|
- epoch: 1 step: 142, loss is 5.33523416519165
- Train epoch time: 466942.745 ms, per step time: 3288.329 ms
- epoch: 2 step: 142, loss is 5.2280778884887695
- Train epoch time: 48865.572 ms, per step time: 344.124 ms
- epoch: 3 step: 142, loss is 5.181534290313721
- Train epoch time: 48260.661 ms, per step time: 339.864 ms
- epoch: 4 step: 142, loss is 5.179335594177246
- Train epoch time: 47188.082 ms, per step time: 332.310 ms
- epoch: 5 step: 142, loss is 5.1155900955200195
- Train epoch time: 46660.846 ms, per step time: 328.598 ms
- epoch: 6 step: 142, loss is 5.134178638458252
- Train epoch time: 48044.050 ms, per step time: 338.338 ms
- epoch: 7 step: 142, loss is 5.171756267547607
- Train epoch time: 47962.436 ms, per step time: 337.764 ms
- epoch: 8 step: 142, loss is 5.24785852432251
- Train epoch time: 47850.357 ms, per step time: 336.974 ms
- epoch: 9 step: 142, loss is 4.878683090209961
- Train epoch time: 47743.643 ms, per step time: 336.223 ms
- epoch: 10 step: 142, loss is 5.039165496826172
- Train epoch time: 46089.171 ms, per step time: 324.572 ms
- Validation performance:
- Epoch: 10, Val Loss: 5.205997117360433, Val Top1-Acc: 0.08385416666666666, Val Top5-Acc: 0.18125
- Validation performance improved! New best epoch: 10
- epoch: 11 step: 142, loss is 5.081485748291016
- Train epoch time: 29640.597 ms, per step time: 208.737 ms
- epoch: 12 step: 142, loss is 4.892566204071045
- Train epoch time: 48504.328 ms, per step time: 341.580 ms
- epoch: 13 step: 142, loss is 4.554628372192383
- Train epoch time: 46734.557 ms, per step time: 329.117 ms
- epoch: 14 step: 142, loss is 4.714110374450684
- Train epoch time: 47698.450 ms, per step time: 335.905 ms
- epoch: 15 step: 142, loss is 4.926340103149414
- Train epoch time: 47582.185 ms, per step time: 335.086 ms
- epoch: 16 step: 142, loss is 4.918840408325195
- Train epoch time: 46415.581 ms, per step time: 326.870 ms
- epoch: 17 step: 142, loss is 5.082817554473877
- Train epoch time: 46295.392 ms, per step time: 326.024 ms
- epoch: 18 step: 142, loss is 4.349315166473389
- Train epoch time: 48657.217 ms, per step time: 342.656 ms
- epoch: 19 step: 142, loss is 4.6006574630737305
- Train epoch time: 46397.027 ms, per step time: 326.740 ms
- epoch: 20 step: 142, loss is 4.543763160705566
- Train epoch time: 47626.111 ms, per step time: 335.395 ms
- Validation performance:
- Epoch: 20, Val Loss: 4.532274278004964, Val Top1-Acc: 0.15208333333333332, Val Top5-Acc: 0.31614583333333335
- Validation performance improved! New best epoch: 20
- epoch: 21 step: 142, loss is 4.383084297180176
- Train epoch time: 30414.539 ms, per step time: 214.187 ms
- epoch: 22 step: 142, loss is 4.218634605407715
- Train epoch time: 47407.274 ms, per step time: 333.854 ms
- epoch: 23 step: 142, loss is 4.482508659362793
- Train epoch time: 47030.947 ms, per step time: 331.204 ms
- epoch: 24 step: 142, loss is 4.274173736572266
- Train epoch time: 47651.918 ms, per step time: 335.577 ms
- epoch: 25 step: 142, loss is 4.176424503326416
- Train epoch time: 47004.044 ms, per step time: 331.014 ms
- epoch: 26 step: 142, loss is 4.204885482788086
- Train epoch time: 47476.218 ms, per step time: 334.340 ms
- epoch: 27 step: 142, loss is 4.199054718017578
- Train epoch time: 47249.446 ms, per step time: 332.743 ms
- epoch: 28 step: 142, loss is 4.146823883056641
- Train epoch time: 47274.369 ms, per step time: 332.918 ms
- epoch: 29 step: 142, loss is 4.071079254150391
- Train epoch time: 46546.470 ms, per step time: 327.792 ms
- epoch: 30 step: 142, loss is 4.1562981605529785
- Train epoch time: 47500.282 ms, per step time: 334.509 ms
- Validation performance:
- Epoch: 30, Val Loss: 4.0407106399536135, Val Top1-Acc: 0.2390625, Val Top5-Acc: 0.434375
- Validation performance improved! New best epoch: 30
- epoch: 31 step: 142, loss is 4.314669609069824
