560/2125] avg loss 0.000511088, throughput 6.02049K wps
[Epoch 30 Batch 1590/2125] avg loss 0.000625754, throughput 6.01743K wps
[Epoch 30 Batch 1620/2125] avg loss 0.000411607, throughput 6.01263K wps
[Epoch 30 Batch 1650/2125] avg loss 0.000534311, throughput 6.01062K wps
[Epoch 30 Batch 1680/2125] avg loss 0.000595546, throughput 6.00283K wps
[Epoch 30 Batch 1710/2125] avg loss 0.000638234, throughput 6.02661K wps
[Epoch 30 Batch 1740/2125] avg loss 0.000732639, throughput 6.02535K wps
[Epoch 30 Batch 1770/2125] avg loss 0.00044589, throughput 6.01569K wps
[Epoch 30 Batch 1800/2125] avg loss 0.000828809, throughput 6.02688K wps
[Epoch 30 Batch 1830/2125] avg loss 0.000619936, throughput 6.02079K wps
[Epoch 30 Batch 1860/2125] avg loss 0.000576655, throughput 6.02263K wps
[Epoch 30 Batch 1890/2125] avg loss 0.000461455, throughput 6.01885K wps
[Epoch 30 Batch 1920/2125] avg loss 0.000504393, throughput 6.013K wps
[Epoch 30 Batch 1950/2125] avg loss 0.000628527, throughput 6.01971K wps
[Epoch 30 Batch 1980/2125] avg loss 0.000288478, throughput 6.00607K wps
[Epoch 30 Batch 2010/2125] avg loss 0.000698254, throughput 6.01435K wps
[Epoch 30 Batch 2040/2125] avg loss 0.00056703, throughput 6.01252K wps
[Epoch 30 Batch 2070/2125] avg loss 0.000622598, throughput 6.00962K wps
[Epoch 30 Batch 2100/2125] avg loss 0.000501783, throughput 6.01235K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 30] train avg loss 0.000481479, test acc 0.9244, test avg loss 0.518721, throughput 6.0125K wps
[Epoch 31 Batch 30/2125] avg loss 0.000362389, throughput 6.15167K wps
[Epoch 31 Batch 60/2125] avg loss 0.000352008, throughput 6.0116K wps
[Epoch 31 Batch 90/2125] avg loss 0.000337979, throughput 6.00983K wps
[Epoch 31 Batch 120/2125] avg loss 0.000425332, throughput 6.02069K wps
[Epoch 31 Batch 150/2125] avg loss 0.000350883, throughput 6.00938K wps
[Epoch 31 Batch 180/2125] avg loss 0.000377834, throughput 6.01792K wps
[Epoch 31 Batch 210/2125] avg loss 0.00035726, throughput 6.00854K wps
[Epoch 31 Batch 240/2125] avg loss 0.000367208, throughput 6.00879K wps
[Epoch 31 Batch 270/2125] avg loss 0.000470727, throughput 6.0165K wps
[Epoch 31 Batch 300/2125] avg loss 0.000348604, throughput 6.01821K wps
[Epoch 31 Batch 330/2125] avg loss 0.00025595, throughput 6.02075K wps
[Epoch 31 Batch 360/2125] avg loss 0.000254262, throughput 6.01474K wps
[Epoch 31 Batch 390/2125] avg loss 0.000286131, throughput 6.01578K wps
[Epoch 31 Batch 420/2125] avg loss 0.000392422, throughput 6.01683K wps
[Epoch 31 Batch 450/2125] avg loss 0.000187274, throughput 5.99828K wps
[Epoch 31 Batch 480/2125] avg loss 0.000286414, throughput 6.01696K wps
[Epoch 31 Batch 510/2125] avg loss 0.000350092, throughput 6.00975K wps
[Epoch 31 Batch 540/2125] avg loss 0.000522796, throughput 6.02507K wps
[Epoch 31 Batch 570/2125] avg loss 0.000345318, throughput 6.03079K wps
[Epoch 31 Batch 600/2125] avg loss 0.000467052, throughput 6.02267K wps
[Epoch 31 Batch 630/2125] avg loss 0.000338258, throughput 6.0014K wps
[Epoch 31 Batch 660/2125] avg loss 0.000374241, throughput 6.00072K wps
[Epoch 31 Batch 690/2125] avg loss 0.000463416, throughput 6.00874K wps
[Epoch 31 Batch 720/2125] avg loss 0.000321377, throughput 6.00293K wps
[Epoch 31 Batch 750/2125] avg loss 0.000362169, throughput 6.00685K wps
[Epoch 31 Batch 780/2125] avg loss 0.000474795, throughput 6.0075K wps
[Epoch 31 Batch 810/2125] avg loss 0.000343329, throughput 6.00507K wps
[Epoch 31 Batch 840/2125] avg loss 0.000497696, throughput 6.00742K wps
[Epoch 31 Batch 870/2125] avg loss 0.000362789, throughput 6.01322K wps
[Epoch 31 Batch 900/2125] avg loss 0.000288264, throughput 6.01758K wps
[Epoch 31 Batch 930/2125] avg loss 0.000457206, throughput 6.009K wps
[Epoch 31 Batch 960/2125] avg loss 0.000418998, throughput 6.01014K wps
[Epoch 31 Batch 990/2125] avg loss 0.000428861, throughput 6.00977K wps
[Epoch 31 Batch 1020/2125] avg loss 0.000730558, throughput 6.01448K wps
[Epoch 31 Batch 1050/2125] avg loss 0.000480018, throughput 6.01456K wps
[Epoch 31 Batch 1080/2125] avg loss 0.000684216, throughput 6.01254K wps
[Epoch 31 Batch 1110/2125] avg loss 0.0005415, throughput 6.00734K wps
[Epoch 31 Batch 1140/2125] avg loss 0.000359297, throughput 6.00406K wps
[Epoch 31 Batch 1170/2125] avg loss 0.000654249, throughput 6.01325K wps
[Epoch 31 Batch 1200/2125] avg loss 0.000389445, throughput 6.01509K wps
[Epoch 31 Batch 1230/2125] avg loss 0.000444326, throughput 6.01147K wps
[Epoch 31 Batch 1260/2125] avg loss 0.000717634, throughput 6.01189K wps
[Epoch 31 Batch 1290/2125] avg loss 0.000491807, throughput 6.01205K wps
[Epoch 31 Batch 1320/2125] avg loss 0.000758059, throughput 6.00724K wps
[Epoch 31 Batch 1350/2125] avg loss 0.000599043, throughput 6.0095K wps
[Epoch 31 Batch 1380/2125] avg loss 0.000687403, throughput 6.0072K wps
[Epoch 31 Batch 1410/2125] avg loss 0.00044189, throughput 6.01253K wps
[Epoch 31 Batch 1440/2125] avg loss 0.000530878, throughput 6.0052K wps
[Epoch 31 Batch 1470/2125] avg loss 0.000692653, throughput 6.00644K wps
[Epoch 31 Batch 1500/2125] avg loss 0.000715063, throughput 6.00975K wps
[Epoch 31 Batch 1530/2125] avg loss 0.000657082, throughput 6.01063K wps
[Epoch 31 Batch 1560/2125] avg loss 0.000505528, throughput 6.01079K wps
[Epoch 31 Batch 1590/2125] avg loss 0.000465958, throughput 6.01204K wps
[Epoch 31 Batch 1620/2125] avg loss 0.000517252, throughput 5.99714K wps
[Epoch 31 Batch 1650/2125] avg loss 0.00049493, throughput 6.00653K wps
[Epoch 31 Batch 1680/2125] avg loss 0.000381782, throughput 6.01303K wps
[Epoch 31 Batch 1710/2125] avg loss 0.000578204, throughput 6.01877K wps
[Epoch 31 Batch 1740/2125] avg loss 0.00040305, throughput 6.01179K wps
[Epoch 31 Batch 1770/2125] avg loss 0.000561021, throughput 6.0144K wps
[Epoch 31 Batch 1800/2125] avg loss 0.000477816, throughput 6.01508K wps
[Epoch 31 Batch 1830/2125] avg loss 0.000599277, throughput 6.01371K wps
[Epoch 31 Batch 1860/2125] avg loss 0.000371665, throughput 6.00774K wps
[Epoch 31 Batch 1890/2125] avg loss 0.000465674, throughput 6.0033K wps
