LR_adaptive_learning.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "ExecuteTime": {
     "start_time": "2019-08-16T06:31:59.472Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/data/yaliu/jupyterbooks/multi-KD/models/teacher/resnet20.py:36: UserWarning: nn.init.kaiming_normal is now deprecated in favor of nn.init.kaiming_normal_.\n",
      "  init.kaiming_normal(m.weight)\n",
      "WARNING:root:Setting up a new session...\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n",
      "StudentNet:\n",
      "\n",
      "ResNet(\n",
      "  (conv1): Conv2d(3, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "  (bn1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "  (layer1): Sequential(\n",
      "    (0): BasicBlock(\n",
      "      (conv1): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "    (1): BasicBlock(\n",
      "      (conv1): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "    (2): BasicBlock(\n",
      "      (conv1): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(16, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(16, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "  )\n",
      "  (layer2): Sequential(\n",
      "    (0): BasicBlock(\n",
      "      (conv1): Conv2d(16, 32, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): LambdaLayer()\n",
      "    )\n",
      "    (1): BasicBlock(\n",
      "      (conv1): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "    (2): BasicBlock(\n",
      "      (conv1): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "  )\n",
      "  (layer3): Sequential(\n",
      "    (0): BasicBlock(\n",
      "      (conv1): Conv2d(32, 64, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): LambdaLayer()\n",
      "    )\n",
      "    (1): BasicBlock(\n",
      "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "    (2): BasicBlock(\n",
      "      (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (shortcut): Sequential()\n",
      "    )\n",
      "  )\n",
      "  (linear): Linear(in_features=64, out_features=10, bias=True)\n",
      ")\n",
      "Training adapter:\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:254: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:185: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:151: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/torch/nn/functional.py:1992: UserWarning: reduction: 'mean' divides the total loss by both the batch size and the support size.'batchmean' divides only by the batch size, and aligns with the KL div math definition.'mean' will be changed to behave the same as 'batchmean' in the next major release.\n",
      "  warnings.warn(\"reduction: 'mean' divides the total loss by both the batch size and the support size.\"\n",
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/torch/nn/_reduction.py:15: UserWarning: reduction='elementwise_mean' is deprecated, please use reduction='mean' instead.\n",
      "  warnings.warn(\"reduction='elementwise_mean' is deprecated, please use reduction='mean' instead.\")\n",
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:269: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch[0/80]adapter Loss: 1.8923\n",
      "epoch[1/80]adapter Loss: 1.2946\n",
      "epoch[2/80]adapter Loss: 1.2766\n",
      "epoch[3/80]adapter Loss: 0.9707\n",
      "epoch[4/80]adapter Loss: 0.8982\n",
      "epoch[5/80]adapter Loss: 1.0691\n",
      "epoch[6/80]adapter Loss: 0.9799\n",
      "epoch[7/80]adapter Loss: 0.8803\n",
      "epoch[8/80]adapter Loss: 0.8061\n",
      "epoch[9/80]adapter Loss: 0.9717\n",
      "epoch[10/80]adapter Loss: 0.7828\n",
      "epoch[11/80]adapter Loss: 0.9247\n",
      "epoch[12/80]adapter Loss: 0.8041\n",
      "epoch[13/80]adapter Loss: 0.7219\n",
      "epoch[14/80]adapter Loss: 0.8872\n",
      "epoch[15/80]adapter Loss: 0.7206\n",
      "epoch[16/80]adapter Loss: 0.7422\n",
      "epoch[17/80]adapter Loss: 0.7651\n",
      "epoch[18/80]adapter Loss: 0.6501\n",
      "epoch[19/80]adapter Loss: 0.6966\n",
      "epoch[20/80]adapter Loss: 0.7639\n",
      "epoch[21/80]adapter Loss: 0.7309\n",
      "epoch[22/80]adapter Loss: 0.6546\n",
      "epoch[23/80]adapter Loss: 0.8422\n",
      "epoch[24/80]adapter Loss: 0.6552\n",
      "epoch[25/80]adapter Loss: 0.8946\n",
      "epoch[26/80]adapter Loss: 0.7836\n",
