genius-7 / ResNet18_CIFAR-10

1 stars 0 forks source link

verse19数据集图像预处理 #2

Open Fucapisun opened 4 months ago

Fucapisun commented 4 months ago

你好,我在PnPNet看到你对作者提出了一些疑问找过来的,想请问一下作者在他的代码库中省略的利用nnUNet的z-score进行归一化,我试着进行了归一化后训练效果有点差,我想知道您是怎么对verse19数据集进行处理的,谢谢

genius-7 commented 4 months ago

我的结果也非常差,和原作者的结果查的很远 ---- 回复的原邮件 ---- @.>发送日期2024年06月21日 13:42 @.> @.>主题[genius-7/ResNet18_CIFAR-10] verse19数据集图像预处理 (Issue #2) 你好,我在PnPNet看到你对作者提出了一些疑问找过来的,想请问一下作者在他的代码库中省略的利用nnUNet的z-score进行归一化,我试着进行了归一化后训练效果有点差,我想知道您是怎么对verse19数据集进行处理的,谢谢 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: @.>

Fucapisun commented 4 months ago

[17:49:32.301] epoch : 215, iteration : 4304, train loss : 0.504301, train loss_ce: 0.359726, train loss_dice: 0.648875, train dice : 0.351125 [17:50:27.135] epoch : 215, iteration : 4312, train loss : 0.497282, train loss_ce: 0.442192, train loss_dice: 0.552371, train dice : 0.447629 [17:51:10.392] epoch : 215, iteration : 4320, train loss : 0.491708, train loss_ce: 0.339461, train loss_dice: 0.643955, train dice : 0.356045 [17:51:10.413] epoch : 215, mean train loss : 0.482834, mean train ce loss: 0.376177, mean train dice : 0.410510 [17:51:11.667] epoch : 215, iteration : 2151, val loss : 0.494327, val loss_ce: 0.404260, val loss_dice: 0.584394, val dice : 0.415606 [17:51:21.039] epoch : 215, iteration : 2154, val loss : 0.462684, val loss_ce: 0.361490, val loss_dice: 0.563878, val dice : 0.436122 [17:51:27.035] epoch : 215, iteration : 2157, val loss : 0.514915, val loss_ce: 0.437994, val loss_dice: 0.591836, val dice : 0.408164 [17:51:34.151] epoch : 215, iteration : 2160, val loss : 0.485785, val loss_ce: 0.402263, val loss_dice: 0.569307, val dice : 0.430693 [17:51:34.152] epoch : 215, mean val loss : 0.575246, mean val ce loss: 0.541292, mean val dice : 0.390800 [17:52:20.930] epoch : 216, iteration : 4328, train loss : 0.515502, train loss_ce: 0.431891, train loss_dice: 0.599113, train dice : 0.400887 [17:53:08.453] epoch : 216, iteration : 4336, train loss : 0.525782, train loss_ce: 0.431233, train loss_dice: 0.620331, train dice : 0.379669 [17:53:32.590] epoch : 216, mean train loss : 0.487081, mean train ce loss: 0.382485, mean train dice : 0.408323 [17:53:42.116] epoch : 216, iteration : 2163, val loss : 0.540408, val loss_ce: 0.447393, val loss_dice: 0.633423, val dice : 0.366577 [17:53:48.341] epoch : 216, iteration : 2166, val loss : 0.502553, val loss_ce: 0.375919, val loss_dice: 0.629186, val dice : 0.370814 [17:53:54.064] epoch : 216, iteration : 2169, val loss : 0.728848, val loss_ce: 0.804907, val loss_dice: 0.652789, val dice : 0.347211 [17:53:56.033] epoch : 216, mean val loss : 0.523348, mean val ce loss: 0.452031, mean val dice : 0.405334 [17:54:22.411] epoch : 217, iteration : 4344, train loss : 0.486062, train loss_ce: 0.344975, train loss_dice: 0.627148, train dice : 0.372852 [17:55:08.915] epoch : 217, iteration : 4352, train loss : 0.488311, train loss_ce: 0.376164, train loss_dice: 0.600459, train dice : 0.399541 [17:55:53.970] epoch : 217, iteration : 4360, train loss : 0.452595, train loss_ce: 0.357177, train loss_dice: 0.548013, train dice : 0.451987 [17:55:53.990] epoch : 217, mean train loss : 0.485286, mean train ce loss: 0.383075, mean train dice : 0.412502 [17:55:59.375] epoch : 217, iteration : 2172, val loss : 0.443410, val loss_ce: 0.312275, val loss_dice: 0.574545, val dice : 0.425455 [17:56:06.960] epoch : 217, iteration : 2175, val loss : 0.439832, val loss_ce: 0.334240, val loss_dice: 0.545425, val dice : 0.454575 [17:56:13.255] epoch : 217, iteration : 2178, val loss : 0.622069, val loss_ce: 0.632170, val loss_dice: 0.611968, val dice : 0.388032 [17:56:17.508] epoch : 217, mean val loss : 0.525838, mean val ce loss: 0.454492, mean val dice : 0.402815 [17:57:02.156] epoch : 218, iteration : 4368, train loss : 0.473139, train loss_ce: 0.358906, train loss_dice: 0.587372, train dice : 0.412628 [17:57:47.087] epoch : 218, iteration : 4376, train loss : 0.450360, train loss_ce: 0.318221, train loss_dice: 0.582498, train dice : 0.417502 [17:58:11.689] epoch : 218, mean train loss : 0.485688, mean train ce loss: 0.376089, mean train dice : 0.404714 [17:58:15.449] epoch : 218, iteration : 2181, val loss : 0.605697, val loss_ce: 0.574678, val loss_dice: 0.636716, val dice : 0.363284 [17:58:20.657] epoch : 218, iteration : 2184, val loss : 0.494207, val loss_ce: 0.424579, val loss_dice: 0.563834, val dice : 0.436166 [17:58:29.918] epoch : 218, iteration : 2187, val loss : 0.502701, val loss_ce: 0.428572, val loss_dice: 0.576829, val dice : 0.423171 [17:58:38.307] epoch : 218, iteration : 2190, val loss : 0.477035, val loss_ce: 0.347020, val loss_dice: 0.607051, val dice : 0.392949 [17:58:38.308] epoch : 218, mean val loss : 0.542393, mean val ce loss: 0.488022, mean val dice : 0.403235 [17:58:53.880] epoch : 219, iteration : 4384, train loss : 0.422732, train loss_ce: 0.295451, train loss_dice: 0.550013, train dice : 0.449987 [17:59:50.115] epoch : 219, iteration : 4392, train loss : 0.569222, train loss_ce: 0.505555, train loss_dice: 0.632888, train dice : 0.367112 [18:00:34.272] epoch : 219, iteration : 4400, train loss : 0.455917, train loss_ce: 0.349803, train loss_dice: 0.562031, train dice : 0.437969 [18:00:34.294] epoch : 219, mean train loss : 0.483951, mean train ce loss: 0.384263, mean train dice : 0.416362 [18:00:43.950] epoch : 219, iteration : 2193, val loss : 0.577745, val loss_ce: 0.575619, val loss_dice: 0.579872, val