[Epoch 31 Batch 1920/2125] avg loss 0.000464639, throughput 6.00641K wps
[Epoch 31 Batch 1950/2125] avg loss 0.000449402, throughput 6.00035K wps
[Epoch 31 Batch 1980/2125] avg loss 0.000637846, throughput 6.00119K wps
[Epoch 31 Batch 2010/2125] avg loss 0.000477015, throughput 6.01225K wps
[Epoch 31 Batch 2040/2125] avg loss 0.000592845, throughput 6.01296K wps
[Epoch 31 Batch 2070/2125] avg loss 0.000457373, throughput 6.01479K wps
[Epoch 31 Batch 2100/2125] avg loss 0.000477741, throughput 6.01119K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 31] train avg loss 0.000460668, test acc 0.9227, test avg loss 0.530348, throughput 6.01299K wps
[Epoch 32 Batch 30/2125] avg loss 0.000315461, throughput 6.14208K wps
[Epoch 32 Batch 60/2125] avg loss 0.000420257, throughput 6.01243K wps
[Epoch 32 Batch 90/2125] avg loss 0.000371659, throughput 6.00979K wps
[Epoch 32 Batch 120/2125] avg loss 0.000365768, throughput 6.00738K wps
[Epoch 32 Batch 150/2125] avg loss 0.000601856, throughput 6.01161K wps
[Epoch 32 Batch 180/2125] avg loss 0.000356195, throughput 6.01251K wps
[Epoch 32 Batch 210/2125] avg loss 0.00027656, throughput 6.00767K wps
[Epoch 32 Batch 240/2125] avg loss 0.000442587, throughput 6.0093K wps
[Epoch 32 Batch 270/2125] avg loss 0.000385117, throughput 6.01416K wps
[Epoch 32 Batch 300/2125] avg loss 0.000317716, throughput 6.00813K wps
[Epoch 32 Batch 330/2125] avg loss 0.000253587, throughput 6.00531K wps
[Epoch 32 Batch 360/2125] avg loss 0.000491862, throughput 5.99409K wps
[Epoch 32 Batch 390/2125] avg loss 0.000448473, throughput 6.0039K wps
[Epoch 32 Batch 420/2125] avg loss 0.000476083, throughput 5.99863K wps
[Epoch 32 Batch 450/2125] avg loss 0.000344106, throughput 6.01037K wps
[Epoch 32 Batch 480/2125] avg loss 0.00034904, throughput 6.00096K wps
[Epoch 32 Batch 510/2125] avg loss 0.00037152, throughput 6.00283K wps
[Epoch 32 Batch 540/2125] avg loss 0.000397034, throughput 6.01303K wps
[Epoch 32 Batch 570/2125] avg loss 0.000479475, throughput 6.0069K wps
[Epoch 32 Batch 600/2125] avg loss 0.000349435, throughput 6.00474K wps
[Epoch 32 Batch 630/2125] avg loss 0.000455354, throughput 6.00359K wps
[Epoch 32 Batch 660/2125] avg loss 0.000474835, throughput 6.00099K wps
[Epoch 32 Batch 690/2125] avg loss 0.000377094, throughput 6.00467K wps
[Epoch 32 Batch 720/2125] avg loss 0.00059228, throughput 6.01922K wps
[Epoch 32 Batch 750/2125] avg loss 0.000293505, throughput 6.0114K wps
[Epoch 32 Batch 780/2125] avg loss 0.00037593, throughput 6.00546K wps
[Epoch 32 Batch 810/2125] avg loss 0.000293208, throughput 6.00876K wps
[Epoch 32 Batch 840/2125] avg loss 0.000504456, throughput 6.01185K wps
[Epoch 32 Batch 870/2125] avg loss 0.000427769, throughput 6.0045K wps
[Epoch 32 Batch 900/2125] avg loss 0.000459941, throughput 6.00652K wps
[Epoch 32 Batch 930/2125] avg loss 0.00067176, throughput 6.00898K wps
[Epoch 32 Batch 960/2125] avg loss 0.000313967, throughput 6.00483K wps
[Epoch 32 Batch 990/2125] avg loss 0.000297065, throughput 6.01307K wps
[Epoch 32 Batch 1020/2125] avg loss 0.000365171, throughput 6.00961K wps
[Epoch 32 Batch 1050/2125] avg loss 0.000415711, throughput 6.01934K wps
[Epoch 32 Batch 1080/2125] avg loss 0.00041938, throughput 6.01035K wps
[Epoch 32 Batch 1110/2125] avg loss 0.00047143, throughput 6.00995K wps
[Epoch 32 Batch 1140/2125] avg loss 0.000352302, throughput 6.00297K wps
[Epoch 32 Batch 1170/2125] avg loss 0.000611022, throughput 6.01918K wps
[Epoch 32 Batch 1200/2125] avg loss 0.000495754, throughput 6.01137K wps
[Epoch 32 Batch 1230/2125] avg loss 0.000622679, throughput 6.01938K wps
[Epoch 32 Batch 1260/2125] avg loss 0.00041043, throughput 6.01306K wps
[Epoch 32 Batch 1290/2125] avg loss 0.000556059, throughput 6.01233K wps
[Epoch 32 Batch 1320/2125] avg loss 0.000342682, throughput 6.01158K wps
[Epoch 32 Batch 1350/2125] avg loss 0.000328941, throughput 6.00972K wps
[Epoch 32 Batch 1380/2125] avg loss 0.000570741, throughput 6.01439K wps
[Epoch 32 Batch 1410/2125] avg loss 0.000560227, throughput 6.00521K wps
[Epoch 32 Batch 1440/2125] avg loss 0.00051843, throughput 6.00119K wps
[Epoch 32 Batch 1470/2125] avg loss 0.000482642, throughput 6.008K wps
[Epoch 32 Batch 1500/2125] avg loss 0.000464309, throughput 6.00143K wps
[Epoch 32 Batch 1530/2125] avg loss 0.000765149, throughput 6.00341K wps
[Epoch 32 Batch 1560/2125] avg loss 0.00061166, throughput 6.0063K wps
[Epoch 32 Batch 1590/2125] avg loss 0.000573475, throughput 6.00442K wps
[Epoch 32 Batch 1620/2125] avg loss 0.000625128, throughput 6.0146K wps
[Epoch 32 Batch 1650/2125] avg loss 0.00033594, throughput 6.01564K wps
[Epoch 32 Batch 1680/2125] avg loss 0.000554814, throughput 6.01771K wps
[Epoch 32 Batch 1710/2125] avg loss 0.000486292, throughput 6.0198K wps
[Epoch 32 Batch 1740/2125] avg loss 0.000467236, throughput 6.01841K wps
[Epoch 32 Batch 1770/2125] avg loss 0.000405691, throughput 6.02525K wps
[Epoch 32 Batch 1800/2125] avg loss 0.000474202, throughput 6.01143K wps
[Epoch 32 Batch 1830/2125] avg loss 0.000347155, throughput 6.00746K wps
[Epoch 32 Batch 1860/2125] avg loss 0.000344121, throughput 6.00849K wps
[Epoch 32 Batch 1890/2125] avg loss 0.000776041, throughput 6.01666K wps
[Epoch 32 Batch 1920/2125] avg loss 0.000403563, throughput 6.03099K wps
[Epoch 32 Batch 1950/2125] avg loss 0.000304814, throughput 6.02229K wps
[Epoch 32 Batch 1980/2125] avg loss 0.000477084, throughput 6.01586K wps
[Epoch 32 Batch 2010/2125] avg loss 0.000531181, throughput 6.00838K wps
[Epoch 32 Batch 2040/2125] avg loss 0.000606539, throughput 6.00742K wps
[Epoch 32 Batch 2070/2125] avg loss 0.000644124, throughput 6.00949K wps
[Epoch 32 Batch 2100/2125] avg loss 0.000497985, throughput 6.0118K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 32] train avg loss 0.000455339, test acc 0.9224, test avg loss 0.531319, throughput 6.01197K wps
[Epoch 33 Batch 30/2125] avg loss 0.000339217, throughput 6.14592K wps
[Epoch 33 Batch 60/2125] avg loss 0.000276444, throughput 6.01294K wps
[Epoch 33 Batch 90/2125] avg loss 0.00024325, throughput 6.01782K wps
