Number of GFLOPs: 46.618931712 Number of million parameters: 89.451406 Start training for 30 epochs Epoch: [0/30] [ 0/5004] eta: 5:40:18 lr: 0.000000 loss: 6.907786 (6.907786) time: 4.080354 data: 1.814401 max mem: 13345 Epoch: [0/30] [ 50/5004] eta: 1:04:52 lr: 0.000000 loss: 6.907732 (6.907744) time: 0.717059 data: 0.000221 max mem: 14335 Epoch: [0/30] [ 100/5004] eta: 1:01:33 lr: 0.000000 loss: 6.907649 (6.907713) time: 0.714567 data: 0.000236 max mem: 14335 Epoch: [0/30] [ 150/5004] eta: 0:59:59 lr: 0.000001 loss: 6.907484 (6.907657) time: 0.715570 data: 0.000224 max mem: 14335 Epoch: [0/30] [ 200/5004] eta: 0:58:53 lr: 0.000001 loss: 6.907295 (6.907586) time: 0.724129 data: 0.000239 max mem: 14335 Epoch: [0/30] [ 250/5004] eta: 0:57:58 lr: 0.000001 loss: 6.907055 (6.907498) time: 0.720119 data: 0.000213 max mem: 14335 Epoch: [0/30] [ 300/5004] eta: 0:57:11 lr: 0.000001 loss: 6.906709 (6.907389) time: 0.722216 data: 0.000154 max mem: 14335 Epoch: [0/30] [ 350/5004] eta: 0:56:24 lr: 0.000001 loss: 6.906426 (6.907265) time: 0.711603 data: 0.000169 max mem: 14335 Epoch: [0/30] [ 400/5004] eta: 0:55:42 lr: 0.000001 loss: 6.905963 (6.907121) time: 0.718247 data: 0.000220 max mem: 14335 Epoch: [0/30] [ 450/5004] eta: 0:54:59 lr: 0.000001 loss: 6.905370 (6.906950) time: 0.714920 data: 0.000215 max mem: 14335 Epoch: [0/30] [ 500/5004] eta: 0:54:18 lr: 0.000002 loss: 6.904907 (6.906761) time: 0.710463 data: 0.000207 max mem: 14335 Epoch: [0/30] [ 550/5004] eta: 0:53:37 lr: 0.000002 loss: 6.904149 (6.906541) time: 0.709973 data: 0.000221 max mem: 14335 Epoch: [0/30] [ 600/5004] eta: 0:52:58 lr: 0.000002 loss: 6.903325 (6.906295) time: 0.715407 data: 0.000216 max mem: 14335 Epoch: [0/30] [ 650/5004] eta: 0:52:20 lr: 0.000002 loss: 6.902551 (6.906034) time: 0.715711 data: 0.000178 max mem: 14335 Epoch: [0/30] [ 700/5004] eta: 0:51:41 lr: 0.000002 loss: 6.901677 (6.905741) time: 0.715612 data: 0.000153 max mem: 14335 Epoch: [0/30] [ 750/5004] eta: 0:51:03 lr: 0.000002 loss: 6.900672 (6.905420) time: 0.714564 data: 0.000225 max mem: 14335 Epoch: [0/30] [ 800/5004] eta: 0:50:25 lr: 0.000003 loss: 6.899216 (6.905070) time: 0.714702 data: 0.000208 max mem: 14335 Epoch: [0/30] [ 850/5004] eta: 0:49:48 lr: 0.000003 loss: 6.897708 (6.904669) time: 0.714492 data: 0.000243 max mem: 14335 Epoch: [0/30] [ 900/5004] eta: 0:49:10 lr: 0.000003 loss: 6.896502 (6.904245) time: 0.711464 data: 0.000202 max mem: 14335 Epoch: [0/30] [ 950/5004] eta: 0:48:33 lr: 0.000003 loss: 6.895230 (6.903800) time: 0.711149 data: 0.000225 max mem: 14335 Epoch: [0/30] [1000/5004] eta: 0:47:56 lr: 0.000003 loss: 6.893285 (6.903294) time: 0.716610 data: 0.000156 max mem: 14335 Epoch: [0/30] [1050/5004] eta: 0:47:20 lr: 0.000003 loss: 6.890790 (6.902730) time: 0.715979 data: 0.000213 max mem: 14335 Epoch: [0/30] [1100/5004] eta: 0:46:43 lr: 0.000004 loss: 6.889252 (6.902149) time: 0.710642 data: 0.000234 max mem: 14335 Epoch: [0/30] [1150/5004] eta: 0:46:06 lr: 0.000004 loss: 6.888077 (6.901533) time: 0.719402 data: 0.000213 max mem: 14335 Epoch: [0/30] [1200/5004] eta: 0:45:30 lr: 0.000004 loss: 6.884116 (6.900856) time: 0.719048 data: 0.000201 max mem: 14335 Epoch: [0/30] [1250/5004] eta: 0:44:54 lr: 0.000004 loss: 6.882926 (6.900154) time: 0.716803 data: 0.000216 max mem: 14335 Epoch: [0/30] [1300/5004] eta: 0:44:17 lr: 0.000004 loss: 6.878216 (6.899365) time: 0.709350 data: 0.000152 max mem: 14335 Epoch: [0/30] [1350/5004] eta: 0:43:41 lr: 0.000004 loss: 6.876118 (6.898532) time: 0.709596 data: 0.000155 max mem: 14335 Epoch: [0/30] [1400/5004] eta: 0:43:04 lr: 0.000005 loss: 6.872086 (6.897623) time: 0.713046 data: 0.000221 max mem: 14335 Epoch: [0/30] [1450/5004] eta: 0:42:28 lr: 0.000005 loss: 6.870232 (6.896696) time: 0.715108 data: 0.000215 max mem: 14335 Epoch: [0/30] [1500/5004] eta: 0:41:52 lr: 0.000005 loss: 6.865791 (6.895692) time: 0.714246 data: 0.000202 max mem: 14335 Epoch: [0/30] [1550/5004] eta: 0:41:15 lr: 0.000005 loss: 6.860647 (6.894637) time: 0.710612 data: 0.000195 max mem: 14335 Epoch: [0/30] [1600/5004] eta: 0:40:40 lr: 0.000005 loss: 6.858417 (6.893496) time: 0.717826 data: 0.000221 max mem: 14335 Epoch: [0/30] [1650/5004] eta: 0:40:03 lr: 0.000005 loss: 6.854877 (6.892306) time: 0.715662 data: 0.000166 max mem: 14335 Epoch: [0/30] [1700/5004] eta: 0:39:27 lr: 0.000005 loss: 6.846450 (6.891031) time: 0.712552 data: 0.000178 max mem: 14335 Epoch: [0/30] [1750/5004] eta: 0:38:51 lr: 0.000006 loss: 6.842788 (6.889681) time: 0.709056 data: 0.000211 max mem: 14335 Epoch: [0/30] [1800/5004] eta: 0:38:15 lr: 0.000006 loss: 6.834389 (6.888245) time: 0.712813 data: 0.000203 max mem: 14335 Epoch: [0/30] [1850/5004] eta: 0:37:39 lr: 0.000006 loss: 6.826809 (6.886740) time: 0.713923 data: 0.000228 max mem: 14335 Epoch: [0/30] [1900/5004] eta: 0:37:03 lr: 0.000006 loss: 6.823882 (6.885176) time: 0.713994 data: 0.000209 max mem: 14335 Epoch: [0/30] [1950/5004] eta: 0:36:27 lr: 0.000006 loss: 6.814876 (6.883488) time: 0.710212 data: 0.000231 max mem: 14335 Epoch: [0/30] [2000/5004] eta: 0:35:51 lr: 0.000006 loss: 6.812983 (6.881719) time: 0.715317 data: 0.000157 max mem: 14335 Epoch: [0/30] [2050/5004] eta: 0:35:15 lr: 0.000007 loss: 6.803906 (6.879897) time: 0.723770 data: 0.000156 max mem: 14335 Epoch: [0/30] [2100/5004] eta: 0:34:39 lr: 0.000007 loss: 6.797008 (6.877944) time: 0.714317 data: 0.000217 max mem: 14335 Epoch: [0/30] [2150/5004] eta: 0:34:03 lr: 0.000007 loss: 6.786987 (6.875895) time: 0.713820 data: 0.000207 max mem: 14335 Epoch: [0/30] [2200/5004] eta: 0:33:27 lr: 0.000007 loss: 6.783235 (6.873819) time: 0.717290 data: 0.000191 max mem: 14335 Epoch: [0/30] [2250/5004] eta: 0:32:51 lr: 0.000007 loss: 6.770944 (6.871589) time: 0.716085 data: 0.000212 max mem: 14335 Epoch: [0/30] [2300/5004] eta: 0:32:15 lr: 0.000007 loss: 6.758085 (6.869199) time: 0.710279 data: 0.000209 max mem: 14335 Epoch: [0/30] [2350/5004] eta: 0:31:39 lr: 0.000008 loss: 6.750230 (6.866818) time: 0.709433 data: 0.000164 max mem: 14335 Epoch: [0/30] [2400/5004] eta: 0:31:04 lr: 0.000008 loss: 6.733223 (6.864221) time: 0.716357 data: 0.000234 max mem: 14335 Epoch: [0/30] [2450/5004] eta: 0:30:28 lr: 0.000008 loss: 6.736514 (6.861597) time: 0.715614 data: 0.000197 max mem: 14335 Epoch: [0/30] [2500/5004] eta: 0:29:52 lr: 0.000008 loss: 6.724088 (6.858922) time: 0.714424 data: 0.000217 max mem: 14335 Epoch: [0/30] [2550/5004] eta: 0:29:16 lr: 0.000008 loss: 6.713208 (6.856141) time: 0.716693 data: 0.000221 max mem: 14335 Epoch: [0/30] [2600/5004] eta: 0:28:40 lr: 0.000008 loss: 6.695549 (6.853100) time: 0.716922 data: 0.000202 max mem: 14335 Epoch: [0/30] [2650/5004] eta: 0:28:04 lr: 0.000009 loss: 6.682677 (6.850114) time: 0.719135 data: 0.000157 max mem: 14335 Epoch: [0/30] [2700/5004] eta: 0:27:29 lr: 0.000009 loss: 6.676820 (6.846965) time: 0.712774 data: 0.000163 max mem: 14335 Epoch: [0/30] [2750/5004] eta: 0:26:53 lr: 0.000009 loss: 6.662771 (6.843753) time: 0.710488 data: 0.000407 max mem: 14335 Epoch: [0/30] [2800/5004] eta: 0:26:17 lr: 0.000009 loss: 6.651351 (6.840402) time: 0.715038 data: 0.000213 max mem: 14335 Epoch: [0/30] [2850/5004] eta: 0:25:41 lr: 0.000009 loss: 6.637753 (6.836922) time: 0.716270 data: 0.000181 max mem: 14335 Epoch: [0/30] [2900/5004] eta: 0:25:05 lr: 0.000009 loss: 6.625386 (6.833386) time: 0.709293 data: 0.000221 max mem: 14335 Epoch: [0/30] [2950/5004] eta: 0:24:29 lr: 0.000009 loss: 6.617451 (6.829715) time: 0.714098 data: 0.000207 max mem: 14335 Epoch: [0/30] [3000/5004] eta: 0:23:54 lr: 0.000010 loss: 6.604140 (6.825945) time: 0.724680 data: 0.000165 max mem: 14335 Epoch: [0/30] [3050/5004] eta: 0:23:18 lr: 0.000010 loss: 6.580612 (6.822035) time: 0.722620 data: 0.000161 max mem: 14335 Epoch: [0/30] [3100/5004] eta: 0:22:42 lr: 0.000010 loss: 6.565612 (6.817990) time: 0.715698 data: 0.000227 max mem: 14335 Epoch: [0/30] [3150/5004] eta: 0:22:06 lr: 0.000010 loss: 6.554471 (6.813805) time: 0.711981 data: 0.000214 max mem: 14335 Epoch: [0/30] [3200/5004] eta: 0:21:31 lr: 0.000010 loss: 6.529529 (6.809529) time: 0.712144 data: 0.000218 max mem: 14335 Epoch: [0/30] [3250/5004] eta: 0:20:55 lr: 0.000010 loss: 6.518351 (6.805156) time: 0.715526 data: 0.000221 max mem: 14335 Epoch: [0/30] [3300/5004] eta: 0:20:19 lr: 0.000011 loss: 6.509091 (6.800808) time: 0.713991 data: 0.000206 max mem: 14335 Epoch: [0/30] [3350/5004] eta: 0:19:43 lr: 0.000011 loss: 6.493838 (6.796252) time: 0.710585 data: 0.000156 max mem: 14335 Epoch: [0/30] [3400/5004] eta: 0:19:07 lr: 0.000011 loss: 6.473322 (6.791594) time: 0.712916 data: 0.000164 max mem: 14335 Epoch: [0/30] [3450/5004] eta: 0:18:32 lr: 0.000011 loss: 6.435613 (6.786785) time: 0.712977 data: 0.000217 max mem: 14335 Epoch: [0/30] [3500/5004] eta: 0:17:56 lr: 0.000011 loss: 6.433422 (6.781983) time: 0.710154 data: 0.000195 max mem: 14335 Epoch: [0/30] [3550/5004] eta: 0:17:20 lr: 0.000011 loss: 6.407966 (6.776944) time: 0.716821 data: 0.000217 max mem: 14335 Epoch: [0/30] [3600/5004] eta: 0:16:44 lr: 0.000012 loss: 6.384393 (6.771836) time: 0.718123 data: 0.000211 max mem: 14335 Epoch: [0/30] [3650/5004] eta: 0:16:08 lr: 0.000012 loss: 6.375049 (6.766563) time: 0.715801 data: 0.000229 max mem: 14335 Epoch: [0/30] [3700/5004] eta: 0:15:33 lr: 0.000012 loss: 6.358595 (6.761357) time: 0.710422 data: 0.000166 max mem: 14335 Epoch: [0/30] [3750/5004] eta: 0:14:57 lr: 0.000012 loss: 6.357800 (6.755999) time: 0.712437 data: 0.000220 max mem: 14335 Epoch: [0/30] [3800/5004] eta: 0:14:21 lr: 0.000012 loss: 6.342391 (6.750698) time: 0.711897 data: 0.000205 max mem: 14335 Epoch: [0/30] [3850/5004] eta: 0:13:45 lr: 0.000012 loss: 6.296119 (6.745126) time: 0.712605 data: 0.000208 max mem: 14335 Epoch: [0/30] [3900/5004] eta: 0:13:09 lr: 0.000013 loss: 6.303308 (6.739498) time: 0.710658 data: 0.000208 max mem: 14335 Epoch: [0/30] [3950/5004] eta: 0:12:33 lr: 0.000013 loss: 6.286768 (6.733792) time: 0.716315 data: 0.000247 max mem: 14335 Epoch: [0/30] [4000/5004] eta: 0:11:58 lr: 0.000013 loss: 6.241007 (6.727889) time: 0.725494 data: 0.000161 max mem: 14335 Epoch: [0/30] [4050/5004] eta: 0:11:22 lr: 0.000013 loss: 6.255827 (6.722054) time: 0.715437 data: 0.000152 max mem: 14335 Epoch: [0/30] [4100/5004] eta: 0:10:46 lr: 0.000013 loss: 6.209891 (6.716202) time: 0.710905 data: 0.000227 max mem: 14335 Epoch: [0/30] [4150/5004] eta: 0:10:10 lr: 0.000013 loss: 6.200284 (6.710114) time: 0.712177 data: 0.000184 max mem: 14335 Epoch: [0/30] [4200/5004] eta: 0:09:35 lr: 0.000013 loss: 6.157529 (6.704011) time: 0.716115 data: 0.000213 max mem: 14335 Epoch: [0/30] [4250/5004] eta: 0:08:59 lr: 0.000014 loss: 6.177060 (6.697684) time: 0.717125 data: 0.000218 max mem: 14335 Epoch: [0/30] [4300/5004] eta: 0:08:23 lr: 0.000014 loss: 6.136187 (6.691396) time: 0.718082 data: 0.000213 max mem: 14335 Epoch: [0/30] [4350/5004] eta: 0:07:47 lr: 0.000014 loss: 6.094280 (6.684767) time: 0.712367 data: 0.000171 max mem: 14335 Epoch: [0/30] [4400/5004] eta: 0:07:12 lr: 0.000014 loss: 6.108954 (6.678216) time: 0.720565 data: 0.000183 max mem: 14335 Epoch: [0/30] [4450/5004] eta: 0:06:36 lr: 0.000014 loss: 6.060546 (6.671562) time: 0.720727 data: 0.000207 max mem: 14335 Epoch: [0/30] [4500/5004] eta: 0:06:00 lr: 0.000014 loss: 6.061740 (6.664787) time: 0.714516 data: 0.000221 max mem: 14335 Epoch: [0/30] [4550/5004] eta: 0:05:24 lr: 0.000015 loss: 6.027140 (6.657942) time: 0.711173 data: 0.000224 max mem: 14335 Epoch: [0/30] [4600/5004] eta: 0:04:48 lr: 0.000015 loss: 6.030268 (6.651194) time: 0.712157 data: 0.000203 max mem: 14335 Epoch: [0/30] [4650/5004] eta: 0:04:13 lr: 0.000015 loss: 5.987349 (6.644241) time: 0.717854 data: 0.000203 max mem: 14335 Epoch: [0/30] [4700/5004] eta: 0:03:37 lr: 0.000015 loss: 5.964944 (6.637235) time: 0.709237 data: 0.000175 max mem: 14335 Epoch: [0/30] [4750/5004] eta: 0:03:01 lr: 0.000015 loss: 5.954533 (6.630122) time: 0.709925 data: 0.000163 max mem: 14335 Epoch: [0/30] [4800/5004] eta: 0:02:25 lr: 0.000015 loss: 5.922452 (6.622862) time: 0.717954 data: 0.000182 max mem: 14335 Epoch: [0/30] [4850/5004] eta: 0:01:50 lr: 0.000016 loss: 5.935663 (6.615802) time: 0.719636 data: 0.000209 max mem: 14335 Epoch: [0/30] [4900/5004] eta: 0:01:14 lr: 0.000016 loss: 5.874578 (6.608462) time: 0.716439 data: 0.000234 max mem: 14335 Epoch: [0/30] [4950/5004] eta: 0:00:38 lr: 0.000016 loss: 5.870518 (6.600914) time: 0.714108 data: 0.000244 max mem: 14335 Epoch: [0/30] [5000/5004] eta: 0:00:02 lr: 0.000016 loss: 5.839691 (6.593429) time: 0.711062 data: 0.000848 max mem: 14335 Epoch: [0/30] [5003/5004] eta: 0:00:00 lr: 0.000016 loss: 5.841025 (6.593012) time: 0.710538 data: 0.000841 max mem: 14335 Epoch: [0/30] Total time: 0:59:39 (0.715340 s / it) Averaged stats: lr: 0.000016 loss: 5.841025 (6.592741) Test: [ 0/196] eta: 0:04:53 loss: 5.547363 (5.547363) acc1: 68.750000 (68.750000) acc5: 87.500000 (87.500000) time: 1.495721 data: 1.100765 max mem: 14335 Test: [ 10/196] eta: 0:01:14 loss: 5.547363 (5.552801) acc1: 62.500000 (63.068182) acc5: 87.500000 (85.227273) time: 0.398362 data: 0.100205 max mem: 14335 Test: [ 20/196] eta: 0:01:00 loss: 5.544434 (5.541225) acc1: 62.500000 (63.988095) acc5: 87.500000 (86.904762) time: 0.288033 data: 0.000127 max mem: 14335 Test: [ 30/196] eta: 0:00:54 loss: 5.542969 (5.536897) acc1: 62.500000 (64.314516) acc5: 87.500000 (86.491935) time: 0.287254 data: 0.000106 max mem: 14335 Test: [ 40/196] eta: 0:00:49 loss: 5.538086 (5.538372) acc1: 62.500000 (63.262195) acc5: 87.500000 (86.432927) time: 0.286332 data: 0.000123 max mem: 14335 Test: [ 50/196] eta: 0:00:45 loss: 5.538086 (5.538186) acc1: 62.500000 (64.338235) acc5: 87.500000 (86.887255) time: 0.286002 data: 0.000126 max mem: 14335 Test: [ 60/196] eta: 0:00:41 loss: 5.557617 (5.541640) acc1: 68.750000 (64.754098) acc5: 87.500000 (86.680328) time: 0.286280 data: 0.000135 max mem: 14335 Test: [ 70/196] eta: 0:00:38 loss: 5.565186 (5.544581) acc1: 68.750000 (63.996479) acc5: 87.500000 (86.795775) time: 0.285847 data: 0.000146 max mem: 14335 Test: [ 80/196] eta: 0:00:34 loss: 5.557861 (5.545724) acc1: 62.500000 (63.888889) acc5: 87.500000 (86.728395) time: 0.286302 data: 0.000128 max mem: 14335 Test: [ 90/196] eta: 0:00:31 loss: 5.526611 (5.546505) acc1: 62.500000 (63.667582) acc5: 87.500000 (86.744505) time: 0.287422 data: 0.000133 max mem: 14335 Test: [100/196] eta: 0:00:28 loss: 5.516602 (5.539991) acc1: 62.500000 (63.613861) acc5: 87.500000 (87.004950) time: 0.286970 data: 0.000137 max mem: 14335 Test: [110/196] eta: 0:00:25 loss: 5.485596 (5.537290) acc1: 62.500000 (63.795045) acc5: 87.500000 (87.162162) time: 0.286580 data: 0.000127 max mem: 14335 Test: [120/196] eta: 0:00:22 loss: 5.504150 (5.536096) acc1: 62.500000 (63.791322) acc5: 87.500000 (87.241736) time: 0.294268 data: 0.000152 max mem: 14335 Test: [130/196] eta: 0:00:19 loss: 5.532959 (5.537963) acc1: 62.500000 (63.120229) acc5: 87.500000 (86.736641) time: 0.294615 data: 0.000160 max mem: 14335 Test: [140/196] eta: 0:00:16 loss: 5.525879 (5.536425) acc1: 62.500000 (63.563830) acc5: 87.500000 (86.879433) time: 0.287854 data: 0.000149 max mem: 14335 Test: [150/196] eta: 0:00:13 loss: 5.518311 (5.537371) acc1: 62.500000 (63.245033) acc5: 87.500000 (86.796358) time: 0.287553 data: 0.000145 max mem: 14335 Test: [160/196] eta: 0:00:10 loss: 5.584717 (5.541279) acc1: 56.250000 (62.965839) acc5: 87.500000 (86.607143) time: 0.286631 data: 0.000127 max mem: 14335 Test: [170/196] eta: 0:00:07 loss: 5.548096 (5.541148) acc1: 62.500000 (62.902047) acc5: 87.500000 (86.622807) time: 0.286386 data: 0.000135 max mem: 14335 Test: [180/196] eta: 0:00:04 loss: 5.538330 (5.541106) acc1: 62.500000 (62.845304) acc5: 87.500000 (86.567680) time: 0.286246 data: 0.000157 max mem: 14335 Test: [190/196] eta: 0:00:01 loss: 5.505615 (5.540072) acc1: 62.500000 (63.089005) acc5: 87.500000 (86.583770) time: 0.283611 data: 0.000123 max mem: 14335 Test: [195/196] eta: 0:00:00 loss: 5.504639 (5.539887) acc1: 62.500000 (63.072000) acc5: 87.500000 (86.400000) time: 0.277289 data: 0.000109 max mem: 14335 Test: Total time: 0:00:57 (0.294108 s / it) * Acc@1 62.982 Acc@5 86.728 loss 5.547 Max accuracy: 62.98% Epoch: [1/30] [ 0/5004] eta: 2:43:58 lr: 0.000016 loss: 5.716255 (5.716255) time: 1.966089 data: 1.210854 max mem: 14338 Epoch: [1/30] [ 50/5004] eta: 1:00:57 lr: 0.000016 loss: 5.832829 (5.818798) time: 0.714321 data: 0.000178 max mem: 14338 Epoch: [1/30] [ 100/5004] eta: 0:59:19 lr: 0.000016 loss: 5.837835 (5.810444) time: 0.713399 data: 0.000225 max mem: 14338 Epoch: [1/30] [ 150/5004] eta: 0:58:21 lr: 0.000017 loss: 5.716791 (5.800807) time: 0.709939 data: 0.000214 max mem: 14338 Epoch: [1/30] [ 200/5004] eta: 0:57:39 lr: 0.000017 loss: 5.717134 (5.791645) time: 0.718012 data: 0.000213 max mem: 14338 Epoch: [1/30] [ 250/5004] eta: 0:57:00 lr: 0.000017 loss: 5.738532 (5.784587) time: 0.714113 data: 0.000179 max mem: 14338 Epoch: [1/30] [ 300/5004] eta: 0:56:22 lr: 0.000017 loss: 5.697884 (5.771494) time: 0.718070 data: 0.000161 max mem: 14338 Epoch: [1/30] [ 350/5004] eta: 0:55:43 lr: 0.000017 loss: 5.703045 (5.758867) time: 0.715638 data: 0.000166 max mem: 14338 Epoch: [1/30] [ 400/5004] eta: 0:55:06 lr: 0.000017 loss: 5.646815 (5.748120) time: 0.719683 data: 0.000225 max mem: 14338 Epoch: [1/30] [ 450/5004] eta: 0:54:29 lr: 0.000017 loss: 5.587826 (5.735514) time: 0.718829 data: 0.000222 max mem: 14338 Epoch: [1/30] [ 500/5004] eta: 0:53:52 lr: 0.000018 loss: 5.588668 (5.723513) time: 0.713746 data: 0.000232 max mem: 14338 Epoch: [1/30] [ 550/5004] eta: 0:53:14 lr: 0.000018 loss: 5.546185 (5.712290) time: 0.710936 data: 0.000202 max mem: 14338 Epoch: [1/30] [ 600/5004] eta: 0:52:36 lr: 0.000018 loss: 5.585028 (5.701526) time: 0.712220 data: 0.000216 max mem: 14338 Epoch: [1/30] [ 650/5004] eta: 0:52:00 lr: 0.000018 loss: 5.528440 (5.688628) time: 0.717756 data: 0.000168 max mem: 14338 Epoch: [1/30] [ 700/5004] eta: 0:51:23 lr: 0.000018 loss: 5.538514 (5.677080) time: 0.712596 data: 0.000174 max mem: 14338 Epoch: [1/30] [ 750/5004] eta: 0:50:47 lr: 0.000018 loss: 5.489222 (5.665273) time: 0.712667 data: 0.000244 max mem: 14338 Epoch: [1/30] [ 800/5004] eta: 0:50:11 lr: 0.000019 loss: 5.450881 (5.652716) time: 0.724122 data: 0.000212 max mem: 14338 Epoch: [1/30] [ 850/5004] eta: 0:49:34 lr: 0.000019 loss: 5.441318 (5.640684) time: 0.711637 data: 0.000219 max mem: 14338 Epoch: [1/30] [ 900/5004] eta: 0:48:58 lr: 0.000019 loss: 5.417715 (5.627656) time: 0.713684 data: 0.000185 max mem: 14338 Epoch: [1/30] [ 950/5004] eta: 0:48:21 lr: 0.000019 loss: 5.382257 (5.615715) time: 0.714277 data: 0.000217 max mem: 14338 Epoch: [1/30] [1000/5004] eta: 0:47:45 lr: 0.000019 loss: 5.373742 (5.604164) time: 0.713505 data: 0.000159 max mem: 14338 Epoch: [1/30] [1050/5004] eta: 0:47:09 lr: 0.000019 loss: 5.353942 (5.592589) time: 0.715928 data: 0.000197 max mem: 14338 Epoch: [1/30] [1100/5004] eta: 0:46:33 lr: 0.000020 loss: 5.297109 (5.579861) time: 0.706858 data: 0.000198 max mem: 14338 Epoch: [1/30] [1150/5004] eta: 0:45:56 lr: 0.000020 loss: 5.228165 (5.566828) time: 0.709647 data: 0.000226 max mem: 14338 Epoch: [1/30] [1200/5004] eta: 0:45:20 lr: 0.000020 loss: 5.123832 (5.553515) time: 0.716791 data: 0.000208 max mem: 14338 Epoch: [1/30] [1250/5004] eta: 0:44:45 lr: 0.000020 loss: 5.236053 (5.541329) time: 0.716286 data: 0.000209 max mem: 14338 Epoch: [1/30] [1300/5004] eta: 0:44:09 lr: 0.000020 loss: 5.180782 (5.529068) time: 0.713488 data: 0.000162 max mem: 14338 Epoch: [1/30] [1350/5004] eta: 0:43:33 lr: 0.000020 loss: 5.139011 (5.515939) time: 0.718722 data: 0.000166 max mem: 14338 Epoch: [1/30] [1400/5004] eta: 0:42:57 lr: 0.000021 loss: 5.148164 (5.503163) time: 0.718970 data: 0.000207 max mem: 14338 Epoch: [1/30] [1450/5004] eta: 0:42:21 lr: 0.000021 loss: 5.154525 (5.491346) time: 0.712849 data: 0.000238 max mem: 14338 Epoch: [1/30] [1500/5004] eta: 0:41:45 lr: 0.000021 loss: 5.118303 (5.479464) time: 0.712218 data: 0.000205 max mem: 14338 Epoch: [1/30] [1550/5004] eta: 0:41:09 lr: 0.000021 loss: 5.065814 (5.467273) time: 0.711707 data: 0.000187 max mem: 14338 Epoch: [1/30] [1600/5004] eta: 0:40:33 lr: 0.000021 loss: 5.004508 (5.454273) time: 0.714764 data: 0.000196 max mem: 14338 Epoch: [1/30] [1650/5004] eta: 0:39:57 lr: 0.000021 loss: 5.070571 (5.441423) time: 0.714066 data: 0.000152 max mem: 14338 Epoch: [1/30] [1700/5004] eta: 0:39:21 lr: 0.000021 loss: 5.005449 (5.428873) time: 0.713872 data: 0.000160 max mem: 14338 Epoch: [1/30] [1750/5004] eta: 0:38:46 lr: 0.000022 loss: 4.953750 (5.416501) time: 0.715154 data: 0.000223 max mem: 14338 Epoch: [1/30] [1800/5004] eta: 0:38:10 lr: 0.000022 loss: 4.863748 (5.403629) time: 0.717552 data: 0.000210 max mem: 14338 Epoch: [1/30] [1850/5004] eta: 0:37:34 lr: 0.000022 loss: 4.834991 (5.390652) time: 0.717934 data: 0.000216 max mem: 14338 Epoch: [1/30] [1900/5004] eta: 0:36:59 lr: 0.000022 loss: 4.880677 (5.377461) time: 0.715776 data: 0.000235 max mem: 14338 Epoch: [1/30] [1950/5004] eta: 0:36:23 lr: 0.000022 loss: 4.769521 (5.363609) time: 0.709095 data: 0.000230 max mem: 14338 Epoch: [1/30] [2000/5004] eta: 0:35:47 lr: 0.000022 loss: 4.803717 (5.349780) time: 0.716507 data: 0.000166 max mem: 14338 Epoch: [1/30] [2050/5004] eta: 0:35:11 lr: 0.000023 loss: 4.755105 (5.336327) time: 0.715903 data: 0.000168 max mem: 14338 Epoch: [1/30] [2100/5004] eta: 0:34:36 lr: 0.000023 loss: 4.777626 (5.323396) time: 0.711415 data: 0.000203 max mem: 14338 Epoch: [1/30] [2150/5004] eta: 0:34:00 lr: 0.000023 loss: 4.691604 (5.309606) time: 0.708194 data: 0.000207 max mem: 14338 Epoch: [1/30] [2200/5004] eta: 0:33:24 lr: 0.000023 loss: 4.654694 (5.297117) time: 0.712919 data: 0.000188 max mem: 14338 Epoch: [1/30] [2250/5004] eta: 0:32:48 lr: 0.000023 loss: 4.607328 (5.283584) time: 0.722246 data: 0.000219 max mem: 14338 Epoch: [1/30] [2300/5004] eta: 0:32:12 lr: 0.000023 loss: 4.642088 (5.270513) time: 0.713239 data: 0.000218 max mem: 14338 Epoch: [1/30] [2350/5004] eta: 0:31:37 lr: 0.000024 loss: 4.582081 (5.257549) time: 0.720685 data: 0.000178 max mem: 14338 Epoch: [1/30] [2400/5004] eta: 0:31:01 lr: 0.000024 loss: 4.551291 (5.244817) time: 0.712498 data: 0.000193 max mem: 14338 Epoch: [1/30] [2450/5004] eta: 0:30:25 lr: 0.000024 loss: 4.547400 (5.231772) time: 0.714647 data: 0.000211 max mem: 14338 Epoch: [1/30] [2500/5004] eta: 0:29:49 lr: 0.000024 loss: 4.526954 (5.219214) time: 0.709738 data: 0.000207 max mem: 14338 Epoch: [1/30] [2550/5004] eta: 0:29:14 lr: 0.000024 loss: 4.496085 (5.206002) time: 0.712652 data: 0.000214 max mem: 14338 Epoch: [1/30] [2600/5004] eta: 0:28:38 lr: 0.000024 loss: 4.466552 (5.192717) time: 0.712521 data: 0.000215 max mem: 14338 Epoch: [1/30] [2650/5004] eta: 0:28:02 lr: 0.000025 loss: 4.475619 (5.179991) time: 0.716213 data: 0.000158 max mem: 14338 Epoch: [1/30] [2700/5004] eta: 0:27:26 lr: 0.000025 loss: 4.444686 (5.165934) time: 0.712365 data: 0.000146 max mem: 14338 Epoch: [1/30] [2750/5004] eta: 0:26:50 lr: 0.000025 loss: 4.379107 (5.152245) time: 0.718365 data: 0.000200 max mem: 14338 Epoch: [1/30] [2800/5004] eta: 0:26:15 lr: 0.000025 loss: 4.471140 (5.139138) time: 0.715398 data: 0.000209 max mem: 14338 Epoch: [1/30] [2850/5004] eta: 0:25:39 lr: 0.000025 loss: 4.322061 (5.126426) time: 0.715049 data: 0.000180 max mem: 14338 Epoch: [1/30] [2900/5004] eta: 0:25:03 lr: 0.000025 loss: 4.354888 (5.113112) time: 0.715926 data: 0.000223 max mem: 14338 Epoch: [1/30] [2950/5004] eta: 0:24:27 lr: 0.000025 loss: 4.315209 (5.099442) time: 0.709070 data: 0.000214 max mem: 14338 Epoch: [1/30] [3000/5004] eta: 0:23:52 lr: 0.000026 loss: 4.246572 (5.085353) time: 0.715441 data: 0.000163 max mem: 14338 Epoch: [1/30] [3050/5004] eta: 0:23:16 lr: 0.000026 loss: 4.214298 (5.071445) time: 0.711255 data: 0.000166 max mem: 14338 Epoch: [1/30] [3100/5004] eta: 0:22:40 lr: 0.000026 loss: 4.178181 (5.058201) time: 0.710359 data: 0.000201 max mem: 14338 Epoch: [1/30] [3150/5004] eta: 0:22:05 lr: 0.000026 loss: 4.172019 (5.044413) time: 0.711686 data: 0.000218 max mem: 14338 Epoch: [1/30] [3200/5004] eta: 0:21:29 lr: 0.000026 loss: 4.227082 (5.031533) time: 0.725365 data: 0.000200 max mem: 14338 Epoch: [1/30] [3250/5004] eta: 0:20:53 lr: 0.000026 loss: 4.111583 (5.018311) time: 0.713771 data: 0.000211 max mem: 14338 Epoch: [1/30] [3300/5004] eta: 0:20:17 lr: 0.000027 loss: 4.096457 (5.004970) time: 0.712126 data: 0.000220 max mem: 14338 Epoch: [1/30] [3350/5004] eta: 0:19:41 lr: 0.000027 loss: 4.051335 (4.991231) time: 0.710409 data: 0.000148 max mem: 14338 Epoch: [1/30] [3400/5004] eta: 0:19:06 lr: 0.000027 loss: 4.007899 (4.977000) time: 0.712513 data: 0.000158 max mem: 14338 Epoch: [1/30] [3450/5004] eta: 0:18:30 lr: 0.000027 loss: 3.957458 (4.963173) time: 0.716536 data: 0.000224 max mem: 14338 Epoch: [1/30] [3500/5004] eta: 0:17:54 lr: 0.000027 loss: 3.907161 (4.950166) time: 0.711859 data: 0.000182 max mem: 14338 Epoch: [1/30] [3550/5004] eta: 0:17:19 lr: 0.000027 loss: 4.005911 (4.936267) time: 0.712999 data: 0.000203 max mem: 14338 Epoch: [1/30] [3600/5004] eta: 0:16:43 lr: 0.000028 loss: 3.870359 (4.922311) time: 0.720966 data: 0.000214 max mem: 14338 Epoch: [1/30] [3650/5004] eta: 0:16:07 lr: 0.000028 loss: 4.040388 (4.909165) time: 0.718963 data: 0.000236 max mem: 14338 Epoch: [1/30] [3700/5004] eta: 0:15:31 lr: 0.000028 loss: 3.886157 (4.895447) time: 0.714692 data: 0.000169 max mem: 14338 Epoch: [1/30] [3750/5004] eta: 0:14:56 lr: 0.000028 loss: 3.915249 (4.882282) time: 0.715435 data: 0.000232 max mem: 14338 Epoch: [1/30] [3800/5004] eta: 0:14:20 lr: 0.000028 loss: 3.804940 (4.868499) time: 0.720811 data: 0.000217 max mem: 14338 Epoch: [1/30] [3850/5004] eta: 0:13:44 lr: 0.000028 loss: 3.785198 (4.854737) time: 0.717040 data: 0.000225 max mem: 14338 Epoch: [1/30] [3900/5004] eta: 0:13:09 lr: 0.000029 loss: 3.813251 (4.841361) time: 0.710575 data: 0.000222 max mem: 14338 Epoch: [1/30] [3950/5004] eta: 0:12:33 lr: 0.000029 loss: 3.674750 (4.827024) time: 0.713626 data: 0.000214 max mem: 14338 Epoch: [1/30] [4000/5004] eta: 0:11:57 lr: 0.000029 loss: 3.756323 (4.814214) time: 0.712005 data: 0.000164 max mem: 14338 Epoch: [1/30] [4050/5004] eta: 0:11:21 lr: 0.000029 loss: 3.683583 (4.800694) time: 0.712596 data: 0.000168 max mem: 14338 Epoch: [1/30] [4100/5004] eta: 0:10:45 lr: 0.000029 loss: 3.736003 (4.787624) time: 0.712787 data: 0.000232 max mem: 14338 Epoch: [1/30] [4150/5004] eta: 0:10:10 lr: 0.000029 loss: 3.595948 (4.774231) time: 0.715808 data: 0.000188 max mem: 14338 Epoch: [1/30] [4200/5004] eta: 0:09:34 lr: 0.000029 loss: 3.657287 (4.761317) time: 0.712843 data: 0.000197 max mem: 14338 Epoch: [1/30] [4250/5004] eta: 0:08:58 lr: 0.000030 loss: 3.642378 (4.748385) time: 0.716357 data: 0.000211 max mem: 14338 Epoch: [1/30] [4300/5004] eta: 0:08:23 lr: 0.000030 loss: 3.471499 (4.734435) time: 0.709569 data: 0.000226 max mem: 14338 Epoch: [1/30] [4350/5004] eta: 0:07:47 lr: 0.000030 loss: 3.403534 (4.720684) time: 0.709796 data: 0.000179 max mem: 14338 Epoch: [1/30] [4400/5004] eta: 0:07:11 lr: 0.000030 loss: 3.534563 (4.707164) time: 0.711825 data: 0.000153 max mem: 14338 Epoch: [1/30] [4450/5004] eta: 0:06:35 lr: 0.000030 loss: 3.469881 (4.693601) time: 0.715898 data: 0.000221 max mem: 14338 Epoch: [1/30] [4500/5004] eta: 0:06:00 lr: 0.000030 loss: 3.427865 (4.679944) time: 0.709703 data: 0.000212 max mem: 14338 Epoch: [1/30] [4550/5004] eta: 0:05:24 lr: 0.000031 loss: 3.339164 (4.665791) time: 0.714038 data: 0.000241 max mem: 14338 Epoch: [1/30] [4600/5004] eta: 0:04:48 lr: 0.000031 loss: 3.262854 (4.651947) time: 0.715252 data: 0.000228 max mem: 14338 Epoch: [1/30] [4650/5004] eta: 0:04:12 lr: 0.000031 loss: 3.364900 (4.638349) time: 0.717753 data: 0.000228 max mem: 14338 Epoch: [1/30] [4700/5004] eta: 0:03:37 lr: 0.000031 loss: 3.276072 (4.624435) time: 0.712014 data: 0.000168 max mem: 14338 Epoch: [1/30] [4750/5004] eta: 0:03:01 lr: 0.000031 loss: 3.203974 (4.610540) time: 0.709550 data: 0.000173 max mem: 14338 Epoch: [1/30] [4800/5004] eta: 0:02:25 lr: 0.000031 loss: 3.282131 (4.596864) time: 0.710893 data: 0.000181 max mem: 14338 Epoch: [1/30] [4850/5004] eta: 0:01:50 lr: 0.000032 loss: 3.316400 (4.583540) time: 0.713366 data: 0.000223 max mem: 14338 Epoch: [1/30] [4900/5004] eta: 0:01:14 lr: 0.000032 loss: 3.157434 (4.569404) time: 0.711998 data: 0.000218 max mem: 14338 Epoch: [1/30] [4950/5004] eta: 0:00:38 lr: 0.000032 loss: 3.036618 (4.555346) time: 0.711371 data: 0.000247 max mem: 14338 Epoch: [1/30] [5000/5004] eta: 0:00:02 lr: 0.000032 loss: 3.274217 (4.541934) time: 0.714091 data: 0.000836 max mem: 14338 Epoch: [1/30] [5003/5004] eta: 0:00:00 lr: 0.000032 loss: 3.241043 (4.541113) time: 0.706377 data: 0.000830 max mem: 14338 Epoch: [1/30] Total time: 0:59:35 (0.714535 s / it) Averaged stats: lr: 0.000032 loss: 3.241043 (4.538030) Test: [ 0/196] eta: 0:05:07 loss: 2.034851 (2.034851) acc1: 68.750000 (68.750000) acc5: 100.000000 (100.000000) time: 1.569167 data: 1.181772 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 2.116486 (2.115387) acc1: 75.000000 (76.136364) acc5: 93.750000 (93.750000) time: 0.404607 data: 0.107556 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 2.116486 (2.113912) acc1: 75.000000 (77.678571) acc5: 93.750000 (94.047619) time: 0.287642 data: 0.000127 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 2.108582 (2.108319) acc1: 75.000000 (76.612903) acc5: 93.750000 (94.153226) time: 0.287463 data: 0.000121 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 2.087708 (2.109059) acc1: 75.000000 (76.676829) acc5: 93.750000 (94.207317) time: 0.288210 data: 0.000143 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 2.074951 (2.103451) acc1: 81.250000 (77.696078) acc5: 93.750000 (94.240196) time: 0.289128 data: 0.000158 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 2.097107 (2.102964) acc1: 81.250000 (77.766393) acc5: 93.750000 (94.262295) time: 0.289652 data: 0.000158 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 2.118774 (2.109256) acc1: 75.000000 (76.584507) acc5: 93.750000 (94.278169) time: 0.294297 data: 0.000143 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 2.109436 (2.111173) acc1: 75.000000 (76.620370) acc5: 93.750000 (94.444444) time: 0.293600 data: 0.000122 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 2.081299 (2.119676) acc1: 75.000000 (76.167582) acc5: 93.750000 (94.299451) time: 0.288333 data: 0.000135 max mem: 14338 Test: [100/196] eta: 0:00:29 loss: 2.063690 (2.106975) acc1: 75.000000 (76.794554) acc5: 93.750000 (94.554455) time: 0.289407 data: 0.000149 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 2.001343 (2.101883) acc1: 81.250000 (76.745495) acc5: 93.750000 (94.594595) time: 0.291435 data: 0.000140 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 2.066040 (2.097671) acc1: 81.250000 (77.014463) acc5: 93.750000 (94.576446) time: 0.291094 data: 0.000141 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 2.153870 (2.111618) acc1: 75.000000 (76.574427) acc5: 93.750000 (94.370229) time: 0.290439 data: 0.000148 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 2.153870 (2.110623) acc1: 75.000000 (76.950355) acc5: 93.750000 (94.414894) time: 0.290768 data: 0.000150 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 2.038696 (2.111802) acc1: 81.250000 (76.862583) acc5: 93.750000 (94.370861) time: 0.290603 data: 0.000157 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 2.120789 (2.120173) acc1: 75.000000 (76.591615) acc5: 93.750000 (94.177019) time: 0.290610 data: 0.000144 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 2.111877 (2.114853) acc1: 75.000000 (76.827485) acc5: 93.750000 (94.188596) time: 0.290618 data: 0.000147 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 1.953033 (2.118282) acc1: 75.000000 (76.553867) acc5: 100.000000 (94.164365) time: 0.289810 data: 0.000171 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 1.994202 (2.112859) acc1: 81.250000 (76.996073) acc5: 100.000000 (94.339005) time: 0.286699 data: 0.000125 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 2.070496 (2.115283) acc1: 81.250000 (76.992000) acc5: 93.750000 (94.240000) time: 0.276830 data: 0.000105 max mem: 14338 Test: Total time: 0:00:58 (0.296711 s / it) * Acc@1 76.154 Acc@5 94.242 loss 2.134 Max accuracy: 76.15% Epoch: [2/30] [ 0/5004] eta: 2:40:30 lr: 0.000032 loss: 3.197627 (3.197627) time: 1.924507 data: 1.196567 max mem: 14338 Epoch: [2/30] [ 50/5004] eta: 1:00:57 lr: 0.000032 loss: 3.071985 (3.182826) time: 0.718278 data: 0.000226 max mem: 14338 Epoch: [2/30] [ 100/5004] eta: 0:59:26 lr: 0.000032 loss: 2.999091 (3.142433) time: 0.713476 data: 0.000213 max mem: 14338 Epoch: [2/30] [ 150/5004] eta: 0:58:28 lr: 0.000033 loss: 3.064190 (3.149972) time: 0.715501 data: 0.000212 max mem: 14338 Epoch: [2/30] [ 200/5004] eta: 0:57:39 lr: 0.000033 loss: 3.002875 (3.132870) time: 0.713250 data: 0.000200 max mem: 14338 Epoch: [2/30] [ 250/5004] eta: 0:56:57 lr: 0.000033 loss: 2.973417 (3.120079) time: 0.712144 data: 0.000188 max mem: 14338 Epoch: [2/30] [ 300/5004] eta: 0:56:18 lr: 0.000033 loss: 3.257875 (3.129637) time: 0.711789 data: 0.000178 max mem: 14338 Epoch: [2/30] [ 350/5004] eta: 0:55:40 lr: 0.000033 loss: 2.965364 (3.115266) time: 0.713956 data: 0.000162 max mem: 14338 Epoch: [2/30] [ 400/5004] eta: 0:55:03 lr: 0.000033 loss: 2.962579 (3.099503) time: 0.713417 data: 0.000209 max mem: 14338 Epoch: [2/30] [ 450/5004] eta: 0:54:25 lr: 0.000033 loss: 2.901317 (3.081469) time: 0.712560 data: 0.000208 max mem: 14338 Epoch: [2/30] [ 500/5004] eta: 0:53:47 lr: 0.000034 loss: 2.914284 (3.069793) time: 0.717598 data: 0.000223 max mem: 14338 Epoch: [2/30] [ 550/5004] eta: 0:53:11 lr: 0.000034 loss: 2.937187 (3.054038) time: 0.717642 data: 0.000217 max mem: 14338 Epoch: [2/30] [ 600/5004] eta: 0:52:34 lr: 0.000034 loss: 2.971992 (3.040496) time: 0.713733 data: 0.000225 max mem: 14338 Epoch: [2/30] [ 650/5004] eta: 0:51:56 lr: 0.000034 loss: 2.879278 (3.031573) time: 0.715626 data: 0.000156 max mem: 14338 Epoch: [2/30] [ 700/5004] eta: 0:51:19 lr: 0.000034 loss: 2.771606 (3.019406) time: 0.708912 data: 0.000160 max mem: 14338 Epoch: [2/30] [ 750/5004] eta: 0:50:44 lr: 0.000034 loss: 2.820019 (3.010145) time: 0.714241 data: 0.000235 max mem: 14338 Epoch: [2/30] [ 800/5004] eta: 0:50:08 lr: 0.000035 loss: 2.828635 (3.000905) time: 0.710698 data: 0.000222 max mem: 14338 Epoch: [2/30] [ 850/5004] eta: 0:49:32 lr: 0.000035 loss: 2.682210 (2.987929) time: 0.713080 data: 0.000207 max mem: 14338 Epoch: [2/30] [ 900/5004] eta: 0:48:56 lr: 0.000035 loss: 2.808015 (2.979073) time: 0.710064 data: 0.000189 max mem: 14338 Epoch: [2/30] [ 950/5004] eta: 0:48:20 lr: 0.000035 loss: 2.709537 (2.969388) time: 0.717100 data: 0.000208 max mem: 14338 Epoch: [2/30] [1000/5004] eta: 0:47:44 lr: 0.000035 loss: 2.676137 (2.958012) time: 0.724701 data: 0.000156 max mem: 14338 Epoch: [2/30] [1050/5004] eta: 0:47:09 lr: 0.000035 loss: 2.699806 (2.950081) time: 0.724318 data: 0.000223 max mem: 14338 Epoch: [2/30] [1100/5004] eta: 0:46:33 lr: 0.000036 loss: 2.585644 (2.939249) time: 0.711093 data: 0.000211 max mem: 14338 Epoch: [2/30] [1150/5004] eta: 0:45:57 lr: 0.000036 loss: 2.622644 (2.931619) time: 0.711369 data: 0.000213 max mem: 14338 Epoch: [2/30] [1200/5004] eta: 0:45:21 lr: 0.000036 loss: 2.627254 (2.922160) time: 0.715152 data: 0.000192 max mem: 14338 Epoch: [2/30] [1250/5004] eta: 0:44:45 lr: 0.000036 loss: 2.675158 (2.911764) time: 0.712826 data: 0.000209 max mem: 14338 Epoch: [2/30] [1300/5004] eta: 0:44:09 lr: 0.000036 loss: 2.756760 (2.903188) time: 0.713158 data: 0.000168 max mem: 14338 Epoch: [2/30] [1350/5004] eta: 0:43:33 lr: 0.000036 loss: 2.626053 (2.893316) time: 0.709611 data: 0.000176 max mem: 14338 Epoch: [2/30] [1400/5004] eta: 0:42:57 lr: 0.000037 loss: 2.617613 (2.888018) time: 0.717065 data: 0.000219 max mem: 14338 Epoch: [2/30] [1450/5004] eta: 0:42:21 lr: 0.000037 loss: 2.619015 (2.877165) time: 0.715999 data: 0.000233 max mem: 14338 Epoch: [2/30] [1500/5004] eta: 0:41:45 lr: 0.000037 loss: 2.360996 (2.864219) time: 0.714729 data: 0.000229 max mem: 14338 Epoch: [2/30] [1550/5004] eta: 0:41:09 lr: 0.000037 loss: 2.573004 (2.854655) time: 0.715540 data: 0.000206 max mem: 14338 Epoch: [2/30] [1600/5004] eta: 0:40:34 lr: 0.000037 loss: 2.472602 (2.846158) time: 0.714412 data: 0.000199 max mem: 14338 Epoch: [2/30] [1650/5004] eta: 0:39:58 lr: 0.000037 loss: 2.511433 (2.836748) time: 0.716260 data: 0.000158 max mem: 14338 Epoch: [2/30] [1700/5004] eta: 0:39:22 lr: 0.000037 loss: 2.581904 (2.829043) time: 0.712422 data: 0.000154 max mem: 14338 Epoch: [2/30] [1750/5004] eta: 0:38:46 lr: 0.000038 loss: 2.537515 (2.820595) time: 0.711372 data: 0.000220 max mem: 14338 Epoch: [2/30] [1800/5004] eta: 0:38:10 lr: 0.000038 loss: 2.427148 (2.811241) time: 0.714337 data: 0.000204 max mem: 14338 Epoch: [2/30] [1850/5004] eta: 0:37:34 lr: 0.000038 loss: 2.485512 (2.803508) time: 0.717417 data: 0.000217 max mem: 14338 Epoch: [2/30] [1900/5004] eta: 0:36:59 lr: 0.000038 loss: 2.285897 (2.795737) time: 0.715547 data: 0.000223 max mem: 14338 Epoch: [2/30] [1950/5004] eta: 0:36:23 lr: 0.000038 loss: 2.495828 (2.787083) time: 0.713918 data: 0.000217 max mem: 14338 Epoch: [2/30] [2000/5004] eta: 0:35:47 lr: 0.000038 loss: 2.500005 (2.778240) time: 0.722988 data: 0.000157 max mem: 14338 Epoch: [2/30] [2050/5004] eta: 0:35:11 lr: 0.000039 loss: 2.576241 (2.772542) time: 0.712379 data: 0.000182 max mem: 14338 Epoch: [2/30] [2100/5004] eta: 0:34:35 lr: 0.000039 loss: 2.390655 (2.762782) time: 0.709460 data: 0.000220 max mem: 14338 Epoch: [2/30] [2150/5004] eta: 0:33:59 lr: 0.000039 loss: 2.340124 (2.754874) time: 0.712768 data: 0.000211 max mem: 14338 Epoch: [2/30] [2200/5004] eta: 0:33:24 lr: 0.000039 loss: 2.309654 (2.747416) time: 0.716008 data: 0.000189 max mem: 14338 Epoch: [2/30] [2250/5004] eta: 0:32:48 lr: 0.000039 loss: 2.320966 (2.739193) time: 0.711777 data: 0.000227 max mem: 14338 Epoch: [2/30] [2300/5004] eta: 0:32:12 lr: 0.000039 loss: 2.248067 (2.730908) time: 0.710929 data: 0.000211 max mem: 14338 Epoch: [2/30] [2350/5004] eta: 0:31:37 lr: 0.000040 loss: 2.396975 (2.722948) time: 0.712539 data: 0.000159 max mem: 14338 Epoch: [2/30] [2400/5004] eta: 0:31:01 lr: 0.000040 loss: 2.470658 (2.717214) time: 0.722017 data: 0.000196 max mem: 14338 Epoch: [2/30] [2450/5004] eta: 0:30:25 lr: 0.000040 loss: 2.249628 (2.709213) time: 0.715926 data: 0.000209 max mem: 14338 Epoch: [2/30] [2500/5004] eta: 0:29:49 lr: 0.000040 loss: 2.290320 (2.700911) time: 0.711735 data: 0.000217 max mem: 14338 Epoch: [2/30] [2550/5004] eta: 0:29:13 lr: 0.000040 loss: 2.200597 (2.694252) time: 0.709548 data: 0.000232 max mem: 14338 Epoch: [2/30] [2600/5004] eta: 0:28:38 lr: 0.000040 loss: 2.312242 (2.686928) time: 0.716839 data: 0.000233 max mem: 14338 Epoch: [2/30] [2650/5004] eta: 0:28:02 lr: 0.000041 loss: 2.196070 (2.678839) time: 0.714359 data: 0.000171 max mem: 14338 Epoch: [2/30] [2700/5004] eta: 0:27:26 lr: 0.000041 loss: 2.259862 (2.670822) time: 0.709804 data: 0.000163 max mem: 14338 Epoch: [2/30] [2750/5004] eta: 0:26:50 lr: 0.000041 loss: 2.263725 (2.663149) time: 0.712876 data: 0.000222 max mem: 14338 Epoch: [2/30] [2800/5004] eta: 0:26:15 lr: 0.000041 loss: 2.136922 (2.657212) time: 0.712018 data: 0.000208 max mem: 14338 Epoch: [2/30] [2850/5004] eta: 0:25:39 lr: 0.000041 loss: 2.331300 (2.650989) time: 0.716528 data: 0.000201 max mem: 14338 Epoch: [2/30] [2900/5004] eta: 0:25:03 lr: 0.000041 loss: 2.332015 (2.644306) time: 0.717215 data: 0.000215 max mem: 14338 Epoch: [2/30] [2950/5004] eta: 0:24:27 lr: 0.000041 loss: 2.218334 (2.637647) time: 0.718031 data: 0.000224 max mem: 14338 Epoch: [2/30] [3000/5004] eta: 0:23:52 lr: 0.000042 loss: 2.154871 (2.630416) time: 0.713564 data: 0.000153 max mem: 14338 Epoch: [2/30] [3050/5004] eta: 0:23:16 lr: 0.000042 loss: 2.246091 (2.623373) time: 0.711572 data: 0.000163 max mem: 14338 Epoch: [2/30] [3100/5004] eta: 0:22:40 lr: 0.000042 loss: 2.278490 (2.618139) time: 0.709481 data: 0.000219 max mem: 14338 Epoch: [2/30] [3150/5004] eta: 0:22:04 lr: 0.000042 loss: 2.189518 (2.612435) time: 0.711050 data: 0.000211 max mem: 14338 Epoch: [2/30] [3200/5004] eta: 0:21:29 lr: 0.000042 loss: 1.889529 (2.604195) time: 0.714451 data: 0.000228 max mem: 14338 Epoch: [2/30] [3250/5004] eta: 0:20:53 lr: 0.000042 loss: 2.135437 (2.597776) time: 0.715647 data: 0.000211 max mem: 14338 Epoch: [2/30] [3300/5004] eta: 0:20:17 lr: 0.000043 loss: 2.154506 (2.591637) time: 0.722911 data: 0.000223 max mem: 14338 Epoch: [2/30] [3350/5004] eta: 0:19:42 lr: 0.000043 loss: 2.104441 (2.586263) time: 0.717748 data: 0.000170 max mem: 14338 Epoch: [2/30] [3400/5004] eta: 0:19:06 lr: 0.000043 loss: 2.109709 (2.580581) time: 0.715022 data: 0.000160 max mem: 14338 Epoch: [2/30] [3450/5004] eta: 0:18:30 lr: 0.000043 loss: 1.979253 (2.573411) time: 0.717697 data: 0.000193 max mem: 14338 Epoch: [2/30] [3500/5004] eta: 0:17:54 lr: 0.000043 loss: 2.176429 (2.567690) time: 0.709681 data: 0.000183 max mem: 14338 Epoch: [2/30] [3550/5004] eta: 0:17:19 lr: 0.000043 loss: 2.142250 (2.562624) time: 0.713507 data: 0.000213 max mem: 14338 Epoch: [2/30] [3600/5004] eta: 0:16:43 lr: 0.000044 loss: 2.178338 (2.556912) time: 0.712176 data: 0.000214 max mem: 14338 Epoch: [2/30] [3650/5004] eta: 0:16:07 lr: 0.000044 loss: 2.285432 (2.551844) time: 0.711880 data: 0.000213 max mem: 14338 Epoch: [2/30] [3700/5004] eta: 0:15:31 lr: 0.000044 loss: 2.184600 (2.547056) time: 0.710846 data: 0.000168 max mem: 14338 Epoch: [2/30] [3750/5004] eta: 0:14:56 lr: 0.000044 loss: 2.055735 (2.541214) time: 0.715118 data: 0.000230 max mem: 14338 Epoch: [2/30] [3800/5004] eta: 0:14:20 lr: 0.000044 loss: 2.032525 (2.535588) time: 0.721867 data: 0.000208 max mem: 14338 Epoch: [2/30] [3850/5004] eta: 0:13:44 lr: 0.000044 loss: 2.202675 (2.531243) time: 0.718366 data: 0.000218 max mem: 14338 Epoch: [2/30] [3900/5004] eta: 0:13:08 lr: 0.000045 loss: 2.213596 (2.526579) time: 0.713035 data: 0.000194 max mem: 14338 Epoch: [2/30] [3950/5004] eta: 0:12:33 lr: 0.000045 loss: 1.922495 (2.520491) time: 0.711265 data: 0.000220 max mem: 14338 Epoch: [2/30] [4000/5004] eta: 0:11:57 lr: 0.000045 loss: 1.896229 (2.515609) time: 0.712324 data: 0.000166 max mem: 14338 Epoch: [2/30] [4050/5004] eta: 0:11:21 lr: 0.000045 loss: 2.103776 (2.510021) time: 0.711634 data: 0.000162 max mem: 14338 Epoch: [2/30] [4100/5004] eta: 0:10:45 lr: 0.000045 loss: 1.951912 (2.504803) time: 0.718154 data: 0.000205 max mem: 14338 Epoch: [2/30] [4150/5004] eta: 0:10:10 lr: 0.000045 loss: 1.899138 (2.499495) time: 0.713304 data: 0.000186 max mem: 14338 Epoch: [2/30] [4200/5004] eta: 0:09:34 lr: 0.000045 loss: 2.020130 (2.495247) time: 0.713710 data: 0.000218 max mem: 14338 Epoch: [2/30] [4250/5004] eta: 0:08:58 lr: 0.000046 loss: 1.889215 (2.490711) time: 0.713692 data: 0.000215 max mem: 14338 Epoch: [2/30] [4300/5004] eta: 0:08:23 lr: 0.000046 loss: 2.155294 (2.485801) time: 0.717678 data: 0.000208 max mem: 14338 Epoch: [2/30] [4350/5004] eta: 0:07:47 lr: 0.000046 loss: 2.022327 (2.480832) time: 0.715996 data: 0.000169 max mem: 14338 Epoch: [2/30] [4400/5004] eta: 0:07:11 lr: 0.000046 loss: 1.922145 (2.475664) time: 0.716315 data: 0.000152 max mem: 14338 Epoch: [2/30] [4450/5004] eta: 0:06:35 lr: 0.000046 loss: 1.945283 (2.469937) time: 0.717731 data: 0.000213 max mem: 14338 Epoch: [2/30] [4500/5004] eta: 0:06:00 lr: 0.000046 loss: 2.011614 (2.465696) time: 0.710097 data: 0.000212 max mem: 14338 Epoch: [2/30] [4550/5004] eta: 0:05:24 lr: 0.000047 loss: 2.084826 (2.461682) time: 0.708430 data: 0.000215 max mem: 14338 Epoch: [2/30] [4600/5004] eta: 0:04:48 lr: 0.000047 loss: 1.879880 (2.457146) time: 0.711363 data: 0.000204 max mem: 14338 Epoch: [2/30] [4650/5004] eta: 0:04:12 lr: 0.000047 loss: 1.855889 (2.452447) time: 0.715629 data: 0.000239 max mem: 14338 Epoch: [2/30] [4700/5004] eta: 0:03:37 lr: 0.000047 loss: 1.923953 (2.447459) time: 0.717965 data: 0.000190 max mem: 14338 Epoch: [2/30] [4750/5004] eta: 0:03:01 lr: 0.000047 loss: 2.090227 (2.443033) time: 0.716218 data: 0.000178 max mem: 14338 Epoch: [2/30] [4800/5004] eta: 0:02:25 lr: 0.000047 loss: 2.109575 (2.438938) time: 0.723145 data: 0.000174 max mem: 14338 Epoch: [2/30] [4850/5004] eta: 0:01:50 lr: 0.000048 loss: 1.858636 (2.434406) time: 0.716693 data: 0.000214 max mem: 14338 Epoch: [2/30] [4900/5004] eta: 0:01:14 lr: 0.000048 loss: 2.009443 (2.429877) time: 0.710869 data: 0.000210 max mem: 14338 Epoch: [2/30] [4950/5004] eta: 0:00:38 lr: 0.000048 loss: 2.067998 (2.426144) time: 0.710392 data: 0.000218 max mem: 14338 Epoch: [2/30] [5000/5004] eta: 0:00:02 lr: 0.000048 loss: 2.006124 (2.422116) time: 0.710172 data: 0.000873 max mem: 14338 Epoch: [2/30] [5003/5004] eta: 0:00:00 lr: 0.000048 loss: 2.009630 (2.422039) time: 0.706642 data: 0.000866 max mem: 14338 Epoch: [2/30] Total time: 0:59:36 (0.714646 s / it) Averaged stats: lr: 0.000048 loss: 2.009630 (2.424513) Test: [ 0/196] eta: 0:05:00 loss: 0.521965 (0.521965) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.531875 data: 1.074470 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.698284 (0.720786) acc1: 87.500000 (88.068182) acc5: 100.000000 (96.590909) time: 0.400683 data: 0.097801 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.748192 (0.727227) acc1: 87.500000 (86.904762) acc5: 93.750000 (96.428571) time: 0.288266 data: 0.000130 max mem: 14338 Test: [ 30/196] eta: 0:00:55 loss: 0.690575 (0.691839) acc1: 87.500000 (86.693548) acc5: 100.000000 (96.975806) time: 0.295104 data: 0.000124 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.584332 (0.690009) acc1: 87.500000 (85.670732) acc5: 100.000000 (96.798780) time: 0.294508 data: 0.000139 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.616028 (0.701545) acc1: 81.250000 (85.539216) acc5: 100.000000 (96.323529) time: 0.287095 data: 0.000143 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.748806 (0.717535) acc1: 81.250000 (85.143443) acc5: 93.750000 (96.311475) time: 0.286469 data: 0.000140 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.753448 (0.725047) acc1: 81.250000 (84.507042) acc5: 100.000000 (96.566901) time: 0.287132 data: 0.000141 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.727322 (0.725321) acc1: 87.500000 (84.567901) acc5: 100.000000 (96.604938) time: 0.286747 data: 0.000120 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.732548 (0.748150) acc1: 81.250000 (83.928571) acc5: 93.750000 (96.497253) time: 0.286374 data: 0.000125 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.732548 (0.734126) acc1: 81.250000 (84.158416) acc5: 100.000000 (96.720297) time: 0.286618 data: 0.000131 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.559330 (0.721380) acc1: 81.250000 (84.009009) acc5: 100.000000 (96.903153) time: 0.286203 data: 0.000124 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.635582 (0.717616) acc1: 81.250000 (84.039256) acc5: 100.000000 (96.849174) time: 0.286227 data: 0.000133 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.770348 (0.731336) acc1: 81.250000 (83.635496) acc5: 93.750000 (96.755725) time: 0.287001 data: 0.000127 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.787491 (0.731306) acc1: 81.250000 (83.687943) acc5: 93.750000 (96.675532) time: 0.287642 data: 0.000133 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.688190 (0.734315) acc1: 87.500000 (83.567881) acc5: 100.000000 (96.605960) time: 0.287719 data: 0.000132 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.688190 (0.738538) acc1: 75.000000 (83.385093) acc5: 100.000000 (96.661491) time: 0.287261 data: 0.000118 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.562729 (0.732968) acc1: 87.500000 (83.662281) acc5: 100.000000 (96.710526) time: 0.294190 data: 0.000136 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.526318 (0.735151) acc1: 87.500000 (83.494475) acc5: 100.000000 (96.685083) time: 0.294006 data: 0.000151 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.526318 (0.726010) acc1: 81.250000 (83.769634) acc5: 100.000000 (96.727749) time: 0.283649 data: 0.000111 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.659073 (0.734516) acc1: 81.250000 (83.520000) acc5: 100.000000 (96.704000) time: 0.273484 data: 0.000102 max mem: 14338 Test: Total time: 0:00:57 (0.294771 s / it) * Acc@1 83.504 Acc@5 96.438 loss 0.753 Max accuracy: 83.50% Epoch: [3/30] [ 0/5004] eta: 2:38:48 lr: 0.000048 loss: 1.472379 (1.472379) time: 1.904204 data: 1.178666 max mem: 14338 Epoch: [3/30] [ 50/5004] eta: 1:00:53 lr: 0.000048 loss: 1.930982 (1.978715) time: 0.713170 data: 0.000191 max mem: 14338 Epoch: [3/30] [ 100/5004] eta: 0:59:23 lr: 0.000048 loss: 1.931608 (1.989103) time: 0.714970 data: 0.000216 max mem: 14338 Epoch: [3/30] [ 150/5004] eta: 0:58:31 lr: 0.000049 loss: 2.018474 (1.991385) time: 0.714343 data: 0.000202 max mem: 14338 Epoch: [3/30] [ 200/5004] eta: 0:57:45 lr: 0.000049 loss: 2.136482 (2.010449) time: 0.716205 data: 0.000204 max mem: 14338 Epoch: [3/30] [ 250/5004] eta: 0:57:04 lr: 0.000049 loss: 1.949974 (2.014997) time: 0.719158 data: 0.000186 max mem: 14338 Epoch: [3/30] [ 300/5004] eta: 0:56:23 lr: 0.000049 loss: 2.007490 (2.024397) time: 0.715816 data: 0.000169 max mem: 14338 Epoch: [3/30] [ 350/5004] eta: 0:55:42 lr: 0.000049 loss: 2.050694 (2.027821) time: 0.711902 data: 0.000165 max mem: 14338 Epoch: [3/30] [ 400/5004] eta: 0:55:04 lr: 0.000049 loss: 1.807955 (2.017525) time: 0.712507 data: 0.000209 max mem: 14338 Epoch: [3/30] [ 450/5004] eta: 0:54:28 lr: 0.000049 loss: 1.925204 (2.012776) time: 0.720762 data: 0.000210 max mem: 14338 Epoch: [3/30] [ 500/5004] eta: 0:53:51 lr: 0.000050 loss: 1.923161 (2.000731) time: 0.711267 data: 0.000215 max mem: 14338 Epoch: [3/30] [ 550/5004] eta: 0:53:15 lr: 0.000050 loss: 1.966613 (2.007312) time: 0.710106 data: 0.000218 max mem: 14338 Epoch: [3/30] [ 600/5004] eta: 0:52:38 lr: 0.000050 loss: 1.837339 (2.002523) time: 0.711051 data: 0.000207 max mem: 14338 Epoch: [3/30] [ 650/5004] eta: 0:52:01 lr: 0.000050 loss: 1.900342 (1.998182) time: 0.714964 data: 0.000163 max mem: 14338 Epoch: [3/30] [ 700/5004] eta: 0:51:24 lr: 0.000050 loss: 2.008621 (1.999197) time: 0.714426 data: 0.000156 max mem: 14338 Epoch: [3/30] [ 750/5004] eta: 0:50:47 lr: 0.000050 loss: 1.809376 (1.988309) time: 0.718132 data: 0.000229 max mem: 14338 Epoch: [3/30] [ 800/5004] eta: 0:50:11 lr: 0.000051 loss: 1.831545 (1.982310) time: 0.713212 data: 0.000200 max mem: 14338 Epoch: [3/30] [ 850/5004] eta: 0:49:34 lr: 0.000051 loss: 2.097471 (1.987460) time: 0.713363 data: 0.000202 max mem: 14338 Epoch: [3/30] [ 900/5004] eta: 0:48:58 lr: 0.000051 loss: 1.897715 (1.983306) time: 0.709963 data: 0.000177 max mem: 14338 Epoch: [3/30] [ 950/5004] eta: 0:48:22 lr: 0.000051 loss: 1.994326 (1.986194) time: 0.713795 data: 0.000196 max mem: 14338 Epoch: [3/30] [1000/5004] eta: 0:47:46 lr: 0.000051 loss: 1.941180 (1.986726) time: 0.718169 data: 0.000154 max mem: 14338 Epoch: [3/30] [1050/5004] eta: 0:47:10 lr: 0.000051 loss: 1.954224 (1.990698) time: 0.712653 data: 0.000209 max mem: 14338 Epoch: [3/30] [1100/5004] eta: 0:46:35 lr: 0.000052 loss: 1.863669 (1.988523) time: 0.716740 data: 0.000210 max mem: 14338 Epoch: [3/30] [1150/5004] eta: 0:45:59 lr: 0.000052 loss: 1.946605 (1.983798) time: 0.720491 data: 0.000213 max mem: 14338 Epoch: [3/30] [1200/5004] eta: 0:45:23 lr: 0.000052 loss: 1.816768 (1.980639) time: 0.720216 data: 0.000202 max mem: 14338 Epoch: [3/30] [1250/5004] eta: 0:44:48 lr: 0.000052 loss: 2.002321 (1.980280) time: 0.720992 data: 0.000199 max mem: 14338 Epoch: [3/30] [1300/5004] eta: 0:44:11 lr: 0.000052 loss: 1.942132 (1.981287) time: 0.710804 data: 0.000188 max mem: 14338 Epoch: [3/30] [1350/5004] eta: 0:43:35 lr: 0.000052 loss: 1.895898 (1.982366) time: 0.710923 data: 0.000173 max mem: 14338 Epoch: [3/30] [1400/5004] eta: 0:42:59 lr: 0.000053 loss: 1.836591 (1.981626) time: 0.711974 data: 0.000204 max mem: 14338 Epoch: [3/30] [1450/5004] eta: 0:42:23 lr: 0.000053 loss: 1.919665 (1.983597) time: 0.711690 data: 0.000222 max mem: 14338 Epoch: [3/30] [1500/5004] eta: 0:41:47 lr: 0.000053 loss: 1.942521 (1.982590) time: 0.712074 data: 0.000210 max mem: 14338 Epoch: [3/30] [1550/5004] eta: 0:41:11 lr: 0.000053 loss: 1.927495 (1.980011) time: 0.713986 data: 0.000240 max mem: 14338 Epoch: [3/30] [1600/5004] eta: 0:40:35 lr: 0.000053 loss: 1.813312 (1.976281) time: 0.715991 data: 0.000198 max mem: 14338 Epoch: [3/30] [1650/5004] eta: 0:39:59 lr: 0.000053 loss: 1.907808 (1.974947) time: 0.713307 data: 0.000155 max mem: 14338 Epoch: [3/30] [1700/5004] eta: 0:39:23 lr: 0.000053 loss: 2.004360 (1.975723) time: 0.717156 data: 0.000175 max mem: 14338 Epoch: [3/30] [1750/5004] eta: 0:38:47 lr: 0.000054 loss: 1.732830 (1.974328) time: 0.717518 data: 0.000213 max mem: 14338 Epoch: [3/30] [1800/5004] eta: 0:38:12 lr: 0.000054 loss: 1.927938 (1.974379) time: 0.718049 data: 0.000198 max mem: 14338 Epoch: [3/30] [1850/5004] eta: 0:37:36 lr: 0.000054 loss: 1.881428 (1.971079) time: 0.719540 data: 0.000219 max mem: 14338 Epoch: [3/30] [1900/5004] eta: 0:37:00 lr: 0.000054 loss: 1.806996 (1.969404) time: 0.709698 data: 0.000229 max mem: 14338 Epoch: [3/30] [1950/5004] eta: 0:36:24 lr: 0.000054 loss: 1.941702 (1.968571) time: 0.709792 data: 0.000223 max mem: 14338 Epoch: [3/30] [2000/5004] eta: 0:35:48 lr: 0.000054 loss: 1.823863 (1.966178) time: 0.713032 data: 0.000158 max mem: 14338 Epoch: [3/30] [2050/5004] eta: 0:35:12 lr: 0.000055 loss: 1.758587 (1.964264) time: 0.713252 data: 0.000180 max mem: 14338 Epoch: [3/30] [2100/5004] eta: 0:34:36 lr: 0.000055 loss: 1.949240 (1.963995) time: 0.714748 data: 0.000217 max mem: 14338 Epoch: [3/30] [2150/5004] eta: 0:34:01 lr: 0.000055 loss: 1.928837 (1.962277) time: 0.715285 data: 0.000219 max mem: 14338 Epoch: [3/30] [2200/5004] eta: 0:33:25 lr: 0.000055 loss: 1.631142 (1.959232) time: 0.715474 data: 0.000167 max mem: 14338 Epoch: [3/30] [2250/5004] eta: 0:32:49 lr: 0.000055 loss: 1.838450 (1.957286) time: 0.720505 data: 0.000219 max mem: 14338 Epoch: [3/30] [2300/5004] eta: 0:32:13 lr: 0.000055 loss: 1.935474 (1.957489) time: 0.711187 data: 0.000203 max mem: 14338 Epoch: [3/30] [2350/5004] eta: 0:31:37 lr: 0.000056 loss: 1.706067 (1.955260) time: 0.710025 data: 0.000169 max mem: 14338 Epoch: [3/30] [2400/5004] eta: 0:31:01 lr: 0.000056 loss: 1.748020 (1.952816) time: 0.714579 data: 0.000224 max mem: 14338 Epoch: [3/30] [2450/5004] eta: 0:30:26 lr: 0.000056 loss: 2.011752 (1.953429) time: 0.715086 data: 0.000215 max mem: 14338 Epoch: [3/30] [2500/5004] eta: 0:29:50 lr: 0.000056 loss: 1.929924 (1.952808) time: 0.716636 data: 0.000376 max mem: 14338 Epoch: [3/30] [2550/5004] eta: 0:29:14 lr: 0.000056 loss: 1.744324 (1.951232) time: 0.717499 data: 0.000223 max mem: 14338 Epoch: [3/30] [2600/5004] eta: 0:28:38 lr: 0.000056 loss: 1.809207 (1.950592) time: 0.718357 data: 0.000233 max mem: 14338 Epoch: [3/30] [2650/5004] eta: 0:28:03 lr: 0.000057 loss: 1.785839 (1.950596) time: 0.725253 data: 0.000163 max mem: 14338 Epoch: [3/30] [2700/5004] eta: 0:27:27 lr: 0.000057 loss: 1.905237 (1.950869) time: 0.712164 data: 0.000165 max mem: 14338 Epoch: [3/30] [2750/5004] eta: 0:26:51 lr: 0.000057 loss: 1.767054 (1.950469) time: 0.711520 data: 0.000200 max mem: 14338 Epoch: [3/30] [2800/5004] eta: 0:26:15 lr: 0.000057 loss: 1.845681 (1.949441) time: 0.714434 data: 0.000202 max mem: 14338 Epoch: [3/30] [2850/5004] eta: 0:25:40 lr: 0.000057 loss: 1.767173 (1.948411) time: 0.715982 data: 0.000187 max mem: 14338 Epoch: [3/30] [2900/5004] eta: 0:25:04 lr: 0.000057 loss: 1.817010 (1.948385) time: 0.709441 data: 0.000202 max mem: 14338 Epoch: [3/30] [2950/5004] eta: 0:24:28 lr: 0.000057 loss: 1.881577 (1.947996) time: 0.711633 data: 0.000183 max mem: 14338 Epoch: [3/30] [3000/5004] eta: 0:23:52 lr: 0.000058 loss: 1.879396 (1.947164) time: 0.716808 data: 0.000166 max mem: 14338 Epoch: [3/30] [3050/5004] eta: 0:23:17 lr: 0.000058 loss: 1.932169 (1.946211) time: 0.717765 data: 0.000164 max mem: 14338 Epoch: [3/30] [3100/5004] eta: 0:22:41 lr: 0.000058 loss: 1.807580 (1.945465) time: 0.716201 data: 0.000211 max mem: 14338 Epoch: [3/30] [3150/5004] eta: 0:22:05 lr: 0.000058 loss: 1.768700 (1.944399) time: 0.713395 data: 0.000221 max mem: 14338 Epoch: [3/30] [3200/5004] eta: 0:21:29 lr: 0.000058 loss: 2.025424 (1.944982) time: 0.720875 data: 0.000210 max mem: 14338 Epoch: [3/30] [3250/5004] eta: 0:20:54 lr: 0.000058 loss: 1.954293 (1.943594) time: 0.717995 data: 0.000204 max mem: 14338 Epoch: [3/30] [3300/5004] eta: 0:20:18 lr: 0.000059 loss: 1.743246 (1.942043) time: 0.710131 data: 0.000222 max mem: 14338 Epoch: [3/30] [3350/5004] eta: 0:19:42 lr: 0.000059 loss: 1.703971 (1.940839) time: 0.711376 data: 0.000168 max mem: 14338 Epoch: [3/30] [3400/5004] eta: 0:19:06 lr: 0.000059 loss: 1.792885 (1.940249) time: 0.712355 data: 0.000180 max mem: 14338 Epoch: [3/30] [3450/5004] eta: 0:18:31 lr: 0.000059 loss: 1.880291 (1.941145) time: 0.719090 data: 0.000209 max mem: 14338 Epoch: [3/30] [3500/5004] eta: 0:17:55 lr: 0.000059 loss: 1.751593 (1.939486) time: 0.716227 data: 0.000178 max mem: 14338 Epoch: [3/30] [3550/5004] eta: 0:17:19 lr: 0.000059 loss: 1.890571 (1.938114) time: 0.715275 data: 0.000222 max mem: 14338 Epoch: [3/30] [3600/5004] eta: 0:16:43 lr: 0.000060 loss: 1.661968 (1.937434) time: 0.718107 data: 0.000219 max mem: 14338 Epoch: [3/30] [3650/5004] eta: 0:16:08 lr: 0.000060 loss: 1.839874 (1.936430) time: 0.718836 data: 0.000224 max mem: 14338 Epoch: [3/30] [3700/5004] eta: 0:15:32 lr: 0.000060 loss: 1.988463 (1.935893) time: 0.711892 data: 0.000179 max mem: 14338 Epoch: [3/30] [3750/5004] eta: 0:14:56 lr: 0.000060 loss: 1.931897 (1.936611) time: 0.709686 data: 0.000227 max mem: 14338 Epoch: [3/30] [3800/5004] eta: 0:14:20 lr: 0.000060 loss: 1.658024 (1.935868) time: 0.716146 data: 0.000200 max mem: 14338 Epoch: [3/30] [3850/5004] eta: 0:13:45 lr: 0.000060 loss: 2.003761 (1.935293) time: 0.715365 data: 0.000214 max mem: 14338 Epoch: [3/30] [3900/5004] eta: 0:13:09 lr: 0.000061 loss: 1.731113 (1.934665) time: 0.712893 data: 0.000206 max mem: 14338 Epoch: [3/30] [3950/5004] eta: 0:12:33 lr: 0.000061 loss: 1.859590 (1.934717) time: 0.717549 data: 0.000214 max mem: 14338 Epoch: [3/30] [4000/5004] eta: 0:11:57 lr: 0.000061 loss: 1.789752 (1.932822) time: 0.714774 data: 0.000170 max mem: 14338 Epoch: [3/30] [4050/5004] eta: 0:11:22 lr: 0.000061 loss: 1.796954 (1.931337) time: 0.721819 data: 0.000169 max mem: 14338 Epoch: [3/30] [4100/5004] eta: 0:10:46 lr: 0.000061 loss: 1.731152 (1.930868) time: 0.717384 data: 0.000232 max mem: 14338 Epoch: [3/30] [4150/5004] eta: 0:10:10 lr: 0.000061 loss: 1.715576 (1.929690) time: 0.713932 data: 0.000199 max mem: 14338 Epoch: [3/30] [4200/5004] eta: 0:09:34 lr: 0.000061 loss: 1.912237 (1.929428) time: 0.711586 data: 0.000234 max mem: 14338 Epoch: [3/30] [4250/5004] eta: 0:08:59 lr: 0.000062 loss: 1.857022 (1.928671) time: 0.712379 data: 0.000202 max mem: 14338 Epoch: [3/30] [4300/5004] eta: 0:08:23 lr: 0.000062 loss: 1.934615 (1.927975) time: 0.710371 data: 0.000216 max mem: 14338 Epoch: [3/30] [4350/5004] eta: 0:07:47 lr: 0.000062 loss: 1.951418 (1.926949) time: 0.709982 data: 0.000174 max mem: 14338 Epoch: [3/30] [4400/5004] eta: 0:07:11 lr: 0.000062 loss: 1.707315 (1.925802) time: 0.716038 data: 0.000169 max mem: 14338 Epoch: [3/30] [4450/5004] eta: 0:06:36 lr: 0.000062 loss: 1.797830 (1.925338) time: 0.721155 data: 0.000218 max mem: 14338 Epoch: [3/30] [4500/5004] eta: 0:06:00 lr: 0.000062 loss: 1.830936 (1.925642) time: 0.715506 data: 0.000225 max mem: 14338 Epoch: [3/30] [4550/5004] eta: 0:05:24 lr: 0.000063 loss: 1.963627 (1.924989) time: 0.711473 data: 0.000246 max mem: 14338 Epoch: [3/30] [4600/5004] eta: 0:04:48 lr: 0.000063 loss: 1.746333 (1.924497) time: 0.712612 data: 0.000210 max mem: 14338 Epoch: [3/30] [4650/5004] eta: 0:04:13 lr: 0.000063 loss: 1.868221 (1.923759) time: 0.714245 data: 0.000226 max mem: 14338 Epoch: [3/30] [4700/5004] eta: 0:03:37 lr: 0.000063 loss: 1.898904 (1.922611) time: 0.711179 data: 0.000189 max mem: 14338 Epoch: [3/30] [4750/5004] eta: 0:03:01 lr: 0.000063 loss: 1.763362 (1.921861) time: 0.710878 data: 0.000170 max mem: 14338 Epoch: [3/30] [4800/5004] eta: 0:02:25 lr: 0.000063 loss: 1.905338 (1.921874) time: 0.712603 data: 0.000192 max mem: 14338 Epoch: [3/30] [4850/5004] eta: 0:01:50 lr: 0.000064 loss: 1.946107 (1.921029) time: 0.717391 data: 0.000222 max mem: 14338 Epoch: [3/30] [4900/5004] eta: 0:01:14 lr: 0.000064 loss: 1.852627 (1.920975) time: 0.714920 data: 0.000224 max mem: 14338 Epoch: [3/30] [4950/5004] eta: 0:00:38 lr: 0.000064 loss: 1.855849 (1.920950) time: 0.722599 data: 0.000236 max mem: 14338 Epoch: [3/30] [5000/5004] eta: 0:00:02 lr: 0.000064 loss: 1.875876 (1.920592) time: 0.718005 data: 0.000830 max mem: 14338 Epoch: [3/30] [5003/5004] eta: 0:00:00 lr: 0.000064 loss: 1.875876 (1.920441) time: 0.714277 data: 0.000819 max mem: 14338 Epoch: [3/30] Total time: 0:59:38 (0.715176 s / it) Averaged stats: lr: 0.000064 loss: 1.875876 (1.912661) Test: [ 0/196] eta: 0:04:50 loss: 0.425079 (0.425079) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.481162 data: 1.110199 max mem: 14338 Test: [ 10/196] eta: 0:01:13 loss: 0.538595 (0.575865) acc1: 87.500000 (86.363636) acc5: 100.000000 (97.727273) time: 0.395243 data: 0.101057 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.555832 (0.583207) acc1: 87.500000 (86.011905) acc5: 100.000000 (97.321429) time: 0.286424 data: 0.000131 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.531948 (0.553605) acc1: 87.500000 (86.693548) acc5: 100.000000 (97.983871) time: 0.286883 data: 0.000114 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.443750 (0.558108) acc1: 87.500000 (86.280488) acc5: 100.000000 (97.865854) time: 0.287908 data: 0.000127 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.501900 (0.574830) acc1: 87.500000 (86.151961) acc5: 100.000000 (97.181373) time: 0.287635 data: 0.000128 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.596806 (0.598445) acc1: 87.500000 (85.450820) acc5: 93.750000 (97.028689) time: 0.286876 data: 0.000124 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.688614 (0.611484) acc1: 81.250000 (84.595070) acc5: 100.000000 (97.183099) time: 0.286849 data: 0.000128 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.593090 (0.615145) acc1: 81.250000 (84.567901) acc5: 100.000000 (97.145062) time: 0.287154 data: 0.000115 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.642403 (0.638212) acc1: 81.250000 (84.065934) acc5: 93.750000 (97.046703) time: 0.287962 data: 0.000121 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.665066 (0.625220) acc1: 81.250000 (84.220297) acc5: 100.000000 (97.215347) time: 0.288062 data: 0.000124 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.540493 (0.613603) acc1: 87.500000 (84.290541) acc5: 100.000000 (97.409910) time: 0.291072 data: 0.000112 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.574411 (0.610910) acc1: 87.500000 (84.452479) acc5: 100.000000 (97.365702) time: 0.290778 data: 0.000123 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.612377 (0.625821) acc1: 87.500000 (84.064885) acc5: 93.750000 (97.185115) time: 0.286697 data: 0.000125 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.641865 (0.627187) acc1: 81.250000 (84.175532) acc5: 93.750000 (97.163121) time: 0.286345 data: 0.000123 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.573046 (0.631804) acc1: 87.500000 (84.147351) acc5: 100.000000 (97.144040) time: 0.286217 data: 0.000125 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.573046 (0.636147) acc1: 81.250000 (84.083851) acc5: 100.000000 (97.166149) time: 0.286274 data: 0.000111 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.566542 (0.631242) acc1: 87.500000 (84.283626) acc5: 100.000000 (97.149123) time: 0.285891 data: 0.000121 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.428368 (0.631261) acc1: 87.500000 (84.219613) acc5: 100.000000 (97.237569) time: 0.285642 data: 0.000126 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.364861 (0.620875) acc1: 87.500000 (84.489529) acc5: 100.000000 (97.284031) time: 0.283503 data: 0.000096 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.504129 (0.630754) acc1: 87.500000 (84.384000) acc5: 100.000000 (97.248000) time: 0.273597 data: 0.000085 max mem: 14338 Test: Total time: 0:00:57 (0.293154 s / it) * Acc@1 84.442 Acc@5 96.908 loss 0.643 Max accuracy: 84.44% Epoch: [4/30] [ 0/5004] eta: 2:57:16 lr: 0.000064 loss: 1.963105 (1.963105) time: 2.125648 data: 1.394732 max mem: 14338 Epoch: [4/30] [ 50/5004] eta: 1:01:11 lr: 0.000064 loss: 1.872586 (1.842714) time: 0.717584 data: 0.000206 max mem: 14338 Epoch: [4/30] [ 100/5004] eta: 0:59:36 lr: 0.000064 loss: 1.838692 (1.850442) time: 0.714791 data: 0.000231 max mem: 14338 Epoch: [4/30] [ 150/5004] eta: 0:58:36 lr: 0.000065 loss: 1.826015 (1.851271) time: 0.709928 data: 0.000226 max mem: 14338 Epoch: [4/30] [ 200/5004] eta: 0:57:49 lr: 0.000065 loss: 1.614131 (1.843182) time: 0.713523 data: 0.000213 max mem: 14338 Epoch: [4/30] [ 250/5004] eta: 0:57:06 lr: 0.000065 loss: 1.688273 (1.829223) time: 0.714568 data: 0.000200 max mem: 14338 Epoch: [4/30] [ 300/5004] eta: 0:56:26 lr: 0.000065 loss: 1.686315 (1.823281) time: 0.712869 data: 0.000163 max mem: 14338 Epoch: [4/30] [ 350/5004] eta: 0:55:45 lr: 0.000065 loss: 1.844437 (1.834717) time: 0.712410 data: 0.000174 max mem: 14338 Epoch: [4/30] [ 400/5004] eta: 0:55:08 lr: 0.000065 loss: 2.001030 (1.843635) time: 0.722660 data: 0.000210 max mem: 14338 Epoch: [4/30] [ 450/5004] eta: 0:54:29 lr: 0.000065 loss: 1.744706 (1.841637) time: 0.715397 data: 0.000218 max mem: 14338 Epoch: [4/30] [ 500/5004] eta: 0:53:51 lr: 0.000066 loss: 1.758747 (1.844055) time: 0.712771 data: 0.000218 max mem: 14338 Epoch: [4/30] [ 550/5004] eta: 0:53:12 lr: 0.000066 loss: 1.789402 (1.847300) time: 0.709900 data: 0.000209 max mem: 14338 Epoch: [4/30] [ 600/5004] eta: 0:52:36 lr: 0.000066 loss: 1.721492 (1.848848) time: 0.714800 data: 0.000230 max mem: 14338 Epoch: [4/30] [ 650/5004] eta: 0:51:59 lr: 0.000066 loss: 1.525844 (1.838430) time: 0.715024 data: 0.000158 max mem: 14338 Epoch: [4/30] [ 700/5004] eta: 0:51:24 lr: 0.000066 loss: 1.760878 (1.837693) time: 0.715018 data: 0.000158 max mem: 14338 Epoch: [4/30] [ 750/5004] eta: 0:50:48 lr: 0.000066 loss: 1.875144 (1.838466) time: 0.712830 data: 0.000200 max mem: 14338 Epoch: [4/30] [ 800/5004] eta: 0:50:12 lr: 0.000067 loss: 1.838897 (1.838101) time: 0.718222 data: 0.000218 max mem: 14338 Epoch: [4/30] [ 850/5004] eta: 0:49:35 lr: 0.000067 loss: 1.810799 (1.836510) time: 0.714699 data: 0.000207 max mem: 14338 Epoch: [4/30] [ 900/5004] eta: 0:48:59 lr: 0.000067 loss: 1.814746 (1.836532) time: 0.717194 data: 0.000207 max mem: 14338 Epoch: [4/30] [ 950/5004] eta: 0:48:22 lr: 0.000067 loss: 1.899711 (1.839100) time: 0.712981 data: 0.000225 max mem: 14338 Epoch: [4/30] [1000/5004] eta: 0:47:46 lr: 0.000067 loss: 1.760559 (1.841689) time: 0.714603 data: 0.000153 max mem: 14338 Epoch: [4/30] [1050/5004] eta: 0:47:10 lr: 0.000067 loss: 1.754683 (1.843165) time: 0.715988 data: 0.000212 max mem: 14338 Epoch: [4/30] [1100/5004] eta: 0:46:34 lr: 0.000068 loss: 1.801675 (1.844144) time: 0.709721 data: 0.000213 max mem: 14338 Epoch: [4/30] [1150/5004] eta: 0:45:58 lr: 0.000068 loss: 1.634319 (1.843563) time: 0.710896 data: 0.000224 max mem: 14338 Epoch: [4/30] [1200/5004] eta: 0:45:21 lr: 0.000068 loss: 1.546961 (1.839934) time: 0.714927 data: 0.000234 max mem: 14338 Epoch: [4/30] [1250/5004] eta: 0:44:45 lr: 0.000068 loss: 1.816103 (1.837760) time: 0.712775 data: 0.000213 max mem: 14338 Epoch: [4/30] [1300/5004] eta: 0:44:09 lr: 0.000068 loss: 1.700367 (1.838057) time: 0.713493 data: 0.000173 max mem: 14338 Epoch: [4/30] [1350/5004] eta: 0:43:33 lr: 0.000068 loss: 1.768719 (1.839109) time: 0.716678 data: 0.000160 max mem: 14338 Epoch: [4/30] [1400/5004] eta: 0:42:57 lr: 0.000069 loss: 1.531962 (1.836883) time: 0.714592 data: 0.000206 max mem: 14338 Epoch: [4/30] [1450/5004] eta: 0:42:21 lr: 0.000069 loss: 1.725195 (1.840168) time: 0.715418 data: 0.000218 max mem: 14338 Epoch: [4/30] [1500/5004] eta: 0:41:45 lr: 0.000069 loss: 1.668049 (1.837495) time: 0.710211 data: 0.000204 max mem: 14338 Epoch: [4/30] [1550/5004] eta: 0:41:10 lr: 0.000069 loss: 1.681602 (1.835956) time: 0.717163 data: 0.000187 max mem: 14338 Epoch: [4/30] [1600/5004] eta: 0:40:34 lr: 0.000069 loss: 1.678653 (1.832155) time: 0.711678 data: 0.000191 max mem: 14338 Epoch: [4/30] [1650/5004] eta: 0:39:58 lr: 0.000069 loss: 1.621413 (1.830970) time: 0.711553 data: 0.000148 max mem: 14338 Epoch: [4/30] [1700/5004] eta: 0:39:22 lr: 0.000069 loss: 1.767374 (1.831974) time: 0.709806 data: 0.000173 max mem: 14338 Epoch: [4/30] [1750/5004] eta: 0:38:46 lr: 0.000070 loss: 1.744769 (1.830980) time: 0.717930 data: 0.000219 max mem: 14338 Epoch: [4/30] [1800/5004] eta: 0:38:11 lr: 0.000070 loss: 1.746312 (1.831200) time: 0.716353 data: 0.000190 max mem: 14338 Epoch: [4/30] [1850/5004] eta: 0:37:35 lr: 0.000070 loss: 1.731482 (1.830039) time: 0.721224 data: 0.000217 max mem: 14338 Epoch: [4/30] [1900/5004] eta: 0:36:59 lr: 0.000070 loss: 1.676550 (1.831015) time: 0.709578 data: 0.000225 max mem: 14338 Epoch: [4/30] [1950/5004] eta: 0:36:23 lr: 0.000070 loss: 1.655558 (1.829491) time: 0.709535 data: 0.000209 max mem: 14338 Epoch: [4/30] [2000/5004] eta: 0:35:47 lr: 0.000070 loss: 1.652724 (1.830172) time: 0.715278 data: 0.000160 max mem: 14338 Epoch: [4/30] [2050/5004] eta: 0:35:11 lr: 0.000071 loss: 1.921136 (1.831013) time: 0.712451 data: 0.000171 max mem: 14338 Epoch: [4/30] [2100/5004] eta: 0:34:36 lr: 0.000071 loss: 1.705720 (1.828812) time: 0.710772 data: 0.000194 max mem: 14338 Epoch: [4/30] [2150/5004] eta: 0:34:00 lr: 0.000071 loss: 1.839913 (1.829476) time: 0.709468 data: 0.000220 max mem: 14338 Epoch: [4/30] [2200/5004] eta: 0:33:24 lr: 0.000071 loss: 1.691820 (1.829437) time: 0.717704 data: 0.000191 max mem: 14338 Epoch: [4/30] [2250/5004] eta: 0:32:48 lr: 0.000071 loss: 1.753697 (1.830058) time: 0.720965 data: 0.000210 max mem: 14338 Epoch: [4/30] [2300/5004] eta: 0:32:13 lr: 0.000071 loss: 1.980252 (1.830742) time: 0.714865 data: 0.000233 max mem: 14338 Epoch: [4/30] [2350/5004] eta: 0:31:37 lr: 0.000072 loss: 1.689514 (1.830139) time: 0.713957 data: 0.000170 max mem: 14338 Epoch: [4/30] [2400/5004] eta: 0:31:01 lr: 0.000072 loss: 1.742181 (1.829615) time: 0.712337 data: 0.000203 max mem: 14338 Epoch: [4/30] [2450/5004] eta: 0:30:25 lr: 0.000072 loss: 1.719815 (1.829045) time: 0.712572 data: 0.000213 max mem: 14338 Epoch: [4/30] [2500/5004] eta: 0:29:49 lr: 0.000072 loss: 1.661935 (1.828951) time: 0.714696 data: 0.000229 max mem: 14338 Epoch: [4/30] [2550/5004] eta: 0:29:14 lr: 0.000072 loss: 1.840586 (1.827881) time: 0.710762 data: 0.000217 max mem: 14338 Epoch: [4/30] [2600/5004] eta: 0:28:38 lr: 0.000072 loss: 1.585503 (1.827373) time: 0.715300 data: 0.000230 max mem: 14338 Epoch: [4/30] [2650/5004] eta: 0:28:02 lr: 0.000073 loss: 1.666278 (1.825267) time: 0.717032 data: 0.000168 max mem: 14338 Epoch: [4/30] [2700/5004] eta: 0:27:27 lr: 0.000073 loss: 1.693954 (1.824150) time: 0.711711 data: 0.000175 max mem: 14338 Epoch: [4/30] [2750/5004] eta: 0:26:51 lr: 0.000073 loss: 1.673810 (1.822816) time: 0.714617 data: 0.000223 max mem: 14338 Epoch: [4/30] [2800/5004] eta: 0:26:15 lr: 0.000073 loss: 1.812744 (1.822442) time: 0.717176 data: 0.000223 max mem: 14338 Epoch: [4/30] [2850/5004] eta: 0:25:39 lr: 0.000073 loss: 1.822923 (1.823871) time: 0.718747 data: 0.000192 max mem: 14338 Epoch: [4/30] [2900/5004] eta: 0:25:03 lr: 0.000073 loss: 1.967314 (1.826016) time: 0.711287 data: 0.000215 max mem: 14338 Epoch: [4/30] [2950/5004] eta: 0:24:27 lr: 0.000073 loss: 1.763600 (1.825800) time: 0.709413 data: 0.000216 max mem: 14338 Epoch: [4/30] [3000/5004] eta: 0:23:52 lr: 0.000074 loss: 1.667954 (1.825049) time: 0.715480 data: 0.000153 max mem: 14338 Epoch: [4/30] [3050/5004] eta: 0:23:16 lr: 0.000074 loss: 1.703204 (1.823368) time: 0.714877 data: 0.000156 max mem: 14338 Epoch: [4/30] [3100/5004] eta: 0:22:40 lr: 0.000074 loss: 1.831585 (1.824016) time: 0.711997 data: 0.000219 max mem: 14338 Epoch: [4/30] [3150/5004] eta: 0:22:05 lr: 0.000074 loss: 1.669392 (1.823362) time: 0.714134 data: 0.000222 max mem: 14338 Epoch: [4/30] [3200/5004] eta: 0:21:29 lr: 0.000074 loss: 1.705956 (1.823457) time: 0.720132 data: 0.000224 max mem: 14338 Epoch: [4/30] [3250/5004] eta: 0:20:53 lr: 0.000074 loss: 1.822549 (1.822579) time: 0.720475 data: 0.000209 max mem: 14338 Epoch: [4/30] [3300/5004] eta: 0:20:17 lr: 0.000075 loss: 1.766116 (1.821410) time: 0.712973 data: 0.000208 max mem: 14338 Epoch: [4/30] [3350/5004] eta: 0:19:42 lr: 0.000075 loss: 1.850383 (1.821842) time: 0.709157 data: 0.000164 max mem: 14338 Epoch: [4/30] [3400/5004] eta: 0:19:06 lr: 0.000075 loss: 1.722302 (1.820523) time: 0.713274 data: 0.000164 max mem: 14338 Epoch: [4/30] [3450/5004] eta: 0:18:30 lr: 0.000075 loss: 1.836478 (1.820552) time: 0.716984 data: 0.000221 max mem: 14338 Epoch: [4/30] [3500/5004] eta: 0:17:54 lr: 0.000075 loss: 1.747066 (1.821274) time: 0.710265 data: 0.000188 max mem: 14338 Epoch: [4/30] [3550/5004] eta: 0:17:19 lr: 0.000075 loss: 1.860980 (1.821701) time: 0.712455 data: 0.000228 max mem: 14338 Epoch: [4/30] [3600/5004] eta: 0:16:43 lr: 0.000076 loss: 1.621094 (1.821226) time: 0.716101 data: 0.000215 max mem: 14338 Epoch: [4/30] [3650/5004] eta: 0:16:07 lr: 0.000076 loss: 1.723202 (1.821162) time: 0.724434 data: 0.000231 max mem: 14338 Epoch: [4/30] [3700/5004] eta: 0:15:32 lr: 0.000076 loss: 1.749356 (1.820688) time: 0.716729 data: 0.000170 max mem: 14338 Epoch: [4/30] [3750/5004] eta: 0:14:56 lr: 0.000076 loss: 1.712858 (1.820635) time: 0.715320 data: 0.000219 max mem: 14338 Epoch: [4/30] [3800/5004] eta: 0:14:20 lr: 0.000076 loss: 1.656404 (1.819782) time: 0.713341 data: 0.000216 max mem: 14338 Epoch: [4/30] [3850/5004] eta: 0:13:44 lr: 0.000076 loss: 1.775677 (1.819007) time: 0.712310 data: 0.000227 max mem: 14338 Epoch: [4/30] [3900/5004] eta: 0:13:09 lr: 0.000076 loss: 1.640967 (1.818394) time: 0.711213 data: 0.000210 max mem: 14338 Epoch: [4/30] [3950/5004] eta: 0:12:33 lr: 0.000077 loss: 1.665979 (1.817975) time: 0.711665 data: 0.000224 max mem: 14338 Epoch: [4/30] [4000/5004] eta: 0:11:57 lr: 0.000077 loss: 1.682908 (1.818010) time: 0.713670 data: 0.000179 max mem: 14338 Epoch: [4/30] [4050/5004] eta: 0:11:21 lr: 0.000077 loss: 1.616287 (1.817807) time: 0.720869 data: 0.000161 max mem: 14338 Epoch: [4/30] [4100/5004] eta: 0:10:46 lr: 0.000077 loss: 1.591196 (1.817512) time: 0.717662 data: 0.000232 max mem: 14338 Epoch: [4/30] [4150/5004] eta: 0:10:10 lr: 0.000077 loss: 1.727292 (1.816968) time: 0.713969 data: 0.000186 max mem: 14338 Epoch: [4/30] [4200/5004] eta: 0:09:34 lr: 0.000077 loss: 1.558698 (1.814904) time: 0.718947 data: 0.000213 max mem: 14338 Epoch: [4/30] [4250/5004] eta: 0:08:58 lr: 0.000078 loss: 2.048512 (1.816317) time: 0.712604 data: 0.000224 max mem: 14338 Epoch: [4/30] [4300/5004] eta: 0:08:23 lr: 0.000078 loss: 1.795835 (1.816848) time: 0.715627 data: 0.000222 max mem: 14338 Epoch: [4/30] [4350/5004] eta: 0:07:47 lr: 0.000078 loss: 1.608547 (1.816005) time: 0.711051 data: 0.000162 max mem: 14338 Epoch: [4/30] [4400/5004] eta: 0:07:11 lr: 0.000078 loss: 1.606153 (1.815274) time: 0.718951 data: 0.000165 max mem: 14338 Epoch: [4/30] [4450/5004] eta: 0:06:35 lr: 0.000078 loss: 1.594722 (1.814372) time: 0.716180 data: 0.000219 max mem: 14338 Epoch: [4/30] [4500/5004] eta: 0:06:00 lr: 0.000078 loss: 1.590356 (1.813839) time: 0.712886 data: 0.000220 max mem: 14338 Epoch: [4/30] [4550/5004] eta: 0:05:24 lr: 0.000079 loss: 1.894182 (1.814607) time: 0.713999 data: 0.000227 max mem: 14338 Epoch: [4/30] [4600/5004] eta: 0:04:48 lr: 0.000079 loss: 1.687945 (1.813957) time: 0.719169 data: 0.000215 max mem: 14338 Epoch: [4/30] [4650/5004] eta: 0:04:13 lr: 0.000079 loss: 1.724699 (1.813623) time: 0.720043 data: 0.000204 max mem: 14338 Epoch: [4/30] [4700/5004] eta: 0:03:37 lr: 0.000079 loss: 1.773097 (1.813923) time: 0.712478 data: 0.000164 max mem: 14338 Epoch: [4/30] [4750/5004] eta: 0:03:01 lr: 0.000079 loss: 1.852848 (1.814351) time: 0.711710 data: 0.000169 max mem: 14338 Epoch: [4/30] [4800/5004] eta: 0:02:25 lr: 0.000079 loss: 1.699617 (1.814530) time: 0.711151 data: 0.000228 max mem: 14338 Epoch: [4/30] [4850/5004] eta: 0:01:50 lr: 0.000080 loss: 1.629515 (1.813986) time: 0.713177 data: 0.000233 max mem: 14338 Epoch: [4/30] [4900/5004] eta: 0:01:14 lr: 0.000080 loss: 1.822907 (1.813994) time: 0.716188 data: 0.000212 max mem: 14338 Epoch: [4/30] [4950/5004] eta: 0:00:38 lr: 0.000080 loss: 1.790215 (1.814337) time: 0.712807 data: 0.000208 max mem: 14338 Epoch: [4/30] [5000/5004] eta: 0:00:02 lr: 0.000075 loss: 1.734416 (1.814116) time: 0.713524 data: 0.000870 max mem: 14338 Epoch: [4/30] [5003/5004] eta: 0:00:00 lr: 0.000075 loss: 1.805951 (1.814063) time: 0.710231 data: 0.000862 max mem: 14338 Epoch: [4/30] Total time: 0:59:37 (0.714936 s / it) Averaged stats: lr: 0.000075 loss: 1.805951 (1.814101) Test: [ 0/196] eta: 0:04:56 loss: 0.412609 (0.412609) acc1: 87.500000 (87.500000) acc5: 100.000000 (100.000000) time: 1.511238 data: 1.062383 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.543348 (0.529905) acc1: 87.500000 (85.795455) acc5: 100.000000 (99.431818) time: 0.398544 data: 0.096722 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.575871 (0.545881) acc1: 87.500000 (86.309524) acc5: 100.000000 (98.511905) time: 0.287041 data: 0.000134 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.561411 (0.523206) acc1: 87.500000 (87.096774) acc5: 100.000000 (98.588710) time: 0.286773 data: 0.000113 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.423171 (0.532213) acc1: 87.500000 (86.737805) acc5: 100.000000 (98.170732) time: 0.287150 data: 0.000130 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.490206 (0.553703) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.671569) time: 0.287968 data: 0.000134 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.553162 (0.583367) acc1: 81.250000 (85.450820) acc5: 100.000000 (97.540984) time: 0.287745 data: 0.000129 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.605990 (0.600488) acc1: 81.250000 (84.859155) acc5: 100.000000 (97.711268) time: 0.295589 data: 0.000133 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.564779 (0.601417) acc1: 81.250000 (85.030864) acc5: 100.000000 (97.685185) time: 0.295682 data: 0.000117 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.592923 (0.620326) acc1: 81.250000 (84.615385) acc5: 93.750000 (97.458791) time: 0.287461 data: 0.000122 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.577122 (0.607314) acc1: 81.250000 (84.839109) acc5: 100.000000 (97.648515) time: 0.287074 data: 0.000125 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.552053 (0.599429) acc1: 81.250000 (84.628378) acc5: 100.000000 (97.804054) time: 0.286517 data: 0.000105 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.578524 (0.597312) acc1: 87.500000 (84.710744) acc5: 100.000000 (97.675620) time: 0.286210 data: 0.000120 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.621407 (0.611412) acc1: 87.500000 (84.541985) acc5: 93.750000 (97.519084) time: 0.286543 data: 0.000126 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.623642 (0.610969) acc1: 81.250000 (84.663121) acc5: 93.750000 (97.517730) time: 0.287731 data: 0.000131 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.640969 (0.617621) acc1: 81.250000 (84.519868) acc5: 100.000000 (97.475166) time: 0.288103 data: 0.000131 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.550680 (0.622122) acc1: 81.250000 (84.355590) acc5: 100.000000 (97.515528) time: 0.287164 data: 0.000113 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.519304 (0.617887) acc1: 87.500000 (84.539474) acc5: 100.000000 (97.514620) time: 0.286356 data: 0.000120 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.503082 (0.617238) acc1: 87.500000 (84.495856) acc5: 100.000000 (97.548343) time: 0.286301 data: 0.000135 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.365713 (0.607948) acc1: 87.500000 (84.685864) acc5: 100.000000 (97.578534) time: 0.283590 data: 0.000109 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.572754 (0.618141) acc1: 87.500000 (84.544000) acc5: 100.000000 (97.504000) time: 0.273522 data: 0.000096 max mem: 14338 Test: Total time: 0:00:57 (0.294045 s / it) * Acc@1 84.726 Acc@5 97.194 loss 0.632 Max accuracy: 84.73% Epoch: [5/30] [ 0/5004] eta: 2:36:13 lr: 0.000075 loss: 1.458258 (1.458258) time: 1.873183 data: 1.139714 max mem: 14338 Epoch: [5/30] [ 50/5004] eta: 1:00:53 lr: 0.000075 loss: 1.842479 (1.735029) time: 0.717301 data: 0.000186 max mem: 14338 Epoch: [5/30] [ 100/5004] eta: 0:59:23 lr: 0.000075 loss: 1.570576 (1.768443) time: 0.717900 data: 0.000242 max mem: 14338 Epoch: [5/30] [ 150/5004] eta: 0:58:27 lr: 0.000075 loss: 1.833262 (1.793216) time: 0.715544 data: 0.000224 max mem: 14338 Epoch: [5/30] [ 200/5004] eta: 0:57:40 lr: 0.000075 loss: 1.587858 (1.779816) time: 0.716495 data: 0.000287 max mem: 14338 Epoch: [5/30] [ 250/5004] eta: 0:56:59 lr: 0.000075 loss: 1.726152 (1.775312) time: 0.715127 data: 0.000189 max mem: 14338 Epoch: [5/30] [ 300/5004] eta: 0:56:20 lr: 0.000075 loss: 1.685932 (1.777230) time: 0.709872 data: 0.000159 max mem: 14338 Epoch: [5/30] [ 350/5004] eta: 0:55:41 lr: 0.000075 loss: 1.785039 (1.769349) time: 0.711654 data: 0.000174 max mem: 14338 Epoch: [5/30] [ 400/5004] eta: 0:55:07 lr: 0.000075 loss: 1.601398 (1.768932) time: 0.723396 data: 0.000189 max mem: 14338 Epoch: [5/30] [ 450/5004] eta: 0:54:28 lr: 0.000075 loss: 1.608125 (1.772832) time: 0.712135 data: 0.000214 max mem: 14338 Epoch: [5/30] [ 500/5004] eta: 0:53:50 lr: 0.000074 loss: 1.793166 (1.777243) time: 0.715089 data: 0.000230 max mem: 14338 Epoch: [5/30] [ 550/5004] eta: 0:53:14 lr: 0.000074 loss: 1.769802 (1.773815) time: 0.716443 data: 0.000221 max mem: 14338 Epoch: [5/30] [ 600/5004] eta: 0:52:37 lr: 0.000074 loss: 1.805784 (1.777533) time: 0.717589 data: 0.000226 max mem: 14338 Epoch: [5/30] [ 650/5004] eta: 0:52:01 lr: 0.000074 loss: 1.735773 (1.782127) time: 0.722396 data: 0.000171 max mem: 14338 Epoch: [5/30] [ 700/5004] eta: 0:51:24 lr: 0.000074 loss: 1.805378 (1.783208) time: 0.714925 data: 0.000179 max mem: 14338 Epoch: [5/30] [ 750/5004] eta: 0:50:48 lr: 0.000074 loss: 1.613538 (1.779708) time: 0.710542 data: 0.000220 max mem: 14338 Epoch: [5/30] [ 800/5004] eta: 0:50:12 lr: 0.000074 loss: 1.770687 (1.781917) time: 0.715626 data: 0.000225 max mem: 14338 Epoch: [5/30] [ 850/5004] eta: 0:49:35 lr: 0.000074 loss: 1.700196 (1.783531) time: 0.714454 data: 0.000212 max mem: 14338 Epoch: [5/30] [ 900/5004] eta: 0:48:58 lr: 0.000074 loss: 1.648745 (1.782887) time: 0.710770 data: 0.000187 max mem: 14338 Epoch: [5/30] [ 950/5004] eta: 0:48:22 lr: 0.000074 loss: 1.828772 (1.781986) time: 0.715138 data: 0.000230 max mem: 14338 Epoch: [5/30] [1000/5004] eta: 0:47:46 lr: 0.000074 loss: 1.643693 (1.780410) time: 0.720377 data: 0.000161 max mem: 14338 Epoch: [5/30] [1050/5004] eta: 0:47:10 lr: 0.000074 loss: 1.825243 (1.784006) time: 0.719510 data: 0.000208 max mem: 14338 Epoch: [5/30] [1100/5004] eta: 0:46:34 lr: 0.000074 loss: 1.707487 (1.787738) time: 0.712565 data: 0.000215 max mem: 14338 Epoch: [5/30] [1150/5004] eta: 0:45:58 lr: 0.000074 loss: 1.706883 (1.786185) time: 0.715824 data: 0.000219 max mem: 14338 Epoch: [5/30] [1200/5004] eta: 0:45:22 lr: 0.000074 loss: 1.668390 (1.786537) time: 0.716257 data: 0.000232 max mem: 14338 Epoch: [5/30] [1250/5004] eta: 0:44:46 lr: 0.000074 loss: 1.743317 (1.786169) time: 0.712938 data: 0.000215 max mem: 14338 Epoch: [5/30] [1300/5004] eta: 0:44:10 lr: 0.000074 loss: 1.733383 (1.786878) time: 0.710081 data: 0.000159 max mem: 14338 Epoch: [5/30] [1350/5004] eta: 0:43:35 lr: 0.000074 loss: 1.787797 (1.787984) time: 0.709640 data: 0.000166 max mem: 14338 Epoch: [5/30] [1400/5004] eta: 0:42:59 lr: 0.000074 loss: 1.698491 (1.788497) time: 0.716581 data: 0.000234 max mem: 14338 Epoch: [5/30] [1450/5004] eta: 0:42:23 lr: 0.000074 loss: 1.703721 (1.789155) time: 0.717808 data: 0.000223 max mem: 14338 Epoch: [5/30] [1500/5004] eta: 0:41:47 lr: 0.000074 loss: 1.632881 (1.789231) time: 0.715005 data: 0.000224 max mem: 14338 Epoch: [5/30] [1550/5004] eta: 0:41:11 lr: 0.000074 loss: 1.596883 (1.786011) time: 0.715431 data: 0.000203 max mem: 14338 Epoch: [5/30] [1600/5004] eta: 0:40:35 lr: 0.000074 loss: 1.588282 (1.785239) time: 0.716051 data: 0.000211 max mem: 14338 Epoch: [5/30] [1650/5004] eta: 0:40:00 lr: 0.000074 loss: 1.664149 (1.783877) time: 0.718476 data: 0.000180 max mem: 14338 Epoch: [5/30] [1700/5004] eta: 0:39:24 lr: 0.000074 loss: 1.597993 (1.782553) time: 0.712043 data: 0.000173 max mem: 14338 Epoch: [5/30] [1750/5004] eta: 0:38:48 lr: 0.000074 loss: 1.625999 (1.782160) time: 0.712359 data: 0.000205 max mem: 14338 Epoch: [5/30] [1800/5004] eta: 0:38:12 lr: 0.000074 loss: 1.722583 (1.783547) time: 0.712883 data: 0.000221 max mem: 14338 Epoch: [5/30] [1850/5004] eta: 0:37:36 lr: 0.000074 loss: 1.764369 (1.782172) time: 0.713264 data: 0.000215 max mem: 14338 Epoch: [5/30] [1900/5004] eta: 0:37:00 lr: 0.000074 loss: 1.862278 (1.783138) time: 0.712089 data: 0.000234 max mem: 14338 Epoch: [5/30] [1950/5004] eta: 0:36:24 lr: 0.000074 loss: 1.622761 (1.782332) time: 0.714928 data: 0.000226 max mem: 14338 Epoch: [5/30] [2000/5004] eta: 0:35:49 lr: 0.000074 loss: 1.643106 (1.782192) time: 0.722522 data: 0.000176 max mem: 14338 Epoch: [5/30] [2050/5004] eta: 0:35:13 lr: 0.000074 loss: 1.595002 (1.783065) time: 0.715588 data: 0.000168 max mem: 14338 Epoch: [5/30] [2100/5004] eta: 0:34:37 lr: 0.000074 loss: 1.750838 (1.783577) time: 0.714677 data: 0.000239 max mem: 14338 Epoch: [5/30] [2150/5004] eta: 0:34:01 lr: 0.000074 loss: 1.650056 (1.784375) time: 0.712380 data: 0.000211 max mem: 14338 Epoch: [5/30] [2200/5004] eta: 0:33:25 lr: 0.000074 loss: 1.661572 (1.783263) time: 0.712468 data: 0.000207 max mem: 14338 Epoch: [5/30] [2250/5004] eta: 0:32:49 lr: 0.000074 loss: 1.756302 (1.783641) time: 0.711438 data: 0.000237 max mem: 14338 Epoch: [5/30] [2300/5004] eta: 0:32:14 lr: 0.000074 loss: 1.606677 (1.782725) time: 0.711358 data: 0.000214 max mem: 14338 Epoch: [5/30] [2350/5004] eta: 0:31:38 lr: 0.000074 loss: 1.769950 (1.783368) time: 0.710119 data: 0.000184 max mem: 14338 Epoch: [5/30] [2400/5004] eta: 0:31:02 lr: 0.000074 loss: 1.589137 (1.783395) time: 0.720451 data: 0.000226 max mem: 14338 Epoch: [5/30] [2450/5004] eta: 0:30:26 lr: 0.000074 loss: 1.817532 (1.784036) time: 0.716989 data: 0.000228 max mem: 14338 Epoch: [5/30] [2500/5004] eta: 0:29:50 lr: 0.000074 loss: 1.703059 (1.783157) time: 0.714977 data: 0.000223 max mem: 14338 Epoch: [5/30] [2550/5004] eta: 0:29:14 lr: 0.000074 loss: 1.672751 (1.781806) time: 0.710052 data: 0.000224 max mem: 14338 Epoch: [5/30] [2600/5004] eta: 0:28:38 lr: 0.000074 loss: 1.666051 (1.781770) time: 0.713136 data: 0.000200 max mem: 14338 Epoch: [5/30] [2650/5004] eta: 0:28:03 lr: 0.000074 loss: 1.854309 (1.781396) time: 0.715218 data: 0.000161 max mem: 14338 Epoch: [5/30] [2700/5004] eta: 0:27:27 lr: 0.000074 loss: 1.719813 (1.781383) time: 0.713089 data: 0.000165 max mem: 14338 Epoch: [5/30] [2750/5004] eta: 0:26:51 lr: 0.000073 loss: 1.683109 (1.779271) time: 0.710875 data: 0.000224 max mem: 14338 Epoch: [5/30] [2800/5004] eta: 0:26:16 lr: 0.000073 loss: 1.666596 (1.778922) time: 0.718042 data: 0.000233 max mem: 14338 Epoch: [5/30] [2850/5004] eta: 0:25:40 lr: 0.000073 loss: 1.716409 (1.778646) time: 0.722165 data: 0.000199 max mem: 14338 Epoch: [5/30] [2900/5004] eta: 0:25:04 lr: 0.000073 loss: 1.655527 (1.776320) time: 0.718443 data: 0.000219 max mem: 14338 Epoch: [5/30] [2950/5004] eta: 0:24:28 lr: 0.000073 loss: 1.727965 (1.776016) time: 0.715299 data: 0.000230 max mem: 14338 Epoch: [5/30] [3000/5004] eta: 0:23:53 lr: 0.000073 loss: 1.567152 (1.775044) time: 0.717054 data: 0.000163 max mem: 14338 Epoch: [5/30] [3050/5004] eta: 0:23:17 lr: 0.000073 loss: 1.626118 (1.774815) time: 0.714793 data: 0.000175 max mem: 14338 Epoch: [5/30] [3100/5004] eta: 0:22:41 lr: 0.000073 loss: 1.647197 (1.775728) time: 0.709602 data: 0.000219 max mem: 14338 Epoch: [5/30] [3150/5004] eta: 0:22:05 lr: 0.000073 loss: 1.720327 (1.775800) time: 0.710209 data: 0.000236 max mem: 14338 Epoch: [5/30] [3200/5004] eta: 0:21:29 lr: 0.000073 loss: 1.554107 (1.774747) time: 0.713745 data: 0.000237 max mem: 14338 Epoch: [5/30] [3250/5004] eta: 0:20:54 lr: 0.000073 loss: 1.833432 (1.775786) time: 0.713229 data: 0.000220 max mem: 14338 Epoch: [5/30] [3300/5004] eta: 0:20:18 lr: 0.000073 loss: 1.557778 (1.776239) time: 0.714282 data: 0.000207 max mem: 14338 Epoch: [5/30] [3350/5004] eta: 0:19:42 lr: 0.000073 loss: 1.772947 (1.775865) time: 0.718399 data: 0.000161 max mem: 14338 Epoch: [5/30] [3400/5004] eta: 0:19:06 lr: 0.000073 loss: 1.745772 (1.777256) time: 0.719828 data: 0.000171 max mem: 14338 Epoch: [5/30] [3450/5004] eta: 0:18:31 lr: 0.000073 loss: 1.813486 (1.777131) time: 0.718536 data: 0.000213 max mem: 14338 Epoch: [5/30] [3500/5004] eta: 0:17:55 lr: 0.000073 loss: 1.940400 (1.778051) time: 0.710252 data: 0.000201 max mem: 14338 Epoch: [5/30] [3550/5004] eta: 0:17:19 lr: 0.000073 loss: 1.870205 (1.778285) time: 0.713107 data: 0.000238 max mem: 14338 Epoch: [5/30] [3600/5004] eta: 0:16:43 lr: 0.000073 loss: 1.601216 (1.776506) time: 0.713691 data: 0.000216 max mem: 14338 Epoch: [5/30] [3650/5004] eta: 0:16:08 lr: 0.000073 loss: 1.492582 (1.775439) time: 0.714471 data: 0.000213 max mem: 14338 Epoch: [5/30] [3700/5004] eta: 0:15:32 lr: 0.000073 loss: 1.639129 (1.775788) time: 0.710402 data: 0.000177 max mem: 14338 Epoch: [5/30] [3750/5004] eta: 0:14:56 lr: 0.000073 loss: 1.796587 (1.776396) time: 0.710456 data: 0.000237 max mem: 14338 Epoch: [5/30] [3800/5004] eta: 0:14:20 lr: 0.000073 loss: 1.639921 (1.776605) time: 0.721218 data: 0.000218 max mem: 14338 Epoch: [5/30] [3850/5004] eta: 0:13:45 lr: 0.000073 loss: 1.641646 (1.775397) time: 0.719314 data: 0.000221 max mem: 14338 Epoch: [5/30] [3900/5004] eta: 0:13:09 lr: 0.000073 loss: 1.622301 (1.775797) time: 0.719006 data: 0.000226 max mem: 14338 Epoch: [5/30] [3950/5004] eta: 0:12:33 lr: 0.000073 loss: 1.598087 (1.775417) time: 0.716240 data: 0.000223 max mem: 14338 Epoch: [5/30] [4000/5004] eta: 0:11:57 lr: 0.000073 loss: 1.745531 (1.776204) time: 0.714994 data: 0.000171 max mem: 14338 Epoch: [5/30] [4050/5004] eta: 0:11:22 lr: 0.000073 loss: 1.667944 (1.775197) time: 0.712442 data: 0.000165 max mem: 14338 Epoch: [5/30] [4100/5004] eta: 0:10:46 lr: 0.000073 loss: 1.729547 (1.775716) time: 0.710833 data: 0.000228 max mem: 14338 Epoch: [5/30] [4150/5004] eta: 0:10:10 lr: 0.000073 loss: 1.545823 (1.775281) time: 0.715853 data: 0.000194 max mem: 14338 Epoch: [5/30] [4200/5004] eta: 0:09:34 lr: 0.000073 loss: 1.557849 (1.775257) time: 0.715407 data: 0.000206 max mem: 14338 Epoch: [5/30] [4250/5004] eta: 0:08:59 lr: 0.000073 loss: 1.540882 (1.774937) time: 0.714604 data: 0.000219 max mem: 14338 Epoch: [5/30] [4300/5004] eta: 0:08:23 lr: 0.000073 loss: 1.770728 (1.775643) time: 0.716991 data: 0.000237 max mem: 14338 Epoch: [5/30] [4350/5004] eta: 0:07:47 lr: 0.000073 loss: 1.767353 (1.775117) time: 0.714319 data: 0.000175 max mem: 14338 Epoch: [5/30] [4400/5004] eta: 0:07:11 lr: 0.000073 loss: 1.742947 (1.775000) time: 0.714524 data: 0.000168 max mem: 14338 Epoch: [5/30] [4450/5004] eta: 0:06:36 lr: 0.000073 loss: 1.514130 (1.775212) time: 0.714308 data: 0.000225 max mem: 14338 Epoch: [5/30] [4500/5004] eta: 0:06:00 lr: 0.000073 loss: 1.500748 (1.774695) time: 0.712808 data: 0.000221 max mem: 14338 Epoch: [5/30] [4550/5004] eta: 0:05:24 lr: 0.000073 loss: 1.619677 (1.774519) time: 0.711156 data: 0.000217 max mem: 14338 Epoch: [5/30] [4600/5004] eta: 0:04:48 lr: 0.000073 loss: 1.566112 (1.773060) time: 0.714187 data: 0.000238 max mem: 14338 Epoch: [5/30] [4650/5004] eta: 0:04:13 lr: 0.000073 loss: 1.670857 (1.772821) time: 0.712956 data: 0.000216 max mem: 14338 Epoch: [5/30] [4700/5004] eta: 0:03:37 lr: 0.000073 loss: 1.748455 (1.772590) time: 0.711912 data: 0.000169 max mem: 14338 Epoch: [5/30] [4750/5004] eta: 0:03:01 lr: 0.000073 loss: 1.693473 (1.772870) time: 0.713417 data: 0.000170 max mem: 14338 Epoch: [5/30] [4800/5004] eta: 0:02:25 lr: 0.000073 loss: 1.675290 (1.772633) time: 0.718085 data: 0.000195 max mem: 14338 Epoch: [5/30] [4850/5004] eta: 0:01:50 lr: 0.000073 loss: 1.710135 (1.772673) time: 0.721537 data: 0.000218 max mem: 14338 Epoch: [5/30] [4900/5004] eta: 0:01:14 lr: 0.000072 loss: 1.658892 (1.772384) time: 0.709766 data: 0.000219 max mem: 14338 Epoch: [5/30] [4950/5004] eta: 0:00:38 lr: 0.000072 loss: 1.654844 (1.771827) time: 0.711995 data: 0.000206 max mem: 14338 Epoch: [5/30] [5000/5004] eta: 0:00:02 lr: 0.000072 loss: 1.688431 (1.771594) time: 0.709533 data: 0.000831 max mem: 14338 Epoch: [5/30] [5003/5004] eta: 0:00:00 lr: 0.000072 loss: 1.777625 (1.771836) time: 0.706345 data: 0.000828 max mem: 14338 Epoch: [5/30] Total time: 0:59:38 (0.715192 s / it) Averaged stats: lr: 0.000072 loss: 1.777625 (1.769313) Test: [ 0/196] eta: 0:04:57 loss: 0.390669 (0.390669) acc1: 87.500000 (87.500000) acc5: 100.000000 (100.000000) time: 1.516977 data: 1.071718 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.486704 (0.526943) acc1: 87.500000 (85.795455) acc5: 100.000000 (99.431818) time: 0.398395 data: 0.097551 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.566320 (0.545590) acc1: 87.500000 (86.309524) acc5: 100.000000 (98.214286) time: 0.293113 data: 0.000123 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.532818 (0.526015) acc1: 87.500000 (87.096774) acc5: 100.000000 (98.387097) time: 0.293719 data: 0.000111 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.416235 (0.528716) acc1: 87.500000 (86.737805) acc5: 100.000000 (98.323171) time: 0.287399 data: 0.000125 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.417946 (0.548791) acc1: 87.500000 (86.642157) acc5: 100.000000 (97.794118) time: 0.286882 data: 0.000130 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.528837 (0.574231) acc1: 87.500000 (86.168033) acc5: 100.000000 (97.745902) time: 0.286686 data: 0.000142 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.600555 (0.596413) acc1: 81.250000 (85.563380) acc5: 100.000000 (97.799296) time: 0.287020 data: 0.000144 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.588795 (0.595512) acc1: 87.500000 (85.725309) acc5: 100.000000 (97.839506) time: 0.287187 data: 0.000117 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.588795 (0.616728) acc1: 87.500000 (85.233516) acc5: 100.000000 (97.596154) time: 0.287147 data: 0.000120 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.592106 (0.605525) acc1: 81.250000 (85.334158) acc5: 100.000000 (97.710396) time: 0.287126 data: 0.000125 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.544542 (0.596282) acc1: 87.500000 (85.472973) acc5: 100.000000 (97.804054) time: 0.286583 data: 0.000122 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.558352 (0.595130) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.675620) time: 0.286971 data: 0.000138 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.607649 (0.610009) acc1: 87.500000 (85.162214) acc5: 93.750000 (97.614504) time: 0.287417 data: 0.000134 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.607649 (0.608670) acc1: 81.250000 (85.239362) acc5: 93.750000 (97.562057) time: 0.286709 data: 0.000131 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.586025 (0.613923) acc1: 87.500000 (85.140728) acc5: 93.750000 (97.516556) time: 0.286513 data: 0.000130 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.551524 (0.616899) acc1: 81.250000 (84.937888) acc5: 100.000000 (97.554348) time: 0.291909 data: 0.000125 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.524715 (0.614193) acc1: 81.250000 (85.087719) acc5: 100.000000 (97.587719) time: 0.292634 data: 0.000138 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.524715 (0.615183) acc1: 81.250000 (84.979282) acc5: 100.000000 (97.651934) time: 0.287383 data: 0.000151 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.350878 (0.604946) acc1: 87.500000 (85.242147) acc5: 100.000000 (97.643979) time: 0.284346 data: 0.000116 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.549393 (0.615715) acc1: 87.500000 (85.120000) acc5: 100.000000 (97.568000) time: 0.274668 data: 0.000096 max mem: 14338 Test: Total time: 0:00:57 (0.294455 s / it) * Acc@1 84.836 Acc@5 97.274 loss 0.630 Max accuracy: 84.84% uploading checkpoint virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0005.pth to hdfs://harunava/user/guoyuanfan/HCSC/virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0005.pth Epoch: [6/30] [ 0/5004] eta: 2:42:05 lr: 0.000072 loss: 1.833464 (1.833464) time: 1.943536 data: 1.180076 max mem: 14338 Epoch: [6/30] [ 50/5004] eta: 1:00:46 lr: 0.000072 loss: 1.675478 (1.751353) time: 0.714962 data: 0.000193 max mem: 14338 Epoch: [6/30] [ 100/5004] eta: 0:59:23 lr: 0.000072 loss: 1.689236 (1.758135) time: 0.720134 data: 0.000238 max mem: 14338 Epoch: [6/30] [ 150/5004] eta: 0:58:29 lr: 0.000072 loss: 1.691665 (1.775137) time: 0.718211 data: 0.000216 max mem: 14338 Epoch: [6/30] [ 200/5004] eta: 0:57:55 lr: 0.000072 loss: 1.659320 (1.749871) time: 0.738185 data: 0.000215 max mem: 14338 Epoch: [6/30] [ 250/5004] eta: 0:57:10 lr: 0.000072 loss: 1.692008 (1.748735) time: 0.717428 data: 0.000199 max mem: 14338 Epoch: [6/30] [ 300/5004] eta: 0:56:28 lr: 0.000072 loss: 1.684032 (1.750099) time: 0.709743 data: 0.000164 max mem: 14338 Epoch: [6/30] [ 350/5004] eta: 0:55:49 lr: 0.000072 loss: 1.860316 (1.757009) time: 0.710328 data: 0.000181 max mem: 14338 Epoch: [6/30] [ 400/5004] eta: 0:55:09 lr: 0.000072 loss: 1.542791 (1.748427) time: 0.713818 data: 0.000232 max mem: 14338 Epoch: [6/30] [ 450/5004] eta: 0:54:29 lr: 0.000072 loss: 1.704924 (1.752092) time: 0.711177 data: 0.000225 max mem: 14338 Epoch: [6/30] [ 500/5004] eta: 0:53:51 lr: 0.000072 loss: 1.806882 (1.764184) time: 0.712661 data: 0.000225 max mem: 14338 Epoch: [6/30] [ 550/5004] eta: 0:53:14 lr: 0.000072 loss: 1.672385 (1.760163) time: 0.717656 data: 0.000232 max mem: 14338 Epoch: [6/30] [ 600/5004] eta: 0:52:37 lr: 0.000072 loss: 1.630625 (1.763459) time: 0.713850 data: 0.000236 max mem: 14338 Epoch: [6/30] [ 650/5004] eta: 0:52:00 lr: 0.000072 loss: 1.539224 (1.765472) time: 0.714679 data: 0.000165 max mem: 14338 Epoch: [6/30] [ 700/5004] eta: 0:51:22 lr: 0.000072 loss: 1.584110 (1.762184) time: 0.708890 data: 0.000171 max mem: 14338 Epoch: [6/30] [ 750/5004] eta: 0:50:46 lr: 0.000072 loss: 1.640563 (1.761821) time: 0.713683 data: 0.000227 max mem: 14338 Epoch: [6/30] [ 800/5004] eta: 0:50:11 lr: 0.000072 loss: 1.542536 (1.762818) time: 0.716384 data: 0.000197 max mem: 14338 Epoch: [6/30] [ 850/5004] eta: 0:49:34 lr: 0.000072 loss: 1.795880 (1.764956) time: 0.717392 data: 0.000202 max mem: 14338 Epoch: [6/30] [ 900/5004] eta: 0:48:58 lr: 0.000072 loss: 1.658067 (1.760353) time: 0.710394 data: 0.000181 max mem: 14338 Epoch: [6/30] [ 950/5004] eta: 0:48:22 lr: 0.000072 loss: 1.550328 (1.756699) time: 0.713936 data: 0.000223 max mem: 14338 Epoch: [6/30] [1000/5004] eta: 0:47:47 lr: 0.000072 loss: 1.663207 (1.759855) time: 0.723307 data: 0.000174 max mem: 14338 Epoch: [6/30] [1050/5004] eta: 0:47:11 lr: 0.000072 loss: 1.725089 (1.761674) time: 0.715689 data: 0.000231 max mem: 14338 Epoch: [6/30] [1100/5004] eta: 0:46:34 lr: 0.000072 loss: 1.788943 (1.761746) time: 0.713745 data: 0.000234 max mem: 14338 Epoch: [6/30] [1150/5004] eta: 0:45:58 lr: 0.000072 loss: 1.752259 (1.760266) time: 0.710264 data: 0.000220 max mem: 14338 Epoch: [6/30] [1200/5004] eta: 0:45:22 lr: 0.000072 loss: 1.618329 (1.758261) time: 0.712747 data: 0.000209 max mem: 14338 Epoch: [6/30] [1250/5004] eta: 0:44:46 lr: 0.000072 loss: 1.699784 (1.757961) time: 0.712430 data: 0.000233 max mem: 14338 Epoch: [6/30] [1300/5004] eta: 0:44:10 lr: 0.000072 loss: 1.750544 (1.758858) time: 0.717813 data: 0.000197 max mem: 14338 Epoch: [6/30] [1350/5004] eta: 0:43:34 lr: 0.000072 loss: 1.506246 (1.758045) time: 0.710761 data: 0.000185 max mem: 14338 Epoch: [6/30] [1400/5004] eta: 0:42:59 lr: 0.000072 loss: 1.536641 (1.755258) time: 0.719018 data: 0.000217 max mem: 14338 Epoch: [6/30] [1450/5004] eta: 0:42:23 lr: 0.000072 loss: 1.687799 (1.753880) time: 0.719799 data: 0.000228 max mem: 14338 Epoch: [6/30] [1500/5004] eta: 0:41:47 lr: 0.000072 loss: 1.676547 (1.754326) time: 0.719540 data: 0.000241 max mem: 14338 Epoch: [6/30] [1550/5004] eta: 0:41:11 lr: 0.000072 loss: 1.562201 (1.753514) time: 0.718444 data: 0.000231 max mem: 14338 Epoch: [6/30] [1600/5004] eta: 0:40:36 lr: 0.000072 loss: 1.667131 (1.754596) time: 0.720366 data: 0.000222 max mem: 14338 Epoch: [6/30] [1650/5004] eta: 0:40:00 lr: 0.000072 loss: 1.724284 (1.754523) time: 0.716109 data: 0.000158 max mem: 14338 Epoch: [6/30] [1700/5004] eta: 0:39:24 lr: 0.000072 loss: 1.867356 (1.756556) time: 0.711455 data: 0.000180 max mem: 14338 Epoch: [6/30] [1750/5004] eta: 0:38:48 lr: 0.000072 loss: 1.890070 (1.759925) time: 0.713508 data: 0.000233 max mem: 14338 Epoch: [6/30] [1800/5004] eta: 0:38:12 lr: 0.000072 loss: 1.796910 (1.760129) time: 0.713633 data: 0.000218 max mem: 14338 Epoch: [6/30] [1850/5004] eta: 0:37:36 lr: 0.000072 loss: 1.840378 (1.761527) time: 0.712849 data: 0.000215 max mem: 14338 Epoch: [6/30] [1900/5004] eta: 0:37:00 lr: 0.000071 loss: 1.713357 (1.760544) time: 0.716492 data: 0.000228 max mem: 14338 Epoch: [6/30] [1950/5004] eta: 0:36:24 lr: 0.000071 loss: 1.644238 (1.759632) time: 0.717100 data: 0.000242 max mem: 14338 Epoch: [6/30] [2000/5004] eta: 0:35:49 lr: 0.000071 loss: 1.657429 (1.760796) time: 0.721088 data: 0.000180 max mem: 14338 Epoch: [6/30] [2050/5004] eta: 0:35:13 lr: 0.000071 loss: 1.825789 (1.760757) time: 0.714530 data: 0.000181 max mem: 14338 Epoch: [6/30] [2100/5004] eta: 0:34:37 lr: 0.000071 loss: 1.520994 (1.758124) time: 0.710432 data: 0.000225 max mem: 14338 Epoch: [6/30] [2150/5004] eta: 0:34:01 lr: 0.000071 loss: 1.517863 (1.756758) time: 0.709076 data: 0.000223 max mem: 14338 Epoch: [6/30] [2200/5004] eta: 0:33:25 lr: 0.000071 loss: 1.727760 (1.755813) time: 0.714135 data: 0.000198 max mem: 14338 Epoch: [6/30] [2250/5004] eta: 0:32:49 lr: 0.000071 loss: 1.535381 (1.753876) time: 0.712914 data: 0.000243 max mem: 14338 Epoch: [6/30] [2300/5004] eta: 0:32:13 lr: 0.000071 loss: 1.692196 (1.752721) time: 0.710345 data: 0.000241 max mem: 14338 Epoch: [6/30] [2350/5004] eta: 0:31:37 lr: 0.000071 loss: 1.780657 (1.753756) time: 0.715983 data: 0.000163 max mem: 14338 Epoch: [6/30] [2400/5004] eta: 0:31:01 lr: 0.000071 loss: 1.776268 (1.753892) time: 0.718030 data: 0.000232 max mem: 14338 Epoch: [6/30] [2450/5004] eta: 0:30:26 lr: 0.000071 loss: 1.768258 (1.754607) time: 0.720586 data: 0.000224 max mem: 14338 Epoch: [6/30] [2500/5004] eta: 0:29:50 lr: 0.000071 loss: 1.696246 (1.755511) time: 0.714416 data: 0.000215 max mem: 14338 Epoch: [6/30] [2550/5004] eta: 0:29:14 lr: 0.000071 loss: 1.589224 (1.754272) time: 0.715767 data: 0.000230 max mem: 14338 Epoch: [6/30] [2600/5004] eta: 0:28:39 lr: 0.000071 loss: 1.715933 (1.753685) time: 0.715077 data: 0.000224 max mem: 14338 Epoch: [6/30] [2650/5004] eta: 0:28:03 lr: 0.000071 loss: 1.823363 (1.754280) time: 0.712235 data: 0.000174 max mem: 14338 Epoch: [6/30] [2700/5004] eta: 0:27:27 lr: 0.000071 loss: 1.823916 (1.754304) time: 0.712088 data: 0.000166 max mem: 14338 Epoch: [6/30] [2750/5004] eta: 0:26:51 lr: 0.000071 loss: 1.649393 (1.754713) time: 0.711528 data: 0.000227 max mem: 14338 Epoch: [6/30] [2800/5004] eta: 0:26:16 lr: 0.000071 loss: 1.738237 (1.755294) time: 0.720183 data: 0.000227 max mem: 14338 Epoch: [6/30] [2850/5004] eta: 0:25:40 lr: 0.000071 loss: 1.675762 (1.754616) time: 0.721848 data: 0.000203 max mem: 14338 Epoch: [6/30] [2900/5004] eta: 0:25:04 lr: 0.000071 loss: 1.595916 (1.754237) time: 0.720611 data: 0.000218 max mem: 14338 Epoch: [6/30] [2950/5004] eta: 0:24:29 lr: 0.000071 loss: 1.655803 (1.754414) time: 0.713528 data: 0.000228 max mem: 14338 Epoch: [6/30] [3000/5004] eta: 0:23:53 lr: 0.000071 loss: 1.679801 (1.753522) time: 0.713745 data: 0.000184 max mem: 14338 Epoch: [6/30] [3050/5004] eta: 0:23:17 lr: 0.000071 loss: 1.740017 (1.753308) time: 0.715043 data: 0.000193 max mem: 14338 Epoch: [6/30] [3100/5004] eta: 0:22:41 lr: 0.000071 loss: 1.657143 (1.754346) time: 0.713225 data: 0.000227 max mem: 14338 Epoch: [6/30] [3150/5004] eta: 0:22:05 lr: 0.000071 loss: 1.770555 (1.754160) time: 0.709452 data: 0.000231 max mem: 14338 Epoch: [6/30] [3200/5004] eta: 0:21:30 lr: 0.000071 loss: 1.625884 (1.754544) time: 0.716604 data: 0.000211 max mem: 14338 Epoch: [6/30] [3250/5004] eta: 0:20:54 lr: 0.000071 loss: 1.674072 (1.753410) time: 0.716612 data: 0.000207 max mem: 14338 Epoch: [6/30] [3300/5004] eta: 0:20:18 lr: 0.000071 loss: 1.705594 (1.753247) time: 0.713034 data: 0.000217 max mem: 14338 Epoch: [6/30] [3350/5004] eta: 0:19:42 lr: 0.000071 loss: 1.660414 (1.751976) time: 0.711785 data: 0.000187 max mem: 14338 Epoch: [6/30] [3400/5004] eta: 0:19:07 lr: 0.000071 loss: 1.621766 (1.751838) time: 0.717588 data: 0.000156 max mem: 14338 Epoch: [6/30] [3450/5004] eta: 0:18:31 lr: 0.000071 loss: 1.647349 (1.753005) time: 0.716480 data: 0.000209 max mem: 14338 Epoch: [6/30] [3500/5004] eta: 0:17:55 lr: 0.000071 loss: 1.631384 (1.752222) time: 0.714502 data: 0.000189 max mem: 14338 Epoch: [6/30] [3550/5004] eta: 0:17:19 lr: 0.000071 loss: 1.753723 (1.752743) time: 0.709500 data: 0.000236 max mem: 14338 Epoch: [6/30] [3600/5004] eta: 0:16:43 lr: 0.000071 loss: 1.619652 (1.751947) time: 0.715114 data: 0.000202 max mem: 14338 Epoch: [6/30] [3650/5004] eta: 0:16:08 lr: 0.000071 loss: 1.872727 (1.752721) time: 0.715995 data: 0.000238 max mem: 14338 Epoch: [6/30] [3700/5004] eta: 0:15:32 lr: 0.000071 loss: 1.633616 (1.752038) time: 0.710558 data: 0.000172 max mem: 14338 Epoch: [6/30] [3750/5004] eta: 0:14:56 lr: 0.000071 loss: 1.747536 (1.751355) time: 0.717748 data: 0.000221 max mem: 14338 Epoch: [6/30] [3800/5004] eta: 0:14:20 lr: 0.000070 loss: 1.633314 (1.751614) time: 0.719861 data: 0.000227 max mem: 14338 Epoch: [6/30] [3850/5004] eta: 0:13:45 lr: 0.000070 loss: 1.651155 (1.751447) time: 0.714481 data: 0.000220 max mem: 14338 Epoch: [6/30] [3900/5004] eta: 0:13:09 lr: 0.000070 loss: 1.761762 (1.751544) time: 0.718257 data: 0.000231 max mem: 14338 Epoch: [6/30] [3950/5004] eta: 0:12:33 lr: 0.000070 loss: 1.700824 (1.751038) time: 0.715461 data: 0.000237 max mem: 14338 Epoch: [6/30] [4000/5004] eta: 0:11:57 lr: 0.000070 loss: 1.518817 (1.750930) time: 0.715020 data: 0.000167 max mem: 14338 Epoch: [6/30] [4050/5004] eta: 0:11:22 lr: 0.000070 loss: 1.707857 (1.750982) time: 0.716917 data: 0.000157 max mem: 14338 Epoch: [6/30] [4100/5004] eta: 0:10:46 lr: 0.000070 loss: 1.733633 (1.750067) time: 0.709708 data: 0.000243 max mem: 14338 Epoch: [6/30] [4150/5004] eta: 0:10:10 lr: 0.000070 loss: 1.697294 (1.749863) time: 0.708625 data: 0.000216 max mem: 14338 Epoch: [6/30] [4200/5004] eta: 0:09:34 lr: 0.000070 loss: 1.723092 (1.749991) time: 0.716868 data: 0.000224 max mem: 14338 Epoch: [6/30] [4250/5004] eta: 0:08:59 lr: 0.000070 loss: 1.682416 (1.749553) time: 0.716294 data: 0.000228 max mem: 14338 Epoch: [6/30] [4300/5004] eta: 0:08:23 lr: 0.000070 loss: 1.663688 (1.750231) time: 0.711002 data: 0.000218 max mem: 14338 Epoch: [6/30] [4350/5004] eta: 0:07:47 lr: 0.000070 loss: 1.755774 (1.750603) time: 0.716255 data: 0.000188 max mem: 14338 Epoch: [6/30] [4400/5004] eta: 0:07:11 lr: 0.000070 loss: 1.561935 (1.750111) time: 0.717620 data: 0.000192 max mem: 14338 Epoch: [6/30] [4450/5004] eta: 0:06:36 lr: 0.000070 loss: 1.672223 (1.749370) time: 0.715895 data: 0.000226 max mem: 14338 Epoch: [6/30] [4500/5004] eta: 0:06:00 lr: 0.000070 loss: 1.665488 (1.749988) time: 0.709977 data: 0.000212 max mem: 14338 Epoch: [6/30] [4550/5004] eta: 0:05:24 lr: 0.000070 loss: 1.646048 (1.749621) time: 0.709315 data: 0.000224 max mem: 14338 Epoch: [6/30] [4600/5004] eta: 0:04:48 lr: 0.000070 loss: 1.692372 (1.749531) time: 0.716056 data: 0.000236 max mem: 14338 Epoch: [6/30] [4650/5004] eta: 0:04:13 lr: 0.000070 loss: 1.740369 (1.748662) time: 0.712301 data: 0.000224 max mem: 14338 Epoch: [6/30] [4700/5004] eta: 0:03:37 lr: 0.000070 loss: 1.693703 (1.748563) time: 0.713779 data: 0.000178 max mem: 14338 Epoch: [6/30] [4750/5004] eta: 0:03:01 lr: 0.000070 loss: 1.723727 (1.748504) time: 0.721284 data: 0.000186 max mem: 14338 Epoch: [6/30] [4800/5004] eta: 0:02:25 lr: 0.000070 loss: 1.503547 (1.747276) time: 0.717760 data: 0.000202 max mem: 14338 Epoch: [6/30] [4850/5004] eta: 0:01:50 lr: 0.000070 loss: 1.613366 (1.747579) time: 0.718091 data: 0.000229 max mem: 14338 Epoch: [6/30] [4900/5004] eta: 0:01:14 lr: 0.000070 loss: 1.614472 (1.747782) time: 0.709373 data: 0.000233 max mem: 14338 Epoch: [6/30] [4950/5004] eta: 0:00:38 lr: 0.000070 loss: 1.639223 (1.747532) time: 0.710087 data: 0.000228 max mem: 14338 Epoch: [6/30] [5000/5004] eta: 0:00:02 lr: 0.000070 loss: 1.595748 (1.746785) time: 0.709585 data: 0.000836 max mem: 14338 Epoch: [6/30] [5003/5004] eta: 0:00:00 lr: 0.000070 loss: 1.595748 (1.746686) time: 0.706701 data: 0.000818 max mem: 14338 Epoch: [6/30] Total time: 0:59:37 (0.714928 s / it) Averaged stats: lr: 0.000070 loss: 1.595748 (1.740572) Test: [ 0/196] eta: 0:05:02 loss: 0.339605 (0.339605) acc1: 87.500000 (87.500000) acc5: 100.000000 (100.000000) time: 1.541733 data: 1.191689 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.489037 (0.531107) acc1: 87.500000 (85.795455) acc5: 100.000000 (99.431818) time: 0.403037 data: 0.108457 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.561899 (0.538815) acc1: 87.500000 (85.714286) acc5: 100.000000 (98.809524) time: 0.288574 data: 0.000121 max mem: 14338 Test: [ 30/196] eta: 0:00:55 loss: 0.510479 (0.516637) acc1: 87.500000 (86.290323) acc5: 100.000000 (98.991935) time: 0.294188 data: 0.000125 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.360714 (0.519784) acc1: 87.500000 (86.128049) acc5: 100.000000 (98.628049) time: 0.293935 data: 0.000151 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.385626 (0.545578) acc1: 87.500000 (86.151961) acc5: 100.000000 (98.039216) time: 0.286624 data: 0.000142 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.560438 (0.571591) acc1: 87.500000 (85.758197) acc5: 100.000000 (98.053279) time: 0.286252 data: 0.000131 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.670092 (0.591280) acc1: 81.250000 (85.211268) acc5: 100.000000 (98.063380) time: 0.287119 data: 0.000135 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.594728 (0.591291) acc1: 87.500000 (85.185185) acc5: 100.000000 (98.070988) time: 0.286854 data: 0.000120 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.594728 (0.612553) acc1: 81.250000 (84.890110) acc5: 100.000000 (97.733516) time: 0.285866 data: 0.000120 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.603468 (0.604445) acc1: 81.250000 (85.024752) acc5: 100.000000 (97.896040) time: 0.285706 data: 0.000125 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.555403 (0.596244) acc1: 87.500000 (85.078829) acc5: 100.000000 (97.972973) time: 0.286090 data: 0.000117 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.527988 (0.592141) acc1: 87.500000 (85.227273) acc5: 100.000000 (97.882231) time: 0.286499 data: 0.000140 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.576639 (0.606269) acc1: 81.250000 (85.066794) acc5: 93.750000 (97.805344) time: 0.287881 data: 0.000134 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.641739 (0.605549) acc1: 81.250000 (85.195035) acc5: 93.750000 (97.739362) time: 0.288275 data: 0.000133 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.635448 (0.612221) acc1: 87.500000 (85.099338) acc5: 93.750000 (97.682119) time: 0.287016 data: 0.000150 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.547562 (0.613562) acc1: 87.500000 (85.131988) acc5: 100.000000 (97.748447) time: 0.286840 data: 0.000133 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.478151 (0.609323) acc1: 81.250000 (85.270468) acc5: 100.000000 (97.770468) time: 0.292858 data: 0.000136 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.478151 (0.610960) acc1: 81.250000 (85.048343) acc5: 100.000000 (97.790055) time: 0.293108 data: 0.000152 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.408869 (0.601775) acc1: 81.250000 (85.209424) acc5: 100.000000 (97.807592) time: 0.284576 data: 0.000119 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.495086 (0.611398) acc1: 81.250000 (85.056000) acc5: 100.000000 (97.728000) time: 0.274903 data: 0.000109 max mem: 14338 Test: Total time: 0:00:57 (0.294535 s / it) * Acc@1 84.982 Acc@5 97.352 loss 0.626 Max accuracy: 84.98% Epoch: [7/30] [ 0/5004] eta: 2:38:58 lr: 0.000070 loss: 1.714142 (1.714142) time: 1.906260 data: 1.148325 max mem: 14338 Epoch: [7/30] [ 50/5004] eta: 1:00:50 lr: 0.000070 loss: 1.555762 (1.682998) time: 0.713515 data: 0.000185 max mem: 14338 Epoch: [7/30] [ 100/5004] eta: 0:59:25 lr: 0.000070 loss: 1.540044 (1.705671) time: 0.713193 data: 0.000223 max mem: 14338 Epoch: [7/30] [ 150/5004] eta: 0:58:32 lr: 0.000070 loss: 1.658773 (1.711147) time: 0.711913 data: 0.000218 max mem: 14338 Epoch: [7/30] [ 200/5004] eta: 0:57:48 lr: 0.000070 loss: 1.765279 (1.720957) time: 0.716939 data: 0.000210 max mem: 14338 Epoch: [7/30] [ 250/5004] eta: 0:57:04 lr: 0.000070 loss: 1.723236 (1.734657) time: 0.718771 data: 0.000206 max mem: 14338 Epoch: [7/30] [ 300/5004] eta: 0:56:23 lr: 0.000070 loss: 1.639750 (1.714080) time: 0.718437 data: 0.000188 max mem: 14338 Epoch: [7/30] [ 350/5004] eta: 0:55:45 lr: 0.000070 loss: 1.569516 (1.704679) time: 0.714150 data: 0.000168 max mem: 14338 Epoch: [7/30] [ 400/5004] eta: 0:55:06 lr: 0.000070 loss: 1.612389 (1.704814) time: 0.715723 data: 0.000217 max mem: 14338 Epoch: [7/30] [ 450/5004] eta: 0:54:27 lr: 0.000070 loss: 1.751727 (1.706078) time: 0.713208 data: 0.000216 max mem: 14338 Epoch: [7/30] [ 500/5004] eta: 0:53:49 lr: 0.000070 loss: 1.534070 (1.697976) time: 0.709489 data: 0.000213 max mem: 14338 Epoch: [7/30] [ 550/5004] eta: 0:53:12 lr: 0.000070 loss: 1.491303 (1.696589) time: 0.712060 data: 0.000221 max mem: 14338 Epoch: [7/30] [ 600/5004] eta: 0:52:35 lr: 0.000069 loss: 1.488151 (1.697199) time: 0.713401 data: 0.000207 max mem: 14338 Epoch: [7/30] [ 650/5004] eta: 0:51:59 lr: 0.000069 loss: 1.653719 (1.697464) time: 0.714531 data: 0.000166 max mem: 14338 Epoch: [7/30] [ 700/5004] eta: 0:51:23 lr: 0.000069 loss: 1.524076 (1.701453) time: 0.714576 data: 0.000160 max mem: 14338 Epoch: [7/30] [ 750/5004] eta: 0:50:46 lr: 0.000069 loss: 1.714113 (1.704283) time: 0.710208 data: 0.000226 max mem: 14338 Epoch: [7/30] [ 800/5004] eta: 0:50:11 lr: 0.000069 loss: 1.781564 (1.706424) time: 0.723652 data: 0.000229 max mem: 14338 Epoch: [7/30] [ 850/5004] eta: 0:49:34 lr: 0.000069 loss: 1.594134 (1.703518) time: 0.714361 data: 0.000221 max mem: 14338 Epoch: [7/30] [ 900/5004] eta: 0:48:58 lr: 0.000069 loss: 1.689470 (1.704198) time: 0.712439 data: 0.000194 max mem: 14338 Epoch: [7/30] [ 950/5004] eta: 0:48:21 lr: 0.000069 loss: 1.444521 (1.706278) time: 0.709956 data: 0.000239 max mem: 14338 Epoch: [7/30] [1000/5004] eta: 0:47:46 lr: 0.000069 loss: 1.580031 (1.708888) time: 0.715786 data: 0.000185 max mem: 14338 Epoch: [7/30] [1050/5004] eta: 0:47:10 lr: 0.000069 loss: 1.625480 (1.706660) time: 0.713972 data: 0.000234 max mem: 14338 Epoch: [7/30] [1100/5004] eta: 0:46:34 lr: 0.000069 loss: 1.634119 (1.706039) time: 0.712844 data: 0.000223 max mem: 14338 Epoch: [7/30] [1150/5004] eta: 0:45:58 lr: 0.000069 loss: 1.791226 (1.706485) time: 0.711958 data: 0.000212 max mem: 14338 Epoch: [7/30] [1200/5004] eta: 0:45:23 lr: 0.000069 loss: 1.651648 (1.709128) time: 0.723180 data: 0.000225 max mem: 14338 Epoch: [7/30] [1250/5004] eta: 0:44:47 lr: 0.000069 loss: 1.512466 (1.707989) time: 0.720311 data: 0.000238 max mem: 14338 Epoch: [7/30] [1300/5004] eta: 0:44:11 lr: 0.000069 loss: 1.662976 (1.707210) time: 0.715623 data: 0.000165 max mem: 14338 Epoch: [7/30] [1350/5004] eta: 0:43:35 lr: 0.000069 loss: 1.694704 (1.708662) time: 0.710699 data: 0.000161 max mem: 14338 Epoch: [7/30] [1400/5004] eta: 0:42:59 lr: 0.000069 loss: 1.618439 (1.707918) time: 0.714433 data: 0.000216 max mem: 14338 Epoch: [7/30] [1450/5004] eta: 0:42:23 lr: 0.000069 loss: 1.647589 (1.708101) time: 0.711166 data: 0.000222 max mem: 14338 Epoch: [7/30] [1500/5004] eta: 0:41:47 lr: 0.000069 loss: 1.649462 (1.708416) time: 0.710894 data: 0.000232 max mem: 14338 Epoch: [7/30] [1550/5004] eta: 0:41:11 lr: 0.000069 loss: 1.546861 (1.707934) time: 0.711534 data: 0.000203 max mem: 14338 Epoch: [7/30] [1600/5004] eta: 0:40:35 lr: 0.000069 loss: 1.719968 (1.707584) time: 0.713694 data: 0.000215 max mem: 14338 Epoch: [7/30] [1650/5004] eta: 0:39:59 lr: 0.000069 loss: 1.600048 (1.706809) time: 0.716714 data: 0.000155 max mem: 14338 Epoch: [7/30] [1700/5004] eta: 0:39:23 lr: 0.000069 loss: 1.656085 (1.706867) time: 0.716672 data: 0.000167 max mem: 14338 Epoch: [7/30] [1750/5004] eta: 0:38:47 lr: 0.000069 loss: 1.728796 (1.707896) time: 0.717784 data: 0.000216 max mem: 14338 Epoch: [7/30] [1800/5004] eta: 0:38:12 lr: 0.000069 loss: 1.757177 (1.709043) time: 0.721017 data: 0.000214 max mem: 14338 Epoch: [7/30] [1850/5004] eta: 0:37:36 lr: 0.000069 loss: 1.676877 (1.709136) time: 0.711624 data: 0.000206 max mem: 14338 Epoch: [7/30] [1900/5004] eta: 0:37:00 lr: 0.000069 loss: 1.739089 (1.710137) time: 0.713224 data: 0.000197 max mem: 14338 Epoch: [7/30] [1950/5004] eta: 0:36:24 lr: 0.000069 loss: 1.612130 (1.709992) time: 0.710297 data: 0.000212 max mem: 14338 Epoch: [7/30] [2000/5004] eta: 0:35:48 lr: 0.000069 loss: 1.616907 (1.710886) time: 0.714925 data: 0.000154 max mem: 14338 Epoch: [7/30] [2050/5004] eta: 0:35:12 lr: 0.000069 loss: 1.644328 (1.709550) time: 0.712115 data: 0.000172 max mem: 14338 Epoch: [7/30] [2100/5004] eta: 0:34:37 lr: 0.000069 loss: 1.783432 (1.710735) time: 0.715686 data: 0.000234 max mem: 14338 Epoch: [7/30] [2150/5004] eta: 0:34:01 lr: 0.000069 loss: 1.707471 (1.712115) time: 0.716123 data: 0.000210 max mem: 14338 Epoch: [7/30] [2200/5004] eta: 0:33:25 lr: 0.000069 loss: 1.821218 (1.713279) time: 0.713588 data: 0.000200 max mem: 14338 Epoch: [7/30] [2250/5004] eta: 0:32:49 lr: 0.000069 loss: 1.715815 (1.714340) time: 0.718020 data: 0.000211 max mem: 14338 Epoch: [7/30] [2300/5004] eta: 0:32:13 lr: 0.000069 loss: 1.732148 (1.714853) time: 0.709512 data: 0.000213 max mem: 14338 Epoch: [7/30] [2350/5004] eta: 0:31:37 lr: 0.000068 loss: 1.791677 (1.715471) time: 0.713709 data: 0.000169 max mem: 14338 Epoch: [7/30] [2400/5004] eta: 0:31:02 lr: 0.000068 loss: 1.598042 (1.715808) time: 0.712532 data: 0.000226 max mem: 14338 Epoch: [7/30] [2450/5004] eta: 0:30:26 lr: 0.000068 loss: 1.685994 (1.716441) time: 0.711796 data: 0.000209 max mem: 14338 Epoch: [7/30] [2500/5004] eta: 0:29:50 lr: 0.000068 loss: 1.726696 (1.716448) time: 0.710605 data: 0.000224 max mem: 14338 Epoch: [7/30] [2550/5004] eta: 0:29:14 lr: 0.000068 loss: 1.670467 (1.715749) time: 0.714573 data: 0.000204 max mem: 14338 Epoch: [7/30] [2600/5004] eta: 0:28:38 lr: 0.000068 loss: 1.701688 (1.715612) time: 0.719876 data: 0.000216 max mem: 14338 Epoch: [7/30] [2650/5004] eta: 0:28:03 lr: 0.000068 loss: 1.605070 (1.716126) time: 0.717650 data: 0.000168 max mem: 14338 Epoch: [7/30] [2700/5004] eta: 0:27:27 lr: 0.000068 loss: 1.805658 (1.718040) time: 0.719572 data: 0.000172 max mem: 14338 Epoch: [7/30] [2750/5004] eta: 0:26:51 lr: 0.000068 loss: 1.576494 (1.717958) time: 0.712556 data: 0.000215 max mem: 14338 Epoch: [7/30] [2800/5004] eta: 0:26:15 lr: 0.000068 loss: 1.816940 (1.718239) time: 0.711494 data: 0.000190 max mem: 14338 Epoch: [7/30] [2850/5004] eta: 0:25:39 lr: 0.000068 loss: 1.524529 (1.717719) time: 0.716362 data: 0.000172 max mem: 14338 Epoch: [7/30] [2900/5004] eta: 0:25:04 lr: 0.000068 loss: 1.641353 (1.717251) time: 0.713143 data: 0.000217 max mem: 14338 Epoch: [7/30] [2950/5004] eta: 0:24:28 lr: 0.000068 loss: 1.554112 (1.715744) time: 0.710154 data: 0.000194 max mem: 14338 Epoch: [7/30] [3000/5004] eta: 0:23:52 lr: 0.000068 loss: 1.820700 (1.717005) time: 0.713929 data: 0.000162 max mem: 14338 Epoch: [7/30] [3050/5004] eta: 0:23:16 lr: 0.000068 loss: 1.599746 (1.716143) time: 0.714779 data: 0.000156 max mem: 14338 Epoch: [7/30] [3100/5004] eta: 0:22:41 lr: 0.000068 loss: 1.833681 (1.717032) time: 0.713010 data: 0.000223 max mem: 14338 Epoch: [7/30] [3150/5004] eta: 0:22:05 lr: 0.000068 loss: 1.649314 (1.716804) time: 0.712683 data: 0.000232 max mem: 14338 Epoch: [7/30] [3200/5004] eta: 0:21:29 lr: 0.000068 loss: 1.715088 (1.716205) time: 0.719206 data: 0.000216 max mem: 14338 Epoch: [7/30] [3250/5004] eta: 0:20:53 lr: 0.000068 loss: 1.549854 (1.714434) time: 0.715635 data: 0.000200 max mem: 14338 Epoch: [7/30] [3300/5004] eta: 0:20:18 lr: 0.000068 loss: 1.759270 (1.713468) time: 0.710864 data: 0.000238 max mem: 14338 Epoch: [7/30] [3350/5004] eta: 0:19:42 lr: 0.000068 loss: 1.439828 (1.712669) time: 0.712912 data: 0.000192 max mem: 14338 Epoch: [7/30] [3400/5004] eta: 0:19:06 lr: 0.000068 loss: 1.698922 (1.712489) time: 0.715994 data: 0.000159 max mem: 14338 Epoch: [7/30] [3450/5004] eta: 