- Train epoch time: 28117.324 ms, per step time: 198.009 ms
- epoch: 32 step: 142, loss is 3.7696971893310547
- Train epoch time: 47769.238 ms, per step time: 336.403 ms
- epoch: 33 step: 142, loss is 3.782928943634033
- Train epoch time: 47134.058 ms, per step time: 331.930 ms
- epoch: 34 step: 142, loss is 3.735499858856201
- Train epoch time: 46403.632 ms, per step time: 326.786 ms
- epoch: 35 step: 142, loss is 4.155681610107422
- Train epoch time: 47942.309 ms, per step time: 337.622 ms
- epoch: 36 step: 142, loss is 3.9728946685791016
- Train epoch time: 47611.195 ms, per step time: 335.290 ms
- epoch: 37 step: 142, loss is 3.6065633296966553
- Train epoch time: 45488.055 ms, per step time: 320.338 ms
- epoch: 38 step: 142, loss is 3.887690305709839
- Train epoch time: 47563.252 ms, per step time: 334.952 ms
- epoch: 39 step: 142, loss is 3.571854829788208
- Train epoch time: 48558.527 ms, per step time: 341.961 ms
- epoch: 40 step: 142, loss is 3.9914932250976562
- Train epoch time: 47325.924 ms, per step time: 333.281 ms
- Validation performance:
- Epoch: 40, Val Loss: 3.8344762961069745, Val Top1-Acc: 0.2734375, Val Top5-Acc: 0.49114583333333334
- Validation performance improved! New best epoch: 40
- epoch: 41 step: 142, loss is 3.8908305168151855
- Train epoch time: 31252.093 ms, per step time: 220.085 ms
- epoch: 42 step: 142, loss is 3.769104242324829
- Train epoch time: 47830.499 ms, per step time: 336.834 ms
- epoch: 43 step: 142, loss is 3.6424167156219482
- Train epoch time: 46492.422 ms, per step time: 327.411 ms
- epoch: 44 step: 142, loss is 3.8570704460144043
- Train epoch time: 47455.459 ms, per step time: 334.193 ms
- epoch: 45 step: 142, loss is 3.802928924560547
- Train epoch time: 47218.324 ms, per step time: 332.523 ms
- epoch: 46 step: 142, loss is 3.583655595779419
- Train epoch time: 46974.631 ms, per step time: 330.807 ms
- epoch: 47 step: 142, loss is 3.7657651901245117
- Train epoch time: 48201.453 ms, per step time: 339.447 ms
- epoch: 48 step: 142, loss is 3.7043159008026123
- Train epoch time: 48076.588 ms, per step time: 338.568 ms
- epoch: 49 step: 142, loss is 3.5437047481536865
- Train epoch time: 48444.659 ms, per step time: 341.160 ms
- epoch: 50 step: 142, loss is 3.4561634063720703
- Train epoch time: 48600.332 ms, per step time: 342.256 ms
- Validation performance:
- Epoch: 50, Val Loss: 3.5991745630900067, Val Top1-Acc: 0.3265625, Val Top5-Acc: 0.5546875
- Validation performance improved! New best epoch: 50
- epoch: 51 step: 142, loss is 3.874518871307373
- Train epoch time: 31854.151 ms, per step time: 224.325 ms
- epoch: 52 step: 142, loss is 3.5904147624969482
- Train epoch time: 47640.754 ms, per step time: 335.498 ms
- epoch: 53 step: 142, loss is 3.374687671661377
- Train epoch time: 46749.419 ms, per step time: 329.221 ms
- epoch: 54 step: 142, loss is 3.4709718227386475
- Train epoch time: 48399.058 ms, per step time: 340.838 ms
- epoch: 55 step: 142, loss is 3.477057933807373
- Train epoch time: 46810.456 ms, per step time: 329.651 ms
- epoch: 56 step: 142, loss is 3.317335367202759
- Train epoch time: 47891.318 ms, per step time: 337.263 ms
- epoch: 57 step: 142, loss is 3.242927312850952
- Train epoch time: 49123.565 ms, per step time: 345.941 ms
- epoch: 58 step: 142, loss is 3.1817688941955566
- Train epoch time: 47277.945 ms, per step time: 332.943 ms
- epoch: 59 step: 142, loss is 3.3107125759124756
- Train epoch time: 49492.650 ms, per step time: 348.540 ms
- epoch: 60 step: 142, loss is 3.301819324493408
- Train epoch time: 47599.864 ms, per step time: 335.210 ms
- Validation performance:
- Epoch: 60, Val Loss: 3.2482171217600504, Val Top1-Acc: 0.40677083333333336, Val Top5-Acc: 0.6317708333333333
- Validation performance improved! New best epoch: 60
- epoch: 61 step: 142, loss is 3.4198808670043945
- Train epoch time: 28841.308 ms, per step time: 203.108 ms