[Epoch 33 Batch 120/2125] avg loss 0.000496051, throughput 6.01003K wps
[Epoch 33 Batch 150/2125] avg loss 0.000330991, throughput 6.00935K wps
[Epoch 33 Batch 180/2125] avg loss 0.000403032, throughput 6.01272K wps
[Epoch 33 Batch 210/2125] avg loss 0.000593048, throughput 5.97958K wps
[Epoch 33 Batch 240/2125] avg loss 0.000236319, throughput 5.98173K wps
[Epoch 33 Batch 270/2125] avg loss 0.000523961, throughput 6.0275K wps
[Epoch 33 Batch 300/2125] avg loss 0.000491843, throughput 6.0238K wps
[Epoch 33 Batch 330/2125] avg loss 0.000268563, throughput 6.01485K wps
[Epoch 33 Batch 360/2125] avg loss 0.000398306, throughput 6.0239K wps
[Epoch 33 Batch 390/2125] avg loss 0.000315339, throughput 6.0212K wps
[Epoch 33 Batch 420/2125] avg loss 0.000276616, throughput 6.02019K wps
[Epoch 33 Batch 450/2125] avg loss 0.000383832, throughput 6.02096K wps
[Epoch 33 Batch 480/2125] avg loss 0.000383325, throughput 6.01891K wps
[Epoch 33 Batch 510/2125] avg loss 0.000324398, throughput 6.02091K wps
[Epoch 33 Batch 540/2125] avg loss 0.000394575, throughput 6.02105K wps
[Epoch 33 Batch 570/2125] avg loss 0.000346895, throughput 6.01563K wps
[Epoch 33 Batch 600/2125] avg loss 0.000450809, throughput 6.0142K wps
[Epoch 33 Batch 630/2125] avg loss 0.000388068, throughput 6.01882K wps
[Epoch 33 Batch 660/2125] avg loss 0.000271728, throughput 6.01245K wps
[Epoch 33 Batch 690/2125] avg loss 0.000352294, throughput 6.02324K wps
[Epoch 33 Batch 720/2125] avg loss 0.000479806, throughput 6.02321K wps
[Epoch 33 Batch 750/2125] avg loss 0.000335624, throughput 6.0245K wps
[Epoch 33 Batch 780/2125] avg loss 0.000537483, throughput 6.02095K wps
[Epoch 33 Batch 810/2125] avg loss 0.000365911, throughput 6.02568K wps
[Epoch 33 Batch 840/2125] avg loss 0.000519935, throughput 6.01974K wps
[Epoch 33 Batch 870/2125] avg loss 0.000286048, throughput 6.01502K wps
[Epoch 33 Batch 900/2125] avg loss 0.000381509, throughput 6.01822K wps
[Epoch 33 Batch 930/2125] avg loss 0.000273712, throughput 6.01085K wps
[Epoch 33 Batch 960/2125] avg loss 0.000601466, throughput 6.00122K wps
[Epoch 33 Batch 990/2125] avg loss 0.000574877, throughput 6.00946K wps
[Epoch 33 Batch 1020/2125] avg loss 0.000377487, throughput 6.01226K wps
[Epoch 33 Batch 1050/2125] avg loss 0.00034413, throughput 6.02165K wps
[Epoch 33 Batch 1080/2125] avg loss 0.000353308, throughput 6.03419K wps
[Epoch 33 Batch 1110/2125] avg loss 0.000520749, throughput 6.0309K wps
[Epoch 33 Batch 1140/2125] avg loss 0.000370962, throughput 6.02278K wps
[Epoch 33 Batch 1170/2125] avg loss 0.000698476, throughput 6.01968K wps
[Epoch 33 Batch 1200/2125] avg loss 0.000440836, throughput 6.01694K wps
[Epoch 33 Batch 1230/2125] avg loss 0.000619661, throughput 6.01216K wps
[Epoch 33 Batch 1260/2125] avg loss 0.000446167, throughput 6.00798K wps
[Epoch 33 Batch 1290/2125] avg loss 0.000363071, throughput 6.02351K wps
[Epoch 33 Batch 1320/2125] avg loss 0.000352961, throughput 6.02714K wps
[Epoch 33 Batch 1350/2125] avg loss 0.000456464, throughput 6.02626K wps
[Epoch 33 Batch 1380/2125] avg loss 0.000761554, throughput 6.01566K wps
[Epoch 33 Batch 1410/2125] avg loss 0.000603379, throughput 6.0247K wps
[Epoch 33 Batch 1440/2125] avg loss 0.000504462, throughput 6.01599K wps
[Epoch 33 Batch 1470/2125] avg loss 0.000555032, throughput 6.02462K wps
[Epoch 33 Batch 1500/2125] avg loss 0.000320491, throughput 6.0261K wps
[Epoch 33 Batch 1530/2125] avg loss 0.000678067, throughput 6.01213K wps
[Epoch 33 Batch 1560/2125] avg loss 0.000462069, throughput 6.0099K wps
[Epoch 33 Batch 1590/2125] avg loss 0.000488282, throughput 6.00498K wps
[Epoch 33 Batch 1620/2125] avg loss 0.00063372, throughput 6.00971K wps
[Epoch 33 Batch 1650/2125] avg loss 0.000371839, throughput 6.02371K wps
[Epoch 33 Batch 1680/2125] avg loss 0.000529249, throughput 6.0175K wps
[Epoch 33 Batch 1710/2125] avg loss 0.000427529, throughput 6.01412K wps
[Epoch 33 Batch 1740/2125] avg loss 0.000349863, throughput 6.02456K wps
[Epoch 33 Batch 1770/2125] avg loss 0.000774648, throughput 6.01804K wps
[Epoch 33 Batch 1800/2125] avg loss 0.000433045, throughput 6.02029K wps
[Epoch 33 Batch 1830/2125] avg loss 0.000451871, throughput 6.01334K wps
[Epoch 33 Batch 1860/2125] avg loss 0.00040038, throughput 6.01853K wps
[Epoch 33 Batch 1890/2125] avg loss 0.000471178, throughput 6.02152K wps
[Epoch 33 Batch 1920/2125] avg loss 0.000411247, throughput 6.01808K wps
[Epoch 33 Batch 1950/2125] avg loss 0.000676557, throughput 6.0163K wps
[Epoch 33 Batch 1980/2125] avg loss 0.000540108, throughput 6.01765K wps
[Epoch 33 Batch 2010/2125] avg loss 0.000958447, throughput 6.00839K wps
[Epoch 33 Batch 2040/2125] avg loss 0.000273879, throughput 6.02825K wps
[Epoch 33 Batch 2070/2125] avg loss 0.000465936, throughput 6.01708K wps
[Epoch 33 Batch 2100/2125] avg loss 0.000535896, throughput 6.01152K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 33] train avg loss 0.000446276, test acc 0.9236, test avg loss 0.543348, throughput 6.01878K wps
[Epoch 34 Batch 30/2125] avg loss 0.000245847, throughput 6.14236K wps
[Epoch 34 Batch 60/2125] avg loss 0.000386291, throughput 6.00801K wps
[Epoch 34 Batch 90/2125] avg loss 0.000358918, throughput 6.01257K wps
[Epoch 34 Batch 120/2125] avg loss 0.000261025, throughput 6.01307K wps
[Epoch 34 Batch 150/2125] avg loss 0.000341013, throughput 6.02525K wps
[Epoch 34 Batch 180/2125] avg loss 0.000428526, throughput 6.01727K wps
[Epoch 34 Batch 210/2125] avg loss 0.000277111, throughput 6.02449K wps
[Epoch 34 Batch 240/2125] avg loss 0.000376086, throughput 6.02183K wps
[Epoch 34 Batch 270/2125] avg loss 0.000399204, throughput 6.00685K wps
[Epoch 34 Batch 300/2125] avg loss 0.000399553, throughput 6.01793K wps
[Epoch 34 Batch 330/2125] avg loss 0.000486479, throughput 6.00937K wps
[Epoch 34 Batch 360/2125] avg loss 0.000432412, throughput 6.02787K wps
[Epoch 34 Batch 390/2125] avg loss 0.000329625, throughput 6.01038K wps
[Epoch 34 Batch 420/2125] avg loss 0.000459426, throughput 6.01146K wps
[Epoch 34 Batch 450/2125] avg loss 0.000365365, throughput 6.02269K wps
