From f4f3f04666b72c81f66edeb613bfecbf9b2e1334 Mon Sep 17 00:00:00 2001 From: Nic Ma Date: Wed, 20 Jul 2022 20:02:44 +0800 Subject: [PATCH 1/5] [DLMED] skip tracking meta in random transforms Signed-off-by: Nic Ma --- acceleration/fast_training_tutorial.ipynb | 300 ++++++++++++++++++++-- 1 file changed, 283 insertions(+), 17 deletions(-) diff --git a/acceleration/fast_training_tutorial.ipynb b/acceleration/fast_training_tutorial.ipynb index 8474ba213a..04743d5777 100644 --- a/acceleration/fast_training_tutorial.ipynb +++ b/acceleration/fast_training_tutorial.ipynb @@ -71,13 +71,55 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 1, "metadata": { "vscode": { "languageId": "python" } }, - "outputs": [], + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/opt/conda/lib/python3.8/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", + " from .autonotebook import tqdm as notebook_tqdm\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "MONAI version: 0+unknown\n", + "Numpy version: 1.22.4\n", + "Pytorch version: 1.13.0a0+340c412\n", + "MONAI flags: HAS_EXT = False, USE_COMPILED = False, USE_META_DICT = False\n", + "MONAI rev id: None\n", + "MONAI __file__: /workspace/data/medical/MONAI/monai/__init__.py\n", + "\n", + "Optional dependencies:\n", + "Pytorch Ignite version: 0.4.9\n", + "Nibabel version: 4.0.1\n", + "scikit-image version: 0.19.3\n", + "Pillow version: 9.0.1\n", + "Tensorboard version: 2.9.1\n", + "gdown version: 4.5.1\n", + "TorchVision version: 0.13.0a0\n", + "tqdm version: 4.64.0\n", + "lmdb version: 1.3.0\n", + "psutil version: 5.9.1\n", + "pandas version: 1.3.5\n", + "einops version: 0.4.1\n", + "transformers version: 4.20.1\n", + "mlflow version: 1.27.0\n", + "pynrrd version: 0.4.3\n", + "\n", + "For details about installing the optional dependencies, please visit:\n", + " https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n", + "\n" + ] + } + ], "source": [ "# Copyright 2020 MONAI Consortium\n", "# Licensed under the Apache License, Version 2.0 (the \"License\");\n", @@ -108,6 +150,8 @@ " ThreadDataLoader,\n", " Dataset,\n", " decollate_batch,\n", + " get_track_meta,\n", + " set_track_meta,\n", ")\n", "from monai.inferers import sliding_window_inference\n", "from monai.losses import DiceLoss, DiceCELoss\n", @@ -119,6 +163,7 @@ " AsDiscrete,\n", " Compose,\n", " CropForegroundd,\n", + " EnsureTyped,\n", " FgBgToIndicesd,\n", " LoadImaged,\n", " Orientationd,\n", @@ -150,13 +195,21 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 2, "metadata": { "vscode": { "languageId": "python" } }, - "outputs": [], + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "root dir is: /workspace/data/medical\n" + ] + } + ], "source": [ "directory = os.environ.get(\"MONAI_DATA_DIRECTORY\")\n", "root_dir = tempfile.mkdtemp() if directory is None else directory\n", @@ -178,7 +231,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 3, "metadata": { "vscode": { "languageId": "python" @@ -232,7 +285,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 4, "metadata": { "vscode": { "languageId": "python" @@ -270,7 +323,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 5, "metadata": { "vscode": { "languageId": "python" @@ -296,7 +349,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 6, "metadata": { "vscode": { "languageId": "python" @@ -326,7 +379,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 7, "metadata": { "vscode": { "languageId": "python" @@ -366,7 +419,10 @@ "\n", " if fast:\n", " # move the data to GPU and cache to avoid CPU -> GPU sync in every epoch\n", - " train_transforms.append(range_func(\"ToDevice\", ToDeviced(keys=[\"image\", \"label\"], device=device)))\n", + " train_transforms.extend([\n", + " range_func(\"EnsureType\", EnsureTyped(keys=[\"image\", \"label\"], track_meta=False)),\n", + " range_func(\"ToDevice\", ToDeviced(keys=[\"image\", \"label\"], device=device)),\n", + " ])\n", "\n", " train_transforms.append(\n", " # randomly crop out patch samples from big\n", @@ -406,9 +462,10 @@ " ]\n", " if fast:\n", " # move the data to GPU and cache to avoid CPU -> GPU sync in every epoch\n", - " val_transforms.append(\n", - " ToDeviced(keys=[\"image\", \"label\"], device=device)\n", - " )\n", + " val_transforms.extend([\n", + " EnsureTyped(keys=[\"image\", \"label\"], track_meta=False),\n", + " ToDeviced(keys=[\"image\", \"label\"], device=device),\n", + " ])\n", "\n", " return Compose(train_transforms), Compose(val_transforms)" ] @@ -433,7 +490,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 8, "metadata": { "vscode": { "languageId": "python" @@ -444,6 +501,7 @@ "def train_process(fast=False):\n", " learning_rate = 2e-4\n", " val_interval = 5 # do validation for every epoch\n", + " set_track_meta(True)\n", "\n", " if torch.cuda.is_available():\n", " device = torch.device(\"cuda:0\")\n", @@ -496,6 +554,7 @@ " bias=True,\n", " dimensions=None,\n", " ).to(device)\n", + " set_track_meta(False)\n", " else:\n", " train_ds = Dataset(data=train_files, transform=train_trans)\n", " val_ds = Dataset(data=val_files, transform=val_trans)\n", @@ -662,7 +721,6 @@ " f\" best mean dice: {best_metric:.4f}\"\n", " f\" at epoch: {best_metric_epoch}\"\n", " )\n", - "\n", " print(\n", " f\"time consuming of epoch {epoch + 1} is:\"\n", " f\" {(time.time() - epoch_start):.4f}\"\n", @@ -728,13 +786,221 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 9, "metadata": { "vscode": { "languageId": "python" } }, - "outputs": [], + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "Loading dataset: 100%|█████████████████████████████████████████████████████████████████████| 32/32 [00:42<00:00, 1.33s/it]\n", + "Loading dataset: 100%|███████████████████████████████████████████████████████████████████████| 9/9 [00:09<00:00, 1.03s/it]\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "----------\n", + "epoch 1/600\n", + "1/8, train_loss: 0.8296 step time: 1.7046\n", + "2/8, train_loss: 0.7641 step time: 0.2115\n", + "3/8, train_loss: 0.5814 step time: 0.2111\n", + "4/8, train_loss: 0.4991 step time: 0.2139\n", + "5/8, train_loss: 0.4853 step time: 0.2131\n", + "6/8, train_loss: 0.4718 step time: 0.2126\n", + "7/8, train_loss: 0.4114 step time: 0.1999\n", + "8/8, train_loss: 0.4430 step time: 0.1998\n", + "epoch 1 average loss: 0.5607\n", + "time consuming of epoch 1 is: 3.1678\n", + "----------\n", + "epoch 2/600\n", + "1/8, train_loss: 0.4435 step time: 0.2696\n", + "2/8, train_loss: 0.4586 step time: 0.2147\n", + "3/8, train_loss: 0.4187 step time: 0.2169\n", + "4/8, train_loss: 0.5169 step time: 0.2079\n", + "5/8, train_loss: 0.5513 step time: 0.2138\n", + "6/8, train_loss: 0.4491 step time: 0.2146\n", + "7/8, train_loss: 0.5062 step time: 0.2003\n", + "8/8, train_loss: 0.5096 step time: 0.2000\n", + "epoch 2 average loss: 0.4817\n", + "time consuming of epoch 2 is: 1.7395\n", + "----------\n", + "epoch 3/600\n", + "1/8, train_loss: 0.4188 step time: 0.2705\n", + "2/8, train_loss: 0.4919 step time: 0.2170\n", + "3/8, train_loss: 0.5508 step time: 0.2171\n", + "4/8, train_loss: 0.4027 step time: 0.2164\n", + "5/8, train_loss: 0.4910 step time: 0.2196\n", + "6/8, train_loss: 0.3804 step time: 0.2190\n", + "7/8, train_loss: 0.3431 step time: 0.1993\n", + "8/8, train_loss: 0.3834 step time: 0.2004\n", + "epoch 3 average loss: 0.4328\n", + "time consuming of epoch 3 is: 1.7609\n", + "----------\n", + "epoch 4/600\n", + "1/8, train_loss: 0.2439 step time: 0.2722\n", + "2/8, train_loss: 0.2532 step time: 0.2148\n", + "3/8, train_loss: 0.2174 step time: 0.2144\n", + "4/8, train_loss: 0.2359 step time: 0.2145\n", + "5/8, train_loss: 0.1941 step time: 0.2153\n", + "6/8, train_loss: 0.2294 step time: 0.2164\n", + "7/8, train_loss: 0.2436 step time: 0.1986\n", + "8/8, train_loss: 0.3432 step time: 0.1989\n", + "epoch 4 average loss: 0.2451\n", + "time consuming of epoch 4 is: 1.7466\n", + "----------\n", + "epoch 5/600\n", + "1/8, train_loss: 0.1843 step time: 0.2684\n", + "2/8, train_loss: 0.2163 step time: 0.2153\n", + "3/8, train_loss: 0.1557 step time: 0.2170\n", + "4/8, train_loss: 0.1434 step time: 0.2166\n", + "5/8, train_loss: 0.1158 step time: 0.2157\n", + "6/8, train_loss: 0.2312 step time: 0.2171\n", + "7/8, train_loss: 0.2958 step time: 0.1981\n", + "8/8, train_loss: 0.1528 step time: 0.1999\n", + "epoch 5 average loss: 0.1869\n", + "saved new best metric model\n", + "current epoch: 5 current mean dice: 0.0626 best mean dice: 0.0626 at epoch: 5\n", + "time consuming of epoch 5 is: 2.8436\n", + "----------\n", + "epoch 6/600\n", + "1/8, train_loss: 0.1136 step time: 0.2639\n", + "2/8, train_loss: 0.1565 step time: 0.2111\n", + "3/8, train_loss: 0.2769 step time: 0.2117\n", + "4/8, train_loss: 0.1773 step time: 0.2094\n", + "5/8, train_loss: 0.2183 step time: 0.2107\n", + "6/8, train_loss: 0.1236 step time: 0.2133\n", + "7/8, train_loss: 0.2068 step time: 0.1992\n", + "8/8, train_loss: 0.1687 step time: 0.1977\n", + "epoch 6 average loss: 0.1802\n", + "time consuming of epoch 6 is: 1.7192\n", + "----------\n", + "epoch 7/600\n", + "1/8, train_loss: 0.1709 step time: 0.2565\n", + "2/8, train_loss: 0.2031 step time: 0.2167\n", + "3/8, train_loss: 0.1453 step time: 0.2137\n", + "4/8, train_loss: 0.2224 step time: 0.2145\n", + "5/8, train_loss: 0.1693 step time: 0.2150\n", + "6/8, train_loss: 0.3097 step time: 0.2107\n", + "7/8, train_loss: 0.3081 step time: 0.1987\n", + "8/8, train_loss: 0.1908 step time: 0.1984\n", + "epoch 7 average loss: 0.2150\n", + "time consuming of epoch 7 is: 1.7256\n", + "----------\n", + "epoch 8/600\n", + "1/8, train_loss: 0.1996 step time: 0.2717\n", + "2/8, train_loss: 0.2088 step time: 0.2170\n", + "3/8, train_loss: 0.1894 step time: 0.2191\n", + "4/8, train_loss: 0.2062 step time: 0.2193\n", + "5/8, train_loss: 0.1687 step time: 0.2193\n", + "6/8, train_loss: 0.1327 step time: 0.2152\n", + "7/8, train_loss: 0.1556 step time: 0.2005\n", + "8/8, train_loss: 0.1491 step time: 0.2002\n", + "epoch 8 average loss: 0.1762\n", + "time consuming of epoch 8 is: 1.7639\n", + "----------\n", + "epoch 9/600\n", + "1/8, train_loss: 0.1722 step time: 0.2708\n", + "2/8, train_loss: 0.0934 step time: 0.2165\n", + "3/8, train_loss: 0.1434 step time: 0.2189\n", + "4/8, train_loss: 0.0838 step time: 0.2178\n", + "5/8, train_loss: 0.1419 step time: 0.2159\n", + "6/8, train_loss: 0.1540 step time: 0.2136\n", + "7/8, train_loss: 0.1397 step time: 0.1987\n", + "8/8, train_loss: 0.1324 step time: 0.1980\n", + "epoch 9 average loss: 0.1326\n", + "time consuming of epoch 9 is: 1.7519\n", + "----------\n", + "epoch 10/600\n", + "1/8, train_loss: 0.1358 step time: 0.2682\n", + "2/8, train_loss: 0.0968 step time: 0.2160\n", + "3/8, train_loss: 0.1124 step time: 0.2181\n", + "4/8, train_loss: 0.1009 step time: 0.2163\n", + "5/8, train_loss: 0.1119 step time: 0.2168\n", + "6/8, train_loss: 0.1132 step time: 0.2163\n", + "7/8, train_loss: 0.0996 step time: 0.1994\n", + "8/8, train_loss: 0.1189 step time: 0.1997\n", + "epoch 10 average loss: 0.1112\n", + "current epoch: 10 current mean dice: 0.0000 best mean dice: 0.0626 at epoch: 5\n", + "time consuming of epoch 10 is: 2.5712\n", + "----------\n", + "epoch 11/600\n", + "1/8, train_loss: 0.0958 step time: 0.2679\n", + "2/8, train_loss: 0.0749 step time: 0.2134\n", + "3/8, train_loss: 0.1171 step time: 0.2122\n", + "4/8, train_loss: 0.1097 step time: 0.2140\n", + "5/8, train_loss: 0.1265 step time: 0.2134\n", + "6/8, train_loss: 0.1216 step time: 0.2140\n", + "7/8, train_loss: 0.1101 step time: 0.1985\n", + "8/8, train_loss: 0.1316 step time: 0.1983\n", + "epoch 11 average loss: 0.1109\n", + "time consuming of epoch 11 is: 1.7330\n", + "----------\n", + "epoch 12/600\n", + "1/8, train_loss: 0.1670 step time: 0.2706\n", + "2/8, train_loss: 0.0785 step time: 0.2156\n", + "3/8, train_loss: 0.1389 step time: 0.2166\n", + "4/8, train_loss: 0.1350 step time: 0.2182\n", + "5/8, train_loss: 0.1447 step time: 0.2166\n", + "6/8, train_loss: 0.1352 step time: 0.2146\n", + "7/8, train_loss: 0.0995 step time: 0.1989\n", + "8/8, train_loss: 0.0594 step time: 0.1991\n", + "epoch 12 average loss: 0.1198\n", + "time consuming of epoch 12 is: 1.7517\n", + "----------\n", + "epoch 13/600\n", + "1/8, train_loss: 0.0674 step time: 0.2714\n", + "2/8, train_loss: 0.1410 step time: 0.2171\n", + "3/8, train_loss: 0.1068 step time: 0.2182\n", + "4/8, train_loss: 0.1073 step time: 0.2165\n", + "5/8, train_loss: 0.0862 step time: 0.2147\n", + "6/8, train_loss: 0.0855 step time: 0.2153\n", + "7/8, train_loss: 0.0986 step time: 0.1986\n", + "8/8, train_loss: 0.2790 step time: 0.1977\n", + "epoch 13 average loss: 0.1215\n", + "time consuming of epoch 13 is: 1.7513\n", + "----------\n", + "epoch 14/600\n", + "1/8, train_loss: 0.0853 step time: 0.2713\n", + "2/8, train_loss: 0.0896 step time: 0.2175\n", + "3/8, train_loss: 0.1138 step time: 0.2190\n", + "4/8, train_loss: 0.0790 step time: 0.2178\n", + "5/8, train_loss: 0.0676 step time: 0.2144\n", + "6/8, train_loss: 0.1342 step time: 0.2163\n", + "7/8, train_loss: 0.1085 step time: 0.1994\n", + "8/8, train_loss: 0.0938 step time: 0.1996\n", + "epoch 14 average loss: 0.0965\n", + "time consuming of epoch 14 is: 1.7572\n", + "----------\n", + "epoch 15/600\n", + "1/8, train_loss: 0.0492 step time: 0.2705\n", + "2/8, train_loss: 0.0648 step time: 0.2182\n", + "3/8, train_loss: 0.0709 step time: 0.2169\n", + "4/8, train_loss: 0.0982 step time: 0.2140\n", + "5/8, train_loss: 0.0664 step time: 0.2174\n", + "6/8, train_loss: 0.1038 step time: 0.2182\n" + ] + }, + { + "ename": "KeyboardInterrupt", + "evalue": "", + "output_type": "error", + "traceback": [ + "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", + "\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)", + "Input \u001b[0;32mIn [9]\u001b[0m, in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[1;32m 1\u001b[0m set_determinism(seed\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0\u001b[39m)\n\u001b[1;32m 2\u001b[0m monai_start \u001b[38;5;241m=\u001b[39m time\u001b[38;5;241m.\u001b[39mtime()\n\u001b[1;32m 3\u001b[0m (\n\u001b[1;32m 4\u001b[0m epoch_num,\n\u001b[1;32m 5\u001b[0m m_epoch_loss_values,\n\u001b[1;32m 6\u001b[0m m_metric_values,\n\u001b[1;32m 7\u001b[0m m_epoch_times,\n\u001b[1;32m 8\u001b[0m m_best,\n\u001b[1;32m 9\u001b[0m m_train_time,\n\u001b[0;32m---> 10\u001b[0m ) \u001b[38;5;241m=\u001b[39m \u001b[43mtrain_process\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfast\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m)\u001b[49m\n\u001b[1;32m 11\u001b[0m m_total_time \u001b[38;5;241m=\u001b[39m time\u001b[38;5;241m.\u001b[39mtime() \u001b[38;5;241m-\u001b[39m monai_start\n\u001b[1;32m 12\u001b[0m \u001b[38;5;28mprint\u001b[39m(\n\u001b[1;32m 13\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtotal time of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mepoch_num\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m epochs with MONAI fast training: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mm_train_time\u001b[38;5;132;01m:\u001b[39;00m\u001b[38;5;124m.4f\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m,\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 14\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m time of preparing cache: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m(m_total_time \u001b[38;5;241m-\u001b[39m m_train_time)\u001b[38;5;132;01m:\u001b[39;00m\u001b[38;5;124m.4f\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 15\u001b[0m )\n", + "Input \u001b[0;32mIn [8]\u001b[0m, in \u001b[0;36mtrain_process\u001b[0;34m(fast)\u001b[0m\n\u001b[1;32m 134\u001b[0m \u001b[38;5;66;03m# profiling: backward\u001b[39;00m\n\u001b[1;32m 135\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m nvtx\u001b[38;5;241m.\u001b[39mannotate(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mbackward\u001b[39m\u001b[38;5;124m\"\u001b[39m, color\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblue\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m profiling \u001b[38;5;28;01melse\u001b[39;00m no_profiling:\n\u001b[0;32m--> 136\u001b[0m \u001b[43mscaler\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscale\u001b[49m\u001b[43m(\u001b[49m\u001b[43mloss\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 138\u001b[0m \u001b[38;5;66;03m# profiling: update\u001b[39;00m\n\u001b[1;32m 139\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m nvtx\u001b[38;5;241m.\u001b[39mannotate(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mupdate\u001b[39m\u001b[38;5;124m\"\u001b[39m, color\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124myellow\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m profiling \u001b[38;5;28;01melse\u001b[39;00m no_profiling:\n", + "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/_tensor.py:396\u001b[0m, in \u001b[0;36mTensor.backward\u001b[0;34m(self, gradient, retain_graph, create_graph, inputs)\u001b[0m\n\u001b[1;32m 387\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m has_torch_function_unary(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m 388\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m handle_torch_function(\n\u001b[1;32m 389\u001b[0m Tensor\u001b[38;5;241m.\u001b[39mbackward,\n\u001b[1;32m 390\u001b[0m (\u001b[38;5;28mself\u001b[39m,),\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 394\u001b[0m create_graph\u001b[38;5;241m=\u001b[39mcreate_graph,\n\u001b[1;32m 395\u001b[0m inputs\u001b[38;5;241m=\u001b[39minputs)\n\u001b[0;32m--> 396\u001b[0m \u001b[43mtorch\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mautograd\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgradient\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43minputs\u001b[49m\u001b[43m)\u001b[49m\n", + "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/autograd/__init__.py:173\u001b[0m, in \u001b[0;36mbackward\u001b[0;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)\u001b[0m\n\u001b[1;32m 168\u001b[0m retain_graph \u001b[38;5;241m=\u001b[39m create_graph\n\u001b[1;32m 170\u001b[0m \u001b[38;5;66;03m# The reason we repeat same the comment below is that\u001b[39;00m\n\u001b[1;32m 171\u001b[0m \u001b[38;5;66;03m# some Python versions print out the first line of a multi-line function\u001b[39;00m\n\u001b[1;32m 172\u001b[0m \u001b[38;5;66;03m# calls in the traceback and some print out the last line\u001b[39;00m\n\u001b[0;32m--> 173\u001b[0m \u001b[43mVariable\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_execution_engine\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun_backward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;66;43;03m# Calls into the C++ engine to run the backward pass\u001b[39;49;00m\n\u001b[1;32m 174\u001b[0m \u001b[43m \u001b[49m\u001b[43mtensors\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgrad_tensors_\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 175\u001b[0m \u001b[43m \u001b[49m\u001b[43mallow_unreachable\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maccumulate_grad\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m)\u001b[49m\n", + "\u001b[0;31mKeyboardInterrupt\u001b[0m: " + ] + } + ], "source": [ "set_determinism(seed=0)\n", "monai_start = time.time()\n", From cc3089088a82e9e6b3fdccb6c5cd41302626cec1 Mon Sep 17 00:00:00 2001 From: Nic Ma Date: Wed, 20 Jul 2022 20:22:50 +0800 Subject: [PATCH 2/5] [DLMED] update doc Signed-off-by: Nic Ma --- acceleration/fast_training_tutorial.ipynb | 3007 +++++++++++++++++++-- 1 file changed, 2801 insertions(+), 206 deletions(-) diff --git a/acceleration/fast_training_tutorial.ipynb b/acceleration/fast_training_tutorial.ipynb index 04743d5777..9f79a6c58f 100644 --- a/acceleration/fast_training_tutorial.ipynb +++ b/acceleration/fast_training_tutorial.ipynb @@ -16,9 +16,10 @@ "1. AMP (Auto mixed precision).\n", "2. CacheDataset for deterministic transforms.\n", "3. Move data to GPU and cache, then execute random transforms on GPU.\n", - "4. multi-threads `ThreadDataLoader` is faster than PyTorch DataLoader in light-weight task.\n", - "5. Use MONAI `DiceCE` loss instead of regular `Dice` loss.\n", - "6. Analyzed training curve and tuned algorithm: Use `SGD` optimizer, different network parameters, etc.\n", + "4. Disable meta tracking in the random transforms to avoid unnecessary computation.\n", + "5. multi-threads `ThreadDataLoader` is faster than PyTorch DataLoader in light-weight task.\n", + "6. Use MONAI `DiceCE` loss instead of regular `Dice` loss.\n", + "7. Analyzed training curve and tuned algorithm: Use `SGD` optimizer, different network parameters, etc.\n", "\n", "With a V100 GPU and the target validation `mean dice = 0.94` of the `forground` channel only, it's more than `100x` speedup compared with the Pytorch regular implementation when achieving the same metric. And every epoch is `20x` faster than regular training.\n", "\n", @@ -71,55 +72,13 @@ }, { "cell_type": "code", - "execution_count": 1, + "execution_count": null, "metadata": { "vscode": { "languageId": "python" } }, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "/opt/conda/lib/python3.8/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", - " from .autonotebook import tqdm as notebook_tqdm\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "MONAI version: 0+unknown\n", - "Numpy version: 1.22.4\n", - "Pytorch version: 1.13.0a0+340c412\n", - "MONAI flags: HAS_EXT = False, USE_COMPILED = False, USE_META_DICT = False\n", - "MONAI rev id: None\n", - "MONAI __file__: /workspace/data/medical/MONAI/monai/__init__.py\n", - "\n", - "Optional dependencies:\n", - "Pytorch Ignite version: 0.4.9\n", - "Nibabel version: 4.0.1\n", - "scikit-image version: 0.19.3\n", - "Pillow version: 9.0.1\n", - "Tensorboard version: 2.9.1\n", - "gdown version: 4.5.1\n", - "TorchVision version: 0.13.0a0\n", - "tqdm version: 4.64.0\n", - "lmdb version: 1.3.0\n", - "psutil version: 5.9.1\n", - "pandas version: 1.3.5\n", - "einops version: 0.4.1\n", - "transformers version: 4.20.1\n", - "mlflow version: 1.27.0\n", - "pynrrd version: 0.4.3\n", - "\n", - "For details about installing the optional dependencies, please visit:\n", - " https://docs.monai.io/en/latest/installation.html#installing-the-recommended-dependencies\n", - "\n" - ] - } - ], + "outputs": [], "source": [ "# Copyright 2020 MONAI Consortium\n", "# Licensed under the Apache License, Version 2.0 (the \"License\");\n", @@ -418,11 +377,10 @@ " ]\n", "\n", " if fast:\n", - " # move the data to GPU and cache to avoid CPU -> GPU sync in every epoch\n", - " train_transforms.extend([\n", - " range_func(\"EnsureType\", EnsureTyped(keys=[\"image\", \"label\"], track_meta=False)),\n", - " range_func(\"ToDevice\", ToDeviced(keys=[\"image\", \"label\"], device=device)),\n", - " ])\n", + " # convert the data to Tensor without meta, move to GPU and cache to avoid CPU -> GPU sync in every epoch\n", + " train_transforms.append(\n", + " range_func(\"EnsureType\", EnsureTyped(keys=[\"image\", \"label\"], device=device, track_meta=False))\n", + " )\n", "\n", " train_transforms.append(\n", " # randomly crop out patch samples from big\n", @@ -461,11 +419,10 @@ " CropForegroundd(keys=[\"image\", \"label\"], source_key=\"image\"),\n", " ]\n", " if fast:\n", - " # move the data to GPU and cache to avoid CPU -> GPU sync in every epoch\n", - " val_transforms.extend([\n", - " EnsureTyped(keys=[\"image\", \"label\"], track_meta=False),\n", - " ToDeviced(keys=[\"image\", \"label\"], device=device),\n", - " ])\n", + " # convert the data to Tensor without meta, move to GPU and cache to avoid CPU -> GPU sync in every epoch\n", + " val_transforms.append(\n", + " EnsureTyped(keys=[\"image\", \"label\"], device=device, track_meta=False)\n", + " )\n", "\n", " return Compose(train_transforms), Compose(val_transforms)" ] @@ -481,9 +438,10 @@ "1. `AMP` (auto mixed precision): AMP is an important feature released in PyTorch v1.6, NVIDIA CUDA 11 added strong support for AMP and significantly improved training speed.\n", "2. `CacheDataset`: Dataset with the cache mechanism that can load data and cache deterministic transforms' result during training.\n", "3. `ToDeviced` transform: to move data to GPU and cache with `CacheDataset`, then execute random transforms on GPU directly, avoid CPU -> GPU sync in every epoch. Please note that not all the MONAI transforms support GPU operation so far, still working in progress.\n", - "4. `ThreadDataLoader`: uses multi-threads instead of multi-processing, faster than `DataLoader` in light-weight task as we already cached the results of most computation.\n", - "5. `DiceCE` loss function: computes Dice loss and Cross Entropy Loss, returns the weighted sum of these two losses.\n", - "6. Analyzed the training curve and tuned algorithm: Use `SGD` optimizer, different network parameters, etc.\n", + "4. `set_track_meta(False)`: to disable meta tracking in the random transforms to avoid unnecessary computation.\n", + "5. `ThreadDataLoader`: uses multi-threads instead of multi-processing, faster than `DataLoader` in light-weight task as we already cached the results of most computation.\n", + "6. `DiceCE` loss function: computes Dice loss and Cross Entropy Loss, returns the weighted sum of these two losses.\n", + "7. Analyzed the training curve and tuned algorithm: Use `SGD` optimizer, different network parameters, etc.\n", "\n", "(A note on code: to improve readability and support the profiling flag, we used the `with nvtx(...) if profiling else no_profiling` context pattern, where `no_profiling` is a null context from Python's native `contextlib` with no effect on the code. An acknowledgement is provided here[1](#fn1).)" ] @@ -554,6 +512,7 @@ " bias=True,\n", " dimensions=None,\n", " ).to(device)\n", + " # avoid the computation of meta information in random transforms\n", " set_track_meta(False)\n", " else:\n", " train_ds = Dataset(data=train_files, transform=train_trans)\n", @@ -786,7 +745,7 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": null, "metadata": { "vscode": { "languageId": "python" @@ -797,8 +756,8 @@ "name": "stderr", "output_type": "stream", "text": [ - "Loading dataset: 100%|█████████████████████████████████████████████████████████████████████| 32/32 [00:42<00:00, 1.33s/it]\n", - "Loading dataset: 100%|███████████████████████████████████████████████████████████████████████| 9/9 [00:09<00:00, 1.03s/it]\n" + "Loading dataset: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 32/32 [00:42<00:00, 1.33s/it]\n", + "Loading dataset: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:09<00:00, 1.05s/it]\n" ] }, { @@ -807,197 +766,2833 @@ "text": [ "----------\n", "epoch 1/600\n", - "1/8, train_loss: 0.8296 step time: 1.7046\n", - "2/8, train_loss: 0.7641 step time: 0.2115\n", - "3/8, train_loss: 0.5814 step time: 0.2111\n", - "4/8, train_loss: 0.4991 step time: 0.2139\n", - "5/8, train_loss: 0.4853 step time: 0.2131\n", - "6/8, train_loss: 0.4718 step time: 0.2126\n", - "7/8, train_loss: 0.4114 step time: 0.1999\n", - "8/8, train_loss: 0.4430 step time: 0.1998\n", + "1/8, train_loss: 0.8296 step time: 1.6296\n", + "2/8, train_loss: 0.7641 step time: 0.2118\n", + "3/8, train_loss: 0.5814 step time: 0.2088\n", + "4/8, train_loss: 0.4991 step time: 0.2105\n", + "5/8, train_loss: 0.4853 step time: 0.2091\n", + "6/8, train_loss: 0.4718 step time: 0.2094\n", + "7/8, train_loss: 0.4114 step time: 0.1994\n", + "8/8, train_loss: 0.4430 step time: 0.2013\n", "epoch 1 average loss: 0.5607\n", - "time consuming of epoch 1 is: 3.1678\n", + "time consuming of epoch 1 is: 3.0811\n", "----------\n", "epoch 2/600\n", - "1/8, train_loss: 0.4435 step time: 0.2696\n", - "2/8, train_loss: 0.4586 step time: 0.2147\n", - "3/8, train_loss: 0.4187 step time: 0.2169\n", - "4/8, train_loss: 0.5169 step time: 0.2079\n", - "5/8, train_loss: 0.5513 step time: 0.2138\n", - "6/8, train_loss: 0.4491 step time: 0.2146\n", - "7/8, train_loss: 0.5062 step time: 0.2003\n", - "8/8, train_loss: 0.5096 step time: 0.2000\n", + "1/8, train_loss: 0.4435 step time: 0.2669\n", + "2/8, train_loss: 0.4586 step time: 0.2144\n", + "3/8, train_loss: 0.4187 step time: 0.2161\n", + "4/8, train_loss: 0.5169 step time: 0.2075\n", + "5/8, train_loss: 0.5513 step time: 0.2170\n", + "6/8, train_loss: 0.4491 step time: 0.2170\n", + "7/8, train_loss: 0.5062 step time: 0.2004\n", + "8/8, train_loss: 0.5096 step time: 0.2005\n", "epoch 2 average loss: 0.4817\n", - "time consuming of epoch 2 is: 1.7395\n", + "time consuming of epoch 2 is: 1.7416\n", "----------\n", "epoch 3/600\n", - "1/8, train_loss: 0.4188 step time: 0.2705\n", - "2/8, train_loss: 0.4919 step time: 0.2170\n", - "3/8, train_loss: 0.5508 step time: 0.2171\n", - "4/8, train_loss: 0.4027 step time: 0.2164\n", - "5/8, train_loss: 0.4910 step time: 0.2196\n", - "6/8, train_loss: 0.3804 step time: 0.2190\n", - "7/8, train_loss: 0.3431 step time: 0.1993\n", - "8/8, train_loss: 0.3834 step time: 0.2004\n", + "1/8, train_loss: 0.4188 step time: 0.2660\n", + "2/8, train_loss: 0.4919 step time: 0.2132\n", + "3/8, train_loss: 0.5508 step time: 0.2119\n", + "4/8, train_loss: 0.4027 step time: 0.2136\n", + "5/8, train_loss: 0.4910 step time: 0.2153\n", + "6/8, train_loss: 0.3804 step time: 0.2130\n", + "7/8, train_loss: 0.3431 step time: 0.2019\n", + "8/8, train_loss: 0.3834 step time: 0.2000\n", "epoch 3 average loss: 0.4328\n", - "time consuming of epoch 3 is: 1.7609\n", + "time consuming of epoch 3 is: 1.7364\n", "----------\n", "epoch 4/600\n", - "1/8, train_loss: 0.2439 step time: 0.2722\n", - "2/8, train_loss: 0.2532 step time: 0.2148\n", - "3/8, train_loss: 0.2174 step time: 0.2144\n", - "4/8, train_loss: 0.2359 step time: 0.2145\n", - "5/8, train_loss: 0.1941 step time: 0.2153\n", - "6/8, train_loss: 0.2294 step time: 0.2164\n", - "7/8, train_loss: 0.2436 step time: 0.1986\n", - "8/8, train_loss: 0.3432 step time: 0.1989\n", + "1/8, train_loss: 0.2439 step time: 0.2637\n", + "2/8, train_loss: 0.2532 step time: 0.2147\n", + "3/8, train_loss: 0.2174 step time: 0.2192\n", + "4/8, train_loss: 0.2359 step time: 0.2136\n", + "5/8, train_loss: 0.1941 step time: 0.2161\n", + "6/8, train_loss: 0.2294 step time: 0.2175\n", + "7/8, train_loss: 0.2436 step time: 0.2000\n", + "8/8, train_loss: 0.3432 step time: 0.2017\n", "epoch 4 average loss: 0.2451\n", - "time consuming of epoch 4 is: 1.7466\n", + "time consuming of epoch 4 is: 1.7480\n", "----------\n", "epoch 5/600\n", - "1/8, train_loss: 0.1843 step time: 0.2684\n", - "2/8, train_loss: 0.2163 step time: 0.2153\n", - "3/8, train_loss: 0.1557 step time: 0.2170\n", - "4/8, train_loss: 0.1434 step time: 0.2166\n", - "5/8, train_loss: 0.1158 step time: 0.2157\n", - "6/8, train_loss: 0.2312 step time: 0.2171\n", - "7/8, train_loss: 0.2958 step time: 0.1981\n", - "8/8, train_loss: 0.1528 step time: 0.1999\n", + "1/8, train_loss: 0.1843 step time: 0.2634\n", + "2/8, train_loss: 0.2163 step time: 0.2137\n", + "3/8, train_loss: 0.1557 step time: 0.2131\n", + "4/8, train_loss: 0.1434 step time: 0.2127\n", + "5/8, train_loss: 0.1158 step time: 0.2117\n", + "6/8, train_loss: 0.2312 step time: 0.2108\n", + "7/8, train_loss: 0.2958 step time: 0.2000\n", + "8/8, train_loss: 0.1528 step time: 0.2006\n", "epoch 5 average loss: 0.1869\n", "saved new best metric model\n", "current epoch: 5 current mean dice: 0.0626 best mean dice: 0.0626 at epoch: 5\n", - "time consuming of epoch 5 is: 2.8436\n", + "time consuming of epoch 5 is: 2.8338\n", "----------\n", "epoch 6/600\n", - "1/8, train_loss: 0.1136 step time: 0.2639\n", - "2/8, train_loss: 0.1565 step time: 0.2111\n", - "3/8, train_loss: 0.2769 step time: 0.2117\n", - "4/8, train_loss: 0.1773 step time: 0.2094\n", - "5/8, train_loss: 0.2183 step time: 0.2107\n", - "6/8, train_loss: 0.1236 step time: 0.2133\n", - "7/8, train_loss: 0.2068 step time: 0.1992\n", - "8/8, train_loss: 0.1687 step time: 0.1977\n", + "1/8, train_loss: 0.1136 step time: 0.2713\n", + "2/8, train_loss: 0.1565 step time: 0.2132\n", + "3/8, train_loss: 0.2769 step time: 0.2103\n", + "4/8, train_loss: 0.1773 step time: 0.2101\n", + "5/8, train_loss: 0.2183 step time: 0.2102\n", + "6/8, train_loss: 0.1236 step time: 0.2118\n", + "7/8, train_loss: 0.2068 step time: 0.2001\n", + "8/8, train_loss: 0.1687 step time: 0.2004\n", "epoch 6 average loss: 0.1802\n", - "time consuming of epoch 6 is: 1.7192\n", + "time consuming of epoch 6 is: 1.7299\n", "----------\n", "epoch 7/600\n", - "1/8, train_loss: 0.1709 step time: 0.2565\n", - "2/8, train_loss: 0.2031 step time: 0.2167\n", - "3/8, train_loss: 0.1453 step time: 0.2137\n", - "4/8, train_loss: 0.2224 step time: 0.2145\n", - "5/8, train_loss: 0.1693 step time: 0.2150\n", - "6/8, train_loss: 0.3097 step time: 0.2107\n", - "7/8, train_loss: 0.3081 step time: 0.1987\n", - "8/8, train_loss: 0.1908 step time: 0.1984\n", + "1/8, train_loss: 0.1709 step time: 0.2542\n", + "2/8, train_loss: 0.2031 step time: 0.2163\n", + "3/8, train_loss: 0.1453 step time: 0.2128\n", + "4/8, train_loss: 0.2224 step time: 0.2095\n", + "5/8, train_loss: 0.1693 step time: 0.2096\n", + "6/8, train_loss: 0.3097 step time: 0.2094\n", + "7/8, train_loss: 0.3081 step time: 0.1962\n", + "8/8, train_loss: 0.1908 step time: 0.1955\n", "epoch 7 average loss: 0.2150\n", - "time consuming of epoch 7 is: 1.7256\n", + "time consuming of epoch 7 is: 1.7049\n", "----------\n", "epoch 8/600\n", - "1/8, train_loss: 0.1996 step time: 0.2717\n", - "2/8, train_loss: 0.2088 step time: 0.2170\n", - "3/8, train_loss: 0.1894 step time: 0.2191\n", - "4/8, train_loss: 0.2062 step time: 0.2193\n", - "5/8, train_loss: 0.1687 step time: 0.2193\n", - "6/8, train_loss: 0.1327 step time: 0.2152\n", - "7/8, train_loss: 0.1556 step time: 0.2005\n", - "8/8, train_loss: 0.1491 step time: 0.2002\n", + "1/8, train_loss: 0.1996 step time: 0.2565\n", + "2/8, train_loss: 0.2088 step time: 0.2068\n", + "3/8, train_loss: 0.1894 step time: 0.2094\n", + "4/8, train_loss: 0.2062 step time: 0.2084\n", + "5/8, train_loss: 0.1687 step time: 0.2070\n", + "6/8, train_loss: 0.1327 step time: 0.2076\n", + "7/8, train_loss: 0.1556 step time: 0.1961\n", + "8/8, train_loss: 0.1491 step time: 0.1957\n", "epoch 8 average loss: 0.1762\n", - "time consuming of epoch 8 is: 1.7639\n", + "time consuming of epoch 8 is: 1.6886\n", "----------\n", "epoch 9/600\n", - "1/8, train_loss: 0.1722 step time: 0.2708\n", - "2/8, train_loss: 0.0934 step time: 0.2165\n", - "3/8, train_loss: 0.1434 step time: 0.2189\n", - "4/8, train_loss: 0.0838 step time: 0.2178\n", - "5/8, train_loss: 0.1419 step time: 0.2159\n", - "6/8, train_loss: 0.1540 step time: 0.2136\n", - "7/8, train_loss: 0.1397 step time: 0.1987\n", - "8/8, train_loss: 0.1324 step time: 0.1980\n", + "1/8, train_loss: 0.1722 step time: 0.2545\n", + "2/8, train_loss: 0.0934 step time: 0.2082\n", + "3/8, train_loss: 0.1434 step time: 0.2081\n", + "4/8, train_loss: 0.0838 step time: 0.2071\n", + "5/8, train_loss: 0.1419 step time: 0.2077\n", + "6/8, train_loss: 0.1540 step time: 0.2072\n", + "7/8, train_loss: 0.1397 step time: 0.1963\n", + "8/8, train_loss: 0.1324 step time: 0.1954\n", "epoch 9 average loss: 0.1326\n", - "time consuming of epoch 9 is: 1.7519\n", + "time consuming of epoch 9 is: 1.6858\n", "----------\n", "epoch 10/600\n", - "1/8, train_loss: 0.1358 step time: 0.2682\n", - "2/8, train_loss: 0.0968 step time: 0.2160\n", - "3/8, train_loss: 0.1124 step time: 0.2181\n", - "4/8, train_loss: 0.1009 step time: 0.2163\n", - "5/8, train_loss: 0.1119 step time: 0.2168\n", - "6/8, train_loss: 0.1132 step time: 0.2163\n", - "7/8, train_loss: 0.0996 step time: 0.1994\n", - "8/8, train_loss: 0.1189 step time: 0.1997\n", + "1/8, train_loss: 0.1358 step time: 0.2547\n", + "2/8, train_loss: 0.0968 step time: 0.2084\n", + "3/8, train_loss: 0.1124 step time: 0.2084\n", + "4/8, train_loss: 0.1009 step time: 0.2071\n", + "5/8, train_loss: 0.1119 step time: 0.2074\n", + "6/8, train_loss: 0.1132 step time: 0.2074\n", + "7/8, train_loss: 0.0996 step time: 0.1964\n", + "8/8, train_loss: 0.1189 step time: 0.1948\n", "epoch 10 average loss: 0.1112\n", "current epoch: 10 current mean dice: 0.0000 best mean dice: 0.0626 at epoch: 5\n", - "time consuming of epoch 10 is: 2.5712\n", + "time consuming of epoch 10 is: 2.4952\n", "----------\n", "epoch 11/600\n", - "1/8, train_loss: 0.0958 step time: 0.2679\n", - "2/8, train_loss: 0.0749 step time: 0.2134\n", - "3/8, train_loss: 0.1171 step time: 0.2122\n", - "4/8, train_loss: 0.1097 step time: 0.2140\n", - "5/8, train_loss: 0.1265 step time: 0.2134\n", - "6/8, train_loss: 0.1216 step time: 0.2140\n", - "7/8, train_loss: 0.1101 step time: 0.1985\n", - "8/8, train_loss: 0.1316 step time: 0.1983\n", + "1/8, train_loss: 0.0958 step time: 0.2555\n", + "2/8, train_loss: 0.0749 step time: 0.2080\n", + "3/8, train_loss: 0.1171 step time: 0.2074\n", + "4/8, train_loss: 0.1097 step time: 0.2070\n", + "5/8, train_loss: 0.1265 step time: 0.2080\n", + "6/8, train_loss: 0.1216 step time: 0.2077\n", + "7/8, train_loss: 0.1101 step time: 0.1962\n", + "8/8, train_loss: 0.1316 step time: 0.1954\n", "epoch 11 average loss: 0.1109\n", - "time consuming of epoch 11 is: 1.7330\n", + "time consuming of epoch 11 is: 1.6864\n", "----------\n", "epoch 12/600\n", - "1/8, train_loss: 0.1670 step time: 0.2706\n", - "2/8, train_loss: 0.0785 step time: 0.2156\n", - "3/8, train_loss: 0.1389 step time: 0.2166\n", - "4/8, train_loss: 0.1350 step time: 0.2182\n", - "5/8, train_loss: 0.1447 step time: 0.2166\n", - "6/8, train_loss: 0.1352 step time: 0.2146\n", - "7/8, train_loss: 0.0995 step time: 0.1989\n", - "8/8, train_loss: 0.0594 step time: 0.1991\n", + "1/8, train_loss: 0.1670 step time: 0.2563\n", + "2/8, train_loss: 0.0785 step time: 0.2070\n", + "3/8, train_loss: 0.1389 step time: 0.2078\n", + "4/8, train_loss: 0.1350 step time: 0.2070\n", + "5/8, train_loss: 0.1447 step time: 0.2077\n", + "6/8, train_loss: 0.1352 step time: 0.2084\n", + "7/8, train_loss: 0.0995 step time: 0.1960\n", + "8/8, train_loss: 0.0594 step time: 0.1954\n", "epoch 12 average loss: 0.1198\n", - "time consuming of epoch 12 is: 1.7517\n", + "time consuming of epoch 12 is: 1.6868\n", "----------\n", "epoch 13/600\n", - "1/8, train_loss: 0.0674 step time: 0.2714\n", - "2/8, train_loss: 0.1410 step time: 0.2171\n", - "3/8, train_loss: 0.1068 step time: 0.2182\n", - "4/8, train_loss: 0.1073 step time: 0.2165\n", - "5/8, train_loss: 0.0862 step time: 0.2147\n", - "6/8, train_loss: 0.0855 step time: 0.2153\n", - "7/8, train_loss: 0.0986 step time: 0.1986\n", - "8/8, train_loss: 0.2790 step time: 0.1977\n", + "1/8, train_loss: 0.0674 step time: 0.2465\n", + "2/8, train_loss: 0.1410 step time: 0.2086\n", + "3/8, train_loss: 0.1068 step time: 0.2081\n", + "4/8, train_loss: 0.1073 step time: 0.2073\n", + "5/8, train_loss: 0.0862 step time: 0.2081\n", + "6/8, train_loss: 0.0855 step time: 0.2074\n", + "7/8, train_loss: 0.0986 step time: 0.1966\n", + "8/8, train_loss: 0.2790 step time: 0.1955\n", "epoch 13 average loss: 0.1215\n", - "time consuming of epoch 13 is: 1.7513\n", + "time consuming of epoch 13 is: 1.6793\n", "----------\n", "epoch 14/600\n", - "1/8, train_loss: 0.0853 step time: 0.2713\n", - "2/8, train_loss: 0.0896 step time: 0.2175\n", - "3/8, train_loss: 0.1138 step time: 0.2190\n", - "4/8, train_loss: 0.0790 step time: 0.2178\n", - "5/8, train_loss: 0.0676 step time: 0.2144\n", - "6/8, train_loss: 0.1342 step time: 0.2163\n", - "7/8, train_loss: 0.1085 step time: 0.1994\n", - "8/8, train_loss: 0.0938 step time: 0.1996\n", + "1/8, train_loss: 0.0853 step time: 0.2546\n", + "2/8, train_loss: 0.0896 step time: 0.2092\n", + "3/8, train_loss: 0.1138 step time: 0.2074\n", + "4/8, train_loss: 0.0790 step time: 0.2069\n", + "5/8, train_loss: 0.0676 step time: 0.2093\n", + "6/8, train_loss: 0.1342 step time: 0.2084\n", + "7/8, train_loss: 0.1085 step time: 0.1961\n", + "8/8, train_loss: 0.0938 step time: 0.1957\n", "epoch 14 average loss: 0.0965\n", - "time consuming of epoch 14 is: 1.7572\n", + "time consuming of epoch 14 is: 1.6888\n", "----------\n", "epoch 15/600\n", - "1/8, train_loss: 0.0492 step time: 0.2705\n", - "2/8, train_loss: 0.0648 step time: 0.2182\n", - "3/8, train_loss: 0.0709 step time: 0.2169\n", - "4/8, train_loss: 0.0982 step time: 0.2140\n", - "5/8, train_loss: 0.0664 step time: 0.2174\n", - "6/8, train_loss: 0.1038 step time: 0.2182\n" + "1/8, train_loss: 0.0492 step time: 0.2534\n", + "2/8, train_loss: 0.0648 step time: 0.2077\n", + "3/8, train_loss: 0.0709 step time: 0.2081\n", + "4/8, train_loss: 0.0982 step time: 0.2076\n", + "5/8, train_loss: 0.0664 step time: 0.2070\n", + "6/8, train_loss: 0.1038 step time: 0.2073\n", + "7/8, train_loss: 0.1088 step time: 0.1967\n", + "8/8, train_loss: 0.0679 step time: 0.1957\n", + "epoch 15 average loss: 0.0788\n", + "saved new best metric model\n", + "current epoch: 15 current mean dice: 0.6743 best mean dice: 0.6743 at epoch: 15\n", + "time consuming of epoch 15 is: 2.7746\n", + "----------\n", + "epoch 16/600\n", + "1/8, train_loss: 0.1002 step time: 0.2685\n", + "2/8, train_loss: 0.0568 step time: 0.2078\n", + "3/8, train_loss: 0.0559 step time: 0.2077\n", + "4/8, train_loss: 0.0857 step time: 0.2094\n", + "5/8, train_loss: 0.0572 step time: 0.2089\n", + "6/8, train_loss: 0.0645 step time: 0.2072\n", + "7/8, train_loss: 0.0604 step time: 0.1962\n", + "8/8, train_loss: 0.0804 step time: 0.1956\n", + "epoch 16 average loss: 0.0701\n", + "time consuming of epoch 16 is: 1.7039\n", + "----------\n", + "epoch 17/600\n", + "1/8, train_loss: 0.0465 step time: 0.2479\n", + "2/8, train_loss: 0.1140 step time: 0.2075\n", + "3/8, train_loss: 0.2151 step time: 0.2083\n", + "4/8, train_loss: 0.0604 step time: 0.2086\n", + "5/8, train_loss: 0.0690 step time: 0.2093\n", + "6/8, train_loss: 0.0898 step time: 0.2075\n", + "7/8, train_loss: 0.0799 step time: 0.1961\n", + "8/8, train_loss: 0.1355 step time: 0.1950\n", + "epoch 17 average loss: 0.1013\n", + "time consuming of epoch 17 is: 1.6814\n", + "----------\n", + "epoch 18/600\n", + "1/8, train_loss: 0.0773 step time: 0.2523\n", + "2/8, train_loss: 0.0589 step time: 0.2069\n", + "3/8, train_loss: 0.0740 step time: 0.2085\n", + "4/8, train_loss: 0.0662 step time: 0.2082\n", + "5/8, train_loss: 0.1036 step time: 0.2079\n", + "6/8, train_loss: 0.0778 step time: 0.2071\n", + "7/8, train_loss: 0.1039 step time: 0.1962\n", + "8/8, train_loss: 0.0743 step time: 0.1956\n", + "epoch 18 average loss: 0.0795\n", + "time consuming of epoch 18 is: 1.6840\n", + "----------\n", + "epoch 19/600\n", + "1/8, train_loss: 0.1254 step time: 0.2494\n", + "2/8, train_loss: 0.0945 step time: 0.2074\n", + "3/8, train_loss: 0.0875 step time: 0.2076\n", + "4/8, train_loss: 0.0549 step time: 0.2085\n", + "5/8, train_loss: 0.0530 step time: 0.2076\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "6/8, train_loss: 0.0795 step time: 0.2082\n", + "7/8, train_loss: 0.0487 step time: 0.1961\n", + "8/8, train_loss: 0.0933 step time: 0.1946\n", + "epoch 19 average loss: 0.0796\n", + "time consuming of epoch 19 is: 1.6806\n", + "----------\n", + "epoch 20/600\n", + "1/8, train_loss: 0.0522 step time: 0.2536\n", + "2/8, train_loss: 0.0380 step time: 0.2069\n", + "3/8, train_loss: 0.0756 step time: 0.2076\n", + "4/8, train_loss: 0.0545 step time: 0.2078\n", + "5/8, train_loss: 0.0803 step time: 0.2085\n", + "6/8, train_loss: 0.0568 step time: 0.2081\n", + "7/8, train_loss: 0.0509 step time: 0.1962\n", + "8/8, train_loss: 0.0532 step time: 0.1956\n", + "epoch 20 average loss: 0.0577\n", + "saved new best metric model\n", + "current epoch: 20 current mean dice: 0.8040 best mean dice: 0.8040 at epoch: 20\n", + "time consuming of epoch 20 is: 2.7756\n", + "----------\n", + "epoch 21/600\n", + "1/8, train_loss: 0.1026 step time: 0.2645\n", + "2/8, train_loss: 0.0634 step time: 0.2071\n", + "3/8, train_loss: 0.0936 step time: 0.2077\n", + "4/8, train_loss: 0.0665 step time: 0.2069\n", + "5/8, train_loss: 0.0788 step time: 0.2093\n", + "6/8, train_loss: 0.0623 step time: 0.2091\n", + "7/8, train_loss: 0.0451 step time: 0.1961\n", + "8/8, train_loss: 0.0487 step time: 0.1953\n", + "epoch 21 average loss: 0.0701\n", + "time consuming of epoch 21 is: 1.6982\n", + "----------\n", + "epoch 22/600\n", + "1/8, train_loss: 0.0657 step time: 0.2550\n", + "2/8, train_loss: 0.0808 step time: 0.2090\n", + "3/8, train_loss: 0.0591 step time: 0.2091\n", + "4/8, train_loss: 0.0695 step time: 0.2076\n", + "5/8, train_loss: 0.0728 step time: 0.2071\n", + "6/8, train_loss: 0.0493 step time: 0.2070\n", + "7/8, train_loss: 0.0728 step time: 0.1961\n", + "8/8, train_loss: 0.0518 step time: 0.1954\n", + "epoch 22 average loss: 0.0652\n", + "time consuming of epoch 22 is: 1.6874\n", + "----------\n", + "epoch 23/600\n", + "1/8, train_loss: 0.0551 step time: 0.2557\n", + "2/8, train_loss: 0.0412 step time: 0.2078\n", + "3/8, train_loss: 0.1448 step time: 0.2073\n", + "4/8, train_loss: 0.0512 step time: 0.2072\n", + "5/8, train_loss: 0.0528 step time: 0.2074\n", + "6/8, train_loss: 0.0461 step time: 0.2083\n", + "7/8, train_loss: 0.0509 step time: 0.1951\n", + "8/8, train_loss: 0.0810 step time: 0.1956\n", + "epoch 23 average loss: 0.0654\n", + "time consuming of epoch 23 is: 1.6856\n", + "----------\n", + "epoch 24/600\n", + "1/8, train_loss: 0.0402 step time: 0.2584\n", + "2/8, train_loss: 0.0642 step time: 0.2071\n", + "3/8, train_loss: 0.0470 step time: 0.2073\n", + "4/8, train_loss: 0.0444 step time: 0.2075\n", + "5/8, train_loss: 0.0337 step time: 0.2068\n", + "6/8, train_loss: 0.0400 step time: 0.2069\n", + "7/8, train_loss: 0.0680 step time: 0.1965\n", + "8/8, train_loss: 0.0397 step time: 0.1948\n", + "epoch 24 average loss: 0.0471\n", + "time consuming of epoch 24 is: 1.6863\n", + "----------\n", + "epoch 25/600\n", + "1/8, train_loss: 0.0394 step time: 0.2559\n", + "2/8, train_loss: 0.0383 step time: 0.2071\n", + "3/8, train_loss: 0.0451 step time: 0.2075\n", + "4/8, train_loss: 0.0898 step time: 0.2069\n", + "5/8, train_loss: 0.0248 step time: 0.2073\n", + "6/8, train_loss: 0.0701 step time: 0.2074\n", + "7/8, train_loss: 0.0732 step time: 0.1955\n", + "8/8, train_loss: 0.0477 step time: 0.1956\n", + "epoch 25 average loss: 0.0536\n", + "saved new best metric model\n", + "current epoch: 25 current mean dice: 0.8969 best mean dice: 0.8969 at epoch: 25\n", + "time consuming of epoch 25 is: 2.7585\n", + "----------\n", + "epoch 26/600\n", + "1/8, train_loss: 0.0439 step time: 0.2651\n", + "2/8, train_loss: 0.0592 step time: 0.2083\n", + "3/8, train_loss: 0.0554 step time: 0.2070\n", + "4/8, train_loss: 0.0395 step time: 0.2078\n", + "5/8, train_loss: 0.0399 step time: 0.2075\n", + "6/8, train_loss: 0.0490 step time: 0.2103\n", + "7/8, train_loss: 0.0506 step time: 0.1965\n", + "8/8, train_loss: 0.0554 step time: 0.1956\n", + "epoch 26 average loss: 0.0491\n", + "time consuming of epoch 26 is: 1.7006\n", + "----------\n", + "epoch 27/600\n", + "1/8, train_loss: 0.0537 step time: 0.2538\n", + "2/8, train_loss: 0.1149 step time: 0.2075\n", + "3/8, train_loss: 0.0783 step time: 0.2071\n", + "4/8, train_loss: 0.0425 step time: 0.2069\n", + "5/8, train_loss: 0.0416 step time: 0.2094\n", + "6/8, train_loss: 0.0262 step time: 0.2069\n", + "7/8, train_loss: 0.0727 step time: 0.1963\n", + "8/8, train_loss: 0.0590 step time: 0.1955\n", + "epoch 27 average loss: 0.0611\n", + "time consuming of epoch 27 is: 1.6845\n", + "----------\n", + "epoch 28/600\n", + "1/8, train_loss: 0.0345 step time: 0.2569\n", + "2/8, train_loss: 0.0424 step time: 0.2083\n", + "3/8, train_loss: 0.0444 step time: 0.2083\n", + "4/8, train_loss: 0.0409 step time: 0.2071\n", + "5/8, train_loss: 0.0490 step time: 0.2081\n", + "6/8, train_loss: 0.0333 step time: 0.2083\n", + "7/8, train_loss: 0.0774 step time: 0.1968\n", + "8/8, train_loss: 0.0903 step time: 0.1956\n", + "epoch 28 average loss: 0.0515\n", + "time consuming of epoch 28 is: 1.6906\n", + "----------\n", + "epoch 29/600\n", + "1/8, train_loss: 0.0437 step time: 0.2563\n", + "2/8, train_loss: 0.0311 step time: 0.2085\n", + "3/8, train_loss: 0.0665 step time: 0.2088\n", + "4/8, train_loss: 0.0528 step time: 0.2080\n", + "5/8, train_loss: 0.0491 step time: 0.2069\n", + "6/8, train_loss: 0.0761 step time: 0.2091\n", + "7/8, train_loss: 0.0543 step time: 0.1958\n", + "8/8, train_loss: 0.0518 step time: 0.1956\n", + "epoch 29 average loss: 0.0532\n", + "time consuming of epoch 29 is: 1.6901\n", + "----------\n", + "epoch 30/600\n", + "1/8, train_loss: 0.0395 step time: 0.2567\n", + "2/8, train_loss: 0.0454 step time: 0.2082\n", + "3/8, train_loss: 0.0517 step time: 0.2072\n", + "4/8, train_loss: 0.0901 step time: 0.2079\n", + "5/8, train_loss: 0.0522 step time: 0.2079\n", + "6/8, train_loss: 0.0425 step time: 0.2095\n", + "7/8, train_loss: 0.0439 step time: 0.1958\n", + "8/8, train_loss: 0.0371 step time: 0.1955\n", + "epoch 30 average loss: 0.0503\n", + "current epoch: 30 current mean dice: 0.8890 best mean dice: 0.8969 at epoch: 25\n", + "time consuming of epoch 30 is: 2.4989\n", + "----------\n", + "epoch 31/600\n", + "1/8, train_loss: 0.0338 step time: 0.2536\n", + "2/8, train_loss: 0.0410 step time: 0.2070\n", + "3/8, train_loss: 0.0410 step time: 0.2071\n", + "4/8, train_loss: 0.0382 step time: 0.2077\n", + "5/8, train_loss: 0.0504 step time: 0.2082\n", + "6/8, train_loss: 0.0302 step time: 0.2079\n", + "7/8, train_loss: 0.0510 step time: 0.1963\n", + "8/8, train_loss: 0.0406 step time: 0.1955\n", + "epoch 31 average loss: 0.0408\n", + "time consuming of epoch 31 is: 1.6845\n", + "----------\n", + "epoch 32/600\n", + "1/8, train_loss: 0.0327 step time: 0.2470\n", + "2/8, train_loss: 0.0291 step time: 0.2068\n", + "3/8, train_loss: 0.1370 step time: 0.2069\n", + "4/8, train_loss: 0.0312 step time: 0.2073\n", + "5/8, train_loss: 0.0306 step time: 0.2081\n", + "6/8, train_loss: 0.0694 step time: 0.2079\n", + "7/8, train_loss: 0.0372 step time: 0.1965\n", + "8/8, train_loss: 0.0602 step time: 0.1956\n", + "epoch 32 average loss: 0.0534\n", + "time consuming of epoch 32 is: 1.6772\n", + "----------\n", + "epoch 33/600\n", + "1/8, train_loss: 0.0540 step time: 0.2468\n", + "2/8, train_loss: 0.1219 step time: 0.2080\n", + "3/8, train_loss: 0.0577 step time: 0.2091\n", + "4/8, train_loss: 0.0325 step time: 0.2073\n", + "5/8, train_loss: 0.0371 step time: 0.2088\n", + "6/8, train_loss: 0.0549 step time: 0.2078\n", + "7/8, train_loss: 0.0983 step time: 0.1960\n", + "8/8, train_loss: 0.0661 step time: 0.1947\n", + "epoch 33 average loss: 0.0653\n", + "time consuming of epoch 33 is: 1.6797\n", + "----------\n", + "epoch 34/600\n", + "1/8, train_loss: 0.0739 step time: 0.2551\n", + "2/8, train_loss: 0.0329 step time: 0.2077\n", + "3/8, train_loss: 0.0439 step time: 0.2073\n", + "4/8, train_loss: 0.0749 step time: 0.2068\n", + "5/8, train_loss: 0.0679 step time: 0.2073\n", + "6/8, train_loss: 0.0517 step time: 0.2075\n", + "7/8, train_loss: 0.0597 step time: 0.1961\n", + "8/8, train_loss: 0.0958 step time: 0.1943\n", + "epoch 34 average loss: 0.0626\n", + "time consuming of epoch 34 is: 1.6832\n", + "----------\n", + "epoch 35/600\n", + "1/8, train_loss: 0.0601 step time: 0.2554\n", + "2/8, train_loss: 0.0365 step time: 0.2076\n", + "3/8, train_loss: 0.0801 step time: 0.2073\n", + "4/8, train_loss: 0.0347 step time: 0.2090\n", + "5/8, train_loss: 0.0524 step time: 0.2083\n", + "6/8, train_loss: 0.0358 step time: 0.2071\n", + "7/8, train_loss: 0.1216 step time: 0.1961\n", + "8/8, train_loss: 0.0970 step time: 0.1956\n", + "epoch 35 average loss: 0.0648\n", + "current epoch: 35 current mean dice: 0.8884 best mean dice: 0.8969 at epoch: 25\n", + "time consuming of epoch 35 is: 2.4965\n", + "----------\n", + "epoch 36/600\n", + "1/8, train_loss: 0.0784 step time: 0.2549\n", + "2/8, train_loss: 0.0462 step time: 0.2097\n", + "3/8, train_loss: 0.0405 step time: 0.2075\n", + "4/8, train_loss: 0.0292 step time: 0.2071\n", + "5/8, train_loss: 0.0605 step time: 0.2092\n", + "6/8, train_loss: 0.0441 step time: 0.2088\n", + "7/8, train_loss: 0.0521 step time: 0.1961\n", + "8/8, train_loss: 0.0539 step time: 0.1954\n", + "epoch 36 average loss: 0.0506\n", + "time consuming of epoch 36 is: 1.6897\n", + "----------\n", + "epoch 37/600\n", + "1/8, train_loss: 0.0474 step time: 0.2541\n", + "2/8, train_loss: 0.0376 step time: 0.2079\n", + "3/8, train_loss: 0.0295 step time: 0.2084\n", + "4/8, train_loss: 0.0289 step time: 0.2081\n", + "5/8, train_loss: 0.0343 step time: 0.2100\n", + "6/8, train_loss: 0.0374 step time: 0.2082\n", + "7/8, train_loss: 0.0438 step time: 0.1961\n", + "8/8, train_loss: 0.0394 step time: 0.1956\n", + "epoch 37 average loss: 0.0373\n", + "time consuming of epoch 37 is: 1.6893\n", + "----------\n", + "epoch 38/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0459 step time: 0.2534\n", + "2/8, train_loss: 0.0381 step time: 0.2073\n", + "3/8, train_loss: 0.0447 step time: 0.2075\n", + "4/8, train_loss: 0.0310 step time: 0.2079\n", + "5/8, train_loss: 0.0437 step time: 0.2174\n", + "6/8, train_loss: 0.0325 step time: 0.2153\n", + "7/8, train_loss: 0.0828 step time: 0.1994\n", + "8/8, train_loss: 0.0459 step time: 0.2006\n", + "epoch 38 average loss: 0.0456\n", + "time consuming of epoch 38 is: 1.7099\n", + "----------\n", + "epoch 39/600\n", + "1/8, train_loss: 0.0330 step time: 0.2644\n", + "2/8, train_loss: 0.0325 step time: 0.2139\n", + "3/8, train_loss: 0.1914 step time: 0.2139\n", + "4/8, train_loss: 0.0379 step time: 0.2179\n", + "5/8, train_loss: 0.0535 step time: 0.2160\n", + "6/8, train_loss: 0.0579 step time: 0.2132\n", + "7/8, train_loss: 0.0368 step time: 0.1992\n", + "8/8, train_loss: 0.0444 step time: 0.1988\n", + "epoch 39 average loss: 0.0609\n", + "time consuming of epoch 39 is: 1.7388\n", + "----------\n", + "epoch 40/600\n", + "1/8, train_loss: 0.0358 step time: 0.2653\n", + "2/8, train_loss: 0.0395 step time: 0.2163\n", + "3/8, train_loss: 0.0406 step time: 0.2156\n", + "4/8, train_loss: 0.0566 step time: 0.2150\n", + "5/8, train_loss: 0.0485 step time: 0.2136\n", + "6/8, train_loss: 0.0495 step time: 0.2155\n", + "7/8, train_loss: 0.0652 step time: 0.2003\n", + "8/8, train_loss: 0.0606 step time: 0.1990\n", + "epoch 40 average loss: 0.0495\n", + "current epoch: 40 current mean dice: 0.8737 best mean dice: 0.8969 at epoch: 25\n", + "time consuming of epoch 40 is: 2.5544\n", + "----------\n", + "epoch 41/600\n", + "1/8, train_loss: 0.0453 step time: 0.2615\n", + "2/8, train_loss: 0.0459 step time: 0.2130\n", + "3/8, train_loss: 0.0305 step time: 0.2125\n", + "4/8, train_loss: 0.0292 step time: 0.2113\n", + "5/8, train_loss: 0.0505 step time: 0.2114\n", + "6/8, train_loss: 0.1944 step time: 0.2138\n", + "7/8, train_loss: 0.0352 step time: 0.1994\n", + "8/8, train_loss: 0.0756 step time: 0.2002\n", + "epoch 41 average loss: 0.0633\n", + "time consuming of epoch 41 is: 1.7243\n", + "----------\n", + "epoch 42/600\n", + "1/8, train_loss: 0.0470 step time: 0.2634\n", + "2/8, train_loss: 0.0408 step time: 0.2150\n", + "3/8, train_loss: 0.0403 step time: 0.2150\n", + "4/8, train_loss: 0.0490 step time: 0.2133\n", + "5/8, train_loss: 0.0520 step time: 0.2117\n", + "6/8, train_loss: 0.0816 step time: 0.2123\n", + "7/8, train_loss: 0.0631 step time: 0.1988\n", + "8/8, train_loss: 0.0490 step time: 0.2003\n", + "epoch 42 average loss: 0.0529\n", + "time consuming of epoch 42 is: 1.7313\n", + "----------\n", + "epoch 43/600\n", + "1/8, train_loss: 0.0699 step time: 0.2516\n", + "2/8, train_loss: 0.0403 step time: 0.2117\n", + "3/8, train_loss: 0.0285 step time: 0.2111\n", + "4/8, train_loss: 0.0575 step time: 0.2104\n", + "5/8, train_loss: 0.0381 step time: 0.2106\n", + "6/8, train_loss: 0.0826 step time: 0.2148\n", + "7/8, train_loss: 0.0470 step time: 0.1998\n", + "8/8, train_loss: 0.0569 step time: 0.1999\n", + "epoch 43 average loss: 0.0526\n", + "time consuming of epoch 43 is: 1.7114\n", + "----------\n", + "epoch 44/600\n", + "1/8, train_loss: 0.0431 step time: 0.2555\n", + "2/8, train_loss: 0.0348 step time: 0.2139\n", + "3/8, train_loss: 0.0449 step time: 0.2146\n", + "4/8, train_loss: 0.0440 step time: 0.2175\n", + "5/8, train_loss: 0.0260 step time: 0.2131\n", + "6/8, train_loss: 0.0405 step time: 0.2142\n", + "7/8, train_loss: 0.0707 step time: 0.1990\n", + "8/8, train_loss: 0.0481 step time: 0.1990\n", + "epoch 44 average loss: 0.0440\n", + "time consuming of epoch 44 is: 1.7280\n", + "----------\n", + "epoch 45/600\n", + "1/8, train_loss: 0.0600 step time: 0.2597\n", + "2/8, train_loss: 0.0377 step time: 0.2154\n", + "3/8, train_loss: 0.0344 step time: 0.2147\n", + "4/8, train_loss: 0.0456 step time: 0.2144\n", + "5/8, train_loss: 0.0550 step time: 0.2152\n", + "6/8, train_loss: 0.0331 step time: 0.2147\n", + "7/8, train_loss: 0.0502 step time: 0.1995\n", + "8/8, train_loss: 0.0285 step time: 0.2007\n", + "epoch 45 average loss: 0.0431\n", + "current epoch: 45 current mean dice: 0.7649 best mean dice: 0.8969 at epoch: 25\n", + "time consuming of epoch 45 is: 2.5463\n", + "----------\n", + "epoch 46/600\n", + "1/8, train_loss: 0.0279 step time: 0.2610\n", + "2/8, train_loss: 0.0401 step time: 0.2126\n", + "3/8, train_loss: 0.0624 step time: 0.2137\n", + "4/8, train_loss: 0.0394 step time: 0.2139\n", + "5/8, train_loss: 0.0491 step time: 0.2124\n", + "6/8, train_loss: 0.0262 step time: 0.2168\n", + "7/8, train_loss: 0.0287 step time: 0.2000\n", + "8/8, train_loss: 0.0410 step time: 0.1991\n", + "epoch 46 average loss: 0.0394\n", + "time consuming of epoch 46 is: 1.7306\n", + "----------\n", + "epoch 47/600\n", + "1/8, train_loss: 0.0326 step time: 0.2660\n", + "2/8, train_loss: 0.0317 step time: 0.2143\n", + "3/8, train_loss: 0.0365 step time: 0.2157\n", + "4/8, train_loss: 0.0346 step time: 0.2152\n", + "5/8, train_loss: 0.0522 step time: 0.2160\n", + "6/8, train_loss: 0.0299 step time: 0.2157\n", + "7/8, train_loss: 0.0280 step time: 0.1992\n", + "8/8, train_loss: 0.0358 step time: 0.2000\n", + "epoch 47 average loss: 0.0352\n", + "time consuming of epoch 47 is: 1.7437\n", + "----------\n", + "epoch 48/600\n", + "1/8, train_loss: 0.0249 step time: 0.2627\n", + "2/8, train_loss: 0.0335 step time: 0.2143\n", + "3/8, train_loss: 0.0271 step time: 0.2151\n", + "4/8, train_loss: 0.0370 step time: 0.2155\n", + "5/8, train_loss: 0.0523 step time: 0.2152\n", + "6/8, train_loss: 0.0408 step time: 0.2155\n", + "7/8, train_loss: 0.0279 step time: 0.1995\n", + "8/8, train_loss: 0.0394 step time: 0.2000\n", + "epoch 48 average loss: 0.0354\n", + "time consuming of epoch 48 is: 1.7393\n", + "----------\n", + "epoch 49/600\n", + "1/8, train_loss: 0.0296 step time: 0.2613\n", + "2/8, train_loss: 0.0325 step time: 0.2138\n", + "3/8, train_loss: 0.0291 step time: 0.2156\n", + "4/8, train_loss: 0.0398 step time: 0.2145\n", + "5/8, train_loss: 0.0238 step time: 0.2163\n", + "6/8, train_loss: 0.0266 step time: 0.2149\n", + "7/8, train_loss: 0.0307 step time: 0.1994\n", + "8/8, train_loss: 0.0251 step time: 0.1996\n", + "epoch 49 average loss: 0.0297\n", + "time consuming of epoch 49 is: 1.7368\n", + "----------\n", + "epoch 50/600\n", + "1/8, train_loss: 0.0264 step time: 0.2658\n", + "2/8, train_loss: 0.0286 step time: 0.2118\n", + "3/8, train_loss: 0.0388 step time: 0.2131\n", + "4/8, train_loss: 0.0331 step time: 0.2138\n", + "5/8, train_loss: 0.0249 step time: 0.2128\n", + "6/8, train_loss: 0.0205 step time: 0.2092\n", + "7/8, train_loss: 0.0279 step time: 0.1995\n", + "8/8, train_loss: 0.0259 step time: 0.1998\n", + "epoch 50 average loss: 0.0283\n", + "saved new best metric model\n", + "current epoch: 50 current mean dice: 0.9166 best mean dice: 0.9166 at epoch: 50\n", + "time consuming of epoch 50 is: 2.7850\n", + "----------\n", + "epoch 51/600\n", + "1/8, train_loss: 0.0233 step time: 0.2631\n", + "2/8, train_loss: 0.0252 step time: 0.2129\n", + "3/8, train_loss: 0.0534 step time: 0.2108\n", + "4/8, train_loss: 0.0390 step time: 0.2091\n", + "5/8, train_loss: 0.0315 step time: 0.2099\n", + "6/8, train_loss: 0.0293 step time: 0.2118\n", + "7/8, train_loss: 0.0325 step time: 0.1992\n", + "8/8, train_loss: 0.0299 step time: 0.1991\n", + "epoch 51 average loss: 0.0330\n", + "time consuming of epoch 51 is: 1.7183\n", + "----------\n", + "epoch 52/600\n", + "1/8, train_loss: 0.0258 step time: 0.2639\n", + "2/8, train_loss: 0.0286 step time: 0.2146\n", + "3/8, train_loss: 0.0314 step time: 0.2149\n", + "4/8, train_loss: 0.0234 step time: 0.2148\n", + "5/8, train_loss: 0.0277 step time: 0.2153\n", + "6/8, train_loss: 0.0254 step time: 0.2144\n", + "7/8, train_loss: 0.0215 step time: 0.2002\n", + "8/8, train_loss: 0.0252 step time: 0.1999\n", + "epoch 52 average loss: 0.0261\n", + "time consuming of epoch 52 is: 1.7396\n", + "----------\n", + "epoch 53/600\n", + "1/8, train_loss: 0.0266 step time: 0.2635\n", + "2/8, train_loss: 0.0226 step time: 0.2135\n", + "3/8, train_loss: 0.0399 step time: 0.2167\n", + "4/8, train_loss: 0.0304 step time: 0.2143\n", + "5/8, train_loss: 0.0394 step time: 0.2148\n", + "6/8, train_loss: 0.0293 step time: 0.2162\n", + "7/8, train_loss: 0.0522 step time: 0.2001\n", + "8/8, train_loss: 0.0267 step time: 0.1995\n", + "epoch 53 average loss: 0.0334\n", + "time consuming of epoch 53 is: 1.7400\n", + "----------\n", + "epoch 54/600\n", + "1/8, train_loss: 0.0273 step time: 0.2665\n", + "2/8, train_loss: 0.0352 step time: 0.2130\n", + "3/8, train_loss: 0.0361 step time: 0.2162\n", + "4/8, train_loss: 0.0304 step time: 0.2162\n", + "5/8, train_loss: 0.0294 step time: 0.2153\n", + "6/8, train_loss: 0.0244 step time: 0.2143\n", + "7/8, train_loss: 0.0247 step time: 0.1995\n", + "8/8, train_loss: 0.0346 step time: 0.2002\n", + "epoch 54 average loss: 0.0303\n", + "time consuming of epoch 54 is: 1.7427\n", + "----------\n", + "epoch 55/600\n", + "1/8, train_loss: 0.0260 step time: 0.2631\n", + "2/8, train_loss: 0.0319 step time: 0.2139\n", + "3/8, train_loss: 0.0277 step time: 0.2184\n", + "4/8, train_loss: 0.0241 step time: 0.2138\n", + "5/8, train_loss: 0.0284 step time: 0.2145\n", + "6/8, train_loss: 0.0248 step time: 0.2152\n", + "7/8, train_loss: 0.0329 step time: 0.2002\n", + "8/8, train_loss: 0.0232 step time: 0.2015\n", + "epoch 55 average loss: 0.0274\n", + "current epoch: 55 current mean dice: 0.8907 best mean dice: 0.9166 at epoch: 50\n", + "time consuming of epoch 55 is: 2.5562\n", + "----------\n", + "epoch 56/600\n", + "1/8, train_loss: 0.0286 step time: 0.2587\n", + "2/8, train_loss: 0.0349 step time: 0.2125\n", + "3/8, train_loss: 0.0282 step time: 0.2091\n", + "4/8, train_loss: 0.0329 step time: 0.2093\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "5/8, train_loss: 0.0265 step time: 0.2106\n", + "6/8, train_loss: 0.0286 step time: 0.2101\n", + "7/8, train_loss: 0.0286 step time: 0.1974\n", + "8/8, train_loss: 0.0239 step time: 0.1990\n", + "epoch 56 average loss: 0.0290\n", + "time consuming of epoch 56 is: 1.7080\n", + "----------\n", + "epoch 57/600\n", + "1/8, train_loss: 0.0247 step time: 0.2537\n", + "2/8, train_loss: 0.0245 step time: 0.2130\n", + "3/8, train_loss: 0.0266 step time: 0.2115\n", + "4/8, train_loss: 0.0267 step time: 0.2122\n", + "5/8, train_loss: 0.0326 step time: 0.2122\n", + "6/8, train_loss: 0.0305 step time: 0.2122\n", + "7/8, train_loss: 0.0281 step time: 0.2015\n", + "8/8, train_loss: 0.0291 step time: 0.1998\n", + "epoch 57 average loss: 0.0278\n", + "time consuming of epoch 57 is: 1.7178\n", + "----------\n", + "epoch 58/600\n", + "1/8, train_loss: 0.0268 step time: 0.2641\n", + "2/8, train_loss: 0.0225 step time: 0.2155\n", + "3/8, train_loss: 0.0350 step time: 0.2116\n", + "4/8, train_loss: 0.0246 step time: 0.2119\n", + "5/8, train_loss: 0.0280 step time: 0.2137\n", + "6/8, train_loss: 0.0243 step time: 0.2139\n", + "7/8, train_loss: 0.0209 step time: 0.1998\n", + "8/8, train_loss: 0.0230 step time: 0.1997\n", + "epoch 58 average loss: 0.0256\n", + "time consuming of epoch 58 is: 1.7315\n", + "----------\n", + "epoch 59/600\n", + "1/8, train_loss: 0.0223 step time: 0.2535\n", + "2/8, train_loss: 0.0225 step time: 0.2142\n", + "3/8, train_loss: 0.0250 step time: 0.2131\n", + "4/8, train_loss: 0.0336 step time: 0.2130\n", + "5/8, train_loss: 0.0186 step time: 0.2136\n", + "6/8, train_loss: 0.0263 step time: 0.2143\n", + "7/8, train_loss: 0.0198 step time: 0.2004\n", + "8/8, train_loss: 0.0233 step time: 0.2000\n", + "epoch 59 average loss: 0.0239\n", + "time consuming of epoch 59 is: 1.7236\n", + "----------\n", + "epoch 60/600\n", + "1/8, train_loss: 0.0260 step time: 0.2645\n", + "2/8, train_loss: 0.0206 step time: 0.2167\n", + "3/8, train_loss: 0.0321 step time: 0.2155\n", + "4/8, train_loss: 0.0288 step time: 0.2156\n", + "5/8, train_loss: 0.0210 step time: 0.2147\n", + "6/8, train_loss: 0.0257 step time: 0.2144\n", + "7/8, train_loss: 0.0254 step time: 0.1994\n", + "8/8, train_loss: 0.0263 step time: 0.1994\n", + "epoch 60 average loss: 0.0257\n", + "saved new best metric model\n", + "current epoch: 60 current mean dice: 0.9340 best mean dice: 0.9340 at epoch: 60\n", + "time consuming of epoch 60 is: 2.8237\n", + "----------\n", + "epoch 61/600\n", + "1/8, train_loss: 0.0231 step time: 0.2700\n", + "2/8, train_loss: 0.0236 step time: 0.2117\n", + "3/8, train_loss: 0.0218 step time: 0.2103\n", + "4/8, train_loss: 0.0253 step time: 0.2102\n", + "5/8, train_loss: 0.0215 step time: 0.2099\n", + "6/8, train_loss: 0.0256 step time: 0.2100\n", + "7/8, train_loss: 0.0253 step time: 0.2000\n", + "8/8, train_loss: 0.0198 step time: 0.2001\n", + "epoch 61 average loss: 0.0233\n", + "time consuming of epoch 61 is: 1.7247\n", + "----------\n", + "epoch 62/600\n", + "1/8, train_loss: 0.0309 step time: 0.2551\n", + "2/8, train_loss: 0.0211 step time: 0.2152\n", + "3/8, train_loss: 0.0231 step time: 0.2132\n", + "4/8, train_loss: 0.0230 step time: 0.2111\n", + "5/8, train_loss: 0.0279 step time: 0.2107\n", + "6/8, train_loss: 0.0284 step time: 0.2136\n", + "7/8, train_loss: 0.0229 step time: 0.1996\n", + "8/8, train_loss: 0.0249 step time: 0.1996\n", + "epoch 62 average loss: 0.0253\n", + "time consuming of epoch 62 is: 1.7195\n", + "----------\n", + "epoch 63/600\n", + "1/8, train_loss: 0.0260 step time: 0.2641\n", + "2/8, train_loss: 0.0300 step time: 0.2141\n", + "3/8, train_loss: 0.0219 step time: 0.2145\n", + "4/8, train_loss: 0.0306 step time: 0.2156\n", + "5/8, train_loss: 0.0251 step time: 0.2157\n", + "6/8, train_loss: 0.0228 step time: 0.2145\n", + "7/8, train_loss: 0.0319 step time: 0.2003\n", + "8/8, train_loss: 0.0199 step time: 0.2000\n", + "epoch 63 average loss: 0.0260\n", + "time consuming of epoch 63 is: 1.7401\n", + "----------\n", + "epoch 64/600\n", + "1/8, train_loss: 0.0230 step time: 0.2620\n", + "2/8, train_loss: 0.0186 step time: 0.2145\n", + "3/8, train_loss: 0.0237 step time: 0.2137\n", + "4/8, train_loss: 0.0239 step time: 0.2137\n", + "5/8, train_loss: 0.0239 step time: 0.2150\n", + "6/8, train_loss: 0.0222 step time: 0.2168\n", + "7/8, train_loss: 0.0272 step time: 0.1988\n", + "8/8, train_loss: 0.0274 step time: 0.2002\n", + "epoch 64 average loss: 0.0237\n", + "time consuming of epoch 64 is: 1.7362\n", + "----------\n", + "epoch 65/600\n", + "1/8, train_loss: 0.0212 step time: 0.2633\n", + "2/8, train_loss: 0.0229 step time: 0.2160\n", + "3/8, train_loss: 0.0218 step time: 0.2163\n", + "4/8, train_loss: 0.0261 step time: 0.2144\n", + "5/8, train_loss: 0.0350 step time: 0.2135\n", + "6/8, train_loss: 0.0251 step time: 0.2160\n", + "7/8, train_loss: 0.0207 step time: 0.1995\n", + "8/8, train_loss: 0.0186 step time: 0.1999\n", + "epoch 65 average loss: 0.0239\n", + "saved new best metric model\n", + "current epoch: 65 current mean dice: 0.9441 best mean dice: 0.9441 at epoch: 65\n", + "time consuming of epoch 65 is: 2.8164\n", + "----------\n", + "epoch 66/600\n", + "1/8, train_loss: 0.0229 step time: 0.2710\n", + "2/8, train_loss: 0.0252 step time: 0.2139\n", + "3/8, train_loss: 0.0238 step time: 0.2132\n", + "4/8, train_loss: 0.0202 step time: 0.2150\n", + "5/8, train_loss: 0.0270 step time: 0.2163\n", + "6/8, train_loss: 0.0259 step time: 0.2157\n", + "7/8, train_loss: 0.0254 step time: 0.2001\n", + "8/8, train_loss: 0.0247 step time: 0.2000\n", + "epoch 66 average loss: 0.0244\n", + "time consuming of epoch 66 is: 1.7476\n", + "----------\n", + "epoch 67/600\n", + "1/8, train_loss: 0.0306 step time: 0.2618\n", + "2/8, train_loss: 0.0243 step time: 0.2172\n", + "3/8, train_loss: 0.0256 step time: 0.2141\n", + "4/8, train_loss: 0.0245 step time: 0.2139\n", + "5/8, train_loss: 0.0223 step time: 0.2141\n", + "6/8, train_loss: 0.0202 step time: 0.2179\n", + "7/8, train_loss: 0.0213 step time: 0.1997\n", + "8/8, train_loss: 0.0214 step time: 0.2006\n", + "epoch 67 average loss: 0.0238\n", + "time consuming of epoch 67 is: 1.7406\n", + "----------\n", + "epoch 68/600\n", + "1/8, train_loss: 0.0344 step time: 0.2626\n", + "2/8, train_loss: 0.0276 step time: 0.2143\n", + "3/8, train_loss: 0.0228 step time: 0.2165\n", + "4/8, train_loss: 0.0265 step time: 0.2150\n", + "5/8, train_loss: 0.0231 step time: 0.2144\n", + "6/8, train_loss: 0.0216 step time: 0.2150\n", + "7/8, train_loss: 0.0171 step time: 0.2006\n", + "8/8, train_loss: 0.0206 step time: 0.1996\n", + "epoch 68 average loss: 0.0242\n", + "time consuming of epoch 68 is: 1.7395\n", + "----------\n", + "epoch 69/600\n", + "1/8, train_loss: 0.0241 step time: 0.2634\n", + "2/8, train_loss: 0.0240 step time: 0.2152\n", + "3/8, train_loss: 0.0196 step time: 0.2108\n", + "4/8, train_loss: 0.0308 step time: 0.2142\n", + "5/8, train_loss: 0.0244 step time: 0.2136\n", + "6/8, train_loss: 0.0219 step time: 0.2118\n", + "7/8, train_loss: 0.0187 step time: 0.2003\n", + "8/8, train_loss: 0.0226 step time: 0.2003\n", + "epoch 69 average loss: 0.0233\n", + "time consuming of epoch 69 is: 1.7313\n", + "----------\n", + "epoch 70/600\n", + "1/8, train_loss: 0.0220 step time: 0.2519\n", + "2/8, train_loss: 0.0245 step time: 0.2118\n", + "3/8, train_loss: 0.0237 step time: 0.2124\n", + "4/8, train_loss: 0.0215 step time: 0.2133\n", + "5/8, train_loss: 0.0228 step time: 0.2136\n", + "6/8, train_loss: 0.0180 step time: 0.2129\n", + "7/8, train_loss: 0.0273 step time: 0.1992\n", + "8/8, train_loss: 0.0235 step time: 0.2002\n", + "epoch 70 average loss: 0.0229\n", + "current epoch: 70 current mean dice: 0.9428 best mean dice: 0.9441 at epoch: 65\n", + "time consuming of epoch 70 is: 2.5292\n", + "----------\n", + "epoch 71/600\n", + "1/8, train_loss: 0.0241 step time: 0.2597\n", + "2/8, train_loss: 0.0210 step time: 0.2105\n", + "3/8, train_loss: 0.0268 step time: 0.2129\n", + "4/8, train_loss: 0.0257 step time: 0.2114\n", + "5/8, train_loss: 0.0234 step time: 0.2124\n", + "6/8, train_loss: 0.0260 step time: 0.2175\n", + "7/8, train_loss: 0.0198 step time: 0.1998\n", + "8/8, train_loss: 0.0336 step time: 0.2009\n", + "epoch 71 average loss: 0.0251\n", + "time consuming of epoch 71 is: 1.7262\n", + "----------\n", + "epoch 72/600\n", + "1/8, train_loss: 0.0223 step time: 0.2650\n", + "2/8, train_loss: 0.0294 step time: 0.2163\n", + "3/8, train_loss: 0.0219 step time: 0.2150\n", + "4/8, train_loss: 0.0195 step time: 0.2133\n", + "5/8, train_loss: 0.0244 step time: 0.2151\n", + "6/8, train_loss: 0.0209 step time: 0.2150\n", + "7/8, train_loss: 0.0294 step time: 0.1997\n", + "8/8, train_loss: 0.0273 step time: 0.2005\n", + "epoch 72 average loss: 0.0244\n", + "time consuming of epoch 72 is: 1.7413\n", + "----------\n", + "epoch 73/600\n", + "1/8, train_loss: 0.0266 step time: 0.2641\n", + "2/8, train_loss: 0.0209 step time: 0.2138\n", + "3/8, train_loss: 0.0210 step time: 0.2138\n", + "4/8, train_loss: 0.0233 step time: 0.2130\n", + "5/8, train_loss: 0.0212 step time: 0.2125\n", + "6/8, train_loss: 0.0252 step time: 0.2139\n", + "7/8, train_loss: 0.0219 step time: 0.1994\n", + "8/8, train_loss: 0.0212 step time: 0.2005\n", + "epoch 73 average loss: 0.0227\n", + "time consuming of epoch 73 is: 1.7325\n", + "----------\n", + "epoch 74/600\n", + "1/8, train_loss: 0.0249 step time: 0.2539\n", + "2/8, train_loss: 0.0219 step time: 0.2138\n", + "3/8, train_loss: 0.0246 step time: 0.2109\n", + "4/8, train_loss: 0.0348 step time: 0.2120\n", + "5/8, train_loss: 0.0210 step time: 0.2121\n", + "6/8, train_loss: 0.0284 step time: 0.2113\n", + "7/8, train_loss: 0.0210 step time: 0.1997\n", + "8/8, train_loss: 0.0221 step time: 0.2006\n", + "epoch 74 average loss: 0.0249\n", + "time consuming of epoch 74 is: 1.7158\n", + "----------\n", + "epoch 75/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0263 step time: 0.2637\n", + "2/8, train_loss: 0.0225 step time: 0.2122\n", + "3/8, train_loss: 0.0216 step time: 0.2104\n", + "4/8, train_loss: 0.0209 step time: 0.2126\n", + "5/8, train_loss: 0.0223 step time: 0.2172\n", + "6/8, train_loss: 0.0213 step time: 0.2157\n", + "7/8, train_loss: 0.0204 step time: 0.1993\n", + "8/8, train_loss: 0.0226 step time: 0.1998\n", + "epoch 75 average loss: 0.0222\n", + "current epoch: 75 current mean dice: 0.9405 best mean dice: 0.9441 at epoch: 65\n", + "time consuming of epoch 75 is: 2.5434\n", + "----------\n", + "epoch 76/600\n", + "1/8, train_loss: 0.0244 step time: 0.2627\n", + "2/8, train_loss: 0.0240 step time: 0.2120\n", + "3/8, train_loss: 0.0239 step time: 0.2123\n", + "4/8, train_loss: 0.0245 step time: 0.2118\n", + "5/8, train_loss: 0.0350 step time: 0.2132\n", + "6/8, train_loss: 0.0216 step time: 0.2146\n", + "7/8, train_loss: 0.0302 step time: 0.1996\n", + "8/8, train_loss: 0.0239 step time: 0.2002\n", + "epoch 76 average loss: 0.0259\n", + "time consuming of epoch 76 is: 1.7276\n", + "----------\n", + "epoch 77/600\n", + "1/8, train_loss: 0.0213 step time: 0.2631\n", + "2/8, train_loss: 0.0239 step time: 0.2147\n", + "3/8, train_loss: 0.0262 step time: 0.2152\n", + "4/8, train_loss: 0.0238 step time: 0.2149\n", + "5/8, train_loss: 0.0223 step time: 0.2146\n", + "6/8, train_loss: 0.0209 step time: 0.2172\n", + "7/8, train_loss: 0.0242 step time: 0.1995\n", + "8/8, train_loss: 0.0226 step time: 0.2000\n", + "epoch 77 average loss: 0.0231\n", + "time consuming of epoch 77 is: 1.7408\n", + "----------\n", + "epoch 78/600\n", + "1/8, train_loss: 0.0219 step time: 0.2660\n", + "2/8, train_loss: 0.0231 step time: 0.2157\n", + "3/8, train_loss: 0.0206 step time: 0.2151\n", + "4/8, train_loss: 0.0336 step time: 0.2141\n", + "5/8, train_loss: 0.0209 step time: 0.2141\n", + "6/8, train_loss: 0.0341 step time: 0.2139\n", + "7/8, train_loss: 0.0215 step time: 0.1995\n", + "8/8, train_loss: 0.0205 step time: 0.2000\n", + "epoch 78 average loss: 0.0245\n", + "time consuming of epoch 78 is: 1.7397\n", + "----------\n", + "epoch 79/600\n", + "1/8, train_loss: 0.0223 step time: 0.2637\n", + "2/8, train_loss: 0.0211 step time: 0.2143\n", + "3/8, train_loss: 0.0186 step time: 0.2136\n", + "4/8, train_loss: 0.0274 step time: 0.2168\n", + "5/8, train_loss: 0.0225 step time: 0.2158\n", + "6/8, train_loss: 0.0278 step time: 0.2155\n", + "7/8, train_loss: 0.0265 step time: 0.1990\n", + "8/8, train_loss: 0.0178 step time: 0.1999\n", + "epoch 79 average loss: 0.0230\n", + "time consuming of epoch 79 is: 1.7401\n", + "----------\n", + "epoch 80/600\n", + "1/8, train_loss: 0.0207 step time: 0.2629\n", + "2/8, train_loss: 0.0219 step time: 0.2165\n", + "3/8, train_loss: 0.0223 step time: 0.2153\n", + "4/8, train_loss: 0.0204 step time: 0.2147\n", + "5/8, train_loss: 0.0242 step time: 0.2150\n", + "6/8, train_loss: 0.0271 step time: 0.2180\n", + "7/8, train_loss: 0.0237 step time: 0.1994\n", + "8/8, train_loss: 0.0229 step time: 0.1998\n", + "epoch 80 average loss: 0.0229\n", + "current epoch: 80 current mean dice: 0.9250 best mean dice: 0.9441 at epoch: 65\n", + "time consuming of epoch 80 is: 2.5564\n", + "----------\n", + "epoch 81/600\n", + "1/8, train_loss: 0.0247 step time: 0.2582\n", + "2/8, train_loss: 0.0294 step time: 0.2130\n", + "3/8, train_loss: 0.0269 step time: 0.2147\n", + "4/8, train_loss: 0.0254 step time: 0.2126\n", + "5/8, train_loss: 0.0182 step time: 0.2139\n", + "6/8, train_loss: 0.0227 step time: 0.2180\n", + "7/8, train_loss: 0.0178 step time: 0.1994\n", + "8/8, train_loss: 0.0295 step time: 0.2014\n", + "epoch 81 average loss: 0.0243\n", + "time consuming of epoch 81 is: 1.7325\n", + "----------\n", + "epoch 82/600\n", + "1/8, train_loss: 0.0277 step time: 0.2627\n", + "2/8, train_loss: 0.0201 step time: 0.2169\n", + "3/8, train_loss: 0.0207 step time: 0.2155\n", + "4/8, train_loss: 0.0211 step time: 0.2182\n", + "5/8, train_loss: 0.0254 step time: 0.2138\n", + "6/8, train_loss: 0.0185 step time: 0.2151\n", + "7/8, train_loss: 0.0179 step time: 0.1993\n", + "8/8, train_loss: 0.0251 step time: 0.1981\n", + "epoch 82 average loss: 0.0221\n", + "time consuming of epoch 82 is: 1.7411\n", + "----------\n", + "epoch 83/600\n", + "1/8, train_loss: 0.0221 step time: 0.2650\n", + "2/8, train_loss: 0.0225 step time: 0.2130\n", + "3/8, train_loss: 0.0208 step time: 0.2152\n", + "4/8, train_loss: 0.0240 step time: 0.2155\n", + "5/8, train_loss: 0.0203 step time: 0.2165\n", + "6/8, train_loss: 0.0196 step time: 0.2176\n", + "7/8, train_loss: 0.0183 step time: 0.1999\n", + "8/8, train_loss: 0.0195 step time: 0.2011\n", + "epoch 83 average loss: 0.0209\n", + "time consuming of epoch 83 is: 1.7456\n", + "----------\n", + "epoch 84/600\n", + "1/8, train_loss: 0.0205 step time: 0.2642\n", + "2/8, train_loss: 0.0234 step time: 0.2147\n", + "3/8, train_loss: 0.0188 step time: 0.2159\n", + "4/8, train_loss: 0.0216 step time: 0.2175\n", + "5/8, train_loss: 0.0210 step time: 0.2155\n", + "6/8, train_loss: 0.0211 step time: 0.2153\n", + "7/8, train_loss: 0.0207 step time: 0.1993\n", + "8/8, train_loss: 0.0242 step time: 0.1996\n", + "epoch 84 average loss: 0.0214\n", + "time consuming of epoch 84 is: 1.7435\n", + "----------\n", + "epoch 85/600\n", + "1/8, train_loss: 0.0187 step time: 0.2659\n", + "2/8, train_loss: 0.0222 step time: 0.2143\n", + "3/8, train_loss: 0.0252 step time: 0.2167\n", + "4/8, train_loss: 0.0215 step time: 0.2142\n", + "5/8, train_loss: 0.0249 step time: 0.2142\n", + "6/8, train_loss: 0.0362 step time: 0.2178\n", + "7/8, train_loss: 0.0182 step time: 0.2005\n", + "8/8, train_loss: 0.0219 step time: 0.2006\n", + "epoch 85 average loss: 0.0236\n", + "saved new best metric model\n", + "current epoch: 85 current mean dice: 0.9482 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 85 is: 2.8376\n", + "----------\n", + "epoch 86/600\n", + "1/8, train_loss: 0.0202 step time: 0.2659\n", + "2/8, train_loss: 0.0165 step time: 0.2085\n", + "3/8, train_loss: 0.0258 step time: 0.2103\n", + "4/8, train_loss: 0.0192 step time: 0.2095\n", + "5/8, train_loss: 0.0221 step time: 0.2104\n", + "6/8, train_loss: 0.0199 step time: 0.2119\n", + "7/8, train_loss: 0.0185 step time: 0.2012\n", + "8/8, train_loss: 0.0274 step time: 0.2001\n", + "epoch 86 average loss: 0.0212\n", + "time consuming of epoch 86 is: 1.7204\n", + "----------\n", + "epoch 87/600\n", + "1/8, train_loss: 0.0320 step time: 0.2659\n", + "2/8, train_loss: 0.0207 step time: 0.2158\n", + "3/8, train_loss: 0.0216 step time: 0.2172\n", + "4/8, train_loss: 0.0200 step time: 0.2153\n", + "5/8, train_loss: 0.0255 step time: 0.2151\n", + "6/8, train_loss: 0.0215 step time: 0.2174\n", + "7/8, train_loss: 0.0228 step time: 0.1996\n", + "8/8, train_loss: 0.0189 step time: 0.1993\n", + "epoch 87 average loss: 0.0229\n", + "time consuming of epoch 87 is: 1.7470\n", + "----------\n", + "epoch 88/600\n", + "1/8, train_loss: 0.0213 step time: 0.2663\n", + "2/8, train_loss: 0.0215 step time: 0.2165\n", + "3/8, train_loss: 0.0184 step time: 0.2113\n", + "4/8, train_loss: 0.0249 step time: 0.2116\n", + "5/8, train_loss: 0.0192 step time: 0.2108\n", + "6/8, train_loss: 0.0292 step time: 0.2094\n", + "7/8, train_loss: 0.0183 step time: 0.1981\n", + "8/8, train_loss: 0.0235 step time: 0.1988\n", + "epoch 88 average loss: 0.0220\n", + "time consuming of epoch 88 is: 1.7243\n", + "----------\n", + "epoch 89/600\n", + "1/8, train_loss: 0.0199 step time: 0.2630\n", + "2/8, train_loss: 0.0237 step time: 0.2133\n", + "3/8, train_loss: 0.0219 step time: 0.2140\n", + "4/8, train_loss: 0.0259 step time: 0.2161\n", + "5/8, train_loss: 0.0306 step time: 0.2153\n", + "6/8, train_loss: 0.0219 step time: 0.2138\n", + "7/8, train_loss: 0.0214 step time: 0.1993\n", + "8/8, train_loss: 0.0251 step time: 0.1987\n", + "epoch 89 average loss: 0.0238\n", + "time consuming of epoch 89 is: 1.7349\n", + "----------\n", + "epoch 90/600\n", + "1/8, train_loss: 0.0178 step time: 0.2633\n", + "2/8, train_loss: 0.0203 step time: 0.2144\n", + "3/8, train_loss: 0.0187 step time: 0.2149\n", + "4/8, train_loss: 0.0280 step time: 0.2154\n", + "5/8, train_loss: 0.0238 step time: 0.2150\n", + "6/8, train_loss: 0.0319 step time: 0.2163\n", + "7/8, train_loss: 0.0220 step time: 0.2003\n", + "8/8, train_loss: 0.0233 step time: 0.2000\n", + "epoch 90 average loss: 0.0232\n", + "current epoch: 90 current mean dice: 0.8999 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 90 is: 2.5539\n", + "----------\n", + "epoch 91/600\n", + "1/8, train_loss: 0.0173 step time: 0.2606\n", + "2/8, train_loss: 0.0312 step time: 0.2131\n", + "3/8, train_loss: 0.0205 step time: 0.2120\n", + "4/8, train_loss: 0.0229 step time: 0.2136\n", + "5/8, train_loss: 0.0366 step time: 0.2129\n", + "6/8, train_loss: 0.0243 step time: 0.2119\n", + "7/8, train_loss: 0.0223 step time: 0.1997\n", + "8/8, train_loss: 0.0689 step time: 0.2008\n", + "epoch 91 average loss: 0.0305\n", + "time consuming of epoch 91 is: 1.7257\n", + "----------\n", + "epoch 92/600\n", + "1/8, train_loss: 0.0223 step time: 0.2624\n", + "2/8, train_loss: 0.0247 step time: 0.2162\n", + "3/8, train_loss: 0.0181 step time: 0.2154\n", + "4/8, train_loss: 0.0256 step time: 0.2150\n", + "5/8, train_loss: 0.0712 step time: 0.2141\n", + "6/8, train_loss: 0.0249 step time: 0.2162\n", + "7/8, train_loss: 0.0207 step time: 0.1999\n", + "8/8, train_loss: 0.0309 step time: 0.1998\n", + "epoch 92 average loss: 0.0298\n", + "time consuming of epoch 92 is: 1.7405\n", + "----------\n", + "epoch 93/600\n", + "1/8, train_loss: 0.0741 step time: 0.2632\n", + "2/8, train_loss: 0.0662 step time: 0.2137\n", + "3/8, train_loss: 0.0379 step time: 0.2155\n", + "4/8, train_loss: 0.0371 step time: 0.2159\n" ] }, { - "ename": "KeyboardInterrupt", - "evalue": "", - "output_type": "error", - "traceback": [ - "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m", - "\u001b[0;31mKeyboardInterrupt\u001b[0m Traceback (most recent call last)", - "Input \u001b[0;32mIn [9]\u001b[0m, in \u001b[0;36m\u001b[0;34m()\u001b[0m\n\u001b[1;32m 1\u001b[0m set_determinism(seed\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m0\u001b[39m)\n\u001b[1;32m 2\u001b[0m monai_start \u001b[38;5;241m=\u001b[39m time\u001b[38;5;241m.\u001b[39mtime()\n\u001b[1;32m 3\u001b[0m (\n\u001b[1;32m 4\u001b[0m epoch_num,\n\u001b[1;32m 5\u001b[0m m_epoch_loss_values,\n\u001b[1;32m 6\u001b[0m m_metric_values,\n\u001b[1;32m 7\u001b[0m m_epoch_times,\n\u001b[1;32m 8\u001b[0m m_best,\n\u001b[1;32m 9\u001b[0m m_train_time,\n\u001b[0;32m---> 10\u001b[0m ) \u001b[38;5;241m=\u001b[39m \u001b[43mtrain_process\u001b[49m\u001b[43m(\u001b[49m\u001b[43mfast\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m)\u001b[49m\n\u001b[1;32m 11\u001b[0m m_total_time \u001b[38;5;241m=\u001b[39m time\u001b[38;5;241m.\u001b[39mtime() \u001b[38;5;241m-\u001b[39m monai_start\n\u001b[1;32m 12\u001b[0m \u001b[38;5;28mprint\u001b[39m(\n\u001b[1;32m 13\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtotal time of \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mepoch_num\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m epochs with MONAI fast training: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mm_train_time\u001b[38;5;132;01m:\u001b[39;00m\u001b[38;5;124m.4f\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m,\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 14\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m time of preparing cache: \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m(m_total_time \u001b[38;5;241m-\u001b[39m m_train_time)\u001b[38;5;132;01m:\u001b[39;00m\u001b[38;5;124m.4f\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m 15\u001b[0m )\n", - "Input \u001b[0;32mIn [8]\u001b[0m, in \u001b[0;36mtrain_process\u001b[0;34m(fast)\u001b[0m\n\u001b[1;32m 134\u001b[0m \u001b[38;5;66;03m# profiling: backward\u001b[39;00m\n\u001b[1;32m 135\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m nvtx\u001b[38;5;241m.\u001b[39mannotate(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mbackward\u001b[39m\u001b[38;5;124m\"\u001b[39m, color\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mblue\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m profiling \u001b[38;5;28;01melse\u001b[39;00m no_profiling:\n\u001b[0;32m--> 136\u001b[0m \u001b[43mscaler\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mscale\u001b[49m\u001b[43m(\u001b[49m\u001b[43mloss\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m 138\u001b[0m \u001b[38;5;66;03m# profiling: update\u001b[39;00m\n\u001b[1;32m 139\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m nvtx\u001b[38;5;241m.\u001b[39mannotate(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mupdate\u001b[39m\u001b[38;5;124m\"\u001b[39m, color\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124myellow\u001b[39m\u001b[38;5;124m\"\u001b[39m) \u001b[38;5;28;01mif\u001b[39;00m profiling \u001b[38;5;28;01melse\u001b[39;00m no_profiling:\n", - "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/_tensor.py:396\u001b[0m, in \u001b[0;36mTensor.backward\u001b[0;34m(self, gradient, retain_graph, create_graph, inputs)\u001b[0m\n\u001b[1;32m 387\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m has_torch_function_unary(\u001b[38;5;28mself\u001b[39m):\n\u001b[1;32m 388\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m handle_torch_function(\n\u001b[1;32m 389\u001b[0m Tensor\u001b[38;5;241m.