- epoch: 62 step: 142, loss is 3.377803087234497
- Train epoch time: 47083.023 ms, per step time: 331.571 ms
- epoch: 63 step: 142, loss is 3.205322742462158
- Train epoch time: 49805.760 ms, per step time: 350.745 ms
- epoch: 64 step: 142, loss is 3.1459641456604004
- Train epoch time: 47144.933 ms, per step time: 332.007 ms
- epoch: 65 step: 142, loss is 3.152859926223755
- Train epoch time: 47961.768 ms, per step time: 337.759 ms
- epoch: 66 step: 142, loss is 3.083037853240967
- Train epoch time: 49825.539 ms, per step time: 350.884 ms
- epoch: 67 step: 142, loss is 3.0961647033691406
- Train epoch time: 47880.375 ms, per step time: 337.186 ms
- epoch: 68 step: 142, loss is 3.1469907760620117
- Train epoch time: 48790.952 ms, per step time: 343.598 ms
- epoch: 69 step: 142, loss is 3.1474645137786865
- Train epoch time: 47990.250 ms, per step time: 337.960 ms
- epoch: 70 step: 142, loss is 3.1076958179473877
- Train epoch time: 48004.387 ms, per step time: 338.059 ms
- Validation performance:
- Epoch: 70, Val Loss: 3.0459800561269126, Val Top1-Acc: 0.4578125, Val Top5-Acc: 0.684375
- Validation performance improved! New best epoch: 70
- epoch: 71 step: 142, loss is 3.36309552192688
- Train epoch time: 29052.627 ms, per step time: 204.596 ms
- epoch: 72 step: 142, loss is 3.144310235977173
- Train epoch time: 48909.786 ms, per step time: 344.435 ms
- epoch: 73 step: 142, loss is 2.924407720565796
- Train epoch time: 48041.722 ms, per step time: 338.322 ms
- epoch: 74 step: 142, loss is 3.0622570514678955
- Train epoch time: 47722.392 ms, per step time: 336.073 ms
- epoch: 75 step: 142, loss is 2.8230950832366943
- Train epoch time: 48693.466 ms, per step time: 342.912 ms
- epoch: 76 step: 142, loss is 3.159212112426758
- Train epoch time: 46501.656 ms, per step time: 327.476 ms
- epoch: 77 step: 142, loss is 2.819096326828003
- Train epoch time: 48330.338 ms, per step time: 340.354 ms
- epoch: 78 step: 142, loss is 2.9183661937713623
- Train epoch time: 48592.997 ms, per step time: 342.204 ms
- epoch: 79 step: 142, loss is 3.0823919773101807
- Train epoch time: 47853.901 ms, per step time: 336.999 ms
- epoch: 80 step: 142, loss is 2.7350447177886963
- Train epoch time: 48405.649 ms, per step time: 340.885 ms
- Validation performance:
- Epoch: 80, Val Loss: 2.857469352086385, Val Top1-Acc: 0.4895833333333333, Val Top5-Acc: 0.7052083333333333
- Validation performance improved! New best epoch: 80
- epoch: 81 step: 142, loss is 2.971315860748291
- Train epoch time: 31463.858 ms, per step time: 221.576 ms
- epoch: 82 step: 142, loss is 3.0406911373138428
- Train epoch time: 48092.924 ms, per step time: 338.683 ms
- epoch: 83 step: 142, loss is 2.890390157699585
- Train epoch time: 47296.615 ms, per step time: 333.075 ms
- epoch: 84 step: 142, loss is 2.5798401832580566
- Train epoch time: 52180.730 ms, per step time: 367.470 ms
- epoch: 85 step: 142, loss is 2.8666367530822754
- Train epoch time: 43709.981 ms, per step time: 307.817 ms
- epoch: 86 step: 142, loss is 2.6662282943725586
- Train epoch time: 48566.078 ms, per step time: 342.015 ms
- epoch: 87 step: 142, loss is 3.015873432159424
- Train epoch time: 48711.894 ms, per step time: 343.042 ms
- epoch: 88 step: 142, loss is 2.6233267784118652
- Train epoch time: 47437.531 ms, per step time: 334.067 ms
- epoch: 89 step: 142, loss is 2.405930757522583
- Train epoch time: 47838.955 ms, per step time: 336.894 ms
- epoch: 90 step: 142, loss is 2.631948232650757
- Train epoch time: 47692.713 ms, per step time: 335.864 ms
- Validation performance:
- Epoch: 90, Val Loss: 2.6245423475901286, Val Top1-Acc: 0.553125, Val Top5-Acc: 0.76875
- Validation performance improved! New best epoch: 90
- epoch: 91 step: 142, loss is 2.962907075881958
- Train epoch time: 30009.695 ms, per step time: 211.336 ms
- epoch: 92 step: 142, loss is 2.872561454772949