[Epoch 34 Batch 480/2125] avg loss 0.000247011, throughput 6.01543K wps
[Epoch 34 Batch 510/2125] avg loss 0.000282136, throughput 6.01808K wps
[Epoch 34 Batch 540/2125] avg loss 0.000431947, throughput 6.00954K wps
[Epoch 34 Batch 570/2125] avg loss 0.000425648, throughput 6.01338K wps
[Epoch 34 Batch 600/2125] avg loss 0.000371698, throughput 6.00493K wps
[Epoch 34 Batch 630/2125] avg loss 0.000311701, throughput 6.01086K wps
[Epoch 34 Batch 660/2125] avg loss 0.000382754, throughput 6.02237K wps
[Epoch 34 Batch 690/2125] avg loss 0.000288856, throughput 6.00504K wps
[Epoch 34 Batch 720/2125] avg loss 0.000362773, throughput 6.01216K wps
[Epoch 34 Batch 750/2125] avg loss 0.000443382, throughput 6.02702K wps
[Epoch 34 Batch 780/2125] avg loss 0.000433283, throughput 6.01722K wps
[Epoch 34 Batch 810/2125] avg loss 0.00031031, throughput 6.02368K wps
[Epoch 34 Batch 840/2125] avg loss 0.000298652, throughput 6.01884K wps
[Epoch 34 Batch 870/2125] avg loss 0.000349797, throughput 6.01942K wps
[Epoch 34 Batch 900/2125] avg loss 0.000419927, throughput 5.99966K wps
[Epoch 34 Batch 930/2125] avg loss 0.000508522, throughput 6.002K wps
[Epoch 34 Batch 960/2125] avg loss 0.000431251, throughput 5.99822K wps
[Epoch 34 Batch 990/2125] avg loss 0.000469011, throughput 5.99634K wps
[Epoch 34 Batch 1020/2125] avg loss 0.00047503, throughput 6.01495K wps
[Epoch 34 Batch 1050/2125] avg loss 0.000541245, throughput 6.02737K wps
[Epoch 34 Batch 1080/2125] avg loss 0.000496673, throughput 6.01166K wps
[Epoch 34 Batch 1110/2125] avg loss 0.000223538, throughput 6.00799K wps
[Epoch 34 Batch 1140/2125] avg loss 0.000543796, throughput 6.01205K wps
[Epoch 34 Batch 1170/2125] avg loss 0.000384565, throughput 6.01684K wps
[Epoch 34 Batch 1200/2125] avg loss 0.000612296, throughput 6.01397K wps
[Epoch 34 Batch 1230/2125] avg loss 0.000388527, throughput 6.01946K wps
[Epoch 34 Batch 1260/2125] avg loss 0.000443104, throughput 6.02076K wps
[Epoch 34 Batch 1290/2125] avg loss 0.000411106, throughput 6.00818K wps
[Epoch 34 Batch 1320/2125] avg loss 0.000648135, throughput 6.01727K wps
[Epoch 34 Batch 1350/2125] avg loss 0.000295454, throughput 6.02153K wps
[Epoch 34 Batch 1380/2125] avg loss 0.000297176, throughput 6.01951K wps
[Epoch 34 Batch 1410/2125] avg loss 0.000604023, throughput 6.02712K wps
[Epoch 34 Batch 1440/2125] avg loss 0.000495838, throughput 6.01826K wps
[Epoch 34 Batch 1470/2125] avg loss 0.000551604, throughput 6.02606K wps
[Epoch 34 Batch 1500/2125] avg loss 0.000620532, throughput 6.01739K wps
[Epoch 34 Batch 1530/2125] avg loss 0.000493108, throughput 6.01717K wps
[Epoch 34 Batch 1560/2125] avg loss 0.00056732, throughput 6.02309K wps
[Epoch 34 Batch 1590/2125] avg loss 0.000397159, throughput 6.02555K wps
[Epoch 34 Batch 1620/2125] avg loss 0.000388599, throughput 6.01987K wps
[Epoch 34 Batch 1650/2125] avg loss 0.0005047, throughput 6.01752K wps
[Epoch 34 Batch 1680/2125] avg loss 0.000421643, throughput 6.01113K wps
[Epoch 34 Batch 1710/2125] avg loss 0.000346231, throughput 6.02143K wps
[Epoch 34 Batch 1740/2125] avg loss 0.000599997, throughput 6.02336K wps
[Epoch 34 Batch 1770/2125] avg loss 0.000833491, throughput 6.02374K wps
[Epoch 34 Batch 1800/2125] avg loss 0.000456612, throughput 6.02483K wps
[Epoch 34 Batch 1830/2125] avg loss 0.000589963, throughput 6.02561K wps
[Epoch 34 Batch 1860/2125] avg loss 0.000466622, throughput 6.0237K wps
[Epoch 34 Batch 1890/2125] avg loss 0.000472787, throughput 6.03435K wps
[Epoch 34 Batch 1920/2125] avg loss 0.000458897, throughput 6.01239K wps
[Epoch 34 Batch 1950/2125] avg loss 0.000345074, throughput 6.01532K wps
[Epoch 34 Batch 1980/2125] avg loss 0.000349887, throughput 6.01774K wps
[Epoch 34 Batch 2010/2125] avg loss 0.000357488, throughput 6.01372K wps
[Epoch 34 Batch 2040/2125] avg loss 0.000521795, throughput 6.01721K wps
[Epoch 34 Batch 2070/2125] avg loss 0.000494484, throughput 6.01995K wps
[Epoch 34 Batch 2100/2125] avg loss 0.000489118, throughput 6.01861K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 34] train avg loss 0.000428381, test acc 0.9233, test avg loss 0.552511, throughput 6.01841K wps
[Epoch 35 Batch 30/2125] avg loss 0.000217066, throughput 6.156K wps
[Epoch 35 Batch 60/2125] avg loss 0.00025674, throughput 6.02571K wps
[Epoch 35 Batch 90/2125] avg loss 0.000337513, throughput 6.02136K wps
[Epoch 35 Batch 120/2125] avg loss 0.000256226, throughput 6.0113K wps
[Epoch 35 Batch 150/2125] avg loss 0.000318267, throughput 6.01141K wps
[Epoch 35 Batch 180/2125] avg loss 0.00033708, throughput 6.01791K wps
[Epoch 35 Batch 210/2125] avg loss 0.000292609, throughput 6.02126K wps
[Epoch 35 Batch 240/2125] avg loss 0.000206058, throughput 6.02768K wps
[Epoch 35 Batch 270/2125] avg loss 0.000522004, throughput 6.0182K wps
[Epoch 35 Batch 300/2125] avg loss 0.00036419, throughput 6.01459K wps
[Epoch 35 Batch 330/2125] avg loss 0.000295741, throughput 6.0137K wps
[Epoch 35 Batch 360/2125] avg loss 0.000305105, throughput 6.01901K wps
[Epoch 35 Batch 390/2125] avg loss 0.000347714, throughput 6.01632K wps
[Epoch 35 Batch 420/2125] avg loss 0.000202683, throughput 6.01231K wps
[Epoch 35 Batch 450/2125] avg loss 0.000461207, throughput 6.01767K wps
[Epoch 35 Batch 480/2125] avg loss 0.000455959, throughput 6.0208K wps
[Epoch 35 Batch 510/2125] avg loss 0.000341094, throughput 6.01249K wps
[Epoch 35 Batch 540/2125] avg loss 0.000376393, throughput 6.01781K wps
[Epoch 35 Batch 570/2125] avg loss 0.000181017, throughput 6.01273K wps
[Epoch 35 Batch 600/2125] avg loss 0.000376486, throughput 6.01492K wps
[Epoch 35 Batch 630/2125] avg loss 0.000354988, throughput 6.00933K wps
[Epoch 35 Batch 660/2125] avg loss 0.000397301, throughput 6.00761K wps
[Epoch 35 Batch 690/2125] avg loss 0.000622583, throughput 6.01442K wps
[Epoch 35 Batch 720/2125] avg loss 0.000343117, throughput 6.01748K wps
[Epoch 35 Batch 750/2125] avg loss 0.000281876, throughput 6.02376K wps
[Epoch 35 Batch 780/2125] avg loss 0.000333863, throughput 6.02871K wps
[Epoch 35 Batch 810/2125] avg loss 0.000431623, throughput 6.01274K wps