\u001b[39mbackward,\n\u001b[1;32m 390\u001b[0m (\u001b[38;5;28mself\u001b[39m,),\n\u001b[0;32m (...)\u001b[0m\n\u001b[1;32m 394\u001b[0m create_graph\u001b[38;5;241m=\u001b[39mcreate_graph,\n\u001b[1;32m 395\u001b[0m inputs\u001b[38;5;241m=\u001b[39minputs)\n\u001b[0;32m--> 396\u001b[0m \u001b[43mtorch\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mautograd\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mbackward\u001b[49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgradient\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43minputs\u001b[49m\u001b[43m)\u001b[49m\n", - "File \u001b[0;32m/opt/conda/lib/python3.8/site-packages/torch/autograd/__init__.py:173\u001b[0m, in \u001b[0;36mbackward\u001b[0;34m(tensors, grad_tensors, retain_graph, create_graph, grad_variables, inputs)\u001b[0m\n\u001b[1;32m 168\u001b[0m retain_graph \u001b[38;5;241m=\u001b[39m create_graph\n\u001b[1;32m 170\u001b[0m \u001b[38;5;66;03m# The reason we repeat same the comment below is that\u001b[39;00m\n\u001b[1;32m 171\u001b[0m \u001b[38;5;66;03m# some Python versions print out the first line of a multi-line function\u001b[39;00m\n\u001b[1;32m 172\u001b[0m \u001b[38;5;66;03m# calls in the traceback and some print out the last line\u001b[39;00m\n\u001b[0;32m--> 173\u001b[0m \u001b[43mVariable\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_execution_engine\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun_backward\u001b[49m\u001b[43m(\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;66;43;03m# Calls into the C++ engine to run the backward pass\u001b[39;49;00m\n\u001b[1;32m 174\u001b[0m \u001b[43m \u001b[49m\u001b[43mtensors\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mgrad_tensors_\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mretain_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcreate_graph\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[1;32m 175\u001b[0m \u001b[43m \u001b[49m\u001b[43mallow_unreachable\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43maccumulate_grad\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43;01mTrue\u001b[39;49;00m\u001b[43m)\u001b[49m\n", - "\u001b[0;31mKeyboardInterrupt\u001b[0m: " + "name": "stdout", + "output_type": "stream", + "text": [ + "5/8, train_loss: 0.0414 step time: 0.2156\n", + "6/8, train_loss: 0.0527 step time: 0.2157\n", + "7/8, train_loss: 0.0365 step time: 0.1993\n", + "8/8, train_loss: 0.0393 step time: 0.2000\n", + "epoch 93 average loss: 0.0481\n", + "time consuming of epoch 93 is: 1.7403\n", + "----------\n", + "epoch 94/600\n", + "1/8, train_loss: 0.0240 step time: 0.2646\n", + "2/8, train_loss: 0.0398 step time: 0.2138\n", + "3/8, train_loss: 0.0549 step time: 0.2149\n", + "4/8, train_loss: 0.0601 step time: 0.2168\n", + "5/8, train_loss: 0.0696 step time: 0.2138\n", + "6/8, train_loss: 0.0344 step time: 0.2156\n", + "7/8, train_loss: 0.0508 step time: 0.1990\n", + "8/8, train_loss: 0.0672 step time: 0.2003\n", + "epoch 94 average loss: 0.0501\n", + "time consuming of epoch 94 is: 1.7402\n", + "----------\n", + "epoch 95/600\n", + "1/8, train_loss: 0.0508 step time: 0.2629\n", + "2/8, train_loss: 0.0413 step time: 0.2136\n", + "3/8, train_loss: 0.0478 step time: 0.2157\n", + "4/8, train_loss: 0.0528 step time: 0.2155\n", + "5/8, train_loss: 0.0248 step time: 0.2148\n", + "6/8, train_loss: 0.0582 step time: 0.2145\n", + "7/8, train_loss: 0.0490 step time: 0.1993\n", + "8/8, train_loss: 0.0247 step time: 0.2014\n", + "epoch 95 average loss: 0.0437\n", + "current epoch: 95 current mean dice: 0.1925 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 95 is: 2.5512\n", + "----------\n", + "epoch 96/600\n", + "1/8, train_loss: 0.0288 step time: 0.2611\n", + "2/8, train_loss: 0.0319 step time: 0.2110\n", + "3/8, train_loss: 0.0300 step time: 0.2119\n", + "4/8, train_loss: 0.0356 step time: 0.2126\n", + "5/8, train_loss: 0.0322 step time: 0.2152\n", + "6/8, train_loss: 0.0352 step time: 0.2150\n", + "7/8, train_loss: 0.0530 step time: 0.2006\n", + "8/8, train_loss: 0.0252 step time: 0.1999\n", + "epoch 96 average loss: 0.0340\n", + "time consuming of epoch 96 is: 1.7285\n", + "----------\n", + "epoch 97/600\n", + "1/8, train_loss: 0.0357 step time: 0.2636\n", + "2/8, train_loss: 0.0318 step time: 0.2148\n", + "3/8, train_loss: 0.0343 step time: 0.2168\n", + "4/8, train_loss: 0.0234 step time: 0.2165\n", + "5/8, train_loss: 0.0269 step time: 0.2130\n", + "6/8, train_loss: 0.0466 step time: 0.2141\n", + "7/8, train_loss: 0.0283 step time: 0.1998\n", + "8/8, train_loss: 0.0359 step time: 0.1998\n", + "epoch 97 average loss: 0.0328\n", + "time consuming of epoch 97 is: 1.7398\n", + "----------\n", + "epoch 98/600\n", + "1/8, train_loss: 0.0201 step time: 0.2641\n", + "2/8, train_loss: 0.0382 step time: 0.2137\n", + "3/8, train_loss: 0.0219 step time: 0.2141\n", + "4/8, train_loss: 0.0327 step time: 0.2139\n", + "5/8, train_loss: 0.0207 step time: 0.2167\n", + "6/8, train_loss: 0.0336 step time: 0.2170\n", + "7/8, train_loss: 0.0258 step time: 0.1995\n", + "8/8, train_loss: 0.0345 step time: 0.1999\n", + "epoch 98 average loss: 0.0284\n", + "time consuming of epoch 98 is: 1.7406\n", + "----------\n", + "epoch 99/600\n", + "1/8, train_loss: 0.0243 step time: 0.2632\n", + "2/8, train_loss: 0.0259 step time: 0.2152\n", + "3/8, train_loss: 0.0277 step time: 0.2139\n", + "4/8, train_loss: 0.0353 step time: 0.2142\n", + "5/8, train_loss: 0.0259 step time: 0.2167\n", + "6/8, train_loss: 0.0255 step time: 0.2153\n", + "7/8, train_loss: 0.0613 step time: 0.2002\n", + "8/8, train_loss: 0.0238 step time: 0.1997\n", + "epoch 99 average loss: 0.0312\n", + "time consuming of epoch 99 is: 1.7397\n", + "----------\n", + "epoch 100/600\n", + "1/8, train_loss: 0.0289 step time: 0.2635\n", + "2/8, train_loss: 0.0191 step time: 0.2150\n", + "3/8, train_loss: 0.0901 step time: 0.2146\n", + "4/8, train_loss: 0.0261 step time: 0.2152\n", + "5/8, train_loss: 0.0320 step time: 0.2176\n", + "6/8, train_loss: 0.0262 step time: 0.2147\n", + "7/8, train_loss: 0.0279 step time: 0.1993\n", + "8/8, train_loss: 0.0354 step time: 0.2003\n", + "epoch 100 average loss: 0.0357\n", + "current epoch: 100 current mean dice: 0.9219 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 100 is: 2.5543\n", + "----------\n", + "epoch 101/600\n", + "1/8, train_loss: 0.0337 step time: 0.3933\n", + "2/8, train_loss: 0.0557 step time: 0.2106\n", + "3/8, train_loss: 0.0516 step time: 0.2073\n", + "4/8, train_loss: 0.0333 step time: 0.2090\n", + "5/8, train_loss: 0.0349 step time: 0.2087\n", + "6/8, train_loss: 0.0440 step time: 0.2080\n", + "7/8, train_loss: 0.0278 step time: 0.1991\n", + "8/8, train_loss: 0.0217 step time: 0.2000\n", + "epoch 101 average loss: 0.0378\n", + "time consuming of epoch 101 is: 1.8373\n", + "----------\n", + "epoch 102/600\n", + "1/8, train_loss: 0.0314 step time: 0.2663\n", + "2/8, train_loss: 0.0334 step time: 0.2169\n", + "3/8, train_loss: 0.0302 step time: 0.2137\n", + "4/8, train_loss: 0.0319 step time: 0.2141\n", + "5/8, train_loss: 0.0315 step time: 0.2155\n", + "6/8, train_loss: 0.0349 step time: 0.2156\n", + "7/8, train_loss: 0.0276 step time: 0.2020\n", + "8/8, train_loss: 0.0277 step time: 0.1991\n", + "epoch 102 average loss: 0.0311\n", + "time consuming of epoch 102 is: 1.7447\n", + "----------\n", + "epoch 103/600\n", + "1/8, train_loss: 0.0380 step time: 0.2642\n", + "2/8, train_loss: 0.0316 step time: 0.2176\n", + "3/8, train_loss: 0.0293 step time: 0.2129\n", + "4/8, train_loss: 0.0413 step time: 0.2127\n", + "5/8, train_loss: 0.0253 step time: 0.2146\n", + "6/8, train_loss: 0.0378 step time: 0.2127\n", + "7/8, train_loss: 0.0198 step time: 0.2004\n", + "8/8, train_loss: 0.0234 step time: 0.2003\n", + "epoch 103 average loss: 0.0308\n", + "time consuming of epoch 103 is: 1.7368\n", + "----------\n", + "epoch 104/600\n", + "1/8, train_loss: 0.0286 step time: 0.2617\n", + "2/8, train_loss: 0.0317 step time: 0.2149\n", + "3/8, train_loss: 0.0281 step time: 0.2140\n", + "4/8, train_loss: 0.0213 step time: 0.2165\n", + "5/8, train_loss: 0.0243 step time: 0.2131\n", + "6/8, train_loss: 0.0266 step time: 0.2173\n", + "7/8, train_loss: 0.0404 step time: 0.1982\n", + "8/8, train_loss: 0.0252 step time: 0.1991\n", + "epoch 104 average loss: 0.0283\n", + "time consuming of epoch 104 is: 1.7364\n", + "----------\n", + "epoch 105/600\n", + "1/8, train_loss: 0.0254 step time: 0.2645\n", + "2/8, train_loss: 0.0208 step time: 0.2172\n", + "3/8, train_loss: 0.0324 step time: 0.2149\n", + "4/8, train_loss: 0.0226 step time: 0.2141\n", + "5/8, train_loss: 0.0243 step time: 0.2148\n", + "6/8, train_loss: 0.0319 step time: 0.2163\n", + "7/8, train_loss: 0.0245 step time: 0.1998\n", + "8/8, train_loss: 0.0257 step time: 0.1995\n", + "epoch 105 average loss: 0.0259\n", + "current epoch: 105 current mean dice: 0.9335 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 105 is: 2.5723\n", + "----------\n", + "epoch 106/600\n", + "1/8, train_loss: 0.0211 step time: 0.2766\n", + "2/8, train_loss: 0.0219 step time: 0.2258\n", + "3/8, train_loss: 0.0244 step time: 0.2242\n", + "4/8, train_loss: 0.0243 step time: 0.2183\n", + "5/8, train_loss: 0.0236 step time: 0.2154\n", + "6/8, train_loss: 0.0244 step time: 0.2137\n", + "7/8, train_loss: 0.0258 step time: 0.2032\n", + "8/8, train_loss: 0.0235 step time: 0.2022\n", + "epoch 106 average loss: 0.0237\n", + "time consuming of epoch 106 is: 1.7806\n", + "----------\n", + "epoch 107/600\n", + "1/8, train_loss: 0.0262 step time: 0.2660\n", + "2/8, train_loss: 0.0273 step time: 0.2159\n", + "3/8, train_loss: 0.0186 step time: 0.2174\n", + "4/8, train_loss: 0.0224 step time: 0.2149\n", + "5/8, train_loss: 0.0279 step time: 0.2148\n", + "6/8, train_loss: 0.0216 step time: 0.2155\n", + "7/8, train_loss: 0.0237 step time: 0.2001\n", + "8/8, train_loss: 0.0244 step time: 0.2037\n", + "epoch 107 average loss: 0.0240\n", + "time consuming of epoch 107 is: 1.7499\n", + "----------\n", + "epoch 108/600\n", + "1/8, train_loss: 0.0229 step time: 0.2651\n", + "2/8, train_loss: 0.0203 step time: 0.2147\n", + "3/8, train_loss: 0.0184 step time: 0.2107\n", + "4/8, train_loss: 0.0293 step time: 0.2118\n", + "5/8, train_loss: 0.0233 step time: 0.2125\n", + "6/8, train_loss: 0.0240 step time: 0.2123\n", + "7/8, train_loss: 0.0281 step time: 0.2007\n", + "8/8, train_loss: 0.0232 step time: 0.2057\n", + "epoch 108 average loss: 0.0237\n", + "time consuming of epoch 108 is: 1.7350\n", + "----------\n", + "epoch 109/600\n", + "1/8, train_loss: 0.0215 step time: 0.2678\n", + "2/8, train_loss: 0.0231 step time: 0.2189\n", + "3/8, train_loss: 0.0246 step time: 0.2177\n", + "4/8, train_loss: 0.0215 step time: 0.2189\n", + "5/8, train_loss: 0.0234 step time: 0.2160\n", + "6/8, train_loss: 0.0181 step time: 0.2145\n", + "7/8, train_loss: 0.0210 step time: 0.2022\n", + "8/8, train_loss: 0.0285 step time: 0.1994\n", + "epoch 109 average loss: 0.0227\n", + "time consuming of epoch 109 is: 1.7568\n", + "----------\n", + "epoch 110/600\n", + "1/8, train_loss: 0.0167 step time: 0.2637\n", + "2/8, train_loss: 0.0200 step time: 0.2150\n", + "3/8, train_loss: 0.0215 step time: 0.2162\n", + "4/8, train_loss: 0.0264 step time: 0.2205\n", + "5/8, train_loss: 0.0186 step time: 0.2225\n", + "6/8, train_loss: 0.0220 step time: 0.2279\n", + "7/8, train_loss: 0.0206 step time: 0.2394\n", + "8/8, train_loss: 0.0223 step time: 0.2422\n", + "epoch 110 average loss: 0.0210\n", + "current epoch: 110 current mean dice: 0.9441 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 110 is: 3.0572\n", + "----------\n", + "epoch 111/600\n", + "1/8, train_loss: 0.0231 step time: 0.3005\n", + "2/8, train_loss: 0.0214 step time: 0.2418\n", + "3/8, train_loss: 0.0227 step time: 0.2369\n", + "4/8, train_loss: 0.0293 step time: 0.2297\n", + "5/8, train_loss: 0.0196 step time: 0.2283\n", + "6/8, train_loss: 0.0322 step time: 0.2267\n", + "7/8, train_loss: 0.0214 step time: 0.2151\n", + "8/8, train_loss: 0.0283 step time: 0.2111\n", + "epoch 111 average loss: 0.0247\n", + "time consuming of epoch 111 is: 1.8914\n", + "----------\n", + "epoch 112/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0210 step time: 0.2729\n", + "2/8, train_loss: 0.0199 step time: 0.2252\n", + "3/8, train_loss: 0.0228 step time: 0.2188\n", + "4/8, train_loss: 0.0183 step time: 0.2215\n", + "5/8, train_loss: 0.0208 step time: 0.2196\n", + "6/8, train_loss: 0.0243 step time: 0.2173\n", + "7/8, train_loss: 0.0227 step time: 0.2083\n", + "8/8, train_loss: 0.0229 step time: 0.2040\n", + "epoch 112 average loss: 0.0216\n", + "time consuming of epoch 112 is: 1.7892\n", + "----------\n", + "epoch 113/600\n", + "1/8, train_loss: 0.0230 step time: 0.2714\n", + "2/8, train_loss: 0.0279 step time: 0.2171\n", + "3/8, train_loss: 0.0225 step time: 0.2201\n", + "4/8, train_loss: 0.0293 step time: 0.2186\n", + "5/8, train_loss: 0.0216 step time: 0.2176\n", + "6/8, train_loss: 0.0257 step time: 0.2168\n", + "7/8, train_loss: 0.0293 step time: 0.2037\n", + "8/8, train_loss: 0.0204 step time: 0.2029\n", + "epoch 113 average loss: 0.0250\n", + "time consuming of epoch 113 is: 1.7697\n", + "----------\n", + "epoch 114/600\n", + "1/8, train_loss: 0.0214 step time: 0.2658\n", + "2/8, train_loss: 0.0239 step time: 0.2184\n", + "3/8, train_loss: 0.0221 step time: 0.2154\n", + "4/8, train_loss: 0.0214 step time: 0.2202\n", + "5/8, train_loss: 0.0236 step time: 0.2184\n", + "6/8, train_loss: 0.0205 step time: 0.2178\n", + "7/8, train_loss: 0.0225 step time: 0.2119\n", + "8/8, train_loss: 0.0242 step time: 0.2511\n", + "epoch 114 average loss: 0.0225\n", + "time consuming of epoch 114 is: 1.8205\n", + "----------\n", + "epoch 115/600\n", + "1/8, train_loss: 0.0266 step time: 0.3097\n", + "2/8, train_loss: 0.0227 step time: 0.2486\n", + "3/8, train_loss: 0.0227 step time: 0.2385\n", + "4/8, train_loss: 0.0214 step time: 0.2344\n", + "5/8, train_loss: 0.0171 step time: 0.2311\n", + "6/8, train_loss: 0.0180 step time: 0.2297\n", + "7/8, train_loss: 0.0211 step time: 0.2150\n", + "8/8, train_loss: 0.0245 step time: 0.2139\n", + "epoch 115 average loss: 0.0217\n", + "current epoch: 115 current mean dice: 0.9427 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 115 is: 3.1218\n", + "----------\n", + "epoch 116/600\n", + "1/8, train_loss: 0.0221 step time: 0.2995\n", + "2/8, train_loss: 0.0187 step time: 0.2502\n", + "3/8, train_loss: 0.0235 step time: 0.2394\n", + "4/8, train_loss: 0.0203 step time: 0.2371\n", + "5/8, train_loss: 0.0244 step time: 0.2363\n", + "6/8, train_loss: 0.0237 step time: 0.2340\n", + "7/8, train_loss: 0.0238 step time: 0.2189\n", + "8/8, train_loss: 0.0228 step time: 0.2172\n", + "epoch 116 average loss: 0.0224\n", + "time consuming of epoch 116 is: 1.9339\n", + "----------\n", + "epoch 117/600\n", + "1/8, train_loss: 0.0215 step time: 0.2804\n", + "2/8, train_loss: 0.0222 step time: 0.2304\n", + "3/8, train_loss: 0.0163 step time: 0.2279\n", + "4/8, train_loss: 0.0185 step time: 0.2276\n", + "5/8, train_loss: 0.0219 step time: 0.2289\n", + "6/8, train_loss: 0.0298 step time: 0.2257\n", + "7/8, train_loss: 0.0226 step time: 0.2153\n", + "8/8, train_loss: 0.0233 step time: 0.2145\n", + "epoch 117 average loss: 0.0220\n", + "time consuming of epoch 117 is: 1.8522\n", + "----------\n", + "epoch 118/600\n", + "1/8, train_loss: 0.0242 step time: 0.2758\n", + "2/8, train_loss: 0.0248 step time: 0.2279\n", + "3/8, train_loss: 0.0180 step time: 0.3280\n", + "4/8, train_loss: 0.0191 step time: 0.4553\n", + "5/8, train_loss: 0.0229 step time: 0.2192\n", + "6/8, train_loss: 0.0205 step time: 0.2147\n", + "7/8, train_loss: 0.0187 step time: 0.2005\n", + "8/8, train_loss: 0.0226 step time: 0.2005\n", + "epoch 118 average loss: 0.0214\n", + "time consuming of epoch 118 is: 2.1235\n", + "----------\n", + "epoch 119/600\n", + "1/8, train_loss: 0.0212 step time: 0.2639\n", + "2/8, train_loss: 0.0215 step time: 0.2128\n", + "3/8, train_loss: 0.0216 step time: 0.2226\n", + "4/8, train_loss: 0.0202 step time: 0.2243\n", + "5/8, train_loss: 0.0204 step time: 0.2245\n", + "6/8, train_loss: 0.0227 step time: 0.2271\n", + "7/8, train_loss: 0.0194 step time: 0.2089\n", + "8/8, train_loss: 0.0202 step time: 0.2101\n", + "epoch 119 average loss: 0.0209\n", + "time consuming of epoch 119 is: 1.7956\n", + "----------\n", + "epoch 120/600\n", + "1/8, train_loss: 0.0178 step time: 0.2716\n", + "2/8, train_loss: 0.0190 step time: 0.2268\n", + "3/8, train_loss: 0.0209 step time: 0.2515\n", + "4/8, train_loss: 0.0243 step time: 0.2591\n", + "5/8, train_loss: 0.0194 step time: 0.2536\n", + "6/8, train_loss: 0.0204 step time: 0.2850\n", + "7/8, train_loss: 0.0233 step time: 0.3002\n", + "8/8, train_loss: 0.0201 step time: 0.2680\n", + "epoch 120 average loss: 0.0207\n", + "current epoch: 120 current mean dice: 0.9477 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 120 is: 3.2718\n", + "----------\n", + "epoch 121/600\n", + "1/8, train_loss: 0.0204 step time: 0.6131\n", + "2/8, train_loss: 0.0204 step time: 0.2160\n", + "3/8, train_loss: 0.0192 step time: 0.2140\n", + "4/8, train_loss: 0.0168 step time: 0.2134\n", + "5/8, train_loss: 0.0171 step time: 0.2161\n", + "6/8, train_loss: 0.0208 step time: 0.2174\n", + "7/8, train_loss: 0.0225 step time: 0.2095\n", + "8/8, train_loss: 0.0220 step time: 0.2179\n", + "epoch 121 average loss: 0.0199\n", + "time consuming of epoch 121 is: 2.1187\n", + "----------\n", + "epoch 122/600\n", + "1/8, train_loss: 0.0174 step time: 0.2792\n", + "2/8, train_loss: 0.0189 step time: 0.2303\n", + "3/8, train_loss: 0.0160 step time: 0.2365\n", + "4/8, train_loss: 0.0231 step time: 0.2357\n", + "5/8, train_loss: 0.0218 step time: 0.2906\n", + "6/8, train_loss: 0.0185 step time: 0.2737\n", + "7/8, train_loss: 0.0206 step time: 0.2583\n", + "8/8, train_loss: 0.0195 step time: 0.2632\n", + "epoch 122 average loss: 0.0195\n", + "time consuming of epoch 122 is: 2.0690\n", + "----------\n", + "epoch 123/600\n", + "1/8, train_loss: 0.0230 step time: 0.2869\n", + "2/8, train_loss: 0.0209 step time: 0.2344\n", + "3/8, train_loss: 0.0280 step time: 0.2307\n", + "4/8, train_loss: 0.0246 step time: 0.2303\n", + "5/8, train_loss: 0.0218 step time: 0.2353\n", + "6/8, train_loss: 0.0244 step time: 0.2301\n", + "7/8, train_loss: 0.0150 step time: 0.2154\n", + "8/8, train_loss: 0.0191 step time: 0.2214\n", + "epoch 123 average loss: 0.0221\n", + "time consuming of epoch 123 is: 1.8861\n", + "----------\n", + "epoch 124/600\n", + "1/8, train_loss: 0.0213 step time: 0.2820\n", + "2/8, train_loss: 0.0249 step time: 0.2331\n", + "3/8, train_loss: 0.0180 step time: 0.2308\n", + "4/8, train_loss: 0.0196 step time: 0.2279\n", + "5/8, train_loss: 0.0197 step time: 0.2346\n", + "6/8, train_loss: 0.0222 step time: 0.2475\n", + "7/8, train_loss: 0.0210 step time: 0.2397\n", + "8/8, train_loss: 0.0195 step time: 0.2430\n", + "epoch 124 average loss: 0.0208\n", + "time consuming of epoch 124 is: 1.9402\n", + "----------\n", + "epoch 125/600\n", + "1/8, train_loss: 0.0232 step time: 0.2910\n", + "2/8, train_loss: 0.0201 step time: 0.2362\n", + "3/8, train_loss: 0.0192 step time: 0.2311\n", + "4/8, train_loss: 0.0206 step time: 0.2287\n", + "5/8, train_loss: 0.0208 step time: 0.2397\n", + "6/8, train_loss: 0.0226 step time: 0.2572\n", + "7/8, train_loss: 0.0209 step time: 0.2512\n", + "8/8, train_loss: 0.0179 step time: 0.2687\n", + "epoch 125 average loss: 0.0207\n", + "current epoch: 125 current mean dice: 0.9476 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 125 is: 3.2610\n", + "----------\n", + "epoch 126/600\n", + "1/8, train_loss: 0.0191 step time: 0.3064\n", + "2/8, train_loss: 0.0198 step time: 0.2526\n", + "3/8, train_loss: 0.0196 step time: 0.2497\n", + "4/8, train_loss: 0.0254 step time: 0.2512\n", + "5/8, train_loss: 0.0194 step time: 0.2375\n", + "6/8, train_loss: 0.0220 step time: 0.2305\n", + "7/8, train_loss: 0.0193 step time: 0.2231\n", + "8/8, train_loss: 0.0207 step time: 0.2207\n", + "epoch 126 average loss: 0.0207\n", + "time consuming of epoch 126 is: 1.9730\n", + "----------\n", + "epoch 127/600\n", + "1/8, train_loss: 0.0232 step time: 0.2841\n", + "2/8, train_loss: 0.0197 step time: 0.2353\n", + "3/8, train_loss: 0.0185 step time: 0.2341\n", + "4/8, train_loss: 0.0225 step time: 0.2313\n", + "5/8, train_loss: 0.0187 step time: 0.2293\n", + "6/8, train_loss: 0.0223 step time: 0.2294\n", + "7/8, train_loss: 0.0180 step time: 0.2609\n", + "8/8, train_loss: 0.0170 step time: 0.2969\n", + "epoch 127 average loss: 0.0200\n", + "time consuming of epoch 127 is: 2.0029\n", + "----------\n", + "epoch 128/600\n", + "1/8, train_loss: 0.0208 step time: 0.3316\n", + "2/8, train_loss: 0.0232 step time: 0.2600\n", + "3/8, train_loss: 0.0173 step time: 0.2530\n", + "4/8, train_loss: 0.0185 step time: 0.2456\n", + "5/8, train_loss: 0.0185 step time: 0.2420\n", + "6/8, train_loss: 0.0174 step time: 0.2396\n", + "7/8, train_loss: 0.0243 step time: 0.2248\n", + "8/8, train_loss: 0.0193 step time: 0.2170\n", + "epoch 128 average loss: 0.0199\n", + "time consuming of epoch 128 is: 2.0151\n", + "----------\n", + "epoch 129/600\n", + "1/8, train_loss: 0.0191 step time: 0.2806\n", + "2/8, train_loss: 0.0201 step time: 0.2297\n", + "3/8, train_loss: 0.0203 step time: 0.2294\n", + "4/8, train_loss: 0.0209 step time: 0.2306\n", + "5/8, train_loss: 0.0182 step time: 0.2314\n", + "6/8, train_loss: 0.0216 step time: 0.2288\n", + "7/8, train_loss: 0.0151 step time: 0.2192\n", + "8/8, train_loss: 0.0207 step time: 0.2188\n", + "epoch 129 average loss: 0.0195\n", + "time consuming of epoch 129 is: 1.8699\n", + "----------\n", + "epoch 130/600\n", + "1/8, train_loss: 0.0208 step time: 0.2805\n", + "2/8, train_loss: 0.0225 step time: 0.2280\n", + "3/8, train_loss: 0.0156 step time: 0.2325\n", + "4/8, train_loss: 0.0190 step time: 0.2392\n", + "5/8, train_loss: 0.0211 step time: 0.2309\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "6/8, train_loss: 0.0212 step time: 0.3590\n", + "7/8, train_loss: 0.0184 step time: 0.4614\n", + "8/8, train_loss: 0.0262 step time: 0.2619\n", + "epoch 130 average loss: 0.0206\n", + "current epoch: 130 current mean dice: 0.9471 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 130 is: 3.4656\n", + "----------\n", + "epoch 131/600\n", + "1/8, train_loss: 0.0291 step time: 0.2864\n", + "2/8, train_loss: 0.0266 step time: 0.2428\n", + "3/8, train_loss: 0.0209 step time: 0.2429\n", + "4/8, train_loss: 0.0181 step time: 0.2571\n", + "5/8, train_loss: 0.0184 step time: 0.2564\n", + "6/8, train_loss: 0.0183 step time: 0.2501\n", + "7/8, train_loss: 0.0196 step time: 0.2410\n", + "8/8, train_loss: 0.0166 step time: 0.2282\n", + "epoch 131 average loss: 0.0209\n", + "time consuming of epoch 131 is: 2.0061\n", + "----------\n", + "epoch 132/600\n", + "1/8, train_loss: 0.0220 step time: 0.2844\n", + "2/8, train_loss: 0.0206 step time: 0.2313\n", + "3/8, train_loss: 0.0181 step time: 0.2287\n", + "4/8, train_loss: 0.0180 step time: 0.2276\n", + "5/8, train_loss: 0.0221 step time: 0.2250\n", + "6/8, train_loss: 0.0238 step time: 0.2252\n", + "7/8, train_loss: 0.0185 step time: 0.2222\n", + "8/8, train_loss: 0.0199 step time: 0.2649\n", + "epoch 132 average loss: 0.0204\n", + "time consuming of epoch 132 is: 1.9109\n", + "----------\n", + "epoch 133/600\n", + "1/8, train_loss: 0.0211 step time: 0.3262\n", + "2/8, train_loss: 0.0201 step time: 0.2779\n", + "3/8, train_loss: 0.0175 step time: 0.2805\n", + "4/8, train_loss: 0.0195 step time: 0.2788\n", + "5/8, train_loss: 0.0156 step time: 0.2808\n", + "6/8, train_loss: 0.0207 step time: 0.2754\n", + "7/8, train_loss: 0.0170 step time: 0.2473\n", + "8/8, train_loss: 0.0198 step time: 0.2316\n", + "epoch 133 average loss: 0.0189\n", + "time consuming of epoch 133 is: 2.2002\n", + "----------\n", + "epoch 134/600\n", + "1/8, train_loss: 0.0187 step time: 0.2844\n", + "2/8, train_loss: 0.0196 step time: 0.2382\n", + "3/8, train_loss: 0.0194 step time: 0.2323\n", + "4/8, train_loss: 0.0172 step time: 0.2874\n", + "5/8, train_loss: 0.0209 step time: 0.3443\n", + "6/8, train_loss: 0.0186 step time: 0.2775\n", + "7/8, train_loss: 0.0215 step time: 0.2533\n", + "8/8, train_loss: 0.0262 step time: 0.2407\n", + "epoch 134 average loss: 0.0203\n", + "time consuming of epoch 134 is: 2.1596\n", + "----------\n", + "epoch 135/600\n", + "1/8, train_loss: 0.0226 step time: 0.2912\n", + "2/8, train_loss: 0.0206 step time: 0.2377\n", + "3/8, train_loss: 0.0199 step time: 0.2368\n", + "4/8, train_loss: 0.0233 step time: 0.2442\n", + "5/8, train_loss: 0.0160 step time: 0.2434\n", + "6/8, train_loss: 0.0186 step time: 0.2377\n", + "7/8, train_loss: 0.0200 step time: 0.2240\n", + "8/8, train_loss: 0.0176 step time: 0.2266\n", + "epoch 135 average loss: 0.0198\n", + "current epoch: 135 current mean dice: 0.9466 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 135 is: 3.2663\n", + "----------\n", + "epoch 136/600\n", + "1/8, train_loss: 0.0209 step time: 0.3197\n", + "2/8, train_loss: 0.0221 step time: 0.2581\n", + "3/8, train_loss: 0.0196 step time: 0.2428\n", + "4/8, train_loss: 0.0203 step time: 0.2376\n", + "5/8, train_loss: 0.0193 step time: 0.2717\n", + "6/8, train_loss: 0.0218 step time: 0.3029\n", + "7/8, train_loss: 0.0176 step time: 0.2722\n", + "8/8, train_loss: 0.0170 step time: 0.2534\n", + "epoch 136 average loss: 0.0198\n", + "time consuming of epoch 136 is: 2.1596\n", + "----------\n", + "epoch 137/600\n", + "1/8, train_loss: 0.0220 step time: 0.2881\n", + "2/8, train_loss: 0.0183 step time: 0.2384\n", + "3/8, train_loss: 0.0175 step time: 0.2355\n", + "4/8, train_loss: 0.0180 step time: 0.2394\n", + "5/8, train_loss: 0.0189 step time: 0.2311\n", + "6/8, train_loss: 0.0173 step time: 0.2291\n", + "7/8, train_loss: 0.0183 step time: 0.2170\n", + "8/8, train_loss: 0.0205 step time: 0.2151\n", + "epoch 137 average loss: 0.0189\n", + "time consuming of epoch 137 is: 1.8952\n", + "----------\n", + "epoch 138/600\n", + "1/8, train_loss: 0.0191 step time: 0.2777\n", + "2/8, train_loss: 0.0270 step time: 0.2284\n", + "3/8, train_loss: 0.0209 step time: 0.2284\n", + "4/8, train_loss: 0.0200 step time: 0.2294\n", + "5/8, train_loss: 0.0199 step time: 0.2257\n", + "6/8, train_loss: 0.0167 step time: 0.2283\n", + "7/8, train_loss: 0.0179 step time: 0.2384\n", + "8/8, train_loss: 0.0231 step time: 0.3161\n", + "epoch 138 average loss: 0.0206\n", + "time consuming of epoch 138 is: 1.9739\n", + "----------\n", + "epoch 139/600\n", + "1/8, train_loss: 0.0240 step time: 0.3044\n", + "2/8, train_loss: 0.0212 step time: 0.2574\n", + "3/8, train_loss: 0.0202 step time: 0.2473\n", + "4/8, train_loss: 0.0170 step time: 0.2423\n", + "5/8, train_loss: 0.0182 step time: 0.2384\n", + "6/8, train_loss: 0.0172 step time: 0.2609\n", + "7/8, train_loss: 0.0451 step time: 0.3056\n", + "8/8, train_loss: 0.0214 step time: 0.2532\n", + "epoch 139 average loss: 0.0230\n", + "time consuming of epoch 139 is: 2.1108\n", + "----------\n", + "epoch 140/600\n", + "1/8, train_loss: 0.0203 step time: 0.2944\n", + "2/8, train_loss: 0.0183 step time: 0.2452\n", + "3/8, train_loss: 0.0229 step time: 0.2454\n", + "4/8, train_loss: 0.0211 step time: 0.2451\n", + "5/8, train_loss: 0.0179 step time: 0.2434\n", + "6/8, train_loss: 0.0258 step time: 0.2364\n", + "7/8, train_loss: 0.0220 step time: 0.2185\n", + "8/8, train_loss: 0.0227 step time: 0.2204\n", + "epoch 140 average loss: 0.0214\n", + "current epoch: 140 current mean dice: 0.9294 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 140 is: 3.4298\n", + "----------\n", + "epoch 141/600\n", + "1/8, train_loss: 0.0185 step time: 0.3186\n", + "2/8, train_loss: 0.0314 step time: 0.2535\n", + "3/8, train_loss: 0.0264 step time: 0.2387\n", + "4/8, train_loss: 0.0225 step time: 0.2328\n", + "5/8, train_loss: 0.0378 step time: 0.2302\n", + "6/8, train_loss: 0.0199 step time: 0.2331\n", + "7/8, train_loss: 0.0142 step time: 0.2206\n", + "8/8, train_loss: 0.0228 step time: 0.2198\n", + "epoch 141 average loss: 0.0242\n", + "time consuming of epoch 141 is: 1.9486\n", + "----------\n", + "epoch 142/600\n", + "1/8, train_loss: 0.0251 step time: 0.2790\n", + "2/8, train_loss: 0.0175 step time: 0.2442\n", + "3/8, train_loss: 0.0218 step time: 0.2596\n", + "4/8, train_loss: 0.0164 step time: 0.2549\n", + "5/8, train_loss: 0.0249 step time: 0.2457\n", + "6/8, train_loss: 0.0186 step time: 0.2438\n", + "7/8, train_loss: 0.0329 step time: 0.2257\n", + "8/8, train_loss: 0.0243 step time: 0.2198\n", + "epoch 142 average loss: 0.0227\n", + "time consuming of epoch 142 is: 1.9742\n", + "----------\n", + "epoch 143/600\n", + "1/8, train_loss: 0.0178 step time: 0.2775\n", + "2/8, train_loss: 0.0203 step time: 0.2299\n", + "3/8, train_loss: 0.0239 step time: 0.2283\n", + "4/8, train_loss: 0.0260 step time: 0.2296\n", + "5/8, train_loss: 0.0190 step time: 0.2273\n", + "6/8, train_loss: 0.0200 step time: 0.2282\n", + "7/8, train_loss: 0.0188 step time: 0.2777\n", + "8/8, train_loss: 0.0293 step time: 0.3090\n", + "epoch 143 average loss: 0.0219\n", + "time consuming of epoch 143 is: 2.0091\n", + "----------\n", + "epoch 144/600\n", + "1/8, train_loss: 0.0224 step time: 0.3165\n", + "2/8, train_loss: 0.0225 step time: 0.2725\n", + "3/8, train_loss: 0.0291 step time: 0.2632\n", + "4/8, train_loss: 0.0212 step time: 0.2556\n", + "5/8, train_loss: 0.0214 step time: 0.2497\n", + "6/8, train_loss: 0.0228 step time: 0.2457\n", + "7/8, train_loss: 0.0169 step time: 0.2364\n", + "8/8, train_loss: 0.0209 step time: 0.2319\n", + "epoch 144 average loss: 0.0222\n", + "time consuming of epoch 144 is: 2.0729\n", + "----------\n", + "epoch 145/600\n", + "1/8, train_loss: 0.0193 step time: 0.2852\n", + "2/8, train_loss: 0.0267 step time: 0.2338\n", + "3/8, train_loss: 0.0279 step time: 0.3805\n", + "4/8, train_loss: 0.0277 step time: 0.4950\n", + "5/8, train_loss: 0.0195 step time: 0.2680\n", + "6/8, train_loss: 0.0193 step time: 0.2482\n", + "7/8, train_loss: 0.0179 step time: 0.2401\n", + "8/8, train_loss: 0.0231 step time: 0.2343\n", + "epoch 145 average loss: 0.0227\n", + "current epoch: 145 current mean dice: 0.9417 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 145 is: 3.7547\n", + "----------\n", + "epoch 146/600\n", + "1/8, train_loss: 0.0642 step time: 0.3173\n", + "2/8, train_loss: 0.0190 step time: 0.2507\n", + "3/8, train_loss: 0.0257 step time: 0.2442\n", + "4/8, train_loss: 0.0213 step time: 0.2439\n", + "5/8, train_loss: 0.0514 step time: 0.2456\n", + "6/8, train_loss: 0.0286 step time: 0.2433\n", + "7/8, train_loss: 0.0311 step time: 0.2393\n", + "8/8, train_loss: 0.0181 step time: 0.2273\n", + "epoch 146 average loss: 0.0324\n", + "time consuming of epoch 146 is: 2.0128\n", + "----------\n", + "epoch 147/600\n", + "1/8, train_loss: 0.0180 step time: 0.2805\n", + "2/8, train_loss: 0.0517 step time: 0.2341\n", + "3/8, train_loss: 0.0350 step time: 0.2633\n", + "4/8, train_loss: 0.0252 step time: 0.2701\n", + "5/8, train_loss: 0.0327 step time: 0.2507\n", + "6/8, train_loss: 0.0262 step time: 0.2449\n", + "7/8, train_loss: 0.0222 step time: 0.2359\n", + "8/8, train_loss: 0.0190 step time: 0.2303\n", + "epoch 147 average loss: 0.0288\n", + "time consuming of epoch 147 is: 2.0113\n", + "----------\n", + "epoch 148/600\n", + "1/8, train_loss: 0.0262 step time: 0.2836\n", + "2/8, train_loss: 0.0271 step time: 0.2313\n", + "3/8, train_loss: 0.0248 step time: 0.2311\n", + "4/8, train_loss: 0.0293 step time: 0.2323\n", + "5/8, train_loss: 0.0271 step time: 0.2420\n", + "6/8, train_loss: 0.0196 step time: 0.2353\n", + "7/8, train_loss: 0.0384 step time: 0.2180\n", + "8/8, train_loss: 0.0247 step time: 0.2626\n", + "epoch 148 average loss: 0.0272\n", + "time consuming of epoch 148 is: 1.9376\n", + "----------\n", + "epoch 149/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0279 step time: 0.3637\n", + "2/8, train_loss: 0.0226 step time: 0.2666\n", + "3/8, train_loss: 0.0229 step time: 0.2559\n", + "4/8, train_loss: 0.0234 step time: 0.2476\n", + "5/8, train_loss: 0.0203 step time: 0.2396\n", + "6/8, train_loss: 0.0442 step time: 0.2475\n", + "7/8, train_loss: 0.0232 step time: 0.2661\n", + "8/8, train_loss: 0.0215 step time: 0.2627\n", + "epoch 149 average loss: 0.0257\n", + "time consuming of epoch 149 is: 2.1513\n", + "----------\n", + "epoch 150/600\n", + "1/8, train_loss: 0.0392 step time: 0.2907\n", + "2/8, train_loss: 0.0269 step time: 0.2484\n", + "3/8, train_loss: 0.0256 step time: 0.2481\n", + "4/8, train_loss: 0.0186 step time: 0.2415\n", + "5/8, train_loss: 0.0289 step time: 0.2323\n", + "6/8, train_loss: 0.0218 step time: 0.2349\n", + "7/8, train_loss: 0.0286 step time: 0.2921\n", + "8/8, train_loss: 0.0311 step time: 0.2697\n", + "epoch 150 average loss: 0.0276\n", + "current epoch: 150 current mean dice: 0.9290 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 150 is: 3.5274\n", + "----------\n", + "epoch 151/600\n", + "1/8, train_loss: 0.0421 step time: 0.2959\n", + "2/8, train_loss: 0.0250 step time: 0.2376\n", + "3/8, train_loss: 0.0240 step time: 0.2412\n", + "4/8, train_loss: 0.0242 step time: 0.2427\n", + "5/8, train_loss: 0.0247 step time: 0.2429\n", + "6/8, train_loss: 0.0245 step time: 0.2930\n", + "7/8, train_loss: 0.0272 step time: 0.2762\n", + "8/8, train_loss: 0.0200 step time: 0.2532\n", + "epoch 151 average loss: 0.0265\n", + "time consuming of epoch 151 is: 2.0839\n", + "----------\n", + "epoch 152/600\n", + "1/8, train_loss: 0.0186 step time: 0.2877\n", + "2/8, train_loss: 0.0211 step time: 0.2318\n", + "3/8, train_loss: 0.0257 step time: 0.2368\n", + "4/8, train_loss: 0.0229 step time: 0.2518\n", + "5/8, train_loss: 0.0183 step time: 0.3069\n", + "6/8, train_loss: 0.0205 step time: 0.2862\n", + "7/8, train_loss: 0.0208 step time: 0.2707\n", + "8/8, train_loss: 0.0196 step time: 0.2550\n", + "epoch 152 average loss: 0.0209\n", + "time consuming of epoch 152 is: 2.1283\n", + "----------\n", + "epoch 153/600\n", + "1/8, train_loss: 0.0242 step time: 0.2957\n", + "2/8, train_loss: 0.0288 step time: 0.2501\n", + "3/8, train_loss: 0.0192 step time: 0.2427\n", + "4/8, train_loss: 0.0193 step time: 0.2368\n", + "5/8, train_loss: 0.0186 step time: 0.2340\n", + "6/8, train_loss: 0.0196 step time: 0.2377\n", + "7/8, train_loss: 0.0225 step time: 0.2245\n", + "8/8, train_loss: 0.0212 step time: 0.3610\n", + "epoch 153 average loss: 0.0217\n", + "time consuming of epoch 153 is: 2.0840\n", + "----------\n", + "epoch 154/600\n", + "1/8, train_loss: 0.0246 step time: 0.4169\n", + "2/8, train_loss: 0.0249 step time: 0.2705\n", + "3/8, train_loss: 0.0208 step time: 0.2581\n", + "4/8, train_loss: 0.0253 step time: 0.2428\n", + "5/8, train_loss: 0.0211 step time: 0.2389\n", + "6/8, train_loss: 0.0176 step time: 0.2344\n", + "7/8, train_loss: 0.0160 step time: 0.2158\n", + "8/8, train_loss: 0.0190 step time: 0.2167\n", + "epoch 154 average loss: 0.0212\n", + "time consuming of epoch 154 is: 2.0955\n", + "----------\n", + "epoch 155/600\n", + "1/8, train_loss: 0.0239 step time: 0.2791\n", + "2/8, train_loss: 0.0194 step time: 0.2277\n", + "3/8, train_loss: 0.0178 step time: 0.2332\n", + "4/8, train_loss: 0.0214 step time: 0.2603\n", + "5/8, train_loss: 0.0191 step time: 0.2691\n", + "6/8, train_loss: 0.0214 step time: 0.2501\n", + "7/8, train_loss: 0.0219 step time: 0.2403\n", + "8/8, train_loss: 0.0278 step time: 0.2408\n", + "epoch 155 average loss: 0.0216\n", + "current epoch: 155 current mean dice: 0.9469 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 155 is: 3.4074\n", + "----------\n", + "epoch 156/600\n", + "1/8, train_loss: 0.0206 step time: 0.2933\n", + "2/8, train_loss: 0.0171 step time: 0.2437\n", + "3/8, train_loss: 0.0169 step time: 0.2352\n", + "4/8, train_loss: 0.0222 step time: 0.2299\n", + "5/8, train_loss: 0.0201 step time: 0.2299\n", + "6/8, train_loss: 0.0198 step time: 0.2305\n", + "7/8, train_loss: 0.0248 step time: 0.2190\n", + "8/8, train_loss: 0.0206 step time: 0.2185\n", + "epoch 156 average loss: 0.0203\n", + "time consuming of epoch 156 is: 1.9013\n", + "----------\n", + "epoch 157/600\n", + "1/8, train_loss: 0.0231 step time: 0.2772\n", + "2/8, train_loss: 0.0178 step time: 0.2292\n", + "3/8, train_loss: 0.0290 step time: 0.3113\n", + "4/8, train_loss: 0.0194 step time: 0.4848\n", + "5/8, train_loss: 0.0177 step time: 0.2184\n", + "6/8, train_loss: 0.0358 step time: 0.2142\n", + "7/8, train_loss: 0.0161 step time: 0.1997\n", + "8/8, train_loss: 0.0227 step time: 0.2015\n", + "epoch 157 average loss: 0.0227\n", + "time consuming of epoch 157 is: 2.1379\n", + "----------\n", + "epoch 158/600\n", + "1/8, train_loss: 0.0184 step time: 0.6316\n", + "2/8, train_loss: 0.0321 step time: 0.2196\n", + "3/8, train_loss: 0.0238 step time: 0.2141\n", + "4/8, train_loss: 0.0169 step time: 0.2157\n", + "5/8, train_loss: 0.0209 step time: 0.2182\n", + "6/8, train_loss: 0.0176 step time: 0.2165\n", + "7/8, train_loss: 0.0297 step time: 0.2098\n", + "8/8, train_loss: 0.0224 step time: 0.2378\n", + "epoch 158 average loss: 0.0227\n", + "time consuming of epoch 158 is: 2.1649\n", + "----------\n", + "epoch 159/600\n", + "1/8, train_loss: 0.0227 step time: 0.3156\n", + "2/8, train_loss: 0.0205 step time: 0.2572\n", + "3/8, train_loss: 0.0251 step time: 0.2584\n", + "4/8, train_loss: 0.0200 step time: 0.2521\n", + "5/8, train_loss: 0.0270 step time: 0.2520\n", + "6/8, train_loss: 0.0168 step time: 0.2447\n", + "7/8, train_loss: 0.0233 step time: 0.2416\n", + "8/8, train_loss: 0.0226 step time: 0.2383\n", + "epoch 159 average loss: 0.0223\n", + "time consuming of epoch 159 is: 2.0613\n", + "----------\n", + "epoch 160/600\n", + "1/8, train_loss: 0.0201 step time: 0.2868\n", + "2/8, train_loss: 0.0187 step time: 0.2407\n", + "3/8, train_loss: 0.0212 step time: 0.2413\n", + "4/8, train_loss: 0.0312 step time: 0.2422\n", + "5/8, train_loss: 0.0193 step time: 0.3374\n", + "6/8, train_loss: 0.0186 step time: 0.3616\n", + "7/8, train_loss: 0.0211 step time: 0.2701\n", + "8/8, train_loss: 0.0200 step time: 0.2424\n", + "epoch 160 average loss: 0.0213\n", + "current epoch: 160 current mean dice: 0.9427 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 160 is: 3.7543\n", + "----------\n", + "epoch 161/600\n", + "1/8, train_loss: 0.0186 step time: 0.2969\n", + "2/8, train_loss: 0.0196 step time: 0.2519\n", + "3/8, train_loss: 0.0212 step time: 0.2500\n", + "4/8, train_loss: 0.0252 step time: 0.2415\n", + "5/8, train_loss: 0.0225 step time: 0.2373\n", + "6/8, train_loss: 0.0230 step time: 0.2382\n", + "7/8, train_loss: 0.0212 step time: 0.2277\n", + "8/8, train_loss: 0.0417 step time: 0.2294\n", + "epoch 161 average loss: 0.0241\n", + "time consuming of epoch 161 is: 1.9741\n", + "----------\n", + "epoch 162/600\n", + "1/8, train_loss: 0.0197 step time: 0.2843\n", + "2/8, train_loss: 0.0233 step time: 0.2471\n", + "3/8, train_loss: 0.0310 step time: 0.2827\n", + "4/8, train_loss: 0.0212 step time: 0.2658\n", + "5/8, train_loss: 0.0225 step time: 0.2538\n", + "6/8, train_loss: 0.0183 step time: 0.2406\n", + "7/8, train_loss: 0.0454 step time: 0.2343\n", + "8/8, train_loss: 0.0217 step time: 0.2321\n", + "epoch 162 average loss: 0.0254\n", + "time consuming of epoch 162 is: 2.0421\n", + "----------\n", + "epoch 163/600\n", + "1/8, train_loss: 0.0188 step time: 0.2879\n", + "2/8, train_loss: 0.0225 step time: 0.2323\n", + "3/8, train_loss: 0.0372 step time: 0.3098\n", + "4/8, train_loss: 0.0234 step time: 0.3268\n", + "5/8, train_loss: 0.0397 step time: 0.2984\n", + "6/8, train_loss: 0.0481 step time: 0.2936\n", + "7/8, train_loss: 0.0245 step time: 0.2780\n", + "8/8, train_loss: 0.0279 step time: 0.2593\n", + "epoch 163 average loss: 0.0303\n", + "time consuming of epoch 163 is: 2.2876\n", + "----------\n", + "epoch 164/600\n", + "1/8, train_loss: 0.0218 step time: 0.3042\n", + "2/8, train_loss: 0.0279 step time: 0.2497\n", + "3/8, train_loss: 0.0235 step time: 0.2509\n", + "4/8, train_loss: 0.0274 step time: 0.2495\n", + "5/8, train_loss: 0.0414 step time: 0.2418\n", + "6/8, train_loss: 0.0170 step time: 0.2381\n", + "7/8, train_loss: 0.0209 step time: 0.2311\n", + "8/8, train_loss: 0.0440 step time: 0.2422\n", + "epoch 164 average loss: 0.0280\n", + "time consuming of epoch 164 is: 2.0089\n", + "----------\n", + "epoch 165/600\n", + "1/8, train_loss: 0.0242 step time: 0.2959\n", + "2/8, train_loss: 0.0278 step time: 0.2412\n", + "3/8, train_loss: 0.0246 step time: 0.2380\n", + "4/8, train_loss: 0.0279 step time: 0.2333\n", + "5/8, train_loss: 0.0258 step time: 0.2348\n", + "6/8, train_loss: 0.0816 step time: 0.2363\n", + "7/8, train_loss: 0.0208 step time: 0.2324\n", + "8/8, train_loss: 0.0205 step time: 0.2429\n", + "epoch 165 average loss: 0.0316\n", + "current epoch: 165 current mean dice: 0.8847 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 165 is: 3.2058\n", + "----------\n", + "epoch 166/600\n", + "1/8, train_loss: 0.0211 step time: 0.3104\n", + "2/8, train_loss: 0.0354 step time: 0.2633\n", + "3/8, train_loss: 0.0271 step time: 0.2558\n", + "4/8, train_loss: 0.0229 step time: 0.2486\n", + "5/8, train_loss: 0.0472 step time: 0.2471\n", + "6/8, train_loss: 0.0257 step time: 0.2406\n", + "7/8, train_loss: 0.0270 step time: 0.2303\n", + "8/8, train_loss: 0.0246 step time: 0.2275\n", + "epoch 166 average loss: 0.0289\n", + "time consuming of epoch 166 is: 2.0250\n", + "----------\n", + "epoch 167/600\n", + "1/8, train_loss: 0.0265 step time: 0.5427\n", + "2/8, train_loss: 0.0294 step time: 0.2162\n", + "3/8, train_loss: 0.0436 step time: 0.2245\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "4/8, train_loss: 0.0251 step time: 0.2375\n", + "5/8, train_loss: 0.0256 step time: 0.2377\n", + "6/8, train_loss: 0.0314 step time: 0.2338\n", + "7/8, train_loss: 0.0255 step time: 0.2285\n", + "8/8, train_loss: 0.0233 step time: 0.2531\n", + "epoch 167 average loss: 0.0288\n", + "time consuming of epoch 167 is: 2.1755\n", + "----------\n", + "epoch 168/600\n", + "1/8, train_loss: 0.0260 step time: 0.3011\n", + "2/8, train_loss: 0.0221 step time: 0.2569\n", + "3/8, train_loss: 0.0297 step time: 0.2618\n", + "4/8, train_loss: 0.0245 step time: 0.2788\n", + "5/8, train_loss: 0.0356 step time: 0.2793\n", + "6/8, train_loss: 0.0343 step time: 0.2673\n", + "7/8, train_loss: 0.0266 step time: 0.2555\n", + "8/8, train_loss: 0.0189 step time: 0.2384\n", + "epoch 168 average loss: 0.0272\n", + "time consuming of epoch 168 is: 2.1406\n", + "----------\n", + "epoch 169/600\n", + "1/8, train_loss: 0.0234 step time: 0.2891\n", + "2/8, train_loss: 0.0229 step time: 0.2429\n", + "3/8, train_loss: 0.0242 step time: 0.2404\n", + "4/8, train_loss: 0.0240 step time: 0.2381\n", + "5/8, train_loss: 0.0227 step time: 0.2401\n", + "6/8, train_loss: 0.0205 step time: 0.2616\n", + "7/8, train_loss: 0.0202 step time: 0.3767\n", + "8/8, train_loss: 0.0216 step time: 0.5020\n", + "epoch 169 average loss: 0.0225\n", + "time consuming of epoch 169 is: 2.3924\n", + "----------\n", + "epoch 170/600\n", + "1/8, train_loss: 0.0215 step time: 0.3012\n", + "2/8, train_loss: 0.0210 step time: 0.2553\n", + "3/8, train_loss: 0.0222 step time: 0.2391\n", + "4/8, train_loss: 0.0177 step time: 0.2395\n", + "5/8, train_loss: 0.0252 step time: 0.2324\n", + "6/8, train_loss: 0.0197 step time: 0.3642\n", + "7/8, train_loss: 0.0197 step time: 0.4686\n", + "8/8, train_loss: 0.0192 step time: 0.2604\n", + "epoch 170 average loss: 0.0208\n", + "current epoch: 170 current mean dice: 0.9454 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 170 is: 3.5385\n", + "----------\n", + "epoch 171/600\n", + "1/8, train_loss: 0.0207 step time: 0.3184\n", + "2/8, train_loss: 0.0224 step time: 0.2674\n", + "3/8, train_loss: 0.0183 step time: 0.2468\n", + "4/8, train_loss: 0.0191 step time: 0.2421\n", + "5/8, train_loss: 0.0207 step time: 0.2400\n", + "6/8, train_loss: 0.0310 step time: 0.3741\n", + "7/8, train_loss: 0.0199 step time: 0.5432\n", + "8/8, train_loss: 0.0185 step time: 0.2021\n", + "epoch 171 average loss: 0.0213\n", + "time consuming of epoch 171 is: 2.4353\n", + "----------\n", + "epoch 172/600\n", + "1/8, train_loss: 0.0227 step time: 0.2634\n", + "2/8, train_loss: 0.0191 step time: 0.2143\n", + "3/8, train_loss: 0.0235 step time: 0.2330\n", + "4/8, train_loss: 0.0332 step time: 0.2387\n", + "5/8, train_loss: 0.0178 step time: 0.2357\n", + "6/8, train_loss: 0.0204 step time: 0.2312\n", + "7/8, train_loss: 0.0204 step time: 0.2812\n", + "8/8, train_loss: 0.0236 step time: 0.3108\n", + "epoch 172 average loss: 0.0226\n", + "time consuming of epoch 172 is: 2.0099\n", + "----------\n", + "epoch 173/600\n", + "1/8, train_loss: 0.0220 step time: 0.3096\n", + "2/8, train_loss: 0.0192 step time: 0.2629\n", + "3/8, train_loss: 0.0188 step time: 0.2627\n", + "4/8, train_loss: 0.0214 step time: 0.2670\n", + "5/8, train_loss: 0.0178 step time: 0.2602\n", + "6/8, train_loss: 0.0219 step time: 0.2610\n", + "7/8, train_loss: 0.0205 step time: 0.2494\n", + "8/8, train_loss: 0.0172 step time: 0.2357\n", + "epoch 173 average loss: 0.0199\n", + "time consuming of epoch 173 is: 2.1102\n", + "----------\n", + "epoch 174/600\n", + "1/8, train_loss: 0.0179 step time: 0.2870\n", + "2/8, train_loss: 0.0210 step time: 0.2391\n", + "3/8, train_loss: 0.0191 step time: 0.2368\n", + "4/8, train_loss: 0.0165 step time: 0.2350\n", + "5/8, train_loss: 0.0249 step time: 0.3245\n", + "6/8, train_loss: 0.0201 step time: 0.4838\n", + "7/8, train_loss: 0.0211 step time: 0.2674\n", + "8/8, train_loss: 0.0237 step time: 0.2636\n", + "epoch 174 average loss: 0.0205\n", + "time consuming of epoch 174 is: 2.3386\n", + "----------\n", + "epoch 175/600\n", + "1/8, train_loss: 0.0206 step time: 0.2998\n", + "2/8, train_loss: 0.0154 step time: 0.2486\n", + "3/8, train_loss: 0.0198 step time: 0.2419\n", + "4/8, train_loss: 0.0206 step time: 0.2320\n", + "5/8, train_loss: 0.0170 step time: 0.2328\n", + "6/8, train_loss: 0.0182 step time: 0.2303\n", + "7/8, train_loss: 0.0252 step time: 0.2255\n", + "8/8, train_loss: 0.0254 step time: 0.3328\n", + "epoch 175 average loss: 0.0203\n", + "current epoch: 175 current mean dice: 0.9476 best mean dice: 0.9482 at epoch: 85\n", + "time consuming of epoch 175 is: 3.3557\n", + "----------\n", + "epoch 176/600\n", + "1/8, train_loss: 0.0172 step time: 0.2848\n", + "2/8, train_loss: 0.0231 step time: 0.3700\n", + "3/8, train_loss: 0.0234 step time: 0.3314\n", + "4/8, train_loss: 0.0198 step time: 0.2648\n", + "5/8, train_loss: 0.0223 step time: 0.2491\n", + "6/8, train_loss: 0.0179 step time: 0.2579\n", + "7/8, train_loss: 0.0183 step time: 0.2632\n", + "8/8, train_loss: 0.0197 step time: 0.2486\n", + "epoch 176 average loss: 0.0202\n", + "time consuming of epoch 176 is: 2.2711\n", + "----------\n", + "epoch 177/600\n", + "1/8, train_loss: 0.0205 step time: 0.2930\n", + "2/8, train_loss: 0.0198 step time: 0.2364\n", + "3/8, train_loss: 0.0198 step time: 0.2359\n", + "4/8, train_loss: 0.0193 step time: 0.2319\n", + "5/8, train_loss: 0.0182 step time: 0.2347\n", + "6/8, train_loss: 0.0223 step time: 0.2347\n", + "7/8, train_loss: 0.0183 step time: 0.2254\n", + "8/8, train_loss: 0.0172 step time: 0.2318\n", + "epoch 177 average loss: 0.0194\n", + "time consuming of epoch 177 is: 1.9252\n", + "----------\n", + "epoch 178/600\n", + "1/8, train_loss: 0.0194 step time: 0.2929\n", + "2/8, train_loss: 0.0187 step time: 0.2520\n", + "3/8, train_loss: 0.0192 step time: 0.2539\n", + "4/8, train_loss: 0.0169 step time: 0.2576\n", + "5/8, train_loss: 0.0192 step time: 0.2503\n", + "6/8, train_loss: 0.0203 step time: 0.2500\n", + "7/8, train_loss: 0.0187 step time: 0.2442\n", + "8/8, train_loss: 0.0170 step time: 0.2411\n", + "epoch 178 average loss: 0.0187\n", + "time consuming of epoch 178 is: 2.0436\n", + "----------\n", + "epoch 179/600\n", + "1/8, train_loss: 0.0220 step time: 0.2719\n", + "2/8, train_loss: 0.0172 step time: 0.2361\n", + "3/8, train_loss: 0.0172 step time: 0.2313\n", + "4/8, train_loss: 0.0170 step time: 0.3401\n", + "5/8, train_loss: 0.0190 step time: 0.5083\n", + "6/8, train_loss: 0.0188 step time: 0.2738\n", + "7/8, train_loss: 0.0213 step time: 0.2551\n", + "8/8, train_loss: 0.0217 step time: 0.2439\n", + "epoch 179 average loss: 0.0193\n", + "time consuming of epoch 179 is: 2.3620\n", + "----------\n", + "epoch 180/600\n", + "1/8, train_loss: 0.0212 step time: 0.2961\n", + "2/8, train_loss: 0.0210 step time: 0.2526\n", + "3/8, train_loss: 0.0196 step time: 0.2526\n", + "4/8, train_loss: 0.0247 step time: 0.2531\n", + "5/8, train_loss: 0.0150 step time: 0.2432\n", + "6/8, train_loss: 0.0175 step time: 0.2448\n", + "7/8, train_loss: 0.0203 step time: 0.2391\n", + "8/8, train_loss: 0.0187 step time: 0.2327\n", + "epoch 180 average loss: 0.0197\n", + "saved new best metric model\n", + "current epoch: 180 current mean dice: 0.9485 best mean dice: 0.9485 at epoch: 180\n", + "time consuming of epoch 180 is: 5.9023\n", + "----------\n", + "epoch 181/600\n", + "1/8, train_loss: 0.0204 step time: 0.2773\n", + "2/8, train_loss: 0.0207 step time: 0.2124\n", + "3/8, train_loss: 0.0226 step time: 0.2155\n", + "4/8, train_loss: 0.0160 step time: 0.2172\n", + "5/8, train_loss: 0.0184 step time: 0.2161\n", + "6/8, train_loss: 0.0167 step time: 0.2226\n", + "7/8, train_loss: 0.0230 step time: 0.2158\n", + "8/8, train_loss: 0.0192 step time: 0.2149\n", + "epoch 181 average loss: 0.0196\n", + "time consuming of epoch 181 is: 1.7943\n", + "----------\n", + "epoch 182/600\n", + "1/8, train_loss: 0.0191 step time: 0.2755\n", + "2/8, train_loss: 0.0186 step time: 0.2259\n", + "3/8, train_loss: 0.0185 step time: 0.2262\n", + "4/8, train_loss: 0.0181 step time: 0.2245\n", + "5/8, train_loss: 0.0211 step time: 0.2247\n", + "6/8, train_loss: 0.0203 step time: 0.2240\n", + "7/8, train_loss: 0.0181 step time: 0.2072\n", + "8/8, train_loss: 0.0180 step time: 0.2187\n", + "epoch 182 average loss: 0.0190\n", + "time consuming of epoch 182 is: 1.8283\n", + "----------\n", + "epoch 183/600\n", + "1/8, train_loss: 0.0181 step time: 0.2943\n", + "2/8, train_loss: 0.0254 step time: 0.2459\n", + "3/8, train_loss: 0.0192 step time: 0.2396\n", + "4/8, train_loss: 0.0203 step time: 0.2375\n", + "5/8, train_loss: 0.0176 step time: 0.2294\n", + "6/8, train_loss: 0.0180 step time: 0.3321\n", + "7/8, train_loss: 0.0195 step time: 0.4235\n", + "8/8, train_loss: 0.0181 step time: 0.2668\n", + "epoch 183 average loss: 0.0195\n", + "time consuming of epoch 183 is: 2.2706\n", + "----------\n", + "epoch 184/600\n", + "1/8, train_loss: 0.0177 step time: 0.2903\n", + "2/8, train_loss: 0.0182 step time: 0.2352\n", + "3/8, train_loss: 0.0178 step time: 0.2336\n", + "4/8, train_loss: 0.0193 step time: 0.2390\n", + "5/8, train_loss: 0.0226 step time: 0.2369\n", + "6/8, train_loss: 0.0185 step time: 0.2496\n", + "7/8, train_loss: 0.0191 step time: 0.2908\n", + "8/8, train_loss: 0.0177 step time: 0.2839\n", + "epoch 184 average loss: 0.0189\n", + "time consuming of epoch 184 is: 2.0607\n", + "----------\n", + "epoch 185/600\n", + "1/8, train_loss: 0.0211 step time: 0.3040\n", + "2/8, train_loss: 0.0228 step time: 0.2631\n", + "3/8, train_loss: 0.0188 step time: 0.2527\n", + "4/8, train_loss: 0.0178 step time: 0.2477\n", + "5/8, train_loss: 0.0179 step time: 0.2535\n", + "6/8, train_loss: 0.0207 step time: 0.2410\n", + "7/8, train_loss: 0.0157 step time: 0.2336\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "8/8, train_loss: 0.0200 step time: 0.2310\n", + "epoch 185 average loss: 0.0194\n", + "saved new best metric model\n", + "current epoch: 185 current mean dice: 0.9491 best mean dice: 0.9491 at epoch: 185\n", + "time consuming of epoch 185 is: 3.9216\n", + "----------\n", + "epoch 186/600\n", + "1/8, train_loss: 0.0146 step time: 0.2700\n", + "2/8, train_loss: 0.0174 step time: 0.2123\n", + "3/8, train_loss: 0.0191 step time: 0.2170\n", + "4/8, train_loss: 0.0185 step time: 0.2514\n", + "5/8, train_loss: 0.0220 step time: 0.2464\n", + "6/8, train_loss: 0.0255 step time: 0.2551\n", + "7/8, train_loss: 0.0193 step time: 0.2429\n", + "8/8, train_loss: 0.0209 step time: 0.3733\n", + "epoch 186 average loss: 0.0197\n", + "time consuming of epoch 186 is: 2.0708\n", + "----------\n", + "epoch 187/600\n", + "1/8, train_loss: 0.0196 step time: 0.3326\n", + "2/8, train_loss: 0.0184 step time: 0.2166\n", + "3/8, train_loss: 0.0176 step time: 0.2135\n", + "4/8, train_loss: 0.0173 step time: 0.2725\n", + "5/8, train_loss: 0.0189 step time: 0.3467\n", + "6/8, train_loss: 0.0186 step time: 0.2724\n", + "7/8, train_loss: 0.0183 step time: 0.2451\n", + "8/8, train_loss: 0.0173 step time: 0.2744\n", + "epoch 187 average loss: 0.0182\n", + "time consuming of epoch 187 is: 2.1753\n", + "----------\n", + "epoch 188/600\n", + "1/8, train_loss: 0.0176 step time: 0.3402\n", + "2/8, train_loss: 0.0194 step time: 0.2621\n", + "3/8, train_loss: 0.0163 step time: 0.2472\n", + "4/8, train_loss: 0.0184 step time: 0.2425\n", + "5/8, train_loss: 0.0184 step time: 0.2406\n", + "6/8, train_loss: 0.0236 step time: 0.2375\n", + "7/8, train_loss: 0.0211 step time: 0.2284\n", + "8/8, train_loss: 0.0202 step time: 0.2251\n", + "epoch 188 average loss: 0.0194\n", + "time consuming of epoch 188 is: 2.0250\n", + "----------\n", + "epoch 189/600\n", + "1/8, train_loss: 0.0178 step time: 0.2813\n", + "2/8, train_loss: 0.0152 step time: 0.2356\n", + "3/8, train_loss: 0.0190 step time: 0.2313\n", + "4/8, train_loss: 0.0196 step time: 0.3860\n", + "5/8, train_loss: 0.0176 step time: 0.6015\n", + "6/8, train_loss: 0.0189 step time: 0.2198\n", + "7/8, train_loss: 0.0193 step time: 0.1999\n", + "8/8, train_loss: 0.0167 step time: 0.2031\n", + "epoch 189 average loss: 0.0180\n", + "time consuming of epoch 189 is: 2.3600\n", + "----------\n", + "epoch 190/600\n", + "1/8, train_loss: 0.0196 step time: 0.3176\n", + "2/8, train_loss: 0.0165 step time: 0.2490\n", + "3/8, train_loss: 0.0178 step time: 0.2473\n", + "4/8, train_loss: 0.0163 step time: 0.2499\n", + "5/8, train_loss: 0.0171 step time: 0.2910\n", + "6/8, train_loss: 0.0236 step time: 0.3832\n", + "7/8, train_loss: 0.0175 step time: 0.2640\n", + "8/8, train_loss: 0.0201 step time: 0.2600\n", + "epoch 190 average loss: 0.0186\n", + "current epoch: 190 current mean dice: 0.9481 best mean dice: 0.9491 at epoch: 185\n", + "time consuming of epoch 190 is: 3.6325\n", + "----------\n", + "epoch 191/600\n", + "1/8, train_loss: 0.0155 step time: 0.2996\n", + "2/8, train_loss: 0.0167 step time: 0.2488\n", + "3/8, train_loss: 0.0172 step time: 0.2481\n", + "4/8, train_loss: 0.0184 step time: 0.2413\n", + "5/8, train_loss: 0.0181 step time: 0.2403\n", + "6/8, train_loss: 0.0206 step time: 0.3334\n", + "7/8, train_loss: 0.0187 step time: 0.3205\n", + "8/8, train_loss: 0.0188 step time: 0.2713\n", + "epoch 191 average loss: 0.0180\n", + "time consuming of epoch 191 is: 2.2047\n", + "----------\n", + "epoch 192/600\n", + "1/8, train_loss: 0.0150 step time: 0.2986\n", + "2/8, train_loss: 0.0177 step time: 0.2443\n", + "3/8, train_loss: 0.0155 step time: 0.2406\n", + "4/8, train_loss: 0.0216 step time: 0.2556\n", + "5/8, train_loss: 0.0208 step time: 0.2543\n", + "6/8, train_loss: 0.0203 step time: 0.2464\n", + "7/8, train_loss: 0.0187 step time: 0.2304\n", + "8/8, train_loss: 0.0162 step time: 0.2241\n", + "epoch 192 average loss: 0.0182\n", + "time consuming of epoch 192 is: 1.9957\n", + "----------\n", + "epoch 193/600\n", + "1/8, train_loss: 0.0155 step time: 0.2807\n", + "2/8, train_loss: 0.0216 step time: 0.2395\n", + "3/8, train_loss: 0.0192 step time: 0.2652\n", + "4/8, train_loss: 0.0180 step time: 0.2484\n", + "5/8, train_loss: 0.0189 step time: 0.2504\n", + "6/8, train_loss: 0.0165 step time: 0.2434\n", + "7/8, train_loss: 0.0200 step time: 0.2273\n", + "8/8, train_loss: 0.0164 step time: 0.2290\n", + "epoch 193 average loss: 0.0183\n", + "time consuming of epoch 193 is: 1.9854\n", + "----------\n", + "epoch 194/600\n", + "1/8, train_loss: 0.0219 step time: 0.2828\n", + "2/8, train_loss: 0.0160 step time: 0.2368\n", + "3/8, train_loss: 0.0180 step time: 0.2331\n", + "4/8, train_loss: 0.0202 step time: 0.2305\n", + "5/8, train_loss: 0.0227 step time: 0.2352\n", + "6/8, train_loss: 0.0161 step time: 0.2340\n", + "7/8, train_loss: 0.0172 step time: 0.2731\n", + "8/8, train_loss: 0.0198 step time: 0.3164\n", + "epoch 194 average loss: 0.0190\n", + "time consuming of epoch 194 is: 2.0433\n", + "----------\n", + "epoch 195/600\n", + "1/8, train_loss: 0.0210 step time: 0.3177\n", + "2/8, train_loss: 0.0179 step time: 0.2551\n", + "3/8, train_loss: 0.0163 step time: 0.2602\n", + "4/8, train_loss: 0.0167 step time: 0.2479\n", + "5/8, train_loss: 0.0215 step time: 0.2478\n", + "6/8, train_loss: 0.0194 step time: 0.2613\n", + "7/8, train_loss: 0.0166 step time: 0.2726\n", + "8/8, train_loss: 0.0237 step time: 0.2564\n", + "epoch 195 average loss: 0.0191\n", + "saved new best metric model\n", + "current epoch: 195 current mean dice: 0.9503 best mean dice: 0.9503 at epoch: 195\n", + "time consuming of epoch 195 is: 3.7715\n", + "----------\n", + "epoch 196/600\n", + "1/8, train_loss: 0.0152 step time: 0.2774\n", + "2/8, train_loss: 0.0173 step time: 0.2143\n", + "3/8, train_loss: 0.0215 step time: 0.2228\n", + "4/8, train_loss: 0.0185 step time: 0.2552\n", + "5/8, train_loss: 0.0217 step time: 0.2449\n", + "6/8, train_loss: 0.0178 step time: 0.2473\n", + "7/8, train_loss: 0.0186 step time: 0.2382\n", + "8/8, train_loss: 0.0192 step time: 0.4247\n", + "epoch 196 average loss: 0.0187\n", + "time consuming of epoch 196 is: 2.1274\n", + "----------\n", + "epoch 197/600\n", + "1/8, train_loss: 0.0196 step time: 0.5390\n", + "2/8, train_loss: 0.0189 step time: 0.2146\n", + "3/8, train_loss: 0.0160 step time: 0.2136\n", + "4/8, train_loss: 0.0200 step time: 0.2197\n", + "5/8, train_loss: 0.0188 step time: 0.2407\n", + "6/8, train_loss: 0.0142 step time: 0.2500\n", + "7/8, train_loss: 0.0213 step time: 0.2517\n", + "8/8, train_loss: 0.0190 step time: 0.4503\n", + "epoch 197 average loss: 0.0185\n", + "time consuming of epoch 197 is: 2.3811\n", + "----------\n", + "epoch 198/600\n", + "1/8, train_loss: 0.0231 step time: 0.3474\n", + "2/8, train_loss: 0.0175 step time: 0.2181\n", + "3/8, train_loss: 0.0154 step time: 0.2168\n", + "4/8, train_loss: 0.0175 step time: 0.2170\n", + "5/8, train_loss: 0.0185 step time: 0.2579\n", + "6/8, train_loss: 0.0188 step time: 0.3679\n", + "7/8, train_loss: 0.0199 step time: 0.3204\n", + "8/8, train_loss: 0.0161 step time: 0.2803\n", + "epoch 198 average loss: 0.0183\n", + "time consuming of epoch 198 is: 2.2272\n", + "----------\n", + "epoch 199/600\n", + "1/8, train_loss: 0.0176 step time: 0.3175\n", + "2/8, train_loss: 0.0158 step time: 0.2807\n", + "3/8, train_loss: 0.0207 step time: 0.2600\n", + "4/8, train_loss: 0.0164 step time: 0.2627\n", + "5/8, train_loss: 0.0192 step time: 0.2530\n", + "6/8, train_loss: 0.0179 step time: 0.4336\n", + "7/8, train_loss: 0.0188 step time: 0.4260\n", + "8/8, train_loss: 0.0181 step time: 0.2600\n", + "epoch 199 average loss: 0.0181\n", + "time consuming of epoch 199 is: 2.4949\n", + "----------\n", + "epoch 200/600\n", + "1/8, train_loss: 0.0188 step time: 0.2941\n", + "2/8, train_loss: 0.0226 step time: 0.2405\n", + "3/8, train_loss: 0.0142 step time: 0.2394\n", + "4/8, train_loss: 0.0141 step time: 0.2685\n", + "5/8, train_loss: 0.0188 step time: 0.2979\n", + "6/8, train_loss: 0.0214 step time: 0.2978\n", + "7/8, train_loss: 0.0161 step time: 0.2977\n", + "8/8, train_loss: 0.0180 step time: 0.2540\n", + "epoch 200 average loss: 0.0180\n", + "saved new best metric model\n", + "current epoch: 200 current mean dice: 0.9505 best mean dice: 0.9505 at epoch: 200\n", + "time consuming of epoch 200 is: 3.6138\n", + "----------\n", + "epoch 201/600\n", + "1/8, train_loss: 0.0207 step time: 0.2754\n", + "2/8, train_loss: 0.0188 step time: 0.2102\n", + "3/8, train_loss: 0.0243 step time: 0.2149\n", + "4/8, train_loss: 0.0172 step time: 0.2318\n", + "5/8, train_loss: 0.0169 step time: 0.2402\n", + "6/8, train_loss: 0.0146 step time: 0.2371\n", + "7/8, train_loss: 0.0179 step time: 0.2475\n", + "8/8, train_loss: 0.0167 step time: 0.2765\n", + "epoch 201 average loss: 0.0184\n", + "time consuming of epoch 201 is: 1.9361\n", + "----------\n", + "epoch 202/600\n", + "1/8, train_loss: 0.0180 step time: 0.3037\n", + "2/8, train_loss: 0.0167 step time: 0.2490\n", + "3/8, train_loss: 0.0186 step time: 0.2464\n", + "4/8, train_loss: 0.0205 step time: 0.2504\n", + "5/8, train_loss: 0.0181 step time: 0.3196\n", + "6/8, train_loss: 0.0197 step time: 0.4920\n", + "7/8, train_loss: 0.0214 step time: 0.2727\n", + "8/8, train_loss: 0.0184 step time: 0.2634\n", + "epoch 202 average loss: 0.0189\n", + "time consuming of epoch 202 is: 2.3987\n", + "----------\n", + "epoch 203/600\n", + "1/8, train_loss: 0.0172 step time: 0.3089\n", + "2/8, train_loss: 0.0183 step time: 0.2598\n", + "3/8, train_loss: 0.0237 step time: 0.2640\n", + "4/8, train_loss: 0.0200 step time: 0.2578\n", + "5/8, train_loss: 0.0194 step time: 0.2452\n", + "6/8, train_loss: 0.0171 step time: 0.2436\n", + "7/8, train_loss: 0.0170 step time: 0.2322\n", + "8/8, train_loss: 0.0185 step time: 0.2244\n", + "epoch 203 average loss: 0.0189\n", + "time consuming of epoch 203 is: 2.0373\n", + "----------\n", + "epoch 204/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0204 step time: 0.6395\n", + "2/8, train_loss: 0.0157 step time: 0.2181\n", + "3/8, train_loss: 0.0187 step time: 0.2167\n", + "4/8, train_loss: 0.0159 step time: 0.2137\n", + "5/8, train_loss: 0.0167 step time: 0.2185\n", + "6/8, train_loss: 0.0158 step time: 0.2524\n", + "7/8, train_loss: 0.0184 step time: 0.2762\n", + "8/8, train_loss: 0.0192 step time: 0.2544\n", + "epoch 204 average loss: 0.0176\n", + "time consuming of epoch 204 is: 2.2911\n", + "----------\n", + "epoch 205/600\n", + "1/8, train_loss: 0.0181 step time: 0.3003\n", + "2/8, train_loss: 0.0224 step time: 0.2888\n", + "3/8, train_loss: 0.0147 step time: 0.2984\n", + "4/8, train_loss: 0.0186 step time: 0.2902\n", + "5/8, train_loss: 0.0181 step time: 0.2889\n", + "6/8, train_loss: 0.0158 step time: 0.2817\n", + "7/8, train_loss: 0.0197 step time: 0.2831\n", + "8/8, train_loss: 0.0181 step time: 0.2819\n", + "epoch 205 average loss: 0.0182\n", + "saved new best metric model\n", + "current epoch: 205 current mean dice: 0.9512 best mean dice: 0.9512 at epoch: 205\n", + "time consuming of epoch 205 is: 3.9843\n", + "----------\n", + "epoch 206/600\n", + "1/8, train_loss: 0.0132 step time: 0.2817\n", + "2/8, train_loss: 0.0194 step time: 0.2169\n", + "3/8, train_loss: 0.0232 step time: 0.2142\n", + "4/8, train_loss: 0.0167 step time: 0.2128\n", + "5/8, train_loss: 0.0183 step time: 0.2233\n", + "6/8, train_loss: 0.0192 step time: 0.2298\n", + "7/8, train_loss: 0.0153 step time: 0.2488\n", + "8/8, train_loss: 0.0280 step time: 0.2830\n", + "epoch 206 average loss: 0.0192\n", + "time consuming of epoch 206 is: 1.9130\n", + "----------\n", + "epoch 207/600\n", + "1/8, train_loss: 0.0146 step time: 0.3288\n", + "2/8, train_loss: 0.0230 step time: 0.2632\n", + "3/8, train_loss: 0.0214 step time: 0.2478\n", + "4/8, train_loss: 0.0159 step time: 0.2462\n", + "5/8, train_loss: 0.0189 step time: 0.2439\n", + "6/8, train_loss: 0.0160 step time: 0.2406\n", + "7/8, train_loss: 0.0176 step time: 0.2285\n", + "8/8, train_loss: 0.0176 step time: 0.2704\n", + "epoch 207 average loss: 0.0181\n", + "time consuming of epoch 207 is: 2.0710\n", + "----------\n", + "epoch 208/600\n", + "1/8, train_loss: 0.0169 step time: 0.3138\n", + "2/8, train_loss: 0.0191 step time: 0.2663\n", + "3/8, train_loss: 0.0209 step time: 0.2587\n", + "4/8, train_loss: 0.0205 step time: 0.2532\n", + "5/8, train_loss: 0.0160 step time: 0.4429\n", + "6/8, train_loss: 0.0221 step time: 0.6164\n", + "7/8, train_loss: 0.0178 step time: 0.2041\n", + "8/8, train_loss: 0.0170 step time: 0.2436\n", + "epoch 208 average loss: 0.0188\n", + "time consuming of epoch 208 is: 2.6004\n", + "----------\n", + "epoch 209/600\n", + "1/8, train_loss: 0.0186 step time: 0.3470\n", + "2/8, train_loss: 0.0231 step time: 0.2752\n", + "3/8, train_loss: 0.0154 step time: 0.2653\n", + "4/8, train_loss: 0.0174 step time: 0.2553\n", + "5/8, train_loss: 0.0165 step time: 0.2555\n", + "6/8, train_loss: 0.0194 step time: 0.2444\n", + "7/8, train_loss: 0.0160 step time: 0.2388\n", + "8/8, train_loss: 0.0185 step time: 0.4021\n", + "epoch 209 average loss: 0.0181\n", + "time consuming of epoch 209 is: 2.2850\n", + "----------\n", + "epoch 210/600\n", + "1/8, train_loss: 0.0198 step time: 0.5640\n", + "2/8, train_loss: 0.0200 step time: 0.2927\n", + "3/8, train_loss: 0.0226 step time: 0.2801\n", + "4/8, train_loss: 0.0153 step time: 0.2575\n", + "5/8, train_loss: 0.0172 step time: 0.4163\n", + "6/8, train_loss: 0.0198 step time: 0.3339\n", + "7/8, train_loss: 0.0149 step time: 0.2660\n", + "8/8, train_loss: 0.0171 step time: 0.2661\n", + "epoch 210 average loss: 0.0183\n", + "current epoch: 210 current mean dice: 0.9496 best mean dice: 0.9512 at epoch: 205\n", + "time consuming of epoch 210 is: 3.9462\n", + "----------\n", + "epoch 211/600\n", + "1/8, train_loss: 0.0165 step time: 0.3327\n", + "2/8, train_loss: 0.0184 step time: 0.2584\n", + "3/8, train_loss: 0.0191 step time: 0.2543\n", + "4/8, train_loss: 0.0161 step time: 0.2462\n", + "5/8, train_loss: 0.0154 step time: 0.2437\n", + "6/8, train_loss: 0.0175 step time: 0.2370\n", + "7/8, train_loss: 0.0204 step time: 0.2361\n", + "8/8, train_loss: 0.0282 step time: 0.2549\n", + "epoch 211 average loss: 0.0189\n", + "time consuming of epoch 211 is: 2.0644\n", + "----------\n", + "epoch 212/600\n", + "1/8, train_loss: 0.0189 step time: 0.3485\n", + "2/8, train_loss: 0.0175 step time: 0.2841\n", + "3/8, train_loss: 0.0147 step time: 0.2692\n", + "4/8, train_loss: 0.0168 step time: 0.2609\n", + "5/8, train_loss: 0.0208 step time: 0.2541\n", + "6/8, train_loss: 0.0180 step time: 0.2502\n", + "7/8, train_loss: 0.0184 step time: 0.3909\n", + "8/8, train_loss: 0.0167 step time: 0.6285\n", + "epoch 212 average loss: 0.0177\n", + "time consuming of epoch 212 is: 2.6878\n", + "----------\n", + "epoch 213/600\n", + "1/8, train_loss: 0.0196 step time: 0.3164\n", + "2/8, train_loss: 0.0192 step time: 0.2563\n", + "3/8, train_loss: 0.0184 step time: 0.2587\n", + "4/8, train_loss: 0.0151 step time: 0.2478\n", + "5/8, train_loss: 0.0149 step time: 0.2495\n", + "6/8, train_loss: 0.0204 step time: 0.2389\n", + "7/8, train_loss: 0.0242 step time: 0.2313\n", + "8/8, train_loss: 0.0188 step time: 0.2213\n", + "epoch 213 average loss: 0.0188\n", + "time consuming of epoch 213 is: 2.0218\n", + "----------\n", + "epoch 214/600\n", + "1/8, train_loss: 0.0211 step time: 0.2855\n", + "2/8, train_loss: 0.0190 step time: 0.2399\n", + "3/8, train_loss: 0.0142 step time: 0.2356\n", + "4/8, train_loss: 0.0176 step time: 0.2329\n", + "5/8, train_loss: 0.0159 step time: 0.2299\n", + "6/8, train_loss: 0.0176 step time: 0.2385\n", + "7/8, train_loss: 0.0203 step time: 0.2422\n", + "8/8, train_loss: 0.0154 step time: 0.2486\n", + "epoch 214 average loss: 0.0176\n", + "time consuming of epoch 214 is: 1.9547\n", + "----------\n", + "epoch 215/600\n", + "1/8, train_loss: 0.0180 step time: 0.2918\n", + "2/8, train_loss: 0.0173 step time: 0.2439\n", + "3/8, train_loss: 0.0175 step time: 0.2473\n", + "4/8, train_loss: 0.0172 step time: 0.2629\n", + "5/8, train_loss: 0.0160 step time: 0.2576\n", + "6/8, train_loss: 0.0195 step time: 0.2763\n", + "7/8, train_loss: 0.0172 step time: 0.2765\n", + "8/8, train_loss: 0.0185 step time: 0.2883\n", + "epoch 215 average loss: 0.0177\n", + "saved new best metric model\n", + "current epoch: 215 current mean dice: 0.9523 best mean dice: 0.9523 at epoch: 215\n", + "time consuming of epoch 215 is: 3.6422\n", + "----------\n", + "epoch 216/600\n", + "1/8, train_loss: 0.0229 step time: 0.2737\n", + "2/8, train_loss: 0.0181 step time: 0.2144\n", + "3/8, train_loss: 0.0158 step time: 0.2232\n", + "4/8, train_loss: 0.0153 step time: 0.2271\n", + "5/8, train_loss: 0.0132 step time: 0.2377\n", + "6/8, train_loss: 0.0192 step time: 0.4043\n", + "7/8, train_loss: 0.0193 step time: 0.6272\n", + "8/8, train_loss: 0.0165 step time: 0.2721\n", + "epoch 216 average loss: 0.0175\n", + "time consuming of epoch 216 is: 2.4822\n", + "----------\n", + "epoch 217/600\n", + "1/8, train_loss: 0.0166 step time: 0.2963\n", + "2/8, train_loss: 0.0150 step time: 0.2412\n", + "3/8, train_loss: 0.0247 step time: 0.2368\n", + "4/8, train_loss: 0.0178 step time: 0.2722\n", + "5/8, train_loss: 0.0214 step time: 0.3334\n", + "6/8, train_loss: 0.0183 step time: 0.2955\n", + "7/8, train_loss: 0.0160 step time: 0.2596\n", + "8/8, train_loss: 0.0162 step time: 0.2529\n", + "epoch 217 average loss: 0.0182\n", + "time consuming of epoch 217 is: 2.1895\n", + "----------\n", + "epoch 218/600\n", + "1/8, train_loss: 0.0206 step time: 0.3029\n", + "2/8, train_loss: 0.0201 step time: 0.4029\n", + "3/8, train_loss: 0.0181 step time: 0.5110\n", + "4/8, train_loss: 0.0177 step time: 0.2719\n", + "5/8, train_loss: 0.0147 step time: 0.2611\n", + "6/8, train_loss: 0.0179 step time: 0.2454\n", + "7/8, train_loss: 0.0166 step time: 0.2351\n", + "8/8, train_loss: 0.0166 step time: 0.2386\n", + "epoch 218 average loss: 0.0178\n", + "time consuming of epoch 218 is: 2.4704\n", + "----------\n", + "epoch 219/600\n", + "1/8, train_loss: 0.0208 step time: 0.2872\n", + "2/8, train_loss: 0.0158 step time: 0.2378\n", + "3/8, train_loss: 0.0127 step time: 0.2411\n", + "4/8, train_loss: 0.0164 step time: 0.2443\n", + "5/8, train_loss: 0.0182 step time: 0.2363\n", + "6/8, train_loss: 0.0178 step time: 0.2320\n", + "7/8, train_loss: 0.0170 step time: 0.2181\n", + "8/8, train_loss: 0.0217 step time: 0.2301\n", + "epoch 219 average loss: 0.0175\n", + "time consuming of epoch 219 is: 1.9284\n", + "----------\n", + "epoch 220/600\n", + "1/8, train_loss: 0.0185 step time: 0.3008\n", + "2/8, train_loss: 0.0174 step time: 0.2583\n", + "3/8, train_loss: 0.0179 step time: 0.2566\n", + "4/8, train_loss: 0.0179 step time: 0.2518\n", + "5/8, train_loss: 0.0181 step time: 0.2557\n", + "6/8, train_loss: 0.0175 step time: 0.2507\n", + "7/8, train_loss: 0.0216 step time: 0.2441\n", + "8/8, train_loss: 0.0177 step time: 0.2441\n", + "epoch 220 average loss: 0.0183\n", + "current epoch: 220 current mean dice: 0.9508 best mean dice: 0.9523 at epoch: 215\n", + "time consuming of epoch 220 is: 3.5726\n", + "----------\n", + "epoch 221/600\n", + "1/8, train_loss: 0.0193 step time: 0.3190\n", + "2/8, train_loss: 0.0182 step time: 0.2611\n", + "3/8, train_loss: 0.0164 step time: 0.2818\n", + "4/8, train_loss: 0.0196 step time: 0.2854\n", + "5/8, train_loss: 0.0184 step time: 0.2722\n", + "6/8, train_loss: 0.0194 step time: 0.4753\n", + "7/8, train_loss: 0.0153 step time: 0.3513\n", + "8/8, train_loss: 0.0166 step time: 0.2609\n", + "epoch 221 average loss: 0.0179\n", + "time consuming of epoch 221 is: 2.5082\n", + "----------\n", + "epoch 222/600\n", + "1/8, train_loss: 0.0160 step time: 0.2944\n", + "2/8, train_loss: 0.0184 step time: 0.2445\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "3/8, train_loss: 0.0187 step time: 0.2433\n", + "4/8, train_loss: 0.0170 step time: 0.2419\n", + "5/8, train_loss: 0.0144 step time: 0.2410\n", + "6/8, train_loss: 0.0189 step time: 0.3124\n", + "7/8, train_loss: 0.0182 step time: 0.3431\n", + "8/8, train_loss: 0.0229 step time: 0.3141\n", + "epoch 222 average loss: 0.0181\n", + "time consuming of epoch 222 is: 2.2361\n", + "----------\n", + "epoch 223/600\n", + "1/8, train_loss: 0.0193 step time: 0.3084\n", + "2/8, train_loss: 0.0188 step time: 0.2543\n", + "3/8, train_loss: 0.0159 step time: 0.2503\n", + "4/8, train_loss: 0.0217 step time: 0.2881\n", + "5/8, train_loss: 0.0194 step time: 0.3382\n", + "6/8, train_loss: 0.0190 step time: 0.2867\n", + "7/8, train_loss: 0.0180 step time: 0.2661\n", + "8/8, train_loss: 0.0156 step time: 0.2530\n", + "epoch 223 average loss: 0.0185\n", + "time consuming of epoch 223 is: 2.2467\n", + "----------\n", + "epoch 224/600\n", + "1/8, train_loss: 0.0176 step time: 0.2955\n", + "2/8, train_loss: 0.0201 step time: 0.3744\n", + "3/8, train_loss: 0.0193 step time: 0.4918\n", + "4/8, train_loss: 0.0160 step time: 0.2910\n", + "5/8, train_loss: 0.0166 step time: 0.2793\n", + "6/8, train_loss: 0.0152 step time: 0.2643\n", + "7/8, train_loss: 0.0223 step time: 0.2754\n", + "8/8, train_loss: 0.0193 step time: 0.2646\n", + "epoch 224 average loss: 0.0183\n", + "time consuming of epoch 224 is: 2.5376\n", + "----------\n", + "epoch 225/600\n", + "1/8, train_loss: 0.0138 step time: 0.3011\n", + "2/8, train_loss: 0.0194 step time: 0.2484\n", + "3/8, train_loss: 0.0175 step time: 0.2445\n", + "4/8, train_loss: 0.0174 step time: 0.2750\n", + "5/8, train_loss: 0.0200 step time: 0.2749\n", + "6/8, train_loss: 0.0259 step time: 0.2639\n", + "7/8, train_loss: 0.0187 step time: 0.2572\n", + "8/8, train_loss: 0.0191 step time: 0.2415\n", + "epoch 225 average loss: 0.0190\n" ] } ], From 1180a47029b169586bee4ba9a5a839ffa60561a0 Mon Sep 17 00:00:00 2001 From: Nic Ma Date: Wed, 20 Jul 2022 20:25:44 +0800 Subject: [PATCH 3/5] [DLMED] fix flake8 Signed-off-by: Nic Ma --- acceleration/fast_training_tutorial.ipynb | 992 +++++++++++++++++++++- 1 file changed, 990 insertions(+), 2 deletions(-) diff --git a/acceleration/fast_training_tutorial.ipynb b/acceleration/fast_training_tutorial.ipynb index 9f79a6c58f..c0a7d6f624 100644 --- a/acceleration/fast_training_tutorial.ipynb +++ b/acceleration/fast_training_tutorial.ipynb @@ -129,7 +129,6 @@ " RandCropByPosNegLabeld,\n", " ScaleIntensityRanged,\n", " Spacingd,\n", - " ToDeviced,\n", ")\n", "from monai.utils import set_determinism\n", "\n", @@ -3592,7 +3591,996 @@ "6/8, train_loss: 0.0259 step time: 0.2639\n", "7/8, train_loss: 0.0187 step time: 0.2572\n", "8/8, train_loss: 0.0191 step time: 0.2415\n", - "epoch 225 average loss: 0.0190\n" + "epoch 225 average loss: 0.0190\n", + "saved new best metric model\n", + "current epoch: 225 current mean dice: 0.9527 best mean dice: 0.9527 at epoch: 225\n", + "time consuming of epoch 225 is: 3.7523\n", + "----------\n", + "epoch 226/600\n", + "1/8, train_loss: 0.0157 step time: 0.2848\n", + "2/8, train_loss: 0.0158 step time: 0.2144\n", + "3/8, train_loss: 0.0184 step time: 0.2377\n", + "4/8, train_loss: 0.0169 step time: 0.3241\n", + "5/8, train_loss: 0.0237 step time: 0.3092\n", + "6/8, train_loss: 0.0194 step time: 0.2706\n", + "7/8, train_loss: 0.0167 step time: 0.2892\n", + "8/8, train_loss: 0.0167 step time: 0.2809\n", + "epoch 226 average loss: 0.0179\n", + "time consuming of epoch 226 is: 2.2134\n", + "----------\n", + "epoch 227/600\n", + "1/8, train_loss: 0.0189 step time: 0.2989\n", + "2/8, train_loss: 0.0266 step time: 0.2535\n", + "3/8, train_loss: 0.0174 step time: 0.2655\n", + "4/8, train_loss: 0.0194 step time: 0.3005\n", + "5/8, train_loss: 0.0178 step time: 0.2575\n", + "6/8, train_loss: 0.0159 step time: 0.4424\n", + "7/8, train_loss: 0.0165 step time: 0.4034\n", + "8/8, train_loss: 0.0174 step time: 0.2544\n", + "epoch 227 average loss: 0.0187\n", + "time consuming of epoch 227 is: 2.4776\n", + "----------\n", + "epoch 228/600\n", + "1/8, train_loss: 0.0171 step time: 0.2964\n", + "2/8, train_loss: 0.0231 step time: 0.2431\n", + "3/8, train_loss: 0.0153 step time: 0.2418\n", + "4/8, train_loss: 0.0176 step time: 0.2357\n", + "5/8, train_loss: 0.0234 step time: 0.2374\n", + "6/8, train_loss: 0.0158 step time: 0.2451\n", + "7/8, train_loss: 0.0152 step time: 0.2614\n", + "8/8, train_loss: 0.0155 step time: 0.2896\n", + "epoch 228 average loss: 0.0179\n", + "time consuming of epoch 228 is: 2.0521\n", + "----------\n", + "epoch 229/600\n", + "1/8, train_loss: 0.0176 step time: 0.3414\n", + "2/8, train_loss: 0.0203 step time: 0.2941\n", + "3/8, train_loss: 0.0184 step time: 0.2783\n", + "4/8, train_loss: 0.0171 step time: 0.2708\n", + "5/8, train_loss: 0.0208 step time: 0.2531\n", + "6/8, train_loss: 0.0134 step time: 0.2704\n", + "7/8, train_loss: 0.0147 step time: 0.2787\n", + "8/8, train_loss: 0.0226 step time: 0.5408\n", + "epoch 229 average loss: 0.0181\n", + "time consuming of epoch 229 is: 2.5290\n", + "----------\n", + "epoch 230/600\n", + "1/8, train_loss: 0.0182 step time: 0.3645\n", + "2/8, train_loss: 0.0195 step time: 0.2831\n", + "3/8, train_loss: 0.0220 step time: 0.2589\n", + "4/8, train_loss: 0.0165 step time: 0.2495\n", + "5/8, train_loss: 0.0169 step time: 0.2492\n", + "6/8, train_loss: 0.0168 step time: 0.2402\n", + "7/8, train_loss: 0.0162 step time: 0.3263\n", + "8/8, train_loss: 0.0202 step time: 0.4967\n", + "epoch 230 average loss: 0.0183\n", + "current epoch: 230 current mean dice: 0.9517 best mean dice: 0.9527 at epoch: 225\n", + "time consuming of epoch 230 is: 3.6626\n", + "----------\n", + "epoch 231/600\n", + "1/8, train_loss: 0.0174 step time: 0.3016\n", + "2/8, train_loss: 0.0182 step time: 0.2497\n", + "3/8, train_loss: 0.0159 step time: 0.2410\n", + "4/8, train_loss: 0.0135 step time: 0.2420\n", + "5/8, train_loss: 0.0197 step time: 0.2408\n", + "6/8, train_loss: 0.0145 step time: 0.2492\n", + "7/8, train_loss: 0.0226 step time: 0.2759\n", + "8/8, train_loss: 0.0176 step time: 0.3449\n", + "epoch 231 average loss: 0.0174\n", + "time consuming of epoch 231 is: 2.1463\n", + "----------\n", + "epoch 232/600\n", + "1/8, train_loss: 0.0154 step time: 0.3294\n", + "2/8, train_loss: 0.0182 step time: 0.2655\n", + "3/8, train_loss: 0.0217 step time: 0.2742\n", + "4/8, train_loss: 0.0164 step time: 0.2578\n", + "5/8, train_loss: 0.0170 step time: 0.2576\n", + "6/8, train_loss: 0.0184 step time: 0.2528\n", + "7/8, train_loss: 0.0141 step time: 0.2381\n", + "8/8, train_loss: 0.0214 step time: 0.2324\n", + "epoch 232 average loss: 0.0178\n", + "time consuming of epoch 232 is: 2.1092\n", + "----------\n", + "epoch 233/600\n", + "1/8, train_loss: 0.0221 step time: 0.2817\n", + "2/8, train_loss: 0.0162 step time: 0.2326\n", + "3/8, train_loss: 0.0204 step time: 0.2326\n", + "4/8, train_loss: 0.0179 step time: 0.2353\n", + "5/8, train_loss: 0.0174 step time: 0.2328\n", + "6/8, train_loss: 0.0162 step time: 0.2574\n", + "7/8, train_loss: 0.0173 step time: 0.2748\n", + "8/8, train_loss: 0.0143 step time: 0.2680\n", + "epoch 233 average loss: 0.0177\n", + "time consuming of epoch 233 is: 2.0167\n", + "----------\n", + "epoch 234/600\n", + "1/8, train_loss: 0.0155 step time: 0.2903\n", + "2/8, train_loss: 0.0190 step time: 0.2401\n", + "3/8, train_loss: 0.0198 step time: 0.2365\n", + "4/8, train_loss: 0.0163 step time: 0.3036\n", + "5/8, train_loss: 0.0174 step time: 0.3158\n", + "6/8, train_loss: 0.0166 step time: 0.2728\n", + "7/8, train_loss: 0.0139 step time: 0.2484\n", + "8/8, train_loss: 0.0208 step time: 0.2474\n", + "epoch 234 average loss: 0.0174\n", + "time consuming of epoch 234 is: 2.1563\n", + "----------\n", + "epoch 235/600\n", + "1/8, train_loss: 0.0183 step time: 0.2995\n", + "2/8, train_loss: 0.0177 step time: 0.2728\n", + "3/8, train_loss: 0.0166 step time: 0.3005\n", + "4/8, train_loss: 0.0179 step time: 0.2730\n", + "5/8, train_loss: 0.0135 step time: 0.2619\n", + "6/8, train_loss: 0.0175 step time: 0.2542\n", + "7/8, train_loss: 0.0167 step time: 0.2565\n", + "8/8, train_loss: 0.0160 step time: 0.2511\n", + "epoch 235 average loss: 0.0168\n", + "saved new best metric model\n", + "current epoch: 235 current mean dice: 0.9530 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 235 is: 3.9097\n", + "----------\n", + "epoch 236/600\n", + "1/8, train_loss: 0.0154 step time: 0.2745\n", + "2/8, train_loss: 0.0182 step time: 0.2119\n", + "3/8, train_loss: 0.0199 step time: 0.2090\n", + "4/8, train_loss: 0.0182 step time: 0.2165\n", + "5/8, train_loss: 0.0174 step time: 0.2652\n", + "6/8, train_loss: 0.0173 step time: 0.2616\n", + "7/8, train_loss: 0.0145 step time: 0.2769\n", + "8/8, train_loss: 0.0181 step time: 0.3164\n", + "epoch 236 average loss: 0.0174\n", + "time consuming of epoch 236 is: 2.0344\n", + "----------\n", + "epoch 237/600\n", + "1/8, train_loss: 0.0164 step time: 0.3204\n", + "2/8, train_loss: 0.0147 step time: 0.2686\n", + "3/8, train_loss: 0.0159 step time: 0.4573\n", + "4/8, train_loss: 0.0191 step time: 0.5862\n", + "5/8, train_loss: 0.0179 step time: 0.2667\n", + "6/8, train_loss: 0.0177 step time: 0.2493\n", + "7/8, train_loss: 0.0147 step time: 0.2315\n", + "8/8, train_loss: 0.0180 step time: 0.2342\n", + "epoch 237 average loss: 0.0168\n", + "time consuming of epoch 237 is: 2.6157\n", + "----------\n", + "epoch 238/600\n", + "1/8, train_loss: 0.0190 step time: 0.2912\n", + "2/8, train_loss: 0.0187 step time: 0.2423\n", + "3/8, train_loss: 0.0182 step time: 0.3849\n", + "4/8, train_loss: 0.0187 step time: 0.6047\n", + "5/8, train_loss: 0.0145 step time: 0.2169\n", + "6/8, train_loss: 0.0168 step time: 0.2144\n", + "7/8, train_loss: 0.0170 step time: 0.2068\n", + "8/8, train_loss: 0.0182 step time: 0.2843\n", + "epoch 238 average loss: 0.0176\n", + "time consuming of epoch 238 is: 2.4469\n", + "----------\n", + "epoch 239/600\n", + "1/8, train_loss: 0.0139 step time: 0.3568\n", + "2/8, train_loss: 0.0170 step time: 0.2754\n", + "3/8, train_loss: 0.0209 step time: 0.2814\n", + "4/8, train_loss: 0.0213 step time: 0.2771\n", + "5/8, train_loss: 0.0253 step time: 0.2510\n", + "6/8, train_loss: 0.0182 step time: 0.2459\n", + "7/8, train_loss: 0.0183 step time: 0.4098\n", + "8/8, train_loss: 0.0183 step time: 0.6277\n", + "epoch 239 average loss: 0.0191\n", + "time consuming of epoch 239 is: 2.7266\n", + "----------\n", + "epoch 240/600\n", + "1/8, train_loss: 0.0144 step time: 0.3156\n", + "2/8, train_loss: 0.0212 step time: 0.2539\n", + "3/8, train_loss: 0.0163 step time: 0.2454\n", + "4/8, train_loss: 0.0186 step time: 0.2369\n", + "5/8, train_loss: 0.0139 step time: 0.2403\n", + "6/8, train_loss: 0.0199 step time: 0.2400\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "7/8, train_loss: 0.0181 step time: 0.2271\n", + "8/8, train_loss: 0.0164 step time: 0.2298\n", + "epoch 240 average loss: 0.0174\n", + "current epoch: 240 current mean dice: 0.9499 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 240 is: 3.3993\n", + "----------\n", + "epoch 241/600\n", + "1/8, train_loss: 0.0152 step time: 0.3111\n", + "2/8, train_loss: 0.0172 step time: 0.2533\n", + "3/8, train_loss: 0.0158 step time: 0.2476\n", + "4/8, train_loss: 0.0187 step time: 0.2453\n", + "5/8, train_loss: 0.0156 step time: 0.2429\n", + "6/8, train_loss: 0.0178 step time: 0.2386\n", + "7/8, train_loss: 0.0244 step time: 0.2206\n", + "8/8, train_loss: 0.0206 step time: 0.2238\n", + "epoch 241 average loss: 0.0182\n", + "time consuming of epoch 241 is: 1.9843\n", + "----------\n", + "epoch 242/600\n", + "1/8, train_loss: 0.0195 step time: 0.6323\n", + "2/8, train_loss: 0.0193 step time: 0.2193\n", + "3/8, train_loss: 0.0180 step time: 0.2135\n", + "4/8, train_loss: 0.0143 step time: 0.4217\n", + "5/8, train_loss: 0.0233 step time: 0.3345\n", + "6/8, train_loss: 0.0186 step time: 0.2866\n", + "7/8, train_loss: 0.0194 step time: 0.2551\n", + "8/8, train_loss: 0.0158 step time: 0.2509\n", + "epoch 242 average loss: 0.0185\n", + "time consuming of epoch 242 is: 2.6153\n", + "----------\n", + "epoch 243/600\n", + "1/8, train_loss: 0.0208 step time: 0.3013\n", + "2/8, train_loss: 0.0204 step time: 0.2461\n", + "3/8, train_loss: 0.0175 step time: 0.2471\n", + "4/8, train_loss: 0.0163 step time: 0.3207\n", + "5/8, train_loss: 0.0186 step time: 0.2976\n", + "6/8, train_loss: 0.0234 step time: 0.2776\n", + "7/8, train_loss: 0.0142 step time: 0.2535\n", + "8/8, train_loss: 0.0227 step time: 0.2491\n", + "epoch 243 average loss: 0.0192\n", + "time consuming of epoch 243 is: 2.1944\n", + "----------\n", + "epoch 244/600\n", + "1/8, train_loss: 0.0184 step time: 0.3041\n", + "2/8, train_loss: 0.0180 step time: 0.2516\n", + "3/8, train_loss: 0.0193 step time: 0.2477\n", + "4/8, train_loss: 0.0213 step time: 0.2443\n", + "5/8, train_loss: 0.0125 step time: 0.2433\n", + "6/8, train_loss: 0.0174 step time: 0.2472\n", + "7/8, train_loss: 0.0170 step time: 0.3434\n", + "8/8, train_loss: 0.0278 step time: 0.4411\n", + "epoch 244 average loss: 0.0190\n", + "time consuming of epoch 244 is: 2.3241\n", + "----------\n", + "epoch 245/600\n", + "1/8, train_loss: 0.0164 step time: 0.3014\n", + "2/8, train_loss: 0.0157 step time: 0.2591\n", + "3/8, train_loss: 0.0196 step time: 0.2474\n", + "4/8, train_loss: 0.0169 step time: 0.2433\n", + "5/8, train_loss: 0.0170 step time: 0.2402\n", + "6/8, train_loss: 0.0165 step time: 0.2396\n", + "7/8, train_loss: 0.0179 step time: 0.2311\n", + "8/8, train_loss: 0.0182 step time: 0.2908\n", + "epoch 245 average loss: 0.0173\n", + "current epoch: 245 current mean dice: 0.9440 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 245 is: 3.5798\n", + "----------\n", + "epoch 246/600\n", + "1/8, train_loss: 0.0163 step time: 0.3161\n", + "2/8, train_loss: 0.0204 step time: 0.2524\n", + "3/8, train_loss: 0.0179 step time: 0.2376\n", + "4/8, train_loss: 0.0184 step time: 0.2357\n", + "5/8, train_loss: 0.0207 step time: 0.2375\n", + "6/8, train_loss: 0.0187 step time: 0.4057\n", + "7/8, train_loss: 0.0181 step time: 0.5498\n", + "8/8, train_loss: 0.0203 step time: 0.2054\n", + "epoch 246 average loss: 0.0188\n", + "time consuming of epoch 246 is: 