- Train epoch time: 48812.669 ms, per step time: 343.751 ms
- epoch: 93 step: 142, loss is 2.655607223510742
- Train epoch time: 47804.398 ms, per step time: 336.651 ms
- epoch: 94 step: 142, loss is 2.9415791034698486
- Train epoch time: 47309.913 ms, per step time: 333.168 ms
- epoch: 95 step: 142, loss is 2.8364064693450928
- Train epoch time: 46692.713 ms, per step time: 328.822 ms
- epoch: 96 step: 142, loss is 2.7743468284606934
- Train epoch time: 47513.019 ms, per step time: 334.599 ms
- epoch: 97 step: 142, loss is 2.684408664703369
- Train epoch time: 47221.238 ms, per step time: 332.544 ms
- epoch: 98 step: 142, loss is 2.447328567504883
- Train epoch time: 48561.620 ms, per step time: 341.983 ms
- epoch: 99 step: 142, loss is 2.685009002685547
- Train epoch time: 49687.076 ms, per step time: 349.909 ms
- epoch: 100 step: 142, loss is 2.6864912509918213
- Train epoch time: 48648.433 ms, per step time: 342.595 ms
- Validation performance:
- Epoch: 100, Val Loss: 2.5233370780944826, Val Top1-Acc: 0.5880208333333333, Val Top5-Acc: 0.7895833333333333
- Validation performance improved! New best epoch: 100
- epoch: 101 step: 142, loss is 2.74818754196167
- Train epoch time: 30682.722 ms, per step time: 216.076 ms
- epoch: 102 step: 142, loss is 2.7232558727264404
- Train epoch time: 48996.795 ms, per step time: 345.048 ms
- epoch: 103 step: 142, loss is 2.5135061740875244
- Train epoch time: 48232.663 ms, per step time: 339.667 ms
- epoch: 104 step: 142, loss is 2.651782751083374
- Train epoch time: 48698.959 ms, per step time: 342.950 ms
- epoch: 105 step: 142, loss is 2.5414721965789795
- Train epoch time: 49776.054 ms, per step time: 350.536 ms
- epoch: 106 step: 142, loss is 2.598870277404785
- Train epoch time: 48725.737 ms, per step time: 343.139 ms
- epoch: 107 step: 142, loss is 2.461397647857666
- Train epoch time: 49466.973 ms, per step time: 348.359 ms
- epoch: 108 step: 142, loss is 2.367835283279419
- Train epoch time: 50582.364 ms, per step time: 356.214 ms
- epoch: 109 step: 142, loss is 2.4824957847595215
- Train epoch time: 47439.699 ms, per step time: 334.082 ms
- epoch: 110 step: 142, loss is 2.6350526809692383
- Train epoch time: 47954.675 ms, per step time: 337.709 ms
- Validation performance:
- Epoch: 110, Val Loss: 2.3500694910685223, Val Top1-Acc: 0.6239583333333333, Val Top5-Acc: 0.8291666666666667
- Validation performance improved! New best epoch: 110
- epoch: 111 step: 142, loss is 2.4544355869293213
- Train epoch time: 32365.710 ms, per step time: 227.928 ms
- epoch: 112 step: 142, loss is 2.4396369457244873
- Train epoch time: 47761.024 ms, per step time: 336.345 ms
- epoch: 113 step: 142, loss is 2.7672829627990723
- Train epoch time: 51452.710 ms, per step time: 362.343 ms
- epoch: 114 step: 142, loss is 2.418727159500122
- Train epoch time: 49222.626 ms, per step time: 346.638 ms
- epoch: 115 step: 142, loss is 2.344849109649658
- Train epoch time: 49173.009 ms, per step time: 346.289 ms
- epoch: 116 step: 142, loss is 2.791077136993408
- Train epoch time: 48322.322 ms, per step time: 340.298 ms
- epoch: 117 step: 142, loss is 2.289138078689575
- Train epoch time: 48319.849 ms, per step time: 340.281 ms
- epoch: 118 step: 142, loss is 2.3315396308898926
- Train epoch time: 49572.731 ms, per step time: 349.104 ms
- epoch: 119 step: 142, loss is 2.1655375957489014
- Train epoch time: 48592.318 ms, per step time: 342.199 ms
- epoch: 120 step: 142, loss is 2.2034788131713867
- Train epoch time: 49539.374 ms, per step time: 348.869 ms
- Validation performance:
- Epoch: 120, Val Loss: 2.2073226451873778, Val Top1-Acc: 0.6739583333333333, Val Top5-Acc: 0.8458333333333333
- Validation performance improved! New best epoch: 120
- epoch: 121 step: 142, loss is 2.3499624729156494
- Train epoch time: 32001.158 ms, per step time: 225.360 ms
- epoch: 122 step: 142, loss is 2.312112808227539