[Epoch 35 Batch 840/2125] avg loss 0.000272843, throughput 6.01604K wps
[Epoch 35 Batch 870/2125] avg loss 0.000344887, throughput 6.00984K wps
[Epoch 35 Batch 900/2125] avg loss 0.000312015, throughput 6.01054K wps
[Epoch 35 Batch 930/2125] avg loss 0.000318909, throughput 6.00843K wps
[Epoch 35 Batch 960/2125] avg loss 0.00033601, throughput 6.01199K wps
[Epoch 35 Batch 990/2125] avg loss 0.000387683, throughput 6.01514K wps
[Epoch 35 Batch 1020/2125] avg loss 0.000415479, throughput 6.02305K wps
[Epoch 35 Batch 1050/2125] avg loss 0.000372766, throughput 6.02212K wps
[Epoch 35 Batch 1080/2125] avg loss 0.000511803, throughput 6.01708K wps
[Epoch 35 Batch 1110/2125] avg loss 0.000338046, throughput 6.01016K wps
[Epoch 35 Batch 1140/2125] avg loss 0.000536972, throughput 6.02086K wps
[Epoch 35 Batch 1170/2125] avg loss 0.000328107, throughput 6.01831K wps
[Epoch 35 Batch 1200/2125] avg loss 0.000532036, throughput 6.01582K wps
[Epoch 35 Batch 1230/2125] avg loss 0.000605506, throughput 5.99858K wps
[Epoch 35 Batch 1260/2125] avg loss 0.000404814, throughput 5.988K wps
[Epoch 35 Batch 1290/2125] avg loss 0.00043888, throughput 5.98089K wps
[Epoch 35 Batch 1320/2125] avg loss 0.000364809, throughput 6.00429K wps
[Epoch 35 Batch 1350/2125] avg loss 0.000630035, throughput 6.0029K wps
[Epoch 35 Batch 1380/2125] avg loss 0.000339437, throughput 6.01141K wps
[Epoch 35 Batch 1410/2125] avg loss 0.00030224, throughput 6.01267K wps
[Epoch 35 Batch 1440/2125] avg loss 0.000473559, throughput 6.00925K wps
[Epoch 35 Batch 1470/2125] avg loss 0.000282134, throughput 6.00993K wps
[Epoch 35 Batch 1500/2125] avg loss 0.000363023, throughput 6.01266K wps
[Epoch 35 Batch 1530/2125] avg loss 0.00048706, throughput 6.02033K wps
[Epoch 35 Batch 1560/2125] avg loss 0.000509265, throughput 6.0103K wps
[Epoch 35 Batch 1590/2125] avg loss 0.000277534, throughput 6.01053K wps
[Epoch 35 Batch 1620/2125] avg loss 0.000406323, throughput 6.00943K wps
[Epoch 35 Batch 1650/2125] avg loss 0.000374591, throughput 6.01399K wps
[Epoch 35 Batch 1680/2125] avg loss 0.00072605, throughput 6.00935K wps
[Epoch 35 Batch 1710/2125] avg loss 0.000399042, throughput 6.0147K wps
[Epoch 35 Batch 1740/2125] avg loss 0.000633206, throughput 6.0111K wps
[Epoch 35 Batch 1770/2125] avg loss 0.000409386, throughput 6.01484K wps
[Epoch 35 Batch 1800/2125] avg loss 0.000255314, throughput 6.00159K wps
[Epoch 35 Batch 1830/2125] avg loss 0.000391296, throughput 6.00876K wps
[Epoch 35 Batch 1860/2125] avg loss 0.000431775, throughput 6.00136K wps
[Epoch 35 Batch 1890/2125] avg loss 0.000736677, throughput 6.00153K wps
[Epoch 35 Batch 1920/2125] avg loss 0.000389928, throughput 6.00411K wps
[Epoch 35 Batch 1950/2125] avg loss 0.000355664, throughput 6.01437K wps
[Epoch 35 Batch 1980/2125] avg loss 0.00058352, throughput 6.0216K wps
[Epoch 35 Batch 2010/2125] avg loss 0.000519954, throughput 6.00886K wps
[Epoch 35 Batch 2040/2125] avg loss 0.000554936, throughput 6.00879K wps
[Epoch 35 Batch 2070/2125] avg loss 0.000296095, throughput 6.0043K wps
[Epoch 35 Batch 2100/2125] avg loss 0.00052905, throughput 6.00895K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 35] train avg loss 0.000398545, test acc 0.9255, test avg loss 0.551359, throughput 6.01459K wps
[Epoch 36 Batch 30/2125] avg loss 0.0002725, throughput 6.14663K wps
[Epoch 36 Batch 60/2125] avg loss 0.000233631, throughput 6.00809K wps
[Epoch 36 Batch 90/2125] avg loss 0.000309849, throughput 6.01011K wps
[Epoch 36 Batch 120/2125] avg loss 0.000346157, throughput 6.00217K wps
[Epoch 36 Batch 150/2125] avg loss 0.000255922, throughput 6.01548K wps
[Epoch 36 Batch 180/2125] avg loss 0.000360388, throughput 6.00143K wps
[Epoch 36 Batch 210/2125] avg loss 0.00033246, throughput 6.01271K wps
[Epoch 36 Batch 240/2125] avg loss 0.000273495, throughput 6.02099K wps
[Epoch 36 Batch 270/2125] avg loss 0.000304999, throughput 6.00707K wps
[Epoch 36 Batch 300/2125] avg loss 0.000368881, throughput 6.00473K wps
[Epoch 36 Batch 330/2125] avg loss 0.00032701, throughput 6.00288K wps
[Epoch 36 Batch 360/2125] avg loss 0.00022502, throughput 6.00661K wps
[Epoch 36 Batch 390/2125] avg loss 0.000438704, throughput 6.01047K wps
[Epoch 36 Batch 420/2125] avg loss 0.000353101, throughput 6.00591K wps
[Epoch 36 Batch 450/2125] avg loss 0.000468635, throughput 6.01899K wps
[Epoch 36 Batch 480/2125] avg loss 0.000324474, throughput 6.01271K wps
[Epoch 36 Batch 510/2125] avg loss 0.000258731, throughput 6.00424K wps
[Epoch 36 Batch 540/2125] avg loss 0.000398004, throughput 6.01928K wps
[Epoch 36 Batch 570/2125] avg loss 0.000559418, throughput 6.016K wps
[Epoch 36 Batch 600/2125] avg loss 0.000309792, throughput 6.02085K wps
[Epoch 36 Batch 630/2125] avg loss 0.000346535, throughput 6.01182K wps
[Epoch 36 Batch 660/2125] avg loss 0.000384029, throughput 6.01338K wps
[Epoch 36 Batch 690/2125] avg loss 0.00020942, throughput 6.00985K wps
[Epoch 36 Batch 720/2125] avg loss 0.000407614, throughput 5.99893K wps
[Epoch 36 Batch 750/2125] avg loss 0.000323848, throughput 6.02261K wps
[Epoch 36 Batch 780/2125] avg loss 0.000419682, throughput 6.01153K wps
[Epoch 36 Batch 810/2125] avg loss 0.000442576, throughput 6.01778K wps
[Epoch 36 Batch 840/2125] avg loss 0.000335384, throughput 6.01187K wps
[Epoch 36 Batch 870/2125] avg loss 0.000368238, throughput 6.01095K wps
[Epoch 36 Batch 900/2125] avg loss 0.00038497, throughput 6.01197K wps
[Epoch 36 Batch 930/2125] avg loss 0.000424942, throughput 6.01139K wps
[Epoch 36 Batch 960/2125] avg loss 0.000589165, throughput 6.00958K wps
[Epoch 36 Batch 990/2125] avg loss 0.000382513, throughput 6.00593K wps
[Epoch 36 Batch 1020/2125] avg loss 0.000730101, throughput 6.01973K wps
[Epoch 36 Batch 1050/2125] avg loss 0.000615391, throughput 6.00402K wps
[Epoch 36 Batch 1080/2125] avg loss 0.000506356, throughput 6.0159K wps
[Epoch 36 Batch 1110/2125] avg loss 0.000187653, throughput 6.00993K wps
[Epoch 36 Batch 1140/2125] avg loss 0.000422279, throughput 6.01161K wps
[Epoch 36 Batch 1170/2125] avg loss 0.000487638, throughput 6.02906K wps
[Epoch 36 Batch 1200/2125] avg loss 0.000534036, throughput 6.02396K wps