2.4416\n", + "----------\n", + "epoch 247/600\n", + "1/8, train_loss: 0.0183 step time: 0.2626\n", + "2/8, train_loss: 0.0194 step time: 0.2116\n", + "3/8, train_loss: 0.0195 step time: 0.2131\n", + "4/8, train_loss: 0.0204 step time: 0.2156\n", + "5/8, train_loss: 0.0204 step time: 0.2394\n", + "6/8, train_loss: 0.0163 step time: 0.3872\n", + "7/8, train_loss: 0.0174 step time: 0.4289\n", + "8/8, train_loss: 0.0215 step time: 0.2575\n", + "epoch 247 average loss: 0.0191\n", + "time consuming of epoch 247 is: 2.2176\n", + "----------\n", + "epoch 248/600\n", + "1/8, train_loss: 0.0158 step time: 0.2884\n", + "2/8, train_loss: 0.0178 step time: 0.3003\n", + "3/8, train_loss: 0.0223 step time: 0.2871\n", + "4/8, train_loss: 0.0186 step time: 0.2734\n", + "5/8, train_loss: 0.0171 step time: 0.2558\n", + "6/8, train_loss: 0.0179 step time: 0.2763\n", + "7/8, train_loss: 0.0152 step time: 0.2638\n", + "8/8, train_loss: 0.0168 step time: 0.2499\n", + "epoch 248 average loss: 0.0177\n", + "time consuming of epoch 248 is: 2.1966\n", + "----------\n", + "epoch 249/600\n", + "1/8, train_loss: 0.0162 step time: 0.2879\n", + "2/8, train_loss: 0.0171 step time: 0.2334\n", + "3/8, train_loss: 0.0153 step time: 0.2393\n", + "4/8, train_loss: 0.0142 step time: 0.2375\n", + "5/8, train_loss: 0.0244 step time: 0.2543\n", + "6/8, train_loss: 0.0173 step time: 0.3753\n", + "7/8, train_loss: 0.0172 step time: 0.4558\n", + "8/8, train_loss: 0.0191 step time: 0.2553\n", + "epoch 249 average loss: 0.0176\n", + "time consuming of epoch 249 is: 2.3404\n", + "----------\n", + "epoch 250/600\n", + "1/8, train_loss: 0.0167 step time: 0.3045\n", + "2/8, train_loss: 0.0163 step time: 0.2436\n", + "3/8, train_loss: 0.0175 step time: 0.2423\n", + "4/8, train_loss: 0.0190 step time: 0.2403\n", + "5/8, train_loss: 0.0184 step time: 0.2319\n", + "6/8, train_loss: 0.0196 step time: 0.2281\n", + "7/8, train_loss: 0.0190 step time: 0.2154\n", + "8/8, train_loss: 0.0166 step time: 0.3999\n", + "epoch 250 average loss: 0.0179\n", + "current epoch: 250 current mean dice: 0.9483 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 250 is: 3.5188\n", + "----------\n", + "epoch 251/600\n", + "1/8, train_loss: 0.0210 step time: 0.2986\n", + "2/8, train_loss: 0.0161 step time: 0.2749\n", + "3/8, train_loss: 0.0161 step time: 0.2771\n", + "4/8, train_loss: 0.0168 step time: 0.2590\n", + "5/8, train_loss: 0.0176 step time: 0.2909\n", + "6/8, train_loss: 0.0179 step time: 0.2959\n", + "7/8, train_loss: 0.0194 step time: 0.2758\n", + "8/8, train_loss: 0.0183 step time: 0.2550\n", + "epoch 251 average loss: 0.0179\n", + "time consuming of epoch 251 is: 2.2285\n", + "----------\n", + "epoch 252/600\n", + "1/8, train_loss: 0.0144 step time: 0.2929\n", + "2/8, train_loss: 0.0210 step time: 0.2449\n", + "3/8, train_loss: 0.0166 step time: 0.2416\n", + "4/8, train_loss: 0.0183 step time: 0.4123\n", + "5/8, train_loss: 0.0182 step time: 0.6209\n", + "6/8, train_loss: 0.0155 step time: 0.2642\n", + "7/8, train_loss: 0.0183 step time: 0.2396\n", + "8/8, train_loss: 0.0150 step time: 0.2298\n", + "epoch 252 average loss: 0.0171\n", + "time consuming of epoch 252 is: 2.5475\n", + "----------\n", + "epoch 253/600\n", + "1/8, train_loss: 0.0175 step time: 0.2784\n", + "2/8, train_loss: 0.0176 step time: 0.2334\n", + "3/8, train_loss: 0.0178 step time: 0.2316\n", + "4/8, train_loss: 0.0184 step time: 0.2283\n", + "5/8, train_loss: 0.0178 step time: 0.2287\n", + "6/8, train_loss: 0.0153 step time: 0.2319\n", + "7/8, train_loss: 0.0201 step time: 0.2520\n", + "8/8, train_loss: 0.0154 step time: 0.3560\n", + "epoch 253 average loss: 0.0175\n", + "time consuming of epoch 253 is: 2.0418\n", + "----------\n", + "epoch 254/600\n", + "1/8, train_loss: 0.0148 step time: 0.3476\n", + "2/8, train_loss: 0.0177 step time: 0.2642\n", + "3/8, train_loss: 0.0182 step time: 0.2503\n", + "4/8, train_loss: 0.0181 step time: 0.2475\n", + "5/8, train_loss: 0.0150 step time: 0.2960\n", + "6/8, train_loss: 0.0205 step time: 0.3475\n", + "7/8, train_loss: 0.0178 step time: 0.2778\n", + "8/8, train_loss: 0.0181 step time: 0.2618\n", + "epoch 254 average loss: 0.0175\n", + "time consuming of epoch 254 is: 2.2941\n", + "----------\n", + "epoch 255/600\n", + "1/8, train_loss: 0.0200 step time: 0.3103\n", + "2/8, train_loss: 0.0145 step time: 0.2528\n", + "3/8, train_loss: 0.0155 step time: 0.2943\n", + "4/8, train_loss: 0.0162 step time: 0.3328\n", + "5/8, train_loss: 0.0163 step time: 0.2843\n", + "6/8, train_loss: 0.0184 step time: 0.2759\n", + "7/8, train_loss: 0.0149 step time: 0.2707\n", + "8/8, train_loss: 0.0198 step time: 0.2616\n", + "epoch 255 average loss: 0.0170\n", + "current epoch: 255 current mean dice: 0.9523 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 255 is: 3.7885\n", + "----------\n", + "epoch 256/600\n", + "1/8, train_loss: 0.0167 step time: 0.3043\n", + "2/8, train_loss: 0.0195 step time: 0.2627\n", + "3/8, train_loss: 0.0170 step time: 0.2620\n", + "4/8, train_loss: 0.0185 step time: 0.3830\n", + "5/8, train_loss: 0.0193 step time: 0.3290\n", + "6/8, train_loss: 0.0159 step time: 0.3143\n", + "7/8, train_loss: 0.0191 step time: 0.2651\n", + "8/8, train_loss: 0.0190 step time: 0.2613\n", + "epoch 256 average loss: 0.0181\n", + "time consuming of epoch 256 is: 2.3828\n", + "----------\n", + "epoch 257/600\n", + "1/8, train_loss: 0.0176 step time: 0.2935\n", + "2/8, train_loss: 0.0174 step time: 0.2490\n", + "3/8, train_loss: 0.0183 step time: 0.2557\n", + "4/8, train_loss: 0.0187 step time: 0.2483\n", + "5/8, train_loss: 0.0174 step time: 0.2429\n", + "6/8, train_loss: 0.0182 step time: 0.2475\n", + "7/8, train_loss: 0.0185 step time: 0.2239\n", + "8/8, train_loss: 0.0170 step time: 0.2273\n", + "epoch 257 average loss: 0.0179\n", + "time consuming of epoch 257 is: 1.9897\n", + "----------\n", + "epoch 258/600\n", + "1/8, train_loss: 0.0186 step time: 0.2825\n", + "2/8, train_loss: 0.0188 step time: 0.2329\n", + "3/8, train_loss: 0.0144 step time: 0.2324\n", + "4/8, train_loss: 0.0157 step time: 0.2316\n", + "5/8, train_loss: 0.0193 step time: 0.2416\n", + "6/8, train_loss: 0.0156 step time: 0.2640\n", + "7/8, train_loss: 0.0155 step time: 0.2920\n", + "8/8, train_loss: 0.0171 step time: 0.2978\n", + "epoch 258 average loss: 0.0169\n", + "time consuming of epoch 258 is: 2.0764\n", + "----------\n", + "epoch 259/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0167 step time: 0.3402\n", + "2/8, train_loss: 0.0161 step time: 0.2924\n", + "3/8, train_loss: 0.0143 step time: 0.3054\n", + "4/8, train_loss: 0.0189 step time: 0.2922\n", + "5/8, train_loss: 0.0170 step time: 0.2683\n", + "6/8, train_loss: 0.0154 step time: 0.2512\n", + "7/8, train_loss: 0.0181 step time: 0.2494\n", + "8/8, train_loss: 0.0188 step time: 0.2938\n", + "epoch 259 average loss: 0.0169\n", + "time consuming of epoch 259 is: 2.2943\n", + "----------\n", + "epoch 260/600\n", + "1/8, train_loss: 0.0160 step time: 0.3622\n", + "2/8, train_loss: 0.0169 step time: 0.2763\n", + "3/8, train_loss: 0.0165 step time: 0.2651\n", + "4/8, train_loss: 0.0165 step time: 0.2616\n", + "5/8, train_loss: 0.0166 step time: 0.2502\n", + "6/8, train_loss: 0.0140 step time: 0.2532\n", + "7/8, train_loss: 0.0160 step time: 0.2442\n", + "8/8, train_loss: 0.0178 step time: 0.4176\n", + "epoch 260 average loss: 0.0163\n", + "current epoch: 260 current mean dice: 0.9523 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 260 is: 3.6430\n", + "----------\n", + "epoch 261/600\n", + "1/8, train_loss: 0.0160 step time: 0.3416\n", + "2/8, train_loss: 0.0201 step time: 0.3069\n", + "3/8, train_loss: 0.0178 step time: 0.3125\n", + "4/8, train_loss: 0.0176 step time: 0.3092\n", + "5/8, train_loss: 0.0136 step time: 0.2916\n", + "6/8, train_loss: 0.0184 step time: 0.2795\n", + "7/8, train_loss: 0.0156 step time: 0.2717\n", + "8/8, train_loss: 0.0166 step time: 0.2499\n", + "epoch 261 average loss: 0.0170\n", + "time consuming of epoch 261 is: 2.3641\n", + "----------\n", + "epoch 262/600\n", + "1/8, train_loss: 0.0168 step time: 0.2938\n", + "2/8, train_loss: 0.0179 step time: 0.2449\n", + "3/8, train_loss: 0.0180 step time: 0.3911\n", + "4/8, train_loss: 0.0141 step time: 0.4980\n", + "5/8, train_loss: 0.0156 step time: 0.2696\n", + "6/8, train_loss: 0.0208 step time: 0.2525\n", + "7/8, train_loss: 0.0164 step time: 0.2392\n", + "8/8, train_loss: 0.0163 step time: 0.2289\n", + "epoch 262 average loss: 0.0170\n", + "time consuming of epoch 262 is: 2.4194\n", + "----------\n", + "epoch 263/600\n", + "1/8, train_loss: 0.0162 step time: 0.2865\n", + "2/8, train_loss: 0.0134 step time: 0.2324\n", + "3/8, train_loss: 0.0199 step time: 0.2383\n", + "4/8, train_loss: 0.0157 step time: 0.2404\n", + "5/8, train_loss: 0.0182 step time: 0.2315\n", + "6/8, train_loss: 0.0154 step time: 0.2312\n", + "7/8, train_loss: 0.0176 step time: 0.2177\n", + "8/8, train_loss: 0.0188 step time: 0.2492\n", + "epoch 263 average loss: 0.0169\n", + "time consuming of epoch 263 is: 1.9288\n", + "----------\n", + "epoch 264/600\n", + "1/8, train_loss: 0.0206 step time: 0.3219\n", + "2/8, train_loss: 0.0181 step time: 0.2678\n", + "3/8, train_loss: 0.0167 step time: 0.2560\n", + "4/8, train_loss: 0.0149 step time: 0.2520\n", + "5/8, train_loss: 0.0162 step time: 0.2385\n", + "6/8, train_loss: 0.0139 step time: 0.2319\n", + "7/8, train_loss: 0.0163 step time: 0.2212\n", + "8/8, train_loss: 0.0145 step time: 0.2158\n", + "epoch 264 average loss: 0.0164\n", + "time consuming of epoch 264 is: 2.0066\n", + "----------\n", + "epoch 265/600\n", + "1/8, train_loss: 0.0159 step time: 0.2807\n", + "2/8, train_loss: 0.0163 step time: 0.2269\n", + "3/8, train_loss: 0.0149 step time: 0.2322\n", + "4/8, train_loss: 0.0181 step time: 0.2257\n", + "5/8, train_loss: 0.0171 step time: 0.2455\n", + "6/8, train_loss: 0.0195 step time: 0.2532\n", + "7/8, train_loss: 0.0163 step time: 0.2464\n", + "8/8, train_loss: 0.0157 step time: 0.2430\n", + "epoch 265 average loss: 0.0167\n", + "current epoch: 265 current mean dice: 0.9524 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 265 is: 3.3196\n", + "----------\n", + "epoch 266/600\n", + "1/8, train_loss: 0.0138 step time: 0.3507\n", + "2/8, train_loss: 0.0173 step time: 0.2978\n", + "3/8, train_loss: 0.0201 step time: 0.2698\n", + "4/8, train_loss: 0.0172 step time: 0.2602\n", + "5/8, train_loss: 0.0153 step time: 0.2612\n", + "6/8, train_loss: 0.0169 step time: 0.2544\n", + "7/8, train_loss: 0.0156 step time: 0.2506\n", + "8/8, train_loss: 0.0283 step time: 0.2605\n", + "epoch 266 average loss: 0.0181\n", + "time consuming of epoch 266 is: 2.2064\n", + "----------\n", + "epoch 267/600\n", + "1/8, train_loss: 0.0169 step time: 0.2887\n", + "2/8, train_loss: 0.0155 step time: 0.2661\n", + "3/8, train_loss: 0.0199 step time: 0.2780\n", + "4/8, train_loss: 0.0160 step time: 0.2589\n", + "5/8, train_loss: 0.0168 step time: 0.2629\n", + "6/8, train_loss: 0.0177 step time: 0.2508\n", + "7/8, train_loss: 0.0148 step time: 0.2433\n", + "8/8, train_loss: 0.0167 step time: 0.2440\n", + "epoch 267 average loss: 0.0168\n", + "time consuming of epoch 267 is: 2.0940\n", + "----------\n", + "epoch 268/600\n", + "1/8, train_loss: 0.0171 step time: 0.6238\n", + "2/8, train_loss: 0.0209 step time: 0.2147\n", + "3/8, train_loss: 0.0174 step time: 0.2239\n", + "4/8, train_loss: 0.0163 step time: 0.4438\n", + "5/8, train_loss: 0.0202 step time: 0.2934\n", + "6/8, train_loss: 0.0167 step time: 0.2667\n", + "7/8, train_loss: 0.0151 step time: 0.2512\n", + "8/8, train_loss: 0.0163 step time: 0.2483\n", + "epoch 268 average loss: 0.0175\n", + "time consuming of epoch 268 is: 2.5670\n", + "----------\n", + "epoch 269/600\n", + "1/8, train_loss: 0.0202 step time: 0.2834\n", + "2/8, train_loss: 0.0189 step time: 0.4035\n", + "3/8, train_loss: 0.0157 step time: 0.4582\n", + "4/8, train_loss: 0.0157 step time: 0.2140\n", + "5/8, train_loss: 0.0163 step time: 0.2112\n", + "6/8, train_loss: 0.0142 step time: 0.2138\n", + "7/8, train_loss: 0.0191 step time: 0.2103\n", + "8/8, train_loss: 0.0189 step time: 0.2163\n", + "epoch 269 average loss: 0.0174\n", + "time consuming of epoch 269 is: 2.2119\n", + "----------\n", + "epoch 270/600\n", + "1/8, train_loss: 0.0173 step time: 0.2888\n", + "2/8, train_loss: 0.0152 step time: 0.2455\n", + "3/8, train_loss: 0.0166 step time: 0.2437\n", + "4/8, train_loss: 0.0192 step time: 0.2452\n", + "5/8, train_loss: 0.0167 step time: 0.2447\n", + "6/8, train_loss: 0.0170 step time: 0.3010\n", + "7/8, train_loss: 0.0186 step time: 0.4381\n", + "8/8, train_loss: 0.0184 step time: 0.2728\n", + "epoch 270 average loss: 0.0174\n", + "current epoch: 270 current mean dice: 0.9522 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 270 is: 3.7816\n", + "----------\n", + "epoch 271/600\n", + "1/8, train_loss: 0.0167 step time: 0.3107\n", + "2/8, train_loss: 0.0193 step time: 0.2550\n", + "3/8, train_loss: 0.0212 step time: 0.2459\n", + "4/8, train_loss: 0.0168 step time: 0.2522\n", + "5/8, train_loss: 0.0169 step time: 0.2564\n", + "6/8, train_loss: 0.0133 step time: 0.2783\n", + "7/8, train_loss: 0.0160 step time: 0.2631\n", + "8/8, train_loss: 0.0147 step time: 0.2509\n", + "epoch 271 average loss: 0.0169\n", + "time consuming of epoch 271 is: 2.1135\n", + "----------\n", + "epoch 272/600\n", + "1/8, train_loss: 0.0166 step time: 0.3020\n", + "2/8, train_loss: 0.0170 step time: 0.2575\n", + "3/8, train_loss: 0.0182 step time: 0.2498\n", + "4/8, train_loss: 0.0152 step time: 0.2477\n", + "5/8, train_loss: 0.0165 step time: 0.2435\n", + "6/8, train_loss: 0.0203 step time: 0.2448\n", + "7/8, train_loss: 0.0135 step time: 0.2396\n", + "8/8, train_loss: 0.0222 step time: 0.2286\n", + "epoch 272 average loss: 0.0174\n", + "time consuming of epoch 272 is: 2.0147\n", + "----------\n", + "epoch 273/600\n", + "1/8, train_loss: 0.0158 step time: 0.2764\n", + "2/8, train_loss: 0.0171 step time: 0.2319\n", + "3/8, train_loss: 0.0150 step time: 0.2354\n", + "4/8, train_loss: 0.0189 step time: 0.3629\n", + "5/8, train_loss: 0.0161 step time: 0.3019\n", + "6/8, train_loss: 0.0148 step time: 0.2657\n", + "7/8, train_loss: 0.0174 step time: 0.2588\n", + "8/8, train_loss: 0.0185 step time: 0.2437\n", + "epoch 273 average loss: 0.0167\n", + "time consuming of epoch 273 is: 2.1778\n", + "----------\n", + "epoch 274/600\n", + "1/8, train_loss: 0.0209 step time: 0.2909\n", + "2/8, train_loss: 0.0169 step time: 0.2702\n", + "3/8, train_loss: 0.0157 step time: 0.3241\n", + "4/8, train_loss: 0.0162 step time: 0.2899\n", + "5/8, train_loss: 0.0181 step time: 0.2572\n", + "6/8, train_loss: 0.0182 step time: 0.2526\n", + "7/8, train_loss: 0.0211 step time: 0.2482\n", + "8/8, train_loss: 0.0155 step time: 0.2476\n", + "epoch 274 average loss: 0.0178\n", + "time consuming of epoch 274 is: 2.1821\n", + "----------\n", + "epoch 275/600\n", + "1/8, train_loss: 0.0169 step time: 0.2977\n", + "2/8, train_loss: 0.0184 step time: 0.2489\n", + "3/8, train_loss: 0.0168 step time: 0.2445\n", + "4/8, train_loss: 0.0191 step time: 0.2474\n", + "5/8, train_loss: 0.0153 step time: 0.2376\n", + "6/8, train_loss: 0.0180 step time: 0.3873\n", + "7/8, train_loss: 0.0157 step time: 0.6222\n", + "8/8, train_loss: 0.0166 step time: 0.2717\n", + "epoch 275 average loss: 0.0171\n", + "current epoch: 275 current mean dice: 0.9518 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 275 is: 3.6783\n", + "----------\n", + "epoch 276/600\n", + "1/8, train_loss: 0.0218 step time: 0.3200\n", + "2/8, train_loss: 0.0154 step time: 0.2693\n", + "3/8, train_loss: 0.0168 step time: 0.2602\n", + "4/8, train_loss: 0.0155 step time: 0.2495\n", + "5/8, train_loss: 0.0177 step time: 0.2433\n", + "6/8, train_loss: 0.0182 step time: 0.2493\n", + "7/8, train_loss: 0.0190 step time: 0.2406\n", + "8/8, train_loss: 0.0171 step time: 0.2339\n", + "epoch 276 average loss: 0.0177\n", + "time consuming of epoch 276 is: 2.0672\n", + "----------\n", + "epoch 277/600\n", + "1/8, train_loss: 0.0214 step time: 0.2885\n", + "2/8, train_loss: 0.0170 step time: 0.2579\n", + "3/8, train_loss: 0.0255 step time: 0.2771\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "4/8, train_loss: 0.0164 step time: 0.2882\n", + "5/8, train_loss: 0.0149 step time: 0.2747\n", + "6/8, train_loss: 0.0201 step time: 0.2671\n", + "7/8, train_loss: 0.0133 step time: 0.2469\n", + "8/8, train_loss: 0.0156 step time: 0.2377\n", + "epoch 277 average loss: 0.0180\n", + "time consuming of epoch 277 is: 2.1396\n", + "----------\n", + "epoch 278/600\n", + "1/8, train_loss: 0.0165 step time: 0.2887\n", + "2/8, train_loss: 0.0179 step time: 0.2401\n", + "3/8, train_loss: 0.0163 step time: 0.2426\n", + "4/8, train_loss: 0.0193 step time: 0.2507\n", + "5/8, train_loss: 0.0146 step time: 0.4353\n", + "6/8, train_loss: 0.0171 step time: 0.4717\n", + "7/8, train_loss: 0.0204 step time: 0.2626\n", + "8/8, train_loss: 0.0167 step time: 0.2533\n", + "epoch 278 average loss: 0.0173\n", + "time consuming of epoch 278 is: 2.4465\n", + "----------\n", + "epoch 279/600\n", + "1/8, train_loss: 0.0170 step time: 0.3003\n", + "2/8, train_loss: 0.0173 step time: 0.2510\n", + "3/8, train_loss: 0.0159 step time: 0.2473\n", + "4/8, train_loss: 0.0156 step time: 0.2394\n", + "5/8, train_loss: 0.0220 step time: 0.2391\n", + "6/8, train_loss: 0.0184 step time: 0.3946\n", + "7/8, train_loss: 0.0168 step time: 0.5849\n", + "8/8, train_loss: 0.0179 step time: 0.2092\n", + "epoch 279 average loss: 0.0176\n", + "time consuming of epoch 279 is: 2.4672\n", + "----------\n", + "epoch 280/600\n", + "1/8, train_loss: 0.0217 step time: 0.2647\n", + "2/8, train_loss: 0.0201 step time: 0.2179\n", + "3/8, train_loss: 0.0188 step time: 0.4956\n", + "4/8, train_loss: 0.0155 step time: 0.2193\n", + "5/8, train_loss: 0.0150 step time: 0.2143\n", + "6/8, train_loss: 0.0204 step time: 0.2149\n", + "7/8, train_loss: 0.0182 step time: 0.1999\n", + "8/8, train_loss: 0.0139 step time: 0.2043\n", + "epoch 280 average loss: 0.0180\n", + "current epoch: 280 current mean dice: 0.9525 best mean dice: 0.9530 at epoch: 235\n", + "time consuming of epoch 280 is: 3.3596\n", + "----------\n", + "epoch 281/600\n", + "1/8, train_loss: 0.0162 step time: 0.3524\n", + "2/8, train_loss: 0.0148 step time: 0.3024\n", + "3/8, train_loss: 0.0173 step time: 0.2768\n", + "4/8, train_loss: 0.0188 step time: 0.2698\n", + "5/8, train_loss: 0.0175 step time: 0.2598\n", + "6/8, train_loss: 0.0174 step time: 0.2558\n", + "7/8, train_loss: 0.0184 step time: 0.2739\n", + "8/8, train_loss: 0.0157 step time: 0.3169\n", + "epoch 281 average loss: 0.0170\n", + "time consuming of epoch 281 is: 2.3091\n", + "----------\n", + "epoch 282/600\n", + "1/8, train_loss: 0.0154 step time: 0.3300\n", + "2/8, train_loss: 0.0188 step time: 0.2780\n", + "3/8, train_loss: 0.0207 step time: 0.2542\n", + "4/8, train_loss: 0.0164 step time: 0.2520\n", + "5/8, train_loss: 0.0172 step time: 0.2524\n", + "6/8, train_loss: 0.0144 step time: 0.2522\n", + "7/8, train_loss: 0.0171 step time: 0.2993\n", + "8/8, train_loss: 0.0165 step time: 0.3187\n", + "epoch 282 average loss: 0.0171\n", + "time consuming of epoch 282 is: 2.2381\n", + "----------\n", + "epoch 283/600\n", + "1/8, train_loss: 0.0154 step time: 0.3011\n", + "2/8, train_loss: 0.0170 step time: 0.2539\n", + "3/8, train_loss: 0.0165 step time: 0.2533\n", + "4/8, train_loss: 0.0226 step time: 0.3942\n", + "5/8, train_loss: 0.0172 step time: 0.3403\n", + "6/8, train_loss: 0.0172 step time: 0.2794\n", + "7/8, train_loss: 0.0190 step time: 0.2684\n", + "8/8, train_loss: 0.0144 step time: 0.2546\n", + "epoch 283 average loss: 0.0174\n", + "time consuming of epoch 283 is: 2.3466\n", + "----------\n", + "epoch 284/600\n", + "1/8, train_loss: 0.0156 step time: 0.2897\n", + "2/8, train_loss: 0.0156 step time: 0.2358\n", + "3/8, train_loss: 0.0179 step time: 0.2372\n", + "4/8, train_loss: 0.0193 step time: 0.2351\n", + "5/8, train_loss: 0.0215 step time: 0.2469\n", + "6/8, train_loss: 0.0138 step time: 0.2587\n", + "7/8, train_loss: 0.0170 step time: 0.2535\n", + "8/8, train_loss: 0.0182 step time: 0.2444\n", + "epoch 284 average loss: 0.0174\n", + "time consuming of epoch 284 is: 2.0028\n", + "----------\n", + "epoch 285/600\n", + "1/8, train_loss: 0.0168 step time: 0.2826\n", + "2/8, train_loss: 0.0166 step time: 0.2353\n", + "3/8, train_loss: 0.0196 step time: 0.2316\n", + "4/8, train_loss: 0.0147 step time: 0.2328\n", + "5/8, train_loss: 0.0202 step time: 0.2665\n", + "6/8, train_loss: 0.0179 step time: 0.3206\n", + "7/8, train_loss: 0.0152 step time: 0.2972\n", + "8/8, train_loss: 0.0161 step time: 0.2724\n", + "epoch 285 average loss: 0.0171\n", + "saved new best metric model\n", + "current epoch: 285 current mean dice: 0.9541 best mean dice: 0.9541 at epoch: 285\n", + "time consuming of epoch 285 is: 3.9340\n", + "----------\n", + "epoch 286/600\n", + "1/8, train_loss: 0.0167 step time: 0.2709\n", + "2/8, train_loss: 0.0158 step time: 0.2123\n", + "3/8, train_loss: 0.0206 step time: 0.2117\n", + "4/8, train_loss: 0.0170 step time: 0.2120\n", + "5/8, train_loss: 0.0185 step time: 0.2126\n", + "6/8, train_loss: 0.0161 step time: 0.2240\n", + "7/8, train_loss: 0.0190 step time: 0.4714\n", + "8/8, train_loss: 0.0220 step time: 0.4731\n", + "epoch 286 average loss: 0.0182\n", + "time consuming of epoch 286 is: 2.2905\n", + "----------\n", + "epoch 287/600\n", + "1/8, train_loss: 0.0173 step time: 0.3103\n", + "2/8, train_loss: 0.0166 step time: 0.2463\n", + "3/8, train_loss: 0.0245 step time: 0.2529\n", + "4/8, train_loss: 0.0168 step time: 0.2548\n", + "5/8, train_loss: 0.0195 step time: 0.2619\n", + "6/8, train_loss: 0.0180 step time: 0.2582\n", + "7/8, train_loss: 0.0185 step time: 0.2520\n", + "8/8, train_loss: 0.0155 step time: 0.3327\n", + "epoch 287 average loss: 0.0183\n", + "time consuming of epoch 287 is: 2.1705\n", + "----------\n", + "epoch 288/600\n", + "1/8, train_loss: 0.0198 step time: 0.3769\n", + "2/8, train_loss: 0.0200 step time: 0.2838\n", + "3/8, train_loss: 0.0154 step time: 0.2612\n", + "4/8, train_loss: 0.0159 step time: 0.2530\n", + "5/8, train_loss: 0.0202 step time: 0.2480\n", + "6/8, train_loss: 0.0153 step time: 0.3317\n", + "7/8, train_loss: 0.0137 step time: 0.4539\n", + "8/8, train_loss: 0.0187 step time: 0.2603\n", + "epoch 288 average loss: 0.0174\n", + "time consuming of epoch 288 is: 2.4702\n", + "----------\n", + "epoch 289/600\n", + "1/8, train_loss: 0.0175 step time: 0.2974\n", + "2/8, train_loss: 0.0169 step time: 0.2471\n", + "3/8, train_loss: 0.0154 step time: 0.2410\n", + "4/8, train_loss: 0.0157 step time: 0.2406\n", + "5/8, train_loss: 0.0176 step time: 0.2342\n", + "6/8, train_loss: 0.0170 step time: 0.2422\n", + "7/8, train_loss: 0.0151 step time: 0.2295\n", + "8/8, train_loss: 0.0160 step time: 0.2651\n", + "epoch 289 average loss: 0.0164\n", + "time consuming of epoch 289 is: 1.9986\n", + "----------\n", + "epoch 290/600\n", + "1/8, train_loss: 0.0171 step time: 0.2982\n", + "2/8, train_loss: 0.0164 step time: 0.2498\n", + "3/8, train_loss: 0.0170 step time: 0.2519\n", + "4/8, train_loss: 0.0202 step time: 0.2522\n", + "5/8, train_loss: 0.0163 step time: 0.2598\n", + "6/8, train_loss: 0.0162 step time: 0.2684\n", + "7/8, train_loss: 0.0176 step time: 0.2716\n", + "8/8, train_loss: 0.0144 step time: 0.2434\n", + "epoch 290 average loss: 0.0169\n", + "current epoch: 290 current mean dice: 0.9529 best mean dice: 0.9541 at epoch: 285\n", + "time consuming of epoch 290 is: 3.5462\n", + "----------\n", + "epoch 291/600\n", + "1/8, train_loss: 0.0181 step time: 0.3293\n", + "2/8, train_loss: 0.0193 step time: 0.2683\n", + "3/8, train_loss: 0.0190 step time: 0.2571\n", + "4/8, train_loss: 0.0153 step time: 0.2443\n", + "5/8, train_loss: 0.0148 step time: 0.2410\n", + "6/8, train_loss: 0.0156 step time: 0.2413\n", + "7/8, train_loss: 0.0175 step time: 0.2322\n", + "8/8, train_loss: 0.0188 step time: 0.2321\n", + "epoch 291 average loss: 0.0173\n", + "time consuming of epoch 291 is: 2.0467\n", + "----------\n", + "epoch 292/600\n", + "1/8, train_loss: 0.0162 step time: 0.2883\n", + "2/8, train_loss: 0.0131 step time: 0.2419\n", + "3/8, train_loss: 0.0165 step time: 0.3568\n", + "4/8, train_loss: 0.0175 step time: 0.3275\n", + "5/8, train_loss: 0.0181 step time: 0.2761\n", + "6/8, train_loss: 0.0253 step time: 0.2580\n", + "7/8, train_loss: 0.0206 step time: 0.2573\n", + "8/8, train_loss: 0.0170 step time: 0.2450\n", + "epoch 292 average loss: 0.0180\n", + "time consuming of epoch 292 is: 2.2523\n", + "----------\n", + "epoch 293/600\n", + "1/8, train_loss: 0.0167 step time: 0.2945\n", + "2/8, train_loss: 0.0185 step time: 0.2572\n", + "3/8, train_loss: 0.0178 step time: 0.2454\n", + "4/8, train_loss: 0.0195 step time: 0.2474\n", + "5/8, train_loss: 0.0168 step time: 0.2437\n", + "6/8, train_loss: 0.0181 step time: 0.2968\n", + "7/8, train_loss: 0.0155 step time: 0.3773\n", + "8/8, train_loss: 0.0149 step time: 0.2694\n", + "epoch 293 average loss: 0.0172\n", + "time consuming of epoch 293 is: 2.2331\n", + "----------\n", + "epoch 294/600\n", + "1/8, train_loss: 0.0162 step time: 0.3046\n", + "2/8, train_loss: 0.0157 step time: 0.2498\n", + "3/8, train_loss: 0.0160 step time: 0.2423\n", + "4/8, train_loss: 0.0161 step time: 0.2383\n", + "5/8, train_loss: 0.0196 step time: 0.2379\n", + "6/8, train_loss: 0.0155 step time: 0.2371\n", + "7/8, train_loss: 0.0186 step time: 0.2278\n", + "8/8, train_loss: 0.0125 step time: 0.2296\n", + "epoch 294 average loss: 0.0163\n", + "time consuming of epoch 294 is: 1.9689\n", + "----------\n", + "epoch 295/600\n", + "1/8, train_loss: 0.0153 step time: 0.2822\n", + "2/8, train_loss: 0.0172 step time: 0.2353\n", + "3/8, train_loss: 0.0182 step time: 0.2341\n", + "4/8, train_loss: 0.0168 step time: 0.2309\n", + "5/8, train_loss: 0.0161 step time: 0.2656\n", + "6/8, train_loss: 0.0179 step time: 0.2736\n", + "7/8, train_loss: 0.0159 step time: 0.2755\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "8/8, train_loss: 0.0189 step time: 0.2737\n", + "epoch 295 average loss: 0.0170\n", + "current epoch: 295 current mean dice: 0.9537 best mean dice: 0.9541 at epoch: 285\n", + "time consuming of epoch 295 is: 3.3778\n", + "----------\n", + "epoch 296/600\n", + "1/8, train_loss: 0.0175 step time: 0.3695\n", + "2/8, train_loss: 0.0159 step time: 0.2958\n", + "3/8, train_loss: 0.0173 step time: 0.2571\n", + "4/8, train_loss: 0.0137 step time: 0.2648\n", + "5/8, train_loss: 0.0160 step time: 0.2611\n", + "6/8, train_loss: 0.0146 step time: 0.2675\n", + "7/8, train_loss: 0.0180 step time: 0.2689\n", + "8/8, train_loss: 0.0170 step time: 0.2603\n", + "epoch 296 average loss: 0.0163\n", + "time consuming of epoch 296 is: 2.2462\n", + "----------\n", + "epoch 297/600\n", + "1/8, train_loss: 0.0160 step time: 0.2973\n", + "2/8, train_loss: 0.0151 step time: 0.2545\n", + "3/8, train_loss: 0.0159 step time: 0.2406\n", + "4/8, train_loss: 0.0165 step time: 0.2579\n", + "5/8, train_loss: 0.0144 step time: 0.2978\n", + "6/8, train_loss: 0.0167 step time: 0.2661\n", + "7/8, train_loss: 0.0215 step time: 0.2629\n", + "8/8, train_loss: 0.0147 step time: 0.2643\n", + "epoch 297 average loss: 0.0164\n", + "time consuming of epoch 297 is: 2.1429\n", + "----------\n", + "epoch 298/600\n", + "1/8, train_loss: 0.0196 step time: 0.3127\n", + "2/8, train_loss: 0.0140 step time: 0.2590\n", + "3/8, train_loss: 0.0166 step time: 0.2526\n", + "4/8, train_loss: 0.0162 step time: 0.2556\n", + "5/8, train_loss: 0.0158 step time: 0.2503\n", + "6/8, train_loss: 0.0167 step time: 0.2462\n", + "7/8, train_loss: 0.0158 step time: 0.2248\n", + "8/8, train_loss: 0.0173 step time: 0.2221\n", + "epoch 298 average loss: 0.0165\n", + "time consuming of epoch 298 is: 2.0247\n", + "----------\n", + "epoch 299/600\n", + "1/8, train_loss: 0.0217 step time: 0.2875\n", + "2/8, train_loss: 0.0150 step time: 0.2312\n", + "3/8, train_loss: 0.0150 step time: 0.2379\n", + "4/8, train_loss: 0.0153 step time: 0.2430\n", + "5/8, train_loss: 0.0141 step time: 0.3139\n", + "6/8, train_loss: 0.0145 step time: 0.4030\n", + "7/8, train_loss: 0.0163 step time: 0.2852\n", + "8/8, train_loss: 0.0212 step time: 0.2546\n", + "epoch 299 average loss: 0.0166\n", + "time consuming of epoch 299 is: 2.2578\n", + "----------\n", + "epoch 300/600\n", + "1/8, train_loss: 0.0147 step time: 0.2949\n", + "2/8, train_loss: 0.0168 step time: 0.2415\n", + "3/8, train_loss: 0.0163 step time: 0.2408\n", + "4/8, train_loss: 0.0154 step time: 0.2446\n", + "5/8, train_loss: 0.0192 step time: 0.2445\n", + "6/8, train_loss: 0.0190 step time: 0.2964\n", + "7/8, train_loss: 0.0165 step time: 0.2926\n", + "8/8, train_loss: 0.0159 step time: 0.2782\n", + "epoch 300 average loss: 0.0167\n", + "current epoch: 300 current mean dice: 0.9538 best mean dice: 0.9541 at epoch: 285\n", + "time consuming of epoch 300 is: 3.3776\n", + "----------\n", + "epoch 301/600\n", + "1/8, train_loss: 0.0161 step time: 0.3009\n", + "2/8, train_loss: 0.0165 step time: 0.2559\n", + "3/8, train_loss: 0.0122 step time: 0.2397\n", + "4/8, train_loss: 0.0192 step time: 0.2401\n", + "5/8, train_loss: 0.0194 step time: 0.2419\n", + "6/8, train_loss: 0.0158 step time: 0.2382\n", + "7/8, train_loss: 0.0160 step time: 0.2207\n", + "8/8, train_loss: 0.0188 step time: 0.3734\n", + "epoch 301 average loss: 0.0168\n", + "time consuming of epoch 301 is: 2.1122\n", + "----------\n", + "epoch 302/600\n", + "1/8, train_loss: 0.0152 step time: 0.5640\n", + "2/8, train_loss: 0.0168 step time: 0.2746\n", + "3/8, train_loss: 0.0235 step time: 0.2620\n", + "4/8, train_loss: 0.0159 step time: 0.2526\n", + "5/8, train_loss: 0.0197 step time: 0.2429\n", + "6/8, train_loss: 0.0177 step time: 0.2453\n", + "7/8, train_loss: 0.0171 step time: 0.2309\n", + "8/8, train_loss: 0.0157 step time: 0.2691\n", + "epoch 302 average loss: 0.0177\n", + "time consuming of epoch 302 is: 2.3429\n", + "----------\n", + "epoch 303/600\n", + "1/8, train_loss: 0.0141 step time: 0.3742\n", + "2/8, train_loss: 0.0144 step time: 0.2751\n", + "3/8, train_loss: 0.0173 step time: 0.2541\n", + "4/8, train_loss: 0.0172 step time: 0.2597\n", + "5/8, train_loss: 0.0220 step time: 0.2565\n", + "6/8, train_loss: 0.0202 step time: 0.2549\n", + "7/8, train_loss: 0.0167 step time: 0.2558\n", + "8/8, train_loss: 0.0165 step time: 0.2521\n", + "epoch 303 average loss: 0.0173\n", + "time consuming of epoch 303 is: 2.1840\n", + "----------\n", + "epoch 304/600\n", + "1/8, train_loss: 0.0157 step time: 0.2934\n", + "2/8, train_loss: 0.0179 step time: 0.2370\n", + "3/8, train_loss: 0.0154 step time: 0.2413\n", + "4/8, train_loss: 0.0168 step time: 0.2559\n", + "5/8, train_loss: 0.0147 step time: 0.2454\n", + "6/8, train_loss: 0.0186 step time: 0.2472\n", + "7/8, train_loss: 0.0229 step time: 0.2388\n" ] } ], From 7b55b46a41469cf4afeac5e9304f83b1fec506a9 Mon Sep 17 00:00:00 2001 From: Nic Ma Date: Wed, 20 Jul 2022 20:39:16 +0800 Subject: [PATCH 4/5] [DLMED] fix flake8 Signed-off-by: Nic Ma --- acceleration/fast_training_tutorial.ipynb | 3727 ++++++++++++++++++++- 1 file changed, 3724 insertions(+), 3 deletions(-) diff --git a/acceleration/fast_training_tutorial.ipynb b/acceleration/fast_training_tutorial.ipynb index c0a7d6f624..2456fca175 100644 --- a/acceleration/fast_training_tutorial.ipynb +++ b/acceleration/fast_training_tutorial.ipynb @@ -109,7 +109,6 @@ " ThreadDataLoader,\n", " Dataset,\n", " decollate_batch,\n", - " get_track_meta,\n", " set_track_meta,\n", ")\n", "from monai.inferers import sliding_window_inference\n", @@ -744,7 +743,7 @@ }, { "cell_type": "code", - "execution_count": null, + "execution_count": 9, "metadata": { "vscode": { "languageId": "python" @@ -4580,7 +4579,3729 @@ "4/8, train_loss: 0.0168 step time: 0.2559\n", "5/8, train_loss: 0.0147 step time: 0.2454\n", "6/8, train_loss: 0.0186 step time: 0.2472\n", - "7/8, train_loss: 0.0229 step time: 0.2388\n" + "7/8, train_loss: 0.0229 step time: 0.2388\n", + "8/8, train_loss: 0.0166 step time: 0.2462\n", + "epoch 304 average loss: 0.0173\n", + "time consuming of epoch 304 is: 2.0067\n", + "----------\n", + "epoch 305/600\n", + "1/8, train_loss: 0.0160 step time: 0.2998\n", + "2/8, train_loss: 0.0174 step time: 0.2526\n", + "3/8, train_loss: 0.0161 step time: 0.2402\n", + "4/8, train_loss: 0.0168 step time: 0.2392\n", + "5/8, train_loss: 0.0152 step time: 0.2368\n", + "6/8, train_loss: 0.0170 step time: 0.2338\n", + "7/8, train_loss: 0.0180 step time: 0.2223\n", + "8/8, train_loss: 0.0161 step time: 0.2206\n", + "epoch 305 average loss: 0.0166\n", + "current epoch: 305 current mean dice: 0.9530 best mean dice: 0.9541 at epoch: 285\n", + "time consuming of epoch 305 is: 3.1726\n", + "----------\n", + "epoch 306/600\n", + "1/8, train_loss: 0.0197 step time: 0.3283\n", + "2/8, train_loss: 0.0146 step time: 0.2576\n", + "3/8, train_loss: 0.0135 step time: 0.2712\n", + "4/8, train_loss: 0.0166 step time: 0.2786\n", + "5/8, train_loss: 0.0179 step time: 0.2582\n", + "6/8, train_loss: 0.0175 step time: 0.2541\n", + "7/8, train_loss: 0.0174 step time: 0.2383\n", + "8/8, train_loss: 0.0201 step time: 0.2313\n", + "epoch 306 average loss: 0.0172\n", + "time consuming of epoch 306 is: 2.1187\n", + "----------\n", + "epoch 307/600\n", + "1/8, train_loss: 0.0204 step time: 0.2800\n", + "2/8, train_loss: 0.0149 step time: 0.3638\n", + "3/8, train_loss: 0.0176 step time: 0.4746\n", + "4/8, train_loss: 0.0163 step time: 0.2594\n", + "5/8, train_loss: 0.0160 step time: 0.2584\n", + "6/8, train_loss: 0.0163 step time: 0.2605\n", + "7/8, train_loss: 0.0190 step time: 0.2415\n", + "8/8, train_loss: 0.0158 step time: 0.2334\n", + "epoch 307 average loss: 0.0170\n", + "time consuming of epoch 307 is: 2.3731\n", + "----------\n", + "epoch 308/600\n", + "1/8, train_loss: 0.0213 step time: 0.2851\n", + "2/8, train_loss: 0.0172 step time: 0.2316\n", + "3/8, train_loss: 0.0150 step time: 0.2283\n", + "4/8, train_loss: 0.0133 step time: 0.2507\n", + "5/8, train_loss: 0.0168 step time: 0.2919\n", + "6/8, train_loss: 0.0184 step time: 0.2921\n", + "7/8, train_loss: 0.0117 step time: 0.2865\n", + "8/8, train_loss: 0.0168 step time: 0.2777\n", + "epoch 308 average loss: 0.0163\n", + "time consuming of epoch 308 is: 2.1453\n", + "----------\n", + "epoch 309/600\n", + "1/8, train_loss: 0.0156 step time: 0.3098\n", + "2/8, train_loss: 0.0174 step time: 0.2606\n", + "3/8, train_loss: 0.0164 step time: 0.2545\n", + "4/8, train_loss: 0.0116 step time: 0.2799\n", + "5/8, train_loss: 0.0159 step time: 0.2804\n", + "6/8, train_loss: 0.0149 step time: 0.2641\n", + "7/8, train_loss: 0.0187 step time: 0.2621\n", + "8/8, train_loss: 0.0168 step time: 0.2468\n", + "epoch 309 average loss: 0.0159\n", + "time consuming of epoch 309 is: 2.1597\n", + "----------\n", + "epoch 310/600\n", + "1/8, train_loss: 0.0155 step time: 0.3055\n", + "2/8, train_loss: 0.0181 step time: 0.2536\n", + "3/8, train_loss: 0.0196 step time: 0.2480\n", + "4/8, train_loss: 0.0167 step time: 0.2469\n", + "5/8, train_loss: 0.0177 step time: 0.2479\n", + "6/8, train_loss: 0.0130 step time: 0.2501\n", + "7/8, train_loss: 0.0160 step time: 0.4057\n", + "8/8, train_loss: 0.0157 step time: 0.3632\n", + "epoch 310 average loss: 0.0165\n", + "current epoch: 310 current mean dice: 0.9537 best mean dice: 0.9541 at epoch: 285\n", + "time consuming of epoch 310 is: 3.4805\n", + "----------\n", + "epoch 311/600\n", + "1/8, train_loss: 0.0160 step time: 0.4076\n", + "2/8, train_loss: 0.0193 step time: 0.4807\n", + "3/8, train_loss: 0.0171 step time: 0.2203\n", + "4/8, train_loss: 0.0160 step time: 0.2149\n", + "5/8, train_loss: 0.0179 step time: 0.2135\n", + "6/8, train_loss: 0.0146 step time: 0.2152\n", + "7/8, train_loss: 0.0190 step time: 0.1999\n", + "8/8, train_loss: 0.0166 step time: 0.2003\n", + "epoch 311 average loss: 0.0171\n", + "time consuming of epoch 311 is: 2.1537\n", + "----------\n", + "epoch 312/600\n", + "1/8, train_loss: 0.0149 step time: 0.2668\n", + "2/8, train_loss: 0.0186 step time: 0.2793\n", + "3/8, train_loss: 0.0125 step time: 0.4171\n", + "4/8, train_loss: 0.0154 step time: 0.3036\n", + "5/8, train_loss: 0.0230 step time: 0.2712\n", + "6/8, train_loss: 0.0165 step time: 0.2650\n", + "7/8, train_loss: 0.0160 step time: 0.2622\n", + "8/8, train_loss: 0.0159 step time: 0.2467\n", + "epoch 312 average loss: 0.0166\n", + "time consuming of epoch 312 is: 2.3135\n", + "----------\n", + "epoch 313/600\n", + "1/8, train_loss: 0.0169 step time: 0.2951\n", + "2/8, train_loss: 0.0170 step time: 0.2387\n", + "3/8, train_loss: 0.0159 step time: 0.2358\n", + "4/8, train_loss: 0.0178 step time: 0.2340\n", + "5/8, train_loss: 0.0189 step time: 0.3767\n", + "6/8, train_loss: 0.0134 step time: 0.4929\n", + "7/8, train_loss: 0.0143 step time: 0.2616\n", + "8/8, train_loss: 0.0143 step time: 0.2551\n", + "epoch 313 average loss: 0.0161\n", + "time consuming of epoch 313 is: 2.3915\n", + "----------\n", + "epoch 314/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0171 step time: 0.2927\n", + "2/8, train_loss: 0.0146 step time: 0.2702\n", + "3/8, train_loss: 0.0204 step time: 0.2722\n", + "4/8, train_loss: 0.0194 step time: 0.3976\n", + "5/8, train_loss: 0.0154 step time: 0.2169\n", + "6/8, train_loss: 0.0149 step time: 0.2154\n", + "7/8, train_loss: 0.0165 step time: 0.2091\n", + "8/8, train_loss: 0.0138 step time: 0.2281\n", + "epoch 314 average loss: 0.0165\n", + "time consuming of epoch 314 is: 2.1039\n", + "----------\n", + "epoch 315/600\n", + "1/8, train_loss: 0.0177 step time: 0.3028\n", + "2/8, train_loss: 0.0208 step time: 0.2496\n", + "3/8, train_loss: 0.0155 step time: 0.2453\n", + "4/8, train_loss: 0.0150 step time: 0.2448\n", + "5/8, train_loss: 0.0174 step time: 0.2421\n", + "6/8, train_loss: 0.0142 step time: 0.2440\n", + "7/8, train_loss: 0.0169 step time: 0.2326\n", + "8/8, train_loss: 0.0142 step time: 0.2302\n", + "epoch 315 average loss: 0.0164\n", + "saved new best metric model\n", + "current epoch: 315 current mean dice: 0.9550 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 315 is: 3.7923\n", + "----------\n", + "epoch 316/600\n", + "1/8, train_loss: 0.0169 step time: 0.2719\n", + "2/8, train_loss: 0.0187 step time: 0.2129\n", + "3/8, train_loss: 0.0163 step time: 0.2146\n", + "4/8, train_loss: 0.0146 step time: 0.2142\n", + "5/8, train_loss: 0.0172 step time: 0.2230\n", + "6/8, train_loss: 0.0150 step time: 0.2375\n", + "7/8, train_loss: 0.0180 step time: 0.2617\n", + "8/8, train_loss: 0.0131 step time: 0.3430\n", + "epoch 316 average loss: 0.0162\n", + "time consuming of epoch 316 is: 1.9814\n", + "----------\n", + "epoch 317/600\n", + "1/8, train_loss: 0.0174 step time: 0.3751\n", + "2/8, train_loss: 0.0163 step time: 0.2657\n", + "3/8, train_loss: 0.0153 step time: 0.2534\n", + "4/8, train_loss: 0.0157 step time: 0.2607\n", + "5/8, train_loss: 0.0156 step time: 0.2451\n", + "6/8, train_loss: 0.0135 step time: 0.2443\n", + "7/8, train_loss: 0.0163 step time: 0.2325\n", + "8/8, train_loss: 0.0159 step time: 0.2319\n", + "epoch 317 average loss: 0.0158\n", + "time consuming of epoch 317 is: 2.1101\n", + "----------\n", + "epoch 318/600\n", + "1/8, train_loss: 0.0145 step time: 0.2847\n", + "2/8, train_loss: 0.0181 step time: 0.2329\n", + "3/8, train_loss: 0.0149 step time: 0.2685\n", + "4/8, train_loss: 0.0159 step time: 0.2734\n", + "5/8, train_loss: 0.0176 step time: 0.2621\n", + "6/8, train_loss: 0.0168 step time: 0.2933\n", + "7/8, train_loss: 0.0175 step time: 0.3421\n", + "8/8, train_loss: 0.0148 step time: 0.2931\n", + "epoch 318 average loss: 0.0163\n", + "time consuming of epoch 318 is: 2.2515\n", + "----------\n", + "epoch 319/600\n", + "1/8, train_loss: 0.0158 step time: 0.3126\n", + "2/8, train_loss: 0.0170 step time: 0.2499\n", + "3/8, train_loss: 0.0183 step time: 0.2404\n", + "4/8, train_loss: 0.0178 step time: 0.2350\n", + "5/8, train_loss: 0.0198 step time: 0.2410\n", + "6/8, train_loss: 0.0129 step time: 0.2548\n", + "7/8, train_loss: 0.0148 step time: 0.2483\n", + "8/8, train_loss: 0.0170 step time: 0.2496\n", + "epoch 319 average loss: 0.0167\n", + "time consuming of epoch 319 is: 2.0329\n", + "----------\n", + "epoch 320/600\n", + "1/8, train_loss: 0.0160 step time: 0.2908\n", + "2/8, train_loss: 0.0151 step time: 0.2473\n", + "3/8, train_loss: 0.0149 step time: 0.2403\n", + "4/8, train_loss: 0.0149 step time: 0.2351\n", + "5/8, train_loss: 0.0171 step time: 0.2989\n", + "6/8, train_loss: 0.0187 step time: 0.3598\n", + "7/8, train_loss: 0.0195 step time: 0.2782\n", + "8/8, train_loss: 0.0151 step time: 0.2536\n", + "epoch 320 average loss: 0.0164\n", + "current epoch: 320 current mean dice: 0.9527 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 320 is: 3.6388\n", + "----------\n", + "epoch 321/600\n", + "1/8, train_loss: 0.0140 step time: 0.3192\n", + "2/8, train_loss: 0.0159 step time: 0.2678\n", + "3/8, train_loss: 0.0167 step time: 0.2498\n", + "4/8, train_loss: 0.0167 step time: 0.2566\n", + "5/8, train_loss: 0.0137 step time: 0.2551\n", + "6/8, train_loss: 0.0157 step time: 0.2534\n", + "7/8, train_loss: 0.0161 step time: 0.2344\n", + "8/8, train_loss: 0.0190 step time: 0.2361\n", + "epoch 321 average loss: 0.0160\n", + "time consuming of epoch 321 is: 2.0737\n", + "----------\n", + "epoch 322/600\n", + "1/8, train_loss: 0.0157 step time: 0.2869\n", + "2/8, train_loss: 0.0178 step time: 0.2380\n", + "3/8, train_loss: 0.0160 step time: 0.2330\n", + "4/8, train_loss: 0.0151 step time: 0.3347\n", + "5/8, train_loss: 0.0164 step time: 0.3913\n", + "6/8, train_loss: 0.0184 step time: 0.2656\n", + "7/8, train_loss: 0.0178 step time: 0.2558\n", + "8/8, train_loss: 0.0177 step time: 0.2494\n", + "epoch 322 average loss: 0.0168\n", + "time consuming of epoch 322 is: 2.2563\n", + "----------\n", + "epoch 323/600\n", + "1/8, train_loss: 0.0186 step time: 0.2993\n", + "2/8, train_loss: 0.0141 step time: 0.2474\n", + "3/8, train_loss: 0.0154 step time: 0.3336\n", + "4/8, train_loss: 0.0176 step time: 0.4839\n", + "5/8, train_loss: 0.0148 step time: 0.2655\n", + "6/8, train_loss: 0.0173 step time: 0.2528\n", + "7/8, train_loss: 0.0132 step time: 0.2476\n", + "8/8, train_loss: 0.0180 step time: 0.2365\n", + "epoch 323 average loss: 0.0161\n", + "time consuming of epoch 323 is: 2.3681\n", + "----------\n", + "epoch 324/600\n", + "1/8, train_loss: 0.0127 step time: 0.2881\n", + "2/8, train_loss: 0.0199 step time: 0.2379\n", + "3/8, train_loss: 0.0163 step time: 0.2308\n", + "4/8, train_loss: 0.0144 step time: 0.2319\n", + "5/8, train_loss: 0.0135 step time: 0.2285\n", + "6/8, train_loss: 0.0154 step time: 0.2321\n", + "7/8, train_loss: 0.0195 step time: 0.2185\n", + "8/8, train_loss: 0.0156 step time: 0.2772\n", + "epoch 324 average loss: 0.0159\n", + "time consuming of epoch 324 is: 1.9464\n", + "----------\n", + "epoch 325/600\n", + "1/8, train_loss: 0.0173 step time: 0.3509\n", + "2/8, train_loss: 0.0162 step time: 0.2768\n", + "3/8, train_loss: 0.0163 step time: 0.2567\n", + "4/8, train_loss: 0.0179 step time: 0.2521\n", + "5/8, train_loss: 0.0138 step time: 0.2471\n", + "6/8, train_loss: 0.0147 step time: 0.2418\n", + "7/8, train_loss: 0.0147 step time: 0.2329\n", + "8/8, train_loss: 0.0142 step time: 0.2329\n", + "epoch 325 average loss: 0.0156\n", + "current epoch: 325 current mean dice: 0.9544 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 325 is: 3.7651\n", + "----------\n", + "epoch 326/600\n", + "1/8, train_loss: 0.0158 step time: 0.2944\n", + "2/8, train_loss: 0.0157 step time: 0.2398\n", + "3/8, train_loss: 0.0177 step time: 0.2308\n", + "4/8, train_loss: 0.0148 step time: 0.3600\n", + "5/8, train_loss: 0.0159 step time: 0.4795\n", + "6/8, train_loss: 0.0137 step time: 0.2579\n", + "7/8, train_loss: 0.0184 step time: 0.2400\n", + "8/8, train_loss: 0.0134 step time: 0.2436\n", + "epoch 326 average loss: 0.0157\n", + "time consuming of epoch 326 is: 2.3473\n", + "----------\n", + "epoch 327/600\n", + "1/8, train_loss: 0.0155 step time: 0.2893\n", + "2/8, train_loss: 0.0125 step time: 0.2354\n", + "3/8, train_loss: 0.0173 step time: 0.2760\n", + "4/8, train_loss: 0.0199 step time: 0.2700\n", + "5/8, train_loss: 0.0173 step time: 0.2555\n", + "6/8, train_loss: 0.0164 step time: 0.2525\n", + "7/8, train_loss: 0.0175 step time: 0.2369\n", + "8/8, train_loss: 0.0158 step time: 0.2886\n", + "epoch 327 average loss: 0.0165\n", + "time consuming of epoch 327 is: 2.1057\n", + "----------\n", + "epoch 328/600\n", + "1/8, train_loss: 0.0152 step time: 0.3375\n", + "2/8, train_loss: 0.0133 step time: 0.2672\n", + "3/8, train_loss: 0.0166 step time: 0.2572\n", + "4/8, train_loss: 0.0162 step time: 0.2499\n", + "5/8, train_loss: 0.0165 step time: 0.2447\n", + "6/8, train_loss: 0.0133 step time: 0.2419\n", + "7/8, train_loss: 0.0179 step time: 0.2904\n", + "8/8, train_loss: 0.0143 step time: 0.3346\n", + "epoch 328 average loss: 0.0154\n", + "time consuming of epoch 328 is: 2.2249\n", + "----------\n", + "epoch 329/600\n", + "1/8, train_loss: 0.0164 step time: 0.3121\n", + "2/8, train_loss: 0.0167 step time: 0.2550\n", + "3/8, train_loss: 0.0163 step time: 0.2479\n", + "4/8, train_loss: 0.0192 step time: 0.2420\n", + "5/8, train_loss: 0.0174 step time: 0.2370\n", + "6/8, train_loss: 0.0122 step time: 0.2340\n", + "7/8, train_loss: 0.0178 step time: 0.2268\n", + "8/8, train_loss: 0.0131 step time: 0.2508\n", + "epoch 329 average loss: 0.0161\n", + "time consuming of epoch 329 is: 2.0071\n", + "----------\n", + "epoch 330/600\n", + "1/8, train_loss: 0.0164 step time: 0.3362\n", + "2/8, train_loss: 0.0172 step time: 0.2683\n", + "3/8, train_loss: 0.0160 step time: 0.2529\n", + "4/8, train_loss: 0.0160 step time: 0.2488\n", + "5/8, train_loss: 0.0129 step time: 0.2550\n", + "6/8, train_loss: 0.0190 step time: 0.2469\n", + "7/8, train_loss: 0.0157 step time: 0.2330\n", + "8/8, train_loss: 0.0165 step time: 0.2333\n", + "epoch 330 average loss: 0.0162\n", + "current epoch: 330 current mean dice: 0.9533 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 330 is: 3.0798\n", + "----------\n", + "epoch 331/600\n", + "1/8, train_loss: 0.0145 step time: 0.3729\n", + "2/8, train_loss: 0.0171 step time: 0.3524\n", + "3/8, train_loss: 0.0145 step time: 0.2932\n", + "4/8, train_loss: 0.0136 step time: 0.2801\n", + "5/8, train_loss: 0.0178 step time: 0.2653\n", + "6/8, train_loss: 0.0194 step time: 0.2656\n", + "7/8, train_loss: 0.0145 step time: 0.2476\n", + "8/8, train_loss: 0.0141 step time: 0.2394\n", + "epoch 331 average loss: 0.0157\n", + "time consuming of epoch 331 is: 2.3178\n", + "----------\n", + "epoch 332/600\n", + "1/8, train_loss: 0.0141 step time: 0.2866\n", + "2/8, train_loss: 0.0201 step time: 0.2372\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "3/8, train_loss: 0.0139 step time: 0.2520\n", + "4/8, train_loss: 0.0156 step time: 0.2529\n", + "5/8, train_loss: 0.0185 step time: 0.2522\n", + "6/8, train_loss: 0.0156 step time: 0.3346\n", + "7/8, train_loss: 0.0156 step time: 0.3277\n", + "8/8, train_loss: 0.0153 step time: 0.2651\n", + "epoch 332 average loss: 0.0161\n", + "time consuming of epoch 332 is: 2.2097\n", + "----------\n", + "epoch 333/600\n", + "1/8, train_loss: 0.0139 step time: 0.2742\n", + "2/8, train_loss: 0.0158 step time: 0.2405\n", + "3/8, train_loss: 0.0165 step time: 0.2417\n", + "4/8, train_loss: 0.0157 step time: 0.2326\n", + "5/8, train_loss: 0.0162 step time: 0.2290\n", + "6/8, train_loss: 0.0197 step time: 0.2328\n", + "7/8, train_loss: 0.0162 step time: 0.3553\n", + "8/8, train_loss: 0.0162 step time: 0.6037\n", + "epoch 333 average loss: 0.0163\n", + "time consuming of epoch 333 is: 2.4113\n", + "----------\n", + "epoch 334/600\n", + "1/8, train_loss: 0.0141 step time: 0.2671\n", + "2/8, train_loss: 0.0155 step time: 0.2154\n", + "3/8, train_loss: 0.0162 step time: 0.2203\n", + "4/8, train_loss: 0.0134 step time: 0.2596\n", + "5/8, train_loss: 0.0144 step time: 0.3132\n", + "6/8, train_loss: 0.0173 step time: 0.3066\n", + "7/8, train_loss: 0.0153 step time: 0.2831\n", + "8/8, train_loss: 0.0187 step time: 0.2832\n", + "epoch 334 average loss: 0.0156\n", + "time consuming of epoch 334 is: 2.1500\n", + "----------\n", + "epoch 335/600\n", + "1/8, train_loss: 0.0144 step time: 0.3054\n", + "2/8, train_loss: 0.0220 step time: 0.2596\n", + "3/8, train_loss: 0.0162 step time: 0.2517\n", + "4/8, train_loss: 0.0132 step time: 0.2401\n", + "5/8, train_loss: 0.0161 step time: 0.2900\n", + "6/8, train_loss: 0.0161 step time: 0.2702\n", + "7/8, train_loss: 0.0156 step time: 0.2607\n", + "8/8, train_loss: 0.0181 step time: 0.2477\n", + "epoch 335 average loss: 0.0164\n", + "current epoch: 335 current mean dice: 0.9537 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 335 is: 3.4367\n", + "----------\n", + "epoch 336/600\n", + "1/8, train_loss: 0.0160 step time: 0.3162\n", + "2/8, train_loss: 0.0165 step time: 0.2510\n", + "3/8, train_loss: 0.0172 step time: 0.2521\n", + "4/8, train_loss: 0.0179 step time: 0.2455\n", + "5/8, train_loss: 0.0166 step time: 0.2512\n", + "6/8, train_loss: 0.0127 step time: 0.3738\n", + "7/8, train_loss: 0.0192 step time: 0.3476\n", + "8/8, train_loss: 0.0137 step time: 0.2580\n", + "epoch 336 average loss: 0.0162\n", + "time consuming of epoch 336 is: 2.2966\n", + "----------\n", + "epoch 337/600\n", + "1/8, train_loss: 0.0142 step time: 0.2913\n", + "2/8, train_loss: 0.0168 step time: 0.2475\n", + "3/8, train_loss: 0.0161 step time: 0.2487\n", + "4/8, train_loss: 0.0171 step time: 0.2383\n", + "5/8, train_loss: 0.0149 step time: 0.2416\n", + "6/8, train_loss: 0.0179 step time: 0.2405\n", + "7/8, train_loss: 0.0138 step time: 0.2315\n", + "8/8, train_loss: 0.0193 step time: 0.2330\n", + "epoch 337 average loss: 0.0163\n", + "time consuming of epoch 337 is: 1.9738\n", + "----------\n", + "epoch 338/600\n", + "1/8, train_loss: 0.0165 step time: 0.5242\n", + "2/8, train_loss: 0.0171 step time: 0.2181\n", + "3/8, train_loss: 0.0163 step time: 0.2143\n", + "4/8, train_loss: 0.0150 step time: 0.2149\n", + "5/8, train_loss: 0.0154 step time: 0.2147\n", + "6/8, train_loss: 0.0135 step time: 0.2334\n", + "7/8, train_loss: 0.0184 step time: 0.3538\n", + "8/8, train_loss: 0.0158 step time: 0.3503\n", + "epoch 338 average loss: 0.0160\n", + "time consuming of epoch 338 is: 2.3252\n", + "----------\n", + "epoch 339/600\n", + "1/8, train_loss: 0.0151 step time: 0.3370\n", + "2/8, train_loss: 0.0154 step time: 0.2630\n", + "3/8, train_loss: 0.0173 step time: 0.2522\n", + "4/8, train_loss: 0.0184 step time: 0.2497\n", + "5/8, train_loss: 0.0174 step time: 0.2505\n", + "6/8, train_loss: 0.0159 step time: 0.2455\n", + "7/8, train_loss: 0.0145 step time: 0.2425\n", + "8/8, train_loss: 0.0152 step time: 0.2837\n", + "epoch 339 average loss: 0.0162\n", + "time consuming of epoch 339 is: 2.1256\n", + "----------\n", + "epoch 340/600\n", + "1/8, train_loss: 0.0174 step time: 0.3269\n", + "2/8, train_loss: 0.0138 step time: 0.2648\n", + "3/8, train_loss: 0.0191 step time: 0.2641\n", + "4/8, train_loss: 0.0162 step time: 0.2527\n", + "5/8, train_loss: 0.0152 step time: 0.2547\n", + "6/8, train_loss: 0.0199 step time: 0.2498\n", + "7/8, train_loss: 0.0167 step time: 0.3554\n", + "8/8, train_loss: 0.0163 step time: 0.3365\n", + "epoch 340 average loss: 0.0168\n", + "current epoch: 340 current mean dice: 0.9534 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 340 is: 3.6277\n", + "----------\n", + "epoch 341/600\n", + "1/8, train_loss: 0.0184 step time: 0.2948\n", + "2/8, train_loss: 0.0166 step time: 0.2596\n", + "3/8, train_loss: 0.0163 step time: 0.2471\n", + "4/8, train_loss: 0.0148 step time: 0.2442\n", + "5/8, train_loss: 0.0167 step time: 0.2459\n", + "6/8, train_loss: 0.0172 step time: 0.2356\n", + "7/8, train_loss: 0.0171 step time: 0.2740\n", + "8/8, train_loss: 0.0145 step time: 0.3404\n", + "epoch 341 average loss: 0.0164\n", + "time consuming of epoch 341 is: 2.1428\n", + "----------\n", + "epoch 342/600\n", + "1/8, train_loss: 0.0162 step time: 0.3398\n", + "2/8, train_loss: 0.0165 step time: 0.2592\n", + "3/8, train_loss: 0.0172 step time: 0.2526\n", + "4/8, train_loss: 0.0145 step time: 0.2521\n", + "5/8, train_loss: 0.0173 step time: 0.2464\n", + "6/8, train_loss: 0.0120 step time: 0.2444\n", + "7/8, train_loss: 0.0157 step time: 0.2364\n", + "8/8, train_loss: 0.0165 step time: 0.2738\n", + "epoch 342 average loss: 0.0157\n", + "time consuming of epoch 342 is: 2.1063\n", + "----------\n", + "epoch 343/600\n", + "1/8, train_loss: 0.0153 step time: 0.3473\n", + "2/8, train_loss: 0.0161 step time: 0.2703\n", + "3/8, train_loss: 0.0170 step time: 0.2537\n", + "4/8, train_loss: 0.0177 step time: 0.2515\n", + "5/8, train_loss: 0.0159 step time: 0.2361\n", + "6/8, train_loss: 0.0166 step time: 0.2300\n", + "7/8, train_loss: 0.0142 step time: 0.2173\n", + "8/8, train_loss: 0.0166 step time: 0.2206\n", + "epoch 343 average loss: 0.0162\n", + "time consuming of epoch 343 is: 2.0282\n", + "----------\n", + "epoch 344/600\n", + "1/8, train_loss: 0.0152 step time: 0.3021\n", + "2/8, train_loss: 0.0162 step time: 0.2561\n", + "3/8, train_loss: 0.0157 step time: 0.2471\n", + "4/8, train_loss: 0.0153 step time: 0.2511\n", + "5/8, train_loss: 0.0216 step time: 0.3351\n", + "6/8, train_loss: 0.0147 step time: 0.4696\n", + "7/8, train_loss: 0.0160 step time: 0.2619\n", + "8/8, train_loss: 0.0145 step time: 0.2536\n", + "epoch 344 average loss: 0.0161\n", + "time consuming of epoch 344 is: 2.3780\n", + "----------\n", + "epoch 345/600\n", + "1/8, train_loss: 0.0152 step time: 0.2827\n", + "2/8, train_loss: 0.0150 step time: 0.2927\n", + "3/8, train_loss: 0.0146 step time: 0.2986\n", + "4/8, train_loss: 0.0175 step time: 0.2623\n", + "5/8, train_loss: 0.0193 step time: 0.2488\n", + "6/8, train_loss: 0.0144 step time: 0.2504\n", + "7/8, train_loss: 0.0168 step time: 0.2526\n", + "8/8, train_loss: 0.0168 step time: 0.2481\n", + "epoch 345 average loss: 0.0162\n", + "current epoch: 345 current mean dice: 0.9530 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 345 is: 3.2603\n", + "----------\n", + "epoch 346/600\n", + "1/8, train_loss: 0.0169 step time: 0.2932\n", + "2/8, train_loss: 0.0132 step time: 0.2426\n", + "3/8, train_loss: 0.0151 step time: 0.2454\n", + "4/8, train_loss: 0.0193 step time: 0.2484\n", + "5/8, train_loss: 0.0170 step time: 0.2389\n", + "6/8, train_loss: 0.0161 step time: 0.2422\n", + "7/8, train_loss: 0.0161 step time: 0.2411\n", + "8/8, train_loss: 0.0142 step time: 0.2534\n", + "epoch 346 average loss: 0.0160\n", + "time consuming of epoch 346 is: 2.0063\n", + "----------\n", + "epoch 347/600\n", + "1/8, train_loss: 0.0157 step time: 0.2950\n", + "2/8, train_loss: 0.0168 step time: 0.2553\n", + "3/8, train_loss: 0.0159 step time: 0.2553\n", + "4/8, train_loss: 0.0139 step time: 0.2546\n", + "5/8, train_loss: 0.0198 step time: 0.2574\n", + "6/8, train_loss: 0.0141 step time: 0.2659\n", + "7/8, train_loss: 0.0139 step time: 0.2475\n", + "8/8, train_loss: 0.0166 step time: 0.2465\n", + "epoch 347 average loss: 0.0158\n", + "time consuming of epoch 347 is: 2.0789\n", + "----------\n", + "epoch 348/600\n", + "1/8, train_loss: 0.0156 step time: 0.2964\n", + "2/8, train_loss: 0.0160 step time: 0.2456\n", + "3/8, train_loss: 0.0152 step time: 0.3931\n", + "4/8, train_loss: 0.0148 step time: 0.5779\n", + "5/8, train_loss: 0.0168 step time: 0.2180\n", + "6/8, train_loss: 0.0157 step time: 0.2139\n", + "7/8, train_loss: 0.0147 step time: 0.2019\n", + "8/8, train_loss: 0.0199 step time: 0.2113\n", + "epoch 348 average loss: 0.0161\n", + "time consuming of epoch 348 is: 2.3596\n", + "----------\n", + "epoch 349/600\n", + "1/8, train_loss: 0.0143 step time: 0.6323\n", + "2/8, train_loss: 0.0146 step time: 0.2167\n", + "3/8, train_loss: 0.0200 step time: 0.2155\n", + "4/8, train_loss: 0.0203 step time: 0.3181\n", + "5/8, train_loss: 0.0152 step time: 0.3854\n", + "6/8, train_loss: 0.0162 step time: 0.2761\n", + "7/8, train_loss: 0.0188 step time: 0.2570\n", + "8/8, train_loss: 0.0144 step time: 0.2450\n", + "epoch 349 average loss: 0.0167\n", + "time consuming of epoch 349 is: 2.5480\n", + "----------\n", + "epoch 350/600\n", + "1/8, train_loss: 0.0165 step time: 0.2941\n", + "2/8, train_loss: 0.0118 step time: 0.2465\n", + "3/8, train_loss: 0.0155 step time: 0.2401\n", + "4/8, train_loss: 0.0139 step time: 0.2421\n", + "5/8, train_loss: 0.0177 step time: 0.2392\n", + "6/8, train_loss: 0.0178 step time: 0.2430\n", + "7/8, train_loss: 0.0164 step time: 0.2344\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "8/8, train_loss: 0.0171 step time: 0.2785\n", + "epoch 350 average loss: 0.0159\n", + "current epoch: 350 current mean dice: 0.9548 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 350 is: 3.4705\n", + "----------\n", + "epoch 351/600\n", + "1/8, train_loss: 0.0127 step time: 0.3383\n", + "2/8, train_loss: 0.0161 step time: 0.2705\n", + "3/8, train_loss: 0.0163 step time: 0.2592\n", + "4/8, train_loss: 0.0145 step time: 0.2517\n", + "5/8, train_loss: 0.0181 step time: 0.2441\n", + "6/8, train_loss: 0.0205 step time: 0.2453\n", + "7/8, train_loss: 0.0144 step time: 0.2579\n", + "8/8, train_loss: 0.0169 step time: 0.3034\n", + "epoch 351 average loss: 0.0162\n", + "time consuming of epoch 351 is: 2.1716\n", + "----------\n", + "epoch 352/600\n", + "1/8, train_loss: 0.0144 step time: 0.3206\n", + "2/8, train_loss: 0.0163 step time: 0.2784\n", + "3/8, train_loss: 0.0163 step time: 0.2682\n", + "4/8, train_loss: 0.0162 step time: 0.2678\n", + "5/8, train_loss: 0.0144 step time: 0.2602\n", + "6/8, train_loss: 0.0144 step time: 0.2583\n", + "7/8, train_loss: 0.0142 step time: 0.2523\n", + "8/8, train_loss: 0.0209 step time: 0.2469\n", + "epoch 352 average loss: 0.0159\n", + "time consuming of epoch 352 is: 2.1543\n", + "----------\n", + "epoch 353/600\n", + "1/8, train_loss: 0.0188 step time: 0.2977\n", + "2/8, train_loss: 0.0169 step time: 0.2969\n", + "3/8, train_loss: 0.0138 step time: 0.2977\n", + "4/8, train_loss: 0.0177 step time: 0.2680\n", + "5/8, train_loss: 0.0151 step time: 0.2523\n", + "6/8, train_loss: 0.0154 step time: 0.2526\n", + "7/8, train_loss: 0.0144 step time: 0.2365\n", + "8/8, train_loss: 0.0120 step time: 0.2307\n", + "epoch 353 average loss: 0.0155\n", + "time consuming of epoch 353 is: 2.1340\n", + "----------\n", + "epoch 354/600\n", + "1/8, train_loss: 0.0181 step time: 0.2850\n", + "2/8, train_loss: 0.0187 step time: 0.2351\n", + "3/8, train_loss: 0.0168 step time: 0.2401\n", + "4/8, train_loss: 0.0161 step time: 0.2391\n", + "5/8, train_loss: 0.0142 step time: 0.2367\n", + "6/8, train_loss: 0.0149 step time: 0.2568\n", + "7/8, train_loss: 0.0151 step time: 0.2754\n", + "8/8, train_loss: 0.0157 step time: 0.2873\n", + "epoch 354 average loss: 0.0162\n", + "time consuming of epoch 354 is: 2.0569\n", + "----------\n", + "epoch 355/600\n", + "1/8, train_loss: 0.0150 step time: 0.3079\n", + "2/8, train_loss: 0.0146 step time: 0.2564\n", + "3/8, train_loss: 0.0161 step time: 0.2485\n", + "4/8, train_loss: 0.0148 step time: 0.5369\n", + "5/8, train_loss: 0.0155 step time: 0.3764\n", + "6/8, train_loss: 0.0165 step time: 0.2687\n", + "7/8, train_loss: 0.0188 step time: 0.2582\n", + "8/8, train_loss: 0.0156 step time: 0.2456\n", + "epoch 355 average loss: 0.0159\n", + "current epoch: 355 current mean dice: 0.9534 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 355 is: 3.9862\n", + "----------\n", + "epoch 356/600\n", + "1/8, train_loss: 0.0189 step time: 0.3007\n", + "2/8, train_loss: 0.0161 step time: 0.2521\n", + "3/8, train_loss: 0.0180 step time: 0.2422\n", + "4/8, train_loss: 0.0167 step time: 0.2424\n", + "5/8, train_loss: 0.0133 step time: 0.2357\n", + "6/8, train_loss: 0.0127 step time: 0.3176\n", + "7/8, train_loss: 0.0152 step time: 0.4719\n", + "8/8, train_loss: 0.0139 step time: 0.2672\n", + "epoch 356 average loss: 0.0156\n", + "time consuming of epoch 356 is: 2.3309\n", + "----------\n", + "epoch 357/600\n", + "1/8, train_loss: 0.0151 step time: 0.3156\n", + "2/8, train_loss: 0.0196 step time: 0.2620\n", + "3/8, train_loss: 0.0160 step time: 0.2544\n", + "4/8, train_loss: 0.0148 step time: 0.2505\n", + "5/8, train_loss: 0.0143 step time: 0.2466\n", + "6/8, train_loss: 0.0153 step time: 0.2397\n", + "7/8, train_loss: 0.0173 step time: 0.2333\n", + "8/8, train_loss: 0.0147 step time: 0.3979\n", + "epoch 357 average loss: 0.0159\n", + "time consuming of epoch 357 is: 2.2016\n", + "----------\n", + "epoch 358/600\n", + "1/8, train_loss: 0.0156 step time: 0.5759\n", + "2/8, train_loss: 0.0139 step time: 0.2902\n", + "3/8, train_loss: 0.0171 step time: 0.2640\n", + "4/8, train_loss: 0.0129 step time: 0.2571\n", + "5/8, train_loss: 0.0179 step time: 0.2455\n", + "6/8, train_loss: 0.0182 step time: 0.2353\n", + "7/8, train_loss: 0.0153 step time: 0.3447\n", + "8/8, train_loss: 0.0142 step time: 0.4597\n", + "epoch 358 average loss: 0.0156\n", + "time consuming of epoch 358 is: 2.6741\n", + "----------\n", + "epoch 359/600\n", + "1/8, train_loss: 0.0192 step time: 0.2903\n", + "2/8, train_loss: 0.0136 step time: 0.2468\n", + "3/8, train_loss: 0.0160 step time: 0.2450\n", + "4/8, train_loss: 0.0159 step time: 0.2581\n", + "5/8, train_loss: 0.0136 step time: 0.2629\n", + "6/8, train_loss: 0.0160 step time: 0.2627\n", + "7/8, train_loss: 0.0133 step time: 0.2610\n", + "8/8, train_loss: 0.0170 step time: 0.2444\n", + "epoch 359 average loss: 0.0156\n", + "time consuming of epoch 359 is: 2.0726\n", + "----------\n", + "epoch 360/600\n", + "1/8, train_loss: 0.0155 step time: 0.2931\n", + "2/8, train_loss: 0.0164 step time: 0.3565\n", + "3/8, train_loss: 0.0112 step time: 0.4710\n", + "4/8, train_loss: 0.0165 step time: 0.2581\n", + "5/8, train_loss: 0.0175 step time: 0.2525\n", + "6/8, train_loss: 0.0157 step time: 0.2507\n", + "7/8, train_loss: 0.0156 step time: 0.2530\n", + "8/8, train_loss: 0.0225 step time: 0.2759\n", + "epoch 360 average loss: 0.0164\n", + "current epoch: 360 current mean dice: 0.9540 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 360 is: 3.6059\n", + "----------\n", + "epoch 361/600\n", + "1/8, train_loss: 0.0155 step time: 0.3069\n", + "2/8, train_loss: 0.0168 step time: 0.2614\n", + "3/8, train_loss: 0.0169 step time: 0.2484\n", + "4/8, train_loss: 0.0216 step time: 0.2491\n", + "5/8, train_loss: 0.0132 step time: 0.2387\n", + "6/8, train_loss: 0.0142 step time: 0.2353\n", + "7/8, train_loss: 0.0143 step time: 0.2256\n", + "8/8, train_loss: 0.0148 step time: 0.2249\n", + "epoch 361 average loss: 0.0159\n", + "time consuming of epoch 361 is: 1.9916\n", + "----------\n", + "epoch 362/600\n", + "1/8, train_loss: 0.0190 step time: 0.2729\n", + "2/8, train_loss: 0.0123 step time: 0.2334\n", + "3/8, train_loss: 0.0180 step time: 0.2414\n", + "4/8, train_loss: 0.0178 step time: 0.2703\n", + "5/8, train_loss: 0.0147 step time: 0.2680\n", + "6/8, train_loss: 0.0179 step time: 0.2673\n", + "7/8, train_loss: 0.0157 step time: 0.2512\n", + "8/8, train_loss: 0.0161 step time: 0.2718\n", + "epoch 362 average loss: 0.0165\n", + "time consuming of epoch 362 is: 2.0778\n", + "----------\n", + "epoch 363/600\n", + "1/8, train_loss: 0.0170 step time: 0.3243\n", + "2/8, train_loss: 0.0147 step time: 0.2720\n", + "3/8, train_loss: 0.0162 step time: 0.2528\n", + "4/8, train_loss: 0.0167 step time: 0.2618\n", + "5/8, train_loss: 0.0138 step time: 0.2741\n", + "6/8, train_loss: 0.0141 step time: 0.2561\n", + "7/8, train_loss: 0.0156 step time: 0.2416\n", + "8/8, train_loss: 0.0197 step time: 0.2436\n", + "epoch 363 average loss: 0.0160\n", + "time consuming of epoch 363 is: 2.1277\n", + "----------\n", + "epoch 364/600\n", + "1/8, train_loss: 0.0140 step time: 0.2933\n", + "2/8, train_loss: 0.0153 step time: 0.2355\n", + "3/8, train_loss: 0.0158 step time: 0.2371\n", + "4/8, train_loss: 0.0169 step time: 0.2482\n", + "5/8, train_loss: 0.0133 step time: 0.2458\n", + "6/8, train_loss: 0.0156 step time: 0.2478\n", + "7/8, train_loss: 0.0151 step time: 0.2672\n", + "8/8, train_loss: 0.0207 step time: 0.2886\n", + "epoch 364 average loss: 0.0158\n", + "time consuming of epoch 364 is: 2.0649\n", + "----------\n", + "epoch 365/600\n", + "1/8, train_loss: 0.0177 step time: 0.3120\n", + "2/8, train_loss: 0.0124 step time: 0.2520\n", + "3/8, train_loss: 0.0177 step time: 0.2564\n", + "4/8, train_loss: 0.0159 step time: 0.2647\n", + "5/8, train_loss: 0.0164 step time: 0.2635\n", + "6/8, train_loss: 0.0159 step time: 0.2508\n", + "7/8, train_loss: 0.0186 step time: 0.2472\n", + "8/8, train_loss: 0.0144 step time: 0.2352\n", + "epoch 365 average loss: 0.0161\n", + "current epoch: 365 current mean dice: 0.9534 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 365 is: 3.5257\n", + "----------\n", + "epoch 366/600\n", + "1/8, train_loss: 0.0164 step time: 0.3106\n", + "2/8, train_loss: 0.0127 step time: 0.2712\n", + "3/8, train_loss: 0.0148 step time: 0.2620\n", + "4/8, train_loss: 0.0162 step time: 0.2638\n", + "5/8, train_loss: 0.0154 step time: 0.2612\n", + "6/8, train_loss: 0.0181 step time: 0.2608\n", + "7/8, train_loss: 0.0167 step time: 0.2630\n", + "8/8, train_loss: 0.0166 step time: 0.3877\n", + "epoch 366 average loss: 0.0159\n", + "time consuming of epoch 366 is: 2.2816\n", + "----------\n", + "epoch 367/600\n", + "1/8, train_loss: 0.0120 step time: 0.4936\n", + "2/8, train_loss: 0.0153 step time: 0.3088\n", + "3/8, train_loss: 0.0182 step time: 0.2684\n", + "4/8, train_loss: 0.0179 step time: 0.2731\n", + "5/8, train_loss: 0.0180 step time: 0.2584\n", + "6/8, train_loss: 0.0171 step time: 0.2477\n", + "7/8, train_loss: 0.0152 step time: 0.2840\n", + "8/8, train_loss: 0.0134 step time: 0.3050\n", + "epoch 367 average loss: 0.0159\n", + "time consuming of epoch 367 is: 2.4404\n", + "----------\n", + "epoch 368/600\n", + "1/8, train_loss: 0.0165 step time: 0.3092\n", + "2/8, train_loss: 0.0167 step time: 0.2499\n", + "3/8, train_loss: 0.0155 step time: 0.2436\n", + "4/8, train_loss: 0.0180 step time: 0.2481\n", + "5/8, train_loss: 0.0167 step time: 0.2472\n", + "6/8, train_loss: 0.0156 step time: 0.2426\n", + "7/8, train_loss: 0.0168 step time: 0.2357\n", + "8/8, train_loss: 0.0157 step time: 0.2213\n", + "epoch 368 average loss: 0.0164\n", + "time consuming of epoch 368 is: 1.9990\n", + "----------\n", + "epoch 369/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0142 step time: 0.5772\n", + "2/8, train_loss: 0.0164 step time: 0.2198\n", + "3/8, train_loss: 0.0154 step time: 0.2141\n", + "4/8, train_loss: 0.0189 step time: 0.2100\n", + "5/8, train_loss: 0.0154 step time: 0.2246\n", + "6/8, train_loss: 0.0169 step time: 0.2283\n", + "7/8, train_loss: 0.0148 step time: 0.2202\n", + "8/8, train_loss: 0.0162 step time: 0.2609\n", + "epoch 369 average loss: 0.0160\n", + "time consuming of epoch 369 is: 2.1565\n", + "----------\n", + "epoch 370/600\n", + "1/8, train_loss: 0.0152 step time: 0.3266\n", + "2/8, train_loss: 0.0165 step time: 0.2921\n", + "3/8, train_loss: 0.0181 step time: 0.2822\n", + "4/8, train_loss: 0.0143 step time: 0.2577\n", + "5/8, train_loss: 0.0144 step time: 0.3602\n", + "6/8, train_loss: 0.0171 step time: 0.3438\n", + "7/8, train_loss: 0.0167 step time: 0.3003\n", + "8/8, train_loss: 0.0130 step time: 0.2783\n", + "epoch 370 average loss: 0.0157\n", + "current epoch: 370 current mean dice: 0.9530 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 370 is: 3.7704\n", + "----------\n", + "epoch 371/600\n", + "1/8, train_loss: 0.0139 step time: 0.3272\n", + "2/8, train_loss: 0.0133 step time: 0.2745\n", + "3/8, train_loss: 0.0170 step time: 0.2601\n", + "4/8, train_loss: 0.0125 step time: 0.3054\n", + "5/8, train_loss: 0.0148 step time: 0.3146\n", + "6/8, train_loss: 0.0193 step time: 0.2895\n", + "7/8, train_loss: 0.0179 step time: 0.2939\n", + "8/8, train_loss: 0.0131 step time: 0.2654\n", + "epoch 371 average loss: 0.0152\n", + "time consuming of epoch 371 is: 2.3320\n", + "----------\n", + "epoch 372/600\n", + "1/8, train_loss: 0.0135 step time: 0.3035\n", + "2/8, train_loss: 0.0154 step time: 0.2586\n", + "3/8, train_loss: 0.0117 step time: 0.2927\n", + "4/8, train_loss: 0.0157 step time: 0.2849\n", + "5/8, train_loss: 0.0159 step time: 0.2856\n", + "6/8, train_loss: 0.0149 step time: 0.2898\n", + "7/8, train_loss: 0.0159 step time: 0.2658\n", + "8/8, train_loss: 0.0188 step time: 0.2715\n", + "epoch 372 average loss: 0.0152\n", + "time consuming of epoch 372 is: 2.2539\n", + "----------\n", + "epoch 373/600\n", + "1/8, train_loss: 0.0172 step time: 0.3073\n", + "2/8, train_loss: 0.0122 step time: 0.2773\n", + "3/8, train_loss: 0.0157 step time: 0.2702\n", + "4/8, train_loss: 0.0156 step time: 0.2581\n", + "5/8, train_loss: 0.0168 step time: 0.2567\n", + "6/8, train_loss: 0.0180 step time: 0.2499\n", + "7/8, train_loss: 0.0151 step time: 0.2600\n", + "8/8, train_loss: 0.0158 step time: 0.2707\n", + "epoch 373 average loss: 0.0158\n", + "time consuming of epoch 373 is: 2.1516\n", + "----------\n", + "epoch 374/600\n", + "1/8, train_loss: 0.0151 step time: 0.3141\n", + "2/8, train_loss: 0.0154 step time: 0.2563\n", + "3/8, train_loss: 0.0140 step time: 0.2523\n", + "4/8, train_loss: 0.0147 step time: 0.2393\n", + "5/8, train_loss: 0.0143 step time: 0.2388\n", + "6/8, train_loss: 0.0166 step time: 0.2484\n", + "7/8, train_loss: 0.0163 step time: 0.2914\n", + "8/8, train_loss: 0.0182 step time: 0.3397\n", + "epoch 374 average loss: 0.0156\n", + "time consuming of epoch 374 is: 2.1819\n", + "----------\n", + "epoch 375/600\n", + "1/8, train_loss: 0.0183 step time: 0.3307\n", + "2/8, train_loss: 0.0165 step time: 0.2600\n", + "3/8, train_loss: 0.0154 step time: 0.4071\n", + "4/8, train_loss: 0.0150 step time: 0.4694\n", + "5/8, train_loss: 0.0187 step time: 0.2161\n", + "6/8, train_loss: 0.0158 step time: 0.2159\n", + "7/8, train_loss: 0.0141 step time: 0.2003\n", + "8/8, train_loss: 0.0118 step time: 0.2076\n", + "epoch 375 average loss: 0.0157\n", + "current epoch: 375 current mean dice: 0.9549 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 375 is: 3.6765\n", + "----------\n", + "epoch 376/600\n", + "1/8, train_loss: 0.0157 step time: 0.3503\n", + "2/8, train_loss: 0.0171 step time: 0.2703\n", + "3/8, train_loss: 0.0148 step time: 0.2553\n", + "4/8, train_loss: 0.0153 step time: 0.2566\n", + "5/8, train_loss: 0.0144 step time: 0.2491\n", + "6/8, train_loss: 0.0123 step time: 0.2520\n", + "7/8, train_loss: 0.0136 step time: 0.2461\n", + "8/8, train_loss: 0.0163 step time: 0.2437\n", + "epoch 376 average loss: 0.0149\n", + "time consuming of epoch 376 is: 2.1247\n", + "----------\n", + "epoch 377/600\n", + "1/8, train_loss: 0.0179 step time: 0.2929\n", + "2/8, train_loss: 0.0148 step time: 0.2495\n", + "3/8, train_loss: 0.0162 step time: 0.3006\n", + "4/8, train_loss: 0.0142 step time: 0.3436\n", + "5/8, train_loss: 0.0152 step time: 0.2696\n", + "6/8, train_loss: 0.0179 step time: 0.2552\n", + "7/8, train_loss: 0.0133 step time: 0.2392\n", + "8/8, train_loss: 0.0131 step time: 0.2409\n", + "epoch 377 average loss: 0.0153\n", + "time consuming of epoch 377 is: 2.1929\n", + "----------\n", + "epoch 378/600\n", + "1/8, train_loss: 0.0138 step time: 0.2944\n", + "2/8, train_loss: 0.0154 step time: 0.2719\n", + "3/8, train_loss: 0.0181 step time: 0.3119\n", + "4/8, train_loss: 0.0142 step time: 0.2746\n", + "5/8, train_loss: 0.0131 step time: 0.2830\n", + "6/8, train_loss: 0.0164 step time: 0.2782\n", + "7/8, train_loss: 0.0134 step time: 0.2583\n", + "8/8, train_loss: 0.0179 step time: 0.2814\n", + "epoch 378 average loss: 0.0153\n", + "time consuming of epoch 378 is: 2.2552\n", + "----------\n", + "epoch 379/600\n", + "1/8, train_loss: 0.0138 step time: 0.3219\n", + "2/8, train_loss: 0.0168 step time: 0.2821\n", + "3/8, train_loss: 0.0136 step time: 0.2629\n", + "4/8, train_loss: 0.0135 step time: 0.2545\n", + "5/8, train_loss: 0.0160 step time: 0.2511\n", + "6/8, train_loss: 0.0171 step time: 0.3398\n", + "7/8, train_loss: 0.0154 step time: 0.3492\n", + "8/8, train_loss: 0.0151 step time: 0.2742\n", + "epoch 379 average loss: 0.0152\n", + "time consuming of epoch 379 is: 2.3372\n", + "----------\n", + "epoch 380/600\n", + "1/8, train_loss: 0.0128 step time: 0.3108\n", + "2/8, train_loss: 0.0136 step time: 0.3037\n", + "3/8, train_loss: 0.0159 step time: 0.3099\n", + "4/8, train_loss: 0.0168 step time: 0.2803\n", + "5/8, train_loss: 0.0142 step time: 0.2685\n", + "6/8, train_loss: 0.0154 step time: 0.2699\n", + "7/8, train_loss: 0.0159 step time: 0.2937\n", + "8/8, train_loss: 0.0177 step time: 0.3155\n", + "epoch 380 average loss: 0.0153\n", + "current epoch: 380 current mean dice: 0.9545 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 380 is: 3.7774\n", + "----------\n", + "epoch 381/600\n", + "1/8, train_loss: 0.0160 step time: 0.3051\n", + "2/8, train_loss: 0.0135 step time: 0.4550\n", + "3/8, train_loss: 0.0181 step time: 0.5773\n", + "4/8, train_loss: 0.0123 step time: 0.2581\n", + "5/8, train_loss: 0.0144 step time: 0.2494\n", + "6/8, train_loss: 0.0186 step time: 0.2434\n", + "7/8, train_loss: 0.0157 step time: 0.2368\n", + "8/8, train_loss: 0.0176 step time: 0.2387\n", + "epoch 381 average loss: 0.0158\n", + "time consuming of epoch 381 is: 2.5650\n", + "----------\n", + "epoch 382/600\n", + "1/8, train_loss: 0.0165 step time: 0.2908\n", + "2/8, train_loss: 0.0132 step time: 0.2466\n", + "3/8, train_loss: 0.0162 step time: 0.2643\n", + "4/8, train_loss: 0.0153 step time: 0.2550\n", + "5/8, train_loss: 0.0165 step time: 0.2875\n", + "6/8, train_loss: 0.0176 step time: 0.3074\n", + "7/8, train_loss: 0.0146 step time: 0.2976\n", + "8/8, train_loss: 0.0146 step time: 0.2763\n", + "epoch 382 average loss: 0.0156\n", + "time consuming of epoch 382 is: 2.2271\n", + "----------\n", + "epoch 383/600\n", + "1/8, train_loss: 0.0186 step time: 0.3163\n", + "2/8, train_loss: 0.0128 step time: 0.2694\n", + "3/8, train_loss: 0.0150 step time: 0.2650\n", + "4/8, train_loss: 0.0154 step time: 0.2550\n", + "5/8, train_loss: 0.0163 step time: 0.2577\n", + "6/8, train_loss: 0.0153 step time: 0.3754\n", + "7/8, train_loss: 0.0184 step time: 0.5047\n", + "8/8, train_loss: 0.0149 step time: 0.2791\n", + "epoch 383 average loss: 0.0159\n", + "time consuming of epoch 383 is: 2.5240\n", + "----------\n", + "epoch 384/600\n", + "1/8, train_loss: 0.0152 step time: 0.3062\n", + "2/8, train_loss: 0.0159 step time: 0.2450\n", + "3/8, train_loss: 0.0123 step time: 0.2423\n", + "4/8, train_loss: 0.0216 step time: 0.2420\n", + "5/8, train_loss: 0.0156 step time: 0.2399\n", + "6/8, train_loss: 0.0138 step time: 0.2488\n", + "7/8, train_loss: 0.0155 step time: 0.2780\n", + "8/8, train_loss: 0.0146 step time: 0.2648\n", + "epoch 384 average loss: 0.0156\n", + "time consuming of epoch 384 is: 2.0685\n", + "----------\n", + "epoch 385/600\n", + "1/8, train_loss: 0.0128 step time: 0.3028\n", + "2/8, train_loss: 0.0172 step time: 0.2529\n", + "3/8, train_loss: 0.0147 step time: 0.2602\n", + "4/8, train_loss: 0.0134 step time: 0.3407\n", + "5/8, train_loss: 0.0208 step time: 0.3611\n", + "6/8, train_loss: 0.0173 step time: 0.2977\n", + "7/8, train_loss: 0.0155 step time: 0.2753\n", + "8/8, train_loss: 0.0152 step time: 0.2572\n", + "epoch 385 average loss: 0.0159\n", + "current epoch: 385 current mean dice: 0.9546 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 385 is: 3.8613\n", + "----------\n", + "epoch 386/600\n", + "1/8, train_loss: 0.0148 step time: 0.3023\n", + "2/8, train_loss: 0.0167 step time: 0.2471\n", + "3/8, train_loss: 0.0160 step time: 0.3439\n", + "4/8, train_loss: 0.0155 step time: 0.4452\n", + "5/8, train_loss: 0.0144 step time: 0.2517\n", + "6/8, train_loss: 0.0145 step time: 0.2453\n", + "7/8, train_loss: 0.0180 step time: 0.2347\n", + "8/8, train_loss: 0.0133 step time: 0.2623\n", + "epoch 386 average loss: 0.0154\n", + "time consuming of epoch 386 is: 2.3338\n", + "----------\n", + "epoch 387/600\n", + "1/8, train_loss: 0.0180 step time: 0.3221\n", + "2/8, train_loss: 0.0147 step time: 0.2618\n", + "3/8, train_loss: 0.0136 step time: 0.2546\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "4/8, train_loss: 0.0203 step time: 0.2452\n", + "5/8, train_loss: 0.0159 step time: 0.3166\n", + "6/8, train_loss: 0.0128 step time: 0.3154\n", + "7/8, train_loss: 0.0160 step time: 0.2815\n", + "8/8, train_loss: 0.0161 step time: 0.3321\n", + "epoch 387 average loss: 0.0159\n", + "time consuming of epoch 387 is: 2.3308\n", + "----------\n", + "epoch 388/600\n", + "1/8, train_loss: 0.0143 step time: 0.3659\n", + "2/8, train_loss: 0.0180 step time: 0.2696\n", + "3/8, train_loss: 0.0151 step time: 0.2619\n", + "4/8, train_loss: 0.0135 step time: 0.2646\n", + "5/8, train_loss: 0.0189 step time: 0.2643\n", + "6/8, train_loss: 0.0159 step time: 0.2616\n", + "7/8, train_loss: 0.0147 step time: 0.2615\n", + "8/8, train_loss: 0.0142 step time: 0.2431\n", + "epoch 388 average loss: 0.0156\n", + "time consuming of epoch 388 is: 2.1941\n", + "----------\n", + "epoch 389/600\n", + "1/8, train_loss: 0.0150 step time: 0.2977\n", + "2/8, train_loss: 0.0139 step time: 0.2454\n", + "3/8, train_loss: 0.0157 step time: 0.2575\n", + "4/8, train_loss: 0.0188 step time: 0.2648\n", + "5/8, train_loss: 0.0157 step time: 0.2622\n", + "6/8, train_loss: 0.0143 step time: 0.2621\n", + "7/8, train_loss: 0.0131 step time: 0.2654\n", + "8/8, train_loss: 0.0143 step time: 0.2777\n", + "epoch 389 average loss: 0.0151\n", + "time consuming of epoch 389 is: 2.1343\n", + "----------\n", + "epoch 390/600\n", + "1/8, train_loss: 0.0139 step time: 0.3109\n", + "2/8, train_loss: 0.0151 step time: 0.2510\n", + "3/8, train_loss: 0.0164 step time: 0.2516\n", + "4/8, train_loss: 0.0132 step time: 0.2506\n", + "5/8, train_loss: 0.0175 step time: 0.2685\n", + "6/8, train_loss: 0.0152 step time: 0.2872\n", + "7/8, train_loss: 0.0134 step time: 0.2670\n", + "8/8, train_loss: 0.0178 step time: 0.4122\n", + "epoch 390 average loss: 0.0153\n", + "current epoch: 390 current mean dice: 0.9549 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 390 is: 3.4081\n", + "----------\n", + "epoch 391/600\n", + "1/8, train_loss: 0.0146 step time: 0.4010\n", + "2/8, train_loss: 0.0143 step time: 0.3074\n", + "3/8, train_loss: 0.0158 step time: 0.2771\n", + "4/8, train_loss: 0.0152 step time: 0.2679\n", + "5/8, train_loss: 0.0132 step time: 0.2797\n", + "6/8, train_loss: 0.0165 step time: 0.2778\n", + "7/8, train_loss: 0.0163 step time: 0.2668\n", + "8/8, train_loss: 0.0151 step time: 0.2764\n", + "epoch 391 average loss: 0.0151\n", + "time consuming of epoch 391 is: 2.3552\n", + "----------\n", + "epoch 392/600\n", + "1/8, train_loss: 0.0160 step time: 0.3073\n", + "2/8, train_loss: 0.0168 step time: 0.2589\n", + "3/8, train_loss: 0.0158 step time: 0.2681\n", + "4/8, train_loss: 0.0184 step time: 0.3522\n", + "5/8, train_loss: 0.0145 step time: 0.3616\n", + "6/8, train_loss: 0.0144 step time: 0.3071\n", + "7/8, train_loss: 0.0168 step time: 0.2716\n", + "8/8, train_loss: 0.0171 step time: 0.2585\n", + "epoch 392 average loss: 0.0162\n", + "time consuming of epoch 392 is: 2.3868\n", + "----------\n", + "epoch 393/600\n", + "1/8, train_loss: 0.0156 step time: 0.2998\n", + "2/8, train_loss: 0.0158 step time: 0.2499\n", + "3/8, train_loss: 0.0144 step time: 0.3094\n", + "4/8, train_loss: 0.0144 step time: 0.4679\n", + "5/8, train_loss: 0.0167 step time: 0.2621\n", + "6/8, train_loss: 0.0147 step time: 0.2595\n", + "7/8, train_loss: 0.0142 step time: 0.2515\n", + "8/8, train_loss: 0.0174 step time: 0.2473\n", + "epoch 393 average loss: 0.0154\n", + "time consuming of epoch 393 is: 2.3489\n", + "----------\n", + "epoch 394/600\n", + "1/8, train_loss: 0.0144 step time: 0.2989\n", + "2/8, train_loss: 0.0170 step time: 0.2422\n", + "3/8, train_loss: 0.0166 step time: 0.4072\n", + "4/8, train_loss: 0.0143 step time: 0.6108\n", + "5/8, train_loss: 0.0178 step time: 0.2654\n", + "6/8, train_loss: 0.0148 step time: 0.2608\n", + "7/8, train_loss: 0.0126 step time: 0.2480\n", + "8/8, train_loss: 0.0146 step time: 0.2453\n", + "epoch 394 average loss: 0.0153\n", + "time consuming of epoch 394 is: 2.5800\n", + "----------\n", + "epoch 395/600\n", + "1/8, train_loss: 0.0168 step time: 0.2927\n", + "2/8, train_loss: 0.0160 step time: 0.2369\n", + "3/8, train_loss: 0.0159 step time: 0.2586\n", + "4/8, train_loss: 0.0149 step time: 0.3128\n", + "5/8, train_loss: 0.0159 step time: 0.2732\n", + "6/8, train_loss: 0.0145 step time: 0.2625\n", + "7/8, train_loss: 0.0180 step time: 0.2609\n", + "8/8, train_loss: 0.0121 step time: 0.2754\n", + "epoch 395 average loss: 0.0155\n", + "current epoch: 395 current mean dice: 0.9545 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 395 is: 3.5276\n", + "----------\n", + "epoch 396/600\n", + "1/8, train_loss: 0.0161 step time: 0.3185\n", + "2/8, train_loss: 0.0118 step time: 0.2689\n", + "3/8, train_loss: 0.0145 step time: 0.2528\n", + "4/8, train_loss: 0.0133 step time: 0.2825\n", + "5/8, train_loss: 0.0147 step time: 0.2658\n", + "6/8, train_loss: 0.0180 step time: 0.2538\n", + "7/8, train_loss: 0.0151 step time: 0.2560\n", + "8/8, train_loss: 0.0159 step time: 0.2392\n", + "epoch 396 average loss: 0.0149\n", + "time consuming of epoch 396 is: 2.1388\n", + "----------\n", + "epoch 397/600\n", + "1/8, train_loss: 0.0143 step time: 0.2840\n", + "2/8, train_loss: 0.0139 step time: 0.2395\n", + "3/8, train_loss: 0.0180 step time: 0.2782\n", + "4/8, train_loss: 0.0160 step time: 0.2807\n", + "5/8, train_loss: 0.0153 step time: 0.2899\n", + "6/8, train_loss: 0.0160 step time: 0.2610\n", + "7/8, train_loss: 0.0184 step time: 0.2605\n", + "8/8, train_loss: 0.0128 step time: 0.4343\n", + "epoch 397 average loss: 0.0156\n", + "time consuming of epoch 397 is: 2.3295\n", + "----------\n", + "epoch 398/600\n", + "1/8, train_loss: 0.0146 step time: 0.3537\n", + "2/8, train_loss: 0.0164 step time: 0.2670\n", + "3/8, train_loss: 0.0145 step time: 0.2563\n", + "4/8, train_loss: 0.0141 step time: 0.2525\n", + "5/8, train_loss: 0.0166 step time: 0.3733\n", + "6/8, train_loss: 0.0231 step time: 0.5910\n", + "7/8, train_loss: 0.0114 step time: 0.2710\n", + "8/8, train_loss: 0.0158 step time: 0.2522\n", + "epoch 398 average loss: 0.0158\n", + "time consuming of epoch 398 is: 2.6184\n", + "----------\n", + "epoch 399/600\n", + "1/8, train_loss: 0.0131 step time: 0.2914\n", + "2/8, train_loss: 0.0147 step time: 0.2418\n", + "3/8, train_loss: 0.0189 step time: 0.2429\n", + "4/8, train_loss: 0.0134 step time: 0.2470\n", + "5/8, train_loss: 0.0157 step time: 0.2693\n", + "6/8, train_loss: 0.0181 step time: 0.2739\n", + "7/8, train_loss: 0.0166 step time: 0.2638\n", + "8/8, train_loss: 0.0175 step time: 0.3156\n", + "epoch 399 average loss: 0.0160\n", + "time consuming of epoch 399 is: 2.1471\n", + "----------\n", + "epoch 400/600\n", + "1/8, train_loss: 0.0162 step time: 0.3245\n", + "2/8, train_loss: 0.0130 step time: 0.2655\n", + "3/8, train_loss: 0.0130 step time: 0.2615\n", + "4/8, train_loss: 0.0174 step time: 0.2600\n", + "5/8, train_loss: 0.0157 step time: 0.2575\n", + "6/8, train_loss: 0.0132 step time: 0.2484\n", + "7/8, train_loss: 0.0160 step time: 0.2443\n", + "8/8, train_loss: 0.0173 step time: 0.2419\n", + "epoch 400 average loss: 0.0152\n", + "current epoch: 400 current mean dice: 0.9546 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 400 is: 3.3862\n", + "----------\n", + "epoch 401/600\n", + "1/8, train_loss: 0.0148 step time: 0.3138\n", + "2/8, train_loss: 0.0151 step time: 0.2786\n", + "3/8, train_loss: 0.0184 step time: 0.2563\n", + "4/8, train_loss: 0.0149 step time: 0.2416\n", + "5/8, train_loss: 0.0114 step time: 0.2539\n", + "6/8, train_loss: 0.0172 step time: 0.2786\n", + "7/8, train_loss: 0.0167 step time: 0.2984\n", + "8/8, train_loss: 0.0186 step time: 0.2876\n", + "epoch 401 average loss: 0.0159\n", + "time consuming of epoch 401 is: 2.2100\n", + "----------\n", + "epoch 402/600\n", + "1/8, train_loss: 0.0173 step time: 0.3061\n", + "2/8, train_loss: 0.0171 step time: 0.2479\n", + "3/8, train_loss: 0.0170 step time: 0.3799\n", + "4/8, train_loss: 0.0115 step time: 0.3250\n", + "5/8, train_loss: 0.0136 step time: 0.2643\n", + "6/8, train_loss: 0.0192 step time: 0.2593\n", + "7/8, train_loss: 0.0153 step time: 0.2490\n", + "8/8, train_loss: 0.0140 step time: 0.2538\n", + "epoch 402 average loss: 0.0156\n", + "time consuming of epoch 402 is: 2.2865\n", + "----------\n", + "epoch 403/600\n", + "1/8, train_loss: 0.0134 step time: 0.3023\n", + "2/8, train_loss: 0.0169 step time: 0.2708\n", + "3/8, train_loss: 0.0168 step time: 0.2569\n", + "4/8, train_loss: 0.0180 step time: 0.2553\n", + "5/8, train_loss: 0.0171 step time: 0.2545\n", + "6/8, train_loss: 0.0131 step time: 0.2498\n", + "7/8, train_loss: 0.0131 step time: 0.2310\n", + "8/8, train_loss: 0.0160 step time: 0.2655\n", + "epoch 403 average loss: 0.0155\n", + "time consuming of epoch 403 is: 2.0875\n", + "----------\n", + "epoch 404/600\n", + "1/8, train_loss: 0.0151 step time: 0.3040\n", + "2/8, train_loss: 0.0144 step time: 0.2564\n", + "3/8, train_loss: 0.0174 step time: 0.2561\n", + "4/8, train_loss: 0.0144 step time: 0.2462\n", + "5/8, train_loss: 0.0127 step time: 0.2466\n", + "6/8, train_loss: 0.0204 step time: 0.3603\n", + "7/8, train_loss: 0.0223 step time: 0.4051\n", + "8/8, train_loss: 0.0140 step time: 0.2636\n", + "epoch 404 average loss: 0.0163\n", + "time consuming of epoch 404 is: 2.3398\n", + "----------\n", + "epoch 405/600\n", + "1/8, train_loss: 0.0136 step time: 0.2991\n", + "2/8, train_loss: 0.0185 step time: 0.2408\n", + "3/8, train_loss: 0.0145 step time: 0.2357\n", + "4/8, train_loss: 0.0145 step time: 0.2390\n", + "5/8, train_loss: 0.0171 step time: 