- Train epoch time: 47793.946 ms, per step time: 336.577 ms
- epoch: 123 step: 142, loss is 2.529247522354126
- Train epoch time: 52669.065 ms, per step time: 370.909 ms
- epoch: 124 step: 142, loss is 2.365990161895752
- Train epoch time: 45828.814 ms, per step time: 322.738 ms
- epoch: 125 step: 142, loss is 2.474100351333618
- Train epoch time: 49000.255 ms, per step time: 345.072 ms
- epoch: 126 step: 142, loss is 2.3713316917419434
- Train epoch time: 49216.161 ms, per step time: 346.593 ms
- epoch: 127 step: 142, loss is 2.63864803314209
- Train epoch time: 49387.635 ms, per step time: 347.800 ms
- epoch: 128 step: 142, loss is 2.4020345211029053
- Train epoch time: 48507.741 ms, per step time: 341.604 ms
- epoch: 129 step: 142, loss is 2.2594010829925537
- Train epoch time: 49206.866 ms, per step time: 346.527 ms
- epoch: 130 step: 142, loss is 2.1809945106506348
- Train epoch time: 49099.895 ms, per step time: 345.774 ms
- Validation performance:
- Epoch: 130, Val Loss: 2.167334047953288, Val Top1-Acc: 0.6703125, Val Top5-Acc: 0.8614583333333333
- Validation performance improved! New best epoch: 130
- epoch: 131 step: 142, loss is 2.20681095123291
- Train epoch time: 30977.443 ms, per step time: 218.151 ms
- epoch: 132 step: 142, loss is 2.232734203338623
- Train epoch time: 48839.664 ms, per step time: 343.941 ms
- epoch: 133 step: 142, loss is 2.3276264667510986
- Train epoch time: 49260.607 ms, per step time: 346.906 ms
- epoch: 134 step: 142, loss is 2.316516637802124
- Train epoch time: 47833.181 ms, per step time: 336.853 ms
- epoch: 135 step: 142, loss is 2.257622718811035
- Train epoch time: 51196.556 ms, per step time: 360.539 ms
- epoch: 136 step: 142, loss is 2.4097483158111572
- Train epoch time: 47542.823 ms, per step time: 334.809 ms
- epoch: 137 step: 142, loss is 2.045778274536133
- Train epoch time: 49754.846 ms, per step time: 350.386 ms
- epoch: 138 step: 142, loss is 2.206174373626709
- Train epoch time: 49142.109 ms, per step time: 346.071 ms
- epoch: 139 step: 142, loss is 2.0907602310180664
- Train epoch time: 49547.503 ms, per step time: 348.926 ms
- epoch: 140 step: 142, loss is 2.0746617317199707
- Train epoch time: 48057.363 ms, per step time: 338.432 ms
- Validation performance:
- Epoch: 140, Val Loss: 2.0793466567993164, Val Top1-Acc: 0.7140625, Val Top5-Acc: 0.8640625
- Validation performance improved! New best epoch: 140
- epoch: 141 step: 142, loss is 2.094987392425537
- Train epoch time: 32611.583 ms, per step time: 229.659 ms
- epoch: 142 step: 142, loss is 2.235243558883667
- Train epoch time: 48800.598 ms, per step time: 343.666 ms
- epoch: 143 step: 142, loss is 2.4842042922973633
- Train epoch time: 47121.535 ms, per step time: 331.842 ms
- epoch: 144 step: 142, loss is 2.1225781440734863
- Train epoch time: 50968.411 ms, per step time: 358.932 ms
- epoch: 145 step: 142, loss is 2.2660281658172607
- Train epoch time: 48805.739 ms, per step time: 343.702 ms
- epoch: 146 step: 142, loss is 2.077753782272339
- Train epoch time: 48110.682 ms, per step time: 338.808 ms
- epoch: 147 step: 142, loss is 2.118350028991699
- Train epoch time: 49300.710 ms, per step time: 347.188 ms
- epoch: 148 step: 142, loss is 2.1737613677978516
- Train epoch time: 48501.087 ms, per step time: 341.557 ms
- epoch: 149 step: 142, loss is 2.15474534034729
- Train epoch time: 48201.218 ms, per step time: 339.445 ms
- epoch: 150 step: 142, loss is 2.2499501705169678
- Train epoch time: 52376.877 ms, per step time: 368.851 ms
- Validation performance:
- Epoch: 150, Val Loss: 1.9730801661809285, Val Top1-Acc: 0.7359375, Val Top5-Acc: 0.8859375
- Validation performance improved! New best epoch: 150
- epoch: 151 step: 142, loss is 2.1676011085510254
- Train epoch time: 29629.752 ms, per step time: 208.660 ms
- epoch: 152 step: 142, loss is 2.0163378715515137
- Train epoch time: 48150.703 ms, per step time: 339.089 ms