      "epoch[27/80]adapter Loss: 0.7034\n",
      "epoch[28/80]adapter Loss: 0.6236\n",
      "epoch[29/80]adapter Loss: 0.7160\n",
      "epoch[30/80]adapter Loss: 0.7357\n",
      "epoch[31/80]adapter Loss: 0.7657\n",
      "epoch[32/80]adapter Loss: 0.7219\n",
      "epoch[33/80]adapter Loss: 0.7153\n",
      "epoch[34/80]adapter Loss: 0.5891\n",
      "epoch[35/80]adapter Loss: 0.8619\n",
      "epoch[36/80]adapter Loss: 0.7001\n",
      "epoch[37/80]adapter Loss: 0.6439\n",
      "epoch[38/80]adapter Loss: 0.7583\n",
      "epoch[39/80]adapter Loss: 0.7245\n",
      "epoch[40/80]adapter Loss: 0.9473\n",
      "epoch[41/80]adapter Loss: 0.6758\n",
      "epoch[42/80]adapter Loss: 0.5753\n",
      "epoch[43/80]adapter Loss: 0.5383\n",
      "epoch[44/80]adapter Loss: 0.7821\n",
      "epoch[45/80]adapter Loss: 0.5553\n",
      "epoch[46/80]adapter Loss: 0.6444\n",
      "epoch[47/80]adapter Loss: 0.5563\n",
      "epoch[48/80]adapter Loss: 0.7357\n",
      "epoch[49/80]adapter Loss: 0.7372\n",
      "epoch[50/80]adapter Loss: 0.5705\n",
      "epoch[51/80]adapter Loss: 0.6225\n",
      "epoch[52/80]adapter Loss: 0.7482\n",
      "epoch[53/80]adapter Loss: 0.6195\n",
      "epoch[54/80]adapter Loss: 0.7055\n",
      "epoch[55/80]adapter Loss: 0.6531\n",
      "epoch[56/80]adapter Loss: 0.7794\n",
      "epoch[57/80]adapter Loss: 0.6703\n",
      "epoch[58/80]adapter Loss: 0.5540\n",
      "epoch[59/80]adapter Loss: 0.6401\n",
      "epoch[60/80]adapter Loss: 0.5266\n",
      "epoch[61/80]adapter Loss: 0.6510\n",
      "epoch[62/80]adapter Loss: 0.7359\n",
      "epoch[63/80]adapter Loss: 0.7207\n",
      "epoch[64/80]adapter Loss: 0.7860\n",
      "epoch[65/80]adapter Loss: 0.6376\n",
      "epoch[66/80]adapter Loss: 0.7005\n",
      "epoch[67/80]adapter Loss: 0.6211\n",
      "epoch[68/80]adapter Loss: 0.7102\n",
      "epoch[69/80]adapter Loss: 0.7402\n",
      "epoch[70/80]adapter Loss: 0.5999\n",
      "epoch[71/80]adapter Loss: 0.7565\n",
      "epoch[72/80]adapter Loss: 0.6990\n",
      "epoch[73/80]adapter Loss: 0.6810\n",
      "epoch[74/80]adapter Loss: 0.5474\n",
      "epoch[75/80]adapter Loss: 0.7612\n",
      "epoch[76/80]adapter Loss: 0.6874\n",
      "epoch[77/80]adapter Loss: 0.8057\n",
      "epoch[78/80]adapter Loss: 0.8718\n",
      "epoch[79/80]adapter Loss: 0.7209\n",
      "--- adapter training cost 88.161 mins ---\n",
      "\n",
      "===> epoch: 1/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.195 (0.195)\tData 0.020 (0.020)\tLoss 0.6173 (0.6173)\tPrec@1 89.062 (89.062)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:319: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n",
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:332: UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10/391]\tTime 0.184 (0.186)\tData 0.020 (0.021)\tLoss 0.5601 (0.6160)\tPrec@1 89.062 (87.500)\n",
      "[20/391]\tTime 0.179 (0.184)\tData 0.020 (0.020)\tLoss 0.6172 (0.6336)\tPrec@1 85.156 (86.570)\n",
      "[30/391]\tTime 0.181 (0.183)\tData 0.020 (0.020)\tLoss 0.6505 (0.6236)\tPrec@1 84.375 (86.492)\n",
      "[40/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5673 (0.6204)\tPrec@1 90.625 (86.662)\n",
      "[50/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6040 (0.6169)\tPrec@1 84.375 (86.765)\n",
      "[60/391]\tTime 0.145 (0.181)\tData 0.020 (0.020)\tLoss 0.6210 (0.6187)\tPrec@1 85.156 (86.680)\n",
      "[70/391]\tTime 0.143 (0.176)\tData 0.020 (0.020)\tLoss 0.5430 (0.6168)\tPrec@1 89.062 (86.818)\n",
      "[80/391]\tTime 0.178 (0.176)\tData 0.019 (0.020)\tLoss 0.5997 (0.6120)\tPrec@1 85.938 (86.979)\n",
      "[90/391]\tTime 0.186 (0.177)\tData 0.019 (0.020)\tLoss 0.7550 (0.6183)\tPrec@1 82.031 (86.779)\n",
      "[100/391]\tTime 0.183 (0.177)\tData 0.022 (0.020)\tLoss 0.4827 (0.6167)\tPrec@1 92.969 (86.819)\n",
      "[110/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.5332 (0.6182)\tPrec@1 90.625 (86.838)\n",
      "[120/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6325 (0.6181)\tPrec@1 86.719 (86.822)\n",
      "[130/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.7508 (0.6186)\tPrec@1 81.250 (86.844)\n",
      "[140/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6535 (0.6223)\tPrec@1 86.719 (86.741)\n",