dice : 0.420128 [18:00:49.764] epoch : 219, iteration : 2196, val loss : 0.643961, val loss_ce: 0.671307, val loss_dice: 0.616615, val dice : 0.383385 [18:00:55.931] epoch : 219, iteration : 2199, val loss : 0.510778, val loss_ce: 0.426397, val loss_dice: 0.595159, val dice : 0.404841 [18:00:57.038] epoch : 219, mean val loss : 0.536568, mean val ce loss: 0.477079, mean val dice : 0.403942 [18:01:32.002] epoch : 220, iteration : 4408, train loss : 0.459404, train loss_ce: 0.322050, train loss_dice: 0.596759, train dice : 0.403241 [18:02:21.242] epoch : 220, iteration : 4416, train loss : 0.422637, train loss_ce: 0.275079, train loss_dice: 0.570196, train dice : 0.429804 [18:02:52.965] epoch : 220, mean train loss : 0.478490, mean train ce loss: 0.371975, mean train dice : 0.414995 [18:02:57.400] epoch : 220, iteration : 2202, val loss : 0.555597, val loss_ce: 0.495404, val loss_dice: 0.615790, val dice : 0.384210 [18:03:08.629] epoch : 220, iteration : 2205, val loss : 0.480225, val loss_ce: 0.353570, val loss_dice: 0.606880, val dice : 0.393120 [18:03:13.957] epoch : 220, iteration : 2208, val loss : 0.433605, val loss_ce: 0.269190, val loss_dice: 0.598019, val dice : 0.401981 [18:03:17.644] epoch : 220, mean val loss : 0.523180, mean val ce loss: 0.444338, mean val dice : 0.397977 [18:03:35.352] epoch : 221, iteration : 4424, train loss : 0.546785, train loss_ce: 0.439320, train loss_dice: 0.654251, train dice : 0.345749 [18:04:26.023] epoch : 221, iteration : 4432, train loss : 0.557667, train loss_ce: 0.517154, train loss_dice: 0.598179, train dice : 0.401821 [18:05:13.353] epoch : 221, iteration : 4440, train loss : 0.476203, train loss_ce: 0.395016, train loss_dice: 0.557390, train dice : 0.442610 [18:05:13.374] epoch : 221, mean train loss : 0.486508, mean train ce loss: 0.384227, mean train dice : 0.411210 [18:05:15.666] epoch : 221, iteration : 2211, val loss : 0.452751, val loss_ce: 0.388601, val loss_dice: 0.516900, val dice : 0.483100 [18:05:22.142] epoch : 221, iteration : 2214, val loss : 0.592010, val loss_ce: 0.545301, val loss_dice: 0.638719, val dice : 0.361281 [18:05:26.846] epoch : 221, iteration : 2217, val loss : 0.419700, val loss_ce: 0.252930, val loss_dice: 0.586470, val dice : 0.413530 [18:05:37.409] epoch : 221, iteration : 2220, val loss : 0.456372, val loss_ce: 0.348983, val loss_dice: 0.563762, val dice : 0.436238 [18:05:37.410] epoch : 221, mean val loss : 0.503025, mean val ce loss: 0.423131, mean val dice : 0.417082 [18:05:37.636] save best model to ../model/TU_VerSe[128, 160, 96]/TU_pretrain_R50-ViT-B_16_skip3_10k_epo1000bs2[128, 160, 96]/best_model.pth at epoch 222, val dice: 0.41708168387413025 [18:06:18.396] epoch : 222, iteration : 4448, train loss : 0.501505, train loss_ce: 0.450121, train loss_dice: 0.552888, train dice : 0.447112 [18:07:12.904] epoch : 222, iteration : 4456, train loss : 0.439732, train loss_ce: 0.362819, train loss_dice: 0.516645, train dice : 0.483355 [18:07:37.398] epoch : 222, mean train loss : 0.462312, mean train ce loss: 0.362351, mean train dice : 0.437726 [18:07:43.656] epoch : 222, iteration : 2223, val loss : 0.429344, val loss_ce: 0.343766, val loss_dice: 0.514921, val dice : 0.485079 [18:07:52.023] epoch : 222, iteration : 2226, val loss : 0.450813, val loss_ce: 0.314080, val loss_dice: 0.587547, val dice : 0.412453 [18:07:59.829] epoch : 222, iteration : 2229, val loss : 0.468553, val loss_ce: 0.345811, val loss_dice: 0.591295, val dice : 0.408705 [18:08:03.618] epoch : 222, mean val loss : 0.508876, mean val ce loss: 0.431067, mean val dice : 0.413314 [18:08:28.978] epoch : 223, iteration : 4464, train loss : 0.428743, train loss_ce: 0.304983, train loss_dice: 0.552504, train dice : 0.447496 [18:09:13.370] epoch : 223, iteration : 4472, train loss : 0.458237, train loss_ce: 0.339192, train loss_dice: 0.577282, train dice : 0.422718 [18:10:01.834] epoch : 223, iteration : 4480, train loss : 0.452512, train loss_ce: 0.331324, train loss_dice: 0.573700, train dice : 0.426300 [18:10:01.858] epoch : 223, mean train loss : 0.461883, mean train ce loss: 0.350164, mean train dice : 0.426399 [18:10:07.140] epoch : 223, iteration : 2232, val loss : 0.592158, val loss_ce: 0.597700, val loss_dice: 0.586616, val dice : 0.413384 [18:10:18.543] epoch : 223, iteration : 2235, val loss : 0.505030, val loss_ce: 0.449292, val loss_dice: 0.560767, val dice : 0.439233 [18:10:23.812] epoch : 223, iteration : 2238, val loss : 0.627458, val loss_ce: 0.607972, val loss_dice: 0.646944, val dice : 0.353056 [18:10:27.624] epoch : 223, mean val loss : 0.553252, mean val ce loss: 0.501436, mean val dice : 0.394932 [18:11:17.640] epoch : 224, iteration : 4488, train loss : 0.417495, train loss_ce: 0.317694, train loss_dice: 0.517296, train dice : 0.482704 [18:12:09.565] epoch : 224, iteration : 4496, train loss : 0.419497, train loss_ce: 0.281620, train loss_dice: 0.557374, train dice : 0.442626 [18:12:24.608] epoch : 224, mean train loss : 0.451224, mean train ce loss: 0.334347, mean train dice : 0.431899 [18:12:27.670] epoch : 224, iteration : 2241, val loss : 0.430589, val loss_ce: 0.256593, val loss_dice: 0.604584, val dice : 0.395416 [18:12:32.179] epoch : 224, iteration : 2244, val loss : 0.451699, val loss_ce: 0.344186, val loss_dice: 0.559212, val dice : 0.440788 [18:12:38.796] epoch : 224, iteration : 2247, val loss : 0.676152, val loss_ce: 0.715927, val loss_dice: 0.636378, val dice : 0.363622 [18:12:52.579] epoch : 224, iteration : 2250, val loss : 0.569324, val loss_ce: 0.499385, val loss_dice: 0.639263, val dice : 0.360737 [18:12:52.580] epoch : 224, mean val loss : 0.524610, mean val ce loss: 0.450817, mean val dice : 0.401596 [18:13:23.043] epoch : 225, iteration : 4504, train loss : 0.541395, train loss_ce: 0.460706, train loss_dice: 0.622083, train dice : 0.377917 [18:14:07.196] epoch : 225, iteration : 4512, train loss : 0.518595, train loss_ce: 0.423366, train loss_dice: 0.613824, train dice : 0.386176 [18:14:49.405] epoch : 225, iteration : 4520, train loss : 0.465621, train loss_ce: 0.363847, train loss_dice: 0.567396, train dice : 0.432604 [18:14:49.427] epoch : 225, mean train loss : 0.491807, mean train ce loss: 0.398470, mean train dice : 0.414857 [18:14:56.323] epoch : 225, iteration : 2253, val loss : 0.476509, val loss_ce: 0.392392, val loss_dice: 0.560626, val dice : 0.439374 [18:15:00.765] epoch : 225, iteration : 2256, val loss : 0.552940, val loss_ce: 0.456129, val loss_dice: 0.649752, val dice : 0.350248 [18:15:10.509] epoch : 225, iteration : 2259, val loss : 0.600356, val loss_ce: 0.616387, val loss_dice: 0.584324, val dice : 0.415676 [18:15:12.931] epoch : 225, mean val loss : 0.649480, mean val ce loss: 0.671157, mean val dice : 0.372197 [18:15:57.594] epoch : 226, iteration : 4528, train loss : 0.516697, train loss_ce: 0.392832, train loss_dice: 0.640562, train dice : 0.359438 [18:16:40.341] epoch : 226, iteration : 4536, train loss : 0.596247, train loss_ce: 0.562226, train loss_dice: 0.630269, train dice : 0.369731 [18:17:09.088] epoch : 226, mean train loss : 0.530151, mean train ce loss: 0.455341, mean train dice : 0.395039 [18:17:15.884] epoch : 226, iteration : 2262, val loss : 0.560005, val loss_ce: 0.514260, val loss_dice: 0.605751, val dice : 0.394249 [18:17:23.693] epoch : 226, iteration : 2265, val loss : 0.558334, val loss_ce: 0.501184, val loss_dice: 0.615485, val dice : 0.384515 [18:17:29.118] epoch : 226, iteration : 2268, val loss : 0.620930, val loss_ce: 0.604458, val loss_dice: 0.637402, val dice : 0.362598 [18:17:36.532] epoch : 226, mean val loss : 0.572833, mean val ce loss: 0.532701, mean val dice : 0.387034 [18:17:59.095] epoch : 227, iteration : 4544, train loss : 0.449529, train loss_ce: 0.327360, train loss_dice: 0.571697, train dice : 0.428303 [18:18:48.991] epoch : 227, iteration : 4552, train loss : 0.637420, train loss_ce: 0.612826, train loss_dice: 0.662014, train dice : 0.337986 [18:19:36.365] epoch : 227, iteration : 4560, train loss : 0.561894, train loss_ce: 0.479890, train loss_dice: 0.643898, train dice : 0.356102 [18:19:36.387] epoch : 227, mean train loss : 0.523357, mean train ce loss: 0.435905, mean train dice : 0.389192 [18:19:42.842] epoch : 227, iteration : 2271, val loss : 0.518858, val loss_ce: 0.389061, val loss_dice: 0.648655, val dice : 0.351345 [18:19:47.209] epoch : 227, iteration : 2274, val loss : 0.828391, val loss_ce: 0.982525, val loss_dice: 0.674256, val dice : 0.325744 [18:19:53.307] epoch : 227, iteration : 2277, val loss : 0.609270, val loss_ce: 0.565387, val loss_dice: 0.653154, val dice : 0.346846 [18:20:00.585] epoch : 227, iteration : 2280, val loss : 0.568079, val loss_ce: 0.505059, val loss_dice: 0.631098, val dice : 0.368902 [18:20:00.585] epoch : 227, mean val loss : 0.677241, mean val ce loss: 0.699294, mean val dice : 0.344811 [18:20:56.939] epoch : 228, iteration : 4568, train loss : 0.514576, train loss_ce: 0.389284, train loss_dice: 0.639867, train dice : 0.360133 [18:21:46.751] epoch : 228, iteration : 4576, train loss : 0.598621, train loss_ce: 0.561616, train loss_dice: 0.635626, train dice : 0.364374 [18:22:04.017] epoch : 228, mean train loss : 0.514617, mean train ce loss: 0.419013, mean train dice : 0.389779 [18:22:11.722] epoch : 228, iteration : 2283, val loss : 0.567952, val loss_ce: 0.488728, val loss_dice: 0.647176, val dice : 0.352824 [18:22:21.935] epoch : 228, iteration : 2286, val loss : 0.587671, val loss_ce: 0.555302, val loss_dice: 0.620039, val dice : 0.379961 [18:22:29.538] epoch : 228, iteration : 2289, val loss : 0.734499, val loss_ce: 0.843359, val loss_dice: 0.625639, val dice : 0.374361 [18:22:30.434] epoch : 228, mean val loss : 0.611014, mean val ce loss: 0.588067, mean val dice : 0.366039 [18:22:57.508] epoch : 229, iteration : 4584, train loss : 0.694545, train loss_ce: 0.759115, train loss_dice: 0.629975, train dice : 0.370025 [18:23:42.637] epoch : 229, iteration : 4592, train loss : 0.480800, train loss_ce: 0.395011, train loss_dice: 0.566589, train dice : 0.433411 [18:24:29.996] epoch : 229, iteration : 4600, train loss : 0.564059, train loss_ce: 0.468773, train loss_dice: 0.659344, train dice : 0.340656 [18:24:30.022] epoch : 229, mean train loss : 0.519690, mean train ce loss: 0.438993, mean train dice : 0.399612 [18:24:37.936] epoch : 229, iteration : 2292, val loss : 0.772898, val loss_ce: 0.900249, val loss_dice: 0.645548, val dice : 0.354452 [18:24:43.545] epoch : 229, iteration : 2295, val loss : 0.521943, val loss_ce: 0.442946, val loss_dice: 0.600941, val dice : 0.399059 [18:24:57.073] epoch : 229, iteration : 2298, val loss : 0.522272, val loss_ce: 0.439036, val loss_dice: 0.605508, val dice : 0.394492 [18:24:59.112] epoch : 229, mean val loss : 0.578968, mean val ce loss: 0.540477, mean val dice : 0.382541 [18:25:47.226] epoch : 230, iteration : 4608, train loss : 0.481783, train loss_ce: 0.401483, train loss_dice: 0.562083, train dice : 0.437917 [18:26:39.516] epoch : 230, iteration : 4616, train loss : 0.542423, train loss_ce: 0.492045, train loss_dice: 0.592801, train dice : 0.407199 [18:27:04.219] epoch : 230, mean train loss : 0.502393, mean train ce loss: 0.411145, mean train dice : 0.406359 [18:27:05.805] epoch : 230, iteration : 2301, val loss : 0.598254, val loss_ce: 0.593082, val loss_dice: 0.603425, val dice : 0.396575 [18:27:14.524] epoch : 230, iteration : 2304, val loss : 0.535704, val loss_ce: 0.503137, val loss_dice: 0.568271, val dice : 0.431729 [18:27:22.233] epoch : 230, iteration : 2307, val loss : 0.684290, val loss_ce: 0.720438, val loss_dice: 0.648141, val dice : 0.351859 [18:27:27.503] epoch : 230, iteration : 2310, val loss : 0.584771, val loss_ce: 0.527792, val loss_dice: 0.641751, val dice : 0.358249 [18:27:27.504] epoch : 230, mean val loss : 0.595500, mean val ce loss: 0.582176, mean val dice : 0.391176 [18:27:48.921] epoch : 231, iteration : 4624, train loss : 0.480631, train loss_ce: 0.362801, train loss_dice: 0.598461, train dice : 0.401539 [18:28:41.242] epoch : 231, iteration : 4632, train loss : 0.501608, train loss_ce: 0.462571, train loss_dice: 0.540645, train dice : 0.459355 [18:29:26.918] epoch : 231, iteration : 4640, train loss : 0.493280, train loss_ce: 0.372176, train loss_dice: 0.614384, train dice : 0.385616 [18:29:26.952] epoch : 231, mean train loss : 0.503448, mean train ce loss: 0.415676, mean train dice : 0.408780 [18:29:36.561] epoch : 231, iteration : 2313, val loss : 0.526928, val loss_ce: 0.438591, val loss_dice: 0.615265, val dice : 0.384735 [18:29:43.192] epoch : 231, iteration : 2316, val loss : 0.526142, val loss_ce: 0.448854, val loss_dice: 0.603431, val dice : 0.396569 [18:29:49.987] epoch : 231, iteration : 2319, val loss : 0.608999, val loss_ce: 0.624379, val loss_dice: 0.593619, val dice : 0.406381 [18:29:51.559] epoch : 231, mean val loss : 0.561597, mean val ce loss: 0.518472, mean val dice : 0.395277 [18:30:44.913] epoch : 232, iteration : 4648, train loss : 0.520054, train loss_ce: 0.407415, train loss_dice: 0.632693, train dice : 0.367307 [18:31:26.248] epoch : 232, iteration : 4656, train loss : 0.501767, train loss_ce: 0.422071, train loss_dice: 0.581464, train dice : 0.418536 [18:31:51.054] epoch : 232, mean train loss : 0.492240, mean train ce loss: 0.389874, mean train dice : 0.405394 [18:31:55.226] epoch : 232, iteration : 2322, val loss : 0.564805, val loss_ce: 0.496107, val loss_dice: 0.633503, val dice : 0.366497 [18:31:59.861] epoch : 232, iteration : 2325, val loss : 0.674503, val loss_ce: 0.714499, val loss_dice: 0.634507, val dice : 0.365493 [18:32:08.130] epoch : 232, iteration : 2328, val loss : 0.496335, val loss_ce: 0.430784, val loss_dice: 0.561887, val dice : 0.438113 [18:32:14.984] epoch : 232, mean val loss : 0.544342, mean val ce loss: 0.488605, mean val dice : 0.399921 [18:32:38.661] epoch : 233, iteration : 4664, train loss : 0.456872, train loss_ce: 0.363619, train loss_dice: 0.550125, train dice : 0.449875 [18:33:28.944] epoch : 233, iteration : 4672, train loss : 0.539905, train loss_ce: 0.529488, train loss_dice: 0.550323, train dice : 0.449677 [18:34:14.196] epoch : 233, iteration : 4680, train loss : 0.650916, train loss_ce: 0.654232, train loss_dice: 0.647601, train dice : 0.352399 [18:34:14.221] epoch : 233, mean train loss : 0.492515, mean train ce loss: 0.395226, mean train dice : 0.410197 [18:34:16.629] epoch : 233, iteration : 2331, val loss : 0.636225, val loss_ce: 0.652039, val loss_dice: 0.620411, val dice : 0.379589 [18:34:26.546] epoch : 233, iteration : 2334, val loss : 0.703947, val loss_ce: 0.748865, val loss_dice: 0.659028, val dice : 0.340972 [18:34:32.437] epoch : 233, iteration : 2337, val loss : 0.464166, val loss_ce: 0.348868, val loss_dice: 0.579464, val dice : 0.420536 [18:34:37.552] epoch : 233, iteration : 2340, val loss : 0.453787, val loss_ce: 0.364526, val loss_dice: 0.543048, val dice : 0.456952 [18:34:37.552] epoch : 233, mean val loss : 0.551774, mean val ce loss: 0.497856, mean val dice : 0.394308 [18:35:14.225] epoch : 234, iteration : 4688, train loss : 0.528716, train loss_ce: 0.438593, train loss_dice: 0.618840, train dice : 0.381160 [18:36:10.982] epoch : 234, iteration : 4696, train loss : 0.469372, train loss_ce: 0.352471, train loss_dice: 0.586274, train dice : 0.413726 [18:36:37.302] epoch : 234, mean train loss : 0.486046, mean train ce loss: 0.382140, mean train dice : 0.410049 [18:36:45.506] epoch : 234, iteration : 2343, val loss : 0.761266, val loss_ce: 0.860709, val loss_dice: 0.661822, val dice : 0.338178 [18:36:52.103] epoch : 234, iteration : 2346, val loss : 0.463206, val loss_ce: 0.331433, val loss_dice: 0.594979, val dice : 0.405021 [18:37:04.661] epoch : 234, iteration : 2349, val loss : 0.521984, val loss_ce: 0.485001, val loss_dice: 0.558967, val dice : 0.441033 [18:37:06.136] epoch : 234, mean val loss : 0.574063, mean val ce loss: 0.532109, mean val dice : 0.383984 [18:37:38.122] epoch : 235, iteration : 4704, train loss : 0.495101, train loss_ce: 0.368246, train loss_dice: 0.621956, train dice : 0.378044 [18:38:22.462] epoch : 235, iteration : 4712, train loss : 0.470470, train loss_ce: 0.348858, train loss_dice: 0.592082, train dice : 0.407918 [18:39:05.506] epoch : 235, iteration : 4720, train loss : 0.434300, train loss_ce: 0.343883, train loss_dice: 0.524717, train dice : 0.475283 [18:39:05.528] epoch : 235, mean train loss : 0.482130, mean train ce loss: 0.381842, mean train dice : 