- epoch: 153 step: 142, loss is 1.9814424514770508
- Train epoch time: 49383.247 ms, per step time: 347.769 ms
- epoch: 154 step: 142, loss is 2.1515121459960938
- Train epoch time: 47715.411 ms, per step time: 336.024 ms
- epoch: 155 step: 142, loss is 1.8720626831054688
- Train epoch time: 47898.050 ms, per step time: 337.310 ms
- epoch: 156 step: 142, loss is 2.0324220657348633
- Train epoch time: 49373.619 ms, per step time: 347.702 ms
- epoch: 157 step: 142, loss is 1.9200868606567383
- Train epoch time: 49588.653 ms, per step time: 349.216 ms
- epoch: 158 step: 142, loss is 2.1014750003814697
- Train epoch time: 46835.649 ms, per step time: 329.829 ms
- epoch: 159 step: 142, loss is 2.00113844871521
- Train epoch time: 48077.916 ms, per step time: 338.577 ms
- epoch: 160 step: 142, loss is 1.9134899377822876
- Train epoch time: 48468.437 ms, per step time: 341.327 ms
- Validation performance:
- Epoch: 160, Val Loss: 1.8441479921340942, Val Top1-Acc: 0.76875, Val Top5-Acc: 0.9020833333333333
- Validation performance improved! New best epoch: 160
- epoch: 161 step: 142, loss is 1.9963997602462769
- Train epoch time: 30017.009 ms, per step time: 211.387 ms
- epoch: 162 step: 142, loss is 2.163062572479248
- Train epoch time: 50951.247 ms, per step time: 358.812 ms
- epoch: 163 step: 142, loss is 1.8607350587844849
- Train epoch time: 48958.783 ms, per step time: 344.780 ms
- epoch: 164 step: 142, loss is 2.0999763011932373
- Train epoch time: 47940.937 ms, per step time: 337.612 ms
- epoch: 165 step: 142, loss is 1.6864500045776367
- Train epoch time: 49373.943 ms, per step time: 347.704 ms
- epoch: 166 step: 142, loss is 1.8133057355880737
- Train epoch time: 49552.401 ms, per step time: 348.961 ms
- epoch: 167 step: 142, loss is 1.800733208656311
- Train epoch time: 47571.864 ms, per step time: 335.013 ms
- epoch: 168 step: 142, loss is 1.8098015785217285
- Train epoch time: 48489.035 ms, per step time: 341.472 ms
- epoch: 169 step: 142, loss is 1.9268313646316528
- Train epoch time: 48554.117 ms, per step time: 341.930 ms
- epoch: 170 step: 142, loss is 1.754817008972168
- Train epoch time: 48399.243 ms, per step time: 340.840 ms
- Validation performance:
- Epoch: 170, Val Loss: 1.8360045353571575, Val Top1-Acc: 0.7729166666666667, Val Top5-Acc: 0.9104166666666667
- Validation performance improved! New best epoch: 170
- epoch: 171 step: 142, loss is 2.026780843734741
- Train epoch time: 31821.839 ms, per step time: 224.097 ms
- epoch: 172 step: 142, loss is 2.0710320472717285
- Train epoch time: 49192.427 ms, per step time: 346.426 ms
- epoch: 173 step: 142, loss is 1.934451699256897
- Train epoch time: 48363.150 ms, per step time: 340.586 ms
- epoch: 174 step: 142, loss is 1.869722604751587
- Train epoch time: 52275.100 ms, per step time: 368.135 ms
- epoch: 175 step: 142, loss is 1.8464833498001099
- Train epoch time: 45709.756 ms, per step time: 321.900 ms
- epoch: 176 step: 142, loss is 1.8638594150543213
- Train epoch time: 48502.998 ms, per step time: 341.570 ms
- epoch: 177 step: 142, loss is 2.0383617877960205
- Train epoch time: 49895.036 ms, per step time: 351.373 ms
- epoch: 178 step: 142, loss is 1.9164249897003174
- Train epoch time: 48001.027 ms, per step time: 338.035 ms
- epoch: 179 step: 142, loss is 1.9494534730911255
- Train epoch time: 47701.810 ms, per step time: 335.928 ms
- epoch: 180 step: 142, loss is 1.8991755247116089
- Train epoch time: 52794.952 ms, per step time: 371.795 ms
- Validation performance:
- Epoch: 180, Val Loss: 1.7383160591125488, Val Top1-Acc: 0.7963541666666667, Val Top5-Acc: 0.9234375
- Validation performance improved! New best epoch: 180
- epoch: 181 step: 142, loss is 1.8521008491516113
- Train epoch time: 28071.832 ms, per step time: 197.689 ms
- epoch: 182 step: 142, loss is 1.6764440536499023
- Train epoch time: 49556.381 ms, per step time: 348.989 ms