0:18:30 lr: 0.000068 loss: 1.652468 (1.712085) time: 0.715832 data: 0.000216 max mem: 14338 Epoch: [7/30] [3500/5004] eta: 0:17:55 lr: 0.000068 loss: 1.654415 (1.712561) time: 0.719703 data: 0.000181 max mem: 14338 Epoch: [7/30] [3550/5004] eta: 0:17:19 lr: 0.000068 loss: 1.677221 (1.712909) time: 0.713222 data: 0.000217 max mem: 14338 Epoch: [7/30] [3600/5004] eta: 0:16:43 lr: 0.000068 loss: 1.594719 (1.712383) time: 0.716692 data: 0.000220 max mem: 14338 Epoch: [7/30] [3650/5004] eta: 0:16:07 lr: 0.000068 loss: 1.607131 (1.713130) time: 0.715257 data: 0.000227 max mem: 14338 Epoch: [7/30] [3700/5004] eta: 0:15:32 lr: 0.000068 loss: 1.795154 (1.713405) time: 0.710841 data: 0.000159 max mem: 14338 Epoch: [7/30] [3750/5004] eta: 0:14:56 lr: 0.000068 loss: 1.793600 (1.714870) time: 0.714590 data: 0.000213 max mem: 14338 Epoch: [7/30] [3800/5004] eta: 0:14:20 lr: 0.000068 loss: 1.733329 (1.714970) time: 0.712535 data: 0.000204 max mem: 14338 Epoch: [7/30] [3850/5004] eta: 0:13:44 lr: 0.000068 loss: 1.554078 (1.714146) time: 0.713762 data: 0.000218 max mem: 14338 Epoch: [7/30] [3900/5004] eta: 0:13:09 lr: 0.000068 loss: 1.561893 (1.713601) time: 0.716228 data: 0.000220 max mem: 14338 Epoch: [7/30] [3950/5004] eta: 0:12:33 lr: 0.000068 loss: 1.567192 (1.713277) time: 0.714194 data: 0.000219 max mem: 14338 Epoch: [7/30] [4000/5004] eta: 0:11:57 lr: 0.000067 loss: 1.582514 (1.712914) time: 0.722586 data: 0.000167 max mem: 14338 Epoch: [7/30] [4050/5004] eta: 0:11:21 lr: 0.000067 loss: 1.623614 (1.713675) time: 0.718198 data: 0.000164 max mem: 14338 Epoch: [7/30] [4100/5004] eta: 0:10:46 lr: 0.000067 loss: 1.593626 (1.713185) time: 0.712526 data: 0.000223 max mem: 14338 Epoch: [7/30] [4150/5004] eta: 0:10:10 lr: 0.000067 loss: 1.773690 (1.714550) time: 0.710283 data: 0.000217 max mem: 14338 Epoch: [7/30] [4200/5004] eta: 0:09:34 lr: 0.000067 loss: 1.577827 (1.713996) time: 0.718787 data: 0.000233 max mem: 14338 Epoch: [7/30] [4250/5004] eta: 0:08:59 lr: 0.000067 loss: 1.609870 (1.714949) time: 0.711988 data: 0.000226 max mem: 14338 Epoch: [7/30] [4300/5004] eta: 0:08:23 lr: 0.000067 loss: 1.530784 (1.715010) time: 0.710074 data: 0.000237 max mem: 14338 Epoch: [7/30] [4350/5004] eta: 0:07:47 lr: 0.000067 loss: 1.790496 (1.715952) time: 0.713432 data: 0.000188 max mem: 14338 Epoch: [7/30] [4400/5004] eta: 0:07:11 lr: 0.000067 loss: 1.753627 (1.715054) time: 0.711490 data: 0.000178 max mem: 14338 Epoch: [7/30] [4450/5004] eta: 0:06:36 lr: 0.000067 loss: 1.673427 (1.715642) time: 0.718643 data: 0.000214 max mem: 14338 Epoch: [7/30] [4500/5004] eta: 0:06:00 lr: 0.000067 loss: 1.802437 (1.715820) time: 0.719033 data: 0.000226 max mem: 14338 Epoch: [7/30] [4550/5004] eta: 0:05:24 lr: 0.000067 loss: 1.795646 (1.715752) time: 0.712764 data: 0.000214 max mem: 14338 Epoch: [7/30] [4600/5004] eta: 0:04:48 lr: 0.000067 loss: 1.594028 (1.715509) time: 0.717368 data: 0.000224 max mem: 14338 Epoch: [7/30] [4650/5004] eta: 0:04:13 lr: 0.000067 loss: 1.745592 (1.715329) time: 0.714027 data: 0.000214 max mem: 14338 Epoch: [7/30] [4700/5004] eta: 0:03:37 lr: 0.000067 loss: 1.676465 (1.715304) time: 0.710070 data: 0.000161 max mem: 14338 Epoch: [7/30] [4750/5004] eta: 0:03:01 lr: 0.000067 loss: 1.756702 (1.715314) time: 0.713088 data: 0.000158 max mem: 14338 Epoch: [7/30] [4800/5004] eta: 0:02:25 lr: 0.000067 loss: 1.678719 (1.715501) time: 0.712267 data: 0.000193 max mem: 14338 Epoch: [7/30] [4850/5004] eta: 0:01:50 lr: 0.000067 loss: 1.732187 (1.715974) time: 0.719546 data: 0.000223 max mem: 14338 Epoch: [7/30] [4900/5004] eta: 0:01:14 lr: 0.000067 loss: 1.741136 (1.715981) time: 0.712814 data: 0.000225 max mem: 14338 Epoch: [7/30] [4950/5004] eta: 0:00:38 lr: 0.000067 loss: 1.752499 (1.716061) time: 0.717910 data: 0.000218 max mem: 14338 Epoch: [7/30] [5000/5004] eta: 0:00:02 lr: 0.000067 loss: 1.548110 (1.715470) time: 0.717128 data: 0.000864 max mem: 14338 Epoch: [7/30] [5003/5004] eta: 0:00:00 lr: 0.000067 loss: 1.548110 (1.715349) time: 0.714059 data: 0.000857 max mem: 14338 Epoch: [7/30] Total time: 0:59:37 (0.714974 s / it) Averaged stats: lr: 0.000067 loss: 1.548110 (1.719480) Test: [ 0/196] eta: 0:04:51 loss: 0.293736 (0.293736) acc1: 87.500000 (87.500000) acc5: 100.000000 (100.000000) time: 1.489251 data: 1.105714 max mem: 14338 Test: [ 10/196] eta: 0:01:13 loss: 0.563589 (0.524768) acc1: 87.500000 (85.227273) acc5: 100.000000 (98.863636) time: 0.397322 data: 0.100639 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.586975 (0.556183) acc1: 87.500000 (85.416667) acc5: 100.000000 (98.511905) time: 0.287950 data: 0.000117 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.492460 (0.532116) acc1: 87.500000 (86.693548) acc5: 100.000000 (98.588710) time: 0.287951 data: 0.000110 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.444671 (0.534666) acc1: 87.500000 (86.737805) acc5: 100.000000 (98.323171) time: 0.288269 data: 0.000143 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.438715 (0.555794) acc1: 87.500000 (86.764706) acc5: 100.000000 (97.671569) time: 0.288820 data: 0.000146 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.604376 (0.583541) acc1: 87.500000 (86.168033) acc5: 93.750000 (97.540984) time: 0.288169 data: 0.000133 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.611814 (0.598421) acc1: 81.250000 (85.739437) acc5: 100.000000 (97.711268) time: 0.286636 data: 0.000128 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.522173 (0.599112) acc1: 87.500000 (85.802469) acc5: 100.000000 (97.762346) time: 0.286431 data: 0.000117 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.552610 (0.622331) acc1: 81.250000 (85.370879) acc5: 100.000000 (97.458791) time: 0.286856 data: 0.000135 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.600962 (0.611045) acc1: 81.250000 (85.457921) acc5: 100.000000 (97.648515) time: 0.294079 data: 0.000134 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.572224 (0.604675) acc1: 87.500000 (85.360360) acc5: 100.000000 (97.747748) time: 0.294393 data: 0.000114 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.583494 (0.603264) acc1: 87.500000 (85.433884) acc5: 100.000000 (97.623967) time: 0.287448 data: 0.000128 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.637943 (0.618314) acc1: 81.250000 (85.019084) acc5: 93.750000 (97.566794) time: 0.287260 data: 0.000135 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.631844 (0.615658) acc1: 81.250000 (85.195035) acc5: 93.750000 (97.517730) time: 0.287128 data: 0.000127 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.585272 (0.621029) acc1: 87.500000 (85.057947) acc5: 93.750000 (97.475166) time: 0.286920 data: 0.000120 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.534566 (0.625159) acc1: 81.250000 (84.976708) acc5: 100.000000 (97.437888) time: 0.286621 data: 0.000112 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.530231 (0.620786) acc1: 81.250000 (85.087719) acc5: 100.000000 (97.478070) time: 0.287256 data: 0.000121 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.462954 (0.621311) acc1: 81.250000 (84.875691) acc5: 100.000000 (97.479282) time: 0.286575 data: 0.000142 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.365690 (0.608388) acc1: 87.500000 (85.143979) acc5: 100.000000 (97.513089) time: 0.283413 data: 0.000119 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.486443 (0.617658) acc1: 87.500000 (84.992000) acc5: 100.000000 (97.440000) time: 0.274185 data: 0.000107 max mem: 14338 Test: Total time: 0:00:57 (0.293870 s / it) * Acc@1 84.956 Acc@5 97.290 loss 0.629 Max accuracy: 84.98% Epoch: [8/30] [ 0/5004] eta: 2:38:01 lr: 0.000067 loss: 2.107752 (2.107752) time: 1.894872 data: 1.144595 max mem: 14338 Epoch: [8/30] [ 50/5004] eta: 1:00:49 lr: 0.000067 loss: 1.538147 (1.684198) time: 0.715843 data: 0.000198 max mem: 14338 Epoch: [8/30] [ 100/5004] eta: 0:59:13 lr: 0.000067 loss: 1.836245 (1.715034) time: 0.712532 data: 0.000272 max mem: 14338 Epoch: [8/30] [ 150/5004] eta: 0:58:18 lr: 0.000067 loss: 1.728541 (1.728820) time: 0.710384 data: 0.000219 max mem: 14338 Epoch: [8/30] [ 200/5004] eta: 0:57:36 lr: 0.000067 loss: 1.533132 (1.699089) time: 0.712740 data: 0.000231 max mem: 14338 Epoch: [8/30] [ 250/5004] eta: 0:56:54 lr: 0.000067 loss: 1.595151 (1.705880) time: 0.713232 data: 0.000203 max mem: 14338 Epoch: [8/30] [ 300/5004] eta: 0:56:13 lr: 0.000067 loss: 1.533081 (1.698086) time: 0.710946 data: 0.000170 max mem: 14338 Epoch: [8/30] [ 350/5004] eta: 0:55:34 lr: 0.000067 loss: 1.775732 (1.706223) time: 0.714838 data: 0.000177 max mem: 14338 Epoch: [8/30] [ 400/5004] eta: 0:54:57 lr: 0.000067 loss: 1.772590 (1.708773) time: 0.717043 data: 0.000207 max mem: 14338 Epoch: [8/30] [ 450/5004] eta: 0:54:21 lr: 0.000067 loss: 1.684665 (1.707874) time: 0.718952 data: 0.000202 max mem: 14338 Epoch: [8/30] [ 500/5004] eta: 0:53:44 lr: 0.000067 loss: 1.820493 (1.705768) time: 0.714525 data: 0.000220 max mem: 14338 Epoch: [8/30] [ 550/5004] eta: 0:53:07 lr: 0.000067 loss: 1.696012 (1.707718) time: 0.715092 data: 0.000230 max mem: 14338 Epoch: [8/30] [ 600/5004] eta: 0:52:31 lr: 0.000067 loss: 1.604885 (1.703215) time: 0.713952 data: 0.000224 max mem: 14338 Epoch: [8/30] [ 650/5004] eta: 0:51:55 lr: 0.000066 loss: 1.772102 (1.709957) time: 0.714094 data: 0.000159 max mem: 14338 Epoch: [8/30] [ 700/5004] eta: 0:51:18 lr: 0.000066 loss: 1.579863 (1.709486) time: 0.711692 data: 0.000159 max mem: 14338 Epoch: [8/30] [ 750/5004] eta: 0:50:42 lr: 0.000066 loss: 1.710663 (1.716783) time: 0.711067 data: 0.000225 max mem: 14338 Epoch: [8/30] [ 800/5004] eta: 0:50:06 lr: 0.000066 loss: 1.800246 (1.719304) time: 0.717458 data: 0.000236 max mem: 14338 Epoch: [8/30] [ 850/5004] eta: 0:49:31 lr: 0.000066 loss: 1.721717 (1.719519) time: 0.719927 data: 0.000223 max mem: 14338 Epoch: [8/30] [ 900/5004] eta: 0:48:55 lr: 0.000066 loss: 1.587371 (1.721247) time: 0.713115 data: 0.000201 max mem: 14338 Epoch: [8/30] [ 950/5004] eta: 0:48:18 lr: 0.000066 loss: 1.707502 (1.717838) time: 0.713020 data: 0.000211 max mem: 14338 Epoch: [8/30] [1000/5004] eta: 0:47:43 lr: 0.000066 loss: 1.691186 (1.717229) time: 0.719737 data: 0.000176 max mem: 14338 Epoch: [8/30] [1050/5004] eta: 0:47:07 lr: 0.000066 loss: 1.629734 (1.716675) time: 0.715205 data: 0.000233 max mem: 14338 Epoch: [8/30] [1100/5004] eta: 0:46:31 lr: 0.000066 loss: 1.614457 (1.712724) time: 0.710666 data: 0.000240 max mem: 14338 Epoch: [8/30] [1150/5004] eta: 0:45:56 lr: 0.000066 loss: 1.681209 (1.713805) time: 0.714188 data: 0.000217 max mem: 14338 Epoch: [8/30] [1200/5004] eta: 0:45:20 lr: 0.000066 loss: 1.748109 (1.713053) time: 0.720249 data: 0.000204 max mem: 14338 Epoch: [8/30] [1250/5004] eta: 0:44:44 lr: 0.000066 loss: 1.546382 (1.712242) time: 0.714755 data: 0.000201 max mem: 14338 Epoch: [8/30] [1300/5004] eta: 0:44:09 lr: 0.000066 loss: 1.580421 (1.712459) time: 0.710856 data: 0.000173 max mem: 14338 Epoch: [8/30] [1350/5004] eta: 0:43:33 lr: 0.000066 loss: 1.728768 (1.709789) time: 0.720363 data: 0.000159 max mem: 14338 Epoch: [8/30] [1400/5004] eta: 0:42:57 lr: 0.000066 loss: 1.585678 (1.706977) time: 0.716478 data: 0.000209 max mem: 14338 Epoch: [8/30] [1450/5004] eta: 0:42:21 lr: 0.000066 loss: 1.615825 (1.706442) time: 0.714935 data: 0.000205 max mem: 14338 Epoch: [8/30] [1500/5004] eta: 0:41:45 lr: 0.000066 loss: 1.773682 (1.708631) time: 0.714119 data: 0.000223 max mem: 14338 Epoch: [8/30] [1550/5004] eta: 0:41:09 lr: 0.000066 loss: 1.618771 (1.708311) time: 0.711827 data: 0.000208 max mem: 14338 Epoch: [8/30] [1600/5004] eta: 0:40:34 lr: 0.000066 loss: 1.787032 (1.712159) time: 0.713886 data: 0.000216 max mem: 14338 Epoch: [8/30] [1650/5004] eta: 0:39:58 lr: 0.000066 loss: 1.721110 (1.710811) time: 0.713123 data: 0.000163 max mem: 14338 Epoch: [8/30] [1700/5004] eta: 0:39:22 lr: 0.000066 loss: 1.791893 (1.709690) time: 0.711244 data: 0.000180 max mem: 14338 Epoch: [8/30] [1750/5004] eta: 0:38:46 lr: 0.000066 loss: 1.846066 (1.711503) time: 0.708916 data: 0.000226 max mem: 14338 Epoch: [8/30] [1800/5004] eta: 0:38:10 lr: 0.000066 loss: 1.483627 (1.710628) time: 0.714328 data: 0.000217 max mem: 14338 Epoch: [8/30] [1850/5004] eta: 0:37:35 lr: 0.000066 loss: 1.577484 (1.710066) time: 0.720451 data: 0.000230 max mem: 14338 Epoch: [8/30] [1900/5004] eta: 0:36:59 lr: 0.000066 loss: 1.603445 (1.710262) time: 0.715874 data: 0.000226 max mem: 14338 Epoch: [8/30] [1950/5004] eta: 0:36:23 lr: 0.000066 loss: 1.628259 (1.711492) time: 0.716265 data: 0.000196 max mem: 14338 Epoch: [8/30] [2000/5004] eta: 0:35:47 lr: 0.000066 loss: 1.695544 (1.712136) time: 0.713943 data: 0.000161 max mem: 14338 Epoch: [8/30] [2050/5004] eta: 0:35:12 lr: 0.000066 loss: 1.613125 (1.711119) time: 0.719456 data: 0.000173 max mem: 14338 Epoch: [8/30] [2100/5004] eta: 0:34:36 lr: 0.000066 loss: 1.619825 (1.711508) time: 0.713098 data: 0.000215 max mem: 14338 Epoch: [8/30] [2150/5004] eta: 0:34:00 lr: 0.000066 loss: 1.751162 (1.712075) time: 0.710401 data: 0.000232 max mem: 14338 Epoch: [8/30] [2200/5004] eta: 0:33:24 lr: 0.000065 loss: 1.566505 (1.710206) time: 0.712701 data: 0.000197 max mem: 14338 Epoch: [8/30] [2250/5004] eta: 0:32:48 lr: 0.000065 loss: 1.899865 (1.711107) time: 0.712355 data: 0.000236 max mem: 14338 Epoch: [8/30] [2300/5004] eta: 0:32:13 lr: 0.000065 loss: 1.691464 (1.711422) time: 0.717125 data: 0.000222 max mem: 14338 Epoch: [8/30] [2350/5004] eta: 0:31:37 lr: 0.000065 loss: 1.664936 (1.710515) time: 0.715987 data: 0.000176 max mem: 14338 Epoch: [8/30] [2400/5004] eta: 0:31:01 lr: 0.000065 loss: 1.769166 (1.710510) time: 0.719559 data: 0.000225 max mem: 14338 Epoch: [8/30] [2450/5004] eta: 0:30:25 lr: 0.000065 loss: 1.643121 (1.711673) time: 0.715108 data: 0.000231 max mem: 14338 Epoch: [8/30] [2500/5004] eta: 0:29:49 lr: 0.000065 loss: 1.934157 (1.712526) time: 0.710728 data: 0.000230 max mem: 14338 Epoch: [8/30] [2550/5004] eta: 0:29:14 lr: 0.000065 loss: 1.705088 (1.712554) time: 0.711502 data: 0.000212 max mem: 14338 Epoch: [8/30] [2600/5004] eta: 0:28:38 lr: 0.000065 loss: 1.719138 (1.712653) time: 0.718017 data: 0.000233 max mem: 14338 Epoch: [8/30] [2650/5004] eta: 0:28:02 lr: 0.000065 loss: 1.591966 (1.713079) time: 0.711959 data: 0.000152 max mem: 14338 Epoch: [8/30] [2700/5004] eta: 0:27:27 lr: 0.000065 loss: 1.650734 (1.712191) time: 0.711669 data: 0.000174 max mem: 14338 Epoch: [8/30] [2750/5004] eta: 0:26:51 lr: 0.000065 loss: 1.738472 (1.713574) time: 0.716332 data: 0.000232 max mem: 14338 Epoch: [8/30] [2800/5004] eta: 0:26:15 lr: 0.000065 loss: 1.626695 (1.713381) time: 0.721275 data: 0.000227 max mem: 14338 Epoch: [8/30] [2850/5004] eta: 0:25:40 lr: 0.000065 loss: 1.648463 (1.712973) time: 0.723026 data: 0.000211 max mem: 14338 Epoch: [8/30] [2900/5004] eta: 0:25:04 lr: 0.000065 loss: 1.527600 (1.712110) time: 0.712946 data: 0.000231 max mem: 14338 Epoch: [8/30] [2950/5004] eta: 0:24:28 lr: 0.000065 loss: 1.701939 (1.713108) time: 0.709525 data: 0.000237 max mem: 14338 Epoch: [8/30] [3000/5004] eta: 0:23:52 lr: 0.000065 loss: 1.721849 (1.713514) time: 0.719789 data: 0.000162 max mem: 14338 Epoch: [8/30] [3050/5004] eta: 0:23:17 lr: 0.000065 loss: 1.738991 (1.714046) time: 0.717439 data: 0.000158 max mem: 14338 Epoch: [8/30] [3100/5004] eta: 0:22:41 lr: 0.000065 loss: 1.639409 (1.714148) time: 0.715041 data: 0.000227 max mem: 14338 Epoch: [8/30] [3150/5004] eta: 0:22:05 lr: 0.000065 loss: 1.550242 (1.714734) time: 0.710204 data: 0.000204 max mem: 14338 Epoch: [8/30] [3200/5004] eta: 0:21:29 lr: 0.000065 loss: 1.835037 (1.715278) time: 0.715465 data: 0.000240 max mem: 14338 Epoch: [8/30] [3250/5004] eta: 0:20:54 lr: 0.000065 loss: 1.579949 (1.714039) time: 0.714151 data: 0.000239 max mem: 14338 Epoch: [8/30] [3300/5004] eta: 0:20:18 lr: 0.000065 loss: 1.628970 (1.714807) time: 0.717558 data: 0.000222 max mem: 14338 Epoch: [8/30] [3350/5004] eta: 0:19:42 lr: 0.000065 loss: 1.628936 (1.714901) time: 0.717649 data: 0.000175 max mem: 14338 Epoch: [8/30] [3400/5004] eta: 0:19:06 lr: 0.000065 loss: 1.500931 (1.713751) time: 0.712238 data: 0.000173 max mem: 14338 Epoch: [8/30] [3450/5004] eta: 0:18:31 lr: 0.000065 loss: 1.515966 (1.713450) time: 0.712121 data: 0.000231 max mem: 14338 Epoch: [8/30] [3500/5004] eta: 0:17:55 lr: 0.000065 loss: 1.766354 (1.714648) time: 0.713084 data: 0.000191 max mem: 14338 Epoch: [8/30] [3550/5004] eta: 0:17:19 lr: 0.000065 loss: 1.531221 (1.714353) time: 0.714500 data: 0.000216 max mem: 14338 Epoch: [8/30] [3600/5004] eta: 0:16:43 lr: 0.000065 loss: 1.608438 (1.714278) time: 0.712864 data: 0.000213 max mem: 14338 Epoch: [8/30] [3650/5004] eta: 0:16:08 lr: 0.000065 loss: 1.695811 (1.715107) time: 0.712081 data: 0.000228 max mem: 14338 Epoch: [8/30] [3700/5004] eta: 0:15:32 lr: 0.000065 loss: 1.549339 (1.715032) time: 0.712681 data: 0.000171 max mem: 14338 Epoch: [8/30] [3750/5004] eta: 0:14:56 lr: 0.000064 loss: 1.713624 (1.715489) time: 0.716180 data: 0.000222 max mem: 14338 Epoch: [8/30] [3800/5004] eta: 0:14:20 lr: 0.000064 loss: 1.687984 (1.715700) time: 0.718461 data: 0.000217 max mem: 14338 Epoch: [8/30] [3850/5004] eta: 0:13:45 lr: 0.000064 loss: 1.644940 (1.715946) time: 0.714476 data: 0.000223 max mem: 14338 Epoch: [8/30] [3900/5004] eta: 0:13:09 lr: 0.000064 loss: 1.441852 (1.715264) time: 0.712705 data: 0.000213 max mem: 14338 Epoch: [8/30] [3950/5004] eta: 0:12:33 lr: 0.000064 loss: 1.770128 (1.715165) time: 0.710221 data: 0.000231 max mem: 14338 Epoch: [8/30] [4000/5004] eta: 0:11:57 lr: 0.000064 loss: 1.619225 (1.715113) time: 0.712531 data: 0.000163 max mem: 14338 Epoch: [8/30] [4050/5004] eta: 0:11:22 lr: 0.000064 loss: 1.696747 (1.714704) time: 0.715408 data: 0.000170 max mem: 14338 Epoch: [8/30] [4100/5004] eta: 0:10:46 lr: 0.000064 loss: 1.705403 (1.714624) time: 0.710356 data: 0.000240 max mem: 14338 Epoch: [8/30] [4150/5004] eta: 0:10:10 lr: 0.000064 loss: 1.633900 (1.714786) time: 0.713195 data: 0.000192 max mem: 14338 Epoch: [8/30] [4200/5004] eta: 0:09:34 lr: 0.000064 loss: 1.670347 (1.714726) time: 0.717363 data: 0.000234 max mem: 14338 Epoch: [8/30] [4250/5004] eta: 0:08:59 lr: 0.000064 loss: 1.584484 (1.713647) time: 0.722159 data: 0.000256 max mem: 14338 Epoch: [8/30] [4300/5004] eta: 0:08:23 lr: 0.000064 loss: 1.602081 (1.713280) time: 0.713829 data: 0.000226 max mem: 14338 Epoch: [8/30] [4350/5004] eta: 0:07:47 lr: 0.000064 loss: 1.647028 (1.713364) time: 0.713707 data: 0.000168 max mem: 14338 Epoch: [8/30] [4400/5004] eta: 0:07:11 lr: 0.000064 loss: 1.639291 (1.713402) time: 0.716257 data: 0.000174 max mem: 14338 Epoch: [8/30] [4450/5004] eta: 0:06:36 lr: 0.000064 loss: 1.703962 (1.713085) time: 0.719626 data: 0.000233 max mem: 14338 Epoch: [8/30] [4500/5004] eta: 0:06:00 lr: 0.000064 loss: 1.567561 (1.713304) time: 0.709738 data: 0.000243 max mem: 14338 Epoch: [8/30] [4550/5004] eta: 0:05:24 lr: 0.000064 loss: 1.435780 (1.712313) time: 0.710325 data: 0.000226 max mem: 14338 Epoch: [8/30] [4600/5004] eta: 0:04:48 lr: 0.000064 loss: 1.689862 (1.711992) time: 0.717106 data: 0.000213 max mem: 14338 Epoch: [8/30] [4650/5004] eta: 0:04:13 lr: 0.000064 loss: 1.671716 (1.711248) time: 0.718130 data: 0.000219 max mem: 14338 Epoch: [8/30] [4700/5004] eta: 0:03:37 lr: 0.000064 loss: 1.673771 (1.711204) time: 0.712642 data: 0.000170 max mem: 14338 Epoch: [8/30] [4750/5004] eta: 0:03:01 lr: 0.000064 loss: 1.491569 (1.710483) time: 0.716301 data: 0.000167 max mem: 14338 Epoch: [8/30] [4800/5004] eta: 0:02:25 lr: 0.000064 loss: 1.687018 (1.711048) time: 0.715409 data: 0.000202 max mem: 14338 Epoch: [8/30] [4850/5004] eta: 0:01:50 lr: 0.000064 loss: 1.508954 (1.711022) time: 0.712199 data: 0.000208 max mem: 14338 Epoch: [8/30] [4900/5004] eta: 0:01:14 lr: 0.000064 loss: 1.817857 (1.711364) time: 0.710422 data: 0.000199 max mem: 14338 Epoch: [8/30] [4950/5004] eta: 0:00:38 lr: 0.000064 loss: 1.692672 (1.710643) time: 0.712716 data: 0.000211 max mem: 14338 Epoch: [8/30] [5000/5004] eta: 0:00:02 lr: 0.000064 loss: 1.673325 (1.710454) time: 0.710885 data: 0.000828 max mem: 14338 Epoch: [8/30] [5003/5004] eta: 0:00:00 lr: 0.000064 loss: 1.673325 (1.710504) time: 0.708621 data: 0.000815 max mem: 14338 Epoch: [8/30] Total time: 0:59:38 (0.715118 s / it) Averaged stats: lr: 0.000064 loss: 1.673325 (1.708121) Test: [ 0/196] eta: 0:04:56 loss: 0.319616 (0.319616) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.514718 data: 1.076423 max mem: 14338 Test: [ 10/196] eta: 0:01:13 loss: 0.540167 (0.541683) acc1: 87.500000 (85.795455) acc5: 100.000000 (98.863636) time: 0.397845 data: 0.097978 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.574371 (0.544779) acc1: 87.500000 (85.416667) acc5: 100.000000 (98.214286) time: 0.285833 data: 0.000115 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.495108 (0.515746) acc1: 87.500000 (87.298387) acc5: 100.000000 (98.588710) time: 0.285838 data: 0.000108 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.409893 (0.521387) acc1: 87.500000 (87.042683) acc5: 100.000000 (98.170732) time: 0.287374 data: 0.000131 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.409893 (0.548299) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.549020) time: 0.296482 data: 0.000133 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.594223 (0.577708) acc1: 81.250000 (85.758197) acc5: 93.750000 (97.438525) time: 0.295615 data: 0.000126 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.636633 (0.595319) acc1: 81.250000 (85.299296) acc5: 100.000000 (97.623239) time: 0.287398 data: 0.000125 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.547011 (0.593102) acc1: 87.500000 (85.416667) acc5: 100.000000 (97.762346) time: 0.287484 data: 0.000116 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.556860 (0.618596) acc1: 81.250000 (84.752747) acc5: 100.000000 (97.527473) time: 0.286640 data: 0.000132 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.578304 (0.606405) acc1: 81.250000 (85.024752) acc5: 100.000000 (97.710396) time: 0.286899 data: 0.000138 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.524798 (0.596943) acc1: 87.500000 (84.966216) acc5: 100.000000 (97.860360) time: 0.286729 data: 0.000129 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.524798 (0.595006) acc1: 87.500000 (85.072314) acc5: 100.000000 (97.778926) time: 0.286286 data: 0.000145 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.647929 (0.609471) acc1: 81.250000 (84.828244) acc5: 93.750000 (97.709924) time: 0.287355 data: 0.000137 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.647929 (0.607725) acc1: 81.250000 (84.973404) acc5: 100.000000 (97.695035) time: 0.287150 data: 0.000132 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.514162 (0.612100) acc1: 87.500000 (84.850993) acc5: 100.000000 (97.640728) time: 0.286683 data: 0.000129 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.495672 (0.616510) acc1: 81.250000 (84.860248) acc5: 100.000000 (97.670807) time: 0.286933 data: 0.000112 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.498122 (0.613104) acc1: 87.500000 (85.014620) acc5: 100.000000 (97.697368) time: 0.286443 data: 0.000127 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.490625 (0.611875) acc1: 87.500000 (84.910221) acc5: 100.000000 (97.686464) time: 0.286373 data: 0.000140 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.394106 (0.601106) acc1: 81.250000 (85.078534) acc5: 100.000000 (97.709424) time: 0.291954 data: 0.000107 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.434338 (0.613555) acc1: 81.250000 (84.960000) acc5: 100.000000 (97.600000) time: 0.282441 data: 0.000096 max mem: 14338 Test: Total time: 0:00:57 (0.294752 s / it) * Acc@1 85.020 Acc@5 97.386 loss 0.625 Max accuracy: 85.02% Epoch: [9/30] [ 0/5004] eta: 2:37:40 lr: 0.000064 loss: 1.728308 (1.728308) time: 1.890650 data: 1.150367 max mem: 14338 Epoch: [9/30] [ 50/5004] eta: 1:00:49 lr: 0.000064 loss: 1.668098 (1.689742) time: 0.713616 data: 0.000180 max mem: 14338 Epoch: [9/30] [ 100/5004] eta: 0:59:20 lr: 0.000064 loss: 1.592435 (1.666906) time: 0.716703 data: 0.000223 max mem: 14338 Epoch: [9/30] [ 150/5004] eta: 0:58:25 lr: 0.000064 loss: 1.861360 (1.710680) time: 0.717845 data: 0.000202 max mem: 14338 Epoch: [9/30] [ 200/5004] eta: 0:57:44 lr: 0.000064 loss: 1.597075 (1.704397) time: 0.721902 data: 0.000221 max mem: 14338 Epoch: [9/30] [ 250/5004] eta: 0:57:02 lr: 0.000063 loss: 1.500029 (1.694876) time: 0.718422 data: 0.000188 max mem: 14338 Epoch: [9/30] [ 300/5004] eta: 0:56:21 lr: 0.000063 loss: 1.519306 (1.695463) time: 0.713350 data: 0.000169 max mem: 14338 Epoch: [9/30] [ 350/5004] eta: 0:55:41 lr: 0.000063 loss: 1.702717 (1.705288) time: 0.711260 data: 0.000176 max mem: 14338 Epoch: [9/30] [ 400/5004] eta: 0:55:02 lr: 0.000063 loss: 1.606603 (1.706162) time: 0.713137 data: 0.000225 max mem: 14338 Epoch: [9/30] [ 450/5004] eta: 0:54:25 lr: 0.000063 loss: 1.683404 (1.705050) time: 0.716711 data: 0.000215 max mem: 14338 Epoch: [9/30] [ 500/5004] eta: 0:53:47 lr: 0.000063 loss: 1.841181 (1.713567) time: 0.713299 data: 0.000208 max mem: 14338 Epoch: [9/30] [ 550/5004] eta: 0:53:11 lr: 0.000063 loss: 1.596392 (1.709758) time: 0.714447 data: 0.000220 max mem: 14338 Epoch: [9/30] [ 600/5004] eta: 0:52:34 lr: 0.000063 loss: 1.588545 (1.708782) time: 0.712346 data: 0.000238 max mem: 14338 Epoch: [9/30] [ 650/5004] eta: 0:51:57 lr: 0.000063 loss: 1.563546 (1.706105) time: 0.721596 data: 0.000156 max mem: 14338 Epoch: [9/30] [ 700/5004] eta: 0:51:22 lr: 0.000063 loss: 1.722661 (1.710791) time: 0.724289 data: 0.000174 max mem: 14338 Epoch: [9/30] [ 750/5004] eta: 0:50:46 lr: 0.000063 loss: 1.615112 (1.707710) time: 0.715507 data: 0.000216 max mem: 14338 Epoch: [9/30] [ 800/5004] eta: 0:50:11 lr: 0.000063 loss: 1.502254 (1.702091) time: 0.716563 data: 0.000214 max mem: 14338 Epoch: [9/30] [ 850/5004] eta: 0:49:34 lr: 0.000063 loss: 1.698083 (1.701225) time: 0.715739 data: 0.000223 max mem: 14338 Epoch: [9/30] [ 900/5004] eta: 0:48:59 lr: 0.000063 loss: 1.461262 (1.696760) time: 0.714324 data: 0.000194 max mem: 14338 Epoch: [9/30] [ 950/5004] eta: 0:48:23 lr: 0.000063 loss: 1.650891 (1.700015) time: 0.711044 data: 0.000211 max mem: 14338 Epoch: [9/30] [1000/5004] eta: 0:47:47 lr: 0.000063 loss: 1.736081 (1.700398) time: 0.715082 data: 0.000169 max mem: 14338 Epoch: [9/30] [1050/5004] eta: 0:47:11 lr: 0.000063 loss: 1.648453 (1.700007) time: 0.718440 data: 0.000218 max mem: 14338 Epoch: [9/30] [1100/5004] eta: 0:46:35 lr: 0.000063 loss: 1.594537 (1.698048) time: 0.717884 data: 0.000223 max mem: 14338 Epoch: [9/30] [1150/5004] eta: 0:45:59 lr: 0.000063 loss: 1.611496 (1.696127) time: 0.714979 data: 0.000219 max mem: 14338 Epoch: [9/30] [1200/5004] eta: 0:45:23 lr: 0.000063 loss: 1.622136 (1.699201) time: 0.715130 data: 0.000226 max mem: 14338 Epoch: [9/30] [1250/5004] eta: 0:44:46 lr: 0.000063 loss: 1.603156 (1.701924) time: 0.712827 data: 0.000215 max mem: 14338 Epoch: [9/30] [1300/5004] eta: 0:44:10 lr: 0.000063 loss: 1.607096 (1.701191) time: 0.710138 data: 0.000159 max mem: 14338 Epoch: [9/30] [1350/5004] eta: 0:43:34 lr: 0.000063 loss: 1.583720 (1.700178) time: 0.713130 data: 0.000179 max mem: 14338 Epoch: [9/30] [1400/5004] eta: 0:42:59 lr: 0.000063 loss: 1.605330 (1.700919) time: 0.715507 data: 0.000201 max mem: 14338 Epoch: [9/30] [1450/5004] eta: 0:42:22 lr: 0.000063 loss: 1.621984 (1.702294) time: 0.713255 data: 0.000223 max mem: 14338 Epoch: [9/30] [1500/5004] eta: 0:41:47 lr: 0.000063 loss: 1.729863 (1.702738) time: 0.709956 data: 0.000213 max mem: 14338 Epoch: [9/30] [1550/5004] eta: 0:41:10 lr: 0.000063 loss: 1.826115 (1.705819) time: 0.712789 data: 0.000206 max mem: 14338 Epoch: [9/30] [1600/5004] eta: 0:40:35 lr: 0.000063 loss: 1.614142 (1.705077) time: 0.720643 data: 0.000234 max mem: 14338 Epoch: [9/30] [1650/5004] eta: 0:39:59 lr: 0.000063 loss: 1.602298 (1.704926) time: 0.719062 data: 0.000172 max mem: 14338 Epoch: [9/30] [1700/5004] eta: 0:39:23 lr: 0.000063 loss: 1.769029 (1.706241) time: 0.713328 data: 0.000163 max mem: 14338 Epoch: [9/30] [1750/5004] eta: 0:38:47 lr: 0.000062 loss: 1.774929 (1.708860) time: 0.712526 data: 0.000224 max mem: 14338 Epoch: [9/30] [1800/5004] eta: 0:38:11 lr: 0.000062 loss: 1.619391 (1.707107) time: 0.716929 data: 0.000218 max mem: 14338 Epoch: [9/30] [1850/5004] eta: 0:37:36 lr: 0.000062 loss: 1.549242 (1.706071) time: 0.712930 data: 0.000241 max mem: 14338 Epoch: [9/30] [1900/5004] eta: 0:37:00 lr: 0.000062 loss: 1.727884 (1.707387) time: 0.713431 data: 0.000233 max mem: 14338 Epoch: [9/30] [1950/5004] eta: 0:36:24 lr: 0.000062 loss: 1.595080 (1.708058) time: 0.710357 data: 0.000235 max mem: 14338 Epoch: [9/30] [2000/5004] eta: 0:35:48 lr: 0.000062 loss: 1.788346 (1.710435) time: 0.718066 data: 0.000172 max mem: 14338 Epoch: [9/30] [2050/5004] eta: 0:35:13 lr: 0.000062 loss: 1.491721 (1.708255) time: 0.716603 data: 0.000164 max mem: 14338 Epoch: [9/30] [2100/5004] eta: 0:34:37 lr: 0.000062 loss: 1.477139 (1.705670) time: 0.721286 data: 0.000238 max mem: 14338 Epoch: [9/30] [2150/5004] eta: 0:34:01 lr: 0.000062 loss: 1.584271 (1.705558) time: 0.717551 data: 0.000224 max mem: 14338 Epoch: [9/30] [2200/5004] eta: 0:33:25 lr: 0.000062 loss: 1.709885 (1.705301) time: 0.714752 data: 0.000187 max mem: 14338 Epoch: [9/30] [2250/5004] eta: 0:32:50 lr: 0.000062 loss: 1.546327 (1.705327) time: 0.717075 data: 0.000234 max mem: 14338 Epoch: [9/30] [2300/5004] eta: 0:32:14 lr: 0.000062 loss: 1.736483 (1.706132) time: 0.713323 data: 0.000235 max mem: 14338 Epoch: [9/30] [2350/5004] eta: 0:31:38 lr: 0.000062 loss: 1.565423 (1.706208) time: 0.712056 data: 0.000181 max mem: 14338 Epoch: [9/30] [2400/5004] eta: 0:31:02 lr: 0.000062 loss: 1.630883 (1.705882) time: 0.716313 data: 0.000231 max mem: 14338 Epoch: [9/30] [2450/5004] eta: 0:30:26 lr: 0.000062 loss: 1.598070 (1.706742) time: 0.712430 data: 0.000235 max mem: 14338 Epoch: [9/30] [2500/5004] eta: 0:29:51 lr: 0.000062 loss: 1.613067 (1.705210) time: 0.716607 data: 0.000214 max mem: 14338 Epoch: [9/30] [2550/5004] eta: 0:29:15 lr: 0.000062 loss: 1.654758 (1.704397) time: 0.714287 data: 0.000227 max mem: 14338 Epoch: [9/30] [2600/5004] eta: 0:28:39 lr: 0.000062 loss: 1.781564 (1.704986) time: 0.720449 data: 0.000224 max mem: 14338 Epoch: [9/30] [2650/5004] eta: 0:28:03 lr: 0.000062 loss: 1.602203 (1.705150) time: 0.718300 data: 0.000155 max mem: 14338 Epoch: [9/30] [2700/5004] eta: 0:27:27 lr: 0.000062 loss: 1.629358 (1.706460) time: 0.709076 data: 0.000183 max mem: 14338 Epoch: [9/30] [2750/5004] eta: 0:26:52 lr: 0.000062 loss: 1.572100 (1.705530) time: 0.716024 data: 0.000228 max mem: 14338 Epoch: [9/30] [2800/5004] eta: 0:26:16 lr: 0.000062 loss: 1.762804 (1.706823) time: 0.715812 data: 0.000212 max mem: 14338 Epoch: [9/30] [2850/5004] eta: 0:25:40 lr: 0.000062 loss: 1.649618 (1.706669) time: 0.713179 data: 0.000196 max mem: 14338 Epoch: [9/30] [2900/5004] eta: 0:25:04 lr: 0.000062 loss: 1.667343 (1.706279) time: 0.709325 data: 0.000228 max mem: 14338 Epoch: [9/30] [2950/5004] eta: 0:24:29 lr: 0.000062 loss: 1.754345 (1.706603) time: 0.711203 data: 0.000243 max mem: 14338 Epoch: [9/30] [3000/5004] eta: 0:23:53 lr: 0.000062 loss: 1.803908 (1.707850) time: 0.721192 data: 0.000162 max mem: 14338 Epoch: [9/30] [3050/5004] eta: 0:23:17 lr: 0.000062 loss: 1.475334 (1.706759) time: 0.724937 data: 0.000165 max mem: 14338 Epoch: [9/30] [3100/5004] eta: 0:22:41 lr: 0.000062 loss: 1.523975 (1.705988) time: 0.717036 data: 0.000209 max mem: 14338 Epoch: [9/30] [3150/5004] eta: 0:22:06 lr: 0.000061 loss: 1.748653 (1.705632) time: 0.712509 data: 0.000219 max mem: 14338 Epoch: [9/30] [3200/5004] eta: 0:21:30 lr: 0.000061 loss: 1.623012 (1.704745) time: 0.715213 data: 0.000229 max mem: 14338 Epoch: [9/30] [3250/5004] eta: 0:20:54 lr: 0.000061 loss: 1.735863 (1.704299) time: 0.714920 data: 0.000219 max mem: 14338 Epoch: [9/30] [3300/5004] eta: 0:20:18 lr: 0.000061 loss: 1.684180 (1.703781) time: 0.710621 data: 0.000220 max mem: 14338 Epoch: [9/30] [3350/5004] eta: 0:19:43 lr: 0.000061 loss: 1.706304 (1.704512) time: 0.709725 data: 0.000164 max mem: 14338 Epoch: [9/30] [3400/5004] eta: 0:19:07 lr: 0.000061 loss: 1.727264 (1.705599) time: 0.714744 data: 0.000173 max mem: 14338 Epoch: [9/30] [3450/5004] eta: 0:18:31 lr: 0.000061 loss: 1.735980 (1.705417) time: 0.722421 data: 0.000215 max mem: 14338 Epoch: [9/30] [3500/5004] eta: 0:17:55 lr: 0.000061 loss: 1.439966 (1.704710) time: 0.719012 data: 0.000190 max mem: 14338 Epoch: [9/30] [3550/5004] eta: 0:17:19 lr: 0.000061 loss: 1.833164 (1.706228) time: 0.715620 data: 0.000232 max mem: 14338 Epoch: [9/30] [3600/5004] eta: 0:16:44 lr: 0.000061 loss: 1.624461 (1.706470) time: 0.718423 data: 0.000240 max mem: 14338 Epoch: [9/30] [3650/5004] eta: 0:16:08 lr: 0.000061 loss: 1.692472 (1.707023) time: 0.713177 data: 0.000224 max mem: 14338 Epoch: [9/30] [3700/5004] eta: 0:15:32 lr: 0.000061 loss: 1.759477 (1.707850) time: 0.710130 data: 0.000185 max mem: 14338 Epoch: [9/30] [3750/5004] eta: 0:14:56 lr: 0.000061 loss: 1.441344 (1.706977) time: 0.709848 data: 0.000212 max mem: 14338 Epoch: [9/30] [3800/5004] eta: 0:14:21 lr: 0.000061 loss: 1.515844 (1.705444) time: 0.713331 data: 0.000222 max mem: 14338 Epoch: [9/30] [3850/5004] eta: 0:13:45 lr: 0.000061 loss: 1.669135 (1.705282) time: 0.711582 data: 0.000218 max mem: 14338 Epoch: [9/30] [3900/5004] eta: 0:13:09 lr: 0.000061 loss: 1.603653 (1.704633) time: 0.716386 data: 0.000239 max mem: 14338 Epoch: [9/30] [3950/5004] eta: 0:12:33 lr: 0.000061 loss: 1.637067 (1.704450) time: 0.719411 data: 0.000227 max mem: 14338 Epoch: [9/30] [4000/5004] eta: 0:11:58 lr: 0.000061 loss: 1.402765 (1.703216) time: 0.718782 data: 0.000174 max mem: 14338 Epoch: [9/30] [4050/5004] eta: 0:11:22 lr: 0.000061 loss: 1.687690 (1.703648) time: 0.719382 data: 0.000165 max mem: 14338 Epoch: [9/30] [4100/5004] eta: 0:10:46 lr: 0.000061 loss: 1.813633 (1.703631) time: 0.710560 data: 0.000223 max mem: 14338 Epoch: [9/30] [4150/5004] eta: 0:10:10 lr: 0.000061 loss: 1.504745 (1.702640) time: 0.709819 data: 0.000196 max mem: 14338 Epoch: [9/30] [4200/5004] eta: 0:09:34 lr: 0.000061 loss: 1.680883 (1.702813) time: 0.712129 data: 0.000236 max mem: 14338 Epoch: [9/30] [4250/5004] eta: 0:08:59 lr: 0.000061 loss: 1.654143 (1.702576) time: 0.716800 data: 0.000203 max mem: 14338 Epoch: [9/30] [4300/5004] eta: 0:08:23 lr: 0.000061 loss: 1.693489 (1.702585) time: 0.710706 data: 0.000230 max mem: 14338 Epoch: [9/30] [4350/5004] eta: 0:07:47 lr: 0.000061 loss: 1.794862 (1.702159) time: 0.718854 data: 0.000173 max mem: 14338 Epoch: [9/30] [4400/5004] eta: 0:07:11 lr: 0.000061 loss: 1.708294 (1.702805) time: 0.721369 data: 0.000170 max mem: 14338 Epoch: [9/30] [4450/5004] eta: 0:06:36 lr: 0.000061 loss: 1.600422 (1.702541) time: 0.720636 data: 0.000229 max mem: 14338 Epoch: [9/30] [4500/5004] eta: 0:06:00 lr: 0.000061 loss: 1.716965 (1.702435) time: 0.721097 data: 0.000239 max mem: 14338 Epoch: [9/30] [4550/5004] eta: 0:05:24 lr: 0.000061 loss: 1.680469 (1.702658) time: 0.709835 data: 0.000226 max mem: 14338 Epoch: [9/30] [4600/5004] eta: 0:04:48 lr: 0.000060 loss: 1.545811 (1.702254) time: 0.713325 data: 0.000208 max mem: 14338 Epoch: [9/30] [4650/5004] eta: 0:04:13 lr: 0.000060 loss: 1.658967 (1.702800) time: 0.710642 data: 0.000216 max mem: 14338 Epoch: [9/30] [4700/5004] eta: 0:03:37 lr: 0.000060 loss: 1.502329 (1.702375) time: 0.717914 data: 0.000182 max mem: 14338 Epoch: [9/30] [4750/5004] eta: 0:03:01 lr: 0.000060 loss: 1.663690 (1.702894) time: 0.710912 data: 0.000169 max mem: 14338 Epoch: [9/30] [4800/5004] eta: 0:02:25 lr: 0.000060 loss: 1.580704 (1.702433) time: 0.712650 data: 0.000197 max mem: 14338 Epoch: [9/30] [4850/5004] eta: 0:01:50 lr: 0.000060 loss: 1.553692 (1.702665) time: 0.714498 data: 0.000221 max mem: 14338 Epoch: [9/30] [4900/5004] eta: 0:01:14 lr: 0.000060 loss: 1.641931 (1.702573) time: 0.710770 data: 0.000224 max mem: 14338 Epoch: [9/30] [4950/5004] eta: 0:00:38 lr: 0.000060 loss: 1.633686 (1.702830) time: 0.716167 data: 0.000229 max mem: 14338 Epoch: [9/30] [5000/5004] eta: 0:00:02 lr: 0.000060 loss: 1.643301 (1.702279) time: 0.711445 data: 0.000851 max mem: 14338 Epoch: [9/30] [5003/5004] eta: 0:00:00 lr: 0.000060 loss: 1.643301 (1.702489) time: 0.711573 data: 0.000844 max mem: 14338 Epoch: [9/30] Total time: 0:59:38 (0.715203 s / it) Averaged stats: lr: 0.000060 loss: 1.643301 (1.695565) Test: [ 0/196] eta: 0:05:17 loss: 0.405773 (0.405773) acc1: 87.500000 (87.500000) acc5: 100.000000 (100.000000) time: 1.619652 data: 1.251574 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 0.479929 (0.533826) acc1: 87.500000 (85.795455) acc5: 100.000000 (98.863636) time: 0.406025 data: 0.113903 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.586061 (0.531139) acc1: 87.500000 (86.309524) acc5: 100.000000 (98.214286) time: 0.286466 data: 0.000122 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.523628 (0.511714) acc1: 87.500000 (87.500000) acc5: 100.000000 (98.387097) time: 0.287503 data: 0.000121 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.392662 (0.515972) acc1: 87.500000 (86.890244) acc5: 100.000000 (98.323171) time: 0.287026 data: 0.000143 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.446760 (0.543632) acc1: 87.500000 (86.764706) acc5: 100.000000 (97.671569) time: 0.287728 data: 0.000137 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.541447 (0.572616) acc1: 87.500000 (86.270492) acc5: 100.000000 (97.745902) time: 0.288020 data: 0.000141 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.642902 (0.590749) acc1: 87.500000 (85.915493) acc5: 100.000000 (97.887324) time: 0.287784 data: 0.000145 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.559021 (0.594742) acc1: 87.500000 (85.648148) acc5: 100.000000 (97.916667) time: 0.287137 data: 0.000132 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.559021 (0.616273) acc1: 81.250000 (85.233516) acc5: 100.000000 (97.664835) time: 0.286585 data: 0.000142 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.582435 (0.606555) acc1: 81.250000 (85.272277) acc5: 100.000000 (97.834158) time: 0.286226 data: 0.000145 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.567976 (0.598495) acc1: 87.500000 (85.247748) acc5: 100.000000 (97.916667) time: 0.285686 data: 0.000140 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.523033 (0.595360) acc1: 87.500000 (85.330579) acc5: 100.000000 (97.830579) time: 0.285991 data: 0.000142 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.587949 (0.611707) acc1: 87.500000 (84.971374) acc5: 100.000000 (97.805344) time: 0.293291 data: 0.000142 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.693192 (0.610948) acc1: 81.250000 (85.106383) acc5: 100.000000 (97.783688) time: 0.293278 data: 0.000149 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.536901 (0.616417) acc1: 87.500000 (85.016556) acc5: 93.750000 (97.682119) time: 0.286731 data: 0.000144 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.525068 (0.620954) acc1: 87.500000 (84.976708) acc5: 100.000000 (97.670807) time: 0.286997 data: 0.000128 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.517663 (0.617079) acc1: 81.250000 (85.124269) acc5: 100.000000 (97.733918) time: 0.286999 data: 0.000148 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.486197 (0.616757) acc1: 81.250000 (85.013812) acc5: 100.000000 (97.755525) time: 0.286026 data: 0.000160 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.370224 (0.605632) acc1: 81.250000 (85.209424) acc5: 100.000000 (97.774869) time: 0.283235 data: 0.000114 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.527756 (0.614790) acc1: 81.250000 (85.088000) acc5: 100.000000 (97.696000) time: 0.273485 data: 0.000096 max mem: 14338 Test: Total time: 0:00:57 (0.294170 s / it) * Acc@1 85.014 Acc@5 97.358 loss 0.624 Max accuracy: 85.02% Epoch: [10/30] [ 0/5004] eta: 2:40:15 lr: 0.000060 loss: 1.630713 (1.630713) time: 1.921615 data: 1.192449 max mem: 14338 Epoch: [10/30] [ 50/5004] eta: 1:00:48 lr: 0.000060 loss: 1.629109 (1.723294) time: 0.712703 data: 0.000196 max mem: 14338 Epoch: [10/30] [ 100/5004] eta: 0:59:18 lr: 0.000060 loss: 1.562114 (1.674665) time: 0.709921 data: 0.000229 max mem: 14338 Epoch: [10/30] [ 150/5004] eta: 0:58:24 lr: 0.000060 loss: 1.683280 (1.682858) time: 0.713503 data: 0.000219 max mem: 14338 Epoch: [10/30] [ 200/5004] eta: 0:57:42 lr: 0.000060 loss: 1.570379 (1.683776) time: 0.714987 data: 0.000217 max mem: 14338 Epoch: [10/30] [ 250/5004] eta: 0:57:04 lr: 0.000060 loss: 1.778551 (1.676687) time: 0.717104 data: 0.000213 max mem: 14338 Epoch: [10/30] [ 300/5004] eta: 0:56:23 lr: 0.000060 loss: 1.484686 (1.664549) time: 0.713579 data: 0.000184 max mem: 14338 Epoch: [10/30] [ 350/5004] eta: 0:55:44 lr: 0.000060 loss: 1.756883 (1.679745) time: 0.717231 data: 0.000176 max mem: 14338 Epoch: [10/30] [ 400/5004] eta: 0:55:06 lr: 0.000060 loss: 1.719236 (1.679674) time: 0.718032 data: 0.000213 max mem: 14338 Epoch: [10/30] [ 450/5004] eta: 0:54:28 lr: 0.000060 loss: 1.570732 (1.674328) time: 0.719070 data: 0.000221 max mem: 14338 Epoch: [10/30] [ 500/5004] eta: 0:53:50 lr: 0.000060 loss: 1.530621 (1.668648) time: 0.713569 data: 0.000229 max mem: 14338 Epoch: [10/30] [ 550/5004] eta: 0:53:12 lr: 0.000060 loss: 1.608488 (1.668312) time: 0.709643 data: 0.000243 max mem: 14338 Epoch: [10/30] [ 600/5004] eta: 0:52:35 lr: 0.000060 loss: 1.656449 (1.673038) time: 0.714643 data: 0.000220 max mem: 14338 Epoch: [10/30] [ 650/5004] eta: 0:51:59 lr: 0.000060 loss: 1.613646 (1.676608) time: 0.713930 data: 0.000168 max mem: 14338 Epoch: [10/30] [ 700/5004] eta: 0:51:22 lr: 0.000060 loss: 1.696039 (1.680962) time: 0.711437 data: 0.000173 max mem: 14338 Epoch: [10/30] [ 750/5004] eta: 0:50:46 lr: 0.000060 loss: 1.649240 (1.679401) time: 0.712215 data: 0.000226 max mem: 14338 Epoch: [10/30] [ 800/5004] eta: 0:50:10 lr: 0.000060 loss: 1.691376 (1.678958) time: 0.719799 data: 0.000236 max mem: 14338 Epoch: [10/30] [ 850/5004] eta: 0:49:34 lr: 0.000060 loss: 1.647568 (1.676514) time: 0.718053 data: 0.000228 max mem: 14338 Epoch: [10/30] [ 900/5004] eta: 0:48:58 lr: 0.000060 loss: 1.816931 (1.677776) time: 0.717652 data: 0.000225 max mem: 14338 Epoch: [10/30] [ 950/5004] eta: 0:48:21 lr: 0.000059 loss: 1.622029 (1.679690) time: 0.710754 data: 0.000215 max mem: 14338 Epoch: [10/30] [1000/5004] eta: 0:47:46 lr: 0.000059 loss: 1.631694 (1.682401) time: 0.717364 data: 0.000168 max mem: 14338 Epoch: [10/30] [1050/5004] eta: 0:47:10 lr: 0.000059 loss: 1.560976 (1.682408) time: 0.711953 data: 0.000213 max mem: 14338 Epoch: [10/30] [1100/5004] eta: 0:46:34 lr: 0.000059 loss: 1.586191 (1.680884) time: 0.713567 data: 0.000233 max mem: 14338 Epoch: [10/30] [1150/5004] eta: 0:45:58 lr: 0.000059 loss: 1.586325 (1.682309) time: 0.711490 data: 0.000224 max mem: 14338 Epoch: [10/30] [1200/5004] eta: 0:45:22 lr: 0.000059 loss: 1.632839 (1.681548) time: 0.715172 data: 0.000219 max mem: 14338 Epoch: [10/30] [1250/5004] eta: 0:44:46 lr: 0.000059 loss: 1.533666 (1.685192) time: 0.716138 data: 0.000205 max mem: 14338 Epoch: [10/30] [1300/5004] eta: 0:44:10 lr: 0.000059 loss: 1.727190 (1.688117) time: 0.713268 data: 0.000178 max mem: 14338 Epoch: [10/30] [1350/5004] eta: 0:43:34 lr: 0.000059 loss: 1.647962 (1.687336) time: 0.713764 data: 0.000170 max mem: 14338 Epoch: [10/30] [1400/5004] eta: 0:42:58 lr: 0.000059 loss: 1.720113 (1.686531) time: 0.721427 data: 0.000215 max mem: 14338 Epoch: [10/30] [1450/5004] eta: 0:42:22 lr: 0.000059 loss: 1.593861 (1.685886) time: 0.717845 data: 0.000202 max mem: 14338 Epoch: [10/30] [1500/5004] eta: 0:41:46 lr: 0.000059 loss: 1.577293 (1.686521) time: 0.712734 data: 0.000216 max mem: 14338 Epoch: [10/30] [1550/5004] eta: 0:41:10 lr: 0.000059 loss: 1.575126 (1.685524) time: 0.710217 data: 0.000198 max mem: 14338 Epoch: [10/30] [1600/5004] eta: 0:40:34 lr: 0.000059 loss: 1.751493 (1.685394) time: 0.711406 data: 0.000217 max mem: 14338 Epoch: [10/30] [1650/5004] eta: 0:39:58 lr: 0.000059 loss: 1.557963 (1.686209) time: 0.713002 data: 0.000167 max mem: 14338 Epoch: [10/30] [1700/5004] eta: 0:39:22 lr: 0.000059 loss: 1.669563 (1.686753) time: 0.709711 data: 0.000171 max mem: 14338 Epoch: [10/30] [1750/5004] eta: 0:38:46 lr: 0.000059 loss: 1.719224 (1.688270) time: 0.716836 data: 0.000232 max mem: 14338 Epoch: [10/30] [1800/5004] eta: 0:38:11 lr: 0.000059 loss: 1.653767 (1.686612) time: 0.715446 data: 0.000239 max mem: 14338 Epoch: [10/30] [1850/5004] eta: 0:37:35 lr: 0.000059 loss: 1.594961 (1.686883) time: 0.722494 data: 0.000238 max mem: 14338 Epoch: [10/30] [1900/5004] eta: 0:36:59 lr: 0.000059 loss: 1.619878 (1.684866) time: 0.716887 data: 0.000219 max mem: 14338 Epoch: [10/30] [1950/5004] eta: 0:36:23 lr: 0.000059 loss: 1.640826 (1.686169) time: 0.711165 data: 0.000214 max mem: 14338 Epoch: [10/30] [2000/5004] eta: 0:35:48 lr: 0.000059 loss: 1.621806 (1.683651) time: 0.714865 data: 0.000160 max mem: 14338 Epoch: [10/30] [2050/5004] eta: 0:35:12 lr: 0.000059 loss: 1.454257 (1.683364) time: 0.711052 data: 0.000172 max mem: 14338 Epoch: [10/30] [2100/5004] eta: 0:34:36 lr: 0.000059 loss: 1.780848 (1.684678) time: 0.712047 data: 0.000217 max mem: 14338 Epoch: [10/30] [2150/5004] eta: 0:34:00 lr: 0.000059 loss: 1.646283 (1.684839) time: 0.710969 data: 0.000210 max mem: 14338 Epoch: [10/30] [2200/5004] eta: 0:33:24 lr: 0.000059 loss: 1.533767 (1.683083) time: 0.716705 data: 0.000186 max mem: 14338 Epoch: [10/30] [2250/5004] eta: 0:32:48 lr: 0.000059 loss: 1.786681 (1.683993) time: 0.718465 data: 0.000220 max mem: 14338 Epoch: [10/30] [2300/5004] eta: 0:32:13 lr: 0.000059 loss: 1.568945 (1.682739) time: 0.712030 data: 0.000230 max mem: 14338 Epoch: [10/30] [2350/5004] eta: 0:31:37 lr: 0.000058 loss: 1.597768 (1.681990) time: 0.718299 data: 0.000171 max mem: 14338 Epoch: [10/30] [2400/5004] eta: 0:31:01 lr: 0.000058 loss: 1.732717 (1.683342) time: 0.712862 data: 0.000211 max mem: 14338 Epoch: [10/30] [2450/5004] eta: 0:30:25 lr: 0.000058 loss: 1.640756 (1.683772) time: 0.715021 data: 0.000238 max mem: 14338 Epoch: [10/30] [2500/5004] eta: 0:29:49 lr: 0.000058 loss: 1.657047 (1.684492) time: 0.710483 data: 0.000225 max mem: 14338 Epoch: [10/30] [2550/5004] eta: 0:29:14 lr: 0.000058 loss: 1.461697 (1.684954) time: 0.714334 data: 0.000215 max mem: 14338 Epoch: [10/30] [2600/5004] eta: 0:28:38 lr: 0.000058 loss: 1.523706 (1.684070) time: 0.712525 data: 0.000211 max mem: 14338 Epoch: [10/30] [2650/5004] eta: 0:28:02 lr: 0.000058 loss: 1.513686 (1.683872) time: 0.712797 data: 0.000177 max mem: 14338 Epoch: [10/30] [2700/5004] eta: 0:27:27 lr: 0.000058 loss: 1.507550 (1.683245) time: 0.723589 data: 0.000180 max mem: 14338 Epoch: [10/30] [2750/5004] eta: 0:26:51 lr: 0.000058 loss: 1.734096 (1.684278) time: 0.711763 data: 0.000234 max mem: 14338 Epoch: [10/30] [2800/5004] eta: 0:26:15 lr: 0.000058 loss: 1.735302 (1.684814) time: 0.721698 data: 0.000213 max mem: 14338 Epoch: [10/30] [2850/5004] eta: 0:25:39 lr: 0.000058 loss: 1.522721 (1.683817) time: 0.712211 data: 0.000194 max mem: 14338 Epoch: [10/30] [2900/5004] eta: 0:25:03 lr: 0.000058 loss: 1.506666 (1.682641) time: 0.711037 data: 0.000227 max mem: 14338 Epoch: [10/30] [2950/5004] eta: 0:24:28 lr: 0.000058 loss: 1.462942 (1.682776) time: 0.715310 data: 0.000240 max mem: 14338 Epoch: [10/30] [3000/5004] eta: 0:23:52 lr: 0.000058 loss: 1.607620 (1.682503) time: 0.716542 data: 0.000158 max mem: 14338 Epoch: [10/30] [3050/5004] eta: 0:23:16 lr: 0.000058 loss: 1.529574 (1.682220) time: 0.713665 data: 0.000164 max mem: 14338 Epoch: [10/30] [3100/5004] eta: 0:22:41 lr: 0.000058 loss: 1.735891 (1.682515) time: 0.712201 data: 0.000247 max mem: 14338 Epoch: [10/30] [3150/5004] eta: 0:22:05 lr: 0.000058 loss: 1.617072 (1.682376) time: 0.711853 data: 0.000238 max mem: 14338 Epoch: [10/30] [3200/5004] eta: 0:21:29 lr: 0.000058 loss: 1.739059 (1.683054) time: 0.721286 data: 0.000228 max mem: 14338 Epoch: [10/30] [3250/5004] eta: 0:20:53 lr: 0.000058 loss: 1.511822 (1.682551) time: 0.718683 data: 0.000214 max mem: 14338 Epoch: [10/30] [3300/5004] eta: 0:20:18 lr: 0.000058 loss: 1.744834 (1.683205) time: 0.715359 data: 0.000242 max mem: 14338 Epoch: [10/30] [3350/5004] eta: 0:19:42 lr: 0.000058 loss: 1.750079 (1.682594) time: 0.714009 data: 0.000163 max mem: 14338 Epoch: [10/30] [3400/5004] eta: 0:19:06 lr: 0.000058 loss: 1.658815 (1.683303) time: 0.712061 data: 0.000171 max mem: 14338 Epoch: [10/30] [3450/5004] eta: 0:18:30 lr: 0.000058 loss: 1.671760 (1.683555) time: 0.718187 data: 0.000247 max mem: 14338 Epoch: [10/30] [3500/5004] eta: 0:17:55 lr: 0.000058 loss: 1.689981 (1.683764) time: 0.711833 data: 0.000211 max mem: 14338 Epoch: [10/30] [3550/5004] eta: 0:17:19 lr: 0.000058 loss: 1.705732 (1.684915) time: 0.710170 data: 0.000239 max mem: 14338 Epoch: [10/30] [3600/5004] eta: 0:16:43 lr: 0.000058 loss: 1.695608 (1.685069) time: 0.716307 data: 0.000222 max mem: 14338 Epoch: [10/30] [3650/5004] eta: 0:16:08 lr: 0.000058 loss: 1.714080 (1.685753) time: 0.722010 data: 0.000216 max mem: 14338 Epoch: [10/30] [3700/5004] eta: 0:15:32 lr: 0.000057 loss: 1.636497 (1.685952) time: 0.718848 data: 0.000178 max mem: 14338 Epoch: [10/30] [3750/5004] eta: 0:14:56 lr: 0.000057 loss: 1.759746 (1.686687) time: 0.714948 data: 0.000227 max mem: 14338 Epoch: [10/30] [3800/5004] eta: 0:14:20 lr: 0.000057 loss: 1.625411 (1.686848) time: 0.714527 data: 0.000231 max mem: 14338 Epoch: [10/30] [3850/5004] eta: 0:13:44 lr: 0.000057 loss: 1.534062 (1.685792) time: 0.711652 data: 0.000214 max mem: 14338 Epoch: [10/30] [3900/5004] eta: 0:13:09 lr: 0.000057 loss: 1.490011 (1.685731) time: 0.710359 data: 0.000216 max mem: 14338 Epoch: [10/30] [3950/5004] eta: 0:12:33 lr: 0.000057 loss: 1.477668 (1.685213) time: 0.714430 data: 0.000231 max mem: 14338 Epoch: [10/30] [4000/5004] eta: 0:11:57 lr: 0.000057 loss: 1.696051 (1.685556) time: 0.712553 data: 0.000184 max mem: 14338 Epoch: [10/30] [4050/5004] eta: 0:11:21 lr: 0.000057 loss: 1.579759 (1.685757) time: 0.712044 data: 0.000148 max mem: 14338 Epoch: [10/30] [4100/5004] eta: 0:10:46 lr: 0.000057 loss: 1.563867 (1.685575) time: 0.717509 data: 0.000219 max mem: 14338 Epoch: [10/30] [4150/5004] eta: 0:10:10 lr: 0.000057 loss: 1.680888 (1.685782) time: 0.713092 data: 0.000193 max mem: 14338 Epoch: [10/30] [4200/5004] eta: 0:09:34 lr: 0.000057 loss: 1.559120 (1.686000) time: 0.722390 data: 0.000220 max mem: 14338 Epoch: [10/30] [4250/5004] eta: 0:08:59 lr: 0.000057 loss: 1.650254 (1.686421) time: 0.719768 data: 0.000220 max mem: 14338 Epoch: [10/30] [4300/5004] eta: 0:08:23 lr: 0.000057 loss: 1.624281 (1.686120) time: 0.709903 data: 0.000219 max mem: 14338 Epoch: [10/30] [4350/5004] eta: 0:07:47 lr: 0.000057 loss: 1.576585 (1.685460) time: 0.713962 data: 0.000173 max mem: 14338 Epoch: [10/30] [4400/5004] eta: 0:07:11 lr: 0.000057 loss: 1.622488 (1.685519) time: 0.714335 data: 0.000158 max mem: 14338 Epoch: [10/30] [4450/5004] eta: 0:06:36 lr: 0.000057 loss: 1.565745 (1.685809) time: 0.718575 data: 0.000216 max mem: 14338 Epoch: [10/30] [4500/5004] eta: 0:06:00 lr: 0.000057 loss: 1.606867 (1.685179) time: 0.713513 data: 0.000244 max mem: 14338 Epoch: [10/30] [4550/5004] eta: 0:05:24 lr: 0.000057 loss: 1.430223 (1.684687) time: 0.711263 data: 0.000223 max mem: 14338 Epoch: [10/30] [4600/5004] eta: 0:04:48 lr: 0.000057 loss: 1.581086 (1.685032) time: 0.716225 data: 0.000208 max mem: 14338 Epoch: [10/30] [4650/5004] eta: 0:04:13 lr: 0.000057 loss: 1.549208 (1.684771) time: 0.714930 data: 0.000216 max mem: 14338 Epoch: [10/30] [4700/5004] eta: 0:03:37 lr: 0.000057 loss: 1.700619 (1.685531) time: 0.711166 data: 0.000179 max mem: 14338 Epoch: [10/30] [4750/5004] eta: 0:03:01 lr: 0.000057 loss: 1.788686 (1.686378) time: 0.710324 data: 0.000169 max mem: 14338 Epoch: [10/30] [4800/5004] eta: 0:02:25 lr: 0.000057 loss: 1.555478 (1.685780) time: 0.712145 data: 0.000198 max mem: 14338 Epoch: [10/30] [4850/5004] eta: 0:01:50 lr: 0.000057 loss: 1.708162 (1.685573) time: 0.711175 data: 0.000215 max mem: 14338 Epoch: [10/30] [4900/5004] eta: 0:01:14 lr: 0.000057 loss: 1.639844 (1.685605) time: 0.713173 data: 0.000226 max mem: 14338 Epoch: [10/30] [4950/5004] eta: 0:00:38 lr: 0.000057 loss: 1.613880 (1.685847) time: 0.710317 data: 0.000217 max mem: 14338 Epoch: [10/30] [5000/5004] eta: 0:00:02 lr: 0.000056 loss: 1.485074 (1.684619) time: 0.710133 data: 0.000835 max mem: 14338 Epoch: [10/30] [5003/5004] eta: 0:00:00 lr: 0.000056 loss: 1.485785 (1.684565) time: 0.706457 data: 0.000824 max mem: 14338 Epoch: [10/30] Total time: 0:59:37 (0.714994 s / it) Averaged stats: lr: 0.000056 loss: 1.485785 (1.684362) Test: [ 0/196] eta: 0:05:05 loss: 0.350448 (0.350448) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.558717 data: 1.142058 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.454856 (0.542636) acc1: 87.500000 (86.363636) acc5: 100.000000 (98.863636) time: 0.401479 data: 0.103946 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.518972 (0.540317) acc1: 87.500000 (86.607143) acc5: 100.000000 (98.214286) time: 0.286638 data: 0.000125 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.492855 (0.514861) acc1: 87.500000 (87.701613) acc5: 100.000000 (98.185484) time: 0.287458 data: 0.000121 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.386699 (0.521683) acc1: 87.500000 (87.195122) acc5: 100.000000 (97.865854) time: 0.286708 data: 0.000135 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.420385 (0.545698) acc1: 87.500000 (86.887255) acc5: 93.750000 (97.426471) time: 0.286423 data: 0.000133 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.557569 (0.572914) acc1: 87.500000 (86.372951) acc5: 93.750000 (97.438525) time: 0.287040 data: 0.000129 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.645292 (0.590438) acc1: 81.250000 (85.915493) acc5: 100.000000 (97.535211) time: 0.294421 data: 0.000128 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.579210 (0.590355) acc1: 87.500000 (86.111111) acc5: 100.000000 (97.685185) time: 0.294974 data: 0.000125 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.579210 (0.615204) acc1: 87.500000 (85.714286) acc5: 100.000000 (97.458791) time: 0.287653 data: 0.000131 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.560677 (0.602954) acc1: 81.250000 (85.829208) acc5: 100.000000 (97.648515) time: 0.286837 data: 0.000133 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.528101 (0.597118) acc1: 87.500000 (85.641892) acc5: 100.000000 (97.747748) time: 0.286355 data: 0.000128 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.541949 (0.595631) acc1: 87.500000 (85.640496) acc5: 100.000000 (97.727273) time: 0.286285 data: 0.000136 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.592619 (0.610624) acc1: 81.250000 (85.257634) acc5: 100.000000 (97.662214) time: 0.286517 data: 0.000135 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.627697 (0.607289) acc1: 81.250000 (85.460993) acc5: 100.000000 (97.606383) time: 0.287194 data: 0.000134 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.540980 (0.613594) acc1: 87.500000 (85.347682) acc5: 100.000000 (97.599338) time: 0.287039 data: 0.000138 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.500509 (0.617125) acc1: 81.250000 (85.364907) acc5: 100.000000 (97.631988) time: 0.286385 data: 0.000133 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.493014 (0.614403) acc1: 81.250000 (85.453216) acc5: 100.000000 (97.697368) time: 0.286700 data: 0.000145 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.466595 (0.615481) acc1: 81.250000 (85.324586) acc5: 100.000000 (97.686464) time: 0.286134 data: 0.000162 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.415297 (0.604218) acc1: 87.500000 (85.569372) acc5: 100.000000 (97.742147) time: 0.283220 data: 0.000121 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.415297 (0.612635) acc1: 87.500000 (85.408000) acc5: 100.000000 (97.664000) time: 0.273663 data: 0.000102 max mem: 14338 Test: Total time: 0:00:57 (0.294189 s / it) * Acc@1 85.138 Acc@5 97.332 loss 0.623 Max accuracy: 85.14% uploading checkpoint virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0010.pth to hdfs://harunava/user/guoyuanfan/HCSC/virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0010.pth Epoch: [11/30] [ 0/5004] eta: 2:37:43 lr: 0.000056 loss: 1.809589 (1.809589) time: 1.891270 data: 1.160622 max mem: 14338 Epoch: [11/30] [ 50/5004] eta: 1:00:47 lr: 0.000056 loss: 1.708443 (1.637913) time: 0.713605 data: 0.000208 max mem: 14338 Epoch: [11/30] [ 100/5004] eta: 0:59:14 lr: 0.000056 loss: 1.555079 (1.643158) time: 0.712364 data: 0.000229 max mem: 14338 Epoch: [11/30] [ 150/5004] eta: 0:58:23 lr: 0.000056 loss: 1.651373 (1.676195) time: 0.714876 data: 0.000222 max mem: 14338 Epoch: [11/30] [ 200/5004] eta: 0:57:36 lr: 0.000056 loss: 1.834460 (1.679037) time: 0.714088 data: 0.000225 max mem: 14338 Epoch: [11/30] [ 250/5004] eta: 0:56:57 lr: 0.000056 loss: 1.634205 (1.675514) time: 0.719221 data: 0.000183 max mem: 14338 Epoch: [11/30] [ 300/5004] eta: 0:56:18 lr: 0.000056 loss: 1.714322 (1.666323) time: 0.716171 data: 0.000169 max mem: 14338 Epoch: [11/30] [ 350/5004] eta: 0:55:38 lr: 0.000056 loss: 1.646795 (1.673425) time: 0.713777 data: 0.000177 max mem: 14338 Epoch: [11/30] [ 400/5004] eta: 0:55:02 lr: 0.000056 loss: 1.605464 (1.668684) time: 0.720435 data: 0.000211 max mem: 14338 Epoch: [11/30] [ 450/5004] eta: 0:54:24 lr: 0.000056 loss: 1.735010 (1.671848) time: 0.712760 data: 0.000218 max mem: 14338 Epoch: [11/30] [ 500/5004] eta: 0:53:47 lr: 0.000056 loss: 1.692688 (1.679941) time: 0.712325 data: 0.000229 max mem: 14338 Epoch: [11/30] [ 550/5004] eta: 0:53:09 lr: 0.000056 loss: 1.594578 (1.683551) time: 0.711282 data: 0.000239 max mem: 14338 Epoch: [11/30] [ 600/5004] eta: 0:52:33 lr: 0.000056 loss: 1.582191 (1.677964) time: 0.712325 data: 0.000216 max mem: 14338 Epoch: [11/30] [ 650/5004] eta: 0:51:56 lr: 0.000056 loss: 1.777295 (1.676389) time: 0.714857 data: 0.000159 max mem: 14338 Epoch: [11/30] [ 700/5004] eta: 0:51:20 lr: 0.000056 loss: 1.571329 (1.670532) time: 0.709548 data: 0.000161 max mem: 14338 Epoch: [11/30] [ 750/5004] eta: 0:50:44 lr: 0.000056 loss: 1.522972 (1.668523) time: 0.714993 data: 0.000230 max mem: 14338 Epoch: [11/30] [ 800/5004] eta: 0:50:08 lr: 0.000056 loss: 1.721521 (1.669711) time: 0.719376 data: 0.000211 max mem: 14338 Epoch: [11/30] [ 850/5004] eta: 0:49:32 lr: 0.000056 loss: 1.555962 (1.667894) time: 0.717293 data: 0.000210 max mem: 14338 Epoch: [11/30] [ 900/5004] eta: 0:48:57 lr: 0.000056 loss: 1.524048 (1.666397) time: 0.716839 data: 0.000210 max mem: 14338 Epoch: [11/30] [ 950/5004] eta: 0:48:21 lr: 0.000056 loss: 1.681489 (1.665609) time: 0.714373 data: 0.000228 max mem: 14338 Epoch: [11/30] [1000/5004] eta: 0:47:46 lr: 0.000056 loss: 1.716887 (1.665726) time: 0.717026 data: 0.000162 max mem: 14338 Epoch: [11/30] [1050/5004] eta: 0:47:10 lr: 0.000056 loss: 1.693931 (1.664520) time: 0.711405 data: 0.000220 max mem: 14338 Epoch: [11/30] [1100/5004] eta: 0:46:34 lr: 0.000056 loss: 1.678042 (1.665278) time: 0.711394 data: 0.000218 max mem: 14338 Epoch: [11/30] [1150/5004] eta: 0:45:58 lr: 0.000056 loss: 1.622406 (1.666801) time: 0.711619 data: 0.000236 max mem: 14338 Epoch: [11/30] [1200/5004] eta: 0:45:22 lr: 0.000056 loss: 1.667448 (1.668664) time: 0.716122 data: 0.000224 max mem: 14338 Epoch: [11/30] [1250/5004] eta: 0:44:46 lr: 0.000056 loss: 1.694138 (1.669728) time: 0.717174 data: 0.000231 max mem: 14338 Epoch: [11/30] [1300/5004] eta: 0:44:10 lr: 0.000055 loss: 1.507595 (1.666998) time: 0.715838 data: 0.000161 max mem: 14338 Epoch: [11/30] [1350/5004] eta: 0:43:34 lr: 0.000055 loss: 1.653573 (1.666297) time: 0.714561 data: 0.000171 max mem: 14338 Epoch: [11/30] [1400/5004] eta: 0:42:58 lr: 0.000055 loss: 1.583204 (1.666935) time: 0.711377 data: 0.000213 max mem: 14338 Epoch: [11/30] [1450/5004] eta: 0:42:22 lr: 0.000055 loss: 1.523488 (1.666002) time: 0.712544 data: 0.000228 max mem: 14338 Epoch: [11/30] [1500/5004] eta: 0:41:46 lr: 0.000055 loss: 1.654644 (1.665183) time: 0.712555 data: 0.000221 max mem: 14338 Epoch: [11/30] [1550/5004] eta: 0:41:10 lr: 0.000055 loss: 1.561095 (1.664774) time: 0.709706 data: 0.000201 max mem: 14338 Epoch: [11/30] [1600/5004] eta: 0:40:34 lr: 0.000055 loss: 1.671732 (1.665614) time: 0.711438 data: 0.000217 max mem: 14338 Epoch: [11/30] [1650/5004] eta: 0:39:58 lr: 0.000055 loss: 1.662313 (1.666891) time: 0.714079 data: 0.000169 max mem: 14338 Epoch: [11/30] [1700/5004] eta: 0:39:22 lr: 0.000055 loss: 1.610441 (1.666301) time: 0.716491 data: 0.000176 max mem: 14338 Epoch: [11/30] [1750/5004] eta: 0:38:46 lr: 0.000055 loss: 1.649326 (1.665953) time: 0.714660 data: 0.000224 max mem: 14338 Epoch: [11/30] [1800/5004] eta: 0:38:11 lr: 0.000055 loss: 1.711128 (1.669565) time: 0.717991 data: 0.000223 