[Epoch 36 Batch 1230/2125] avg loss 0.000334362, throughput 6.02071K wps
[Epoch 36 Batch 1260/2125] avg loss 0.000341338, throughput 6.02407K wps
[Epoch 36 Batch 1290/2125] avg loss 0.000293192, throughput 6.00485K wps
[Epoch 36 Batch 1320/2125] avg loss 0.000508239, throughput 6.01246K wps
[Epoch 36 Batch 1350/2125] avg loss 0.000427394, throughput 6.02104K wps
[Epoch 36 Batch 1380/2125] avg loss 0.000461452, throughput 6.01207K wps
[Epoch 36 Batch 1410/2125] avg loss 0.000398281, throughput 6.0204K wps
[Epoch 36 Batch 1440/2125] avg loss 0.000356849, throughput 6.02114K wps
[Epoch 36 Batch 1470/2125] avg loss 0.000418999, throughput 6.01692K wps
[Epoch 36 Batch 1500/2125] avg loss 0.000388094, throughput 6.01112K wps
[Epoch 36 Batch 1530/2125] avg loss 0.000500717, throughput 6.02651K wps
[Epoch 36 Batch 1560/2125] avg loss 0.000378582, throughput 6.02704K wps
[Epoch 36 Batch 1590/2125] avg loss 0.000662705, throughput 6.01276K wps
[Epoch 36 Batch 1620/2125] avg loss 0.000635782, throughput 6.0176K wps
[Epoch 36 Batch 1650/2125] avg loss 0.000542101, throughput 6.0191K wps
[Epoch 36 Batch 1680/2125] avg loss 0.00057881, throughput 6.01933K wps
[Epoch 36 Batch 1710/2125] avg loss 0.000465591, throughput 6.01647K wps
[Epoch 36 Batch 1740/2125] avg loss 0.00040366, throughput 6.01882K wps
[Epoch 36 Batch 1770/2125] avg loss 0.000384022, throughput 6.01207K wps
[Epoch 36 Batch 1800/2125] avg loss 0.000454175, throughput 6.01505K wps
[Epoch 36 Batch 1830/2125] avg loss 0.000349745, throughput 6.00668K wps
[Epoch 36 Batch 1860/2125] avg loss 0.000381443, throughput 6.01268K wps
[Epoch 36 Batch 1890/2125] avg loss 0.000518132, throughput 6.00798K wps
[Epoch 36 Batch 1920/2125] avg loss 0.000554961, throughput 6.01495K wps
[Epoch 36 Batch 1950/2125] avg loss 0.000465523, throughput 6.01559K wps
[Epoch 36 Batch 1980/2125] avg loss 0.000413217, throughput 6.01452K wps
[Epoch 36 Batch 2010/2125] avg loss 0.000398464, throughput 6.01241K wps
[Epoch 36 Batch 2040/2125] avg loss 0.000362492, throughput 6.01713K wps
[Epoch 36 Batch 2070/2125] avg loss 0.000392127, throughput 6.00997K wps
[Epoch 36 Batch 2100/2125] avg loss 0.000332667, throughput 6.01362K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 36] train avg loss 0.000404294, test acc 0.9227, test avg loss 0.562625, throughput 6.01549K wps
[Epoch 37 Batch 30/2125] avg loss 0.000372018, throughput 6.16569K wps
[Epoch 37 Batch 60/2125] avg loss 0.000334157, throughput 6.01878K wps
[Epoch 37 Batch 90/2125] avg loss 0.000327675, throughput 6.00777K wps
[Epoch 37 Batch 120/2125] avg loss 0.000351701, throughput 6.00594K wps
[Epoch 37 Batch 150/2125] avg loss 0.000179639, throughput 6.00453K wps
[Epoch 37 Batch 180/2125] avg loss 0.00029556, throughput 6.00787K wps
[Epoch 37 Batch 210/2125] avg loss 0.000229949, throughput 6.0155K wps
[Epoch 37 Batch 240/2125] avg loss 0.000341687, throughput 6.01568K wps
[Epoch 37 Batch 270/2125] avg loss 0.000192782, throughput 6.00551K wps
[Epoch 37 Batch 300/2125] avg loss 0.000210072, throughput 6.00646K wps
[Epoch 37 Batch 330/2125] avg loss 0.000401501, throughput 6.00842K wps
[Epoch 37 Batch 360/2125] avg loss 0.000191091, throughput 5.99937K wps
[Epoch 37 Batch 390/2125] avg loss 0.000376885, throughput 6.01029K wps
[Epoch 37 Batch 420/2125] avg loss 0.000351278, throughput 6.01708K wps
[Epoch 37 Batch 450/2125] avg loss 0.000235137, throughput 6.00938K wps
[Epoch 37 Batch 480/2125] avg loss 0.000287694, throughput 5.99209K wps
[Epoch 37 Batch 510/2125] avg loss 0.000268587, throughput 6.003K wps
[Epoch 37 Batch 540/2125] avg loss 0.000399089, throughput 5.99365K wps
[Epoch 37 Batch 570/2125] avg loss 0.000277292, throughput 6.01897K wps
[Epoch 37 Batch 600/2125] avg loss 0.000521492, throughput 6.02104K wps
[Epoch 37 Batch 630/2125] avg loss 0.000191098, throughput 6.0178K wps
[Epoch 37 Batch 660/2125] avg loss 0.00035387, throughput 5.99908K wps
[Epoch 37 Batch 690/2125] avg loss 0.000319843, throughput 6.0038K wps
[Epoch 37 Batch 720/2125] avg loss 0.00047892, throughput 6.01662K wps
[Epoch 37 Batch 750/2125] avg loss 0.000454883, throughput 6.0171K wps
[Epoch 37 Batch 780/2125] avg loss 0.00030741, throughput 6.01044K wps
[Epoch 37 Batch 810/2125] avg loss 0.000473792, throughput 6.00577K wps
[Epoch 37 Batch 840/2125] avg loss 0.000408863, throughput 6.00851K wps
[Epoch 37 Batch 870/2125] avg loss 0.000359343, throughput 6.00528K wps
[Epoch 37 Batch 900/2125] avg loss 0.000349063, throughput 6.02626K wps
[Epoch 37 Batch 930/2125] avg loss 0.000427903, throughput 6.01373K wps
[Epoch 37 Batch 960/2125] avg loss 0.000704589, throughput 6.02179K wps
[Epoch 37 Batch 990/2125] avg loss 0.000444677, throughput 6.02322K wps
[Epoch 37 Batch 1020/2125] avg loss 0.000345921, throughput 6.00978K wps
[Epoch 37 Batch 1050/2125] avg loss 0.000486986, throughput 6.00488K wps
[Epoch 37 Batch 1080/2125] avg loss 0.00033203, throughput 6.01116K wps
[Epoch 37 Batch 1110/2125] avg loss 0.000748765, throughput 6.00328K wps
[Epoch 37 Batch 1140/2125] avg loss 0.000497602, throughput 6.01661K wps
[Epoch 37 Batch 1170/2125] avg loss 0.000296741, throughput 6.01686K wps
[Epoch 37 Batch 1200/2125] avg loss 0.000166469, throughput 6.01715K wps
[Epoch 37 Batch 1230/2125] avg loss 0.000371399, throughput 6.01333K wps
[Epoch 37 Batch 1260/2125] avg loss 0.000400917, throughput 6.01124K wps
[Epoch 37 Batch 1290/2125] avg loss 0.000462243, throughput 6.01141K wps
[Epoch 37 Batch 1320/2125] avg loss 0.000391722, throughput 6.01174K wps
[Epoch 37 Batch 1350/2125] avg loss 0.000320287, throughput 6.01986K wps
[Epoch 37 Batch 1380/2125] avg loss 0.000537525, throughput 6.00986K wps
[Epoch 37 Batch 1410/2125] avg loss 0.000510831, throughput 6.00562K wps
[Epoch 37 Batch 1440/2125] avg loss 0.000647884, throughput 6.01536K wps
[Epoch 37 Batch 1470/2125] avg loss 0.00027876, throughput 6.01862K wps
[Epoch 37 Batch 1500/2125] avg loss 0.000444519, throughput 6.01643K wps
[Epoch 37 Batch 1530/2125] avg loss 0.000578073, throughput 6.01446K wps