- epoch: 183 step: 142, loss is 1.820200800895691
- Train epoch time: 48406.325 ms, per step time: 340.890 ms
- epoch: 184 step: 142, loss is 1.671720266342163
- Train epoch time: 48001.429 ms, per step time: 338.038 ms
- epoch: 185 step: 142, loss is 1.7586582899093628
- Train epoch time: 46833.795 ms, per step time: 329.815 ms
- epoch: 186 step: 142, loss is 1.8482362031936646
- Train epoch time: 49889.911 ms, per step time: 351.337 ms
- epoch: 187 step: 142, loss is 1.777928113937378
- Train epoch time: 48675.972 ms, per step time: 342.789 ms
- epoch: 188 step: 142, loss is 1.7924835681915283
- Train epoch time: 47733.815 ms, per step time: 336.154 ms
- epoch: 189 step: 142, loss is 1.7896496057510376
- Train epoch time: 48358.642 ms, per step time: 340.554 ms
- epoch: 190 step: 142, loss is 1.746803879737854
- Train epoch time: 48544.003 ms, per step time: 341.859 ms
- Validation performance:
- Epoch: 190, Val Loss: 1.739528743426005, Val Top1-Acc: 0.7953125, Val Top5-Acc: 0.9192708333333334
- Validation performance did not improved. Current best epoch: 180
- epoch: 191 step: 142, loss is 2.0373024940490723
- Train epoch time: 31356.363 ms, per step time: 220.819 ms
- epoch: 192 step: 142, loss is 1.7843517065048218
- Train epoch time: 49461.867 ms, per step time: 348.323 ms
- epoch: 193 step: 142, loss is 1.6440223455429077
- Train epoch time: 47911.703 ms, per step time: 337.406 ms
- epoch: 194 step: 142, loss is 1.7683343887329102
- Train epoch time: 48822.916 ms, per step time: 343.823 ms
- epoch: 195 step: 142, loss is 1.8447833061218262
- Train epoch time: 51074.448 ms, per step time: 359.679 ms
- epoch: 196 step: 142, loss is 1.882637858390808
- Train epoch time: 47401.848 ms, per step time: 333.816 ms
- epoch: 197 step: 142, loss is 1.7472056150436401
- Train epoch time: 50202.288 ms, per step time: 353.537 ms
- epoch: 198 step: 142, loss is 2.0267553329467773
- Train epoch time: 48996.804 ms, per step time: 345.048 ms
- epoch: 199 step: 142, loss is 1.6978826522827148
- Train epoch time: 48898.534 ms, per step time: 344.356 ms
- epoch: 200 step: 142, loss is 1.8158314228057861
- Train epoch time: 47102.878 ms, per step time: 331.710 ms
- Validation performance:
- Epoch: 200, Val Loss: 1.6507750272750854, Val Top1-Acc: 0.825, Val Top5-Acc: 0.9364583333333333
- Validation performance improved! New best epoch: 200
- epoch: 201 step: 142, loss is 2.1883652210235596
- Train epoch time: 31482.231 ms, per step time: 221.706 ms
- epoch: 202 step: 142, loss is 1.7768665552139282
- Train epoch time: 49320.903 ms, per step time: 347.330 ms
- epoch: 203 step: 142, loss is 1.6073081493377686
- Train epoch time: 47986.109 ms, per step time: 337.930 ms
- epoch: 204 step: 142, loss is 1.8855197429656982
- Train epoch time: 49149.597 ms, per step time: 346.124 ms
- epoch: 205 step: 142, loss is 1.7606561183929443
- Train epoch time: 50472.152 ms, per step time: 355.438 ms
- epoch: 206 step: 142, loss is 1.7706910371780396
- Train epoch time: 47843.498 ms, per step time: 336.926 ms
- epoch: 207 step: 142, loss is 1.5602138042449951
- Train epoch time: 50745.376 ms, per step time: 357.362 ms
- epoch: 208 step: 142, loss is 1.6820634603500366
- Train epoch time: 47382.332 ms, per step time: 333.678 ms
- epoch: 209 step: 142, loss is 1.6964693069458008
- Train epoch time: 49515.815 ms, per step time: 348.703 ms
- epoch: 210 step: 142, loss is 1.7031025886535645
- Train epoch time: 48811.430 ms, per step time: 343.742 ms
- Validation performance:
- Epoch: 210, Val Loss: 1.6079153696695963, Val Top1-Acc: 0.8359375, Val Top5-Acc: 0.9364583333333333
- Validation performance improved! New best epoch: 210
- epoch: 211 step: 142, loss is 1.9112911224365234
- Train epoch time: 30729.769 ms, per step time: 216.407 ms
- epoch: 212 step: 142, loss is 1.8493471145629883
- Train epoch time: 48662.219 ms, per step time: 342.692 ms