0.417581 [18:39:12.256] epoch : 235, iteration : 2352, val loss : 0.542943, val loss_ce: 0.454461, val loss_dice: 0.631426, val dice : 0.368574 [18:39:19.761] epoch : 235, iteration : 2355, val loss : 0.623973, val loss_ce: 0.604272, val loss_dice: 0.643674, val dice : 0.356326 [18:39:25.642] epoch : 235, iteration : 2358, val loss : 0.508808, val loss_ce: 0.438871, val loss_dice: 0.578745, val dice : 0.421255 [18:39:30.371] epoch : 235, mean val loss : 0.515969, mean val ce loss: 0.436540, mean val dice : 0.404602 [18:40:20.041] epoch : 236, iteration : 4728, train loss : 0.565068, train loss_ce: 0.500321, train loss_dice: 0.629815, train dice : 0.370185 [18:41:18.691] epoch : 236, iteration : 4736, train loss : 0.483675, train loss_ce: 0.383398, train loss_dice: 0.583951, train dice : 0.416049 [18:41:37.495] epoch : 236, mean train loss : 0.480595, mean train ce loss: 0.380793, mean train dice : 0.419602 [18:41:39.325] epoch : 236, iteration : 2361, val loss : 0.666330, val loss_ce: 0.733737, val loss_dice: 0.598922, val dice : 0.401078 [18:41:49.810] epoch : 236, iteration : 2364, val loss : 0.520099, val loss_ce: 0.440779, val loss_dice: 0.599419, val dice : 0.400581 [18:41:58.795] epoch : 236, iteration : 2367, val loss : 0.472289, val loss_ce: 0.369144, val loss_dice: 0.575434, val dice : 0.424566 [18:42:04.049] epoch : 236, iteration : 2370, val loss : 0.467794, val loss_ce: 0.315754, val loss_dice: 0.619834, val dice : 0.380166 [18:42:04.049] epoch : 236, mean val loss : 0.515964, mean val ce loss: 0.444774, mean val dice : 0.412847 [18:42:32.112] epoch : 237, iteration : 4744, train loss : 0.625036, train loss_ce: 0.630728, train loss_dice: 0.619344, train dice : 0.380656 [18:43:14.384] epoch : 237, iteration : 4752, train loss : 0.443366, train loss_ce: 0.344574, train loss_dice: 0.542159, train dice : 0.457841 [18:44:09.768] epoch : 237, iteration : 4760, train loss : 0.399659, train loss_ce: 0.281919, train loss_dice: 0.517399, train dice : 0.482601 [18:44:09.799] epoch : 237, mean train loss : 0.482625, mean train ce loss: 0.384729, mean train dice : 0.419480 [18:44:15.583] epoch : 237, iteration : 2373, val loss : 0.767368, val loss_ce: 0.906697, val loss_dice: 0.628039, val dice : 0.371961 [18:44:27.240] epoch : 237, iteration : 2376, val loss : 0.517018, val loss_ce: 0.441419, val loss_dice: 0.592618, val dice : 0.407382 [18:44:34.295] epoch : 237, iteration : 2379, val loss : 0.420480, val loss_ce: 0.315259, val loss_dice: 0.525700, val dice : 0.474300 [18:44:36.778] epoch : 237, mean val loss : 0.531334, mean val ce loss: 0.470146, mean val dice : 0.407479 [18:45:42.595] epoch : 238, iteration : 4768, train loss : 0.508246, train loss_ce: 0.404869, train loss_dice: 0.611624, train dice : 0.388376 [18:46:33.062] epoch : 238, iteration : 4776, train loss : 0.656637, train loss_ce: 0.697100, train loss_dice: 0.616173, train dice : 0.383827 [18:46:54.416] epoch : 238, mean train loss : 0.477434, mean train ce loss: 0.380670, mean train dice : 0.425801 [18:46:58.727] epoch : 238, iteration : 2382, val loss : 0.465136, val loss_ce: 0.340943, val loss_dice: 0.589329, val dice : 0.410671 [18:47:07.442] epoch : 238, iteration : 2385, val loss : 0.583951, val loss_ce: 0.619825, val loss_dice: 0.548077, val dice : 0.451923 [18:47:17.502] epoch : 238, iteration : 2388, val loss : 0.526839, val loss_ce: 0.467401, val loss_dice: 0.586276, val dice : 0.413724 [18:47:22.166] epoch : 238, mean val loss : 0.520230, mean val ce loss: 0.478050, mean val dice : 0.437590 [18:47:22.459] save best model to ../model/TU_VerSe[128, 160, 96]/TU_pretrain_R50-ViT-B_16_skip3_10k_epo1000bs2[128, 160, 96]/best_model.pth at epoch 239, val dice: 0.4375903606414795 [18:47:54.119] epoch : 239, iteration : 4784, train loss : 0.414092, train loss_ce: 0.303901, train loss_dice: 0.524283, train dice : 0.475717 [18:48:38.371] epoch : 239, iteration : 4792, train loss : 0.437197, train loss_ce: 0.349556, train loss_dice: 0.524837, train dice : 0.475163 [18:49:31.479] epoch : 239, iteration : 4800, train loss : 0.420807, train loss_ce: 0.291968, train loss_dice: 0.549647, train dice : 0.450353 [18:49:31.519] epoch : 239, mean train loss : 0.458118, mean train ce loss: 0.348684, mean train dice : 0.432448 [18:49:34.360] epoch : 239, iteration : 2391, val loss : 0.456692, val loss_ce: 0.342687, val loss_dice: 0.570696, val dice : 0.429304 [18:49:39.565] epoch : 239, iteration : 2394, val loss : 0.493549, val loss_ce: 0.377184, val loss_dice: 0.609913, val dice : 0.390087 [18:49:46.971] epoch : 239, iteration : 2397, val loss : 0.694030, val loss_ce: 0.720825, val loss_dice: 0.667234, val dice : 0.332766 [18:49:56.729] epoch : 239, iteration : 2400, val loss : 0.682885, val loss_ce: 0.735146, val loss_dice: 0.630625, val dice : 0.369375 [18:49:56.730] epoch : 239, mean val loss : 0.511857, mean val ce loss: 0.438366, mean val dice : 0.414652 [18:50:57.019] epoch : 240, iteration : 4808, train loss : 0.427881, train loss_ce: 0.366126, train loss_dice: 0.489636, train dice : 0.510364 [18:51:36.738] epoch : 240, iteration : 4816, train loss : 0.434528, train loss_ce: 0.270140, train loss_dice: 0.598917, train dice : 0.401083 [18:51:59.410] epoch : 240, mean train loss : 0.474088, mean train ce loss: 0.374113, mean train dice : 0.425937 [18:52:10.083] epoch : 240, iteration : 2403, val loss : 0.535857, val loss_ce: 0.461933, val loss_dice: 0.609782, val dice : 0.390218 [18:52:15.069] epoch : 240, iteration : 2406, val