max mem: 14338 Epoch: [11/30] [1850/5004] eta: 0:37:35 lr: 0.000055 loss: 1.654791 (1.671208) time: 0.713119 data: 0.000214 max mem: 14338 Epoch: [11/30] [1900/5004] eta: 0:36:59 lr: 0.000055 loss: 1.596917 (1.671187) time: 0.710979 data: 0.000233 max mem: 14338 Epoch: [11/30] [1950/5004] eta: 0:36:23 lr: 0.000055 loss: 1.529909 (1.671313) time: 0.709744 data: 0.000217 max mem: 14338 Epoch: [11/30] [2000/5004] eta: 0:35:47 lr: 0.000055 loss: 1.628377 (1.672639) time: 0.713544 data: 0.000166 max mem: 14338 Epoch: [11/30] [2050/5004] eta: 0:35:11 lr: 0.000055 loss: 1.564036 (1.670684) time: 0.715309 data: 0.000178 max mem: 14338 Epoch: [11/30] [2100/5004] eta: 0:34:36 lr: 0.000055 loss: 1.621316 (1.671456) time: 0.710395 data: 0.000238 max mem: 14338 Epoch: [11/30] [2150/5004] eta: 0:34:00 lr: 0.000055 loss: 1.622010 (1.671831) time: 0.709346 data: 0.000227 max mem: 14338 Epoch: [11/30] [2200/5004] eta: 0:33:24 lr: 0.000055 loss: 1.562140 (1.670773) time: 0.725710 data: 0.000193 max mem: 14338 Epoch: [11/30] [2250/5004] eta: 0:32:49 lr: 0.000055 loss: 1.673877 (1.671926) time: 0.723024 data: 0.000234 max mem: 14338 Epoch: [11/30] [2300/5004] eta: 0:32:13 lr: 0.000055 loss: 1.505008 (1.671753) time: 0.710949 data: 0.000222 max mem: 14338 Epoch: [11/30] [2350/5004] eta: 0:31:37 lr: 0.000055 loss: 1.549617 (1.670567) time: 0.713985 data: 0.000178 max mem: 14338 Epoch: [11/30] [2400/5004] eta: 0:31:01 lr: 0.000055 loss: 1.477696 (1.669908) time: 0.711325 data: 0.000210 max mem: 14338 Epoch: [11/30] [2450/5004] eta: 0:30:25 lr: 0.000055 loss: 1.663950 (1.668867) time: 0.712607 data: 0.000191 max mem: 14338 Epoch: [11/30] [2500/5004] eta: 0:29:50 lr: 0.000055 loss: 1.916567 (1.670965) time: 0.710110 data: 0.000214 max mem: 14338 Epoch: [11/30] [2550/5004] eta: 0:29:14 lr: 0.000055 loss: 1.750467 (1.672219) time: 0.709476 data: 0.000212 max mem: 14338 Epoch: [11/30] [2600/5004] eta: 0:28:38 lr: 0.000054 loss: 1.559229 (1.671772) time: 0.720685 data: 0.000208 max mem: 14338 Epoch: [11/30] [2650/5004] eta: 0:28:02 lr: 0.000054 loss: 1.627760 (1.670981) time: 0.719093 data: 0.000163 max mem: 14338 Epoch: [11/30] [2700/5004] eta: 0:27:27 lr: 0.000054 loss: 1.621184 (1.671226) time: 0.718363 data: 0.000163 max mem: 14338 Epoch: [11/30] [2750/5004] eta: 0:26:51 lr: 0.000054 loss: 1.662152 (1.671595) time: 0.719646 data: 0.000220 max mem: 14338 Epoch: [11/30] [2800/5004] eta: 0:26:15 lr: 0.000054 loss: 1.603782 (1.672297) time: 0.712200 data: 0.000221 max mem: 14338 Epoch: [11/30] [2850/5004] eta: 0:25:39 lr: 0.000054 loss: 1.826690 (1.673051) time: 0.714010 data: 0.000198 max mem: 14338 Epoch: [11/30] [2900/5004] eta: 0:25:04 lr: 0.000054 loss: 1.581471 (1.672013) time: 0.710709 data: 0.000211 max mem: 14338 Epoch: [11/30] [2950/5004] eta: 0:24:28 lr: 0.000054 loss: 1.634861 (1.670848) time: 0.711497 data: 0.000211 max mem: 14338 Epoch: [11/30] [3000/5004] eta: 0:23:52 lr: 0.000054 loss: 1.673688 (1.671414) time: 0.715953 data: 0.000167 max mem: 14338 Epoch: [11/30] [3050/5004] eta: 0:23:16 lr: 0.000054 loss: 1.668448 (1.672246) time: 0.716699 data: 0.000180 max mem: 14338 Epoch: [11/30] [3100/5004] eta: 0:22:41 lr: 0.000054 loss: 1.645064 (1.672911) time: 0.717885 data: 0.000231 max mem: 14338 Epoch: [11/30] [3150/5004] eta: 0:22:05 lr: 0.000054 loss: 1.810072 (1.673679) time: 0.717549 data: 0.000223 max mem: 14338 Epoch: [11/30] [3200/5004] eta: 0:21:29 lr: 0.000054 loss: 1.738257 (1.674909) time: 0.716448 data: 0.000208 max mem: 14338 Epoch: [11/30] [3250/5004] eta: 0:20:54 lr: 0.000054 loss: 1.603270 (1.673693) time: 0.719734 data: 0.000224 max mem: 14338 Epoch: [11/30] [3300/5004] eta: 0:20:18 lr: 0.000054 loss: 1.474030 (1.673059) time: 0.709577 data: 0.000226 max mem: 14338 Epoch: [11/30] [3350/5004] eta: 0:19:42 lr: 0.000054 loss: 1.576179 (1.672129) time: 0.714339 data: 0.000191 max mem: 14338 Epoch: [11/30] [3400/5004] eta: 0:19:06 lr: 0.000054 loss: 1.636984 (1.673123) time: 0.717463 data: 0.000171 max mem: 14338 Epoch: [11/30] [3450/5004] eta: 0:18:30 lr: 0.000054 loss: 1.567619 (1.673474) time: 0.714891 data: 0.000205 max mem: 14338 Epoch: [11/30] [3500/5004] eta: 0:17:55 lr: 0.000054 loss: 1.813225 (1.675030) time: 0.709876 data: 0.000196 max mem: 14338 Epoch: [11/30] [3550/5004] eta: 0:17:19 lr: 0.000054 loss: 1.654676 (1.675077) time: 0.715125 data: 0.000220 max mem: 14338 Epoch: [11/30] [3600/5004] eta: 0:16:43 lr: 0.000054 loss: 1.448046 (1.674452) time: 0.716613 data: 0.000215 max mem: 14338 Epoch: [11/30] [3650/5004] eta: 0:16:07 lr: 0.000054 loss: 1.615872 (1.674957) time: 0.718979 data: 0.000230 max mem: 14338 Epoch: [11/30] [3700/5004] eta: 0:15:32 lr: 0.000054 loss: 1.726577 (1.675221) time: 0.717675 data: 0.000188 max mem: 14338 Epoch: [11/30] [3750/5004] eta: 0:14:56 lr: 0.000054 loss: 1.566291 (1.675828) time: 0.710179 data: 0.000229 max mem: 14338 Epoch: [11/30] [3800/5004] eta: 0:14:20 lr: 0.000054 loss: 1.731675 (1.676284) time: 0.714674 data: 0.000213 max mem: 14338 Epoch: [11/30] [3850/5004] eta: 0:13:44 lr: 0.000054 loss: 1.578676 (1.676262) time: 0.717810 data: 0.000228 max mem: 14338 Epoch: [11/30] [3900/5004] eta: 0:13:09 lr: 0.000053 loss: 1.716034 (1.676471) time: 0.708835 data: 0.000220 max mem: 14338 Epoch: [11/30] [3950/5004] eta: 0:12:33 lr: 0.000053 loss: 1.574342 (1.676178) time: 0.712521 data: 0.000216 max mem: 14338 Epoch: [11/30] [4000/5004] eta: 0:11:57 lr: 0.000053 loss: 1.662908 (1.675954) time: 0.715730 data: 0.000149 max mem: 14338 Epoch: [11/30] [4050/5004] eta: 0:11:21 lr: 0.000053 loss: 1.461929 (1.675303) time: 0.722797 data: 0.000168 max mem: 14338 Epoch: [11/30] [4100/5004] eta: 0:10:46 lr: 0.000053 loss: 1.577102 (1.675609) time: 0.715170 data: 0.000224 max mem: 14338 Epoch: [11/30] [4150/5004] eta: 0:10:10 lr: 0.000053 loss: 1.658545 (1.676319) time: 0.720453 data: 0.000204 max mem: 14338 Epoch: [11/30] [4200/5004] eta: 0:09:34 lr: 0.000053 loss: 1.514004 (1.675870) time: 0.715095 data: 0.000214 max mem: 14338 Epoch: [11/30] [4250/5004] eta: 0:08:59 lr: 0.000053 loss: 2.004959 (1.677331) time: 0.717022 data: 0.000206 max mem: 14338 Epoch: [11/30] [4300/5004] eta: 0:08:23 lr: 0.000053 loss: 1.591868 (1.677206) time: 0.709802 data: 0.000219 max mem: 14338 Epoch: [11/30] [4350/5004] eta: 0:07:47 lr: 0.000053 loss: 1.738526 (1.678080) time: 0.711554 data: 0.000163 max mem: 14338 Epoch: [11/30] [4400/5004] eta: 0:07:11 lr: 0.000053 loss: 1.679929 (1.677674) time: 0.715945 data: 0.000164 max mem: 14338 Epoch: [11/30] [4450/5004] eta: 0:06:36 lr: 0.000053 loss: 1.551667 (1.678165) time: 0.715360 data: 0.000228 max mem: 14338 Epoch: [11/30] [4500/5004] eta: 0:06:00 lr: 0.000053 loss: 1.748601 (1.678448) time: 0.712795 data: 0.000235 max mem: 14338 Epoch: [11/30] [4550/5004] eta: 0:05:24 lr: 0.000053 loss: 1.658238 (1.679136) time: 0.718140 data: 0.000226 max mem: 14338 Epoch: [11/30] [4600/5004] eta: 0:04:48 lr: 0.000053 loss: 1.568470 (1.678719) time: 0.714977 data: 0.000228 max mem: 14338 Epoch: [11/30] [4650/5004] eta: 0:04:13 lr: 0.000053 loss: 1.761310 (1.679470) time: 0.710792 data: 0.000224 max mem: 14338 Epoch: [11/30] [4700/5004] eta: 0:03:37 lr: 0.000053 loss: 1.634207 (1.679302) time: 0.712265 data: 0.000182 max mem: 14338 Epoch: [11/30] [4750/5004] eta: 0:03:01 lr: 0.000053 loss: 1.440160 (1.678650) time: 0.709426 data: 0.000180 max mem: 14338 Epoch: [11/30] [4800/5004] eta: 0:02:25 lr: 0.000053 loss: 1.664376 (1.678835) time: 0.716194 data: 0.000181 max mem: 14338 Epoch: [11/30] [4850/5004] eta: 0:01:50 lr: 0.000053 loss: 1.578311 (1.678725) time: 0.716743 data: 0.000220 max mem: 14338 Epoch: [11/30] [4900/5004] eta: 0:01:14 lr: 0.000053 loss: 1.434922 (1.678063) time: 0.709165 data: 0.000232 max mem: 14338 Epoch: [11/30] [4950/5004] eta: 0:00:38 lr: 0.000053 loss: 1.596825 (1.677997) time: 0.715101 data: 0.000231 max mem: 14338 Epoch: [11/30] [5000/5004] eta: 0:00:02 lr: 0.000053 loss: 1.573340 (1.677386) time: 0.710593 data: 0.000831 max mem: 14338 Epoch: [11/30] [5003/5004] eta: 0:00:00 lr: 0.000053 loss: 1.611359 (1.677315) time: 0.707373 data: 0.000824 max mem: 14338 Epoch: [11/30] Total time: 0:59:37 (0.715009 s / it) Averaged stats: lr: 0.000053 loss: 1.611359 (1.678567) Test: [ 0/196] eta: 0:05:08 loss: 0.390607 (0.390607) acc1: 87.500000 (87.500000) acc5: 100.000000 (100.000000) time: 1.575689 data: 1.198929 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 0.480951 (0.552072) acc1: 87.500000 (85.795455) acc5: 100.000000 (98.295455) time: 0.404881 data: 0.109122 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.568273 (0.542945) acc1: 87.500000 (86.309524) acc5: 100.000000 (97.916667) time: 0.288154 data: 0.000136 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.509075 (0.520078) acc1: 87.500000 (87.096774) acc5: 100.000000 (97.983871) time: 0.289141 data: 0.000135 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.421188 (0.523446) acc1: 87.500000 (87.042683) acc5: 100.000000 (97.865854) time: 0.288968 data: 0.000139 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.447361 (0.549049) acc1: 87.500000 (86.887255) acc5: 100.000000 (97.426471) time: 0.287503 data: 0.000133 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.546507 (0.572989) acc1: 87.500000 (86.372951) acc5: 93.750000 (97.438525) time: 0.286860 data: 0.000147 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.665039 (0.589362) acc1: 81.250000 (85.915493) acc5: 100.000000 (97.623239) time: 0.287337 data: 0.000145 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.541271 (0.589043) acc1: 87.500000 (85.879630) acc5: 100.000000 (97.762346) time: 0.288592 data: 0.000122 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.525113 (0.611833) acc1: 81.250000 (85.439560) acc5: 100.000000 (97.527473) time: 0.294578 data: 0.000137 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.598154 (0.600326) acc1: 81.250000 (85.643564) acc5: 100.000000 (97.710396) time: 0.293278 data: 0.000139 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.521089 (0.591930) acc1: 87.500000 (85.585586) acc5: 100.000000 (97.860360) time: 0.286831 data: 0.000128 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.495961 (0.589038) acc1: 87.500000 (85.640496) acc5: 100.000000 (97.778926) time: 0.286467 data: 0.000134 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.550065 (0.603625) acc1: 81.250000 (85.305344) acc5: 93.750000 (97.709924) time: 0.286686 data: 0.000141 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.628687 (0.601118) acc1: 81.250000 (85.460993) acc5: 100.000000 (97.695035) time: 0.286550 data: 0.000143 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.580034 (0.608616) acc1: 87.500000 (85.347682) acc5: 100.000000 (97.640728) time: 0.286261 data: 0.000139 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.503694 (0.612159) acc1: 81.250000 (85.248447) acc5: 100.000000 (97.631988) time: 0.286741 data: 0.000131 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.472095 (0.608996) acc1: 81.250000 (85.343567) acc5: 100.000000 (97.697368) time: 0.286755 data: 0.000144 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.462818 (0.610187) acc1: 81.250000 (85.082873) acc5: 100.000000 (97.686464) time: 0.286591 data: 0.000142 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.386427 (0.598370) acc1: 87.500000 (85.242147) acc5: 100.000000 (97.709424) time: 0.283957 data: 0.000098 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.475946 (0.609123) acc1: 81.250000 (85.088000) acc5: 100.000000 (97.632000) time: 0.274166 data: 0.000089 max mem: 14338 Test: Total time: 0:00:57 (0.294662 s / it) * Acc@1 85.002 Acc@5 97.258 loss 0.628 Max accuracy: 85.14% Epoch: [12/30] [ 0/5004] eta: 2:35:28 lr: 0.000053 loss: 1.699534 (1.699534) time: 1.864158 data: 1.126462 max mem: 14338 Epoch: [12/30] [ 50/5004] eta: 1:00:58 lr: 0.000053 loss: 1.608850 (1.614245) time: 0.718881 data: 0.000193 max mem: 14338 Epoch: [12/30] [ 100/5004] eta: 0:59:18 lr: 0.000053 loss: 1.751242 (1.651518) time: 0.713299 data: 0.000213 max mem: 14338 Epoch: [12/30] [ 150/5004] eta: 0:58:20 lr: 0.000052 loss: 1.592486 (1.637336) time: 0.713340 data: 0.000219 max mem: 14338 Epoch: [12/30] [ 200/5004] eta: 0:57:35 lr: 0.000052 loss: 1.617622 (1.631071) time: 0.712343 data: 0.000212 max mem: 14338 Epoch: [12/30] [ 250/5004] eta: 0:56:54 lr: 0.000052 loss: 1.782470 (1.650377) time: 0.716631 data: 0.000176 max mem: 14338 Epoch: [12/30] [ 300/5004] eta: 0:56:15 lr: 0.000052 loss: 1.554929 (1.642958) time: 0.713103 data: 0.000153 max mem: 14338 Epoch: [12/30] [ 350/5004] eta: 0:55:37 lr: 0.000052 loss: 1.561173 (1.642548) time: 0.711728 data: 0.000174 max mem: 14338 Epoch: [12/30] [ 400/5004] eta: 0:55:00 lr: 0.000052 loss: 1.574606 (1.652815) time: 0.715374 data: 0.000205 max mem: 14338 Epoch: [12/30] [ 450/5004] eta: 0:54:24 lr: 0.000052 loss: 1.583624 (1.655351) time: 0.721538 data: 0.000220 max mem: 14338 Epoch: [12/30] [ 500/5004] eta: 0:53:46 lr: 0.000052 loss: 1.513202 (1.657080) time: 0.715166 data: 0.000225 max mem: 14338 Epoch: [12/30] [ 550/5004] eta: 0:53:10 lr: 0.000052 loss: 1.511733 (1.655482) time: 0.715764 data: 0.000209 max mem: 14338 Epoch: [12/30] [ 600/5004] eta: 0:52:33 lr: 0.000052 loss: 1.766089 (1.666665) time: 0.714908 data: 0.000214 max mem: 14338 Epoch: [12/30] [ 650/5004] eta: 0:51:55 lr: 0.000052 loss: 1.571087 (1.668025) time: 0.712150 data: 0.000148 max mem: 14338 Epoch: [12/30] [ 700/5004] eta: 0:51:19 lr: 0.000052 loss: 1.539442 (1.663805) time: 0.709777 data: 0.000155 max mem: 14338 Epoch: [12/30] [ 750/5004] eta: 0:50:43 lr: 0.000052 loss: 1.680150 (1.664048) time: 0.720453 data: 0.000210 max mem: 14338 Epoch: [12/30] [ 800/5004] eta: 0:50:08 lr: 0.000052 loss: 1.639044 (1.664501) time: 0.717368 data: 0.000205 max mem: 14338 Epoch: [12/30] [ 850/5004] eta: 0:49:32 lr: 0.000052 loss: 1.677255 (1.666783) time: 0.715040 data: 0.000214 max mem: 14338 Epoch: [12/30] [ 900/5004] eta: 0:48:56 lr: 0.000052 loss: 1.514800 (1.666664) time: 0.712308 data: 0.000183 max mem: 14338 Epoch: [12/30] [ 950/5004] eta: 0:48:19 lr: 0.000052 loss: 1.717372 (1.666455) time: 0.712625 data: 0.000205 max mem: 14338 Epoch: [12/30] [1000/5004] eta: 0:47:44 lr: 0.000052 loss: 1.480514 (1.662854) time: 0.722445 data: 0.000174 max mem: 14338 Epoch: [12/30] [1050/5004] eta: 0:47:08 lr: 0.000052 loss: 1.673196 (1.662192) time: 0.719532 data: 0.000211 max mem: 14338 Epoch: [12/30] [1100/5004] eta: 0:46:32 lr: 0.000052 loss: 1.441019 (1.660707) time: 0.711890 data: 0.000244 max mem: 14338 Epoch: [12/30] [1150/5004] eta: 0:45:56 lr: 0.000052 loss: 1.558412 (1.662850) time: 0.711089 data: 0.000235 max mem: 14338 Epoch: [12/30] [1200/5004] eta: 0:45:21 lr: 0.000052 loss: 1.664092 (1.659901) time: 0.716176 data: 0.000220 max mem: 14338 Epoch: [12/30] [1250/5004] eta: 0:44:45 lr: 0.000052 loss: 1.530946 (1.658670) time: 0.712510 data: 0.000227 max mem: 14338 Epoch: [12/30] [1300/5004] eta: 0:44:09 lr: 0.000052 loss: 1.525607 (1.659585) time: 0.713928 data: 0.000173 max mem: 14338 Epoch: [12/30] [1350/5004] eta: 0:43:33 lr: 0.000052 loss: 1.663228 (1.660469) time: 0.712568 data: 0.000188 max mem: 14338 Epoch: [12/30] [1400/5004] eta: 0:42:57 lr: 0.000051 loss: 1.532089 (1.660392) time: 0.714880 data: 0.000205 max mem: 14338 Epoch: [12/30] [1450/5004] eta: 0:42:21 lr: 0.000051 loss: 1.595117 (1.660203) time: 0.715240 data: 0.000210 max mem: 14338 Epoch: [12/30] [1500/5004] eta: 0:41:45 lr: 0.000051 loss: 1.581711 (1.658941) time: 0.714784 data: 0.000206 max mem: 14338 Epoch: [12/30] [1550/5004] eta: 0:41:09 lr: 0.000051 loss: 1.560092 (1.656659) time: 0.709175 data: 0.000210 max mem: 14338 Epoch: [12/30] [1600/5004] eta: 0:40:33 lr: 0.000051 loss: 1.670167 (1.657934) time: 0.713368 data: 0.000216 max mem: 14338 Epoch: [12/30] [1650/5004] eta: 0:39:57 lr: 0.000051 loss: 1.622700 (1.659986) time: 0.711545 data: 0.000170 max mem: 14338 Epoch: [12/30] [1700/5004] eta: 0:39:21 lr: 0.000051 loss: 1.601518 (1.659639) time: 0.709373 data: 0.000191 max mem: 14338 Epoch: [12/30] [1750/5004] eta: 0:38:45 lr: 0.000051 loss: 1.711173 (1.661633) time: 0.711713 data: 0.000216 max mem: 14338 Epoch: [12/30] [1800/5004] eta: 0:38:10 lr: 0.000051 loss: 1.625395 (1.661096) time: 0.714669 data: 0.000224 max mem: 14338 Epoch: [12/30] [1850/5004] eta: 0:37:34 lr: 0.000051 loss: 1.580896 (1.660336) time: 0.720043 data: 0.000224 max mem: 14338 Epoch: [12/30] [1900/5004] eta: 0:36:58 lr: 0.000051 loss: 1.510883 (1.659525) time: 0.722565 data: 0.000228 max mem: 14338 Epoch: [12/30] [1950/5004] eta: 0:36:23 lr: 0.000051 loss: 1.744672 (1.659623) time: 0.727747 data: 0.000219 max mem: 14338 Epoch: [12/30] [2000/5004] eta: 0:35:47 lr: 0.000051 loss: 1.712238 (1.660700) time: 0.713772 data: 0.000156 max mem: 14338 Epoch: [12/30] [2050/5004] eta: 0:35:11 lr: 0.000051 loss: 1.726886 (1.663647) time: 0.712716 data: 0.000176 max mem: 14338 Epoch: [12/30] [2100/5004] eta: 0:34:36 lr: 0.000051 loss: 1.519860 (1.661204) time: 0.710669 data: 0.000227 max mem: 14338 Epoch: [12/30] [2150/5004] eta: 0:34:00 lr: 0.000051 loss: 1.702485 (1.661732) time: 0.712675 data: 0.000222 max mem: 14338 Epoch: [12/30] [2200/5004] eta: 0:33:24 lr: 0.000051 loss: 1.723544 (1.662951) time: 0.713576 data: 0.000187 max mem: 14338 Epoch: [12/30] [2250/5004] eta: 0:32:48 lr: 0.000051 loss: 1.619360 (1.663285) time: 0.713267 data: 0.000217 max mem: 14338 Epoch: [12/30] [2300/5004] eta: 0:32:12 lr: 0.000051 loss: 1.636936 (1.664423) time: 0.710419 data: 0.000211 max mem: 14338 Epoch: [12/30] [2350/5004] eta: 0:31:36 lr: 0.000051 loss: 1.478368 (1.663501) time: 0.711224 data: 0.000174 max mem: 14338 Epoch: [12/30] [2400/5004] eta: 0:31:01 lr: 0.000051 loss: 1.568962 (1.663577) time: 0.718733 data: 0.000223 max mem: 14338 Epoch: [12/30] [2450/5004] eta: 0:30:25 lr: 0.000051 loss: 1.668955 (1.664641) time: 0.715104 data: 0.000199 max mem: 14338 Epoch: [12/30] [2500/5004] eta: 0:29:49 lr: 0.000051 loss: 1.664174 (1.665465) time: 0.709973 data: 0.000215 max mem: 14338 Epoch: [12/30] [2550/5004] eta: 0:29:13 lr: 0.000051 loss: 1.532741 (1.665875) time: 0.711654 data: 0.000215 max mem: 14338 Epoch: [12/30] [2600/5004] eta: 0:28:37 lr: 0.000051 loss: 1.567680 (1.665940) time: 0.713364 data: 0.000218 max mem: 14338 Epoch: [12/30] [2650/5004] eta: 0:28:02 lr: 0.000050 loss: 1.586448 (1.665813) time: 0.713539 data: 0.000167 max mem: 14338 Epoch: [12/30] [2700/5004] eta: 0:27:26 lr: 0.000050 loss: 1.530449 (1.665487) time: 0.709577 data: 0.000175 max mem: 14338 Epoch: [12/30] [2750/5004] eta: 0:26:50 lr: 0.000050 loss: 1.588613 (1.666423) time: 0.718448 data: 0.000223 max mem: 14338 Epoch: [12/30] [2800/5004] eta: 0:26:15 lr: 0.000050 loss: 1.628772 (1.666464) time: 0.713446 data: 0.000223 max mem: 14338 Epoch: [12/30] [2850/5004] eta: 0:25:39 lr: 0.000050 loss: 1.476548 (1.665887) time: 0.715539 data: 0.000192 max mem: 14338 Epoch: [12/30] [2900/5004] eta: 0:25:03 lr: 0.000050 loss: 1.556221 (1.667319) time: 0.717079 data: 0.000232 max mem: 14338 Epoch: [12/30] [2950/5004] eta: 0:24:27 lr: 0.000050 loss: 1.665028 (1.668296) time: 0.710483 data: 0.000210 max mem: 14338 Epoch: [12/30] [3000/5004] eta: 0:23:52 lr: 0.000050 loss: 1.495319 (1.668131) time: 0.714013 data: 0.000160 max mem: 14338 Epoch: [12/30] [3050/5004] eta: 0:23:16 lr: 0.000050 loss: 1.619451 (1.667289) time: 0.714914 data: 0.000178 max mem: 14338 Epoch: [12/30] [3100/5004] eta: 0:22:40 lr: 0.000050 loss: 1.490851 (1.665568) time: 0.710614 data: 0.000223 max mem: 14338 Epoch: [12/30] [3150/5004] eta: 0:22:04 lr: 0.000050 loss: 1.659494 (1.665360) time: 0.709731 data: 0.000224 max mem: 14338 Epoch: [12/30] [3200/5004] eta: 0:21:29 lr: 0.000050 loss: 1.570936 (1.665264) time: 0.713358 data: 0.000225 max mem: 14338 Epoch: [12/30] [3250/5004] eta: 0:20:53 lr: 0.000050 loss: 1.580884 (1.665178) time: 0.721770 data: 0.000218 max mem: 14338 Epoch: [12/30] [3300/5004] eta: 0:20:17 lr: 0.000050 loss: 1.654381 (1.665501) time: 0.713423 data: 0.000208 max mem: 14338 Epoch: [12/30] [3350/5004] eta: 0:19:41 lr: 0.000050 loss: 1.408699 (1.663726) time: 0.711315 data: 0.000173 max mem: 14338 Epoch: [12/30] [3400/5004] eta: 0:19:06 lr: 0.000050 loss: 1.589018 (1.664190) time: 0.719295 data: 0.000158 max mem: 14338 Epoch: [12/30] [3450/5004] eta: 0:18:30 lr: 0.000050 loss: 1.542534 (1.663727) time: 0.712434 data: 0.000212 max mem: 14338 Epoch: [12/30] [3500/5004] eta: 0:17:54 lr: 0.000050 loss: 1.501680 (1.663440) time: 0.709463 data: 0.000190 max mem: 14338 Epoch: [12/30] [3550/5004] eta: 0:17:18 lr: 0.000050 loss: 1.613618 (1.662640) time: 0.710704 data: 0.000225 max mem: 14338 Epoch: [12/30] [3600/5004] eta: 0:16:43 lr: 0.000050 loss: 1.844194 (1.664038) time: 0.715412 data: 0.000221 max mem: 14338 Epoch: [12/30] [3650/5004] eta: 0:16:07 lr: 0.000050 loss: 1.561046 (1.663540) time: 0.716009 data: 0.000223 max mem: 14338 Epoch: [12/30] [3700/5004] eta: 0:15:31 lr: 0.000050 loss: 1.582361 (1.663914) time: 0.714218 data: 0.000173 max mem: 14338 Epoch: [12/30] [3750/5004] eta: 0:14:56 lr: 0.000050 loss: 1.703756 (1.664233) time: 0.718205 data: 0.000225 max mem: 14338 Epoch: [12/30] [3800/5004] eta: 0:14:20 lr: 0.000050 loss: 1.601375 (1.664715) time: 0.725197 data: 0.000222 max mem: 14338 Epoch: [12/30] [3850/5004] eta: 0:13:44 lr: 0.000050 loss: 1.574901 (1.665946) time: 0.715451 data: 0.000225 max mem: 14338 Epoch: [12/30] [3900/5004] eta: 0:13:08 lr: 0.000049 loss: 1.734175 (1.666811) time: 0.708404 data: 0.000237 max mem: 14338 Epoch: [12/30] [3950/5004] eta: 0:12:33 lr: 0.000049 loss: 1.703314 (1.667408) time: 0.712690 data: 0.000245 max mem: 14338 Epoch: [12/30] [4000/5004] eta: 0:11:57 lr: 0.000049 loss: 1.672326 (1.666976) time: 0.711321 data: 0.000167 max mem: 14338 Epoch: [12/30] [4050/5004] eta: 0:11:21 lr: 0.000049 loss: 1.507921 (1.666206) time: 0.714288 data: 0.000155 max mem: 14338 Epoch: [12/30] [4100/5004] eta: 0:10:45 lr: 0.000049 loss: 1.683756 (1.666581) time: 0.709428 data: 0.000220 max mem: 14338 Epoch: [12/30] [4150/5004] eta: 0:10:10 lr: 0.000049 loss: 1.664541 (1.666612) time: 0.715983 data: 0.000191 max mem: 14338 Epoch: [12/30] [4200/5004] eta: 0:09:34 lr: 0.000049 loss: 1.686653 (1.666413) time: 0.723322 data: 0.000217 max mem: 14338 Epoch: [12/30] [4250/5004] eta: 0:08:58 lr: 0.000049 loss: 1.725835 (1.666928) time: 0.716575 data: 0.000238 max mem: 14338 Epoch: [12/30] [4300/5004] eta: 0:08:23 lr: 0.000049 loss: 1.539449 (1.667014) time: 0.719486 data: 0.000219 max mem: 14338 Epoch: [12/30] [4350/5004] eta: 0:07:47 lr: 0.000049 loss: 1.514185 (1.666710) time: 0.715208 data: 0.000162 max mem: 14338 Epoch: [12/30] [4400/5004] eta: 0:07:11 lr: 0.000049 loss: 1.778076 (1.667221) time: 0.711642 data: 0.000165 max mem: 14338 Epoch: [12/30] [4450/5004] eta: 0:06:35 lr: 0.000049 loss: 1.481302 (1.666735) time: 0.712353 data: 0.000221 max mem: 14338 Epoch: [12/30] [4500/5004] eta: 0:06:00 lr: 0.000049 loss: 1.595688 (1.667072) time: 0.711967 data: 0.000219 max mem: 14338 Epoch: [12/30] [4550/5004] eta: 0:05:24 lr: 0.000049 loss: 1.572690 (1.666842) time: 0.711281 data: 0.000226 max mem: 14338 Epoch: [12/30] [4600/5004] eta: 0:04:48 lr: 0.000049 loss: 1.629566 (1.667184) time: 0.717675 data: 0.000224 max mem: 14338 Epoch: [12/30] [4650/5004] eta: 0:04:12 lr: 0.000049 loss: 1.703547 (1.667953) time: 0.716383 data: 0.000193 max mem: 14338 Epoch: [12/30] [4700/5004] eta: 0:03:37 lr: 0.000049 loss: 1.546614 (1.667048) time: 0.715696 data: 0.000164 max mem: 14338 Epoch: [12/30] [4750/5004] eta: 0:03:01 lr: 0.000049 loss: 1.424719 (1.666173) time: 0.718143 data: 0.000169 max mem: 14338 Epoch: [12/30] [4800/5004] eta: 0:02:25 lr: 0.000049 loss: 1.640025 (1.666287) time: 0.715921 data: 0.000196 max mem: 14338 Epoch: [12/30] [4850/5004] eta: 0:01:50 lr: 0.000049 loss: 1.575860 (1.666050) time: 0.712471 data: 0.000214 max mem: 14338 Epoch: [12/30] [4900/5004] eta: 0:01:14 lr: 0.000049 loss: 1.524659 (1.665331) time: 0.712469 data: 0.000250 max mem: 14338 Epoch: [12/30] [4950/5004] eta: 0:00:38 lr: 0.000049 loss: 1.762270 (1.665351) time: 0.713642 data: 0.000242 max mem: 14338 Epoch: [12/30] [5000/5004] eta: 0:00:02 lr: 0.000049 loss: 1.575466 (1.665560) time: 0.713056 data: 0.000830 max mem: 14338 Epoch: [12/30] [5003/5004] eta: 0:00:00 lr: 0.000049 loss: 1.565442 (1.665486) time: 0.710030 data: 0.000818 max mem: 14338 Epoch: [12/30] Total time: 0:59:36 (0.714798 s / it) Averaged stats: lr: 0.000049 loss: 1.565442 (1.670484) Test: [ 0/196] eta: 0:05:18 loss: 0.307149 (0.307149) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.625720 data: 1.275644 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 0.457304 (0.521622) acc1: 87.500000 (85.795455) acc5: 100.000000 (98.863636) time: 0.408232 data: 0.116088 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.491423 (0.524425) acc1: 87.500000 (86.309524) acc5: 100.000000 (98.214286) time: 0.286471 data: 0.000124 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.450603 (0.501346) acc1: 87.500000 (87.903226) acc5: 100.000000 (98.387097) time: 0.286436 data: 0.000116 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.420151 (0.507728) acc1: 87.500000 (87.500000) acc5: 100.000000 (98.170732) time: 0.293290 data: 0.000135 max mem: 14338 Test: [ 50/196] eta: 0:00:46 loss: 0.423681 (0.534959) acc1: 87.500000 (87.132353) acc5: 100.000000 (97.671569) time: 0.293489 data: 0.000138 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.568687 (0.564567) acc1: 81.250000 (86.168033) acc5: 93.750000 (97.540984) time: 0.287752 data: 0.000137 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.679911 (0.582838) acc1: 81.250000 (85.739437) acc5: 100.000000 (97.711268) time: 0.287542 data: 0.000152 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.525750 (0.582855) acc1: 87.500000 (85.725309) acc5: 100.000000 (97.762346) time: 0.286698 data: 0.000137 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.525750 (0.604372) acc1: 87.500000 (85.302198) acc5: 100.000000 (97.458791) time: 0.286567 data: 0.000135 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.547973 (0.594845) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.586634) time: 0.286342 data: 0.000135 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.545447 (0.587161) acc1: 87.500000 (85.360360) acc5: 100.000000 (97.691441) time: 0.286836 data: 0.000120 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.545447 (0.585417) acc1: 87.500000 (85.330579) acc5: 100.000000 (97.675620) time: 0.286493 data: 0.000132 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.606449 (0.600386) acc1: 87.500000 (85.114504) acc5: 100.000000 (97.662214) time: 0.286196 data: 0.000136 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.636353 (0.598815) acc1: 87.500000 (85.195035) acc5: 100.000000 (97.650709) time: 0.286255 data: 0.000141 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.525587 (0.606452) acc1: 87.500000 (85.182119) acc5: 100.000000 (97.640728) time: 0.285978 data: 0.000144 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.525587 (0.613000) acc1: 81.250000 (85.054348) acc5: 100.000000 (97.631988) time: 0.285556 data: 0.000128 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.484195 (0.609633) acc1: 87.500000 (85.160819) acc5: 100.000000 (97.660819) time: 0.285896 data: 0.000145 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.433834 (0.609523) acc1: 87.500000 (85.048343) acc5: 100.000000 (97.686464) time: 0.292494 data: 0.000158 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.327815 (0.598184) acc1: 87.500000 (85.274869) acc5: 100.000000 (97.709424) time: 0.289819 data: 0.000112 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.465627 (0.610004) acc1: 87.500000 (85.152000) acc5: 100.000000 (97.600000) time: 0.280296 data: 0.000101 max mem: 14338 Test: Total time: 0:00:57 (0.294592 s / it) * Acc@1 85.036 Acc@5 97.374 loss 0.626 Max accuracy: 85.14% Epoch: [13/30] [ 0/5004] eta: 2:35:52 lr: 0.000049 loss: 1.805119 (1.805119) time: 1.869029 data: 1.120936 max mem: 14338 Epoch: [13/30] [ 50/5004] eta: 1:00:44 lr: 0.000049 loss: 1.692475 (1.700053) time: 0.714477 data: 0.000201 max mem: 14338 Epoch: [13/30] [ 100/5004] eta: 0:59:22 lr: 0.000049 loss: 1.615361 (1.704764) time: 0.719124 data: 0.000221 max mem: 14338 Epoch: [13/30] [ 150/5004] eta: 0:58:30 lr: 0.000048 loss: 1.759033 (1.728423) time: 0.719758 data: 0.000209 max mem: 14338 Epoch: [13/30] [ 200/5004] eta: 0:57:45 lr: 0.000048 loss: 1.565758 (1.699491) time: 0.718902 data: 0.000223 max mem: 14338 Epoch: [13/30] [ 250/5004] eta: 0:57:01 lr: 0.000048 loss: 1.709745 (1.704102) time: 0.716979 data: 0.000187 max mem: 14338 Epoch: [13/30] [ 300/5004] eta: 0:56:20 lr: 0.000048 loss: 1.721010 (1.697175) time: 0.711926 data: 0.000160 max mem: 14338 Epoch: [13/30] [ 350/5004] eta: 0:55:40 lr: 0.000048 loss: 1.608094 (1.688853) time: 0.710460 data: 0.000168 max mem: 14338 Epoch: [13/30] [ 400/5004] eta: 0:55:02 lr: 0.000048 loss: 1.579702 (1.688307) time: 0.717766 data: 0.000197 max mem: 14338 Epoch: [13/30] [ 450/5004] eta: 0:54:25 lr: 0.000048 loss: 1.684779 (1.684460) time: 0.718218 data: 0.000206 max mem: 14338 Epoch: [13/30] [ 500/5004] eta: 0:53:48 lr: 0.000048 loss: 1.615393 (1.684625) time: 0.711343 data: 0.000209 max mem: 14338 Epoch: [13/30] [ 550/5004] eta: 0:53:10 lr: 0.000048 loss: 1.751433 (1.682129) time: 0.710767 data: 0.000210 max mem: 14338 Epoch: [13/30] [ 600/5004] eta: 0:52:35 lr: 0.000048 loss: 1.727882 (1.683645) time: 0.722473 data: 0.000213 max mem: 14338 Epoch: [13/30] [ 650/5004] eta: 0:51:59 lr: 0.000048 loss: 1.690828 (1.686633) time: 0.721164 data: 0.000176 max mem: 14338 Epoch: [13/30] [ 700/5004] eta: 0:51:22 lr: 0.000048 loss: 1.561903 (1.682769) time: 0.713247 data: 0.000164 max mem: 14338 Epoch: [13/30] [ 750/5004] eta: 0:50:45 lr: 0.000048 loss: 1.538521 (1.680774) time: 0.710173 data: 0.000211 max mem: 14338 Epoch: [13/30] [ 800/5004] eta: 0:50:08 lr: 0.000048 loss: 1.510509 (1.680692) time: 0.713226 data: 0.000224 max mem: 14338 Epoch: [13/30] [ 850/5004] eta: 0:49:32 lr: 0.000048 loss: 1.378955 (1.677818) time: 0.716097 data: 0.000246 max mem: 14338 Epoch: [13/30] [ 900/5004] eta: 0:48:56 lr: 0.000048 loss: 1.496350 (1.674344) time: 0.709523 data: 0.000191 max mem: 14338 Epoch: [13/30] [ 950/5004] eta: 0:48:20 lr: 0.000048 loss: 1.592189 (1.670607) time: 0.709386 data: 0.000202 max mem: 14338 Epoch: [13/30] [1000/5004] eta: 0:47:44 lr: 0.000048 loss: 1.707941 (1.668964) time: 0.714490 data: 0.000167 max mem: 14338 Epoch: [13/30] [1050/5004] eta: 0:47:08 lr: 0.000048 loss: 1.436257 (1.672351) time: 0.715105 data: 0.000219 max mem: 14338 Epoch: [13/30] [1100/5004] eta: 0:46:32 lr: 0.000048 loss: 1.555271 (1.665650) time: 0.715368 data: 0.000222 max mem: 14338 Epoch: [13/30] [1150/5004] eta: 0:45:55 lr: 0.000048 loss: 1.638645 (1.664678) time: 0.711897 data: 0.000224 max mem: 14338 Epoch: [13/30] [1200/5004] eta: 0:45:20 lr: 0.000048 loss: 1.528499 (1.663592) time: 0.715578 data: 0.000229 max mem: 14338 Epoch: [13/30] [1250/5004] eta: 0:44:43 lr: 0.000048 loss: 1.650506 (1.663069) time: 0.712663 data: 0.000223 max mem: 14338 Epoch: [13/30] [1300/5004] eta: 0:44:08 lr: 0.000048 loss: 1.559153 (1.660156) time: 0.715726 data: 0.000176 max mem: 14338 Epoch: [13/30] [1350/5004] eta: 0:43:32 lr: 0.000048 loss: 1.648341 (1.663706) time: 0.715820 data: 0.000179 max mem: 14338 Epoch: [13/30] [1400/5004] eta: 0:42:57 lr: 0.000047 loss: 1.658808 (1.665364) time: 0.716288 data: 0.000228 max mem: 14338 Epoch: [13/30] [1450/5004] eta: 0:42:21 lr: 0.000047 loss: 1.647966 (1.666544) time: 0.712879 data: 0.000216 max mem: 14338 Epoch: [13/30] [1500/5004] eta: 0:41:45 lr: 0.000047 loss: 1.570342 (1.666334) time: 0.711887 data: 0.000216 max mem: 14338 Epoch: [13/30] [1550/5004] eta: 0:41:09 lr: 0.000047 loss: 1.683126 (1.665347) time: 0.714434 data: 0.000199 max mem: 14338 Epoch: [13/30] [1600/5004] eta: 0:40:33 lr: 0.000047 loss: 1.675529 (1.667793) time: 0.721039 data: 0.000216 max mem: 14338 Epoch: [13/30] [1650/5004] eta: 0:39:57 lr: 0.000047 loss: 1.588814 (1.666287) time: 0.722726 data: 0.000170 max mem: 14338 Epoch: [13/30] [1700/5004] eta: 0:39:22 lr: 0.000047 loss: 1.688778 (1.667365) time: 0.713478 data: 0.000159 max mem: 14338 Epoch: [13/30] [1750/5004] eta: 0:38:46 lr: 0.000047 loss: 1.661582 (1.668479) time: 0.717108 data: 0.000218 max mem: 14338 Epoch: [13/30] [1800/5004] eta: 0:38:11 lr: 0.000047 loss: 1.521669 (1.668497) time: 0.713082 data: 0.000229 max mem: 14338 Epoch: [13/30] [1850/5004] eta: 0:37:35 lr: 0.000047 loss: 1.716717 (1.670528) time: 0.714923 data: 0.000223 max mem: 14338 Epoch: [13/30] [1900/5004] eta: 0:36:59 lr: 0.000047 loss: 1.598695 (1.667629) time: 0.712381 data: 0.000204 max mem: 14338 Epoch: [13/30] [1950/5004] eta: 0:36:24 lr: 0.000047 loss: 1.625900 (1.667408) time: 0.714114 data: 0.000223 max mem: 14338 Epoch: [13/30] [2000/5004] eta: 0:35:48 lr: 0.000047 loss: 1.592767 (1.666668) time: 0.714504 data: 0.000169 max mem: 14338 Epoch: [13/30] [2050/5004] eta: 0:35:12 lr: 0.000047 loss: 1.651690 (1.666120) time: 0.716982 data: 0.000161 max mem: 14338 Epoch: [13/30] [2100/5004] eta: 0:34:36 lr: 0.000047 loss: 1.549139 (1.665471) time: 0.718521 data: 0.000229 max mem: 14338 Epoch: [13/30] [2150/5004] eta: 0:34:00 lr: 0.000047 loss: 1.744930 (1.667049) time: 0.708850 data: 0.000230 max mem: 14338 Epoch: [13/30] [2200/5004] eta: 0:33:24 lr: 0.000047 loss: 1.683098 (1.666893) time: 0.714131 data: 0.000215 max mem: 14338 Epoch: [13/30] [2250/5004] eta: 0:32:49 lr: 0.000047 loss: 1.732942 (1.668973) time: 0.725068 data: 0.000223 max mem: 14338 Epoch: [13/30] [2300/5004] eta: 0:32:13 lr: 0.000047 loss: 1.671881 (1.669592) time: 0.709625 data: 0.000228 max mem: 14338 Epoch: [13/30] [2350/5004] eta: 0:31:37 lr: 0.000047 loss: 1.748934 (1.669827) time: 0.713793 data: 0.000171 max mem: 14338 Epoch: [13/30] [2400/5004] eta: 0:31:02 lr: 0.000047 loss: 1.616176 (1.669669) time: 0.714958 data: 0.000234 max mem: 14338 Epoch: [13/30] [2450/5004] eta: 0:30:26 lr: 0.000047 loss: 1.498810 (1.667857) time: 0.715789 data: 0.000230 max mem: 14338 Epoch: [13/30] [2500/5004] eta: 0:29:50 lr: 0.000047 loss: 1.761503 (1.668209) time: 0.713397 data: 0.000224 max mem: 14338 Epoch: [13/30] [2550/5004] eta: 0:29:14 lr: 0.000047 loss: 1.396626 (1.667995) time: 0.711136 data: 0.000212 max mem: 14338 Epoch: [13/30] [2600/5004] eta: 0:28:38 lr: 0.000046 loss: 1.564201 (1.666872) time: 0.719026 data: 0.000210 max mem: 14338 Epoch: [13/30] [2650/5004] eta: 0:28:02 lr: 0.000046 loss: 1.696283 (1.666812) time: 0.713816 data: 0.000163 max mem: 14338 Epoch: [13/30] [2700/5004] eta: 0:27:27 lr: 0.000046 loss: 1.585458 (1.667972) time: 0.709711 data: 0.000165 max mem: 14338 Epoch: [13/30] [2750/5004] eta: 0:26:51 lr: 0.000046 loss: 1.596393 (1.667201) time: 0.710145 data: 0.000208 max mem: 14338 Epoch: [13/30] [2800/5004] eta: 0:26:15 lr: 0.000046 loss: 1.596386 (1.667018) time: 0.715901 data: 0.000203 max mem: 14338 Epoch: [13/30] [2850/5004] eta: 0:25:39 lr: 0.000046 loss: 1.495766 (1.666303) time: 0.712321 data: 0.000185 max mem: 14338 Epoch: [13/30] [2900/5004] eta: 0:25:04 lr: 0.000046 loss: 1.551566 (1.665957) time: 0.713153 data: 0.000222 max mem: 14338 Epoch: [13/30] [2950/5004] eta: 0:24:28 lr: 0.000046 loss: 1.557655 (1.665191) time: 0.712649 data: 0.000232 max mem: 14338 Epoch: [13/30] [3000/5004] eta: 0:23:52 lr: 0.000046 loss: 1.622688 (1.664939) time: 0.720814 data: 0.000162 max mem: 14338 Epoch: [13/30] [3050/5004] eta: 0:23:16 lr: 0.000046 loss: 1.668079 (1.664833) time: 0.717443 data: 0.000165 max mem: 14338 Epoch: [13/30] [3100/5004] eta: 0:22:41 lr: 0.000046 loss: 1.622861 (1.665188) time: 0.710646 data: 0.000209 max mem: 14338 Epoch: [13/30] [3150/5004] eta: 0:22:05 lr: 0.000046 loss: 1.656478 (1.664635) time: 0.711762 data: 0.000222 max mem: 14338 Epoch: [13/30] [3200/5004] eta: 0:21:29 lr: 0.000046 loss: 1.536054 (1.664212) time: 0.721086 data: 0.000219 max mem: 14338 Epoch: [13/30] [3250/5004] eta: 0:20:53 lr: 0.000046 loss: 1.598356 (1.664233) time: 0.712874 data: 0.000223 max mem: 14338 Epoch: [13/30] [3300/5004] eta: 0:20:17 lr: 0.000046 loss: 1.602831 (1.664466) time: 0.709605 data: 0.000232 max mem: 14338 Epoch: [13/30] [3350/5004] eta: 0:19:42 lr: 0.000046 loss: 1.640709 (1.664673) time: 0.710133 data: 0.000159 max mem: 14338 Epoch: [13/30] [3400/5004] eta: 0:19:06 lr: 0.000046 loss: 1.797431 (1.665886) time: 0.720971 data: 0.000157 max mem: 14338 Epoch: [13/30] [3450/5004] eta: 0:18:30 lr: 0.000046 loss: 1.450420 (1.665689) time: 0.718153 data: 0.000217 max mem: 14338 Epoch: [13/30] [3500/5004] eta: 0:17:55 lr: 0.000046 loss: 1.627615 (1.665420) time: 0.720780 data: 0.000192 max mem: 14338 Epoch: [13/30] [3550/5004] eta: 0:17:19 lr: 0.000046 loss: 1.483886 (1.665593) time: 0.713046 data: 0.000228 max mem: 14338 Epoch: [13/30] [3600/5004] eta: 0:16:43 lr: 0.000046 loss: 1.543262 (1.664922) time: 0.715055 data: 0.000194 max mem: 14338 Epoch: [13/30] [3650/5004] eta: 0:16:07 lr: 0.000046 loss: 1.515905 (1.664924) time: 0.712503 data: 0.000216 max mem: 14338 Epoch: [13/30] [3700/5004] eta: 0:15:32 lr: 0.000046 loss: 1.603795 (1.665233) time: 0.711746 data: 0.000179 max mem: 14338 Epoch: [13/30] [3750/5004] eta: 0:14:56 lr: 0.000046 loss: 1.550881 (1.665132) time: 0.709609 data: 0.000226 max mem: 14338 Epoch: [13/30] [3800/5004] eta: 0:14:20 lr: 0.000045 loss: 1.619153 (1.665375) time: 0.721074 data: 0.000204 max mem: 14338 Epoch: [13/30] [3850/5004] eta: 0:13:44 lr: 0.000045 loss: 1.484655 (1.664781) time: 0.711565 data: 0.000199 max mem: 14338 Epoch: [13/30] [3900/5004] eta: 0:13:09 lr: 0.000045 loss: 1.534169 (1.664363) time: 0.714532 data: 0.000216 max mem: 14338 Epoch: [13/30] [3950/5004] eta: 0:12:33 lr: 0.000045 loss: 1.432688 (1.665194) time: 0.710776 data: 0.000227 max mem: 14338 Epoch: [13/30] [4000/5004] eta: 0:11:57 lr: 0.000045 loss: 1.601492 (1.665909) time: 0.721502 data: 0.000171 max mem: 14338 Epoch: [13/30] [4050/5004] eta: 0:11:21 lr: 0.000045 loss: 1.556861 (1.666015) time: 0.719940 data: 0.000166 max mem: 14338 Epoch: [13/30] [4100/5004] eta: 0:10:46 lr: 0.000045 loss: 1.691033 (1.666120) time: 0.709460 data: 0.000219 max mem: 14338 Epoch: [13/30] [4150/5004] eta: 0:10:10 lr: 0.000045 loss: 1.599211 (1.666471) time: 0.710364 data: 0.000203 max mem: 14338 Epoch: [13/30] [4200/5004] eta: 0:09:34 lr: 0.000045 loss: 1.642809 (1.665982) time: 0.715626 data: 0.000206 max mem: 14338 Epoch: [13/30] [4250/5004] eta: 0:08:58 lr: 0.000045 loss: 1.686147 (1.666477) time: 0.712902 data: 0.000235 max mem: 14338 Epoch: [13/30] [4300/5004] eta: 0:08:23 lr: 0.000045 loss: 1.526918 (1.667145) time: 0.713897 data: 0.000215 max mem: 14338 Epoch: [13/30] [4350/5004] eta: 0:07:47 lr: 0.000045 loss: 1.654580 (1.667502) time: 0.712144 data: 0.000164 max mem: 14338 Epoch: [13/30] [4400/5004] eta: 0:07:11 lr: 0.000045 loss: 1.708200 (1.667300) time: 0.721779 data: 0.000168 max mem: 14338 Epoch: [13/30] [4450/5004] eta: 0:06:35 lr: 0.000045 loss: 1.581746 (1.667600) time: 0.716981 data: 0.000217 max mem: 14338 Epoch: [13/30] [4500/5004] eta: 0:06:00 lr: 0.000045 loss: 1.714013 (1.667797) time: 0.717218 data: 0.000216 max mem: 14338 Epoch: [13/30] [4550/5004] eta: 0:05:24 lr: 0.000045 loss: 1.710682 (1.668593) time: 0.713690 data: 0.000229 max mem: 14338 Epoch: [13/30] [4600/5004] eta: 0:04:48 lr: 0.000045 loss: 1.676850 (1.668351) time: 0.712627 data: 0.000230 max mem: 14338 Epoch: [13/30] [4650/5004] eta: 0:04:13 lr: 0.000045 loss: 1.618598 (1.668235) time: 0.711071 data: 0.000209 max mem: 14338 Epoch: [13/30] [4700/5004] eta: 0:03:37 lr: 0.000045 loss: 1.641698 (1.668330) time: 0.714954 data: 0.000182 max mem: 14338 Epoch: [13/30] [4750/5004] eta: 0:03:01 lr: 0.000045 loss: 1.678672 (1.668161) time: 0.709385 data: 0.000168 max mem: 14338 Epoch: [13/30] [4800/5004] eta: 0:02:25 lr: 0.000045 loss: 1.620970 (1.667613) time: 0.714375 data: 0.000201 max mem: 14338 Epoch: [13/30] [4850/5004] eta: 0:01:50 lr: 0.000045 loss: 1.496670 (1.666770) time: 0.717394 data: 0.000220 max mem: 14338 Epoch: [13/30] [4900/5004] eta: 0:01:14 lr: 0.000045 loss: 1.595376 (1.666723) time: 0.718191 data: 0.000220 max mem: 14338 Epoch: [13/30] [4950/5004] eta: 0:00:38 lr: 0.000045 loss: 1.543342 (1.666539) time: 0.713610 data: 0.000223 max mem: 14338 Epoch: [13/30] [5000/5004] eta: 0:00:02 lr: 0.000044 loss: 1.434552 (1.666053) time: 0.713838 data: 0.000830 max mem: 14338 Epoch: [13/30] [5003/5004] eta: 0:00:00 lr: 0.000044 loss: 1.434552 (1.665984) time: 0.711097 data: 0.000818 max mem: 14338 Epoch: [13/30] Total time: 0:59:37 (0.714915 s / it) Averaged stats: lr: 0.000044 loss: 1.434552 (1.665228) Test: [ 0/196] eta: 0:05:20 loss: 0.302411 (0.302411) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.633579 data: 1.231838 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 0.464153 (0.540336) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.408576 data: 0.112124 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.522300 (0.543092) acc1: 87.500000 (85.714286) acc5: 100.000000 (98.214286) time: 0.285942 data: 0.000128 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.503824 (0.512154) acc1: 87.500000 (87.500000) acc5: 100.000000 (98.387097) time: 0.285933 data: 0.000108 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.394103 (0.514793) acc1: 87.500000 (87.195122) acc5: 100.000000 (98.018293) time: 0.286239 data: 0.000123 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.469360 (0.542803) acc1: 87.500000 (87.132353) acc5: 93.750000 (97.426471) time: 0.286501 data: 0.000127 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.628931 (0.570593) acc1: 81.250000 (86.475410) acc5: 93.750000 (97.438525) time: 0.286724 data: 0.000129 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.676204 (0.587225) acc1: 81.250000 (86.091549) acc5: 100.000000 (97.623239) time: 0.286878 data: 0.000127 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.525217 (0.585799) acc1: 87.500000 (86.111111) acc5: 100.000000 (97.685185) time: 0.287226 data: 0.000120 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.501750 (0.605996) acc1: 81.250000 (85.576923) acc5: 100.000000 (97.390110) time: 0.287098 data: 0.000144 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.570803 (0.595012) acc1: 81.250000 (85.643564) acc5: 100.000000 (97.586634) time: 0.286830 data: 0.000147 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.562811 (0.586336) acc1: 87.500000 (85.641892) acc5: 100.000000 (97.691441) time: 0.287078 data: 0.000135 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.499270 (0.583245) acc1: 87.500000 (85.692149) acc5: 100.000000 (97.675620) time: 0.286574 data: 0.000143 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.582930 (0.597402) acc1: 81.250000 (85.400763) acc5: 100.000000 (97.662214) time: 0.288920 data: 0.000133 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.624793 (0.595839) acc1: 81.250000 (85.505319) acc5: 100.000000 (97.606383) time: 0.293113 data: 0.000134 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.535626 (0.602924) acc1: 87.500000 (85.430464) acc5: 100.000000 (97.640728) time: 0.290812 data: 0.000139 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.484962 (0.609180) acc1: 81.250000 (85.326087) acc5: 100.000000 (97.631988) time: 0.287460 data: 0.000121 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.471170 (0.606281) acc1: 81.250000 (85.380117) acc5: 100.000000 (97.697368) time: 0.286553 data: 0.000135 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.462122 (0.607178) acc1: 87.500000 (85.186464) acc5: 100.000000 (97.720994) time: 0.285418 data: 0.000160 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.346872 (0.596450) acc1: 87.500000 (85.438482) acc5: 100.000000 (97.742147) time: 0.283244 data: 0.000119 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.534611 (0.608246) acc1: 87.500000 (85.344000) acc5: 100.000000 (97.632000) time: 0.273869 data: 0.000112 max mem: 14338 Test: Total time: 0:00:57 (0.294238 s / it) * Acc@1 85.044 Acc@5 97.300 loss 0.628 Max accuracy: 85.14% Epoch: [14/30] [ 0/5004] eta: 2:39:23 lr: 0.000044 loss: 1.688413 (1.688413) time: 1.911106 data: 1.169099 max mem: 14338 Epoch: [14/30] [ 50/5004] eta: 1:01:01 lr: 0.000044 loss: 1.435373 (1.647142) time: 0.718641 data: 0.000197 max mem: 14338 Epoch: [14/30] [ 100/5004] eta: 0:59:19 lr: 0.000044 loss: 1.454641 (1.629078) time: 0.712377 data: 0.000215 max mem: 14338 Epoch: [14/30] [ 150/5004] eta: 0:58:29 lr: 0.000044 loss: 1.680276 (1.647123) time: 0.712911 data: 0.000224 max mem: 14338 Epoch: [14/30] [ 200/5004] eta: 0:57:50 lr: 0.000044 loss: 1.503142 (1.634626) time: 0.712892 data: 0.000215 max mem: 14338 Epoch: [14/30] [ 250/5004] eta: 0:57:09 lr: 0.000044 loss: 1.564090 (1.645895) time: 0.719395 data: 0.000195 max mem: 14338 Epoch: [14/30] [ 300/5004] eta: 0:56:28 lr: 0.000044 loss: 1.677744 (1.654647) time: 0.710270 data: 0.000170 max mem: 14338 Epoch: [14/30] [ 350/5004] eta: 0:55:48 lr: 0.000044 loss: 1.555493 (1.656076) time: 0.715379 data: 0.000181 max mem: 14338 Epoch: [14/30] [ 400/5004] eta: 0:55:11 lr: 0.000044 loss: 1.553275 (1.655849) time: 0.723759 data: 0.000213 max mem: 14338 Epoch: [14/30] [ 450/5004] eta: 0:54:31 lr: 0.000044 loss: 1.518868 (1.657966) time: 0.714287 data: 0.000211 max mem: 14338 Epoch: [14/30] [ 500/5004] eta: 0:53:53 lr: 0.000044 loss: 1.608453 (1.659906) time: 0.711961 data: 0.000226 max mem: 14338 Epoch: [14/30] [ 550/5004] eta: 0:53:15 lr: 0.000044 loss: 1.523533 (1.656096) time: 0.710242 data: 0.000233 max mem: 14338 Epoch: [14/30] [ 600/5004] eta: 0:52:39 lr: 0.000044 loss: 1.663339 (1.650230) time: 0.714764 data: 0.000224 max mem: 14338 Epoch: [14/30] [ 650/5004] eta: 0:52:02 lr: 0.000044 loss: 1.516591 (1.647374) time: 0.712874 data: 0.000150 max mem: 14338 Epoch: [14/30] [ 700/5004] eta: 0:51:26 lr: 0.000044 loss: 1.649752 (1.647489) time: 0.713134 data: 0.000174 max mem: 14338 Epoch: [14/30] [ 750/5004] eta: 0:50:49 lr: 0.000044 loss: 1.577197 (1.647070) time: 0.710066 data: 0.000221 max mem: 14338 Epoch: [14/30] [ 800/5004] eta: 0:50:13 lr: 0.000044 loss: 1.650684 (1.649014) time: 0.717322 data: 0.000224 max mem: 14338 Epoch: [14/30] [ 850/5004] eta: 0:49:36 lr: 0.000044 loss: 1.695529 (1.650094) time: 0.713370 data: 0.000225 max mem: 14338 Epoch: [14/30] [ 900/5004] eta: 0:49:00 lr: 0.000044 loss: 1.473742 (1.649018) time: 0.715110 data: 0.000193 max mem: 14338 Epoch: [14/30] [ 950/5004] eta: 0:48:23 lr: 0.000044 loss: 1.590124 (1.649442) time: 0.713104 data: 0.000213 max mem: 14338 Epoch: [14/30] [1000/5004] eta: 0:47:47 lr: 0.000044 loss: 1.674327 (1.649238) time: 0.718165 data: 0.000161 max mem: 14338 Epoch: [14/30] [1050/5004] eta: 0:47:11 lr: 0.000044 loss: 1.853410 (1.651690) time: 0.714970 data: 0.000207 max mem: 14338 Epoch: [14/30] [1100/5004] eta: 0:46:34 lr: 0.000044 loss: 1.884391 (1.655524) time: 0.710264 data: 0.000220 max mem: 14338 Epoch: [14/30] [1150/5004] eta: 0:45:58 lr: 0.000044 loss: 1.680394 (1.655803) time: 0.711399 data: 0.000238 max mem: 14338 Epoch: [14/30] [1200/5004] eta: 0:45:22 lr: 0.000044 loss: 1.515895 (1.655816) time: 0.714713 data: 0.000205 max mem: 14338 Epoch: [14/30] [1250/5004] eta: 0:44:46 lr: 0.000043 loss: 1.567997 (1.656084) time: 0.714181 data: 0.000211 max mem: 14338 Epoch: [14/30] [1300/5004] eta: 0:44:10 lr: 0.000043 loss: 1.485236 (1.656289) time: 0.713688 data: 0.000173 max mem: 14338 Epoch: [14/30] [1350/5004] eta: 0:43:34 lr: 0.000043 loss: 1.656545 (1.657639) time: 0.713062 data: 0.000187 max mem: 14338 Epoch: [14/30] [1400/5004] eta: 0:42:59 lr: 0.000043 loss: 1.507159 (1.656044) time: 0.720784 data: 0.000240 max mem: 14338 Epoch: [14/30] [1450/5004] eta: 0:42:22 lr: 0.000043 loss: 1.484974 (1.654434) time: 0.713036 data: 0.000231 max mem: 14338 Epoch: [14/30] [1500/5004] eta: 0:41:46 lr: 0.000043 loss: 1.652689 (1.652775) time: 0.710118 data: 0.000220 max mem: 14338 Epoch: [14/30] [1550/5004] eta: 0:41:10 lr: 0.000043 loss: 1.581920 (1.651428) time: 0.710917 data: 0.000192 max mem: 14338 Epoch: [14/30] [1600/5004] eta: 0:40:34 lr: 0.000043 loss: 1.455222 (1.652026) time: 0.714341 data: 0.000199 max mem: 14338 Epoch: [14/30] [1650/5004] eta: 0:39:59 lr: 0.000043 loss: 1.485329 (1.652619) time: 0.711173 data: 0.000177 max mem: 14338 Epoch: [14/30] [1700/5004] eta: 0:39:23 lr: 0.000043 loss: 1.565211 (1.650904) time: 0.711294 data: 0.000177 max mem: 14338 Epoch: [14/30] [1750/5004] eta: 0:38:47 lr: 0.000043 loss: 1.628010 (1.651000) time: 0.716097 data: 0.000212 max mem: 14338 Epoch: [14/30] [1800/5004] eta: 0:38:11 lr: 0.000043 loss: 1.629633 (1.650444) time: 0.717742 data: 0.000227 max mem: 14338 Epoch: [14/30] [1850/5004] eta: 0:37:35 lr: 0.000043 loss: 1.721472 (1.654511) time: 0.721018 data: 0.000242 max mem: 14338 Epoch: [14/30] [1900/5004] eta: 0:36:59 lr: 0.000043 loss: 1.714756 (1.657156) time: 0.711297 data: 0.000216 max mem: 14338 Epoch: [14/30] [1950/5004] eta: 0:36:24 lr: 0.000043 loss: 1.653433 (1.658418) time: 0.709275 data: 0.000233 max mem: 14338 Epoch: [14/30] [2000/5004] eta: 0:35:48 lr: 0.000043 loss: 1.533867 (1.657232) time: 0.713187 data: 0.000161 max mem: 14338 Epoch: [14/30] [2050/5004] eta: 0:35:12 lr: 0.000043 loss: 1.667087 (1.657694) time: 0.717979 data: 0.000169 max mem: 14338 Epoch: [14/30] [2100/5004] eta: 0:34:36 lr: 0.000043 loss: 1.531834 (1.658016) time: 0.710495 data: 0.000246 max mem: 14338 Epoch: [14/30] [2150/5004] eta: 0:34:00 lr: 0.000043 loss: 1.624372 (1.659168) time: 0.713248 data: 0.000240 max mem: 14338 Epoch: [14/30] [2200/5004] eta: 0:33:25 lr: 0.000043 loss: 1.595701 (1.659208) time: 0.712281 data: 0.000189 max mem: 14338 Epoch: [14/30] [2250/5004] eta: 0:32:49 lr: 0.000043 loss: 1.688939 (1.659554) time: 0.719641 data: 0.000207 max mem: 14338 Epoch: [14/30] [2300/5004] eta: 0:32:13 lr: 0.000043 loss: 1.701530 (1.659531) time: 0.716382 data: 0.000220 max mem: 14338 Epoch: [14/30] [2350/5004] eta: 0:31:37 lr: 0.000043 loss: 1.573187 (1.658546) time: 0.713179 data: 0.000160 max mem: 14338 Epoch: [14/30] [2400/5004] eta: 0:31:01 lr: 0.000043 loss: 1.592421 (1.658915) time: 0.715546 data: 0.000228 max mem: 14338 Epoch: [14/30] [2450/5004] eta: 0:30:26 lr: 0.000042 loss: 1.608792 (1.658611) time: 0.715831 data: 0.000201 max mem: 14338 Epoch: [14/30] [2500/5004] eta: 0:29:50 lr: 0.000042 loss: 1.695989 (1.657302) time: 0.714788 data: 0.000219 max mem: 14338 Epoch: [14/30] [2550/5004] eta: 0:29:14 lr: 0.000042 loss: 1.613140 (1.658227) time: 0.709221 data: 0.000218 max mem: 14338 Epoch: [14/30] [2600/5004] eta: 0:28:38 lr: 0.000042 loss: 1.663899 (1.657691) time: 0.712268 data: 0.000229 max mem: 14338 Epoch: [14/30] [2650/5004] eta: 0:28:02 lr: 0.000042 loss: 1.614260 (1.658116) time: 0.715675 data: 0.000167 max mem: 14338 Epoch: [14/30] [2700/5004] eta: 0:27:27 lr: 0.000042 loss: 1.598822 (1.658174) time: 0.713838 data: 0.000153 max mem: 14338 Epoch: [14/30] [2750/5004] eta: 0:26:51 lr: 0.000042 loss: 1.596247 (1.657688) time: 0.716819 data: 0.000215 max mem: 14338 Epoch: [14/30] [2800/5004] eta: 0:26:15 lr: 0.000042 loss: 1.623644 (1.656954) time: 0.717404 data: 0.000219 max mem: 14338 Epoch: [14/30] [2850/5004] eta: 0:25:39 lr: 0.000042 loss: 1.596707 (1.656216) time: 0.715021 data: 0.000201 max mem: 14338 Epoch: [14/30] [2900/5004] eta: 0:25:03 lr: 0.000042 loss: 1.610496 (1.656844) time: 0.710882 data: 0.000221 max mem: 14338 Epoch: [14/30] [2950/5004] eta: 0:24:28 lr: 0.000042 loss: 1.577476 (1.657764) time: 0.712567 data: 0.000225 max mem: 14338 Epoch: [14/30] [3000/5004] eta: 0:23:52 lr: 0.000042 loss: 1.581213 (1.657836) time: 0.715529 data: 0.000162 max mem: 14338 Epoch: [14/30] [3050/5004] eta: 0:23:16 lr: 0.000042 loss: 1.600768 (1.657393) time: 0.713253 data: 0.000156 max mem: 14338 Epoch: [14/30] [3100/5004] eta: 0:22:40 lr: 0.000042 loss: 1.624528 (1.657541) time: 0.712687 data: 0.000218 max mem: 14338 Epoch: [14/30] [3150/5004] eta: 0:22:05 lr: 0.000042 loss: 1.723835 (1.657471) time: 0.714392 data: 0.000226 max mem: 14338 Epoch: [14/30] [3200/5004] eta: 0:21:29 lr: 0.000042 loss: 1.686426 (1.657114) time: 0.720618 data: 0.000206 max mem: 14338 Epoch: [14/30] [3250/5004] eta: 0:20:53 lr: 0.000042 loss: 1.581972 (1.657861) time: 0.718531 data: 0.000223 max mem: 14338 Epoch: [14/30] [3300/5004] eta: 0:20:17 lr: 0.000042 loss: 1.484738 (1.656889) time: 0.713916 data: 0.000220 max mem: 14338 Epoch: [14/30] [3350/5004] eta: 0:19:42 lr: 0.000042 loss: 1.708881 (1.657653) time: 0.710450 data: 0.000163 max mem: 14338 Epoch: [14/30] [3400/5004] eta: 0:19:06 lr: 0.000042 loss: 1.656197 (1.658039) time: 0.719638 data: 0.000180 max mem: 14338 Epoch: [14/30] [3450/5004] eta: 0:18:30 lr: 0.000042 loss: 1.576092 (1.657246) time: 0.718186 data: 0.000217 max mem: 14338 Epoch: [14/30] [3500/5004] eta: 0:17:55 lr: 0.000042 loss: 1.512943 (1.657197) time: 0.713518 data: 0.000183 max mem: 14338 Epoch: [14/30] [3550/5004] eta: 0:17:19 lr: 0.000042 loss: 1.449627 (1.656916) time: 0.709906 data: 0.000217 max mem: 14338 Epoch: [14/30] [3600/5004] eta: 0:16:43 lr: 0.000042 loss: 1.636256 (1.657059) time: 0.718366 data: 0.000215 max mem: 14338 Epoch: [14/30] [3650/5004] eta: 0:16:07 lr: 0.000041 loss: 1.715236 (1.657399) time: 0.720771 data: 0.000222 max mem: 14338 Epoch: [14/30] [3700/5004] eta: 0:15:32 lr: 0.000041 loss: 1.659131 (1.657923) time: 0.719757 data: 0.000187 max mem: 14338 Epoch: [14/30] [3750/5004] eta: 0:14:56 lr: 0.000041 loss: 1.504773 (1.656740) time: 0.714492 data: 0.000241 max mem: 14338 Epoch: [14/30] [3800/5004] eta: 0:14:20 lr: 0.000041 loss: 1.656533 (1.656716) time: 0.717345 data: 0.000217 max mem: 14338 Epoch: [14/30] [3850/5004] eta: 0:13:44 lr: 0.000041 loss: 1.660492 (1.655635) time: 0.715740 data: 0.000197 max mem: 14338 Epoch: [14/30] [3900/5004] eta: 0:13:09 lr: 0.000041 loss: 1.589075 (1.655933) time: 0.711958 data: 0.000234 max mem: 14338 Epoch: [14/30] [3950/5004] eta: 0:12:33 lr: 0.000041 loss: 1.536697 (1.655571) time: 0.709338 data: 0.000232 max mem: 14338 Epoch: [14/30] [4000/5004] eta: 0:11:57 lr: 0.000041 loss: 1.534685 (1.654988) time: 0.713533 data: 0.000177 max mem: 14338 Epoch: [14/30] [4050/5004] eta: 0:11:21 lr: 0.000041 loss: 1.622554 (1.655629) time: 0.718384 data: 0.000172 max mem: 14338 Epoch: [14/30] [4100/5004] eta: 0:10:46 lr: 0.000041 loss: 1.616448 (1.655477) time: 0.709579 data: 0.000208 max mem: 14338 Epoch: [14/30] [4150/5004] eta: 0:10:10 lr: 0.000041 loss: 1.537821 (1.655249) time: 0.716938 data: 0.000185 max mem: 14338 Epoch: [14/30] [4200/5004] eta: 0:09:34 lr: 0.000041 loss: 1.554974 (1.655449) time: 0.715011 data: 0.000209 max mem: 14338 Epoch: [14/30] [4250/5004] eta: 0:08:58 lr: 0.000041 loss: 1.620106 (1.655799) time: 0.723381 data: 0.000217 max mem: 14338 Epoch: [14/30] [4300/5004] eta: 0:08:23 lr: 0.000041 loss: 1.738000 (1.656017) time: 0.714737 data: 0.000207 max mem: 14338 Epoch: [14/30] [4350/5004] eta: 0:07:47 lr: 0.000041 loss: 1.632178 (1.655841) time: 0.712936 data: 0.000165 max mem: 14338 Epoch: [14/30] [4400/5004] eta: 0:07:11 lr: 0.000041 loss: 1.408936 (1.655720) time: 0.718866 data: 0.000152 max mem: 14338 Epoch: [14/30] [4450/5004] eta: 0:06:36 lr: 0.000041 loss: 1.569605 (1.655428) time: 0.715246 data: 0.000240 max mem: 14338 Epoch: [14/30] [4500/5004] eta: 0:06:00 lr: 0.000041 loss: 1.476384 (1.655047) time: 0.709685 data: 0.000225 max mem: 14338 Epoch: [14/30] [4550/5004] eta: 0:05:24 lr: 0.000041 loss: 1.467110 (1.654806) time: 0.710075 data: 0.000220 max mem: 14338 Epoch: [14/30] [4600/5004] eta: 0:04:48 lr: 0.000041 loss: 1.733781 (1.655014) time: 0.720953 data: 0.000208 max mem: 14338 Epoch: [14/30] [4650/5004] eta: 0:04:13 lr: 0.000041 loss: 1.642038 (1.655084) time: 0.716136 data: 0.000209 max mem: 14338 Epoch: [14/30] [4700/5004] eta: 0:03:37 lr: 0.000041 loss: 1.507614 (1.654896) time: 0.715017 data: 0.000149 max mem: 14338 Epoch: [14/30] [4750/5004] eta: 0:03:01 lr: 0.000041 loss: 1.606164 (1.655145) time: 0.710382 data: 0.000178 max mem: 14338 Epoch: [14/30] [4800/5004] eta: 0:02:25 lr: 0.000041 loss: 1.544526 (1.655911) time: 0.713062 data: 0.000199 max mem: 14338 Epoch: [14/30] [4850/5004] eta: 0:01:50 lr: 0.000040 loss: 1.518656 (1.655197) time: 0.711227 data: 0.000221 max mem: 14338 Epoch: [14/30] [4900/5004] eta: 0:01:14 lr: 0.000040 loss: 1.664819 (1.655443) time: 0.711640 data: 0.000224 max mem: 14338 Epoch: [14/30] [4950/5004] eta: 0:00:38 lr: 0.000040 loss: 1.486432 (1.654865) time: 0.708887 data: 0.000236 max mem: 14338 Epoch: [14/30] [5000/5004] eta: 0:00:02 lr: 0.000040 loss: 1.717629 (1.655905) time: 0.709324 data: 0.000838 max mem: 14338 Epoch: [14/30] [5003/5004] eta: 0:00:00 lr: 0.000040 loss: 1.717629 (1.655804) time: 0.706442 data: 0.000830 max mem: 14338 Epoch: [14/30] Total time: 0:59:37 (0.714905 s / it) Averaged stats: lr: 0.000040 loss: 1.717629 (1.655937) Test: [ 0/196] eta: 0:04:58 loss: 0.272455 (0.272455) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.520696 data: 1.124168 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.520761 (0.554215) acc1: 87.500000 (85.227273) acc5: 100.000000 (98.863636) time: 0.398903 data: 0.102335 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.553091 (0.552151) acc1: 87.500000 (85.714286) acc5: 100.000000 (98.214286) time: 0.286492 data: 0.000134 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.502470 (0.520041) acc1: 87.500000 (87.298387) acc5: 100.000000 (98.387097) time: 0.286769 data: 0.000119 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.412808 (0.524590) acc1: 87.500000 (87.500000) acc5: 100.000000 (98.018293) time: 0.286969 data: 0.000129 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.433035 (0.553144) acc1: 87.500000 (87.254902) acc5: 100.000000 (97.426471) time: 0.287098 data: 0.000121 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.609138 (0.581445) acc1: 87.500000 (86.782787) acc5: 93.750000 (97.336066) time: 0.294907 data: 0.000118 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.625679 (0.598129) acc1: 81.250000 (86.443662) acc5: 100.000000 (97.447183) time: 0.295125 data: 0.000133 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.517659 (0.600308) acc1: 87.500000 (86.188272) acc5: 100.000000 (97.530864) time: 0.287872 data: 0.000128 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.539953 (0.621385) acc1: 87.500000 (85.851648) acc5: 100.000000 (97.252747) time: 0.287389 data: 0.000129 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.565202 (0.609682) acc1: 81.250000 (85.891089) acc5: 100.000000 (97.462871) time: 0.287701 data: 0.000143 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.547247 (0.599861) acc1: 87.500000 (85.754505) acc5: 100.000000 (97.578829) time: 0.287766 data: 0.000136 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.504158 (0.596202) acc1: 87.500000 (85.743802) acc5: 100.000000 (97.572314) time: 0.287319 data: 0.000148 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.661959 (0.609542) acc1: 87.500000 (85.543893) acc5: 100.000000 (97.566794) time: 0.286997 data: 0.000154 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.696539 (0.608402) acc1: 87.500000 (85.638298) acc5: 100.000000 (97.562057) time: 0.286298 data: 0.000152 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.531434 (0.616706) acc1: 87.500000 (85.596026) acc5: 100.000000 (97.599338) time: 0.286171 data: 0.000158 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.504810 (0.620451) acc1: 81.250000 (85.520186) acc5: 100.000000 (97.631988) time: 0.287387 data: 0.000143 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.468355 (0.616524) acc1: 81.250000 (85.635965) acc5: 100.000000 (97.697368) time: 0.287432 data: 0.000140 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.468355 (0.616600) acc1: 87.500000 (85.531768) acc5: 100.000000 (97.686464) time: 0.286008 data: 0.000147 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.366802 (0.606126) acc1: 87.500000 (85.732984) acc5: 100.000000 (97.709424) time: 0.283785 data: 0.000112 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.527404 (0.617560) acc1: 87.500000 (85.568000) acc5: 100.000000 (97.632000) time: 0.273655 data: 0.000098 max mem: 14338 Test: Total time: 0:00:57 (0.294345 s / it) * Acc@1 85.000 Acc@5 97.306 loss 0.630 Max accuracy: 85.14% Epoch: [15/30] [ 0/5004] eta: 2:44:09 lr: 0.000040 loss: 1.899029 (1.899029) time: 1.968252 data: 1.233292 max mem: 14338 Epoch: [15/30] [ 50/5004] eta: 1:01:06 lr: 0.000040 loss: 1.769525 (1.714939) time: 0.721226 data: 0.000200 max mem: 14338 Epoch: [15/30] [ 100/5004] eta: 0:59:38 lr: 0.000040 loss: 1.507130 (1.683189) time: 0.721576 data: 0.000241 max mem: 14338 Epoch: [15/30] [ 150/5004] eta: 0:58:54 lr: 0.000040 loss: 1.541915 (1.675596) time: 0.712610 data: 0.000240 max mem: 14338 Epoch: [15/30] [ 200/5004] eta: 