0.2425\n", + "6/8, train_loss: 0.0149 step time: 0.2451\n", + "7/8, train_loss: 0.0173 step time: 0.2384\n", + "8/8, train_loss: 0.0158 step time: 0.2662\n", + "epoch 405 average loss: 0.0158\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "current epoch: 405 current mean dice: 0.9521 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 405 is: 3.3828\n", + "----------\n", + "epoch 406/600\n", + "1/8, train_loss: 0.0149 step time: 0.3092\n", + "2/8, train_loss: 0.0130 step time: 0.2433\n", + "3/8, train_loss: 0.0150 step time: 0.2442\n", + "4/8, train_loss: 0.0133 step time: 0.2565\n", + "5/8, train_loss: 0.0169 step time: 0.3303\n", + "6/8, train_loss: 0.0163 step time: 0.3752\n", + "7/8, train_loss: 0.0158 step time: 0.2776\n", + "8/8, train_loss: 0.0177 step time: 0.2467\n", + "epoch 406 average loss: 0.0154\n", + "time consuming of epoch 406 is: 2.2842\n", + "----------\n", + "epoch 407/600\n", + "1/8, train_loss: 0.0133 step time: 0.2920\n", + "2/8, train_loss: 0.0134 step time: 0.2402\n", + "3/8, train_loss: 0.0151 step time: 0.2528\n", + "4/8, train_loss: 0.0154 step time: 0.2808\n", + "5/8, train_loss: 0.0164 step time: 0.2793\n", + "6/8, train_loss: 0.0223 step time: 0.2849\n", + "7/8, train_loss: 0.0142 step time: 0.2818\n", + "8/8, train_loss: 0.0132 step time: 0.2767\n", + "epoch 407 average loss: 0.0154\n", + "time consuming of epoch 407 is: 2.1900\n", + "----------\n", + "epoch 408/600\n", + "1/8, train_loss: 0.0112 step time: 0.3067\n", + "2/8, train_loss: 0.0164 step time: 0.2575\n", + "3/8, train_loss: 0.0169 step time: 0.2536\n", + "4/8, train_loss: 0.0149 step time: 0.2518\n", + "5/8, train_loss: 0.0193 step time: 0.2544\n", + "6/8, train_loss: 0.0131 step time: 0.2848\n", + "7/8, train_loss: 0.0155 step time: 0.2828\n", + "8/8, train_loss: 0.0150 step time: 0.2550\n", + "epoch 408 average loss: 0.0153\n", + "time consuming of epoch 408 is: 2.1480\n", + "----------\n", + "epoch 409/600\n", + "1/8, train_loss: 0.0151 step time: 0.2982\n", + "2/8, train_loss: 0.0107 step time: 0.2420\n", + "3/8, train_loss: 0.0151 step time: 0.2432\n", + "4/8, train_loss: 0.0167 step time: 0.3666\n", + "5/8, train_loss: 0.0173 step time: 0.4909\n", + "6/8, train_loss: 0.0131 step time: 0.2915\n", + "7/8, train_loss: 0.0162 step time: 0.2730\n", + "8/8, train_loss: 0.0162 step time: 0.2562\n", + "epoch 409 average loss: 0.0150\n", + "time consuming of epoch 409 is: 2.4631\n", + "----------\n", + "epoch 410/600\n", + "1/8, train_loss: 0.0143 step time: 0.2960\n", + "2/8, train_loss: 0.0140 step time: 0.2377\n", + "3/8, train_loss: 0.0205 step time: 0.2342\n", + "4/8, train_loss: 0.0143 step time: 0.3479\n", + "5/8, train_loss: 0.0143 step time: 0.4900\n", + "6/8, train_loss: 0.0160 step time: 0.2794\n", + "7/8, train_loss: 0.0153 step time: 0.2531\n", + "8/8, train_loss: 0.0141 step time: 0.2404\n", + "epoch 410 average loss: 0.0153\n", + "current epoch: 410 current mean dice: 0.9546 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 410 is: 3.6006\n", + "----------\n", + "epoch 411/600\n", + "1/8, train_loss: 0.0115 step time: 0.3314\n", + "2/8, train_loss: 0.0139 step time: 0.2685\n", + "3/8, train_loss: 0.0159 step time: 0.2588\n", + "4/8, train_loss: 0.0147 step time: 0.2547\n", + "5/8, train_loss: 0.0175 step time: 0.2501\n", + "6/8, train_loss: 0.0153 step time: 0.2433\n", + "7/8, train_loss: 0.0148 step time: 0.2320\n", + "8/8, train_loss: 0.0169 step time: 0.2365\n", + "epoch 411 average loss: 0.0151\n", + "time consuming of epoch 411 is: 2.0764\n", + "----------\n", + "epoch 412/600\n", + "1/8, train_loss: 0.0132 step time: 0.5415\n", + "2/8, train_loss: 0.0141 step time: 0.2185\n", + "3/8, train_loss: 0.0175 step time: 0.2162\n", + "4/8, train_loss: 0.0159 step time: 0.2166\n", + "5/8, train_loss: 0.0152 step time: 0.2214\n", + "6/8, train_loss: 0.0145 step time: 0.2739\n", + "7/8, train_loss: 0.0118 step time: 0.4605\n", + "8/8, train_loss: 0.0188 step time: 0.2085\n", + "epoch 412 average loss: 0.0151\n", + "time consuming of epoch 412 is: 2.3586\n", + "----------\n", + "epoch 413/600\n", + "1/8, train_loss: 0.0150 step time: 0.2650\n", + "2/8, train_loss: 0.0145 step time: 0.2171\n", + "3/8, train_loss: 0.0162 step time: 0.2408\n", + "4/8, train_loss: 0.0153 step time: 0.2795\n", + "5/8, train_loss: 0.0157 step time: 0.3120\n", + "6/8, train_loss: 0.0178 step time: 0.2940\n", + "7/8, train_loss: 0.0133 step time: 0.2836\n", + "8/8, train_loss: 0.0161 step time: 0.2812\n", + "epoch 413 average loss: 0.0155\n", + "time consuming of epoch 413 is: 2.1747\n", + "----------\n", + "epoch 414/600\n", + "1/8, train_loss: 0.0143 step time: 0.3054\n", + "2/8, train_loss: 0.0153 step time: 0.2810\n", + "3/8, train_loss: 0.0158 step time: 0.3170\n", + "4/8, train_loss: 0.0125 step time: 0.3107\n", + "5/8, train_loss: 0.0205 step time: 0.3241\n", + "6/8, train_loss: 0.0150 step time: 0.2761\n", + "7/8, train_loss: 0.0122 step time: 0.2590\n", + "8/8, train_loss: 0.0152 step time: 0.2486\n", + "epoch 414 average loss: 0.0151\n", + "time consuming of epoch 414 is: 2.3235\n", + "----------\n", + "epoch 415/600\n", + "1/8, train_loss: 0.0180 step time: 0.3040\n", + "2/8, train_loss: 0.0165 step time: 0.2497\n", + "3/8, train_loss: 0.0133 step time: 0.2473\n", + "4/8, train_loss: 0.0198 step time: 0.2479\n", + "5/8, train_loss: 0.0140 step time: 0.2520\n", + "6/8, train_loss: 0.0130 step time: 0.4107\n", + "7/8, train_loss: 0.0145 step time: 0.4534\n", + "8/8, train_loss: 0.0182 step time: 0.2678\n", + "epoch 415 average loss: 0.0159\n", + "current epoch: 415 current mean dice: 0.9539 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 415 is: 3.7101\n", + "----------\n", + "epoch 416/600\n", + "1/8, train_loss: 0.0132 step time: 0.3039\n", + "2/8, train_loss: 0.0144 step time: 0.2558\n", + "3/8, train_loss: 0.0138 step time: 0.2475\n", + "4/8, train_loss: 0.0135 step time: 0.2887\n", + "5/8, train_loss: 0.0163 step time: 0.3703\n", + "6/8, train_loss: 0.0135 step time: 0.2854\n", + "7/8, train_loss: 0.0151 step time: 0.2610\n", + "8/8, train_loss: 0.0150 step time: 0.2545\n", + "epoch 416 average loss: 0.0144\n", + "time consuming of epoch 416 is: 2.2683\n", + "----------\n", + "epoch 417/600\n", + "1/8, train_loss: 0.0121 step time: 0.2942\n", + "2/8, train_loss: 0.0177 step time: 0.2966\n", + "3/8, train_loss: 0.0144 step time: 0.4228\n", + "4/8, train_loss: 0.0129 step time: 0.2869\n", + "5/8, train_loss: 0.0159 step time: 0.2677\n", + "6/8, train_loss: 0.0151 step time: 0.2543\n", + "7/8, train_loss: 0.0164 step time: 0.2362\n", + "8/8, train_loss: 0.0217 step time: 0.2395\n", + "epoch 417 average loss: 0.0158\n", + "time consuming of epoch 417 is: 2.2996\n", + "----------\n", + "epoch 418/600\n", + "1/8, train_loss: 0.0164 step time: 0.2955\n", + "2/8, train_loss: 0.0148 step time: 0.2506\n", + "3/8, train_loss: 0.0125 step time: 0.3544\n", + "4/8, train_loss: 0.0158 step time: 0.4843\n", + "5/8, train_loss: 0.0169 step time: 0.2687\n", + "6/8, train_loss: 0.0152 step time: 0.2518\n", + "7/8, train_loss: 0.0142 step time: 0.2494\n", + "8/8, train_loss: 0.0167 step time: 0.2522\n", + "epoch 418 average loss: 0.0153\n", + "time consuming of epoch 418 is: 2.4084\n", + "----------\n", + "epoch 419/600\n", + "1/8, train_loss: 0.0126 step time: 0.2995\n", + "2/8, train_loss: 0.0151 step time: 0.2557\n", + "3/8, train_loss: 0.0140 step time: 0.3048\n", + "4/8, train_loss: 0.0169 step time: 0.3084\n", + "5/8, train_loss: 0.0151 step time: 0.3005\n", + "6/8, train_loss: 0.0169 step time: 0.2846\n", + "7/8, train_loss: 0.0148 step time: 0.2979\n", + "8/8, train_loss: 0.0149 step time: 0.2789\n", + "epoch 419 average loss: 0.0151\n", + "time consuming of epoch 419 is: 2.3318\n", + "----------\n", + "epoch 420/600\n", + "1/8, train_loss: 0.0184 step time: 0.3022\n", + "2/8, train_loss: 0.0158 step time: 0.2566\n", + "3/8, train_loss: 0.0148 step time: 0.2549\n", + "4/8, train_loss: 0.0159 step time: 0.2566\n", + "5/8, train_loss: 0.0150 step time: 0.2777\n", + "6/8, train_loss: 0.0170 step time: 0.2618\n", + "7/8, train_loss: 0.0149 step time: 0.3722\n", + "8/8, train_loss: 0.0151 step time: 0.3634\n", + "epoch 420 average loss: 0.0159\n", + "current epoch: 420 current mean dice: 0.9538 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 420 is: 3.5460\n", + "----------\n", + "epoch 421/600\n", + "1/8, train_loss: 0.0139 step time: 0.4007\n", + "2/8, train_loss: 0.0139 step time: 0.2889\n", + "3/8, train_loss: 0.0164 step time: 0.2788\n", + "4/8, train_loss: 0.0160 step time: 0.2597\n", + "5/8, train_loss: 0.0154 step time: 0.2583\n", + "6/8, train_loss: 0.0151 step time: 0.2503\n", + "7/8, train_loss: 0.0144 step time: 0.2639\n", + "8/8, train_loss: 0.0162 step time: 0.2794\n", + "epoch 421 average loss: 0.0152\n", + "time consuming of epoch 421 is: 2.2812\n", + "----------\n", + "epoch 422/600\n", + "1/8, train_loss: 0.0130 step time: 0.3034\n", + "2/8, train_loss: 0.0243 step time: 0.2396\n", + "3/8, train_loss: 0.0155 step time: 0.2344\n", + "4/8, train_loss: 0.0177 step time: 0.2337\n", + "5/8, train_loss: 0.0148 step time: 0.2311\n", + "6/8, train_loss: 0.0141 step time: 0.2356\n", + "7/8, train_loss: 0.0180 step time: 0.2608\n", + "8/8, train_loss: 0.0161 step time: 0.2528\n", + "epoch 422 average loss: 0.0167\n", + "time consuming of epoch 422 is: 1.9928\n", + "----------\n", + "epoch 423/600\n", + "1/8, train_loss: 0.0166 step time: 0.2942\n", + "2/8, train_loss: 0.0147 step time: 0.2445\n", + "3/8, train_loss: 0.0166 step time: 0.2408\n", + "4/8, train_loss: 0.0165 step time: 0.3769\n", + "5/8, train_loss: 0.0153 step time: 0.6298\n", + "6/8, train_loss: 0.0160 step time: 0.2706\n", + "7/8, train_loss: 0.0131 step time: 0.2460\n", + "8/8, train_loss: 0.0142 step time: 0.2255\n", + "epoch 423 average loss: 0.0154\n", + "time consuming of epoch 423 is: 2.5299\n", + "----------\n", + "epoch 424/600\n", + "1/8, train_loss: 0.0150 step time: 0.2779\n", + "2/8, train_loss: 0.0140 step time: 0.2395\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "3/8, train_loss: 0.0176 step time: 0.2502\n", + "4/8, train_loss: 0.0157 step time: 0.3444\n", + "5/8, train_loss: 0.0136 step time: 0.3390\n", + "6/8, train_loss: 0.0149 step time: 0.2949\n", + "7/8, train_loss: 0.0151 step time: 0.2662\n", + "8/8, train_loss: 0.0133 step time: 0.2698\n", + "epoch 424 average loss: 0.0149\n", + "time consuming of epoch 424 is: 2.2832\n", + "----------\n", + "epoch 425/600\n", + "1/8, train_loss: 0.0156 step time: 0.3079\n", + "2/8, train_loss: 0.0144 step time: 0.2622\n", + "3/8, train_loss: 0.0165 step time: 0.2548\n", + "4/8, train_loss: 0.0147 step time: 0.2527\n", + "5/8, train_loss: 0.0132 step time: 0.2480\n", + "6/8, train_loss: 0.0136 step time: 0.2483\n", + "7/8, train_loss: 0.0184 step time: 0.2449\n", + "8/8, train_loss: 0.0135 step time: 0.2905\n", + "epoch 425 average loss: 0.0150\n", + "current epoch: 425 current mean dice: 0.9543 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 425 is: 3.3486\n", + "----------\n", + "epoch 426/600\n", + "1/8, train_loss: 0.0135 step time: 0.2981\n", + "2/8, train_loss: 0.0142 step time: 0.2455\n", + "3/8, train_loss: 0.0152 step time: 0.2692\n", + "4/8, train_loss: 0.0154 step time: 0.2814\n", + "5/8, train_loss: 0.0204 step time: 0.2820\n", + "6/8, train_loss: 0.0145 step time: 0.2821\n", + "7/8, train_loss: 0.0158 step time: 0.2842\n", + "8/8, train_loss: 0.0144 step time: 0.2805\n", + "epoch 426 average loss: 0.0154\n", + "time consuming of epoch 426 is: 2.2241\n", + "----------\n", + "epoch 427/600\n", + "1/8, train_loss: 0.0109 step time: 0.3085\n", + "2/8, train_loss: 0.0147 step time: 0.2674\n", + "3/8, train_loss: 0.0160 step time: 0.2685\n", + "4/8, train_loss: 0.0136 step time: 0.2656\n", + "5/8, train_loss: 0.0149 step time: 0.2656\n", + "6/8, train_loss: 0.0187 step time: 0.2549\n", + "7/8, train_loss: 0.0130 step time: 0.2887\n", + "8/8, train_loss: 0.0144 step time: 0.2772\n", + "epoch 427 average loss: 0.0145\n", + "time consuming of epoch 427 is: 2.1979\n", + "----------\n", + "epoch 428/600\n", + "1/8, train_loss: 0.0127 step time: 0.3149\n", + "2/8, train_loss: 0.0151 step time: 0.2489\n", + "3/8, train_loss: 0.0135 step time: 0.3219\n", + "4/8, train_loss: 0.0190 step time: 0.3073\n", + "5/8, train_loss: 0.0151 step time: 0.2821\n", + "6/8, train_loss: 0.0161 step time: 0.2717\n", + "7/8, train_loss: 0.0153 step time: 0.2699\n", + "8/8, train_loss: 0.0147 step time: 0.2578\n", + "epoch 428 average loss: 0.0152\n", + "time consuming of epoch 428 is: 2.2760\n", + "----------\n", + "epoch 429/600\n", + "1/8, train_loss: 0.0167 step time: 0.2932\n", + "2/8, train_loss: 0.0160 step time: 0.2385\n", + "3/8, train_loss: 0.0152 step time: 0.2378\n", + "4/8, train_loss: 0.0172 step time: 0.3513\n", + "5/8, train_loss: 0.0149 step time: 0.4932\n", + "6/8, train_loss: 0.0133 step time: 0.2753\n", + "7/8, train_loss: 0.0149 step time: 0.2603\n", + "8/8, train_loss: 0.0119 step time: 0.2633\n", + "epoch 429 average loss: 0.0150\n", + "time consuming of epoch 429 is: 2.4145\n", + "----------\n", + "epoch 430/600\n", + "1/8, train_loss: 0.0167 step time: 0.2973\n", + "2/8, train_loss: 0.0162 step time: 0.2453\n", + "3/8, train_loss: 0.0128 step time: 0.2447\n", + "4/8, train_loss: 0.0137 step time: 0.2363\n", + "5/8, train_loss: 0.0149 step time: 0.2365\n", + "6/8, train_loss: 0.0131 step time: 0.2310\n", + "7/8, train_loss: 0.0137 step time: 0.2197\n", + "8/8, train_loss: 0.0153 step time: 0.2207\n", + "epoch 430 average loss: 0.0146\n", + "current epoch: 430 current mean dice: 0.9530 best mean dice: 0.9550 at epoch: 315\n", + "time consuming of epoch 430 is: 3.2540\n", + "----------\n", + "epoch 431/600\n", + "1/8, train_loss: 0.0147 step time: 0.3151\n", + "2/8, train_loss: 0.0159 step time: 0.2642\n", + "3/8, train_loss: 0.0126 step time: 0.2592\n", + "4/8, train_loss: 0.0167 step time: 0.2524\n", + "5/8, train_loss: 0.0168 step time: 0.2520\n", + "6/8, train_loss: 0.0154 step time: 0.2407\n", + "7/8, train_loss: 0.0144 step time: 0.2356\n", + "8/8, train_loss: 0.0143 step time: 0.3086\n", + "epoch 431 average loss: 0.0151\n", + "time consuming of epoch 431 is: 2.1290\n", + "----------\n", + "epoch 432/600\n", + "1/8, train_loss: 0.0147 step time: 0.3433\n", + "2/8, train_loss: 0.0152 step time: 0.2782\n", + "3/8, train_loss: 0.0181 step time: 0.2757\n", + "4/8, train_loss: 0.0135 step time: 0.2513\n", + "5/8, train_loss: 0.0149 step time: 0.2508\n", + "6/8, train_loss: 0.0141 step time: 0.2487\n", + "7/8, train_loss: 0.0157 step time: 0.2408\n", + "8/8, train_loss: 0.0186 step time: 0.2404\n", + "epoch 432 average loss: 0.0156\n", + "time consuming of epoch 432 is: 2.1308\n", + "----------\n", + "epoch 433/600\n", + "1/8, train_loss: 0.0159 step time: 0.2956\n", + "2/8, train_loss: 0.0174 step time: 0.2459\n", + "3/8, train_loss: 0.0179 step time: 0.2408\n", + "4/8, train_loss: 0.0136 step time: 0.2308\n", + "5/8, train_loss: 0.0127 step time: 0.2376\n", + "6/8, train_loss: 0.0145 step time: 0.2377\n", + "7/8, train_loss: 0.0148 step time: 0.2567\n", + "8/8, train_loss: 0.0122 step time: 0.2996\n", + "epoch 433 average loss: 0.0149\n", + "time consuming of epoch 433 is: 2.0462\n", + "----------\n", + "epoch 434/600\n", + "1/8, train_loss: 0.0120 step time: 0.3040\n", + "2/8, train_loss: 0.0140 step time: 0.2563\n", + "3/8, train_loss: 0.0168 step time: 0.2452\n", + "4/8, train_loss: 0.0123 step time: 0.2444\n", + "5/8, train_loss: 0.0144 step time: 0.2381\n", + "6/8, train_loss: 0.0184 step time: 0.2342\n", + "7/8, train_loss: 0.0124 step time: 0.2645\n", + "8/8, train_loss: 0.0160 step time: 0.3644\n", + "epoch 434 average loss: 0.0145\n", + "time consuming of epoch 434 is: 2.1525\n", + "----------\n", + "epoch 435/600\n", + "1/8, train_loss: 0.0175 step time: 0.3285\n", + "2/8, train_loss: 0.0150 step time: 0.2554\n", + "3/8, train_loss: 0.0154 step time: 0.2569\n", + "4/8, train_loss: 0.0161 step time: 0.2704\n", + "5/8, train_loss: 0.0138 step time: 0.2686\n", + "6/8, train_loss: 0.0123 step time: 0.2566\n", + "7/8, train_loss: 0.0158 step time: 0.2467\n", + "8/8, train_loss: 0.0139 step time: 0.2413\n", + "epoch 435 average loss: 0.0150\n", + "saved new best metric model\n", + "current epoch: 435 current mean dice: 0.9551 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 435 is: 4.0019\n", + "----------\n", + "epoch 436/600\n", + "1/8, train_loss: 0.0128 step time: 0.2755\n", + "2/8, train_loss: 0.0154 step time: 0.2164\n", + "3/8, train_loss: 0.0165 step time: 0.2153\n", + "4/8, train_loss: 0.0135 step time: 0.2155\n", + "5/8, train_loss: 0.0172 step time: 0.2204\n", + "6/8, train_loss: 0.0161 step time: 0.2201\n", + "7/8, train_loss: 0.0115 step time: 0.2429\n", + "8/8, train_loss: 0.0152 step time: 0.2824\n", + "epoch 436 average loss: 0.0148\n", + "time consuming of epoch 436 is: 1.8910\n", + "----------\n", + "epoch 437/600\n", + "1/8, train_loss: 0.0153 step time: 0.3138\n", + "2/8, train_loss: 0.0139 step time: 0.2802\n", + "3/8, train_loss: 0.0163 step time: 0.2838\n", + "4/8, train_loss: 0.0133 step time: 0.2676\n", + "5/8, train_loss: 0.0153 step time: 0.2767\n", + "6/8, train_loss: 0.0128 step time: 0.2657\n", + "7/8, train_loss: 0.0143 step time: 0.2649\n", + "8/8, train_loss: 0.0161 step time: 0.2528\n", + "epoch 437 average loss: 0.0147\n", + "time consuming of epoch 437 is: 2.2069\n", + "----------\n", + "epoch 438/600\n", + "1/8, train_loss: 0.0148 step time: 0.2992\n", + "2/8, train_loss: 0.0140 step time: 0.2510\n", + "3/8, train_loss: 0.0145 step time: 0.2475\n", + "4/8, train_loss: 0.0145 step time: 0.2520\n", + "5/8, train_loss: 0.0174 step time: 0.2459\n", + "6/8, train_loss: 0.0160 step time: 0.2473\n", + "7/8, train_loss: 0.0164 step time: 0.2692\n", + "8/8, train_loss: 0.0157 step time: 0.2750\n", + "epoch 438 average loss: 0.0154\n", + "time consuming of epoch 438 is: 2.0885\n", + "----------\n", + "epoch 439/600\n", + "1/8, train_loss: 0.0155 step time: 0.2974\n", + "2/8, train_loss: 0.0166 step time: 0.2586\n", + "3/8, train_loss: 0.0117 step time: 0.2505\n", + "4/8, train_loss: 0.0130 step time: 0.2452\n", + "5/8, train_loss: 0.0146 step time: 0.2674\n", + "6/8, train_loss: 0.0170 step time: 0.3161\n", + "7/8, train_loss: 0.0134 step time: 0.2947\n", + "8/8, train_loss: 0.0169 step time: 0.2598\n", + "epoch 439 average loss: 0.0148\n", + "time consuming of epoch 439 is: 2.1912\n", + "----------\n", + "epoch 440/600\n", + "1/8, train_loss: 0.0135 step time: 0.3042\n", + "2/8, train_loss: 0.0127 step time: 0.2468\n", + "3/8, train_loss: 0.0182 step time: 0.2411\n", + "4/8, train_loss: 0.0136 step time: 0.2335\n", + "5/8, train_loss: 0.0149 step time: 0.2332\n", + "6/8, train_loss: 0.0169 step time: 0.2288\n", + "7/8, train_loss: 0.0148 step time: 0.3976\n", + "8/8, train_loss: 0.0157 step time: 0.4508\n", + "epoch 440 average loss: 0.0150\n", + "current epoch: 440 current mean dice: 0.9544 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 440 is: 3.5602\n", + "----------\n", + "epoch 441/600\n", + "1/8, train_loss: 0.0156 step time: 0.3199\n", + "2/8, train_loss: 0.0156 step time: 0.2643\n", + "3/8, train_loss: 0.0128 step time: 0.2510\n", + "4/8, train_loss: 0.0156 step time: 0.2456\n", + "5/8, train_loss: 0.0149 step time: 0.2475\n", + "6/8, train_loss: 0.0112 step time: 0.2463\n", + "7/8, train_loss: 0.0148 step time: 0.2667\n", + "8/8, train_loss: 0.0148 step time: 0.2973\n", + "epoch 441 average loss: 0.0144\n", + "time consuming of epoch 441 is: 2.1400\n", + "----------\n", + "epoch 442/600\n", + "1/8, train_loss: 0.0141 step time: 0.3177\n", + "2/8, train_loss: 0.0174 step time: 0.2627\n", + "3/8, train_loss: 0.0146 step time: 0.2617\n", + "4/8, train_loss: 0.0122 step time: 0.2622\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "5/8, train_loss: 0.0135 step time: 0.2667\n", + "6/8, train_loss: 0.0187 step time: 0.2940\n", + "7/8, train_loss: 0.0163 step time: 0.2752\n", + "8/8, train_loss: 0.0147 step time: 0.2571\n", + "epoch 442 average loss: 0.0152\n", + "time consuming of epoch 442 is: 2.1988\n", + "----------\n", + "epoch 443/600\n", + "1/8, train_loss: 0.0129 step time: 0.2875\n", + "2/8, train_loss: 0.0172 step time: 0.2430\n", + "3/8, train_loss: 0.0157 step time: 0.3421\n", + "4/8, train_loss: 0.0150 step time: 0.3522\n", + "5/8, train_loss: 0.0125 step time: 0.2767\n", + "6/8, train_loss: 0.0149 step time: 0.2731\n", + "7/8, train_loss: 0.0144 step time: 0.2533\n", + "8/8, train_loss: 0.0139 step time: 0.2480\n", + "epoch 443 average loss: 0.0146\n", + "time consuming of epoch 443 is: 2.2774\n", + "----------\n", + "epoch 444/600\n", + "1/8, train_loss: 0.0174 step time: 0.2874\n", + "2/8, train_loss: 0.0152 step time: 0.2458\n", + "3/8, train_loss: 0.0173 step time: 0.2509\n", + "4/8, train_loss: 0.0135 step time: 0.2483\n", + "5/8, train_loss: 0.0138 step time: 0.2536\n", + "6/8, train_loss: 0.0136 step time: 0.2517\n", + "7/8, train_loss: 0.0143 step time: 0.2445\n", + "8/8, train_loss: 0.0179 step time: 0.2450\n", + "epoch 444 average loss: 0.0154\n", + "time consuming of epoch 444 is: 2.0286\n", + "----------\n", + "epoch 445/600\n", + "1/8, train_loss: 0.0156 step time: 0.2932\n", + "2/8, train_loss: 0.0155 step time: 0.2613\n", + "3/8, train_loss: 0.0158 step time: 0.2615\n", + "4/8, train_loss: 0.0142 step time: 0.2587\n", + "5/8, train_loss: 0.0164 step time: 0.2558\n", + "6/8, train_loss: 0.0124 step time: 0.2498\n", + "7/8, train_loss: 0.0148 step time: 0.2315\n", + "8/8, train_loss: 0.0141 step time: 0.3062\n", + "epoch 445 average loss: 0.0149\n", + "current epoch: 445 current mean dice: 0.9539 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 445 is: 3.6090\n", + "----------\n", + "epoch 446/600\n", + "1/8, train_loss: 0.0160 step time: 0.3007\n", + "2/8, train_loss: 0.0156 step time: 0.2535\n", + "3/8, train_loss: 0.0141 step time: 0.2464\n", + "4/8, train_loss: 0.0136 step time: 0.2431\n", + "5/8, train_loss: 0.0149 step time: 0.3652\n", + "6/8, train_loss: 0.0138 step time: 0.3962\n", + "7/8, train_loss: 0.0151 step time: 0.2716\n", + "8/8, train_loss: 0.0130 step time: 0.2539\n", + "epoch 446 average loss: 0.0145\n", + "time consuming of epoch 446 is: 2.3322\n", + "----------\n", + "epoch 447/600\n", + "1/8, train_loss: 0.0155 step time: 0.2902\n", + "2/8, train_loss: 0.0149 step time: 0.2396\n", + "3/8, train_loss: 0.0144 step time: 0.2692\n", + "4/8, train_loss: 0.0157 step time: 0.2675\n", + "5/8, train_loss: 0.0134 step time: 0.2539\n", + "6/8, train_loss: 0.0163 step time: 0.2527\n", + "7/8, train_loss: 0.0154 step time: 0.2323\n", + "8/8, train_loss: 0.0128 step time: 0.2568\n", + "epoch 447 average loss: 0.0148\n", + "time consuming of epoch 447 is: 2.0636\n", + "----------\n", + "epoch 448/600\n", + "1/8, train_loss: 0.0169 step time: 0.3244\n", + "2/8, train_loss: 0.0145 step time: 0.2560\n", + "3/8, train_loss: 0.0173 step time: 0.2607\n", + "4/8, train_loss: 0.0136 step time: 0.3511\n", + "5/8, train_loss: 0.0164 step time: 0.3774\n", + "6/8, train_loss: 0.0142 step time: 0.2722\n", + "7/8, train_loss: 0.0143 step time: 0.2581\n", + "8/8, train_loss: 0.0148 step time: 0.2500\n", + "epoch 448 average loss: 0.0152\n", + "time consuming of epoch 448 is: 2.3513\n", + "----------\n", + "epoch 449/600\n", + "1/8, train_loss: 0.0132 step time: 0.2998\n", + "2/8, train_loss: 0.0159 step time: 0.2523\n", + "3/8, train_loss: 0.0144 step time: 0.2458\n", + "4/8, train_loss: 0.0173 step time: 0.2503\n", + "5/8, train_loss: 0.0146 step time: 0.2416\n", + "6/8, train_loss: 0.0156 step time: 0.2418\n", + "7/8, train_loss: 0.0142 step time: 0.2915\n", + "8/8, train_loss: 0.0155 step time: 0.3002\n", + "epoch 449 average loss: 0.0151\n", + "time consuming of epoch 449 is: 2.1249\n", + "----------\n", + "epoch 450/600\n", + "1/8, train_loss: 0.0168 step time: 0.3177\n", + "2/8, train_loss: 0.0133 step time: 0.2538\n", + "3/8, train_loss: 0.0136 step time: 0.2469\n", + "4/8, train_loss: 0.0160 step time: 0.2632\n", + "5/8, train_loss: 0.0156 step time: 0.2656\n", + "6/8, train_loss: 0.0158 step time: 0.4643\n", + "7/8, train_loss: 0.0148 step time: 0.4271\n", + "8/8, train_loss: 0.0142 step time: 0.2558\n", + "epoch 450 average loss: 0.0150\n", + "current epoch: 450 current mean dice: 0.9550 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 450 is: 3.9072\n", + "----------\n", + "epoch 451/600\n", + "1/8, train_loss: 0.0138 step time: 0.2930\n", + "2/8, train_loss: 0.0134 step time: 0.2369\n", + "3/8, train_loss: 0.0140 step time: 0.2385\n", + "4/8, train_loss: 0.0144 step time: 0.2341\n", + "5/8, train_loss: 0.0161 step time: 0.2397\n", + "6/8, train_loss: 0.0155 step time: 0.2360\n", + "7/8, train_loss: 0.0128 step time: 0.2232\n", + "8/8, train_loss: 0.0157 step time: 0.2177\n", + "epoch 451 average loss: 0.0145\n", + "time consuming of epoch 451 is: 1.9204\n", + "----------\n", + "epoch 452/600\n", + "1/8, train_loss: 0.0185 step time: 0.2838\n", + "2/8, train_loss: 0.0132 step time: 0.2305\n", + "3/8, train_loss: 0.0178 step time: 0.2306\n", + "4/8, train_loss: 0.0114 step time: 0.2289\n", + "5/8, train_loss: 0.0157 step time: 0.2292\n", + "6/8, train_loss: 0.0139 step time: 0.2275\n", + "7/8, train_loss: 0.0147 step time: 0.2183\n", + "8/8, train_loss: 0.0160 step time: 0.2197\n", + "epoch 452 average loss: 0.0152\n", + "time consuming of epoch 452 is: 1.8699\n", + "----------\n", + "epoch 453/600\n", + "1/8, train_loss: 0.0220 step time: 0.2805\n", + "2/8, train_loss: 0.0121 step time: 0.2313\n", + "3/8, train_loss: 0.0132 step time: 0.2277\n", + "4/8, train_loss: 0.0163 step time: 0.2330\n", + "5/8, train_loss: 0.0140 step time: 0.2762\n", + "6/8, train_loss: 0.0162 step time: 0.2841\n", + "7/8, train_loss: 0.0132 step time: 0.2785\n", + "8/8, train_loss: 0.0149 step time: 0.2854\n", + "epoch 453 average loss: 0.0152\n", + "time consuming of epoch 453 is: 2.0980\n", + "----------\n", + "epoch 454/600\n", + "1/8, train_loss: 0.0159 step time: 0.3185\n", + "2/8, train_loss: 0.0116 step time: 0.2604\n", + "3/8, train_loss: 0.0153 step time: 0.2616\n", + "4/8, train_loss: 0.0198 step time: 0.2611\n", + "5/8, train_loss: 0.0135 step time: 0.2537\n", + "6/8, train_loss: 0.0182 step time: 0.2511\n", + "7/8, train_loss: 0.0131 step time: 0.3452\n", + "8/8, train_loss: 0.0128 step time: 0.3181\n", + "epoch 454 average loss: 0.0150\n", + "time consuming of epoch 454 is: 2.2712\n", + "----------\n", + "epoch 455/600\n", + "1/8, train_loss: 0.0145 step time: 0.3069\n", + "2/8, train_loss: 0.0173 step time: 0.2569\n", + "3/8, train_loss: 0.0124 step time: 0.2507\n", + "4/8, train_loss: 0.0152 step time: 0.2495\n", + "5/8, train_loss: 0.0145 step time: 0.2803\n", + "6/8, train_loss: 0.0149 step time: 0.2717\n", + "7/8, train_loss: 0.0170 step time: 0.2583\n", + "8/8, train_loss: 0.0133 step time: 0.2614\n", + "epoch 455 average loss: 0.0149\n", + "current epoch: 455 current mean dice: 0.9541 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 455 is: 3.4658\n", + "----------\n", + "epoch 456/600\n", + "1/8, train_loss: 0.0156 step time: 0.2978\n", + "2/8, train_loss: 0.0141 step time: 0.2519\n", + "3/8, train_loss: 0.0168 step time: 0.2469\n", + "4/8, train_loss: 0.0153 step time: 0.2395\n", + "5/8, train_loss: 0.0148 step time: 0.2400\n", + "6/8, train_loss: 0.0157 step time: 0.2455\n", + "7/8, train_loss: 0.0157 step time: 0.2372\n", + "8/8, train_loss: 0.0146 step time: 0.2534\n", + "epoch 456 average loss: 0.0153\n", + "time consuming of epoch 456 is: 2.0133\n", + "----------\n", + "epoch 457/600\n", + "1/8, train_loss: 0.0164 step time: 0.3359\n", + "2/8, train_loss: 0.0154 step time: 0.2799\n", + "3/8, train_loss: 0.0153 step time: 0.2511\n", + "4/8, train_loss: 0.0131 step time: 0.2475\n", + "5/8, train_loss: 0.0133 step time: 0.2502\n", + "6/8, train_loss: 0.0147 step time: 0.2409\n", + "7/8, train_loss: 0.0101 step time: 0.2336\n", + "8/8, train_loss: 0.0181 step time: 0.2324\n", + "epoch 457 average loss: 0.0146\n", + "time consuming of epoch 457 is: 2.0730\n", + "----------\n", + "epoch 458/600\n", + "1/8, train_loss: 0.0142 step time: 0.2889\n", + "2/8, train_loss: 0.0136 step time: 0.2387\n", + "3/8, train_loss: 0.0171 step time: 0.2382\n", + "4/8, train_loss: 0.0164 step time: 0.2359\n", + "5/8, train_loss: 0.0132 step time: 0.2333\n", + "6/8, train_loss: 0.0115 step time: 0.3759\n", + "7/8, train_loss: 0.0150 step time: 0.4552\n", + "8/8, train_loss: 0.0172 step time: 0.2565\n", + "epoch 458 average loss: 0.0148\n", + "time consuming of epoch 458 is: 2.3238\n", + "----------\n", + "epoch 459/600\n", + "1/8, train_loss: 0.0161 step time: 0.3029\n", + "2/8, train_loss: 0.0162 step time: 0.2362\n", + "3/8, train_loss: 0.0144 step time: 0.2384\n", + "4/8, train_loss: 0.0136 step time: 0.2400\n", + "5/8, train_loss: 0.0126 step time: 0.2406\n", + "6/8, train_loss: 0.0134 step time: 0.2319\n", + "7/8, train_loss: 0.0187 step time: 0.2360\n", + "8/8, train_loss: 0.0154 step time: 0.2644\n", + "epoch 459 average loss: 0.0151\n", + "time consuming of epoch 459 is: 1.9919\n", + "----------\n", + "epoch 460/600\n", + "1/8, train_loss: 0.0145 step time: 0.3175\n", + "2/8, train_loss: 0.0147 step time: 0.2627\n", + "3/8, train_loss: 0.0152 step time: 0.2569\n", + "4/8, train_loss: 0.0135 step time: 0.2575\n", + "5/8, train_loss: 0.0142 step time: 0.2485\n", + "6/8, train_loss: 0.0147 step time: 0.2396\n", + "7/8, train_loss: 0.0160 step time: 0.2328\n", + "8/8, train_loss: 0.0155 step time: 0.2266\n", + "epoch 460 average loss: 0.0148\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "current epoch: 460 current mean dice: 0.9547 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 460 is: 3.2773\n", + "----------\n", + "epoch 461/600\n", + "1/8, train_loss: 0.0155 step time: 0.3138\n", + "2/8, train_loss: 0.0145 step time: 0.2652\n", + "3/8, train_loss: 0.0187 step time: 0.2638\n", + "4/8, train_loss: 0.0169 step time: 0.2636\n", + "5/8, train_loss: 0.0117 step time: 0.2605\n", + "6/8, train_loss: 0.0140 step time: 0.2580\n", + "7/8, train_loss: 0.0134 step time: 0.2541\n", + "8/8, train_loss: 0.0143 step time: 0.2554\n", + "epoch 461 average loss: 0.0149\n", + "time consuming of epoch 461 is: 2.1355\n", + "----------\n", + "epoch 462/600\n", + "1/8, train_loss: 0.0144 step time: 0.3150\n", + "2/8, train_loss: 0.0146 step time: 0.2540\n", + "3/8, train_loss: 0.0113 step time: 0.2555\n", + "4/8, train_loss: 0.0152 step time: 0.2430\n", + "5/8, train_loss: 0.0136 step time: 0.2400\n", + "6/8, train_loss: 0.0152 step time: 0.2412\n", + "7/8, train_loss: 0.0165 step time: 0.2667\n", + "8/8, train_loss: 0.0149 step time: 0.3267\n", + "epoch 462 average loss: 0.0145\n", + "time consuming of epoch 462 is: 2.1435\n", + "----------\n", + "epoch 463/600\n", + "1/8, train_loss: 0.0165 step time: 0.3185\n", + "2/8, train_loss: 0.0137 step time: 0.2549\n", + "3/8, train_loss: 0.0159 step time: 0.2576\n", + "4/8, train_loss: 0.0140 step time: 0.2502\n", + "5/8, train_loss: 0.0155 step time: 0.2506\n", + "6/8, train_loss: 0.0159 step time: 0.2482\n", + "7/8, train_loss: 0.0136 step time: 0.2421\n", + "8/8, train_loss: 0.0179 step time: 0.2317\n", + "epoch 463 average loss: 0.0154\n", + "time consuming of epoch 463 is: 2.0553\n", + "----------\n", + "epoch 464/600\n", + "1/8, train_loss: 0.0175 step time: 0.2874\n", + "2/8, train_loss: 0.0153 step time: 0.2330\n", + "3/8, train_loss: 0.0157 step time: 0.2649\n", + "4/8, train_loss: 0.0144 step time: 0.2802\n", + "5/8, train_loss: 0.0145 step time: 0.2749\n", + "6/8, train_loss: 0.0139 step time: 0.2755\n", + "7/8, train_loss: 0.0136 step time: 0.3063\n", + "8/8, train_loss: 0.0159 step time: 0.2904\n", + "epoch 464 average loss: 0.0151\n", + "time consuming of epoch 464 is: 2.2140\n", + "----------\n", + "epoch 465/600\n", + "1/8, train_loss: 0.0143 step time: 0.3027\n", + "2/8, train_loss: 0.0157 step time: 0.2467\n", + "3/8, train_loss: 0.0138 step time: 0.2521\n", + "4/8, train_loss: 0.0137 step time: 0.2473\n", + "5/8, train_loss: 0.0145 step time: 0.2420\n", + "6/8, train_loss: 0.0174 step time: 0.2459\n", + "7/8, train_loss: 0.0162 step time: 0.2326\n", + "8/8, train_loss: 0.0121 step time: 0.2325\n", + "epoch 465 average loss: 0.0147\n", + "current epoch: 465 current mean dice: 0.9540 best mean dice: 0.9551 at epoch: 435\n", + "time consuming of epoch 465 is: 3.3323\n", + "----------\n", + "epoch 466/600\n", + "1/8, train_loss: 0.0150 step time: 0.3229\n", + "2/8, train_loss: 0.0142 step time: 0.2761\n", + "3/8, train_loss: 0.0145 step time: 0.2561\n", + "4/8, train_loss: 0.0148 step time: 0.2588\n", + "5/8, train_loss: 0.0150 step time: 0.4052\n", + "6/8, train_loss: 0.0128 step time: 0.4585\n", + "7/8, train_loss: 0.0128 step time: 0.2029\n", + "8/8, train_loss: 0.0149 step time: 0.1997\n", + "epoch 466 average loss: 0.0142\n", + "time consuming of epoch 466 is: 2.3815\n", + "----------\n", + "epoch 467/600\n", + "1/8, train_loss: 0.0144 step time: 0.2667\n", + "2/8, train_loss: 0.0141 step time: 0.2169\n", + "3/8, train_loss: 0.0141 step time: 0.2123\n", + "4/8, train_loss: 0.0163 step time: 0.2608\n", + "5/8, train_loss: 0.0119 step time: 0.3160\n", + "6/8, train_loss: 0.0148 step time: 0.3083\n", + "7/8, train_loss: 0.0156 step time: 0.2919\n", + "8/8, train_loss: 0.0162 step time: 0.2674\n", + "epoch 467 average loss: 0.0147\n", + "time consuming of epoch 467 is: 2.1417\n", + "----------\n", + "epoch 468/600\n", + "1/8, train_loss: 0.0110 step time: 0.2963\n", + "2/8, train_loss: 0.0139 step time: 0.2463\n", + "3/8, train_loss: 0.0144 step time: 0.2426\n", + "4/8, train_loss: 0.0155 step time: 0.2400\n", + "5/8, train_loss: 0.0142 step time: 0.2429\n", + "6/8, train_loss: 0.0156 step time: 0.2340\n", + "7/8, train_loss: 0.0179 step time: 0.2190\n", + "8/8, train_loss: 0.0128 step time: 0.2212\n", + "epoch 468 average loss: 0.0144\n", + "time consuming of epoch 468 is: 1.9437\n", + "----------\n", + "epoch 469/600\n", + "1/8, train_loss: 0.0164 step time: 0.2822\n", + "2/8, train_loss: 0.0145 step time: 0.2337\n", + "3/8, train_loss: 0.0184 step time: 0.2304\n", + "4/8, train_loss: 0.0143 step time: 0.2280\n", + "5/8, train_loss: 0.0149 step time: 0.2290\n", + "6/8, train_loss: 0.0137 step time: 0.2348\n", + "7/8, train_loss: 0.0129 step time: 0.2291\n", + "8/8, train_loss: 0.0126 step time: 0.2767\n", + "epoch 469 average loss: 0.0147\n", + "time consuming of epoch 469 is: 1.9454\n", + "----------\n", + "epoch 470/600\n", + "1/8, train_loss: 0.0141 step time: 0.3345\n", + "2/8, train_loss: 0.0139 step time: 0.2614\n", + "3/8, train_loss: 0.0166 step time: 0.2536\n", + "4/8, train_loss: 0.0158 step time: 0.2520\n", + "5/8, train_loss: 0.0141 step time: 0.2932\n", + "6/8, train_loss: 0.0155 step time: 0.3049\n", + "7/8, train_loss: 0.0108 step time: 0.2592\n", + "8/8, train_loss: 0.0138 step time: 0.2553\n", + "epoch 470 average loss: 0.0143\n", + "saved new best metric model\n", + "current epoch: 470 current mean dice: 0.9554 best mean dice: 0.9554 at epoch: 470\n", + "time consuming of epoch 470 is: 3.8738\n", + "----------\n", + "epoch 471/600\n", + "1/8, train_loss: 0.0156 step time: 0.2769\n", + "2/8, train_loss: 0.0140 step time: 0.2128\n", + "3/8, train_loss: 0.0215 step time: 0.2125\n", + "4/8, train_loss: 0.0133 step time: 0.2178\n", + "5/8, train_loss: 0.0154 step time: 0.2344\n", + "6/8, train_loss: 0.0155 step time: 0.2448\n", + "7/8, train_loss: 0.0148 step time: 0.2474\n", + "8/8, train_loss: 0.0133 step time: 0.2392\n", + "epoch 471 average loss: 0.0154\n", + "time consuming of epoch 471 is: 1.8882\n", + "----------\n", + "epoch 472/600\n", + "1/8, train_loss: 0.0155 step time: 0.2938\n", + "2/8, train_loss: 0.0183 step time: 0.2460\n", + "3/8, train_loss: 0.0109 step time: 0.2394\n", + "4/8, train_loss: 0.0164 step time: 0.2422\n", + "5/8, train_loss: 0.0161 step time: 0.2393\n", + "6/8, train_loss: 0.0162 step time: 0.2398\n", + "7/8, train_loss: 0.0170 step time: 0.2231\n", + "8/8, train_loss: 0.0164 step time: 0.2213\n", + "epoch 472 average loss: 0.0158\n", + "time consuming of epoch 472 is: 1.9466\n", + "----------\n", + "epoch 473/600\n", + "1/8, train_loss: 0.0136 step time: 0.2790\n", + "2/8, train_loss: 0.0160 step time: 0.2329\n", + "3/8, train_loss: 0.0183 step time: 0.2446\n", + "4/8, train_loss: 0.0143 step time: 0.2610\n", + "5/8, train_loss: 0.0155 step time: 0.3160\n", + "6/8, train_loss: 0.0143 step time: 0.3516\n", + "7/8, train_loss: 0.0112 step time: 0.2948\n", + "8/8, train_loss: 0.0187 step time: 0.2710\n", + "epoch 473 average loss: 0.0152\n", + "time consuming of epoch 473 is: 2.2521\n", + "----------\n", + "epoch 474/600\n", + "1/8, train_loss: 0.0133 step time: 0.3126\n", + "2/8, train_loss: 0.0144 step time: 0.3610\n", + "3/8, train_loss: 0.0172 step time: 0.3268\n", + "4/8, train_loss: 0.0153 step time: 0.2781\n", + "5/8, train_loss: 0.0149 step time: 0.2557\n", + "6/8, train_loss: 0.0165 step time: 0.2544\n", + "7/8, train_loss: 0.0112 step time: 0.2359\n", + "8/8, train_loss: 0.0182 step time: 0.2390\n", + "epoch 474 average loss: 0.0151\n", + "time consuming of epoch 474 is: 2.2651\n", + "----------\n", + "epoch 475/600\n", + "1/8, train_loss: 0.0144 step time: 0.3054\n", + "2/8, train_loss: 0.0160 step time: 0.2510\n", + "3/8, train_loss: 0.0160 step time: 0.2531\n", + "4/8, train_loss: 0.0133 step time: 0.2580\n", + "5/8, train_loss: 0.0165 step time: 0.2775\n", + "6/8, train_loss: 0.0136 step time: 0.2529\n", + "7/8, train_loss: 0.0162 step time: 0.2443\n", + "8/8, train_loss: 0.0149 step time: 0.2431\n", + "epoch 475 average loss: 0.0151\n", + "current epoch: 475 current mean dice: 0.9552 best mean dice: 0.9554 at epoch: 470\n", + "time consuming of epoch 475 is: 3.2465\n", + "----------\n", + "epoch 476/600\n", + "1/8, train_loss: 0.0147 step time: 0.3742\n", + "2/8, train_loss: 0.0196 step time: 0.2866\n", + "3/8, train_loss: 0.0133 step time: 0.2521\n", + "4/8, train_loss: 0.0140 step time: 0.2535\n", + "5/8, train_loss: 0.0149 step time: 0.2420\n", + "6/8, train_loss: 0.0147 step time: 0.2487\n", + "7/8, train_loss: 0.0185 step time: 0.2804\n", + "8/8, train_loss: 0.0132 step time: 0.2710\n", + "epoch 476 average loss: 0.0154\n", + "time consuming of epoch 476 is: 2.2096\n", + "----------\n", + "epoch 477/600\n", + "1/8, train_loss: 0.0204 step time: 0.3028\n", + "2/8, train_loss: 0.0142 step time: 0.2947\n", + "3/8, train_loss: 0.0129 step time: 0.3529\n", + "4/8, train_loss: 0.0152 step time: 0.2810\n", + "5/8, train_loss: 0.0164 step time: 0.2595\n", + "6/8, train_loss: 0.0141 step time: 0.2475\n", + "7/8, train_loss: 0.0135 step time: 0.2326\n", + "8/8, train_loss: 0.0161 step time: 0.2363\n", + "epoch 477 average loss: 0.0154\n", + "time consuming of epoch 477 is: 2.2088\n", + "----------\n", + "epoch 478/600\n", + "1/8, train_loss: 0.0150 step time: 0.2893\n", + "2/8, train_loss: 0.0141 step time: 0.2333\n", + "3/8, train_loss: 0.0189 step time: 0.2824\n", + "4/8, train_loss: 0.0146 step time: 0.3096\n", + "5/8, train_loss: 0.0166 step time: 0.2638\n", + "6/8, train_loss: 0.0136 step time: 0.2552\n", + "7/8, train_loss: 0.0135 step time: 0.3019\n", + "8/8, train_loss: 0.0152 step time: 0.3165\n", + "epoch 478 average loss: 0.0152\n", + "time consuming of epoch 478 is: 2.2534\n", + "----------\n", + "epoch 479/600\n", + "1/8, train_loss: 0.0141 step time: 0.3209\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "2/8, train_loss: 0.0139 step time: 0.2661\n", + "3/8, train_loss: 0.0142 step time: 0.2618\n", + "4/8, train_loss: 0.0177 step time: 0.2494\n", + "5/8, train_loss: 0.0144 step time: 0.2530\n", + "6/8, train_loss: 0.0143 step time: 0.3314\n", + "7/8, train_loss: 0.0150 step time: 0.2974\n", + "8/8, train_loss: 0.0146 step time: 0.2646\n", + "epoch 479 average loss: 0.0148\n", + "time consuming of epoch 479 is: 2.2460\n", + "----------\n", + "epoch 480/600\n", + "1/8, train_loss: 0.0157 step time: 0.2989\n", + "2/8, train_loss: 0.0144 step time: 0.2578\n", + "3/8, train_loss: 0.0134 step time: 0.2620\n", + "4/8, train_loss: 0.0149 step time: 0.2540\n", + "5/8, train_loss: 0.0133 step time: 0.2488\n", + "6/8, train_loss: 0.0150 step time: 0.2362\n", + "7/8, train_loss: 0.0144 step time: 0.2291\n", + "8/8, train_loss: 0.0184 step time: 0.2233\n", + "epoch 480 average loss: 0.0149\n", + "current epoch: 480 current mean dice: 0.9554 best mean dice: 0.9554 at epoch: 470\n", + "time consuming of epoch 480 is: 3.3287\n", + "----------\n", + "epoch 481/600\n", + "1/8, train_loss: 0.0141 step time: 0.3229\n", + "2/8, train_loss: 0.0150 step time: 0.2605\n", + "3/8, train_loss: 0.0170 step time: 0.2603\n", + "4/8, train_loss: 0.0146 step time: 0.2558\n", + "5/8, train_loss: 0.0134 step time: 0.2550\n", + "6/8, train_loss: 0.0136 step time: 0.2429\n", + "7/8, train_loss: 0.0145 step time: 0.2360\n", + "8/8, train_loss: 0.0149 step time: 0.2314\n", + "epoch 481 average loss: 0.0146\n", + "time consuming of epoch 481 is: 2.0661\n", + "----------\n", + "epoch 482/600\n", + "1/8, train_loss: 0.0166 step time: 0.2823\n", + "2/8, train_loss: 0.0171 step time: 0.2336\n", + "3/8, train_loss: 0.0153 step time: 0.2496\n", + "4/8, train_loss: 0.0132 step time: 0.2803\n", + "5/8, train_loss: 0.0144 step time: 0.2598\n", + "6/8, train_loss: 0.0136 step time: 0.2554\n", + "7/8, train_loss: 0.0143 step time: 0.2433\n", + "8/8, train_loss: 0.0137 step time: 0.2441\n", + "epoch 482 average loss: 0.0148\n", + "time consuming of epoch 482 is: 2.0500\n", + "----------\n", + "epoch 483/600\n", + "1/8, train_loss: 0.0133 step time: 0.2910\n", + "2/8, train_loss: 0.0160 step time: 0.2315\n", + "3/8, train_loss: 0.0136 step time: 0.2338\n", + "4/8, train_loss: 0.0178 step time: 0.2300\n", + "5/8, train_loss: 0.0131 step time: 0.2344\n", + "6/8, train_loss: 0.0148 step time: 0.2686\n", + "7/8, train_loss: 0.0134 step time: 0.2845\n", + "8/8, train_loss: 0.0143 step time: 0.2772\n", + "epoch 483 average loss: 0.0145\n", + "time consuming of epoch 483 is: 2.0524\n", + "----------\n", + "epoch 484/600\n", + "1/8, train_loss: 0.0154 step time: 0.3146\n", + "2/8, train_loss: 0.0158 step time: 0.2606\n", + "3/8, train_loss: 0.0171 step time: 0.2619\n", + "4/8, train_loss: 0.0133 step time: 0.2604\n", + "5/8, train_loss: 0.0166 step time: 0.2625\n", + "6/8, train_loss: 0.0139 step time: 0.2575\n", + "7/8, train_loss: 0.0139 step time: 0.2417\n", + "8/8, train_loss: 0.0143 step time: 0.2429\n", + "epoch 484 average loss: 0.0150\n", + "time consuming of epoch 484 is: 2.1036\n", + "----------\n", + "epoch 485/600\n", + "1/8, train_loss: 0.0133 step time: 0.2964\n", + "2/8, train_loss: 0.0141 step time: 0.2497\n", + "3/8, train_loss: 0.0145 step time: 0.2524\n", + "4/8, train_loss: 0.0153 step time: 0.2486\n", + "5/8, train_loss: 0.0159 step time: 0.2936\n", + "6/8, train_loss: 0.0156 step time: 0.3017\n", + "7/8, train_loss: 0.0129 step time: 0.2895\n", + "8/8, train_loss: 0.0143 step time: 0.2630\n", + "epoch 485 average loss: 0.0145\n", + "current epoch: 485 current mean dice: 0.9552 best mean dice: 0.9554 at epoch: 470\n", + "time consuming of epoch 485 is: 3.3863\n", + "----------\n", + "epoch 486/600\n", + "1/8, train_loss: 0.0112 step time: 0.2874\n", + "2/8, train_loss: 0.0152 step time: 0.2324\n", + "3/8, train_loss: 0.0185 step time: 0.2861\n", + "4/8, train_loss: 0.0140 step time: 0.3710\n", + "5/8, train_loss: 0.0131 step time: 0.2781\n", + "6/8, train_loss: 0.0123 step time: 0.2649\n", + "7/8, train_loss: 0.0169 step time: 0.2526\n", + "8/8, train_loss: 0.0161 step time: 0.2593\n", + "epoch 486 average loss: 0.0147\n", + "time consuming of epoch 486 is: 2.2330\n", + "----------\n", + "epoch 487/600\n", + "1/8, train_loss: 0.0132 step time: 0.3407\n", + "2/8, train_loss: 0.0131 step time: 0.2681\n", + "3/8, train_loss: 0.0151 step time: 0.2612\n", + "4/8, train_loss: 0.0144 step time: 0.2514\n", + "5/8, train_loss: 0.0139 step time: 0.2487\n", + "6/8, train_loss: 0.0144 step time: 0.2951\n", + "7/8, train_loss: 0.0148 step time: 0.3359\n", + "8/8, train_loss: 0.0123 step time: 0.2927\n", + "epoch 487 average loss: 0.0139\n", + "time consuming of epoch 487 is: 2.2952\n", + "----------\n", + "epoch 488/600\n", + "1/8, train_loss: 0.0154 step time: 0.3051\n", + "2/8, train_loss: 0.0140 step time: 0.2491\n", + "3/8, train_loss: 0.0119 step time: 0.2532\n", + "4/8, train_loss: 0.0114 step time: 0.2449\n", + "5/8, train_loss: 0.0137 step time: 0.2664\n", + "6/8, train_loss: 0.0147 step time: 0.2636\n", + "7/8, train_loss: 0.0192 step time: 0.2845\n", + "8/8, train_loss: 0.0153 step time: 0.2975\n", + "epoch 488 average loss: 0.0145\n", + "time consuming of epoch 488 is: 2.1659\n", + "----------\n", + "epoch 489/600\n", + "1/8, train_loss: 0.0128 step time: 0.3175\n", + "2/8, train_loss: 0.0162 step time: 0.2649\n", + "3/8, train_loss: 0.0149 step time: 0.2530\n", + "4/8, train_loss: 0.0152 step time: 0.2496\n", + "5/8, train_loss: 0.0182 step time: 0.2377\n", + "6/8, train_loss: 0.0150 step time: 0.2313\n", + "7/8, train_loss: 0.0142 step time: 0.2258\n", + "8/8, train_loss: 0.0139 step time: 0.2285\n", + "epoch 489 average loss: 0.0150\n", + "time consuming of epoch 489 is: 2.0099\n", + "----------\n", + "epoch 490/600\n", + "1/8, train_loss: 0.0161 step time: 0.2843\n", + "2/8, train_loss: 0.0134 step time: 0.2335\n", + "3/8, train_loss: 0.0166 step time: 0.2409\n", + "4/8, train_loss: 0.0156 step time: 0.3146\n", + "5/8, train_loss: 0.0153 step time: 0.2950\n", + "6/8, train_loss: 0.0151 step time: 0.2785\n", + "7/8, train_loss: 0.0142 step time: 0.4672\n", + "8/8, train_loss: 0.0126 step time: 0.6426\n", + "epoch 490 average loss: 0.0149\n", + "current epoch: 490 current mean dice: 0.9550 best mean dice: 0.9554 at epoch: 470\n", + "time consuming of epoch 490 is: 4.0932\n", + "----------\n", + "epoch 491/600\n", + "1/8, train_loss: 0.0113 step time: 0.3642\n", + "2/8, train_loss: 0.0137 step time: 0.2706\n", + "3/8, train_loss: 0.0140 step time: 0.2574\n", + "4/8, train_loss: 0.0156 step time: 0.2587\n", + "5/8, train_loss: 0.0149 step time: 0.2486\n", + "6/8, train_loss: 0.0192 step time: 0.2684\n", + "7/8, train_loss: 0.0176 step time: 0.2597\n", + "8/8, train_loss: 0.0153 step time: 0.2509\n", + "epoch 491 average loss: 0.0152\n", + "time consuming of epoch 491 is: 2.1798\n", + "----------\n", + "epoch 492/600\n", + "1/8, train_loss: 0.0141 step time: 0.2929\n", + "2/8, train_loss: 0.0143 step time: 0.2390\n", + "3/8, train_loss: 0.0141 step time: 0.2363\n", + "4/8, train_loss: 0.0138 step time: 0.2392\n", + "5/8, train_loss: 0.0132 step time: 0.3037\n", + "6/8, train_loss: 0.0164 step time: 0.3584\n", + "7/8, train_loss: 0.0184 step time: 0.2681\n", + "8/8, train_loss: 0.0140 step time: 0.2407\n", + "epoch 492 average loss: 0.0148\n", + "time consuming of epoch 492 is: 2.1797\n", + "----------\n", + "epoch 493/600\n", + "1/8, train_loss: 0.0148 step time: 0.2872\n", + "2/8, train_loss: 0.0150 step time: 0.2293\n", + "3/8, train_loss: 0.0154 step time: 0.2318\n", + "4/8, train_loss: 0.0136 step time: 0.2306\n", + "5/8, train_loss: 0.0151 step time: 0.2340\n", + "6/8, train_loss: 0.0165 step time: 0.2295\n", + "7/8, train_loss: 0.0143 step time: 0.2187\n", + "8/8, train_loss: 0.0168 step time: 0.2166\n", + "epoch 493 average loss: 0.0152\n", + "time consuming of epoch 493 is: 1.8794\n", + "----------\n", + "epoch 494/600\n", + "1/8, train_loss: 0.0158 step time: 0.3014\n", + "2/8, train_loss: 0.0155 step time: 0.2646\n", + "3/8, train_loss: 0.0136 step time: 0.2612\n", + "4/8, train_loss: 0.0157 step time: 0.2506\n", + "5/8, train_loss: 0.0143 step time: 0.2482\n", + "6/8, train_loss: 0.0143 step time: 0.2508\n", + "7/8, train_loss: 0.0161 step time: 0.2457\n", + "8/8, train_loss: 0.0141 step time: 0.3862\n", + "epoch 494 average loss: 0.0149\n", + "time consuming of epoch 494 is: 2.2103\n", + "----------\n", + "epoch 495/600\n", + "1/8, train_loss: 0.0126 step time: 0.3945\n", + "2/8, train_loss: 0.0140 step time: 0.2974\n", + "3/8, train_loss: 0.0155 step time: 0.2699\n", + "4/8, train_loss: 0.0136 step time: 0.2544\n", + "5/8, train_loss: 0.0191 step time: 0.2495\n", + "6/8, train_loss: 0.0148 step time: 0.2482\n", + "7/8, train_loss: 0.0175 step time: 0.3006\n", + "8/8, train_loss: 0.0132 step time: 0.3796\n", + "epoch 495 average loss: 0.0150\n", + "saved new best metric model\n", + "current epoch: 495 current mean dice: 0.9558 best mean dice: 0.9558 at epoch: 495\n", + "time consuming of epoch 495 is: 3.9349\n", + "----------\n", + "epoch 496/600\n", + "1/8, train_loss: 0.0128 step time: 0.2782\n", + "2/8, train_loss: 0.0174 step time: 0.2116\n", + "3/8, train_loss: 0.0162 step time: 0.2069\n", + "4/8, train_loss: 0.0131 step time: 0.2154\n", + "5/8, train_loss: 0.0160 step time: 0.2506\n", + "6/8, train_loss: 0.0127 step time: 0.2818\n", + "7/8, train_loss: 0.0156 step time: 0.2524\n", + "8/8, train_loss: 0.0159 step time: 0.2488\n", + "epoch 496 average loss: 0.0150\n", + "time consuming of epoch 496 is: 1.9483\n", + "----------\n", + "epoch 497/600\n", + "1/8, train_loss: 0.0153 step time: 0.2958\n", + "2/8, train_loss: 0.0128 step time: 0.2593\n", + "3/8, train_loss: 0.0159 step time: 0.2472\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "4/8, train_loss: 0.0114 step time: 0.2497\n", + "5/8, train_loss: 0.0156 step time: 0.2485\n", + "6/8, train_loss: 0.0144 step time: 0.2585\n", + "7/8, train_loss: 0.0140 step time: 0.2685\n", + "8/8, train_loss: 0.0156 step time: 0.2734\n", + "epoch 497 average loss: 0.0144\n", + "time consuming of epoch 497 is: 2.1022\n", + "----------\n", + "epoch 498/600\n", + "1/8, train_loss: 0.0150 step time: 0.3006\n", + "2/8, train_loss: 0.0155 step time: 0.2549\n", + "3/8, train_loss: 0.0117 step time: 0.2456\n", + "4/8, train_loss: 0.0151 step time: 0.2401\n", + "5/8, train_loss: 0.0133 step time: 0.2396\n", + "6/8, train_loss: 0.0135 step time: 0.2358\n", + "7/8, train_loss: 0.0201 step time: 0.2249\n", + "8/8, train_loss: 0.0140 step time: 0.2329\n", + "epoch 498 average loss: 0.0148\n", + "time consuming of epoch 498 is: 1.9756\n", + "----------\n", + "epoch 499/600\n", + "1/8, train_loss: 0.0135 step time: 0.3000\n", + "2/8, train_loss: 0.0138 step time: 0.2484\n", + "3/8, train_loss: 0.0130 step time: 0.3354\n", + "4/8, train_loss: 0.0145 step time: 0.4866\n", + "5/8, train_loss: 0.0150 step time: 0.2613\n", + "6/8, train_loss: 0.0178 step time: 0.2568\n", + "7/8, train_loss: 0.0122 step time: 0.2515\n", + "8/8, train_loss: 0.0160 step time: 0.2575\n", + "epoch 499 average loss: 0.0145\n", + "time consuming of epoch 499 is: 2.3988\n", + "----------\n", + "epoch 500/600\n", + "1/8, train_loss: 0.0145 step time: 0.3066\n", + "2/8, train_loss: 0.0153 step time: 0.2517\n", + "3/8, train_loss: 0.0149 step time: 0.2671\n", + "4/8, train_loss: 0.0135 step time: 0.2731\n", + "5/8, train_loss: 0.0133 step time: 0.2580\n", + "6/8, train_loss: 0.0122 step time: 0.2616\n", + "7/8, train_loss: 0.0151 step time: 0.2389\n", + "8/8, train_loss: 0.0154 step time: 0.2866\n", + "epoch 500 average loss: 0.0143\n", + "current epoch: 500 current mean dice: 0.9547 best mean dice: 0.9558 at epoch: 495\n", + "time consuming of epoch 500 is: 3.7080\n", + "----------\n", + "epoch 501/600\n", + "1/8, train_loss: 0.0162 step time: 0.3120\n", + "2/8, train_loss: 0.0138 step time: 0.2527\n", + "3/8, train_loss: 0.0140 step time: 0.2570\n", + "4/8, train_loss: 0.0160 step time: 0.2436\n", + "5/8, train_loss: 0.0125 step time: 0.2445\n", + "6/8, train_loss: 0.0173 step time: 0.3748\n", + "7/8, train_loss: 0.0142 step time: 0.4953\n", + "8/8, train_loss: 0.0148 step time: 0.2738\n", + "epoch 501 average loss: 0.0149\n", + "time consuming of epoch 501 is: 2.4549\n", + "----------\n", + "epoch 502/600\n", + "1/8, train_loss: 0.0157 step time: 0.3025\n", + "2/8, train_loss: 0.0163 step time: 0.2399\n", + "3/8, train_loss: 0.0129 step time: 0.2424\n", + "4/8, train_loss: 0.0128 step time: 0.2418\n", + "5/8, train_loss: 0.0131 step time: 0.2419\n", + "6/8, train_loss: 0.0146 step time: 0.3292\n", + "7/8, train_loss: 0.0198 step time: 0.3586\n", + "8/8, train_loss: 0.0146 step time: 0.2797\n", + "epoch 502 average loss: 0.0150\n", + "time consuming of epoch 502 is: 2.2376\n", + "----------\n", + "epoch 503/600\n", + "1/8, train_loss: 0.0132 step time: 0.3024\n", + "2/8, train_loss: 0.0161 step time: 0.2529\n", + "3/8, train_loss: 0.0146 step time: 0.2614\n", + "4/8, train_loss: 0.0145 step time: 0.2627\n", + "5/8, train_loss: 0.0162 step time: 0.2586\n", + "6/8, train_loss: 0.0156 step time: 0.2477\n", + "7/8, train_loss: 0.0167 step time: 0.2716\n", + "8/8, train_loss: 0.0145 step time: 0.2760\n", + "epoch 503 average loss: 0.0152\n", + "time consuming of epoch 503 is: 2.1349\n", + "----------\n", + "epoch 504/600\n", + "1/8, train_loss: 0.0152 step time: 0.3164\n", + "2/8, train_loss: 0.0172 step time: 0.2593\n", + "3/8, train_loss: 0.0109 step time: 0.3106\n", + "4/8, train_loss: 0.0162 step time: 0.3506\n", + "5/8, train_loss: 0.0154 step time: 0.2706\n", + "6/8, train_loss: 0.0151 step time: 0.2478\n", + "7/8, train_loss: 0.0167 step time: 0.2346\n", + "8/8, train_loss: 0.0136 step time: 0.2328\n", + "epoch 504 average loss: 0.0150\n", + "time consuming of epoch 504 is: 2.2241\n", + "----------\n", + "epoch 505/600\n", + "1/8, train_loss: 0.0166 step time: 0.2880\n", + "2/8, train_loss: 0.0154 step time: 0.2453\n", + "3/8, train_loss: 0.0115 step time: 0.2354\n", + "4/8, train_loss: 0.0132 step time: 0.2396\n", + "5/8, train_loss: 0.0133 step time: 0.3407\n", + "6/8, train_loss: 0.0160 step time: 0.3373\n", + "7/8, train_loss: 0.0152 step time: 0.2788\n", + "8/8, train_loss: 0.0179 step time: 0.2603\n", + "epoch 505 average loss: 0.0149\n", + "current epoch: 505 current mean dice: 0.9556 best mean dice: 0.9558 at epoch: 495\n", + "time consuming of epoch 505 is: 3.6885\n", + "----------\n", + "epoch 506/600\n", + "1/8, train_loss: 0.0163 step time: 0.3048\n", + "2/8, train_loss: 0.0147 step time: 0.2487\n", + "3/8, train_loss: 0.0128 step time: 0.2435\n", + "4/8, train_loss: 0.0156 step time: 0.2428\n", + "5/8, train_loss: 0.0142 step time: 0.2368\n", + "6/8, train_loss: 0.0135 step time: 0.2397\n", + "7/8, train_loss: 0.0139 step time: 0.2236\n", + "8/8, train_loss: 0.0153 step time: 0.2272\n", + "epoch 506 average loss: 0.0145\n", + "time consuming of epoch 506 is: 1.9684\n", + "----------\n", + "epoch 507/600\n", + "1/8, train_loss: 0.0147 step time: 0.5454\n", + "2/8, train_loss: 0.0137 step time: 0.2189\n", + "3/8, train_loss: 0.0162 step time: 0.2151\n", + "4/8, train_loss: 0.0214 step time: 0.2182\n", + "5/8, train_loss: 0.0153 step time: 0.2238\n", + "6/8, train_loss: 0.0172 step time: 0.2299\n", + "7/8, train_loss: 0.0147 step time: 0.2155\n", + "8/8, train_loss: 0.0170 step time: 0.3197\n", + "epoch 507 average loss: 0.0163\n", + "time consuming of epoch 507 is: 2.1880\n", + "----------\n", + "epoch 508/600\n", + "1/8, train_loss: 0.0144 step time: 0.4254\n", + "2/8, train_loss: 0.0150 step time: 0.2917\n", + "3/8, train_loss: 0.0141 step time: 0.2839\n", + "4/8, train_loss: 0.0136 step time: 0.2631\n", + "5/8, train_loss: 0.0169 step time: 0.2656\n", + "6/8, train_loss: 0.0183 step time: 0.2600\n", + "7/8, train_loss: 0.0148 step time: 0.2536\n", + "8/8, train_loss: 0.0149 step time: 0.2494\n", + "epoch 508 average loss: 0.0152\n", + "time consuming of epoch 508 is: 2.2942\n", + "----------\n", + "epoch 509/600\n", + "1/8, train_loss: 0.0139 step time: 0.2943\n", + "2/8, train_loss: 0.0147 step time: 0.2521\n", + "3/8, train_loss: 0.0144 step time: 0.2479\n", + "4/8, train_loss: 0.0144 step time: 0.2400\n", + "5/8, train_loss: 0.0154 step time: 0.2331\n", + "6/8, train_loss: 0.0129 step time: 0.2358\n", + "7/8, train_loss: 0.0142 step time: 0.3023\n", + "8/8, train_loss: 0.0129 step time: 0.2952\n", + "epoch 509 average loss: 0.0141\n", + "time consuming of epoch 509 is: 2.1022\n", + "----------\n", + "epoch 510/600\n", + "1/8, train_loss: 0.0140 step time: 0.3167\n", + "2/8, train_loss: 0.0162 step time: 0.2516\n", + "3/8, train_loss: 0.0140 step time: 0.2411\n", + "4/8, train_loss: 0.0147 step time: 0.2425\n", + "5/8, train_loss: 0.0166 step time: 0.2614\n", + "6/8, train_loss: 0.0152 step time: 0.2643\n", + "7/8, train_loss: 0.0120 step time: 0.4660\n", + "8/8, train_loss: 0.0160 step time: 0.4494\n", + "epoch 510 average loss: 0.0148\n", + "saved new best metric model\n", + "current epoch: 510 current mean dice: 0.9562 best mean dice: 0.9562 at epoch: 510\n", + "time consuming of epoch 510 is: 4.1072\n", + "----------\n", + "epoch 511/600\n", + "1/8, train_loss: 0.0143 step time: 0.2783\n", + "2/8, train_loss: 0.0127 step time: 0.2176\n", + "3/8, train_loss: 0.0147 step time: 0.3002\n", + "4/8, train_loss: 0.0159 step time: 0.4107\n", + "5/8, train_loss: 0.0166 step time: 0.2625\n", + "6/8, train_loss: 0.0171 step time: 0.2496\n", + "7/8, train_loss: 0.0126 step time: 0.2334\n", + "8/8, train_loss: 0.0153 step time: 0.2311\n", + "epoch 511 average loss: 0.0149\n", + "time consuming of epoch 511 is: 2.1861\n", + "----------\n", + "epoch 512/600\n", + "1/8, train_loss: 0.0134 step time: 0.2813\n", + "2/8, train_loss: 0.0180 step time: 0.2325\n", + "3/8, train_loss: 0.0167 step time: 0.2362\n", + "4/8, train_loss: 0.0139 step time: 0.2354\n", + "5/8, train_loss: 0.0146 step time: 0.5940\n", + "6/8, train_loss: 0.0121 step time: 0.2183\n", + "7/8, train_loss: 0.0138 step time: 0.2914\n", + "8/8, train_loss: 0.0096 step time: 0.3570\n", + "epoch 512 average loss: 0.0140\n", + "time consuming of epoch 512 is: 2.4477\n", + "----------\n", + "epoch 513/600\n", + "1/8, train_loss: 0.0143 step time: 0.3087\n", + "2/8, train_loss: 0.0123 step time: 0.2559\n", + "3/8, train_loss: 0.0117 step time: 0.2573\n", + "4/8, train_loss: 0.0167 step time: 0.2510\n", + "5/8, train_loss: 0.0161 step time: 0.2473\n", + "6/8, train_loss: 0.0151 step time: 0.2621\n", + "7/8, train_loss: 0.0146 step time: 0.2758\n", + "8/8, train_loss: 0.0154 step time: 0.2657\n", + "epoch 513 average loss: 0.0145\n", + "time consuming of epoch 513 is: 2.1253\n", + "----------\n", + "epoch 514/600\n", + "1/8, train_loss: 0.0143 step time: 0.3010\n", + "2/8, train_loss: 0.0131 step time: 0.2495\n", + "3/8, train_loss: 0.0143 step time: 0.2481\n", + "4/8, train_loss: 0.0136 step time: 0.2445\n", + "5/8, train_loss: 0.0141 step time: 0.2456\n", + "6/8, train_loss: 0.0165 step time: 0.2794\n", + "7/8, train_loss: 0.0157 step time: 0.3413\n", + "8/8, train_loss: 0.0139 step time: 0.2712\n", + "epoch 514 average loss: 0.0144\n", + "time consuming of epoch 514 is: 2.1821\n", + "----------\n", + "epoch 515/600\n", + "1/8, train_loss: 0.0141 step time: 0.3116\n", + "2/8, train_loss: 0.0137 step time: 0.2439\n", + "3/8, train_loss: 0.0168 step time: 0.3822\n", + "4/8, train_loss: 0.0177 step time: 0.5838\n", + "5/8, train_loss: 0.0122 step time: 0.2639\n", + "6/8, train_loss: 0.0138 step time: 0.2514\n", + "7/8, train_loss: 0.0132 step time: 0.2406\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "8/8, train_loss: 0.0165 step time: 0.2313\n", + "epoch 515 average loss: 0.0148\n", + "saved new best metric model\n", + "current epoch: 515 current mean dice: 0.9565 best mean dice: 0.9565 at epoch: 515\n", + "time consuming of epoch 515 is: 4.0181\n", + "----------\n", + "epoch 516/600\n", + "1/8, train_loss: 0.0108 step time: 0.2806\n", + "2/8, train_loss: 0.0134 step time: 0.2180\n", + "3/8, train_loss: 0.0137 step time: 0.2208\n", + "4/8, train_loss: 0.0148 step time: 0.2289\n", + "5/8, train_loss: 0.0139 step time: 0.3556\n", + "6/8, train_loss: 0.0127 step time: 0.3864\n", + "7/8, train_loss: 0.0161 step time: 0.2664\n", + "8/8, train_loss: 0.0208 step time: 0.2403\n", + "epoch 516 average loss: 0.0145\n", + "time consuming of epoch 516 is: 2.1994\n", + "----------\n", + "epoch 517/600\n", + "1/8, train_loss: 0.0149 step time: 0.2833\n", + "2/8, train_loss: 0.0119 step time: 0.2339\n", + "3/8, train_loss: 0.0152 step time: 0.2331\n", + "4/8, train_loss: 0.0142 step time: 0.2363\n", + "5/8, train_loss: 0.0149 step time: 0.2309\n", + "6/8, train_loss: 0.0143 step time: 0.5837\n", + "7/8, train_loss: 0.0154 step time: 0.2048\n", + "8/8, train_loss: 0.0143 step time: 0.2162\n", + "epoch 517 average loss: 0.0144\n", + "time consuming of epoch 517 is: 2.2236\n", + "----------\n", + "epoch 518/600\n", + "1/8, train_loss: 0.0146 step time: 0.2916\n", + "2/8, train_loss: 0.0159 step time: 0.2403\n", + "3/8, train_loss: 0.0137 step time: 0.2399\n", + "4/8, train_loss: 0.0136 step time: 0.2409\n", + "5/8, train_loss: 0.0155 step time: 0.2421\n", + "6/8, train_loss: 0.0138 step time: 0.2406\n", + "7/8, train_loss: 0.0149 step time: 0.3824\n", + "8/8, train_loss: 0.0137 step time: 0.4227\n", + "epoch 518 average loss: 0.0145\n", + "time consuming of epoch 518 is: 2.3021\n", + "----------\n", + "epoch 519/600\n", + "1/8, train_loss: 0.0111 step time: 0.2984\n", + "2/8, train_loss: 0.0159 step time: 0.2545\n", + "3/8, train_loss: 0.0151 step time: 0.2518\n", + "4/8, train_loss: 0.0141 step time: 0.2689\n", + "5/8, train_loss: 0.0140 step time: 0.2885\n", + "6/8, train_loss: 0.0155 step time: 0.2850\n", + "7/8, train_loss: 0.0148 step time: 0.2648\n", + "8/8, train_loss: 0.0153 step time: 0.2524\n", + "epoch 519 average loss: 0.0145\n", + "time consuming of epoch 519 is: 2.1658\n", + "----------\n", + "epoch 520/600\n", + "1/8, train_loss: 0.0125 step time: 0.2932\n", + "2/8, train_loss: 0.0151 step time: 0.2636\n", + "3/8, train_loss: 0.0172 step time: 0.2738\n", + "4/8, train_loss: 0.0140 step time: 0.2517\n", + "5/8, train_loss: 0.0151 step time: 0.2490\n", + "6/8, train_loss: 0.0132 step time: 0.2492\n", + "7/8, train_loss: 0.0149 step time: 0.2344\n", + "8/8, train_loss: 0.0120 step time: 0.2347\n", + "epoch 520 average loss: 0.0142\n", + "current epoch: 520 current mean dice: 0.9562 best mean dice: 0.9565 at epoch: 515\n", + "time consuming of epoch 520 is: 3.6764\n", + "----------\n", + "epoch 521/600\n", + "1/8, train_loss: 0.0156 step time: 0.2901\n", + "2/8, train_loss: 0.0123 step time: 0.2437\n", + "3/8, train_loss: 0.0133 step time: 0.2409\n", + "4/8, train_loss: 0.0120 step time: 0.2425\n", + "5/8, train_loss: 0.0148 step time: 0.2423\n", + "6/8, train_loss: 0.0149 step time: 0.2399\n", + "7/8, train_loss: 0.0151 step time: 0.2385\n", + "8/8, train_loss: 0.0267 step time: 0.2370\n", + "epoch 521 average loss: 0.0156\n", + "time consuming of epoch 521 is: 1.9763\n", + "----------\n", + "epoch 522/600\n", + "1/8, train_loss: 0.0130 step time: 0.2929\n", + "2/8, train_loss: 0.0153 step time: 0.2497\n", + "3/8, train_loss: 0.0166 step time: 0.2513\n", + "4/8, train_loss: 0.0161 step time: 0.2485\n", + "5/8, train_loss: 0.0120 step time: 0.3345\n", + "6/8, train_loss: 0.0143 step time: 0.5173\n", + "7/8, train_loss: 0.0162 step time: 0.2787\n", + "8/8, train_loss: 0.0190 step time: 0.2540\n", + "epoch 522 average loss: 0.0153\n", + "time consuming of epoch 522 is: 2.4284\n", + "----------\n", + "epoch 523/600\n", + "1/8, train_loss: 0.0159 step time: 0.2937\n", + "2/8, train_loss: 0.0130 step time: 0.2518\n", + "3/8, train_loss: 0.0133 step time: 0.2571\n", + "4/8, train_loss: 0.0123 step time: 0.2578\n", + "5/8, train_loss: 0.0173 step time: 0.2509\n", + "6/8, train_loss: 0.0138 step time: 0.2434\n", + "7/8, train_loss: 0.0157 step time: 0.2387\n", + "8/8, train_loss: 0.0133 step time: 0.2357\n", + "epoch 523 average loss: 0.0143\n", + "time consuming of epoch 523 is: 2.0307\n", + "----------\n", + "epoch 524/600\n", + "1/8, train_loss: 0.0162 step time: 0.2824\n", + "2/8, train_loss: 0.0153 step time: 0.2725\n", + "3/8, train_loss: 0.0135 step time: 0.2768\n", + "4/8, train_loss: 0.0141 step time: 0.2673\n", + "5/8, train_loss: 0.0158 step time: 0.2587\n", + "6/8, train_loss: 0.0186 step time: 0.2584\n", + "7/8, train_loss: 0.0150 step time: 0.2603\n", + "8/8, train_loss: 0.0157 step time: 0.2508\n", + "epoch 524 average loss: 0.0155\n", + "time consuming of epoch 524 is: 2.1288\n", + "----------\n", + "epoch 525/600\n", + "1/8, train_loss: 0.0152 step time: 0.3036\n", + "2/8, train_loss: 0.0139 step time: 0.2506\n", + "3/8, train_loss: 0.0150 step time: 0.3531\n", + "4/8, train_loss: 0.0150 step time: 0.3199\n", + "5/8, train_loss: 0.0117 step time: 0.3079\n", + "6/8, train_loss: 0.0149 step time: 0.2763\n", + "7/8, train_loss: 0.0127 step time: 0.2791\n", + "8/8, train_loss: 0.0143 step time: 0.2566\n", + "epoch 525 average loss: 0.0141\n", + "saved new best metric model\n", + "current epoch: 525 current mean dice: 0.9567 best mean dice: 0.9567 at epoch: 525\n", + "time consuming of epoch 525 is: 3.9988\n", + "----------\n", + "epoch 526/600\n", + "1/8, train_loss: 0.0133 step time: 0.2811\n", + "2/8, train_loss: 0.0160 step time: 0.2136\n", + "3/8, train_loss: 0.0159 step time: 0.2140\n", + "4/8, train_loss: 0.0150 step time: 0.2393\n", + "5/8, train_loss: 0.0118 step time: 0.2959\n", + "6/8, train_loss: 0.0137 step time: 0.2766\n", + "7/8, train_loss: 0.0136 step time: 0.2584\n", + "8/8, train_loss: 0.0156 step time: 0.2799\n", + "epoch 526 average loss: 0.0144\n", + "time consuming of epoch 526 is: 2.0613\n", + "----------\n", + "epoch 527/600\n", + "1/8, train_loss: 0.0127 step time: 0.3468\n", + "2/8, train_loss: 0.0163 step time: 0.2859\n", + "3/8, train_loss: 0.0124 step time: 0.2578\n", + "4/8, train_loss: 0.0145 step time: 0.2914\n", + "5/8, train_loss: 0.0156 step time: 0.2973\n", + "6/8, train_loss: 0.0109 step time: 0.2932\n", + "7/8, train_loss: 0.0143 step time: 0.2718\n", + "8/8, train_loss: 0.0172 step time: 0.2638\n", + "epoch 527 average loss: 0.0142\n", + "time consuming of epoch 527 is: 2.3094\n", + "----------\n", + "epoch 528/600\n", + "1/8, train_loss: 0.0100 step time: 0.3013\n", + "2/8, train_loss: 0.0156 step time: 0.3216\n", + "3/8, train_loss: 0.0144 step time: 0.4727\n", + "4/8, train_loss: 0.0135 step time: 0.2812\n", + "5/8, train_loss: 0.0187 step time: 0.2765\n", + "6/8, train_loss: 0.0151 step time: 0.2646\n", + "7/8, train_loss: 0.0148 step time: 0.2627\n", + "8/8, train_loss: 0.0128 step time: 0.2539\n", + "epoch 528 average loss: 0.0144\n", + "time consuming of epoch 528 is: 2.4362\n", + "----------\n", + "epoch 529/600\n", + "1/8, train_loss: 0.0166 step time: 0.2836\n", + "2/8, train_loss: 0.0134 step time: 0.3092\n", + "3/8, train_loss: 0.0122 step time: 0.3309\n", + "4/8, train_loss: 0.0147 step time: 0.3043\n", + "5/8, train_loss: 0.0148 step time: 0.2759\n", + "6/8, train_loss: 0.0144 step time: 0.2740\n", + "7/8, train_loss: 0.0147 step time: 0.2561\n", + "8/8, train_loss: 0.0126 step time: 0.2645\n", + "epoch 529 average loss: 0.0142\n", + "time consuming of epoch 529 is: 2.3000\n", + "----------\n", + "epoch 530/600\n", + "1/8, train_loss: 0.0147 step time: 0.2985\n", + "2/8, train_loss: 0.0144 step time: 0.2450\n", + "3/8, train_loss: 0.0142 step time: 0.2609\n", + "4/8, train_loss: 0.0128 step time: 0.2716\n", + "5/8, train_loss: 0.0149 step time: 0.2654\n", + "6/8, train_loss: 0.0130 step time: 0.2532\n", + "7/8, train_loss: 0.0161 step time: 0.2523\n", + "8/8, train_loss: 0.0157 step time: 0.2405\n", + "epoch 530 average loss: 0.0145\n", + "current epoch: 530 current mean dice: 0.9563 best mean dice: 0.9567 at epoch: 525\n", + "time consuming of epoch 530 is: 3.6364\n", + "----------\n", + "epoch 531/600\n", + "1/8, train_loss: 0.0153 step time: 0.4621\n", + "2/8, train_loss: 0.0133 step time: 0.3092\n", + "3/8, train_loss: 0.0168 step time: 0.2747\n", + "4/8, train_loss: 0.0157 step time: 0.2507\n", + "5/8, train_loss: 0.0122 step time: 0.2570\n", + "6/8, train_loss: 0.0143 step time: 0.2439\n", + "7/8, train_loss: 0.0171 step time: 0.2315\n", + "8/8, train_loss: 0.0143 step time: 0.2616\n", + "epoch 531 average loss: 0.0149\n", + "time consuming of epoch 531 is: 2.2918\n", + "----------\n", + "epoch 532/600\n", + "1/8, train_loss: 0.0097 step time: 0.3294\n", + "2/8, train_loss: 0.0155 step time: 0.2748\n", + "3/8, train_loss: 0.0137 step time: 0.2602\n", + "4/8, train_loss: 0.0135 step time: 0.2605\n", + "5/8, train_loss: 0.0149 step time: 0.2521\n", + "6/8, train_loss: 0.0145 step time: 0.2545\n", + "7/8, train_loss: 0.0159 step time: 0.2502\n", + "8/8, train_loss: 0.0139 step time: 0.2766\n", + "epoch 532 average loss: 0.0139\n", + "time consuming of epoch 532 is: 2.1597\n", + "----------\n", + "epoch 533/600\n", + "1/8, train_loss: 0.0132 step time: 0.3204\n", + "2/8, train_loss: 0.0169 step time: 0.2607\n", + "3/8, train_loss: 0.0168 step time: 0.2449\n", + "4/8, train_loss: 0.0188 step time: 0.2478\n", + "5/8, train_loss: 0.0121 step time: 0.2413\n", + "6/8, train_loss: 0.0124 step time: 0.2325\n", + "7/8, train_loss: 0.0142 step time: 0.2253\n", + "8/8, train_loss: 0.0104 step time: 0.2318\n", + "epoch 533 average loss: 0.0144\n", + "time consuming of epoch 533 is: 2.0062\n", + "----------\n", + "epoch 534/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0158 step time: 0.2971\n", + "2/8, train_loss: 0.0121 step time: 0.2513\n", + "3/8, train_loss: 0.0148 step time: 0.2558\n", + "4/8, train_loss: 0.0152 step time: 0.2820\n", + "5/8, train_loss: 0.0173 step time: 0.2766\n", + "6/8, train_loss: 0.0168 step time: 0.2653\n", + "7/8, train_loss: 0.0126 step time: 0.4649\n", + "8/8, train_loss: 0.0146 step time: 0.4537\n", + "epoch 534 average loss: 0.0149\n", + "time consuming of epoch 534 is: 2.5483\n", + "----------\n", + "epoch 535/600\n", + "1/8, train_loss: 0.0147 step time: 0.2978\n", + "2/8, train_loss: 0.0157 step time: 0.2453\n", + "3/8, train_loss: 0.0137 step time: 0.2440\n", + "4/8, train_loss: 0.0155 step time: 0.2463\n", + "5/8, train_loss: 0.0124 step time: 0.2448\n", + "6/8, train_loss: 0.0135 step time: 0.2398\n", + "7/8, train_loss: 0.0128 step time: 0.2258\n", + "8/8, train_loss: 0.0161 step time: 0.2454\n", + "epoch 535 average loss: 0.0143\n", + "current epoch: 535 current mean dice: 0.9561 best mean dice: 0.9567 at epoch: 525\n", + "time consuming of epoch 535 is: 3.4063\n", + "----------\n", + "epoch 536/600\n", + "1/8, train_loss: 0.0149 step time: 0.3464\n", + "2/8, train_loss: 0.0120 step time: 0.2958\n", + "3/8, train_loss: 0.0137 step time: 0.2792\n", + "4/8, train_loss: 0.0161 step time: 0.2764\n", + "5/8, train_loss: 0.0138 step time: 0.2547\n", + "6/8, train_loss: 0.0139 step time: 0.2499\n", + "7/8, train_loss: 0.0117 step time: 0.2587\n", + "8/8, train_loss: 0.0128 step time: 0.2594\n", + "epoch 536 average loss: 0.0136\n", + "time consuming of epoch 536 is: 2.2218\n", + "----------\n", + "epoch 537/600\n", + "1/8, train_loss: 0.0133 step time: 0.2981\n", + "2/8, train_loss: 0.0163 step time: 0.2580\n", + "3/8, train_loss: 0.0111 step time: 0.2510\n", + "4/8, train_loss: 0.0145 step time: 0.2650\n", + "5/8, train_loss: 0.0140 step time: 0.2841\n", + "6/8, train_loss: 0.0159 step time: 0.2684\n", + "7/8, train_loss: 0.0131 step time: 0.5561\n", + "8/8, train_loss: 0.0128 step time: 0.4515\n", + "epoch 537 average loss: 0.0139\n", + "time consuming of epoch 537 is: 2.6338\n", + "----------\n", + "epoch 538/600\n", + "1/8, train_loss: 0.0143 step time: 0.3007\n", + "2/8, train_loss: 0.0180 step time: 0.2419\n", + "3/8, train_loss: 0.0130 step time: 0.2413\n", + "4/8, train_loss: 0.0115 step time: 0.2409\n", + "5/8, train_loss: 0.0165 step time: 0.2401\n", + "6/8, train_loss: 0.0132 step time: 0.2738\n", + "7/8, train_loss: 0.0173 step time: 0.2786\n", + "8/8, train_loss: 0.0109 step time: 0.2742\n", + "epoch 538 average loss: 0.0144\n", + "time consuming of epoch 538 is: 2.0930\n", + "----------\n", + "epoch 539/600\n", + "1/8, train_loss: 0.0160 step time: 0.3116\n", + "2/8, train_loss: 0.0138 step time: 0.2459\n", + "3/8, train_loss: 0.0181 step time: 0.2418\n", + "4/8, train_loss: 0.0139 step time: 0.2407\n", + "5/8, train_loss: 0.0139 step time: 0.2310\n", + "6/8, train_loss: 0.0117 step time: 0.2599\n", + "7/8, train_loss: 0.0144 step time: 0.3260\n", + "8/8, train_loss: 0.0163 step time: 0.2811\n", + "epoch 539 average loss: 0.0148\n", + "time consuming of epoch 539 is: 2.1394\n", + "----------\n", + "epoch 540/600\n", + "1/8, train_loss: 0.0136 step time: 0.3150\n", + "2/8, train_loss: 0.0142 step time: 0.2706\n", + "3/8, train_loss: 0.0168 step time: 0.2724\n", + "4/8, train_loss: 0.0151 step time: 0.2811\n", + "5/8, train_loss: 0.0137 step time: 0.2603\n", + "6/8, train_loss: 0.0168 step time: 0.2532\n", + "7/8, train_loss: 0.0154 step time: 0.2510\n", + "8/8, train_loss: 0.0128 step time: 0.2392\n", + "epoch 540 average loss: 0.0148\n", + "current epoch: 540 current mean dice: 0.9520 best mean dice: 0.9567 at epoch: 525\n", + "time consuming of epoch 540 is: 3.5460\n", + "----------\n", + "epoch 541/600\n", + "1/8, train_loss: 0.0160 step time: 0.3325\n", + "2/8, train_loss: 0.0154 step time: 0.2838\n", + "3/8, train_loss: 0.0149 step time: 0.2659\n", + "4/8, train_loss: 0.0152 step time: 0.2480\n", + "5/8, train_loss: 0.0168 step time: 0.2434\n", + "6/8, train_loss: 0.0129 step time: 0.2442\n", + "7/8, train_loss: 0.0129 step time: 0.2238\n", + "8/8, train_loss: 0.0142 step time: 0.2262\n", + "epoch 541 average loss: 0.0148\n", + "time consuming of epoch 541 is: 2.0691\n", + "----------\n", + "epoch 542/600\n", + "1/8, train_loss: 0.0162 step time: 0.4539\n", + "2/8, train_loss: 0.0144 step time: 0.5945\n", + "3/8, train_loss: 0.0120 step time: 0.2692\n", + "4/8, train_loss: 0.0174 step time: 0.2546\n", + "5/8, train_loss: 0.0120 step time: 0.2419\n", + "6/8, train_loss: 0.0150 step time: 0.2485\n", + "7/8, train_loss: 0.0168 step time: 0.5282\n", + "8/8, train_loss: 0.0154 step time: 0.4443\n", + "epoch 542 average loss: 0.0149\n", + "time consuming of epoch 542 is: 3.0366\n", + "----------\n", + "epoch 543/600\n", + "1/8, train_loss: 0.0147 step time: 0.3248\n", + "2/8, train_loss: 0.0155 step time: 0.2644\n", + "3/8, train_loss: 0.0154 step time: 0.2435\n", + "4/8, train_loss: 0.0148 step time: 0.2387\n", + "5/8, train_loss: 0.0112 step time: 0.2401\n", + "6/8, train_loss: 0.0125 step time: 0.2385\n", + "7/8, train_loss: 0.0144 step time: 0.2328\n", + "8/8, train_loss: 0.0149 step time: 0.2566\n", + "epoch 543 average loss: 0.0142\n", + "time consuming of epoch 543 is: 2.0409\n", + "----------\n", + "epoch 544/600\n", + "1/8, train_loss: 0.0142 step time: 0.3262\n", + "2/8, train_loss: 0.0176 step time: 0.2701\n", + "3/8, train_loss: 0.0137 step time: 0.2718\n", + "4/8, train_loss: 0.0149 step time: 0.2686\n", + "5/8, train_loss: 0.0150 step time: 0.2699\n", + "6/8, train_loss: 0.0131 step time: 0.2661\n", + "7/8, train_loss: 0.0108 step time: 0.2779\n", + "8/8, train_loss: 0.0128 step time: 0.2981\n", + "epoch 544 average loss: 0.0140\n", + "time consuming of epoch 544 is: 2.2502\n", + "----------\n", + "epoch 545/600\n", + "1/8, train_loss: 0.0117 step time: 0.3072\n", + "2/8, train_loss: 0.0153 step time: 0.2550\n", + "3/8, train_loss: 0.0175 step time: 0.2835\n", + "4/8, train_loss: 0.0144 step time: 0.2772\n", + "5/8, train_loss: 0.0126 step time: 0.2636\n", + "6/8, train_loss: 0.0117 step time: 0.2735\n", + "7/8, train_loss: 0.0133 step time: 0.2633\n", + "8/8, train_loss: 0.0122 step time: 0.2554\n", + "epoch 545 average loss: 0.0136\n", + "saved new best metric model\n", + "current epoch: 545 current mean dice: 0.9570 best mean dice: 0.9570 at epoch: 545\n", + "time consuming of epoch 545 is: 3.9191\n", + "----------\n", + "epoch 546/600\n", + "1/8, train_loss: 0.0119 step time: 0.2768\n", + "2/8, train_loss: 0.0178 step time: 0.2553\n", + "3/8, train_loss: 0.0161 step time: 0.3216\n", + "4/8, train_loss: 0.0151 step time: 0.2712\n", + "5/8, train_loss: 0.0112 step time: 0.2617\n", + "6/8, train_loss: 0.0170 step time: 0.2595\n", + "7/8, train_loss: 0.0137 step time: 0.2862\n", + "8/8, train_loss: 0.0126 step time: 0.3048\n", + "epoch 546 average loss: 0.0144\n", + "time consuming of epoch 546 is: 2.2397\n", + "----------\n", + "epoch 547/600\n", + "1/8, train_loss: 0.0135 step time: 0.3232\n", + "2/8, train_loss: 0.0176 step time: 0.2631\n", + "3/8, train_loss: 0.0130 step time: 0.2953\n", + "4/8, train_loss: 0.0164 step time: 0.3203\n", + "5/8, train_loss: 0.0130 step time: 0.3093\n", + "6/8, train_loss: 0.0148 step time: 0.3132\n", + "7/8, train_loss: 0.0120 step time: 0.2893\n", + "8/8, train_loss: 0.0111 step time: 0.2786\n", + "epoch 547 average loss: 0.0139\n", + "time consuming of epoch 547 is: 2.3938\n", + "----------\n", + "epoch 548/600\n", + "1/8, train_loss: 0.0156 step time: 0.3058\n", + "2/8, train_loss: 0.0136 step time: 0.2517\n", + "3/8, train_loss: 0.0149 step time: 0.2477\n", + "4/8, train_loss: 0.0135 step time: 0.2518\n", + "5/8, train_loss: 0.0112 step time: 0.4019\n", + "6/8, train_loss: 0.0144 step time: 0.4811\n", + "7/8, train_loss: 0.0131 step time: 0.2657\n", + "8/8, train_loss: 0.0139 step time: 0.2435\n", + "epoch 548 average loss: 0.0138\n", + "time consuming of epoch 548 is: 2.4506\n", + "----------\n", + "epoch 549/600\n", + "1/8, train_loss: 0.0132 step time: 0.2858\n", + "2/8, train_loss: 0.0144 step time: 0.2677\n", + "3/8, train_loss: 0.0137 step time: 0.2752\n", + "4/8, train_loss: 0.0094 step time: 0.2639\n", + "5/8, train_loss: 0.0139 step time: 0.2572\n", + "6/8, train_loss: 0.0166 step time: 0.2466\n", + "7/8, train_loss: 0.0140 step time: 0.2240\n", + "8/8, train_loss: 0.0124 step time: 0.2238\n", + "epoch 549 average loss: 0.0135\n", + "time consuming of epoch 549 is: 2.0456\n", + "----------\n", + "epoch 550/600\n", + "1/8, train_loss: 0.0144 step time: 0.2811\n", + "2/8, train_loss: 0.0146 step time: 0.2308\n", + "3/8, train_loss: 0.0164 step time: 0.2300\n", + "4/8, train_loss: 0.0159 step time: 0.2301\n", + "5/8, train_loss: 0.0127 step time: 0.2309\n", + "6/8, train_loss: 0.0140 step time: 0.3292\n", + "7/8, train_loss: 0.0127 step time: 0.3068\n", + "8/8, train_loss: 0.0141 step time: 0.2757\n", + "epoch 550 average loss: 0.0143\n", + "current epoch: 550 current mean dice: 0.9567 best mean dice: 0.9570 at epoch: 545\n", + "time consuming of epoch 550 is: 3.7092\n", + "----------\n", + "epoch 551/600\n", + "1/8, train_loss: 0.0152 step time: 0.3052\n", + "2/8, train_loss: 0.0152 step time: 0.2695\n", + "3/8, train_loss: 0.0142 step time: 0.2651\n", + "4/8, train_loss: 0.0142 step time: 0.2739\n", + "5/8, train_loss: 0.0092 step time: 0.2701\n", + "6/8, train_loss: 0.0133 step time: 0.2813\n", + "7/8, train_loss: 0.0148 step time: 0.2793\n", + "8/8, train_loss: 0.0147 step time: 0.2570\n", + "epoch 551 average loss: 0.0138\n", + "time consuming of epoch 551 is: 2.2028\n", + "----------\n", + "epoch 552/600\n", + "1/8, train_loss: 0.0122 step time: 0.3062\n", + "2/8, train_loss: 0.0151 step time: 0.2590\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "3/8, train_loss: 0.0152 step time: 0.2525\n", + "4/8, train_loss: 0.0137 step time: 0.2521\n", + "5/8, train_loss: 0.0170 step time: 0.2521\n", + "6/8, train_loss: 0.0131 step time: 0.2461\n", + "7/8, train_loss: 0.0143 step time: 0.2536\n", + "8/8, train_loss: 0.0126 step time: 0.2838\n", + "epoch 552 average loss: 0.0141\n", + "time consuming of epoch 552 is: 2.1070\n", + "----------\n", + "epoch 553/600\n", + "1/8, train_loss: 0.0155 step time: 0.3193\n", + "2/8, train_loss: 0.0113 step time: 0.2586\n", + "3/8, train_loss: 0.0137 step time: 0.2527\n", + "4/8, train_loss: 0.0173 step time: 0.2613\n", + "5/8, train_loss: 0.0149 step time: 0.3232\n", + "6/8, train_loss: 0.0147 step time: 0.3309\n", + "7/8, train_loss: 0.0111 step time: 0.2825\n", + "8/8, train_loss: 0.0148 step time: 0.2544\n", + "epoch 553 average loss: 0.0142\n", + "time consuming of epoch 553 is: 2.2843\n", + "----------\n", + "epoch 554/600\n", + "1/8, train_loss: 0.0168 step time: 0.2938\n", + "2/8, train_loss: 0.0137 step time: 0.3006\n", + "3/8, train_loss: 0.0152 step time: 0.3725\n", + "4/8, train_loss: 0.0110 step time: 0.2663\n", + "5/8, train_loss: 0.0150 step time: 0.2566\n", + "6/8, train_loss: 0.0131 step time: 0.2496\n", + "7/8, train_loss: 0.0132 step time: 0.3977\n", + "8/8, train_loss: 0.0126 step time: 0.5971\n", + "epoch 554 average loss: 0.0138\n", + "time consuming of epoch 554 is: 2.7356\n", + "----------\n", + "epoch 555/600\n", + "1/8, train_loss: 0.0121 step time: 0.2695\n", + "2/8, train_loss: 0.0147 step time: 0.2194\n", + "3/8, train_loss: 0.0141 step time: 0.2225\n", + "4/8, train_loss: 0.0153 step time: 0.2231\n", + "5/8, train_loss: 0.0139 step time: 0.2223\n", + "6/8, train_loss: 0.0122 step time: 0.3301\n", + "7/8, train_loss: 0.0157 step time: 0.4462\n", + "8/8, train_loss: 0.0144 step time: 0.2575\n", + "epoch 555 average loss: 0.0141\n", + "current epoch: 555 current mean dice: 0.9569 best mean dice: 0.9570 at epoch: 545\n", + "time consuming of epoch 555 is: 3.6251\n", + "----------\n", + "epoch 556/600\n", + "1/8, train_loss: 0.0140 step time: 0.3296\n", + "2/8, train_loss: 0.0150 step time: 0.2743\n", + "3/8, train_loss: 0.0118 step time: 0.2643\n", + "4/8, train_loss: 0.0165 step time: 0.2546\n", + "5/8, train_loss: 0.0166 step time: 0.2500\n", + "6/8, train_loss: 0.0132 step time: 0.4506\n", + "7/8, train_loss: 0.0122 step time: 0.4417\n", + "8/8, train_loss: 0.0139 step time: 0.2567\n", + "epoch 556 average loss: 0.0141\n", + "time consuming of epoch 556 is: 2.5230\n", + "----------\n", + "epoch 557/600\n", + "1/8, train_loss: 0.0129 step time: 0.2935\n", + "2/8, train_loss: 0.0127 step time: 0.2377\n", + "3/8, train_loss: 0.0134 step time: 0.2375\n", + "4/8, train_loss: 0.0135 step time: 0.2400\n", + "5/8, train_loss: 0.0176 step time: 0.2340\n", + "6/8, train_loss: 0.0171 step time: 0.2342\n", + "7/8, train_loss: 0.0112 step time: 0.4926\n", + "8/8, train_loss: 0.0159 step time: 0.3657\n", + "epoch 557 average loss: 0.0143\n", + "time consuming of epoch 557 is: 2.3367\n", + "----------\n", + "epoch 558/600\n", + "1/8, train_loss: 0.0134 step time: 0.3071\n", + "2/8, train_loss: 0.0115 step time: 0.2493\n", + "3/8, train_loss: 0.0142 step time: 0.2489\n", + "4/8, train_loss: 0.0151 step time: 0.2708\n", + "5/8, train_loss: 0.0174 step time: 0.2718\n", + "6/8, train_loss: 0.0146 step time: 0.2617\n", + "7/8, train_loss: 0.0144 step time: 0.2643\n", + "8/8, train_loss: 0.0155 step time: 0.2598\n", + "epoch 558 average loss: 0.0145\n", + "time consuming of epoch 558 is: 2.1351\n", + "----------\n", + "epoch 559/600\n", + "1/8, train_loss: 0.0117 step time: 0.3035\n", + "2/8, train_loss: 0.0169 step time: 0.2492\n", + "3/8, train_loss: 0.0131 step time: 0.2437\n", + "4/8, train_loss: 0.0159 step time: 0.2446\n", + "5/8, train_loss: 0.0140 step time: 0.3306\n", + "6/8, train_loss: 0.0154 step time: 0.4943\n", + "7/8, train_loss: 0.0132 step time: 0.2757\n", + "8/8, train_loss: 0.0124 step time: 0.2856\n", + "epoch 559 average loss: 0.0141\n", + "time consuming of epoch 559 is: 2.4287\n", + "----------\n", + "epoch 560/600\n", + "1/8, train_loss: 0.0107 step time: 0.3177\n", + "2/8, train_loss: 0.0169 step time: 0.2547\n", + "3/8, train_loss: 0.0145 step time: 0.2553\n", + "4/8, train_loss: 0.0126 step time: 0.2534\n", + "5/8, train_loss: 0.0151 step time: 0.2493\n", + "6/8, train_loss: 0.0133 step time: 0.2442\n", + "7/8, train_loss: 0.0151 step time: 0.2384\n", + "8/8, train_loss: 0.0150 step time: 0.2188\n", + "epoch 560 average loss: 0.0142\n", + "current epoch: 560 current mean dice: 0.9564 best mean dice: 0.9570 at epoch: 545\n", + "time consuming of epoch 560 is: 3.5012\n", + "----------\n", + "epoch 561/600\n", + "1/8, train_loss: 0.0149 step time: 0.3237\n", + "2/8, train_loss: 0.0152 step time: 0.2688\n", + "3/8, train_loss: 0.0124 step time: 0.2498\n", + "4/8, train_loss: 0.0126 step time: 0.2531\n", + "5/8, train_loss: 0.0139 step time: 0.2501\n", + "6/8, train_loss: 0.0130 step time: 0.3305\n", + "7/8, train_loss: 0.0140 step time: 0.3248\n", + "8/8, train_loss: 0.0138 step time: 0.2744\n", + "epoch 561 average loss: 0.0137\n", + "time consuming of epoch 561 is: 2.2764\n", + "----------\n", + "epoch 562/600\n", + "1/8, train_loss: 0.0125 step time: 0.3088\n", + "2/8, train_loss: 0.0105 step time: 0.2554\n", + "3/8, train_loss: 0.0156 step time: 0.2535\n", + "4/8, train_loss: 0.0155 step time: 0.2513\n", + "5/8, train_loss: 0.0151 step time: 0.2617\n", + "6/8, train_loss: 0.0162 step time: 0.2479\n", + "7/8, train_loss: 0.0127 step time: 0.2259\n", + "8/8, train_loss: 0.0131 step time: 0.2304\n", + "epoch 562 average loss: 0.0139\n", + "time consuming of epoch 562 is: 2.0363\n", + "----------\n", + "epoch 563/600\n", + "1/8, train_loss: 0.0133 step time: 0.2802\n", + "2/8, train_loss: 0.0133 step time: 0.2375\n", + "3/8, train_loss: 0.0159 step time: 0.2416\n", + "4/8, train_loss: 0.0129 step time: 0.2395\n", + "5/8, train_loss: 0.0140 step time: 0.2387\n", + "6/8, train_loss: 0.0141 step time: 0.2907\n", + "7/8, train_loss: 0.0102 step time: 0.3092\n", + "8/8, train_loss: 0.0155 step time: 0.3050\n", + "epoch 563 average loss: 0.0137\n", + "time consuming of epoch 563 is: 2.1439\n", + "----------\n", + "epoch 564/600\n", + "1/8, train_loss: 0.0149 step time: 0.3201\n", + "2/8, train_loss: 0.0139 step time: 0.2792\n", + "3/8, train_loss: 0.0146 step time: 0.2759\n", + "4/8, train_loss: 0.0112 step time: 0.2639\n", + "5/8, train_loss: 0.0136 step time: 0.2607\n", + "6/8, train_loss: 0.0146 step time: 0.2570\n", + "7/8, train_loss: 0.0122 step time: 0.2500\n", + "8/8, train_loss: 0.0139 step time: 0.2361\n", + "epoch 564 average loss: 0.0136\n", + "time consuming of epoch 564 is: 2.1447\n", + "----------\n", + "epoch 565/600\n", + "1/8, train_loss: 0.0132 step time: 0.2835\n", + "2/8, train_loss: 0.0138 step