loss : 0.526950, val loss_ce: 0.435285, val loss_dice: 0.618614, val dice : 0.381386 [18:52:19.271] epoch : 240, iteration : 2409, val loss : 0.668363, val loss_ce: 0.719207, val loss_dice: 0.617519, val dice : 0.382481 [18:52:22.647] epoch : 240, mean val loss : 0.502829, mean val ce loss: 0.422568, mean val dice : 0.416909 [18:52:58.453] epoch : 241, iteration : 4824, train loss : 0.463510, train loss_ce: 0.373335, train loss_dice: 0.553684, train dice : 0.446316 [18:53:43.593] epoch : 241, iteration : 4832, train loss : 0.443457, train loss_ce: 0.343557, train loss_dice: 0.543357, train dice : 0.456643 [18:54:27.758] epoch : 241, iteration : 4840, train loss : 0.550547, train loss_ce: 0.474306, train loss_dice: 0.626788, train dice : 0.373212 [18:54:27.784] epoch : 241, mean train loss : 0.465996, mean train ce loss: 0.363293, mean train dice : 0.431300 [18:54:33.344] epoch : 241, iteration : 2412, val loss : 0.735449, val loss_ce: 0.790262, val loss_dice: 0.680636, val dice : 0.319364 [18:54:36.947] epoch : 241, iteration : 2415, val loss : 0.411936, val loss_ce: 0.282863, val loss_dice: 0.541008, val dice : 0.458992 [18:54:47.223] epoch : 241, iteration : 2418, val loss : 0.473754, val loss_ce: 0.370743, val loss_dice: 0.576765, val dice : 0.423235 [18:54:51.143] epoch : 241, mean val loss : 0.515456, mean val ce loss: 0.441622, mean val dice : 0.410711 [18:55:53.875] epoch : 242, iteration : 4848, train loss : 0.587043, train loss_ce: 0.532318, train loss_dice: 0.641768, train dice : 0.358232 [18:56:32.215] epoch : 242, iteration : 4856, train loss : 0.385783, train loss_ce: 0.248208, train loss_dice: 0.523358, train dice : 0.476642 [18:56:59.341] epoch : 242, mean train loss : 0.452507, mean train ce loss: 0.339839, mean train dice : 0.434826 [18:57:03.041] epoch : 242, iteration : 2421, val loss : 0.523436, val loss_ce: 0.404102, val loss_dice: 0.642770, val dice : 0.357230 [18:57:08.109] epoch : 242, iteration : 2424, val loss : 0.505186, val loss_ce: 0.382706, val loss_dice: 0.627666, val dice : 0.372334 [18:57:14.088] epoch : 242, iteration : 2427, val loss : 0.528673, val loss_ce: 0.428347, val loss_dice: 0.628999, val dice : 0.371001 [18:57:22.954] epoch : 242, iteration : 2430, val loss : 0.459949, val loss_ce: 0.375033, val loss_dice: 0.544864, val dice : 0.455136 [18:57:22.955] epoch : 242, mean val loss : 0.502970, mean val ce loss: 0.424470, mean val dice : 0.418530 [18:57:48.696] epoch : 243, iteration : 4864, train loss : 0.510277, train loss_ce: 0.423874, train loss_dice: 0.596680, train dice : 0.403320 [18:58:34.904] epoch : 243, iteration : 4872, train loss : 0.457913, train loss_ce: 0.322100, train loss_dice: 0.593727, train dice : 0.406273 [18:59:22.787] epoch : 243, iteration : 4880, train loss : 0.382589, train loss_ce: 0.245173, train loss_dice: 0.520005, train dice : 0.479995 [18:59:22.816] epoch : 243, mean train loss : 0.458545, mean train ce loss: 0.349726, mean train dice : 0.432635 [18:59:35.360] epoch : 243, iteration : 2433, val loss : 0.635984, val loss_ce: 0.674637, val loss_dice: 0.597332, val dice : 0.402668 [18:59:41.607] epoch : 243, iteration : 2436, val loss : 0.492778, val loss_ce: 0.423875, val loss_dice: 0.561681, val dice : 0.438319 [18:59:46.882] epoch : 243, iteration : 2439, val loss : 0.445572, val loss_ce: 0.346660, val loss_dice: 0.544485, val dice : 0.455515 [18:59:48.200] epoch : 243, mean val loss : 0.487326, mean val ce loss: 0.420461, mean val dice : 0.445809 [18:59:48.419] save best model to ../model/TU_VerSe[128, 160, 96]/TU_pretrain_R50-ViT-B_16_skip3_10k_epo1000bs2[128, 160, 96]/best_model.pth at epoch 244, val dice: 0.44580888748168945 [19:00:33.848] epoch : 244, iteration : 4888, train loss : 0.441147, train loss_ce: 0.370021, train loss_dice: 0.512273, train dice : 0.487727 [19:01:25.764] epoch : 244, iteration : 4896, train loss : 0.433941, train loss_ce: 0.278367, train loss_dice: 0.589515, train dice : 0.410485 [19:01:49.082] epoch : 244, mean train loss : 0.448368, mean train ce loss: 0.330373, mean train dice : 0.433638 [19:01:53.376] epoch : 244, iteration : 2442, val loss : 0.644373, val loss_ce: 0.655967, val loss_dice: 0.632779, val dice : 0.367221 [19:01:59.157] epoch : 244, iteration : 2445, val loss : 0.410745, val loss_ce: 0.308276, val loss_dice: 0.513213, val dice : 0.486787 [19:02:07.735] epoch : 244, iteration : 2448, val loss : 0.406374, val loss_ce: 0.292414, val loss_dice: 0.520333, val dice : 0.479667 [19:02:14.126] epoch : 244, mean val loss : 0.489104, mean val ce loss: 0.408199, mean val dice : 0.429991 [19:02:39.345] epoch : 245, iteration : 4904, train loss : 0.434040, train loss_ce: 0.327944, train loss_dice: 0.540137, train dice : 0.459863 [19:03:24.388] epoch : 245, iteration : 4912, train loss : 0.541118, train loss_ce: 0.450737, train loss_dice: 0.631499, train dice : 0.368501 [19:04:13.051] epoch : 245, iteration : 4920, train loss : 0.420718, train loss_ce: 0.354349, train loss_dice: 0.487087, train dice : 0.512913 [19:04:13.075] epoch : 245, mean train loss : 0.455525, mean train ce loss: 0.336811, mean train dice : 0.425760 [19:04:15.939] epoch : 245, iteration : 2451, val loss : 0.456028, val loss_ce: 0.365019, val loss_dice: 0.547037, val dice : 0.452963 [19:04:22.168] epoch : 245, iteration : 2454, val loss : 0.494869, val loss_ce: 0.396538, val loss_dice: 0.593199, val dice : 0.406801 [19:04:31.010] epoch : 245, iteration : 2457, val loss : 0.451313, val loss_ce: 0.374646, val loss_dice: 