0:58:01 lr: 0.000040 loss: 1.508105 (1.660668) time: 0.716503 data: 0.000220 max mem: 14338 Epoch: [15/30] [ 250/5004] eta: 0:57:15 lr: 0.000040 loss: 1.356341 (1.637798) time: 0.715947 data: 0.000202 max mem: 14338 Epoch: [15/30] [ 300/5004] eta: 0:56:31 lr: 0.000040 loss: 1.649057 (1.636333) time: 0.709198 data: 0.000188 max mem: 14338 Epoch: [15/30] [ 350/5004] eta: 0:55:50 lr: 0.000040 loss: 1.573815 (1.643273) time: 0.712033 data: 0.000171 max mem: 14338 Epoch: [15/30] [ 400/5004] eta: 0:55:11 lr: 0.000040 loss: 1.549758 (1.639753) time: 0.711486 data: 0.000221 max mem: 14338 Epoch: [15/30] [ 450/5004] eta: 0:54:32 lr: 0.000040 loss: 1.510568 (1.640241) time: 0.713760 data: 0.000225 max mem: 14338 Epoch: [15/30] [ 500/5004] eta: 0:53:55 lr: 0.000040 loss: 1.662169 (1.641441) time: 0.713295 data: 0.000219 max mem: 14338 Epoch: [15/30] [ 550/5004] eta: 0:53:18 lr: 0.000040 loss: 1.521056 (1.636319) time: 0.716286 data: 0.000215 max mem: 14338 Epoch: [15/30] [ 600/5004] eta: 0:52:42 lr: 0.000040 loss: 1.595556 (1.640591) time: 0.722523 data: 0.000224 max mem: 14338 Epoch: [15/30] [ 650/5004] eta: 0:52:04 lr: 0.000040 loss: 1.653503 (1.643131) time: 0.720966 data: 0.000169 max mem: 14338 Epoch: [15/30] [ 700/5004] eta: 0:51:27 lr: 0.000040 loss: 1.505626 (1.639237) time: 0.713125 data: 0.000189 max mem: 14338 Epoch: [15/30] [ 750/5004] eta: 0:50:50 lr: 0.000040 loss: 1.609621 (1.632158) time: 0.709880 data: 0.000219 max mem: 14338 Epoch: [15/30] [ 800/5004] eta: 0:50:13 lr: 0.000040 loss: 1.834999 (1.633118) time: 0.713803 data: 0.000221 max mem: 14338 Epoch: [15/30] [ 850/5004] eta: 0:49:37 lr: 0.000040 loss: 1.545490 (1.633337) time: 0.712646 data: 0.000233 max mem: 14338 Epoch: [15/30] [ 900/5004] eta: 0:49:01 lr: 0.000040 loss: 1.698112 (1.636010) time: 0.710261 data: 0.000203 max mem: 14338 Epoch: [15/30] [ 950/5004] eta: 0:48:25 lr: 0.000040 loss: 1.725124 (1.639063) time: 0.713286 data: 0.000215 max mem: 14338 Epoch: [15/30] [1000/5004] eta: 0:47:49 lr: 0.000040 loss: 1.581653 (1.638934) time: 0.722665 data: 0.000161 max mem: 14338 Epoch: [15/30] [1050/5004] eta: 0:47:14 lr: 0.000039 loss: 1.480254 (1.639357) time: 0.715405 data: 0.000231 max mem: 14338 Epoch: [15/30] [1100/5004] eta: 0:46:38 lr: 0.000039 loss: 1.608415 (1.639100) time: 0.720498 data: 0.000238 max mem: 14338 Epoch: [15/30] [1150/5004] eta: 0:46:02 lr: 0.000039 loss: 1.563909 (1.641154) time: 0.720722 data: 0.000228 max mem: 14338 Epoch: [15/30] [1200/5004] eta: 0:45:26 lr: 0.000039 loss: 1.666382 (1.641788) time: 0.713888 data: 0.000242 max mem: 14338 Epoch: [15/30] [1250/5004] eta: 0:44:50 lr: 0.000039 loss: 1.661088 (1.641938) time: 0.713745 data: 0.000226 max mem: 14338 Epoch: [15/30] [1300/5004] eta: 0:44:14 lr: 0.000039 loss: 1.465661 (1.640522) time: 0.711544 data: 0.000171 max mem: 14338 Epoch: [15/30] [1350/5004] eta: 0:43:38 lr: 0.000039 loss: 1.593551 (1.639237) time: 0.713086 data: 0.000172 max mem: 14338 Epoch: [15/30] [1400/5004] eta: 0:43:02 lr: 0.000039 loss: 1.730665 (1.640997) time: 0.713069 data: 0.000234 max mem: 14338 Epoch: [15/30] [1450/5004] eta: 0:42:26 lr: 0.000039 loss: 1.627741 (1.642094) time: 0.713620 data: 0.000207 max mem: 14338 Epoch: [15/30] [1500/5004] eta: 0:41:50 lr: 0.000039 loss: 1.512101 (1.640388) time: 0.718285 data: 0.000219 max mem: 14338 Epoch: [15/30] [1550/5004] eta: 0:41:14 lr: 0.000039 loss: 1.533438 (1.638619) time: 0.714454 data: 0.000195 max mem: 14338 Epoch: [15/30] [1600/5004] eta: 0:40:38 lr: 0.000039 loss: 1.531522 (1.636976) time: 0.721222 data: 0.000210 max mem: 14338 Epoch: [15/30] [1650/5004] eta: 0:40:02 lr: 0.000039 loss: 1.648282 (1.636002) time: 0.713539 data: 0.000163 max mem: 14338 Epoch: [15/30] [1700/5004] eta: 0:39:26 lr: 0.000039 loss: 1.620707 (1.635880) time: 0.709770 data: 0.000162 max mem: 14338 Epoch: [15/30] [1750/5004] eta: 0:38:50 lr: 0.000039 loss: 1.672028 (1.640088) time: 0.712872 data: 0.000215 max mem: 14338 Epoch: [15/30] [1800/5004] eta: 0:38:14 lr: 0.000039 loss: 1.679241 (1.641186) time: 0.712411 data: 0.000206 max mem: 14338 Epoch: [15/30] [1850/5004] eta: 0:37:38 lr: 0.000039 loss: 1.663923 (1.643471) time: 0.713693 data: 0.000228 max mem: 14338 Epoch: [15/30] [1900/5004] eta: 0:37:02 lr: 0.000039 loss: 1.618594 (1.644234) time: 0.712819 data: 0.000212 max mem: 14338 Epoch: [15/30] [1950/5004] eta: 0:36:26 lr: 0.000039 loss: 1.720502 (1.645324) time: 0.715307 data: 0.000219 max mem: 14338 Epoch: [15/30] [2000/5004] eta: 0:35:50 lr: 0.000039 loss: 1.733033 (1.646864) time: 0.716487 data: 0.000163 max mem: 14338 Epoch: [15/30] [2050/5004] eta: 0:35:14 lr: 0.000039 loss: 1.386085 (1.646262) time: 0.712776 data: 0.000153 max mem: 14338 Epoch: [15/30] [2100/5004] eta: 0:34:38 lr: 0.000039 loss: 1.644455 (1.646954) time: 0.711924 data: 0.000212 max mem: 14338 Epoch: [15/30] [2150/5004] eta: 0:34:02 lr: 0.000039 loss: 1.648932 (1.647612) time: 0.710715 data: 0.000224 max mem: 14338 Epoch: [15/30] [2200/5004] eta: 0:33:26 lr: 0.000039 loss: 1.525014 (1.647042) time: 0.719597 data: 0.000193 max mem: 14338 Epoch: [15/30] [2250/5004] eta: 0:32:51 lr: 0.000038 loss: 1.603115 (1.646243) time: 0.715888 data: 0.000227 max mem: 14338 Epoch: [15/30] [2300/5004] eta: 0:32:15 lr: 0.000038 loss: 1.668925 (1.647400) time: 0.711660 data: 0.000227 max mem: 14338 Epoch: [15/30] [2350/5004] eta: 0:31:39 lr: 0.000038 loss: 1.716802 (1.646732) time: 0.709422 data: 0.000156 max mem: 14338 Epoch: [15/30] [2400/5004] eta: 0:31:03 lr: 0.000038 loss: 1.486753 (1.646576) time: 0.718097 data: 0.000233 max mem: 14338 Epoch: [15/30] [2450/5004] eta: 0:30:28 lr: 0.000038 loss: 1.613029 (1.647752) time: 0.721277 data: 0.000240 max mem: 14338 Epoch: [15/30] [2500/5004] eta: 0:29:52 lr: 0.000038 loss: 1.470528 (1.648056) time: 0.713618 data: 0.000230 max mem: 14338 Epoch: [15/30] [2550/5004] eta: 0:29:16 lr: 0.000038 loss: 1.700609 (1.649736) time: 0.711910 data: 0.000221 max mem: 14338 Epoch: [15/30] [2600/5004] eta: 0:28:40 lr: 0.000038 loss: 1.586785 (1.650626) time: 0.714792 data: 0.000223 max mem: 14338 Epoch: [15/30] [2650/5004] eta: 0:28:04 lr: 0.000038 loss: 1.646684 (1.650747) time: 0.713673 data: 0.000166 max mem: 14338 Epoch: [15/30] [2700/5004] eta: 0:27:28 lr: 0.000038 loss: 1.610414 (1.650693) time: 0.711504 data: 0.000169 max mem: 14338 Epoch: [15/30] [2750/5004] eta: 0:26:52 lr: 0.000038 loss: 1.543592 (1.649512) time: 0.711027 data: 0.000221 max mem: 14338 Epoch: [15/30] [2800/5004] eta: 0:26:17 lr: 0.000038 loss: 1.535724 (1.648555) time: 0.716441 data: 0.000206 max mem: 14338 Epoch: [15/30] [2850/5004] eta: 0:25:41 lr: 0.000038 loss: 1.547430 (1.648482) time: 0.716116 data: 0.000184 max mem: 14338 Epoch: [15/30] [2900/5004] eta: 0:25:05 lr: 0.000038 loss: 1.541548 (1.647891) time: 0.713904 data: 0.000219 max mem: 14338 Epoch: [15/30] [2950/5004] eta: 0:24:29 lr: 0.000038 loss: 1.642249 (1.647837) time: 0.714678 data: 0.000227 max mem: 14338 Epoch: [15/30] [3000/5004] eta: 0:23:53 lr: 0.000038 loss: 1.599725 (1.647652) time: 0.715967 data: 0.000153 max mem: 14338 Epoch: [15/30] [3050/5004] eta: 0:23:17 lr: 0.000038 loss: 1.599362 (1.647261) time: 0.713651 data: 0.000161 max mem: 14338 Epoch: [15/30] [3100/5004] eta: 0:22:42 lr: 0.000038 loss: 1.695429 (1.647405) time: 0.716027 data: 0.000227 max mem: 14338 Epoch: [15/30] [3150/5004] eta: 0:22:06 lr: 0.000038 loss: 1.451990 (1.645901) time: 0.711226 data: 0.000211 max mem: 14338 Epoch: [15/30] [3200/5004] eta: 0:21:30 lr: 0.000038 loss: 1.707045 (1.647137) time: 0.712651 data: 0.000218 max mem: 14338 Epoch: [15/30] [3250/5004] eta: 0:20:54 lr: 0.000038 loss: 1.509840 (1.646314) time: 0.716962 data: 0.000219 max mem: 14338 Epoch: [15/30] [3300/5004] eta: 0:20:19 lr: 0.000038 loss: 1.652366 (1.646261) time: 0.710498 data: 0.000234 max mem: 14338 Epoch: [15/30] [3350/5004] eta: 0:19:43 lr: 0.000038 loss: 1.863044 (1.647945) time: 0.716252 data: 0.000169 max mem: 14338 Epoch: [15/30] [3400/5004] eta: 0:19:07 lr: 0.000038 loss: 1.589054 (1.647822) time: 0.720091 data: 0.000167 max mem: 14338 Epoch: [15/30] [3450/5004] eta: 0:18:31 lr: 0.000037 loss: 1.600251 (1.647648) time: 0.716421 data: 0.000229 max mem: 14338 Epoch: [15/30] [3500/5004] eta: 0:17:55 lr: 0.000037 loss: 1.584361 (1.648144) time: 0.711851 data: 0.000203 max mem: 14338 Epoch: [15/30] [3550/5004] eta: 0:17:20 lr: 0.000037 loss: 1.507917 (1.647778) time: 0.710147 data: 0.000218 max mem: 14338 Epoch: [15/30] [3600/5004] eta: 0:16:44 lr: 0.000037 loss: 1.706993 (1.648952) time: 0.720528 data: 0.000206 max mem: 14338 Epoch: [15/30] [3650/5004] eta: 0:16:08 lr: 0.000037 loss: 1.628115 (1.648874) time: 0.719824 data: 0.000192 max mem: 14338 Epoch: [15/30] [3700/5004] eta: 0:15:32 lr: 0.000037 loss: 1.619259 (1.648940) time: 0.710906 data: 0.000162 max mem: 14338 Epoch: [15/30] [3750/5004] eta: 0:14:57 lr: 0.000037 loss: 1.695255 (1.648754) time: 0.712603 data: 0.000204 max mem: 14338 Epoch: [15/30] [3800/5004] eta: 0:14:21 lr: 0.000037 loss: 1.628420 (1.649562) time: 0.722022 data: 0.000224 max mem: 14338 Epoch: [15/30] [3850/5004] eta: 0:13:45 lr: 0.000037 loss: 1.604085 (1.649828) time: 0.717664 data: 0.000197 max mem: 14338 Epoch: [15/30] [3900/5004] eta: 0:13:09 lr: 0.000037 loss: 1.536826 (1.649633) time: 0.718555 data: 0.000229 max mem: 14338 Epoch: [15/30] [3950/5004] eta: 0:12:34 lr: 0.000037 loss: 1.603566 (1.649497) time: 0.713888 data: 0.000222 max mem: 14338 Epoch: [15/30] [4000/5004] eta: 0:11:58 lr: 0.000037 loss: 1.543793 (1.649997) time: 0.711637 data: 0.000166 max mem: 14338 Epoch: [15/30] [4050/5004] eta: 0:11:22 lr: 0.000037 loss: 1.542924 (1.650133) time: 0.717116 data: 0.000185 max mem: 14338 Epoch: [15/30] [4100/5004] eta: 0:10:46 lr: 0.000037 loss: 1.660500 (1.651100) time: 0.712500 data: 0.000223 max mem: 14338 Epoch: [15/30] [4150/5004] eta: 0:10:10 lr: 0.000037 loss: 1.570700 (1.650743) time: 0.717075 data: 0.000191 max mem: 14338 Epoch: [15/30] [4200/5004] eta: 0:09:35 lr: 0.000037 loss: 1.702520 (1.650601) time: 0.714020 data: 0.000228 max mem: 14338 Epoch: [15/30] [4250/5004] eta: 0:08:59 lr: 0.000037 loss: 1.569213 (1.650943) time: 0.711682 data: 0.000214 max mem: 14338 Epoch: [15/30] [4300/5004] eta: 0:08:23 lr: 0.000037 loss: 1.577110 (1.650846) time: 0.713044 data: 0.000221 max mem: 14338 Epoch: [15/30] [4350/5004] eta: 0:07:47 lr: 0.000037 loss: 1.666158 (1.650716) time: 0.711878 data: 0.000175 max mem: 14338 Epoch: [15/30] [4400/5004] eta: 0:07:12 lr: 0.000037 loss: 1.551794 (1.650285) time: 0.718295 data: 0.000161 max mem: 14338 Epoch: [15/30] [4450/5004] eta: 0:06:36 lr: 0.000037 loss: 1.548423 (1.650557) time: 0.714646 data: 0.000199 max mem: 14338 Epoch: [15/30] [4500/5004] eta: 0:06:00 lr: 0.000037 loss: 1.769777 (1.650957) time: 0.709495 data: 0.000232 max mem: 14338 Epoch: [15/30] [4550/5004] eta: 0:05:24 lr: 0.000037 loss: 1.706659 (1.651088) time: 0.709827 data: 0.000222 max mem: 14338 Epoch: [15/30] [4600/5004] eta: 0:04:48 lr: 0.000037 loss: 1.698283 (1.651185) time: 0.712364 data: 0.000214 max mem: 14338 Epoch: [15/30] [4650/5004] eta: 0:04:13 lr: 0.000037 loss: 1.694396 (1.651239) time: 0.715674 data: 0.000207 max mem: 14338 Epoch: [15/30] [4700/5004] eta: 0:03:37 lr: 0.000036 loss: 1.464135 (1.651303) time: 0.711091 data: 0.000158 max mem: 14338 Epoch: [15/30] [4750/5004] eta: 0:03:01 lr: 0.000036 loss: 1.609189 (1.652544) time: 0.715619 data: 0.000163 max mem: 14338 Epoch: [15/30] [4800/5004] eta: 0:02:25 lr: 0.000036 loss: 1.618013 (1.652651) time: 0.721615 data: 0.000192 max mem: 14338 Epoch: [15/30] [4850/5004] eta: 0:01:50 lr: 0.000036 loss: 1.694291 (1.652802) time: 0.720143 data: 0.000221 max mem: 14338 Epoch: [15/30] [4900/5004] eta: 0:01:14 lr: 0.000036 loss: 1.750037 (1.653497) time: 0.713416 data: 0.000232 max mem: 14338 Epoch: [15/30] [4950/5004] eta: 0:00:38 lr: 0.000036 loss: 1.722690 (1.653728) time: 0.708883 data: 0.000202 max mem: 14338 Epoch: [15/30] [5000/5004] eta: 0:00:02 lr: 0.000036 loss: 1.801891 (1.654342) time: 0.713044 data: 0.000833 max mem: 14338 Epoch: [15/30] [5003/5004] eta: 0:00:00 lr: 0.000036 loss: 1.680684 (1.654411) time: 0.710610 data: 0.000828 max mem: 14338 Epoch: [15/30] Total time: 0:59:39 (0.715230 s / it) Averaged stats: lr: 0.000036 loss: 1.680684 (1.652512) Test: [ 0/196] eta: 0:04:51 loss: 0.261856 (0.261856) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.484786 data: 1.076195 max mem: 14338 Test: [ 10/196] eta: 0:01:16 loss: 0.443556 (0.530092) acc1: 87.500000 (85.227273) acc5: 100.000000 (98.863636) time: 0.411012 data: 0.097968 max mem: 14338 Test: [ 20/196] eta: 0:01:02 loss: 0.514458 (0.530869) acc1: 87.500000 (85.119048) acc5: 100.000000 (98.214286) time: 0.296124 data: 0.000136 max mem: 14338 Test: [ 30/196] eta: 0:00:55 loss: 0.495548 (0.503096) acc1: 87.500000 (86.895161) acc5: 100.000000 (98.387097) time: 0.288898 data: 0.000134 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.399017 (0.508630) acc1: 87.500000 (86.585366) acc5: 100.000000 (98.170732) time: 0.287935 data: 0.000137 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.400785 (0.538871) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.671569) time: 0.287100 data: 0.000132 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.578647 (0.567407) acc1: 87.500000 (85.963115) acc5: 100.000000 (97.643443) time: 0.287949 data: 0.000142 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.651890 (0.586734) acc1: 81.250000 (85.563380) acc5: 100.000000 (97.711268) time: 0.287431 data: 0.000144 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.540377 (0.586663) acc1: 87.500000 (85.570988) acc5: 100.000000 (97.762346) time: 0.287812 data: 0.000137 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.540377 (0.607408) acc1: 87.500000 (85.302198) acc5: 100.000000 (97.458791) time: 0.289206 data: 0.000146 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.557677 (0.596930) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.648515) time: 0.288274 data: 0.000142 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.530467 (0.587107) acc1: 87.500000 (85.416667) acc5: 100.000000 (97.804054) time: 0.287291 data: 0.000137 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.515690 (0.584571) acc1: 87.500000 (85.433884) acc5: 100.000000 (97.778926) time: 0.286923 data: 0.000143 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.658383 (0.599815) acc1: 81.250000 (85.066794) acc5: 100.000000 (97.757634) time: 0.286119 data: 0.000136 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.670524 (0.598626) acc1: 81.250000 (85.195035) acc5: 100.000000 (97.739362) time: 0.286040 data: 0.000142 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.556432 (0.606239) acc1: 87.500000 (85.099338) acc5: 100.000000 (97.764901) time: 0.292934 data: 0.000147 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.556432 (0.611941) acc1: 81.250000 (85.015528) acc5: 100.000000 (97.787267) time: 0.293460 data: 0.000138 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.489160 (0.608677) acc1: 81.250000 (85.160819) acc5: 100.000000 (97.843567) time: 0.286928 data: 0.000141 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.467047 (0.609268) acc1: 87.500000 (85.082873) acc5: 100.000000 (97.859116) time: 0.285834 data: 0.000153 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.363710 (0.599209) acc1: 87.500000 (85.373037) acc5: 100.000000 (97.873037) time: 0.283039 data: 0.000114 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.573648 (0.609625) acc1: 87.500000 (85.248000) acc5: 100.000000 (97.792000) time: 0.273925 data: 0.000104 max mem: 14338 Test: Total time: 0:00:57 (0.294974 s / it) * Acc@1 85.122 Acc@5 97.296 loss 0.629 Max accuracy: 85.14% uploading checkpoint virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0015.pth to hdfs://harunava/user/guoyuanfan/HCSC/virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0015.pth Epoch: [16/30] [ 0/5004] eta: 2:36:34 lr: 0.000036 loss: 1.098880 (1.098880) time: 1.877317 data: 1.129951 max mem: 14338 Epoch: [16/30] [ 50/5004] eta: 1:00:48 lr: 0.000036 loss: 1.528346 (1.589907) time: 0.716007 data: 0.000209 max mem: 14338 Epoch: [16/30] [ 100/5004] eta: 0:59:21 lr: 0.000036 loss: 1.618277 (1.594778) time: 0.712334 data: 0.000216 max mem: 14338 Epoch: [16/30] [ 150/5004] eta: 0:58:24 lr: 0.000036 loss: 1.685268 (1.626477) time: 0.710070 data: 0.000218 max mem: 14338 Epoch: [16/30] [ 200/5004] eta: 0:57:41 lr: 0.000036 loss: 1.770798 (1.632766) time: 0.719632 data: 0.000208 max mem: 14338 Epoch: [16/30] [ 250/5004] eta: 0:57:00 lr: 0.000036 loss: 1.562667 (1.647087) time: 0.716424 data: 0.000189 max mem: 14338 Epoch: [16/30] [ 300/5004] eta: 0:56:20 lr: 0.000036 loss: 1.766582 (1.652240) time: 0.720724 data: 0.000176 max mem: 14338 Epoch: [16/30] [ 350/5004] eta: 0:55:41 lr: 0.000036 loss: 1.707393 (1.653535) time: 0.710139 data: 0.000194 max mem: 14338 Epoch: [16/30] [ 400/5004] eta: 0:55:06 lr: 0.000036 loss: 1.574230 (1.652953) time: 0.719997 data: 0.000232 max mem: 14338 Epoch: [16/30] [ 450/5004] eta: 0:54:28 lr: 0.000036 loss: 1.562868 (1.650551) time: 0.713945 data: 0.000242 max mem: 14338 Epoch: [16/30] [ 500/5004] eta: 0:53:50 lr: 0.000036 loss: 1.566493 (1.656919) time: 0.714099 data: 0.000214 max mem: 14338 Epoch: [16/30] [ 550/5004] eta: 0:53:14 lr: 0.000036 loss: 1.498424 (1.649193) time: 0.711482 data: 0.000215 max mem: 14338 Epoch: [16/30] [ 600/5004] eta: 0:52:37 lr: 0.000036 loss: 1.596139 (1.651536) time: 0.714511 data: 0.000226 max mem: 14338 Epoch: [16/30] [ 650/5004] eta: 0:52:00 lr: 0.000036 loss: 1.646474 (1.648750) time: 0.718933 data: 0.000187 max mem: 14338 Epoch: [16/30] [ 700/5004] eta: 0:51:23 lr: 0.000036 loss: 1.517656 (1.649284) time: 0.713689 data: 0.000160 max mem: 14338 Epoch: [16/30] [ 750/5004] eta: 0:50:47 lr: 0.000036 loss: 1.444242 (1.645393) time: 0.716003 data: 0.000221 max mem: 14338 Epoch: [16/30] [ 800/5004] eta: 0:50:10 lr: 0.000036 loss: 1.533223 (1.648465) time: 0.714784 data: 0.000213 max mem: 14338 Epoch: [16/30] [ 850/5004] eta: 0:49:33 lr: 0.000036 loss: 1.667142 (1.646904) time: 0.712461 data: 0.000222 max mem: 14338 Epoch: [16/30] [ 900/5004] eta: 0:48:57 lr: 0.000035 loss: 1.525644 (1.647200) time: 0.711357 data: 0.000214 max mem: 14338 Epoch: [16/30] [ 950/5004] eta: 0:48:21 lr: 0.000035 loss: 1.649113 (1.647201) time: 0.710580 data: 0.000230 max mem: 14338 Epoch: [16/30] [1000/5004] eta: 0:47:45 lr: 0.000035 loss: 1.572452 (1.644255) time: 0.715744 data: 0.000177 max mem: 14338 Epoch: [16/30] [1050/5004] eta: 0:47:09 lr: 0.000035 loss: 1.550451 (1.644929) time: 0.713560 data: 0.000220 max mem: 14338 Epoch: [16/30] [1100/5004] eta: 0:46:33 lr: 0.000035 loss: 1.659972 (1.645617) time: 0.709993 data: 0.000214 max mem: 14338 Epoch: [16/30] [1150/5004] eta: 0:45:57 lr: 0.000035 loss: 1.784838 (1.646889) time: 0.721733 data: 0.000239 max mem: 14338 Epoch: [16/30] [1200/5004] eta: 0:45:22 lr: 0.000035 loss: 1.589776 (1.645861) time: 0.719940 data: 0.000202 max mem: 14338 Epoch: [16/30] [1250/5004] eta: 0:44:46 lr: 0.000035 loss: 1.499437 (1.643383) time: 0.711878 data: 0.000213 max mem: 14338 Epoch: [16/30] [1300/5004] eta: 0:44:10 lr: 0.000035 loss: 1.535242 (1.645087) time: 0.709925 data: 0.000162 max mem: 14338 Epoch: [16/30] [1350/5004] eta: 0:43:34 lr: 0.000035 loss: 1.524518 (1.642568) time: 0.712617 data: 0.000164 max mem: 14338 Epoch: [16/30] [1400/5004] eta: 0:42:58 lr: 0.000035 loss: 1.610161 (1.644904) time: 0.715661 data: 0.000216 max mem: 14338 Epoch: [16/30] [1450/5004] eta: 0:42:22 lr: 0.000035 loss: 1.507581 (1.643530) time: 0.713521 data: 0.000214 max mem: 14338 Epoch: [16/30] [1500/5004] eta: 0:41:45 lr: 0.000035 loss: 1.592341 (1.643049) time: 0.709116 data: 0.000236 max mem: 14338 Epoch: [16/30] [1550/5004] eta: 0:41:10 lr: 0.000035 loss: 1.486595 (1.640910) time: 0.716058 data: 0.000190 max mem: 14338 Epoch: [16/30] [1600/5004] eta: 0:40:34 lr: 0.000035 loss: 1.495645 (1.642306) time: 0.721874 data: 0.000213 max mem: 14338 Epoch: [16/30] [1650/5004] eta: 0:39:58 lr: 0.000035 loss: 1.657988 (1.645137) time: 0.720015 data: 0.000176 max mem: 14338 Epoch: [16/30] [1700/5004] eta: 0:39:22 lr: 0.000035 loss: 1.494858 (1.645704) time: 0.717696 data: 0.000162 max mem: 14338 Epoch: [16/30] [1750/5004] eta: 0:38:46 lr: 0.000035 loss: 1.427815 (1.645296) time: 0.710908 data: 0.000213 max mem: 14338 Epoch: [16/30] [1800/5004] eta: 0:38:10 lr: 0.000035 loss: 1.588123 (1.645844) time: 0.711195 data: 0.000219 max mem: 14338 Epoch: [16/30] [1850/5004] eta: 0:37:34 lr: 0.000035 loss: 1.508290 (1.644125) time: 0.712227 data: 0.000208 max mem: 14338 Epoch: [16/30] [1900/5004] eta: 0:36:58 lr: 0.000035 loss: 1.531757 (1.644743) time: 0.709667 data: 0.000227 max mem: 14338 Epoch: [16/30] [1950/5004] eta: 0:36:22 lr: 0.000035 loss: 1.591898 (1.645891) time: 0.710968 data: 0.000226 max mem: 14338 Epoch: [16/30] [2000/5004] eta: 0:35:47 lr: 0.000035 loss: 1.608970 (1.647708) time: 0.713757 data: 0.000165 max mem: 14338 Epoch: [16/30] [2050/5004] eta: 0:35:11 lr: 0.000035 loss: 1.554020 (1.647480) time: 0.717663 data: 0.000169 max mem: 14338 Epoch: [16/30] [2100/5004] eta: 0:34:35 lr: 0.000034 loss: 1.586591 (1.647745) time: 0.715025 data: 0.000207 max mem: 14338 Epoch: [16/30] [2150/5004] eta: 0:34:00 lr: 0.000034 loss: 1.637913 (1.647239) time: 0.717695 data: 0.000224 max mem: 14338 Epoch: [16/30] [2200/5004] eta: 0:33:24 lr: 0.000034 loss: 1.598897 (1.646723) time: 0.713690 data: 0.000190 max mem: 14338 Epoch: [16/30] [2250/5004] eta: 0:32:48 lr: 0.000034 loss: 1.627574 (1.646393) time: 0.712100 data: 0.000221 max mem: 14338 Epoch: [16/30] [2300/5004] eta: 0:32:12 lr: 0.000034 loss: 1.631173 (1.645977) time: 0.712171 data: 0.000209 max mem: 14338 Epoch: [16/30] [2350/5004] eta: 0:31:36 lr: 0.000034 loss: 1.503651 (1.645879) time: 0.709185 data: 0.000162 max mem: 14338 Epoch: [16/30] [2400/5004] eta: 0:31:00 lr: 0.000034 loss: 1.481494 (1.646102) time: 0.711750 data: 0.000212 max mem: 14338 Epoch: [16/30] [2450/5004] eta: 0:30:25 lr: 0.000034 loss: 1.572997 (1.647171) time: 0.714059 data: 0.000224 max mem: 14338 Epoch: [16/30] [2500/5004] eta: 0:29:49 lr: 0.000034 loss: 1.410050 (1.645650) time: 0.710839 data: 0.000221 max mem: 14338 Epoch: [16/30] [2550/5004] eta: 0:29:13 lr: 0.000034 loss: 1.587036 (1.645037) time: 0.713924 data: 0.000227 max mem: 14338 Epoch: [16/30] [2600/5004] eta: 0:28:37 lr: 0.000034 loss: 1.554532 (1.643949) time: 0.719103 data: 0.000220 max mem: 14338 Epoch: [16/30] [2650/5004] eta: 0:28:01 lr: 0.000034 loss: 1.540383 (1.642507) time: 0.713563 data: 0.000160 max mem: 14338 Epoch: [16/30] [2700/5004] eta: 0:27:26 lr: 0.000034 loss: 1.528081 (1.643490) time: 0.710965 data: 0.000169 max mem: 14338 Epoch: [16/30] [2750/5004] eta: 0:26:50 lr: 0.000034 loss: 1.510945 (1.641287) time: 0.710633 data: 0.000212 max mem: 14338 Epoch: [16/30] [2800/5004] eta: 0:26:14 lr: 0.000034 loss: 1.657624 (1.641859) time: 0.711791 data: 0.000226 max mem: 14338 Epoch: [16/30] [2850/5004] eta: 0:25:38 lr: 0.000034 loss: 1.616153 (1.642103) time: 0.714740 data: 0.000188 max mem: 14338 Epoch: [16/30] [2900/5004] eta: 0:25:03 lr: 0.000034 loss: 1.554845 (1.641721) time: 0.713321 data: 0.000240 max mem: 14338 Epoch: [16/30] [2950/5004] eta: 0:24:27 lr: 0.000034 loss: 1.651332 (1.642952) time: 0.713592 data: 0.000236 max mem: 14338 Epoch: [16/30] [3000/5004] eta: 0:23:51 lr: 0.000034 loss: 1.405567 (1.642385) time: 0.721634 data: 0.000185 max mem: 14338 Epoch: [16/30] [3050/5004] eta: 0:23:16 lr: 0.000034 loss: 1.757337 (1.643765) time: 0.717947 data: 0.000165 max mem: 14338 Epoch: [16/30] [3100/5004] eta: 0:22:40 lr: 0.000034 loss: 1.639906 (1.644033) time: 0.713315 data: 0.000215 max mem: 14338 Epoch: [16/30] [3150/5004] eta: 0:22:04 lr: 0.000034 loss: 1.539464 (1.643829) time: 0.713832 data: 0.000223 max mem: 14338 Epoch: [16/30] [3200/5004] eta: 0:21:28 lr: 0.000034 loss: 1.525941 (1.642982) time: 0.713442 data: 0.000221 max mem: 14338 Epoch: [16/30] [3250/5004] eta: 0:20:53 lr: 0.000034 loss: 1.736175 (1.643953) time: 0.711796 data: 0.000224 max mem: 14338 Epoch: [16/30] [3300/5004] eta: 0:20:17 lr: 0.000034 loss: 1.533863 (1.644436) time: 0.708991 data: 0.000212 max mem: 14338 Epoch: [16/30] [3350/5004] eta: 0:19:41 lr: 0.000033 loss: 1.676792 (1.644976) time: 0.710939 data: 0.000166 max mem: 14338 Epoch: [16/30] [3400/5004] eta: 0:19:05 lr: 0.000033 loss: 1.745829 (1.645890) time: 0.712552 data: 0.000154 max mem: 14338 Epoch: [16/30] [3450/5004] eta: 0:18:30 lr: 0.000033 loss: 1.554372 (1.645056) time: 0.720744 data: 0.000224 max mem: 14338 Epoch: [16/30] [3500/5004] eta: 0:17:54 lr: 0.000033 loss: 1.547876 (1.645260) time: 0.717789 data: 0.000214 max mem: 14338 Epoch: [16/30] [3550/5004] eta: 0:17:18 lr: 0.000033 loss: 1.746473 (1.645584) time: 0.713163 data: 0.000218 max mem: 14338 Epoch: [16/30] [3600/5004] eta: 0:16:42 lr: 0.000033 loss: 1.659220 (1.645837) time: 0.717595 data: 0.000212 max mem: 14338 Epoch: [16/30] [3650/5004] eta: 0:16:07 lr: 0.000033 loss: 1.488999 (1.644412) time: 0.711429 data: 0.000207 max mem: 14338 Epoch: [16/30] [3700/5004] eta: 0:15:31 lr: 0.000033 loss: 1.654333 (1.644724) time: 0.711127 data: 0.000174 max mem: 14338 Epoch: [16/30] [3750/5004] eta: 0:14:55 lr: 0.000033 loss: 1.497150 (1.643844) time: 0.709963 data: 0.000233 max mem: 14338 Epoch: [16/30] [3800/5004] eta: 0:14:20 lr: 0.000033 loss: 1.628863 (1.645104) time: 0.713839 data: 0.000212 max mem: 14338 Epoch: [16/30] [3850/5004] eta: 0:13:44 lr: 0.000033 loss: 1.648721 (1.644543) time: 0.711543 data: 0.000209 max mem: 14338 Epoch: [16/30] [3900/5004] eta: 0:13:08 lr: 0.000033 loss: 1.552433 (1.644863) time: 0.711740 data: 0.000207 max mem: 14338 Epoch: [16/30] [3950/5004] eta: 0:12:32 lr: 0.000033 loss: 1.545022 (1.644693) time: 0.726439 data: 0.000211 max mem: 14338 Epoch: [16/30] [4000/5004] eta: 0:11:57 lr: 0.000033 loss: 1.491845 (1.644002) time: 0.715147 data: 0.000159 max mem: 14338 Epoch: [16/30] [4050/5004] eta: 0:11:21 lr: 0.000033 loss: 1.425332 (1.643560) time: 0.713309 data: 0.000161 max mem: 14338 Epoch: [16/30] [4100/5004] eta: 0:10:45 lr: 0.000033 loss: 1.643387 (1.643375) time: 0.714601 data: 0.000233 max mem: 14338 Epoch: [16/30] [4150/5004] eta: 0:10:10 lr: 0.000033 loss: 1.581467 (1.643427) time: 0.709709 data: 0.000193 max mem: 14338 Epoch: [16/30] [4200/5004] eta: 0:09:34 lr: 0.000033 loss: 1.508613 (1.643529) time: 0.713745 data: 0.000212 max mem: 14338 Epoch: [16/30] [4250/5004] eta: 0:08:58 lr: 0.000033 loss: 1.627374 (1.643568) time: 0.710854 data: 0.000229 max mem: 14338 Epoch: [16/30] [4300/5004] eta: 0:08:22 lr: 0.000033 loss: 1.519456 (1.643464) time: 0.712677 data: 0.000209 max mem: 14338 Epoch: [16/30] [4350/5004] eta: 0:07:47 lr: 0.000033 loss: 1.490634 (1.643169) time: 0.715330 data: 0.000170 max mem: 14338 Epoch: [16/30] [4400/5004] eta: 0:07:11 lr: 0.000033 loss: 1.595795 (1.643987) time: 0.724118 data: 0.000160 max mem: 14338 Epoch: [16/30] [4450/5004] eta: 0:06:35 lr: 0.000033 loss: 1.481961 (1.644099) time: 0.715500 data: 0.000198 max mem: 14338 Epoch: [16/30] [4500/5004] eta: 0:06:00 lr: 0.000033 loss: 1.581766 (1.644641) time: 0.715276 data: 0.000204 max mem: 14338 Epoch: [16/30] [4550/5004] eta: 0:05:24 lr: 0.000032 loss: 1.673853 (1.644545) time: 0.712766 data: 0.000234 max mem: 14338 Epoch: [16/30] [4600/5004] eta: 0:04:48 lr: 0.000032 loss: 1.570228 (1.644512) time: 0.711968 data: 0.000221 max mem: 14338 Epoch: [16/30] [4650/5004] eta: 0:04:12 lr: 0.000032 loss: 1.561867 (1.644402) time: 0.712157 data: 0.000216 max mem: 14338 Epoch: [16/30] [4700/5004] eta: 0:03:37 lr: 0.000032 loss: 1.678550 (1.644403) time: 0.710152 data: 0.000187 max mem: 14338 Epoch: [16/30] [4750/5004] eta: 0:03:01 lr: 0.000032 loss: 1.493011 (1.643846) time: 0.713266 data: 0.000168 max mem: 14338 Epoch: [16/30] [4800/5004] eta: 0:02:25 lr: 0.000032 loss: 1.734724 (1.644046) time: 0.712151 data: 0.000191 max mem: 14338 Epoch: [16/30] [4850/5004] eta: 0:01:50 lr: 0.000032 loss: 1.664808 (1.643533) time: 0.719233 data: 0.000225 max mem: 14338 Epoch: [16/30] [4900/5004] eta: 0:01:14 lr: 0.000032 loss: 1.585336 (1.643455) time: 0.718694 data: 0.000222 max mem: 14338 Epoch: [16/30] [4950/5004] eta: 0:00:38 lr: 0.000032 loss: 1.511282 (1.642407) time: 0.715150 data: 0.000211 max mem: 14338 Epoch: [16/30] [5000/5004] eta: 0:00:02 lr: 0.000032 loss: 1.559997 (1.641961) time: 0.712674 data: 0.000828 max mem: 14338 Epoch: [16/30] [5003/5004] eta: 0:00:00 lr: 0.000032 loss: 1.576165 (1.641986) time: 0.709924 data: 0.000823 max mem: 14338 Epoch: [16/30] Total time: 0:59:35 (0.714540 s / it) Averaged stats: lr: 0.000032 loss: 1.576165 (1.647564) Test: [ 0/196] eta: 0:05:04 loss: 0.279543 (0.279543) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.552709 data: 1.159472 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.435867 (0.548244) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.401498 data: 0.105532 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.577848 (0.547875) acc1: 87.500000 (85.714286) acc5: 100.000000 (98.214286) time: 0.286948 data: 0.000136 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.513350 (0.520411) acc1: 87.500000 (86.895161) acc5: 100.000000 (98.387097) time: 0.286962 data: 0.000127 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.423487 (0.526586) acc1: 87.500000 (86.737805) acc5: 100.000000 (98.018293) time: 0.287642 data: 0.000135 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.423487 (0.551071) acc1: 87.500000 (86.397059) acc5: 100.000000 (97.549020) time: 0.288089 data: 0.000150 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.613784 (0.578102) acc1: 87.500000 (85.860656) acc5: 93.750000 (97.438525) time: 0.287282 data: 0.000153 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.658664 (0.593821) acc1: 81.250000 (85.563380) acc5: 100.000000 (97.623239) time: 0.286706 data: 0.000147 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.511056 (0.592973) acc1: 87.500000 (85.648148) acc5: 100.000000 (97.685185) time: 0.285587 data: 0.000127 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.536191 (0.615194) acc1: 81.250000 (85.370879) acc5: 100.000000 (97.390110) time: 0.285723 data: 0.000136 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.558767 (0.603585) acc1: 81.250000 (85.457921) acc5: 100.000000 (97.586634) time: 0.286739 data: 0.000146 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.539591 (0.593274) acc1: 87.500000 (85.472973) acc5: 100.000000 (97.691441) time: 0.286806 data: 0.000143 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.478296 (0.589302) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.727273) time: 0.287175 data: 0.000151 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.638171 (0.605846) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.709924) time: 0.294532 data: 0.000144 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.702862 (0.603483) acc1: 81.250000 (85.328014) acc5: 100.000000 (97.695035) time: 0.294089 data: 0.000141 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.568714 (0.610198) acc1: 87.500000 (85.264901) acc5: 100.000000 (97.723510) time: 0.286448 data: 0.000134 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.505378 (0.615076) acc1: 81.250000 (85.131988) acc5: 100.000000 (97.748447) time: 0.286574 data: 0.000126 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.505378 (0.611516) acc1: 81.250000 (85.160819) acc5: 100.000000 (97.807018) time: 0.286763 data: 0.000148 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.479206 (0.612848) acc1: 81.250000 (84.979282) acc5: 100.000000 (97.824586) time: 0.286227 data: 0.000137 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.366947 (0.601119) acc1: 87.500000 (85.274869) acc5: 100.000000 (97.807592) time: 0.283928 data: 0.000096 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.460436 (0.611808) acc1: 87.500000 (85.216000) acc5: 100.000000 (97.728000) time: 0.274732 data: 0.000087 max mem: 14338 Test: Total time: 0:00:57 (0.294150 s / it) * Acc@1 85.096 Acc@5 97.282 loss 0.630 Max accuracy: 85.14% Epoch: [17/30] [ 0/5004] eta: 2:41:25 lr: 0.000032 loss: 1.481055 (1.481055) time: 1.935573 data: 1.206459 max mem: 14338 Epoch: [17/30] [ 50/5004] eta: 1:00:58 lr: 0.000032 loss: 1.443002 (1.572094) time: 0.719074 data: 0.000199 max mem: 14338 Epoch: [17/30] [ 100/5004] eta: 0:59:23 lr: 0.000032 loss: 1.582131 (1.616949) time: 0.715483 data: 0.000259 max mem: 14338 Epoch: [17/30] [ 150/5004] eta: 0:58:29 lr: 0.000032 loss: 1.818997 (1.618041) time: 0.720766 data: 0.000224 max mem: 14338 Epoch: [17/30] [ 200/5004] eta: 0:57:42 lr: 0.000032 loss: 1.444657 (1.624985) time: 0.716563 data: 0.000230 max mem: 14338 Epoch: [17/30] [ 250/5004] eta: 0:57:01 lr: 0.000032 loss: 1.574594 (1.625537) time: 0.718370 data: 0.000216 max mem: 14338 Epoch: [17/30] [ 300/5004] eta: 0:56:22 lr: 0.000032 loss: 1.552862 (1.631371) time: 0.716533 data: 0.000162 max mem: 14338 Epoch: [17/30] [ 350/5004] eta: 0:55:45 lr: 0.000032 loss: 1.596060 (1.640107) time: 0.714201 data: 0.000153 max mem: 14338 Epoch: [17/30] [ 400/5004] eta: 0:55:09 lr: 0.000032 loss: 1.564473 (1.639149) time: 0.717664 data: 0.000218 max mem: 14338 Epoch: [17/30] [ 450/5004] eta: 0:54:31 lr: 0.000032 loss: 1.472044 (1.638689) time: 0.712972 data: 0.000199 max mem: 14338 Epoch: [17/30] [ 500/5004] eta: 0:53:53 lr: 0.000032 loss: 1.693747 (1.644216) time: 0.713677 data: 0.000205 max mem: 14338 Epoch: [17/30] [ 550/5004] eta: 0:53:16 lr: 0.000032 loss: 1.712776 (1.641378) time: 0.718972 data: 0.000221 max mem: 14338 Epoch: [17/30] [ 600/5004] eta: 0:52:39 lr: 0.000032 loss: 1.544022 (1.638729) time: 0.717401 data: 0.000218 max mem: 14338 Epoch: [17/30] [ 650/5004] eta: 0:52:02 lr: 0.000032 loss: 1.573044 (1.635809) time: 0.716981 data: 0.000159 max mem: 14338 Epoch: [17/30] [ 700/5004] eta: 0:51:25 lr: 0.000032 loss: 1.573290 (1.633837) time: 0.714713 data: 0.000189 max mem: 14338 Epoch: [17/30] [ 750/5004] eta: 0:50:49 lr: 0.000032 loss: 1.667141 (1.631841) time: 0.709643 data: 0.000222 max mem: 14338 Epoch: [17/30] [ 800/5004] eta: 0:50:12 lr: 0.000031 loss: 1.532913 (1.631857) time: 0.712253 data: 0.000219 max mem: 14338 Epoch: [17/30] [ 850/5004] eta: 0:49:36 lr: 0.000031 loss: 1.667187 (1.631410) time: 0.715302 data: 0.000220 max mem: 14338 Epoch: [17/30] [ 900/5004] eta: 0:49:00 lr: 0.000031 loss: 1.472407 (1.627678) time: 0.711999 data: 0.000305 max mem: 14338 Epoch: [17/30] [ 950/5004] eta: 0:48:24 lr: 0.000031 loss: 1.567535 (1.626249) time: 0.711128 data: 0.000216 max mem: 14338 Epoch: [17/30] [1000/5004] eta: 0:47:48 lr: 0.000031 loss: 1.463826 (1.627948) time: 0.720019 data: 0.000165 max mem: 14338 Epoch: [17/30] [1050/5004] eta: 0:47:12 lr: 0.000031 loss: 1.744930 (1.633035) time: 0.721341 data: 0.000239 max mem: 14338 Epoch: [17/30] [1100/5004] eta: 0:46:35 lr: 0.000031 loss: 1.678471 (1.632086) time: 0.715945 data: 0.000240 max mem: 14338 Epoch: [17/30] [1150/5004] eta: 0:45:59 lr: 0.000031 loss: 1.429651 (1.630525) time: 0.712965 data: 0.000227 max mem: 14338 Epoch: [17/30] [1200/5004] eta: 0:45:23 lr: 0.000031 loss: 1.554913 (1.630320) time: 0.716281 data: 0.000223 max mem: 14338 Epoch: [17/30] [1250/5004] eta: 0:44:47 lr: 0.000031 loss: 1.652172 (1.632302) time: 0.712875 data: 0.000216 max mem: 14338 Epoch: [17/30] [1300/5004] eta: 0:44:11 lr: 0.000031 loss: 1.429685 (1.633948) time: 0.711275 data: 0.000174 max mem: 14338 Epoch: [17/30] [1350/5004] eta: 0:43:35 lr: 0.000031 loss: 1.614887 (1.635112) time: 0.710424 data: 0.000178 max mem: 14338 Epoch: [17/30] [1400/5004] eta: 0:42:59 lr: 0.000031 loss: 1.502262 (1.634337) time: 0.713368 data: 0.000224 max mem: 14338 Epoch: [17/30] [1450/5004] eta: 0:42:23 lr: 0.000031 loss: 1.478722 (1.635311) time: 0.720534 data: 0.000225 max mem: 14338 Epoch: [17/30] [1500/5004] eta: 0:41:47 lr: 0.000031 loss: 1.457772 (1.633219) time: 0.713633 data: 0.000204 max mem: 14338 Epoch: [17/30] [1550/5004] eta: 0:41:11 lr: 0.000031 loss: 1.581914 (1.633162) time: 0.718623 data: 0.000195 max mem: 14338 Epoch: [17/30] [1600/5004] eta: 0:40:35 lr: 0.000031 loss: 1.516365 (1.633343) time: 0.712705 data: 0.000222 max mem: 14338 Epoch: [17/30] [1650/5004] eta: 0:39:59 lr: 0.000031 loss: 1.593161 (1.632557) time: 0.711776 data: 0.000163 max mem: 14338 Epoch: [17/30] [1700/5004] eta: 0:39:23 lr: 0.000031 loss: 1.502291 (1.632268) time: 0.708399 data: 0.000166 max mem: 14338 Epoch: [17/30] [1750/5004] eta: 0:38:47 lr: 0.000031 loss: 1.554795 (1.631339) time: 0.711937 data: 0.000211 max mem: 14338 Epoch: [17/30] [1800/5004] eta: 0:38:11 lr: 0.000031 loss: 1.609442 (1.632143) time: 0.715467 data: 0.000226 max mem: 14338 Epoch: [17/30] [1850/5004] eta: 0:37:35 lr: 0.000031 loss: 1.624514 (1.632068) time: 0.713624 data: 0.000241 max mem: 14338 Epoch: [17/30] [1900/5004] eta: 0:36:59 lr: 0.000031 loss: 1.489522 (1.632875) time: 0.713028 data: 0.000236 max mem: 14338 Epoch: [17/30] [1950/5004] eta: 0:36:24 lr: 0.000031 loss: 1.616620 (1.633542) time: 0.717035 data: 0.000214 max mem: 14338 Epoch: [17/30] [2000/5004] eta: 0:35:48 lr: 0.000031 loss: 1.519856 (1.632476) time: 0.718462 data: 0.000157 max mem: 14338 Epoch: [17/30] [2050/5004] eta: 0:35:12 lr: 0.000030 loss: 1.547691 (1.632826) time: 0.717762 data: 0.000158 max mem: 14338 Epoch: [17/30] [2100/5004] eta: 0:34:36 lr: 0.000030 loss: 1.674151 (1.633972) time: 0.715653 data: 0.000206 max mem: 14338 Epoch: [17/30] [2150/5004] eta: 0:34:01 lr: 0.000030 loss: 1.713887 (1.635639) time: 0.716746 data: 0.000236 max mem: 14338 Epoch: [17/30] [2200/5004] eta: 0:33:25 lr: 0.000030 loss: 1.668383 (1.636945) time: 0.713498 data: 0.000198 max mem: 14338 Epoch: [17/30] [2250/5004] eta: 0:32:49 lr: 0.000030 loss: 1.559199 (1.637593) time: 0.714966 data: 0.000206 max mem: 14338 Epoch: [17/30] [2300/5004] eta: 0:32:14 lr: 0.000030 loss: 1.791328 (1.638838) time: 0.711309 data: 0.000225 max mem: 14338 Epoch: [17/30] [2350/5004] eta: 0:31:38 lr: 0.000030 loss: 1.547347 (1.639032) time: 0.708746 data: 0.000174 max mem: 14338 Epoch: [17/30] [2400/5004] eta: 0:31:02 lr: 0.000030 loss: 1.719987 (1.640733) time: 0.720394 data: 0.000221 max mem: 14338 Epoch: [17/30] [2450/5004] eta: 0:30:26 lr: 0.000030 loss: 1.476955 (1.639707) time: 0.725238 data: 0.000232 max mem: 14338 Epoch: [17/30] [2500/5004] eta: 0:29:50 lr: 0.000030 loss: 1.444068 (1.638620) time: 0.715642 data: 0.000207 max mem: 14338 Epoch: [17/30] [2550/5004] eta: 0:29:15 lr: 0.000030 loss: 1.698409 (1.639151) time: 0.712729 data: 0.000200 max mem: 14338 Epoch: [17/30] [2600/5004] eta: 0:28:39 lr: 0.000030 loss: 1.549187 (1.637982) time: 0.711789 data: 0.000209 max mem: 14338 Epoch: [17/30] [2650/5004] eta: 0:28:03 lr: 0.000030 loss: 1.534497 (1.637943) time: 0.714950 data: 0.000160 max mem: 14338 Epoch: [17/30] [2700/5004] eta: 0:27:27 lr: 0.000030 loss: 1.610782 (1.638041) time: 0.714058 data: 0.000168 max mem: 14338 Epoch: [17/30] [2750/5004] eta: 0:26:52 lr: 0.000030 loss: 1.467757 (1.637985) time: 0.717025 data: 0.000216 max mem: 14338 Epoch: [17/30] [2800/5004] eta: 0:26:16 lr: 0.000030 loss: 1.578565 (1.638484) time: 0.712165 data: 0.000225 max mem: 14338 Epoch: [17/30] [2850/5004] eta: 0:25:40 lr: 0.000030 loss: 1.421736 (1.638201) time: 0.720002 data: 0.000186 max mem: 14338 Epoch: [17/30] [2900/5004] eta: 0:25:04 lr: 0.000030 loss: 1.408987 (1.637712) time: 0.718353 data: 0.000216 max mem: 14338 Epoch: [17/30] [2950/5004] eta: 0:24:29 lr: 0.000030 loss: 1.536036 (1.638634) time: 0.714284 data: 0.000234 max mem: 14338 Epoch: [17/30] [3000/5004] eta: 0:23:53 lr: 0.000030 loss: 1.596214 (1.638913) time: 0.715604 data: 0.000163 max mem: 14338 Epoch: [17/30] [3050/5004] eta: 0:23:17 lr: 0.000030 loss: 1.606188 (1.639217) time: 0.715226 data: 0.000155 max mem: 14338 Epoch: [17/30] [3100/5004] eta: 0:22:41 lr: 0.000030 loss: 1.626519 (1.640070) time: 0.712465 data: 0.000220 max mem: 14338 Epoch: [17/30] [3150/5004] eta: 0:22:05 lr: 0.000030 loss: 1.467721 (1.638574) time: 0.713930 data: 0.000215 max mem: 14338 Epoch: [17/30] [3200/5004] eta: 0:21:30 lr: 0.000030 loss: 1.687152 (1.639178) time: 0.712624 data: 0.000213 max mem: 14338 Epoch: [17/30] [3250/5004] eta: 0:20:54 lr: 0.000030 loss: 1.554401 (1.639604) time: 0.715778 data: 0.000208 max mem: 14338 Epoch: [17/30] [3300/5004] eta: 0:20:18 lr: 0.000029 loss: 1.551325 (1.639595) time: 0.717575 data: 0.000213 max mem: 14338 Epoch: [17/30] [3350/5004] eta: 0:19:42 lr: 0.000029 loss: 1.579709 (1.640413) time: 0.711366 data: 0.000176 max mem: 14338 Epoch: [17/30] [3400/5004] eta: 0:19:07 lr: 0.000029 loss: 1.524267 (1.640017) time: 0.716354 data: 0.000185 max mem: 14338 Epoch: [17/30] [3450/5004] eta: 0:18:31 lr: 0.000029 loss: 1.555745 (1.639787) time: 0.719053 data: 0.000215 max mem: 14338 Epoch: [17/30] [3500/5004] eta: 0:17:55 lr: 0.000029 loss: 1.545492 (1.639447) time: 0.713096 data: 0.000219 max mem: 14338 Epoch: [17/30] [3550/5004] eta: 0:17:19 lr: 0.000029 loss: 1.704940 (1.640577) time: 0.709109 data: 0.000222 max mem: 14338 Epoch: [17/30] [3600/5004] eta: 0:16:43 lr: 0.000029 loss: 1.537592 (1.640387) time: 0.713953 data: 0.000206 max mem: 14338 Epoch: [17/30] [3650/5004] eta: 0:16:08 lr: 0.000029 loss: 1.557639 (1.640327) time: 0.713450 data: 0.000213 max mem: 14338 Epoch: [17/30] [3700/5004] eta: 0:15:32 lr: 0.000029 loss: 1.610528 (1.641452) time: 0.711319 data: 0.000175 max mem: 14338 Epoch: [17/30] [3750/5004] eta: 0:14:56 lr: 0.000029 loss: 1.560292 (1.640624) time: 0.709686 data: 0.000204 max mem: 14338 Epoch: [17/30] [3800/5004] eta: 0:14:20 lr: 0.000029 loss: 1.533543 (1.640008) time: 0.718216 data: 0.000205 max mem: 14338 Epoch: [17/30] [3850/5004] eta: 0:13:45 lr: 0.000029 loss: 1.503973 (1.639245) time: 0.719568 data: 0.000214 max mem: 14338 Epoch: [17/30] [3900/5004] eta: 0:13:09 lr: 0.000029 loss: 1.525769 (1.638832) time: 0.718009 data: 0.000234 max mem: 14338 Epoch: [17/30] [3950/5004] eta: 0:12:33 lr: 0.000029 loss: 1.687073 (1.639347) time: 0.718401 data: 0.000206 max mem: 14338 Epoch: [17/30] [4000/5004] eta: 0:11:57 lr: 0.000029 loss: 1.616915 (1.639123) time: 0.716041 data: 0.000167 max mem: 14338 Epoch: [17/30] [4050/5004] eta: 0:11:22 lr: 0.000029 loss: 1.723311 (1.640140) time: 0.714313 data: 0.000169 max mem: 14338 Epoch: [17/30] [4100/5004] eta: 0:10:46 lr: 0.000029 loss: 1.604059 (1.639498) time: 0.710175 data: 0.000214 max mem: 14338 Epoch: [17/30] [4150/5004] eta: 0:10:10 lr: 0.000029 loss: 1.536100 (1.638788) time: 0.708814 data: 0.000199 max mem: 14338 Epoch: [17/30] [4200/5004] eta: 0:09:34 lr: 0.000029 loss: 1.432306 (1.638991) time: 0.713593 data: 0.000220 max mem: 14338 Epoch: [17/30] [4250/5004] eta: 0:08:59 lr: 0.000029 loss: 1.602169 (1.638884) time: 0.713822 data: 0.000219 max mem: 14338 Epoch: [17/30] [4300/5004] eta: 0:08:23 lr: 0.000029 loss: 1.565135 (1.638520) time: 0.715753 data: 0.000216 max mem: 14338 Epoch: [17/30] [4350/5004] eta: 0:07:47 lr: 0.000029 loss: 1.604349 (1.638321) time: 0.717514 data: 0.000185 max mem: 14338 Epoch: [17/30] [4400/5004] eta: 0:07:11 lr: 0.000029 loss: 1.609867 (1.638642) time: 0.717307 data: 0.000161 max mem: 14338 Epoch: [17/30] [4450/5004] eta: 0:06:36 lr: 0.000029 loss: 1.692200 (1.638684) time: 0.714213 data: 0.000217 max mem: 14338 Epoch: [17/30] [4500/5004] eta: 0:06:00 lr: 0.000029 loss: 1.579968 (1.638379) time: 0.709894 data: 0.000216 max mem: 14338 Epoch: [17/30] [4550/5004] eta: 0:05:24 lr: 0.000028 loss: 1.612967 (1.638630) time: 0.713840 data: 0.000217 max mem: 14338 Epoch: [17/30] [4600/5004] eta: 0:04:48 lr: 0.000028 loss: 1.510667 (1.638475) time: 0.713926 data: 0.000214 max mem: 14338 Epoch: [17/30] [4650/5004] eta: 0:04:13 lr: 0.000028 loss: 1.498674 (1.638150) time: 0.720708 data: 0.000219 max mem: 14338 Epoch: [17/30] [4700/5004] eta: 0:03:37 lr: 0.000028 loss: 1.687593 (1.639046) time: 0.710014 data: 0.000174 max mem: 14338 Epoch: [17/30] [4750/5004] eta: 0:03:01 lr: 0.000028 loss: 1.747919 (1.640049) time: 0.713192 data: 0.000175 max mem: 14338 Epoch: [17/30] [4800/5004] eta: 0:02:25 lr: 0.000028 loss: 1.639039 (1.639520) time: 0.716541 data: 0.000190 max mem: 14338 Epoch: [17/30] [4850/5004] eta: 0:01:50 lr: 0.000028 loss: 1.578186 (1.638966) time: 0.722281 data: 0.000211 max mem: 14338 Epoch: [17/30] [4900/5004] eta: 0:01:14 lr: 0.000028 loss: 1.747254 (1.639164) time: 0.708691 data: 0.000211 max mem: 14338 Epoch: [17/30] [4950/5004] eta: 0:00:38 lr: 0.000028 loss: 1.492238 (1.639166) time: 0.710714 data: 0.000229 max mem: 14338 Epoch: [17/30] [5000/5004] eta: 0:00:02 lr: 0.000028 loss: 1.502274 (1.639431) time: 0.708484 data: 0.000841 max mem: 14338 Epoch: [17/30] [5003/5004] eta: 0:00:00 lr: 0.000028 loss: 1.463324 (1.639402) time: 0.705461 data: 0.000830 max mem: 14338 Epoch: [17/30] Total time: 0:59:37 (0.714956 s / it) Averaged stats: lr: 0.000028 loss: 1.463324 (1.645057) Test: [ 0/196] eta: 0:04:51 loss: 0.285137 (0.285137) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.486812 data: 1.099473 max mem: 14338 Test: [ 10/196] eta: 0:01:13 loss: 0.449142 (0.537108) acc1: 87.500000 (85.227273) acc5: 100.000000 (98.863636) time: 0.396199 data: 0.100071 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.579093 (0.544826) acc1: 87.500000 (85.416667) acc5: 100.000000 (98.214286) time: 0.291045 data: 0.000116 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.536318 (0.521023) acc1: 87.500000 (86.895161) acc5: 100.000000 (98.387097) time: 0.290892 data: 0.000113 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.437328 (0.528834) acc1: 87.500000 (86.737805) acc5: 100.000000 (98.018293) time: 0.286346 data: 0.000128 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.442387 (0.557214) acc1: 87.500000 (86.764706) acc5: 93.750000 (97.426471) time: 0.286640 data: 0.000122 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.575664 (0.582204) acc1: 87.500000 (86.168033) acc5: 93.750000 (97.336066) time: 0.286512 data: 0.000127 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.659402 (0.599451) acc1: 81.250000 (85.827465) acc5: 100.000000 (97.359155) time: 0.285889 data: 0.000131 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.513357 (0.598376) acc1: 87.500000 (85.802469) acc5: 100.000000 (97.453704) time: 0.286657 data: 0.000114 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.513357 (0.619467) acc1: 81.250000 (85.302198) acc5: 100.000000 (97.184066) time: 0.287065 data: 0.000127 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.573491 (0.609132) acc1: 81.250000 (85.457921) acc5: 100.000000 (97.400990) time: 0.286337 data: 0.000130 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.529078 (0.597598) acc1: 87.500000 (85.529279) acc5: 100.000000 (97.466216) time: 0.286487 data: 0.000119 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.484966 (0.593432) acc1: 87.500000 (85.537190) acc5: 100.000000 (97.520661) time: 0.286554 data: 0.000127 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.640052 (0.609047) acc1: 87.500000 (85.209924) acc5: 100.000000 (97.471374) time: 0.286453 data: 0.000127 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.669456 (0.606856) acc1: 81.250000 (85.283688) acc5: 100.000000 (97.473404) time: 0.286570 data: 0.000132 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.503701 (0.615033) acc1: 87.500000 (85.182119) acc5: 100.000000 (97.475166) time: 0.286602 data: 0.000127 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.502985 (0.619054) acc1: 81.250000 (85.170807) acc5: 100.000000 (97.476708) time: 0.293192 data: 0.000135 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.495453 (0.615158) acc1: 81.250000 (85.160819) acc5: 100.000000 (97.551170) time: 0.293279 data: 0.000145 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.461151 (0.615382) acc1: 81.250000 (85.048343) acc5: 100.000000 (97.582873) time: 0.286968 data: 0.000143 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.349967 (0.603861) acc1: 87.500000 (85.274869) acc5: 100.000000 (97.611257) time: 0.284413 data: 0.000116 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.507110 (0.615181) acc1: 87.500000 (85.184000) acc5: 100.000000 (97.536000) time: 0.274678 data: 0.000103 max mem: 14338 Test: Total time: 0:00:57 (0.293926 s / it) * Acc@1 85.052 Acc@5 97.290 loss 0.632 Max accuracy: 85.14% Epoch: [18/30] [ 0/5004] eta: 2:40:38 lr: 0.000028 loss: 2.226136 (2.226136) time: 1.926180 data: 1.179346 max mem: 14338 Epoch: [18/30] [ 50/5004] eta: 1:00:58 lr: 0.000028 loss: 1.548257 (1.633453) time: 0.712806 data: 0.000184 max mem: 14338 Epoch: [18/30] [ 100/5004] eta: 0:59:20 lr: 0.000028 loss: 1.497287 (1.624761) time: 0.711046 data: 0.000213 max mem: 14338 Epoch: [18/30] [ 150/5004] eta: 0:58:26 lr: 0.000028 loss: 1.587249 (1.596960) time: 0.710353 data: 0.000200 max mem: 14338 Epoch: [18/30] [ 200/5004] eta: 0:57:42 lr: 0.000028 loss: 1.521321 (1.594953) time: 0.717342 data: 0.000195 max mem: 14338 Epoch: [18/30] [ 250/5004] eta: 0:57:06 lr: 0.000028 loss: 1.820512 (1.616334) time: 0.723604 data: 0.000186 max mem: 14338 Epoch: [18/30] [ 300/5004] eta: 0:56:26 lr: 0.000028 loss: 1.724183 (1.623053) time: 0.719308 data: 0.000172 max mem: 14338 Epoch: [18/30] [ 350/5004] eta: 0:55:46 lr: 0.000028 loss: 1.582653 (1.618156) time: 0.714709 data: 0.000184 max mem: 14338 Epoch: [18/30] [ 400/5004] eta: 0:55:07 lr: 0.000028 loss: 1.477416 (1.619381) time: 0.714799 data: 0.000209 max mem: 14338 Epoch: [18/30] [ 450/5004] eta: 0:54:28 lr: 0.000028 loss: 1.606032 (1.620474) time: 0.712498 data: 0.000205 max mem: 14338 Epoch: [18/30] [ 500/5004] eta: 0:53:52 lr: 0.000028 loss: 1.529052 (1.620523) time: 0.714537 data: 0.000229 max mem: 14338 Epoch: [18/30] [ 550/5004] eta: 0:53:15 lr: 0.000028 loss: 1.663415 (1.623024) time: 0.709865 data: 0.000207 max mem: 14338 Epoch: [18/30] [ 600/5004] eta: 0:52:37 lr: 0.000028 loss: 1.815927 (1.630667) time: 0.713232 data: 0.000214 max mem: 14338 Epoch: [18/30] [ 650/5004] eta: 0:52:00 lr: 0.000028 loss: 1.610892 (1.632782) time: 0.715669 data: 0.000159 max mem: 14338 Epoch: [18/30] [ 700/5004] eta: 0:51:25 lr: 0.000028 loss: 1.624608 (1.634087) time: 0.717478 data: 0.000171 max mem: 14338 Epoch: [18/30] [ 750/5004] eta: 0:50:48 lr: 0.000028 loss: 1.587761 (1.636531) time: 0.711972 data: 0.000223 max mem: 14338 Epoch: [18/30] [ 800/5004] eta: 0:50:11 lr: 0.000027 loss: 1.600172 (1.637796) time: 0.717992 data: 0.000216 max mem: 14338 Epoch: [18/30] [ 850/5004] eta: 0:49:35 lr: 0.000027 loss: 1.445887 (1.636097) time: 0.720907 data: 0.000216 max mem: 14338 Epoch: [18/30] [ 900/5004] eta: 0:49:00 lr: 0.000027 loss: 1.523986 (1.636808) time: 0.717178 data: 0.000203 max mem: 14338 Epoch: [18/30] [ 950/5004] eta: 0:48:24 lr: 0.000027 loss: 1.648120 (1.635572) time: 0.714254 data: 0.000225 max mem: 14338 Epoch: [18/30] [1000/5004] eta: 0:47:48 lr: 0.000027 loss: 1.589348 (1.634313) time: 0.712001 data: 0.000175 max mem: 14338 Epoch: [18/30] [1050/5004] eta: 0:47:12 lr: 0.000027 loss: 1.482164 (1.632480) time: 0.714348 data: 0.000218 max mem: 14338 Epoch: [18/30] [1100/5004] eta: 0:46:36 lr: 0.000027 loss: 1.481503 (1.633772) time: 0.711897 data: 0.000226 max mem: 14338 Epoch: [18/30] [1150/5004] eta: 0:46:00 lr: 0.000027 loss: 1.512143 (1.633525) time: 0.711277 data: 0.000233 max mem: 14338 Epoch: [18/30] [1200/5004] eta: 0:45:24 lr: 0.000027 loss: 1.566711 (1.634098) time: 0.719188 data: 0.000228 max mem: 14338 Epoch: [18/30] [1250/5004] eta: 0:44:48 lr: 0.000027 loss: 1.615671 (1.631275) time: 0.718531 data: 0.000220 max mem: 14338 Epoch: [18/30] [1300/5004] eta: 0:44:12 lr: 0.000027 loss: 1.569777 (1.630482) time: 0.715918 data: 0.000167 max mem: 14338 Epoch: [18/30] [1350/5004] eta: 0:43:36 lr: 0.000027 loss: 1.651037 (1.630593) time: 0.710900 data: 0.000163 max mem: 14338 Epoch: [18/30] [1400/5004] eta: 0:43:00 lr: 0.000027 loss: 1.723508 (1.631640) time: 0.717563 data: 0.000218 max mem: 14338 Epoch: [18/30] [1450/5004] eta: 0:42:24 lr: 0.000027 loss: 1.480111 (1.630539) time: 0.715650 data: 0.000227 max mem: 14338 Epoch: [18/30] [1500/5004] eta: 0:41:49 lr: 0.000027 loss: 1.651357 (1.632005) time: 0.715484 data: 0.000224 max mem: 14338 Epoch: [18/30] [1550/5004] eta: 0:41:12 lr: 0.000027 loss: 1.535190 (1.632680) time: 0.712844 data: 0.000190 max mem: 14338 Epoch: [18/30] [1600/5004] eta: 0:40:37 lr: 0.000027 loss: 1.635288 (1.632557) time: 0.718047 data: 0.000231 max mem: 14338 Epoch: [18/30] [1650/5004] eta: 0:40:01 lr: 0.000027 loss: 1.568454 (1.631378) time: 0.725351 data: 0.000173 max mem: 14338 Epoch: [18/30] [1700/5004] eta: 0:39:25 lr: 0.000027 loss: 1.514886 (1.630732) time: 0.717795 data: 0.000162 max mem: 14338 Epoch: [18/30] [1750/5004] eta: 0:38:49 lr: 0.000027 loss: 1.674574 (1.631564) time: 0.713519 data: 0.000221 max mem: 14338 Epoch: [18/30] [1800/5004] eta: 0:38:13 lr: 0.000027 loss: 1.519853 (1.630857) time: 0.716139 data: 0.000209 max mem: 14338 Epoch: [18/30] [1850/5004] eta: 0:37:37 lr: 0.000027 loss: 1.604558 (1.628074) time: 0.711159 data: 0.000214 max mem: 14338 Epoch: [18/30] [1900/5004] eta: 0:37:01 lr: 0.000027 loss: 1.627426 (1.628657) time: 0.715174 data: 0.000210 max mem: 14338 Epoch: [18/30] [1950/5004] eta: 0:36:25 lr: 0.000027 loss: 1.539217 (1.629150) time: 0.708469 data: 0.000211 max mem: 14338 Epoch: [18/30] [2000/5004] eta: 0:35:50 lr: 0.000027 loss: 1.565972 (1.630694) time: 0.714512 data: 0.000161 max mem: 14338 Epoch: [18/30] [2050/5004] eta: 0:35:13 lr: 0.000027 loss: 1.509476 (1.630406) time: 0.713069 data: 0.000153 max mem: 14338 Epoch: [18/30] [2100/5004] eta: 0:34:38 lr: 0.000026 loss: 1.564150 (1.630262) time: 0.715755 data: 0.000241 max mem: 14338 Epoch: [18/30] [2150/5004] eta: 0:34:02 lr: 0.000026 loss: 1.559449 (1.630483) time: 0.720874 data: 0.000230 max mem: 14338 Epoch: [18/30] [2200/5004] eta: 0:33:26 lr: 0.000026 loss: 1.656198 (1.631647) time: 0.715532 data: 0.000192 max mem: 14338 Epoch: [18/30] [2250/5004] eta: 0:32:50 lr: 0.000026 loss: 1.547181 (1.630356) time: 0.718687 data: 0.000215 max mem: 14338 Epoch: [18/30] [2300/5004] eta: 0:32:14 lr: 0.000026 loss: 1.681793 (1.630153) time: 0.709031 data: 0.000225 max mem: 14338 Epoch: [18/30] [2350/5004] eta: 0:31:38 lr: 0.000026 loss: 1.609723 (1.629035) time: 0.709318 data: 0.000164 max mem: 14338 Epoch: [18/30] [2400/5004] eta: 0:31:02 lr: 0.000026 loss: 1.726661 (1.630131) time: 0.714913 data: 0.000215 max mem: 14338 Epoch: [18/30] [2450/5004] eta: 0:30:26 lr: 0.000026 loss: 1.564614 (1.630925) time: 0.710736 data: 0.000210 max mem: 14338 Epoch: [18/30] [2500/5004] eta: 0:29:50 lr: 0.000026 loss: 1.603855 (1.630325) time: 0.712143 data: 0.000222 max mem: 14338 Epoch: [18/30] [2550/5004] eta: 0:29:15 lr: 0.000026 loss: 1.719365 (1.630922) time: 0.713892 data: 0.000224 max mem: 14338 Epoch: [18/30] [2600/5004] eta: 0:28:39 lr: 0.000026 loss: 1.535348 (1.631959) time: 0.717659 data: 0.000202 max mem: 14338 Epoch: [18/30] [2650/5004] eta: 0:28:03 lr: 0.000026 loss: 1.511909 (1.632483) time: 0.717812 data: 0.000174 max mem: 14338 Epoch: [18/30] [2700/5004] eta: 0:27:27 lr: 0.000026 loss: 1.413300 (1.631385) time: 0.713413 data: 0.000171 max mem: 14338 Epoch: [18/30] [2750/5004] eta: 0:26:51 lr: 0.000026 loss: 1.554294 (1.632100) time: 0.711397 data: 0.000219 max mem: 14338 Epoch: [18/30] [2800/5004] eta: 0:26:15 lr: 0.000026 loss: 1.551217 (1.630744) time: 0.711561 data: 0.000212 max mem: 14338 Epoch: [18/30] [2850/5004] eta: 0:25:40 lr: 0.000026 loss: 1.674935 (1.630881) time: 0.715940 data: 0.000185 max mem: 14338 Epoch: [18/30] [2900/5004] eta: 0:25:04 lr: 0.000026 loss: 1.763829 (1.630519) time: 0.708930 data: 0.000216 max mem: 14338 Epoch: [18/30] [2950/5004] eta: 0:24:28 lr: 0.000026 loss: 1.589357 (1.630078) time: 0.710267 data: 0.000216 max mem: 14338 Epoch: [18/30] [3000/5004] eta: 0:23:52 lr: 0.000026 loss: 1.593639 (1.630148) time: 0.712987 data: 0.000163 max mem: 14338 Epoch: [18/30] [3050/5004] eta: 0:23:17 lr: 0.000026 loss: 1.639228 (1.630213) time: 0.715870 data: 0.000167 max mem: 14338 Epoch: [18/30] [3100/5004] eta: 0:22:41 lr: 0.000026 loss: 1.558375 (1.630371) time: 0.719709 data: 0.000210 max mem: 14338 Epoch: [18/30] [3150/5004] eta: 0:22:05 lr: 0.000026 loss: 1.668252 (1.630575) time: 0.713760 data: 0.000241 max mem: 14338 Epoch: [18/30] [3200/5004] eta: 0:21:29 lr: 0.000026 loss: 1.522414 (1.629475) time: 0.711383 data: 0.000222 max mem: 14338 Epoch: [18/30] [3250/5004] eta: 0:20:54 lr: 0.000026 loss: 1.605571 (1.630101) time: 0.715268 data: 0.000214 max mem: 14338 Epoch: [18/30] [3300/5004] eta: 0:20:18 lr: 0.000026 loss: 1.614290 (1.630379) time: 0.712281 data: 0.000218 max mem: 14338 Epoch: [18/30] [3350/5004] eta: 0:19:42 lr: 0.000026 loss: 1.793787 (1.631358) time: 0.709627 data: 0.000165 max mem: 14338 Epoch: [18/30] [3400/5004] eta: 0:19:06 lr: 0.000025 loss: 1.575620 (1.631100) time: 0.712452 data: 0.000166 max mem: 14338 Epoch: [18/30] [3450/5004] eta: 0:18:30 lr: 0.000025 loss: 1.611177 (1.630975) time: 0.714430 data: 0.000233 max mem: 14338 Epoch: [18/30] [3500/5004] eta: 0:17:55 lr: 0.000025 loss: 1.646447 (1.630976) time: 0.712696 data: 0.000201 max mem: 14338 Epoch: [18/30] [3550/5004] eta: 0:17:19 lr: 0.000025 loss: 1.494984 (1.631475) time: 0.710710 data: 0.000228 max mem: 14338 Epoch: [18/30] [3600/5004] eta: 0:16:43 lr: 0.000025 loss: 1.515247 (1.632037) time: 0.720233 data: 0.000224 max mem: 14338 Epoch: [18/30] [3650/5004] eta: 0:16:07 lr: 0.000025 loss: 1.408892 (1.631580) time: 0.717458 data: 0.000214 max mem: 14338 Epoch: [18/30] [3700/5004] eta: 0:15:32 lr: 0.000025 loss: 1.709122 (1.631976) time: 0.710815 data: 0.000171 max mem: 14338 Epoch: [18/30] [3750/5004] eta: 0:14:56 lr: 0.000025 loss: 1.650715 (1.631656) time: 0.712210 data: 0.000240 max mem: 14338 Epoch: [18/30] [3800/5004] eta: 0:14:20 lr: 0.000025 loss: 1.478886 (1.631816) time: 0.711846 data: 0.000238 max mem: 14338 Epoch: [18/30] [3850/5004] eta: 0:13:44 lr: 0.000025 loss: 1.691270 (1.632733) time: 0.715087 data: 0.000217 max mem: 14338 Epoch: [18/30] [3900/5004] eta: 0:13:09 lr: 0.000025 loss: 1.456919 (1.631774) time: 0.709448 data: 0.000226 max mem: 14338 Epoch: [18/30] [3950/5004] eta: 0:12:33 lr: 0.000025 loss: 1.722786 (1.632583) time: 0.711698 data: 0.000239 max mem: 14338 Epoch: [18/30] [4000/5004] eta: 0:11:57 lr: 0.000025 loss: 1.507535 (1.633341) time: 0.721634 data: 0.000165 max mem: 14338 Epoch: [18/30] [4050/5004] eta: 0:11:21 lr: 0.000025 loss: 1.698242 (1.633502) time: 0.715639 data: 0.000176 max mem: 14338 Epoch: [18/30] [4100/5004] eta: 0:10:46 lr: 0.000025 loss: 1.514578 (1.633473) time: 0.709587 data: 0.000212 max mem: 14338 Epoch: [18/30] [4150/5004] eta: 0:10:10 lr: 0.000025 loss: 1.688913 (1.633398) time: 0.709755 data: 0.000190 max mem: 14338 Epoch: [18/30] [4200/5004] eta: 0:09:34 lr: 0.000025 loss: 1.459331 (1.632894) time: 0.715411 data: 0.000235 max mem: 14338 Epoch: [18/30] [4250/5004] eta: 0:08:58 lr: 0.000025 loss: 1.767980 (1.634006) time: 0.713501 data: 0.000229 max mem: 14338 Epoch: [18/30] [4300/5004] eta: 0:08:23 lr: 0.000025 loss: 1.593641 (1.633691) time: 0.712201 data: 0.000205 max mem: 14338 Epoch: [18/30] [4350/5004] eta: 0:07:47 lr: 0.000025 loss: 1.519916 (1.634500) time: 0.710149 data: 0.000167 max mem: 14338 Epoch: [18/30] [4400/5004] eta: 0:07:11 lr: 0.000025 loss: 1.621758 (1.634344) time: 0.718504 data: 0.000164 max mem: 14338 Epoch: [18/30] [4450/5004] eta: 0:06:35 lr: 0.000025 loss: 1.567655 (1.635326) time: 0.713723 data: 0.000219 max mem: 14338 Epoch: [18/30] [4500/5004] eta: 0:06:00 lr: 0.000025 loss: 1.611841 (1.636017) time: 0.711283 data: 0.000218 max mem: 14338 Epoch: [18/30] [4550/5004] eta: 0:05:24 lr: 0.000025 loss: 1.603579 (1.636697) time: 0.713335 data: 0.000220 max mem: 14338 Epoch: [18/30] [4600/5004] eta: 0:04:48 lr: 0.000025 loss: 1.702183 (1.636312) time: 0.711198 data: 0.000215 max mem: 14338 Epoch: [18/30] [4650/5004] eta: 0:04:12 lr: 0.000025 loss: 1.485417 (1.636607) time: 0.712333 data: 0.000209 max mem: 14338 Epoch: [18/30] [4700/5004] eta: 0:03:37 lr: 0.000024 loss: 1.510979 (1.636174) time: 0.708896 data: 0.000169 max mem: 14338 Epoch: [18/30] [4750/5004] eta: 0:03:01 lr: 0.000024 loss: 1.668512 (1.636801) time: 0.710483 data: 0.000164 max mem: 14338 Epoch: [18/30] [4800/5004] eta: 0:02:25 lr: 0.000024 loss: 1.581796 (1.636178) time: 0.711811 data: 0.000202 max mem: 14338 Epoch: [18/30] [4850/5004] eta: 0:01:50 lr: 0.000024 loss: 1.516546 (1.636311) time: 0.723235 data: 0.000202 max mem: 14338 Epoch: [18/30] [4900/5004] eta: 0:01:14 lr: 0.000024 loss: 1.563868 (1.636103) time: 0.715383 data: 0.000216 max mem: 14338 Epoch: [18/30] [4950/5004] eta: 0:00:38 lr: 0.000024 loss: 1.544757 (1.636086) time: 0.721264 data: 0.000217 max mem: 14338 Epoch: [18/30] [5000/5004] eta: 0:00:02 lr: 0.000024 loss: 1.679847 (1.635987) time: 