time: 0.2398\n", + "3/8, train_loss: 0.0137 step time: 0.2372\n", + "4/8, train_loss: 0.0111 step time: 0.2353\n", + "5/8, train_loss: 0.0129 step time: 0.2749\n", + "6/8, train_loss: 0.0166 step time: 0.2783\n", + "7/8, train_loss: 0.0156 step time: 0.2740\n", + "8/8, train_loss: 0.0144 step time: 0.2854\n", + "epoch 565 average loss: 0.0139\n", + "saved new best metric model\n", + "current epoch: 565 current mean dice: 0.9570 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 565 is: 3.8064\n", + "----------\n", + "epoch 566/600\n", + "1/8, train_loss: 0.0144 step time: 0.2843\n", + "2/8, train_loss: 0.0140 step time: 0.2134\n", + "3/8, train_loss: 0.0144 step time: 0.2217\n", + "4/8, train_loss: 0.0146 step time: 0.2218\n", + "5/8, train_loss: 0.0103 step time: 0.2293\n", + "6/8, train_loss: 0.0145 step time: 0.2388\n", + "7/8, train_loss: 0.0115 step time: 0.2437\n", + "8/8, train_loss: 0.0121 step time: 0.3030\n", + "epoch 566 average loss: 0.0132\n", + "time consuming of epoch 566 is: 1.9587\n", + "----------\n", + "epoch 567/600\n", + "1/8, train_loss: 0.0141 step time: 0.3355\n", + "2/8, train_loss: 0.0138 step time: 0.2807\n", + "3/8, train_loss: 0.0110 step time: 0.2573\n", + "4/8, train_loss: 0.0156 step time: 0.2518\n", + "5/8, train_loss: 0.0144 step time: 0.2596\n", + "6/8, train_loss: 0.0136 step time: 0.2557\n", + "7/8, train_loss: 0.0145 step time: 0.2572\n", + "8/8, train_loss: 0.0145 step time: 0.2509\n", + "epoch 567 average loss: 0.0139\n", + "time consuming of epoch 567 is: 2.1502\n", + "----------\n", + "epoch 568/600\n", + "1/8, train_loss: 0.0137 step time: 0.2900\n", + "2/8, train_loss: 0.0146 step time: 0.3190\n", + "3/8, train_loss: 0.0164 step time: 0.3527\n", + "4/8, train_loss: 0.0145 step time: 0.3196\n", + "5/8, train_loss: 0.0112 step time: 0.2951\n", + "6/8, train_loss: 0.0150 step time: 0.2814\n", + "7/8, train_loss: 0.0127 step time: 0.2766\n", + "8/8, train_loss: 0.0132 step time: 0.2762\n", + "epoch 568 average loss: 0.0139\n", + "time consuming of epoch 568 is: 2.4121\n", + "----------\n", + "epoch 569/600\n", + "1/8, train_loss: 0.0147 step time: 0.2996\n", + "2/8, train_loss: 0.0119 step time: 0.2535\n", + "3/8, train_loss: 0.0121 step time: 0.2492\n", + "4/8, train_loss: 0.0154 step time: 0.2533\n", + "5/8, train_loss: 0.0122 step time: 0.3694\n", + "6/8, train_loss: 0.0164 step time: 0.3542\n", + "7/8, train_loss: 0.0131 step time: 0.2772\n", + "8/8, train_loss: 0.0141 step time: 0.2744\n", + "epoch 569 average loss: 0.0137\n", + "time consuming of epoch 569 is: 2.3321\n", + "----------\n", + "epoch 570/600\n", + "1/8, train_loss: 0.0115 step time: 0.3015\n", + "2/8, train_loss: 0.0143 step time: 0.2503\n", + "3/8, train_loss: 0.0147 step time: 0.2460\n", + "4/8, train_loss: 0.0158 step time: 0.2478\n", + "5/8, train_loss: 0.0140 step time: 0.2503\n", + "6/8, train_loss: 0.0114 step time: 0.2428\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "7/8, train_loss: 0.0130 step time: 0.2395\n", + "8/8, train_loss: 0.0131 step time: 0.2599\n", + "epoch 570 average loss: 0.0135\n", + "current epoch: 570 current mean dice: 0.9564 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 570 is: 3.4808\n", + "----------\n", + "epoch 571/600\n", + "1/8, train_loss: 0.0124 step time: 0.3131\n", + "2/8, train_loss: 0.0163 step time: 0.2542\n", + "3/8, train_loss: 0.0127 step time: 0.2498\n", + "4/8, train_loss: 0.0159 step time: 0.2447\n", + "5/8, train_loss: 0.0161 step time: 0.2467\n", + "6/8, train_loss: 0.0150 step time: 0.2346\n", + "7/8, train_loss: 0.0142 step time: 0.2817\n", + "8/8, train_loss: 0.0121 step time: 0.3427\n", + "epoch 571 average loss: 0.0143\n", + "time consuming of epoch 571 is: 2.1687\n", + "----------\n", + "epoch 572/600\n", + "1/8, train_loss: 0.0134 step time: 0.3317\n", + "2/8, train_loss: 0.0140 step time: 0.2901\n", + "3/8, train_loss: 0.0131 step time: 0.2800\n", + "4/8, train_loss: 0.0135 step time: 0.2641\n", + "5/8, train_loss: 0.0140 step time: 0.2651\n", + "6/8, train_loss: 0.0156 step time: 0.2617\n", + "7/8, train_loss: 0.0179 step time: 0.2425\n", + "8/8, train_loss: 0.0138 step time: 0.2324\n", + "epoch 572 average loss: 0.0144\n", + "time consuming of epoch 572 is: 2.1691\n", + "----------\n", + "epoch 573/600\n", + "1/8, train_loss: 0.0136 step time: 0.2835\n", + "2/8, train_loss: 0.0142 step time: 0.2430\n", + "3/8, train_loss: 0.0126 step time: 0.2444\n", + "4/8, train_loss: 0.0149 step time: 0.4018\n", + "5/8, train_loss: 0.0173 step time: 0.4729\n", + "6/8, train_loss: 0.0149 step time: 0.2708\n", + "7/8, train_loss: 0.0118 step time: 0.2485\n", + "8/8, train_loss: 0.0157 step time: 0.2524\n", + "epoch 573 average loss: 0.0144\n", + "time consuming of epoch 573 is: 2.4188\n", + "----------\n", + "epoch 574/600\n", + "1/8, train_loss: 0.0165 step time: 0.3020\n", + "2/8, train_loss: 0.0143 step time: 0.2468\n", + "3/8, train_loss: 0.0129 step time: 0.2630\n", + "4/8, train_loss: 0.0154 step time: 0.2666\n", + "5/8, train_loss: 0.0175 step time: 0.3344\n", + "6/8, train_loss: 0.0159 step time: 0.3198\n", + "7/8, train_loss: 0.0119 step time: 0.3130\n", + "8/8, train_loss: 0.0155 step time: 0.3013\n", + "epoch 574 average loss: 0.0150\n", + "time consuming of epoch 574 is: 2.3484\n", + "----------\n", + "epoch 575/600\n", + "1/8, train_loss: 0.0162 step time: 0.3148\n", + "2/8, train_loss: 0.0146 step time: 0.2578\n", + "3/8, train_loss: 0.0196 step time: 0.2583\n", + "4/8, train_loss: 0.0176 step time: 0.2645\n", + "5/8, train_loss: 0.0278 step time: 0.2587\n", + "6/8, train_loss: 0.0203 step time: 0.2587\n", + "7/8, train_loss: 0.0189 step time: 0.3994\n", + "8/8, train_loss: 0.0173 step time: 0.3447\n", + "epoch 575 average loss: 0.0190\n", + "current epoch: 575 current mean dice: 0.6157 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 575 is: 3.5838\n", + "----------\n", + "epoch 576/600\n", + "1/8, train_loss: 0.0267 step time: 0.3742\n", + "2/8, train_loss: 0.0423 step time: 0.3487\n", + "3/8, train_loss: 0.0208 step time: 0.3219\n", + "4/8, train_loss: 0.0596 step time: 0.3010\n", + "5/8, train_loss: 0.0293 step time: 0.2699\n", + "6/8, train_loss: 0.0247 step time: 0.2537\n", + "7/8, train_loss: 0.0616 step time: 0.2398\n", + "8/8, train_loss: 0.0180 step time: 0.2390\n", + "epoch 576 average loss: 0.0354\n", + "time consuming of epoch 576 is: 2.3494\n", + "----------\n", + "epoch 577/600\n", + "1/8, train_loss: 0.0172 step time: 0.2959\n", + "2/8, train_loss: 0.0322 step time: 0.2391\n", + "3/8, train_loss: 0.0407 step time: 0.2459\n", + "4/8, train_loss: 0.0196 step time: 0.3281\n", + "5/8, train_loss: 0.0171 step time: 0.3268\n", + "6/8, train_loss: 0.0202 step time: 0.2792\n", + "7/8, train_loss: 0.0321 step time: 0.2779\n", + "8/8, train_loss: 0.0475 step time: 0.2798\n", + "epoch 577 average loss: 0.0283\n", + "time consuming of epoch 577 is: 2.2742\n", + "----------\n", + "epoch 578/600\n", + "1/8, train_loss: 0.0216 step time: 0.3330\n", + "2/8, train_loss: 0.0193 step time: 0.2659\n", + "3/8, train_loss: 0.0246 step time: 0.2686\n", + "4/8, train_loss: 0.0217 step time: 0.2494\n", + "5/8, train_loss: 0.0611 step time: 0.2525\n", + "6/8, train_loss: 0.0256 step time: 0.3037\n", + "7/8, train_loss: 0.0258 step time: 0.3017\n", + "8/8, train_loss: 0.0495 step time: 0.2911\n", + "epoch 578 average loss: 0.0312\n", + "time consuming of epoch 578 is: 2.2673\n", + "----------\n", + "epoch 579/600\n", + "1/8, train_loss: 0.0287 step time: 0.3131\n", + "2/8, train_loss: 0.0192 step time: 0.2610\n", + "3/8, train_loss: 0.0218 step time: 0.2511\n", + "4/8, train_loss: 0.0208 step time: 0.2446\n", + "5/8, train_loss: 0.0256 step time: 0.2426\n", + "6/8, train_loss: 0.0236 step time: 0.2344\n", + "7/8, train_loss: 0.0237 step time: 0.2271\n", + "8/8, train_loss: 0.0341 step time: 0.2308\n", + "epoch 579 average loss: 0.0247\n", + "time consuming of epoch 579 is: 2.0062\n", + "----------\n", + "epoch 580/600\n", + "1/8, train_loss: 0.0241 step time: 0.2838\n", + "2/8, train_loss: 0.0213 step time: 0.2402\n", + "3/8, train_loss: 0.0162 step time: 0.2398\n", + "4/8, train_loss: 0.0223 step time: 0.2467\n", + "5/8, train_loss: 0.0239 step time: 0.2537\n", + "6/8, train_loss: 0.0266 step time: 0.2510\n", + "7/8, train_loss: 0.0197 step time: 0.2480\n", + "8/8, train_loss: 0.0190 step time: 0.2935\n", + "epoch 580 average loss: 0.0216\n", + "current epoch: 580 current mean dice: 0.5842 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 580 is: 3.3831\n", + "----------\n", + "epoch 581/600\n", + "1/8, train_loss: 0.0208 step time: 0.3043\n", + "2/8, train_loss: 0.0243 step time: 0.2567\n", + "3/8, train_loss: 0.0150 step time: 0.2624\n", + "4/8, train_loss: 0.0211 step time: 0.2601\n", + "5/8, train_loss: 0.0207 step time: 0.2537\n", + "6/8, train_loss: 0.0200 step time: 0.2360\n", + "7/8, train_loss: 0.0142 step time: 0.2273\n", + "8/8, train_loss: 0.0199 step time: 0.2274\n", + "epoch 581 average loss: 0.0195\n", + "time consuming of epoch 581 is: 2.0291\n", + "----------\n", + "epoch 582/600\n", + "1/8, train_loss: 0.0205 step time: 0.5857\n", + "2/8, train_loss: 0.0220 step time: 0.2857\n", + "3/8, train_loss: 0.0212 step time: 0.2568\n", + "4/8, train_loss: 0.0202 step time: 0.2480\n", + "5/8, train_loss: 0.0169 step time: 0.2397\n", + "6/8, train_loss: 0.0185 step time: 0.2460\n", + "7/8, train_loss: 0.0153 step time: 0.2693\n", + "8/8, train_loss: 0.0191 step time: 0.2764\n", + "epoch 582 average loss: 0.0192\n", + "time consuming of epoch 582 is: 2.4090\n", + "----------\n", + "epoch 583/600\n", + "1/8, train_loss: 0.0169 step time: 0.3217\n", + "2/8, train_loss: 0.0161 step time: 0.2769\n", + "3/8, train_loss: 0.0189 step time: 0.2645\n", + "4/8, train_loss: 0.0221 step time: 0.2606\n", + "5/8, train_loss: 0.0195 step time: 0.2523\n", + "6/8, train_loss: 0.0188 step time: 0.2425\n", + "7/8, train_loss: 0.0171 step time: 0.2274\n", + "8/8, train_loss: 0.0226 step time: 0.2284\n", + "epoch 583 average loss: 0.0190\n", + "time consuming of epoch 583 is: 2.0756\n", + "----------\n", + "epoch 584/600\n", + "1/8, train_loss: 0.0178 step time: 0.2821\n", + "2/8, train_loss: 0.0172 step time: 0.2427\n", + "3/8, train_loss: 0.0163 step time: 0.2449\n", + "4/8, train_loss: 0.0233 step time: 0.3345\n", + "5/8, train_loss: 0.0173 step time: 0.3845\n", + "6/8, train_loss: 0.0200 step time: 0.2720\n", + "7/8, train_loss: 0.0156 step time: 0.2522\n", + "8/8, train_loss: 0.0182 step time: 0.2409\n", + "epoch 584 average loss: 0.0182\n", + "time consuming of epoch 584 is: 2.2554\n", + "----------\n", + "epoch 585/600\n", + "1/8, train_loss: 0.0216 step time: 0.2776\n", + "2/8, train_loss: 0.0234 step time: 0.2341\n", + "3/8, train_loss: 0.0189 step time: 0.2320\n", + "4/8, train_loss: 0.0155 step time: 0.2310\n", + "5/8, train_loss: 0.0181 step time: 0.2312\n", + "6/8, train_loss: 0.0184 step time: 0.5762\n", + "7/8, train_loss: 0.0172 step time: 0.2038\n", + "8/8, train_loss: 0.0141 step time: 0.2035\n", + "epoch 585 average loss: 0.0184\n", + "current epoch: 585 current mean dice: 0.9476 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 585 is: 3.8190\n", + "----------\n", + "epoch 586/600\n", + "1/8, train_loss: 0.0221 step time: 0.3041\n", + "2/8, train_loss: 0.0193 step time: 0.2530\n", + "3/8, train_loss: 0.0194 step time: 0.2511\n", + "4/8, train_loss: 0.0144 step time: 0.2564\n", + "5/8, train_loss: 0.0224 step time: 0.2544\n", + "6/8, train_loss: 0.0169 step time: 0.2564\n", + "7/8, train_loss: 0.0175 step time: 0.2531\n", + "8/8, train_loss: 0.0150 step time: 0.2549\n", + "epoch 586 average loss: 0.0184\n", + "time consuming of epoch 586 is: 2.0845\n", + "----------\n", + "epoch 587/600\n", + "1/8, train_loss: 0.0180 step time: 0.3033\n", + "2/8, train_loss: 0.0138 step time: 0.2540\n", + "3/8, train_loss: 0.0162 step time: 0.3777\n", + "4/8, train_loss: 0.0169 step time: 0.3005\n", + "5/8, train_loss: 0.0182 step time: 0.2977\n", + "6/8, train_loss: 0.0183 step time: 0.2835\n", + "7/8, train_loss: 0.0141 step time: 0.2805\n", + "8/8, train_loss: 0.0169 step time: 0.2580\n", + "epoch 587 average loss: 0.0165\n", + "time consuming of epoch 587 is: 2.3567\n", + "----------\n", + "epoch 588/600\n", + "1/8, train_loss: 0.0169 step time: 0.3040\n", + "2/8, train_loss: 0.0157 step time: 0.2603\n", + "3/8, train_loss: 0.0196 step time: 0.2434\n", + "4/8, train_loss: 0.0199 step time: 0.2664\n", + "5/8, train_loss: 0.0147 step time: 0.3162\n", + "6/8, train_loss: 0.0151 step time: 0.2892\n", + "7/8, train_loss: 0.0154 step time: 0.2603\n", + "8/8, train_loss: 0.0134 step time: 0.2456\n", + "epoch 588 average loss: 0.0163\n", + "time consuming of epoch 588 is: 2.1869\n", + "----------\n", + "epoch 589/600\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "1/8, train_loss: 0.0148 step time: 0.2963\n", + "2/8, train_loss: 0.0142 step time: 0.2491\n", + "3/8, train_loss: 0.0160 step time: 0.2398\n", + "4/8, train_loss: 0.0185 step time: 0.2387\n", + "5/8, train_loss: 0.0187 step time: 0.2353\n", + "6/8, train_loss: 0.0134 step time: 0.2410\n", + "7/8, train_loss: 0.0173 step time: 0.2309\n", + "8/8, train_loss: 0.0172 step time: 0.2321\n", + "epoch 589 average loss: 0.0162\n", + "time consuming of epoch 589 is: 1.9646\n", + "----------\n", + "epoch 590/600\n", + "1/8, train_loss: 0.0149 step time: 0.2860\n", + "2/8, train_loss: 0.0176 step time: 0.3303\n", + "3/8, train_loss: 0.0145 step time: 0.3770\n", + "4/8, train_loss: 0.0157 step time: 0.3063\n", + "5/8, train_loss: 0.0167 step time: 0.2762\n", + "6/8, train_loss: 0.0138 step time: 0.2607\n", + "7/8, train_loss: 0.0205 step time: 0.2597\n", + "8/8, train_loss: 0.0162 step time: 0.2618\n", + "epoch 590 average loss: 0.0162\n", + "current epoch: 590 current mean dice: 0.9524 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 590 is: 3.6541\n", + "----------\n", + "epoch 591/600\n", + "1/8, train_loss: 0.0179 step time: 0.3185\n", + "2/8, train_loss: 0.0162 step time: 0.2596\n", + "3/8, train_loss: 0.0159 step time: 0.2525\n", + "4/8, train_loss: 0.0172 step time: 0.2594\n", + "5/8, train_loss: 0.0150 step time: 0.2445\n", + "6/8, train_loss: 0.0203 step time: 0.2416\n", + "7/8, train_loss: 0.0155 step time: 0.2225\n", + "8/8, train_loss: 0.0192 step time: 0.2232\n", + "epoch 591 average loss: 0.0171\n", + "time consuming of epoch 591 is: 2.0231\n", + "----------\n", + "epoch 592/600\n", + "1/8, train_loss: 0.0133 step time: 0.5008\n", + "2/8, train_loss: 0.0136 step time: 0.2161\n", + "3/8, train_loss: 0.0166 step time: 0.2123\n", + "4/8, train_loss: 0.0156 step time: 0.2163\n", + "5/8, train_loss: 0.0162 step time: 0.2322\n", + "6/8, train_loss: 0.0157 step time: 0.2843\n", + "7/8, train_loss: 0.0126 step time: 0.3000\n", + "8/8, train_loss: 0.0233 step time: 0.2872\n", + "epoch 592 average loss: 0.0159\n", + "time consuming of epoch 592 is: 2.2507\n", + "----------\n", + "epoch 593/600\n", + "1/8, train_loss: 0.0268 step time: 0.3071\n", + "2/8, train_loss: 0.0133 step time: 0.2545\n", + "3/8, train_loss: 0.0190 step time: 0.2471\n", + "4/8, train_loss: 0.0175 step time: 0.2415\n", + "5/8, train_loss: 0.0192 step time: 0.2407\n", + "6/8, train_loss: 0.0144 step time: 0.3130\n", + "7/8, train_loss: 0.0137 step time: 0.4752\n", + "8/8, train_loss: 0.0175 step time: 0.2735\n", + "epoch 593 average loss: 0.0177\n", + "time consuming of epoch 593 is: 2.3541\n", + "----------\n", + "epoch 594/600\n", + "1/8, train_loss: 0.0182 step time: 0.3081\n", + "2/8, train_loss: 0.0204 step time: 0.2519\n", + "3/8, train_loss: 0.0159 step time: 0.2555\n", + "4/8, train_loss: 0.0162 step time: 0.2554\n", + "5/8, train_loss: 0.0161 step time: 0.2515\n", + "6/8, train_loss: 0.0191 step time: 0.2488\n", + "7/8, train_loss: 0.0175 step time: 0.2403\n", + "8/8, train_loss: 0.0121 step time: 0.2411\n", + "epoch 594 average loss: 0.0169\n", + "time consuming of epoch 594 is: 2.0542\n", + "----------\n", + "epoch 595/600\n", + "1/8, train_loss: 0.0151 step time: 0.2936\n", + "2/8, train_loss: 0.0138 step time: 0.2961\n", + "3/8, train_loss: 0.0150 step time: 0.3332\n", + "4/8, train_loss: 0.0222 step time: 0.2901\n", + "5/8, train_loss: 0.0152 step time: 0.2760\n", + "6/8, train_loss: 0.0160 step time: 0.2778\n", + "7/8, train_loss: 0.0187 step time: 0.2739\n", + "8/8, train_loss: 0.0175 step time: 0.2603\n", + "epoch 595 average loss: 0.0167\n", + "current epoch: 595 current mean dice: 0.9536 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 595 is: 3.5425\n", + "----------\n", + "epoch 596/600\n", + "1/8, train_loss: 0.0151 step time: 0.3051\n", + "2/8, train_loss: 0.0151 step time: 0.2521\n", + "3/8, train_loss: 0.0162 step time: 0.2437\n", + "4/8, train_loss: 0.0165 step time: 0.2434\n", + "5/8, train_loss: 0.0154 step time: 0.3578\n", + "6/8, train_loss: 0.0138 step time: 0.4849\n", + "7/8, train_loss: 0.0148 step time: 0.2703\n", + "8/8, train_loss: 0.0142 step time: 0.2516\n", + "epoch 596 average loss: 0.0151\n", + "time consuming of epoch 596 is: 2.4104\n", + "----------\n", + "epoch 597/600\n", + "1/8, train_loss: 0.0157 step time: 0.2957\n", + "2/8, train_loss: 0.0170 step time: 0.2438\n", + "3/8, train_loss: 0.0154 step time: 0.2434\n", + "4/8, train_loss: 0.0147 step time: 0.2382\n", + "5/8, train_loss: 0.0172 step time: 0.2416\n", + "6/8, train_loss: 0.0157 step time: 0.2393\n", + "7/8, train_loss: 0.0124 step time: 0.2195\n", + "8/8, train_loss: 0.0173 step time: 0.3617\n", + "epoch 597 average loss: 0.0157\n", + "time consuming of epoch 597 is: 2.0847\n", + "----------\n", + "epoch 598/600\n", + "1/8, train_loss: 0.0168 step time: 0.4376\n", + "2/8, train_loss: 0.0163 step time: 0.2782\n", + "3/8, train_loss: 0.0144 step time: 0.2617\n", + "4/8, train_loss: 0.0183 step time: 0.2626\n", + "5/8, train_loss: 0.0160 step time: 0.2489\n", + "6/8, train_loss: 0.0135 step time: 0.2485\n", + "7/8, train_loss: 0.0159 step time: 0.2269\n", + "8/8, train_loss: 0.0129 step time: 0.2705\n", + "epoch 598 average loss: 0.0155\n", + "time consuming of epoch 598 is: 2.2365\n", + "----------\n", + "epoch 599/600\n", + "1/8, train_loss: 0.0137 step time: 0.3769\n", + "2/8, train_loss: 0.0160 step time: 0.2753\n", + "3/8, train_loss: 0.0170 step time: 0.2577\n", + "4/8, train_loss: 0.0162 step time: 0.2495\n", + "5/8, train_loss: 0.0192 step time: 0.2774\n", + "6/8, train_loss: 0.0162 step time: 0.3099\n", + "7/8, train_loss: 0.0140 step time: 0.2596\n", + "8/8, train_loss: 0.0174 step time: 0.2547\n", + "epoch 599 average loss: 0.0162\n", + "time consuming of epoch 599 is: 2.2626\n", + "----------\n", + "epoch 600/600\n", + "1/8, train_loss: 0.0149 step time: 0.3104\n", + "2/8, train_loss: 0.0137 step time: 0.2571\n", + "3/8, train_loss: 0.0172 step time: 0.2480\n", + "4/8, train_loss: 0.0157 step time: 0.2512\n", + "5/8, train_loss: 0.0133 step time: 0.2491\n", + "6/8, train_loss: 0.0148 step time: 0.3255\n", + "7/8, train_loss: 0.0131 step time: 0.3099\n", + "8/8, train_loss: 0.0143 step time: 0.2755\n", + "epoch 600 average loss: 0.0146\n", + "current epoch: 600 current mean dice: 0.9553 best mean dice: 0.9570 at epoch: 565\n", + "time consuming of epoch 600 is: 3.5567\n", + "train completed, best_metric: 0.9570 at epoch: 565 total time: 1423.5083\n", + "total time of 600 epochs with MONAI fast training: 1423.5083, time of preparing cache: 52.4564\n" ] } ], From af0d17c3567d2b377c3721c794a4530b85003661 Mon Sep 17 00:00:00 2001 From: Nic Ma Date: Thu, 21 Jul 2022 13:42:55 +0800 Subject: [PATCH 5/5] [DLMED] update charts Signed-off-by: Nic Ma --- acceleration/fast_training_tutorial.ipynb | 7571 +-------------------- 1 file changed, 8 insertions(+), 7563 deletions(-) diff --git a/acceleration/fast_training_tutorial.ipynb b/acceleration/fast_training_tutorial.ipynb index 2456fca175..9efd12fb66 100644 --- a/acceleration/fast_training_tutorial.ipynb +++ b/acceleration/fast_training_tutorial.ipynb @@ -743,7568 +743,13 @@ }, { "cell_type": "code", - "execution_count": 9, + "execution_count": null, "metadata": { "vscode": { "languageId": "python" } }, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "Loading dataset: 100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 32/32 [00:42<00:00, 1.33s/it]\n", - "Loading dataset: 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 9/9 [00:09<00:00, 1.05s/it]\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "----------\n", - "epoch 1/600\n", - "1/8, train_loss: 0.8296 step time: 1.6296\n", - "2/8, train_loss: 0.7641 step time: 0.2118\n", - "3/8, train_loss: 0.5814 step time: 0.2088\n", - "4/8, train_loss: 0.4991 step time: 0.2105\n", - "5/8, train_loss: 0.4853 step time: 0.2091\n", - "6/8, train_loss: 0.4718 step time: 0.2094\n", - "7/8, train_loss: 0.4114 step time: 0.1994\n", - "8/8, train_loss: 0.4430 step time: 0.2013\n", - "epoch 1 average loss: 0.5607\n", - "time consuming of epoch 1 is: 3.0811\n", - "----------\n", - "epoch 2/600\n", - "1/8, train_loss: 0.4435 step time: 0.2669\n", - "2/8, train_loss: 0.4586 step time: 0.2144\n", - "3/8, train_loss: 0.4187 step time: 0.2161\n", - "4/8, train_loss: 0.5169 step time: 0.2075\n", - "5/8, train_loss: 0.5513 step time: 0.2170\n", - "6/8, train_loss: 0.4491 step time: 0.2170\n", - "7/8, train_loss: 0.5062 step time: 0.2004\n", - "8/8, train_loss: 0.5096 step time: 0.2005\n", - "epoch 2 average loss: 0.4817\n", - "time consuming of epoch 2 is: 1.7416\n", - "----------\n", - "epoch 3/600\n", - "1/8, train_loss: 0.4188 step time: 0.2660\n", - "2/8, train_loss: 0.4919 step time: 0.2132\n", - "3/8, train_loss: 0.5508 step time: 0.2119\n", - "4/8, train_loss: 0.4027 step time: 0.2136\n", - "5/8, train_loss: 0.4910 step time: 0.2153\n", - "6/8, train_loss: 0.3804 step time: 0.2130\n", - "7/8, train_loss: 0.3431 step time: 0.2019\n", - "8/8, train_loss: 0.3834 step time: 0.2000\n", - "epoch 3 average loss: 0.4328\n", - "time consuming of epoch 3 is: 1.7364\n", - "----------\n", - "epoch 4/600\n", - "1/8, train_loss: 0.2439 step time: 0.2637\n", - "2/8, train_loss: 0.2532 step time: 0.2147\n", - "3/8, train_loss: 0.2174 step time: 0.2192\n", - "4/8, train_loss: 0.2359 step time: 0.2136\n", - "5/8, train_loss: 0.1941 step time: 0.2161\n", - "6/8, train_loss: 0.2294 step time: 0.2175\n", - "7/8, train_loss: 0.2436 step time: 0.2000\n", - "8/8, train_loss: 0.3432 step time: 0.2017\n", - "epoch 4 average loss: 0.2451\n", - "time consuming of epoch 4 is: 1.7480\n", - "----------\n", - "epoch 5/600\n", - "1/8, train_loss: 0.1843 step time: 0.2634\n", - "2/8, train_loss: 0.2163 step time: 0.2137\n", - "3/8, train_loss: 0.1557 step time: 0.2131\n", - "4/8, train_loss: 0.1434 step time: 0.2127\n", - "5/8, train_loss: 0.1158 step time: 0.2117\n", - "6/8, train_loss: 0.2312 step time: 0.2108\n", - "7/8, train_loss: 0.2958 step time: 0.2000\n", - "8/8, train_loss: 0.1528 step time: 0.2006\n", - "epoch 5 average loss: 0.1869\n", - "saved new best metric model\n", - "current epoch: 5 current mean dice: 0.0626 best mean dice: 0.0626 at epoch: 5\n", - "time consuming of epoch 5 is: 2.8338\n", - "----------\n", - "epoch 6/600\n", - "1/8, train_loss: 0.1136 step time: 0.2713\n", - "2/8, train_loss: 0.1565 step time: 0.2132\n", - "3/8, train_loss: 0.2769 step time: 0.2103\n", - "4/8, train_loss: 0.1773 step time: 0.2101\n", - "5/8, train_loss: 0.2183 step time: 0.2102\n", - "6/8, train_loss: 0.1236 step time: 0.2118\n", - "7/8, train_loss: 0.2068 step time: 0.2001\n", - "8/8, train_loss: 0.1687 step time: 0.2004\n", - "epoch 6 average loss: 0.1802\n", - "time consuming of epoch 6 is: 1.7299\n", - "----------\n", - "epoch 7/600\n", - "1/8, train_loss: 0.1709 step time: 0.2542\n", - "2/8, train_loss: 0.2031 step time: 0.2163\n", - "3/8, train_loss: 0.1453 step time: 0.2128\n", - "4/8, train_loss: 0.2224 step time: 0.2095\n", - "5/8, train_loss: 0.1693 step time: 0.2096\n", - "6/8, train_loss: 0.3097 step time: 0.2094\n", - "7/8, train_loss: 0.3081 step time: 0.1962\n", - "8/8, train_loss: 0.1908 step time: 0.1955\n", - "epoch 7 average loss: 0.2150\n", - "time consuming of epoch 7 is: 1.7049\n", - "----------\n", - "epoch 8/600\n", - "1/8, train_loss: 0.1996 step time: 0.2565\n", - "2/8, train_loss: 0.2088 step time: 0.2068\n", - "3/8, train_loss: 0.1894 step time: 0.2094\n", - "4/8, train_loss: 0.2062 step time: 0.2084\n", - "5/8, train_loss: 0.1687 step time: 0.2070\n", - "6/8, train_loss: 0.1327 step time: 0.2076\n", - "7/8, train_loss: 0.1556 step time: 0.1961\n", - "8/8, train_loss: 0.1491 step time: 0.1957\n", - "epoch 8 average loss: 0.1762\n", - "time consuming of epoch 8 is: 1.6886\n", - "----------\n", - "epoch 9/600\n", - "1/8, train_loss: 0.1722 step time: 0.2545\n", - "2/8, train_loss: 0.0934 step time: 0.2082\n", - "3/8, train_loss: 0.1434 step time: 0.2081\n", - "4/8, train_loss: 0.0838 step time: 0.2071\n", - "5/8, train_loss: 0.1419 step time: 0.2077\n", - "6/8, train_loss: 0.1540 step time: 0.2072\n", - "7/8, train_loss: 0.1397 step time: 0.1963\n", - "8/8, train_loss: 0.1324 step time: 0.1954\n", - "epoch 9 average loss: 0.1326\n", - "time consuming of epoch 9 is: 1.6858\n", - "----------\n", - "epoch 10/600\n", - "1/8, train_loss: 0.1358 step time: 0.2547\n", - "2/8, train_loss: 0.0968 step time: 0.2084\n", - "3/8, train_loss: 0.1124 step time: 0.2084\n", - "4/8, train_loss: 0.1009 step time: 0.2071\n", - "5/8, train_loss: 0.1119 step time: 0.2074\n", - "6/8, train_loss: 0.1132 step time: 0.2074\n", - "7/8, train_loss: 0.0996 step time: 0.1964\n", - "8/8, train_loss: 0.1189 step time: 0.1948\n", - "epoch 10 average loss: 0.1112\n", - "current epoch: 10 current mean dice: 0.0000 best mean dice: 0.0626 at epoch: 5\n", - "time consuming of epoch 10 is: 2.4952\n", - "----------\n", - "epoch 11/600\n", - "1/8, train_loss: 0.0958 step time: 0.2555\n", - "2/8, train_loss: 0.0749 step time: 0.2080\n", - "3/8, train_loss: 0.1171 step time: 0.2074\n", - "4/8, train_loss: 0.1097 step time: 0.2070\n", - "5/8, train_loss: 0.1265 step time: 0.2080\n", - "6/8, train_loss: 0.1216 step time: 0.2077\n", - "7/8, train_loss: 0.1101 step time: 0.1962\n", - "8/8, train_loss: 0.1316 step time: 0.1954\n", - "epoch 11 average loss: 0.1109\n", - "time consuming of epoch 11 is: 1.6864\n", - "----------\n", - "epoch 12/600\n", - "1/8, train_loss: 0.1670 step time: 0.2563\n", - "2/8, train_loss: 0.0785 step time: 0.2070\n", - "3/8, train_loss: 0.1389 step time: 0.2078\n", - "4/8, train_loss: 0.1350 step time: 0.2070\n", - "5/8, train_loss: 0.1447 step time: 0.2077\n", - "6/8, train_loss: 0.1352 step time: 0.2084\n", - "7/8, train_loss: 0.0995 step time: 0.1960\n", - "8/8, train_loss: 0.0594 step time: 0.1954\n", - "epoch 12 average loss: 0.1198\n", - "time consuming of epoch 12 is: 1.6868\n", - "----------\n", - "epoch 13/600\n", - "1/8, train_loss: 0.0674 step time: 0.2465\n", - "2/8, train_loss: 0.1410 step time: 0.2086\n", - "3/8, train_loss: 0.1068 step time: 0.2081\n", - "4/8, train_loss: 0.1073 step time: 0.2073\n", - "5/8, train_loss: 0.0862 step time: 0.2081\n", - "6/8, train_loss: 0.0855 step time: 0.2074\n", - "7/8, train_loss: 0.0986 step time: 0.1966\n", - "8/8, train_loss: 0.2790 step time: 0.1955\n", - "epoch 13 average loss: 0.1215\n", - "time consuming of epoch 13 is: 1.6793\n", - "----------\n", - "epoch 14/600\n", - "1/8, train_loss: 0.0853 step time: 0.2546\n", - "2/8, train_loss: 0.0896 step time: 0.2092\n", - "3/8, train_loss: 0.1138 step time: 0.2074\n", - "4/8, train_loss: 0.0790 step time: 0.2069\n", - "5/8, train_loss: 0.0676 step time: 0.2093\n", - "6/8, train_loss: 0.1342 step time: 0.2084\n", - "7/8, train_loss: 0.1085 step time: 0.1961\n", - "8/8, train_loss: 0.0938 step time: 0.1957\n", - "epoch 14 average loss: 0.0965\n", - "time consuming of epoch 14 is: 1.6888\n", - "----------\n", - "epoch 15/600\n", - "1/8, train_loss: 0.0492 step time: 0.2534\n", - "2/8, train_loss: 0.0648 step time: 0.2077\n", - "3/8, train_loss: 0.0709 step time: 0.2081\n", - "4/8, train_loss: 0.0982 step time: 0.2076\n", - "5/8, train_loss: 0.0664 step time: 0.2070\n", - "6/8, train_loss: 0.1038 step time: 0.2073\n", - "7/8, train_loss: 0.1088 step time: 0.1967\n", - "8/8, train_loss: 0.0679 step time: 0.1957\n", - "epoch 15 average loss: 0.0788\n", - "saved new best metric model\n", - "current epoch: 15 current mean dice: 0.6743 best mean dice: 0.6743 at epoch: 15\n", - "time consuming of epoch 15 is: 2.7746\n", - "----------\n", - "epoch 16/600\n", - "1/8, train_loss: 0.1002 step time: 0.2685\n", - "2/8, train_loss: 0.0568 step time: 0.2078\n", - "3/8, train_loss: 0.0559 step time: 0.2077\n", - "4/8, train_loss: 0.0857 step time: 0.2094\n", - "5/8, train_loss: 0.0572 step time: 0.2089\n", - "6/8, train_loss: 0.0645 step time: 0.2072\n", - "7/8, train_loss: 0.0604 step time: 0.1962\n", - "8/8, train_loss: 0.0804 step time: 0.1956\n", - "epoch 16 average loss: 0.0701\n", - "time consuming of epoch 16 is: 1.7039\n", - "----------\n", - "epoch 17/600\n", - "1/8, train_loss: 0.0465 step time: 0.2479\n", - "2/8, train_loss: 0.1140 step time: 0.2075\n", - "3/8, train_loss: 0.2151 step time: 0.2083\n", - "4/8, train_loss: 0.0604 step time: 0.2086\n", - "5/8, train_loss: 0.0690 step time: 0.2093\n", - "6/8, train_loss: 0.0898 step time: 0.2075\n", - "7/8, train_loss: 0.0799 step time: 0.1961\n", - "8/8, train_loss: 0.1355 step time: 0.1950\n", - "epoch 17 average loss: 0.1013\n", - "time consuming of epoch 17 is: 1.6814\n", - "----------\n", - "epoch 18/600\n", - "1/8, train_loss: 0.0773 step time: 0.2523\n", - "2/8, train_loss: 0.0589 step time: 0.2069\n", - "3/8, train_loss: 0.0740 step time: 0.2085\n", - "4/8, train_loss: 0.0662 step time: 0.2082\n", - "5/8, train_loss: 0.1036 step time: 0.2079\n", - "6/8, train_loss: 0.0778 step time: 0.2071\n", - "7/8, train_loss: 0.1039 step time: 0.1962\n", - "8/8, train_loss: 0.0743 step time: 0.1956\n", - "epoch 18 average loss: 0.0795\n", - "time consuming of epoch 18 is: 1.6840\n", - "----------\n", - "epoch 19/600\n", - "1/8, train_loss: 0.1254 step time: 0.2494\n", - "2/8, train_loss: 0.0945 step time: 0.2074\n", - "3/8, train_loss: 0.0875 step time: 0.2076\n", - "4/8, train_loss: 0.0549 step time: 0.2085\n", - "5/8, train_loss: 0.0530 step time: 0.2076\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "6/8, train_loss: 0.0795 step time: 0.2082\n", - "7/8, train_loss: 0.0487 step time: 0.1961\n", - "8/8, train_loss: 0.0933 step time: 0.1946\n", - "epoch 19 average loss: 0.0796\n", - "time consuming of epoch 19 is: 1.6806\n", - "----------\n", - "epoch 20/600\n", - "1/8, train_loss: 0.0522 step time: 0.2536\n", - "2/8, train_loss: 0.0380 step time: 0.2069\n", - "3/8, train_loss: 0.0756 step time: 0.2076\n", - "4/8, train_loss: 0.0545 step time: 0.2078\n", - "5/8, train_loss: 0.0803 step time: 0.2085\n", - "6/8, train_loss: 0.0568 step time: 0.2081\n", - "7/8, train_loss: 0.0509 step time: 0.1962\n", - "8/8, train_loss: 0.0532 step time: 0.1956\n", - "epoch 20 average loss: 0.0577\n", - "saved new best metric model\n", - "current epoch: 20 current mean dice: 0.8040 best mean dice: 0.8040 at epoch: 20\n", - "time consuming of epoch 20 is: 2.7756\n", - "----------\n", - "epoch 21/600\n", - "1/8, train_loss: 0.1026 step time: 0.2645\n", - "2/8, train_loss: 0.0634 step time: 0.2071\n", - "3/8, train_loss: 0.0936 step time: 0.2077\n", - "4/8, train_loss: 0.0665 step time: 0.2069\n", - "5/8, train_loss: 0.0788 step time: 0.2093\n", - "6/8, train_loss: 0.0623 step time: 0.2091\n", - "7/8, train_loss: 0.0451 step time: 0.1961\n", - "8/8, train_loss: 0.0487 step time: 0.1953\n", - "epoch 21 average loss: 0.0701\n", - "time consuming of epoch 21 is: 1.6982\n", - "----------\n", - "epoch 22/600\n", - "1/8, train_loss: 0.0657 step time: 0.2550\n", - "2/8, train_loss: 0.0808 step time: 0.2090\n", - "3/8, train_loss: 0.0591 step time: 0.2091\n", - "4/8, train_loss: 0.0695 step time: 0.2076\n", - "5/8, train_loss: 0.0728 step time: 0.2071\n", - "6/8, train_loss: 0.0493 step time: 0.2070\n", - "7/8, train_loss: 0.0728 step time: 0.1961\n", - "8/8, train_loss: 0.0518 step time: 0.1954\n", - "epoch 22 average loss: 0.0652\n", - "time consuming of epoch 22 is: 1.6874\n", - "----------\n", - "epoch 23/600\n", - "1/8, train_loss: 0.0551 step time: 0.2557\n", - "2/8, train_loss: 0.0412 step time: 0.2078\n", - "3/8, train_loss: 0.1448 step time: 0.2073\n", - "4/8, train_loss: 0.0512 step time: 0.2072\n", - "5/8, train_loss: 0.0528 step time: 0.2074\n", - "6/8, train_loss: 0.0461 step time: 0.2083\n", - "7/8, train_loss: 0.0509 step time: 0.1951\n", - "8/8, train_loss: 0.0810 step time: 0.1956\n", - "epoch 23 average loss: 0.0654\n", - "time consuming of epoch 23 is: 1.6856\n", - "----------\n", - "epoch 24/600\n", - "1/8, train_loss: 0.0402 step time: 0.2584\n", - "2/8, train_loss: 0.0642 step time: 0.2071\n", - "3/8, train_loss: 0.0470 step time: 0.2073\n", - "4/8, train_loss: 0.0444 step time: 0.2075\n", - "5/8, train_loss: 0.0337 step time: 0.2068\n", - "6/8, train_loss: 0.0400 step time: 0.2069\n", - "7/8, train_loss: 0.0680 step time: 0.1965\n", - "8/8, train_loss: 0.0397 step time: 0.1948\n", - "epoch 24 average loss: 0.0471\n", - "time consuming of epoch 24 is: 1.6863\n", - "----------\n", - "epoch 25/600\n", - "1/8, train_loss: 0.0394 step time: 0.2559\n", - "2/8, train_loss: 0.0383 step time: 0.2071\n", - "3/8, train_loss: 0.0451 step time: 0.2075\n", - "4/8, train_loss: 0.0898 step time: 0.2069\n", - "5/8, train_loss: 0.0248 step time: 0.2073\n", - "6/8, train_loss: 0.0701 step time: 0.2074\n", - "7/8, train_loss: 0.0732 step time: 0.1955\n", - "8/8, train_loss: 0.0477 step time: 0.1956\n", - "epoch 25 average loss: 0.0536\n", - "saved new best metric model\n", - "current epoch: 25 current mean dice: 0.8969 best mean dice: 0.8969 at epoch: 25\n", - "time consuming of epoch 25 is: 2.7585\n", - "----------\n", - "epoch 26/600\n", - "1/8, train_loss: 0.0439 step time: 0.2651\n", - "2/8, train_loss: 0.0592 step time: 0.2083\n", - "3/8, train_loss: 0.0554 step time: 0.2070\n", - "4/8, train_loss: 0.0395 step time: 0.2078\n", - "5/8, train_loss: 0.0399 step time: 0.2075\n", - "6/8, train_loss: 0.0490 step time: 0.2103\n", - "7/8, train_loss: 0.0506 step time: 0.1965\n", - "8/8, train_loss: 0.0554 step time: 0.1956\n", - "epoch 26 average loss: 0.0491\n", - "time consuming of epoch 26 is: 1.7006\n", - "----------\n", - "epoch 27/600\n", - "1/8, train_loss: 0.0537 step time: 0.2538\n", - "2/8, train_loss: 0.1149 step time: 0.2075\n", - "3/8, train_loss: 0.0783 step time: 0.2071\n", - "4/8, train_loss: 0.0425 step time: 0.2069\n", - "5/8, train_loss: 0.0416 step time: 0.2094\n", - "6/8, train_loss: 0.0262 step time: 0.2069\n", - "7/8, train_loss: 0.0727 step time: 0.1963\n", - "8/8, train_loss: 0.0590 step time: 0.1955\n", - "epoch 27 average loss: 0.0611\n", - "time consuming of epoch 27 is: 1.6845\n", - "----------\n", - "epoch 28/600\n", - "1/8, train_loss: 0.0345 step time: 0.2569\n", - "2/8, train_loss: 0.0424 step time: 0.2083\n", - "3/8, train_loss: 0.0444 step time: 0.2083\n", - "4/8, train_loss: 0.0409 step time: 0.2071\n", - "5/8, train_loss: 0.0490 step time: 0.2081\n", - "6/8, train_loss: 0.0333 step time: 0.2083\n", - "7/8, train_loss: 0.0774 step time: 0.1968\n", - "8/8, train_loss: 0.0903 step time: 0.1956\n", - "epoch 28 average loss: 0.0515\n", - "time consuming of epoch 28 is: 1.6906\n", - "----------\n", - "epoch 29/600\n", - "1/8, train_loss: 0.0437 step time: 0.2563\n", - "2/8, train_loss: 0.0311 step time: 0.2085\n", - "3/8, train_loss: 0.0665 step time: 0.2088\n", - "4/8, train_loss: 0.0528 step time: 0.2080\n", - "5/8, train_loss: 0.0491 step time: 0.2069\n", - "6/8, train_loss: 0.0761 step time: 0.2091\n", - "7/8, train_loss: 0.0543 step time: 0.1958\n", - "8/8, train_loss: 0.0518 step time: 0.1956\n", - "epoch 29 average loss: 0.0532\n", - "time consuming of epoch 29 is: 1.6901\n", - "----------\n", - "epoch 30/600\n", - "1/8, train_loss: 0.0395 step time: 0.2567\n", - "2/8, train_loss: 0.0454 step time: 0.2082\n", - "3/8, train_loss: 0.0517 step time: 0.2072\n", - "4/8, train_loss: 0.0901 step time: 0.2079\n", - "5/8, train_loss: 0.0522 step time: 0.2079\n", - "6/8, train_loss: 0.0425 step time: 0.2095\n", - "7/8, train_loss: 0.0439 step time: 0.1958\n", - "8/8, train_loss: 0.0371 step time: 0.1955\n", - "epoch 30 average loss: 0.0503\n", - "current epoch: 30 current mean dice: 0.8890 best mean dice: 0.8969 at epoch: 25\n", - "time consuming of epoch 30 is: 2.4989\n", - "----------\n", - "epoch 31/600\n", - "1/8, train_loss: 0.0338 step time: 0.2536\n", - "2/8, train_loss: 0.0410 step time: 0.2070\n", - "3/8, train_loss: 0.0410 step time: 0.2071\n", - "4/8, train_loss: 0.0382 step time: 0.2077\n", - "5/8, train_loss: 0.0504 step time: 0.2082\n", - "6/8, train_loss: 0.0302 step time: 0.2079\n", - "7/8, train_loss: 0.0510 step time: 0.1963\n", - "8/8, train_loss: 0.0406 step time: 0.1955\n", - "epoch 31 average loss: 0.0408\n", - "time consuming of epoch 31 is: 1.6845\n", - "----------\n", - "epoch 32/600\n", - "1/8, train_loss: 0.0327 step time: 0.2470\n", - "2/8, train_loss: 0.0291 step time: 0.2068\n", - "3/8, train_loss: 0.1370 step time: 0.2069\n", - "4/8, train_loss: 0.0312 step time: 0.2073\n", - "5/8, train_loss: 0.0306 step time: 0.2081\n", - "6/8, train_loss: 0.0694 step time: 0.2079\n", - "7/8, train_loss: 0.0372 step time: 0.1965\n", - "8/8, train_loss: 0.0602 step time: 0.1956\n", - "epoch 32 average loss: 0.0534\n", - "time consuming of epoch 32 is: 1.6772\n", - "----------\n", - "epoch 33/600\n", - "1/8, train_loss: 0.0540 step time: 0.2468\n", - "2/8, train_loss: 0.1219 step time: 0.2080\n", - "3/8, train_loss: 0.0577 step time: 0.2091\n", - "4/8, train_loss: 0.0325 step time: 0.2073\n", - "5/8, train_loss: 0.0371 step time: 0.2088\n", - "6/8, train_loss: 0.0549 step time: 0.2078\n", - "7/8, train_loss: 0.0983 step time: 0.1960\n", - "8/8, train_loss: 0.0661 step time: 0.1947\n", - "epoch 33 average loss: 0.0653\n", - "time consuming of epoch 33 is: 1.6797\n", - "----------\n", - "epoch 34/600\n", - "1/8, train_loss: 0.0739 step time: 0.2551\n", - "2/8, train_loss: 0.0329 step time: 0.2077\n", - "3/8, train_loss: 0.0439 step time: 0.2073\n", - "4/8, train_loss: 0.0749 step time: 0.2068\n", - "5/8, train_loss: 0.0679 step time: 0.2073\n", - "6/8, train_loss: 0.0517 step time: 0.2075\n", - "7/8, train_loss: 0.0597 step time: 0.1961\n", - "8/8, train_loss: 0.0958 step time: 0.1943\n", - "epoch 34 average loss: 0.0626\n", - "time consuming of epoch 34 is: 1.6832\n", - "----------\n", - "epoch 35/600\n", - "1/8, train_loss: 0.0601 step time: 0.2554\n", - "2/8, train_loss: 0.0365 step time: 0.2076\n", - "3/8, train_loss: 0.0801 step time: 0.2073\n", - "4/8, train_loss: 0.0347 step time: 0.2090\n", - "5/8, train_loss: 0.0524 step time: 0.2083\n", - "6/8, train_loss: 0.0358 step time: 0.2071\n", - "7/8, train_loss: 0.1216 step time: 0.1961\n", - "8/8, train_loss: 0.0970 step time: 0.1956\n", - "epoch 35 average loss: 0.0648\n", - "current epoch: 35 current mean dice: 0.8884 best mean dice: 0.8969 at epoch: 25\n", - "time consuming of epoch 35 is: 2.4965\n", - "----------\n", - "epoch 36/600\n", - "1/8, train_loss: 0.0784 step time: 0.2549\n", - "2/8, train_loss: 0.0462 step time: 0.2097\n", - "3/8, train_loss: 0.0405 step time: 0.2075\n", - "4/8, train_loss: 0.0292 step time: 0.2071\n", - "5/8, train_loss: 0.0605 step time: 0.2092\n", - "6/8, train_loss: 0.0441 step time: 0.2088\n", - "7/8, train_loss: 0.0521 step time: 0.1961\n", - "8/8, train_loss: 0.0539 step time: 0.1954\n", - "epoch 36 average loss: 0.0506\n", - "time consuming of epoch 36 is: 1.6897\n", - "----------\n", - "epoch 37/600\n", - "1/8, train_loss: 0.0474 step time: 0.2541\n", - "2/8, train_loss: 0.0376 step time: 0.2079\n", - "3/8, train_loss: 0.0295 step time: 0.2084\n", - "4/8, train_loss: 0.0289 step time: 0.2081\n", - "5/8, train_loss: 0.0343 step time: 0.2100\n", - "6/8, train_loss: 0.0374 step time: 0.2082\n", - "7/8, train_loss: 0.0438 step time: 0.1961\n", - "8/8, train_loss: 0.0394 step time: 0.1956\n", - "epoch 37 average loss: 0.0373\n", - "time consuming of epoch 37 is: 1.6893\n", - "----------\n", - "epoch 38/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0459 step time: 0.2534\n", - "2/8, train_loss: 0.0381 step time: 0.2073\n", - "3/8, train_loss: 0.0447 step time: 0.2075\n", - "4/8, train_loss: 0.0310 step time: 0.2079\n", - "5/8, train_loss: 0.0437 step time: 0.2174\n", - "6/8, train_loss: 0.0325 step time: 0.2153\n", - "7/8, train_loss: 0.0828 step time: 0.1994\n", - "8/8, train_loss: 0.0459 step time: 0.2006\n", - "epoch 38 average loss: 0.0456\n", - "time consuming of epoch 38 is: 1.7099\n", - "----------\n", - "epoch 39/600\n", - "1/8, train_loss: 0.0330 step time: 0.2644\n", - "2/8, train_loss: 0.0325 step time: 0.2139\n", - "3/8, train_loss: 0.1914 step time: 0.2139\n", - "4/8, train_loss: 0.0379 step time: 0.2179\n", - "5/8, train_loss: 0.0535 step time: 0.2160\n", - "6/8, train_loss: 0.0579 step time: 0.2132\n", - "7/8, train_loss: 0.0368 step time: 0.1992\n", - "8/8, train_loss: 0.0444 step time: 0.1988\n", - "epoch 39 average loss: 0.0609\n", - "time consuming of epoch 39 is: 1.7388\n", - "----------\n", - "epoch 40/600\n", - "1/8, train_loss: 0.0358 step time: 0.2653\n", - "2/8, train_loss: 0.0395 step time: 0.2163\n", - "3/8, train_loss: 0.0406 step time: 0.2156\n", - "4/8, train_loss: 0.0566 step time: 0.2150\n", - "5/8, train_loss: 0.0485 step time: 0.2136\n", - "6/8, train_loss: 0.0495 step time: 0.2155\n", - "7/8, train_loss: 0.0652 step time: 0.2003\n", - "8/8, train_loss: 0.0606 step time: 0.1990\n", - "epoch 40 average loss: 0.0495\n", - "current epoch: 40 current mean dice: 0.8737 best mean dice: 0.8969 at epoch: 25\n", - "time consuming of epoch 40 is: 2.5544\n", - "----------\n", - "epoch 41/600\n", - "1/8, train_loss: 0.0453 step time: 0.2615\n", - "2/8, train_loss: 0.0459 step time: 0.2130\n", - "3/8, train_loss: 0.0305 step time: 0.2125\n", - "4/8, train_loss: 0.0292 step time: 0.2113\n", - "5/8, train_loss: 0.0505 step time: 0.2114\n", - "6/8, train_loss: 0.1944 step time: 0.2138\n", - "7/8, train_loss: 0.0352 step time: 0.1994\n", - "8/8, train_loss: 0.0756 step time: 0.2002\n", - "epoch 41 average loss: 0.0633\n", - "time consuming of epoch 41 is: 1.7243\n", - "----------\n", - "epoch 42/600\n", - "1/8, train_loss: 0.0470 step time: 0.2634\n", - "2/8, train_loss: 0.0408 step time: 0.2150\n", - "3/8, train_loss: 0.0403 step time: 0.2150\n", - "4/8, train_loss: 0.0490 step time: 0.2133\n", - "5/8, train_loss: 0.0520 step time: 0.2117\n", - "6/8, train_loss: 0.0816 step time: 0.2123\n", - "7/8, train_loss: 0.0631 step time: 0.1988\n", - "8/8, train_loss: 0.0490 step time: 0.2003\n", - "epoch 42 average loss: 0.0529\n", - "time consuming of epoch 42 is: 1.7313\n", - "----------\n", - "epoch 43/600\n", - "1/8, train_loss: 0.0699 step time: 0.2516\n", - "2/8, train_loss: 0.0403 step time: 0.2117\n", - "3/8, train_loss: 0.0285 step time: 0.2111\n", - "4/8, train_loss: 0.0575 step time: 0.2104\n", - "5/8, train_loss: 0.0381 step time: 0.2106\n", - "6/8, train_loss: 0.0826 step time: 0.2148\n", - "7/8, train_loss: 0.0470 step time: 0.1998\n", - "8/8, train_loss: 0.0569 step time: 0.1999\n", - "epoch 43 average loss: 0.0526\n", - "time consuming of epoch 43 is: 1.7114\n", - "----------\n", - "epoch 44/600\n", - "1/8, train_loss: 0.0431 step time: 0.2555\n", - "2/8, train_loss: 0.0348 step time: 0.2139\n", - "3/8, train_loss: 0.0449 step time: 0.2146\n", - "4/8, train_loss: 0.0440 step time: 0.2175\n", - "5/8, train_loss: 0.0260 step time: 0.2131\n", - "6/8, train_loss: 0.0405 step time: 0.2142\n", - "7/8, train_loss: 0.0707 step time: 0.1990\n", - "8/8, train_loss: 0.0481 step time: 0.1990\n", - "epoch 44 average loss: 0.0440\n", - "time consuming of epoch 44 is: 1.7280\n", - "----------\n", - "epoch 45/600\n", - "1/8, train_loss: 0.0600 step time: 0.2597\n", - "2/8, train_loss: 0.0377 step time: 0.2154\n", - "3/8, train_loss: 0.0344 step time: 0.2147\n", - "4/8, train_loss: 0.0456 step time: 0.2144\n", - "5/8, train_loss: 0.0550 step time: 0.2152\n", - "6/8, train_loss: 0.0331 step time: 0.2147\n", - "7/8, train_loss: 0.0502 step time: 0.1995\n", - "8/8, train_loss: 0.0285 step time: 0.2007\n", - "epoch 45 average loss: 0.0431\n", - "current epoch: 45 current mean dice: 0.7649 best mean dice: 0.8969 at epoch: 25\n", - "time consuming of epoch 45 is: 2.5463\n", - "----------\n", - "epoch 46/600\n", - "1/8, train_loss: 0.0279 step time: 0.2610\n", - "2/8, train_loss: 0.0401 step time: 0.2126\n", - "3/8, train_loss: 0.0624 step time: 0.2137\n", - "4/8, train_loss: 0.0394 step time: 0.2139\n", - "5/8, train_loss: 0.0491 step time: 0.2124\n", - "6/8, train_loss: 0.0262 step time: 0.2168\n", - "7/8, train_loss: 0.0287 step time: 0.2000\n", - "8/8, train_loss: 0.0410 step time: 0.1991\n", - "epoch 46 average loss: 0.0394\n", - "time consuming of epoch 46 is: 1.7306\n", - "----------\n", - "epoch 47/600\n", - "1/8, train_loss: 0.0326 step time: 0.2660\n", - "2/8, train_loss: 0.0317 step time: 0.2143\n", - "3/8, train_loss: 0.0365 step time: 0.2157\n", - "4/8, train_loss: 0.0346 step time: 0.2152\n", - "5/8, train_loss: 0.0522 step time: 0.2160\n", - "6/8, train_loss: 0.0299 step time: 0.2157\n", - "7/8, train_loss: 0.0280 step time: 0.1992\n", - "8/8, train_loss: 0.0358 step time: 0.2000\n", - "epoch 47 average loss: 0.0352\n", - "time consuming of epoch 47 is: 1.7437\n", - "----------\n", - "epoch 48/600\n", - "1/8, train_loss: 0.0249 step time: 0.2627\n", - "2/8, train_loss: 0.0335 step time: 0.2143\n", - "3/8, train_loss: 0.0271 step time: 0.2151\n", - "4/8, train_loss: 0.0370 step time: 0.2155\n", - "5/8, train_loss: 0.0523 step time: 0.2152\n", - "6/8, train_loss: 0.0408 step time: 0.2155\n", - "7/8, train_loss: 0.0279 step time: 0.1995\n", - "8/8, train_loss: 0.0394 step time: 0.2000\n", - "epoch 48 average loss: 0.0354\n", - "time consuming of epoch 48 is: 1.7393\n", - "----------\n", - "epoch 49/600\n", - "1/8, train_loss: 0.0296 step time: 0.2613\n", - "2/8, train_loss: 0.0325 step time: 0.2138\n", - "3/8, train_loss: 0.0291 step time: 0.2156\n", - "4/8, train_loss: 0.0398 step time: 0.2145\n", - "5/8, train_loss: 0.0238 step time: 0.2163\n", - "6/8, train_loss: 0.0266 step time: 0.2149\n", - "7/8, train_loss: 0.0307 step time: 0.1994\n", - "8/8, train_loss: 0.0251 step time: 0.1996\n", - "epoch 49 average loss: 0.0297\n", - "time consuming of epoch 49 is: 1.7368\n", - "----------\n", - "epoch 50/600\n", - "1/8, train_loss: 0.0264 step time: 0.2658\n", - "2/8, train_loss: 0.0286 step time: 0.2118\n", - "3/8, train_loss: 0.0388 step time: 0.2131\n", - "4/8, train_loss: 0.0331 step time: 0.2138\n", - "5/8, train_loss: 0.0249 step time: 0.2128\n", - "6/8, train_loss: 0.0205 step time: 0.2092\n", - "7/8, train_loss: 0.0279 step time: 0.1995\n", - "8/8, train_loss: 0.0259 step time: 0.1998\n", - "epoch 50 average loss: 0.0283\n", - "saved new best metric model\n", - "current epoch: 50 current mean dice: 0.9166 best mean dice: 0.9166 at epoch: 50\n", - "time consuming of epoch 50 is: 2.7850\n", - "----------\n", - "epoch 51/600\n", - "1/8, train_loss: 0.0233 step time: 0.2631\n", - "2/8, train_loss: 0.0252 step time: 0.2129\n", - "3/8, train_loss: 0.0534 step time: 0.2108\n", - "4/8, train_loss: 0.0390 step time: 0.2091\n", - "5/8, train_loss: 0.0315 step time: 0.2099\n", - "6/8, train_loss: 0.0293 step time: 0.2118\n", - "7/8, train_loss: 0.0325 step time: 0.1992\n", - "8/8, train_loss: 0.0299 step time: 0.1991\n", - "epoch 51 average loss: 0.0330\n", - "time consuming of epoch 51 is: 1.7183\n", - "----------\n", - "epoch 52/600\n", - "1/8, train_loss: 0.0258 step time: 0.2639\n", - "2/8, train_loss: 0.0286 step time: 0.2146\n", - "3/8, train_loss: 0.0314 step time: 0.2149\n", - "4/8, train_loss: 0.0234 step time: 0.2148\n", - "5/8, train_loss: 0.0277 step time: 0.2153\n", - "6/8, train_loss: 0.0254 step time: 0.2144\n", - "7/8, train_loss: 0.0215 step time: 0.2002\n", - "8/8, train_loss: 0.0252 step time: 0.1999\n", - "epoch 52 average loss: 0.0261\n", - "time consuming of epoch 52 is: 1.7396\n", - "----------\n", - "epoch 53/600\n", - "1/8, train_loss: 0.0266 step time: 0.2635\n", - "2/8, train_loss: 0.0226 step time: 0.2135\n", - "3/8, train_loss: 0.0399 step time: 0.2167\n", - "4/8, train_loss: 0.0304 step time: 0.2143\n", - "5/8, train_loss: 0.0394 step time: 0.2148\n", - "6/8, train_loss: 0.0293 step time: 0.2162\n", - "7/8, train_loss: 0.0522 step time: 0.2001\n", - "8/8, train_loss: 0.0267 step time: 0.1995\n", - "epoch 53 average loss: 0.0334\n", - "time consuming of epoch 53 is: 1.7400\n", - "----------\n", - "epoch 54/600\n", - "1/8, train_loss: 0.0273 step time: 0.2665\n", - "2/8, train_loss: 0.0352 step time: 0.2130\n", - "3/8, train_loss: 0.0361 step time: 0.2162\n", - "4/8, train_loss: 0.0304 step time: 0.2162\n", - "5/8, train_loss: 0.0294 step time: 0.2153\n", - "6/8, train_loss: 0.0244 step time: 0.2143\n", - "7/8, train_loss: 0.0247 step time: 0.1995\n", - "8/8, train_loss: 0.0346 step time: 0.2002\n", - "epoch 54 average loss: 0.0303\n", - "time consuming of epoch 54 is: 1.7427\n", - "----------\n", - "epoch 55/600\n", - "1/8, train_loss: 0.0260 step time: 0.2631\n", - "2/8, train_loss: 0.0319 step time: 0.2139\n", - "3/8, train_loss: 0.0277 step time: 0.2184\n", - "4/8, train_loss: 0.0241 step time: 0.2138\n", - "5/8, train_loss: 0.0284 step time: 0.2145\n", - "6/8, train_loss: 0.0248 step time: 0.2152\n", - "7/8, train_loss: 0.0329 step time: 0.2002\n", - "8/8, train_loss: 0.0232 step time: 0.2015\n", - "epoch 55 average loss: 0.0274\n", - "current epoch: 55 current mean dice: 0.8907 best mean dice: 0.9166 at epoch: 50\n", - "time consuming of epoch 55 is: 2.5562\n", - "----------\n", - "epoch 56/600\n", - "1/8, train_loss: 0.0286 step time: 0.2587\n", - "2/8, train_loss: 0.0349 step time: 0.2125\n", - "3/8, train_loss: 0.0282 step time: 0.2091\n", - "4/8, train_loss: 0.0329 step time: 0.2093\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "5/8, train_loss: 0.0265 step time: 0.2106\n", - "6/8, train_loss: 0.0286 step time: 0.2101\n", - "7/8, train_loss: 0.0286 step time: 0.1974\n", - "8/8, train_loss: 0.0239 step time: 0.1990\n", - "epoch 56 average loss: 0.0290\n", - "time consuming of epoch 56 is: 1.7080\n", - "----------\n", - "epoch 57/600\n", - "1/8, train_loss: 0.0247 step time: 0.2537\n", - "2/8, train_loss: 0.0245 step time: 0.2130\n", - "3/8, train_loss: 0.0266 step time: 0.2115\n", - "4/8, train_loss: 0.0267 step time: 0.2122\n", - "5/8, train_loss: 0.0326 step time: 0.2122\n", - "6/8, train_loss: 0.0305 step time: 0.2122\n", - "7/8, train_loss: 0.0281 step time: 0.2015\n", - "8/8, train_loss: 0.0291 step time: 0.1998\n", - "epoch 57 average loss: 0.0278\n", - "time consuming of epoch 57 is: 1.7178\n", - "----------\n", - "epoch 58/600\n", - "1/8, train_loss: 0.0268 step time: 0.2641\n", - "2/8, train_loss: 0.0225 step time: 0.2155\n", - "3/8, train_loss: 0.0350 step time: 0.2116\n", - "4/8, train_loss: 0.0246 step time: 0.2119\n", - "5/8, train_loss: 0.0280 step time: 0.2137\n", - "6/8, train_loss: 0.0243 step time: 0.2139\n", - "7/8, train_loss: 0.0209 step time: 0.1998\n", - "8/8, train_loss: 0.0230 step time: 0.1997\n", - "epoch 58 average loss: 0.0256\n", - "time consuming of epoch 58 is: 1.7315\n", - "----------\n", - "epoch 59/600\n", - "1/8, train_loss: 0.0223 step time: 0.2535\n", - "2/8, train_loss: 0.0225 step time: 0.2142\n", - "3/8, train_loss: 0.0250 step time: 0.2131\n", - "4/8, train_loss: 0.0336 step time: 0.2130\n", - "5/8, train_loss: 0.0186 step time: 0.2136\n", - "6/8, train_loss: 0.0263 step time: 0.2143\n", - "7/8, train_loss: 0.0198 step time: 0.2004\n", - "8/8, train_loss: 0.0233 step time: 0.2000\n", - "epoch 59 average loss: 0.0239\n", - "time consuming of epoch 59 is: 1.7236\n", - "----------\n", - "epoch 60/600\n", - "1/8, train_loss: 0.0260 step time: 0.2645\n", - "2/8, train_loss: 0.0206 step time: 0.2167\n", - "3/8, train_loss: 0.0321 step time: 0.2155\n", - "4/8, train_loss: 0.0288 step time: 0.2156\n", - "5/8, train_loss: 0.0210 step time: 0.2147\n", - "6/8, train_loss: 0.0257 step time: 0.2144\n", - "7/8, train_loss: 0.0254 step time: 0.1994\n", - "8/8, train_loss: 0.0263 step time: 0.1994\n", - "epoch 60 average loss: 0.0257\n", - "saved new best metric model\n", - "current epoch: 60 current mean dice: 0.9340 best mean dice: 0.9340 at epoch: 60\n", - "time consuming of epoch 60 is: 2.8237\n", - "----------\n", - "epoch 61/600\n", - "1/8, train_loss: 0.0231 step time: 0.2700\n", - "2/8, train_loss: 0.0236 step time: 0.2117\n", - "3/8, train_loss: 0.0218 step time: 0.2103\n", - "4/8, train_loss: 0.0253 step time: 0.2102\n", - "5/8, train_loss: 0.0215 step time: 0.2099\n", - "6/8, train_loss: 0.0256 step time: 0.2100\n", - "7/8, train_loss: 0.0253 step time: 0.2000\n", - "8/8, train_loss: 0.0198 step time: 0.2001\n", - "epoch 61 average loss: 0.0233\n", - "time consuming of epoch 61 is: 1.7247\n", - "----------\n", - "epoch 62/600\n", - "1/8, train_loss: 0.0309 step time: 0.2551\n", - "2/8, train_loss: 0.0211 step time: 0.2152\n", - "3/8, train_loss: 0.0231 step time: 0.2132\n", - "4/8, train_loss: 0.0230 step time: 0.2111\n", - "5/8, train_loss: 0.0279 step time: 0.2107\n", - "6/8, train_loss: 0.0284 step time: 0.2136\n", - "7/8, train_loss: 0.0229 step time: 0.1996\n", - "8/8, train_loss: 0.0249 step time: 0.1996\n", - "epoch 62 average loss: 0.0253\n", - "time consuming of epoch 62 is: 1.7195\n", - "----------\n", - "epoch 63/600\n", - "1/8, train_loss: 0.0260 step time: 0.2641\n", - "2/8, train_loss: 0.0300 step time: 0.2141\n", - "3/8, train_loss: 0.0219 step time: 0.2145\n", - "4/8, train_loss: 0.0306 step time: 0.2156\n", - "5/8, train_loss: 0.0251 step time: 0.2157\n", - "6/8, train_loss: 0.0228 step time: 0.2145\n", - "7/8, train_loss: 0.0319 step time: 0.2003\n", - "8/8, train_loss: 0.0199 step time: 0.2000\n", - "epoch 63 average loss: 0.0260\n", - "time consuming of epoch 63 is: 1.7401\n", - "----------\n", - "epoch 64/600\n", - "1/8, train_loss: 0.0230 step time: 0.2620\n", - "2/8, train_loss: 0.0186 step time: 0.2145\n", - "3/8, train_loss: 0.0237 step time: 0.2137\n", - "4/8, train_loss: 0.0239 step time: 0.2137\n", - "5/8, train_loss: 0.0239 step time: 0.2150\n", - "6/8, train_loss: 0.0222 step time: 0.2168\n", - "7/8, train_loss: 0.0272 step time: 0.1988\n", - "8/8, train_loss: 0.0274 step time: 0.2002\n", - "epoch 64 average loss: 0.0237\n", - "time consuming of epoch 64 is: 1.7362\n", - "----------\n", - "epoch 65/600\n", - "1/8, train_loss: 0.0212 step time: 0.2633\n", - "2/8, train_loss: 0.0229 step time: 0.2160\n", - "3/8, train_loss: 0.0218 step time: 0.2163\n", - "4/8, train_loss: 0.0261 step time: 0.2144\n", - "5/8, train_loss: 0.0350 step time: 0.2135\n", - "6/8, train_loss: 0.0251 step time: 0.2160\n", - "7/8, train_loss: 0.0207 step time: 0.1995\n", - "8/8, train_loss: 0.0186 step time: 0.1999\n", - "epoch 65 average loss: 0.0239\n", - "saved new best metric model\n", - "current epoch: 65 current mean dice: 0.9441 best mean dice: 0.9441 at epoch: 65\n", - "time consuming of epoch 65 is: 2.8164\n", - "----------\n", - "epoch 66/600\n", - "1/8, train_loss: 0.0229 step time: 0.2710\n", - "2/8, train_loss: 0.0252 step time: 0.2139\n", - "3/8, train_loss: 0.0238 step time: 0.2132\n", - "4/8, train_loss: 0.0202 step time: 0.2150\n", - "5/8, train_loss: 0.0270 step time: 0.2163\n", - "6/8, train_loss: 0.0259 step time: 0.2157\n", - "7/8, train_loss: 0.0254 step time: 0.2001\n", - "8/8, train_loss: 0.0247 step time: 0.2000\n", - "epoch 66 average loss: 0.0244\n", - "time consuming of epoch 66 is: 1.7476\n", - "----------\n", - "epoch 67/600\n", - "1/8, train_loss: 0.0306 step time: 0.2618\n", - "2/8, train_loss: 0.0243 step time: 0.2172\n", - "3/8, train_loss: 0.0256 step time: 0.2141\n", - "4/8, train_loss: 0.0245 step time: 0.2139\n", - "5/8, train_loss: 0.0223 step time: 0.2141\n", - "6/8, train_loss: 0.0202 step time: 0.2179\n", - "7/8, train_loss: 0.0213 step time: 0.1997\n", - "8/8, train_loss: 0.0214 step time: 0.2006\n", - "epoch 67 average loss: 0.0238\n", - "time consuming of epoch 67 is: 1.7406\n", - "----------\n", - "epoch 68/600\n", - "1/8, train_loss: 0.0344 step time: 0.2626\n", - "2/8, train_loss: 0.0276 step time: 0.2143\n", - "3/8, train_loss: 0.0228 step time: 0.2165\n", - "4/8, train_loss: 0.0265 step time: 0.2150\n", - "5/8, train_loss: 0.0231 step time: 0.2144\n", - "6/8, train_loss: 0.0216 step time: 0.2150\n", - "7/8, train_loss: 0.0171 step time: 0.2006\n", - "8/8, train_loss: 0.0206 step time: 0.1996\n", - "epoch 68 average loss: 0.0242\n", - "time consuming of epoch 68 is: 1.7395\n", - "----------\n", - "epoch 69/600\n", - "1/8, train_loss: 0.0241 step time: 0.2634\n", - "2/8, train_loss: 0.0240 step time: 0.2152\n", - "3/8, train_loss: 0.0196 step time: 0.2108\n", - "4/8, train_loss: 0.0308 step time: 0.2142\n", - "5/8, train_loss: 0.0244 step time: 0.2136\n", - "6/8, train_loss: 0.0219 step time: 0.2118\n", - "7/8, train_loss: 0.0187 step time: 0.2003\n", - "8/8, train_loss: 0.0226 step time: 0.2003\n", - "epoch 69 average loss: 0.0233\n", - "time consuming of epoch 69 is: 1.7313\n", - "----------\n", - "epoch 70/600\n", - "1/8, train_loss: 0.0220 step time: 0.2519\n", - "2/8, train_loss: 0.0245 step time: 0.2118\n", - "3/8, train_loss: 0.0237 step time: 0.2124\n", - "4/8, train_loss: 0.0215 step time: 0.2133\n", - "5/8, train_loss: 0.0228 step time: 0.2136\n", - "6/8, train_loss: 0.0180 step time: 0.2129\n", - "7/8, train_loss: 0.0273 step time: 0.1992\n", - "8/8, train_loss: 0.0235 step time: 0.2002\n", - "epoch 70 average loss: 0.0229\n", - "current epoch: 70 current mean dice: 0.9428 best mean dice: 0.9441 at epoch: 65\n", - "time consuming of epoch 70 is: 2.5292\n", - "----------\n", - "epoch 71/600\n", - "1/8, train_loss: 0.0241 step time: 0.2597\n", - "2/8, train_loss: 0.0210 step time: 0.2105\n", - "3/8, train_loss: 0.0268 step time: 0.2129\n", - "4/8, train_loss: 0.0257 step time: 0.2114\n", - "5/8, train_loss: 0.0234 step time: 0.2124\n", - "6/8, train_loss: 0.0260 step time: 0.2175\n", - "7/8, train_loss: 0.0198 step time: 0.1998\n", - "8/8, train_loss: 0.0336 step time: 0.2009\n", - "epoch 71 average loss: 0.0251\n", - "time consuming of epoch 71 is: 1.7262\n", - "----------\n", - "epoch 72/600\n", - "1/8, train_loss: 0.0223 step time: 0.2650\n", - "2/8, train_loss: 0.0294 step time: 0.2163\n", - "3/8, train_loss: 0.0219 step time: 0.2150\n", - "4/8, train_loss: 0.0195 step time: 0.2133\n", - "5/8, train_loss: 0.0244 step time: 0.2151\n", - "6/8, train_loss: 0.0209 step time: 0.2150\n", - "7/8, train_loss: 0.0294 step time: 0.1997\n", - "8/8, train_loss: 0.0273 step time: 0.2005\n", - "epoch 72 average loss: 0.0244\n", - "time consuming of epoch 72 is: 1.7413\n", - "----------\n", - "epoch 73/600\n", - "1/8, train_loss: 0.0266 step time: 0.2641\n", - "2/8, train_loss: 0.0209 step time: 0.2138\n", - "3/8, train_loss: 0.0210 step time: 0.2138\n", - "4/8, train_loss: 0.0233 step time: 0.2130\n", - "5/8, train_loss: 0.0212 step time: 0.2125\n", - "6/8, train_loss: 0.0252 step time: 0.2139\n", - "7/8, train_loss: 0.0219 step time: 0.1994\n", - "8/8, train_loss: 0.0212 step time: 0.2005\n", - "epoch 73 average loss: 0.0227\n", - "time consuming of epoch 73 is: 1.7325\n", - "----------\n", - "epoch 74/600\n", - "1/8, train_loss: 0.0249 step time: 0.2539\n", - "2/8, train_loss: 0.0219 step time: 0.2138\n", - "3/8, train_loss: 0.0246 step time: 0.2109\n", - "4/8, train_loss: 0.0348 step time: 0.2120\n", - "5/8, train_loss: 0.0210 step time: 0.2121\n", - "6/8, train_loss: 0.0284 step time: 0.2113\n", - "7/8, train_loss: 0.0210 step time: 0.1997\n", - "8/8, train_loss: 0.0221 step time: 0.2006\n", - "epoch 74 average loss: 0.0249\n", - "time consuming of epoch 74 is: 1.7158\n", - "----------\n", - "epoch 75/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0263 step time: 0.2637\n", - "2/8, train_loss: 0.0225 step time: 0.2122\n", - "3/8, train_loss: 0.0216 step time: 0.2104\n", - "4/8, train_loss: 0.0209 step time: 0.2126\n", - "5/8, train_loss: 0.0223 step time: 0.2172\n", - "6/8, train_loss: 0.0213 step time: 0.2157\n", - "7/8, train_loss: 0.0204 step time: 0.1993\n", - "8/8, train_loss: 0.0226 step time: 0.1998\n", - "epoch 75 average loss: 0.0222\n", - "current epoch: 75 current mean dice: 0.9405 best mean dice: 0.9441 at epoch: 65\n", - "time consuming of epoch 75 is: 2.5434\n", - "----------\n", - "epoch 76/600\n", - "1/8, train_loss: 0.0244 step time: 0.2627\n", - "2/8, train_loss: 0.0240 step time: 0.2120\n", - "3/8, train_loss: 0.0239 step time: 0.2123\n", - "4/8, train_loss: 0.0245 step time: 0.2118\n", - "5/8, train_loss: 0.0350 step time: 0.2132\n", - "6/8, train_loss: 0.0216 step time: 0.2146\n", - "7/8, train_loss: 0.0302 step time: 0.1996\n", - "8/8, train_loss: 0.0239 step time: 0.2002\n", - "epoch 76 average loss: 0.0259\n", - "time consuming of epoch 76 is: 1.7276\n", - "----------\n", - "epoch 77/600\n", - "1/8, train_loss: 0.0213 step time: 0.2631\n", - "2/8, train_loss: 0.0239 step time: 0.2147\n", - "3/8, train_loss: 0.0262 step time: 0.2152\n", - "4/8, train_loss: 0.0238 step time: 0.2149\n", - "5/8, train_loss: 0.0223 step time: 0.2146\n", - "6/8, train_loss: 0.0209 step time: 0.2172\n", - "7/8, train_loss: 0.0242 step time: 0.1995\n", - "8/8, train_loss: 0.0226 step time: 0.2000\n", - "epoch 77 average loss: 0.0231\n", - "time consuming of epoch 77 is: 1.7408\n", - "----------\n", - "epoch 78/600\n", - "1/8, train_loss: 0.0219 step time: 0.2660\n", - "2/8, train_loss: 0.0231 step time: 0.2157\n", - "3/8, train_loss: 0.0206 step time: 0.2151\n", - "4/8, train_loss: 0.0336 step time: 0.2141\n", - "5/8, train_loss: 0.0209 step time: 0.2141\n", - "6/8, train_loss: 0.0341 step time: 0.2139\n", - "7/8, train_loss: 0.0215 step time: 0.1995\n", - "8/8, train_loss: 0.0205 step time: 0.2000\n", - "epoch 78 average loss: 0.0245\n", - "time consuming of epoch 78 is: 1.7397\n", - "----------\n", - "epoch 79/600\n", - "1/8, train_loss: 0.0223 step time: 0.2637\n", - "2/8, train_loss: 0.0211 step time: 0.2143\n", - "3/8, train_loss: 0.0186 step time: 0.2136\n", - "4/8, train_loss: 0.0274 step time: 0.2168\n", - "5/8, train_loss: 0.0225 step time: 0.2158\n", - "6/8, train_loss: 0.0278 step time: 0.2155\n", - "7/8, train_loss: 0.0265 step time: 0.1990\n", - "8/8, train_loss: 0.0178 step time: 0.1999\n", - "epoch 79 average loss: 0.0230\n", - "time consuming of epoch 79 is: 1.7401\n", - "----------\n", - "epoch 80/600\n", - "1/8, train_loss: 0.0207 step time: 0.2629\n", - "2/8, train_loss: 0.0219 step time: 0.2165\n", - "3/8, train_loss: 0.0223 step time: 0.2153\n", - "4/8, train_loss: 0.0204 step time: 0.2147\n", - "5/8, train_loss: 0.0242 step time: 0.2150\n", - "6/8, train_loss: 0.0271 step time: 0.2180\n", - "7/8, train_loss: 0.0237 step time: 0.1994\n", - "8/8, train_loss: 0.0229 step time: 0.1998\n", - "epoch 80 average loss: 0.0229\n", - "current epoch: 80 current mean dice: 0.9250 best mean dice: 0.9441 at epoch: 65\n", - "time consuming of epoch 80 is: 2.5564\n", - "----------\n", - "epoch 81/600\n", - "1/8, train_loss: 0.0247 step time: 0.2582\n", - "2/8, train_loss: 0.0294 step time: 0.2130\n", - "3/8, train_loss: 0.0269 step time: 0.2147\n", - "4/8, train_loss: 0.0254 step time: 0.2126\n", - "5/8, train_loss: 0.0182 step time: 0.2139\n", - "6/8, train_loss: 0.0227 step time: 0.2180\n", - "7/8, train_loss: 0.0178 step time: 0.1994\n", - "8/8, train_loss: 0.0295 step time: 0.2014\n", - "epoch 81 average loss: 0.0243\n", - "time consuming of epoch 81 is: 1.7325\n", - "----------\n", - "epoch 82/600\n", - "1/8, train_loss: 0.0277 step time: 0.2627\n", - "2/8, train_loss: 0.0201 step time: 0.2169\n", - "3/8, train_loss: 0.0207 step time: 0.2155\n", - "4/8, train_loss: 0.0211 step time: 0.2182\n", - "5/8, train_loss: 0.0254 step time: 0.2138\n", - "6/8, train_loss: 0.0185 step time: 0.2151\n", - "7/8, train_loss: 0.0179 step time: 0.1993\n", - "8/8, train_loss: 0.0251 step time: 0.1981\n", - "epoch 82 average loss: 0.0221\n", - "time consuming of epoch 82 is: 1.7411\n", - "----------\n", - "epoch 83/600\n", - "1/8, train_loss: 0.0221 step time: 0.2650\n", - "2/8, train_loss: 0.0225 step time: 0.2130\n", - "3/8, train_loss: 0.0208 step time: 0.2152\n", - "4/8, train_loss: 0.0240 step time: 0.2155\n", - "5/8, train_loss: 0.0203 step time: 0.2165\n", - "6/8, train_loss: 0.0196 step time: 0.2176\n", - "7/8, train_loss: 0.0183 step time: 0.1999\n", - "8/8, train_loss: 0.0195 step time: 0.2011\n", - "epoch 83 average loss: 0.0209\n", - "time consuming of epoch 83 is: 1.7456\n", - "----------\n", - "epoch 84/600\n", - "1/8, train_loss: 0.0205 step time: 0.2642\n", - "2/8, train_loss: 0.0234 step time: 0.2147\n", - "3/8, train_loss: 0.0188 step time: 0.2159\n", - "4/8, train_loss: 0.0216 step time: 0.2175\n", - "5/8, train_loss: 0.0210 step time: 0.2155\n", - "6/8, train_loss: 0.0211 step time: 0.2153\n", - "7/8, train_loss: 0.0207 step time: 0.1993\n", - "8/8, train_loss: 0.0242 step time: 0.1996\n", - "epoch 84 average loss: 0.0214\n", - "time consuming of epoch 84 is: 1.7435\n", - "----------\n", - "epoch 85/600\n", - "1/8, train_loss: 0.0187 step time: 0.2659\n", - "2/8, train_loss: 0.0222 step time: 0.2143\n", - "3/8, train_loss: 0.0252 step time: 0.2167\n", - "4/8, train_loss: 0.0215 step time: 0.2142\n", - "5/8, train_loss: 0.0249 step time: 0.2142\n", - "6/8, train_loss: 0.0362 step time: 0.2178\n", - "7/8, train_loss: 0.0182 step time: 0.2005\n", - "8/8, train_loss: 0.0219 step time: 0.2006\n", - "epoch 85 average loss: 0.0236\n", - "saved new best metric model\n", - "current epoch: 85 current mean dice: 0.9482 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 85 is: 2.8376\n", - "----------\n", - "epoch 86/600\n", - "1/8, train_loss: 0.0202 step time: 0.2659\n", - "2/8, train_loss: 0.0165 step time: 0.2085\n", - "3/8, train_loss: 0.0258 step time: 0.2103\n", - "4/8, train_loss: 0.0192 step time: 0.2095\n", - "5/8, train_loss: 0.0221 step time: 0.2104\n", - "6/8, train_loss: 0.0199 step time: 0.2119\n", - "7/8, train_loss: 0.0185 step time: 0.2012\n", - "8/8, train_loss: 0.0274 step time: 0.2001\n", - "epoch 86 average loss: 0.0212\n", - "time consuming of epoch 86 is: 1.7204\n", - "----------\n", - "epoch 87/600\n", - "1/8, train_loss: 0.0320 step time: 0.2659\n", - "2/8, train_loss: 0.0207 step time: 0.2158\n", - "3/8, train_loss: 0.0216 step time: 0.2172\n", - "4/8, train_loss: 0.0200 step time: 0.2153\n", - "5/8, train_loss: 0.0255 step time: 0.2151\n", - "6/8, train_loss: 0.0215 step time: 0.2174\n", - "7/8, train_loss: 0.0228 step time: 0.1996\n", - "8/8, train_loss: 0.0189 step time: 0.1993\n", - "epoch 87 average loss: 0.0229\n", - "time consuming of epoch 87 is: 1.7470\n", - "----------\n", - "epoch 88/600\n", - "1/8, train_loss: 0.0213 step time: 0.2663\n", - "2/8, train_loss: 0.0215 step time: 0.2165\n", - "3/8, train_loss: 0.0184 step time: 0.2113\n", - "4/8, train_loss: 0.0249 step time: 0.2116\n", - "5/8, train_loss: 0.0192 step time: 0.2108\n", - "6/8, train_loss: 0.0292 step time: 0.2094\n", - "7/8, train_loss: 0.0183 step time: 0.1981\n", - "8/8, train_loss: 0.0235 step time: 0.1988\n", - "epoch 88 average loss: 0.0220\n", - "time consuming of epoch 88 is: 1.7243\n", - "----------\n", - "epoch 89/600\n", - "1/8, train_loss: 0.0199 step time: 0.2630\n", - "2/8, train_loss: 0.0237 step time: 0.2133\n", - "3/8, train_loss: 0.0219 step time: 0.2140\n", - "4/8, train_loss: 0.0259 step time: 0.2161\n", - "5/8, train_loss: 0.0306 step time: 0.2153\n", - "6/8, train_loss: 0.0219 step time: 0.2138\n", - "7/8, train_loss: 0.0214 step time: 0.1993\n", - "8/8, train_loss: 0.0251 step time: 0.1987\n", - "epoch 89 average loss: 0.0238\n", - "time consuming of epoch 89 is: 1.7349\n", - "----------\n", - "epoch 90/600\n", - "1/8, train_loss: 0.0178 step time: 0.2633\n", - "2/8, train_loss: 0.0203 step time: 0.2144\n", - "3/8, train_loss: 0.0187 step time: 0.2149\n", - "4/8, train_loss: 0.0280 step time: 0.2154\n", - "5/8, train_loss: 0.0238 step time: 0.2150\n", - "6/8, train_loss: 0.0319 step time: 0.2163\n", - "7/8, train_loss: 0.0220 step time: 0.2003\n", - "8/8, train_loss: 0.0233 step time: 0.2000\n", - "epoch 90 average loss: 0.0232\n", - "current epoch: 90 current mean dice: 0.8999 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 90 is: 2.5539\n", - "----------\n", - "epoch 91/600\n", - "1/8, train_loss: 0.0173 step time: 0.2606\n", - "2/8, train_loss: 0.0312 step time: 0.2131\n", - "3/8, train_loss: 0.0205 step time: 0.2120\n", - "4/8, train_loss: 0.0229 step time: 0.2136\n", - "5/8, train_loss: 0.0366 step time: 0.2129\n", - "6/8, train_loss: 0.0243 step time: 0.2119\n", - "7/8, train_loss: 0.0223 step time: 0.1997\n", - "8/8, train_loss: 0.0689 step time: 0.2008\n", - "epoch 91 average loss: 0.0305\n", - "time consuming of epoch 91 is: 1.7257\n", - "----------\n", - "epoch 92/600\n", - "1/8, train_loss: 0.0223 step time: 0.2624\n", - "2/8, train_loss: 0.0247 step time: 0.2162\n", - "3/8, train_loss: 0.0181 step time: 0.2154\n", - "4/8, train_loss: 0.0256 step time: 0.2150\n", - "5/8, train_loss: 0.0712 step time: 0.2141\n", - "6/8, train_loss: 0.0249 step time: 0.2162\n", - "7/8, train_loss: 0.0207 step time: 0.1999\n", - "8/8, train_loss: 0.0309 step time: 0.1998\n", - "epoch 92 average loss: 0.0298\n", - "time consuming of epoch 92 is: 1.7405\n", - "----------\n", - "epoch 93/600\n", - "1/8, train_loss: 0.0741 step time: 0.2632\n", - "2/8, train_loss: 0.0662 step time: 0.2137\n", - "3/8, train_loss: 0.0379 step time: 0.2155\n", - "4/8, train_loss: 0.0371 step time: 0.2159\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "5/8, train_loss: 0.0414 step time: 0.2156\n", - "6/8, train_loss: 0.0527 step time: 0.2157\n", - "7/8, train_loss: 0.0365 step time: 0.1993\n", - "8/8, train_loss: 0.0393 step time: 0.2000\n", - "epoch 93 average loss: 0.0481\n", - "time consuming of epoch 93 is: 1.7403\n", - "----------\n", - "epoch 94/600\n", - "1/8, train_loss: 0.0240 step time: 0.2646\n", - "2/8, train_loss: 0.0398 step time: 0.2138\n", - "3/8, train_loss: 0.0549 step time: 0.2149\n", - "4/8, train_loss: 0.0601 step time: 0.2168\n", - "5/8, train_loss: 0.0696 step time: 0.2138\n", - "6/8, train_loss: 0.0344 step time: 0.2156\n", - "7/8, train_loss: 0.0508 step time: 0.1990\n", - "8/8, train_loss: 0.0672 step time: 0.2003\n", - "epoch 94 average loss: 0.0501\n", - "time consuming of epoch 94 is: 1.7402\n", - "----------\n", - "epoch 95/600\n", - "1/8, train_loss: 0.0508 step time: 0.2629\n", - "2/8, train_loss: 0.0413 step time: 0.2136\n", - "3/8, train_loss: 0.0478 step time: 0.2157\n", - "4/8, train_loss: 0.0528 step time: 0.2155\n", - "5/8, train_loss: 0.0248 step time: 0.2148\n", - "6/8, train_loss: 0.0582 step time: 0.2145\n", - "7/8, train_loss: 0.0490 step time: 0.1993\n", - "8/8, train_loss: 0.0247 step time: 0.2014\n", - "epoch 95 average loss: 0.0437\n", - "current epoch: 95 current mean dice: 0.1925 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 95 is: 2.5512\n", - "----------\n", - "epoch 96/600\n", - "1/8, train_loss: 0.0288 step time: 0.2611\n", - "2/8, train_loss: 0.0319 step time: 0.2110\n", - "3/8, train_loss: 0.0300 step time: 0.2119\n", - "4/8, train_loss: 0.0356 step time: 0.2126\n", - "5/8, train_loss: 0.0322 step time: 0.2152\n", - "6/8, train_loss: 0.0352 step time: 0.2150\n", - "7/8, train_loss: 0.0530 step time: 0.2006\n", - "8/8, train_loss: 0.0252 step time: 0.1999\n", - "epoch 96 average loss: 0.0340\n", - "time consuming of epoch 96 is: 1.7285\n", - "----------\n", - "epoch 97/600\n", - "1/8, train_loss: 0.0357 step time: 0.2636\n", - "2/8, train_loss: 0.0318 step time: 0.2148\n", - "3/8, train_loss: 0.0343 step time: 0.2168\n", - "4/8, train_loss: 0.0234 step time: 0.2165\n", - "5/8, train_loss: 0.0269 step time: 0.2130\n", - "6/8, train_loss: 0.0466 step time: 0.2141\n", - "7/8, train_loss: 0.0283 step time: 0.1998\n", - "8/8, train_loss: 0.0359 step time: 0.1998\n", - "epoch 97 average loss: 0.0328\n", - "time consuming of epoch 97 is: 1.7398\n", - "----------\n", - "epoch 98/600\n", - "1/8, train_loss: 0.0201 step time: 0.2641\n", - "2/8, train_loss: 0.0382 step time: 0.2137\n", - "3/8, train_loss: 0.0219 step time: 0.2141\n", - "4/8, train_loss: 0.0327 step time: 0.2139\n", - "5/8, train_loss: 0.0207 step time: 0.2167\n", - "6/8, train_loss: 0.0336 step time: 0.2170\n", - "7/8, train_loss: 0.0258 step time: 0.1995\n", - "8/8, train_loss: 0.0345 step time: 0.1999\n", - "epoch 98 average loss: 0.0284\n", - "time consuming of epoch 98 is: 1.7406\n", - "----------\n", - "epoch 99/600\n", - "1/8, train_loss: 0.0243 step time: 0.2632\n", - "2/8, train_loss: 0.0259 step time: 0.2152\n", - "3/8, train_loss: 0.0277 step time: 0.2139\n", - "4/8, train_loss: 0.0353 step time: 0.2142\n", - "5/8, train_loss: 0.0259 step time: 0.2167\n", - "6/8, train_loss: 0.0255 step time: 0.2153\n", - "7/8, train_loss: 0.0613 step time: 0.2002\n", - "8/8, train_loss: 0.0238 step time: 0.1997\n", - "epoch 99 average loss: 0.0312\n", - "time consuming of epoch 99 is: 1.7397\n", - "----------\n", - "epoch 100/600\n", - "1/8, train_loss: 0.0289 step time: 0.2635\n", - "2/8, train_loss: 0.0191 step time: 0.2150\n", - "3/8, train_loss: 0.0901 step time: 0.2146\n", - "4/8, train_loss: 0.0261 step time: 0.2152\n", - "5/8, train_loss: 0.0320 step time: 0.2176\n", - "6/8, train_loss: 0.0262 step time: 0.2147\n", - "7/8, train_loss: 0.0279 step time: 0.1993\n", - "8/8, train_loss: 0.0354 step time: 0.2003\n", - "epoch 100 average loss: 0.0357\n", - "current epoch: 100 current mean dice: 0.9219 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 100 is: 2.5543\n", - "----------\n", - "epoch 101/600\n", - "1/8, train_loss: 0.0337 step time: 0.3933\n", - "2/8, train_loss: 0.0557 step time: 0.2106\n", - "3/8, train_loss: 0.0516 step time: 0.2073\n", - "4/8, train_loss: 0.0333 step time: 0.2090\n", - "5/8, train_loss: 0.0349 step time: 0.2087\n", - "6/8, train_loss: 0.0440 step time: 0.2080\n", - "7/8, train_loss: 0.0278 step time: 0.1991\n", - "8/8, train_loss: 0.0217 step time: 0.2000\n", - "epoch 101 average loss: 0.0378\n", - "time consuming of epoch 101 is: 1.8373\n", - "----------\n", - "epoch 102/600\n", - "1/8, train_loss: 0.0314 step time: 0.2663\n", - "2/8, train_loss: 0.0334 step time: 0.2169\n", - "3/8, train_loss: 0.0302 step time: 0.2137\n", - "4/8, train_loss: 0.0319 step time: 0.2141\n", - "5/8, train_loss: 0.0315 step time: 0.2155\n", - "6/8, train_loss: 0.0349 step time: 0.2156\n", - "7/8, train_loss: 0.0276 step time: 0.2020\n", - "8/8, train_loss: 0.0277 step time: 0.1991\n", - "epoch 102 average loss: 0.0311\n", - "time consuming of epoch 102 is: 1.7447\n", - "----------\n", - "epoch 103/600\n", - "1/8, train_loss: 0.0380 step time: 0.2642\n", - "2/8, train_loss: 0.0316 step time: 0.2176\n", - "3/8, train_loss: 0.0293 step time: 0.2129\n", - "4/8, train_loss: 0.0413 step time: 0.2127\n", - "5/8, train_loss: 0.0253 step time: 0.2146\n", - "6/8, train_loss: 0.0378 step time: 0.2127\n", - "7/8, train_loss: 0.0198 step time: 0.2004\n", - "8/8, train_loss: 0.0234 step time: 0.2003\n", - "epoch 103 average loss: 0.0308\n", - "time consuming of epoch 103 is: 1.7368\n", - "----------\n", - "epoch 104/600\n", - "1/8, train_loss: 0.0286 step time: 0.2617\n", - "2/8, train_loss: 0.0317 step time: 0.2149\n", - "3/8, train_loss: 0.0281 step time: 0.2140\n", - "4/8, train_loss: 0.0213 step time: 0.2165\n", - "5/8, train_loss: 0.0243 step time: 0.2131\n", - "6/8, train_loss: 0.0266 step time: 0.2173\n", - "7/8, train_loss: 0.0404 step time: 0.1982\n", - "8/8, train_loss: 0.0252 step time: 0.1991\n", - "epoch 104 average loss: 0.0283\n", - "time consuming of epoch 104 is: 1.7364\n", - "----------\n", - "epoch 105/600\n", - "1/8, train_loss: 0.0254 step time: 0.2645\n", - "2/8, train_loss: 0.0208 step time: 0.2172\n", - "3/8, train_loss: 0.0324 step time: 0.2149\n", - "4/8, train_loss: 0.0226 step time: 0.2141\n", - "5/8, train_loss: 0.0243 step time: 0.2148\n", - "6/8, train_loss: 0.0319 step time: 0.2163\n", - "7/8, train_loss: 0.0245 step time: 0.1998\n", - "8/8, train_loss: 0.0257 step time: 0.1995\n", - "epoch 105 average loss: 0.0259\n", - "current epoch: 105 current mean dice: 0.9335 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 105 is: 2.5723\n", - "----------\n", - "epoch 106/600\n", - "1/8, train_loss: 0.0211 step time: 0.2766\n", - "2/8, train_loss: 0.0219 step time: 0.2258\n", - "3/8, train_loss: 0.0244 step time: 0.2242\n", - "4/8, train_loss: 0.0243 step time: 0.2183\n", - "5/8, train_loss: 0.0236 step time: 0.2154\n", - "6/8, train_loss: 0.0244 step time: 0.2137\n", - "7/8, train_loss: 0.0258 step time: 0.2032\n", - "8/8, train_loss: 0.0235 step time: 0.2022\n", - "epoch 106 average loss: 0.0237\n", - "time consuming of epoch 106 is: 1.7806\n", - "----------\n", - "epoch 107/600\n", - "1/8, train_loss: 0.0262 step time: 0.2660\n", - "2/8, train_loss: 0.0273 step time: 0.2159\n", - "3/8, train_loss: 0.0186 step time: 0.2174\n", - "4/8, train_loss: 0.0224 step time: 0.2149\n", - "5/8, train_loss: 0.0279 step time: 0.2148\n", - "6/8, train_loss: 0.0216 step time: 0.2155\n", - "7/8, train_loss: 0.0237 step time: 0.2001\n", - "8/8, train_loss: 0.0244 step time: 0.2037\n", - "epoch 107 average loss: 0.0240\n", - "time consuming of epoch 107 is: 1.7499\n", - "----------\n", - "epoch 108/600\n", - "1/8, train_loss: 0.0229 step time: 0.2651\n", - "2/8, train_loss: 0.0203 step time: 0.2147\n", - "3/8, train_loss: 0.0184 step time: 0.2107\n", - "4/8, train_loss: 0.0293 step time: 0.2118\n", - "5/8, train_loss: 0.0233 step time: 0.2125\n", - "6/8, train_loss: 0.0240 step time: 0.2123\n", - "7/8, train_loss: 0.0281 step time: 0.2007\n", - "8/8, train_loss: 0.0232 step time: 0.2057\n", - "epoch 108 average loss: 0.0237\n", - "time consuming of epoch 108 is: 1.7350\n", - "----------\n", - "epoch 109/600\n", - "1/8, train_loss: 0.0215 step time: 0.2678\n", - "2/8, train_loss: 0.0231 step time: 0.2189\n", - "3/8, train_loss: 0.0246 step time: 0.2177\n", - "4/8, train_loss: 0.0215 step time: 0.2189\n", - "5/8, train_loss: 0.0234 step time: 0.2160\n", - "6/8, train_loss: 0.0181 step time: 0.2145\n", - "7/8, train_loss: 0.0210 step time: 0.2022\n", - "8/8, train_loss: 0.0285 step time: 0.1994\n", - "epoch 109 average loss: 0.0227\n", - "time consuming of epoch 109 is: 1.7568\n", - "----------\n", - "epoch 110/600\n", - "1/8, train_loss: 0.0167 step time: 0.2637\n", - "2/8, train_loss: 0.0200 step time: 0.2150\n", - "3/8, train_loss: 0.0215 step time: 0.2162\n", - "4/8, train_loss: 0.0264 step time: 0.2205\n", - "5/8, train_loss: 0.0186 step time: 0.2225\n", - "6/8, train_loss: 0.0220 step time: 0.2279\n", - "7/8, train_loss: 0.0206 step time: 0.2394\n", - "8/8, train_loss: 0.0223 step time: 0.2422\n", - "epoch 110 average loss: 0.0210\n", - "current epoch: 110 current mean dice: 0.9441 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 110 is: 3.0572\n", - "----------\n", - "epoch 111/600\n", - "1/8, train_loss: 0.0231 step time: 0.3005\n", - "2/8, train_loss: 0.0214 step time: 0.2418\n", - "3/8, train_loss: 0.0227 step time: 0.2369\n", - "4/8, train_loss: 0.0293 step time: 0.2297\n", - "5/8, train_loss: 0.0196 step time: 0.2283\n", - "6/8, train_loss: 0.0322 step time: 0.2267\n", - "7/8, train_loss: 0.0214 step time: 0.2151\n", - "8/8, train_loss: 0.0283 step time: 0.2111\n", - "epoch 111 average loss: 0.0247\n", - "time consuming of epoch 111 is: 1.8914\n", - "----------\n", - "epoch 112/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0210 step time: 0.2729\n", - "2/8, train_loss: 0.0199 step time: 0.2252\n", - "3/8, train_loss: 0.0228 step time: 0.2188\n", - "4/8, train_loss: 0.0183 step time: 0.2215\n", - "5/8, train_loss: 0.0208 step time: 0.2196\n", - "6/8, train_loss: 0.0243 step time: 0.2173\n", - "7/8, train_loss: 0.0227 step time: 0.2083\n", - "8/8, train_loss: 0.0229 step time: 0.2040\n", - "epoch 112 average loss: 0.0216\n", - "time consuming of epoch 112 is: 1.7892\n", - "----------\n", - "epoch 113/600\n", - "1/8, train_loss: 0.0230 step time: 0.2714\n", - "2/8, train_loss: 0.0279 step time: 0.2171\n", - "3/8, train_loss: 0.0225 step time: 0.2201\n", - "4/8, train_loss: 0.0293 step time: 0.2186\n", - "5/8, train_loss: 0.0216 step time: 0.2176\n", - "6/8, train_loss: 0.0257 step time: 0.2168\n", - "7/8, train_loss: 0.0293 step time: 0.2037\n", - "8/8, train_loss: 0.0204 step time: 0.2029\n", - "epoch 113 average loss: 0.0250\n", - "time consuming of epoch 113 is: 1.7697\n", - "----------\n", - "epoch 114/600\n", - "1/8, train_loss: 0.0214 step time: 0.2658\n", - "2/8, train_loss: 0.0239 step time: 0.2184\n", - "3/8, train_loss: 0.0221 step time: 0.2154\n", - "4/8, train_loss: 0.0214 step time: 0.2202\n", - "5/8, train_loss: 0.0236 step time: 0.2184\n", - "6/8, train_loss: 0.0205 step time: 0.2178\n", - "7/8, train_loss: 0.0225 step time: 0.2119\n", - "8/8, train_loss: 0.0242 step time: 0.2511\n", - "epoch 114 average loss: 0.0225\n", - "time consuming of epoch 114 is: 1.8205\n", - "----------\n", - "epoch 115/600\n", - "1/8, train_loss: 0.0266 step time: 0.3097\n", - "2/8, train_loss: 0.0227 step time: 0.2486\n", - "3/8, train_loss: 0.0227 step time: 0.2385\n", - "4/8, train_loss: 0.0214 step time: 0.2344\n", - "5/8, train_loss: 0.0171 step time: 0.2311\n", - "6/8, train_loss: 0.0180 step time: 0.2297\n", - "7/8, train_loss: 0.0211 step time: 0.2150\n", - "8/8, train_loss: 0.0245 step time: 0.2139\n", - "epoch 115 average loss: 0.0217\n", - "current epoch: 115 current mean dice: 0.9427 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 115 is: 3.1218\n", - "----------\n", - "epoch 116/600\n", - "1/8, train_loss: 0.0221 step time: 0.2995\n", - "2/8, train_loss: 0.0187 step time: 0.2502\n", - "3/8, train_loss: 0.0235 step time: 0.2394\n", - "4/8, train_loss: 0.0203 step time: 0.2371\n", - "5/8, train_loss: 0.0244 step time: 0.2363\n", - "6/8, train_loss: 0.0237 step time: 0.2340\n", - "7/8, train_loss: 0.0238 step time: 0.2189\n", - "8/8, train_loss: 0.0228 step time: 0.2172\n", - "epoch 116 average loss: 0.0224\n", - "time consuming of epoch 116 is: 1.9339\n", - "----------\n", - "epoch 117/600\n", - "1/8, train_loss: 0.0215 step time: 0.2804\n", - "2/8, train_loss: 0.0222 step time: 0.2304\n", - "3/8, train_loss: 0.0163 step time: 0.2279\n", - "4/8, train_loss: 0.0185 step time: 0.2276\n", - "5/8, train_loss: 0.0219 step time: 0.2289\n", - "6/8, train_loss: 0.0298 step time: 0.2257\n", - "7/8, train_loss: 0.0226 step time: 0.2153\n", - "8/8, train_loss: 0.0233 step time: 0.2145\n", - "epoch 117 average loss: 0.0220\n", - "time consuming of epoch 117 is: 1.8522\n", - "----------\n", - "epoch 118/600\n", - "1/8, train_loss: 0.0242 step time: 0.2758\n", - "2/8, train_loss: 0.0248 step time: 0.2279\n", - "3/8, train_loss: 0.0180 step time: 0.3280\n", - "4/8, train_loss: 0.0191 step time: 0.4553\n", - "5/8, train_loss: 0.0229 step time: 0.2192\n", - "6/8, train_loss: 0.0205 step time: 0.2147\n", - "7/8, train_loss: 0.0187 step time: 0.2005\n", - "8/8, train_loss: 0.0226 step time: 0.2005\n", - "epoch 118 average loss: 0.0214\n", - "time consuming of epoch 118 is: 2.1235\n", - "----------\n", - "epoch 119/600\n", - "1/8, train_loss: 0.0212 step time: 0.2639\n", - "2/8, train_loss: 0.0215 step time: 0.2128\n", - "3/8, train_loss: 0.0216 step time: 0.2226\n", - "4/8, train_loss: 0.0202 step time: 0.2243\n", - "5/8, train_loss: 0.0204 step time: 0.2245\n", - "6/8, train_loss: 0.0227 step time: 0.2271\n", - "7/8, train_loss: 0.0194 step time: 0.2089\n", - "8/8, train_loss: 0.0202 step time: 0.2101\n", - "epoch 119 average loss: 0.0209\n", - "time consuming of epoch 119 is: 1.7956\n", - "----------\n", - "epoch 120/600\n", - "1/8, train_loss: 0.0178 step time: 0.2716\n", - "2/8, train_loss: 0.0190 step time: 0.2268\n", - "3/8, train_loss: 0.0209 step time: 0.2515\n", - "4/8, train_loss: 0.0243 step time: 0.2591\n", - "5/8, train_loss: 0.0194 step time: 0.2536\n", - "6/8, train_loss: 0.0204 step time: 0.2850\n", - "7/8, train_loss: 0.0233 step time: 0.3002\n", - "8/8, train_loss: 0.0201 step time: 0.2680\n", - "epoch 120 average loss: 0.0207\n", - "current epoch: 120 current mean dice: 0.9477 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 120 is: 3.2718\n", - "----------\n", - "epoch 121/600\n", - "1/8, train_loss: 0.0204 step time: 0.6131\n", - "2/8, train_loss: 0.0204 step time: 0.2160\n", - "3/8, train_loss: 0.0192 step time: 0.2140\n", - "4/8, train_loss: 0.0168 step time: 0.2134\n", - "5/8, train_loss: 0.0171 step time: 0.2161\n", - "6/8, train_loss: 0.0208 step time: 0.2174\n", - "7/8, train_loss: 0.0225 step time: 0.2095\n", - "8/8, train_loss: 0.0220 step time: 0.2179\n", - "epoch 121 average loss: 0.0199\n", - "time consuming of epoch 121 is: 2.1187\n", - "----------\n", - "epoch 122/600\n", - "1/8, train_loss: 0.0174 step time: 0.2792\n", - "2/8, train_loss: 0.0189 step time: 0.2303\n", - "3/8, train_loss: 0.0160 step time: 0.2365\n", - "4/8, train_loss: 0.0231 step time: 0.2357\n", - "5/8, train_loss: 0.0218 step time: 0.2906\n", - "6/8, train_loss: 0.0185 step time: 0.2737\n", - "7/8, train_loss: 0.0206 step time: 0.2583\n", - "8/8, train_loss: 0.0195 step time: 0.2632\n", - "epoch 122 average loss: 0.0195\n", - "time consuming of epoch 122 is: 2.0690\n", - "----------\n", - "epoch 123/600\n", - "1/8, train_loss: 0.0230 step time: 0.2869\n", - "2/8, train_loss: 0.0209 step time: 0.2344\n", - "3/8, train_loss: 0.0280 step time: 0.2307\n", - "4/8, train_loss: 0.0246 step time: 0.2303\n", - "5/8, train_loss: 