0.527979, val dice : 0.472021 [19:04:37.335] epoch : 245, iteration : 2460, val loss : 0.719799, val loss_ce: 0.816516, val loss_dice: 0.623082, val dice : 0.376918 [19:04:37.336] epoch : 245, mean val loss : 0.484856, mean val ce loss: 0.415119, mean val dice : 0.445408 [19:05:30.464] epoch : 246, iteration : 4928, train loss : 0.464424, train loss_ce: 0.337978, train loss_dice: 0.590871, train dice : 0.409129 [19:06:11.832] epoch : 246, iteration : 4936, train loss : 0.491802, train loss_ce: 0.411296, train loss_dice: 0.572308, train dice : 0.427692 [19:06:39.037] epoch : 246, mean train loss : 0.447735, mean train ce loss: 0.336406, mean train dice : 0.440936 [19:06:48.789] epoch : 246, iteration : 2463, val loss : 0.513220, val loss_ce: 0.384329, val loss_dice: 0.642111, val dice : 0.357889 [19:06:54.939] epoch : 246, iteration : 2466, val loss : 0.588692, val loss_ce: 0.525361, val loss_dice: 0.652023, val dice : 0.347977 [19:07:01.251] epoch : 246, iteration : 2469, val loss : 0.660719, val loss_ce: 0.650600, val loss_dice: 0.670838, val dice : 0.329162 [19:07:02.380] epoch : 246, mean val loss : 0.526715, mean val ce loss: 0.448278, mean val dice : 0.394849 [19:07:24.011] epoch : 247, iteration : 4944, train loss : 0.432931, train loss_ce: 0.320900, train loss_dice: 0.544961, train dice : 0.455039 [19:08:13.272] epoch : 247, iteration : 4952, train loss : 0.373250, train loss_ce: 0.268839, train loss_dice: 0.477661, train dice : 0.522339 [19:09:01.493] epoch : 247, iteration : 4960, train loss : 0.401550, train loss_ce: 0.322363, train loss_dice: 0.480737, train dice : 0.519263 [19:09:01.517] epoch : 247, mean train loss : 0.429379, mean train ce loss: 0.321564, mean train dice : 0.462806 [19:09:04.437] epoch : 247, iteration : 2472, val loss : 0.411354, val loss_ce: 0.303853, val loss_dice: 0.518854, val dice : 0.481146 [19:09:11.852] epoch : 247, iteration : 2475, val loss : 0.381061, val loss_ce: 0.292348, val loss_dice: 0.469774, val dice : 0.530226 [19:09:17.670] epoch : 247, iteration : 2478, val loss : 0.838209, val loss_ce: 1.064741, val loss_dice: 0.611678, val dice : 0.388322 [19:09:26.208] epoch : 247, mean val loss : 0.505544, mean val ce loss: 0.433717, mean val dice : 0.422630 [19:10:16.752] epoch : 248, iteration : 4968, train loss : 0.405634, train loss_ce: 0.326550, train loss_dice: 0.484717, train dice : 0.515283 [19:11:03.025] epoch : 248, iteration : 4976, train loss : 0.420996, train loss_ce: 0.322849, train loss_dice: 0.519142, train dice : 0.480858 [19:11:22.781] epoch : 248, mean train loss : 0.442251, mean train ce loss: 0.324446, mean train dice : 0.439945 [19:11:24.308] epoch : 248, iteration : 2481, val loss : 0.465499, val loss_ce: 0.309315, val loss_dice: 0.621683, val dice : 0.378317 [19:11:34.771] epoch : 248, iteration : 2484, val loss : 0.512207, val loss_ce: 0.444130, val loss_dice: 0.580285, val dice : 0.419715 [19:11:40.883] epoch : 248, iteration : 2487, val loss : 0.460554, val loss_ce: 0.358344, val loss_dice: 0.562764, val dice : 0.437236 [19:11:46.525] epoch : 248, iteration : 2490, val loss : 0.498986, val loss_ce: 0.391894, val loss_dice: 0.606077, val dice : 0.393923 [19:11:46.525] epoch : 248, mean val loss : 0.506528, mean val ce loss: 0.436100, mean val dice : 0.423045 [19:12:11.085] epoch : 249, iteration : 4984, train loss : 0.436950, train loss_ce: 0.371329, train loss_dice: 0.502571, train dice : 0.497429 [19:13:03.192] epoch : 249, iteration : 4992, train loss : 0.409231, train loss_ce: 0.271198, train loss_dice: 0.547265, train dice : 0.452735 [19:13:45.601] epoch : 249, iteration : 5000, train loss : 0.411030, train loss_ce: 0.359788, train loss_dice: 0.462272, train dice : 0.537728 [19:13:45.634] epoch : 249, mean train loss : 0.440938, mean train ce loss: 0.336297, mean train dice : 0.454421 [19:13:52.075] epoch : 249, iteration : 2493, val loss : 0.405502, val loss_ce: 0.311153, val loss_dice: 0.499852, val dice : 0.500148 [19:13:56.591] epoch : 249, iteration : 2496, val loss : 0.408573, val loss_ce: 0.300315, val loss_dice: 0.516831, val dice : 0.483169 [19:14:04.865] epoch : 249, iteration : 2499, val loss : 0.518696, val loss_ce: 0.478452, val loss_dice: 0.558940, val dice : 0.441060 [19:14:09.304] epoch : 249, mean val loss : 0.506586, mean val ce loss: 0.454135, mean val dice : 0.440962 [19:14:52.369] epoch : 250, iteration : 5008, train loss : 0.518009, train loss_ce: 0.446680, train loss_dice: 0.589339, train dice : 0.410661 [19:15:43.745] epoch : 250, iteration : 5016, train loss : 0.530675, train loss_ce: 0.461523, train loss_dice: 0.599827, train dice : 0.400173 [19:16:07.768] epoch : 250, mean train loss : 0.468734, mean train ce loss: 0.364105, mean train dice : 0.426636 [19:16:11.402] epoch : 250, iteration : 2502, val loss : 0.520929, val loss_ce: 0.443499, val loss_dice: 0.598359, val dice : 0.401641 [19:16:16.964] epoch : 250, iteration : 2505, val loss : 0.419263, val loss_ce: 0.264663, val loss_dice: 0.573862, val dice : 0.426138 [19:16:27.341] epoch : 250, iteration : 2508, val loss : 0.573564, val loss_ce: 0.558201, val loss_dice: 0.588928, val dice : 0.411072 [19:16:30.805] epoch : 250, mean val loss : 0.532790, mean val ce loss: 0.462567, mean val dice : 0.396986 [19:17:01.832] epoch : 251, iteration : 5024, train loss : 0.472343, train loss_ce: 0.348604, train loss_dice: 0.596083, train dice : 0.403917 [19:17:49.348] epoch : 251, iteration : 5032, train loss : 0.396743, train loss_ce: 