0.715817 data: 0.000816 max mem: 14338 Epoch: [18/30] [5003/5004] eta: 0:00:00 lr: 0.000024 loss: 1.596456 (1.635950) time: 0.712701 data: 0.000807 max mem: 14338 Epoch: [18/30] Total time: 0:59:36 (0.714722 s / it) Averaged stats: lr: 0.000024 loss: 1.596456 (1.639706) Test: [ 0/196] eta: 0:05:19 loss: 0.310463 (0.310463) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.628510 data: 1.236604 max mem: 14338 Test: [ 10/196] eta: 0:01:16 loss: 0.494454 (0.539983) acc1: 87.500000 (85.795455) acc5: 100.000000 (98.863636) time: 0.409987 data: 0.112550 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.519804 (0.534347) acc1: 87.500000 (86.607143) acc5: 100.000000 (98.511905) time: 0.287138 data: 0.000128 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.511503 (0.511026) acc1: 87.500000 (87.903226) acc5: 100.000000 (98.588710) time: 0.286191 data: 0.000122 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.442305 (0.520225) acc1: 87.500000 (87.347561) acc5: 100.000000 (98.170732) time: 0.286631 data: 0.000132 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.447231 (0.549309) acc1: 87.500000 (87.254902) acc5: 100.000000 (97.671569) time: 0.286994 data: 0.000154 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.591964 (0.574675) acc1: 87.500000 (86.885246) acc5: 100.000000 (97.643443) time: 0.286511 data: 0.000163 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.652613 (0.593900) acc1: 81.250000 (86.443662) acc5: 100.000000 (97.623239) time: 0.285859 data: 0.000139 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.540306 (0.593174) acc1: 87.500000 (86.342593) acc5: 100.000000 (97.762346) time: 0.285846 data: 0.000130 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.534670 (0.615487) acc1: 87.500000 (85.851648) acc5: 100.000000 (97.458791) time: 0.286523 data: 0.000142 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.573864 (0.605722) acc1: 81.250000 (85.891089) acc5: 100.000000 (97.648515) time: 0.286423 data: 0.000140 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.561937 (0.594814) acc1: 87.500000 (85.867117) acc5: 100.000000 (97.804054) time: 0.291412 data: 0.000120 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.458541 (0.590945) acc1: 87.500000 (85.847107) acc5: 100.000000 (97.830579) time: 0.292061 data: 0.000131 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.648009 (0.606321) acc1: 87.500000 (85.639313) acc5: 100.000000 (97.805344) time: 0.286630 data: 0.000135 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.648009 (0.604642) acc1: 87.500000 (85.726950) acc5: 100.000000 (97.783688) time: 0.287072 data: 0.000156 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.523033 (0.613064) acc1: 87.500000 (85.554636) acc5: 100.000000 (97.806291) time: 0.287221 data: 0.000168 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.505877 (0.617607) acc1: 81.250000 (85.364907) acc5: 100.000000 (97.826087) time: 0.286871 data: 0.000135 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.505877 (0.614409) acc1: 81.250000 (85.489766) acc5: 100.000000 (97.880117) time: 0.286906 data: 0.000147 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.454365 (0.614996) acc1: 81.250000 (85.359116) acc5: 100.000000 (97.893646) time: 0.285956 data: 0.000171 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.399379 (0.603755) acc1: 87.500000 (85.536649) acc5: 100.000000 (97.905759) time: 0.283622 data: 0.000128 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.497104 (0.615236) acc1: 87.500000 (85.408000) acc5: 100.000000 (97.824000) time: 0.275002 data: 0.000111 max mem: 14338 Test: Total time: 0:00:57 (0.294013 s / it) * Acc@1 85.150 Acc@5 97.270 loss 0.629 Max accuracy: 85.15% Epoch: [19/30] [ 0/5004] eta: 2:40:38 lr: 0.000024 loss: 1.757291 (1.757291) time: 1.926194 data: 1.190181 max mem: 14338 Epoch: [19/30] [ 50/5004] eta: 1:01:00 lr: 0.000024 loss: 1.522135 (1.593598) time: 0.721037 data: 0.000197 max mem: 14338 Epoch: [19/30] [ 100/5004] eta: 0:59:18 lr: 0.000024 loss: 1.422457 (1.602004) time: 0.710084 data: 0.000230 max mem: 14338 Epoch: [19/30] [ 150/5004] eta: 0:58:25 lr: 0.000024 loss: 1.625113 (1.629376) time: 0.711849 data: 0.000226 max mem: 14338 Epoch: [19/30] [ 200/5004] eta: 0:57:39 lr: 0.000024 loss: 1.511184 (1.610592) time: 0.713733 data: 0.000225 max mem: 14338 Epoch: [19/30] [ 250/5004] eta: 0:57:00 lr: 0.000024 loss: 1.606895 (1.616287) time: 0.716176 data: 0.000185 max mem: 14338 Epoch: [19/30] [ 300/5004] eta: 0:56:19 lr: 0.000024 loss: 1.577969 (1.629950) time: 0.711178 data: 0.000170 max mem: 14338 Epoch: [19/30] [ 350/5004] eta: 0:55:40 lr: 0.000024 loss: 1.578415 (1.629578) time: 0.708612 data: 0.000168 max mem: 14338 Epoch: [19/30] [ 400/5004] eta: 0:55:02 lr: 0.000024 loss: 1.680045 (1.627305) time: 0.714075 data: 0.000238 max mem: 14338 Epoch: [19/30] [ 450/5004] eta: 0:54:24 lr: 0.000024 loss: 1.536441 (1.624834) time: 0.716552 data: 0.000222 max mem: 14338 Epoch: [19/30] [ 500/5004] eta: 0:53:49 lr: 0.000024 loss: 1.612986 (1.626462) time: 0.719959 data: 0.000211 max mem: 14338 Epoch: [19/30] [ 550/5004] eta: 0:53:11 lr: 0.000024 loss: 1.740998 (1.630281) time: 0.713964 data: 0.000203 max mem: 14338 Epoch: [19/30] [ 600/5004] eta: 0:52:34 lr: 0.000024 loss: 1.555820 (1.627239) time: 0.712365 data: 0.000201 max mem: 14338 Epoch: [19/30] [ 650/5004] eta: 0:51:58 lr: 0.000024 loss: 1.456426 (1.623856) time: 0.715207 data: 0.000156 max mem: 14338 Epoch: [19/30] [ 700/5004] eta: 0:51:22 lr: 0.000024 loss: 1.619904 (1.625345) time: 0.709437 data: 0.000168 max mem: 14338 Epoch: [19/30] [ 750/5004] eta: 0:50:45 lr: 0.000024 loss: 1.577836 (1.626207) time: 0.713706 data: 0.000202 max mem: 14338 Epoch: [19/30] [ 800/5004] eta: 0:50:10 lr: 0.000024 loss: 1.486946 (1.629047) time: 0.714288 data: 0.000225 max mem: 14338 Epoch: [19/30] [ 850/5004] eta: 0:49:33 lr: 0.000024 loss: 1.500913 (1.633896) time: 0.713370 data: 0.000198 max mem: 14338 Epoch: [19/30] [ 900/5004] eta: 0:48:57 lr: 0.000024 loss: 1.493472 (1.629260) time: 0.712776 data: 0.000192 max mem: 14338 Epoch: [19/30] [ 950/5004] eta: 0:48:21 lr: 0.000024 loss: 1.548249 (1.626170) time: 0.715408 data: 0.000206 max mem: 14338 Epoch: [19/30] [1000/5004] eta: 0:47:45 lr: 0.000023 loss: 1.797567 (1.630776) time: 0.721846 data: 0.000167 max mem: 14338 Epoch: [19/30] [1050/5004] eta: 0:47:09 lr: 0.000023 loss: 1.507087 (1.631364) time: 0.713735 data: 0.000214 max mem: 14338 Epoch: [19/30] [1100/5004] eta: 0:46:33 lr: 0.000023 loss: 1.443861 (1.631851) time: 0.713055 data: 0.000204 max mem: 14338 Epoch: [19/30] [1150/5004] eta: 0:45:57 lr: 0.000023 loss: 1.511962 (1.633596) time: 0.711747 data: 0.000225 max mem: 14338 Epoch: [19/30] [1200/5004] eta: 0:45:21 lr: 0.000023 loss: 1.426063 (1.634505) time: 0.716092 data: 0.000209 max mem: 14338 Epoch: [19/30] [1250/5004] eta: 0:44:45 lr: 0.000023 loss: 1.548604 (1.632440) time: 0.715664 data: 0.000213 max mem: 14338 Epoch: [19/30] [1300/5004] eta: 0:44:09 lr: 0.000023 loss: 1.532556 (1.632244) time: 0.710731 data: 0.000158 max mem: 14338 Epoch: [19/30] [1350/5004] eta: 0:43:33 lr: 0.000023 loss: 1.803751 (1.633282) time: 0.711542 data: 0.000164 max mem: 14338 Epoch: [19/30] [1400/5004] eta: 0:42:58 lr: 0.000023 loss: 1.572396 (1.632454) time: 0.717617 data: 0.000202 max mem: 14338 Epoch: [19/30] [1450/5004] eta: 0:42:21 lr: 0.000023 loss: 1.537300 (1.630573) time: 0.717160 data: 0.000238 max mem: 14338 Epoch: [19/30] [1500/5004] eta: 0:41:46 lr: 0.000023 loss: 1.509534 (1.629200) time: 0.712077 data: 0.000220 max mem: 14338 Epoch: [19/30] [1550/5004] eta: 0:41:10 lr: 0.000023 loss: 1.662295 (1.629490) time: 0.711371 data: 0.000186 max mem: 14338 Epoch: [19/30] [1600/5004] eta: 0:40:34 lr: 0.000023 loss: 1.532800 (1.629510) time: 0.716941 data: 0.000225 max mem: 14338 Epoch: [19/30] [1650/5004] eta: 0:39:58 lr: 0.000023 loss: 1.528073 (1.628423) time: 0.716720 data: 0.000178 max mem: 14338 Epoch: [19/30] [1700/5004] eta: 0:39:22 lr: 0.000023 loss: 1.502676 (1.626955) time: 0.710895 data: 0.000167 max mem: 14338 Epoch: [19/30] [1750/5004] eta: 0:38:47 lr: 0.000023 loss: 1.408320 (1.625228) time: 0.710821 data: 0.000225 max mem: 14338 Epoch: [19/30] [1800/5004] eta: 0:38:11 lr: 0.000023 loss: 1.737825 (1.626580) time: 0.714296 data: 0.000207 max mem: 14338 Epoch: [19/30] [1850/5004] eta: 0:37:35 lr: 0.000023 loss: 1.691551 (1.627783) time: 0.714727 data: 0.000226 max mem: 14338 Epoch: [19/30] [1900/5004] eta: 0:36:59 lr: 0.000023 loss: 1.733426 (1.630422) time: 0.720428 data: 0.000232 max mem: 14338 Epoch: [19/30] [1950/5004] eta: 0:36:23 lr: 0.000023 loss: 1.654748 (1.630543) time: 0.715781 data: 0.000220 max mem: 14338 Epoch: [19/30] [2000/5004] eta: 0:35:48 lr: 0.000023 loss: 1.654751 (1.631110) time: 0.713019 data: 0.000166 max mem: 14338 Epoch: [19/30] [2050/5004] eta: 0:35:12 lr: 0.000023 loss: 1.707664 (1.631270) time: 0.712558 data: 0.000152 max mem: 14338 Epoch: [19/30] [2100/5004] eta: 0:34:36 lr: 0.000023 loss: 1.691555 (1.631671) time: 0.709984 data: 0.000211 max mem: 14338 Epoch: [19/30] [2150/5004] eta: 0:34:00 lr: 0.000023 loss: 1.460665 (1.630358) time: 0.708446 data: 0.000232 max mem: 14338 Epoch: [19/30] [2200/5004] eta: 0:33:24 lr: 0.000023 loss: 1.590720 (1.629650) time: 0.716765 data: 0.000179 max mem: 14338 Epoch: [19/30] [2250/5004] eta: 0:32:49 lr: 0.000023 loss: 1.462143 (1.629817) time: 0.715219 data: 0.000216 max mem: 14338 Epoch: [19/30] [2300/5004] eta: 0:32:13 lr: 0.000023 loss: 1.532818 (1.629215) time: 0.718528 data: 0.000213 max mem: 14338 Epoch: [19/30] [2350/5004] eta: 0:31:37 lr: 0.000022 loss: 1.599150 (1.631446) time: 0.717143 data: 0.000158 max mem: 14338 Epoch: [19/30] [2400/5004] eta: 0:31:01 lr: 0.000022 loss: 1.546380 (1.632106) time: 0.718467 data: 0.000220 max mem: 14338 Epoch: [19/30] [2450/5004] eta: 0:30:25 lr: 0.000022 loss: 1.526185 (1.630427) time: 0.711827 data: 0.000230 max mem: 14338 Epoch: [19/30] [2500/5004] eta: 0:29:50 lr: 0.000022 loss: 1.589345 (1.629962) time: 0.712633 data: 0.000235 max mem: 14338 Epoch: [19/30] [2550/5004] eta: 0:29:14 lr: 0.000022 loss: 1.603858 (1.630802) time: 0.709938 data: 0.000219 max mem: 14338 Epoch: [19/30] [2600/5004] eta: 0:28:38 lr: 0.000022 loss: 1.666759 (1.630622) time: 0.711556 data: 0.000211 max mem: 14338 Epoch: [19/30] [2650/5004] eta: 0:28:02 lr: 0.000022 loss: 1.506872 (1.629140) time: 0.718437 data: 0.000151 max mem: 14338 Epoch: [19/30] [2700/5004] eta: 0:27:27 lr: 0.000022 loss: 1.441203 (1.627974) time: 0.713195 data: 0.000160 max mem: 14338 Epoch: [19/30] [2750/5004] eta: 0:26:51 lr: 0.000022 loss: 1.681775 (1.628956) time: 0.716979 data: 0.000235 max mem: 14338 Epoch: [19/30] [2800/5004] eta: 0:26:15 lr: 0.000022 loss: 1.530369 (1.628472) time: 0.714141 data: 0.000211 max mem: 14338 Epoch: [19/30] [2850/5004] eta: 0:25:40 lr: 0.000022 loss: 1.677870 (1.629207) time: 0.721824 data: 0.000186 max mem: 14338 Epoch: [19/30] [2900/5004] eta: 0:25:04 lr: 0.000022 loss: 1.476064 (1.628541) time: 0.710704 data: 0.000194 max mem: 14338 Epoch: [19/30] [2950/5004] eta: 0:24:28 lr: 0.000022 loss: 1.622975 (1.628469) time: 0.708973 data: 0.000221 max mem: 14338 Epoch: [19/30] [3000/5004] eta: 0:23:52 lr: 0.000022 loss: 1.526968 (1.628191) time: 0.717070 data: 0.000162 max mem: 14338 Epoch: [19/30] [3050/5004] eta: 0:23:16 lr: 0.000022 loss: 1.562667 (1.628818) time: 0.712356 data: 0.000175 max mem: 14338 Epoch: [19/30] [3100/5004] eta: 0:22:41 lr: 0.000022 loss: 1.647498 (1.629936) time: 0.712706 data: 0.000228 max mem: 14338 Epoch: [19/30] [3150/5004] eta: 0:22:05 lr: 0.000022 loss: 1.408859 (1.629371) time: 0.713448 data: 0.000210 max mem: 14338 Epoch: [19/30] [3200/5004] eta: 0:21:29 lr: 0.000022 loss: 1.403803 (1.629311) time: 0.713189 data: 0.000204 max mem: 14338 Epoch: [19/30] [3250/5004] eta: 0:20:53 lr: 0.000022 loss: 1.522496 (1.630112) time: 0.718359 data: 0.000215 max mem: 14338 Epoch: [19/30] [3300/5004] eta: 0:20:18 lr: 0.000022 loss: 1.570589 (1.630411) time: 0.718815 data: 0.000222 max mem: 14338 Epoch: [19/30] [3350/5004] eta: 0:19:42 lr: 0.000022 loss: 1.611257 (1.629843) time: 0.715237 data: 0.000165 max mem: 14338 Epoch: [19/30] [3400/5004] eta: 0:19:06 lr: 0.000022 loss: 1.515494 (1.629322) time: 0.716026 data: 0.000160 max mem: 14338 Epoch: [19/30] [3450/5004] eta: 0:18:30 lr: 0.000022 loss: 1.670666 (1.629572) time: 0.713997 data: 0.000207 max mem: 14338 Epoch: [19/30] [3500/5004] eta: 0:17:55 lr: 0.000022 loss: 1.576389 (1.629762) time: 0.710510 data: 0.000216 max mem: 14338 Epoch: [19/30] [3550/5004] eta: 0:17:19 lr: 0.000022 loss: 1.567732 (1.630060) time: 0.708993 data: 0.000214 max mem: 14338 Epoch: [19/30] [3600/5004] eta: 0:16:43 lr: 0.000022 loss: 1.495928 (1.630051) time: 0.713389 data: 0.000209 max mem: 14338 Epoch: [19/30] [3650/5004] eta: 0:16:07 lr: 0.000022 loss: 1.530602 (1.630008) time: 0.712538 data: 0.000217 max mem: 14338 Epoch: [19/30] [3700/5004] eta: 0:15:32 lr: 0.000021 loss: 1.677958 (1.630097) time: 0.714413 data: 0.000162 max mem: 14338 Epoch: [19/30] [3750/5004] eta: 0:14:56 lr: 0.000021 loss: 1.458819 (1.628982) time: 0.715755 data: 0.000239 max mem: 14338 Epoch: [19/30] [3800/5004] eta: 0:14:20 lr: 0.000021 loss: 1.709841 (1.629849) time: 0.712553 data: 0.000215 max mem: 14338 Epoch: [19/30] [3850/5004] eta: 0:13:44 lr: 0.000021 loss: 1.640860 (1.630721) time: 0.715139 data: 0.000227 max mem: 14338 Epoch: [19/30] [3900/5004] eta: 0:13:09 lr: 0.000021 loss: 1.675804 (1.631485) time: 0.709541 data: 0.000211 max mem: 14338 Epoch: [19/30] [3950/5004] eta: 0:12:33 lr: 0.000021 loss: 1.652548 (1.631792) time: 0.714875 data: 0.000213 max mem: 14338 Epoch: [19/30] [4000/5004] eta: 0:11:57 lr: 0.000021 loss: 1.623933 (1.632317) time: 0.714127 data: 0.000175 max mem: 14338 Epoch: [19/30] [4050/5004] eta: 0:11:21 lr: 0.000021 loss: 1.481045 (1.631900) time: 0.715435 data: 0.000165 max mem: 14338 Epoch: [19/30] [4100/5004] eta: 0:10:46 lr: 0.000021 loss: 1.511773 (1.631527) time: 0.711038 data: 0.000203 max mem: 14338 Epoch: [19/30] [4150/5004] eta: 0:10:10 lr: 0.000021 loss: 1.684436 (1.631887) time: 0.715303 data: 0.000191 max mem: 14338 Epoch: [19/30] [4200/5004] eta: 0:09:34 lr: 0.000021 loss: 1.621225 (1.631406) time: 0.717343 data: 0.000205 max mem: 14338 Epoch: [19/30] [4250/5004] eta: 0:08:58 lr: 0.000021 loss: 1.522949 (1.630794) time: 0.718572 data: 0.000230 max mem: 14338 Epoch: [19/30] [4300/5004] eta: 0:08:23 lr: 0.000021 loss: 1.550690 (1.630931) time: 0.710561 data: 0.000220 max mem: 14338 Epoch: [19/30] [4350/5004] eta: 0:07:47 lr: 0.000021 loss: 1.632074 (1.631214) time: 0.709131 data: 0.000165 max mem: 14338 Epoch: [19/30] [4400/5004] eta: 0:07:11 lr: 0.000021 loss: 1.597720 (1.630956) time: 0.717407 data: 0.000166 max mem: 14338 Epoch: [19/30] [4450/5004] eta: 0:06:35 lr: 0.000021 loss: 1.512039 (1.630766) time: 0.712770 data: 0.000237 max mem: 14338 Epoch: [19/30] [4500/5004] eta: 0:06:00 lr: 0.000021 loss: 1.604181 (1.631008) time: 0.710501 data: 0.000219 max mem: 14338 Epoch: [19/30] [4550/5004] eta: 0:05:24 lr: 0.000021 loss: 1.482404 (1.630819) time: 0.709475 data: 0.000227 max mem: 14338 Epoch: [19/30] [4600/5004] eta: 0:04:48 lr: 0.000021 loss: 1.599393 (1.631534) time: 0.713344 data: 0.000221 max mem: 14338 Epoch: [19/30] [4650/5004] eta: 0:04:12 lr: 0.000021 loss: 1.623516 (1.631272) time: 0.716144 data: 0.000224 max mem: 14338 Epoch: [19/30] [4700/5004] eta: 0:03:37 lr: 0.000021 loss: 1.500008 (1.630780) time: 0.711439 data: 0.000175 max mem: 14338 Epoch: [19/30] [4750/5004] eta: 0:03:01 lr: 0.000021 loss: 1.527430 (1.630484) time: 0.711487 data: 0.000177 max mem: 14338 Epoch: [19/30] [4800/5004] eta: 0:02:25 lr: 0.000021 loss: 1.464037 (1.630499) time: 0.717987 data: 0.000197 max mem: 14338 Epoch: [19/30] [4850/5004] eta: 0:01:50 lr: 0.000021 loss: 1.490110 (1.630280) time: 0.712268 data: 0.000228 max mem: 14338 Epoch: [19/30] [4900/5004] eta: 0:01:14 lr: 0.000021 loss: 1.460284 (1.630066) time: 0.709748 data: 0.000204 max mem: 14338 Epoch: [19/30] [4950/5004] eta: 0:00:38 lr: 0.000021 loss: 1.642701 (1.630032) time: 0.711875 data: 0.000227 max mem: 14338 Epoch: [19/30] [5000/5004] eta: 0:00:02 lr: 0.000021 loss: 1.588768 (1.630226) time: 0.709967 data: 0.000839 max mem: 14338 Epoch: [19/30] [5003/5004] eta: 0:00:00 lr: 0.000021 loss: 1.521333 (1.630133) time: 0.706397 data: 0.000830 max mem: 14338 Epoch: [19/30] Total time: 0:59:36 (0.714686 s / it) Averaged stats: lr: 0.000021 loss: 1.521333 (1.634616) Test: [ 0/196] eta: 0:05:01 loss: 0.284120 (0.284120) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.536043 data: 1.126544 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.533331 (0.542021) acc1: 87.500000 (85.795455) acc5: 100.000000 (98.863636) time: 0.400630 data: 0.102540 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.543386 (0.541949) acc1: 87.500000 (86.309524) acc5: 100.000000 (98.214286) time: 0.287317 data: 0.000127 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.499034 (0.516919) acc1: 87.500000 (87.298387) acc5: 100.000000 (98.387097) time: 0.286871 data: 0.000111 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.427405 (0.522617) acc1: 87.500000 (87.042683) acc5: 100.000000 (98.170732) time: 0.286134 data: 0.000119 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.434639 (0.551306) acc1: 87.500000 (87.009804) acc5: 100.000000 (97.549020) time: 0.289664 data: 0.000123 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.619280 (0.578054) acc1: 87.500000 (86.372951) acc5: 93.750000 (97.540984) time: 0.294578 data: 0.000123 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.635780 (0.593386) acc1: 81.250000 (86.003521) acc5: 100.000000 (97.623239) time: 0.291675 data: 0.000125 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.517666 (0.592604) acc1: 87.500000 (85.956790) acc5: 100.000000 (97.762346) time: 0.287597 data: 0.000117 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.522539 (0.614070) acc1: 87.500000 (85.508242) acc5: 100.000000 (97.458791) time: 0.287473 data: 0.000123 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.591172 (0.604595) acc1: 81.250000 (85.581683) acc5: 100.000000 (97.648515) time: 0.287402 data: 0.000122 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.538330 (0.593116) acc1: 87.500000 (85.585586) acc5: 100.000000 (97.747748) time: 0.287326 data: 0.000123 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.509884 (0.589241) acc1: 87.500000 (85.640496) acc5: 100.000000 (97.727273) time: 0.287024 data: 0.000143 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.646746 (0.604544) acc1: 87.500000 (85.353053) acc5: 100.000000 (97.709924) time: 0.287317 data: 0.000152 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.650179 (0.603856) acc1: 87.500000 (85.416667) acc5: 100.000000 (97.695035) time: 0.287422 data: 0.000140 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.538666 (0.611727) acc1: 87.500000 (85.306291) acc5: 100.000000 (97.682119) time: 0.287254 data: 0.000119 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.491783 (0.615997) acc1: 81.250000 (85.248447) acc5: 100.000000 (97.709627) time: 0.286877 data: 0.000114 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.463631 (0.612980) acc1: 81.250000 (85.343567) acc5: 100.000000 (97.770468) time: 0.286270 data: 0.000126 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.463631 (0.612994) acc1: 87.500000 (85.220994) acc5: 100.000000 (97.755525) time: 0.285828 data: 0.000125 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.365318 (0.601348) acc1: 87.500000 (85.438482) acc5: 100.000000 (97.774869) time: 0.288585 data: 0.000096 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.494993 (0.612007) acc1: 87.500000 (85.312000) acc5: 100.000000 (97.696000) time: 0.281478 data: 0.000085 max mem: 14338 Test: Total time: 0:00:57 (0.294853 s / it) * Acc@1 85.246 Acc@5 97.298 loss 0.629 Max accuracy: 85.25% Epoch: [20/30] [ 0/5004] eta: 2:36:50 lr: 0.000021 loss: 1.243587 (1.243587) time: 1.880570 data: 1.148767 max mem: 14338 Epoch: [20/30] [ 50/5004] eta: 1:00:50 lr: 0.000021 loss: 1.481200 (1.593360) time: 0.713051 data: 0.000190 max mem: 14338 Epoch: [20/30] [ 100/5004] eta: 0:59:16 lr: 0.000020 loss: 1.668703 (1.602997) time: 0.711846 data: 0.000208 max mem: 14338 Epoch: [20/30] [ 150/5004] eta: 0:58:20 lr: 0.000020 loss: 1.562581 (1.622187) time: 0.712302 data: 0.000223 max mem: 14338 Epoch: [20/30] [ 200/5004] eta: 0:57:40 lr: 0.000020 loss: 1.436925 (1.612591) time: 0.723467 data: 0.000217 max mem: 14338 Epoch: [20/30] [ 250/5004] eta: 0:56:56 lr: 0.000020 loss: 1.522457 (1.611813) time: 0.712952 data: 0.000185 max mem: 14338 Epoch: [20/30] [ 300/5004] eta: 0:56:16 lr: 0.000020 loss: 1.592193 (1.611694) time: 0.713935 data: 0.000152 max mem: 14338 Epoch: [20/30] [ 350/5004] eta: 0:55:38 lr: 0.000020 loss: 1.654271 (1.618981) time: 0.709792 data: 0.000156 max mem: 14338 Epoch: [20/30] [ 400/5004] eta: 0:54:59 lr: 0.000020 loss: 1.493143 (1.622559) time: 0.712491 data: 0.000226 max mem: 14338 Epoch: [20/30] [ 450/5004] eta: 0:54:21 lr: 0.000020 loss: 1.592975 (1.622361) time: 0.712476 data: 0.000218 max mem: 14338 Epoch: [20/30] [ 500/5004] eta: 0:53:44 lr: 0.000020 loss: 1.584781 (1.626350) time: 0.710601 data: 0.000222 max mem: 14338 Epoch: [20/30] [ 550/5004] eta: 0:53:09 lr: 0.000020 loss: 1.443495 (1.623184) time: 0.715021 data: 0.000220 max mem: 14338 Epoch: [20/30] [ 600/5004] eta: 0:52:35 lr: 0.000020 loss: 1.600014 (1.627832) time: 0.726542 data: 0.000206 max mem: 14338 Epoch: [20/30] [ 650/5004] eta: 0:51:59 lr: 0.000020 loss: 1.481356 (1.623443) time: 0.721699 data: 0.000161 max mem: 14338 Epoch: [20/30] [ 700/5004] eta: 0:51:23 lr: 0.000020 loss: 1.514516 (1.620866) time: 0.710816 data: 0.000170 max mem: 14338 Epoch: [20/30] [ 750/5004] eta: 0:50:46 lr: 0.000020 loss: 1.859726 (1.628602) time: 0.714421 data: 0.000218 max mem: 14338 Epoch: [20/30] [ 800/5004] eta: 0:50:10 lr: 0.000020 loss: 1.504114 (1.625531) time: 0.713788 data: 0.000211 max mem: 14338 Epoch: [20/30] [ 850/5004] eta: 0:49:34 lr: 0.000020 loss: 1.596480 (1.623449) time: 0.711755 data: 0.000227 max mem: 14338 Epoch: [20/30] [ 900/5004] eta: 0:48:58 lr: 0.000020 loss: 1.543769 (1.624501) time: 0.713212 data: 0.000214 max mem: 14338 Epoch: [20/30] [ 950/5004] eta: 0:48:22 lr: 0.000020 loss: 1.435535 (1.621699) time: 0.710952 data: 0.000217 max mem: 14338 Epoch: [20/30] [1000/5004] eta: 0:47:46 lr: 0.000020 loss: 1.560793 (1.622581) time: 0.715403 data: 0.000163 max mem: 14338 Epoch: [20/30] [1050/5004] eta: 0:47:10 lr: 0.000020 loss: 1.538110 (1.620090) time: 0.718106 data: 0.000221 max mem: 14338 Epoch: [20/30] [1100/5004] eta: 0:46:34 lr: 0.000020 loss: 1.669391 (1.620418) time: 0.713419 data: 0.000225 max mem: 14338 Epoch: [20/30] [1150/5004] eta: 0:45:58 lr: 0.000020 loss: 1.554723 (1.620551) time: 0.714393 data: 0.000231 max mem: 14338 Epoch: [20/30] [1200/5004] eta: 0:45:22 lr: 0.000020 loss: 1.656574 (1.624216) time: 0.717095 data: 0.000220 max mem: 14338 Epoch: [20/30] [1250/5004] eta: 0:44:46 lr: 0.000020 loss: 1.600631 (1.626038) time: 0.715344 data: 0.000213 max mem: 14338 Epoch: [20/30] [1300/5004] eta: 0:44:10 lr: 0.000020 loss: 1.613326 (1.626099) time: 0.710696 data: 0.000158 max mem: 14338 Epoch: [20/30] [1350/5004] eta: 0:43:34 lr: 0.000020 loss: 1.588683 (1.626183) time: 0.709391 data: 0.000159 max mem: 14338 Epoch: [20/30] [1400/5004] eta: 0:42:58 lr: 0.000020 loss: 1.478225 (1.626490) time: 0.715281 data: 0.000221 max mem: 14338 Epoch: [20/30] [1450/5004] eta: 0:42:22 lr: 0.000020 loss: 1.672295 (1.625475) time: 0.712076 data: 0.000210 max mem: 14338 Epoch: [20/30] [1500/5004] eta: 0:41:46 lr: 0.000019 loss: 1.463728 (1.624117) time: 0.713088 data: 0.000245 max mem: 14338 Epoch: [20/30] [1550/5004] eta: 0:41:10 lr: 0.000019 loss: 1.660993 (1.625416) time: 0.716562 data: 0.000223 max mem: 14338 Epoch: [20/30] [1600/5004] eta: 0:40:34 lr: 0.000019 loss: 1.510607 (1.625863) time: 0.721026 data: 0.000216 max mem: 14338 Epoch: [20/30] [1650/5004] eta: 0:39:59 lr: 0.000019 loss: 1.568556 (1.624888) time: 0.721650 data: 0.000179 max mem: 14338 Epoch: [20/30] [1700/5004] eta: 0:39:23 lr: 0.000019 loss: 1.668748 (1.625764) time: 0.710575 data: 0.000166 max mem: 14338 Epoch: [20/30] [1750/5004] eta: 0:38:47 lr: 0.000019 loss: 1.666700 (1.625260) time: 0.710263 data: 0.000222 max mem: 14338 Epoch: [20/30] [1800/5004] eta: 0:38:11 lr: 0.000019 loss: 1.704932 (1.626091) time: 0.711664 data: 0.000205 max mem: 14338 Epoch: [20/30] [1850/5004] eta: 0:37:35 lr: 0.000019 loss: 1.418113 (1.625995) time: 0.712722 data: 0.000217 max mem: 14338 Epoch: [20/30] [1900/5004] eta: 0:36:59 lr: 0.000019 loss: 1.589414 (1.625672) time: 0.710748 data: 0.000227 max mem: 14338 Epoch: [20/30] [1950/5004] eta: 0:36:23 lr: 0.000019 loss: 1.604978 (1.627055) time: 0.713271 data: 0.000217 max mem: 14338 Epoch: [20/30] [2000/5004] eta: 0:35:47 lr: 0.000019 loss: 1.587725 (1.626898) time: 0.718181 data: 0.000165 max mem: 14338 Epoch: [20/30] [2050/5004] eta: 0:35:12 lr: 0.000019 loss: 1.514830 (1.626977) time: 0.720573 data: 0.000159 max mem: 14338 Epoch: [20/30] [2100/5004] eta: 0:34:36 lr: 0.000019 loss: 1.542492 (1.627559) time: 0.718595 data: 0.000218 max mem: 14338 Epoch: [20/30] [2150/5004] eta: 0:34:00 lr: 0.000019 loss: 1.608555 (1.626222) time: 0.711094 data: 0.000242 max mem: 14338 Epoch: [20/30] [2200/5004] eta: 0:33:24 lr: 0.000019 loss: 1.424201 (1.625974) time: 0.711463 data: 0.000180 max mem: 14338 Epoch: [20/30] [2250/5004] eta: 0:32:48 lr: 0.000019 loss: 1.680228 (1.626448) time: 0.714117 data: 0.000219 max mem: 14338 Epoch: [20/30] [2300/5004] eta: 0:32:13 lr: 0.000019 loss: 1.470610 (1.625392) time: 0.709793 data: 0.000239 max mem: 14338 Epoch: [20/30] [2350/5004] eta: 0:31:37 lr: 0.000019 loss: 1.692351 (1.627488) time: 0.709710 data: 0.000168 max mem: 14338 Epoch: [20/30] [2400/5004] eta: 0:31:01 lr: 0.000019 loss: 1.557335 (1.628456) time: 0.716364 data: 0.000233 max mem: 14338 Epoch: [20/30] [2450/5004] eta: 0:30:25 lr: 0.000019 loss: 1.608453 (1.628635) time: 0.713737 data: 0.000232 max mem: 14338 Epoch: [20/30] [2500/5004] eta: 0:29:49 lr: 0.000019 loss: 1.521258 (1.629272) time: 0.717883 data: 0.000229 max mem: 14338 Epoch: [20/30] [2550/5004] eta: 0:29:14 lr: 0.000019 loss: 1.637218 (1.628954) time: 0.715221 data: 0.000223 max mem: 14338 Epoch: [20/30] [2600/5004] eta: 0:28:38 lr: 0.000019 loss: 1.629798 (1.629705) time: 0.718922 data: 0.000230 max mem: 14338 Epoch: [20/30] [2650/5004] eta: 0:28:02 lr: 0.000019 loss: 1.624393 (1.631281) time: 0.717393 data: 0.000158 max mem: 14338 Epoch: [20/30] [2700/5004] eta: 0:27:26 lr: 0.000019 loss: 1.625443 (1.632456) time: 0.712133 data: 0.000163 max mem: 14338 Epoch: [20/30] [2750/5004] eta: 0:26:51 lr: 0.000019 loss: 1.491951 (1.632140) time: 0.712696 data: 0.000218 max mem: 14338 Epoch: [20/30] [2800/5004] eta: 0:26:15 lr: 0.000019 loss: 1.509073 (1.632010) time: 0.711437 data: 0.000211 max mem: 14338 Epoch: [20/30] [2850/5004] eta: 0:25:39 lr: 0.000019 loss: 1.506619 (1.631797) time: 0.714189 data: 0.000205 max mem: 14338 Epoch: [20/30] [2900/5004] eta: 0:25:04 lr: 0.000019 loss: 1.532585 (1.631574) time: 0.715848 data: 0.000208 max mem: 14338 Epoch: [20/30] [2950/5004] eta: 0:24:28 lr: 0.000018 loss: 1.566964 (1.631545) time: 0.714938 data: 0.000233 max mem: 14338 Epoch: [20/30] [3000/5004] eta: 0:23:52 lr: 0.000018 loss: 1.596918 (1.630564) time: 0.717573 data: 0.000157 max mem: 14338 Epoch: [20/30] [3050/5004] eta: 0:23:16 lr: 0.000018 loss: 1.545411 (1.630231) time: 0.717970 data: 0.000160 max mem: 14338 Epoch: [20/30] [3100/5004] eta: 0:22:41 lr: 0.000018 loss: 1.586367 (1.631083) time: 0.710827 data: 0.000196 max mem: 14338 Epoch: [20/30] [3150/5004] eta: 0:22:05 lr: 0.000018 loss: 1.573364 (1.631057) time: 0.712753 data: 0.000209 max mem: 14338 Epoch: [20/30] [3200/5004] eta: 0:21:29 lr: 0.000018 loss: 1.543285 (1.631610) time: 0.715200 data: 0.000226 max mem: 14338 Epoch: [20/30] [3250/5004] eta: 0:20:53 lr: 0.000018 loss: 1.497616 (1.632039) time: 0.712431 data: 0.000222 max mem: 14338 Epoch: [20/30] [3300/5004] eta: 0:20:17 lr: 0.000018 loss: 1.644146 (1.632502) time: 0.710810 data: 0.000288 max mem: 14338 Epoch: [20/30] [3350/5004] eta: 0:19:42 lr: 0.000018 loss: 1.524021 (1.632034) time: 0.712165 data: 0.000177 max mem: 14338 Epoch: [20/30] [3400/5004] eta: 0:19:06 lr: 0.000018 loss: 1.515455 (1.632223) time: 0.720133 data: 0.000155 max mem: 14338 Epoch: [20/30] [3450/5004] eta: 0:18:30 lr: 0.000018 loss: 1.530032 (1.632257) time: 0.720231 data: 0.000210 max mem: 14338 Epoch: [20/30] [3500/5004] eta: 0:17:54 lr: 0.000018 loss: 1.487139 (1.631554) time: 0.715672 data: 0.000192 max mem: 14338 Epoch: [20/30] [3550/5004] eta: 0:17:19 lr: 0.000018 loss: 1.634084 (1.631566) time: 0.717655 data: 0.000220 max mem: 14338 Epoch: [20/30] [3600/5004] eta: 0:16:43 lr: 0.000018 loss: 1.704039 (1.632459) time: 0.712309 data: 0.000218 max mem: 14338 Epoch: [20/30] [3650/5004] eta: 0:16:07 lr: 0.000018 loss: 1.461901 (1.632768) time: 0.713087 data: 0.000238 max mem: 14338 Epoch: [20/30] [3700/5004] eta: 0:15:31 lr: 0.000018 loss: 1.539840 (1.632562) time: 0.709770 data: 0.000159 max mem: 14338 Epoch: [20/30] [3750/5004] eta: 0:14:56 lr: 0.000018 loss: 1.750656 (1.633182) time: 0.715124 data: 0.000214 max mem: 14338 Epoch: [20/30] [3800/5004] eta: 0:14:20 lr: 0.000018 loss: 1.462262 (1.632911) time: 0.710960 data: 0.000216 max mem: 14338 Epoch: [20/30] [3850/5004] eta: 0:13:44 lr: 0.000018 loss: 1.733336 (1.632439) time: 0.718149 data: 0.000221 max mem: 14338 Epoch: [20/30] [3900/5004] eta: 0:13:08 lr: 0.000018 loss: 1.465170 (1.631081) time: 0.714146 data: 0.000223 max mem: 14338 Epoch: [20/30] [3950/5004] eta: 0:12:33 lr: 0.000018 loss: 1.637479 (1.631139) time: 0.713884 data: 0.000230 max mem: 14338 Epoch: [20/30] [4000/5004] eta: 0:11:57 lr: 0.000018 loss: 1.607516 (1.631625) time: 0.714762 data: 0.000172 max mem: 14338 Epoch: [20/30] [4050/5004] eta: 0:11:21 lr: 0.000018 loss: 1.566398 (1.631566) time: 0.715260 data: 0.000175 max mem: 14338 Epoch: [20/30] [4100/5004] eta: 0:10:45 lr: 0.000018 loss: 1.592410 (1.631077) time: 0.710854 data: 0.000226 max mem: 14338 Epoch: [20/30] [4150/5004] eta: 0:10:10 lr: 0.000018 loss: 1.514702 (1.630954) time: 0.709941 data: 0.000191 max mem: 14338 Epoch: [20/30] [4200/5004] eta: 0:09:34 lr: 0.000018 loss: 1.530966 (1.631172) time: 0.716155 data: 0.000199 max mem: 14338 Epoch: [20/30] [4250/5004] eta: 0:08:58 lr: 0.000018 loss: 1.660468 (1.632322) time: 0.713967 data: 0.000219 max mem: 14338 Epoch: [20/30] [4300/5004] eta: 0:08:23 lr: 0.000018 loss: 1.586298 (1.632238) time: 0.714441 data: 0.000215 max mem: 14338 Epoch: [20/30] [4350/5004] eta: 0:07:47 lr: 0.000018 loss: 1.568073 (1.632939) time: 0.713941 data: 0.000175 max mem: 14338 Epoch: [20/30] [4400/5004] eta: 0:07:11 lr: 0.000017 loss: 1.443732 (1.632898) time: 0.718067 data: 0.000169 max mem: 14338 Epoch: [20/30] [4450/5004] eta: 0:06:35 lr: 0.000017 loss: 1.712388 (1.633491) time: 0.713553 data: 0.000213 max mem: 14338 Epoch: [20/30] [4500/5004] eta: 0:06:00 lr: 0.000017 loss: 1.625944 (1.633925) time: 0.713813 data: 0.000213 max mem: 14338 Epoch: [20/30] [4550/5004] eta: 0:05:24 lr: 0.000017 loss: 1.648180 (1.633762) time: 0.708802 data: 0.000214 max mem: 14338 Epoch: [20/30] [4600/5004] eta: 0:04:48 lr: 0.000017 loss: 1.665299 (1.633971) time: 0.715196 data: 0.000226 max mem: 14338 Epoch: [20/30] [4650/5004] eta: 0:04:12 lr: 0.000017 loss: 1.622804 (1.634382) time: 0.714088 data: 0.000218 max mem: 14338 Epoch: [20/30] [4700/5004] eta: 0:03:37 lr: 0.000017 loss: 1.563224 (1.635025) time: 0.708761 data: 0.000169 max mem: 14338 Epoch: [20/30] [4750/5004] eta: 0:03:01 lr: 0.000017 loss: 1.545580 (1.634599) time: 0.714308 data: 0.000178 max mem: 14338 Epoch: [20/30] [4800/5004] eta: 0:02:25 lr: 0.000017 loss: 1.645677 (1.634289) time: 0.717645 data: 0.000198 max mem: 14338 Epoch: [20/30] [4850/5004] eta: 0:01:50 lr: 0.000017 loss: 1.433502 (1.633431) time: 0.714430 data: 0.000234 max mem: 14338 Epoch: [20/30] [4900/5004] eta: 0:01:14 lr: 0.000017 loss: 1.537142 (1.632972) time: 0.713186 data: 0.000222 max mem: 14338 Epoch: [20/30] [4950/5004] eta: 0:00:38 lr: 0.000017 loss: 1.571634 (1.633575) time: 0.712334 data: 0.000218 max mem: 14338 Epoch: [20/30] [5000/5004] eta: 0:00:02 lr: 0.000017 loss: 1.631669 (1.634091) time: 0.708342 data: 0.000823 max mem: 14338 Epoch: [20/30] [5003/5004] eta: 0:00:00 lr: 0.000017 loss: 1.567468 (1.634123) time: 0.705290 data: 0.000818 max mem: 14338 Epoch: [20/30] Total time: 0:59:36 (0.714694 s / it) Averaged stats: lr: 0.000017 loss: 1.567468 (1.632581) Test: [ 0/196] eta: 0:05:19 loss: 0.285763 (0.285763) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.627905 data: 1.176251 max mem: 14338 Test: [ 10/196] eta: 0:01:17 loss: 0.479425 (0.549962) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.418524 data: 0.107053 max mem: 14338 Test: [ 20/196] eta: 0:01:02 loss: 0.515881 (0.540403) acc1: 87.500000 (85.119048) acc5: 100.000000 (98.214286) time: 0.292965 data: 0.000132 max mem: 14338 Test: [ 30/196] eta: 0:00:55 loss: 0.485915 (0.516309) acc1: 87.500000 (86.491935) acc5: 100.000000 (98.387097) time: 0.287975 data: 0.000120 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.435365 (0.524618) acc1: 87.500000 (86.280488) acc5: 100.000000 (98.170732) time: 0.286869 data: 0.000126 max mem: 14338 Test: [ 50/196] eta: 0:00:46 loss: 0.446392 (0.551633) acc1: 87.500000 (86.397059) acc5: 100.000000 (97.671569) time: 0.286245 data: 0.000130 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.611903 (0.579389) acc1: 87.500000 (85.758197) acc5: 93.750000 (97.540984) time: 0.286211 data: 0.000133 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.667468 (0.595872) acc1: 81.250000 (85.299296) acc5: 100.000000 (97.711268) time: 0.286270 data: 0.000150 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.492313 (0.593900) acc1: 87.500000 (85.339506) acc5: 100.000000 (97.839506) time: 0.285987 data: 0.000146 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.492313 (0.615661) acc1: 87.500000 (85.096154) acc5: 100.000000 (97.596154) time: 0.286899 data: 0.000148 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.590508 (0.605321) acc1: 81.250000 (85.272277) acc5: 100.000000 (97.772277) time: 0.287242 data: 0.000135 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.557402 (0.594152) acc1: 87.500000 (85.360360) acc5: 100.000000 (97.916667) time: 0.285924 data: 0.000113 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.498995 (0.590232) acc1: 87.500000 (85.485537) acc5: 100.000000 (97.882231) time: 0.285873 data: 0.000127 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.602773 (0.605567) acc1: 87.500000 (85.209924) acc5: 100.000000 (97.853053) time: 0.285968 data: 0.000131 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.664605 (0.604721) acc1: 87.500000 (85.372340) acc5: 100.000000 (97.828014) time: 0.290783 data: 0.000131 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.565902 (0.613316) acc1: 87.500000 (85.264901) acc5: 100.000000 (97.847682) time: 0.292411 data: 0.000129 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.507546 (0.617630) acc1: 81.250000 (85.248447) acc5: 100.000000 (97.903727) time: 0.288208 data: 0.000118 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.504450 (0.614221) acc1: 81.250000 (85.343567) acc5: 100.000000 (97.953216) time: 0.286577 data: 0.000135 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.504450 (0.614155) acc1: 87.500000 (85.255525) acc5: 100.000000 (97.962707) time: 0.285477 data: 0.000139 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.366638 (0.602486) acc1: 87.500000 (85.438482) acc5: 100.000000 (97.971204) time: 0.283215 data: 0.000102 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.502576 (0.613305) acc1: 87.500000 (85.344000) acc5: 100.000000 (97.888000) time: 0.278462 data: 0.000093 max mem: 14338 Test: Total time: 0:00:57 (0.294836 s / it) * Acc@1 85.084 Acc@5 97.302 loss 0.631 Max accuracy: 85.25% uploading checkpoint virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0020.pth to hdfs://harunava/user/guoyuanfan/HCSC/virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0020.pth Epoch: [21/30] [ 0/5004] eta: 2:39:54 lr: 0.000017 loss: 1.442816 (1.442816) time: 1.917463 data: 1.193341 max mem: 14338 Epoch: [21/30] [ 50/5004] eta: 1:00:49 lr: 0.000017 loss: 1.626540 (1.626657) time: 0.711343 data: 0.000186 max mem: 14338 Epoch: [21/30] [ 100/5004] eta: 0:59:13 lr: 0.000017 loss: 1.571124 (1.641835) time: 0.712734 data: 0.000232 max mem: 14338 Epoch: [21/30] [ 150/5004] eta: 0:58:22 lr: 0.000017 loss: 1.570637 (1.633048) time: 0.717922 data: 0.000225 max mem: 14338 Epoch: [21/30] [ 200/5004] eta: 0:57:43 lr: 0.000017 loss: 1.612946 (1.644312) time: 0.723806 data: 0.000223 max mem: 14338 Epoch: [21/30] [ 250/5004] eta: 0:57:02 lr: 0.000017 loss: 1.582896 (1.641732) time: 0.718026 data: 0.000193 max mem: 14338 Epoch: [21/30] [ 300/5004] eta: 0:56:21 lr: 0.000017 loss: 1.693466 (1.643297) time: 0.715106 data: 0.000178 max mem: 14338 Epoch: [21/30] [ 350/5004] eta: 0:55:46 lr: 0.000017 loss: 1.534916 (1.632132) time: 0.715703 data: 0.000166 max mem: 14338 Epoch: [21/30] [ 400/5004] eta: 0:55:06 lr: 0.000017 loss: 1.585119 (1.629204) time: 0.711818 data: 0.000211 max mem: 14338 Epoch: [21/30] [ 450/5004] eta: 0:54:28 lr: 0.000017 loss: 1.658910 (1.635768) time: 0.716244 data: 0.000214 max mem: 14338 Epoch: [21/30] [ 500/5004] eta: 0:53:51 lr: 0.000017 loss: 1.700596 (1.633688) time: 0.711009 data: 0.000223 max mem: 14338 Epoch: [21/30] [ 550/5004] eta: 0:53:13 lr: 0.000017 loss: 1.492628 (1.628121) time: 0.711139 data: 0.000219 max mem: 14338 Epoch: [21/30] [ 600/5004] eta: 0:52:36 lr: 0.000017 loss: 1.582085 (1.623463) time: 0.721409 data: 0.000224 max mem: 14338 Epoch: [21/30] [ 650/5004] eta: 0:51:59 lr: 0.000017 loss: 1.637547 (1.623794) time: 0.717883 data: 0.000164 max mem: 14338 Epoch: [21/30] [ 700/5004] eta: 0:51:22 lr: 0.000017 loss: 1.449801 (1.617875) time: 0.715814 data: 0.000185 max mem: 14338 Epoch: [21/30] [ 750/5004] eta: 0:50:45 lr: 0.000017 loss: 1.725960 (1.622591) time: 0.711753 data: 0.000233 max mem: 14338 Epoch: [21/30] [ 800/5004] eta: 0:50:09 lr: 0.000017 loss: 1.582722 (1.620978) time: 0.713013 data: 0.000226 max mem: 14338 Epoch: [21/30] [ 850/5004] eta: 0:49:34 lr: 0.000017 loss: 1.680151 (1.624010) time: 0.713974 data: 0.000227 max mem: 14338 Epoch: [21/30] [ 900/5004] eta: 0:48:57 lr: 0.000016 loss: 1.501003 (1.621689) time: 0.711656 data: 0.000206 max mem: 14338 Epoch: [21/30] [ 950/5004] eta: 0:48:22 lr: 0.000016 loss: 1.522012 (1.625507) time: 0.714963 data: 0.000219 max mem: 14338 Epoch: [21/30] [1000/5004] eta: 0:47:46 lr: 0.000016 loss: 1.616468 (1.626017) time: 0.711099 data: 0.000174 max mem: 14338 Epoch: [21/30] [1050/5004] eta: 0:47:09 lr: 0.000016 loss: 1.433013 (1.623041) time: 0.720015 data: 0.000223 max mem: 14338 Epoch: [21/30] [1100/5004] eta: 0:46:33 lr: 0.000016 loss: 1.477006 (1.620162) time: 0.716618 data: 0.000243 max mem: 14338 Epoch: [21/30] [1150/5004] eta: 0:45:57 lr: 0.000016 loss: 1.641354 (1.618863) time: 0.714968 data: 0.000222 max mem: 14338 Epoch: [21/30] [1200/5004] eta: 0:45:21 lr: 0.000016 loss: 1.450208 (1.620580) time: 0.715371 data: 0.000221 max mem: 14338 Epoch: [21/30] [1250/5004] eta: 0:44:45 lr: 0.000016 loss: 1.500887 (1.621833) time: 0.712544 data: 0.000221 max mem: 14338 Epoch: [21/30] [1300/5004] eta: 0:44:09 lr: 0.000016 loss: 1.404936 (1.622570) time: 0.715383 data: 0.000183 max mem: 14338 Epoch: [21/30] [1350/5004] eta: 0:43:34 lr: 0.000016 loss: 1.525227 (1.621967) time: 0.712056 data: 0.000190 max mem: 14338 Epoch: [21/30] [1400/5004] eta: 0:42:58 lr: 0.000016 loss: 1.535869 (1.623752) time: 0.712886 data: 0.000211 max mem: 14338 Epoch: [21/30] [1450/5004] eta: 0:42:22 lr: 0.000016 loss: 1.569924 (1.622815) time: 0.717154 data: 0.000216 max mem: 14338 Epoch: [21/30] [1500/5004] eta: 0:41:46 lr: 0.000016 loss: 1.540996 (1.622143) time: 0.714843 data: 0.000232 max mem: 14338 Epoch: [21/30] [1550/5004] eta: 0:41:10 lr: 0.000016 loss: 1.585377 (1.620646) time: 0.716888 data: 0.000197 max mem: 14338 Epoch: [21/30] [1600/5004] eta: 0:40:35 lr: 0.000016 loss: 1.477860 (1.621172) time: 0.718413 data: 0.000222 max mem: 14338 Epoch: [21/30] [1650/5004] eta: 0:39:59 lr: 0.000016 loss: 1.596159 (1.622593) time: 0.722387 data: 0.000171 max mem: 14338 Epoch: [21/30] [1700/5004] eta: 0:39:23 lr: 0.000016 loss: 1.614119 (1.622492) time: 0.710856 data: 0.000168 max mem: 14338 Epoch: [21/30] [1750/5004] eta: 0:38:47 lr: 0.000016 loss: 1.563056 (1.620527) time: 0.710560 data: 0.000230 max mem: 14338 Epoch: [21/30] [1800/5004] eta: 0:38:11 lr: 0.000016 loss: 1.486767 (1.621687) time: 0.719350 data: 0.000232 max mem: 14338 Epoch: [21/30] [1850/5004] eta: 0:37:36 lr: 0.000016 loss: 1.350244 (1.620880) time: 0.714405 data: 0.000232 max mem: 14338 Epoch: [21/30] [1900/5004] eta: 0:37:00 lr: 0.000016 loss: 1.526952 (1.619830) time: 0.716796 data: 0.000225 max mem: 14338 Epoch: [21/30] [1950/5004] eta: 0:36:24 lr: 0.000016 loss: 1.563543 (1.619409) time: 0.711790 data: 0.000221 max mem: 14338 Epoch: [21/30] [2000/5004] eta: 0:35:48 lr: 0.000016 loss: 1.733942 (1.619863) time: 0.717415 data: 0.000171 max mem: 14338 Epoch: [21/30] [2050/5004] eta: 0:35:13 lr: 0.000016 loss: 1.733416 (1.620547) time: 0.718150 data: 0.000165 max mem: 14338 Epoch: [21/30] [2100/5004] eta: 0:34:37 lr: 0.000016 loss: 1.538707 (1.620808) time: 0.715392 data: 0.000228 max mem: 14338 Epoch: [21/30] [2150/5004] eta: 0:34:01 lr: 0.000016 loss: 1.699565 (1.620848) time: 0.712524 data: 0.000235 max mem: 14338 Epoch: [21/30] [2200/5004] eta: 0:33:25 lr: 0.000016 loss: 1.493796 (1.621743) time: 0.715744 data: 0.000193 max mem: 14338 Epoch: [21/30] [2250/5004] eta: 0:32:49 lr: 0.000016 loss: 1.391433 (1.621413) time: 0.712095 data: 0.000232 max mem: 14338 Epoch: [21/30] [2300/5004] eta: 0:32:13 lr: 0.000016 loss: 1.556767 (1.621783) time: 0.709121 data: 0.000215 max mem: 14338 Epoch: [21/30] [2350/5004] eta: 0:31:37 lr: 0.000016 loss: 1.688785 (1.622666) time: 0.709322 data: 0.000163 max mem: 14338 Epoch: [21/30] [2400/5004] eta: 0:31:01 lr: 0.000015 loss: 1.541547 (1.622270) time: 0.712451 data: 0.000213 max mem: 14338 Epoch: [21/30] [2450/5004] eta: 0:30:26 lr: 0.000015 loss: 1.785671 (1.623330) time: 0.716867 data: 0.000220 max mem: 14338 Epoch: [21/30] [2500/5004] eta: 0:29:50 lr: 0.000015 loss: 1.418679 (1.622233) time: 0.716217 data: 0.000234 max mem: 14338 Epoch: [21/30] [2550/5004] eta: 0:29:14 lr: 0.000015 loss: 1.619157 (1.622420) time: 0.719793 data: 0.000237 max mem: 14338 Epoch: [21/30] [2600/5004] eta: 0:28:39 lr: 0.000015 loss: 1.728344 (1.624302) time: 0.718465 data: 0.000237 max mem: 14338 Epoch: [21/30] [2650/5004] eta: 0:28:03 lr: 0.000015 loss: 1.561358 (1.626181) time: 0.713433 data: 0.000152 max mem: 14338 Epoch: [21/30] [2700/5004] eta: 0:27:27 lr: 0.000015 loss: 1.444173 (1.625357) time: 0.713517 data: 0.000157 max mem: 14338 Epoch: [21/30] [2750/5004] eta: 0:26:51 lr: 0.000015 loss: 1.564238 (1.626719) time: 0.708928 data: 0.000224 max mem: 14338 Epoch: [21/30] [2800/5004] eta: 0:26:15 lr: 0.000015 loss: 1.699754 (1.627398) time: 0.717307 data: 0.000230 max mem: 14338 Epoch: [21/30] [2850/5004] eta: 0:25:40 lr: 0.000015 loss: 1.480378 (1.626918) time: 0.720999 data: 0.000194 max mem: 14338 Epoch: [21/30] [2900/5004] eta: 0:25:04 lr: 0.000015 loss: 1.490822 (1.625892) time: 0.709640 data: 0.000206 max mem: 14338 Epoch: [21/30] [2950/5004] eta: 0:24:28 lr: 0.000015 loss: 1.553315 (1.626060) time: 0.714468 data: 0.000220 max mem: 14338 Epoch: [21/30] [3000/5004] eta: 0:23:52 lr: 0.000015 loss: 1.622499 (1.626293) time: 0.718536 data: 0.000178 max mem: 14338 Epoch: [21/30] [3050/5004] eta: 0:23:17 lr: 0.000015 loss: 1.535881 (1.627204) time: 0.719399 data: 0.000170 max mem: 14338 Epoch: [21/30] [3100/5004] eta: 0:22:41 lr: 0.000015 loss: 1.560348 (1.627032) time: 0.711990 data: 0.000231 max mem: 14338 Epoch: [21/30] [3150/5004] eta: 0:22:05 lr: 0.000015 loss: 1.550076 (1.626118) time: 0.709512 data: 0.000221 max mem: 14338 Epoch: [21/30] [3200/5004] eta: 0:21:29 lr: 0.000015 loss: 1.548186 (1.626252) time: 0.718938 data: 0.000230 max mem: 14338 Epoch: [21/30] [3250/5004] eta: 0:20:54 lr: 0.000015 loss: 1.485461 (1.626114) time: 0.711478 data: 0.000225 max mem: 14338 Epoch: [21/30] [3300/5004] eta: 0:20:18 lr: 0.000015 loss: 1.800777 (1.627290) time: 0.711734 data: 0.000214 max mem: 14338 Epoch: [21/30] [3350/5004] eta: 0:19:42 lr: 0.000015 loss: 1.414500 (1.626156) time: 0.717032 data: 0.000163 max mem: 14338 Epoch: [21/30] [3400/5004] eta: 0:19:06 lr: 0.000015 loss: 1.566016 (1.625445) time: 0.713848 data: 0.000178 max mem: 14338 Epoch: [21/30] [3450/5004] eta: 0:18:31 lr: 0.000015 loss: 1.568759 (1.625170) time: 0.721027 data: 0.000219 max mem: 14338 Epoch: [21/30] [3500/5004] eta: 0:17:55 lr: 0.000015 loss: 1.749899 (1.626810) time: 0.721775 data: 0.000191 max mem: 14338 Epoch: [21/30] [3550/5004] eta: 0:17:19 lr: 0.000015 loss: 1.581519 (1.626245) time: 0.717255 data: 0.000211 max mem: 14338 Epoch: [21/30] [3600/5004] eta: 0:16:43 lr: 0.000015 loss: 1.542310 (1.626037) time: 0.712046 data: 0.000224 max mem: 14338 Epoch: [21/30] [3650/5004] eta: 0:16:08 lr: 0.000015 loss: 1.719569 (1.627250) time: 0.712534 data: 0.000209 max mem: 14338 Epoch: [21/30] [3700/5004] eta: 0:15:32 lr: 0.000015 loss: 1.689745 (1.627464) time: 0.709969 data: 0.000166 max mem: 14338 Epoch: [21/30] [3750/5004] eta: 0:14:56 lr: 0.000015 loss: 1.516097 (1.627348) time: 0.713430 data: 0.000212 max mem: 14338 Epoch: [21/30] [3800/5004] eta: 0:14:20 lr: 0.000015 loss: 1.586220 (1.627761) time: 0.711574 data: 0.000204 max mem: 14338 Epoch: [21/30] [3850/5004] eta: 0:13:45 lr: 0.000015 loss: 1.546512 (1.626609) time: 0.717200 data: 0.000212 max mem: 14338 Epoch: [21/30] [3900/5004] eta: 0:13:09 lr: 0.000015 loss: 1.516649 (1.626089) time: 0.715925 data: 0.000216 max mem: 14338 Epoch: [21/30] [3950/5004] eta: 0:12:33 lr: 0.000015 loss: 1.535042 (1.626101) time: 0.724002 data: 0.000224 max mem: 14338 Epoch: [21/30] [4000/5004] eta: 0:11:57 lr: 0.000014 loss: 1.572721 (1.626068) time: 0.716872 data: 0.000158 max mem: 14338 Epoch: [21/30] [4050/5004] eta: 0:11:22 lr: 0.000014 loss: 1.620965 (1.626715) time: 0.714378 data: 0.000160 max mem: 14338 Epoch: [21/30] [4100/5004] eta: 0:10:46 lr: 0.000014 loss: 1.713662 (1.626466) time: 0.709290 data: 0.000228 max mem: 14338 Epoch: [21/30] [4150/5004] eta: 0:10:10 lr: 0.000014 loss: 1.575061 (1.626255) time: 0.710270 data: 0.000191 max mem: 14338 Epoch: [21/30] [4200/5004] eta: 0:09:34 lr: 0.000014 loss: 1.545651 (1.626037) time: 0.715554 data: 0.000209 max mem: 14338 Epoch: [21/30] [4250/5004] eta: 0:08:59 lr: 0.000014 loss: 1.660849 (1.625468) time: 0.720039 data: 0.000213 max mem: 14338 Epoch: [21/30] [4300/5004] eta: 0:08:23 lr: 0.000014 loss: 1.533781 (1.625561) time: 0.713910 data: 0.000221 max mem: 14338 Epoch: [21/30] [4350/5004] eta: 0:07:47 lr: 0.000014 loss: 1.600201 (1.625644) time: 0.714960 data: 0.000171 max mem: 14338 Epoch: [21/30] [4400/5004] eta: 0:07:11 lr: 0.000014 loss: 1.616240 (1.625854) time: 0.723387 data: 0.000157 max mem: 14338 Epoch: [21/30] [4450/5004] eta: 0:06:36 lr: 0.000014 loss: 1.705767 (1.626386) time: 0.712977 data: 0.000210 max mem: 14338 Epoch: [21/30] [4500/5004] eta: 0:06:00 lr: 0.000014 loss: 1.577862 (1.626622) time: 0.715800 data: 0.000198 max mem: 14338 Epoch: [21/30] [4550/5004] eta: 0:05:24 lr: 0.000014 loss: 1.420822 (1.626271) time: 0.711851 data: 0.000234 max mem: 14338 Epoch: [21/30] [4600/5004] eta: 0:04:48 lr: 0.000014 loss: 1.557591 (1.626357) time: 0.714123 data: 0.000215 max mem: 14338 Epoch: [21/30] [4650/5004] eta: 0:04:13 lr: 0.000014 loss: 1.455956 (1.625949) time: 0.715165 data: 0.000220 max mem: 14338 Epoch: [21/30] [4700/5004] eta: 0:03:37 lr: 0.000014 loss: 1.536869 (1.626078) time: 0.708021 data: 0.000153 max mem: 14338 Epoch: [21/30] [4750/5004] eta: 0:03:01 lr: 0.000014 loss: 1.532426 (1.626191) time: 0.709971 data: 0.000172 max mem: 14338 Epoch: [21/30] [4800/5004] eta: 0:02:25 lr: 0.000014 loss: 1.531802 (1.625964) time: 0.717550 data: 0.000182 max mem: 14338 Epoch: [21/30] [4850/5004] eta: 0:01:50 lr: 0.000014 loss: 1.529436 (1.625707) time: 0.719327 data: 0.000219 max mem: 14338 Epoch: [21/30] [4900/5004] eta: 0:01:14 lr: 0.000014 loss: 1.405619 (1.624829) time: 0.724979 data: 0.000218 max mem: 14338 Epoch: [21/30] [4950/5004] eta: 0:00:38 lr: 0.000014 loss: 1.548325 (1.624575) time: 0.714333 data: 0.000209 max mem: 14338 Epoch: [21/30] [5000/5004] eta: 0:00:02 lr: 0.000014 loss: 1.739262 (1.625506) time: 0.708718 data: 0.000834 max mem: 14338 Epoch: [21/30] [5003/5004] eta: 0:00:00 lr: 0.000014 loss: 1.739262 (1.625620) time: 0.706779 data: 0.000821 max mem: 14338 Epoch: [21/30] Total time: 0:59:38 (0.715152 s / it) Averaged stats: lr: 0.000014 loss: 1.739262 (1.629532) Test: [ 0/196] eta: 0:05:04 loss: 0.307317 (0.307317) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.552116 data: 1.155786 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 0.440833 (0.548173) acc1: 87.500000 (85.227273) acc5: 100.000000 (98.863636) time: 0.404194 data: 0.105198 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.555346 (0.546426) acc1: 87.500000 (85.714286) acc5: 100.000000 (98.214286) time: 0.287971 data: 0.000123 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.516585 (0.519600) acc1: 87.500000 (87.298387) acc5: 100.000000 (98.387097) time: 0.286467 data: 0.000122 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.440210 (0.527249) acc1: 87.500000 (87.042683) acc5: 100.000000 (98.170732) time: 0.286683 data: 0.000144 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.458711 (0.554602) acc1: 87.500000 (86.887255) acc5: 100.000000 (97.671569) time: 0.287429 data: 0.000134 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.621686 (0.582557) acc1: 87.500000 (86.168033) acc5: 93.750000 (97.540984) time: 0.287901 data: 0.000137 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.649784 (0.597605) acc1: 81.250000 (85.739437) acc5: 100.000000 (97.623239) time: 0.287645 data: 0.000164 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.512945 (0.596863) acc1: 87.500000 (85.725309) acc5: 100.000000 (97.762346) time: 0.286981 data: 0.000150 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.520380 (0.619805) acc1: 81.250000 (85.302198) acc5: 100.000000 (97.527473) time: 0.286913 data: 0.000137 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.594408 (0.608906) acc1: 81.250000 (85.457921) acc5: 100.000000 (97.710396) time: 0.287171 data: 0.000139 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.511359 (0.597501) acc1: 87.500000 (85.472973) acc5: 100.000000 (97.860360) time: 0.286864 data: 0.000131 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.496346 (0.593244) acc1: 87.500000 (85.537190) acc5: 100.000000 (97.882231) time: 0.286575 data: 0.000139 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.628433 (0.609927) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.805344) time: 0.293945 data: 0.000140 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.673042 (0.608993) acc1: 87.500000 (85.372340) acc5: 100.000000 (97.783688) time: 0.294466 data: 0.000148 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.547803 (0.617639) acc1: 87.500000 (85.223510) acc5: 100.000000 (97.806291) time: 0.287543 data: 0.000158 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.514553 (0.621993) acc1: 81.250000 (85.170807) acc5: 100.000000 (97.787267) time: 0.287955 data: 0.000139 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.511093 (0.618939) acc1: 81.250000 (85.270468) acc5: 100.000000 (97.843567) time: 0.287167 data: 0.000130 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.476793 (0.618211) acc1: 87.500000 (85.186464) acc5: 100.000000 (97.893646) time: 0.286227 data: 0.000150 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.348849 (0.606425) acc1: 87.500000 (85.373037) acc5: 100.000000 (97.905759) time: 0.283812 data: 0.000122 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.470046 (0.616684) acc1: 87.500000 (85.280000) acc5: 100.000000 (97.824000) time: 0.274117 data: 0.000109 max mem: 14338 Test: Total time: 0:00:57 (0.294498 s / it) * Acc@1 85.134 Acc@5 97.306 loss 0.632 Max accuracy: 85.25% Epoch: [22/30] [ 0/5004] eta: 2:41:47 lr: 0.000014 loss: 1.405092 (1.405092) time: 1.939871 data: 1.209541 max mem: 14338 Epoch: [22/30] [ 50/5004] eta: 1:01:02 lr: 0.000014 loss: 1.640937 (1.664370) time: 0.713522 data: 0.000204 max mem: 14338 Epoch: [22/30] [ 100/5004] eta: 0:59:21 lr: 0.000014 loss: 1.780771 (1.673080) time: 0.711392 data: 0.000216 max mem: 14338 Epoch: [22/30] [ 150/5004] eta: 0:58:23 lr: 0.000014 loss: 1.518356 (1.668900) time: 0.710464 data: 0.000202 max mem: 14338 Epoch: [22/30] [ 200/5004] eta: 0:57:41 lr: 0.000014 loss: 1.522069 (1.643725) time: 0.715150 data: 0.000199 max mem: 14338 Epoch: [22/30] [ 250/5004] eta: 0:56:58 lr: 0.000014 loss: 1.484897 (1.629150) time: 0.713639 data: 0.000184 max mem: 14338 Epoch: [22/30] [ 300/5004] eta: 0:56:17 lr: 0.000014 loss: 1.511395 (1.625040) time: 0.711999 data: 0.000172 max mem: 14338 Epoch: [22/30] [ 350/5004] eta: 0:55:40 lr: 0.000014 loss: 1.440620 (1.619634) time: 0.718008 data: 0.000170 max mem: 14338 Epoch: [22/30] [ 400/5004] eta: 0:55:04 lr: 0.000014 loss: 1.530140 (1.624041) time: 0.720560 data: 0.000215 max mem: 14338 Epoch: [22/30] [ 450/5004] eta: 0:54:25 lr: 0.000014 loss: 1.563061 (1.625430) time: 0.712728 data: 0.000217 max mem: 14338 Epoch: [22/30] [ 500/5004] eta: 0:53:49 lr: 0.000014 loss: 1.454160 (1.621249) time: 0.713545 data: 0.000228 max mem: 14338 Epoch: [22/30] [ 550/5004] eta: 0:53:12 lr: 0.000014 loss: 1.549244 (1.621907) time: 0.712718 data: 0.000232 max mem: 14338 Epoch: [22/30] [ 600/5004] eta: 0:52:35 lr: 0.000013 loss: 1.793501 (1.628636) time: 0.714306 data: 0.000240 max mem: 14338 Epoch: [22/30] [ 650/5004] eta: 0:51:59 lr: 0.000013 loss: 1.575053 (1.628344) time: 0.712353 data: 0.000165 max mem: 14338 Epoch: [22/30] [ 700/5004] eta: 0:51:22 lr: 0.000013 loss: 1.530926 (1.626620) time: 0.713827 data: 0.000173 max mem: 14338 Epoch: [22/30] [ 750/5004] eta: 0:50:46 lr: 0.000013 loss: 1.777105 (1.633240) time: 0.715577 data: 0.000223 max mem: 14338 Epoch: [22/30] [ 800/5004] eta: 0:50:10 lr: 0.000013 loss: 1.533013 (1.631054) time: 0.722940 data: 0.000220 max mem: 14338 Epoch: [22/30] [ 850/5004] eta: 0:49:34 lr: 0.000013 loss: 1.563165 (1.629135) time: 0.718178 data: 0.000222 max mem: 14338 Epoch: [22/30] [ 900/5004] eta: 0:48:58 lr: 0.000013 loss: 1.575510 (1.625712) time: 0.715098 data: 0.000184 max mem: 14338 Epoch: [22/30] [ 950/5004] eta: 0:48:22 lr: 0.000013 loss: 1.334507 (1.625066) time: 0.712923 data: 0.000246 max mem: 14338 Epoch: [22/30] [1000/5004] eta: 0:47:46 lr: 0.000013 loss: 1.440060 (1.625170) time: 0.713320 data: 0.000169 max mem: 14338 Epoch: [22/30] [1050/5004] eta: 0:47:10 lr: 0.000013 loss: 1.643813 (1.623144) time: 0.712459 data: 0.000232 max mem: 14338 Epoch: [22/30] [1100/5004] eta: 0:46:34 lr: 0.000013 loss: 1.558981 (1.622343) time: 0.713421 data: 0.000222 max mem: 14338 Epoch: [22/30] [1150/5004] eta: 0:45:58 lr: 0.000013 loss: 1.663633 (1.620813) time: 0.710958 data: 0.000228 max mem: 14338 Epoch: [22/30] [1200/5004] eta: 0:45:22 lr: 0.000013 loss: 1.605021 (1.622328) time: 0.716610 data: 0.000227 max mem: 14338 Epoch: [22/30] [1250/5004] eta: 0:44:46 lr: 0.000013 loss: 1.716916 (1.625711) time: 0.716797 data: 0.000214 max mem: 14338 Epoch: [22/30] [1300/5004] eta: 0:44:10 lr: 0.000013 loss: 1.414499 (1.621851) time: 0.720781 data: 0.000181 max mem: 14338 Epoch: [22/30] [1350/5004] eta: 0:43:35 lr: 0.000013 loss: 1.452060 (1.621196) time: 0.716862 data: 0.000178 max mem: 14338 Epoch: [22/30] [1400/5004] eta: 0:42:59 lr: 0.000013 loss: 1.438941 (1.619338) time: 0.713558 data: 0.000220 max mem: 14338 Epoch: [22/30] [1450/5004] eta: 0:42:23 lr: 0.000013 loss: 1.586412 (1.620350) time: 0.716889 data: 0.000221 max mem: 14338 Epoch: [22/30] [1500/5004] eta: 0:41:47 lr: 0.000013 loss: 1.481344 (1.621111) time: 0.713863 data: 0.000222 max mem: 14338 Epoch: [22/30] [1550/5004] eta: 0:41:11 lr: 0.000013 loss: 1.531241 (1.619789) time: 0.709948 data: 0.000201 max mem: 14338 Epoch: [22/30] [1600/5004] eta: 0:40:35 lr: 0.000013 loss: 1.577525 (1.620902) time: 0.711986 data: 0.000229 max mem: 14338 Epoch: [22/30] [1650/5004] eta: 0:39:59 lr: 0.000013 loss: 1.453577 (1.619703) time: 0.713572 data: 0.000180 max mem: 14338 Epoch: [22/30] [1700/5004] eta: 0:39:23 lr: 0.000013 loss: 1.663758 (1.620332) time: 0.713661 data: 0.000174 max mem: 14338 Epoch: [22/30] [1750/5004] eta: 0:38:47 lr: 0.000013 loss: 1.464871 (1.619234) time: 0.711727 data: 0.000225 max mem: 14338 Epoch: [22/30] [1800/5004] eta: 0:38:11 lr: 0.000013 loss: 1.539359 (1.620398) time: 0.722477 data: 0.000241 max mem: 14338 Epoch: [22/30] [1850/5004] eta: 0:37:35 lr: 0.000013 loss: 1.548200 (1.620573) time: 0.713094 data: 0.000214 max mem: 14338 Epoch: [22/30] [1900/5004] eta: 0:36:59 lr: 0.000013 loss: 1.497726 (1.620562) time: 0.712451 data: 0.000241 max mem: 14338 Epoch: [22/30] [1950/5004] eta: 0:36:23 lr: 0.000013 loss: 1.532669 (1.621567) time: 0.711914 data: 0.000221 max mem: 14338 Epoch: [22/30] [2000/5004] eta: 0:35:48 lr: 0.000013 loss: 1.524411 (1.621184) time: 0.713540 data: 0.000174 max mem: 14338 Epoch: [22/30] [2050/5004] eta: 0:35:12 lr: 0.000013 loss: 1.513627 (1.621242) time: 0.711386 data: 0.000155 max mem: 14338 Epoch: [22/30] [2100/5004] eta: 0:34:36 lr: 0.000013 loss: 1.629565 (1.621684) time: 0.710708 data: 0.000230 max mem: 14338 Epoch: [22/30] [2150/5004] eta: 0:34:00 lr: 0.000013 loss: 1.551020 (1.620006) time: 0.713050 data: 0.000232 max mem: 14338 Epoch: [22/30] [2200/5004] eta: 0:33:24 lr: 0.000013 loss: 1.577996 (1.618779) time: 0.720606 data: 0.000201 max mem: 14338 Epoch: [22/30] [2250/5004] eta: 0:32:49 lr: 0.000012 loss: 1.599802 (1.620693) time: 0.719037 data: 0.000227 max mem: 14338 Epoch: [22/30] [2300/5004] eta: 0:32:13 lr: 0.000012 loss: 1.489405 (1.621270) time: 0.713099 data: 0.000230 max mem: 14338 Epoch: [22/30] [2350/5004] eta: 0:31:37 lr: 0.000012 loss: 1.518259 (1.621622) time: 0.713810 data: 0.000158 max mem: 14338 Epoch: [22/30] [2400/5004] eta: 0:31:02 lr: 0.000012 loss: 1.648176 (1.622517) time: 0.719720 data: 0.000238 max mem: 14338 Epoch: [22/30] [2450/5004] eta: 0:30:26 lr: 0.000012 loss: 1.434991 (1.621698) time: 0.721869 data: 0.000210 max mem: 14338 Epoch: [22/30] [2500/5004] eta: 0:29:50 lr: 0.000012 loss: 1.730615 (1.623751) time: 0.713284 data: 0.000214 max mem: 14338 Epoch: [22/30] [2550/5004] eta: 0:29:14 lr: 0.000012 loss: 1.646644 (1.623571) time: 0.710393 data: 0.000217 max mem: 14338 Epoch: [22/30] [2600/5004] eta: 0:28:38 lr: 0.000012 loss: 1.502331 (1.622924) time: 0.713970 data: 0.000208 max mem: 14338 Epoch: [22/30] [2650/5004] eta: 0:28:03 lr: 0.000012 loss: 1.483042 (1.622168) time: 0.723363 data: 0.000181 max mem: 14338 Epoch: [22/30] [2700/5004] eta: 0:27:27 lr: 0.000012 loss: 1.644965 (1.622295) time: 0.715085 data: 0.000171 max mem: 14338 Epoch: [22/30] [2750/5004] eta: 0:26:51 lr: 0.000012 loss: 1.645727 (1.622994) time: 0.716913 data: 0.000218 max mem: 14338 Epoch: [22/30] [2800/5004] eta: 0:26:15 lr: 0.000012 loss: 1.518186 (1.621883) time: 0.712379 data: 0.000207 max mem: 14338 Epoch: [22/30] [2850/5004] eta: 0:25:40 lr: 0.000012 loss: 1.527345 (1.621896) time: 0.718720 data: 0.000191 max mem: 14338 Epoch: [22/30] [2900/5004] eta: 0:25:04 lr: 0.000012 loss: 1.649114 (1.623093) time: 0.717847 data: 0.000227 max mem: 14338 Epoch: [22/30] [2950/5004] eta: 0:24:28 lr: 0.000012 loss: 1.532813 (1.622450) time: 0.711791 data: 0.000216 max mem: 14338 Epoch: [22/30] [3000/5004] eta: 0:23:53 lr: 0.000012 loss: 1.490536 (1.622213) time: 0.714430 data: 0.000170 max mem: 14338 Epoch: [22/30] [3050/5004] eta: 0:23:17 lr: 0.000012 loss: 1.494580 (1.620862) time: 0.715669 data: 0.000157 max mem: 14338 Epoch: [22/30] [3100/5004] eta: 0:22:41 lr: 0.000012 loss: 1.624358 (1.621301) time: 0.712309 data: 0.000210 max mem: 14338 Epoch: [22/30] [3150/5004] eta: 0:22:05 lr: 0.000012 loss: 1.535194 (1.621238) time: 0.720023 data: 0.000241 max mem: 14338 Epoch: [22/30] [3200/5004] eta: 0:21:29 lr: 0.000012 loss: 1.533412 (1.620259) time: 0.719983 data: 0.000233 max mem: 14338 Epoch: [22/30] [3250/5004] eta: 0:20:54 lr: 0.000012 loss: 1.627658 (1.620838) time: 0.712806 data: 0.000236 max mem: 14338 Epoch: [22/30] [3300/5004] eta: 0:20:18 lr: 0.000012 loss: 1.671725 (1.620745) time: 