[Epoch 37 Batch 1560/2125] avg loss 0.000500076, throughput 6.00124K wps
[Epoch 37 Batch 1590/2125] avg loss 0.000408727, throughput 6.0088K wps
[Epoch 37 Batch 1620/2125] avg loss 0.000324031, throughput 6.01258K wps
[Epoch 37 Batch 1650/2125] avg loss 0.000364547, throughput 6.0162K wps
[Epoch 37 Batch 1680/2125] avg loss 0.000385806, throughput 6.01485K wps
[Epoch 37 Batch 1710/2125] avg loss 0.000440024, throughput 6.02164K wps
[Epoch 37 Batch 1740/2125] avg loss 0.000447023, throughput 6.02365K wps
[Epoch 37 Batch 1770/2125] avg loss 0.000325128, throughput 6.03037K wps
[Epoch 37 Batch 1800/2125] avg loss 0.000239, throughput 6.01502K wps
[Epoch 37 Batch 1830/2125] avg loss 0.000521663, throughput 6.01597K wps
[Epoch 37 Batch 1860/2125] avg loss 0.000410447, throughput 6.00844K wps
[Epoch 37 Batch 1890/2125] avg loss 0.000381562, throughput 6.01808K wps
[Epoch 37 Batch 1920/2125] avg loss 0.000320958, throughput 6.01703K wps
[Epoch 37 Batch 1950/2125] avg loss 0.000478991, throughput 6.00627K wps
[Epoch 37 Batch 1980/2125] avg loss 0.000392992, throughput 6.01197K wps
[Epoch 37 Batch 2010/2125] avg loss 0.000370871, throughput 6.01284K wps
[Epoch 37 Batch 2040/2125] avg loss 0.000605713, throughput 6.00718K wps
[Epoch 37 Batch 2070/2125] avg loss 0.000799619, throughput 6.02424K wps
[Epoch 37 Batch 2100/2125] avg loss 0.000555969, throughput 6.01373K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 37] train avg loss 0.000394628, test acc 0.9223, test avg loss 0.558042, throughput 6.01437K wps
[Epoch 38 Batch 30/2125] avg loss 0.000356316, throughput 6.15139K wps
[Epoch 38 Batch 60/2125] avg loss 0.000487049, throughput 6.01803K wps
[Epoch 38 Batch 90/2125] avg loss 0.000177, throughput 5.98888K wps
[Epoch 38 Batch 120/2125] avg loss 0.000313671, throughput 5.99443K wps
[Epoch 38 Batch 150/2125] avg loss 0.000349249, throughput 6.01161K wps
[Epoch 38 Batch 180/2125] avg loss 0.00042928, throughput 6.02154K wps
[Epoch 38 Batch 210/2125] avg loss 0.000260377, throughput 6.01876K wps
[Epoch 38 Batch 240/2125] avg loss 0.000326358, throughput 6.00994K wps
[Epoch 38 Batch 270/2125] avg loss 0.000256297, throughput 6.01246K wps
[Epoch 38 Batch 300/2125] avg loss 0.000180864, throughput 6.0217K wps
[Epoch 38 Batch 330/2125] avg loss 0.000429099, throughput 6.00417K wps
[Epoch 38 Batch 360/2125] avg loss 0.000459487, throughput 6.02131K wps
[Epoch 38 Batch 390/2125] avg loss 0.000307329, throughput 6.01399K wps
[Epoch 38 Batch 420/2125] avg loss 0.000331998, throughput 6.00871K wps
[Epoch 38 Batch 450/2125] avg loss 0.000476715, throughput 6.01379K wps
[Epoch 38 Batch 480/2125] avg loss 0.000313983, throughput 6.0162K wps
[Epoch 38 Batch 510/2125] avg loss 0.000381617, throughput 6.02323K wps
[Epoch 38 Batch 540/2125] avg loss 0.00035003, throughput 6.01986K wps
[Epoch 38 Batch 570/2125] avg loss 0.000425695, throughput 6.01354K wps
[Epoch 38 Batch 600/2125] avg loss 0.000383114, throughput 6.01506K wps
[Epoch 38 Batch 630/2125] avg loss 0.000339299, throughput 6.02137K wps
[Epoch 38 Batch 660/2125] avg loss 0.000399361, throughput 6.02202K wps
[Epoch 38 Batch 690/2125] avg loss 0.000212351, throughput 6.01218K wps
[Epoch 38 Batch 720/2125] avg loss 0.000340067, throughput 6.01388K wps
[Epoch 38 Batch 750/2125] avg loss 0.000265307, throughput 6.01551K wps
[Epoch 38 Batch 780/2125] avg loss 0.000386527, throughput 6.0146K wps
[Epoch 38 Batch 810/2125] avg loss 0.000307832, throughput 6.02616K wps
[Epoch 38 Batch 840/2125] avg loss 0.000520985, throughput 6.01158K wps
[Epoch 38 Batch 870/2125] avg loss 0.000375319, throughput 6.01235K wps
[Epoch 38 Batch 900/2125] avg loss 0.000383625, throughput 6.0162K wps
[Epoch 38 Batch 930/2125] avg loss 0.00037645, throughput 6.02328K wps
[Epoch 38 Batch 960/2125] avg loss 0.000410997, throughput 6.019K wps
[Epoch 38 Batch 990/2125] avg loss 0.000333164, throughput 6.02023K wps
[Epoch 38 Batch 1020/2125] avg loss 0.000346319, throughput 6.02123K wps
[Epoch 38 Batch 1050/2125] avg loss 0.000423562, throughput 6.01954K wps
[Epoch 38 Batch 1080/2125] avg loss 0.000243454, throughput 6.02101K wps
[Epoch 38 Batch 1110/2125] avg loss 0.000455139, throughput 6.0134K wps
[Epoch 38 Batch 1140/2125] avg loss 0.000453554, throughput 6.01355K wps
[Epoch 38 Batch 1170/2125] avg loss 0.000515781, throughput 6.01008K wps
[Epoch 38 Batch 1200/2125] avg loss 0.000406491, throughput 6.02223K wps
[Epoch 38 Batch 1230/2125] avg loss 0.000552252, throughput 6.01219K wps
[Epoch 38 Batch 1260/2125] avg loss 0.000369538, throughput 6.00885K wps
[Epoch 38 Batch 1290/2125] avg loss 0.0003597, throughput 6.01896K wps
[Epoch 38 Batch 1320/2125] avg loss 0.000290938, throughput 6.01839K wps
[Epoch 38 Batch 1350/2125] avg loss 0.000366695, throughput 6.01689K wps
[Epoch 38 Batch 1380/2125] avg loss 0.000492468, throughput 6.0268K wps
[Epoch 38 Batch 1410/2125] avg loss 0.000325794, throughput 6.02623K wps
[Epoch 38 Batch 1440/2125] avg loss 0.000565653, throughput 6.01664K wps
[Epoch 38 Batch 1470/2125] avg loss 0.000515704, throughput 6.01499K wps
[Epoch 38 Batch 1500/2125] avg loss 0.000428748, throughput 6.01883K wps
[Epoch 38 Batch 1530/2125] avg loss 0.000421593, throughput 6.02199K wps
[Epoch 38 Batch 1560/2125] avg loss 0.000297441, throughput 6.01937K wps
[Epoch 38 Batch 1590/2125] avg loss 0.000421248, throughput 6.02225K wps
[Epoch 38 Batch 1620/2125] avg loss 0.000520042, throughput 6.02044K wps
[Epoch 38 Batch 1650/2125] avg loss 0.000467813, throughput 6.02627K wps
[Epoch 38 Batch 1680/2125] avg loss 0.000526308, throughput 6.01158K wps
[Epoch 38 Batch 1710/2125] avg loss 0.000266067, throughput 6.02522K wps
[Epoch 38 Batch 1740/2125] avg loss 0.00035336, throughput 6.02325K wps
[Epoch 38 Batch 1770/2125] avg loss 0.00030342, throughput 6.01721K wps
[Epoch 38 Batch 1800/2125] avg loss 0.00060782, throughput 6.01419K wps
[Epoch 38 Batch 1830/2125] avg loss 0.000406801, throughput 6.0238K wps
[Epoch 38 Batch 1860/2125] avg loss 0.000236425, throughput 6.01348K wps
[Epoch 38 Batch 1890/2125] avg loss 0.000569834, throughput 6.01755K wps