0.0218 step time: 0.2353\n", - "6/8, train_loss: 0.0244 step time: 0.2301\n", - "7/8, train_loss: 0.0150 step time: 0.2154\n", - "8/8, train_loss: 0.0191 step time: 0.2214\n", - "epoch 123 average loss: 0.0221\n", - "time consuming of epoch 123 is: 1.8861\n", - "----------\n", - "epoch 124/600\n", - "1/8, train_loss: 0.0213 step time: 0.2820\n", - "2/8, train_loss: 0.0249 step time: 0.2331\n", - "3/8, train_loss: 0.0180 step time: 0.2308\n", - "4/8, train_loss: 0.0196 step time: 0.2279\n", - "5/8, train_loss: 0.0197 step time: 0.2346\n", - "6/8, train_loss: 0.0222 step time: 0.2475\n", - "7/8, train_loss: 0.0210 step time: 0.2397\n", - "8/8, train_loss: 0.0195 step time: 0.2430\n", - "epoch 124 average loss: 0.0208\n", - "time consuming of epoch 124 is: 1.9402\n", - "----------\n", - "epoch 125/600\n", - "1/8, train_loss: 0.0232 step time: 0.2910\n", - "2/8, train_loss: 0.0201 step time: 0.2362\n", - "3/8, train_loss: 0.0192 step time: 0.2311\n", - "4/8, train_loss: 0.0206 step time: 0.2287\n", - "5/8, train_loss: 0.0208 step time: 0.2397\n", - "6/8, train_loss: 0.0226 step time: 0.2572\n", - "7/8, train_loss: 0.0209 step time: 0.2512\n", - "8/8, train_loss: 0.0179 step time: 0.2687\n", - "epoch 125 average loss: 0.0207\n", - "current epoch: 125 current mean dice: 0.9476 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 125 is: 3.2610\n", - "----------\n", - "epoch 126/600\n", - "1/8, train_loss: 0.0191 step time: 0.3064\n", - "2/8, train_loss: 0.0198 step time: 0.2526\n", - "3/8, train_loss: 0.0196 step time: 0.2497\n", - "4/8, train_loss: 0.0254 step time: 0.2512\n", - "5/8, train_loss: 0.0194 step time: 0.2375\n", - "6/8, train_loss: 0.0220 step time: 0.2305\n", - "7/8, train_loss: 0.0193 step time: 0.2231\n", - "8/8, train_loss: 0.0207 step time: 0.2207\n", - "epoch 126 average loss: 0.0207\n", - "time consuming of epoch 126 is: 1.9730\n", - "----------\n", - "epoch 127/600\n", - "1/8, train_loss: 0.0232 step time: 0.2841\n", - "2/8, train_loss: 0.0197 step time: 0.2353\n", - "3/8, train_loss: 0.0185 step time: 0.2341\n", - "4/8, train_loss: 0.0225 step time: 0.2313\n", - "5/8, train_loss: 0.0187 step time: 0.2293\n", - "6/8, train_loss: 0.0223 step time: 0.2294\n", - "7/8, train_loss: 0.0180 step time: 0.2609\n", - "8/8, train_loss: 0.0170 step time: 0.2969\n", - "epoch 127 average loss: 0.0200\n", - "time consuming of epoch 127 is: 2.0029\n", - "----------\n", - "epoch 128/600\n", - "1/8, train_loss: 0.0208 step time: 0.3316\n", - "2/8, train_loss: 0.0232 step time: 0.2600\n", - "3/8, train_loss: 0.0173 step time: 0.2530\n", - "4/8, train_loss: 0.0185 step time: 0.2456\n", - "5/8, train_loss: 0.0185 step time: 0.2420\n", - "6/8, train_loss: 0.0174 step time: 0.2396\n", - "7/8, train_loss: 0.0243 step time: 0.2248\n", - "8/8, train_loss: 0.0193 step time: 0.2170\n", - "epoch 128 average loss: 0.0199\n", - "time consuming of epoch 128 is: 2.0151\n", - "----------\n", - "epoch 129/600\n", - "1/8, train_loss: 0.0191 step time: 0.2806\n", - "2/8, train_loss: 0.0201 step time: 0.2297\n", - "3/8, train_loss: 0.0203 step time: 0.2294\n", - "4/8, train_loss: 0.0209 step time: 0.2306\n", - "5/8, train_loss: 0.0182 step time: 0.2314\n", - "6/8, train_loss: 0.0216 step time: 0.2288\n", - "7/8, train_loss: 0.0151 step time: 0.2192\n", - "8/8, train_loss: 0.0207 step time: 0.2188\n", - "epoch 129 average loss: 0.0195\n", - "time consuming of epoch 129 is: 1.8699\n", - "----------\n", - "epoch 130/600\n", - "1/8, train_loss: 0.0208 step time: 0.2805\n", - "2/8, train_loss: 0.0225 step time: 0.2280\n", - "3/8, train_loss: 0.0156 step time: 0.2325\n", - "4/8, train_loss: 0.0190 step time: 0.2392\n", - "5/8, train_loss: 0.0211 step time: 0.2309\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "6/8, train_loss: 0.0212 step time: 0.3590\n", - "7/8, train_loss: 0.0184 step time: 0.4614\n", - "8/8, train_loss: 0.0262 step time: 0.2619\n", - "epoch 130 average loss: 0.0206\n", - "current epoch: 130 current mean dice: 0.9471 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 130 is: 3.4656\n", - "----------\n", - "epoch 131/600\n", - "1/8, train_loss: 0.0291 step time: 0.2864\n", - "2/8, train_loss: 0.0266 step time: 0.2428\n", - "3/8, train_loss: 0.0209 step time: 0.2429\n", - "4/8, train_loss: 0.0181 step time: 0.2571\n", - "5/8, train_loss: 0.0184 step time: 0.2564\n", - "6/8, train_loss: 0.0183 step time: 0.2501\n", - "7/8, train_loss: 0.0196 step time: 0.2410\n", - "8/8, train_loss: 0.0166 step time: 0.2282\n", - "epoch 131 average loss: 0.0209\n", - "time consuming of epoch 131 is: 2.0061\n", - "----------\n", - "epoch 132/600\n", - "1/8, train_loss: 0.0220 step time: 0.2844\n", - "2/8, train_loss: 0.0206 step time: 0.2313\n", - "3/8, train_loss: 0.0181 step time: 0.2287\n", - "4/8, train_loss: 0.0180 step time: 0.2276\n", - "5/8, train_loss: 0.0221 step time: 0.2250\n", - "6/8, train_loss: 0.0238 step time: 0.2252\n", - "7/8, train_loss: 0.0185 step time: 0.2222\n", - "8/8, train_loss: 0.0199 step time: 0.2649\n", - "epoch 132 average loss: 0.0204\n", - "time consuming of epoch 132 is: 1.9109\n", - "----------\n", - "epoch 133/600\n", - "1/8, train_loss: 0.0211 step time: 0.3262\n", - "2/8, train_loss: 0.0201 step time: 0.2779\n", - "3/8, train_loss: 0.0175 step time: 0.2805\n", - "4/8, train_loss: 0.0195 step time: 0.2788\n", - "5/8, train_loss: 0.0156 step time: 0.2808\n", - "6/8, train_loss: 0.0207 step time: 0.2754\n", - "7/8, train_loss: 0.0170 step time: 0.2473\n", - "8/8, train_loss: 0.0198 step time: 0.2316\n", - "epoch 133 average loss: 0.0189\n", - "time consuming of epoch 133 is: 2.2002\n", - "----------\n", - "epoch 134/600\n", - "1/8, train_loss: 0.0187 step time: 0.2844\n", - "2/8, train_loss: 0.0196 step time: 0.2382\n", - "3/8, train_loss: 0.0194 step time: 0.2323\n", - "4/8, train_loss: 0.0172 step time: 0.2874\n", - "5/8, train_loss: 0.0209 step time: 0.3443\n", - "6/8, train_loss: 0.0186 step time: 0.2775\n", - "7/8, train_loss: 0.0215 step time: 0.2533\n", - "8/8, train_loss: 0.0262 step time: 0.2407\n", - "epoch 134 average loss: 0.0203\n", - "time consuming of epoch 134 is: 2.1596\n", - "----------\n", - "epoch 135/600\n", - "1/8, train_loss: 0.0226 step time: 0.2912\n", - "2/8, train_loss: 0.0206 step time: 0.2377\n", - "3/8, train_loss: 0.0199 step time: 0.2368\n", - "4/8, train_loss: 0.0233 step time: 0.2442\n", - "5/8, train_loss: 0.0160 step time: 0.2434\n", - "6/8, train_loss: 0.0186 step time: 0.2377\n", - "7/8, train_loss: 0.0200 step time: 0.2240\n", - "8/8, train_loss: 0.0176 step time: 0.2266\n", - "epoch 135 average loss: 0.0198\n", - "current epoch: 135 current mean dice: 0.9466 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 135 is: 3.2663\n", - "----------\n", - "epoch 136/600\n", - "1/8, train_loss: 0.0209 step time: 0.3197\n", - "2/8, train_loss: 0.0221 step time: 0.2581\n", - "3/8, train_loss: 0.0196 step time: 0.2428\n", - "4/8, train_loss: 0.0203 step time: 0.2376\n", - "5/8, train_loss: 0.0193 step time: 0.2717\n", - "6/8, train_loss: 0.0218 step time: 0.3029\n", - "7/8, train_loss: 0.0176 step time: 0.2722\n", - "8/8, train_loss: 0.0170 step time: 0.2534\n", - "epoch 136 average loss: 0.0198\n", - "time consuming of epoch 136 is: 2.1596\n", - "----------\n", - "epoch 137/600\n", - "1/8, train_loss: 0.0220 step time: 0.2881\n", - "2/8, train_loss: 0.0183 step time: 0.2384\n", - "3/8, train_loss: 0.0175 step time: 0.2355\n", - "4/8, train_loss: 0.0180 step time: 0.2394\n", - "5/8, train_loss: 0.0189 step time: 0.2311\n", - "6/8, train_loss: 0.0173 step time: 0.2291\n", - "7/8, train_loss: 0.0183 step time: 0.2170\n", - "8/8, train_loss: 0.0205 step time: 0.2151\n", - "epoch 137 average loss: 0.0189\n", - "time consuming of epoch 137 is: 1.8952\n", - "----------\n", - "epoch 138/600\n", - "1/8, train_loss: 0.0191 step time: 0.2777\n", - "2/8, train_loss: 0.0270 step time: 0.2284\n", - "3/8, train_loss: 0.0209 step time: 0.2284\n", - "4/8, train_loss: 0.0200 step time: 0.2294\n", - "5/8, train_loss: 0.0199 step time: 0.2257\n", - "6/8, train_loss: 0.0167 step time: 0.2283\n", - "7/8, train_loss: 0.0179 step time: 0.2384\n", - "8/8, train_loss: 0.0231 step time: 0.3161\n", - "epoch 138 average loss: 0.0206\n", - "time consuming of epoch 138 is: 1.9739\n", - "----------\n", - "epoch 139/600\n", - "1/8, train_loss: 0.0240 step time: 0.3044\n", - "2/8, train_loss: 0.0212 step time: 0.2574\n", - "3/8, train_loss: 0.0202 step time: 0.2473\n", - "4/8, train_loss: 0.0170 step time: 0.2423\n", - "5/8, train_loss: 0.0182 step time: 0.2384\n", - "6/8, train_loss: 0.0172 step time: 0.2609\n", - "7/8, train_loss: 0.0451 step time: 0.3056\n", - "8/8, train_loss: 0.0214 step time: 0.2532\n", - "epoch 139 average loss: 0.0230\n", - "time consuming of epoch 139 is: 2.1108\n", - "----------\n", - "epoch 140/600\n", - "1/8, train_loss: 0.0203 step time: 0.2944\n", - "2/8, train_loss: 0.0183 step time: 0.2452\n", - "3/8, train_loss: 0.0229 step time: 0.2454\n", - "4/8, train_loss: 0.0211 step time: 0.2451\n", - "5/8, train_loss: 0.0179 step time: 0.2434\n", - "6/8, train_loss: 0.0258 step time: 0.2364\n", - "7/8, train_loss: 0.0220 step time: 0.2185\n", - "8/8, train_loss: 0.0227 step time: 0.2204\n", - "epoch 140 average loss: 0.0214\n", - "current epoch: 140 current mean dice: 0.9294 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 140 is: 3.4298\n", - "----------\n", - "epoch 141/600\n", - "1/8, train_loss: 0.0185 step time: 0.3186\n", - "2/8, train_loss: 0.0314 step time: 0.2535\n", - "3/8, train_loss: 0.0264 step time: 0.2387\n", - "4/8, train_loss: 0.0225 step time: 0.2328\n", - "5/8, train_loss: 0.0378 step time: 0.2302\n", - "6/8, train_loss: 0.0199 step time: 0.2331\n", - "7/8, train_loss: 0.0142 step time: 0.2206\n", - "8/8, train_loss: 0.0228 step time: 0.2198\n", - "epoch 141 average loss: 0.0242\n", - "time consuming of epoch 141 is: 1.9486\n", - "----------\n", - "epoch 142/600\n", - "1/8, train_loss: 0.0251 step time: 0.2790\n", - "2/8, train_loss: 0.0175 step time: 0.2442\n", - "3/8, train_loss: 0.0218 step time: 0.2596\n", - "4/8, train_loss: 0.0164 step time: 0.2549\n", - "5/8, train_loss: 0.0249 step time: 0.2457\n", - "6/8, train_loss: 0.0186 step time: 0.2438\n", - "7/8, train_loss: 0.0329 step time: 0.2257\n", - "8/8, train_loss: 0.0243 step time: 0.2198\n", - "epoch 142 average loss: 0.0227\n", - "time consuming of epoch 142 is: 1.9742\n", - "----------\n", - "epoch 143/600\n", - "1/8, train_loss: 0.0178 step time: 0.2775\n", - "2/8, train_loss: 0.0203 step time: 0.2299\n", - "3/8, train_loss: 0.0239 step time: 0.2283\n", - "4/8, train_loss: 0.0260 step time: 0.2296\n", - "5/8, train_loss: 0.0190 step time: 0.2273\n", - "6/8, train_loss: 0.0200 step time: 0.2282\n", - "7/8, train_loss: 0.0188 step time: 0.2777\n", - "8/8, train_loss: 0.0293 step time: 0.3090\n", - "epoch 143 average loss: 0.0219\n", - "time consuming of epoch 143 is: 2.0091\n", - "----------\n", - "epoch 144/600\n", - "1/8, train_loss: 0.0224 step time: 0.3165\n", - "2/8, train_loss: 0.0225 step time: 0.2725\n", - "3/8, train_loss: 0.0291 step time: 0.2632\n", - "4/8, train_loss: 0.0212 step time: 0.2556\n", - "5/8, train_loss: 0.0214 step time: 0.2497\n", - "6/8, train_loss: 0.0228 step time: 0.2457\n", - "7/8, train_loss: 0.0169 step time: 0.2364\n", - "8/8, train_loss: 0.0209 step time: 0.2319\n", - "epoch 144 average loss: 0.0222\n", - "time consuming of epoch 144 is: 2.0729\n", - "----------\n", - "epoch 145/600\n", - "1/8, train_loss: 0.0193 step time: 0.2852\n", - "2/8, train_loss: 0.0267 step time: 0.2338\n", - "3/8, train_loss: 0.0279 step time: 0.3805\n", - "4/8, train_loss: 0.0277 step time: 0.4950\n", - "5/8, train_loss: 0.0195 step time: 0.2680\n", - "6/8, train_loss: 0.0193 step time: 0.2482\n", - "7/8, train_loss: 0.0179 step time: 0.2401\n", - "8/8, train_loss: 0.0231 step time: 0.2343\n", - "epoch 145 average loss: 0.0227\n", - "current epoch: 145 current mean dice: 0.9417 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 145 is: 3.7547\n", - "----------\n", - "epoch 146/600\n", - "1/8, train_loss: 0.0642 step time: 0.3173\n", - "2/8, train_loss: 0.0190 step time: 0.2507\n", - "3/8, train_loss: 0.0257 step time: 0.2442\n", - "4/8, train_loss: 0.0213 step time: 0.2439\n", - "5/8, train_loss: 0.0514 step time: 0.2456\n", - "6/8, train_loss: 0.0286 step time: 0.2433\n", - "7/8, train_loss: 0.0311 step time: 0.2393\n", - "8/8, train_loss: 0.0181 step time: 0.2273\n", - "epoch 146 average loss: 0.0324\n", - "time consuming of epoch 146 is: 2.0128\n", - "----------\n", - "epoch 147/600\n", - "1/8, train_loss: 0.0180 step time: 0.2805\n", - "2/8, train_loss: 0.0517 step time: 0.2341\n", - "3/8, train_loss: 0.0350 step time: 0.2633\n", - "4/8, train_loss: 0.0252 step time: 0.2701\n", - "5/8, train_loss: 0.0327 step time: 0.2507\n", - "6/8, train_loss: 0.0262 step time: 0.2449\n", - "7/8, train_loss: 0.0222 step time: 0.2359\n", - "8/8, train_loss: 0.0190 step time: 0.2303\n", - "epoch 147 average loss: 0.0288\n", - "time consuming of epoch 147 is: 2.0113\n", - "----------\n", - "epoch 148/600\n", - "1/8, train_loss: 0.0262 step time: 0.2836\n", - "2/8, train_loss: 0.0271 step time: 0.2313\n", - "3/8, train_loss: 0.0248 step time: 0.2311\n", - "4/8, train_loss: 0.0293 step time: 0.2323\n", - "5/8, train_loss: 0.0271 step time: 0.2420\n", - "6/8, train_loss: 0.0196 step time: 0.2353\n", - "7/8, train_loss: 0.0384 step time: 0.2180\n", - "8/8, train_loss: 0.0247 step time: 0.2626\n", - "epoch 148 average loss: 0.0272\n", - "time consuming of epoch 148 is: 1.9376\n", - "----------\n", - "epoch 149/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0279 step time: 0.3637\n", - "2/8, train_loss: 0.0226 step time: 0.2666\n", - "3/8, train_loss: 0.0229 step time: 0.2559\n", - "4/8, train_loss: 0.0234 step time: 0.2476\n", - "5/8, train_loss: 0.0203 step time: 0.2396\n", - "6/8, train_loss: 0.0442 step time: 0.2475\n", - "7/8, train_loss: 0.0232 step time: 0.2661\n", - "8/8, train_loss: 0.0215 step time: 0.2627\n", - "epoch 149 average loss: 0.0257\n", - "time consuming of epoch 149 is: 2.1513\n", - "----------\n", - "epoch 150/600\n", - "1/8, train_loss: 0.0392 step time: 0.2907\n", - "2/8, train_loss: 0.0269 step time: 0.2484\n", - "3/8, train_loss: 0.0256 step time: 0.2481\n", - "4/8, train_loss: 0.0186 step time: 0.2415\n", - "5/8, train_loss: 0.0289 step time: 0.2323\n", - "6/8, train_loss: 0.0218 step time: 0.2349\n", - "7/8, train_loss: 0.0286 step time: 0.2921\n", - "8/8, train_loss: 0.0311 step time: 0.2697\n", - "epoch 150 average loss: 0.0276\n", - "current epoch: 150 current mean dice: 0.9290 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 150 is: 3.5274\n", - "----------\n", - "epoch 151/600\n", - "1/8, train_loss: 0.0421 step time: 0.2959\n", - "2/8, train_loss: 0.0250 step time: 0.2376\n", - "3/8, train_loss: 0.0240 step time: 0.2412\n", - "4/8, train_loss: 0.0242 step time: 0.2427\n", - "5/8, train_loss: 0.0247 step time: 0.2429\n", - "6/8, train_loss: 0.0245 step time: 0.2930\n", - "7/8, train_loss: 0.0272 step time: 0.2762\n", - "8/8, train_loss: 0.0200 step time: 0.2532\n", - "epoch 151 average loss: 0.0265\n", - "time consuming of epoch 151 is: 2.0839\n", - "----------\n", - "epoch 152/600\n", - "1/8, train_loss: 0.0186 step time: 0.2877\n", - "2/8, train_loss: 0.0211 step time: 0.2318\n", - "3/8, train_loss: 0.0257 step time: 0.2368\n", - "4/8, train_loss: 0.0229 step time: 0.2518\n", - "5/8, train_loss: 0.0183 step time: 0.3069\n", - "6/8, train_loss: 0.0205 step time: 0.2862\n", - "7/8, train_loss: 0.0208 step time: 0.2707\n", - "8/8, train_loss: 0.0196 step time: 0.2550\n", - "epoch 152 average loss: 0.0209\n", - "time consuming of epoch 152 is: 2.1283\n", - "----------\n", - "epoch 153/600\n", - "1/8, train_loss: 0.0242 step time: 0.2957\n", - "2/8, train_loss: 0.0288 step time: 0.2501\n", - "3/8, train_loss: 0.0192 step time: 0.2427\n", - "4/8, train_loss: 0.0193 step time: 0.2368\n", - "5/8, train_loss: 0.0186 step time: 0.2340\n", - "6/8, train_loss: 0.0196 step time: 0.2377\n", - "7/8, train_loss: 0.0225 step time: 0.2245\n", - "8/8, train_loss: 0.0212 step time: 0.3610\n", - "epoch 153 average loss: 0.0217\n", - "time consuming of epoch 153 is: 2.0840\n", - "----------\n", - "epoch 154/600\n", - "1/8, train_loss: 0.0246 step time: 0.4169\n", - "2/8, train_loss: 0.0249 step time: 0.2705\n", - "3/8, train_loss: 0.0208 step time: 0.2581\n", - "4/8, train_loss: 0.0253 step time: 0.2428\n", - "5/8, train_loss: 0.0211 step time: 0.2389\n", - "6/8, train_loss: 0.0176 step time: 0.2344\n", - "7/8, train_loss: 0.0160 step time: 0.2158\n", - "8/8, train_loss: 0.0190 step time: 0.2167\n", - "epoch 154 average loss: 0.0212\n", - "time consuming of epoch 154 is: 2.0955\n", - "----------\n", - "epoch 155/600\n", - "1/8, train_loss: 0.0239 step time: 0.2791\n", - "2/8, train_loss: 0.0194 step time: 0.2277\n", - "3/8, train_loss: 0.0178 step time: 0.2332\n", - "4/8, train_loss: 0.0214 step time: 0.2603\n", - "5/8, train_loss: 0.0191 step time: 0.2691\n", - "6/8, train_loss: 0.0214 step time: 0.2501\n", - "7/8, train_loss: 0.0219 step time: 0.2403\n", - "8/8, train_loss: 0.0278 step time: 0.2408\n", - "epoch 155 average loss: 0.0216\n", - "current epoch: 155 current mean dice: 0.9469 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 155 is: 3.4074\n", - "----------\n", - "epoch 156/600\n", - "1/8, train_loss: 0.0206 step time: 0.2933\n", - "2/8, train_loss: 0.0171 step time: 0.2437\n", - "3/8, train_loss: 0.0169 step time: 0.2352\n", - "4/8, train_loss: 0.0222 step time: 0.2299\n", - "5/8, train_loss: 0.0201 step time: 0.2299\n", - "6/8, train_loss: 0.0198 step time: 0.2305\n", - "7/8, train_loss: 0.0248 step time: 0.2190\n", - "8/8, train_loss: 0.0206 step time: 0.2185\n", - "epoch 156 average loss: 0.0203\n", - "time consuming of epoch 156 is: 1.9013\n", - "----------\n", - "epoch 157/600\n", - "1/8, train_loss: 0.0231 step time: 0.2772\n", - "2/8, train_loss: 0.0178 step time: 0.2292\n", - "3/8, train_loss: 0.0290 step time: 0.3113\n", - "4/8, train_loss: 0.0194 step time: 0.4848\n", - "5/8, train_loss: 0.0177 step time: 0.2184\n", - "6/8, train_loss: 0.0358 step time: 0.2142\n", - "7/8, train_loss: 0.0161 step time: 0.1997\n", - "8/8, train_loss: 0.0227 step time: 0.2015\n", - "epoch 157 average loss: 0.0227\n", - "time consuming of epoch 157 is: 2.1379\n", - "----------\n", - "epoch 158/600\n", - "1/8, train_loss: 0.0184 step time: 0.6316\n", - "2/8, train_loss: 0.0321 step time: 0.2196\n", - "3/8, train_loss: 0.0238 step time: 0.2141\n", - "4/8, train_loss: 0.0169 step time: 0.2157\n", - "5/8, train_loss: 0.0209 step time: 0.2182\n", - "6/8, train_loss: 0.0176 step time: 0.2165\n", - "7/8, train_loss: 0.0297 step time: 0.2098\n", - "8/8, train_loss: 0.0224 step time: 0.2378\n", - "epoch 158 average loss: 0.0227\n", - "time consuming of epoch 158 is: 2.1649\n", - "----------\n", - "epoch 159/600\n", - "1/8, train_loss: 0.0227 step time: 0.3156\n", - "2/8, train_loss: 0.0205 step time: 0.2572\n", - "3/8, train_loss: 0.0251 step time: 0.2584\n", - "4/8, train_loss: 0.0200 step time: 0.2521\n", - "5/8, train_loss: 0.0270 step time: 0.2520\n", - "6/8, train_loss: 0.0168 step time: 0.2447\n", - "7/8, train_loss: 0.0233 step time: 0.2416\n", - "8/8, train_loss: 0.0226 step time: 0.2383\n", - "epoch 159 average loss: 0.0223\n", - "time consuming of epoch 159 is: 2.0613\n", - "----------\n", - "epoch 160/600\n", - "1/8, train_loss: 0.0201 step time: 0.2868\n", - "2/8, train_loss: 0.0187 step time: 0.2407\n", - "3/8, train_loss: 0.0212 step time: 0.2413\n", - "4/8, train_loss: 0.0312 step time: 0.2422\n", - "5/8, train_loss: 0.0193 step time: 0.3374\n", - "6/8, train_loss: 0.0186 step time: 0.3616\n", - "7/8, train_loss: 0.0211 step time: 0.2701\n", - "8/8, train_loss: 0.0200 step time: 0.2424\n", - "epoch 160 average loss: 0.0213\n", - "current epoch: 160 current mean dice: 0.9427 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 160 is: 3.7543\n", - "----------\n", - "epoch 161/600\n", - "1/8, train_loss: 0.0186 step time: 0.2969\n", - "2/8, train_loss: 0.0196 step time: 0.2519\n", - "3/8, train_loss: 0.0212 step time: 0.2500\n", - "4/8, train_loss: 0.0252 step time: 0.2415\n", - "5/8, train_loss: 0.0225 step time: 0.2373\n", - "6/8, train_loss: 0.0230 step time: 0.2382\n", - "7/8, train_loss: 0.0212 step time: 0.2277\n", - "8/8, train_loss: 0.0417 step time: 0.2294\n", - "epoch 161 average loss: 0.0241\n", - "time consuming of epoch 161 is: 1.9741\n", - "----------\n", - "epoch 162/600\n", - "1/8, train_loss: 0.0197 step time: 0.2843\n", - "2/8, train_loss: 0.0233 step time: 0.2471\n", - "3/8, train_loss: 0.0310 step time: 0.2827\n", - "4/8, train_loss: 0.0212 step time: 0.2658\n", - "5/8, train_loss: 0.0225 step time: 0.2538\n", - "6/8, train_loss: 0.0183 step time: 0.2406\n", - "7/8, train_loss: 0.0454 step time: 0.2343\n", - "8/8, train_loss: 0.0217 step time: 0.2321\n", - "epoch 162 average loss: 0.0254\n", - "time consuming of epoch 162 is: 2.0421\n", - "----------\n", - "epoch 163/600\n", - "1/8, train_loss: 0.0188 step time: 0.2879\n", - "2/8, train_loss: 0.0225 step time: 0.2323\n", - "3/8, train_loss: 0.0372 step time: 0.3098\n", - "4/8, train_loss: 0.0234 step time: 0.3268\n", - "5/8, train_loss: 0.0397 step time: 0.2984\n", - "6/8, train_loss: 0.0481 step time: 0.2936\n", - "7/8, train_loss: 0.0245 step time: 0.2780\n", - "8/8, train_loss: 0.0279 step time: 0.2593\n", - "epoch 163 average loss: 0.0303\n", - "time consuming of epoch 163 is: 2.2876\n", - "----------\n", - "epoch 164/600\n", - "1/8, train_loss: 0.0218 step time: 0.3042\n", - "2/8, train_loss: 0.0279 step time: 0.2497\n", - "3/8, train_loss: 0.0235 step time: 0.2509\n", - "4/8, train_loss: 0.0274 step time: 0.2495\n", - "5/8, train_loss: 0.0414 step time: 0.2418\n", - "6/8, train_loss: 0.0170 step time: 0.2381\n", - "7/8, train_loss: 0.0209 step time: 0.2311\n", - "8/8, train_loss: 0.0440 step time: 0.2422\n", - "epoch 164 average loss: 0.0280\n", - "time consuming of epoch 164 is: 2.0089\n", - "----------\n", - "epoch 165/600\n", - "1/8, train_loss: 0.0242 step time: 0.2959\n", - "2/8, train_loss: 0.0278 step time: 0.2412\n", - "3/8, train_loss: 0.0246 step time: 0.2380\n", - "4/8, train_loss: 0.0279 step time: 0.2333\n", - "5/8, train_loss: 0.0258 step time: 0.2348\n", - "6/8, train_loss: 0.0816 step time: 0.2363\n", - "7/8, train_loss: 0.0208 step time: 0.2324\n", - "8/8, train_loss: 0.0205 step time: 0.2429\n", - "epoch 165 average loss: 0.0316\n", - "current epoch: 165 current mean dice: 0.8847 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 165 is: 3.2058\n", - "----------\n", - "epoch 166/600\n", - "1/8, train_loss: 0.0211 step time: 0.3104\n", - "2/8, train_loss: 0.0354 step time: 0.2633\n", - "3/8, train_loss: 0.0271 step time: 0.2558\n", - "4/8, train_loss: 0.0229 step time: 0.2486\n", - "5/8, train_loss: 0.0472 step time: 0.2471\n", - "6/8, train_loss: 0.0257 step time: 0.2406\n", - "7/8, train_loss: 0.0270 step time: 0.2303\n", - "8/8, train_loss: 0.0246 step time: 0.2275\n", - "epoch 166 average loss: 0.0289\n", - "time consuming of epoch 166 is: 2.0250\n", - "----------\n", - "epoch 167/600\n", - "1/8, train_loss: 0.0265 step time: 0.5427\n", - "2/8, train_loss: 0.0294 step time: 0.2162\n", - "3/8, train_loss: 0.0436 step time: 0.2245\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "4/8, train_loss: 0.0251 step time: 0.2375\n", - "5/8, train_loss: 0.0256 step time: 0.2377\n", - "6/8, train_loss: 0.0314 step time: 0.2338\n", - "7/8, train_loss: 0.0255 step time: 0.2285\n", - "8/8, train_loss: 0.0233 step time: 0.2531\n", - "epoch 167 average loss: 0.0288\n", - "time consuming of epoch 167 is: 2.1755\n", - "----------\n", - "epoch 168/600\n", - "1/8, train_loss: 0.0260 step time: 0.3011\n", - "2/8, train_loss: 0.0221 step time: 0.2569\n", - "3/8, train_loss: 0.0297 step time: 0.2618\n", - "4/8, train_loss: 0.0245 step time: 0.2788\n", - "5/8, train_loss: 0.0356 step time: 0.2793\n", - "6/8, train_loss: 0.0343 step time: 0.2673\n", - "7/8, train_loss: 0.0266 step time: 0.2555\n", - "8/8, train_loss: 0.0189 step time: 0.2384\n", - "epoch 168 average loss: 0.0272\n", - "time consuming of epoch 168 is: 2.1406\n", - "----------\n", - "epoch 169/600\n", - "1/8, train_loss: 0.0234 step time: 0.2891\n", - "2/8, train_loss: 0.0229 step time: 0.2429\n", - "3/8, train_loss: 0.0242 step time: 0.2404\n", - "4/8, train_loss: 0.0240 step time: 0.2381\n", - "5/8, train_loss: 0.0227 step time: 0.2401\n", - "6/8, train_loss: 0.0205 step time: 0.2616\n", - "7/8, train_loss: 0.0202 step time: 0.3767\n", - "8/8, train_loss: 0.0216 step time: 0.5020\n", - "epoch 169 average loss: 0.0225\n", - "time consuming of epoch 169 is: 2.3924\n", - "----------\n", - "epoch 170/600\n", - "1/8, train_loss: 0.0215 step time: 0.3012\n", - "2/8, train_loss: 0.0210 step time: 0.2553\n", - "3/8, train_loss: 0.0222 step time: 0.2391\n", - "4/8, train_loss: 0.0177 step time: 0.2395\n", - "5/8, train_loss: 0.0252 step time: 0.2324\n", - "6/8, train_loss: 0.0197 step time: 0.3642\n", - "7/8, train_loss: 0.0197 step time: 0.4686\n", - "8/8, train_loss: 0.0192 step time: 0.2604\n", - "epoch 170 average loss: 0.0208\n", - "current epoch: 170 current mean dice: 0.9454 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 170 is: 3.5385\n", - "----------\n", - "epoch 171/600\n", - "1/8, train_loss: 0.0207 step time: 0.3184\n", - "2/8, train_loss: 0.0224 step time: 0.2674\n", - "3/8, train_loss: 0.0183 step time: 0.2468\n", - "4/8, train_loss: 0.0191 step time: 0.2421\n", - "5/8, train_loss: 0.0207 step time: 0.2400\n", - "6/8, train_loss: 0.0310 step time: 0.3741\n", - "7/8, train_loss: 0.0199 step time: 0.5432\n", - "8/8, train_loss: 0.0185 step time: 0.2021\n", - "epoch 171 average loss: 0.0213\n", - "time consuming of epoch 171 is: 2.4353\n", - "----------\n", - "epoch 172/600\n", - "1/8, train_loss: 0.0227 step time: 0.2634\n", - "2/8, train_loss: 0.0191 step time: 0.2143\n", - "3/8, train_loss: 0.0235 step time: 0.2330\n", - "4/8, train_loss: 0.0332 step time: 0.2387\n", - "5/8, train_loss: 0.0178 step time: 0.2357\n", - "6/8, train_loss: 0.0204 step time: 0.2312\n", - "7/8, train_loss: 0.0204 step time: 0.2812\n", - "8/8, train_loss: 0.0236 step time: 0.3108\n", - "epoch 172 average loss: 0.0226\n", - "time consuming of epoch 172 is: 2.0099\n", - "----------\n", - "epoch 173/600\n", - "1/8, train_loss: 0.0220 step time: 0.3096\n", - "2/8, train_loss: 0.0192 step time: 0.2629\n", - "3/8, train_loss: 0.0188 step time: 0.2627\n", - "4/8, train_loss: 0.0214 step time: 0.2670\n", - "5/8, train_loss: 0.0178 step time: 0.2602\n", - "6/8, train_loss: 0.0219 step time: 0.2610\n", - "7/8, train_loss: 0.0205 step time: 0.2494\n", - "8/8, train_loss: 0.0172 step time: 0.2357\n", - "epoch 173 average loss: 0.0199\n", - "time consuming of epoch 173 is: 2.1102\n", - "----------\n", - "epoch 174/600\n", - "1/8, train_loss: 0.0179 step time: 0.2870\n", - "2/8, train_loss: 0.0210 step time: 0.2391\n", - "3/8, train_loss: 0.0191 step time: 0.2368\n", - "4/8, train_loss: 0.0165 step time: 0.2350\n", - "5/8, train_loss: 0.0249 step time: 0.3245\n", - "6/8, train_loss: 0.0201 step time: 0.4838\n", - "7/8, train_loss: 0.0211 step time: 0.2674\n", - "8/8, train_loss: 0.0237 step time: 0.2636\n", - "epoch 174 average loss: 0.0205\n", - "time consuming of epoch 174 is: 2.3386\n", - "----------\n", - "epoch 175/600\n", - "1/8, train_loss: 0.0206 step time: 0.2998\n", - "2/8, train_loss: 0.0154 step time: 0.2486\n", - "3/8, train_loss: 0.0198 step time: 0.2419\n", - "4/8, train_loss: 0.0206 step time: 0.2320\n", - "5/8, train_loss: 0.0170 step time: 0.2328\n", - "6/8, train_loss: 0.0182 step time: 0.2303\n", - "7/8, train_loss: 0.0252 step time: 0.2255\n", - "8/8, train_loss: 0.0254 step time: 0.3328\n", - "epoch 175 average loss: 0.0203\n", - "current epoch: 175 current mean dice: 0.9476 best mean dice: 0.9482 at epoch: 85\n", - "time consuming of epoch 175 is: 3.3557\n", - "----------\n", - "epoch 176/600\n", - "1/8, train_loss: 0.0172 step time: 0.2848\n", - "2/8, train_loss: 0.0231 step time: 0.3700\n", - "3/8, train_loss: 0.0234 step time: 0.3314\n", - "4/8, train_loss: 0.0198 step time: 0.2648\n", - "5/8, train_loss: 0.0223 step time: 0.2491\n", - "6/8, train_loss: 0.0179 step time: 0.2579\n", - "7/8, train_loss: 0.0183 step time: 0.2632\n", - "8/8, train_loss: 0.0197 step time: 0.2486\n", - "epoch 176 average loss: 0.0202\n", - "time consuming of epoch 176 is: 2.2711\n", - "----------\n", - "epoch 177/600\n", - "1/8, train_loss: 0.0205 step time: 0.2930\n", - "2/8, train_loss: 0.0198 step time: 0.2364\n", - "3/8, train_loss: 0.0198 step time: 0.2359\n", - "4/8, train_loss: 0.0193 step time: 0.2319\n", - "5/8, train_loss: 0.0182 step time: 0.2347\n", - "6/8, train_loss: 0.0223 step time: 0.2347\n", - "7/8, train_loss: 0.0183 step time: 0.2254\n", - "8/8, train_loss: 0.0172 step time: 0.2318\n", - "epoch 177 average loss: 0.0194\n", - "time consuming of epoch 177 is: 1.9252\n", - "----------\n", - "epoch 178/600\n", - "1/8, train_loss: 0.0194 step time: 0.2929\n", - "2/8, train_loss: 0.0187 step time: 0.2520\n", - "3/8, train_loss: 0.0192 step time: 0.2539\n", - "4/8, train_loss: 0.0169 step time: 0.2576\n", - "5/8, train_loss: 0.0192 step time: 0.2503\n", - "6/8, train_loss: 0.0203 step time: 0.2500\n", - "7/8, train_loss: 0.0187 step time: 0.2442\n", - "8/8, train_loss: 0.0170 step time: 0.2411\n", - "epoch 178 average loss: 0.0187\n", - "time consuming of epoch 178 is: 2.0436\n", - "----------\n", - "epoch 179/600\n", - "1/8, train_loss: 0.0220 step time: 0.2719\n", - "2/8, train_loss: 0.0172 step time: 0.2361\n", - "3/8, train_loss: 0.0172 step time: 0.2313\n", - "4/8, train_loss: 0.0170 step time: 0.3401\n", - "5/8, train_loss: 0.0190 step time: 0.5083\n", - "6/8, train_loss: 0.0188 step time: 0.2738\n", - "7/8, train_loss: 0.0213 step time: 0.2551\n", - "8/8, train_loss: 0.0217 step time: 0.2439\n", - "epoch 179 average loss: 0.0193\n", - "time consuming of epoch 179 is: 2.3620\n", - "----------\n", - "epoch 180/600\n", - "1/8, train_loss: 0.0212 step time: 0.2961\n", - "2/8, train_loss: 0.0210 step time: 0.2526\n", - "3/8, train_loss: 0.0196 step time: 0.2526\n", - "4/8, train_loss: 0.0247 step time: 0.2531\n", - "5/8, train_loss: 0.0150 step time: 0.2432\n", - "6/8, train_loss: 0.0175 step time: 0.2448\n", - "7/8, train_loss: 0.0203 step time: 0.2391\n", - "8/8, train_loss: 0.0187 step time: 0.2327\n", - "epoch 180 average loss: 0.0197\n", - "saved new best metric model\n", - "current epoch: 180 current mean dice: 0.9485 best mean dice: 0.9485 at epoch: 180\n", - "time consuming of epoch 180 is: 5.9023\n", - "----------\n", - "epoch 181/600\n", - "1/8, train_loss: 0.0204 step time: 0.2773\n", - "2/8, train_loss: 0.0207 step time: 0.2124\n", - "3/8, train_loss: 0.0226 step time: 0.2155\n", - "4/8, train_loss: 0.0160 step time: 0.2172\n", - "5/8, train_loss: 0.0184 step time: 0.2161\n", - "6/8, train_loss: 0.0167 step time: 0.2226\n", - "7/8, train_loss: 0.0230 step time: 0.2158\n", - "8/8, train_loss: 0.0192 step time: 0.2149\n", - "epoch 181 average loss: 0.0196\n", - "time consuming of epoch 181 is: 1.7943\n", - "----------\n", - "epoch 182/600\n", - "1/8, train_loss: 0.0191 step time: 0.2755\n", - "2/8, train_loss: 0.0186 step time: 0.2259\n", - "3/8, train_loss: 0.0185 step time: 0.2262\n", - "4/8, train_loss: 0.0181 step time: 0.2245\n", - "5/8, train_loss: 0.0211 step time: 0.2247\n", - "6/8, train_loss: 0.0203 step time: 0.2240\n", - "7/8, train_loss: 0.0181 step time: 0.2072\n", - "8/8, train_loss: 0.0180 step time: 0.2187\n", - "epoch 182 average loss: 0.0190\n", - "time consuming of epoch 182 is: 1.8283\n", - "----------\n", - "epoch 183/600\n", - "1/8, train_loss: 0.0181 step time: 0.2943\n", - "2/8, train_loss: 0.0254 step time: 0.2459\n", - "3/8, train_loss: 0.0192 step time: 0.2396\n", - "4/8, train_loss: 0.0203 step time: 0.2375\n", - "5/8, train_loss: 0.0176 step time: 0.2294\n", - "6/8, train_loss: 0.0180 step time: 0.3321\n", - "7/8, train_loss: 0.0195 step time: 0.4235\n", - "8/8, train_loss: 0.0181 step time: 0.2668\n", - "epoch 183 average loss: 0.0195\n", - "time consuming of epoch 183 is: 2.2706\n", - "----------\n", - "epoch 184/600\n", - "1/8, train_loss: 0.0177 step time: 0.2903\n", - "2/8, train_loss: 0.0182 step time: 0.2352\n", - "3/8, train_loss: 0.0178 step time: 0.2336\n", - "4/8, train_loss: 0.0193 step time: 0.2390\n", - "5/8, train_loss: 0.0226 step time: 0.2369\n", - "6/8, train_loss: 0.0185 step time: 0.2496\n", - "7/8, train_loss: 0.0191 step time: 0.2908\n", - "8/8, train_loss: 0.0177 step time: 0.2839\n", - "epoch 184 average loss: 0.0189\n", - "time consuming of epoch 184 is: 2.0607\n", - "----------\n", - "epoch 185/600\n", - "1/8, train_loss: 0.0211 step time: 0.3040\n", - "2/8, train_loss: 0.0228 step time: 0.2631\n", - "3/8, train_loss: 0.0188 step time: 0.2527\n", - "4/8, train_loss: 0.0178 step time: 0.2477\n", - "5/8, train_loss: 0.0179 step time: 0.2535\n", - "6/8, train_loss: 0.0207 step time: 0.2410\n", - "7/8, train_loss: 0.0157 step time: 0.2336\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "8/8, train_loss: 0.0200 step time: 0.2310\n", - "epoch 185 average loss: 0.0194\n", - "saved new best metric model\n", - "current epoch: 185 current mean dice: 0.9491 best mean dice: 0.9491 at epoch: 185\n", - "time consuming of epoch 185 is: 3.9216\n", - "----------\n", - "epoch 186/600\n", - "1/8, train_loss: 0.0146 step time: 0.2700\n", - "2/8, train_loss: 0.0174 step time: 0.2123\n", - "3/8, train_loss: 0.0191 step time: 0.2170\n", - "4/8, train_loss: 0.0185 step time: 0.2514\n", - "5/8, train_loss: 0.0220 step time: 0.2464\n", - "6/8, train_loss: 0.0255 step time: 0.2551\n", - "7/8, train_loss: 0.0193 step time: 0.2429\n", - "8/8, train_loss: 0.0209 step time: 0.3733\n", - "epoch 186 average loss: 0.0197\n", - "time consuming of epoch 186 is: 2.0708\n", - "----------\n", - "epoch 187/600\n", - "1/8, train_loss: 0.0196 step time: 0.3326\n", - "2/8, train_loss: 0.0184 step time: 0.2166\n", - "3/8, train_loss: 0.0176 step time: 0.2135\n", - "4/8, train_loss: 0.0173 step time: 0.2725\n", - "5/8, train_loss: 0.0189 step time: 0.3467\n", - "6/8, train_loss: 0.0186 step time: 0.2724\n", - "7/8, train_loss: 0.0183 step time: 0.2451\n", - "8/8, train_loss: 0.0173 step time: 0.2744\n", - "epoch 187 average loss: 0.0182\n", - "time consuming of epoch 187 is: 2.1753\n", - "----------\n", - "epoch 188/600\n", - "1/8, train_loss: 0.0176 step time: 0.3402\n", - "2/8, train_loss: 0.0194 step time: 0.2621\n", - "3/8, train_loss: 0.0163 step time: 0.2472\n", - "4/8, train_loss: 0.0184 step time: 0.2425\n", - "5/8, train_loss: 0.0184 step time: 0.2406\n", - "6/8, train_loss: 0.0236 step time: 0.2375\n", - "7/8, train_loss: 0.0211 step time: 0.2284\n", - "8/8, train_loss: 0.0202 step time: 0.2251\n", - "epoch 188 average loss: 0.0194\n", - "time consuming of epoch 188 is: 2.0250\n", - "----------\n", - "epoch 189/600\n", - "1/8, train_loss: 0.0178 step time: 0.2813\n", - "2/8, train_loss: 0.0152 step time: 0.2356\n", - "3/8, train_loss: 0.0190 step time: 0.2313\n", - "4/8, train_loss: 0.0196 step time: 0.3860\n", - "5/8, train_loss: 0.0176 step time: 0.6015\n", - "6/8, train_loss: 0.0189 step time: 0.2198\n", - "7/8, train_loss: 0.0193 step time: 0.1999\n", - "8/8, train_loss: 0.0167 step time: 0.2031\n", - "epoch 189 average loss: 0.0180\n", - "time consuming of epoch 189 is: 2.3600\n", - "----------\n", - "epoch 190/600\n", - "1/8, train_loss: 0.0196 step time: 0.3176\n", - "2/8, train_loss: 0.0165 step time: 0.2490\n", - "3/8, train_loss: 0.0178 step time: 0.2473\n", - "4/8, train_loss: 0.0163 step time: 0.2499\n", - "5/8, train_loss: 0.0171 step time: 0.2910\n", - "6/8, train_loss: 0.0236 step time: 0.3832\n", - "7/8, train_loss: 0.0175 step time: 0.2640\n", - "8/8, train_loss: 0.0201 step time: 0.2600\n", - "epoch 190 average loss: 0.0186\n", - "current epoch: 190 current mean dice: 0.9481 best mean dice: 0.9491 at epoch: 185\n", - "time consuming of epoch 190 is: 3.6325\n", - "----------\n", - "epoch 191/600\n", - "1/8, train_loss: 0.0155 step time: 0.2996\n", - "2/8, train_loss: 0.0167 step time: 0.2488\n", - "3/8, train_loss: 0.0172 step time: 0.2481\n", - "4/8, train_loss: 0.0184 step time: 0.2413\n", - "5/8, train_loss: 0.0181 step time: 0.2403\n", - "6/8, train_loss: 0.0206 step time: 0.3334\n", - "7/8, train_loss: 0.0187 step time: 0.3205\n", - "8/8, train_loss: 0.0188 step time: 0.2713\n", - "epoch 191 average loss: 0.0180\n", - "time consuming of epoch 191 is: 2.2047\n", - "----------\n", - "epoch 192/600\n", - "1/8, train_loss: 0.0150 step time: 0.2986\n", - "2/8, train_loss: 0.0177 step time: 0.2443\n", - "3/8, train_loss: 0.0155 step time: 0.2406\n", - "4/8, train_loss: 0.0216 step time: 0.2556\n", - "5/8, train_loss: 0.0208 step time: 0.2543\n", - "6/8, train_loss: 0.0203 step time: 0.2464\n", - "7/8, train_loss: 0.0187 step time: 0.2304\n", - "8/8, train_loss: 0.0162 step time: 0.2241\n", - "epoch 192 average loss: 0.0182\n", - "time consuming of epoch 192 is: 1.9957\n", - "----------\n", - "epoch 193/600\n", - "1/8, train_loss: 0.0155 step time: 0.2807\n", - "2/8, train_loss: 0.0216 step time: 0.2395\n", - "3/8, train_loss: 0.0192 step time: 0.2652\n", - "4/8, train_loss: 0.0180 step time: 0.2484\n", - "5/8, train_loss: 0.0189 step time: 0.2504\n", - "6/8, train_loss: 0.0165 step time: 0.2434\n", - "7/8, train_loss: 0.0200 step time: 0.2273\n", - "8/8, train_loss: 0.0164 step time: 0.2290\n", - "epoch 193 average loss: 0.0183\n", - "time consuming of epoch 193 is: 1.9854\n", - "----------\n", - "epoch 194/600\n", - "1/8, train_loss: 0.0219 step time: 0.2828\n", - "2/8, train_loss: 0.0160 step time: 0.2368\n", - "3/8, train_loss: 0.0180 step time: 0.2331\n", - "4/8, train_loss: 0.0202 step time: 0.2305\n", - "5/8, train_loss: 0.0227 step time: 0.2352\n", - "6/8, train_loss: 0.0161 step time: 0.2340\n", - "7/8, train_loss: 0.0172 step time: 0.2731\n", - "8/8, train_loss: 0.0198 step time: 0.3164\n", - "epoch 194 average loss: 0.0190\n", - "time consuming of epoch 194 is: 2.0433\n", - "----------\n", - "epoch 195/600\n", - "1/8, train_loss: 0.0210 step time: 0.3177\n", - "2/8, train_loss: 0.0179 step time: 0.2551\n", - "3/8, train_loss: 0.0163 step time: 0.2602\n", - "4/8, train_loss: 0.0167 step time: 0.2479\n", - "5/8, train_loss: 0.0215 step time: 0.2478\n", - "6/8, train_loss: 0.0194 step time: 0.2613\n", - "7/8, train_loss: 0.0166 step time: 0.2726\n", - "8/8, train_loss: 0.0237 step time: 0.2564\n", - "epoch 195 average loss: 0.0191\n", - "saved new best metric model\n", - "current epoch: 195 current mean dice: 0.9503 best mean dice: 0.9503 at epoch: 195\n", - "time consuming of epoch 195 is: 3.7715\n", - "----------\n", - "epoch 196/600\n", - "1/8, train_loss: 0.0152 step time: 0.2774\n", - "2/8, train_loss: 0.0173 step time: 0.2143\n", - "3/8, train_loss: 0.0215 step time: 0.2228\n", - "4/8, train_loss: 0.0185 step time: 0.2552\n", - "5/8, train_loss: 0.0217 step time: 0.2449\n", - "6/8, train_loss: 0.0178 step time: 0.2473\n", - "7/8, train_loss: 0.0186 step time: 0.2382\n", - "8/8, train_loss: 0.0192 step time: 0.4247\n", - "epoch 196 average loss: 0.0187\n", - "time consuming of epoch 196 is: 2.1274\n", - "----------\n", - "epoch 197/600\n", - "1/8, train_loss: 0.0196 step time: 0.5390\n", - "2/8, train_loss: 0.0189 step time: 0.2146\n", - "3/8, train_loss: 0.0160 step time: 0.2136\n", - "4/8, train_loss: 0.0200 step time: 0.2197\n", - "5/8, train_loss: 0.0188 step time: 0.2407\n", - "6/8, train_loss: 0.0142 step time: 0.2500\n", - "7/8, train_loss: 0.0213 step time: 0.2517\n", - "8/8, train_loss: 0.0190 step time: 0.4503\n", - "epoch 197 average loss: 0.0185\n", - "time consuming of epoch 197 is: 2.3811\n", - "----------\n", - "epoch 198/600\n", - "1/8, train_loss: 0.0231 step time: 0.3474\n", - "2/8, train_loss: 0.0175 step time: 0.2181\n", - "3/8, train_loss: 0.0154 step time: 0.2168\n", - "4/8, train_loss: 0.0175 step time: 0.2170\n", - "5/8, train_loss: 0.0185 step time: 0.2579\n", - "6/8, train_loss: 0.0188 step time: 0.3679\n", - "7/8, train_loss: 0.0199 step time: 0.3204\n", - "8/8, train_loss: 0.0161 step time: 0.2803\n", - "epoch 198 average loss: 0.0183\n", - "time consuming of epoch 198 is: 2.2272\n", - "----------\n", - "epoch 199/600\n", - "1/8, train_loss: 0.0176 step time: 0.3175\n", - "2/8, train_loss: 0.0158 step time: 0.2807\n", - "3/8, train_loss: 0.0207 step time: 0.2600\n", - "4/8, train_loss: 0.0164 step time: 0.2627\n", - "5/8, train_loss: 0.0192 step time: 0.2530\n", - "6/8, train_loss: 0.0179 step time: 0.4336\n", - "7/8, train_loss: 0.0188 step time: 0.4260\n", - "8/8, train_loss: 0.0181 step time: 0.2600\n", - "epoch 199 average loss: 0.0181\n", - "time consuming of epoch 199 is: 2.4949\n", - "----------\n", - "epoch 200/600\n", - "1/8, train_loss: 0.0188 step time: 0.2941\n", - "2/8, train_loss: 0.0226 step time: 0.2405\n", - "3/8, train_loss: 0.0142 step time: 0.2394\n", - "4/8, train_loss: 0.0141 step time: 0.2685\n", - "5/8, train_loss: 0.0188 step time: 0.2979\n", - "6/8, train_loss: 0.0214 step time: 0.2978\n", - "7/8, train_loss: 0.0161 step time: 0.2977\n", - "8/8, train_loss: 0.0180 step time: 0.2540\n", - "epoch 200 average loss: 0.0180\n", - "saved new best metric model\n", - "current epoch: 200 current mean dice: 0.9505 best mean dice: 0.9505 at epoch: 200\n", - "time consuming of epoch 200 is: 3.6138\n", - "----------\n", - "epoch 201/600\n", - "1/8, train_loss: 0.0207 step time: 0.2754\n", - "2/8, train_loss: 0.0188 step time: 0.2102\n", - "3/8, train_loss: 0.0243 step time: 0.2149\n", - "4/8, train_loss: 0.0172 step time: 0.2318\n", - "5/8, train_loss: 0.0169 step time: 0.2402\n", - "6/8, train_loss: 0.0146 step time: 0.2371\n", - "7/8, train_loss: 0.0179 step time: 0.2475\n", - "8/8, train_loss: 0.0167 step time: 0.2765\n", - "epoch 201 average loss: 0.0184\n", - "time consuming of epoch 201 is: 1.9361\n", - "----------\n", - "epoch 202/600\n", - "1/8, train_loss: 0.0180 step time: 0.3037\n", - "2/8, train_loss: 0.0167 step time: 0.2490\n", - "3/8, train_loss: 0.0186 step time: 0.2464\n", - "4/8, train_loss: 0.0205 step time: 0.2504\n", - "5/8, train_loss: 0.0181 step time: 0.3196\n", - "6/8, train_loss: 0.0197 step time: 0.4920\n", - "7/8, train_loss: 0.0214 step time: 0.2727\n", - "8/8, train_loss: 0.0184 step time: 0.2634\n", - "epoch 202 average loss: 0.0189\n", - "time consuming of epoch 202 is: 2.3987\n", - "----------\n", - "epoch 203/600\n", - "1/8, train_loss: 0.0172 step time: 0.3089\n", - "2/8, train_loss: 0.0183 step time: 0.2598\n", - "3/8, train_loss: 0.0237 step time: 0.2640\n", - "4/8, train_loss: 0.0200 step time: 0.2578\n", - "5/8, train_loss: 0.0194 step time: 0.2452\n", - "6/8, train_loss: 0.0171 step time: 0.2436\n", - "7/8, train_loss: 0.0170 step time: 0.2322\n", - "8/8, train_loss: 0.0185 step time: 0.2244\n", - "epoch 203 average loss: 0.0189\n", - "time consuming of epoch 203 is: 2.0373\n", - "----------\n", - "epoch 204/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0204 step time: 0.6395\n", - "2/8, train_loss: 0.0157 step time: 0.2181\n", - "3/8, train_loss: 0.0187 step time: 0.2167\n", - "4/8, train_loss: 0.0159 step time: 0.2137\n", - "5/8, train_loss: 0.0167 step time: 0.2185\n", - "6/8, train_loss: 0.0158 step time: 0.2524\n", - "7/8, train_loss: 0.0184 step time: 0.2762\n", - "8/8, train_loss: 0.0192 step time: 0.2544\n", - "epoch 204 average loss: 0.0176\n", - "time consuming of epoch 204 is: 2.2911\n", - "----------\n", - "epoch 205/600\n", - "1/8, train_loss: 0.0181 step time: 0.3003\n", - "2/8, train_loss: 0.0224 step time: 0.2888\n", - "3/8, train_loss: 0.0147 step time: 0.2984\n", - "4/8, train_loss: 0.0186 step time: 0.2902\n", - "5/8, train_loss: 0.0181 step time: 0.2889\n", - "6/8, train_loss: 0.0158 step time: 0.2817\n", - "7/8, train_loss: 0.0197 step time: 0.2831\n", - "8/8, train_loss: 0.0181 step time: 0.2819\n", - "epoch 205 average loss: 0.0182\n", - "saved new best metric model\n", - "current epoch: 205 current mean dice: 0.9512 best mean dice: 0.9512 at epoch: 205\n", - "time consuming of epoch 205 is: 3.9843\n", - "----------\n", - "epoch 206/600\n", - "1/8, train_loss: 0.0132 step time: 0.2817\n", - "2/8, train_loss: 0.0194 step time: 0.2169\n", - "3/8, train_loss: 0.0232 step time: 0.2142\n", - "4/8, train_loss: 0.0167 step time: 0.2128\n", - "5/8, train_loss: 0.0183 step time: 0.2233\n", - "6/8, train_loss: 0.0192 step time: 0.2298\n", - "7/8, train_loss: 0.0153 step time: 0.2488\n", - "8/8, train_loss: 0.0280 step time: 0.2830\n", - "epoch 206 average loss: 0.0192\n", - "time consuming of epoch 206 is: 1.9130\n", - "----------\n", - "epoch 207/600\n", - "1/8, train_loss: 0.0146 step time: 0.3288\n", - "2/8, train_loss: 0.0230 step time: 0.2632\n", - "3/8, train_loss: 0.0214 step time: 0.2478\n", - "4/8, train_loss: 0.0159 step time: 0.2462\n", - "5/8, train_loss: 0.0189 step time: 0.2439\n", - "6/8, train_loss: 0.0160 step time: 0.2406\n", - "7/8, train_loss: 0.0176 step time: 0.2285\n", - "8/8, train_loss: 0.0176 step time: 0.2704\n", - "epoch 207 average loss: 0.0181\n", - "time consuming of epoch 207 is: 2.0710\n", - "----------\n", - "epoch 208/600\n", - "1/8, train_loss: 0.0169 step time: 0.3138\n", - "2/8, train_loss: 0.0191 step time: 0.2663\n", - "3/8, train_loss: 0.0209 step time: 0.2587\n", - "4/8, train_loss: 0.0205 step time: 0.2532\n", - "5/8, train_loss: 0.0160 step time: 0.4429\n", - "6/8, train_loss: 0.0221 step time: 0.6164\n", - "7/8, train_loss: 0.0178 step time: 0.2041\n", - "8/8, train_loss: 0.0170 step time: 0.2436\n", - "epoch 208 average loss: 0.0188\n", - "time consuming of epoch 208 is: 2.6004\n", - "----------\n", - "epoch 209/600\n", - "1/8, train_loss: 0.0186 step time: 0.3470\n", - "2/8, train_loss: 0.0231 step time: 0.2752\n", - "3/8, train_loss: 0.0154 step time: 0.2653\n", - "4/8, train_loss: 0.0174 step time: 0.2553\n", - "5/8, train_loss: 0.0165 step time: 0.2555\n", - "6/8, train_loss: 0.0194 step time: 0.2444\n", - "7/8, train_loss: 0.0160 step time: 0.2388\n", - "8/8, train_loss: 0.0185 step time: 0.4021\n", - "epoch 209 average loss: 0.0181\n", - "time consuming of epoch 209 is: 2.2850\n", - "----------\n", - "epoch 210/600\n", - "1/8, train_loss: 0.0198 step time: 0.5640\n", - "2/8, train_loss: 0.0200 step time: 0.2927\n", - "3/8, train_loss: 0.0226 step time: 0.2801\n", - "4/8, train_loss: 0.0153 step time: 0.2575\n", - "5/8, train_loss: 0.0172 step time: 0.4163\n", - "6/8, train_loss: 0.0198 step time: 0.3339\n", - "7/8, train_loss: 0.0149 step time: 0.2660\n", - "8/8, train_loss: 0.0171 step time: 0.2661\n", - "epoch 210 average loss: 0.0183\n", - "current epoch: 210 current mean dice: 0.9496 best mean dice: 0.9512 at epoch: 205\n", - "time consuming of epoch 210 is: 3.9462\n", - "----------\n", - "epoch 211/600\n", - "1/8, train_loss: 0.0165 step time: 0.3327\n", - "2/8, train_loss: 0.0184 step time: 0.2584\n", - "3/8, train_loss: 0.0191 step time: 0.2543\n", - "4/8, train_loss: 0.0161 step time: 0.2462\n", - "5/8, train_loss: 0.0154 step time: 0.2437\n", - "6/8, train_loss: 0.0175 step time: 0.2370\n", - "7/8, train_loss: 0.0204 step time: 0.2361\n", - "8/8, train_loss: 0.0282 step time: 0.2549\n", - "epoch 211 average loss: 0.0189\n", - "time consuming of epoch 211 is: 2.0644\n", - "----------\n", - "epoch 212/600\n", - "1/8, train_loss: 0.0189 step time: 0.3485\n", - "2/8, train_loss: 0.0175 step time: 0.2841\n", - "3/8, train_loss: 0.0147 step time: 0.2692\n", - "4/8, train_loss: 0.0168 step time: 0.2609\n", - "5/8, train_loss: 0.0208 step time: 0.2541\n", - "6/8, train_loss: 0.0180 step time: 0.2502\n", - "7/8, train_loss: 0.0184 step time: 0.3909\n", - "8/8, train_loss: 0.0167 step time: 0.6285\n", - "epoch 212 average loss: 0.0177\n", - "time consuming of epoch 212 is: 2.6878\n", - "----------\n", - "epoch 213/600\n", - "1/8, train_loss: 0.0196 step time: 0.3164\n", - "2/8, train_loss: 0.0192 step time: 0.2563\n", - "3/8, train_loss: 0.0184 step time: 0.2587\n", - "4/8, train_loss: 0.0151 step time: 0.2478\n", - "5/8, train_loss: 0.0149 step time: 0.2495\n", - "6/8, train_loss: 0.0204 step time: 0.2389\n", - "7/8, train_loss: 0.0242 step time: 0.2313\n", - "8/8, train_loss: 0.0188 step time: 0.2213\n", - "epoch 213 average loss: 0.0188\n", - "time consuming of epoch 213 is: 2.0218\n", - "----------\n", - "epoch 214/600\n", - "1/8, train_loss: 0.0211 step time: 0.2855\n", - "2/8, train_loss: 0.0190 step time: 0.2399\n", - "3/8, train_loss: 0.0142 step time: 0.2356\n", - "4/8, train_loss: 0.0176 step time: 0.2329\n", - "5/8, train_loss: 0.0159 step time: 0.2299\n", - "6/8, train_loss: 0.0176 step time: 0.2385\n", - "7/8, train_loss: 0.0203 step time: 0.2422\n", - "8/8, train_loss: 0.0154 step time: 0.2486\n", - "epoch 214 average loss: 0.0176\n", - "time consuming of epoch 214 is: 1.9547\n", - "----------\n", - "epoch 215/600\n", - "1/8, train_loss: 0.0180 step time: 0.2918\n", - "2/8, train_loss: 0.0173 step time: 0.2439\n", - "3/8, train_loss: 0.0175 step time: 0.2473\n", - "4/8, train_loss: 0.0172 step time: 0.2629\n", - "5/8, train_loss: 0.0160 step time: 0.2576\n", - "6/8, train_loss: 0.0195 step time: 0.2763\n", - "7/8, train_loss: 0.0172 step time: 0.2765\n", - "8/8, train_loss: 0.0185 step time: 0.2883\n", - "epoch 215 average loss: 0.0177\n", - "saved new best metric model\n", - "current epoch: 215 current mean dice: 0.9523 best mean dice: 0.9523 at epoch: 215\n", - "time consuming of epoch 215 is: 3.6422\n", - "----------\n", - "epoch 216/600\n", - "1/8, train_loss: 0.0229 step time: 0.2737\n", - "2/8, train_loss: 0.0181 step time: 0.2144\n", - "3/8, train_loss: 0.0158 step time: 0.2232\n", - "4/8, train_loss: 0.0153 step time: 0.2271\n", - "5/8, train_loss: 0.0132 step time: 0.2377\n", - "6/8, train_loss: 0.0192 step time: 0.4043\n", - "7/8, train_loss: 0.0193 step time: 0.6272\n", - "8/8, train_loss: 0.0165 step time: 0.2721\n", - "epoch 216 average loss: 0.0175\n", - "time consuming of epoch 216 is: 2.4822\n", - "----------\n", - "epoch 217/600\n", - "1/8, train_loss: 0.0166 step time: 0.2963\n", - "2/8, train_loss: 0.0150 step time: 0.2412\n", - "3/8, train_loss: 0.0247 step time: 0.2368\n", - "4/8, train_loss: 0.0178 step time: 0.2722\n", - "5/8, train_loss: 0.0214 step time: 0.3334\n", - "6/8, train_loss: 0.0183 step time: 0.2955\n", - "7/8, train_loss: 0.0160 step time: 0.2596\n", - "8/8, train_loss: 0.0162 step time: 0.2529\n", - "epoch 217 average loss: 0.0182\n", - "time consuming of epoch 217 is: 2.1895\n", - "----------\n", - "epoch 218/600\n", - "1/8, train_loss: 0.0206 step time: 0.3029\n", - "2/8, train_loss: 0.0201 step time: 0.4029\n", - "3/8, train_loss: 0.0181 step time: 0.5110\n", - "4/8, train_loss: 0.0177 step time: 0.2719\n", - "5/8, train_loss: 0.0147 step time: 0.2611\n", - "6/8, train_loss: 0.0179 step time: 0.2454\n", - "7/8, train_loss: 0.0166 step time: 0.2351\n", - "8/8, train_loss: 0.0166 step time: 0.2386\n", - "epoch 218 average loss: 0.0178\n", - "time consuming of epoch 218 is: 2.4704\n", - "----------\n", - "epoch 219/600\n", - "1/8, train_loss: 0.0208 step time: 0.2872\n", - "2/8, train_loss: 0.0158 step time: 0.2378\n", - "3/8, train_loss: 0.0127 step time: 0.2411\n", - "4/8, train_loss: 0.0164 step time: 0.2443\n", - "5/8, train_loss: 0.0182 step time: 0.2363\n", - "6/8, train_loss: 0.0178 step time: 0.2320\n", - "7/8, train_loss: 0.0170 step time: 0.2181\n", - "8/8, train_loss: 0.0217 step time: 0.2301\n", - "epoch 219 average loss: 0.0175\n", - "time consuming of epoch 219 is: 1.9284\n", - "----------\n", - "epoch 220/600\n", - "1/8, train_loss: 0.0185 step time: 0.3008\n", - "2/8, train_loss: 0.0174 step time: 0.2583\n", - "3/8, train_loss: 0.0179 step time: 0.2566\n", - "4/8, train_loss: 0.0179 step time: 0.2518\n", - "5/8, train_loss: 0.0181 step time: 0.2557\n", - "6/8, train_loss: 0.0175 step time: 0.2507\n", - "7/8, train_loss: 0.0216 step time: 0.2441\n", - "8/8, train_loss: 0.0177 step time: 0.2441\n", - "epoch 220 average loss: 0.0183\n", - "current epoch: 220 current mean dice: 0.9508 best mean dice: 0.9523 at epoch: 215\n", - "time consuming of epoch 220 is: 3.5726\n", - "----------\n", - "epoch 221/600\n", - "1/8, train_loss: 0.0193 step time: 0.3190\n", - "2/8, train_loss: 0.0182 step time: 0.2611\n", - "3/8, train_loss: 0.0164 step time: 0.2818\n", - "4/8, train_loss: 0.0196 step time: 0.2854\n", - "5/8, train_loss: 0.0184 step time: 0.2722\n", - "6/8, train_loss: 0.0194 step time: 0.4753\n", - "7/8, train_loss: 0.0153 step time: 0.3513\n", - "8/8, train_loss: 0.0166 step time: 0.2609\n", - "epoch 221 average loss: 0.0179\n", - "time consuming of epoch 221 is: 2.5082\n", - "----------\n", - "epoch 222/600\n", - "1/8, train_loss: 0.0160 step time: 0.2944\n", - "2/8, train_loss: 0.0184 step time: 0.2445\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "3/8, train_loss: 0.0187 step time: 0.2433\n", - "4/8, train_loss: 0.0170 step time: 0.2419\n", - "5/8, train_loss: 0.0144 step time: 0.2410\n", - "6/8, train_loss: 0.0189 step time: 0.3124\n", - "7/8, train_loss: 0.0182 step time: 0.3431\n", - "8/8, train_loss: 0.0229 step time: 0.3141\n", - "epoch 222 average loss: 0.0181\n", - "time consuming of epoch 222 is: 2.2361\n", - "----------\n", - "epoch 223/600\n", - "1/8, train_loss: 0.0193 step time: 0.3084\n", - "2/8, train_loss: 0.0188 step time: 0.2543\n", - "3/8, train_loss: 0.0159 step time: 0.2503\n", - "4/8, train_loss: 0.0217 step time: 0.2881\n", - "5/8, train_loss: 0.0194 step time: 0.3382\n", - "6/8, train_loss: 0.0190 step time: 0.2867\n", - "7/8, train_loss: 0.0180 step time: 0.2661\n", - "8/8, train_loss: 0.0156 step time: 0.2530\n", - "epoch 223 average loss: 0.0185\n", - "time consuming of epoch 223 is: 2.2467\n", - "----------\n", - "epoch 224/600\n", - "1/8, train_loss: 0.0176 step time: 0.2955\n", - "2/8, train_loss: 0.0201 step time: 0.3744\n", - "3/8, train_loss: 0.0193 step time: 0.4918\n", - "4/8, train_loss: 0.0160 step time: 0.2910\n", - "5/8, train_loss: 0.0166 step time: 0.2793\n", - "6/8, train_loss: 0.0152 step time: 0.2643\n", - "7/8, train_loss: 0.0223 step time: 0.2754\n", - "8/8, train_loss: 0.0193 step time: 0.2646\n", - "epoch 224 average loss: 0.0183\n", - "time consuming of epoch 224 is: 2.5376\n", - "----------\n", - "epoch 225/600\n", - "1/8, train_loss: 0.0138 step time: 0.3011\n", - "2/8, train_loss: 0.0194 step time: 0.2484\n", - "3/8, train_loss: 0.0175 step time: 0.2445\n", - "4/8, train_loss: 0.0174 step time: 0.2750\n", - "5/8, train_loss: 0.0200 step time: 0.2749\n", - "6/8, train_loss: 0.0259 step time: 0.2639\n", - "7/8, train_loss: 0.0187 step time: 0.2572\n", - "8/8, train_loss: 0.0191 step time: 0.2415\n", - "epoch 225 average loss: 0.0190\n", - "saved new best metric model\n", - "current epoch: 225 current mean dice: 0.9527 best mean dice: 0.9527 at epoch: 225\n", - "time consuming of epoch 225 is: 3.7523\n", - "----------\n", - "epoch 226/600\n", - "1/8, train_loss: 0.0157 step time: 0.2848\n", - "2/8, train_loss: 0.0158 step time: 0.2144\n", - "3/8, train_loss: 0.0184 step time: 0.2377\n", - "4/8, train_loss: 0.0169 step time: 0.3241\n", - "5/8, train_loss: 0.0237 step time: 0.3092\n", - "6/8, train_loss: 0.0194 step time: 0.2706\n", - "7/8, train_loss: 0.0167 step time: 0.2892\n", - "8/8, train_loss: 0.0167 step time: 0.2809\n", - "epoch 226 average loss: 0.0179\n", - "time consuming of epoch 226 is: 2.2134\n", - "----------\n", - "epoch 227/600\n", - "1/8, train_loss: 0.0189 step time: 0.2989\n", - "2/8, train_loss: 0.0266 step time: 0.2535\n", - "3/8, train_loss: 0.0174 step time: 0.2655\n", - "4/8, train_loss: 0.0194 step time: 0.3005\n", - "5/8, train_loss: 0.0178 step time: 0.2575\n", - "6/8, train_loss: 0.0159 step time: 0.4424\n", - "7/8, train_loss: 0.0165 step time: 0.4034\n", - "8/8, train_loss: 0.0174 step time: 0.2544\n", - "epoch 227 average loss: 0.0187\n", - "time consuming of epoch 227 is: 2.4776\n", - "----------\n", - "epoch 228/600\n", - "1/8, train_loss: 0.0171 step time: 0.2964\n", - "2/8, train_loss: 0.0231 step time: 0.2431\n", - "3/8, train_loss: 0.0153 step time: 0.2418\n", - "4/8, train_loss: 0.0176 step time: 0.2357\n", - "5/8, train_loss: 0.0234 step time: 0.2374\n", - "6/8, train_loss: 0.0158 step time: 0.2451\n", - "7/8, train_loss: 0.0152 step time: 0.2614\n", - "8/8, train_loss: 0.0155 step time: 0.2896\n", - "epoch 228 average loss: 0.0179\n", - "time consuming of epoch 228 is: 2.0521\n", - "----------\n", - "epoch 229/600\n", - "1/8, train_loss: 0.0176 step time: 0.3414\n", - "2/8, train_loss: 0.0203 step time: 0.2941\n", - "3/8, train_loss: 0.0184 step time: 0.2783\n", - "4/8, train_loss: 0.0171 step time: 0.2708\n", - "5/8, train_loss: 0.0208 step time: 0.2531\n", - "6/8, train_loss: 0.0134 step time: 0.2704\n", - "7/8, train_loss: 0.0147 step time: 0.2787\n", - "8/8, train_loss: 0.0226 step time: 0.5408\n", - "epoch 229 average loss: 0.0181\n", - "time consuming of epoch 229 is: 2.5290\n", - "----------\n", - "epoch 230/600\n", - "1/8, train_loss: 0.0182 step time: 0.3645\n", - "2/8, train_loss: 0.0195 step time: 0.2831\n", - "3/8, train_loss: 0.0220 step time: 0.2589\n", - "4/8, train_loss: 0.0165 step time: 0.2495\n", - "5/8, train_loss: 0.0169 step time: 0.2492\n", - "6/8, train_loss: 0.0168 step time: 0.2402\n", - "7/8, train_loss: 0.0162 step time: 0.3263\n", - "8/8, train_loss: 0.0202 step time: 0.4967\n", - "epoch 230 average loss: 0.0183\n", - "current epoch: 230 current mean dice: 0.9517 best mean dice: 0.9527 at epoch: 225\n", - "time consuming of epoch 230 is: 3.6626\n", - "----------\n", - "epoch 231/600\n", - "1/8, train_loss: 0.0174 step time: 0.3016\n", - "2/8, train_loss: 0.0182 step time: 0.2497\n", - "3/8, train_loss: 0.0159 step time: 0.2410\n", - "4/8, train_loss: 0.0135 step time: 0.2420\n", - "5/8, train_loss: 0.0197 step time: 0.2408\n", - "6/8, train_loss: 0.0145 step time: 0.2492\n", - "7/8, train_loss: 0.0226 step time: 0.2759\n", - "8/8, train_loss: 0.0176 step time: 0.3449\n", - "epoch 231 average loss: 0.0174\n", - "time consuming of epoch 231 is: 2.1463\n", - "----------\n", - "epoch 232/600\n", - "1/8, train_loss: 0.0154 step time: 0.3294\n", - "2/8, train_loss: 0.0182 step time: 0.2655\n", - "3/8, train_loss: 0.0217 step time: 0.2742\n", - "4/8, train_loss: 0.0164 step time: 0.2578\n", - "5/8, train_loss: 0.0170 step time: 0.2576\n", - "6/8, train_loss: 0.0184 step time: 0.2528\n", - "7/8, train_loss: 0.0141 step time: 0.2381\n", - "8/8, train_loss: 0.0214 step time: 0.2324\n", - "epoch 232 average loss: 0.0178\n", - "time consuming of epoch 232 is: 2.1092\n", - "----------\n", - "epoch 233/600\n", - "1/8, train_loss: 0.0221 step time: 0.2817\n", - "2/8, train_loss: 0.0162 step time: 0.2326\n", - "3/8, train_loss: 0.0204 step time: 0.2326\n", - "4/8, train_loss: 0.0179 step time: 0.2353\n", - "5/8, train_loss: 0.0174 step time: 0.2328\n", - "6/8, train_loss: 0.0162 step time: 0.2574\n", - "7/8, train_loss: 0.0173 step time: 0.2748\n", - "8/8, train_loss: 0.0143 step time: 0.2680\n", - "epoch 233 average loss: 0.0177\n", - "time consuming of epoch 233 is: 2.0167\n", - "----------\n", - "epoch 234/600\n", - "1/8, train_loss: 0.0155 step time: 0.2903\n", - "2/8, train_loss: 0.0190 step time: 0.2401\n", - "3/8, train_loss: 0.0198 step time: 0.2365\n", - "4/8, train_loss: 0.0163 step time: 0.3036\n", - "5/8, train_loss: 0.0174 step time: 0.3158\n", - "6/8, train_loss: 0.0166 step time: 0.2728\n", - "7/8, train_loss: 0.0139 step time: 0.2484\n", - "8/8, train_loss: 0.0208 step time: 0.2474\n", - "epoch 234 average loss: 0.0174\n", - "time consuming of epoch 234 is: 2.1563\n", - "----------\n", - "epoch 235/600\n", - "1/8, train_loss: 0.0183 step time: 0.2995\n", - "2/8, train_loss: 0.0177 step time: 0.2728\n", - "3/8, train_loss: 0.0166 step time: 0.3005\n", - "4/8, train_loss: 0.0179 step time: 0.2730\n", - "5/8, train_loss: 0.0135 step time: 0.2619\n", - "6/8, train_loss: 0.0175 step time: 0.2542\n", - "7/8, train_loss: 0.0167 step time: 0.2565\n", - "8/8, train_loss: 0.0160 step time: 0.2511\n", - "epoch 235 average loss: 0.0168\n", - "saved new best metric model\n", - "current epoch: 235 current mean dice: 0.9530 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 235 is: 3.9097\n", - "----------\n", - "epoch 236/600\n", - "1/8, train_loss: 0.0154 step time: 0.2745\n", - "2/8, train_loss: 0.0182 step time: 0.2119\n", - "3/8, train_loss: 0.0199 step time: 0.2090\n", - "4/8, train_loss: 0.0182 step time: 0.2165\n", - "5/8, train_loss: 0.0174 step time: 0.2652\n", - "6/8, train_loss: 0.0173 step time: 0.2616\n", - "7/8, train_loss: 0.0145 step time: 0.2769\n", - "8/8, train_loss: 0.0181 step time: 0.3164\n", - "epoch 236 average loss: 0.0174\n", - "time consuming of epoch 236 is: 2.0344\n", - "----------\n", - "epoch 237/600\n", - "1/8, train_loss: 0.0164 step time: 0.3204\n", - "2/8, train_loss: 0.0147 step time: 0.2686\n", - "3/8, train_loss: 0.0159 step time: 0.4573\n", - "4/8, train_loss: 0.0191 step time: 0.5862\n", - "5/8, train_loss: 0.0179 step time: 0.2667\n", - "6/8, train_loss: 0.0177 step time: 0.2493\n", - "7/8, train_loss: 0.0147 step time: 0.2315\n", - "8/8, train_loss: 0.0180 step time: 0.2342\n", - "epoch 237 average loss: 0.0168\n", - "time consuming of epoch 237 is: 2.6157\n", - "----------\n", - "epoch 238/600\n", - "1/8, train_loss: 0.0190 step time: 0.2912\n", - "2/8, train_loss: 0.0187 step time: 0.2423\n", - "3/8, train_loss: 0.0182 step time: 0.3849\n", - "4/8, train_loss: 0.0187 step time: 0.6047\n", - "5/8, train_loss: 0.0145 step time: 0.2169\n", - "6/8, train_loss: 0.0168 step time: 0.2144\n", - "7/8, train_loss: 0.0170 step time: 0.2068\n", - "8/8, train_loss: 0.0182 step time: 0.2843\n", - "epoch 238 average loss: 0.0176\n", - "time consuming of epoch 238 is: 2.4469\n", - "----------\n", - "epoch 239/600\n", - "1/8, train_loss: 0.0139 step time: 0.3568\n", - "2/8, train_loss: 0.0170 step time: 0.2754\n", - "3/8, train_loss: 0.0209 step time: 0.2814\n", - "4/8, train_loss: 0.0213 step time: 0.2771\n", - "5/8, train_loss: 0.0253 step time: 0.2510\n", - "6/8, train_loss: 0.0182 step time: 0.2459\n", - "7/8, train_loss: 0.0183 step time: 0.4098\n", - "8/8, train_loss: 0.0183 step time: 0.6277\n", - "epoch 239 average loss: 0.0191\n", - "time consuming of epoch 239 is: 2.7266\n", - "----------\n", - "epoch 240/600\n", - "1/8, train_loss: 0.0144 step time: 0.3156\n", - "2/8, train_loss: 0.0212 step time: 0.2539\n", - "3/8, train_loss: 0.0163 step time: 0.2454\n", - "4/8, train_loss: 0.0186 step time: 0.2369\n", - "5/8, train_loss: 0.0139 step time: 0.2403\n", - "6/8, train_loss: 0.0199 step time: 0.2400\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "7/8, train_loss: 0.0181 step time: 0.2271\n", - "8/8, train_loss: 0.0164 step time: 0.2298\n", - "epoch 240 average loss: 0.0174\n", - "current epoch: 240 current mean dice: 0.9499 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 240 is: 3.3993\n", - "----------\n", - "epoch 241/600\n", - "1/8, train_loss: 0.0152 step time: 0.3111\n", - "2/8, train_loss: 0.0172 step time: 0.2533\n", - "3/8, train_loss: 0.0158 step time: 0.2476\n", - "4/8, train_loss: 0.0187 step time: 0.2453\n", - "5/8, train_loss: 0.0156 step time: 0.2429\n", - "6/8, train_loss: 0.0178 step time: 0.2386\n", - "7/8, train_loss: 0.0244 step time: 0.2206\n", - "8/8, train_loss: 0.0206 step time: 0.2238\n", - "epoch 241 average loss: 0.0182\n", - "time consuming of epoch 241 is: 1.9843\n", - "----------\n", - "epoch 242/600\n", - "1/8, train_loss: 0.0195 step time: 0.6323\n", - "2/8, train_loss: 0.0193 step time: 0.2193\n", - "3/8, train_loss: 0.0180 step time: 0.2135\n", - "4/8, train_loss: 0.0143 step time: 0.4217\n", - "5/8, train_loss: 0.0233 step time: 0.3345\n", - "6/8, train_loss: 0.0186 step time: 0.2866\n", - "7/8, train_loss: 0.0194 step time: 0.2551\n", - "8/8, train_loss: 0.0158 step time: 0.2509\n", - "epoch 242 average loss: 0.0185\n", - "time consuming of epoch 242 is: 2.6153\n", - "----------\n", - "epoch 243/600\n", - "1/8, train_loss: 0.0208 step time: 0.3013\n", - "2/8, train_loss: 0.0204 step time: 0.2461\n", - "3/8, train_loss: 0.0175 step time: 0.2471\n", - "4/8, train_loss: 0.0163 step time: 0.3207\n", - "5/8, train_loss: 0.0186 step time: 0.2976\n", - "6/8, train_loss: 0.0234 step time: 0.2776\n", - "7/8, train_loss: 0.0142 step time: 0.2535\n", - "8/8, train_loss: 0.0227 step time: 0.2491\n", - "epoch 243 average loss: 0.0192\n", - "time consuming of epoch 243 is: 2.1944\n", - "----------\n", - "epoch 244/600\n", - "1/8, train_loss: 0.0184 step time: 0.3041\n", - "2/8, train_loss: 0.0180 step time: 0.2516\n", - "3/8, train_loss: 0.0193 step time: 0.2477\n", - "4/8, train_loss: 0.0213 step time: 0.2443\n", - "5/8, train_loss: 0.0125 step time: 0.2433\n", - "6/8, train_loss: 0.0174 step time: 0.2472\n", - "7/8, train_loss: 0.0170 step time: 0.3434\n", - "8/8, train_loss: 0.0278 step time: 0.4411\n", - "epoch 244 average loss: 0.0190\n", - "time consuming of epoch 244 is: 2.3241\n", - "----------\n", - "epoch 245/600\n", - "1/8, train_loss: 0.0164 step time: 0.3014\n", - "2/8, train_loss: 0.0157 step time: 0.2591\n", - "3/8, train_loss: 0.0196 step time: 0.2474\n", - "4/8, train_loss: 0.0169 step time: 0.2433\n", - "5/8, train_loss: 0.0170 step time: 0.2402\n", - "6/8, train_loss: 0.0165 step time: 0.2396\n", - "7/8, train_loss: 0.0179 step time: 0.2311\n", - "8/8, train_loss: 0.0182 step time: 0.2908\n", - "epoch 245 average loss: 0.0173\n", - "current epoch: 245 current mean dice: 0.9440 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 245 is: 3.5798\n", - "----------\n", - "epoch 246/600\n", - "1/8, train_loss: 0.0163 step time: 0.3161\n", - "2/8, train_loss: 0.0204 step time: 0.2524\n", - "3/8, train_loss: 0.0179 step time: 0.2376\n", - "4/8, train_loss: 0.0184 step time: 0.2357\n", - "5/8, train_loss: 0.0207 step time: 0.2375\n", - "6/8, train_loss: 0.0187 step time: 0.4057\n", - "7/8, train_loss: 0.0181 step time: 0.5498\n", - "8/8, train_loss: 0.0203 step time: 0.2054\n", - "epoch 246 average loss: 0.0188\n", - "time consuming of epoch 246 is: 2.4416\n", - "----------\n", - "epoch 247/600\n", - "1/8, train_loss: 0.0183 step time: 0.2626\n", - "2/8, train_loss: 0.0194 step time: 0.2116\n", - "3/8, train_loss: 0.0195 step time: 0.2131\n", - "4/8, train_loss: 0.0204 step time: 0.2156\n", - "5/8, train_loss: 0.0204 step time: 0.2394\n", - "6/8, train_loss: 0.0163 step time: 0.3872\n", - "7/8, train_loss: 0.0174 step time: 0.4289\n", - "8/8, train_loss: 0.0215 step time: 0.2575\n", - "epoch 247 average loss: 0.0191\n", - "time consuming of epoch 247 is: 2.2176\n", - "----------\n", - "epoch 248/600\n", - "1/8, train_loss: 0.0158 step time: 0.2884\n", - "2/8, train_loss: 0.0178 step time: 0.3003\n", - "3/8, train_loss: 0.0223 step time: 0.2871\n", - "4/8, train_loss: 0.0186 step time: 0.2734\n", - "5/8, train_loss: 0.0171 step time: 0.2558\n", - "6/8, train_loss: 0.0179 step time: 0.2763\n", - "7/8, train_loss: 0.0152 step time: 0.2638\n", - "8/8, train_loss: 0.0168 step time: 0.2499\n", - "epoch 248 average loss: 0.0177\n", - "time consuming of epoch 248 is: 2.1966\n", - "----------\n", - "epoch 249/600\n", - "1/8, train_loss: 0.0162 step time: 0.2879\n", - "2/8, train_loss: 0.0171 step time: 0.2334\n", - "3/8, train_loss: 0.0153 step time: 0.2393\n", - "4/8, train_loss: 0.0142 step time: 0.2375\n", - "5/8, train_loss: 0.0244 step time: 0.2543\n", - "6/8, train_loss: 0.0173 step time: 0.3753\n", - "7/8, train_loss: 0.0172 step time: 0.4558\n", - "8/8, train_loss: 0.0191 step time: 0.2553\n", - "epoch 249 average loss: 0.0176\n", - "time consuming of epoch 249 is: 2.3404\n", - "----------\n", - "epoch 250/600\n", - "1/8, train_loss: 0.0167 step time: 0.3045\n", - "2/8, train_loss: 0.0163 step time: 0.2436\n", - "3/8, train_loss: 0.0175 step time: 0.2423\n", - "4/8, train_loss: 0.0190 step time: 0.2403\n", - "5/8, train_loss: 0.0184 step time: 0.2319\n", - "6/8, train_loss: 0.0196 step time: 0.2281\n", - "7/8, train_loss: 0.0190 step time: 0.2154\n", - "8/8, train_loss: 0.0166 step time: 0.3999\n", - "epoch 250 average loss: 0.0179\n", - "current epoch: 250 current mean dice: 0.9483 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 250 is: 3.5188\n", - "----------\n", - "epoch 251/600\n", - "1/8, train_loss: 0.0210 step time: 0.2986\n", - "2/8, train_loss: 0.0161 step time: 0.2749\n", - "3/8, train_loss: 0.0161 step time: 0.2771\n", - "4/8, train_loss: 0.0168 step time: 0.2590\n", - "5/8, train_loss: 0.0176 step time: 0.2909\n", - "6/8, train_loss: 0.0179 step time: 0.2959\n", - "7/8, train_loss: 0.0194 step time: 0.2758\n", - "8/8, train_loss: 0.0183 step time: 0.2550\n", - "epoch 251 average loss: 0.0179\n", - "time consuming of epoch 251 is: 2.2285\n", - "----------\n", - "epoch 252/600\n", - "1/8, train_loss: 0.0144 step time: 0.2929\n", - "2/8, train_loss: 0.0210 step time: 0.2449\n", - "3/8, train_loss: 0.0166 step time: 0.2416\n", - "4/8, train_loss: 0.0183 step time: 0.4123\n", - "5/8, train_loss: 0.0182 step time: 0.6209\n", - "6/8, train_loss: 0.0155 step time: 0.2642\n", - "7/8, train_loss: 0.0183 step time: 0.2396\n", - "8/8, train_loss: 0.0150 step time: 0.2298\n", - "epoch 252 average loss: 0.0171\n", - "time consuming of epoch 252 is: 2.5475\n", - "----------\n", - "epoch 253/600\n", - "1/8, train_loss: 0.0175 step time: 0.2784\n", - "2/8, train_loss: 0.0176 step time: 0.2334\n", - "3/8, train_loss: 0.0178 step time: 0.2316\n", - "4/8, train_loss: 0.0184 step time: 0.2283\n", - "5/8, train_loss: 0.0178 step time: 0.2287\n", - "6/8, train_loss: 0.0153 step time: 0.2319\n", - "7/8, train_loss: 0.0201 step time: 0.2520\n", - "8/8, train_loss: 0.0154 step time: 0.3560\n", - "epoch 253 average loss: 0.0175\n", - "time consuming of epoch 253 is: 2.0418\n", - "----------\n", - "epoch 254/600\n", - "1/8, train_loss: 0.0148 step time: 0.3476\n", - "2/8, train_loss: 0.0177 step time: 0.2642\n", - "3/8, train_loss: 0.0182 step time: 0.2503\n", - "4/8, train_loss: 0.0181 step time: 0.2475\n", - "5/8, train_loss: 0.0150 step time: 0.2960\n", - "6/8, train_loss: 0.0205 step time: 0.3475\n", - "7/8, train_loss: 0.0178 step time: 0.2778\n", - "8/8, train_loss: 0.0181 step time: 0.2618\n", - "epoch 254 average loss: 0.0175\n", - "time consuming of epoch 254 is: 2.2941\n", - "----------\n", - "epoch 255/600\n", - "1/8, train_loss: 0.0200 step time: 0.3103\n", - "2/8, train_loss: 0.0145 step time: 0.2528\n", - "3/8, train_loss: 0.0155 step time: 0.2943\n", - "4/8, train_loss: 0.0162 step time: 0.3328\n", - "5/8, train_loss: 0.0163 step time: 0.2843\n", - "6/8, train_loss: 0.0184 step time: 0.2759\n", - "7/8, train_loss: 0.0149 step time: 0.2707\n", - "8/8, train_loss: 0.0198 step time: 0.2616\n", - "epoch 255 average loss: 0.0170\n", - "current epoch: 255 current mean dice: 0.9523 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 255 is: 3.7885\n", - "----------\n", - "epoch 256/600\n", - "1/8, train_loss: 0.0167 step time: 0.3043\n", - "2/8, train_loss: 0.0195 step time: 0.2627\n", - "3/8, train_loss: 0.0170 step time: 0.2620\n", - "4/8, train_loss: 0.0185 step time: 0.3830\n", - "5/8, train_loss: 0.0193 step time: 0.3290\n", - "6/8, train_loss: 0.0159 step time: 0.3143\n", - "7/8, train_loss: 0.0191 step time: 0.2651\n", - "8/8, train_loss: 0.0190 step time: 0.2613\n", - "epoch 256 average loss: 0.0181\n", - "time consuming of epoch 256 is: 2.3828\n", - "----------\n", - "epoch 257/600\n", - "1/8, train_loss: 0.0176 step time: 0.2935\n", - "2/8, train_loss: 0.0174 step time: 0.2490\n", - "3/8, train_loss: 0.0183 step time: 0.2557\n", - "4/8, train_loss: 0.0187 step time: 0.2483\n", - "5/8, train_loss: 0.0174 step time: 0.2429\n", - "6/8, train_loss: 0.0182 step time: 0.2475\n", - "7/8, train_loss: 0.0185 step time: 0.2239\n", - "8/8, train_loss: 0.0170 step time: 0.2273\n", - "epoch 257 average loss: 0.0179\n", - "time consuming of epoch 257 is: 1.9897\n", - "----------\n", - "epoch 258/600\n", - "1/8, train_loss: 0.0186 step time: 0.2825\n", - "2/8, train_loss: 0.0188 step time: 0.2329\n", - "3/8, train_loss: 0.0144 step time: 0.2324\n", - "4/8, train_loss: 0.0157 step time: 0.2316\n", - "5/8, train_loss: 0.0193 step time: 0.2416\n", - "6/8, train_loss: 0.0156 step time: 0.2640\n", - "7/8, train_loss: 0.0155 step time: 0.2920\n", - "8/8, train_loss: 0.0171 step time: 0.2978\n", - "epoch 258 average loss: 0.0169\n", - "time consuming of epoch 258 is: 2.0764\n", - "----------\n", - "epoch 259/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0167 step time: 0.3402\n", - "2/8, train_loss: 0.0161 step time: 0.2924\n", - "3/8, train_loss: 0.0143 step time: 0.3054\n", - "4/8, train_loss: 0.0189 step time: 0.2922\n", - "5/8, train_loss: 0.0170 step time: 0.2683\n", - "6/8, train_loss: 0.0154 step time: 0.2512\n", - "7/8, train_loss: 0.0181 step time: 0.2494\n", - "8/8, train_loss: 0.0188 step time: 0.2938\n", - "epoch 259 average loss: 0.0169\n", - "time consuming of epoch 259 is: 2.2943\n", - "----------\n", - "epoch 260/600\n", - "1/8, train_loss: 0.0160 step time: 0.3622\n", - "2/8, train_loss: 0.0169 step time: 0.2763\n", - "3/8, train_loss: 0.0165 step time: 0.2651\n", - "4/8, train_loss: 0.0165 step time: 0.2616\n", - "5/8, train_loss: 0.0166 step time: 0.2502\n", - "6/8, train_loss: 0.0140 step time: 0.2532\n", - "7/8, train_loss: 0.0160 step time: 0.2442\n", - "8/8, train_loss: 0.0178 step time: 0.4176\n", - "epoch 260 average loss: 0.0163\n", - "current epoch: 260 current mean dice: 0.9523 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 260 is: 3.6430\n", - "----------\n", - "epoch 261/600\n", - "1/8, train_loss: 0.0160 step time: 0.3416\n", - "2/8, train_loss: 0.0201 step time: 0.3069\n", - "3/8, train_loss: 0.0178 step time: 0.3125\n", - "4/8, train_loss: 0.0176 step time: 0.3092\n", - "5/8, train_loss: 0.0136 step time: 0.2916\n", - "6/8, train_loss: 0.0184 step time: 0.2795\n", - "7/8, train_loss: 0.0156 step time: 0.2717\n", - "8/8, train_loss: 0.0166 step time: 0.2499\n", - "epoch 261 average loss: 0.0170\n", - "time consuming of epoch 261 is: 2.3641\n", - "----------\n", - "epoch 262/600\n", - "1/8, train_loss: 0.0168 step time: 0.2938\n", - "2/8, train_loss: 0.0179 step time: 0.2449\n", - "3/8, train_loss: 0.0180 step time: 0.3911\n", - "4/8, train_loss: 0.0141 step time: 0.4980\n", - "5/8, train_loss: 0.0156 step time: 0.2696\n", - "6/8, train_loss: 0.0208 step time: 0.2525\n", - "7/8, train_loss: 0.0164 step time: 0.2392\n", - "8/8, train_loss: 0.0163 step time: 0.2289\n", - "epoch 262 average loss: 0.0170\n", - "time consuming of epoch 262 is: 2.4194\n", - "----------\n", - "epoch 263/600\n", - "1/8, train_loss: 0.0162 step time: 0.2865\n", - "2/8, train_loss: 0.0134 step time: 0.2324\n", - "3/8, train_loss: 0.0199 step time: 0.2383\n", - "4/8, train_loss: 0.0157 step time: 0.2404\n", - "5/8, train_loss: 0.0182 step time: 0.2315\n", - "6/8, train_loss: 0.0154 step time: 0.2312\n", - "7/8, train_loss: 0.0176 step time: 0.2177\n", - "8/8, train_loss: 0.0188 step time: 0.2492\n", - "epoch 263 average loss: 0.0169\n", - "time consuming of epoch 263 is: 1.9288\n", - "----------\n", - "epoch 264/600\n", - "1/8, train_loss: 0.0206 step time: 0.3219\n", - "2/8, train_loss: 0.0181 step time: 0.2678\n", - "3/8, train_loss: 0.0167 step time: 0.2560\n", - "4/8, train_loss: 0.0149 step time: 0.2520\n", - "5/8, train_loss: 0.0162 step time: 0.2385\n", - "6/8, train_loss: 0.0139 step time: 0.2319\n", - "7/8, train_loss: 0.0163 step time: 0.2212\n", - "8/8, train_loss: 0.0145 step time: 0.2158\n", - "epoch 264 average loss: 0.0164\n", - "time consuming of epoch 264 is: 2.0066\n", - "----------\n", - "epoch 265/600\n", - "1/8, train_loss: 0.0159 step time: 0.2807\n", - "2/8, train_loss: 0.0163 step time: 0.2269\n", - "3/8, train_loss: 0.0149 step time: 0.2322\n", - "4/8, train_loss: 0.0181 step time: 0.2257\n", - "5/8, train_loss: 0.0171 step time: 0.2455\n", - "6/8, train_loss: 0.0195 step time: 0.2532\n", - "7/8, train_loss: 0.0163 step time: 0.2464\n", - "8/8, train_loss: 0.0157 step time: 0.2430\n", - "epoch 265 average loss: 0.0167\n", - "current epoch: 265 current mean dice: 0.9524 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 265 is: 3.3196\n", - "----------\n", - "epoch 266/600\n", - "1/8, train_loss: 0.0138 step time: 0.3507\n", - "2/8, train_loss: 0.0173 step time: 0.2978\n", - "3/8, train_loss: 0.0201 step time: 0.2698\n", - "4/8, train_loss: 0.0172 step time: 0.2602\n", - "5/8, train_loss: 0.0153 step time: 0.2612\n", - "6/8, train_loss: 0.0169 step time: 0.2544\n", - "7/8, train_loss: 0.0156 step time: 0.2506\n", - "8/8, train_loss: 0.0283 step time: 0.2605\n", - "epoch 266 average loss: 0.0181\n", - "time consuming of epoch 266 is: 2.2064\n", - "----------\n", - "epoch 267/600\n", - "1/8, train_loss: 0.0169 step time: 0.2887\n", - "2/8, train_loss: 0.0155 step time: 0.2661\n", - "3/8, train_loss: 0.0199 step time: 0.2780\n", - "4/8, train_loss: 0.0160 step time: 0.2589\n", - "5/8, train_loss: 0.0168 step time: 0.2629\n", - "6/8, train_loss: 0.0177 step time: 0.2508\n", - "7/8, train_loss: 0.0148 step time: 0.2433\n", - "8/8, train_loss: 0.0167 step time: 0.2440\n", - "epoch 267 average loss: 0.0168\n", - "time consuming of epoch 267 is: 2.0940\n", - "----------\n", - "epoch 268/600\n", - "1/8, train_loss: 0.0171 step time: 0.6238\n", - "2/8, train_loss: 0.0209 step time: 0.2147\n", - "3/8, train_loss: 0.0174 step time: 0.2239\n", - "4/8, train_loss: 0.0163 step time: 0.4438\n", - "5/8, train_loss: 0.0202 step time: 0.2934\n", - "6/8, train_loss: 0.0167 step time: 0.2667\n", - "7/8, train_loss: 0.0151 step time: 0.2512\n", - "8/8, train_loss: 0.0163 step time: 0.2483\n", - "epoch 268 average loss: 0.0175\n", - "time consuming of epoch 268 is: 2.5670\n", - "----------\n", - "epoch 269/600\n", - "1/8, train_loss: 0.0202 step time: 0.2834\n", - "2/8, train_loss: 0.0189 step time: 0.4035\n", - "3/8, train_loss: 0.0157 step time: 0.4582\n", - "4/8, train_loss: 0.0157 step time: 0.2140\n", - "5/8, train_loss: 0.0163 step time: 0.2112\n", - "6/8, train_loss: 0.0142 step time: 0.2138\n", - "7/8, train_loss: 0.0191 step time: 0.2103\n", - "8/8, train_loss: 0.0189 step time: 0.2163\n", - "epoch 269 average loss: 0.0174\n", - "time consuming of epoch 269 is: 2.2119\n", - "----------\n", - "epoch 270/600\n", - "1/8, train_loss: 0.0173 step time: 0.2888\n", - "2/8, train_loss: 0.0152 step time: 0.2455\n", - "3/8, train_loss: 0.0166 step time: 0.2437\n", - "4/8, train_loss: 0.0192 step time: 0.2452\n", - "5/8, train_loss: 0.0167 step time: 0.2447\n", - "6/8, train_loss: 0.0170 step time: 0.3010\n", - "7/8, train_loss: 0.0186 step time: 0.4381\n", - "8/8, train_loss: 0.0184 step time: 0.2728\n", - "epoch 270 average loss: 0.0174\n", - "current epoch: 270 current mean dice: 0.9522 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 270 is: 3.7816\n", - "----------\n", - "epoch 271/600\n", - "1/8, train_loss: 0.0167 step time: 0.3107\n", - "2/8, train_loss: 0.0193 step time: 0.2550\n", - "3/8, train_loss: 0.0212 step time: 0.2459\n", - "4/8, train_loss: 0.0168 step time: 0.2522\n", - "5/8, train_loss: 0.0169 step time: 0.2564\n", - "6/8, train_loss: 0.0133 step time: 0.2783\n", - "7/8, train_loss: 0.0160 step time: 0.2631\n", - "8/8, train_loss: 0.0147 step time: 0.2509\n", - "epoch 271 average loss: 0.0169\n", - "time consuming of epoch 271 is: 2.1135\n", - "----------\n", - "epoch 272/600\n", - "1/8, train_loss: 0.0166 step time: 0.3020\n", - "2/8, train_loss: 0.0170 step time: 0.2575\n", - "3/8, train_loss: 0.0182 step time: 0.2498\n", - "4/8, train_loss: 0.0152 step time: 0.2477\n", - "5/8, train_loss: 0.0165 step time: 0.2435\n", - "6/8, train_loss: 0.0203 step time: 0.2448\n", - "7/8, train_loss: 0.0135 step time: 0.2396\n", - "8/8, train_loss: 0.0222 step time: 0.2286\n", - "epoch 272 average loss: 0.0174\n", - "time consuming of epoch 272 is: 2.0147\n", - "----------\n", - "epoch 273/600\n", - "1/8, train_loss: 0.0158 step time: 0.2764\n", - "2/8, train_loss: 0.0171 step time: 0.2319\n", - "3/8, train_loss: 0.0150 step time: 0.2354\n", - "4/8, train_loss: 0.0189 step time: 0.3629\n", - "5/8, train_loss: 0.0161 step time: 0.3019\n", - "6/8, train_loss: 0.0148 step time: 0.2657\n", - "7/8, train_loss: 0.0174 step time: 0.2588\n", - "8/8, train_loss: 0.0185 step time: 0.2437\n", - "epoch 273 average loss: 0.0167\n", - "time consuming of epoch 273 is: 2.1778\n", - "----------\n", - "epoch 274/600\n", - "1/8, train_loss: 0.0209 step time: 0.2909\n", - "2/8, train_loss: 0.0169 step time: 0.2702\n", - "3/8, train_loss: 0.0157 step time: 0.3241\n", - "4/8, train_loss: 0.0162 step time: 0.2899\n", - "5/8, train_loss: 0.0181 step time: 0.2572\n", - "6/8, train_loss: 0.0182 step time: 0.2526\n", - "7/8, train_loss: 0.0211 step time: 0.2482\n", - "8/8, train_loss: 0.0155 step time: 0.2476\n", - "epoch 274 average loss: 0.0178\n", - "time consuming of epoch 274 is: 2.1821\n", - "----------\n", - "epoch 275/600\n", - "1/8, train_loss: 0.0169 step time: 0.2977\n", - "2/8, train_loss: 0.0184 step time: 0.2489\n", - "3/8, train_loss: 0.0168 step time: 0.2445\n", - "4/8, train_loss: 0.0191 step time: 0.2474\n", - "5/8, train_loss: 0.0153 step time: 0.2376\n", - "6/8, train_loss: 0.0180 step time: 0.3873\n", - "7/8, train_loss: 0.0157 step time: 0.6222\n", - "8/8, train_loss: 0.0166 step time: 0.2717\n", - "epoch 275 average loss: 0.0171\n", - "current epoch: 275 current mean dice: 0.9518 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 275 is: 3.6783\n", - "----------\n", - "epoch 276/600\n", - "1/8, train_loss: 0.0218 step time: 0.3200\n", - "2/8, train_loss: 0.0154 step time: 0.2693\n", - "3/8, train_loss: 0.0168 step time: 0.2602\n", - "4/8, train_loss: 0.0155 step time: 0.2495\n", - "5/8, train_loss: 0.0177 step time: 0.2433\n", - "6/8, train_loss: 0.0182 step time: 0.2493\n", - "7/8, train_loss: 0.0190 step time: 0.2406\n", - "8/8, train_loss: 0.0171 step time: 0.2339\n", - "epoch 276 average loss: 0.0177\n", - "time consuming of epoch 276 is: 2.0672\n", - "----------\n", - "epoch 277/600\n", - "1/8, train_loss: 0.0214 step time: 0.2885\n", - "2/8, train_loss: 0.0170 step time: 0.2579\n", - "3/8, train_loss: 0.0255 step time: 0.2771\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "4/8, train_loss: 0.0164 step time: 0.2882\n", - "5/8, train_loss: 0.0149 step time: 0.2747\n", - "6/8, train_loss: 0.0201 step time: 0.2671\n", - "7/8, train_loss: 0.0133 step time: 0.2469\n", - "8/8, train_loss: 0.0156 step time: 0.2377\n", - "epoch 277 average loss: 0.0180\n", - "time consuming of epoch 277 is: 2.1396\n", - "----------\n", - "epoch 278/600\n", - "1/8, train_loss: 0.0165 step time: 0.2887\n", - "2/8, train_loss: 0.0179 step time: 0.2401\n", - "3/8, train_loss: 0.0163 step time: 0.2426\n", - "4/8, train_loss: 0.0193 step time: 0.2507\n", - "5/8, train_loss: 0.0146 step time: 0.4353\n", - "6/8, train_loss: 0.0171 step time: 0.4717\n", - "7/8, train_loss: 0.0204 step time: 0.2626\n", - "8/8, train_loss: 0.0167 step time: 0.2533\n", - "epoch 278 average loss: 0.0173\n", - "time consuming of epoch 278 is: 2.4465\n", - "----------\n", - "epoch 279/600\n", - "1/8, train_loss: 0.0170 step time: 0.3003\n", - "2/8, train_loss: 0.0173 step time: 0.2510\n", - "3/8, train_loss: 0.0159 step time: 0.2473\n", - "4/8, train_loss: 0.0156 step time: 0.2394\n", - "5/8, train_loss: 0.0220 step time: 0.2391\n", - "6/8, train_loss: 0.0184 step time: 0.3946\n", - "7/8, train_loss: 0.0168 step time: 0.5849\n", - "8/8, train_loss: 0.0179 step time: 0.2092\n", - "epoch 279 average loss: 0.0176\n", - "time consuming of epoch 279 is: 2.4672\n", - "----------\n", - "epoch 280/600\n", - "1/8, train_loss: 0.0217 step time: 0.2647\n", - "2/8, train_loss: 0.0201 step time: 0.2179\n", - "3/8, train_loss: 0.0188 step time: 0.4956\n", - "4/8, train_loss: 0.0155 step time: 0.2193\n", - "5/8, train_loss: 0.0150 step time: 0.2143\n", - "6/8, train_loss: 0.0204 step time: 0.2149\n", - "7/8, train_loss: 0.0182 step time: 0.1999\n", - "8/8, train_loss: 0.0139 step time: 0.2043\n", - "epoch 280 average loss: 0.0180\n", - "current epoch: 280 current mean dice: 0.9525 best mean dice: 0.9530 at epoch: 235\n", - "time consuming of epoch 280 is: 3.3596\n", - "----------\n", - "epoch 281/600\n", - "1/8, train_loss: 0.0162 step time: 0.3524\n", - "2/8, train_loss: 0.0148 step time: 0.3024\n", - "3/8, train_loss: 0.0173 step time: 0.2768\n", - "4/8, train_loss: 0.0188 step time: 0.2698\n", - "5/8, train_loss: 0.0175 step time: 0.2598\n", - "6/8, train_loss: 0.0174 step time: 0.2558\n", - "7/8, train_loss: 0.0184 step time: 0.2739\n", - "8/8, train_loss: 0.0157 step time: 0.3169\n", - "epoch 281 average loss: 0.0170\n", - "time consuming of epoch 281 is: 2.3091\n", - "----------\n", - "epoch 282/600\n", - "1/8, train_loss: 0.0154 step time: 0.3300\n", - "2/8, train_loss: 0.0188 step time: 0.2780\n", - "3/8, train_loss: 0.0207 step time: 0.2542\n", - "4/8, train_loss: 0.0164 step time: 0.2520\n", - "5/8, train_loss: 0.0172 step time: 0.2524\n", - "6/8, train_loss: 0.0144 step time: 0.2522\n", - "7/8, train_loss: 0.0171 step time: 0.2993\n", - "8/8, train_loss: 0.0165 step time: 0.3187\n", - "epoch 282 average loss: 0.0171\n", - "time consuming of epoch 282 is: 2.2381\n", - "----------\n", - "epoch 283/600\n", - "1/8, train_loss: 0.0154 step time: 0.3011\n", - "2/8, train_loss: 0.0170 step time: 0.2539\n", - "3/8, train_loss: 0.0165 step time: 0.2533\n", - "4/8, train_loss: 0.0226 step time: 