0.709468 data: 0.000216 max mem: 14338 Epoch: [22/30] [3350/5004] eta: 0:19:42 lr: 0.000012 loss: 1.564071 (1.621027) time: 0.709741 data: 0.000173 max mem: 14338 Epoch: [22/30] [3400/5004] eta: 0:19:06 lr: 0.000012 loss: 1.763263 (1.622357) time: 0.716406 data: 0.000160 max mem: 14338 Epoch: [22/30] [3450/5004] eta: 0:18:30 lr: 0.000012 loss: 1.548173 (1.621546) time: 0.710712 data: 0.000217 max mem: 14338 Epoch: [22/30] [3500/5004] eta: 0:17:55 lr: 0.000012 loss: 1.486045 (1.621043) time: 0.711385 data: 0.000202 max mem: 14338 Epoch: [22/30] [3550/5004] eta: 0:17:19 lr: 0.000012 loss: 1.565935 (1.620587) time: 0.713061 data: 0.000211 max mem: 14338 Epoch: [22/30] [3600/5004] eta: 0:16:43 lr: 0.000012 loss: 1.574245 (1.620688) time: 0.722877 data: 0.000221 max mem: 14338 Epoch: [22/30] [3650/5004] eta: 0:16:07 lr: 0.000012 loss: 1.734912 (1.621723) time: 0.721899 data: 0.000219 max mem: 14338 Epoch: [22/30] [3700/5004] eta: 0:15:32 lr: 0.000012 loss: 1.503548 (1.621010) time: 0.717938 data: 0.000192 max mem: 14338 Epoch: [22/30] [3750/5004] eta: 0:14:56 lr: 0.000012 loss: 1.594507 (1.621410) time: 0.709734 data: 0.000244 max mem: 14338 Epoch: [22/30] [3800/5004] eta: 0:14:20 lr: 0.000012 loss: 1.584854 (1.621133) time: 0.712037 data: 0.000227 max mem: 14338 Epoch: [22/30] [3850/5004] eta: 0:13:44 lr: 0.000012 loss: 1.625762 (1.620706) time: 0.714457 data: 0.000217 max mem: 14338 Epoch: [22/30] [3900/5004] eta: 0:13:09 lr: 0.000012 loss: 1.592298 (1.620537) time: 0.710258 data: 0.000228 max mem: 14338 Epoch: [22/30] [3950/5004] eta: 0:12:33 lr: 0.000012 loss: 1.672749 (1.620985) time: 0.711152 data: 0.000233 max mem: 14338 Epoch: [22/30] [4000/5004] eta: 0:11:57 lr: 0.000011 loss: 1.573934 (1.620792) time: 0.719438 data: 0.000184 max mem: 14338 Epoch: [22/30] [4050/5004] eta: 0:11:21 lr: 0.000011 loss: 1.638363 (1.620440) time: 0.719443 data: 0.000166 max mem: 14338 Epoch: [22/30] [4100/5004] eta: 0:10:46 lr: 0.000011 loss: 1.538735 (1.619602) time: 0.713483 data: 0.000234 max mem: 14338 Epoch: [22/30] [4150/5004] eta: 0:10:10 lr: 0.000011 loss: 1.611120 (1.620408) time: 0.714417 data: 0.000181 max mem: 14338 Epoch: [22/30] [4200/5004] eta: 0:09:34 lr: 0.000011 loss: 1.707573 (1.621091) time: 0.713990 data: 0.000225 max mem: 14338 Epoch: [22/30] [4250/5004] eta: 0:08:58 lr: 0.000011 loss: 1.538680 (1.620658) time: 0.717021 data: 0.000216 max mem: 14338 Epoch: [22/30] [4300/5004] eta: 0:08:23 lr: 0.000011 loss: 1.523374 (1.620884) time: 0.709236 data: 0.000225 max mem: 14338 Epoch: [22/30] [4350/5004] eta: 0:07:47 lr: 0.000011 loss: 1.527899 (1.620697) time: 0.710134 data: 0.000167 max mem: 14338 Epoch: [22/30] [4400/5004] eta: 0:07:11 lr: 0.000011 loss: 1.486141 (1.620796) time: 0.713442 data: 0.000158 max mem: 14338 Epoch: [22/30] [4450/5004] eta: 0:06:35 lr: 0.000011 loss: 1.531359 (1.620628) time: 0.716484 data: 0.000220 max mem: 14338 Epoch: [22/30] [4500/5004] eta: 0:06:00 lr: 0.000011 loss: 1.586363 (1.621219) time: 0.713796 data: 0.000254 max mem: 14338 Epoch: [22/30] [4550/5004] eta: 0:05:24 lr: 0.000011 loss: 1.523715 (1.621163) time: 0.714026 data: 0.000212 max mem: 14338 Epoch: [22/30] [4600/5004] eta: 0:04:48 lr: 0.000011 loss: 1.650017 (1.621150) time: 0.719044 data: 0.000222 max mem: 14338 Epoch: [22/30] [4650/5004] eta: 0:04:13 lr: 0.000011 loss: 1.698637 (1.622107) time: 0.717765 data: 0.000209 max mem: 14338 Epoch: [22/30] [4700/5004] eta: 0:03:37 lr: 0.000011 loss: 1.597579 (1.622976) time: 0.712097 data: 0.000160 max mem: 14338 Epoch: [22/30] [4750/5004] eta: 0:03:01 lr: 0.000011 loss: 1.600907 (1.622842) time: 0.713332 data: 0.000156 max mem: 14338 Epoch: [22/30] [4800/5004] eta: 0:02:25 lr: 0.000011 loss: 1.402328 (1.622510) time: 0.717515 data: 0.000193 max mem: 14338 Epoch: [22/30] [4850/5004] eta: 0:01:50 lr: 0.000011 loss: 1.551866 (1.622605) time: 0.716745 data: 0.000227 max mem: 14338 Epoch: [22/30] [4900/5004] eta: 0:01:14 lr: 0.000011 loss: 1.506683 (1.621929) time: 0.710480 data: 0.000233 max mem: 14338 Epoch: [22/30] [4950/5004] eta: 0:00:38 lr: 0.000011 loss: 1.400850 (1.620786) time: 0.715842 data: 0.000238 max mem: 14338 Epoch: [22/30] [5000/5004] eta: 0:00:02 lr: 0.000011 loss: 1.567620 (1.621714) time: 0.710447 data: 0.000851 max mem: 14338 Epoch: [22/30] [5003/5004] eta: 0:00:00 lr: 0.000011 loss: 1.530665 (1.621645) time: 0.707179 data: 0.000840 max mem: 14338 Epoch: [22/30] Total time: 0:59:37 (0.714890 s / it) Averaged stats: lr: 0.000011 loss: 1.530665 (1.626576) Test: [ 0/196] eta: 0:05:06 loss: 0.293124 (0.293124) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.563656 data: 1.216251 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.449516 (0.544585) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.402809 data: 0.110691 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.523725 (0.537191) acc1: 87.500000 (85.119048) acc5: 100.000000 (98.511905) time: 0.286749 data: 0.000121 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.508875 (0.513962) acc1: 87.500000 (86.693548) acc5: 100.000000 (98.588710) time: 0.286815 data: 0.000110 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.440609 (0.522665) acc1: 87.500000 (86.432927) acc5: 100.000000 (98.323171) time: 0.290077 data: 0.000129 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.466339 (0.550035) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.794118) time: 0.290352 data: 0.000126 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.630333 (0.577995) acc1: 87.500000 (85.963115) acc5: 93.750000 (97.643443) time: 0.293042 data: 0.000122 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.653913 (0.595584) acc1: 81.250000 (85.739437) acc5: 100.000000 (97.799296) time: 0.292705 data: 0.000121 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.504845 (0.595143) acc1: 87.500000 (85.725309) acc5: 100.000000 (97.839506) time: 0.286772 data: 0.000110 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.551497 (0.618601) acc1: 81.250000 (85.233516) acc5: 100.000000 (97.596154) time: 0.292571 data: 0.000134 max mem: 14338 Test: [100/196] eta: 0:00:29 loss: 0.575272 (0.607302) acc1: 81.250000 (85.334158) acc5: 100.000000 (97.772277) time: 0.292850 data: 0.000137 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.530441 (0.595396) acc1: 87.500000 (85.360360) acc5: 100.000000 (97.860360) time: 0.286602 data: 0.000113 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.501366 (0.591085) acc1: 87.500000 (85.485537) acc5: 100.000000 (97.882231) time: 0.286215 data: 0.000122 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.618412 (0.607133) acc1: 87.500000 (85.162214) acc5: 100.000000 (97.853053) time: 0.286253 data: 0.000139 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.655050 (0.607041) acc1: 81.250000 (85.283688) acc5: 100.000000 (97.828014) time: 0.286163 data: 0.000139 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.566334 (0.616172) acc1: 87.500000 (85.140728) acc5: 100.000000 (97.847682) time: 0.286465 data: 0.000123 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.532758 (0.620723) acc1: 81.250000 (85.093168) acc5: 100.000000 (97.826087) time: 0.286908 data: 0.000111 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.490421 (0.617872) acc1: 81.250000 (85.197368) acc5: 100.000000 (97.843567) time: 0.287413 data: 0.000135 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.490421 (0.617720) acc1: 87.500000 (85.048343) acc5: 100.000000 (97.859116) time: 0.286969 data: 0.000157 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.371651 (0.605860) acc1: 87.500000 (85.274869) acc5: 100.000000 (97.873037) time: 0.283977 data: 0.000117 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.490000 (0.615917) acc1: 87.500000 (85.184000) acc5: 100.000000 (97.792000) time: 0.274023 data: 0.000107 max mem: 14338 Test: Total time: 0:00:57 (0.295241 s / it) * Acc@1 85.156 Acc@5 97.314 loss 0.631 Max accuracy: 85.25% Epoch: [23/30] [ 0/5004] eta: 2:42:01 lr: 0.000011 loss: 2.495364 (2.495364) time: 1.942751 data: 1.218248 max mem: 14338 Epoch: [23/30] [ 50/5004] eta: 1:00:54 lr: 0.000011 loss: 1.396925 (1.575082) time: 0.718397 data: 0.000210 max mem: 14338 Epoch: [23/30] [ 100/5004] eta: 0:59:19 lr: 0.000011 loss: 1.542391 (1.602271) time: 0.714709 data: 0.000243 max mem: 14338 Epoch: [23/30] [ 150/5004] eta: 0:58:23 lr: 0.000011 loss: 1.524988 (1.626664) time: 0.713643 data: 0.000215 max mem: 14338 Epoch: [23/30] [ 200/5004] eta: 0:57:41 lr: 0.000011 loss: 1.410884 (1.614697) time: 0.717075 data: 0.000233 max mem: 14338 Epoch: [23/30] [ 250/5004] eta: 0:56:58 lr: 0.000011 loss: 1.541039 (1.614887) time: 0.713096 data: 0.000198 max mem: 14338 Epoch: [23/30] [ 300/5004] eta: 0:56:18 lr: 0.000011 loss: 1.552813 (1.616972) time: 0.711043 data: 0.000174 max mem: 14338 Epoch: [23/30] [ 350/5004] eta: 0:55:40 lr: 0.000011 loss: 1.573855 (1.620415) time: 0.711958 data: 0.000189 max mem: 14338 Epoch: [23/30] [ 400/5004] eta: 0:55:02 lr: 0.000011 loss: 1.501465 (1.618088) time: 0.716001 data: 0.000217 max mem: 14338 Epoch: [23/30] [ 450/5004] eta: 0:54:26 lr: 0.000011 loss: 1.603099 (1.613119) time: 0.723711 data: 0.000222 max mem: 14338 Epoch: [23/30] [ 500/5004] eta: 0:53:49 lr: 0.000011 loss: 1.483472 (1.613759) time: 0.717059 data: 0.000208 max mem: 14338 Epoch: [23/30] [ 550/5004] eta: 0:53:13 lr: 0.000011 loss: 1.603423 (1.613857) time: 0.721621 data: 0.000223 max mem: 14338 Epoch: [23/30] [ 600/5004] eta: 0:52:36 lr: 0.000011 loss: 1.434983 (1.614823) time: 0.716860 data: 0.000211 max mem: 14338 Epoch: [23/30] [ 650/5004] eta: 0:52:00 lr: 0.000011 loss: 1.573839 (1.618329) time: 0.713362 data: 0.000177 max mem: 14338 Epoch: [23/30] [ 700/5004] eta: 0:51:23 lr: 0.000011 loss: 1.446734 (1.616001) time: 0.711497 data: 0.000156 max mem: 14338 Epoch: [23/30] [ 750/5004] eta: 0:50:47 lr: 0.000011 loss: 1.549429 (1.616802) time: 0.712175 data: 0.000205 max mem: 14338 Epoch: [23/30] [ 800/5004] eta: 0:50:10 lr: 0.000010 loss: 1.547583 (1.614640) time: 0.712261 data: 0.000229 max mem: 14338 Epoch: [23/30] [ 850/5004] eta: 0:49:34 lr: 0.000010 loss: 1.698683 (1.621173) time: 0.711399 data: 0.000221 max mem: 14338 Epoch: [23/30] [ 900/5004] eta: 0:48:57 lr: 0.000010 loss: 1.480117 (1.621571) time: 0.712036 data: 0.000205 max mem: 14338 Epoch: [23/30] [ 950/5004] eta: 0:48:21 lr: 0.000010 loss: 1.584966 (1.621328) time: 0.714693 data: 0.000214 max mem: 14338 Epoch: [23/30] [1000/5004] eta: 0:47:45 lr: 0.000010 loss: 1.507803 (1.623660) time: 0.723369 data: 0.000158 max mem: 14338 Epoch: [23/30] [1050/5004] eta: 0:47:09 lr: 0.000010 loss: 1.579316 (1.626537) time: 0.722100 data: 0.000222 max mem: 14338 Epoch: [23/30] [1100/5004] eta: 0:46:33 lr: 0.000010 loss: 1.559596 (1.625728) time: 0.708444 data: 0.000214 max mem: 14338 Epoch: [23/30] [1150/5004] eta: 0:45:57 lr: 0.000010 loss: 1.560292 (1.624561) time: 0.712675 data: 0.000217 max mem: 14338 Epoch: [23/30] [1200/5004] eta: 0:45:21 lr: 0.000010 loss: 1.555074 (1.621873) time: 0.712250 data: 0.000212 max mem: 14338 Epoch: [23/30] [1250/5004] eta: 0:44:45 lr: 0.000010 loss: 1.549456 (1.622180) time: 0.715048 data: 0.000214 max mem: 14338 Epoch: [23/30] [1300/5004] eta: 0:44:09 lr: 0.000010 loss: 1.501074 (1.621204) time: 0.709773 data: 0.000163 max mem: 14338 Epoch: [23/30] [1350/5004] eta: 0:43:33 lr: 0.000010 loss: 1.619085 (1.621920) time: 0.714594 data: 0.000170 max mem: 14338 Epoch: [23/30] [1400/5004] eta: 0:42:57 lr: 0.000010 loss: 1.594982 (1.623501) time: 0.717053 data: 0.000223 max mem: 14338 Epoch: [23/30] [1450/5004] eta: 0:42:21 lr: 0.000010 loss: 1.608205 (1.622908) time: 0.719844 data: 0.000220 max mem: 14338 Epoch: [23/30] [1500/5004] eta: 0:41:45 lr: 0.000010 loss: 1.541373 (1.622922) time: 0.709937 data: 0.000222 max mem: 14338 Epoch: [23/30] [1550/5004] eta: 0:41:10 lr: 0.000010 loss: 1.629972 (1.625347) time: 0.709872 data: 0.000185 max mem: 14338 Epoch: [23/30] [1600/5004] eta: 0:40:35 lr: 0.000010 loss: 1.441837 (1.624558) time: 0.725632 data: 0.000211 max mem: 14338 Epoch: [23/30] [1650/5004] eta: 0:39:59 lr: 0.000010 loss: 1.560150 (1.624491) time: 0.710701 data: 0.000177 max mem: 14338 Epoch: [23/30] [1700/5004] eta: 0:39:23 lr: 0.000010 loss: 1.734854 (1.626315) time: 0.710592 data: 0.000173 max mem: 14338 Epoch: [23/30] [1750/5004] eta: 0:38:47 lr: 0.000010 loss: 1.544245 (1.626700) time: 0.711447 data: 0.000224 max mem: 14338 Epoch: [23/30] [1800/5004] eta: 0:38:11 lr: 0.000010 loss: 1.382331 (1.625345) time: 0.711909 data: 0.000225 max mem: 14338 Epoch: [23/30] [1850/5004] eta: 0:37:35 lr: 0.000010 loss: 1.736288 (1.627099) time: 0.717642 data: 0.000215 max mem: 14338 Epoch: [23/30] [1900/5004] eta: 0:36:59 lr: 0.000010 loss: 1.592971 (1.628942) time: 0.714265 data: 0.000213 max mem: 14338 Epoch: [23/30] [1950/5004] eta: 0:36:23 lr: 0.000010 loss: 1.714937 (1.630046) time: 0.710627 data: 0.000231 max mem: 14338 Epoch: [23/30] [2000/5004] eta: 0:35:48 lr: 0.000010 loss: 1.529766 (1.629858) time: 0.724968 data: 0.000164 max mem: 14338 Epoch: [23/30] [2050/5004] eta: 0:35:12 lr: 0.000010 loss: 1.560455 (1.630453) time: 0.716093 data: 0.000147 max mem: 14338 Epoch: [23/30] [2100/5004] eta: 0:34:36 lr: 0.000010 loss: 1.581246 (1.629337) time: 0.710298 data: 0.000215 max mem: 14338 Epoch: [23/30] [2150/5004] eta: 0:34:00 lr: 0.000010 loss: 1.660339 (1.630506) time: 0.712118 data: 0.000199 max mem: 14338 Epoch: [23/30] [2200/5004] eta: 0:33:24 lr: 0.000010 loss: 1.613835 (1.631429) time: 0.712800 data: 0.000185 max mem: 14338 Epoch: [23/30] [2250/5004] eta: 0:32:49 lr: 0.000010 loss: 1.600935 (1.633147) time: 0.717877 data: 0.000193 max mem: 14338 Epoch: [23/30] [2300/5004] eta: 0:32:13 lr: 0.000010 loss: 1.664016 (1.634785) time: 0.713351 data: 0.000230 max mem: 14338 Epoch: [23/30] [2350/5004] eta: 0:31:37 lr: 0.000010 loss: 1.470210 (1.633563) time: 0.715280 data: 0.000177 max mem: 14338 Epoch: [23/30] [2400/5004] eta: 0:31:01 lr: 0.000010 loss: 1.552633 (1.632986) time: 0.714445 data: 0.000213 max mem: 14338 Epoch: [23/30] [2450/5004] eta: 0:30:26 lr: 0.000010 loss: 1.568991 (1.632767) time: 0.720409 data: 0.000226 max mem: 14338 Epoch: [23/30] [2500/5004] eta: 0:29:50 lr: 0.000010 loss: 1.435036 (1.631426) time: 0.713467 data: 0.000209 max mem: 14338 Epoch: [23/30] [2550/5004] eta: 0:29:14 lr: 0.000010 loss: 1.612946 (1.632141) time: 0.710599 data: 0.000216 max mem: 14338 Epoch: [23/30] [2600/5004] eta: 0:28:38 lr: 0.000010 loss: 1.634085 (1.632691) time: 0.715243 data: 0.000213 max mem: 14338 Epoch: [23/30] [2650/5004] eta: 0:28:03 lr: 0.000010 loss: 1.670820 (1.632276) time: 0.715424 data: 0.000168 max mem: 14338 Epoch: [23/30] [2700/5004] eta: 0:27:27 lr: 0.000009 loss: 1.660156 (1.632116) time: 0.710224 data: 0.000177 max mem: 14338 Epoch: [23/30] [2750/5004] eta: 0:26:51 lr: 0.000009 loss: 1.730550 (1.632379) time: 0.713999 data: 0.000225 max mem: 14338 Epoch: [23/30] [2800/5004] eta: 0:26:15 lr: 0.000009 loss: 1.602484 (1.631765) time: 0.719604 data: 0.000209 max mem: 14338 Epoch: [23/30] [2850/5004] eta: 0:25:40 lr: 0.000009 loss: 1.517139 (1.632479) time: 0.719246 data: 0.000183 max mem: 14338 Epoch: [23/30] [2900/5004] eta: 0:25:04 lr: 0.000009 loss: 1.566946 (1.633118) time: 0.716878 data: 0.000218 max mem: 14338 Epoch: [23/30] [2950/5004] eta: 0:24:28 lr: 0.000009 loss: 1.630697 (1.633649) time: 0.708738 data: 0.000224 max mem: 14338 Epoch: [23/30] [3000/5004] eta: 0:23:52 lr: 0.000009 loss: 1.424320 (1.633251) time: 0.710197 data: 0.000183 max mem: 14338 Epoch: [23/30] [3050/5004] eta: 0:23:16 lr: 0.000009 loss: 1.391518 (1.632846) time: 0.710699 data: 0.000176 max mem: 14338 Epoch: [23/30] [3100/5004] eta: 0:22:41 lr: 0.000009 loss: 1.543262 (1.631820) time: 0.713081 data: 0.000206 max mem: 14338 Epoch: [23/30] [3150/5004] eta: 0:22:05 lr: 0.000009 loss: 1.634754 (1.632631) time: 0.711122 data: 0.000214 max mem: 14338 Epoch: [23/30] [3200/5004] eta: 0:21:29 lr: 0.000009 loss: 1.689992 (1.632187) time: 0.712173 data: 0.000214 max mem: 14338 Epoch: [23/30] [3250/5004] eta: 0:20:53 lr: 0.000009 loss: 1.568996 (1.632370) time: 0.717378 data: 0.000214 max mem: 14338 Epoch: [23/30] [3300/5004] eta: 0:20:17 lr: 0.000009 loss: 1.452486 (1.631046) time: 0.716911 data: 0.000223 max mem: 14338 Epoch: [23/30] [3350/5004] eta: 0:19:42 lr: 0.000009 loss: 1.487136 (1.631703) time: 0.712940 data: 0.000171 max mem: 14338 Epoch: [23/30] [3400/5004] eta: 0:19:06 lr: 0.000009 loss: 1.682752 (1.632522) time: 0.718183 data: 0.000176 max mem: 14338 Epoch: [23/30] [3450/5004] eta: 0:18:30 lr: 0.000009 loss: 1.657627 (1.632931) time: 0.717351 data: 0.000210 max mem: 14338 Epoch: [23/30] [3500/5004] eta: 0:17:54 lr: 0.000009 loss: 1.599487 (1.632362) time: 0.713699 data: 0.000182 max mem: 14338 Epoch: [23/30] [3550/5004] eta: 0:17:19 lr: 0.000009 loss: 1.598822 (1.631798) time: 0.710962 data: 0.000197 max mem: 14338 Epoch: [23/30] [3600/5004] eta: 0:16:43 lr: 0.000009 loss: 1.463726 (1.631159) time: 0.716202 data: 0.000209 max mem: 14338 Epoch: [23/30] [3650/5004] eta: 0:16:07 lr: 0.000009 loss: 1.584722 (1.631124) time: 0.712286 data: 0.000202 max mem: 14338 Epoch: [23/30] [3700/5004] eta: 0:15:31 lr: 0.000009 loss: 1.550491 (1.630785) time: 0.712845 data: 0.000164 max mem: 14338 Epoch: [23/30] [3750/5004] eta: 0:14:56 lr: 0.000009 loss: 1.613515 (1.631023) time: 0.718176 data: 0.000231 max mem: 14338 Epoch: [23/30] [3800/5004] eta: 0:14:20 lr: 0.000009 loss: 1.587122 (1.631159) time: 0.719069 data: 0.000224 max mem: 14338 Epoch: [23/30] [3850/5004] eta: 0:13:44 lr: 0.000009 loss: 1.563601 (1.631277) time: 0.721494 data: 0.000217 max mem: 14338 Epoch: [23/30] [3900/5004] eta: 0:13:09 lr: 0.000009 loss: 1.652874 (1.630655) time: 0.715258 data: 0.000224 max mem: 14338 Epoch: [23/30] [3950/5004] eta: 0:12:33 lr: 0.000009 loss: 1.492162 (1.630298) time: 0.712171 data: 0.000221 max mem: 14338 Epoch: [23/30] [4000/5004] eta: 0:11:57 lr: 0.000009 loss: 1.684843 (1.630780) time: 0.715978 data: 0.000160 max mem: 14338 Epoch: [23/30] [4050/5004] eta: 0:11:21 lr: 0.000009 loss: 1.493403 (1.629885) time: 0.711055 data: 0.000147 max mem: 14338 Epoch: [23/30] [4100/5004] eta: 0:10:46 lr: 0.000009 loss: 1.581466 (1.630176) time: 0.710088 data: 0.000212 max mem: 14338 Epoch: [23/30] [4150/5004] eta: 0:10:10 lr: 0.000009 loss: 1.527590 (1.630374) time: 0.715612 data: 0.000177 max mem: 14338 Epoch: [23/30] [4200/5004] eta: 0:09:34 lr: 0.000009 loss: 1.640015 (1.630718) time: 0.720041 data: 0.000230 max mem: 14338 Epoch: [23/30] [4250/5004] eta: 0:08:58 lr: 0.000009 loss: 1.710567 (1.630975) time: 0.720904 data: 0.000213 max mem: 14338 Epoch: [23/30] [4300/5004] eta: 0:08:23 lr: 0.000009 loss: 1.650861 (1.630887) time: 0.714034 data: 0.000227 max mem: 14338 Epoch: [23/30] [4350/5004] eta: 0:07:47 lr: 0.000009 loss: 1.519796 (1.631129) time: 0.711208 data: 0.000166 max mem: 14338 Epoch: [23/30] [4400/5004] eta: 0:07:11 lr: 0.000009 loss: 1.683402 (1.631909) time: 0.714344 data: 0.000162 max mem: 14338 Epoch: [23/30] [4450/5004] eta: 0:06:35 lr: 0.000009 loss: 1.539133 (1.631852) time: 0.714118 data: 0.000203 max mem: 14338 Epoch: [23/30] [4500/5004] eta: 0:06:00 lr: 0.000009 loss: 1.708423 (1.631811) time: 0.710168 data: 0.000241 max mem: 14338 Epoch: [23/30] [4550/5004] eta: 0:05:24 lr: 0.000009 loss: 1.643841 (1.631717) time: 0.713561 data: 0.000230 max mem: 14338 Epoch: [23/30] [4600/5004] eta: 0:04:48 lr: 0.000009 loss: 1.784380 (1.631840) time: 0.716388 data: 0.000240 max mem: 14338 Epoch: [23/30] [4650/5004] eta: 0:04:13 lr: 0.000008 loss: 1.535706 (1.631171) time: 0.715751 data: 0.000205 max mem: 14338 Epoch: [23/30] [4700/5004] eta: 0:03:37 lr: 0.000008 loss: 1.509495 (1.630481) time: 0.721012 data: 0.000170 max mem: 14338 Epoch: [23/30] [4750/5004] eta: 0:03:01 lr: 0.000008 loss: 1.746101 (1.630547) time: 0.710185 data: 0.000202 max mem: 14338 Epoch: [23/30] [4800/5004] eta: 0:02:25 lr: 0.000008 loss: 1.508247 (1.629882) time: 0.723340 data: 0.000210 max mem: 14338 Epoch: [23/30] [4850/5004] eta: 0:01:50 lr: 0.000008 loss: 1.483161 (1.630135) time: 0.716680 data: 0.000223 max mem: 14338 Epoch: [23/30] [4900/5004] eta: 0:01:14 lr: 0.000008 loss: 1.578902 (1.630400) time: 0.710125 data: 0.000215 max mem: 14338 Epoch: [23/30] [4950/5004] eta: 0:00:38 lr: 0.000008 loss: 1.386212 (1.629849) time: 0.710386 data: 0.000207 max mem: 14338 Epoch: [23/30] [5000/5004] eta: 0:00:02 lr: 0.000008 loss: 1.549191 (1.630466) time: 0.707866 data: 0.000838 max mem: 14338 Epoch: [23/30] [5003/5004] eta: 0:00:00 lr: 0.000008 loss: 1.549191 (1.630376) time: 0.705419 data: 0.000832 max mem: 14338 Epoch: [23/30] Total time: 0:59:37 (0.714956 s / it) Averaged stats: lr: 0.000008 loss: 1.549191 (1.626587) Test: [ 0/196] eta: 0:04:56 loss: 0.302551 (0.302551) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.511906 data: 1.132535 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.455346 (0.547110) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.397999 data: 0.103076 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.535601 (0.542862) acc1: 87.500000 (84.821429) acc5: 100.000000 (98.214286) time: 0.286551 data: 0.000113 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.513010 (0.517117) acc1: 87.500000 (86.491935) acc5: 100.000000 (98.387097) time: 0.286541 data: 0.000108 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.439939 (0.525529) acc1: 87.500000 (86.280488) acc5: 100.000000 (98.170732) time: 0.291993 data: 0.000131 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.466249 (0.551775) acc1: 87.500000 (86.397059) acc5: 100.000000 (97.671569) time: 0.292321 data: 0.000127 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.599816 (0.579204) acc1: 87.500000 (85.860656) acc5: 93.750000 (97.540984) time: 0.286616 data: 0.000129 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.655671 (0.595538) acc1: 81.250000 (85.475352) acc5: 100.000000 (97.711268) time: 0.286350 data: 0.000125 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.503748 (0.595292) acc1: 87.500000 (85.493827) acc5: 100.000000 (97.762346) time: 0.286903 data: 0.000110 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.512936 (0.618335) acc1: 87.500000 (85.164835) acc5: 100.000000 (97.527473) time: 0.286335 data: 0.000120 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.560169 (0.606080) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.710396) time: 0.286210 data: 0.000127 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.537053 (0.594957) acc1: 87.500000 (85.360360) acc5: 100.000000 (97.804054) time: 0.286866 data: 0.000123 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.537053 (0.591129) acc1: 87.500000 (85.433884) acc5: 100.000000 (97.830579) time: 0.286671 data: 0.000128 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.611789 (0.606710) acc1: 87.500000 (85.114504) acc5: 100.000000 (97.805344) time: 0.286362 data: 0.000123 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.625248 (0.606125) acc1: 81.250000 (85.239362) acc5: 100.000000 (97.783688) time: 0.286464 data: 0.000119 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.560702 (0.615292) acc1: 87.500000 (85.140728) acc5: 100.000000 (97.806291) time: 0.286602 data: 0.000123 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.520753 (0.620195) acc1: 81.250000 (85.093168) acc5: 100.000000 (97.787267) time: 0.287159 data: 0.000124 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.503601 (0.617049) acc1: 81.250000 (85.160819) acc5: 100.000000 (97.807018) time: 0.286719 data: 0.000131 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.474371 (0.616605) acc1: 81.250000 (85.013812) acc5: 100.000000 (97.790055) time: 0.291731 data: 0.000130 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.360172 (0.604729) acc1: 87.500000 (85.274869) acc5: 100.000000 (97.807592) time: 0.290320 data: 0.000101 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.491629 (0.615137) acc1: 87.500000 (85.152000) acc5: 100.000000 (97.728000) time: 0.280689 data: 0.000094 max mem: 14338 Test: Total time: 0:00:57 (0.293980 s / it) * Acc@1 85.186 Acc@5 97.268 loss 0.631 Max accuracy: 85.25% Epoch: [24/30] [ 0/5004] eta: 2:44:20 lr: 0.000008 loss: 1.386665 (1.386665) time: 1.970555 data: 1.251085 max mem: 14338 Epoch: [24/30] [ 50/5004] eta: 1:00:57 lr: 0.000008 loss: 1.448377 (1.694445) time: 0.713186 data: 0.000178 max mem: 14338 Epoch: [24/30] [ 100/5004] eta: 0:59:23 lr: 0.000008 loss: 1.573707 (1.685728) time: 0.713755 data: 0.000236 max mem: 14338 Epoch: [24/30] [ 150/5004] eta: 0:58:25 lr: 0.000008 loss: 1.648383 (1.684651) time: 0.710561 data: 0.000232 max mem: 14338 Epoch: [24/30] [ 200/5004] eta: 0:57:44 lr: 0.000008 loss: 1.549135 (1.666444) time: 0.723562 data: 0.000216 max mem: 14338 Epoch: [24/30] [ 250/5004] eta: 0:57:02 lr: 0.000008 loss: 1.512701 (1.648418) time: 0.720547 data: 0.000183 max mem: 14338 Epoch: [24/30] [ 300/5004] eta: 0:56:20 lr: 0.000008 loss: 1.455031 (1.637436) time: 0.712492 data: 0.000171 max mem: 14338 Epoch: [24/30] [ 350/5004] eta: 0:55:43 lr: 0.000008 loss: 1.705036 (1.649090) time: 0.716151 data: 0.000186 max mem: 14338 Epoch: [24/30] [ 400/5004] eta: 0:55:05 lr: 0.000008 loss: 1.540204 (1.639535) time: 0.713712 data: 0.000213 max mem: 14338 Epoch: [24/30] [ 450/5004] eta: 0:54:27 lr: 0.000008 loss: 1.665206 (1.635187) time: 0.715635 data: 0.000214 max mem: 14338 Epoch: [24/30] [ 500/5004] eta: 0:53:49 lr: 0.000008 loss: 1.569601 (1.635835) time: 0.709188 data: 0.000215 max mem: 14338 Epoch: [24/30] [ 550/5004] eta: 0:53:14 lr: 0.000008 loss: 1.628506 (1.640300) time: 0.723068 data: 0.000218 max mem: 14338 Epoch: [24/30] [ 600/5004] eta: 0:52:38 lr: 0.000008 loss: 1.580234 (1.639535) time: 0.712913 data: 0.000217 max mem: 14338 Epoch: [24/30] [ 650/5004] eta: 0:52:01 lr: 0.000008 loss: 1.724717 (1.642518) time: 0.718209 data: 0.000169 max mem: 14338 Epoch: [24/30] [ 700/5004] eta: 0:51:25 lr: 0.000008 loss: 1.585672 (1.640581) time: 0.713534 data: 0.000160 max mem: 14338 Epoch: [24/30] [ 750/5004] eta: 0:50:49 lr: 0.000008 loss: 1.575713 (1.638943) time: 0.716066 data: 0.000206 max mem: 14338 Epoch: [24/30] [ 800/5004] eta: 0:50:12 lr: 0.000008 loss: 1.479606 (1.633511) time: 0.717242 data: 0.000200 max mem: 14338 Epoch: [24/30] [ 850/5004] eta: 0:49:36 lr: 0.000008 loss: 1.531215 (1.632171) time: 0.717756 data: 0.000235 max mem: 14338 Epoch: [24/30] [ 900/5004] eta: 0:49:01 lr: 0.000008 loss: 1.465475 (1.629964) time: 0.711812 data: 0.000224 max mem: 14338 Epoch: [24/30] [ 950/5004] eta: 0:48:24 lr: 0.000008 loss: 1.705695 (1.636751) time: 0.709680 data: 0.000221 max mem: 14338 Epoch: [24/30] [1000/5004] eta: 0:47:48 lr: 0.000008 loss: 1.679928 (1.641287) time: 0.713483 data: 0.000151 max mem: 14338 Epoch: [24/30] [1050/5004] eta: 0:47:11 lr: 0.000008 loss: 1.539756 (1.639723) time: 0.711044 data: 0.000343 max mem: 14338 Epoch: [24/30] [1100/5004] eta: 0:46:35 lr: 0.000008 loss: 1.563881 (1.637673) time: 0.709512 data: 0.000218 max mem: 14338 Epoch: [24/30] [1150/5004] eta: 0:45:58 lr: 0.000008 loss: 1.593431 (1.635705) time: 0.714435 data: 0.000221 max mem: 14338 Epoch: [24/30] [1200/5004] eta: 0:45:23 lr: 0.000008 loss: 1.613982 (1.632253) time: 0.720259 data: 0.000220 max mem: 14338 Epoch: [24/30] [1250/5004] eta: 0:44:47 lr: 0.000008 loss: 1.521379 (1.633329) time: 0.716402 data: 0.000212 max mem: 14338 Epoch: [24/30] [1300/5004] eta: 0:44:10 lr: 0.000008 loss: 1.538094 (1.633430) time: 0.712291 data: 0.000154 max mem: 14338 Epoch: [24/30] [1350/5004] eta: 0:43:34 lr: 0.000008 loss: 1.524844 (1.630019) time: 0.709427 data: 0.000165 max mem: 14338 Epoch: [24/30] [1400/5004] eta: 0:42:58 lr: 0.000008 loss: 1.588807 (1.632753) time: 0.715492 data: 0.000198 max mem: 14338 Epoch: [24/30] [1450/5004] eta: 0:42:22 lr: 0.000008 loss: 1.615231 (1.634596) time: 0.714084 data: 0.000218 max mem: 14338 Epoch: [24/30] [1500/5004] eta: 0:41:46 lr: 0.000008 loss: 1.536891 (1.632559) time: 0.709811 data: 0.000225 max mem: 14338 Epoch: [24/30] [1550/5004] eta: 0:41:10 lr: 0.000008 loss: 1.507704 (1.631390) time: 0.710176 data: 0.000193 max mem: 14338 Epoch: [24/30] [1600/5004] eta: 0:40:35 lr: 0.000008 loss: 1.530214 (1.631184) time: 0.719680 data: 0.000212 max mem: 14338 Epoch: [24/30] [1650/5004] eta: 0:39:59 lr: 0.000008 loss: 1.675746 (1.632036) time: 0.719862 data: 0.000174 max mem: 14338 Epoch: [24/30] [1700/5004] eta: 0:39:23 lr: 0.000008 loss: 1.443445 (1.633930) time: 0.715666 data: 0.000180 max mem: 14338 Epoch: [24/30] [1750/5004] eta: 0:38:47 lr: 0.000007 loss: 1.526009 (1.635087) time: 0.709825 data: 0.000211 max mem: 14338 Epoch: [24/30] [1800/5004] eta: 0:38:11 lr: 0.000007 loss: 1.556413 (1.633276) time: 0.713636 data: 0.000208 max mem: 14338 Epoch: [24/30] [1850/5004] eta: 0:37:35 lr: 0.000007 loss: 1.626170 (1.635125) time: 0.712504 data: 0.000205 max mem: 14338 Epoch: [24/30] [1900/5004] eta: 0:36:59 lr: 0.000007 loss: 1.602373 (1.633745) time: 0.709537 data: 0.000223 max mem: 14338 Epoch: [24/30] [1950/5004] eta: 0:36:23 lr: 0.000007 loss: 1.608169 (1.634194) time: 0.709854 data: 0.000211 max mem: 14338 Epoch: [24/30] [2000/5004] eta: 0:35:48 lr: 0.000007 loss: 1.667317 (1.634729) time: 0.710790 data: 0.000172 max mem: 14338 Epoch: [24/30] [2050/5004] eta: 0:35:12 lr: 0.000007 loss: 1.612336 (1.633870) time: 0.718135 data: 0.000161 max mem: 14338 Epoch: [24/30] [2100/5004] eta: 0:34:36 lr: 0.000007 loss: 1.624611 (1.633339) time: 0.722262 data: 0.000214 max mem: 14338 Epoch: [24/30] [2150/5004] eta: 0:34:00 lr: 0.000007 loss: 1.544265 (1.633791) time: 0.719296 data: 0.000225 max mem: 14338 Epoch: [24/30] [2200/5004] eta: 0:33:25 lr: 0.000007 loss: 1.557538 (1.633307) time: 0.717560 data: 0.000197 max mem: 14338 Epoch: [24/30] [2250/5004] eta: 0:32:49 lr: 0.000007 loss: 1.530292 (1.633977) time: 0.711848 data: 0.000224 max mem: 14338 Epoch: [24/30] [2300/5004] eta: 0:32:13 lr: 0.000007 loss: 1.444305 (1.633325) time: 0.710939 data: 0.000213 max mem: 14338 Epoch: [24/30] [2350/5004] eta: 0:31:37 lr: 0.000007 loss: 1.508098 (1.632875) time: 0.711097 data: 0.000166 max mem: 14338 Epoch: [24/30] [2400/5004] eta: 0:31:01 lr: 0.000007 loss: 1.547098 (1.631713) time: 0.716356 data: 0.000212 max mem: 14338 Epoch: [24/30] [2450/5004] eta: 0:30:25 lr: 0.000007 loss: 1.615084 (1.630177) time: 0.710796 data: 0.000217 max mem: 14338 Epoch: [24/30] [2500/5004] eta: 0:29:50 lr: 0.000007 loss: 1.428969 (1.629182) time: 0.711354 data: 0.000242 max mem: 14338 Epoch: [24/30] [2550/5004] eta: 0:29:14 lr: 0.000007 loss: 1.620415 (1.629092) time: 0.714021 data: 0.000231 max mem: 14338 Epoch: [24/30] [2600/5004] eta: 0:28:38 lr: 0.000007 loss: 1.637496 (1.629573) time: 0.722793 data: 0.000224 max mem: 14338 Epoch: [24/30] [2650/5004] eta: 0:28:02 lr: 0.000007 loss: 1.650769 (1.629674) time: 0.721097 data: 0.000168 max mem: 14338 Epoch: [24/30] [2700/5004] eta: 0:27:27 lr: 0.000007 loss: 1.489099 (1.630314) time: 0.714084 data: 0.000197 max mem: 14338 Epoch: [24/30] [2750/5004] eta: 0:26:51 lr: 0.000007 loss: 1.579760 (1.630189) time: 0.714538 data: 0.000219 max mem: 14338 Epoch: [24/30] [2800/5004] eta: 0:26:15 lr: 0.000007 loss: 1.532994 (1.628962) time: 0.717489 data: 0.000217 max mem: 14338 Epoch: [24/30] [2850/5004] eta: 0:25:39 lr: 0.000007 loss: 1.564300 (1.628970) time: 0.711168 data: 0.000175 max mem: 14338 Epoch: [24/30] [2900/5004] eta: 0:25:04 lr: 0.000007 loss: 1.526256 (1.628459) time: 0.709684 data: 0.000222 max mem: 14338 Epoch: [24/30] [2950/5004] eta: 0:24:28 lr: 0.000007 loss: 1.560744 (1.627711) time: 0.713011 data: 0.000196 max mem: 14338 Epoch: [24/30] [3000/5004] eta: 0:23:52 lr: 0.000007 loss: 1.439819 (1.626336) time: 0.713355 data: 0.000176 max mem: 14338 Epoch: [24/30] [3050/5004] eta: 0:23:16 lr: 0.000007 loss: 1.595447 (1.626530) time: 0.718598 data: 0.000161 max mem: 14338 Epoch: [24/30] [3100/5004] eta: 0:22:41 lr: 0.000007 loss: 1.484246 (1.627260) time: 0.715293 data: 0.000208 max mem: 14338 Epoch: [24/30] [3150/5004] eta: 0:22:05 lr: 0.000007 loss: 1.563175 (1.626621) time: 0.710393 data: 0.000206 max mem: 14338 Epoch: [24/30] [3200/5004] eta: 0:21:29 lr: 0.000007 loss: 1.468488 (1.625826) time: 0.714452 data: 0.000220 max mem: 14338 Epoch: [24/30] [3250/5004] eta: 0:20:53 lr: 0.000007 loss: 1.634215 (1.625885) time: 0.711925 data: 0.000227 max mem: 14338 Epoch: [24/30] [3300/5004] eta: 0:20:17 lr: 0.000007 loss: 1.453639 (1.625690) time: 0.711701 data: 0.000202 max mem: 14338 Epoch: [24/30] [3350/5004] eta: 0:19:42 lr: 0.000007 loss: 1.520012 (1.625676) time: 0.710393 data: 0.000154 max mem: 14338 Epoch: [24/30] [3400/5004] eta: 0:19:06 lr: 0.000007 loss: 1.664354 (1.625664) time: 0.713339 data: 0.000178 max mem: 14338 Epoch: [24/30] [3450/5004] eta: 0:18:30 lr: 0.000007 loss: 1.652580 (1.625540) time: 0.718170 data: 0.000241 max mem: 14338 Epoch: [24/30] [3500/5004] eta: 0:17:54 lr: 0.000007 loss: 1.502980 (1.624639) time: 0.714834 data: 0.000195 max mem: 14338 Epoch: [24/30] [3550/5004] eta: 0:17:19 lr: 0.000007 loss: 1.396425 (1.623331) time: 0.712931 data: 0.000227 max mem: 14338 Epoch: [24/30] [3600/5004] eta: 0:16:43 lr: 0.000007 loss: 1.657992 (1.622749) time: 0.717790 data: 0.000232 max mem: 14338 Epoch: [24/30] [3650/5004] eta: 0:16:07 lr: 0.000007 loss: 1.508848 (1.623147) time: 0.716404 data: 0.000209 max mem: 14338 Epoch: [24/30] [3700/5004] eta: 0:15:31 lr: 0.000007 loss: 1.476099 (1.622825) time: 0.711284 data: 0.000160 max mem: 14338 Epoch: [24/30] [3750/5004] eta: 0:14:56 lr: 0.000007 loss: 1.521672 (1.622533) time: 0.710897 data: 0.000221 max mem: 14338 Epoch: [24/30] [3800/5004] eta: 0:14:20 lr: 0.000007 loss: 1.418122 (1.622209) time: 0.715657 data: 0.000230 max mem: 14338 Epoch: [24/30] [3850/5004] eta: 0:13:44 lr: 0.000007 loss: 1.434510 (1.621878) time: 0.712205 data: 0.000214 max mem: 14338 Epoch: [24/30] [3900/5004] eta: 0:13:08 lr: 0.000007 loss: 1.598547 (1.622620) time: 0.713303 data: 0.000228 max mem: 14338 Epoch: [24/30] [3950/5004] eta: 0:12:33 lr: 0.000007 loss: 1.519616 (1.622642) time: 0.715996 data: 0.000228 max mem: 14338 Epoch: [24/30] [4000/5004] eta: 0:11:57 lr: 0.000006 loss: 1.445660 (1.622828) time: 0.723160 data: 0.000165 max mem: 14338 Epoch: [24/30] [4050/5004] eta: 0:11:21 lr: 0.000006 loss: 1.565476 (1.622663) time: 0.716359 data: 0.000162 max mem: 14338 Epoch: [24/30] [4100/5004] eta: 0:10:46 lr: 0.000006 loss: 1.490947 (1.622361) time: 0.715306 data: 0.000207 max mem: 14338 Epoch: [24/30] [4150/5004] eta: 0:10:10 lr: 0.000006 loss: 1.446448 (1.621276) time: 0.711213 data: 0.000201 max mem: 14338 Epoch: [24/30] [4200/5004] eta: 0:09:34 lr: 0.000006 loss: 1.692533 (1.620913) time: 0.715807 data: 0.000211 max mem: 14338 Epoch: [24/30] [4250/5004] eta: 0:08:58 lr: 0.000006 loss: 1.583264 (1.621037) time: 0.720246 data: 0.000210 max mem: 14338 Epoch: [24/30] [4300/5004] eta: 0:08:23 lr: 0.000006 loss: 1.664919 (1.621736) time: 0.710961 data: 0.000237 max mem: 14338 Epoch: [24/30] [4350/5004] eta: 0:07:47 lr: 0.000006 loss: 1.568613 (1.621368) time: 0.714583 data: 0.000171 max mem: 14338 Epoch: [24/30] [4400/5004] eta: 0:07:11 lr: 0.000006 loss: 1.568220 (1.621291) time: 0.718588 data: 0.000154 max mem: 14338 Epoch: [24/30] [4450/5004] eta: 0:06:35 lr: 0.000006 loss: 1.534655 (1.621647) time: 0.716505 data: 0.000222 max mem: 14338 Epoch: [24/30] [4500/5004] eta: 0:06:00 lr: 0.000006 loss: 1.545990 (1.622313) time: 0.723859 data: 0.000225 max mem: 14338 Epoch: [24/30] [4550/5004] eta: 0:05:24 lr: 0.000006 loss: 1.616525 (1.622360) time: 0.716233 data: 0.000226 max mem: 14338 Epoch: [24/30] [4600/5004] eta: 0:04:48 lr: 0.000006 loss: 1.797759 (1.622737) time: 0.715488 data: 0.000223 max mem: 14338 Epoch: [24/30] [4650/5004] eta: 0:04:13 lr: 0.000006 loss: 1.516871 (1.622770) time: 0.718231 data: 0.000225 max mem: 14338 Epoch: [24/30] [4700/5004] eta: 0:03:37 lr: 0.000006 loss: 1.508033 (1.623043) time: 0.709871 data: 0.000157 max mem: 14338 Epoch: [24/30] [4750/5004] eta: 0:03:01 lr: 0.000006 loss: 1.379217 (1.622608) time: 0.709961 data: 0.000163 max mem: 14338 Epoch: [24/30] [4800/5004] eta: 0:02:25 lr: 0.000006 loss: 1.473969 (1.622441) time: 0.711090 data: 0.000188 max mem: 14338 Epoch: [24/30] [4850/5004] eta: 0:01:50 lr: 0.000006 loss: 1.590955 (1.622239) time: 0.711807 data: 0.000226 max mem: 14338 Epoch: [24/30] [4900/5004] eta: 0:01:14 lr: 0.000006 loss: 1.555369 (1.622056) time: 0.711313 data: 0.000218 max mem: 14338 Epoch: [24/30] [4950/5004] eta: 0:00:38 lr: 0.000006 loss: 1.668489 (1.622249) time: 0.717976 data: 0.000234 max mem: 14338 Epoch: [24/30] [5000/5004] eta: 0:00:02 lr: 0.000006 loss: 1.448136 (1.622403) time: 0.714634 data: 0.000829 max mem: 14338 Epoch: [24/30] [5003/5004] eta: 0:00:00 lr: 0.000006 loss: 1.448136 (1.622418) time: 0.711295 data: 0.000822 max mem: 14338 Epoch: [24/30] Total time: 0:59:37 (0.714899 s / it) Averaged stats: lr: 0.000006 loss: 1.448136 (1.621703) Test: [ 0/196] eta: 0:04:50 loss: 0.294232 (0.294232) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.482836 data: 1.056692 max mem: 14338 Test: [ 10/196] eta: 0:01:13 loss: 0.448512 (0.542036) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.397466 data: 0.096183 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.537971 (0.538736) acc1: 87.500000 (85.119048) acc5: 100.000000 (98.214286) time: 0.287596 data: 0.000133 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.514225 (0.515365) acc1: 87.500000 (86.693548) acc5: 100.000000 (98.387097) time: 0.286059 data: 0.000124 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.433483 (0.522414) acc1: 87.500000 (86.585366) acc5: 100.000000 (98.170732) time: 0.286062 data: 0.000121 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.466986 (0.549284) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.671569) time: 0.286200 data: 0.000123 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.614033 (0.577346) acc1: 81.250000 (85.860656) acc5: 93.750000 (97.540984) time: 0.285833 data: 0.000130 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.657902 (0.595175) acc1: 81.250000 (85.475352) acc5: 100.000000 (97.623239) time: 0.285595 data: 0.000135 max mem: 14338 Test: [ 80/196] eta: 0:00:34 loss: 0.513306 (0.595142) acc1: 87.500000 (85.493827) acc5: 100.000000 (97.762346) time: 0.285823 data: 0.000126 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.513306 (0.617911) acc1: 87.500000 (85.233516) acc5: 100.000000 (97.527473) time: 0.285954 data: 0.000132 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.570942 (0.605870) acc1: 81.250000 (85.334158) acc5: 100.000000 (97.710396) time: 0.286408 data: 0.000130 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.542136 (0.595013) acc1: 87.500000 (85.416667) acc5: 100.000000 (97.747748) time: 0.286796 data: 0.000135 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.517976 (0.591118) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.727273) time: 0.293965 data: 0.000157 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.598634 (0.606337) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.662214) time: 0.294245 data: 0.000143 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.633632 (0.605637) acc1: 81.250000 (85.372340) acc5: 100.000000 (97.650709) time: 0.286671 data: 0.000137 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.545743 (0.615021) acc1: 87.500000 (85.223510) acc5: 100.000000 (97.682119) time: 0.287059 data: 0.000133 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.517737 (0.620008) acc1: 81.250000 (85.170807) acc5: 100.000000 (97.709627) time: 0.287266 data: 0.000115 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.490709 (0.616846) acc1: 81.250000 (85.270468) acc5: 100.000000 (97.733918) time: 0.286288 data: 0.000137 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.484877 (0.616410) acc1: 87.500000 (85.186464) acc5: 100.000000 (97.720994) time: 0.285760 data: 0.000158 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.358076 (0.604528) acc1: 87.500000 (85.405759) acc5: 100.000000 (97.742147) time: 0.283194 data: 0.000118 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.472458 (0.615134) acc1: 87.500000 (85.280000) acc5: 100.000000 (97.664000) time: 0.273413 data: 0.000108 max mem: 14338 Test: Total time: 0:00:57 (0.293300 s / it) * Acc@1 85.150 Acc@5 97.276 loss 0.632 Max accuracy: 85.25% Epoch: [25/30] [ 0/5004] eta: 2:37:12 lr: 0.000006 loss: 1.404321 (1.404321) time: 1.884929 data: 1.159348 max mem: 14338 Epoch: [25/30] [ 50/5004] eta: 1:00:48 lr: 0.000006 loss: 1.532105 (1.603847) time: 0.716552 data: 0.000170 max mem: 14338 Epoch: [25/30] [ 100/5004] eta: 0:59:19 lr: 0.000006 loss: 1.694896 (1.637571) time: 0.714050 data: 0.000228 max mem: 14338 Epoch: [25/30] [ 150/5004] eta: 0:58:22 lr: 0.000006 loss: 1.453350 (1.607433) time: 0.710595 data: 0.000236 max mem: 14338 Epoch: [25/30] [ 200/5004] eta: 0:57:40 lr: 0.000006 loss: 1.446496 (1.603335) time: 0.715041 data: 0.000211 max mem: 14338 Epoch: [25/30] [ 250/5004] eta: 0:56:59 lr: 0.000006 loss: 1.655466 (1.621332) time: 0.717075 data: 0.000198 max mem: 14338 Epoch: [25/30] [ 300/5004] eta: 0:56:21 lr: 0.000006 loss: 1.572082 (1.619432) time: 0.714514 data: 0.000163 max mem: 14338 Epoch: [25/30] [ 350/5004] eta: 0:55:40 lr: 0.000006 loss: 1.508773 (1.616299) time: 0.710025 data: 0.000194 max mem: 14338 Epoch: [25/30] [ 400/5004] eta: 0:55:02 lr: 0.000006 loss: 1.524130 (1.616099) time: 0.715118 data: 0.000213 max mem: 14338 Epoch: [25/30] [ 450/5004] eta: 0:54:24 lr: 0.000006 loss: 1.654281 (1.613605) time: 0.720447 data: 0.000224 max mem: 14338 Epoch: [25/30] [ 500/5004] eta: 0:53:47 lr: 0.000006 loss: 1.538505 (1.618401) time: 0.715759 data: 0.000229 max mem: 14338 Epoch: [25/30] [ 550/5004] eta: 0:53:09 lr: 0.000006 loss: 1.535762 (1.614543) time: 0.712283 data: 0.000221 max mem: 14338 Epoch: [25/30] [ 600/5004] eta: 0:52:34 lr: 0.000006 loss: 1.533907 (1.607943) time: 0.713757 data: 0.000227 max mem: 14338 Epoch: [25/30] [ 650/5004] eta: 0:51:57 lr: 0.000006 loss: 1.586107 (1.607117) time: 0.711136 data: 0.000162 max mem: 14338 Epoch: [25/30] [ 700/5004] eta: 0:51:20 lr: 0.000006 loss: 1.640167 (1.609079) time: 0.709613 data: 0.000180 max mem: 14338 Epoch: [25/30] [ 750/5004] eta: 0:50:44 lr: 0.000006 loss: 1.549101 (1.609066) time: 0.709751 data: 0.000227 max mem: 14338 Epoch: [25/30] [ 800/5004] eta: 0:50:07 lr: 0.000006 loss: 1.604928 (1.614942) time: 0.713742 data: 0.000207 max mem: 14338 Epoch: [25/30] [ 850/5004] eta: 0:49:32 lr: 0.000006 loss: 1.584939 (1.621317) time: 0.723732 data: 0.000222 max mem: 14338 Epoch: [25/30] [ 900/5004] eta: 0:48:56 lr: 0.000006 loss: 1.523263 (1.620361) time: 0.717408 data: 0.000195 max mem: 14338 Epoch: [25/30] [ 950/5004] eta: 0:48:19 lr: 0.000006 loss: 1.519579 (1.621142) time: 0.715603 data: 0.000220 max mem: 14338 Epoch: [25/30] [1000/5004] eta: 0:47:43 lr: 0.000006 loss: 1.541301 (1.620309) time: 0.716676 data: 0.000153 max mem: 14338 Epoch: [25/30] [1050/5004] eta: 0:47:07 lr: 0.000006 loss: 1.622647 (1.619008) time: 0.718183 data: 0.000202 max mem: 14338 Epoch: [25/30] [1100/5004] eta: 0:46:32 lr: 0.000006 loss: 1.594011 (1.620732) time: 0.710600 data: 0.000214 max mem: 14338 Epoch: [25/30] [1150/5004] eta: 0:45:56 lr: 0.000006 loss: 1.600494 (1.621117) time: 0.713162 data: 0.000215 max mem: 14338 Epoch: [25/30] [1200/5004] eta: 0:45:21 lr: 0.000006 loss: 1.457629 (1.619306) time: 0.713971 data: 0.000207 max mem: 14338 Epoch: [25/30] [1250/5004] eta: 0:44:45 lr: 0.000006 loss: 1.633652 (1.620901) time: 0.715440 data: 0.000226 max mem: 14338 Epoch: [25/30] [1300/5004] eta: 0:44:09 lr: 0.000006 loss: 1.617620 (1.622442) time: 0.714232 data: 0.000169 max mem: 14338 Epoch: [25/30] [1350/5004] eta: 0:43:33 lr: 0.000006 loss: 1.309219 (1.620340) time: 0.712178 data: 0.000164 max mem: 14338 Epoch: [25/30] [1400/5004] eta: 0:42:57 lr: 0.000006 loss: 1.507856 (1.619371) time: 0.716730 data: 0.000217 max mem: 14338 Epoch: [25/30] [1450/5004] eta: 0:42:21 lr: 0.000005 loss: 1.575563 (1.618000) time: 0.717705 data: 0.000218 max mem: 14338 Epoch: [25/30] [1500/5004] eta: 0:41:45 lr: 0.000005 loss: 1.518744 (1.617005) time: 0.708822 data: 0.000196 max mem: 14338 Epoch: [25/30] [1550/5004] eta: 0:41:09 lr: 0.000005 loss: 1.640283 (1.617164) time: 0.712389 data: 0.000188 max mem: 14338 Epoch: [25/30] [1600/5004] eta: 0:40:33 lr: 0.000005 loss: 1.608207 (1.616869) time: 0.711119 data: 0.000209 max mem: 14338 Epoch: [25/30] [1650/5004] eta: 0:39:58 lr: 0.000005 loss: 1.516659 (1.616016) time: 0.714655 data: 0.000165 max mem: 14338 Epoch: [25/30] [1700/5004] eta: 0:39:22 lr: 0.000005 loss: 1.556176 (1.615201) time: 0.712648 data: 0.000174 max mem: 14338 Epoch: [25/30] [1750/5004] eta: 0:38:46 lr: 0.000005 loss: 1.611233 (1.614104) time: 0.709316 data: 0.000217 max mem: 14338 Epoch: [25/30] [1800/5004] eta: 0:38:10 lr: 0.000005 loss: 1.574418 (1.615959) time: 0.716437 data: 0.000212 max mem: 14338 Epoch: [25/30] [1850/5004] eta: 0:37:35 lr: 0.000005 loss: 1.426815 (1.613293) time: 0.721699 data: 0.000208 max mem: 14338 Epoch: [25/30] [1900/5004] eta: 0:36:59 lr: 0.000005 loss: 1.700435 (1.613601) time: 0.717119 data: 0.000211 max mem: 14338 Epoch: [25/30] [1950/5004] eta: 0:36:23 lr: 0.000005 loss: 1.521833 (1.615100) time: 0.712493 data: 0.000211 max mem: 14338 Epoch: [25/30] [2000/5004] eta: 0:35:47 lr: 0.000005 loss: 1.577439 (1.615678) time: 0.712259 data: 0.000153 max mem: 14338 Epoch: [25/30] [2050/5004] eta: 0:35:11 lr: 0.000005 loss: 1.569666 (1.615256) time: 0.712844 data: 0.000152 max mem: 14338 Epoch: [25/30] [2100/5004] eta: 0:34:35 lr: 0.000005 loss: 1.553081 (1.614245) time: 0.709626 data: 0.000224 max mem: 14338 Epoch: [25/30] [2150/5004] eta: 0:34:00 lr: 0.000005 loss: 1.563871 (1.614739) time: 0.708881 data: 0.000199 max mem: 14338 Epoch: [25/30] [2200/5004] eta: 0:33:24 lr: 0.000005 loss: 1.646520 (1.614453) time: 0.713948 data: 0.000192 max mem: 14338 Epoch: [25/30] [2250/5004] eta: 0:32:48 lr: 0.000005 loss: 1.430022 (1.613107) time: 0.713725 data: 0.000219 max mem: 14338 Epoch: [25/30] [2300/5004] eta: 0:32:12 lr: 0.000005 loss: 1.551489 (1.612854) time: 0.715355 data: 0.000358 max mem: 14338 Epoch: [25/30] [2350/5004] eta: 0:31:36 lr: 0.000005 loss: 1.615123 (1.615295) time: 0.714452 data: 0.000163 max mem: 14338 Epoch: [25/30] [2400/5004] eta: 0:31:01 lr: 0.000005 loss: 1.534092 (1.615387) time: 0.719143 data: 0.000207 max mem: 14338 Epoch: [25/30] [2450/5004] eta: 0:30:25 lr: 0.000005 loss: 1.631238 (1.615608) time: 0.717589 data: 0.000221 max mem: 14338 Epoch: [25/30] [2500/5004] eta: 0:29:49 lr: 0.000005 loss: 1.614198 (1.615547) time: 0.714253 data: 0.000236 max mem: 14338 Epoch: [25/30] [2550/5004] eta: 0:29:14 lr: 0.000005 loss: 1.496567 (1.615272) time: 0.712642 data: 0.000221 max mem: 14338 Epoch: [25/30] [2600/5004] eta: 0:28:38 lr: 0.000005 loss: 1.512486 (1.615847) time: 0.712093 data: 0.000209 max mem: 14338 Epoch: [25/30] [2650/5004] eta: 0:28:02 lr: 0.000005 loss: 1.635775 (1.614598) time: 0.715198 data: 0.000153 max mem: 14338 Epoch: [25/30] [2700/5004] eta: 0:27:26 lr: 0.000005 loss: 1.537727 (1.614375) time: 0.714692 data: 0.000170 max mem: 14338 Epoch: [25/30] [2750/5004] eta: 0:26:50 lr: 0.000005 loss: 1.482988 (1.613609) time: 0.716247 data: 0.000219 max mem: 14338 Epoch: [25/30] [2800/5004] eta: 0:26:15 lr: 0.000005 loss: 1.536471 (1.612712) time: 0.720396 data: 0.000227 max mem: 14338 Epoch: [25/30] [2850/5004] eta: 0:25:39 lr: 0.000005 loss: 1.506651 (1.611656) time: 0.718534 data: 0.000180 max mem: 14338 Epoch: [25/30] [2900/5004] eta: 0:25:03 lr: 0.000005 loss: 1.693662 (1.613136) time: 0.709352 data: 0.000210 max mem: 14338 Epoch: [25/30] [2950/5004] eta: 0:24:27 lr: 0.000005 loss: 1.453430 (1.613563) time: 0.709025 data: 0.000204 max mem: 14338 Epoch: [25/30] [3000/5004] eta: 0:23:52 lr: 0.000005 loss: 1.667725 (1.614107) time: 0.713735 data: 0.000165 max mem: 14338 Epoch: [25/30] [3050/5004] eta: 0:23:16 lr: 0.000005 loss: 1.525175 (1.614210) time: 0.715380 data: 0.000162 max mem: 14338 Epoch: [25/30] [3100/5004] eta: 0:22:40 lr: 0.000005 loss: 1.522645 (1.614264) time: 0.709993 data: 0.000236 max mem: 14338 Epoch: [25/30] [3150/5004] eta: 0:22:04 lr: 0.000005 loss: 1.661938 (1.614935) time: 0.713839 data: 0.000226 max mem: 14338 Epoch: [25/30] [3200/5004] eta: 0:21:29 lr: 0.000005 loss: 1.601639 (1.615935) time: 0.714914 data: 0.000209 max mem: 14338 Epoch: [25/30] [3250/5004] eta: 0:20:53 lr: 0.000005 loss: 1.452591 (1.615956) time: 0.720853 data: 0.000227 max mem: 14338 Epoch: [25/30] [3300/5004] eta: 0:20:17 lr: 0.000005 loss: 1.550416 (1.615852) time: 0.714064 data: 0.000216 max mem: 14338 Epoch: [25/30] [3350/5004] eta: 0:19:41 lr: 0.000005 loss: 1.619002 (1.616244) time: 0.709620 data: 0.000176 max mem: 14338 Epoch: [25/30] [3400/5004] eta: 0:19:06 lr: 0.000005 loss: 1.595014 (1.616945) time: 0.716437 data: 0.000154 max mem: 14338 Epoch: [25/30] [3450/5004] eta: 0:18:30 lr: 0.000005 loss: 1.440513 (1.616531) time: 0.715814 data: 0.000217 max mem: 14338 Epoch: [25/30] [3500/5004] eta: 0:17:54 lr: 0.000005 loss: 1.479711 (1.616621) time: 0.712178 data: 0.000199 max mem: 14338 Epoch: [25/30] [3550/5004] eta: 0:17:19 lr: 0.000005 loss: 1.649796 (1.617287) time: 0.713411 data: 0.000230 max mem: 14338 Epoch: [25/30] [3600/5004] eta: 0:16:43 lr: 0.000005 loss: 1.557882 (1.618194) time: 0.713371 data: 0.000211 max mem: 14338 Epoch: [25/30] [3650/5004] eta: 0:16:07 lr: 0.000005 loss: 1.532816 (1.618023) time: 0.716310 data: 0.000226 max mem: 14338 Epoch: [25/30] [3700/5004] eta: 0:15:31 lr: 0.000005 loss: 1.630183 (1.618330) time: 0.714516 data: 0.000169 max mem: 14338 Epoch: [25/30] [3750/5004] eta: 0:14:56 lr: 0.000005 loss: 1.684929 (1.618342) time: 0.713049 data: 0.000224 max mem: 14338 Epoch: [25/30] [3800/5004] eta: 0:14:20 lr: 0.000005 loss: 1.699941 (1.619828) time: 0.715910 data: 0.000219 max mem: 14338 Epoch: [25/30] [3850/5004] eta: 0:13:44 lr: 0.000005 loss: 1.606910 (1.621369) time: 0.712893 data: 0.000199 max mem: 14338 Epoch: [25/30] [3900/5004] eta: 0:13:08 lr: 0.000005 loss: 1.536834 (1.621162) time: 0.709179 data: 0.000209 max mem: 14338 Epoch: [25/30] [3950/5004] eta: 0:12:33 lr: 0.000005 loss: 1.587503 (1.621119) time: 0.711871 data: 0.000210 max mem: 14338 Epoch: [25/30] [4000/5004] eta: 0:11:57 lr: 0.000005 loss: 1.449859 (1.620681) time: 0.714011 data: 0.000168 max mem: 14338 Epoch: [25/30] [4050/5004] eta: 0:11:21 lr: 0.000005 loss: 1.519917 (1.620052) time: 0.713734 data: 0.000163 max mem: 14338 Epoch: [25/30] [4100/5004] eta: 0:10:45 lr: 0.000005 loss: 1.675267 (1.620333) time: 0.716059 data: 0.000217 max mem: 14338 Epoch: [25/30] [4150/5004] eta: 0:10:10 lr: 0.000004 loss: 1.663946 (1.620557) time: 0.715823 data: 0.000195 max mem: 14338 Epoch: [25/30] [4200/5004] eta: 0:09:34 lr: 0.000004 loss: 1.564927 (1.619851) time: 0.719641 data: 0.000196 max mem: 14338 Epoch: [25/30] [4250/5004] eta: 0:08:58 lr: 0.000004 loss: 1.715831 (1.619944) time: 0.714272 data: 0.000217 max mem: 14338 Epoch: [25/30] [4300/5004] eta: 0:08:23 lr: 0.000004 loss: 1.428882 (1.619468) time: 0.711162 data: 0.000229 max mem: 14338 Epoch: [25/30] [4350/5004] eta: 0:07:47 lr: 0.000004 loss: 1.466125 (1.619364) time: 0.708949 data: 0.000166 max mem: 14338 Epoch: [25/30] [4400/5004] eta: 0:07:11 lr: 0.000004 loss: 1.420209 (1.618466) time: 0.716115 data: 0.000149 max mem: 14338 Epoch: [25/30] [4450/5004] eta: 0:06:35 lr: 0.000004 loss: 1.483145 (1.618236) time: 0.716876 data: 0.000207 max mem: 14338 Epoch: [25/30] [4500/5004] eta: 0:06:00 lr: 0.000004 loss: 1.423725 (1.617806) time: 0.710656 data: 0.000204 max mem: 14338 Epoch: [25/30] [4550/5004] eta: 0:05:24 lr: 0.000004 loss: 1.574487 (1.617799) time: 0.712032 data: 0.000215 max mem: 14338 Epoch: [25/30] [4600/5004] eta: 0:04:48 lr: 0.000004 loss: 1.582821 (1.618117) time: 0.719185 data: 0.000205 max mem: 14338 Epoch: [25/30] [4650/5004] eta: 0:04:12 lr: 0.000004 loss: 1.443675 (1.618118) time: 0.719101 data: 0.000215 max mem: 14338 Epoch: [25/30] [4700/5004] eta: 0:03:37 lr: 0.000004 loss: 1.564631 (1.618652) time: 0.716862 data: 0.000167 max mem: 14338 Epoch: [25/30] [4750/5004] eta: 0:03:01 lr: 0.000004 loss: 1.550287 (1.618448) time: 0.713740 data: 0.000168 max mem: 14338 Epoch: [25/30] [4800/5004] eta: 0:02:25 lr: 0.000004 loss: 1.663143 (1.618712) time: 0.715842 data: 0.000181 max mem: 14338 Epoch: [25/30] [4850/5004] eta: 0:01:50 lr: 0.000004 loss: 1.542405 (1.618982) time: 0.717238 data: 0.000208 max mem: 14338 Epoch: [25/30] [4900/5004] eta: 0:01:14 lr: 0.000004 loss: 1.521648 (1.618342) time: 0.711559 data: 0.000242 max mem: 14338 Epoch: [25/30] [4950/5004] eta: 0:00:38 lr: 0.000004 loss: 1.520194 (1.617991) time: 0.712775 data: 0.000222 max mem: 14338 Epoch: [25/30] [5000/5004] eta: 0:00:02 lr: 0.000004 loss: 1.485003 (1.617229) time: 0.714417 data: 0.000825 max mem: 14338 Epoch: [25/30] [5003/5004] eta: 0:00:00 lr: 0.000004 loss: 1.594542 (1.617328) time: 0.710763 data: 0.000810 max mem: 14338 Epoch: [25/30] Total time: 0:59:36 (0.714669 s / it) Averaged stats: lr: 0.000004 loss: 1.594542 (1.621682) Test: [ 0/196] eta: 0:05:06 loss: 0.278280 (0.278280) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.564400 data: 1.193433 max mem: 14338 Test: [ 10/196] eta: 0:01:15 loss: 0.458672 (0.533671) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.403422 data: 0.108610 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.533266 (0.534656) acc1: 87.500000 (85.416667) acc5: 100.000000 (98.214286) time: 0.286758 data: 0.000138 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.523443 (0.512581) acc1: 87.500000 (86.895161) acc5: 100.000000 (98.387097) time: 0.286725 data: 0.000126 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.439263 (0.520452) acc1: 87.500000 (86.585366) acc5: 100.000000 (98.170732) time: 0.286985 data: 0.000121 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.462421 (0.548900) acc1: 87.500000 (86.642157) acc5: 100.000000 (97.671569) time: 0.286430 data: 0.000125 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.616230 (0.577227) acc1: 87.500000 (86.065574) acc5: 93.750000 (97.540984) time: 0.289326 data: 0.000123 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.658149 (0.595350) acc1: 81.250000 (85.651408) acc5: 100.000000 (97.623239) time: 0.291406 data: 0.000126 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.490980 (0.593938) acc1: 87.500000 (85.648148) acc5: 100.000000 (97.762346) time: 0.288302 data: 0.000121 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.512016 (0.616792) acc1: 87.500000 (85.233516) acc5: 100.000000 (97.527473) time: 0.286083 data: 0.000130 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.579025 (0.605802) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.710396) time: 0.285968 data: 0.000124 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.559700 (0.594842) acc1: 87.500000 (85.416667) acc5: 100.000000 (97.804054) time: 0.285846 data: 0.000117 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.527699 (0.590689) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.830579) time: 0.286250 data: 0.000130 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.612215 (0.606139) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.757634) time: 0.286661 data: 0.000130 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.627374 (0.605764) acc1: 81.250000 (85.372340) acc5: 100.000000 (97.739362) time: 0.286699 data: 0.000132 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.544827 (0.614834) acc1: 87.500000 (85.264901) acc5: 100.000000 (97.764901) time: 0.286976 data: 0.000143 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.529831 (0.619942) acc1: 81.250000 (85.209627) acc5: 100.000000 (97.787267) time: 0.286625 data: 0.000131 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.503055 (0.616898) acc1: 81.250000 (85.307018) acc5: 100.000000 (97.807018) time: 0.285780 data: 0.000125 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.480837 (0.616593) acc1: 87.500000 (85.186464) acc5: 100.000000 (97.790055) time: 0.285554 data: 0.000122 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.362172 (0.604653) acc1: 87.500000 (85.405759) acc5: 100.000000 (97.807592) time: 0.283694 data: 0.000094 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.486032 (0.615306) acc1: 87.500000 (85.312000) acc5: 100.000000 (97.728000) time: 0.274197 data: 0.000087 max mem: 14338 Test: Total time: 0:00:57 (0.293439 s / it) * Acc@1 85.134 Acc@5 97.304 loss 0.631 Max accuracy: 85.25% uploading checkpoint virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0025.pth to hdfs://harunava/user/guoyuanfan/HCSC/virg/experiments/classification/imagenet1k/eurnet_base_224to384_30eps_reproduce/re19/checkpoint_0025.pth Epoch: [26/30] [ 0/5004] eta: 2:40:18 lr: 0.000004 loss: 1.516706 (1.516706) time: 1.922183 data: 1.187063 max mem: 14338 Epoch: [26/30] [ 50/5004] eta: 1:00:53 lr: 0.000004 loss: 1.556948 (1.547248) time: 0.714855 data: 0.000205 max mem: 14338 Epoch: [26/30] [ 100/5004] eta: 0:59:23 lr: 0.000004 loss: 1.633552 (1.591294) time: 0.713221 data: 0.000229 max mem: 14338 Epoch: [26/30] [ 150/5004] eta: 0:58:23 lr: 0.000004 loss: 1.489355 (1.618868) time: 0.711434 data: 0.000215 max mem: 14338 Epoch: [26/30] [ 200/5004] eta: 0:57:43 lr: 0.000004 loss: 1.552238 (1.608618) time: 0.720165 data: 0.000210 max mem: 14338 Epoch: [26/30] [ 250/5004] eta: 0:57:01 lr: 0.000004 loss: 1.515862 (1.606790) time: 0.712562 data: 0.000216 max mem: 14338 Epoch: [26/30] [ 300/5004] eta: 0:56:22 lr: 0.000004 loss: 1.598212 (1.614837) time: 0.718270 data: 0.000191 max mem: 14338 Epoch: [26/30] [ 350/5004] eta: 0:55:43 lr: 0.000004 loss: 1.676739 (1.616350) time: 0.717201 data: 0.000167 max mem: 14338 Epoch: [26/30] [ 400/5004] eta: 0:55:06 lr: 0.000004 loss: 1.509328 (1.619162) time: 0.718970 data: 0.000210 max mem: 14338 Epoch: [26/30] [ 450/5004] eta: 0:54:29 lr: 0.000004 loss: 1.549239 (1.614827) time: 0.722578 data: 0.000219 max mem: 14338 Epoch: [26/30] [ 500/5004] eta: 0:53:51 lr: 0.000004 loss: 1.500683 (1.617394) time: 0.710046 data: 0.000213 max mem: 14338 Epoch: [26/30] [ 550/5004] eta: 0:53:14 lr: 0.000004 loss: 1.525824 (1.617720) time: 0.709612 data: 0.000219 max mem: 14338 Epoch: [26/30] [ 600/5004] eta: 0:52:37 lr: 0.000004 loss: 1.460466 (1.614262) time: 0.712397 data: 0.000218 max mem: 14338 Epoch: [26/30] [ 650/5004] eta: 0:52:00 lr: 0.000004 loss: 1.538424 (1.615286) time: 0.712661 data: 0.000157 max mem: 14338 Epoch: [26/30] [ 700/5004] eta: 0:51:23 lr: 0.000004 loss: 1.535707 (1.615652) time: 0.712032 data: 0.000163 max mem: 14338 Epoch: [26/30] [ 750/5004] eta: 0:50:47 lr: 0.000004 loss: 1.510227 (1.606127) time: 0.711908 data: 0.000226 max mem: 14338 Epoch: [26/30] [ 800/5004] eta: 0:50:11 lr: 0.000004 loss: 1.459795 (1.602301) time: 0.716314 data: 0.000224 max mem: 14338 Epoch: [26/30] [ 850/5004] eta: 0:49:35 lr: 0.000004 loss: 1.614455 (1.601060) time: 0.725764 data: 0.000220 max mem: 14338 Epoch: [26/30] [ 900/5004] eta: 0:48:59 lr: 0.000004 loss: 1.559694 (1.603705) time: 0.715092 data: 0.000186 max mem: 14338 Epoch: [26/30] [ 950/5004] eta: 0:48:23 lr: 0.000004 loss: 1.644880 (1.603296) time: 0.716330 data: 0.000210 max mem: 14338 Epoch: [26/30] [1000/5004] eta: 0:47:47 lr: 0.000004 loss: 1.476107 (1.599408) time: 0.711431 data: 0.000166 max mem: 14338 Epoch: [26/30] [1050/5004] eta: 0:47:10 lr: 0.000004 loss: 1.685350 (1.599996) time: 0.714233 data: 0.000219 max mem: 14338 Epoch: [26/30] [1100/5004] eta: 0:46:34 lr: 0.000004 loss: 1.471915 (1.599910) time: 0.709792 data: 0.000212 max mem: 14338 Epoch: [26/30] [1150/5004] eta: 0:45:58 lr: 0.000004 loss: 1.590828 (1.602762) time: 0.714436 data: 0.000206 max mem: 14338 Epoch: [26/30] [1200/5004] eta: 0:45:23 lr: 0.000004 loss: 1.545546 (1.602295) time: 0.712496 data: 0.000221 max mem: 14338 Epoch: [26/30] [1250/5004] eta: 0:44:47 lr: 0.000004 loss: 1.574252 (1.603720) time: 0.716247 data: 0.000233 max mem: 14338 Epoch: [26/30] [1300/5004] eta: 0:44:11 lr: 0.000004 loss: 1.495414 (1.602582) time: 0.714209 data: 0.000184 max mem: 14338 Epoch: [26/30] [1350/5004] eta: 0:43:34 lr: 0.000004 loss: 1.646009 (1.602164) time: 0.715443 data: 0.000180 max mem: 14338 Epoch: [26/30] [1400/5004] eta: 0:42:58 lr: 0.000004 loss: 1.372282 (1.601924) time: 0.713633 data: 0.000205 max mem: 14338 Epoch: [26/30] [1450/5004] eta: 0:42:22 lr: 0.000004 loss: 1.495555 (1.601958) time: 0.716965 data: 0.000215 max mem: 14338 Epoch: [26/30] [1500/5004] eta: 0:41:46 lr: 0.000004 loss: 1.507345 (1.599597) time: 0.711778 data: 0.000216 max mem: 14338 Epoch: [26/30] [1550/5004] eta: 0:41:10 lr: 0.000004 loss: 1.465399 (1.600889) time: 0.716127 data: 0.000207 max mem: 14338 Epoch: [26/30] [1600/5004] eta: 0:40:35 lr: 