0.223616, train loss_dice: 0.569870, train dice : 0.430130 [19:18:31.640] epoch : 251, iteration : 5040, train loss : 0.462730, train loss_ce: 0.312186, train loss_dice: 0.613273, train dice : 0.386727 [19:18:31.676] epoch : 251, mean train loss : 0.501349, mean train ce loss: 0.411274, mean train dice : 0.408575 [19:18:33.834] epoch : 251, iteration : 2511, val loss : 0.653781, val loss_ce: 0.665973, val loss_dice: 0.641590, val dice : 0.358410 [19:18:40.659] epoch : 251, iteration : 2514, val loss : 0.466057, val loss_ce: 0.400094, val loss_dice: 0.532019, val dice : 0.467981 [19:18:47.745] epoch : 251, iteration : 2517, val loss : 0.582677, val loss_ce: 0.530452, val loss_dice: 0.634902, val dice : 0.365098 [19:18:57.314] epoch : 251, iteration : 2520, val loss : 0.511959, val loss_ce: 0.412172, val loss_dice: 0.611746, val dice : 0.388254 [19:18:57.314] epoch : 251, mean val loss : 0.560462, mean val ce loss: 0.508739, mean val dice : 0.387814 [19:19:41.099] epoch : 252, iteration : 5048, train loss : 0.544034, train loss_ce: 0.492233, train loss_dice: 0.595836, train dice : 0.404164 [19:20:36.533] epoch : 252, iteration : 5056, train loss : 0.451951, train loss_ce: 0.325917, train loss_dice: 0.577984, train dice : 0.422016 [19:20:57.954] epoch : 252, mean train loss : 0.501085, mean train ce loss: 0.409853, mean train dice : 0.407683 [19:21:04.930] epoch : 252, iteration : 2523, val loss : 0.597217, val loss_ce: 0.569529, val loss_dice: 0.624905, val dice : 0.375095 [19:21:14.870] epoch : 252, iteration : 2526, val loss : 0.590290, val loss_ce: 0.604577, val loss_dice: 0.576003, val dice : 0.423997 [19:21:19.956] epoch : 252, iteration : 2529, val loss : 0.816948, val loss_ce: 0.954074, val loss_dice: 0.679823, val dice : 0.320177 [19:21:22.722] epoch : 252, mean val loss : 0.614470, mean val ce loss: 0.604756, mean val dice : 0.375817 [19:21:43.998] epoch : 253, iteration : 5064, train loss : 0.540637, train loss_ce: 0.474261, train loss_dice: 0.607014, train dice : 0.392986 [19:22:36.473] epoch : 253, iteration : 5072, train loss : 0.479556, train loss_ce: 0.361735, train loss_dice: 0.597378, train dice : 0.402622 [19:23:20.281] epoch : 253, iteration : 5080, train loss : 0.444579, train loss_ce: 0.358820, train loss_dice: 0.530338, train dice : 0.469662 [19:23:20.311] epoch : 253, mean train loss : 0.501719, mean train ce loss: 0.399166, mean train dice : 0.395727 [19:23:24.083] epoch : 253, iteration : 2532, val loss : 0.465672, val loss_ce: 0.341756, val loss_dice: 0.589589, val dice : 0.410411 [19:23:30.058] epoch : 253, iteration : 2535, val loss : 0.546804, val loss_ce: 0.491566, val loss_dice: 0.602041, val dice : 0.397959 [19:23:39.399] epoch : 253, iteration : 2538, val loss : 0.492754, val loss_ce: 0.378806, val loss_dice: 0.606701, val dice : 0.393299 [19:23:43.884] epoch : 253, mean val loss : 0.514012, mean val ce loss: 0.445747, mean val dice : 0.417723 [19:24:24.923] epoch : 254, iteration : 5088, train loss : 0.541243, train loss_ce: 0.463417, train loss_dice: 0.619069, train dice : 0.380931 [19:25:15.507] epoch : 254, iteration : 5096, train loss : 0.472618, train loss_ce: 0.335003, train loss_dice: 0.610232, train dice : 0.389768 [19:25:41.821] epoch : 254, mean train loss : 0.477912, mean train ce loss: 0.369288, mean train dice : 0.413465 [19:25:44.233] epoch : 254, iteration : 2541, val loss : 0.425721, val loss_ce: 0.242462, val loss_dice: 0.608979, val dice : 0.391021 [19:25:48.758] epoch : 254, iteration : 2544, val loss : 0.473352, val loss_ce: 0.368890, val loss_dice: 0.577814, val dice : 0.422186 [19:25:58.794] epoch : 254, iteration : 2547, val loss : 0.648355, val loss_ce: 0.659757, val loss_dice: 0.636952, val dice : 0.363048 [19:26:04.920] epoch : 254, iteration : 2550, val loss : 0.486871, val loss_ce: 0.394176, val loss_dice: 0.579566, val dice : 0.420434 [19:26:04.921] epoch : 254, mean val loss : 0.556487, mean val ce loss: 0.506643, mean val dice : 0.393669 [19:26:35.521] epoch : 255, iteration : 5104, train loss : 0.633332, train loss_ce: 0.673483, train loss_dice: 0.593182, train dice : 0.406818 [19:27:23.885] epoch : 255, iteration : 5112, train loss : 0.482773, train loss_ce: 0.383064, train loss_dice: 0.582482, train dice : 0.417518 [19:28:03.421] epoch : 255, iteration : 5120, train loss : 0.491270, train loss_ce: 0.385489, train loss_dice: 0.597052, train dice : 0.402948 [19:28:03.443] epoch : 255, mean train loss : 0.481107, mean train ce loss: 0.384652, mean train dice : 0.422437 [19:28:11.807] epoch : 255, iteration : 2553, val loss : 0.432482, val loss_ce: 0.359811, val loss_dice: 0.505154, val dice : 0.494846 [19:28:18.632] epoch : 255, iteration : 2556, val loss : 0.419343, val loss_ce: 0.324765, val loss_dice: 0.513921, val dice : 0.486079 [19:28:25.926] epoch : 255, iteration : 2559, val loss : 0.694223, val loss_ce: 0.742483, val loss_dice: 0.645963, val dice : 0.354037 [19:28:26.692] epoch : 255, mean val loss : 0.531167, mean val ce loss: 0.476434, mean val dice : 0.414099

Fucapisun commented 4 months ago

原作者在训练到我这个轮次的时候已经有七十多精准度了,我训练了1000个epoch最多就到六十多,我感觉就是图像预处理的问题比较大,而且我用的是4090差距不应该这么大

genius-7 commented 4 months ago

我之前跑的结果应该和你的差不多。预处理影响没这么大吧,我也不太清楚怎么回事 ---- 回复的原邮件 ---- @.>发送日期2024年6月21日 @.>@.>, @.>主题Re: [genius-7/ResNet18_CIFAR-10] verse19数据集图像预处理 (Issue #2) 原作者在训练到我这个轮次的时候已经有七十多精准度了,我训练了1000个epoch最多就到六十多,我感觉就是图像预处理的问题比较大,而且我用的是4090差距不应该这么大 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you commented.Message ID: @.***>