0.3942\n", - "5/8, train_loss: 0.0172 step time: 0.3403\n", - "6/8, train_loss: 0.0172 step time: 0.2794\n", - "7/8, train_loss: 0.0190 step time: 0.2684\n", - "8/8, train_loss: 0.0144 step time: 0.2546\n", - "epoch 283 average loss: 0.0174\n", - "time consuming of epoch 283 is: 2.3466\n", - "----------\n", - "epoch 284/600\n", - "1/8, train_loss: 0.0156 step time: 0.2897\n", - "2/8, train_loss: 0.0156 step time: 0.2358\n", - "3/8, train_loss: 0.0179 step time: 0.2372\n", - "4/8, train_loss: 0.0193 step time: 0.2351\n", - "5/8, train_loss: 0.0215 step time: 0.2469\n", - "6/8, train_loss: 0.0138 step time: 0.2587\n", - "7/8, train_loss: 0.0170 step time: 0.2535\n", - "8/8, train_loss: 0.0182 step time: 0.2444\n", - "epoch 284 average loss: 0.0174\n", - "time consuming of epoch 284 is: 2.0028\n", - "----------\n", - "epoch 285/600\n", - "1/8, train_loss: 0.0168 step time: 0.2826\n", - "2/8, train_loss: 0.0166 step time: 0.2353\n", - "3/8, train_loss: 0.0196 step time: 0.2316\n", - "4/8, train_loss: 0.0147 step time: 0.2328\n", - "5/8, train_loss: 0.0202 step time: 0.2665\n", - "6/8, train_loss: 0.0179 step time: 0.3206\n", - "7/8, train_loss: 0.0152 step time: 0.2972\n", - "8/8, train_loss: 0.0161 step time: 0.2724\n", - "epoch 285 average loss: 0.0171\n", - "saved new best metric model\n", - "current epoch: 285 current mean dice: 0.9541 best mean dice: 0.9541 at epoch: 285\n", - "time consuming of epoch 285 is: 3.9340\n", - "----------\n", - "epoch 286/600\n", - "1/8, train_loss: 0.0167 step time: 0.2709\n", - "2/8, train_loss: 0.0158 step time: 0.2123\n", - "3/8, train_loss: 0.0206 step time: 0.2117\n", - "4/8, train_loss: 0.0170 step time: 0.2120\n", - "5/8, train_loss: 0.0185 step time: 0.2126\n", - "6/8, train_loss: 0.0161 step time: 0.2240\n", - "7/8, train_loss: 0.0190 step time: 0.4714\n", - "8/8, train_loss: 0.0220 step time: 0.4731\n", - "epoch 286 average loss: 0.0182\n", - "time consuming of epoch 286 is: 2.2905\n", - "----------\n", - "epoch 287/600\n", - "1/8, train_loss: 0.0173 step time: 0.3103\n", - "2/8, train_loss: 0.0166 step time: 0.2463\n", - "3/8, train_loss: 0.0245 step time: 0.2529\n", - "4/8, train_loss: 0.0168 step time: 0.2548\n", - "5/8, train_loss: 0.0195 step time: 0.2619\n", - "6/8, train_loss: 0.0180 step time: 0.2582\n", - "7/8, train_loss: 0.0185 step time: 0.2520\n", - "8/8, train_loss: 0.0155 step time: 0.3327\n", - "epoch 287 average loss: 0.0183\n", - "time consuming of epoch 287 is: 2.1705\n", - "----------\n", - "epoch 288/600\n", - "1/8, train_loss: 0.0198 step time: 0.3769\n", - "2/8, train_loss: 0.0200 step time: 0.2838\n", - "3/8, train_loss: 0.0154 step time: 0.2612\n", - "4/8, train_loss: 0.0159 step time: 0.2530\n", - "5/8, train_loss: 0.0202 step time: 0.2480\n", - "6/8, train_loss: 0.0153 step time: 0.3317\n", - "7/8, train_loss: 0.0137 step time: 0.4539\n", - "8/8, train_loss: 0.0187 step time: 0.2603\n", - "epoch 288 average loss: 0.0174\n", - "time consuming of epoch 288 is: 2.4702\n", - "----------\n", - "epoch 289/600\n", - "1/8, train_loss: 0.0175 step time: 0.2974\n", - "2/8, train_loss: 0.0169 step time: 0.2471\n", - "3/8, train_loss: 0.0154 step time: 0.2410\n", - "4/8, train_loss: 0.0157 step time: 0.2406\n", - "5/8, train_loss: 0.0176 step time: 0.2342\n", - "6/8, train_loss: 0.0170 step time: 0.2422\n", - "7/8, train_loss: 0.0151 step time: 0.2295\n", - "8/8, train_loss: 0.0160 step time: 0.2651\n", - "epoch 289 average loss: 0.0164\n", - "time consuming of epoch 289 is: 1.9986\n", - "----------\n", - "epoch 290/600\n", - "1/8, train_loss: 0.0171 step time: 0.2982\n", - "2/8, train_loss: 0.0164 step time: 0.2498\n", - "3/8, train_loss: 0.0170 step time: 0.2519\n", - "4/8, train_loss: 0.0202 step time: 0.2522\n", - "5/8, train_loss: 0.0163 step time: 0.2598\n", - "6/8, train_loss: 0.0162 step time: 0.2684\n", - "7/8, train_loss: 0.0176 step time: 0.2716\n", - "8/8, train_loss: 0.0144 step time: 0.2434\n", - "epoch 290 average loss: 0.0169\n", - "current epoch: 290 current mean dice: 0.9529 best mean dice: 0.9541 at epoch: 285\n", - "time consuming of epoch 290 is: 3.5462\n", - "----------\n", - "epoch 291/600\n", - "1/8, train_loss: 0.0181 step time: 0.3293\n", - "2/8, train_loss: 0.0193 step time: 0.2683\n", - "3/8, train_loss: 0.0190 step time: 0.2571\n", - "4/8, train_loss: 0.0153 step time: 0.2443\n", - "5/8, train_loss: 0.0148 step time: 0.2410\n", - "6/8, train_loss: 0.0156 step time: 0.2413\n", - "7/8, train_loss: 0.0175 step time: 0.2322\n", - "8/8, train_loss: 0.0188 step time: 0.2321\n", - "epoch 291 average loss: 0.0173\n", - "time consuming of epoch 291 is: 2.0467\n", - "----------\n", - "epoch 292/600\n", - "1/8, train_loss: 0.0162 step time: 0.2883\n", - "2/8, train_loss: 0.0131 step time: 0.2419\n", - "3/8, train_loss: 0.0165 step time: 0.3568\n", - "4/8, train_loss: 0.0175 step time: 0.3275\n", - "5/8, train_loss: 0.0181 step time: 0.2761\n", - "6/8, train_loss: 0.0253 step time: 0.2580\n", - "7/8, train_loss: 0.0206 step time: 0.2573\n", - "8/8, train_loss: 0.0170 step time: 0.2450\n", - "epoch 292 average loss: 0.0180\n", - "time consuming of epoch 292 is: 2.2523\n", - "----------\n", - "epoch 293/600\n", - "1/8, train_loss: 0.0167 step time: 0.2945\n", - "2/8, train_loss: 0.0185 step time: 0.2572\n", - "3/8, train_loss: 0.0178 step time: 0.2454\n", - "4/8, train_loss: 0.0195 step time: 0.2474\n", - "5/8, train_loss: 0.0168 step time: 0.2437\n", - "6/8, train_loss: 0.0181 step time: 0.2968\n", - "7/8, train_loss: 0.0155 step time: 0.3773\n", - "8/8, train_loss: 0.0149 step time: 0.2694\n", - "epoch 293 average loss: 0.0172\n", - "time consuming of epoch 293 is: 2.2331\n", - "----------\n", - "epoch 294/600\n", - "1/8, train_loss: 0.0162 step time: 0.3046\n", - "2/8, train_loss: 0.0157 step time: 0.2498\n", - "3/8, train_loss: 0.0160 step time: 0.2423\n", - "4/8, train_loss: 0.0161 step time: 0.2383\n", - "5/8, train_loss: 0.0196 step time: 0.2379\n", - "6/8, train_loss: 0.0155 step time: 0.2371\n", - "7/8, train_loss: 0.0186 step time: 0.2278\n", - "8/8, train_loss: 0.0125 step time: 0.2296\n", - "epoch 294 average loss: 0.0163\n", - "time consuming of epoch 294 is: 1.9689\n", - "----------\n", - "epoch 295/600\n", - "1/8, train_loss: 0.0153 step time: 0.2822\n", - "2/8, train_loss: 0.0172 step time: 0.2353\n", - "3/8, train_loss: 0.0182 step time: 0.2341\n", - "4/8, train_loss: 0.0168 step time: 0.2309\n", - "5/8, train_loss: 0.0161 step time: 0.2656\n", - "6/8, train_loss: 0.0179 step time: 0.2736\n", - "7/8, train_loss: 0.0159 step time: 0.2755\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "8/8, train_loss: 0.0189 step time: 0.2737\n", - "epoch 295 average loss: 0.0170\n", - "current epoch: 295 current mean dice: 0.9537 best mean dice: 0.9541 at epoch: 285\n", - "time consuming of epoch 295 is: 3.3778\n", - "----------\n", - "epoch 296/600\n", - "1/8, train_loss: 0.0175 step time: 0.3695\n", - "2/8, train_loss: 0.0159 step time: 0.2958\n", - "3/8, train_loss: 0.0173 step time: 0.2571\n", - "4/8, train_loss: 0.0137 step time: 0.2648\n", - "5/8, train_loss: 0.0160 step time: 0.2611\n", - "6/8, train_loss: 0.0146 step time: 0.2675\n", - "7/8, train_loss: 0.0180 step time: 0.2689\n", - "8/8, train_loss: 0.0170 step time: 0.2603\n", - "epoch 296 average loss: 0.0163\n", - "time consuming of epoch 296 is: 2.2462\n", - "----------\n", - "epoch 297/600\n", - "1/8, train_loss: 0.0160 step time: 0.2973\n", - "2/8, train_loss: 0.0151 step time: 0.2545\n", - "3/8, train_loss: 0.0159 step time: 0.2406\n", - "4/8, train_loss: 0.0165 step time: 0.2579\n", - "5/8, train_loss: 0.0144 step time: 0.2978\n", - "6/8, train_loss: 0.0167 step time: 0.2661\n", - "7/8, train_loss: 0.0215 step time: 0.2629\n", - "8/8, train_loss: 0.0147 step time: 0.2643\n", - "epoch 297 average loss: 0.0164\n", - "time consuming of epoch 297 is: 2.1429\n", - "----------\n", - "epoch 298/600\n", - "1/8, train_loss: 0.0196 step time: 0.3127\n", - "2/8, train_loss: 0.0140 step time: 0.2590\n", - "3/8, train_loss: 0.0166 step time: 0.2526\n", - "4/8, train_loss: 0.0162 step time: 0.2556\n", - "5/8, train_loss: 0.0158 step time: 0.2503\n", - "6/8, train_loss: 0.0167 step time: 0.2462\n", - "7/8, train_loss: 0.0158 step time: 0.2248\n", - "8/8, train_loss: 0.0173 step time: 0.2221\n", - "epoch 298 average loss: 0.0165\n", - "time consuming of epoch 298 is: 2.0247\n", - "----------\n", - "epoch 299/600\n", - "1/8, train_loss: 0.0217 step time: 0.2875\n", - "2/8, train_loss: 0.0150 step time: 0.2312\n", - "3/8, train_loss: 0.0150 step time: 0.2379\n", - "4/8, train_loss: 0.0153 step time: 0.2430\n", - "5/8, train_loss: 0.0141 step time: 0.3139\n", - "6/8, train_loss: 0.0145 step time: 0.4030\n", - "7/8, train_loss: 0.0163 step time: 0.2852\n", - "8/8, train_loss: 0.0212 step time: 0.2546\n", - "epoch 299 average loss: 0.0166\n", - "time consuming of epoch 299 is: 2.2578\n", - "----------\n", - "epoch 300/600\n", - "1/8, train_loss: 0.0147 step time: 0.2949\n", - "2/8, train_loss: 0.0168 step time: 0.2415\n", - "3/8, train_loss: 0.0163 step time: 0.2408\n", - "4/8, train_loss: 0.0154 step time: 0.2446\n", - "5/8, train_loss: 0.0192 step time: 0.2445\n", - "6/8, train_loss: 0.0190 step time: 0.2964\n", - "7/8, train_loss: 0.0165 step time: 0.2926\n", - "8/8, train_loss: 0.0159 step time: 0.2782\n", - "epoch 300 average loss: 0.0167\n", - "current epoch: 300 current mean dice: 0.9538 best mean dice: 0.9541 at epoch: 285\n", - "time consuming of epoch 300 is: 3.3776\n", - "----------\n", - "epoch 301/600\n", - "1/8, train_loss: 0.0161 step time: 0.3009\n", - "2/8, train_loss: 0.0165 step time: 0.2559\n", - "3/8, train_loss: 0.0122 step time: 0.2397\n", - "4/8, train_loss: 0.0192 step time: 0.2401\n", - "5/8, train_loss: 0.0194 step time: 0.2419\n", - "6/8, train_loss: 0.0158 step time: 0.2382\n", - "7/8, train_loss: 0.0160 step time: 0.2207\n", - "8/8, train_loss: 0.0188 step time: 0.3734\n", - "epoch 301 average loss: 0.0168\n", - "time consuming of epoch 301 is: 2.1122\n", - "----------\n", - "epoch 302/600\n", - "1/8, train_loss: 0.0152 step time: 0.5640\n", - "2/8, train_loss: 0.0168 step time: 0.2746\n", - "3/8, train_loss: 0.0235 step time: 0.2620\n", - "4/8, train_loss: 0.0159 step time: 0.2526\n", - "5/8, train_loss: 0.0197 step time: 0.2429\n", - "6/8, train_loss: 0.0177 step time: 0.2453\n", - "7/8, train_loss: 0.0171 step time: 0.2309\n", - "8/8, train_loss: 0.0157 step time: 0.2691\n", - "epoch 302 average loss: 0.0177\n", - "time consuming of epoch 302 is: 2.3429\n", - "----------\n", - "epoch 303/600\n", - "1/8, train_loss: 0.0141 step time: 0.3742\n", - "2/8, train_loss: 0.0144 step time: 0.2751\n", - "3/8, train_loss: 0.0173 step time: 0.2541\n", - "4/8, train_loss: 0.0172 step time: 0.2597\n", - "5/8, train_loss: 0.0220 step time: 0.2565\n", - "6/8, train_loss: 0.0202 step time: 0.2549\n", - "7/8, train_loss: 0.0167 step time: 0.2558\n", - "8/8, train_loss: 0.0165 step time: 0.2521\n", - "epoch 303 average loss: 0.0173\n", - "time consuming of epoch 303 is: 2.1840\n", - "----------\n", - "epoch 304/600\n", - "1/8, train_loss: 0.0157 step time: 0.2934\n", - "2/8, train_loss: 0.0179 step time: 0.2370\n", - "3/8, train_loss: 0.0154 step time: 0.2413\n", - "4/8, train_loss: 0.0168 step time: 0.2559\n", - "5/8, train_loss: 0.0147 step time: 0.2454\n", - "6/8, train_loss: 0.0186 step time: 0.2472\n", - "7/8, train_loss: 0.0229 step time: 0.2388\n", - "8/8, train_loss: 0.0166 step time: 0.2462\n", - "epoch 304 average loss: 0.0173\n", - "time consuming of epoch 304 is: 2.0067\n", - "----------\n", - "epoch 305/600\n", - "1/8, train_loss: 0.0160 step time: 0.2998\n", - "2/8, train_loss: 0.0174 step time: 0.2526\n", - "3/8, train_loss: 0.0161 step time: 0.2402\n", - "4/8, train_loss: 0.0168 step time: 0.2392\n", - "5/8, train_loss: 0.0152 step time: 0.2368\n", - "6/8, train_loss: 0.0170 step time: 0.2338\n", - "7/8, train_loss: 0.0180 step time: 0.2223\n", - "8/8, train_loss: 0.0161 step time: 0.2206\n", - "epoch 305 average loss: 0.0166\n", - "current epoch: 305 current mean dice: 0.9530 best mean dice: 0.9541 at epoch: 285\n", - "time consuming of epoch 305 is: 3.1726\n", - "----------\n", - "epoch 306/600\n", - "1/8, train_loss: 0.0197 step time: 0.3283\n", - "2/8, train_loss: 0.0146 step time: 0.2576\n", - "3/8, train_loss: 0.0135 step time: 0.2712\n", - "4/8, train_loss: 0.0166 step time: 0.2786\n", - "5/8, train_loss: 0.0179 step time: 0.2582\n", - "6/8, train_loss: 0.0175 step time: 0.2541\n", - "7/8, train_loss: 0.0174 step time: 0.2383\n", - "8/8, train_loss: 0.0201 step time: 0.2313\n", - "epoch 306 average loss: 0.0172\n", - "time consuming of epoch 306 is: 2.1187\n", - "----------\n", - "epoch 307/600\n", - "1/8, train_loss: 0.0204 step time: 0.2800\n", - "2/8, train_loss: 0.0149 step time: 0.3638\n", - "3/8, train_loss: 0.0176 step time: 0.4746\n", - "4/8, train_loss: 0.0163 step time: 0.2594\n", - "5/8, train_loss: 0.0160 step time: 0.2584\n", - "6/8, train_loss: 0.0163 step time: 0.2605\n", - "7/8, train_loss: 0.0190 step time: 0.2415\n", - "8/8, train_loss: 0.0158 step time: 0.2334\n", - "epoch 307 average loss: 0.0170\n", - "time consuming of epoch 307 is: 2.3731\n", - "----------\n", - "epoch 308/600\n", - "1/8, train_loss: 0.0213 step time: 0.2851\n", - "2/8, train_loss: 0.0172 step time: 0.2316\n", - "3/8, train_loss: 0.0150 step time: 0.2283\n", - "4/8, train_loss: 0.0133 step time: 0.2507\n", - "5/8, train_loss: 0.0168 step time: 0.2919\n", - "6/8, train_loss: 0.0184 step time: 0.2921\n", - "7/8, train_loss: 0.0117 step time: 0.2865\n", - "8/8, train_loss: 0.0168 step time: 0.2777\n", - "epoch 308 average loss: 0.0163\n", - "time consuming of epoch 308 is: 2.1453\n", - "----------\n", - "epoch 309/600\n", - "1/8, train_loss: 0.0156 step time: 0.3098\n", - "2/8, train_loss: 0.0174 step time: 0.2606\n", - "3/8, train_loss: 0.0164 step time: 0.2545\n", - "4/8, train_loss: 0.0116 step time: 0.2799\n", - "5/8, train_loss: 0.0159 step time: 0.2804\n", - "6/8, train_loss: 0.0149 step time: 0.2641\n", - "7/8, train_loss: 0.0187 step time: 0.2621\n", - "8/8, train_loss: 0.0168 step time: 0.2468\n", - "epoch 309 average loss: 0.0159\n", - "time consuming of epoch 309 is: 2.1597\n", - "----------\n", - "epoch 310/600\n", - "1/8, train_loss: 0.0155 step time: 0.3055\n", - "2/8, train_loss: 0.0181 step time: 0.2536\n", - "3/8, train_loss: 0.0196 step time: 0.2480\n", - "4/8, train_loss: 0.0167 step time: 0.2469\n", - "5/8, train_loss: 0.0177 step time: 0.2479\n", - "6/8, train_loss: 0.0130 step time: 0.2501\n", - "7/8, train_loss: 0.0160 step time: 0.4057\n", - "8/8, train_loss: 0.0157 step time: 0.3632\n", - "epoch 310 average loss: 0.0165\n", - "current epoch: 310 current mean dice: 0.9537 best mean dice: 0.9541 at epoch: 285\n", - "time consuming of epoch 310 is: 3.4805\n", - "----------\n", - "epoch 311/600\n", - "1/8, train_loss: 0.0160 step time: 0.4076\n", - "2/8, train_loss: 0.0193 step time: 0.4807\n", - "3/8, train_loss: 0.0171 step time: 0.2203\n", - "4/8, train_loss: 0.0160 step time: 0.2149\n", - "5/8, train_loss: 0.0179 step time: 0.2135\n", - "6/8, train_loss: 0.0146 step time: 0.2152\n", - "7/8, train_loss: 0.0190 step time: 0.1999\n", - "8/8, train_loss: 0.0166 step time: 0.2003\n", - "epoch 311 average loss: 0.0171\n", - "time consuming of epoch 311 is: 2.1537\n", - "----------\n", - "epoch 312/600\n", - "1/8, train_loss: 0.0149 step time: 0.2668\n", - "2/8, train_loss: 0.0186 step time: 0.2793\n", - "3/8, train_loss: 0.0125 step time: 0.4171\n", - "4/8, train_loss: 0.0154 step time: 0.3036\n", - "5/8, train_loss: 0.0230 step time: 0.2712\n", - "6/8, train_loss: 0.0165 step time: 0.2650\n", - "7/8, train_loss: 0.0160 step time: 0.2622\n", - "8/8, train_loss: 0.0159 step time: 0.2467\n", - "epoch 312 average loss: 0.0166\n", - "time consuming of epoch 312 is: 2.3135\n", - "----------\n", - "epoch 313/600\n", - "1/8, train_loss: 0.0169 step time: 0.2951\n", - "2/8, train_loss: 0.0170 step time: 0.2387\n", - "3/8, train_loss: 0.0159 step time: 0.2358\n", - "4/8, train_loss: 0.0178 step time: 0.2340\n", - "5/8, train_loss: 0.0189 step time: 0.3767\n", - "6/8, train_loss: 0.0134 step time: 0.4929\n", - "7/8, train_loss: 0.0143 step time: 0.2616\n", - "8/8, train_loss: 0.0143 step time: 0.2551\n", - "epoch 313 average loss: 0.0161\n", - "time consuming of epoch 313 is: 2.3915\n", - "----------\n", - "epoch 314/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0171 step time: 0.2927\n", - "2/8, train_loss: 0.0146 step time: 0.2702\n", - "3/8, train_loss: 0.0204 step time: 0.2722\n", - "4/8, train_loss: 0.0194 step time: 0.3976\n", - "5/8, train_loss: 0.0154 step time: 0.2169\n", - "6/8, train_loss: 0.0149 step time: 0.2154\n", - "7/8, train_loss: 0.0165 step time: 0.2091\n", - "8/8, train_loss: 0.0138 step time: 0.2281\n", - "epoch 314 average loss: 0.0165\n", - "time consuming of epoch 314 is: 2.1039\n", - "----------\n", - "epoch 315/600\n", - "1/8, train_loss: 0.0177 step time: 0.3028\n", - "2/8, train_loss: 0.0208 step time: 0.2496\n", - "3/8, train_loss: 0.0155 step time: 0.2453\n", - "4/8, train_loss: 0.0150 step time: 0.2448\n", - "5/8, train_loss: 0.0174 step time: 0.2421\n", - "6/8, train_loss: 0.0142 step time: 0.2440\n", - "7/8, train_loss: 0.0169 step time: 0.2326\n", - "8/8, train_loss: 0.0142 step time: 0.2302\n", - "epoch 315 average loss: 0.0164\n", - "saved new best metric model\n", - "current epoch: 315 current mean dice: 0.9550 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 315 is: 3.7923\n", - "----------\n", - "epoch 316/600\n", - "1/8, train_loss: 0.0169 step time: 0.2719\n", - "2/8, train_loss: 0.0187 step time: 0.2129\n", - "3/8, train_loss: 0.0163 step time: 0.2146\n", - "4/8, train_loss: 0.0146 step time: 0.2142\n", - "5/8, train_loss: 0.0172 step time: 0.2230\n", - "6/8, train_loss: 0.0150 step time: 0.2375\n", - "7/8, train_loss: 0.0180 step time: 0.2617\n", - "8/8, train_loss: 0.0131 step time: 0.3430\n", - "epoch 316 average loss: 0.0162\n", - "time consuming of epoch 316 is: 1.9814\n", - "----------\n", - "epoch 317/600\n", - "1/8, train_loss: 0.0174 step time: 0.3751\n", - "2/8, train_loss: 0.0163 step time: 0.2657\n", - "3/8, train_loss: 0.0153 step time: 0.2534\n", - "4/8, train_loss: 0.0157 step time: 0.2607\n", - "5/8, train_loss: 0.0156 step time: 0.2451\n", - "6/8, train_loss: 0.0135 step time: 0.2443\n", - "7/8, train_loss: 0.0163 step time: 0.2325\n", - "8/8, train_loss: 0.0159 step time: 0.2319\n", - "epoch 317 average loss: 0.0158\n", - "time consuming of epoch 317 is: 2.1101\n", - "----------\n", - "epoch 318/600\n", - "1/8, train_loss: 0.0145 step time: 0.2847\n", - "2/8, train_loss: 0.0181 step time: 0.2329\n", - "3/8, train_loss: 0.0149 step time: 0.2685\n", - "4/8, train_loss: 0.0159 step time: 0.2734\n", - "5/8, train_loss: 0.0176 step time: 0.2621\n", - "6/8, train_loss: 0.0168 step time: 0.2933\n", - "7/8, train_loss: 0.0175 step time: 0.3421\n", - "8/8, train_loss: 0.0148 step time: 0.2931\n", - "epoch 318 average loss: 0.0163\n", - "time consuming of epoch 318 is: 2.2515\n", - "----------\n", - "epoch 319/600\n", - "1/8, train_loss: 0.0158 step time: 0.3126\n", - "2/8, train_loss: 0.0170 step time: 0.2499\n", - "3/8, train_loss: 0.0183 step time: 0.2404\n", - "4/8, train_loss: 0.0178 step time: 0.2350\n", - "5/8, train_loss: 0.0198 step time: 0.2410\n", - "6/8, train_loss: 0.0129 step time: 0.2548\n", - "7/8, train_loss: 0.0148 step time: 0.2483\n", - "8/8, train_loss: 0.0170 step time: 0.2496\n", - "epoch 319 average loss: 0.0167\n", - "time consuming of epoch 319 is: 2.0329\n", - "----------\n", - "epoch 320/600\n", - "1/8, train_loss: 0.0160 step time: 0.2908\n", - "2/8, train_loss: 0.0151 step time: 0.2473\n", - "3/8, train_loss: 0.0149 step time: 0.2403\n", - "4/8, train_loss: 0.0149 step time: 0.2351\n", - "5/8, train_loss: 0.0171 step time: 0.2989\n", - "6/8, train_loss: 0.0187 step time: 0.3598\n", - "7/8, train_loss: 0.0195 step time: 0.2782\n", - "8/8, train_loss: 0.0151 step time: 0.2536\n", - "epoch 320 average loss: 0.0164\n", - "current epoch: 320 current mean dice: 0.9527 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 320 is: 3.6388\n", - "----------\n", - "epoch 321/600\n", - "1/8, train_loss: 0.0140 step time: 0.3192\n", - "2/8, train_loss: 0.0159 step time: 0.2678\n", - "3/8, train_loss: 0.0167 step time: 0.2498\n", - "4/8, train_loss: 0.0167 step time: 0.2566\n", - "5/8, train_loss: 0.0137 step time: 0.2551\n", - "6/8, train_loss: 0.0157 step time: 0.2534\n", - "7/8, train_loss: 0.0161 step time: 0.2344\n", - "8/8, train_loss: 0.0190 step time: 0.2361\n", - "epoch 321 average loss: 0.0160\n", - "time consuming of epoch 321 is: 2.0737\n", - "----------\n", - "epoch 322/600\n", - "1/8, train_loss: 0.0157 step time: 0.2869\n", - "2/8, train_loss: 0.0178 step time: 0.2380\n", - "3/8, train_loss: 0.0160 step time: 0.2330\n", - "4/8, train_loss: 0.0151 step time: 0.3347\n", - "5/8, train_loss: 0.0164 step time: 0.3913\n", - "6/8, train_loss: 0.0184 step time: 0.2656\n", - "7/8, train_loss: 0.0178 step time: 0.2558\n", - "8/8, train_loss: 0.0177 step time: 0.2494\n", - "epoch 322 average loss: 0.0168\n", - "time consuming of epoch 322 is: 2.2563\n", - "----------\n", - "epoch 323/600\n", - "1/8, train_loss: 0.0186 step time: 0.2993\n", - "2/8, train_loss: 0.0141 step time: 0.2474\n", - "3/8, train_loss: 0.0154 step time: 0.3336\n", - "4/8, train_loss: 0.0176 step time: 0.4839\n", - "5/8, train_loss: 0.0148 step time: 0.2655\n", - "6/8, train_loss: 0.0173 step time: 0.2528\n", - "7/8, train_loss: 0.0132 step time: 0.2476\n", - "8/8, train_loss: 0.0180 step time: 0.2365\n", - "epoch 323 average loss: 0.0161\n", - "time consuming of epoch 323 is: 2.3681\n", - "----------\n", - "epoch 324/600\n", - "1/8, train_loss: 0.0127 step time: 0.2881\n", - "2/8, train_loss: 0.0199 step time: 0.2379\n", - "3/8, train_loss: 0.0163 step time: 0.2308\n", - "4/8, train_loss: 0.0144 step time: 0.2319\n", - "5/8, train_loss: 0.0135 step time: 0.2285\n", - "6/8, train_loss: 0.0154 step time: 0.2321\n", - "7/8, train_loss: 0.0195 step time: 0.2185\n", - "8/8, train_loss: 0.0156 step time: 0.2772\n", - "epoch 324 average loss: 0.0159\n", - "time consuming of epoch 324 is: 1.9464\n", - "----------\n", - "epoch 325/600\n", - "1/8, train_loss: 0.0173 step time: 0.3509\n", - "2/8, train_loss: 0.0162 step time: 0.2768\n", - "3/8, train_loss: 0.0163 step time: 0.2567\n", - "4/8, train_loss: 0.0179 step time: 0.2521\n", - "5/8, train_loss: 0.0138 step time: 0.2471\n", - "6/8, train_loss: 0.0147 step time: 0.2418\n", - "7/8, train_loss: 0.0147 step time: 0.2329\n", - "8/8, train_loss: 0.0142 step time: 0.2329\n", - "epoch 325 average loss: 0.0156\n", - "current epoch: 325 current mean dice: 0.9544 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 325 is: 3.7651\n", - "----------\n", - "epoch 326/600\n", - "1/8, train_loss: 0.0158 step time: 0.2944\n", - "2/8, train_loss: 0.0157 step time: 0.2398\n", - "3/8, train_loss: 0.0177 step time: 0.2308\n", - "4/8, train_loss: 0.0148 step time: 0.3600\n", - "5/8, train_loss: 0.0159 step time: 0.4795\n", - "6/8, train_loss: 0.0137 step time: 0.2579\n", - "7/8, train_loss: 0.0184 step time: 0.2400\n", - "8/8, train_loss: 0.0134 step time: 0.2436\n", - "epoch 326 average loss: 0.0157\n", - "time consuming of epoch 326 is: 2.3473\n", - "----------\n", - "epoch 327/600\n", - "1/8, train_loss: 0.0155 step time: 0.2893\n", - "2/8, train_loss: 0.0125 step time: 0.2354\n", - "3/8, train_loss: 0.0173 step time: 0.2760\n", - "4/8, train_loss: 0.0199 step time: 0.2700\n", - "5/8, train_loss: 0.0173 step time: 0.2555\n", - "6/8, train_loss: 0.0164 step time: 0.2525\n", - "7/8, train_loss: 0.0175 step time: 0.2369\n", - "8/8, train_loss: 0.0158 step time: 0.2886\n", - "epoch 327 average loss: 0.0165\n", - "time consuming of epoch 327 is: 2.1057\n", - "----------\n", - "epoch 328/600\n", - "1/8, train_loss: 0.0152 step time: 0.3375\n", - "2/8, train_loss: 0.0133 step time: 0.2672\n", - "3/8, train_loss: 0.0166 step time: 0.2572\n", - "4/8, train_loss: 0.0162 step time: 0.2499\n", - "5/8, train_loss: 0.0165 step time: 0.2447\n", - "6/8, train_loss: 0.0133 step time: 0.2419\n", - "7/8, train_loss: 0.0179 step time: 0.2904\n", - "8/8, train_loss: 0.0143 step time: 0.3346\n", - "epoch 328 average loss: 0.0154\n", - "time consuming of epoch 328 is: 2.2249\n", - "----------\n", - "epoch 329/600\n", - "1/8, train_loss: 0.0164 step time: 0.3121\n", - "2/8, train_loss: 0.0167 step time: 0.2550\n", - "3/8, train_loss: 0.0163 step time: 0.2479\n", - "4/8, train_loss: 0.0192 step time: 0.2420\n", - "5/8, train_loss: 0.0174 step time: 0.2370\n", - "6/8, train_loss: 0.0122 step time: 0.2340\n", - "7/8, train_loss: 0.0178 step time: 0.2268\n", - "8/8, train_loss: 0.0131 step time: 0.2508\n", - "epoch 329 average loss: 0.0161\n", - "time consuming of epoch 329 is: 2.0071\n", - "----------\n", - "epoch 330/600\n", - "1/8, train_loss: 0.0164 step time: 0.3362\n", - "2/8, train_loss: 0.0172 step time: 0.2683\n", - "3/8, train_loss: 0.0160 step time: 0.2529\n", - "4/8, train_loss: 0.0160 step time: 0.2488\n", - "5/8, train_loss: 0.0129 step time: 0.2550\n", - "6/8, train_loss: 0.0190 step time: 0.2469\n", - "7/8, train_loss: 0.0157 step time: 0.2330\n", - "8/8, train_loss: 0.0165 step time: 0.2333\n", - "epoch 330 average loss: 0.0162\n", - "current epoch: 330 current mean dice: 0.9533 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 330 is: 3.0798\n", - "----------\n", - "epoch 331/600\n", - "1/8, train_loss: 0.0145 step time: 0.3729\n", - "2/8, train_loss: 0.0171 step time: 0.3524\n", - "3/8, train_loss: 0.0145 step time: 0.2932\n", - "4/8, train_loss: 0.0136 step time: 0.2801\n", - "5/8, train_loss: 0.0178 step time: 0.2653\n", - "6/8, train_loss: 0.0194 step time: 0.2656\n", - "7/8, train_loss: 0.0145 step time: 0.2476\n", - "8/8, train_loss: 0.0141 step time: 0.2394\n", - "epoch 331 average loss: 0.0157\n", - "time consuming of epoch 331 is: 2.3178\n", - "----------\n", - "epoch 332/600\n", - "1/8, train_loss: 0.0141 step time: 0.2866\n", - "2/8, train_loss: 0.0201 step time: 0.2372\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "3/8, train_loss: 0.0139 step time: 0.2520\n", - "4/8, train_loss: 0.0156 step time: 0.2529\n", - "5/8, train_loss: 0.0185 step time: 0.2522\n", - "6/8, train_loss: 0.0156 step time: 0.3346\n", - "7/8, train_loss: 0.0156 step time: 0.3277\n", - "8/8, train_loss: 0.0153 step time: 0.2651\n", - "epoch 332 average loss: 0.0161\n", - "time consuming of epoch 332 is: 2.2097\n", - "----------\n", - "epoch 333/600\n", - "1/8, train_loss: 0.0139 step time: 0.2742\n", - "2/8, train_loss: 0.0158 step time: 0.2405\n", - "3/8, train_loss: 0.0165 step time: 0.2417\n", - "4/8, train_loss: 0.0157 step time: 0.2326\n", - "5/8, train_loss: 0.0162 step time: 0.2290\n", - "6/8, train_loss: 0.0197 step time: 0.2328\n", - "7/8, train_loss: 0.0162 step time: 0.3553\n", - "8/8, train_loss: 0.0162 step time: 0.6037\n", - "epoch 333 average loss: 0.0163\n", - "time consuming of epoch 333 is: 2.4113\n", - "----------\n", - "epoch 334/600\n", - "1/8, train_loss: 0.0141 step time: 0.2671\n", - "2/8, train_loss: 0.0155 step time: 0.2154\n", - "3/8, train_loss: 0.0162 step time: 0.2203\n", - "4/8, train_loss: 0.0134 step time: 0.2596\n", - "5/8, train_loss: 0.0144 step time: 0.3132\n", - "6/8, train_loss: 0.0173 step time: 0.3066\n", - "7/8, train_loss: 0.0153 step time: 0.2831\n", - "8/8, train_loss: 0.0187 step time: 0.2832\n", - "epoch 334 average loss: 0.0156\n", - "time consuming of epoch 334 is: 2.1500\n", - "----------\n", - "epoch 335/600\n", - "1/8, train_loss: 0.0144 step time: 0.3054\n", - "2/8, train_loss: 0.0220 step time: 0.2596\n", - "3/8, train_loss: 0.0162 step time: 0.2517\n", - "4/8, train_loss: 0.0132 step time: 0.2401\n", - "5/8, train_loss: 0.0161 step time: 0.2900\n", - "6/8, train_loss: 0.0161 step time: 0.2702\n", - "7/8, train_loss: 0.0156 step time: 0.2607\n", - "8/8, train_loss: 0.0181 step time: 0.2477\n", - "epoch 335 average loss: 0.0164\n", - "current epoch: 335 current mean dice: 0.9537 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 335 is: 3.4367\n", - "----------\n", - "epoch 336/600\n", - "1/8, train_loss: 0.0160 step time: 0.3162\n", - "2/8, train_loss: 0.0165 step time: 0.2510\n", - "3/8, train_loss: 0.0172 step time: 0.2521\n", - "4/8, train_loss: 0.0179 step time: 0.2455\n", - "5/8, train_loss: 0.0166 step time: 0.2512\n", - "6/8, train_loss: 0.0127 step time: 0.3738\n", - "7/8, train_loss: 0.0192 step time: 0.3476\n", - "8/8, train_loss: 0.0137 step time: 0.2580\n", - "epoch 336 average loss: 0.0162\n", - "time consuming of epoch 336 is: 2.2966\n", - "----------\n", - "epoch 337/600\n", - "1/8, train_loss: 0.0142 step time: 0.2913\n", - "2/8, train_loss: 0.0168 step time: 0.2475\n", - "3/8, train_loss: 0.0161 step time: 0.2487\n", - "4/8, train_loss: 0.0171 step time: 0.2383\n", - "5/8, train_loss: 0.0149 step time: 0.2416\n", - "6/8, train_loss: 0.0179 step time: 0.2405\n", - "7/8, train_loss: 0.0138 step time: 0.2315\n", - "8/8, train_loss: 0.0193 step time: 0.2330\n", - "epoch 337 average loss: 0.0163\n", - "time consuming of epoch 337 is: 1.9738\n", - "----------\n", - "epoch 338/600\n", - "1/8, train_loss: 0.0165 step time: 0.5242\n", - "2/8, train_loss: 0.0171 step time: 0.2181\n", - "3/8, train_loss: 0.0163 step time: 0.2143\n", - "4/8, train_loss: 0.0150 step time: 0.2149\n", - "5/8, train_loss: 0.0154 step time: 0.2147\n", - "6/8, train_loss: 0.0135 step time: 0.2334\n", - "7/8, train_loss: 0.0184 step time: 0.3538\n", - "8/8, train_loss: 0.0158 step time: 0.3503\n", - "epoch 338 average loss: 0.0160\n", - "time consuming of epoch 338 is: 2.3252\n", - "----------\n", - "epoch 339/600\n", - "1/8, train_loss: 0.0151 step time: 0.3370\n", - "2/8, train_loss: 0.0154 step time: 0.2630\n", - "3/8, train_loss: 0.0173 step time: 0.2522\n", - "4/8, train_loss: 0.0184 step time: 0.2497\n", - "5/8, train_loss: 0.0174 step time: 0.2505\n", - "6/8, train_loss: 0.0159 step time: 0.2455\n", - "7/8, train_loss: 0.0145 step time: 0.2425\n", - "8/8, train_loss: 0.0152 step time: 0.2837\n", - "epoch 339 average loss: 0.0162\n", - "time consuming of epoch 339 is: 2.1256\n", - "----------\n", - "epoch 340/600\n", - "1/8, train_loss: 0.0174 step time: 0.3269\n", - "2/8, train_loss: 0.0138 step time: 0.2648\n", - "3/8, train_loss: 0.0191 step time: 0.2641\n", - "4/8, train_loss: 0.0162 step time: 0.2527\n", - "5/8, train_loss: 0.0152 step time: 0.2547\n", - "6/8, train_loss: 0.0199 step time: 0.2498\n", - "7/8, train_loss: 0.0167 step time: 0.3554\n", - "8/8, train_loss: 0.0163 step time: 0.3365\n", - "epoch 340 average loss: 0.0168\n", - "current epoch: 340 current mean dice: 0.9534 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 340 is: 3.6277\n", - "----------\n", - "epoch 341/600\n", - "1/8, train_loss: 0.0184 step time: 0.2948\n", - "2/8, train_loss: 0.0166 step time: 0.2596\n", - "3/8, train_loss: 0.0163 step time: 0.2471\n", - "4/8, train_loss: 0.0148 step time: 0.2442\n", - "5/8, train_loss: 0.0167 step time: 0.2459\n", - "6/8, train_loss: 0.0172 step time: 0.2356\n", - "7/8, train_loss: 0.0171 step time: 0.2740\n", - "8/8, train_loss: 0.0145 step time: 0.3404\n", - "epoch 341 average loss: 0.0164\n", - "time consuming of epoch 341 is: 2.1428\n", - "----------\n", - "epoch 342/600\n", - "1/8, train_loss: 0.0162 step time: 0.3398\n", - "2/8, train_loss: 0.0165 step time: 0.2592\n", - "3/8, train_loss: 0.0172 step time: 0.2526\n", - "4/8, train_loss: 0.0145 step time: 0.2521\n", - "5/8, train_loss: 0.0173 step time: 0.2464\n", - "6/8, train_loss: 0.0120 step time: 0.2444\n", - "7/8, train_loss: 0.0157 step time: 0.2364\n", - "8/8, train_loss: 0.0165 step time: 0.2738\n", - "epoch 342 average loss: 0.0157\n", - "time consuming of epoch 342 is: 2.1063\n", - "----------\n", - "epoch 343/600\n", - "1/8, train_loss: 0.0153 step time: 0.3473\n", - "2/8, train_loss: 0.0161 step time: 0.2703\n", - "3/8, train_loss: 0.0170 step time: 0.2537\n", - "4/8, train_loss: 0.0177 step time: 0.2515\n", - "5/8, train_loss: 0.0159 step time: 0.2361\n", - "6/8, train_loss: 0.0166 step time: 0.2300\n", - "7/8, train_loss: 0.0142 step time: 0.2173\n", - "8/8, train_loss: 0.0166 step time: 0.2206\n", - "epoch 343 average loss: 0.0162\n", - "time consuming of epoch 343 is: 2.0282\n", - "----------\n", - "epoch 344/600\n", - "1/8, train_loss: 0.0152 step time: 0.3021\n", - "2/8, train_loss: 0.0162 step time: 0.2561\n", - "3/8, train_loss: 0.0157 step time: 0.2471\n", - "4/8, train_loss: 0.0153 step time: 0.2511\n", - "5/8, train_loss: 0.0216 step time: 0.3351\n", - "6/8, train_loss: 0.0147 step time: 0.4696\n", - "7/8, train_loss: 0.0160 step time: 0.2619\n", - "8/8, train_loss: 0.0145 step time: 0.2536\n", - "epoch 344 average loss: 0.0161\n", - "time consuming of epoch 344 is: 2.3780\n", - "----------\n", - "epoch 345/600\n", - "1/8, train_loss: 0.0152 step time: 0.2827\n", - "2/8, train_loss: 0.0150 step time: 0.2927\n", - "3/8, train_loss: 0.0146 step time: 0.2986\n", - "4/8, train_loss: 0.0175 step time: 0.2623\n", - "5/8, train_loss: 0.0193 step time: 0.2488\n", - "6/8, train_loss: 0.0144 step time: 0.2504\n", - "7/8, train_loss: 0.0168 step time: 0.2526\n", - "8/8, train_loss: 0.0168 step time: 0.2481\n", - "epoch 345 average loss: 0.0162\n", - "current epoch: 345 current mean dice: 0.9530 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 345 is: 3.2603\n", - "----------\n", - "epoch 346/600\n", - "1/8, train_loss: 0.0169 step time: 0.2932\n", - "2/8, train_loss: 0.0132 step time: 0.2426\n", - "3/8, train_loss: 0.0151 step time: 0.2454\n", - "4/8, train_loss: 0.0193 step time: 0.2484\n", - "5/8, train_loss: 0.0170 step time: 0.2389\n", - "6/8, train_loss: 0.0161 step time: 0.2422\n", - "7/8, train_loss: 0.0161 step time: 0.2411\n", - "8/8, train_loss: 0.0142 step time: 0.2534\n", - "epoch 346 average loss: 0.0160\n", - "time consuming of epoch 346 is: 2.0063\n", - "----------\n", - "epoch 347/600\n", - "1/8, train_loss: 0.0157 step time: 0.2950\n", - "2/8, train_loss: 0.0168 step time: 0.2553\n", - "3/8, train_loss: 0.0159 step time: 0.2553\n", - "4/8, train_loss: 0.0139 step time: 0.2546\n", - "5/8, train_loss: 0.0198 step time: 0.2574\n", - "6/8, train_loss: 0.0141 step time: 0.2659\n", - "7/8, train_loss: 0.0139 step time: 0.2475\n", - "8/8, train_loss: 0.0166 step time: 0.2465\n", - "epoch 347 average loss: 0.0158\n", - "time consuming of epoch 347 is: 2.0789\n", - "----------\n", - "epoch 348/600\n", - "1/8, train_loss: 0.0156 step time: 0.2964\n", - "2/8, train_loss: 0.0160 step time: 0.2456\n", - "3/8, train_loss: 0.0152 step time: 0.3931\n", - "4/8, train_loss: 0.0148 step time: 0.5779\n", - "5/8, train_loss: 0.0168 step time: 0.2180\n", - "6/8, train_loss: 0.0157 step time: 0.2139\n", - "7/8, train_loss: 0.0147 step time: 0.2019\n", - "8/8, train_loss: 0.0199 step time: 0.2113\n", - "epoch 348 average loss: 0.0161\n", - "time consuming of epoch 348 is: 2.3596\n", - "----------\n", - "epoch 349/600\n", - "1/8, train_loss: 0.0143 step time: 0.6323\n", - "2/8, train_loss: 0.0146 step time: 0.2167\n", - "3/8, train_loss: 0.0200 step time: 0.2155\n", - "4/8, train_loss: 0.0203 step time: 0.3181\n", - "5/8, train_loss: 0.0152 step time: 0.3854\n", - "6/8, train_loss: 0.0162 step time: 0.2761\n", - "7/8, train_loss: 0.0188 step time: 0.2570\n", - "8/8, train_loss: 0.0144 step time: 0.2450\n", - "epoch 349 average loss: 0.0167\n", - "time consuming of epoch 349 is: 2.5480\n", - "----------\n", - "epoch 350/600\n", - "1/8, train_loss: 0.0165 step time: 0.2941\n", - "2/8, train_loss: 0.0118 step time: 0.2465\n", - "3/8, train_loss: 0.0155 step time: 0.2401\n", - "4/8, train_loss: 0.0139 step time: 0.2421\n", - "5/8, train_loss: 0.0177 step time: 0.2392\n", - "6/8, train_loss: 0.0178 step time: 0.2430\n", - "7/8, train_loss: 0.0164 step time: 0.2344\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "8/8, train_loss: 0.0171 step time: 0.2785\n", - "epoch 350 average loss: 0.0159\n", - "current epoch: 350 current mean dice: 0.9548 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 350 is: 3.4705\n", - "----------\n", - "epoch 351/600\n", - "1/8, train_loss: 0.0127 step time: 0.3383\n", - "2/8, train_loss: 0.0161 step time: 0.2705\n", - "3/8, train_loss: 0.0163 step time: 0.2592\n", - "4/8, train_loss: 0.0145 step time: 0.2517\n", - "5/8, train_loss: 0.0181 step time: 0.2441\n", - "6/8, train_loss: 0.0205 step time: 0.2453\n", - "7/8, train_loss: 0.0144 step time: 0.2579\n", - "8/8, train_loss: 0.0169 step time: 0.3034\n", - "epoch 351 average loss: 0.0162\n", - "time consuming of epoch 351 is: 2.1716\n", - "----------\n", - "epoch 352/600\n", - "1/8, train_loss: 0.0144 step time: 0.3206\n", - "2/8, train_loss: 0.0163 step time: 0.2784\n", - "3/8, train_loss: 0.0163 step time: 0.2682\n", - "4/8, train_loss: 0.0162 step time: 0.2678\n", - "5/8, train_loss: 0.0144 step time: 0.2602\n", - "6/8, train_loss: 0.0144 step time: 0.2583\n", - "7/8, train_loss: 0.0142 step time: 0.2523\n", - "8/8, train_loss: 0.0209 step time: 0.2469\n", - "epoch 352 average loss: 0.0159\n", - "time consuming of epoch 352 is: 2.1543\n", - "----------\n", - "epoch 353/600\n", - "1/8, train_loss: 0.0188 step time: 0.2977\n", - "2/8, train_loss: 0.0169 step time: 0.2969\n", - "3/8, train_loss: 0.0138 step time: 0.2977\n", - "4/8, train_loss: 0.0177 step time: 0.2680\n", - "5/8, train_loss: 0.0151 step time: 0.2523\n", - "6/8, train_loss: 0.0154 step time: 0.2526\n", - "7/8, train_loss: 0.0144 step time: 0.2365\n", - "8/8, train_loss: 0.0120 step time: 0.2307\n", - "epoch 353 average loss: 0.0155\n", - "time consuming of epoch 353 is: 2.1340\n", - "----------\n", - "epoch 354/600\n", - "1/8, train_loss: 0.0181 step time: 0.2850\n", - "2/8, train_loss: 0.0187 step time: 0.2351\n", - "3/8, train_loss: 0.0168 step time: 0.2401\n", - "4/8, train_loss: 0.0161 step time: 0.2391\n", - "5/8, train_loss: 0.0142 step time: 0.2367\n", - "6/8, train_loss: 0.0149 step time: 0.2568\n", - "7/8, train_loss: 0.0151 step time: 0.2754\n", - "8/8, train_loss: 0.0157 step time: 0.2873\n", - "epoch 354 average loss: 0.0162\n", - "time consuming of epoch 354 is: 2.0569\n", - "----------\n", - "epoch 355/600\n", - "1/8, train_loss: 0.0150 step time: 0.3079\n", - "2/8, train_loss: 0.0146 step time: 0.2564\n", - "3/8, train_loss: 0.0161 step time: 0.2485\n", - "4/8, train_loss: 0.0148 step time: 0.5369\n", - "5/8, train_loss: 0.0155 step time: 0.3764\n", - "6/8, train_loss: 0.0165 step time: 0.2687\n", - "7/8, train_loss: 0.0188 step time: 0.2582\n", - "8/8, train_loss: 0.0156 step time: 0.2456\n", - "epoch 355 average loss: 0.0159\n", - "current epoch: 355 current mean dice: 0.9534 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 355 is: 3.9862\n", - "----------\n", - "epoch 356/600\n", - "1/8, train_loss: 0.0189 step time: 0.3007\n", - "2/8, train_loss: 0.0161 step time: 0.2521\n", - "3/8, train_loss: 0.0180 step time: 0.2422\n", - "4/8, train_loss: 0.0167 step time: 0.2424\n", - "5/8, train_loss: 0.0133 step time: 0.2357\n", - "6/8, train_loss: 0.0127 step time: 0.3176\n", - "7/8, train_loss: 0.0152 step time: 0.4719\n", - "8/8, train_loss: 0.0139 step time: 0.2672\n", - "epoch 356 average loss: 0.0156\n", - "time consuming of epoch 356 is: 2.3309\n", - "----------\n", - "epoch 357/600\n", - "1/8, train_loss: 0.0151 step time: 0.3156\n", - "2/8, train_loss: 0.0196 step time: 0.2620\n", - "3/8, train_loss: 0.0160 step time: 0.2544\n", - "4/8, train_loss: 0.0148 step time: 0.2505\n", - "5/8, train_loss: 0.0143 step time: 0.2466\n", - "6/8, train_loss: 0.0153 step time: 0.2397\n", - "7/8, train_loss: 0.0173 step time: 0.2333\n", - "8/8, train_loss: 0.0147 step time: 0.3979\n", - "epoch 357 average loss: 0.0159\n", - "time consuming of epoch 357 is: 2.2016\n", - "----------\n", - "epoch 358/600\n", - "1/8, train_loss: 0.0156 step time: 0.5759\n", - "2/8, train_loss: 0.0139 step time: 0.2902\n", - "3/8, train_loss: 0.0171 step time: 0.2640\n", - "4/8, train_loss: 0.0129 step time: 0.2571\n", - "5/8, train_loss: 0.0179 step time: 0.2455\n", - "6/8, train_loss: 0.0182 step time: 0.2353\n", - "7/8, train_loss: 0.0153 step time: 0.3447\n", - "8/8, train_loss: 0.0142 step time: 0.4597\n", - "epoch 358 average loss: 0.0156\n", - "time consuming of epoch 358 is: 2.6741\n", - "----------\n", - "epoch 359/600\n", - "1/8, train_loss: 0.0192 step time: 0.2903\n", - "2/8, train_loss: 0.0136 step time: 0.2468\n", - "3/8, train_loss: 0.0160 step time: 0.2450\n", - "4/8, train_loss: 0.0159 step time: 0.2581\n", - "5/8, train_loss: 0.0136 step time: 0.2629\n", - "6/8, train_loss: 0.0160 step time: 0.2627\n", - "7/8, train_loss: 0.0133 step time: 0.2610\n", - "8/8, train_loss: 0.0170 step time: 0.2444\n", - "epoch 359 average loss: 0.0156\n", - "time consuming of epoch 359 is: 2.0726\n", - "----------\n", - "epoch 360/600\n", - "1/8, train_loss: 0.0155 step time: 0.2931\n", - "2/8, train_loss: 0.0164 step time: 0.3565\n", - "3/8, train_loss: 0.0112 step time: 0.4710\n", - "4/8, train_loss: 0.0165 step time: 0.2581\n", - "5/8, train_loss: 0.0175 step time: 0.2525\n", - "6/8, train_loss: 0.0157 step time: 0.2507\n", - "7/8, train_loss: 0.0156 step time: 0.2530\n", - "8/8, train_loss: 0.0225 step time: 0.2759\n", - "epoch 360 average loss: 0.0164\n", - "current epoch: 360 current mean dice: 0.9540 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 360 is: 3.6059\n", - "----------\n", - "epoch 361/600\n", - "1/8, train_loss: 0.0155 step time: 0.3069\n", - "2/8, train_loss: 0.0168 step time: 0.2614\n", - "3/8, train_loss: 0.0169 step time: 0.2484\n", - "4/8, train_loss: 0.0216 step time: 0.2491\n", - "5/8, train_loss: 0.0132 step time: 0.2387\n", - "6/8, train_loss: 0.0142 step time: 0.2353\n", - "7/8, train_loss: 0.0143 step time: 0.2256\n", - "8/8, train_loss: 0.0148 step time: 0.2249\n", - "epoch 361 average loss: 0.0159\n", - "time consuming of epoch 361 is: 1.9916\n", - "----------\n", - "epoch 362/600\n", - "1/8, train_loss: 0.0190 step time: 0.2729\n", - "2/8, train_loss: 0.0123 step time: 0.2334\n", - "3/8, train_loss: 0.0180 step time: 0.2414\n", - "4/8, train_loss: 0.0178 step time: 0.2703\n", - "5/8, train_loss: 0.0147 step time: 0.2680\n", - "6/8, train_loss: 0.0179 step time: 0.2673\n", - "7/8, train_loss: 0.0157 step time: 0.2512\n", - "8/8, train_loss: 0.0161 step time: 0.2718\n", - "epoch 362 average loss: 0.0165\n", - "time consuming of epoch 362 is: 2.0778\n", - "----------\n", - "epoch 363/600\n", - "1/8, train_loss: 0.0170 step time: 0.3243\n", - "2/8, train_loss: 0.0147 step time: 0.2720\n", - "3/8, train_loss: 0.0162 step time: 0.2528\n", - "4/8, train_loss: 0.0167 step time: 0.2618\n", - "5/8, train_loss: 0.0138 step time: 0.2741\n", - "6/8, train_loss: 0.0141 step time: 0.2561\n", - "7/8, train_loss: 0.0156 step time: 0.2416\n", - "8/8, train_loss: 0.0197 step time: 0.2436\n", - "epoch 363 average loss: 0.0160\n", - "time consuming of epoch 363 is: 2.1277\n", - "----------\n", - "epoch 364/600\n", - "1/8, train_loss: 0.0140 step time: 0.2933\n", - "2/8, train_loss: 0.0153 step time: 0.2355\n", - "3/8, train_loss: 0.0158 step time: 0.2371\n", - "4/8, train_loss: 0.0169 step time: 0.2482\n", - "5/8, train_loss: 0.0133 step time: 0.2458\n", - "6/8, train_loss: 0.0156 step time: 0.2478\n", - "7/8, train_loss: 0.0151 step time: 0.2672\n", - "8/8, train_loss: 0.0207 step time: 0.2886\n", - "epoch 364 average loss: 0.0158\n", - "time consuming of epoch 364 is: 2.0649\n", - "----------\n", - "epoch 365/600\n", - "1/8, train_loss: 0.0177 step time: 0.3120\n", - "2/8, train_loss: 0.0124 step time: 0.2520\n", - "3/8, train_loss: 0.0177 step time: 0.2564\n", - "4/8, train_loss: 0.0159 step time: 0.2647\n", - "5/8, train_loss: 0.0164 step time: 0.2635\n", - "6/8, train_loss: 0.0159 step time: 0.2508\n", - "7/8, train_loss: 0.0186 step time: 0.2472\n", - "8/8, train_loss: 0.0144 step time: 0.2352\n", - "epoch 365 average loss: 0.0161\n", - "current epoch: 365 current mean dice: 0.9534 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 365 is: 3.5257\n", - "----------\n", - "epoch 366/600\n", - "1/8, train_loss: 0.0164 step time: 0.3106\n", - "2/8, train_loss: 0.0127 step time: 0.2712\n", - "3/8, train_loss: 0.0148 step time: 0.2620\n", - "4/8, train_loss: 0.0162 step time: 0.2638\n", - "5/8, train_loss: 0.0154 step time: 0.2612\n", - "6/8, train_loss: 0.0181 step time: 0.2608\n", - "7/8, train_loss: 0.0167 step time: 0.2630\n", - "8/8, train_loss: 0.0166 step time: 0.3877\n", - "epoch 366 average loss: 0.0159\n", - "time consuming of epoch 366 is: 2.2816\n", - "----------\n", - "epoch 367/600\n", - "1/8, train_loss: 0.0120 step time: 0.4936\n", - "2/8, train_loss: 0.0153 step time: 0.3088\n", - "3/8, train_loss: 0.0182 step time: 0.2684\n", - "4/8, train_loss: 0.0179 step time: 0.2731\n", - "5/8, train_loss: 0.0180 step time: 0.2584\n", - "6/8, train_loss: 0.0171 step time: 0.2477\n", - "7/8, train_loss: 0.0152 step time: 0.2840\n", - "8/8, train_loss: 0.0134 step time: 0.3050\n", - "epoch 367 average loss: 0.0159\n", - "time consuming of epoch 367 is: 2.4404\n", - "----------\n", - "epoch 368/600\n", - "1/8, train_loss: 0.0165 step time: 0.3092\n", - "2/8, train_loss: 0.0167 step time: 0.2499\n", - "3/8, train_loss: 0.0155 step time: 0.2436\n", - "4/8, train_loss: 0.0180 step time: 0.2481\n", - "5/8, train_loss: 0.0167 step time: 0.2472\n", - "6/8, train_loss: 0.0156 step time: 0.2426\n", - "7/8, train_loss: 0.0168 step time: 0.2357\n", - "8/8, train_loss: 0.0157 step time: 0.2213\n", - "epoch 368 average loss: 0.0164\n", - "time consuming of epoch 368 is: 1.9990\n", - "----------\n", - "epoch 369/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0142 step time: 0.5772\n", - "2/8, train_loss: 0.0164 step time: 0.2198\n", - "3/8, train_loss: 0.0154 step time: 0.2141\n", - "4/8, train_loss: 0.0189 step time: 0.2100\n", - "5/8, train_loss: 0.0154 step time: 0.2246\n", - "6/8, train_loss: 0.0169 step time: 0.2283\n", - "7/8, train_loss: 0.0148 step time: 0.2202\n", - "8/8, train_loss: 0.0162 step time: 0.2609\n", - "epoch 369 average loss: 0.0160\n", - "time consuming of epoch 369 is: 2.1565\n", - "----------\n", - "epoch 370/600\n", - "1/8, train_loss: 0.0152 step time: 0.3266\n", - "2/8, train_loss: 0.0165 step time: 0.2921\n", - "3/8, train_loss: 0.0181 step time: 0.2822\n", - "4/8, train_loss: 0.0143 step time: 0.2577\n", - "5/8, train_loss: 0.0144 step time: 0.3602\n", - "6/8, train_loss: 0.0171 step time: 0.3438\n", - "7/8, train_loss: 0.0167 step time: 0.3003\n", - "8/8, train_loss: 0.0130 step time: 0.2783\n", - "epoch 370 average loss: 0.0157\n", - "current epoch: 370 current mean dice: 0.9530 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 370 is: 3.7704\n", - "----------\n", - "epoch 371/600\n", - "1/8, train_loss: 0.0139 step time: 0.3272\n", - "2/8, train_loss: 0.0133 step time: 0.2745\n", - "3/8, train_loss: 0.0170 step time: 0.2601\n", - "4/8, train_loss: 0.0125 step time: 0.3054\n", - "5/8, train_loss: 0.0148 step time: 0.3146\n", - "6/8, train_loss: 0.0193 step time: 0.2895\n", - "7/8, train_loss: 0.0179 step time: 0.2939\n", - "8/8, train_loss: 0.0131 step time: 0.2654\n", - "epoch 371 average loss: 0.0152\n", - "time consuming of epoch 371 is: 2.3320\n", - "----------\n", - "epoch 372/600\n", - "1/8, train_loss: 0.0135 step time: 0.3035\n", - "2/8, train_loss: 0.0154 step time: 0.2586\n", - "3/8, train_loss: 0.0117 step time: 0.2927\n", - "4/8, train_loss: 0.0157 step time: 0.2849\n", - "5/8, train_loss: 0.0159 step time: 0.2856\n", - "6/8, train_loss: 0.0149 step time: 0.2898\n", - "7/8, train_loss: 0.0159 step time: 0.2658\n", - "8/8, train_loss: 0.0188 step time: 0.2715\n", - "epoch 372 average loss: 0.0152\n", - "time consuming of epoch 372 is: 2.2539\n", - "----------\n", - "epoch 373/600\n", - "1/8, train_loss: 0.0172 step time: 0.3073\n", - "2/8, train_loss: 0.0122 step time: 0.2773\n", - "3/8, train_loss: 0.0157 step time: 0.2702\n", - "4/8, train_loss: 0.0156 step time: 0.2581\n", - "5/8, train_loss: 0.0168 step time: 0.2567\n", - "6/8, train_loss: 0.0180 step time: 0.2499\n", - "7/8, train_loss: 0.0151 step time: 0.2600\n", - "8/8, train_loss: 0.0158 step time: 0.2707\n", - "epoch 373 average loss: 0.0158\n", - "time consuming of epoch 373 is: 2.1516\n", - "----------\n", - "epoch 374/600\n", - "1/8, train_loss: 0.0151 step time: 0.3141\n", - "2/8, train_loss: 0.0154 step time: 0.2563\n", - "3/8, train_loss: 0.0140 step time: 0.2523\n", - "4/8, train_loss: 0.0147 step time: 0.2393\n", - "5/8, train_loss: 0.0143 step time: 0.2388\n", - "6/8, train_loss: 0.0166 step time: 0.2484\n", - "7/8, train_loss: 0.0163 step time: 0.2914\n", - "8/8, train_loss: 0.0182 step time: 0.3397\n", - "epoch 374 average loss: 0.0156\n", - "time consuming of epoch 374 is: 2.1819\n", - "----------\n", - "epoch 375/600\n", - "1/8, train_loss: 0.0183 step time: 0.3307\n", - "2/8, train_loss: 0.0165 step time: 0.2600\n", - "3/8, train_loss: 0.0154 step time: 0.4071\n", - "4/8, train_loss: 0.0150 step time: 0.4694\n", - "5/8, train_loss: 0.0187 step time: 0.2161\n", - "6/8, train_loss: 0.0158 step time: 0.2159\n", - "7/8, train_loss: 0.0141 step time: 0.2003\n", - "8/8, train_loss: 0.0118 step time: 0.2076\n", - "epoch 375 average loss: 0.0157\n", - "current epoch: 375 current mean dice: 0.9549 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 375 is: 3.6765\n", - "----------\n", - "epoch 376/600\n", - "1/8, train_loss: 0.0157 step time: 0.3503\n", - "2/8, train_loss: 0.0171 step time: 0.2703\n", - "3/8, train_loss: 0.0148 step time: 0.2553\n", - "4/8, train_loss: 0.0153 step time: 0.2566\n", - "5/8, train_loss: 0.0144 step time: 0.2491\n", - "6/8, train_loss: 0.0123 step time: 0.2520\n", - "7/8, train_loss: 0.0136 step time: 0.2461\n", - "8/8, train_loss: 0.0163 step time: 0.2437\n", - "epoch 376 average loss: 0.0149\n", - "time consuming of epoch 376 is: 2.1247\n", - "----------\n", - "epoch 377/600\n", - "1/8, train_loss: 0.0179 step time: 0.2929\n", - "2/8, train_loss: 0.0148 step time: 0.2495\n", - "3/8, train_loss: 0.0162 step time: 0.3006\n", - "4/8, train_loss: 0.0142 step time: 0.3436\n", - "5/8, train_loss: 0.0152 step time: 0.2696\n", - "6/8, train_loss: 0.0179 step time: 0.2552\n", - "7/8, train_loss: 0.0133 step time: 0.2392\n", - "8/8, train_loss: 0.0131 step time: 0.2409\n", - "epoch 377 average loss: 0.0153\n", - "time consuming of epoch 377 is: 2.1929\n", - "----------\n", - "epoch 378/600\n", - "1/8, train_loss: 0.0138 step time: 0.2944\n", - "2/8, train_loss: 0.0154 step time: 0.2719\n", - "3/8, train_loss: 0.0181 step time: 0.3119\n", - "4/8, train_loss: 0.0142 step time: 0.2746\n", - "5/8, train_loss: 0.0131 step time: 0.2830\n", - "6/8, train_loss: 0.0164 step time: 0.2782\n", - "7/8, train_loss: 0.0134 step time: 0.2583\n", - "8/8, train_loss: 0.0179 step time: 0.2814\n", - "epoch 378 average loss: 0.0153\n", - "time consuming of epoch 378 is: 2.2552\n", - "----------\n", - "epoch 379/600\n", - "1/8, train_loss: 0.0138 step time: 0.3219\n", - "2/8, train_loss: 0.0168 step time: 0.2821\n", - "3/8, train_loss: 0.0136 step time: 0.2629\n", - "4/8, train_loss: 0.0135 step time: 0.2545\n", - "5/8, train_loss: 0.0160 step time: 0.2511\n", - "6/8, train_loss: 0.0171 step time: 0.3398\n", - "7/8, train_loss: 0.0154 step time: 0.3492\n", - "8/8, train_loss: 0.0151 step time: 0.2742\n", - "epoch 379 average loss: 0.0152\n", - "time consuming of epoch 379 is: 2.3372\n", - "----------\n", - "epoch 380/600\n", - "1/8, train_loss: 0.0128 step time: 0.3108\n", - "2/8, train_loss: 0.0136 step time: 0.3037\n", - "3/8, train_loss: 0.0159 step time: 0.3099\n", - "4/8, train_loss: 0.0168 step time: 0.2803\n", - "5/8, train_loss: 0.0142 step time: 0.2685\n", - "6/8, train_loss: 0.0154 step time: 0.2699\n", - "7/8, train_loss: 0.0159 step time: 0.2937\n", - "8/8, train_loss: 0.0177 step time: 0.3155\n", - "epoch 380 average loss: 0.0153\n", - "current epoch: 380 current mean dice: 0.9545 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 380 is: 3.7774\n", - "----------\n", - "epoch 381/600\n", - "1/8, train_loss: 0.0160 step time: 0.3051\n", - "2/8, train_loss: 0.0135 step time: 0.4550\n", - "3/8, train_loss: 0.0181 step time: 0.5773\n", - "4/8, train_loss: 0.0123 step time: 0.2581\n", - "5/8, train_loss: 0.0144 step time: 0.2494\n", - "6/8, train_loss: 0.0186 step time: 0.2434\n", - "7/8, train_loss: 0.0157 step time: 0.2368\n", - "8/8, train_loss: 0.0176 step time: 0.2387\n", - "epoch 381 average loss: 0.0158\n", - "time consuming of epoch 381 is: 2.5650\n", - "----------\n", - "epoch 382/600\n", - "1/8, train_loss: 0.0165 step time: 0.2908\n", - "2/8, train_loss: 0.0132 step time: 0.2466\n", - "3/8, train_loss: 0.0162 step time: 0.2643\n", - "4/8, train_loss: 0.0153 step time: 0.2550\n", - "5/8, train_loss: 0.0165 step time: 0.2875\n", - "6/8, train_loss: 0.0176 step time: 0.3074\n", - "7/8, train_loss: 0.0146 step time: 0.2976\n", - "8/8, train_loss: 0.0146 step time: 0.2763\n", - "epoch 382 average loss: 0.0156\n", - "time consuming of epoch 382 is: 2.2271\n", - "----------\n", - "epoch 383/600\n", - "1/8, train_loss: 0.0186 step time: 0.3163\n", - "2/8, train_loss: 0.0128 step time: 0.2694\n", - "3/8, train_loss: 0.0150 step time: 0.2650\n", - "4/8, train_loss: 0.0154 step time: 0.2550\n", - "5/8, train_loss: 0.0163 step time: 0.2577\n", - "6/8, train_loss: 0.0153 step time: 0.3754\n", - "7/8, train_loss: 0.0184 step time: 0.5047\n", - "8/8, train_loss: 0.0149 step time: 0.2791\n", - "epoch 383 average loss: 0.0159\n", - "time consuming of epoch 383 is: 2.5240\n", - "----------\n", - "epoch 384/600\n", - "1/8, train_loss: 0.0152 step time: 0.3062\n", - "2/8, train_loss: 0.0159 step time: 0.2450\n", - "3/8, train_loss: 0.0123 step time: 0.2423\n", - "4/8, train_loss: 0.0216 step time: 0.2420\n", - "5/8, train_loss: 0.0156 step time: 0.2399\n", - "6/8, train_loss: 0.0138 step time: 0.2488\n", - "7/8, train_loss: 0.0155 step time: 0.2780\n", - "8/8, train_loss: 0.0146 step time: 0.2648\n", - "epoch 384 average loss: 0.0156\n", - "time consuming of epoch 384 is: 2.0685\n", - "----------\n", - "epoch 385/600\n", - "1/8, train_loss: 0.0128 step time: 0.3028\n", - "2/8, train_loss: 0.0172 step time: 0.2529\n", - "3/8, train_loss: 0.0147 step time: 0.2602\n", - "4/8, train_loss: 0.0134 step time: 0.3407\n", - "5/8, train_loss: 0.0208 step time: 0.3611\n", - "6/8, train_loss: 0.0173 step time: 0.2977\n", - "7/8, train_loss: 0.0155 step time: 0.2753\n", - "8/8, train_loss: 0.0152 step time: 0.2572\n", - "epoch 385 average loss: 0.0159\n", - "current epoch: 385 current mean dice: 0.9546 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 385 is: 3.8613\n", - "----------\n", - "epoch 386/600\n", - "1/8, train_loss: 0.0148 step time: 0.3023\n", - "2/8, train_loss: 0.0167 step time: 0.2471\n", - "3/8, train_loss: 0.0160 step time: 0.3439\n", - "4/8, train_loss: 0.0155 step time: 0.4452\n", - "5/8, train_loss: 0.0144 step time: 0.2517\n", - "6/8, train_loss: 0.0145 step time: 0.2453\n", - "7/8, train_loss: 0.0180 step time: 0.2347\n", - "8/8, train_loss: 0.0133 step time: 0.2623\n", - "epoch 386 average loss: 0.0154\n", - "time consuming of epoch 386 is: 2.3338\n", - "----------\n", - "epoch 387/600\n", - "1/8, train_loss: 0.0180 step time: 0.3221\n", - "2/8, train_loss: 0.0147 step time: 0.2618\n", - "3/8, train_loss: 0.0136 step time: 0.2546\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "4/8, train_loss: 0.0203 step time: 0.2452\n", - "5/8, train_loss: 0.0159 step time: 0.3166\n", - "6/8, train_loss: 0.0128 step time: 0.3154\n", - "7/8, train_loss: 0.0160 step time: 0.2815\n", - "8/8, train_loss: 0.0161 step time: 0.3321\n", - "epoch 387 average loss: 0.0159\n", - "time consuming of epoch 387 is: 2.3308\n", - "----------\n", - "epoch 388/600\n", - "1/8, train_loss: 0.0143 step time: 0.3659\n", - "2/8, train_loss: 0.0180 step time: 0.2696\n", - "3/8, train_loss: 0.0151 step time: 0.2619\n", - "4/8, train_loss: 0.0135 step time: 0.2646\n", - "5/8, train_loss: 0.0189 step time: 0.2643\n", - "6/8, train_loss: 0.0159 step time: 0.2616\n", - "7/8, train_loss: 0.0147 step time: 0.2615\n", - "8/8, train_loss: 0.0142 step time: 0.2431\n", - "epoch 388 average loss: 0.0156\n", - "time consuming of epoch 388 is: 2.1941\n", - "----------\n", - "epoch 389/600\n", - "1/8, train_loss: 0.0150 step time: 0.2977\n", - "2/8, train_loss: 0.0139 step time: 0.2454\n", - "3/8, train_loss: 0.0157 step time: 0.2575\n", - "4/8, train_loss: 0.0188 step time: 0.2648\n", - "5/8, train_loss: 0.0157 step time: 0.2622\n", - "6/8, train_loss: 0.0143 step time: 0.2621\n", - "7/8, train_loss: 0.0131 step time: 0.2654\n", - "8/8, train_loss: 0.0143 step time: 0.2777\n", - "epoch 389 average loss: 0.0151\n", - "time consuming of epoch 389 is: 2.1343\n", - "----------\n", - "epoch 390/600\n", - "1/8, train_loss: 0.0139 step time: 0.3109\n", - "2/8, train_loss: 0.0151 step time: 0.2510\n", - "3/8, train_loss: 0.0164 step time: 0.2516\n", - "4/8, train_loss: 0.0132 step time: 0.2506\n", - "5/8, train_loss: 0.0175 step time: 0.2685\n", - "6/8, train_loss: 0.0152 step time: 0.2872\n", - "7/8, train_loss: 0.0134 step time: 0.2670\n", - "8/8, train_loss: 0.0178 step time: 0.4122\n", - "epoch 390 average loss: 0.0153\n", - "current epoch: 390 current mean dice: 0.9549 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 390 is: 3.4081\n", - "----------\n", - "epoch 391/600\n", - "1/8, train_loss: 0.0146 step time: 0.4010\n", - "2/8, train_loss: 0.0143 step time: 0.3074\n", - "3/8, train_loss: 0.0158 step time: 0.2771\n", - "4/8, train_loss: 0.0152 step time: 0.2679\n", - "5/8, train_loss: 0.0132 step time: 0.2797\n", - "6/8, train_loss: 0.0165 step time: 0.2778\n", - "7/8, train_loss: 0.0163 step time: 0.2668\n", - "8/8, train_loss: 0.0151 step time: 0.2764\n", - "epoch 391 average loss: 0.0151\n", - "time consuming of epoch 391 is: 2.3552\n", - "----------\n", - "epoch 392/600\n", - "1/8, train_loss: 0.0160 step time: 0.3073\n", - "2/8, train_loss: 0.0168 step time: 0.2589\n", - "3/8, train_loss: 0.0158 step time: 0.2681\n", - "4/8, train_loss: 0.0184 step time: 0.3522\n", - "5/8, train_loss: 0.0145 step time: 0.3616\n", - "6/8, train_loss: 0.0144 step time: 0.3071\n", - "7/8, train_loss: 0.0168 step time: 0.2716\n", - "8/8, train_loss: 0.0171 step time: 0.2585\n", - "epoch 392 average loss: 0.0162\n", - "time consuming of epoch 392 is: 2.3868\n", - "----------\n", - "epoch 393/600\n", - "1/8, train_loss: 0.0156 step time: 0.2998\n", - "2/8, train_loss: 0.0158 step time: 0.2499\n", - "3/8, train_loss: 0.0144 step time: 0.3094\n", - "4/8, train_loss: 0.0144 step time: 0.4679\n", - "5/8, train_loss: 0.0167 step time: 0.2621\n", - "6/8, train_loss: 0.0147 step time: 0.2595\n", - "7/8, train_loss: 0.0142 step time: 0.2515\n", - "8/8, train_loss: 0.0174 step time: 0.2473\n", - "epoch 393 average loss: 0.0154\n", - "time consuming of epoch 393 is: 2.3489\n", - "----------\n", - "epoch 394/600\n", - "1/8, train_loss: 0.0144 step time: 0.2989\n", - "2/8, train_loss: 0.0170 step time: 0.2422\n", - "3/8, train_loss: 0.0166 step time: 0.4072\n", - "4/8, train_loss: 0.0143 step time: 0.6108\n", - "5/8, train_loss: 0.0178 step time: 0.2654\n", - "6/8, train_loss: 0.0148 step time: 0.2608\n", - "7/8, train_loss: 0.0126 step time: 0.2480\n", - "8/8, train_loss: 0.0146 step time: 0.2453\n", - "epoch 394 average loss: 0.0153\n", - "time consuming of epoch 394 is: 2.5800\n", - "----------\n", - "epoch 395/600\n", - "1/8, train_loss: 0.0168 step time: 0.2927\n", - "2/8, train_loss: 0.0160 step time: 0.2369\n", - "3/8, train_loss: 0.0159 step time: 0.2586\n", - "4/8, train_loss: 0.0149 step time: 0.3128\n", - "5/8, train_loss: 0.0159 step time: 0.2732\n", - "6/8, train_loss: 0.0145 step time: 0.2625\n", - "7/8, train_loss: 0.0180 step time: 0.2609\n", - "8/8, train_loss: 0.0121 step time: 0.2754\n", - "epoch 395 average loss: 0.0155\n", - "current epoch: 395 current mean dice: 0.9545 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 395 is: 3.5276\n", - "----------\n", - "epoch 396/600\n", - "1/8, train_loss: 0.0161 step time: 0.3185\n", - "2/8, train_loss: 0.0118 step time: 0.2689\n", - "3/8, train_loss: 0.0145 step time: 0.2528\n", - "4/8, train_loss: 0.0133 step time: 0.2825\n", - "5/8, train_loss: 0.0147 step time: 0.2658\n", - "6/8, train_loss: 0.0180 step time: 0.2538\n", - "7/8, train_loss: 0.0151 step time: 0.2560\n", - "8/8, train_loss: 0.0159 step time: 0.2392\n", - "epoch 396 average loss: 0.0149\n", - "time consuming of epoch 396 is: 2.1388\n", - "----------\n", - "epoch 397/600\n", - "1/8, train_loss: 0.0143 step time: 0.2840\n", - "2/8, train_loss: 0.0139 step time: 0.2395\n", - "3/8, train_loss: 0.0180 step time: 0.2782\n", - "4/8, train_loss: 0.0160 step time: 0.2807\n", - "5/8, train_loss: 0.0153 step time: 0.2899\n", - "6/8, train_loss: 0.0160 step time: 0.2610\n", - "7/8, train_loss: 0.0184 step time: 0.2605\n", - "8/8, train_loss: 0.0128 step time: 0.4343\n", - "epoch 397 average loss: 0.0156\n", - "time consuming of epoch 397 is: 2.3295\n", - "----------\n", - "epoch 398/600\n", - "1/8, train_loss: 0.0146 step time: 0.3537\n", - "2/8, train_loss: 0.0164 step time: 0.2670\n", - "3/8, train_loss: 0.0145 step time: 0.2563\n", - "4/8, train_loss: 0.0141 step time: 0.2525\n", - "5/8, train_loss: 0.0166 step time: 0.3733\n", - "6/8, train_loss: 0.0231 step time: 0.5910\n", - "7/8, train_loss: 0.0114 step time: 0.2710\n", - "8/8, train_loss: 0.0158 step time: 0.2522\n", - "epoch 398 average loss: 0.0158\n", - "time consuming of epoch 398 is: 2.6184\n", - "----------\n", - "epoch 399/600\n", - "1/8, train_loss: 0.0131 step time: 0.2914\n", - "2/8, train_loss: 0.0147 step time: 0.2418\n", - "3/8, train_loss: 0.0189 step time: 0.2429\n", - "4/8, train_loss: 0.0134 step time: 0.2470\n", - "5/8, train_loss: 0.0157 step time: 0.2693\n", - "6/8, train_loss: 0.0181 step time: 0.2739\n", - "7/8, train_loss: 0.0166 step time: 0.2638\n", - "8/8, train_loss: 0.0175 step time: 0.3156\n", - "epoch 399 average loss: 0.0160\n", - "time consuming of epoch 399 is: 2.1471\n", - "----------\n", - "epoch 400/600\n", - "1/8, train_loss: 0.0162 step time: 0.3245\n", - "2/8, train_loss: 0.0130 step time: 0.2655\n", - "3/8, train_loss: 0.0130 step time: 0.2615\n", - "4/8, train_loss: 0.0174 step time: 0.2600\n", - "5/8, train_loss: 0.0157 step time: 0.2575\n", - "6/8, train_loss: 0.0132 step time: 0.2484\n", - "7/8, train_loss: 0.0160 step time: 0.2443\n", - "8/8, train_loss: 0.0173 step time: 0.2419\n", - "epoch 400 average loss: 0.0152\n", - "current epoch: 400 current mean dice: 0.9546 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 400 is: 3.3862\n", - "----------\n", - "epoch 401/600\n", - "1/8, train_loss: 0.0148 step time: 0.3138\n", - "2/8, train_loss: 0.0151 step time: 0.2786\n", - "3/8, train_loss: 0.0184 step time: 0.2563\n", - "4/8, train_loss: 0.0149 step time: 0.2416\n", - "5/8, train_loss: 0.0114 step time: 0.2539\n", - "6/8, train_loss: 0.0172 step time: 0.2786\n", - "7/8, train_loss: 0.0167 step time: 0.2984\n", - "8/8, train_loss: 0.0186 step time: 0.2876\n", - "epoch 401 average loss: 0.0159\n", - "time consuming of epoch 401 is: 2.2100\n", - "----------\n", - "epoch 402/600\n", - "1/8, train_loss: 0.0173 step time: 0.3061\n", - "2/8, train_loss: 0.0171 step time: 0.2479\n", - "3/8, train_loss: 0.0170 step time: 0.3799\n", - "4/8, train_loss: 0.0115 step time: 0.3250\n", - "5/8, train_loss: 0.0136 step time: 0.2643\n", - "6/8, train_loss: 0.0192 step time: 0.2593\n", - "7/8, train_loss: 0.0153 step time: 0.2490\n", - "8/8, train_loss: 0.0140 step time: 0.2538\n", - "epoch 402 average loss: 0.0156\n", - "time consuming of epoch 402 is: 2.2865\n", - "----------\n", - "epoch 403/600\n", - "1/8, train_loss: 0.0134 step time: 0.3023\n", - "2/8, train_loss: 0.0169 step time: 0.2708\n", - "3/8, train_loss: 0.0168 step time: 0.2569\n", - "4/8, train_loss: 0.0180 step time: 0.2553\n", - "5/8, train_loss: 0.0171 step time: 0.2545\n", - "6/8, train_loss: 0.0131 step time: 0.2498\n", - "7/8, train_loss: 0.0131 step time: 0.2310\n", - "8/8, train_loss: 0.0160 step time: 0.2655\n", - "epoch 403 average loss: 0.0155\n", - "time consuming of epoch 403 is: 2.0875\n", - "----------\n", - "epoch 404/600\n", - "1/8, train_loss: 0.0151 step time: 0.3040\n", - "2/8, train_loss: 0.0144 step time: 0.2564\n", - "3/8, train_loss: 0.0174 step time: 0.2561\n", - "4/8, train_loss: 0.0144 step time: 0.2462\n", - "5/8, train_loss: 0.0127 step time: 0.2466\n", - "6/8, train_loss: 0.0204 step time: 0.3603\n", - "7/8, train_loss: 0.0223 step time: 0.4051\n", - "8/8, train_loss: 0.0140 step time: 0.2636\n", - "epoch 404 average loss: 0.0163\n", - "time consuming of epoch 404 is: 2.3398\n", - "----------\n", - "epoch 405/600\n", - "1/8, train_loss: 0.0136 step time: 0.2991\n", - "2/8, train_loss: 0.0185 step time: 0.2408\n", - "3/8, train_loss: 0.0145 step time: 0.2357\n", - "4/8, train_loss: 0.0145 step time: 0.2390\n", - "5/8, train_loss: 0.0171 step time: 0.2425\n", - "6/8, train_loss: 0.0149 step time: 0.2451\n", - "7/8, train_loss: 0.0173 step time: 0.2384\n", - "8/8, train_loss: 0.0158 step time: 0.2662\n", - "epoch 405 average loss: 0.0158\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "current epoch: 405 current mean dice: 0.9521 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 405 is: 3.3828\n", - "----------\n", - "epoch 406/600\n", - "1/8, train_loss: 0.0149 step time: 0.3092\n", - "2/8, train_loss: 0.0130 step time: 0.2433\n", - "3/8, train_loss: 0.0150 step time: 0.2442\n", - "4/8, train_loss: 0.0133 step time: 0.2565\n", - "5/8, train_loss: 0.0169 step time: 0.3303\n", - "6/8, train_loss: 0.0163 step time: 0.3752\n", - "7/8, train_loss: 0.0158 step time: 0.2776\n", - "8/8, train_loss: 0.0177 step time: 0.2467\n", - "epoch 406 average loss: 0.0154\n", - "time consuming of epoch 406 is: 2.2842\n", - "----------\n", - "epoch 407/600\n", - "1/8, train_loss: 0.0133 step time: 0.2920\n", - "2/8, train_loss: 0.0134 step time: 0.2402\n", - "3/8, train_loss: 0.0151 step time: 0.2528\n", - "4/8, train_loss: 0.0154 step time: 0.2808\n", - "5/8, train_loss: 0.0164 step time: 0.2793\n", - "6/8, train_loss: 0.0223 step time: 0.2849\n", - "7/8, train_loss: 0.0142 step time: 0.2818\n", - "8/8, train_loss: 0.0132 step time: 0.2767\n", - "epoch 407 average loss: 0.0154\n", - "time consuming of epoch 407 is: 2.1900\n", - "----------\n", - "epoch 408/600\n", - "1/8, train_loss: 0.0112 step time: 0.3067\n", - "2/8, train_loss: 0.0164 step time: 0.2575\n", - "3/8, train_loss: 0.0169 step time: 0.2536\n", - "4/8, train_loss: 0.0149 step time: 0.2518\n", - "5/8, train_loss: 0.0193 step time: 0.2544\n", - "6/8, train_loss: 0.0131 step time: 0.2848\n", - "7/8, train_loss: 0.0155 step time: 0.2828\n", - "8/8, train_loss: 0.0150 step time: 0.2550\n", - "epoch 408 average loss: 0.0153\n", - "time consuming of epoch 408 is: 2.1480\n", - "----------\n", - "epoch 409/600\n", - "1/8, train_loss: 0.0151 step time: 0.2982\n", - "2/8, train_loss: 0.0107 step time: 0.2420\n", - "3/8, train_loss: 0.0151 step time: 0.2432\n", - "4/8, train_loss: 0.0167 step time: 0.3666\n", - "5/8, train_loss: 0.0173 step time: 0.4909\n", - "6/8, train_loss: 0.0131 step time: 0.2915\n", - "7/8, train_loss: 0.0162 step time: 0.2730\n", - "8/8, train_loss: 0.0162 step time: 0.2562\n", - "epoch 409 average loss: 0.0150\n", - "time consuming of epoch 409 is: 2.4631\n", - "----------\n", - "epoch 410/600\n", - "1/8, train_loss: 0.0143 step time: 0.2960\n", - "2/8, train_loss: 0.0140 step time: 0.2377\n", - "3/8, train_loss: 0.0205 step time: 0.2342\n", - "4/8, train_loss: 0.0143 step time: 0.3479\n", - "5/8, train_loss: 0.0143 step time: 0.4900\n", - "6/8, train_loss: 0.0160 step time: 0.2794\n", - "7/8, train_loss: 0.0153 step time: 0.2531\n", - "8/8, train_loss: 0.0141 step time: 0.2404\n", - "epoch 410 average loss: 0.0153\n", - "current epoch: 410 current mean dice: 0.9546 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 410 is: 3.6006\n", - "----------\n", - "epoch 411/600\n", - "1/8, train_loss: 0.0115 step time: 0.3314\n", - "2/8, train_loss: 0.0139 step time: 0.2685\n", - "3/8, train_loss: 0.0159 step time: 0.2588\n", - "4/8, train_loss: 0.0147 step time: 0.2547\n", - "5/8, train_loss: 0.0175 step time: 0.2501\n", - "6/8, train_loss: 0.0153 step time: 0.2433\n", - "7/8, train_loss: 0.0148 step time: 0.2320\n", - "8/8, train_loss: 0.0169 step time: 0.2365\n", - "epoch 411 average loss: 0.0151\n", - "time consuming of epoch 411 is: 2.0764\n", - "----------\n", - "epoch 412/600\n", - "1/8, train_loss: 0.0132 step time: 0.5415\n", - "2/8, train_loss: 0.0141 step time: 0.2185\n", - "3/8, train_loss: 0.0175 step time: 0.2162\n", - "4/8, train_loss: 0.0159 step time: 0.2166\n", - "5/8, train_loss: 0.0152 step time: 0.2214\n", - "6/8, train_loss: 0.0145 step time: 0.2739\n", - "7/8, train_loss: 0.0118 step time: 0.4605\n", - "8/8, train_loss: 0.0188 step time: 0.2085\n", - "epoch 412 average loss: 0.0151\n", - "time consuming of epoch 412 is: 2.3586\n", - "----------\n", - "epoch 413/600\n", - "1/8, train_loss: 0.0150 step time: 0.2650\n", - "2/8, train_loss: 0.0145 step time: 0.2171\n", - "3/8, train_loss: 0.0162 step time: 0.2408\n", - "4/8, train_loss: 0.0153 step time: 0.2795\n", - "5/8, train_loss: 0.0157 step time: 0.3120\n", - "6/8, train_loss: 0.0178 step time: 0.2940\n", - "7/8, train_loss: 0.0133 step time: 0.2836\n", - "8/8, train_loss: 0.0161 step time: 0.2812\n", - "epoch 413 average loss: 0.0155\n", - "time consuming of epoch 413 is: 2.1747\n", - "----------\n", - "epoch 414/600\n", - "1/8, train_loss: 0.0143 step time: 0.3054\n", - "2/8, train_loss: 0.0153 step time: 0.2810\n", - "3/8, train_loss: 0.0158 step time: 0.3170\n", - "4/8, train_loss: 0.0125 step time: 0.3107\n", - "5/8, train_loss: 0.0205 step time: 0.3241\n", - "6/8, train_loss: 0.0150 step time: 0.2761\n", - "7/8, train_loss: 0.0122 step time: 0.2590\n", - "8/8, train_loss: 0.0152 step time: 0.2486\n", - "epoch 414 average loss: 0.0151\n", - "time consuming of epoch 414 is: 2.3235\n", - "----------\n", - "epoch 415/600\n", - "1/8, train_loss: 0.0180 step time: 0.3040\n", - "2/8, train_loss: 0.0165 step time: 0.2497\n", - "3/8, train_loss: 0.0133 step time: 0.2473\n", - "4/8, train_loss: 0.0198 step time: 0.2479\n", - "5/8, train_loss: 0.0140 step time: 0.2520\n", - "6/8, train_loss: 0.0130 step time: 0.4107\n", - "7/8, train_loss: 0.0145 step time: 0.4534\n", - "8/8, train_loss: 0.0182 step time: 0.2678\n", - "epoch 415 average loss: 0.0159\n", - "current epoch: 415 current mean dice: 0.9539 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 415 is: 3.7101\n", - "----------\n", - "epoch 416/600\n", - "1/8, train_loss: 0.0132 step time: 0.3039\n", - "2/8, train_loss: 0.0144 step time: 0.2558\n", - "3/8, train_loss: 0.0138 step time: 0.2475\n", - "4/8, train_loss: 0.0135 step time: 0.2887\n", - "5/8, train_loss: 0.0163 step time: 0.3703\n", - "6/8, train_loss: 0.0135 step time: 0.2854\n", - "7/8, train_loss: 0.0151 step time: 0.2610\n", - "8/8, train_loss: 0.0150 step time: 0.2545\n", - "epoch 416 average loss: 0.0144\n", - "time consuming of epoch 416 is: 2.2683\n", - "----------\n", - "epoch 417/600\n", - "1/8, train_loss: 0.0121 step time: 0.2942\n", - "2/8, train_loss: 0.0177 step time: 0.2966\n", - "3/8, train_loss: 0.0144 step time: 0.4228\n", - "4/8, train_loss: 0.0129 step time: 0.2869\n", - "5/8, train_loss: 0.0159 step time: 0.2677\n", - "6/8, train_loss: 0.0151 step time: 0.2543\n", - "7/8, train_loss: 0.0164 step time: 0.2362\n", - "8/8, train_loss: 0.0217 step time: 0.2395\n", - "epoch 417 average loss: 0.0158\n", - "time consuming of epoch 417 is: 2.2996\n", - "----------\n", - "epoch 418/600\n", - "1/8, train_loss: 0.0164 step time: 0.2955\n", - "2/8, train_loss: 0.0148 step time: 0.2506\n", - "3/8, train_loss: 0.0125 step time: 0.3544\n", - "4/8, train_loss: 0.0158 step time: 0.4843\n", - "5/8, train_loss: 0.0169 step time: 0.2687\n", - "6/8, train_loss: 0.0152 step time: 0.2518\n", - "7/8, train_loss: 0.0142 step time: 0.2494\n", - "8/8, train_loss: 0.0167 step time: 0.2522\n", - "epoch 418 average loss: 0.0153\n", - "time consuming of epoch 418 is: 2.4084\n", - "----------\n", - "epoch 419/600\n", - "1/8, train_loss: 0.0126 step time: 0.2995\n", - "2/8, train_loss: 0.0151 step time: 0.2557\n", - "3/8, train_loss: 0.0140 step time: 0.3048\n", - "4/8, train_loss: 0.0169 step time: 0.3084\n", - "5/8, train_loss: 0.0151 step time: 0.3005\n", - "6/8, train_loss: 0.0169 step time: 0.2846\n", - "7/8, train_loss: 0.0148 step time: 0.2979\n", - "8/8, train_loss: 0.0149 step time: 0.2789\n", - "epoch 419 average loss: 0.0151\n", - "time consuming of epoch 419 is: 2.3318\n", - "----------\n", - "epoch 420/600\n", - "1/8, train_loss: 0.0184 step time: 0.3022\n", - "2/8, train_loss: 0.0158 step time: 0.2566\n", - "3/8, train_loss: 0.0148 step time: 0.2549\n", - "4/8, train_loss: 0.0159 step time: 0.2566\n", - "5/8, train_loss: 0.0150 step time: 0.2777\n", - "6/8, train_loss: 0.0170 step time: 0.2618\n", - "7/8, train_loss: 0.0149 step time: 0.3722\n", - "8/8, train_loss: 0.0151 step time: 0.3634\n", - "epoch 420 average loss: 0.0159\n", - "current epoch: 420 current mean dice: 0.9538 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 420 is: 3.5460\n", - "----------\n", - "epoch 421/600\n", - "1/8, train_loss: 0.0139 step time: 0.4007\n", - "2/8, train_loss: 0.0139 step time: 0.2889\n", - "3/8, train_loss: 0.0164 step time: 0.2788\n", - "4/8, train_loss: 0.0160 step time: 0.2597\n", - "5/8, train_loss: 0.0154 step time: 0.2583\n", - "6/8, train_loss: 0.0151 step time: 0.2503\n", - "7/8, train_loss: 0.0144 step time: 0.2639\n", - "8/8, train_loss: 0.0162 step time: 0.2794\n", - "epoch 421 average loss: 0.0152\n", - "time consuming of epoch 421 is: 2.2812\n", - "----------\n", - "epoch 422/600\n", - "1/8, train_loss: 0.0130 step time: 0.3034\n", - "2/8, train_loss: 0.0243 step time: 0.2396\n", - "3/8, train_loss: 0.0155 step time: 0.2344\n", - "4/8, train_loss: 0.0177 step time: 0.2337\n", - "5/8, train_loss: 0.0148 step time: 0.2311\n", - "6/8, train_loss: 0.0141 step time: 0.2356\n", - "7/8, train_loss: 0.0180 step time: 0.2608\n", - "8/8, train_loss: 0.0161 step time: 0.2528\n", - "epoch 422 average loss: 0.0167\n", - "time consuming of epoch 422 is: 1.9928\n", - "----------\n", - "epoch 423/600\n", - "1/8, train_loss: 0.0166 step time: 0.2942\n", - "2/8, train_loss: 0.0147 step time: 0.2445\n", - "3/8, train_loss: 0.0166 step time: 0.2408\n", - "4/8, train_loss: 0.0165 step time: 0.3769\n", - "5/8, train_loss: 0.0153 step time: 0.6298\n", - "6/8, train_loss: 0.0160 step time: 0.2706\n", - "7/8, train_loss: 0.0131 step time: 0.2460\n", - "8/8, train_loss: 0.0142 step time: 0.2255\n", - "epoch 423 average loss: 0.0154\n", - "time consuming of epoch 423 is: 2.5299\n", - "----------\n", - "epoch 424/600\n", - "1/8, train_loss: 0.0150 step time: 0.2779\n", - "2/8, train_loss: 0.0140 step time: 0.2395\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "3/8, train_loss: 0.0176 step time: 0.2502\n", - "4/8, train_loss: 0.0157 step time: 0.3444\n", - "5/8, train_loss: 0.0136 step time: 0.3390\n", - "6/8, train_loss: 0.0149 step time: 0.2949\n", - "7/8, train_loss: 0.0151 step time: 0.2662\n", - "8/8, train_loss: 0.0133 step time: 0.2698\n", - "epoch 424 average loss: 0.0149\n", - "time consuming of epoch 424 is: 2.2832\n", - "----------\n", - "epoch 425/600\n", - "1/8, train_loss: 0.0156 step time: 0.3079\n", - "2/8, train_loss: 0.0144 step time: 0.2622\n", - "3/8, train_loss: 0.0165 step time: 0.2548\n", - "4/8, train_loss: 0.0147 step time: 0.2527\n", - "5/8, train_loss: 0.0132 step time: 0.2480\n", - "6/8, train_loss: 0.0136 step time: 0.2483\n", - "7/8, train_loss: 0.0184 step time: 0.2449\n", - "8/8, train_loss: 0.0135 step time: 0.2905\n", - "epoch 425 average loss: 0.0150\n", - "current epoch: 425 current mean dice: 0.9543 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 425 is: 3.3486\n", - "----------\n", - "epoch 426/600\n", - "1/8, train_loss: 0.0135 step time: 0.2981\n", - "2/8, train_loss: 0.0142 step time: 0.2455\n", - "3/8, train_loss: 0.0152 step time: 0.2692\n", - "4/8, train_loss: 0.0154 step time: 0.2814\n", - "5/8, train_loss: 0.0204 step time: 0.2820\n", - "6/8, train_loss: 0.0145 step time: 0.2821\n", - "7/8, train_loss: 0.0158 step time: 0.2842\n", - "8/8, train_loss: 0.0144 step time: 0.2805\n", - "epoch 426 average loss: 0.0154\n", - "time consuming of epoch 426 is: 2.2241\n", - "----------\n", - "epoch 427/600\n", - "1/8, train_loss: 0.0109 step time: 0.3085\n", - "2/8, train_loss: 0.0147 step time: 0.2674\n", - "3/8, train_loss: 0.0160 step time: 0.2685\n", - "4/8, train_loss: 0.0136 step time: 0.2656\n", - "5/8, train_loss: 0.0149 step time: 0.2656\n", - "6/8, train_loss: 0.0187 step time: 0.2549\n", - "7/8, train_loss: 0.0130 step time: 0.2887\n", - "8/8, train_loss: 0.0144 step time: 0.2772\n", - "epoch 427 average loss: 0.0145\n", - "time consuming of epoch 427 is: 2.1979\n", - "----------\n", - "epoch 428/600\n", - "1/8, train_loss: 0.0127 step time: 0.3149\n", - "2/8, train_loss: 0.0151 step time: 0.2489\n", - "3/8, train_loss: 0.0135 step time: 0.3219\n", - "4/8, train_loss: 0.0190 step time: 0.3073\n", - "5/8, train_loss: 0.0151 step time: 0.2821\n", - "6/8, train_loss: 0.0161 step time: 0.2717\n", - "7/8, train_loss: 0.0153 step time: 0.2699\n", - "8/8, train_loss: 0.0147 step time: 0.2578\n", - "epoch 428 average loss: 0.0152\n", - "time consuming of epoch 428 is: 2.2760\n", - "----------\n", - "epoch 429/600\n", - "1/8, train_loss: 0.0167 step time: 0.2932\n", - "2/8, train_loss: 0.0160 step time: 0.2385\n", - "3/8, train_loss: 0.0152 step time: 0.2378\n", - "4/8, train_loss: 0.0172 step time: 0.3513\n", - "5/8, train_loss: 0.0149 step time: 0.4932\n", - "6/8, train_loss: 0.0133 step time: 0.2753\n", - "7/8, train_loss: 0.0149 step time: 0.2603\n", - "8/8, train_loss: 0.0119 step time: 0.2633\n", - "epoch 429 average loss: 0.0150\n", - "time consuming of epoch 429 is: 2.4145\n", - "----------\n", - "epoch 430/600\n", - "1/8, train_loss: 0.0167 step time: 0.2973\n", - "2/8, train_loss: 0.0162 step time: 0.2453\n", - "3/8, train_loss: 0.0128 step time: 0.2447\n", - "4/8, train_loss: 0.0137 step time: 0.2363\n", - "5/8, train_loss: 0.0149 step time: 0.2365\n", - "6/8, train_loss: 0.0131 step time: 0.2310\n", - "7/8, train_loss: 0.0137 step time: 0.2197\n", - "8/8, train_loss: 0.0153 step time: 0.2207\n", - "epoch 430 average loss: 0.0146\n", - "current epoch: 430 current mean dice: 0.9530 best mean dice: 0.9550 at epoch: 315\n", - "time consuming of epoch 430 is: 3.2540\n", - "----------\n", - "epoch 431/600\n", - "1/8, train_loss: 0.0147 step time: 0.3151\n", - "2/8, train_loss: 0.0159 step time: 0.2642\n", - "3/8, train_loss: 0.0126 step time: 0.2592\n", - "4/8, train_loss: 0.0167 step time: 0.2524\n", - "5/8, train_loss: 0.0168 step time: 0.2520\n", - "6/8, train_loss: 0.0154 step time: 0.2407\n", - "7/8, train_loss: 0.0144 step time: 0.2356\n", - "8/8, train_loss: 0.0143 step time: 0.3086\n", - "epoch 431 average loss: 0.0151\n", - "time consuming of epoch 431 is: 2.1290\n", - "----------\n", - "epoch 432/600\n", - "1/8, train_loss: 0.0147 step time: 0.3433\n", - "2/8, train_loss: 0.0152 step time: 0.2782\n", - "3/8, train_loss: 0.0181 step time: 0.2757\n", - "4/8, train_loss: 0.0135 step time: 0.2513\n", - "5/8, train_loss: 0.0149 step time: 0.2508\n", - "6/8, train_loss: 0.0141 step time: 0.2487\n", - "7/8, train_loss: 0.0157 step time: 0.2408\n", - "8/8, train_loss: 0.0186 step time: 0.2404\n", - "epoch 432 average loss: 0.0156\n", - "time consuming of epoch 432 is: 2.1308\n", - "----------\n", - "epoch 433/600\n", - "1/8, train_loss: 0.0159 step time: 0.2956\n", - "2/8, train_loss: 0.0174 step time: 0.2459\n", - "3/8, train_loss: 0.0179 step time: 0.2408\n", - "4/8, train_loss: 0.0136 step time: 0.2308\n", - "5/8, train_loss: 0.0127 step time: 0.2376\n", - "6/8, train_loss: 0.0145 step time: 0.2377\n", - "7/8, train_loss: 0.0148 step time: 0.2567\n", - "8/8, train_loss: 0.0122 step time: 0.2996\n", - "epoch 433 average loss: 0.0149\n", - "time consuming of epoch 433 is: 2.0462\n", - "----------\n", - "epoch 434/600\n", - "1/8, train_loss: 0.0120 step time: 0.3040\n", - "2/8, train_loss: 0.0140 step time: 0.2563\n", - "3/8, train_loss: 0.0168 step time: 0.2452\n", - "4/8, train_loss: 0.0123 step time: 0.2444\n", - "5/8, train_loss: 0.0144 step time: 0.2381\n", - "6/8, train_loss: 0.0184 step time: 0.2342\n", - "7/8, train_loss: 0.0124 step time: 0.2645\n", - "8/8, train_loss: 0.0160 step time: 0.3644\n", - "epoch 434 average loss: 0.0145\n", - "time consuming of epoch 434 is: 2.1525\n", - "----------\n", - "epoch 435/600\n", - "1/8, train_loss: 0.0175 step time: 0.3285\n", - "2/8, train_loss: 0.0150 step time: 0.2554\n", - "3/8, train_loss: 0.0154 step time: 0.2569\n", - "4/8, train_loss: 0.0161 step time: 0.2704\n", - "5/8, train_loss: 0.0138 step time: 0.2686\n", - "6/8, train_loss: 0.0123 step time: 0.2566\n", - "7/8, train_loss: 0.0158 step time: 0.2467\n", - "8/8, train_loss: 0.0139 step time: 0.2413\n", - "epoch 435 average loss: 0.0150\n", - "saved new best metric model\n", - "current epoch: 435 current mean dice: 0.9551 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 435 is: 4.0019\n", - "----------\n", - "epoch 436/600\n", - "1/8, train_loss: 0.0128 step time: 0.2755\n", - "2/8, train_loss: 0.0154 step time: 0.2164\n", - "3/8, train_loss: 0.0165 step time: 0.2153\n", - "4/8, train_loss: 0.0135 step time: 0.2155\n", - "5/8, train_loss: 0.0172 step time: 0.2204\n", - "6/8, train_loss: 0.0161 step time: 0.2201\n", - "7/8, train_loss: 0.0115 step time: 0.2429\n", - "8/8, train_loss: 0.0152 step time: 0.2824\n", - "epoch 436 average loss: 0.0148\n", - "time consuming of epoch 436 is: 1.8910\n", - "----------\n", - "epoch 437/600\n", - "1/8, train_loss: 0.0153 step time: 0.3138\n", - "2/8, train_loss: 0.0139 step time: 0.2802\n", - "3/8, train_loss: 0.0163 step time: 0.2838\n", - "4/8, train_loss: 0.0133 step time: 0.2676\n", - "5/8, train_loss: 0.0153 step time: 0.2767\n", - "6/8, train_loss: 0.0128 step time: 0.2657\n", - "7/8, train_loss: 0.0143 step time: 0.2649\n", - "8/8, train_loss: 0.0161 step time: 0.2528\n", - "epoch 437 average loss: 0.0147\n", - "time consuming of epoch 437 is: 2.2069\n", - "----------\n", - "epoch 438/600\n", - "1/8, train_loss: 0.0148 step time: 0.2992\n", - "2/8, train_loss: 0.0140 step time: 0.2510\n", - "3/8, train_loss: 0.0145 step time: 0.2475\n", - "4/8, train_loss: 0.0145 step time: 0.2520\n", - "5/8, train_loss: 0.0174 step time: 0.2459\n", - "6/8, train_loss: 0.0160 step time: 0.2473\n", - "7/8, train_loss: 0.0164 step time: 0.2692\n", - "8/8, train_loss: 0.0157 step time: 0.2750\n", - "epoch 438 average loss: 0.0154\n", - "time consuming of epoch 438 is: 2.0885\n", - "----------\n", - "epoch 439/600\n", - "1/8, train_loss: 0.0155 step time: 0.2974\n", - "2/8, train_loss: 0.0166 step time: 0.2586\n", - "3/8, train_loss: 0.0117 step time: 0.2505\n", - "4/8, train_loss: 0.0130 step time: 0.2452\n", - "5/8, train_loss: 0.0146 step time: 0.2674\n", - "6/8, train_loss: 0.0170 step time: 0.3161\n", - "7/8, train_loss: 0.0134 step time: 0.2947\n", - "8/8, train_loss: 0.0169 step time: 0.2598\n", - "epoch 439 average loss: 0.0148\n", - "time consuming of epoch 439 is: 2.1912\n", - "----------\n", - "epoch 440/600\n", - "1/8, train_loss: 0.0135 step time: 0.3042\n", - "2/8, train_loss: 0.0127 step time: 0.2468\n", - "3/8, train_loss: 0.0182 step time: 0.2411\n", - "4/8, train_loss: 0.0136 step time: 0.2335\n", - "5/8, train_loss: 0.0149 step time: 0.2332\n", - "6/8, train_loss: 0.0169 step time: 0.2288\n", - "7/8, train_loss: 0.0148 step time: 0.3976\n", - "8/8, train_loss: 0.0157 step time: 0.4508\n", - "epoch 440 average loss: 0.0150\n", - "current epoch: 440 current mean dice: 0.9544 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 440 is: 3.5602\n", - "----------\n", - "epoch 441/600\n", - "1/8, train_loss: 0.0156 step time: 0.3199\n", - "2/8, train_loss: 0.0156 step time: 0.2643\n", - "3/8, train_loss: 0.0128 step time: 0.2510\n", - "4/8, train_loss: 0.0156 step time: 0.2456\n", - "5/8, train_loss: 0.0149 step time: 0.2475\n", - "6/8, train_loss: 0.0112 step time: 0.2463\n", - "7/8, train_loss: 0.0148 step time: 0.2667\n", - "8/8, train_loss: 0.0148 step time: 0.2973\n", - "epoch 441 average loss: 0.0144\n", - "time consuming of epoch 441 is: 2.1400\n", - "----------\n", - "epoch 442/600\n", - "1/8, train_loss: 0.0141 step time: 0.3177\n", - "2/8, train_loss: 0.0174 step time: 0.2627\n", - "3/8, train_loss: 0.0146 step time: 0.2617\n", - "4/8, train_loss: 0.0122 step time: 0.2622\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "5/8, train_loss: 0.0135 step time: 0.2667\n", - "6/8, train_loss: 0.0187 step time: 0.2940\n", - "7/8, train_loss: 0.0163 step time: 0.2752\n", - "8/8, train_loss: 0.0147 step time: 0.2571\n", - "epoch 442 average loss: 0.0152\n", - "time consuming of epoch 442 is: 2.1988\n", - "----------\n", - "epoch 443/600\n", - "1/8, train_loss: 0.0129 step time: 0.2875\n", - "2/8, train_loss: 0.0172 step time: 0.2430\n", - "3/8, train_loss: 0.0157 step time: 0.3421\n", - "4/8, train_loss: 0.0150 step time: 0.3522\n", - "5/8, train_loss: 0.0125 step time: 0.2767\n", - "6/8, train_loss: 0.0149 step time: 0.2731\n", - "7/8, train_loss: 0.0144 step time: 0.2533\n", - "8/8, train_loss: 0.0139 step time: 0.2480\n", - "epoch 