- epoch: 213 step: 142, loss is 1.722116470336914
- Train epoch time: 49256.747 ms, per step time: 346.878 ms
- epoch: 214 step: 142, loss is 1.6782371997833252
- Train epoch time: 48828.172 ms, per step time: 343.860 ms
- epoch: 215 step: 142, loss is 1.723233699798584
- Train epoch time: 49016.163 ms, per step time: 345.184 ms
- epoch: 216 step: 142, loss is 1.709067463874817
- Train epoch time: 48882.369 ms, per step time: 344.242 ms
- epoch: 217 step: 142, loss is 1.709031105041504
- Train epoch time: 47868.208 ms, per step time: 337.100 ms
- epoch: 218 step: 142, loss is 1.8784955739974976
- Train epoch time: 49842.096 ms, per step time: 351.001 ms
- epoch: 219 step: 142, loss is 1.6540143489837646
- Train epoch time: 49220.608 ms, per step time: 346.624 ms
- epoch: 220 step: 142, loss is 1.7352114915847778
- Train epoch time: 48269.466 ms, per step time: 339.926 ms
- Validation performance:
- Epoch: 220, Val Loss: 1.6116946538289387, Val Top1-Acc: 0.8369791666666667, Val Top5-Acc: 0.934375
- Validation performance did not improved. Current best epoch: 210
- epoch: 221 step: 142, loss is 1.7015295028686523
- Train epoch time: 30229.061 ms, per step time: 212.881 ms
- epoch: 222 step: 142, loss is 1.572005033493042
- Train epoch time: 49715.331 ms, per step time: 350.108 ms
- epoch: 223 step: 142, loss is 1.5999895334243774
- Train epoch time: 47857.121 ms, per step time: 337.022 ms
- epoch: 224 step: 142, loss is 1.6412485837936401
- Train epoch time: 49703.060 ms, per step time: 350.022 ms
- epoch: 225 step: 142, loss is 1.6987757682800293
- Train epoch time: 49323.812 ms, per step time: 347.351 ms
- epoch: 226 step: 142, loss is 1.8112436532974243
- Train epoch time: 48437.352 ms, per step time: 341.108 ms
- epoch: 227 step: 142, loss is 1.7333730459213257
- Train epoch time: 49648.395 ms, per step time: 349.637 ms
- epoch: 228 step: 142, loss is 1.5414342880249023
- Train epoch time: 49104.798 ms, per step time: 345.808 ms
- epoch: 229 step: 142, loss is 1.6562273502349854
- Train epoch time: 49086.706 ms, per step time: 345.681 ms
- epoch: 230 step: 142, loss is 1.6155034303665161
- Train epoch time: 48143.645 ms, per step time: 339.040 ms
- Validation performance:
- Epoch: 230, Val Loss: 1.5507694721221923, Val Top1-Acc: 0.8515625, Val Top5-Acc: 0.9427083333333334
- Validation performance improved! New best epoch: 230
- epoch: 231 step: 142, loss is 1.6809314489364624
- Train epoch time: 31274.221 ms, per step time: 220.241 ms
- epoch: 232 step: 142, loss is 1.8295596837997437
- Train epoch time: 48336.633 ms, per step time: 340.399 ms
- epoch: 233 step: 142, loss is 1.7615573406219482
- Train epoch time: 48800.689 ms, per step time: 343.667 ms
- epoch: 234 step: 142, loss is 1.6227948665618896
- Train epoch time: 50060.053 ms, per step time: 352.536 ms
- epoch: 235 step: 142, loss is 1.6259677410125732
- Train epoch time: 49090.666 ms, per step time: 345.709 ms
- epoch: 236 step: 142, loss is 1.72707200050354
- Train epoch time: 49672.819 ms, per step time: 349.809 ms
- epoch: 237 step: 142, loss is 1.7659276723861694
- Train epoch time: 49030.422 ms, per step time: 345.285 ms
- epoch: 238 step: 142, loss is 1.8375918865203857
- Train epoch time: 48888.538 ms, per step time: 344.285 ms
- epoch: 239 step: 142, loss is 1.4764424562454224
- Train epoch time: 49857.704 ms, per step time: 351.111 ms
- epoch: 240 step: 142, loss is 1.8095428943634033
- Train epoch time: 49700.508 ms, per step time: 350.004 ms
- Validation performance:
- Epoch: 240, Val Loss: 1.5217422644297283, Val Top1-Acc: 0.8578125, Val Top5-Acc: 0.9375
- Validation performance improved! New best epoch: 240
- epoch: 241 step: 142, loss is 1.7403652667999268
- Train epoch time: 29665.143 ms, per step time: 208.909 ms
- epoch: 242 step: 142, loss is 1.6006786823272705
- Train epoch time: 48251.737 ms, per step time: 339.801 ms
|