0.000004 loss: 1.721529 (1.601251) time: 0.717566 data: 0.000222 max mem: 14338 Epoch: [26/30] [1650/5004] eta: 0:39:59 lr: 0.000004 loss: 1.567129 (1.601086) time: 0.711784 data: 0.000168 max mem: 14338 Epoch: [26/30] [1700/5004] eta: 0:39:23 lr: 0.000004 loss: 1.441135 (1.599538) time: 0.710710 data: 0.000172 max mem: 14338 Epoch: [26/30] [1750/5004] eta: 0:38:47 lr: 0.000004 loss: 1.470484 (1.598595) time: 0.717757 data: 0.000225 max mem: 14338 Epoch: [26/30] [1800/5004] eta: 0:38:12 lr: 0.000004 loss: 1.564329 (1.599313) time: 0.721524 data: 0.000230 max mem: 14338 Epoch: [26/30] [1850/5004] eta: 0:37:36 lr: 0.000004 loss: 1.448520 (1.600060) time: 0.718289 data: 0.000216 max mem: 14338 Epoch: [26/30] [1900/5004] eta: 0:37:00 lr: 0.000004 loss: 1.611802 (1.601759) time: 0.709039 data: 0.000221 max mem: 14338 Epoch: [26/30] [1950/5004] eta: 0:36:23 lr: 0.000004 loss: 1.592855 (1.600978) time: 0.708948 data: 0.000221 max mem: 14338 Epoch: [26/30] [2000/5004] eta: 0:35:48 lr: 0.000004 loss: 1.569579 (1.601150) time: 0.711973 data: 0.000168 max mem: 14338 Epoch: [26/30] [2050/5004] eta: 0:35:12 lr: 0.000004 loss: 1.530642 (1.600413) time: 0.710794 data: 0.000161 max mem: 14338 Epoch: [26/30] [2100/5004] eta: 0:34:36 lr: 0.000004 loss: 1.708099 (1.601856) time: 0.708789 data: 0.000215 max mem: 14338 Epoch: [26/30] [2150/5004] eta: 0:34:00 lr: 0.000004 loss: 1.666712 (1.602457) time: 0.711317 data: 0.000209 max mem: 14338 Epoch: [26/30] [2200/5004] eta: 0:33:24 lr: 0.000003 loss: 1.476308 (1.604231) time: 0.713654 data: 0.000193 max mem: 14338 Epoch: [26/30] [2250/5004] eta: 0:32:49 lr: 0.000003 loss: 1.659844 (1.606199) time: 0.726782 data: 0.000198 max mem: 14338 Epoch: [26/30] [2300/5004] eta: 0:32:13 lr: 0.000003 loss: 1.475186 (1.605651) time: 0.717523 data: 0.000234 max mem: 14338 Epoch: [26/30] [2350/5004] eta: 0:31:37 lr: 0.000003 loss: 1.427369 (1.604912) time: 0.710102 data: 0.000159 max mem: 14338 Epoch: [26/30] [2400/5004] eta: 0:31:01 lr: 0.000003 loss: 1.592429 (1.606179) time: 0.711694 data: 0.000217 max mem: 14338 Epoch: [26/30] [2450/5004] eta: 0:30:25 lr: 0.000003 loss: 1.490768 (1.606319) time: 0.711647 data: 0.000217 max mem: 14338 Epoch: [26/30] [2500/5004] eta: 0:29:49 lr: 0.000003 loss: 1.522057 (1.606032) time: 0.710416 data: 0.000220 max mem: 14338 Epoch: [26/30] [2550/5004] eta: 0:29:13 lr: 0.000003 loss: 1.594154 (1.605781) time: 0.710694 data: 0.000241 max mem: 14338 Epoch: [26/30] [2600/5004] eta: 0:28:38 lr: 0.000003 loss: 1.420544 (1.605579) time: 0.718786 data: 0.000232 max mem: 14338 Epoch: [26/30] [2650/5004] eta: 0:28:02 lr: 0.000003 loss: 1.568736 (1.605278) time: 0.714479 data: 0.000173 max mem: 14338 Epoch: [26/30] [2700/5004] eta: 0:27:26 lr: 0.000003 loss: 1.576753 (1.604445) time: 0.717657 data: 0.000176 max mem: 14338 Epoch: [26/30] [2750/5004] eta: 0:26:50 lr: 0.000003 loss: 1.525178 (1.604688) time: 0.711929 data: 0.000209 max mem: 14338 Epoch: [26/30] [2800/5004] eta: 0:26:15 lr: 0.000003 loss: 1.641007 (1.605899) time: 0.714261 data: 0.000220 max mem: 14338 Epoch: [26/30] [2850/5004] eta: 0:25:39 lr: 0.000003 loss: 1.482470 (1.606018) time: 0.715584 data: 0.000190 max mem: 14338 Epoch: [26/30] [2900/5004] eta: 0:25:03 lr: 0.000003 loss: 1.676163 (1.606688) time: 0.712098 data: 0.000230 max mem: 14338 Epoch: [26/30] [2950/5004] eta: 0:24:27 lr: 0.000003 loss: 1.529162 (1.606561) time: 0.712960 data: 0.000226 max mem: 14338 Epoch: [26/30] [3000/5004] eta: 0:23:52 lr: 0.000003 loss: 1.502649 (1.606207) time: 0.712933 data: 0.000167 max mem: 14338 Epoch: [26/30] [3050/5004] eta: 0:23:16 lr: 0.000003 loss: 1.496394 (1.607310) time: 0.715988 data: 0.000153 max mem: 14338 Epoch: [26/30] [3100/5004] eta: 0:22:40 lr: 0.000003 loss: 1.441579 (1.606666) time: 0.717141 data: 0.000243 max mem: 14338 Epoch: [26/30] [3150/5004] eta: 0:22:04 lr: 0.000003 loss: 1.600559 (1.606414) time: 0.709905 data: 0.000233 max mem: 14338 Epoch: [26/30] [3200/5004] eta: 0:21:29 lr: 0.000003 loss: 1.485827 (1.606619) time: 0.721070 data: 0.000227 max mem: 14338 Epoch: [26/30] [3250/5004] eta: 0:20:53 lr: 0.000003 loss: 1.689738 (1.606983) time: 0.712510 data: 0.000234 max mem: 14338 Epoch: [26/30] [3300/5004] eta: 0:20:17 lr: 0.000003 loss: 1.575984 (1.606622) time: 0.709607 data: 0.000218 max mem: 14338 Epoch: [26/30] [3350/5004] eta: 0:19:41 lr: 0.000003 loss: 1.520493 (1.607437) time: 0.709757 data: 0.000162 max mem: 14338 Epoch: [26/30] [3400/5004] eta: 0:19:06 lr: 0.000003 loss: 1.552246 (1.607600) time: 0.716201 data: 0.000166 max mem: 14338 Epoch: [26/30] [3450/5004] eta: 0:18:30 lr: 0.000003 loss: 1.513039 (1.607938) time: 0.715712 data: 0.000205 max mem: 14338 Epoch: [26/30] [3500/5004] eta: 0:17:54 lr: 0.000003 loss: 1.608623 (1.608693) time: 0.713435 data: 0.000190 max mem: 14338 Epoch: [26/30] [3550/5004] eta: 0:17:18 lr: 0.000003 loss: 1.556626 (1.607997) time: 0.709107 data: 0.000213 max mem: 14338 Epoch: [26/30] [3600/5004] eta: 0:16:43 lr: 0.000003 loss: 1.632762 (1.608871) time: 0.720754 data: 0.000218 max mem: 14338 Epoch: [26/30] [3650/5004] eta: 0:16:07 lr: 0.000003 loss: 1.733307 (1.609520) time: 0.716186 data: 0.000218 max mem: 14338 Epoch: [26/30] [3700/5004] eta: 0:15:31 lr: 0.000003 loss: 1.526029 (1.610509) time: 0.712009 data: 0.000170 max mem: 14338 Epoch: [26/30] [3750/5004] eta: 0:14:55 lr: 0.000003 loss: 1.511774 (1.609770) time: 0.709164 data: 0.000194 max mem: 14338 Epoch: [26/30] [3800/5004] eta: 0:14:20 lr: 0.000003 loss: 1.508120 (1.609285) time: 0.712343 data: 0.000209 max mem: 14338 Epoch: [26/30] [3850/5004] eta: 0:13:44 lr: 0.000003 loss: 1.589609 (1.610318) time: 0.712041 data: 0.000204 max mem: 14338 Epoch: [26/30] [3900/5004] eta: 0:13:08 lr: 0.000003 loss: 1.582281 (1.610210) time: 0.709875 data: 0.000224 max mem: 14338 Epoch: [26/30] [3950/5004] eta: 0:12:33 lr: 0.000003 loss: 1.443218 (1.609930) time: 0.708626 data: 0.000213 max mem: 14338 Epoch: [26/30] [4000/5004] eta: 0:11:57 lr: 0.000003 loss: 1.403450 (1.610075) time: 0.714182 data: 0.000172 max mem: 14338 Epoch: [26/30] [4050/5004] eta: 0:11:21 lr: 0.000003 loss: 1.618507 (1.609990) time: 0.717005 data: 0.000164 max mem: 14338 Epoch: [26/30] [4100/5004] eta: 0:10:45 lr: 0.000003 loss: 1.644485 (1.610240) time: 0.712681 data: 0.000232 max mem: 14338 Epoch: [26/30] [4150/5004] eta: 0:10:10 lr: 0.000003 loss: 1.678619 (1.610086) time: 0.719812 data: 0.000212 max mem: 14338 Epoch: [26/30] [4200/5004] eta: 0:09:34 lr: 0.000003 loss: 1.536773 (1.609790) time: 0.711847 data: 0.000210 max mem: 14338 Epoch: [26/30] [4250/5004] eta: 0:08:58 lr: 0.000003 loss: 1.588442 (1.610183) time: 0.714997 data: 0.000202 max mem: 14338 Epoch: [26/30] [4300/5004] eta: 0:08:22 lr: 0.000003 loss: 1.657073 (1.610778) time: 0.711156 data: 0.000229 max mem: 14338 Epoch: [26/30] [4350/5004] eta: 0:07:47 lr: 0.000003 loss: 1.571587 (1.610242) time: 0.708782 data: 0.000169 max mem: 14338 Epoch: [26/30] [4400/5004] eta: 0:07:11 lr: 0.000003 loss: 1.550487 (1.610434) time: 0.717534 data: 0.000169 max mem: 14338 Epoch: [26/30] [4450/5004] eta: 0:06:35 lr: 0.000003 loss: 1.552005 (1.609960) time: 0.715973 data: 0.000217 max mem: 14338 Epoch: [26/30] [4500/5004] eta: 0:06:00 lr: 0.000003 loss: 1.461589 (1.608992) time: 0.711560 data: 0.000218 max mem: 14338 Epoch: [26/30] [4550/5004] eta: 0:05:24 lr: 0.000003 loss: 1.562108 (1.609091) time: 0.714446 data: 0.000218 max mem: 14338 Epoch: [26/30] [4600/5004] eta: 0:04:48 lr: 0.000003 loss: 1.527701 (1.608938) time: 0.719580 data: 0.000229 max mem: 14338 Epoch: [26/30] [4650/5004] eta: 0:04:12 lr: 0.000003 loss: 1.395433 (1.608738) time: 0.711950 data: 0.000221 max mem: 14338 Epoch: [26/30] [4700/5004] eta: 0:03:37 lr: 0.000003 loss: 1.491015 (1.608505) time: 0.709805 data: 0.000161 max mem: 14338 Epoch: [26/30] [4750/5004] eta: 0:03:01 lr: 0.000003 loss: 1.512947 (1.609132) time: 0.710242 data: 0.000183 max mem: 14338 Epoch: [26/30] [4800/5004] eta: 0:02:25 lr: 0.000003 loss: 1.472763 (1.609770) time: 0.715306 data: 0.000189 max mem: 14338 Epoch: [26/30] [4850/5004] eta: 0:01:50 lr: 0.000003 loss: 1.689039 (1.610160) time: 0.714499 data: 0.000204 max mem: 14338 Epoch: [26/30] [4900/5004] eta: 0:01:14 lr: 0.000003 loss: 1.600661 (1.609990) time: 0.713427 data: 0.000209 max mem: 14338 Epoch: [26/30] [4950/5004] eta: 0:00:38 lr: 0.000003 loss: 1.578212 (1.609952) time: 0.711852 data: 0.000227 max mem: 14338 Epoch: [26/30] [5000/5004] eta: 0:00:02 lr: 0.000003 loss: 1.661873 (1.610477) time: 0.715188 data: 0.000823 max mem: 14338 Epoch: [26/30] [5003/5004] eta: 0:00:00 lr: 0.000003 loss: 1.800980 (1.610651) time: 0.712206 data: 0.000815 max mem: 14338 Epoch: [26/30] Total time: 0:59:35 (0.714489 s / it) Averaged stats: lr: 0.000003 loss: 1.800980 (1.621076) Test: [ 0/196] eta: 0:04:51 loss: 0.281495 (0.281495) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.486860 data: 1.091325 max mem: 14338 Test: [ 10/196] eta: 0:01:13 loss: 0.447640 (0.532243) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.396854 data: 0.099319 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.548729 (0.534888) acc1: 87.500000 (85.119048) acc5: 100.000000 (98.214286) time: 0.287162 data: 0.000112 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.521664 (0.512055) acc1: 87.500000 (86.693548) acc5: 100.000000 (98.387097) time: 0.286155 data: 0.000121 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.437252 (0.519467) acc1: 87.500000 (86.432927) acc5: 100.000000 (98.018293) time: 0.285746 data: 0.000140 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.463209 (0.548529) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.549020) time: 0.285592 data: 0.000134 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.608149 (0.576265) acc1: 87.500000 (85.963115) acc5: 93.750000 (97.438525) time: 0.285969 data: 0.000136 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.653787 (0.593876) acc1: 81.250000 (85.563380) acc5: 100.000000 (97.535211) time: 0.286258 data: 0.000140 max mem: 14338 Test: [ 80/196] eta: 0:00:34 loss: 0.492964 (0.593357) acc1: 87.500000 (85.570988) acc5: 100.000000 (97.685185) time: 0.285802 data: 0.000126 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.512917 (0.616317) acc1: 87.500000 (85.233516) acc5: 100.000000 (97.458791) time: 0.293865 data: 0.000130 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.589203 (0.605790) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.648515) time: 0.295235 data: 0.000126 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.551950 (0.594819) acc1: 87.500000 (85.472973) acc5: 100.000000 (97.747748) time: 0.286999 data: 0.000110 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.517067 (0.590916) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.727273) time: 0.285400 data: 0.000128 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.607094 (0.606092) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.709924) time: 0.285345 data: 0.000128 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.631199 (0.605276) acc1: 81.250000 (85.372340) acc5: 100.000000 (97.695035) time: 0.285827 data: 0.000122 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.544243 (0.614107) acc1: 87.500000 (85.223510) acc5: 100.000000 (97.723510) time: 0.286057 data: 0.000123 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.530323 (0.619407) acc1: 81.250000 (85.170807) acc5: 100.000000 (97.748447) time: 0.285733 data: 0.000120 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.508199 (0.616402) acc1: 81.250000 (85.270468) acc5: 100.000000 (97.807018) time: 0.285404 data: 0.000138 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.470354 (0.616257) acc1: 87.500000 (85.151934) acc5: 100.000000 (97.790055) time: 0.285079 data: 0.000132 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.359392 (0.604273) acc1: 87.500000 (85.373037) acc5: 100.000000 (97.807592) time: 0.282971 data: 0.000096 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.483219 (0.614723) acc1: 87.500000 (85.248000) acc5: 100.000000 (97.728000) time: 0.273631 data: 0.000085 max mem: 14338 Test: Total time: 0:00:57 (0.292902 s / it) * Acc@1 85.110 Acc@5 97.298 loss 0.632 Max accuracy: 85.25% Epoch: [27/30] [ 0/5004] eta: 2:38:53 lr: 0.000003 loss: 1.507064 (1.507064) time: 1.905163 data: 1.174964 max mem: 14338 Epoch: [27/30] [ 50/5004] eta: 1:00:59 lr: 0.000003 loss: 1.448813 (1.650106) time: 0.721842 data: 0.000185 max mem: 14338 Epoch: [27/30] [ 100/5004] eta: 0:59:17 lr: 0.000003 loss: 1.674638 (1.664290) time: 0.712183 data: 0.000211 max mem: 14338 Epoch: [27/30] [ 150/5004] eta: 0:58:27 lr: 0.000003 loss: 1.552331 (1.638374) time: 0.716524 data: 0.000230 max mem: 14338 Epoch: [27/30] [ 200/5004] eta: 0:57:42 lr: 0.000003 loss: 1.546680 (1.618854) time: 0.714130 data: 0.000210 max mem: 14338 Epoch: [27/30] [ 250/5004] eta: 0:56:58 lr: 0.000003 loss: 1.527782 (1.617591) time: 0.712697 data: 0.000189 max mem: 14338 Epoch: [27/30] [ 300/5004] eta: 0:56:17 lr: 0.000003 loss: 1.541327 (1.611162) time: 0.709229 data: 0.000168 max mem: 14338 Epoch: [27/30] [ 350/5004] eta: 0:55:39 lr: 0.000003 loss: 1.526647 (1.612835) time: 0.710788 data: 0.000171 max mem: 14338 Epoch: [27/30] [ 400/5004] eta: 0:55:03 lr: 0.000003 loss: 1.542892 (1.606035) time: 0.717639 data: 0.000215 max mem: 14338 Epoch: [27/30] [ 450/5004] eta: 0:54:26 lr: 0.000003 loss: 1.586040 (1.605194) time: 0.714898 data: 0.000203 max mem: 14338 Epoch: [27/30] [ 500/5004] eta: 0:53:49 lr: 0.000003 loss: 1.525066 (1.609879) time: 0.711128 data: 0.000202 max mem: 14338 Epoch: [27/30] [ 550/5004] eta: 0:53:11 lr: 0.000003 loss: 1.721706 (1.615109) time: 0.711076 data: 0.000220 max mem: 14338 Epoch: [27/30] [ 600/5004] eta: 0:52:36 lr: 0.000003 loss: 1.515527 (1.612783) time: 0.723302 data: 0.000225 max mem: 14338 Epoch: [27/30] [ 650/5004] eta: 0:51:59 lr: 0.000003 loss: 1.536729 (1.616656) time: 0.711020 data: 0.000180 max mem: 14338 Epoch: [27/30] [ 700/5004] eta: 0:51:22 lr: 0.000003 loss: 1.497467 (1.613569) time: 0.710417 data: 0.000164 max mem: 14338 Epoch: [27/30] [ 750/5004] eta: 0:50:45 lr: 0.000003 loss: 1.651356 (1.619259) time: 0.711758 data: 0.000232 max mem: 14338 Epoch: [27/30] [ 800/5004] eta: 0:50:09 lr: 0.000003 loss: 1.532396 (1.620454) time: 0.712977 data: 0.000202 max mem: 14338 Epoch: [27/30] [ 850/5004] eta: 0:49:32 lr: 0.000003 loss: 1.673957 (1.623737) time: 0.712128 data: 0.000209 max mem: 14338 Epoch: [27/30] [ 900/5004] eta: 0:48:56 lr: 0.000002 loss: 1.572173 (1.618029) time: 0.709666 data: 0.000183 max mem: 14338 Epoch: [27/30] [ 950/5004] eta: 0:48:20 lr: 0.000002 loss: 1.682228 (1.623332) time: 0.716592 data: 0.000232 max mem: 14338 Epoch: [27/30] [1000/5004] eta: 0:47:44 lr: 0.000002 loss: 1.388208 (1.619740) time: 0.715415 data: 0.000162 max mem: 14338 Epoch: [27/30] [1050/5004] eta: 0:47:08 lr: 0.000002 loss: 1.558052 (1.619093) time: 0.721700 data: 0.000205 max mem: 14338 Epoch: [27/30] [1100/5004] eta: 0:46:31 lr: 0.000002 loss: 1.632299 (1.621156) time: 0.711542 data: 0.000216 max mem: 14338 Epoch: [27/30] [1150/5004] eta: 0:45:55 lr: 0.000002 loss: 1.561636 (1.623264) time: 0.709137 data: 0.000217 max mem: 14338 Epoch: [27/30] [1200/5004] eta: 0:45:19 lr: 0.000002 loss: 1.694642 (1.623730) time: 0.712555 data: 0.000214 max mem: 14338 Epoch: [27/30] [1250/5004] eta: 0:44:43 lr: 0.000002 loss: 1.497610 (1.624881) time: 0.713790 data: 0.000216 max mem: 14338 Epoch: [27/30] [1300/5004] eta: 0:44:07 lr: 0.000002 loss: 1.611342 (1.624680) time: 0.709402 data: 0.000190 max mem: 14338 Epoch: [27/30] [1350/5004] eta: 0:43:31 lr: 0.000002 loss: 1.698581 (1.625730) time: 0.712536 data: 0.000163 max mem: 14338 Epoch: [27/30] [1400/5004] eta: 0:42:56 lr: 0.000002 loss: 1.697281 (1.625947) time: 0.719751 data: 0.000229 max mem: 14338 Epoch: [27/30] [1450/5004] eta: 0:42:20 lr: 0.000002 loss: 1.433712 (1.625044) time: 0.716230 data: 0.000220 max mem: 14338 Epoch: [27/30] [1500/5004] eta: 0:41:44 lr: 0.000002 loss: 1.787348 (1.627102) time: 0.716778 data: 0.000220 max mem: 14338 Epoch: [27/30] [1550/5004] eta: 0:41:08 lr: 0.000002 loss: 1.605742 (1.625392) time: 0.713157 data: 0.000181 max mem: 14338 Epoch: [27/30] [1600/5004] eta: 0:40:33 lr: 0.000002 loss: 1.643455 (1.626604) time: 0.712405 data: 0.000207 max mem: 14338 Epoch: [27/30] [1650/5004] eta: 0:39:57 lr: 0.000002 loss: 1.507811 (1.625426) time: 0.711384 data: 0.000176 max mem: 14338 Epoch: [27/30] [1700/5004] eta: 0:39:21 lr: 0.000002 loss: 1.540282 (1.627194) time: 0.711262 data: 0.000170 max mem: 14338 Epoch: [27/30] [1750/5004] eta: 0:38:45 lr: 0.000002 loss: 1.604069 (1.625391) time: 0.710738 data: 0.000220 max mem: 14338 Epoch: [27/30] [1800/5004] eta: 0:38:10 lr: 0.000002 loss: 1.676387 (1.623954) time: 0.717845 data: 0.000212 max mem: 14338 Epoch: [27/30] [1850/5004] eta: 0:37:34 lr: 0.000002 loss: 1.559911 (1.623451) time: 0.717070 data: 0.000223 max mem: 14338 Epoch: [27/30] [1900/5004] eta: 0:36:58 lr: 0.000002 loss: 1.518449 (1.621608) time: 0.715772 data: 0.000234 max mem: 14338 Epoch: [27/30] [1950/5004] eta: 0:36:23 lr: 0.000002 loss: 1.573588 (1.619862) time: 0.714288 data: 0.000216 max mem: 14338 Epoch: [27/30] [2000/5004] eta: 0:35:47 lr: 0.000002 loss: 1.524551 (1.620420) time: 0.720155 data: 0.000175 max mem: 14338 Epoch: [27/30] [2050/5004] eta: 0:35:11 lr: 0.000002 loss: 1.688169 (1.621181) time: 0.711995 data: 0.000173 max mem: 14338 Epoch: [27/30] [2100/5004] eta: 0:34:35 lr: 0.000002 loss: 1.500714 (1.619868) time: 0.710163 data: 0.000232 max mem: 14338 Epoch: [27/30] [2150/5004] eta: 0:33:59 lr: 0.000002 loss: 1.583911 (1.619973) time: 0.708931 data: 0.000243 max mem: 14338 Epoch: [27/30] [2200/5004] eta: 0:33:24 lr: 0.000002 loss: 1.605802 (1.621084) time: 0.717095 data: 0.000235 max mem: 14338 Epoch: [27/30] [2250/5004] eta: 0:32:48 lr: 0.000002 loss: 1.611234 (1.621853) time: 0.712572 data: 0.000226 max mem: 14338 Epoch: [27/30] [2300/5004] eta: 0:32:12 lr: 0.000002 loss: 1.576302 (1.621345) time: 0.710734 data: 0.000219 max mem: 14338 Epoch: [27/30] [2350/5004] eta: 0:31:36 lr: 0.000002 loss: 1.613531 (1.622064) time: 0.719146 data: 0.000164 max mem: 14338 Epoch: [27/30] [2400/5004] eta: 0:31:01 lr: 0.000002 loss: 1.563398 (1.622156) time: 0.720990 data: 0.000223 max mem: 14338 Epoch: [27/30] [2450/5004] eta: 0:30:25 lr: 0.000002 loss: 1.533032 (1.622544) time: 0.716748 data: 0.000222 max mem: 14338 Epoch: [27/30] [2500/5004] eta: 0:29:49 lr: 0.000002 loss: 1.529714 (1.623535) time: 0.712326 data: 0.000225 max mem: 14338 Epoch: [27/30] [2550/5004] eta: 0:29:13 lr: 0.000002 loss: 1.588923 (1.624157) time: 0.710512 data: 0.000202 max mem: 14338 Epoch: [27/30] [2600/5004] eta: 0:28:37 lr: 0.000002 loss: 1.597971 (1.626042) time: 0.713598 data: 0.000205 max mem: 14338 Epoch: [27/30] [2650/5004] eta: 0:28:02 lr: 0.000002 loss: 1.629772 (1.626814) time: 0.712516 data: 0.000180 max mem: 14338 Epoch: [27/30] [2700/5004] eta: 0:27:26 lr: 0.000002 loss: 1.621994 (1.627524) time: 0.710739 data: 0.000178 max mem: 14338 Epoch: [27/30] [2750/5004] eta: 0:26:50 lr: 0.000002 loss: 1.566149 (1.627594) time: 0.709714 data: 0.000212 max mem: 14338 Epoch: [27/30] [2800/5004] eta: 0:26:14 lr: 0.000002 loss: 1.639861 (1.628918) time: 0.717476 data: 0.000233 max mem: 14338 Epoch: [27/30] [2850/5004] eta: 0:25:39 lr: 0.000002 loss: 1.501237 (1.627834) time: 0.715299 data: 0.000177 max mem: 14338 Epoch: [27/30] [2900/5004] eta: 0:25:03 lr: 0.000002 loss: 1.492710 (1.626740) time: 0.715557 data: 0.000207 max mem: 14338 Epoch: [27/30] [2950/5004] eta: 0:24:27 lr: 0.000002 loss: 1.698293 (1.627022) time: 0.714312 data: 0.000234 max mem: 14338 Epoch: [27/30] [3000/5004] eta: 0:23:51 lr: 0.000002 loss: 1.664051 (1.628458) time: 0.714004 data: 0.000176 max mem: 14338 Epoch: [27/30] [3050/5004] eta: 0:23:16 lr: 0.000002 loss: 1.614672 (1.628970) time: 0.711952 data: 0.000169 max mem: 14338 Epoch: [27/30] [3100/5004] eta: 0:22:40 lr: 0.000002 loss: 1.512072 (1.628222) time: 0.708999 data: 0.000220 max mem: 14338 Epoch: [27/30] [3150/5004] eta: 0:22:04 lr: 0.000002 loss: 1.529423 (1.628465) time: 0.712180 data: 0.000227 max mem: 14338 Epoch: [27/30] [3200/5004] eta: 0:21:28 lr: 0.000002 loss: 1.485063 (1.628990) time: 0.717381 data: 0.000218 max mem: 14338 Epoch: [27/30] [3250/5004] eta: 0:20:53 lr: 0.000002 loss: 1.633859 (1.629877) time: 0.713612 data: 0.000219 max mem: 14338 Epoch: [27/30] [3300/5004] eta: 0:20:17 lr: 0.000002 loss: 1.499607 (1.629147) time: 0.712810 data: 0.000218 max mem: 14338 Epoch: [27/30] [3350/5004] eta: 0:19:41 lr: 0.000002 loss: 1.448373 (1.628817) time: 0.714704 data: 0.000175 max mem: 14338 Epoch: [27/30] [3400/5004] eta: 0:19:05 lr: 0.000002 loss: 1.395787 (1.627706) time: 0.714187 data: 0.000171 max mem: 14338 Epoch: [27/30] [3450/5004] eta: 0:18:30 lr: 0.000002 loss: 1.485234 (1.626468) time: 0.719819 data: 0.000225 max mem: 14338 Epoch: [27/30] [3500/5004] eta: 0:17:54 lr: 0.000002 loss: 1.544154 (1.625992) time: 0.713255 data: 0.000194 max mem: 14338 Epoch: [27/30] [3550/5004] eta: 0:17:18 lr: 0.000002 loss: 1.530910 (1.626104) time: 0.709949 data: 0.000228 max mem: 14338 Epoch: [27/30] [3600/5004] eta: 0:16:43 lr: 0.000002 loss: 1.562551 (1.625250) time: 0.713888 data: 0.000198 max mem: 14338 Epoch: [27/30] [3650/5004] eta: 0:16:07 lr: 0.000002 loss: 1.591907 (1.625057) time: 0.712725 data: 0.000210 max mem: 14338 Epoch: [27/30] [3700/5004] eta: 0:15:31 lr: 0.000002 loss: 1.568493 (1.625732) time: 0.711277 data: 0.000179 max mem: 14338 Epoch: [27/30] [3750/5004] eta: 0:14:55 lr: 0.000002 loss: 1.481198 (1.624870) time: 0.712250 data: 0.000240 max mem: 14338 Epoch: [27/30] [3800/5004] eta: 0:14:20 lr: 0.000002 loss: 1.589458 (1.625044) time: 0.716221 data: 0.000231 max mem: 14338 Epoch: [27/30] [3850/5004] eta: 0:13:44 lr: 0.000002 loss: 1.603585 (1.624816) time: 0.719693 data: 0.000219 max mem: 14338 Epoch: [27/30] [3900/5004] eta: 0:13:08 lr: 0.000002 loss: 1.644297 (1.624915) time: 0.708953 data: 0.000229 max mem: 14338 Epoch: [27/30] [3950/5004] eta: 0:12:32 lr: 0.000002 loss: 1.616355 (1.624699) time: 0.711395 data: 0.000215 max mem: 14338 Epoch: [27/30] [4000/5004] eta: 0:11:57 lr: 0.000002 loss: 1.606088 (1.624299) time: 0.712399 data: 0.000168 max mem: 14338 Epoch: [27/30] [4050/5004] eta: 0:11:21 lr: 0.000002 loss: 1.661604 (1.625089) time: 0.711148 data: 0.000164 max mem: 14338 Epoch: [27/30] [4100/5004] eta: 0:10:45 lr: 0.000002 loss: 1.574348 (1.625065) time: 0.710929 data: 0.000210 max mem: 14338 Epoch: [27/30] [4150/5004] eta: 0:10:10 lr: 0.000002 loss: 1.553481 (1.625213) time: 0.712469 data: 0.000195 max mem: 14338 Epoch: [27/30] [4200/5004] eta: 0:09:34 lr: 0.000002 loss: 1.567757 (1.625129) time: 0.723336 data: 0.000218 max mem: 14338 Epoch: [27/30] [4250/5004] eta: 0:08:58 lr: 0.000002 loss: 1.666232 (1.625464) time: 0.718393 data: 0.000229 max mem: 14338 Epoch: [27/30] [4300/5004] eta: 0:08:22 lr: 0.000002 loss: 1.651260 (1.625582) time: 0.715018 data: 0.000211 max mem: 14338 Epoch: [27/30] [4350/5004] eta: 0:07:47 lr: 0.000002 loss: 1.587831 (1.624976) time: 0.716864 data: 0.000176 max mem: 14338 Epoch: [27/30] [4400/5004] eta: 0:07:11 lr: 0.000002 loss: 1.625063 (1.625086) time: 0.716141 data: 0.000161 max mem: 14338 Epoch: [27/30] [4450/5004] eta: 0:06:35 lr: 0.000002 loss: 1.726891 (1.625104) time: 0.715815 data: 0.000215 max mem: 14338 Epoch: [27/30] [4500/5004] eta: 0:06:00 lr: 0.000002 loss: 1.586280 (1.624951) time: 0.712916 data: 0.000222 max mem: 14338 Epoch: [27/30] [4550/5004] eta: 0:05:24 lr: 0.000002 loss: 1.492129 (1.624055) time: 0.709151 data: 0.000216 max mem: 14338 Epoch: [27/30] [4600/5004] eta: 0:04:48 lr: 0.000002 loss: 1.597897 (1.623685) time: 0.715983 data: 0.000217 max mem: 14338 Epoch: [27/30] [4650/5004] eta: 0:04:12 lr: 0.000002 loss: 1.503386 (1.622977) time: 0.723934 data: 0.000220 max mem: 14338 Epoch: [27/30] [4700/5004] eta: 0:03:37 lr: 0.000002 loss: 1.568921 (1.622674) time: 0.711568 data: 0.000163 max mem: 14338 Epoch: [27/30] [4750/5004] eta: 0:03:01 lr: 0.000002 loss: 1.512019 (1.622093) time: 0.713263 data: 0.000181 max mem: 14338 Epoch: [27/30] [4800/5004] eta: 0:02:25 lr: 0.000002 loss: 1.526299 (1.622079) time: 0.718412 data: 0.000198 max mem: 14338 Epoch: [27/30] [4850/5004] eta: 0:01:50 lr: 0.000002 loss: 1.642101 (1.622383) time: 0.714630 data: 0.000196 max mem: 14338 Epoch: [27/30] [4900/5004] eta: 0:01:14 lr: 0.000002 loss: 1.566368 (1.622629) time: 0.710117 data: 0.000214 max mem: 14338 Epoch: [27/30] [4950/5004] eta: 0:00:38 lr: 0.000002 loss: 1.532657 (1.622688) time: 0.713280 data: 0.000218 max mem: 14338 Epoch: [27/30] [5000/5004] eta: 0:00:02 lr: 0.000002 loss: 1.510676 (1.622626) time: 0.712187 data: 0.000819 max mem: 14338 Epoch: [27/30] [5003/5004] eta: 0:00:00 lr: 0.000002 loss: 1.507292 (1.622385) time: 0.708762 data: 0.000810 max mem: 14338 Epoch: [27/30] Total time: 0:59:35 (0.714529 s / it) Averaged stats: lr: 0.000002 loss: 1.507292 (1.617973) Test: [ 0/196] eta: 0:05:20 loss: 0.274158 (0.274158) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.635705 data: 1.256958 max mem: 14338 Test: [ 10/196] eta: 0:01:16 loss: 0.449497 (0.539642) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.410148 data: 0.114396 max mem: 14338 Test: [ 20/196] eta: 0:01:01 loss: 0.543470 (0.538703) acc1: 87.500000 (85.416667) acc5: 100.000000 (98.214286) time: 0.287419 data: 0.000132 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.521442 (0.514619) acc1: 87.500000 (86.895161) acc5: 100.000000 (98.387097) time: 0.287360 data: 0.000124 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.433374 (0.522265) acc1: 87.500000 (86.585366) acc5: 100.000000 (98.170732) time: 0.292916 data: 0.000148 max mem: 14338 Test: [ 50/196] eta: 0:00:46 loss: 0.471220 (0.550233) acc1: 87.500000 (86.642157) acc5: 100.000000 (97.671569) time: 0.292946 data: 0.000147 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.618668 (0.577958) acc1: 87.500000 (86.065574) acc5: 93.750000 (97.540984) time: 0.287224 data: 0.000138 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.658093 (0.595236) acc1: 81.250000 (85.651408) acc5: 100.000000 (97.623239) time: 0.286917 data: 0.000145 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.498766 (0.594493) acc1: 87.500000 (85.648148) acc5: 100.000000 (97.762346) time: 0.286691 data: 0.000135 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.512835 (0.617801) acc1: 87.500000 (85.233516) acc5: 100.000000 (97.527473) time: 0.286753 data: 0.000149 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.573381 (0.606439) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.710396) time: 0.286646 data: 0.000146 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.540434 (0.595357) acc1: 87.500000 (85.472973) acc5: 100.000000 (97.804054) time: 0.285652 data: 0.000130 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.533278 (0.591367) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.778926) time: 0.286344 data: 0.000142 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.607533 (0.606511) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.757634) time: 0.287584 data: 0.000140 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.618154 (0.605614) acc1: 81.250000 (85.372340) acc5: 100.000000 (97.739362) time: 0.286448 data: 0.000140 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.550240 (0.614689) acc1: 87.500000 (85.223510) acc5: 100.000000 (97.764901) time: 0.285989 data: 0.000140 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.522891 (0.619882) acc1: 81.250000 (85.131988) acc5: 100.000000 (97.787267) time: 0.286401 data: 0.000132 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.510480 (0.616819) acc1: 81.250000 (85.233918) acc5: 100.000000 (97.807018) time: 0.286372 data: 0.000141 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.467432 (0.616757) acc1: 81.250000 (85.082873) acc5: 100.000000 (97.790055) time: 0.291398 data: 0.000156 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.371746 (0.604899) acc1: 87.500000 (85.307592) acc5: 100.000000 (97.807592) time: 0.288700 data: 0.000119 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.493693 (0.615160) acc1: 87.500000 (85.216000) acc5: 100.000000 (97.728000) time: 0.274162 data: 0.000104 max mem: 14338 Test: Total time: 0:00:57 (0.294489 s / it) * Acc@1 85.102 Acc@5 97.306 loss 0.632 Max accuracy: 85.25% Epoch: [28/30] [ 0/5004] eta: 2:38:52 lr: 0.000002 loss: 2.012574 (2.012574) time: 1.905055 data: 1.163281 max mem: 14338 Epoch: [28/30] [ 50/5004] eta: 1:01:14 lr: 0.000002 loss: 1.431764 (1.626945) time: 0.717361 data: 0.000182 max mem: 14338 Epoch: [28/30] [ 100/5004] eta: 0:59:34 lr: 0.000002 loss: 1.771378 (1.633518) time: 0.715146 data: 0.000222 max mem: 14338 Epoch: [28/30] [ 150/5004] eta: 0:58:37 lr: 0.000002 loss: 1.512151 (1.619896) time: 0.718000 data: 0.000224 max mem: 14338 Epoch: [28/30] [ 200/5004] eta: 0:57:47 lr: 0.000002 loss: 1.700583 (1.615361) time: 0.716089 data: 0.000218 max mem: 14338 Epoch: [28/30] [ 250/5004] eta: 0:57:02 lr: 0.000002 loss: 1.643929 (1.631044) time: 0.716512 data: 0.000194 max mem: 14338 Epoch: [28/30] [ 300/5004] eta: 0:56:20 lr: 0.000002 loss: 1.566243 (1.629112) time: 0.709017 data: 0.000160 max mem: 14338 Epoch: [28/30] [ 350/5004] eta: 0:55:40 lr: 0.000002 loss: 1.571598 (1.641522) time: 0.708481 data: 0.000170 max mem: 14338 Epoch: [28/30] [ 400/5004] eta: 0:55:02 lr: 0.000002 loss: 1.666864 (1.635558) time: 0.711529 data: 0.000214 max mem: 14338 Epoch: [28/30] [ 450/5004] eta: 0:54:25 lr: 0.000002 loss: 1.478190 (1.623542) time: 0.716523 data: 0.000209 max mem: 14338 Epoch: [28/30] [ 500/5004] eta: 0:53:49 lr: 0.000002 loss: 1.598492 (1.623629) time: 0.714653 data: 0.000223 max mem: 14338 Epoch: [28/30] [ 550/5004] eta: 0:53:13 lr: 0.000002 loss: 1.555471 (1.620672) time: 0.714812 data: 0.000214 max mem: 14338 Epoch: [28/30] [ 600/5004] eta: 0:52:36 lr: 0.000002 loss: 1.725396 (1.625202) time: 0.714980 data: 0.000214 max mem: 14338 Epoch: [28/30] [ 650/5004] eta: 0:51:59 lr: 0.000002 loss: 1.485202 (1.623129) time: 0.716280 data: 0.000157 max mem: 14338 Epoch: [28/30] [ 700/5004] eta: 0:51:23 lr: 0.000002 loss: 1.717681 (1.625062) time: 0.717203 data: 0.000167 max mem: 14338 Epoch: [28/30] [ 750/5004] eta: 0:50:46 lr: 0.000002 loss: 1.414755 (1.618945) time: 0.711839 data: 0.000212 max mem: 14338 Epoch: [28/30] [ 800/5004] eta: 0:50:09 lr: 0.000002 loss: 1.718189 (1.618221) time: 0.713818 data: 0.000200 max mem: 14338 Epoch: [28/30] [ 850/5004] eta: 0:49:34 lr: 0.000002 loss: 1.557863 (1.619233) time: 0.716494 data: 0.000198 max mem: 14338 Epoch: [28/30] [ 900/5004] eta: 0:48:57 lr: 0.000002 loss: 1.451852 (1.622357) time: 0.708347 data: 0.000177 max mem: 14338 Epoch: [28/30] [ 950/5004] eta: 0:48:21 lr: 0.000001 loss: 1.538222 (1.620056) time: 0.710538 data: 0.000238 max mem: 14338 Epoch: [28/30] [1000/5004] eta: 0:47:45 lr: 0.000001 loss: 1.425386 (1.617038) time: 0.711815 data: 0.000173 max mem: 14338 Epoch: [28/30] [1050/5004] eta: 0:47:09 lr: 0.000001 loss: 1.684212 (1.618978) time: 0.717952 data: 0.000212 max mem: 14338 Epoch: [28/30] [1100/5004] eta: 0:46:33 lr: 0.000001 loss: 1.637275 (1.619518) time: 0.716155 data: 0.000218 max mem: 14338 Epoch: [28/30] [1150/5004] eta: 0:45:57 lr: 0.000001 loss: 1.669621 (1.619517) time: 0.714338 data: 0.000218 max mem: 14338 Epoch: [28/30] [1200/5004] eta: 0:45:21 lr: 0.000001 loss: 1.834023 (1.620111) time: 0.713132 data: 0.000213 max mem: 14338 Epoch: [28/30] [1250/5004] eta: 0:44:45 lr: 0.000001 loss: 1.664095 (1.621226) time: 0.713049 data: 0.000220 max mem: 14338 Epoch: [28/30] [1300/5004] eta: 0:44:09 lr: 0.000001 loss: 1.655364 (1.622441) time: 0.713589 data: 0.000153 max mem: 14338 Epoch: [28/30] [1350/5004] eta: 0:43:33 lr: 0.000001 loss: 1.671374 (1.625146) time: 0.712037 data: 0.000158 max mem: 14338 Epoch: [28/30] [1400/5004] eta: 0:42:58 lr: 0.000001 loss: 1.614349 (1.624351) time: 0.716459 data: 0.000207 max mem: 14338 Epoch: [28/30] [1450/5004] eta: 0:42:22 lr: 0.000001 loss: 1.507069 (1.625570) time: 0.712378 data: 0.000220 max mem: 14338 Epoch: [28/30] [1500/5004] eta: 0:41:46 lr: 0.000001 loss: 1.556594 (1.624884) time: 0.711617 data: 0.000216 max mem: 14338 Epoch: [28/30] [1550/5004] eta: 0:41:10 lr: 0.000001 loss: 1.468138 (1.624242) time: 0.716677 data: 0.000181 max mem: 14338 Epoch: [28/30] [1600/5004] eta: 0:40:34 lr: 0.000001 loss: 1.784765 (1.625662) time: 0.716702 data: 0.000202 max mem: 14338 Epoch: [28/30] [1650/5004] eta: 0:39:59 lr: 0.000001 loss: 1.505161 (1.624755) time: 0.716618 data: 0.000152 max mem: 14338 Epoch: [28/30] [1700/5004] eta: 0:39:22 lr: 0.000001 loss: 1.501991 (1.625514) time: 0.712503 data: 0.000186 max mem: 14338 Epoch: [28/30] [1750/5004] eta: 0:38:47 lr: 0.000001 loss: 1.564344 (1.626549) time: 0.710861 data: 0.000219 max mem: 14338 Epoch: [28/30] [1800/5004] eta: 0:38:11 lr: 0.000001 loss: 1.550685 (1.625582) time: 0.713943 data: 0.000218 max mem: 14338 Epoch: [28/30] [1850/5004] eta: 0:37:35 lr: 0.000001 loss: 1.696864 (1.625246) time: 0.715849 data: 0.000232 max mem: 14338 Epoch: [28/30] [1900/5004] eta: 0:36:59 lr: 0.000001 loss: 1.628232 (1.625333) time: 0.713582 data: 0.000220 max mem: 14338 Epoch: [28/30] [1950/5004] eta: 0:36:23 lr: 0.000001 loss: 1.579390 (1.625880) time: 0.709771 data: 0.000216 max mem: 14338 Epoch: [28/30] [2000/5004] eta: 0:35:48 lr: 0.000001 loss: 1.561249 (1.624468) time: 0.722979 data: 0.000164 max mem: 14338 Epoch: [28/30] [2050/5004] eta: 0:35:12 lr: 0.000001 loss: 1.520945 (1.624250) time: 0.725791 data: 0.000167 max mem: 14338 Epoch: [28/30] [2100/5004] eta: 0:34:36 lr: 0.000001 loss: 1.518311 (1.624429) time: 0.717202 data: 0.000222 max mem: 14338 Epoch: [28/30] [2150/5004] eta: 0:34:00 lr: 0.000001 loss: 1.511390 (1.624109) time: 0.712153 data: 0.000224 max mem: 14338 Epoch: [28/30] [2200/5004] eta: 0:33:25 lr: 0.000001 loss: 1.631081 (1.625164) time: 0.713907 data: 0.000216 max mem: 14338 Epoch: [28/30] [2250/5004] eta: 0:32:49 lr: 0.000001 loss: 1.447729 (1.622510) time: 0.712571 data: 0.000229 max mem: 14338 Epoch: [28/30] [2300/5004] eta: 0:32:13 lr: 0.000001 loss: 1.716878 (1.623126) time: 0.712852 data: 0.000223 max mem: 14338 Epoch: [28/30] [2350/5004] eta: 0:31:37 lr: 0.000001 loss: 1.543813 (1.623313) time: 0.713110 data: 0.000161 max mem: 14338 Epoch: [28/30] [2400/5004] eta: 0:31:02 lr: 0.000001 loss: 1.522614 (1.622325) time: 0.712445 data: 0.000219 max mem: 14338 Epoch: [28/30] [2450/5004] eta: 0:30:26 lr: 0.000001 loss: 1.572659 (1.621335) time: 0.717556 data: 0.000212 max mem: 14338 Epoch: [28/30] [2500/5004] eta: 0:29:50 lr: 0.000001 loss: 1.571254 (1.620350) time: 0.716388 data: 0.000211 max mem: 14338 Epoch: [28/30] [2550/5004] eta: 0:29:14 lr: 0.000001 loss: 1.658728 (1.620220) time: 0.711651 data: 0.000208 max mem: 14338 Epoch: [28/30] [2600/5004] eta: 0:28:38 lr: 0.000001 loss: 1.555033 (1.620289) time: 0.719509 data: 0.000213 max mem: 14338 Epoch: [28/30] [2650/5004] eta: 0:28:03 lr: 0.000001 loss: 1.523146 (1.620487) time: 0.710763 data: 0.000163 max mem: 14338 Epoch: [28/30] [2700/5004] eta: 0:27:27 lr: 0.000001 loss: 1.578528 (1.620166) time: 0.712522 data: 0.000160 max mem: 14338 Epoch: [28/30] [2750/5004] eta: 0:26:51 lr: 0.000001 loss: 1.601392 (1.620833) time: 0.712300 data: 0.000224 max mem: 14338 Epoch: [28/30] [2800/5004] eta: 0:26:15 lr: 0.000001 loss: 1.560650 (1.619483) time: 0.711371 data: 0.000216 max mem: 14338 Epoch: [28/30] [2850/5004] eta: 0:25:39 lr: 0.000001 loss: 1.528902 (1.619267) time: 0.713369 data: 0.000181 max mem: 14338 Epoch: [28/30] [2900/5004] eta: 0:25:04 lr: 0.000001 loss: 1.529838 (1.618272) time: 0.716129 data: 0.000223 max mem: 14338 Epoch: [28/30] [2950/5004] eta: 0:24:28 lr: 0.000001 loss: 1.565722 (1.617554) time: 0.714542 data: 0.000213 max mem: 14338 Epoch: [28/30] [3000/5004] eta: 0:23:52 lr: 0.000001 loss: 1.559128 (1.617682) time: 0.721263 data: 0.000178 max mem: 14338 Epoch: [28/30] [3050/5004] eta: 0:23:17 lr: 0.000001 loss: 1.512159 (1.617992) time: 0.726503 data: 0.000164 max mem: 14338 Epoch: [28/30] [3100/5004] eta: 0:22:41 lr: 0.000001 loss: 1.555789 (1.617476) time: 0.710721 data: 0.000233 max mem: 14338 Epoch: [28/30] [3150/5004] eta: 0:22:05 lr: 0.000001 loss: 1.577583 (1.618556) time: 0.709669 data: 0.000229 max mem: 14338 Epoch: [28/30] [3200/5004] eta: 0:21:29 lr: 0.000001 loss: 1.670319 (1.619132) time: 0.712435 data: 0.000231 max mem: 14338 Epoch: [28/30] [3250/5004] eta: 0:20:53 lr: 0.000001 loss: 1.613020 (1.618918) time: 0.711707 data: 0.000213 max mem: 14338 Epoch: [28/30] [3300/5004] eta: 0:20:18 lr: 0.000001 loss: 1.537759 (1.618125) time: 0.708960 data: 0.000220 max mem: 14338 Epoch: [28/30] [3350/5004] eta: 0:19:42 lr: 0.000001 loss: 1.500152 (1.617910) time: 0.711728 data: 0.000160 max mem: 14338 Epoch: [28/30] [3400/5004] eta: 0:19:06 lr: 0.000001 loss: 1.621902 (1.618429) time: 0.718500 data: 0.000167 max mem: 14338 Epoch: [28/30] [3450/5004] eta: 0:18:30 lr: 0.000001 loss: 1.426085 (1.617719) time: 0.716639 data: 0.000209 max mem: 14338 Epoch: [28/30] [3500/5004] eta: 0:17:55 lr: 0.000001 loss: 1.432570 (1.617836) time: 0.718697 data: 0.000196 max mem: 14338 Epoch: [28/30] [3550/5004] eta: 0:17:19 lr: 0.000001 loss: 1.684463 (1.617709) time: 0.709526 data: 0.000221 max mem: 14338 Epoch: [28/30] [3600/5004] eta: 0:16:43 lr: 0.000001 loss: 1.635715 (1.617608) time: 0.712482 data: 0.000227 max mem: 14338 Epoch: [28/30] [3650/5004] eta: 0:16:07 lr: 0.000001 loss: 1.502928 (1.616307) time: 0.712321 data: 0.000207 max mem: 14338 Epoch: [28/30] [3700/5004] eta: 0:15:31 lr: 0.000001 loss: 1.549051 (1.615723) time: 0.708972 data: 0.000172 max mem: 14338 Epoch: [28/30] [3750/5004] eta: 0:14:56 lr: 0.000001 loss: 1.523617 (1.615173) time: 0.710538 data: 0.000216 max mem: 14338 Epoch: [28/30] [3800/5004] eta: 0:14:20 lr: 0.000001 loss: 1.333903 (1.614412) time: 0.711518 data: 0.000215 max mem: 14338 Epoch: [28/30] [3850/5004] eta: 0:13:44 lr: 0.000001 loss: 1.405279 (1.614468) time: 0.716236 data: 0.000218 max mem: 14338 Epoch: [28/30] [3900/5004] eta: 0:13:08 lr: 0.000001 loss: 1.501419 (1.614012) time: 0.721200 data: 0.000214 max mem: 14338 Epoch: [28/30] [3950/5004] eta: 0:12:33 lr: 0.000001 loss: 1.494205 (1.612942) time: 0.713796 data: 0.000237 max mem: 14338 Epoch: [28/30] [4000/5004] eta: 0:11:57 lr: 0.000001 loss: 1.643992 (1.613677) time: 0.716644 data: 0.000164 max mem: 14338 Epoch: [28/30] [4050/5004] eta: 0:11:21 lr: 0.000001 loss: 1.591316 (1.613968) time: 0.715174 data: 0.000180 max mem: 14338 Epoch: [28/30] [4100/5004] eta: 0:10:45 lr: 0.000001 loss: 1.604196 (1.613942) time: 0.712589 data: 0.000216 max mem: 14338 Epoch: [28/30] [4150/5004] eta: 0:10:10 lr: 0.000001 loss: 1.518838 (1.614008) time: 0.712304 data: 0.000187 max mem: 14338 Epoch: [28/30] [4200/5004] eta: 0:09:34 lr: 0.000001 loss: 1.559255 (1.614466) time: 0.712384 data: 0.000214 max mem: 14338 Epoch: [28/30] [4250/5004] eta: 0:08:58 lr: 0.000001 loss: 1.695498 (1.614270) time: 0.711311 data: 0.000204 max mem: 14338 Epoch: [28/30] [4300/5004] eta: 0:08:23 lr: 0.000001 loss: 1.370938 (1.613674) time: 0.715534 data: 0.000219 max mem: 14338 Epoch: [28/30] [4350/5004] eta: 0:07:47 lr: 0.000001 loss: 1.426717 (1.614258) time: 0.717180 data: 0.000176 max mem: 14338 Epoch: [28/30] [4400/5004] eta: 0:07:11 lr: 0.000001 loss: 1.539711 (1.613401) time: 0.716098 data: 0.000180 max mem: 14338 Epoch: [28/30] [4450/5004] eta: 0:06:35 lr: 0.000001 loss: 1.641856 (1.613977) time: 0.720751 data: 0.000217 max mem: 14338 Epoch: [28/30] [4500/5004] eta: 0:06:00 lr: 0.000001 loss: 1.529660 (1.613158) time: 0.711200 data: 0.000221 max mem: 14338 Epoch: [28/30] [4550/5004] eta: 0:05:24 lr: 0.000001 loss: 1.596951 (1.613282) time: 0.711367 data: 0.000220 max mem: 14338 Epoch: [28/30] [4600/5004] eta: 0:04:48 lr: 0.000001 loss: 1.575180 (1.613132) time: 0.712751 data: 0.000235 max mem: 14338 Epoch: [28/30] [4650/5004] eta: 0:04:12 lr: 0.000001 loss: 1.612248 (1.613125) time: 0.711718 data: 0.000239 max mem: 14338 Epoch: [28/30] [4700/5004] eta: 0:03:37 lr: 0.000001 loss: 1.556310 (1.613187) time: 0.714195 data: 0.000169 max mem: 14338 Epoch: [28/30] [4750/5004] eta: 0:03:01 lr: 0.000001 loss: 1.559078 (1.613819) time: 0.710771 data: 0.000169 max mem: 14338 Epoch: [28/30] [4800/5004] eta: 0:02:25 lr: 0.000001 loss: 1.515295 (1.613458) time: 0.719328 data: 0.000205 max mem: 14338 Epoch: [28/30] [4850/5004] eta: 0:01:50 lr: 0.000001 loss: 1.576199 (1.613494) time: 0.712996 data: 0.000235 max mem: 14338 Epoch: [28/30] [4900/5004] eta: 0:01:14 lr: 0.000001 loss: 1.530847 (1.613869) time: 0.714166 data: 0.000227 max mem: 14338 Epoch: [28/30] [4950/5004] eta: 0:00:38 lr: 0.000001 loss: 1.605028 (1.613864) time: 0.714361 data: 0.000226 max mem: 14338 Epoch: [28/30] [5000/5004] eta: 0:00:02 lr: 0.000001 loss: 1.501729 (1.614159) time: 0.710248 data: 0.000833 max mem: 14338 Epoch: [28/30] [5003/5004] eta: 0:00:00 lr: 0.000001 loss: 1.521672 (1.614105) time: 0.706894 data: 0.000824 max mem: 14338 Epoch: [28/30] Total time: 0:59:36 (0.714652 s / it) Averaged stats: lr: 0.000001 loss: 1.521672 (1.618628) Test: [ 0/196] eta: 0:05:17 loss: 0.277108 (0.277108) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.620258 data: 1.151356 max mem: 14338 Test: [ 10/196] eta: 0:01:16 loss: 0.450932 (0.538479) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.412959 data: 0.104797 max mem: 14338 Test: [ 20/196] eta: 0:01:02 loss: 0.546607 (0.537086) acc1: 87.500000 (85.119048) acc5: 100.000000 (98.214286) time: 0.289990 data: 0.000125 max mem: 14338 Test: [ 30/196] eta: 0:00:55 loss: 0.523958 (0.513910) acc1: 87.500000 (86.693548) acc5: 100.000000 (98.387097) time: 0.287120 data: 0.000113 max mem: 14338 Test: [ 40/196] eta: 0:00:50 loss: 0.435311 (0.521779) acc1: 87.500000 (86.432927) acc5: 100.000000 (98.170732) time: 0.286514 data: 0.000131 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.468033 (0.550166) acc1: 87.500000 (86.519608) acc5: 100.000000 (97.671569) time: 0.287222 data: 0.000141 max mem: 14338 Test: [ 60/196] eta: 0:00:42 loss: 0.616065 (0.578025) acc1: 87.500000 (85.963115) acc5: 93.750000 (97.540984) time: 0.287129 data: 0.000141 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.660769 (0.595075) acc1: 81.250000 (85.651408) acc5: 100.000000 (97.711268) time: 0.286889 data: 0.000142 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.500288 (0.594270) acc1: 87.500000 (85.648148) acc5: 100.000000 (97.839506) time: 0.286785 data: 0.000128 max mem: 14338 Test: [ 90/196] eta: 0:00:32 loss: 0.510654 (0.617387) acc1: 87.500000 (85.233516) acc5: 100.000000 (97.596154) time: 0.285979 data: 0.000128 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.577424 (0.606184) acc1: 81.250000 (85.396040) acc5: 100.000000 (97.772277) time: 0.286180 data: 0.000128 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.549971 (0.595270) acc1: 87.500000 (85.472973) acc5: 100.000000 (97.860360) time: 0.287045 data: 0.000128 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.533677 (0.591209) acc1: 87.500000 (85.588843) acc5: 100.000000 (97.830579) time: 0.286833 data: 0.000141 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.615413 (0.606595) acc1: 87.500000 (85.257634) acc5: 100.000000 (97.805344) time: 0.286502 data: 0.000131 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.635067 (0.605714) acc1: 81.250000 (85.372340) acc5: 100.000000 (97.783688) time: 0.291798 data: 0.000138 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.544625 (0.614655) acc1: 87.500000 (85.264901) acc5: 100.000000 (97.806291) time: 0.291631 data: 0.000158 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.525650 (0.619760) acc1: 81.250000 (85.209627) acc5: 100.000000 (97.787267) time: 0.286293 data: 0.000138 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.504965 (0.616664) acc1: 81.250000 (85.307018) acc5: 100.000000 (97.807018) time: 0.286063 data: 0.000134 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.465029 (0.616555) acc1: 87.500000 (85.186464) acc5: 100.000000 (97.790055) time: 0.286029 data: 0.000152 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.364075 (0.604802) acc1: 87.500000 (85.405759) acc5: 100.000000 (97.807592) time: 0.283899 data: 0.000112 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.494855 (0.615013) acc1: 87.500000 (85.312000) acc5: 100.000000 (97.728000) time: 0.274520 data: 0.000098 max mem: 14338 Test: Total time: 0:00:57 (0.294259 s / it) * Acc@1 85.128 Acc@5 97.300 loss 0.632 Max accuracy: 85.25% Epoch: [29/30] [ 0/5004] eta: 2:42:33 lr: 0.000001 loss: 1.183298 (1.183298) time: 1.949130 data: 1.203243 max mem: 14338 Epoch: [29/30] [ 50/5004] eta: 1:01:07 lr: 0.000001 loss: 1.715035 (1.668943) time: 0.714800 data: 0.000174 max mem: 14338 Epoch: [29/30] [ 100/5004] eta: 0:59:25 lr: 0.000001 loss: 1.514611 (1.647438) time: 0.709661 data: 0.000203 max mem: 14338 Epoch: [29/30] [ 150/5004] eta: 0:58:28 lr: 0.000001 loss: 1.697823 (1.662880) time: 0.709608 data: 0.000223 max mem: 14338 Epoch: [29/30] [ 200/5004] eta: 0:57:46 lr: 0.000001 loss: 1.571378 (1.647056) time: 0.716988 data: 0.000244 max mem: 14338 Epoch: [29/30] [ 250/5004] eta: 0:57:04 lr: 0.000001 loss: 1.539744 (1.654622) time: 0.715838 data: 0.000191 max mem: 14338 Epoch: [29/30] [ 300/5004] eta: 0:56:23 lr: 0.000001 loss: 1.683740 (1.659878) time: 0.712762 data: 0.000176 max mem: 14338 Epoch: [29/30] [ 350/5004] eta: 0:55:42 lr: 0.000001 loss: 1.663350 (1.656335) time: 0.711937 data: 0.000159 max mem: 14338 Epoch: [29/30] [ 400/5004] eta: 0:55:04 lr: 0.000001 loss: 1.460760 (1.652698) time: 0.719772 data: 0.000220 max mem: 14338 Epoch: [29/30] [ 450/5004] eta: 0:54:26 lr: 0.000001 loss: 1.615770 (1.654472) time: 0.713389 data: 0.000210 max mem: 14338 Epoch: [29/30] [ 500/5004] eta: 0:53:48 lr: 0.000001 loss: 1.556334 (1.651878) time: 0.712947 data: 0.000210 max mem: 14338 Epoch: [29/30] [ 550/5004] eta: 0:53:12 lr: 0.000001 loss: 1.600391 (1.650591) time: 0.713852 data: 0.000200 max mem: 14338 Epoch: [29/30] [ 600/5004] eta: 0:52:37 lr: 0.000001 loss: 1.608095 (1.645280) time: 0.712467 data: 0.000225 max mem: 14338 Epoch: [29/30] [ 650/5004] eta: 0:51:59 lr: 0.000001 loss: 1.618143 (1.643123) time: 0.712310 data: 0.000174 max mem: 14338 Epoch: [29/30] [ 700/5004] eta: 0:51:23 lr: 0.000001 loss: 1.616497 (1.641239) time: 0.712866 data: 0.000157 max mem: 14338 Epoch: [29/30] [ 750/5004] eta: 0:50:47 lr: 0.000001 loss: 1.445677 (1.635404) time: 0.713071 data: 0.000215 max mem: 14338 Epoch: [29/30] [ 800/5004] eta: 0:50:10 lr: 0.000001 loss: 1.469091 (1.631125) time: 0.718402 data: 0.000214 max mem: 14338 Epoch: [29/30] [ 850/5004] eta: 0:49:34 lr: 0.000001 loss: 1.692067 (1.633289) time: 0.721416 data: 0.000223 max mem: 14338 Epoch: [29/30] [ 900/5004] eta: 0:48:57 lr: 0.000001 loss: 1.629495 (1.630994) time: 0.716181 data: 0.000183 max mem: 14338 Epoch: [29/30] [ 950/5004] eta: 0:48:21 lr: 0.000001 loss: 1.698674 (1.634330) time: 0.714026 data: 0.000238 max mem: 14338 Epoch: [29/30] [1000/5004] eta: 0:47:46 lr: 0.000001 loss: 1.552731 (1.633345) time: 0.716168 data: 0.000168 max mem: 14338 Epoch: [29/30] [1050/5004] eta: 0:47:10 lr: 0.000001 loss: 1.514468 (1.631605) time: 0.716996 data: 0.000205 max mem: 14338 Epoch: [29/30] [1100/5004] eta: 0:46:34 lr: 0.000001 loss: 1.488642 (1.627862) time: 0.713830 data: 0.000208 max mem: 14338 Epoch: [29/30] [1150/5004] eta: 0:45:58 lr: 0.000001 loss: 1.454966 (1.625097) time: 0.711295 data: 0.000205 max mem: 14338 Epoch: [29/30] [1200/5004] eta: 0:45:22 lr: 0.000001 loss: 1.526196 (1.622028) time: 0.714964 data: 0.000230 max mem: 14338 Epoch: [29/30] [1250/5004] eta: 0:44:46 lr: 0.000001 loss: 1.599595 (1.622557) time: 0.720019 data: 0.000218 max mem: 14338 Epoch: [29/30] [1300/5004] eta: 0:44:09 lr: 0.000001 loss: 1.564749 (1.623453) time: 0.712202 data: 0.000181 max mem: 14338 Epoch: [29/30] [1350/5004] eta: 0:43:33 lr: 0.000001 loss: 1.526507 (1.624684) time: 0.714267 data: 0.000169 max mem: 14338 Epoch: [29/30] [1400/5004] eta: 0:42:57 lr: 0.000001 loss: 1.738271 (1.625346) time: 0.715878 data: 0.000218 max mem: 14338 Epoch: [29/30] [1450/5004] eta: 0:42:21 lr: 0.000001 loss: 1.647951 (1.627083) time: 0.711323 data: 0.000227 max mem: 14338 Epoch: [29/30] [1500/5004] eta: 0:41:45 lr: 0.000001 loss: 1.485150 (1.626823) time: 0.708454 data: 0.000218 max mem: 14338 Epoch: [29/30] [1550/5004] eta: 0:41:09 lr: 0.000001 loss: 1.619632 (1.628320) time: 0.708604 data: 0.000190 max mem: 14338 Epoch: [29/30] [1600/5004] eta: 0:40:33 lr: 0.000001 loss: 1.657457 (1.627656) time: 0.712353 data: 0.000212 max mem: 14338 Epoch: [29/30] [1650/5004] eta: 0:39:57 lr: 0.000001 loss: 1.627874 (1.627657) time: 0.712802 data: 0.000172 max mem: 14338 Epoch: [29/30] [1700/5004] eta: 0:39:21 lr: 0.000001 loss: 1.667137 (1.627998) time: 0.713030 data: 0.000171 max mem: 14338 Epoch: [29/30] [1750/5004] eta: 0:38:45 lr: 0.000001 loss: 1.581225 (1.628393) time: 0.712239 data: 0.000215 max mem: 14338 Epoch: [29/30] [1800/5004] eta: 0:38:10 lr: 0.000001 loss: 1.583932 (1.627189) time: 0.714813 data: 0.000231 max mem: 14338 Epoch: [29/30] [1850/5004] eta: 0:37:34 lr: 0.000001 loss: 1.490381 (1.626626) time: 0.715994 data: 0.000216 max mem: 14338 Epoch: [29/30] [1900/5004] eta: 0:36:58 lr: 0.000001 loss: 1.708403 (1.627018) time: 0.711911 data: 0.000221 max mem: 14338 Epoch: [29/30] [1950/5004] eta: 0:36:22 lr: 0.000001 loss: 1.540250 (1.626242) time: 0.711744 data: 0.000230 max mem: 14338 Epoch: [29/30] [2000/5004] eta: 0:35:47 lr: 0.000001 loss: 1.516943 (1.626608) time: 0.712939 data: 0.000172 max mem: 14338 Epoch: [29/30] [2050/5004] eta: 0:35:11 lr: 0.000001 loss: 1.636388 (1.626967) time: 0.720122 data: 0.000168 max mem: 14338 Epoch: [29/30] [2100/5004] eta: 0:34:35 lr: 0.000001 loss: 1.519403 (1.627657) time: 0.713631 data: 0.000214 max mem: 14338 Epoch: [29/30] [2150/5004] eta: 0:33:59 lr: 0.000001 loss: 1.607357 (1.627611) time: 0.709042 data: 0.000213 max mem: 14338 Epoch: [29/30] [2200/5004] eta: 0:33:24 lr: 0.000001 loss: 1.486750 (1.625447) time: 0.719433 data: 0.000187 max mem: 14338 Epoch: [29/30] [2250/5004] eta: 0:32:48 lr: 0.000001 loss: 1.526716 (1.625055) time: 0.712471 data: 0.000231 max mem: 14338 Epoch: [29/30] [2300/5004] eta: 0:32:12 lr: 0.000001 loss: 1.562554 (1.624981) time: 0.712619 data: 0.000232 max mem: 14338 Epoch: [29/30] [2350/5004] eta: 0:31:36 lr: 0.000001 loss: 1.623207 (1.623911) time: 0.713027 data: 0.000161 max mem: 14338 Epoch: [29/30] [2400/5004] eta: 0:31:00 lr: 0.000001 loss: 1.380879 (1.621914) time: 0.715841 data: 0.000215 max mem: 14338 Epoch: [29/30] [2450/5004] eta: 0:30:25 lr: 0.000001 loss: 1.715394 (1.622674) time: 0.715879 data: 0.000214 max mem: 14338 Epoch: [29/30] [2500/5004] eta: 0:29:49 lr: 0.000001 loss: 1.708675 (1.623888) time: 0.715157 data: 0.000227 max mem: 14338 Epoch: [29/30] [2550/5004] eta: 0:29:13 lr: 0.000001 loss: 1.542135 (1.624494) time: 0.711566 data: 0.000224 max mem: 14338 Epoch: [29/30] [2600/5004] eta: 0:28:37 lr: 0.000001 loss: 1.502298 (1.624275) time: 0.711319 data: 0.000226 max mem: 14338 Epoch: [29/30] [2650/5004] eta: 0:28:02 lr: 0.000001 loss: 1.634861 (1.625240) time: 0.716789 data: 0.000172 max mem: 14338 Epoch: [29/30] [2700/5004] eta: 0:27:26 lr: 0.000001 loss: 1.459554 (1.623746) time: 0.708344 data: 0.000165 max mem: 14338 Epoch: [29/30] [2750/5004] eta: 0:26:50 lr: 0.000001 loss: 1.732358 (1.625453) time: 0.715318 data: 0.000232 max mem: 14338 Epoch: [29/30] [2800/5004] eta: 0:26:14 lr: 0.000001 loss: 1.528918 (1.624138) time: 0.711676 data: 0.000220 max mem: 14338 Epoch: [29/30] [2850/5004] eta: 0:25:38 lr: 0.000001 loss: 1.452119 (1.623490) time: 0.713769 data: 0.000218 max mem: 14338 Epoch: [29/30] [2900/5004] eta: 0:25:03 lr: 0.000001 loss: 1.487191 (1.622877) time: 0.713187 data: 0.000214 max mem: 14338 Epoch: [29/30] [2950/5004] eta: 0:24:27 lr: 0.000001 loss: 1.540376 (1.621973) time: 0.711089 data: 0.000211 max mem: 14338 Epoch: [29/30] [3000/5004] eta: 0:23:51 lr: 0.000001 loss: 1.564168 (1.621578) time: 0.714109 data: 0.000175 max mem: 14338 Epoch: [29/30] [3050/5004] eta: 0:23:16 lr: 0.000001 loss: 1.625545 (1.621291) time: 0.711230 data: 0.000176 max mem: 14338 Epoch: [29/30] [3100/5004] eta: 0:22:40 lr: 0.000001 loss: 1.534221 (1.621802) time: 0.708466 data: 0.000223 max mem: 14338 Epoch: [29/30] [3150/5004] eta: 0:22:04 lr: 0.000001 loss: 1.492316 (1.621602) time: 0.710640 data: 0.000216 max mem: 14338 Epoch: [29/30] [3200/5004] eta: 0:21:28 lr: 0.000001 loss: 1.562758 (1.620853) time: 0.714215 data: 0.000226 max mem: 14338 Epoch: [29/30] [3250/5004] eta: 0:20:52 lr: 0.000001 loss: 1.601949 (1.621677) time: 0.711527 data: 0.000200 max mem: 14338 Epoch: [29/30] [3300/5004] eta: 0:20:17 lr: 0.000001 loss: 1.576535 (1.621978) time: 0.708303 data: 0.000213 max mem: 14338 Epoch: [29/30] [3350/5004] eta: 0:19:41 lr: 0.000001 loss: 1.698223 (1.622108) time: 0.709900 data: 0.000174 max mem: 14338 Epoch: [29/30] [3400/5004] eta: 0:19:05 lr: 0.000001 loss: 1.525722 (1.622373) time: 0.714462 data: 0.000167 max mem: 14338 Epoch: [29/30] [3450/5004] eta: 0:18:29 lr: 0.000001 loss: 1.441473 (1.622122) time: 0.715254 data: 0.000202 max mem: 14338 Epoch: [29/30] [3500/5004] eta: 0:17:54 lr: 0.000001 loss: 1.633543 (1.621786) time: 0.708868 data: 0.000204 max mem: 14338 Epoch: [29/30] [3550/5004] eta: 0:17:18 lr: 0.000001 loss: 1.517968 (1.621846) time: 0.716444 data: 0.000238 max mem: 14338 Epoch: [29/30] [3600/5004] eta: 0:16:42 lr: 0.000001 loss: 1.555205 (1.621910) time: 0.720254 data: 0.000232 max mem: 14338 Epoch: [29/30] [3650/5004] eta: 0:16:07 lr: 0.000001 loss: 1.746854 (1.622620) time: 0.719171 data: 0.000218 max mem: 14338 Epoch: [29/30] [3700/5004] eta: 0:15:31 lr: 0.000001 loss: 1.588486 (1.622518) time: 0.713760 data: 0.000163 max mem: 14338 Epoch: [29/30] [3750/5004] eta: 0:14:55 lr: 0.000001 loss: 1.467651 (1.622455) time: 0.711421 data: 0.000206 max mem: 14338 Epoch: [29/30] [3800/5004] eta: 0:14:20 lr: 0.000001 loss: 1.625897 (1.621830) time: 0.712464 data: 0.000211 max mem: 14338 Epoch: [29/30] [3850/5004] eta: 0:13:44 lr: 0.000001 loss: 1.540770 (1.621458) time: 0.714830 data: 0.000209 max mem: 14338 Epoch: [29/30] [3900/5004] eta: 0:13:08 lr: 0.000001 loss: 1.591989 (1.621998) time: 0.712236 data: 0.000212 max mem: 14338 Epoch: [29/30] [3950/5004] eta: 0:12:32 lr: 0.000001 loss: 1.518505 (1.622209) time: 0.716545 data: 0.000223 max mem: 14338 Epoch: [29/30] [4000/5004] eta: 0:11:57 lr: 0.000001 loss: 1.541827 (1.622054) time: 0.711623 data: 0.000161 max mem: 14338 Epoch: [29/30] [4050/5004] eta: 0:11:21 lr: 0.000001 loss: 1.419040 (1.621968) time: 0.714963 data: 0.000177 max mem: 14338 Epoch: [29/30] [4100/5004] eta: 0:10:45 lr: 0.000001 loss: 1.595120 (1.623167) time: 0.716641 data: 0.000218 max mem: 14338 Epoch: [29/30] [4150/5004] eta: 0:10:10 lr: 0.000001 loss: 1.484529 (1.622245) time: 0.713855 data: 0.000224 max mem: 14338 Epoch: [29/30] [4200/5004] eta: 0:09:34 lr: 0.000001 loss: 1.506624 (1.622178) time: 0.717253 data: 0.000217 max mem: 14338 Epoch: [29/30] [4250/5004] eta: 0:08:58 lr: 0.000001 loss: 1.469930 (1.622767) time: 0.717465 data: 0.000219 max mem: 14338 Epoch: [29/30] [4300/5004] eta: 0:08:22 lr: 0.000001 loss: 1.637033 (1.623117) time: 0.709310 data: 0.000206 max mem: 14338 Epoch: [29/30] [4350/5004] eta: 0:07:47 lr: 0.000001 loss: 1.555038 (1.622915) time: 0.710462 data: 0.000176 max mem: 14338 Epoch: [29/30] [4400/5004] eta: 0:07:11 lr: 0.000001 loss: 1.486203 (1.622627) time: 0.712063 data: 0.000176 max mem: 14338 Epoch: [29/30] [4450/5004] eta: 0:06:35 lr: 0.000001 loss: 1.648222 (1.622073) time: 0.713801 data: 0.000220 max mem: 14338 Epoch: [29/30] [4500/5004] eta: 0:05:59 lr: 0.000001 loss: 1.519882 (1.622435) time: 0.715460 data: 0.000231 max mem: 14338 Epoch: [29/30] [4550/5004] eta: 0:05:24 lr: 0.000001 loss: 1.750841 (1.623975) time: 0.713561 data: 0.000220 max mem: 14338 Epoch: [29/30] [4600/5004] eta: 0:04:48 lr: 0.000001 loss: 1.600878 (1.624562) time: 0.718610 data: 0.000220 max mem: 14338 Epoch: [29/30] [4650/5004] eta: 0:04:12 lr: 0.000001 loss: 1.609109 (1.624812) time: 0.713894 data: 0.000230 max mem: 14338 Epoch: [29/30] [4700/5004] eta: 0:03:37 lr: 0.000001 loss: 1.657404 (1.624807) time: 0.712004 data: 0.000170 max mem: 14338 Epoch: [29/30] [4750/5004] eta: 0:03:01 lr: 0.000001 loss: 1.557512 (1.624695) time: 0.713472 data: 0.000170 max mem: 14338 Epoch: [29/30] [4800/5004] eta: 0:02:25 lr: 0.000001 loss: 1.639773 (1.624671) time: 0.714574 data: 0.000179 max mem: 14338 Epoch: [29/30] [4850/5004] eta: 0:01:50 lr: 0.000001 loss: 1.554693 (1.624784) time: 0.723868 data: 0.000229 max mem: 14338 Epoch: [29/30] [4900/5004] eta: 0:01:14 lr: 0.000001 loss: 1.529730 (1.625102) time: 0.710127 data: 0.000228 max mem: 14338 Epoch: [29/30] [4950/5004] eta: 0:00:38 lr: 0.000001 loss: 1.636863 (1.625556) time: 0.712401 data: 0.000218 max mem: 14338 Epoch: [29/30] [5000/5004] eta: 0:00:02 lr: 0.000001 loss: 1.661436 (1.625642) time: 0.719586 data: 0.000829 max mem: 14338 Epoch: [29/30] [5003/5004] eta: 0:00:00 lr: 0.000001 loss: 1.621220 (1.625746) time: 0.714778 data: 0.000823 max mem: 14338 Epoch: [29/30] Total time: 0:59:35 (0.714477 s / it) Averaged stats: lr: 0.000001 loss: 1.621220 (1.621039) Test: [ 0/196] eta: 0:05:02 loss: 0.278950 (0.278950) acc1: 93.750000 (93.750000) acc5: 100.000000 (100.000000) time: 1.541809 data: 1.124965 max mem: 14338 Test: [ 10/196] eta: 0:01:14 loss: 0.448892 (0.540639) acc1: 87.500000 (84.659091) acc5: 100.000000 (98.863636) time: 0.400261 data: 0.102382 max mem: 14338 Test: [ 20/196] eta: 0:01:00 loss: 0.544837 (0.538789) acc1: 87.500000 (85.714286) acc5: 100.000000 (98.214286) time: 0.286163 data: 0.000115 max mem: 14338 Test: [ 30/196] eta: 0:00:54 loss: 0.524008 (0.514219) acc1: 87.500000 (87.096774) acc5: 100.000000 (98.387097) time: 0.285984 data: 0.000113 max mem: 14338 Test: [ 40/196] eta: 0:00:49 loss: 0.436463 (0.521592) acc1: 87.500000 (86.737805) acc5: 100.000000 (98.170732) time: 0.286522 data: 0.000124 max mem: 14338 Test: [ 50/196] eta: 0:00:45 loss: 0.465740 (0.550015) acc1: 87.500000 (86.764706) acc5: 100.000000 (97.671569) time: 0.286980 data: 0.000121 max mem: 14338 Test: [ 60/196] eta: 0:00:41 loss: 0.609313 (0.577821) acc1: 87.500000 (86.168033) acc5: 93.750000 (97.540984) time: 0.286654 data: 0.000129 max mem: 14338 Test: [ 70/196] eta: 0:00:38 loss: 0.660213 (0.594790) acc1: 81.250000 (85.739437) acc5: 100.000000 (97.711268) time: 0.286850 data: 0.000138 max mem: 14338 Test: [ 80/196] eta: 0:00:35 loss: 0.498135 (0.594162) acc1: 87.500000 (85.725309) acc5: 100.000000 (97.839506) time: 0.290970 data: 0.000137 max mem: 14338 Test: [ 90/196] eta: 0:00:31 loss: 0.514025 (0.617325) acc1: 87.500000 (85.302198) acc5: 100.000000 (97.596154) time: 0.293562 data: 0.000143 max mem: 14338 Test: [100/196] eta: 0:00:28 loss: 0.581197 (0.606141) acc1: 81.250000 (85.457921) acc5: 100.000000 (97.772277) time: 0.289799 data: 0.000133 max mem: 14338 Test: [110/196] eta: 0:00:25 loss: 0.544615 (0.594987) acc1: 87.500000 (85.529279) acc5: 100.000000 (97.860360) time: 0.286822 data: 0.000124 max mem: 14338 Test: [120/196] eta: 0:00:22 loss: 0.534492 (0.591063) acc1: 87.500000 (85.640496) acc5: 100.000000 (97.830579) time: 0.286283 data: 0.000129 max mem: 14338 Test: [130/196] eta: 0:00:19 loss: 0.618479 (0.606393) acc1: 87.500000 (85.353053) acc5: 100.000000 (97.805344) time: 0.286249 data: 0.000132 max mem: 14338 Test: [140/196] eta: 0:00:16 loss: 0.623802 (0.605583) acc1: 87.500000 (85.416667) acc5: 100.000000 (97.783688) time: 0.286597 data: 0.000134 max mem: 14338 Test: [150/196] eta: 0:00:13 loss: 0.547834 (0.614369) acc1: 87.500000 (85.306291) acc5: 100.000000 (97.806291) time: 0.286874 data: 0.000126 max mem: 14338 Test: [160/196] eta: 0:00:10 loss: 0.521175 (0.619488) acc1: 81.250000 (85.248447) acc5: 100.000000 (97.787267) time: 0.286393 data: 0.000122 max mem: 14338 Test: [170/196] eta: 0:00:07 loss: 0.500202 (0.616477) acc1: 81.250000 (85.343567) acc5: 100.000000 (97.807018) time: 0.285747 data: 0.000129 max mem: 14338 Test: [180/196] eta: 0:00:04 loss: 0.466488 (0.616340) acc1: 87.500000 (85.220994) acc5: 100.000000 (97.790055) time: 0.285622 data: 0.000141 max mem: 14338 Test: [190/196] eta: 0:00:01 loss: 0.364544 (0.604606) acc1: 87.500000 (85.438482) acc5: 100.000000 (97.807592) time: 0.283576 data: 0.000111 max mem: 14338 Test: [195/196] eta: 0:00:00 loss: 0.494119 (0.614849) acc1: 87.500000 (85.344000) acc5: 100.000000 (97.728000) time: 0.273570 data: 0.000099 max mem: 14338 Test: Total time: 0:00:57 (0.293388 s / it) * Acc@1 85.144 Acc@5 97.304 loss 0.632 Max accuracy: 85.25% Training time 1 day, 6:29:52