443 average loss: 0.0146\n", - "time consuming of epoch 443 is: 2.2774\n", - "----------\n", - "epoch 444/600\n", - "1/8, train_loss: 0.0174 step time: 0.2874\n", - "2/8, train_loss: 0.0152 step time: 0.2458\n", - "3/8, train_loss: 0.0173 step time: 0.2509\n", - "4/8, train_loss: 0.0135 step time: 0.2483\n", - "5/8, train_loss: 0.0138 step time: 0.2536\n", - "6/8, train_loss: 0.0136 step time: 0.2517\n", - "7/8, train_loss: 0.0143 step time: 0.2445\n", - "8/8, train_loss: 0.0179 step time: 0.2450\n", - "epoch 444 average loss: 0.0154\n", - "time consuming of epoch 444 is: 2.0286\n", - "----------\n", - "epoch 445/600\n", - "1/8, train_loss: 0.0156 step time: 0.2932\n", - "2/8, train_loss: 0.0155 step time: 0.2613\n", - "3/8, train_loss: 0.0158 step time: 0.2615\n", - "4/8, train_loss: 0.0142 step time: 0.2587\n", - "5/8, train_loss: 0.0164 step time: 0.2558\n", - "6/8, train_loss: 0.0124 step time: 0.2498\n", - "7/8, train_loss: 0.0148 step time: 0.2315\n", - "8/8, train_loss: 0.0141 step time: 0.3062\n", - "epoch 445 average loss: 0.0149\n", - "current epoch: 445 current mean dice: 0.9539 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 445 is: 3.6090\n", - "----------\n", - "epoch 446/600\n", - "1/8, train_loss: 0.0160 step time: 0.3007\n", - "2/8, train_loss: 0.0156 step time: 0.2535\n", - "3/8, train_loss: 0.0141 step time: 0.2464\n", - "4/8, train_loss: 0.0136 step time: 0.2431\n", - "5/8, train_loss: 0.0149 step time: 0.3652\n", - "6/8, train_loss: 0.0138 step time: 0.3962\n", - "7/8, train_loss: 0.0151 step time: 0.2716\n", - "8/8, train_loss: 0.0130 step time: 0.2539\n", - "epoch 446 average loss: 0.0145\n", - "time consuming of epoch 446 is: 2.3322\n", - "----------\n", - "epoch 447/600\n", - "1/8, train_loss: 0.0155 step time: 0.2902\n", - "2/8, train_loss: 0.0149 step time: 0.2396\n", - "3/8, train_loss: 0.0144 step time: 0.2692\n", - "4/8, train_loss: 0.0157 step time: 0.2675\n", - "5/8, train_loss: 0.0134 step time: 0.2539\n", - "6/8, train_loss: 0.0163 step time: 0.2527\n", - "7/8, train_loss: 0.0154 step time: 0.2323\n", - "8/8, train_loss: 0.0128 step time: 0.2568\n", - "epoch 447 average loss: 0.0148\n", - "time consuming of epoch 447 is: 2.0636\n", - "----------\n", - "epoch 448/600\n", - "1/8, train_loss: 0.0169 step time: 0.3244\n", - "2/8, train_loss: 0.0145 step time: 0.2560\n", - "3/8, train_loss: 0.0173 step time: 0.2607\n", - "4/8, train_loss: 0.0136 step time: 0.3511\n", - "5/8, train_loss: 0.0164 step time: 0.3774\n", - "6/8, train_loss: 0.0142 step time: 0.2722\n", - "7/8, train_loss: 0.0143 step time: 0.2581\n", - "8/8, train_loss: 0.0148 step time: 0.2500\n", - "epoch 448 average loss: 0.0152\n", - "time consuming of epoch 448 is: 2.3513\n", - "----------\n", - "epoch 449/600\n", - "1/8, train_loss: 0.0132 step time: 0.2998\n", - "2/8, train_loss: 0.0159 step time: 0.2523\n", - "3/8, train_loss: 0.0144 step time: 0.2458\n", - "4/8, train_loss: 0.0173 step time: 0.2503\n", - "5/8, train_loss: 0.0146 step time: 0.2416\n", - "6/8, train_loss: 0.0156 step time: 0.2418\n", - "7/8, train_loss: 0.0142 step time: 0.2915\n", - "8/8, train_loss: 0.0155 step time: 0.3002\n", - "epoch 449 average loss: 0.0151\n", - "time consuming of epoch 449 is: 2.1249\n", - "----------\n", - "epoch 450/600\n", - "1/8, train_loss: 0.0168 step time: 0.3177\n", - "2/8, train_loss: 0.0133 step time: 0.2538\n", - "3/8, train_loss: 0.0136 step time: 0.2469\n", - "4/8, train_loss: 0.0160 step time: 0.2632\n", - "5/8, train_loss: 0.0156 step time: 0.2656\n", - "6/8, train_loss: 0.0158 step time: 0.4643\n", - "7/8, train_loss: 0.0148 step time: 0.4271\n", - "8/8, train_loss: 0.0142 step time: 0.2558\n", - "epoch 450 average loss: 0.0150\n", - "current epoch: 450 current mean dice: 0.9550 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 450 is: 3.9072\n", - "----------\n", - "epoch 451/600\n", - "1/8, train_loss: 0.0138 step time: 0.2930\n", - "2/8, train_loss: 0.0134 step time: 0.2369\n", - "3/8, train_loss: 0.0140 step time: 0.2385\n", - "4/8, train_loss: 0.0144 step time: 0.2341\n", - "5/8, train_loss: 0.0161 step time: 0.2397\n", - "6/8, train_loss: 0.0155 step time: 0.2360\n", - "7/8, train_loss: 0.0128 step time: 0.2232\n", - "8/8, train_loss: 0.0157 step time: 0.2177\n", - "epoch 451 average loss: 0.0145\n", - "time consuming of epoch 451 is: 1.9204\n", - "----------\n", - "epoch 452/600\n", - "1/8, train_loss: 0.0185 step time: 0.2838\n", - "2/8, train_loss: 0.0132 step time: 0.2305\n", - "3/8, train_loss: 0.0178 step time: 0.2306\n", - "4/8, train_loss: 0.0114 step time: 0.2289\n", - "5/8, train_loss: 0.0157 step time: 0.2292\n", - "6/8, train_loss: 0.0139 step time: 0.2275\n", - "7/8, train_loss: 0.0147 step time: 0.2183\n", - "8/8, train_loss: 0.0160 step time: 0.2197\n", - "epoch 452 average loss: 0.0152\n", - "time consuming of epoch 452 is: 1.8699\n", - "----------\n", - "epoch 453/600\n", - "1/8, train_loss: 0.0220 step time: 0.2805\n", - "2/8, train_loss: 0.0121 step time: 0.2313\n", - "3/8, train_loss: 0.0132 step time: 0.2277\n", - "4/8, train_loss: 0.0163 step time: 0.2330\n", - "5/8, train_loss: 0.0140 step time: 0.2762\n", - "6/8, train_loss: 0.0162 step time: 0.2841\n", - "7/8, train_loss: 0.0132 step time: 0.2785\n", - "8/8, train_loss: 0.0149 step time: 0.2854\n", - "epoch 453 average loss: 0.0152\n", - "time consuming of epoch 453 is: 2.0980\n", - "----------\n", - "epoch 454/600\n", - "1/8, train_loss: 0.0159 step time: 0.3185\n", - "2/8, train_loss: 0.0116 step time: 0.2604\n", - "3/8, train_loss: 0.0153 step time: 0.2616\n", - "4/8, train_loss: 0.0198 step time: 0.2611\n", - "5/8, train_loss: 0.0135 step time: 0.2537\n", - "6/8, train_loss: 0.0182 step time: 0.2511\n", - "7/8, train_loss: 0.0131 step time: 0.3452\n", - "8/8, train_loss: 0.0128 step time: 0.3181\n", - "epoch 454 average loss: 0.0150\n", - "time consuming of epoch 454 is: 2.2712\n", - "----------\n", - "epoch 455/600\n", - "1/8, train_loss: 0.0145 step time: 0.3069\n", - "2/8, train_loss: 0.0173 step time: 0.2569\n", - "3/8, train_loss: 0.0124 step time: 0.2507\n", - "4/8, train_loss: 0.0152 step time: 0.2495\n", - "5/8, train_loss: 0.0145 step time: 0.2803\n", - "6/8, train_loss: 0.0149 step time: 0.2717\n", - "7/8, train_loss: 0.0170 step time: 0.2583\n", - "8/8, train_loss: 0.0133 step time: 0.2614\n", - "epoch 455 average loss: 0.0149\n", - "current epoch: 455 current mean dice: 0.9541 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 455 is: 3.4658\n", - "----------\n", - "epoch 456/600\n", - "1/8, train_loss: 0.0156 step time: 0.2978\n", - "2/8, train_loss: 0.0141 step time: 0.2519\n", - "3/8, train_loss: 0.0168 step time: 0.2469\n", - "4/8, train_loss: 0.0153 step time: 0.2395\n", - "5/8, train_loss: 0.0148 step time: 0.2400\n", - "6/8, train_loss: 0.0157 step time: 0.2455\n", - "7/8, train_loss: 0.0157 step time: 0.2372\n", - "8/8, train_loss: 0.0146 step time: 0.2534\n", - "epoch 456 average loss: 0.0153\n", - "time consuming of epoch 456 is: 2.0133\n", - "----------\n", - "epoch 457/600\n", - "1/8, train_loss: 0.0164 step time: 0.3359\n", - "2/8, train_loss: 0.0154 step time: 0.2799\n", - "3/8, train_loss: 0.0153 step time: 0.2511\n", - "4/8, train_loss: 0.0131 step time: 0.2475\n", - "5/8, train_loss: 0.0133 step time: 0.2502\n", - "6/8, train_loss: 0.0147 step time: 0.2409\n", - "7/8, train_loss: 0.0101 step time: 0.2336\n", - "8/8, train_loss: 0.0181 step time: 0.2324\n", - "epoch 457 average loss: 0.0146\n", - "time consuming of epoch 457 is: 2.0730\n", - "----------\n", - "epoch 458/600\n", - "1/8, train_loss: 0.0142 step time: 0.2889\n", - "2/8, train_loss: 0.0136 step time: 0.2387\n", - "3/8, train_loss: 0.0171 step time: 0.2382\n", - "4/8, train_loss: 0.0164 step time: 0.2359\n", - "5/8, train_loss: 0.0132 step time: 0.2333\n", - "6/8, train_loss: 0.0115 step time: 0.3759\n", - "7/8, train_loss: 0.0150 step time: 0.4552\n", - "8/8, train_loss: 0.0172 step time: 0.2565\n", - "epoch 458 average loss: 0.0148\n", - "time consuming of epoch 458 is: 2.3238\n", - "----------\n", - "epoch 459/600\n", - "1/8, train_loss: 0.0161 step time: 0.3029\n", - "2/8, train_loss: 0.0162 step time: 0.2362\n", - "3/8, train_loss: 0.0144 step time: 0.2384\n", - "4/8, train_loss: 0.0136 step time: 0.2400\n", - "5/8, train_loss: 0.0126 step time: 0.2406\n", - "6/8, train_loss: 0.0134 step time: 0.2319\n", - "7/8, train_loss: 0.0187 step time: 0.2360\n", - "8/8, train_loss: 0.0154 step time: 0.2644\n", - "epoch 459 average loss: 0.0151\n", - "time consuming of epoch 459 is: 1.9919\n", - "----------\n", - "epoch 460/600\n", - "1/8, train_loss: 0.0145 step time: 0.3175\n", - "2/8, train_loss: 0.0147 step time: 0.2627\n", - "3/8, train_loss: 0.0152 step time: 0.2569\n", - "4/8, train_loss: 0.0135 step time: 0.2575\n", - "5/8, train_loss: 0.0142 step time: 0.2485\n", - "6/8, train_loss: 0.0147 step time: 0.2396\n", - "7/8, train_loss: 0.0160 step time: 0.2328\n", - "8/8, train_loss: 0.0155 step time: 0.2266\n", - "epoch 460 average loss: 0.0148\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "current epoch: 460 current mean dice: 0.9547 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 460 is: 3.2773\n", - "----------\n", - "epoch 461/600\n", - "1/8, train_loss: 0.0155 step time: 0.3138\n", - "2/8, train_loss: 0.0145 step time: 0.2652\n", - "3/8, train_loss: 0.0187 step time: 0.2638\n", - "4/8, train_loss: 0.0169 step time: 0.2636\n", - "5/8, train_loss: 0.0117 step time: 0.2605\n", - "6/8, train_loss: 0.0140 step time: 0.2580\n", - "7/8, train_loss: 0.0134 step time: 0.2541\n", - "8/8, train_loss: 0.0143 step time: 0.2554\n", - "epoch 461 average loss: 0.0149\n", - "time consuming of epoch 461 is: 2.1355\n", - "----------\n", - "epoch 462/600\n", - "1/8, train_loss: 0.0144 step time: 0.3150\n", - "2/8, train_loss: 0.0146 step time: 0.2540\n", - "3/8, train_loss: 0.0113 step time: 0.2555\n", - "4/8, train_loss: 0.0152 step time: 0.2430\n", - "5/8, train_loss: 0.0136 step time: 0.2400\n", - "6/8, train_loss: 0.0152 step time: 0.2412\n", - "7/8, train_loss: 0.0165 step time: 0.2667\n", - "8/8, train_loss: 0.0149 step time: 0.3267\n", - "epoch 462 average loss: 0.0145\n", - "time consuming of epoch 462 is: 2.1435\n", - "----------\n", - "epoch 463/600\n", - "1/8, train_loss: 0.0165 step time: 0.3185\n", - "2/8, train_loss: 0.0137 step time: 0.2549\n", - "3/8, train_loss: 0.0159 step time: 0.2576\n", - "4/8, train_loss: 0.0140 step time: 0.2502\n", - "5/8, train_loss: 0.0155 step time: 0.2506\n", - "6/8, train_loss: 0.0159 step time: 0.2482\n", - "7/8, train_loss: 0.0136 step time: 0.2421\n", - "8/8, train_loss: 0.0179 step time: 0.2317\n", - "epoch 463 average loss: 0.0154\n", - "time consuming of epoch 463 is: 2.0553\n", - "----------\n", - "epoch 464/600\n", - "1/8, train_loss: 0.0175 step time: 0.2874\n", - "2/8, train_loss: 0.0153 step time: 0.2330\n", - "3/8, train_loss: 0.0157 step time: 0.2649\n", - "4/8, train_loss: 0.0144 step time: 0.2802\n", - "5/8, train_loss: 0.0145 step time: 0.2749\n", - "6/8, train_loss: 0.0139 step time: 0.2755\n", - "7/8, train_loss: 0.0136 step time: 0.3063\n", - "8/8, train_loss: 0.0159 step time: 0.2904\n", - "epoch 464 average loss: 0.0151\n", - "time consuming of epoch 464 is: 2.2140\n", - "----------\n", - "epoch 465/600\n", - "1/8, train_loss: 0.0143 step time: 0.3027\n", - "2/8, train_loss: 0.0157 step time: 0.2467\n", - "3/8, train_loss: 0.0138 step time: 0.2521\n", - "4/8, train_loss: 0.0137 step time: 0.2473\n", - "5/8, train_loss: 0.0145 step time: 0.2420\n", - "6/8, train_loss: 0.0174 step time: 0.2459\n", - "7/8, train_loss: 0.0162 step time: 0.2326\n", - "8/8, train_loss: 0.0121 step time: 0.2325\n", - "epoch 465 average loss: 0.0147\n", - "current epoch: 465 current mean dice: 0.9540 best mean dice: 0.9551 at epoch: 435\n", - "time consuming of epoch 465 is: 3.3323\n", - "----------\n", - "epoch 466/600\n", - "1/8, train_loss: 0.0150 step time: 0.3229\n", - "2/8, train_loss: 0.0142 step time: 0.2761\n", - "3/8, train_loss: 0.0145 step time: 0.2561\n", - "4/8, train_loss: 0.0148 step time: 0.2588\n", - "5/8, train_loss: 0.0150 step time: 0.4052\n", - "6/8, train_loss: 0.0128 step time: 0.4585\n", - "7/8, train_loss: 0.0128 step time: 0.2029\n", - "8/8, train_loss: 0.0149 step time: 0.1997\n", - "epoch 466 average loss: 0.0142\n", - "time consuming of epoch 466 is: 2.3815\n", - "----------\n", - "epoch 467/600\n", - "1/8, train_loss: 0.0144 step time: 0.2667\n", - "2/8, train_loss: 0.0141 step time: 0.2169\n", - "3/8, train_loss: 0.0141 step time: 0.2123\n", - "4/8, train_loss: 0.0163 step time: 0.2608\n", - "5/8, train_loss: 0.0119 step time: 0.3160\n", - "6/8, train_loss: 0.0148 step time: 0.3083\n", - "7/8, train_loss: 0.0156 step time: 0.2919\n", - "8/8, train_loss: 0.0162 step time: 0.2674\n", - "epoch 467 average loss: 0.0147\n", - "time consuming of epoch 467 is: 2.1417\n", - "----------\n", - "epoch 468/600\n", - "1/8, train_loss: 0.0110 step time: 0.2963\n", - "2/8, train_loss: 0.0139 step time: 0.2463\n", - "3/8, train_loss: 0.0144 step time: 0.2426\n", - "4/8, train_loss: 0.0155 step time: 0.2400\n", - "5/8, train_loss: 0.0142 step time: 0.2429\n", - "6/8, train_loss: 0.0156 step time: 0.2340\n", - "7/8, train_loss: 0.0179 step time: 0.2190\n", - "8/8, train_loss: 0.0128 step time: 0.2212\n", - "epoch 468 average loss: 0.0144\n", - "time consuming of epoch 468 is: 1.9437\n", - "----------\n", - "epoch 469/600\n", - "1/8, train_loss: 0.0164 step time: 0.2822\n", - "2/8, train_loss: 0.0145 step time: 0.2337\n", - "3/8, train_loss: 0.0184 step time: 0.2304\n", - "4/8, train_loss: 0.0143 step time: 0.2280\n", - "5/8, train_loss: 0.0149 step time: 0.2290\n", - "6/8, train_loss: 0.0137 step time: 0.2348\n", - "7/8, train_loss: 0.0129 step time: 0.2291\n", - "8/8, train_loss: 0.0126 step time: 0.2767\n", - "epoch 469 average loss: 0.0147\n", - "time consuming of epoch 469 is: 1.9454\n", - "----------\n", - "epoch 470/600\n", - "1/8, train_loss: 0.0141 step time: 0.3345\n", - "2/8, train_loss: 0.0139 step time: 0.2614\n", - "3/8, train_loss: 0.0166 step time: 0.2536\n", - "4/8, train_loss: 0.0158 step time: 0.2520\n", - "5/8, train_loss: 0.0141 step time: 0.2932\n", - "6/8, train_loss: 0.0155 step time: 0.3049\n", - "7/8, train_loss: 0.0108 step time: 0.2592\n", - "8/8, train_loss: 0.0138 step time: 0.2553\n", - "epoch 470 average loss: 0.0143\n", - "saved new best metric model\n", - "current epoch: 470 current mean dice: 0.9554 best mean dice: 0.9554 at epoch: 470\n", - "time consuming of epoch 470 is: 3.8738\n", - "----------\n", - "epoch 471/600\n", - "1/8, train_loss: 0.0156 step time: 0.2769\n", - "2/8, train_loss: 0.0140 step time: 0.2128\n", - "3/8, train_loss: 0.0215 step time: 0.2125\n", - "4/8, train_loss: 0.0133 step time: 0.2178\n", - "5/8, train_loss: 0.0154 step time: 0.2344\n", - "6/8, train_loss: 0.0155 step time: 0.2448\n", - "7/8, train_loss: 0.0148 step time: 0.2474\n", - "8/8, train_loss: 0.0133 step time: 0.2392\n", - "epoch 471 average loss: 0.0154\n", - "time consuming of epoch 471 is: 1.8882\n", - "----------\n", - "epoch 472/600\n", - "1/8, train_loss: 0.0155 step time: 0.2938\n", - "2/8, train_loss: 0.0183 step time: 0.2460\n", - "3/8, train_loss: 0.0109 step time: 0.2394\n", - "4/8, train_loss: 0.0164 step time: 0.2422\n", - "5/8, train_loss: 0.0161 step time: 0.2393\n", - "6/8, train_loss: 0.0162 step time: 0.2398\n", - "7/8, train_loss: 0.0170 step time: 0.2231\n", - "8/8, train_loss: 0.0164 step time: 0.2213\n", - "epoch 472 average loss: 0.0158\n", - "time consuming of epoch 472 is: 1.9466\n", - "----------\n", - "epoch 473/600\n", - "1/8, train_loss: 0.0136 step time: 0.2790\n", - "2/8, train_loss: 0.0160 step time: 0.2329\n", - "3/8, train_loss: 0.0183 step time: 0.2446\n", - "4/8, train_loss: 0.0143 step time: 0.2610\n", - "5/8, train_loss: 0.0155 step time: 0.3160\n", - "6/8, train_loss: 0.0143 step time: 0.3516\n", - "7/8, train_loss: 0.0112 step time: 0.2948\n", - "8/8, train_loss: 0.0187 step time: 0.2710\n", - "epoch 473 average loss: 0.0152\n", - "time consuming of epoch 473 is: 2.2521\n", - "----------\n", - "epoch 474/600\n", - "1/8, train_loss: 0.0133 step time: 0.3126\n", - "2/8, train_loss: 0.0144 step time: 0.3610\n", - "3/8, train_loss: 0.0172 step time: 0.3268\n", - "4/8, train_loss: 0.0153 step time: 0.2781\n", - "5/8, train_loss: 0.0149 step time: 0.2557\n", - "6/8, train_loss: 0.0165 step time: 0.2544\n", - "7/8, train_loss: 0.0112 step time: 0.2359\n", - "8/8, train_loss: 0.0182 step time: 0.2390\n", - "epoch 474 average loss: 0.0151\n", - "time consuming of epoch 474 is: 2.2651\n", - "----------\n", - "epoch 475/600\n", - "1/8, train_loss: 0.0144 step time: 0.3054\n", - "2/8, train_loss: 0.0160 step time: 0.2510\n", - "3/8, train_loss: 0.0160 step time: 0.2531\n", - "4/8, train_loss: 0.0133 step time: 0.2580\n", - "5/8, train_loss: 0.0165 step time: 0.2775\n", - "6/8, train_loss: 0.0136 step time: 0.2529\n", - "7/8, train_loss: 0.0162 step time: 0.2443\n", - "8/8, train_loss: 0.0149 step time: 0.2431\n", - "epoch 475 average loss: 0.0151\n", - "current epoch: 475 current mean dice: 0.9552 best mean dice: 0.9554 at epoch: 470\n", - "time consuming of epoch 475 is: 3.2465\n", - "----------\n", - "epoch 476/600\n", - "1/8, train_loss: 0.0147 step time: 0.3742\n", - "2/8, train_loss: 0.0196 step time: 0.2866\n", - "3/8, train_loss: 0.0133 step time: 0.2521\n", - "4/8, train_loss: 0.0140 step time: 0.2535\n", - "5/8, train_loss: 0.0149 step time: 0.2420\n", - "6/8, train_loss: 0.0147 step time: 0.2487\n", - "7/8, train_loss: 0.0185 step time: 0.2804\n", - "8/8, train_loss: 0.0132 step time: 0.2710\n", - "epoch 476 average loss: 0.0154\n", - "time consuming of epoch 476 is: 2.2096\n", - "----------\n", - "epoch 477/600\n", - "1/8, train_loss: 0.0204 step time: 0.3028\n", - "2/8, train_loss: 0.0142 step time: 0.2947\n", - "3/8, train_loss: 0.0129 step time: 0.3529\n", - "4/8, train_loss: 0.0152 step time: 0.2810\n", - "5/8, train_loss: 0.0164 step time: 0.2595\n", - "6/8, train_loss: 0.0141 step time: 0.2475\n", - "7/8, train_loss: 0.0135 step time: 0.2326\n", - "8/8, train_loss: 0.0161 step time: 0.2363\n", - "epoch 477 average loss: 0.0154\n", - "time consuming of epoch 477 is: 2.2088\n", - "----------\n", - "epoch 478/600\n", - "1/8, train_loss: 0.0150 step time: 0.2893\n", - "2/8, train_loss: 0.0141 step time: 0.2333\n", - "3/8, train_loss: 0.0189 step time: 0.2824\n", - "4/8, train_loss: 0.0146 step time: 0.3096\n", - "5/8, train_loss: 0.0166 step time: 0.2638\n", - "6/8, train_loss: 0.0136 step time: 0.2552\n", - "7/8, train_loss: 0.0135 step time: 0.3019\n", - "8/8, train_loss: 0.0152 step time: 0.3165\n", - "epoch 478 average loss: 0.0152\n", - "time consuming of epoch 478 is: 2.2534\n", - "----------\n", - "epoch 479/600\n", - "1/8, train_loss: 0.0141 step time: 0.3209\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "2/8, train_loss: 0.0139 step time: 0.2661\n", - "3/8, train_loss: 0.0142 step time: 0.2618\n", - "4/8, train_loss: 0.0177 step time: 0.2494\n", - "5/8, train_loss: 0.0144 step time: 0.2530\n", - "6/8, train_loss: 0.0143 step time: 0.3314\n", - "7/8, train_loss: 0.0150 step time: 0.2974\n", - "8/8, train_loss: 0.0146 step time: 0.2646\n", - "epoch 479 average loss: 0.0148\n", - "time consuming of epoch 479 is: 2.2460\n", - "----------\n", - "epoch 480/600\n", - "1/8, train_loss: 0.0157 step time: 0.2989\n", - "2/8, train_loss: 0.0144 step time: 0.2578\n", - "3/8, train_loss: 0.0134 step time: 0.2620\n", - "4/8, train_loss: 0.0149 step time: 0.2540\n", - "5/8, train_loss: 0.0133 step time: 0.2488\n", - "6/8, train_loss: 0.0150 step time: 0.2362\n", - "7/8, train_loss: 0.0144 step time: 0.2291\n", - "8/8, train_loss: 0.0184 step time: 0.2233\n", - "epoch 480 average loss: 0.0149\n", - "current epoch: 480 current mean dice: 0.9554 best mean dice: 0.9554 at epoch: 470\n", - "time consuming of epoch 480 is: 3.3287\n", - "----------\n", - "epoch 481/600\n", - "1/8, train_loss: 0.0141 step time: 0.3229\n", - "2/8, train_loss: 0.0150 step time: 0.2605\n", - "3/8, train_loss: 0.0170 step time: 0.2603\n", - "4/8, train_loss: 0.0146 step time: 0.2558\n", - "5/8, train_loss: 0.0134 step time: 0.2550\n", - "6/8, train_loss: 0.0136 step time: 0.2429\n", - "7/8, train_loss: 0.0145 step time: 0.2360\n", - "8/8, train_loss: 0.0149 step time: 0.2314\n", - "epoch 481 average loss: 0.0146\n", - "time consuming of epoch 481 is: 2.0661\n", - "----------\n", - "epoch 482/600\n", - "1/8, train_loss: 0.0166 step time: 0.2823\n", - "2/8, train_loss: 0.0171 step time: 0.2336\n", - "3/8, train_loss: 0.0153 step time: 0.2496\n", - "4/8, train_loss: 0.0132 step time: 0.2803\n", - "5/8, train_loss: 0.0144 step time: 0.2598\n", - "6/8, train_loss: 0.0136 step time: 0.2554\n", - "7/8, train_loss: 0.0143 step time: 0.2433\n", - "8/8, train_loss: 0.0137 step time: 0.2441\n", - "epoch 482 average loss: 0.0148\n", - "time consuming of epoch 482 is: 2.0500\n", - "----------\n", - "epoch 483/600\n", - "1/8, train_loss: 0.0133 step time: 0.2910\n", - "2/8, train_loss: 0.0160 step time: 0.2315\n", - "3/8, train_loss: 0.0136 step time: 0.2338\n", - "4/8, train_loss: 0.0178 step time: 0.2300\n", - "5/8, train_loss: 0.0131 step time: 0.2344\n", - "6/8, train_loss: 0.0148 step time: 0.2686\n", - "7/8, train_loss: 0.0134 step time: 0.2845\n", - "8/8, train_loss: 0.0143 step time: 0.2772\n", - "epoch 483 average loss: 0.0145\n", - "time consuming of epoch 483 is: 2.0524\n", - "----------\n", - "epoch 484/600\n", - "1/8, train_loss: 0.0154 step time: 0.3146\n", - "2/8, train_loss: 0.0158 step time: 0.2606\n", - "3/8, train_loss: 0.0171 step time: 0.2619\n", - "4/8, train_loss: 0.0133 step time: 0.2604\n", - "5/8, train_loss: 0.0166 step time: 0.2625\n", - "6/8, train_loss: 0.0139 step time: 0.2575\n", - "7/8, train_loss: 0.0139 step time: 0.2417\n", - "8/8, train_loss: 0.0143 step time: 0.2429\n", - "epoch 484 average loss: 0.0150\n", - "time consuming of epoch 484 is: 2.1036\n", - "----------\n", - "epoch 485/600\n", - "1/8, train_loss: 0.0133 step time: 0.2964\n", - "2/8, train_loss: 0.0141 step time: 0.2497\n", - "3/8, train_loss: 0.0145 step time: 0.2524\n", - "4/8, train_loss: 0.0153 step time: 0.2486\n", - "5/8, train_loss: 0.0159 step time: 0.2936\n", - "6/8, train_loss: 0.0156 step time: 0.3017\n", - "7/8, train_loss: 0.0129 step time: 0.2895\n", - "8/8, train_loss: 0.0143 step time: 0.2630\n", - "epoch 485 average loss: 0.0145\n", - "current epoch: 485 current mean dice: 0.9552 best mean dice: 0.9554 at epoch: 470\n", - "time consuming of epoch 485 is: 3.3863\n", - "----------\n", - "epoch 486/600\n", - "1/8, train_loss: 0.0112 step time: 0.2874\n", - "2/8, train_loss: 0.0152 step time: 0.2324\n", - "3/8, train_loss: 0.0185 step time: 0.2861\n", - "4/8, train_loss: 0.0140 step time: 0.3710\n", - "5/8, train_loss: 0.0131 step time: 0.2781\n", - "6/8, train_loss: 0.0123 step time: 0.2649\n", - "7/8, train_loss: 0.0169 step time: 0.2526\n", - "8/8, train_loss: 0.0161 step time: 0.2593\n", - "epoch 486 average loss: 0.0147\n", - "time consuming of epoch 486 is: 2.2330\n", - "----------\n", - "epoch 487/600\n", - "1/8, train_loss: 0.0132 step time: 0.3407\n", - "2/8, train_loss: 0.0131 step time: 0.2681\n", - "3/8, train_loss: 0.0151 step time: 0.2612\n", - "4/8, train_loss: 0.0144 step time: 0.2514\n", - "5/8, train_loss: 0.0139 step time: 0.2487\n", - "6/8, train_loss: 0.0144 step time: 0.2951\n", - "7/8, train_loss: 0.0148 step time: 0.3359\n", - "8/8, train_loss: 0.0123 step time: 0.2927\n", - "epoch 487 average loss: 0.0139\n", - "time consuming of epoch 487 is: 2.2952\n", - "----------\n", - "epoch 488/600\n", - "1/8, train_loss: 0.0154 step time: 0.3051\n", - "2/8, train_loss: 0.0140 step time: 0.2491\n", - "3/8, train_loss: 0.0119 step time: 0.2532\n", - "4/8, train_loss: 0.0114 step time: 0.2449\n", - "5/8, train_loss: 0.0137 step time: 0.2664\n", - "6/8, train_loss: 0.0147 step time: 0.2636\n", - "7/8, train_loss: 0.0192 step time: 0.2845\n", - "8/8, train_loss: 0.0153 step time: 0.2975\n", - "epoch 488 average loss: 0.0145\n", - "time consuming of epoch 488 is: 2.1659\n", - "----------\n", - "epoch 489/600\n", - "1/8, train_loss: 0.0128 step time: 0.3175\n", - "2/8, train_loss: 0.0162 step time: 0.2649\n", - "3/8, train_loss: 0.0149 step time: 0.2530\n", - "4/8, train_loss: 0.0152 step time: 0.2496\n", - "5/8, train_loss: 0.0182 step time: 0.2377\n", - "6/8, train_loss: 0.0150 step time: 0.2313\n", - "7/8, train_loss: 0.0142 step time: 0.2258\n", - "8/8, train_loss: 0.0139 step time: 0.2285\n", - "epoch 489 average loss: 0.0150\n", - "time consuming of epoch 489 is: 2.0099\n", - "----------\n", - "epoch 490/600\n", - "1/8, train_loss: 0.0161 step time: 0.2843\n", - "2/8, train_loss: 0.0134 step time: 0.2335\n", - "3/8, train_loss: 0.0166 step time: 0.2409\n", - "4/8, train_loss: 0.0156 step time: 0.3146\n", - "5/8, train_loss: 0.0153 step time: 0.2950\n", - "6/8, train_loss: 0.0151 step time: 0.2785\n", - "7/8, train_loss: 0.0142 step time: 0.4672\n", - "8/8, train_loss: 0.0126 step time: 0.6426\n", - "epoch 490 average loss: 0.0149\n", - "current epoch: 490 current mean dice: 0.9550 best mean dice: 0.9554 at epoch: 470\n", - "time consuming of epoch 490 is: 4.0932\n", - "----------\n", - "epoch 491/600\n", - "1/8, train_loss: 0.0113 step time: 0.3642\n", - "2/8, train_loss: 0.0137 step time: 0.2706\n", - "3/8, train_loss: 0.0140 step time: 0.2574\n", - "4/8, train_loss: 0.0156 step time: 0.2587\n", - "5/8, train_loss: 0.0149 step time: 0.2486\n", - "6/8, train_loss: 0.0192 step time: 0.2684\n", - "7/8, train_loss: 0.0176 step time: 0.2597\n", - "8/8, train_loss: 0.0153 step time: 0.2509\n", - "epoch 491 average loss: 0.0152\n", - "time consuming of epoch 491 is: 2.1798\n", - "----------\n", - "epoch 492/600\n", - "1/8, train_loss: 0.0141 step time: 0.2929\n", - "2/8, train_loss: 0.0143 step time: 0.2390\n", - "3/8, train_loss: 0.0141 step time: 0.2363\n", - "4/8, train_loss: 0.0138 step time: 0.2392\n", - "5/8, train_loss: 0.0132 step time: 0.3037\n", - "6/8, train_loss: 0.0164 step time: 0.3584\n", - "7/8, train_loss: 0.0184 step time: 0.2681\n", - "8/8, train_loss: 0.0140 step time: 0.2407\n", - "epoch 492 average loss: 0.0148\n", - "time consuming of epoch 492 is: 2.1797\n", - "----------\n", - "epoch 493/600\n", - "1/8, train_loss: 0.0148 step time: 0.2872\n", - "2/8, train_loss: 0.0150 step time: 0.2293\n", - "3/8, train_loss: 0.0154 step time: 0.2318\n", - "4/8, train_loss: 0.0136 step time: 0.2306\n", - "5/8, train_loss: 0.0151 step time: 0.2340\n", - "6/8, train_loss: 0.0165 step time: 0.2295\n", - "7/8, train_loss: 0.0143 step time: 0.2187\n", - "8/8, train_loss: 0.0168 step time: 0.2166\n", - "epoch 493 average loss: 0.0152\n", - "time consuming of epoch 493 is: 1.8794\n", - "----------\n", - "epoch 494/600\n", - "1/8, train_loss: 0.0158 step time: 0.3014\n", - "2/8, train_loss: 0.0155 step time: 0.2646\n", - "3/8, train_loss: 0.0136 step time: 0.2612\n", - "4/8, train_loss: 0.0157 step time: 0.2506\n", - "5/8, train_loss: 0.0143 step time: 0.2482\n", - "6/8, train_loss: 0.0143 step time: 0.2508\n", - "7/8, train_loss: 0.0161 step time: 0.2457\n", - "8/8, train_loss: 0.0141 step time: 0.3862\n", - "epoch 494 average loss: 0.0149\n", - "time consuming of epoch 494 is: 2.2103\n", - "----------\n", - "epoch 495/600\n", - "1/8, train_loss: 0.0126 step time: 0.3945\n", - "2/8, train_loss: 0.0140 step time: 0.2974\n", - "3/8, train_loss: 0.0155 step time: 0.2699\n", - "4/8, train_loss: 0.0136 step time: 0.2544\n", - "5/8, train_loss: 0.0191 step time: 0.2495\n", - "6/8, train_loss: 0.0148 step time: 0.2482\n", - "7/8, train_loss: 0.0175 step time: 0.3006\n", - "8/8, train_loss: 0.0132 step time: 0.3796\n", - "epoch 495 average loss: 0.0150\n", - "saved new best metric model\n", - "current epoch: 495 current mean dice: 0.9558 best mean dice: 0.9558 at epoch: 495\n", - "time consuming of epoch 495 is: 3.9349\n", - "----------\n", - "epoch 496/600\n", - "1/8, train_loss: 0.0128 step time: 0.2782\n", - "2/8, train_loss: 0.0174 step time: 0.2116\n", - "3/8, train_loss: 0.0162 step time: 0.2069\n", - "4/8, train_loss: 0.0131 step time: 0.2154\n", - "5/8, train_loss: 0.0160 step time: 0.2506\n", - "6/8, train_loss: 0.0127 step time: 0.2818\n", - "7/8, train_loss: 0.0156 step time: 0.2524\n", - "8/8, train_loss: 0.0159 step time: 0.2488\n", - "epoch 496 average loss: 0.0150\n", - "time consuming of epoch 496 is: 1.9483\n", - "----------\n", - "epoch 497/600\n", - "1/8, train_loss: 0.0153 step time: 0.2958\n", - "2/8, train_loss: 0.0128 step time: 0.2593\n", - "3/8, train_loss: 0.0159 step time: 0.2472\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "4/8, train_loss: 0.0114 step time: 0.2497\n", - "5/8, train_loss: 0.0156 step time: 0.2485\n", - "6/8, train_loss: 0.0144 step time: 0.2585\n", - "7/8, train_loss: 0.0140 step time: 0.2685\n", - "8/8, train_loss: 0.0156 step time: 0.2734\n", - "epoch 497 average loss: 0.0144\n", - "time consuming of epoch 497 is: 2.1022\n", - "----------\n", - "epoch 498/600\n", - "1/8, train_loss: 0.0150 step time: 0.3006\n", - "2/8, train_loss: 0.0155 step time: 0.2549\n", - "3/8, train_loss: 0.0117 step time: 0.2456\n", - "4/8, train_loss: 0.0151 step time: 0.2401\n", - "5/8, train_loss: 0.0133 step time: 0.2396\n", - "6/8, train_loss: 0.0135 step time: 0.2358\n", - "7/8, train_loss: 0.0201 step time: 0.2249\n", - "8/8, train_loss: 0.0140 step time: 0.2329\n", - "epoch 498 average loss: 0.0148\n", - "time consuming of epoch 498 is: 1.9756\n", - "----------\n", - "epoch 499/600\n", - "1/8, train_loss: 0.0135 step time: 0.3000\n", - "2/8, train_loss: 0.0138 step time: 0.2484\n", - "3/8, train_loss: 0.0130 step time: 0.3354\n", - "4/8, train_loss: 0.0145 step time: 0.4866\n", - "5/8, train_loss: 0.0150 step time: 0.2613\n", - "6/8, train_loss: 0.0178 step time: 0.2568\n", - "7/8, train_loss: 0.0122 step time: 0.2515\n", - "8/8, train_loss: 0.0160 step time: 0.2575\n", - "epoch 499 average loss: 0.0145\n", - "time consuming of epoch 499 is: 2.3988\n", - "----------\n", - "epoch 500/600\n", - "1/8, train_loss: 0.0145 step time: 0.3066\n", - "2/8, train_loss: 0.0153 step time: 0.2517\n", - "3/8, train_loss: 0.0149 step time: 0.2671\n", - "4/8, train_loss: 0.0135 step time: 0.2731\n", - "5/8, train_loss: 0.0133 step time: 0.2580\n", - "6/8, train_loss: 0.0122 step time: 0.2616\n", - "7/8, train_loss: 0.0151 step time: 0.2389\n", - "8/8, train_loss: 0.0154 step time: 0.2866\n", - "epoch 500 average loss: 0.0143\n", - "current epoch: 500 current mean dice: 0.9547 best mean dice: 0.9558 at epoch: 495\n", - "time consuming of epoch 500 is: 3.7080\n", - "----------\n", - "epoch 501/600\n", - "1/8, train_loss: 0.0162 step time: 0.3120\n", - "2/8, train_loss: 0.0138 step time: 0.2527\n", - "3/8, train_loss: 0.0140 step time: 0.2570\n", - "4/8, train_loss: 0.0160 step time: 0.2436\n", - "5/8, train_loss: 0.0125 step time: 0.2445\n", - "6/8, train_loss: 0.0173 step time: 0.3748\n", - "7/8, train_loss: 0.0142 step time: 0.4953\n", - "8/8, train_loss: 0.0148 step time: 0.2738\n", - "epoch 501 average loss: 0.0149\n", - "time consuming of epoch 501 is: 2.4549\n", - "----------\n", - "epoch 502/600\n", - "1/8, train_loss: 0.0157 step time: 0.3025\n", - "2/8, train_loss: 0.0163 step time: 0.2399\n", - "3/8, train_loss: 0.0129 step time: 0.2424\n", - "4/8, train_loss: 0.0128 step time: 0.2418\n", - "5/8, train_loss: 0.0131 step time: 0.2419\n", - "6/8, train_loss: 0.0146 step time: 0.3292\n", - "7/8, train_loss: 0.0198 step time: 0.3586\n", - "8/8, train_loss: 0.0146 step time: 0.2797\n", - "epoch 502 average loss: 0.0150\n", - "time consuming of epoch 502 is: 2.2376\n", - "----------\n", - "epoch 503/600\n", - "1/8, train_loss: 0.0132 step time: 0.3024\n", - "2/8, train_loss: 0.0161 step time: 0.2529\n", - "3/8, train_loss: 0.0146 step time: 0.2614\n", - "4/8, train_loss: 0.0145 step time: 0.2627\n", - "5/8, train_loss: 0.0162 step time: 0.2586\n", - "6/8, train_loss: 0.0156 step time: 0.2477\n", - "7/8, train_loss: 0.0167 step time: 0.2716\n", - "8/8, train_loss: 0.0145 step time: 0.2760\n", - "epoch 503 average loss: 0.0152\n", - "time consuming of epoch 503 is: 2.1349\n", - "----------\n", - "epoch 504/600\n", - "1/8, train_loss: 0.0152 step time: 0.3164\n", - "2/8, train_loss: 0.0172 step time: 0.2593\n", - "3/8, train_loss: 0.0109 step time: 0.3106\n", - "4/8, train_loss: 0.0162 step time: 0.3506\n", - "5/8, train_loss: 0.0154 step time: 0.2706\n", - "6/8, train_loss: 0.0151 step time: 0.2478\n", - "7/8, train_loss: 0.0167 step time: 0.2346\n", - "8/8, train_loss: 0.0136 step time: 0.2328\n", - "epoch 504 average loss: 0.0150\n", - "time consuming of epoch 504 is: 2.2241\n", - "----------\n", - "epoch 505/600\n", - "1/8, train_loss: 0.0166 step time: 0.2880\n", - "2/8, train_loss: 0.0154 step time: 0.2453\n", - "3/8, train_loss: 0.0115 step time: 0.2354\n", - "4/8, train_loss: 0.0132 step time: 0.2396\n", - "5/8, train_loss: 0.0133 step time: 0.3407\n", - "6/8, train_loss: 0.0160 step time: 0.3373\n", - "7/8, train_loss: 0.0152 step time: 0.2788\n", - "8/8, train_loss: 0.0179 step time: 0.2603\n", - "epoch 505 average loss: 0.0149\n", - "current epoch: 505 current mean dice: 0.9556 best mean dice: 0.9558 at epoch: 495\n", - "time consuming of epoch 505 is: 3.6885\n", - "----------\n", - "epoch 506/600\n", - "1/8, train_loss: 0.0163 step time: 0.3048\n", - "2/8, train_loss: 0.0147 step time: 0.2487\n", - "3/8, train_loss: 0.0128 step time: 0.2435\n", - "4/8, train_loss: 0.0156 step time: 0.2428\n", - "5/8, train_loss: 0.0142 step time: 0.2368\n", - "6/8, train_loss: 0.0135 step time: 0.2397\n", - "7/8, train_loss: 0.0139 step time: 0.2236\n", - "8/8, train_loss: 0.0153 step time: 0.2272\n", - "epoch 506 average loss: 0.0145\n", - "time consuming of epoch 506 is: 1.9684\n", - "----------\n", - "epoch 507/600\n", - "1/8, train_loss: 0.0147 step time: 0.5454\n", - "2/8, train_loss: 0.0137 step time: 0.2189\n", - "3/8, train_loss: 0.0162 step time: 0.2151\n", - "4/8, train_loss: 0.0214 step time: 0.2182\n", - "5/8, train_loss: 0.0153 step time: 0.2238\n", - "6/8, train_loss: 0.0172 step time: 0.2299\n", - "7/8, train_loss: 0.0147 step time: 0.2155\n", - "8/8, train_loss: 0.0170 step time: 0.3197\n", - "epoch 507 average loss: 0.0163\n", - "time consuming of epoch 507 is: 2.1880\n", - "----------\n", - "epoch 508/600\n", - "1/8, train_loss: 0.0144 step time: 0.4254\n", - "2/8, train_loss: 0.0150 step time: 0.2917\n", - "3/8, train_loss: 0.0141 step time: 0.2839\n", - "4/8, train_loss: 0.0136 step time: 0.2631\n", - "5/8, train_loss: 0.0169 step time: 0.2656\n", - "6/8, train_loss: 0.0183 step time: 0.2600\n", - "7/8, train_loss: 0.0148 step time: 0.2536\n", - "8/8, train_loss: 0.0149 step time: 0.2494\n", - "epoch 508 average loss: 0.0152\n", - "time consuming of epoch 508 is: 2.2942\n", - "----------\n", - "epoch 509/600\n", - "1/8, train_loss: 0.0139 step time: 0.2943\n", - "2/8, train_loss: 0.0147 step time: 0.2521\n", - "3/8, train_loss: 0.0144 step time: 0.2479\n", - "4/8, train_loss: 0.0144 step time: 0.2400\n", - "5/8, train_loss: 0.0154 step time: 0.2331\n", - "6/8, train_loss: 0.0129 step time: 0.2358\n", - "7/8, train_loss: 0.0142 step time: 0.3023\n", - "8/8, train_loss: 0.0129 step time: 0.2952\n", - "epoch 509 average loss: 0.0141\n", - "time consuming of epoch 509 is: 2.1022\n", - "----------\n", - "epoch 510/600\n", - "1/8, train_loss: 0.0140 step time: 0.3167\n", - "2/8, train_loss: 0.0162 step time: 0.2516\n", - "3/8, train_loss: 0.0140 step time: 0.2411\n", - "4/8, train_loss: 0.0147 step time: 0.2425\n", - "5/8, train_loss: 0.0166 step time: 0.2614\n", - "6/8, train_loss: 0.0152 step time: 0.2643\n", - "7/8, train_loss: 0.0120 step time: 0.4660\n", - "8/8, train_loss: 0.0160 step time: 0.4494\n", - "epoch 510 average loss: 0.0148\n", - "saved new best metric model\n", - "current epoch: 510 current mean dice: 0.9562 best mean dice: 0.9562 at epoch: 510\n", - "time consuming of epoch 510 is: 4.1072\n", - "----------\n", - "epoch 511/600\n", - "1/8, train_loss: 0.0143 step time: 0.2783\n", - "2/8, train_loss: 0.0127 step time: 0.2176\n", - "3/8, train_loss: 0.0147 step time: 0.3002\n", - "4/8, train_loss: 0.0159 step time: 0.4107\n", - "5/8, train_loss: 0.0166 step time: 0.2625\n", - "6/8, train_loss: 0.0171 step time: 0.2496\n", - "7/8, train_loss: 0.0126 step time: 0.2334\n", - "8/8, train_loss: 0.0153 step time: 0.2311\n", - "epoch 511 average loss: 0.0149\n", - "time consuming of epoch 511 is: 2.1861\n", - "----------\n", - "epoch 512/600\n", - "1/8, train_loss: 0.0134 step time: 0.2813\n", - "2/8, train_loss: 0.0180 step time: 0.2325\n", - "3/8, train_loss: 0.0167 step time: 0.2362\n", - "4/8, train_loss: 0.0139 step time: 0.2354\n", - "5/8, train_loss: 0.0146 step time: 0.5940\n", - "6/8, train_loss: 0.0121 step time: 0.2183\n", - "7/8, train_loss: 0.0138 step time: 0.2914\n", - "8/8, train_loss: 0.0096 step time: 0.3570\n", - "epoch 512 average loss: 0.0140\n", - "time consuming of epoch 512 is: 2.4477\n", - "----------\n", - "epoch 513/600\n", - "1/8, train_loss: 0.0143 step time: 0.3087\n", - "2/8, train_loss: 0.0123 step time: 0.2559\n", - "3/8, train_loss: 0.0117 step time: 0.2573\n", - "4/8, train_loss: 0.0167 step time: 0.2510\n", - "5/8, train_loss: 0.0161 step time: 0.2473\n", - "6/8, train_loss: 0.0151 step time: 0.2621\n", - "7/8, train_loss: 0.0146 step time: 0.2758\n", - "8/8, train_loss: 0.0154 step time: 0.2657\n", - "epoch 513 average loss: 0.0145\n", - "time consuming of epoch 513 is: 2.1253\n", - "----------\n", - "epoch 514/600\n", - "1/8, train_loss: 0.0143 step time: 0.3010\n", - "2/8, train_loss: 0.0131 step time: 0.2495\n", - "3/8, train_loss: 0.0143 step time: 0.2481\n", - "4/8, train_loss: 0.0136 step time: 0.2445\n", - "5/8, train_loss: 0.0141 step time: 0.2456\n", - "6/8, train_loss: 0.0165 step time: 0.2794\n", - "7/8, train_loss: 0.0157 step time: 0.3413\n", - "8/8, train_loss: 0.0139 step time: 0.2712\n", - "epoch 514 average loss: 0.0144\n", - "time consuming of epoch 514 is: 2.1821\n", - "----------\n", - "epoch 515/600\n", - "1/8, train_loss: 0.0141 step time: 0.3116\n", - "2/8, train_loss: 0.0137 step time: 0.2439\n", - "3/8, train_loss: 0.0168 step time: 0.3822\n", - "4/8, train_loss: 0.0177 step time: 0.5838\n", - "5/8, train_loss: 0.0122 step time: 0.2639\n", - "6/8, train_loss: 0.0138 step time: 0.2514\n", - "7/8, train_loss: 0.0132 step time: 0.2406\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "8/8, train_loss: 0.0165 step time: 0.2313\n", - "epoch 515 average loss: 0.0148\n", - "saved new best metric model\n", - "current epoch: 515 current mean dice: 0.9565 best mean dice: 0.9565 at epoch: 515\n", - "time consuming of epoch 515 is: 4.0181\n", - "----------\n", - "epoch 516/600\n", - "1/8, train_loss: 0.0108 step time: 0.2806\n", - "2/8, train_loss: 0.0134 step time: 0.2180\n", - "3/8, train_loss: 0.0137 step time: 0.2208\n", - "4/8, train_loss: 0.0148 step time: 0.2289\n", - "5/8, train_loss: 0.0139 step time: 0.3556\n", - "6/8, train_loss: 0.0127 step time: 0.3864\n", - "7/8, train_loss: 0.0161 step time: 0.2664\n", - "8/8, train_loss: 0.0208 step time: 0.2403\n", - "epoch 516 average loss: 0.0145\n", - "time consuming of epoch 516 is: 2.1994\n", - "----------\n", - "epoch 517/600\n", - "1/8, train_loss: 0.0149 step time: 0.2833\n", - "2/8, train_loss: 0.0119 step time: 0.2339\n", - "3/8, train_loss: 0.0152 step time: 0.2331\n", - "4/8, train_loss: 0.0142 step time: 0.2363\n", - "5/8, train_loss: 0.0149 step time: 0.2309\n", - "6/8, train_loss: 0.0143 step time: 0.5837\n", - "7/8, train_loss: 0.0154 step time: 0.2048\n", - "8/8, train_loss: 0.0143 step time: 0.2162\n", - "epoch 517 average loss: 0.0144\n", - "time consuming of epoch 517 is: 2.2236\n", - "----------\n", - "epoch 518/600\n", - "1/8, train_loss: 0.0146 step time: 0.2916\n", - "2/8, train_loss: 0.0159 step time: 0.2403\n", - "3/8, train_loss: 0.0137 step time: 0.2399\n", - "4/8, train_loss: 0.0136 step time: 0.2409\n", - "5/8, train_loss: 0.0155 step time: 0.2421\n", - "6/8, train_loss: 0.0138 step time: 0.2406\n", - "7/8, train_loss: 0.0149 step time: 0.3824\n", - "8/8, train_loss: 0.0137 step time: 0.4227\n", - "epoch 518 average loss: 0.0145\n", - "time consuming of epoch 518 is: 2.3021\n", - "----------\n", - "epoch 519/600\n", - "1/8, train_loss: 0.0111 step time: 0.2984\n", - "2/8, train_loss: 0.0159 step time: 0.2545\n", - "3/8, train_loss: 0.0151 step time: 0.2518\n", - "4/8, train_loss: 0.0141 step time: 0.2689\n", - "5/8, train_loss: 0.0140 step time: 0.2885\n", - "6/8, train_loss: 0.0155 step time: 0.2850\n", - "7/8, train_loss: 0.0148 step time: 0.2648\n", - "8/8, train_loss: 0.0153 step time: 0.2524\n", - "epoch 519 average loss: 0.0145\n", - "time consuming of epoch 519 is: 2.1658\n", - "----------\n", - "epoch 520/600\n", - "1/8, train_loss: 0.0125 step time: 0.2932\n", - "2/8, train_loss: 0.0151 step time: 0.2636\n", - "3/8, train_loss: 0.0172 step time: 0.2738\n", - "4/8, train_loss: 0.0140 step time: 0.2517\n", - "5/8, train_loss: 0.0151 step time: 0.2490\n", - "6/8, train_loss: 0.0132 step time: 0.2492\n", - "7/8, train_loss: 0.0149 step time: 0.2344\n", - "8/8, train_loss: 0.0120 step time: 0.2347\n", - "epoch 520 average loss: 0.0142\n", - "current epoch: 520 current mean dice: 0.9562 best mean dice: 0.9565 at epoch: 515\n", - "time consuming of epoch 520 is: 3.6764\n", - "----------\n", - "epoch 521/600\n", - "1/8, train_loss: 0.0156 step time: 0.2901\n", - "2/8, train_loss: 0.0123 step time: 0.2437\n", - "3/8, train_loss: 0.0133 step time: 0.2409\n", - "4/8, train_loss: 0.0120 step time: 0.2425\n", - "5/8, train_loss: 0.0148 step time: 0.2423\n", - "6/8, train_loss: 0.0149 step time: 0.2399\n", - "7/8, train_loss: 0.0151 step time: 0.2385\n", - "8/8, train_loss: 0.0267 step time: 0.2370\n", - "epoch 521 average loss: 0.0156\n", - "time consuming of epoch 521 is: 1.9763\n", - "----------\n", - "epoch 522/600\n", - "1/8, train_loss: 0.0130 step time: 0.2929\n", - "2/8, train_loss: 0.0153 step time: 0.2497\n", - "3/8, train_loss: 0.0166 step time: 0.2513\n", - "4/8, train_loss: 0.0161 step time: 0.2485\n", - "5/8, train_loss: 0.0120 step time: 0.3345\n", - "6/8, train_loss: 0.0143 step time: 0.5173\n", - "7/8, train_loss: 0.0162 step time: 0.2787\n", - "8/8, train_loss: 0.0190 step time: 0.2540\n", - "epoch 522 average loss: 0.0153\n", - "time consuming of epoch 522 is: 2.4284\n", - "----------\n", - "epoch 523/600\n", - "1/8, train_loss: 0.0159 step time: 0.2937\n", - "2/8, train_loss: 0.0130 step time: 0.2518\n", - "3/8, train_loss: 0.0133 step time: 0.2571\n", - "4/8, train_loss: 0.0123 step time: 0.2578\n", - "5/8, train_loss: 0.0173 step time: 0.2509\n", - "6/8, train_loss: 0.0138 step time: 0.2434\n", - "7/8, train_loss: 0.0157 step time: 0.2387\n", - "8/8, train_loss: 0.0133 step time: 0.2357\n", - "epoch 523 average loss: 0.0143\n", - "time consuming of epoch 523 is: 2.0307\n", - "----------\n", - "epoch 524/600\n", - "1/8, train_loss: 0.0162 step time: 0.2824\n", - "2/8, train_loss: 0.0153 step time: 0.2725\n", - "3/8, train_loss: 0.0135 step time: 0.2768\n", - "4/8, train_loss: 0.0141 step time: 0.2673\n", - "5/8, train_loss: 0.0158 step time: 0.2587\n", - "6/8, train_loss: 0.0186 step time: 0.2584\n", - "7/8, train_loss: 0.0150 step time: 0.2603\n", - "8/8, train_loss: 0.0157 step time: 0.2508\n", - "epoch 524 average loss: 0.0155\n", - "time consuming of epoch 524 is: 2.1288\n", - "----------\n", - "epoch 525/600\n", - "1/8, train_loss: 0.0152 step time: 0.3036\n", - "2/8, train_loss: 0.0139 step time: 0.2506\n", - "3/8, train_loss: 0.0150 step time: 0.3531\n", - "4/8, train_loss: 0.0150 step time: 0.3199\n", - "5/8, train_loss: 0.0117 step time: 0.3079\n", - "6/8, train_loss: 0.0149 step time: 0.2763\n", - "7/8, train_loss: 0.0127 step time: 0.2791\n", - "8/8, train_loss: 0.0143 step time: 0.2566\n", - "epoch 525 average loss: 0.0141\n", - "saved new best metric model\n", - "current epoch: 525 current mean dice: 0.9567 best mean dice: 0.9567 at epoch: 525\n", - "time consuming of epoch 525 is: 3.9988\n", - "----------\n", - "epoch 526/600\n", - "1/8, train_loss: 0.0133 step time: 0.2811\n", - "2/8, train_loss: 0.0160 step time: 0.2136\n", - "3/8, train_loss: 0.0159 step time: 0.2140\n", - "4/8, train_loss: 0.0150 step time: 0.2393\n", - "5/8, train_loss: 0.0118 step time: 0.2959\n", - "6/8, train_loss: 0.0137 step time: 0.2766\n", - "7/8, train_loss: 0.0136 step time: 0.2584\n", - "8/8, train_loss: 0.0156 step time: 0.2799\n", - "epoch 526 average loss: 0.0144\n", - "time consuming of epoch 526 is: 2.0613\n", - "----------\n", - "epoch 527/600\n", - "1/8, train_loss: 0.0127 step time: 0.3468\n", - "2/8, train_loss: 0.0163 step time: 0.2859\n", - "3/8, train_loss: 0.0124 step time: 0.2578\n", - "4/8, train_loss: 0.0145 step time: 0.2914\n", - "5/8, train_loss: 0.0156 step time: 0.2973\n", - "6/8, train_loss: 0.0109 step time: 0.2932\n", - "7/8, train_loss: 0.0143 step time: 0.2718\n", - "8/8, train_loss: 0.0172 step time: 0.2638\n", - "epoch 527 average loss: 0.0142\n", - "time consuming of epoch 527 is: 2.3094\n", - "----------\n", - "epoch 528/600\n", - "1/8, train_loss: 0.0100 step time: 0.3013\n", - "2/8, train_loss: 0.0156 step time: 0.3216\n", - "3/8, train_loss: 0.0144 step time: 0.4727\n", - "4/8, train_loss: 0.0135 step time: 0.2812\n", - "5/8, train_loss: 0.0187 step time: 0.2765\n", - "6/8, train_loss: 0.0151 step time: 0.2646\n", - "7/8, train_loss: 0.0148 step time: 0.2627\n", - "8/8, train_loss: 0.0128 step time: 0.2539\n", - "epoch 528 average loss: 0.0144\n", - "time consuming of epoch 528 is: 2.4362\n", - "----------\n", - "epoch 529/600\n", - "1/8, train_loss: 0.0166 step time: 0.2836\n", - "2/8, train_loss: 0.0134 step time: 0.3092\n", - "3/8, train_loss: 0.0122 step time: 0.3309\n", - "4/8, train_loss: 0.0147 step time: 0.3043\n", - "5/8, train_loss: 0.0148 step time: 0.2759\n", - "6/8, train_loss: 0.0144 step time: 0.2740\n", - "7/8, train_loss: 0.0147 step time: 0.2561\n", - "8/8, train_loss: 0.0126 step time: 0.2645\n", - "epoch 529 average loss: 0.0142\n", - "time consuming of epoch 529 is: 2.3000\n", - "----------\n", - "epoch 530/600\n", - "1/8, train_loss: 0.0147 step time: 0.2985\n", - "2/8, train_loss: 0.0144 step time: 0.2450\n", - "3/8, train_loss: 0.0142 step time: 0.2609\n", - "4/8, train_loss: 0.0128 step time: 0.2716\n", - "5/8, train_loss: 0.0149 step time: 0.2654\n", - "6/8, train_loss: 0.0130 step time: 0.2532\n", - "7/8, train_loss: 0.0161 step time: 0.2523\n", - "8/8, train_loss: 0.0157 step time: 0.2405\n", - "epoch 530 average loss: 0.0145\n", - "current epoch: 530 current mean dice: 0.9563 best mean dice: 0.9567 at epoch: 525\n", - "time consuming of epoch 530 is: 3.6364\n", - "----------\n", - "epoch 531/600\n", - "1/8, train_loss: 0.0153 step time: 0.4621\n", - "2/8, train_loss: 0.0133 step time: 0.3092\n", - "3/8, train_loss: 0.0168 step time: 0.2747\n", - "4/8, train_loss: 0.0157 step time: 0.2507\n", - "5/8, train_loss: 0.0122 step time: 0.2570\n", - "6/8, train_loss: 0.0143 step time: 0.2439\n", - "7/8, train_loss: 0.0171 step time: 0.2315\n", - "8/8, train_loss: 0.0143 step time: 0.2616\n", - "epoch 531 average loss: 0.0149\n", - "time consuming of epoch 531 is: 2.2918\n", - "----------\n", - "epoch 532/600\n", - "1/8, train_loss: 0.0097 step time: 0.3294\n", - "2/8, train_loss: 0.0155 step time: 0.2748\n", - "3/8, train_loss: 0.0137 step time: 0.2602\n", - "4/8, train_loss: 0.0135 step time: 0.2605\n", - "5/8, train_loss: 0.0149 step time: 0.2521\n", - "6/8, train_loss: 0.0145 step time: 0.2545\n", - "7/8, train_loss: 0.0159 step time: 0.2502\n", - "8/8, train_loss: 0.0139 step time: 0.2766\n", - "epoch 532 average loss: 0.0139\n", - "time consuming of epoch 532 is: 2.1597\n", - "----------\n", - "epoch 533/600\n", - "1/8, train_loss: 0.0132 step time: 0.3204\n", - "2/8, train_loss: 0.0169 step time: 0.2607\n", - "3/8, train_loss: 0.0168 step time: 0.2449\n", - "4/8, train_loss: 0.0188 step time: 0.2478\n", - "5/8, train_loss: 0.0121 step time: 0.2413\n", - "6/8, train_loss: 0.0124 step time: 0.2325\n", - "7/8, train_loss: 0.0142 step time: 0.2253\n", - "8/8, train_loss: 0.0104 step time: 0.2318\n", - "epoch 533 average loss: 0.0144\n", - "time consuming of epoch 533 is: 2.0062\n", - "----------\n", - "epoch 534/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0158 step time: 0.2971\n", - "2/8, train_loss: 0.0121 step time: 0.2513\n", - "3/8, train_loss: 0.0148 step time: 0.2558\n", - "4/8, train_loss: 0.0152 step time: 0.2820\n", - "5/8, train_loss: 0.0173 step time: 0.2766\n", - "6/8, train_loss: 0.0168 step time: 0.2653\n", - "7/8, train_loss: 0.0126 step time: 0.4649\n", - "8/8, train_loss: 0.0146 step time: 0.4537\n", - "epoch 534 average loss: 0.0149\n", - "time consuming of epoch 534 is: 2.5483\n", - "----------\n", - "epoch 535/600\n", - "1/8, train_loss: 0.0147 step time: 0.2978\n", - "2/8, train_loss: 0.0157 step time: 0.2453\n", - "3/8, train_loss: 0.0137 step time: 0.2440\n", - "4/8, train_loss: 0.0155 step time: 0.2463\n", - "5/8, train_loss: 0.0124 step time: 0.2448\n", - "6/8, train_loss: 0.0135 step time: 0.2398\n", - "7/8, train_loss: 0.0128 step time: 0.2258\n", - "8/8, train_loss: 0.0161 step time: 0.2454\n", - "epoch 535 average loss: 0.0143\n", - "current epoch: 535 current mean dice: 0.9561 best mean dice: 0.9567 at epoch: 525\n", - "time consuming of epoch 535 is: 3.4063\n", - "----------\n", - "epoch 536/600\n", - "1/8, train_loss: 0.0149 step time: 0.3464\n", - "2/8, train_loss: 0.0120 step time: 0.2958\n", - "3/8, train_loss: 0.0137 step time: 0.2792\n", - "4/8, train_loss: 0.0161 step time: 0.2764\n", - "5/8, train_loss: 0.0138 step time: 0.2547\n", - "6/8, train_loss: 0.0139 step time: 0.2499\n", - "7/8, train_loss: 0.0117 step time: 0.2587\n", - "8/8, train_loss: 0.0128 step time: 0.2594\n", - "epoch 536 average loss: 0.0136\n", - "time consuming of epoch 536 is: 2.2218\n", - "----------\n", - "epoch 537/600\n", - "1/8, train_loss: 0.0133 step time: 0.2981\n", - "2/8, train_loss: 0.0163 step time: 0.2580\n", - "3/8, train_loss: 0.0111 step time: 0.2510\n", - "4/8, train_loss: 0.0145 step time: 0.2650\n", - "5/8, train_loss: 0.0140 step time: 0.2841\n", - "6/8, train_loss: 0.0159 step time: 0.2684\n", - "7/8, train_loss: 0.0131 step time: 0.5561\n", - "8/8, train_loss: 0.0128 step time: 0.4515\n", - "epoch 537 average loss: 0.0139\n", - "time consuming of epoch 537 is: 2.6338\n", - "----------\n", - "epoch 538/600\n", - "1/8, train_loss: 0.0143 step time: 0.3007\n", - "2/8, train_loss: 0.0180 step time: 0.2419\n", - "3/8, train_loss: 0.0130 step time: 0.2413\n", - "4/8, train_loss: 0.0115 step time: 0.2409\n", - "5/8, train_loss: 0.0165 step time: 0.2401\n", - "6/8, train_loss: 0.0132 step time: 0.2738\n", - "7/8, train_loss: 0.0173 step time: 0.2786\n", - "8/8, train_loss: 0.0109 step time: 0.2742\n", - "epoch 538 average loss: 0.0144\n", - "time consuming of epoch 538 is: 2.0930\n", - "----------\n", - "epoch 539/600\n", - "1/8, train_loss: 0.0160 step time: 0.3116\n", - "2/8, train_loss: 0.0138 step time: 0.2459\n", - "3/8, train_loss: 0.0181 step time: 0.2418\n", - "4/8, train_loss: 0.0139 step time: 0.2407\n", - "5/8, train_loss: 0.0139 step time: 0.2310\n", - "6/8, train_loss: 0.0117 step time: 0.2599\n", - "7/8, train_loss: 0.0144 step time: 0.3260\n", - "8/8, train_loss: 0.0163 step time: 0.2811\n", - "epoch 539 average loss: 0.0148\n", - "time consuming of epoch 539 is: 2.1394\n", - "----------\n", - "epoch 540/600\n", - "1/8, train_loss: 0.0136 step time: 0.3150\n", - "2/8, train_loss: 0.0142 step time: 0.2706\n", - "3/8, train_loss: 0.0168 step time: 0.2724\n", - "4/8, train_loss: 0.0151 step time: 0.2811\n", - "5/8, train_loss: 0.0137 step time: 0.2603\n", - "6/8, train_loss: 0.0168 step time: 0.2532\n", - "7/8, train_loss: 0.0154 step time: 0.2510\n", - "8/8, train_loss: 0.0128 step time: 0.2392\n", - "epoch 540 average loss: 0.0148\n", - "current epoch: 540 current mean dice: 0.9520 best mean dice: 0.9567 at epoch: 525\n", - "time consuming of epoch 540 is: 3.5460\n", - "----------\n", - "epoch 541/600\n", - "1/8, train_loss: 0.0160 step time: 0.3325\n", - "2/8, train_loss: 0.0154 step time: 0.2838\n", - "3/8, train_loss: 0.0149 step time: 0.2659\n", - "4/8, train_loss: 0.0152 step time: 0.2480\n", - "5/8, train_loss: 0.0168 step time: 0.2434\n", - "6/8, train_loss: 0.0129 step time: 0.2442\n", - "7/8, train_loss: 0.0129 step time: 0.2238\n", - "8/8, train_loss: 0.0142 step time: 0.2262\n", - "epoch 541 average loss: 0.0148\n", - "time consuming of epoch 541 is: 2.0691\n", - "----------\n", - "epoch 542/600\n", - "1/8, train_loss: 0.0162 step time: 0.4539\n", - "2/8, train_loss: 0.0144 step time: 0.5945\n", - "3/8, train_loss: 0.0120 step time: 0.2692\n", - "4/8, train_loss: 0.0174 step time: 0.2546\n", - "5/8, train_loss: 0.0120 step time: 0.2419\n", - "6/8, train_loss: 0.0150 step time: 0.2485\n", - "7/8, train_loss: 0.0168 step time: 0.5282\n", - "8/8, train_loss: 0.0154 step time: 0.4443\n", - "epoch 542 average loss: 0.0149\n", - "time consuming of epoch 542 is: 3.0366\n", - "----------\n", - "epoch 543/600\n", - "1/8, train_loss: 0.0147 step time: 0.3248\n", - "2/8, train_loss: 0.0155 step time: 0.2644\n", - "3/8, train_loss: 0.0154 step time: 0.2435\n", - "4/8, train_loss: 0.0148 step time: 0.2387\n", - "5/8, train_loss: 0.0112 step time: 0.2401\n", - "6/8, train_loss: 0.0125 step time: 0.2385\n", - "7/8, train_loss: 0.0144 step time: 0.2328\n", - "8/8, train_loss: 0.0149 step time: 0.2566\n", - "epoch 543 average loss: 0.0142\n", - "time consuming of epoch 543 is: 2.0409\n", - "----------\n", - "epoch 544/600\n", - "1/8, train_loss: 0.0142 step time: 0.3262\n", - "2/8, train_loss: 0.0176 step time: 0.2701\n", - "3/8, train_loss: 0.0137 step time: 0.2718\n", - "4/8, train_loss: 0.0149 step time: 0.2686\n", - "5/8, train_loss: 0.0150 step time: 0.2699\n", - "6/8, train_loss: 0.0131 step time: 0.2661\n", - "7/8, train_loss: 0.0108 step time: 0.2779\n", - "8/8, train_loss: 0.0128 step time: 0.2981\n", - "epoch 544 average loss: 0.0140\n", - "time consuming of epoch 544 is: 2.2502\n", - "----------\n", - "epoch 545/600\n", - "1/8, train_loss: 0.0117 step time: 0.3072\n", - "2/8, train_loss: 0.0153 step time: 0.2550\n", - "3/8, train_loss: 0.0175 step time: 0.2835\n", - "4/8, train_loss: 0.0144 step time: 0.2772\n", - "5/8, train_loss: 0.0126 step time: 0.2636\n", - "6/8, train_loss: 0.0117 step time: 0.2735\n", - "7/8, train_loss: 0.0133 step time: 0.2633\n", - "8/8, train_loss: 0.0122 step time: 0.2554\n", - "epoch 545 average loss: 0.0136\n", - "saved new best metric model\n", - "current epoch: 545 current mean dice: 0.9570 best mean dice: 0.9570 at epoch: 545\n", - "time consuming of epoch 545 is: 3.9191\n", - "----------\n", - "epoch 546/600\n", - "1/8, train_loss: 0.0119 step time: 0.2768\n", - "2/8, train_loss: 0.0178 step time: 0.2553\n", - "3/8, train_loss: 0.0161 step time: 0.3216\n", - "4/8, train_loss: 0.0151 step time: 0.2712\n", - "5/8, train_loss: 0.0112 step time: 0.2617\n", - "6/8, train_loss: 0.0170 step time: 0.2595\n", - "7/8, train_loss: 0.0137 step time: 0.2862\n", - "8/8, train_loss: 0.0126 step time: 0.3048\n", - "epoch 546 average loss: 0.0144\n", - "time consuming of epoch 546 is: 2.2397\n", - "----------\n", - "epoch 547/600\n", - "1/8, train_loss: 0.0135 step time: 0.3232\n", - "2/8, train_loss: 0.0176 step time: 0.2631\n", - "3/8, train_loss: 0.0130 step time: 0.2953\n", - "4/8, train_loss: 0.0164 step time: 0.3203\n", - "5/8, train_loss: 0.0130 step time: 0.3093\n", - "6/8, train_loss: 0.0148 step time: 0.3132\n", - "7/8, train_loss: 0.0120 step time: 0.2893\n", - "8/8, train_loss: 0.0111 step time: 0.2786\n", - "epoch 547 average loss: 0.0139\n", - "time consuming of epoch 547 is: 2.3938\n", - "----------\n", - "epoch 548/600\n", - "1/8, train_loss: 0.0156 step time: 0.3058\n", - "2/8, train_loss: 0.0136 step time: 0.2517\n", - "3/8, train_loss: 0.0149 step time: 0.2477\n", - "4/8, train_loss: 0.0135 step time: 0.2518\n", - "5/8, train_loss: 0.0112 step time: 0.4019\n", - "6/8, train_loss: 0.0144 step time: 0.4811\n", - "7/8, train_loss: 0.0131 step time: 0.2657\n", - "8/8, train_loss: 0.0139 step time: 0.2435\n", - "epoch 548 average loss: 0.0138\n", - "time consuming of epoch 548 is: 2.4506\n", - "----------\n", - "epoch 549/600\n", - "1/8, train_loss: 0.0132 step time: 0.2858\n", - "2/8, train_loss: 0.0144 step time: 0.2677\n", - "3/8, train_loss: 0.0137 step time: 0.2752\n", - "4/8, train_loss: 0.0094 step time: 0.2639\n", - "5/8, train_loss: 0.0139 step time: 0.2572\n", - "6/8, train_loss: 0.0166 step time: 0.2466\n", - "7/8, train_loss: 0.0140 step time: 0.2240\n", - "8/8, train_loss: 0.0124 step time: 0.2238\n", - "epoch 549 average loss: 0.0135\n", - "time consuming of epoch 549 is: 2.0456\n", - "----------\n", - "epoch 550/600\n", - "1/8, train_loss: 0.0144 step time: 0.2811\n", - "2/8, train_loss: 0.0146 step time: 0.2308\n", - "3/8, train_loss: 0.0164 step time: 0.2300\n", - "4/8, train_loss: 0.0159 step time: 0.2301\n", - "5/8, train_loss: 0.0127 step time: 0.2309\n", - "6/8, train_loss: 0.0140 step time: 0.3292\n", - "7/8, train_loss: 0.0127 step time: 0.3068\n", - "8/8, train_loss: 0.0141 step time: 0.2757\n", - "epoch 550 average loss: 0.0143\n", - "current epoch: 550 current mean dice: 0.9567 best mean dice: 0.9570 at epoch: 545\n", - "time consuming of epoch 550 is: 3.7092\n", - "----------\n", - "epoch 551/600\n", - "1/8, train_loss: 0.0152 step time: 0.3052\n", - "2/8, train_loss: 0.0152 step time: 0.2695\n", - "3/8, train_loss: 0.0142 step time: 0.2651\n", - "4/8, train_loss: 0.0142 step time: 0.2739\n", - "5/8, train_loss: 0.0092 step time: 0.2701\n", - "6/8, train_loss: 0.0133 step time: 0.2813\n", - "7/8, train_loss: 0.0148 step time: 0.2793\n", - "8/8, train_loss: 0.0147 step time: 0.2570\n", - "epoch 551 average loss: 0.0138\n", - "time consuming of epoch 551 is: 2.2028\n", - "----------\n", - "epoch 552/600\n", - "1/8, train_loss: 0.0122 step time: 0.3062\n", - "2/8, train_loss: 0.0151 step time: 0.2590\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "3/8, train_loss: 0.0152 step time: 0.2525\n", - "4/8, train_loss: 0.0137 step time: 0.2521\n", - "5/8, train_loss: 0.0170 step time: 0.2521\n", - "6/8, train_loss: 0.0131 step time: 0.2461\n", - "7/8, train_loss: 0.0143 step time: 0.2536\n", - "8/8, train_loss: 0.0126 step time: 0.2838\n", - "epoch 552 average loss: 0.0141\n", - "time consuming of epoch 552 is: 2.1070\n", - "----------\n", - "epoch 553/600\n", - "1/8, train_loss: 0.0155 step time: 0.3193\n", - "2/8, train_loss: 0.0113 step time: 0.2586\n", - "3/8, train_loss: 0.0137 step time: 0.2527\n", - "4/8, train_loss: 0.0173 step time: 0.2613\n", - "5/8, train_loss: 0.0149 step time: 0.3232\n", - "6/8, train_loss: 0.0147 step time: 0.3309\n", - "7/8, train_loss: 0.0111 step time: 0.2825\n", - "8/8, train_loss: 0.0148 step time: 0.2544\n", - "epoch 553 average loss: 0.0142\n", - "time consuming of epoch 553 is: 2.2843\n", - "----------\n", - "epoch 554/600\n", - "1/8, train_loss: 0.0168 step time: 0.2938\n", - "2/8, train_loss: 0.0137 step time: 0.3006\n", - "3/8, train_loss: 0.0152 step time: 0.3725\n", - "4/8, train_loss: 0.0110 step time: 0.2663\n", - "5/8, train_loss: 0.0150 step time: 0.2566\n", - "6/8, train_loss: 0.0131 step time: 0.2496\n", - "7/8, train_loss: 0.0132 step time: 0.3977\n", - "8/8, train_loss: 0.0126 step time: 0.5971\n", - "epoch 554 average loss: 0.0138\n", - "time consuming of epoch 554 is: 2.7356\n", - "----------\n", - "epoch 555/600\n", - "1/8, train_loss: 0.0121 step time: 0.2695\n", - "2/8, train_loss: 0.0147 step time: 0.2194\n", - "3/8, train_loss: 0.0141 step time: 0.2225\n", - "4/8, train_loss: 0.0153 step time: 0.2231\n", - "5/8, train_loss: 0.0139 step time: 0.2223\n", - "6/8, train_loss: 0.0122 step time: 0.3301\n", - "7/8, train_loss: 0.0157 step time: 0.4462\n", - "8/8, train_loss: 0.0144 step time: 0.2575\n", - "epoch 555 average loss: 0.0141\n", - "current epoch: 555 current mean dice: 0.9569 best mean dice: 0.9570 at epoch: 545\n", - "time consuming of epoch 555 is: 3.6251\n", - "----------\n", - "epoch 556/600\n", - "1/8, train_loss: 0.0140 step time: 0.3296\n", - "2/8, train_loss: 0.0150 step time: 0.2743\n", - "3/8, train_loss: 0.0118 step time: 0.2643\n", - "4/8, train_loss: 0.0165 step time: 0.2546\n", - "5/8, train_loss: 0.0166 step time: 0.2500\n", - "6/8, train_loss: 0.0132 step time: 0.4506\n", - "7/8, train_loss: 0.0122 step time: 0.4417\n", - "8/8, train_loss: 0.0139 step time: 0.2567\n", - "epoch 556 average loss: 0.0141\n", - "time consuming of epoch 556 is: 2.5230\n", - "----------\n", - "epoch 557/600\n", - "1/8, train_loss: 0.0129 step time: 0.2935\n", - "2/8, train_loss: 0.0127 step time: 0.2377\n", - "3/8, train_loss: 0.0134 step time: 0.2375\n", - "4/8, train_loss: 0.0135 step time: 0.2400\n", - "5/8, train_loss: 0.0176 step time: 0.2340\n", - "6/8, train_loss: 0.0171 step time: 0.2342\n", - "7/8, train_loss: 0.0112 step time: 0.4926\n", - "8/8, train_loss: 0.0159 step time: 0.3657\n", - "epoch 557 average loss: 0.0143\n", - "time consuming of epoch 557 is: 2.3367\n", - "----------\n", - "epoch 558/600\n", - "1/8, train_loss: 0.0134 step time: 0.3071\n", - "2/8, train_loss: 0.0115 step time: 0.2493\n", - "3/8, train_loss: 0.0142 step time: 0.2489\n", - "4/8, train_loss: 0.0151 step time: 0.2708\n", - "5/8, train_loss: 0.0174 step time: 0.2718\n", - "6/8, train_loss: 0.0146 step time: 0.2617\n", - "7/8, train_loss: 0.0144 step time: 0.2643\n", - "8/8, train_loss: 0.0155 step time: 0.2598\n", - "epoch 558 average loss: 0.0145\n", - "time consuming of epoch 558 is: 2.1351\n", - "----------\n", - "epoch 559/600\n", - "1/8, train_loss: 0.0117 step time: 0.3035\n", - "2/8, train_loss: 0.0169 step time: 0.2492\n", - "3/8, train_loss: 0.0131 step time: 0.2437\n", - "4/8, train_loss: 0.0159 step time: 0.2446\n", - "5/8, train_loss: 0.0140 step time: 0.3306\n", - "6/8, train_loss: 0.0154 step time: 0.4943\n", - "7/8, train_loss: 0.0132 step time: 0.2757\n", - "8/8, train_loss: 0.0124 step time: 0.2856\n", - "epoch 559 average loss: 0.0141\n", - "time consuming of epoch 559 is: 2.4287\n", - "----------\n", - "epoch 560/600\n", - "1/8, train_loss: 0.0107 step time: 0.3177\n", - "2/8, train_loss: 0.0169 step time: 0.2547\n", - "3/8, train_loss: 0.0145 step time: 0.2553\n", - "4/8, train_loss: 0.0126 step time: 0.2534\n", - "5/8, train_loss: 0.0151 step time: 0.2493\n", - "6/8, train_loss: 0.0133 step time: 0.2442\n", - "7/8, train_loss: 0.0151 step time: 0.2384\n", - "8/8, train_loss: 0.0150 step time: 0.2188\n", - "epoch 560 average loss: 0.0142\n", - "current epoch: 560 current mean dice: 0.9564 best mean dice: 0.9570 at epoch: 545\n", - "time consuming of epoch 560 is: 3.5012\n", - "----------\n", - "epoch 561/600\n", - "1/8, train_loss: 0.0149 step time: 0.3237\n", - "2/8, train_loss: 0.0152 step time: 0.2688\n", - "3/8, train_loss: 0.0124 step time: 0.2498\n", - "4/8, train_loss: 0.0126 step time: 0.2531\n", - "5/8, train_loss: 0.0139 step time: 0.2501\n", - "6/8, train_loss: 0.0130 step time: 0.3305\n", - "7/8, train_loss: 0.0140 step time: 0.3248\n", - "8/8, train_loss: 0.0138 step time: 0.2744\n", - "epoch 561 average loss: 0.0137\n", - "time consuming of epoch 561 is: 2.2764\n", - "----------\n", - "epoch 562/600\n", - "1/8, train_loss: 0.0125 step time: 0.3088\n", - "2/8, train_loss: 0.0105 step time: 0.2554\n", - "3/8, train_loss: 0.0156 step time: 0.2535\n", - "4/8, train_loss: 0.0155 step time: 0.2513\n", - "5/8, train_loss: 0.0151 step time: 0.2617\n", - "6/8, train_loss: 0.0162 step time: 0.2479\n", - "7/8, train_loss: 0.0127 step time: 0.2259\n", - "8/8, train_loss: 0.0131 step time: 0.2304\n", - "epoch 562 average loss: 0.0139\n", - "time consuming of epoch 562 is: 2.0363\n", - "----------\n", - "epoch 563/600\n", - "1/8, train_loss: 0.0133 step time: 0.2802\n", - "2/8, train_loss: 0.0133 step time: 0.2375\n", - "3/8, train_loss: 0.0159 step time: 0.2416\n", - "4/8, train_loss: 0.0129 step time: 0.2395\n", - "5/8, train_loss: 0.0140 step time: 0.2387\n", - "6/8, train_loss: 0.0141 step time: 0.2907\n", - "7/8, train_loss: 0.0102 step time: 0.3092\n", - "8/8, train_loss: 0.0155 step time: 0.3050\n", - "epoch 563 average loss: 0.0137\n", - "time consuming of epoch 563 is: 2.1439\n", - "----------\n", - "epoch 564/600\n", - "1/8, train_loss: 0.0149 step time: 0.3201\n", - "2/8, train_loss: 0.0139 step time: 0.2792\n", - "3/8, train_loss: 0.0146 step time: 0.2759\n", - "4/8, train_loss: 0.0112 step time: 0.2639\n", - "5/8, train_loss: 0.0136 step time: 0.2607\n", - "6/8, train_loss: 0.0146 step time: 0.2570\n", - "7/8, train_loss: 0.0122 step time: 0.2500\n", - "8/8, train_loss: 0.0139 step time: 0.2361\n", - "epoch 564 average loss: 0.0136\n", - "time consuming of epoch 564 is: 2.1447\n", - "----------\n", - "epoch 565/600\n", - "1/8, train_loss: 0.0132 step time: 0.2835\n", - "2/8, train_loss: 0.0138 step time: 0.2398\n", - "3/8, train_loss: 0.0137 step time: 0.2372\n", - "4/8, train_loss: 0.0111 step time: 0.2353\n", - "5/8, train_loss: 0.0129 step time: 0.2749\n", - "6/8, train_loss: 0.0166 step time: 0.2783\n", - "7/8, train_loss: 0.0156 step time: 0.2740\n", - "8/8, train_loss: 0.0144 step time: 0.2854\n", - "epoch 565 average loss: 0.0139\n", - "saved new best metric model\n", - "current epoch: 565 current mean dice: 0.9570 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 565 is: 3.8064\n", - "----------\n", - "epoch 566/600\n", - "1/8, train_loss: 0.0144 step time: 0.2843\n", - "2/8, train_loss: 0.0140 step time: 0.2134\n", - "3/8, train_loss: 0.0144 step time: 0.2217\n", - "4/8, train_loss: 0.0146 step time: 0.2218\n", - "5/8, train_loss: 0.0103 step time: 0.2293\n", - "6/8, train_loss: 0.0145 step time: 0.2388\n", - "7/8, train_loss: 0.0115 step time: 0.2437\n", - "8/8, train_loss: 0.0121 step time: 0.3030\n", - "epoch 566 average loss: 0.0132\n", - "time consuming of epoch 566 is: 1.9587\n", - "----------\n", - "epoch 567/600\n", - "1/8, train_loss: 0.0141 step time: 0.3355\n", - "2/8, train_loss: 0.0138 step time: 0.2807\n", - "3/8, train_loss: 0.0110 step time: 0.2573\n", - "4/8, train_loss: 0.0156 step time: 0.2518\n", - "5/8, train_loss: 0.0144 step time: 0.2596\n", - "6/8, train_loss: 0.0136 step time: 0.2557\n", - "7/8, train_loss: 0.0145 step time: 0.2572\n", - "8/8, train_loss: 0.0145 step time: 0.2509\n", - "epoch 567 average loss: 0.0139\n", - "time consuming of epoch 567 is: 2.1502\n", - "----------\n", - "epoch 568/600\n", - "1/8, train_loss: 0.0137 step time: 0.2900\n", - "2/8, train_loss: 0.0146 step time: 0.3190\n", - "3/8, train_loss: 0.0164 step time: 0.3527\n", - "4/8, train_loss: 0.0145 step time: 0.3196\n", - "5/8, train_loss: 0.0112 step time: 0.2951\n", - "6/8, train_loss: 0.0150 step time: 0.2814\n", - "7/8, train_loss: 0.0127 step time: 0.2766\n", - "8/8, train_loss: 0.0132 step time: 0.2762\n", - "epoch 568 average loss: 0.0139\n", - "time consuming of epoch 568 is: 2.4121\n", - "----------\n", - "epoch 569/600\n", - "1/8, train_loss: 0.0147 step time: 0.2996\n", - "2/8, train_loss: 0.0119 step time: 0.2535\n", - "3/8, train_loss: 0.0121 step time: 0.2492\n", - "4/8, train_loss: 0.0154 step time: 0.2533\n", - "5/8, train_loss: 0.0122 step time: 0.3694\n", - "6/8, train_loss: 0.0164 step time: 0.3542\n", - "7/8, train_loss: 0.0131 step time: 0.2772\n", - "8/8, train_loss: 0.0141 step time: 0.2744\n", - "epoch 569 average loss: 0.0137\n", - "time consuming of epoch 569 is: 2.3321\n", - "----------\n", - "epoch 570/600\n", - "1/8, train_loss: 0.0115 step time: 0.3015\n", - "2/8, train_loss: 0.0143 step time: 0.2503\n", - "3/8, train_loss: 0.0147 step time: 0.2460\n", - "4/8, train_loss: 0.0158 step time: 0.2478\n", - "5/8, train_loss: 0.0140 step time: 0.2503\n", - "6/8, train_loss: 0.0114 step time: 0.2428\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "7/8, train_loss: 0.0130 step time: 0.2395\n", - "8/8, train_loss: 0.0131 step time: 0.2599\n", - "epoch 570 average loss: 0.0135\n", - "current epoch: 570 current mean dice: 0.9564 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 570 is: 3.4808\n", - "----------\n", - "epoch 571/600\n", - "1/8, train_loss: 0.0124 step time: 0.3131\n", - "2/8, train_loss: 0.0163 step time: 0.2542\n", - "3/8, train_loss: 0.0127 step time: 0.2498\n", - "4/8, train_loss: 0.0159 step time: 0.2447\n", - "5/8, train_loss: 0.0161 step time: 0.2467\n", - "6/8, train_loss: 0.0150 step time: 0.2346\n", - "7/8, train_loss: 0.0142 step time: 0.2817\n", - "8/8, train_loss: 0.0121 step time: 0.3427\n", - "epoch 571 average loss: 0.0143\n", - "time consuming of epoch 571 is: 2.1687\n", - "----------\n", - "epoch 572/600\n", - "1/8, train_loss: 0.0134 step time: 0.3317\n", - "2/8, train_loss: 0.0140 step time: 0.2901\n", - "3/8, train_loss: 0.0131 step time: 0.2800\n", - "4/8, train_loss: 0.0135 step time: 0.2641\n", - "5/8, train_loss: 0.0140 step time: 0.2651\n", - "6/8, train_loss: 0.0156 step time: 0.2617\n", - "7/8, train_loss: 0.0179 step time: 0.2425\n", - "8/8, train_loss: 0.0138 step time: 0.2324\n", - "epoch 572 average loss: 0.0144\n", - "time consuming of epoch 572 is: 2.1691\n", - "----------\n", - "epoch 573/600\n", - "1/8, train_loss: 0.0136 step time: 0.2835\n", - "2/8, train_loss: 0.0142 step time: 0.2430\n", - "3/8, train_loss: 0.0126 step time: 0.2444\n", - "4/8, train_loss: 0.0149 step time: 0.4018\n", - "5/8, train_loss: 0.0173 step time: 0.4729\n", - "6/8, train_loss: 0.0149 step time: 0.2708\n", - "7/8, train_loss: 0.0118 step time: 0.2485\n", - "8/8, train_loss: 0.0157 step time: 0.2524\n", - "epoch 573 average loss: 0.0144\n", - "time consuming of epoch 573 is: 2.4188\n", - "----------\n", - "epoch 574/600\n", - "1/8, train_loss: 0.0165 step time: 0.3020\n", - "2/8, train_loss: 0.0143 step time: 0.2468\n", - "3/8, train_loss: 0.0129 step time: 0.2630\n", - "4/8, train_loss: 0.0154 step time: 0.2666\n", - "5/8, train_loss: 0.0175 step time: 0.3344\n", - "6/8, train_loss: 0.0159 step time: 0.3198\n", - "7/8, train_loss: 0.0119 step time: 0.3130\n", - "8/8, train_loss: 0.0155 step time: 0.3013\n", - "epoch 574 average loss: 0.0150\n", - "time consuming of epoch 574 is: 2.3484\n", - "----------\n", - "epoch 575/600\n", - "1/8, train_loss: 0.0162 step time: 0.3148\n", - "2/8, train_loss: 0.0146 step time: 0.2578\n", - "3/8, train_loss: 0.0196 step time: 0.2583\n", - "4/8, train_loss: 0.0176 step time: 0.2645\n", - "5/8, train_loss: 0.0278 step time: 0.2587\n", - "6/8, train_loss: 0.0203 step time: 0.2587\n", - "7/8, train_loss: 0.0189 step time: 0.3994\n", - "8/8, train_loss: 0.0173 step time: 0.3447\n", - "epoch 575 average loss: 0.0190\n", - "current epoch: 575 current mean dice: 0.6157 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 575 is: 3.5838\n", - "----------\n", - "epoch 576/600\n", - "1/8, train_loss: 0.0267 step time: 0.3742\n", - "2/8, train_loss: 0.0423 step time: 0.3487\n", - "3/8, train_loss: 0.0208 step time: 0.3219\n", - "4/8, train_loss: 0.0596 step time: 0.3010\n", - "5/8, train_loss: 0.0293 step time: 0.2699\n", - "6/8, train_loss: 0.0247 step time: 0.2537\n", - "7/8, train_loss: 0.0616 step time: 0.2398\n", - "8/8, train_loss: 0.0180 step time: 0.2390\n", - "epoch 576 average loss: 0.0354\n", - "time consuming of epoch 576 is: 2.3494\n", - "----------\n", - "epoch 577/600\n", - "1/8, train_loss: 0.0172 step time: 0.2959\n", - "2/8, train_loss: 0.0322 step time: 0.2391\n", - "3/8, train_loss: 0.0407 step time: 0.2459\n", - "4/8, train_loss: 0.0196 step time: 0.3281\n", - "5/8, train_loss: 0.0171 step time: 0.3268\n", - "6/8, train_loss: 0.0202 step time: 0.2792\n", - "7/8, train_loss: 0.0321 step time: 0.2779\n", - "8/8, train_loss: 0.0475 step time: 0.2798\n", - "epoch 577 average loss: 0.0283\n", - "time consuming of epoch 577 is: 2.2742\n", - "----------\n", - "epoch 578/600\n", - "1/8, train_loss: 0.0216 step time: 0.3330\n", - "2/8, train_loss: 0.0193 step time: 0.2659\n", - "3/8, train_loss: 0.0246 step time: 0.2686\n", - "4/8, train_loss: 0.0217 step time: 0.2494\n", - "5/8, train_loss: 0.0611 step time: 0.2525\n", - "6/8, train_loss: 0.0256 step time: 0.3037\n", - "7/8, train_loss: 0.0258 step time: 0.3017\n", - "8/8, train_loss: 0.0495 step time: 0.2911\n", - "epoch 578 average loss: 0.0312\n", - "time consuming of epoch 578 is: 2.2673\n", - "----------\n", - "epoch 579/600\n", - "1/8, train_loss: 0.0287 step time: 0.3131\n", - "2/8, train_loss: 0.0192 step time: 0.2610\n", - "3/8, train_loss: 0.0218 step time: 0.2511\n", - "4/8, train_loss: 0.0208 step time: 0.2446\n", - "5/8, train_loss: 0.0256 step time: 0.2426\n", - "6/8, train_loss: 0.0236 step time: 0.2344\n", - "7/8, train_loss: 0.0237 step time: 0.2271\n", - "8/8, train_loss: 0.0341 step time: 0.2308\n", - "epoch 579 average loss: 0.0247\n", - "time consuming of epoch 579 is: 2.0062\n", - "----------\n", - "epoch 580/600\n", - "1/8, train_loss: 0.0241 step time: 0.2838\n", - "2/8, train_loss: 0.0213 step time: 0.2402\n", - "3/8, train_loss: 0.0162 step time: 0.2398\n", - "4/8, train_loss: 0.0223 step time: 0.2467\n", - "5/8, train_loss: 0.0239 step time: 0.2537\n", - "6/8, train_loss: 0.0266 step time: 0.2510\n", - "7/8, train_loss: 0.0197 step time: 0.2480\n", - "8/8, train_loss: 0.0190 step time: 0.2935\n", - "epoch 580 average loss: 0.0216\n", - "current epoch: 580 current mean dice: 0.5842 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 580 is: 3.3831\n", - "----------\n", - "epoch 581/600\n", - "1/8, train_loss: 0.0208 step time: 0.3043\n", - "2/8, train_loss: 0.0243 step time: 0.2567\n", - "3/8, train_loss: 0.0150 step time: 0.2624\n", - "4/8, train_loss: 0.0211 step time: 0.2601\n", - "5/8, train_loss: 0.0207 step time: 0.2537\n", - "6/8, train_loss: 0.0200 step time: 0.2360\n", - "7/8, train_loss: 0.0142 step time: 0.2273\n", - "8/8, train_loss: 0.0199 step time: 0.2274\n", - "epoch 581 average loss: 0.0195\n", - "time consuming of epoch 581 is: 2.0291\n", - "----------\n", - "epoch 582/600\n", - "1/8, train_loss: 0.0205 step time: 0.5857\n", - "2/8, train_loss: 0.0220 step time: 0.2857\n", - "3/8, train_loss: 0.0212 step time: 0.2568\n", - "4/8, train_loss: 0.0202 step time: 0.2480\n", - "5/8, train_loss: 0.0169 step time: 0.2397\n", - "6/8, train_loss: 0.0185 step time: 0.2460\n", - "7/8, train_loss: 0.0153 step time: 0.2693\n", - "8/8, train_loss: 0.0191 step time: 0.2764\n", - "epoch 582 average loss: 0.0192\n", - "time consuming of epoch 582 is: 2.4090\n", - "----------\n", - "epoch 583/600\n", - "1/8, train_loss: 0.0169 step time: 0.3217\n", - "2/8, train_loss: 0.0161 step time: 0.2769\n", - "3/8, train_loss: 0.0189 step time: 0.2645\n", - "4/8, train_loss: 0.0221 step time: 0.2606\n", - "5/8, train_loss: 0.0195 step time: 0.2523\n", - "6/8, train_loss: 0.0188 step time: 0.2425\n", - "7/8, train_loss: 0.0171 step time: 0.2274\n", - "8/8, train_loss: 0.0226 step time: 0.2284\n", - "epoch 583 average loss: 0.0190\n", - "time consuming of epoch 583 is: 2.0756\n", - "----------\n", - "epoch 584/600\n", - "1/8, train_loss: 0.0178 step time: 0.2821\n", - "2/8, train_loss: 0.0172 step time: 0.2427\n", - "3/8, train_loss: 0.0163 step time: 0.2449\n", - "4/8, train_loss: 0.0233 step time: 0.3345\n", - "5/8, train_loss: 0.0173 step time: 0.3845\n", - "6/8, train_loss: 0.0200 step time: 0.2720\n", - "7/8, train_loss: 0.0156 step time: 0.2522\n", - "8/8, train_loss: 0.0182 step time: 0.2409\n", - "epoch 584 average loss: 0.0182\n", - "time consuming of epoch 584 is: 2.2554\n", - "----------\n", - "epoch 585/600\n", - "1/8, train_loss: 0.0216 step time: 0.2776\n", - "2/8, train_loss: 0.0234 step time: 0.2341\n", - "3/8, train_loss: 0.0189 step time: 0.2320\n", - "4/8, train_loss: 0.0155 step time: 0.2310\n", - "5/8, train_loss: 0.0181 step time: 0.2312\n", - "6/8, train_loss: 0.0184 step time: 0.5762\n", - "7/8, train_loss: 0.0172 step time: 0.2038\n", - "8/8, train_loss: 0.0141 step time: 0.2035\n", - "epoch 585 average loss: 0.0184\n", - "current epoch: 585 current mean dice: 0.9476 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 585 is: 3.8190\n", - "----------\n", - "epoch 586/600\n", - "1/8, train_loss: 0.0221 step time: 0.3041\n", - "2/8, train_loss: 0.0193 step time: 0.2530\n", - "3/8, train_loss: 0.0194 step time: 0.2511\n", - "4/8, train_loss: 0.0144 step time: 0.2564\n", - "5/8, train_loss: 0.0224 step time: 0.2544\n", - "6/8, train_loss: 0.0169 step time: 0.2564\n", - "7/8, train_loss: 0.0175 step time: 0.2531\n", - "8/8, train_loss: 0.0150 step time: 0.2549\n", - "epoch 586 average loss: 0.0184\n", - "time consuming of epoch 586 is: 2.0845\n", - "----------\n", - "epoch 587/600\n", - "1/8, train_loss: 0.0180 step time: 0.3033\n", - "2/8, train_loss: 0.0138 step time: 0.2540\n", - "3/8, train_loss: 0.0162 step time: 0.3777\n", - "4/8, train_loss: 0.0169 step time: 0.3005\n", - "5/8, train_loss: 0.0182 step time: 0.2977\n", - "6/8, train_loss: 0.0183 step time: 0.2835\n", - "7/8, train_loss: 0.0141 step time: 0.2805\n", - "8/8, train_loss: 0.0169 step time: 0.2580\n", - "epoch 587 average loss: 0.0165\n", - "time consuming of epoch 587 is: 2.3567\n", - "----------\n", - "epoch 588/600\n", - "1/8, train_loss: 0.0169 step time: 0.3040\n", - "2/8, train_loss: 0.0157 step time: 0.2603\n", - "3/8, train_loss: 0.0196 step time: 0.2434\n", - "4/8, train_loss: 0.0199 step time: 0.2664\n", - "5/8, train_loss: 0.0147 step time: 0.3162\n", - "6/8, train_loss: 0.0151 step time: 0.2892\n", - "7/8, train_loss: 0.0154 step time: 0.2603\n", - "8/8, train_loss: 0.0134 step time: 0.2456\n", - "epoch 588 average loss: 0.0163\n", - "time consuming of epoch 588 is: 2.1869\n", - "----------\n", - "epoch 589/600\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "1/8, train_loss: 0.0148 step time: 0.2963\n", - "2/8, train_loss: 0.0142 step time: 0.2491\n", - "3/8, train_loss: 0.0160 step time: 0.2398\n", - "4/8, train_loss: 0.0185 step time: 0.2387\n", - "5/8, train_loss: 0.0187 step time: 0.2353\n", - "6/8, train_loss: 0.0134 step time: 0.2410\n", - "7/8, train_loss: 0.0173 step time: 0.2309\n", - "8/8, train_loss: 0.0172 step time: 0.2321\n", - "epoch 589 average loss: 0.0162\n", - "time consuming of epoch 589 is: 1.9646\n", - "----------\n", - "epoch 590/600\n", - "1/8, train_loss: 0.0149 step time: 0.2860\n", - "2/8, train_loss: 0.0176 step time: 0.3303\n", - "3/8, train_loss: 0.0145 step time: 0.3770\n", - "4/8, train_loss: 0.0157 step time: 0.3063\n", - "5/8, train_loss: 0.0167 step time: 0.2762\n", - "6/8, train_loss: 0.0138 step time: 0.2607\n", - "7/8, train_loss: 0.0205 step time: 0.2597\n", - "8/8, train_loss: 0.0162 step time: 0.2618\n", - "epoch 590 average loss: 0.0162\n", - "current epoch: 590 current mean dice: 0.9524 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 590 is: 3.6541\n", - "----------\n", - "epoch 591/600\n", - "1/8, train_loss: 0.0179 step time: 0.3185\n", - "2/8, train_loss: 0.0162 step time: 0.2596\n", - "3/8, train_loss: 0.0159 step time: 0.2525\n", - "4/8, train_loss: 0.0172 step time: 0.2594\n", - "5/8, train_loss: 0.0150 step time: 0.2445\n", - "6/8, train_loss: 0.0203 step time: 0.2416\n", - "7/8, train_loss: 0.0155 step time: 0.2225\n", - "8/8, train_loss: 0.0192 step time: 0.2232\n", - "epoch 591 average loss: 0.0171\n", - "time consuming of epoch 591 is: 2.0231\n", - "----------\n", - "epoch 592/600\n", - "1/8, train_loss: 0.0133 step time: 0.5008\n", - "2/8, train_loss: 0.0136 step time: 0.2161\n", - "3/8, train_loss: 0.0166 step time: 0.2123\n", - "4/8, train_loss: 0.0156 step time: 0.2163\n", - "5/8, train_loss: 0.0162 step time: 0.2322\n", - "6/8, train_loss: 0.0157 step time: 0.2843\n", - "7/8, train_loss: 0.0126 step time: 0.3000\n", - "8/8, train_loss: 0.0233 step time: 0.2872\n", - "epoch 592 average loss: 0.0159\n", - "time consuming of epoch 592 is: 2.2507\n", - "----------\n", - "epoch 593/600\n", - "1/8, train_loss: 0.0268 step time: 0.3071\n", - "2/8, train_loss: 0.0133 step time: 0.2545\n", - "3/8, train_loss: 0.0190 step time: 0.2471\n", - "4/8, train_loss: 0.0175 step time: 0.2415\n", - "5/8, train_loss: 0.0192 step time: 0.2407\n", - "6/8, train_loss: 0.0144 step time: 0.3130\n", - "7/8, train_loss: 0.0137 step time: 0.4752\n", - "8/8, train_loss: 0.0175 step time: 0.2735\n", - "epoch 593 average loss: 0.0177\n", - "time consuming of epoch 593 is: 2.3541\n", - "----------\n", - "epoch 594/600\n", - "1/8, train_loss: 0.0182 step time: 0.3081\n", - "2/8, train_loss: 0.0204 step time: 0.2519\n", - "3/8, train_loss: 0.0159 step time: 0.2555\n", - "4/8, train_loss: 0.0162 step time: 0.2554\n", - "5/8, train_loss: 0.0161 step time: 0.2515\n", - "6/8, train_loss: 0.0191 step time: 0.2488\n", - "7/8, train_loss: 0.0175 step time: 0.2403\n", - "8/8, train_loss: 0.0121 step time: 0.2411\n", - "epoch 594 average loss: 0.0169\n", - "time consuming of epoch 594 is: 2.0542\n", - "----------\n", - "epoch 595/600\n", - "1/8, train_loss: 0.0151 step time: 0.2936\n", - "2/8, train_loss: 0.0138 step time: 0.2961\n", - "3/8, train_loss: 0.0150 step time: 0.3332\n", - "4/8, train_loss: 0.0222 step time: 0.2901\n", - "5/8, train_loss: 0.0152 step time: 0.2760\n", - "6/8, train_loss: 0.0160 step time: 0.2778\n", - "7/8, train_loss: 0.0187 step time: 0.2739\n", - "8/8, train_loss: 0.0175 step time: 0.2603\n", - "epoch 595 average loss: 0.0167\n", - "current epoch: 595 current mean dice: 0.9536 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 595 is: 3.5425\n", - "----------\n", - "epoch 596/600\n", - "1/8, train_loss: 0.0151 step time: 0.3051\n", - "2/8, train_loss: 0.0151 step time: 0.2521\n", - "3/8, train_loss: 0.0162 step time: 0.2437\n", - "4/8, train_loss: 0.0165 step time: 0.2434\n", - "5/8, train_loss: 0.0154 step time: 0.3578\n", - "6/8, train_loss: 0.0138 step time: 0.4849\n", - "7/8, train_loss: 0.0148 step time: 0.2703\n", - "8/8, train_loss: 0.0142 step time: 0.2516\n", - "epoch 596 average loss: 0.0151\n", - "time consuming of epoch 596 is: 2.4104\n", - "----------\n", - "epoch 597/600\n", - "1/8, train_loss: 0.0157 step time: 0.2957\n", - "2/8, train_loss: 0.0170 step time: 0.2438\n", - "3/8, train_loss: 0.0154 step time: 0.2434\n", - "4/8, train_loss: 0.0147 step time: 0.2382\n", - "5/8, train_loss: 0.0172 step time: 0.2416\n", - "6/8, train_loss: 0.0157 step time: 0.2393\n", - "7/8, train_loss: 0.0124 step time: 0.2195\n", - "8/8, train_loss: 0.0173 step time: 0.3617\n", - "epoch 597 average loss: 0.0157\n", - "time consuming of epoch 597 is: 2.0847\n", - "----------\n", - "epoch 598/600\n", - "1/8, train_loss: 0.0168 step time: 0.4376\n", - "2/8, train_loss: 0.0163 step time: 0.2782\n", - "3/8, train_loss: 0.0144 step time: 0.2617\n", - "4/8, train_loss: 0.0183 step time: 0.2626\n", - "5/8, train_loss: 0.0160 step time: 0.2489\n", - "6/8, train_loss: 0.0135 step time: 0.2485\n", - "7/8, train_loss: 0.0159 step time: 0.2269\n", - "8/8, train_loss: 0.0129 step time: 0.2705\n", - "epoch 598 average loss: 0.0155\n", - "time consuming of epoch 598 is: 2.2365\n", - "----------\n", - "epoch 599/600\n", - "1/8, train_loss: 0.0137 step time: 0.3769\n", - "2/8, train_loss: 0.0160 step time: 0.2753\n", - "3/8, train_loss: 0.0170 step time: 0.2577\n", - "4/8, train_loss: 0.0162 step time: 0.2495\n", - "5/8, train_loss: 0.0192 step time: 0.2774\n", - "6/8, train_loss: 0.0162 step time: 0.3099\n", - "7/8, train_loss: 0.0140 step time: 0.2596\n", - "8/8, train_loss: 0.0174 step time: 0.2547\n", - "epoch 599 average loss: 0.0162\n", - "time consuming of epoch 599 is: 2.2626\n", - "----------\n", - "epoch 600/600\n", - "1/8, train_loss: 0.0149 step time: 0.3104\n", - "2/8, train_loss: 0.0137 step time: 0.2571\n", - "3/8, train_loss: 0.0172 step time: 0.2480\n", - "4/8, train_loss: 0.0157 step time: 0.2512\n", - "5/8, train_loss: 0.0133 step time: 0.2491\n", - "6/8, train_loss: 0.0148 step time: 0.3255\n", - "7/8, train_loss: 0.0131 step time: 0.3099\n", - "8/8, train_loss: 0.0143 step time: 0.2755\n", - "epoch 600 average loss: 0.0146\n", - "current epoch: 600 current mean dice: 0.9553 best mean dice: 0.9570 at epoch: 565\n", - "time consuming of epoch 600 is: 3.5567\n", - "train completed, best_metric: 0.9570 at epoch: 565 total time: 1423.5083\n", - "total time of 600 epochs with MONAI fast training: 1423.5083, time of preparing cache: 52.4564\n" - ] - } - ], + "outputs": [], "source": [ "set_determinism(seed=0)\n", "monai_start = time.time()\n", @@ -8369,7 +814,7 @@ }, { "cell_type": "code", - "execution_count": 23, + "execution_count": 11, "metadata": { "vscode": { "languageId": "python" @@ -8378,7 +823,7 @@ "outputs": [ { "data": { - "image/png": "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", + "image/png": "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\n", "text/plain": [ "
" ] @@ -8437,7 +882,7 @@ }, { "cell_type": "code", - "execution_count": 24, + "execution_count": 12, "metadata": { "vscode": { "languageId": "python" @@ -8446,7 +891,7 @@ "outputs": [ { "data": { - "image/png": "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", + "image/png": "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\n", "text/plain": [ "
" ] @@ -8491,7 +936,7 @@ }, { "cell_type": "code", - "execution_count": 25, + "execution_count": 13, "metadata": { "vscode": { "languageId": "python" @@ -8500,7 +945,7 @@ "outputs": [ { "data": { - "image/png": "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", + "image/png": "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\n", "text/plain": [ "
" ]