[Epoch 38 Batch 1920/2125] avg loss 0.000173527, throughput 6.02222K wps
[Epoch 38 Batch 1950/2125] avg loss 0.000328443, throughput 6.01522K wps
[Epoch 38 Batch 1980/2125] avg loss 0.000490441, throughput 6.00556K wps
[Epoch 38 Batch 2010/2125] avg loss 0.000458236, throughput 6.01236K wps
[Epoch 38 Batch 2040/2125] avg loss 0.000323746, throughput 6.00953K wps
[Epoch 38 Batch 2070/2125] avg loss 0.000437577, throughput 6.02142K wps
[Epoch 38 Batch 2100/2125] avg loss 0.000376439, throughput 6.02302K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 38] train avg loss 0.000382876, test acc 0.9234, test avg loss 0.560039, throughput 6.01848K wps
[Epoch 39 Batch 30/2125] avg loss 0.000268129, throughput 6.14206K wps
[Epoch 39 Batch 60/2125] avg loss 0.000304811, throughput 6.00803K wps
[Epoch 39 Batch 90/2125] avg loss 0.000311121, throughput 6.01647K wps
[Epoch 39 Batch 120/2125] avg loss 0.00024747, throughput 6.00453K wps
[Epoch 39 Batch 150/2125] avg loss 0.000204238, throughput 6.00683K wps
[Epoch 39 Batch 180/2125] avg loss 0.000362221, throughput 6.01464K wps
[Epoch 39 Batch 210/2125] avg loss 0.00020425, throughput 6.0106K wps
[Epoch 39 Batch 240/2125] avg loss 0.00030132, throughput 5.99355K wps
[Epoch 39 Batch 270/2125] avg loss 0.000359749, throughput 6.01092K wps
[Epoch 39 Batch 300/2125] avg loss 0.000177062, throughput 6.00766K wps
[Epoch 39 Batch 330/2125] avg loss 0.000231014, throughput 6.00988K wps
[Epoch 39 Batch 360/2125] avg loss 0.000207672, throughput 6.01183K wps
[Epoch 39 Batch 390/2125] avg loss 0.000244367, throughput 6.01251K wps
[Epoch 39 Batch 420/2125] avg loss 0.000265842, throughput 6.01402K wps
[Epoch 39 Batch 450/2125] avg loss 0.000344297, throughput 6.01588K wps
[Epoch 39 Batch 480/2125] avg loss 0.00054962, throughput 6.00921K wps
[Epoch 39 Batch 510/2125] avg loss 0.000246353, throughput 6.02472K wps
[Epoch 39 Batch 540/2125] avg loss 0.000334582, throughput 6.03139K wps
[Epoch 39 Batch 570/2125] avg loss 0.000467465, throughput 6.02228K wps
[Epoch 39 Batch 600/2125] avg loss 0.000269847, throughput 6.01682K wps
[Epoch 39 Batch 630/2125] avg loss 0.000401783, throughput 6.00502K wps
[Epoch 39 Batch 660/2125] avg loss 0.000290321, throughput 6.0078K wps
[Epoch 39 Batch 690/2125] avg loss 0.000391187, throughput 6.01418K wps
[Epoch 39 Batch 720/2125] avg loss 0.00036601, throughput 6.01549K wps
[Epoch 39 Batch 750/2125] avg loss 0.00032815, throughput 6.01757K wps
[Epoch 39 Batch 780/2125] avg loss 0.000246137, throughput 6.01268K wps
[Epoch 39 Batch 810/2125] avg loss 0.000368755, throughput 6.00684K wps
[Epoch 39 Batch 840/2125] avg loss 0.000283056, throughput 6.00697K wps
[Epoch 39 Batch 870/2125] avg loss 0.000314853, throughput 6.01492K wps
[Epoch 39 Batch 900/2125] avg loss 0.000522814, throughput 6.01461K wps
[Epoch 39 Batch 930/2125] avg loss 0.000347165, throughput 6.0131K wps
[Epoch 39 Batch 960/2125] avg loss 0.000395124, throughput 6.02092K wps
[Epoch 39 Batch 990/2125] avg loss 0.000272325, throughput 6.01735K wps
[Epoch 39 Batch 1020/2125] avg loss 0.000271779, throughput 6.01291K wps
[Epoch 39 Batch 1050/2125] avg loss 0.000303936, throughput 6.01612K wps
[Epoch 39 Batch 1080/2125] avg loss 0.000507535, throughput 6.00436K wps
[Epoch 39 Batch 1110/2125] avg loss 0.000294283, throughput 6.02077K wps
[Epoch 39 Batch 1140/2125] avg loss 0.000224579, throughput 6.01446K wps
[Epoch 39 Batch 1170/2125] avg loss 0.000425345, throughput 6.01021K wps
[Epoch 39 Batch 1200/2125] avg loss 0.000227772, throughput 6.02341K wps
[Epoch 39 Batch 1230/2125] avg loss 0.000349748, throughput 6.01729K wps
[Epoch 39 Batch 1260/2125] avg loss 0.000325864, throughput 6.02172K wps
[Epoch 39 Batch 1290/2125] avg loss 0.000542695, throughput 6.02851K wps
[Epoch 39 Batch 1320/2125] avg loss 0.000373671, throughput 6.00715K wps
[Epoch 39 Batch 1350/2125] avg loss 0.00048523, throughput 6.01779K wps
[Epoch 39 Batch 1380/2125] avg loss 0.000407869, throughput 6.01526K wps
[Epoch 39 Batch 1410/2125] avg loss 0.000338561, throughput 6.00899K wps
[Epoch 39 Batch 1440/2125] avg loss 0.000492299, throughput 6.01217K wps
[Epoch 39 Batch 1470/2125] avg loss 0.000523689, throughput 6.02829K wps
[Epoch 39 Batch 1500/2125] avg loss 0.000387721, throughput 6.00879K wps
[Epoch 39 Batch 1530/2125] avg loss 0.000366142, throughput 6.01754K wps
[Epoch 39 Batch 1560/2125] avg loss 0.000334434, throughput 6.01529K wps
[Epoch 39 Batch 1590/2125] avg loss 0.000392557, throughput 6.01489K wps
[Epoch 39 Batch 1620/2125] avg loss 0.00038135, throughput 6.01877K wps
[Epoch 39 Batch 1650/2125] avg loss 0.000482855, throughput 6.01522K wps
[Epoch 39 Batch 1680/2125] avg loss 0.000373921, throughput 6.01898K wps
[Epoch 39 Batch 1710/2125] avg loss 0.00054892, throughput 6.01055K wps
[Epoch 39 Batch 1740/2125] avg loss 0.000517618, throughput 6.0137K wps
[Epoch 39 Batch 1770/2125] avg loss 0.000308003, throughput 6.02021K wps
[Epoch 39 Batch 1800/2125] avg loss 0.000343815, throughput 6.01965K wps
[Epoch 39 Batch 1830/2125] avg loss 0.000486128, throughput 6.0135K wps
[Epoch 39 Batch 1860/2125] avg loss 0.000347147, throughput 6.01077K wps
[Epoch 39 Batch 1890/2125] avg loss 0.000493897, throughput 6.01523K wps
[Epoch 39 Batch 1920/2125] avg loss 0.000451051, throughput 6.01182K wps
[Epoch 39 Batch 1950/2125] avg loss 0.000508861, throughput 6.01266K wps
[Epoch 39 Batch 1980/2125] avg loss 0.000318871, throughput 6.01042K wps
[Epoch 39 Batch 2010/2125] avg loss 0.000557633, throughput 6.02134K wps
[Epoch 39 Batch 2040/2125] avg loss 0.000507708, throughput 6.02737K wps
[Epoch 39 Batch 2070/2125] avg loss 0.000438482, throughput 6.01704K wps
[Epoch 39 Batch 2100/2125] avg loss 0.000324335, throughput 6.0265K wps
Begin Testing...
[Batch 30/237] elapsed 0.29 s
[Batch 60/237] elapsed 0.27 s
[Batch 90/237] elapsed 0.27 s
[Batch 120/237] elapsed 0.27 s
[Batch 150/237] elapsed 0.27 s
[Batch 180/237] elapsed 0.27 s
[Batch 210/237] elapsed 0.27 s
[Epoch 39] train avg loss 0.00036533, test acc 0.9227, test avg loss 0.574377, throughput 6.0165K wps
Test loss 0.185785, test acc 0.9358
Total time cost 2916.38s