      "[150/391]\tTime 0.184 (0.178)\tData 0.021 (0.020)\tLoss 0.6265 (0.6200)\tPrec@1 85.156 (86.843)\n",
      "[160/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.7110 (0.6200)\tPrec@1 85.156 (86.830)\n",
      "[170/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6124 (0.6199)\tPrec@1 89.844 (86.838)\n",
      "[180/391]\tTime 0.182 (0.179)\tData 0.022 (0.020)\tLoss 0.6051 (0.6202)\tPrec@1 87.500 (86.805)\n",
      "[190/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5788 (0.6195)\tPrec@1 89.844 (86.801)\n",
      "[200/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.7786 (0.6212)\tPrec@1 82.031 (86.758)\n",
      "[210/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5764 (0.6230)\tPrec@1 89.062 (86.704)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5566 (0.6238)\tPrec@1 89.062 (86.676)\n",
      "[230/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6274 (0.6247)\tPrec@1 85.938 (86.611)\n",
      "[240/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5896 (0.6246)\tPrec@1 88.281 (86.592)\n",
      "[250/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6102 (0.6245)\tPrec@1 87.500 (86.600)\n",
      "[260/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.6167 (0.6249)\tPrec@1 83.594 (86.566)\n",
      "[270/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.7128 (0.6261)\tPrec@1 77.344 (86.494)\n",
      "[280/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6418 (0.6261)\tPrec@1 86.719 (86.477)\n",
      "[290/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5815 (0.6265)\tPrec@1 88.281 (86.445)\n",
      "[300/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6358 (0.6282)\tPrec@1 83.594 (86.374)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.8307 (0.6301)\tPrec@1 78.906 (86.302)\n",
      "[320/391]\tTime 0.188 (0.180)\tData 0.020 (0.020)\tLoss 0.7201 (0.6315)\tPrec@1 82.031 (86.239)\n",
      "[330/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6366 (0.6326)\tPrec@1 87.500 (86.202)\n",
      "[340/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6135 (0.6322)\tPrec@1 86.719 (86.206)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7602 (0.6320)\tPrec@1 80.469 (86.189)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.6419 (0.6317)\tPrec@1 85.156 (86.178)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5739 (0.6316)\tPrec@1 86.719 (86.199)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5707 (0.6320)\tPrec@1 88.281 (86.179)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.012 (0.020)\tLoss 0.6884 (0.6326)\tPrec@1 85.000 (86.180)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9060 (0.9060)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.9348 (1.1300)\tPrec@1 84.375 (78.409)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0432 (1.1373)\tPrec@1 78.906 (78.460)\n",
      "Test: [30/79]\tTime 0.027 (0.023)\tLoss 0.9692 (1.1056)\tPrec@1 80.469 (79.007)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.3377 (1.0966)\tPrec@1 80.469 (79.554)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 1.3359 (1.0866)\tPrec@1 80.469 (79.550)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.0817 (1.0862)\tPrec@1 78.906 (79.239)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 0.7001 (1.0840)\tPrec@1 83.594 (79.269)\n",
      " * Prec@1 79.360\n",
      "\n",
      "===> epoch: 2/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.154 (0.154)\tData 0.020 (0.020)\tLoss 0.5787 (0.5787)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.182 (0.177)\tData 0.020 (0.020)\tLoss 0.5882 (0.6028)\tPrec@1 87.500 (87.003)\n",
      "[20/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.6167 (0.6255)\tPrec@1 85.938 (86.272)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5106 (0.6155)\tPrec@1 89.844 (86.719)\n",
      "[40/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6244 (0.6139)\tPrec@1 85.156 (86.681)\n",
      "[50/391]\tTime 0.145 (0.180)\tData 0.020 (0.020)\tLoss 0.5306 (0.6186)\tPrec@1 91.406 (86.504)\n",
      "[60/391]\tTime 0.149 (0.174)\tData 0.020 (0.020)\tLoss 0.6123 (0.6185)\tPrec@1 89.844 (86.463)\n",
      "[70/391]\tTime 0.182 (0.174)\tData 0.019 (0.020)\tLoss 0.8165 (0.6266)\tPrec@1 79.688 (86.114)\n",
      "[80/391]\tTime 0.183 (0.175)\tData 0.019 (0.020)\tLoss 0.6364 (0.6272)\tPrec@1 85.156 (86.256)\n",
      "[90/391]\tTime 0.182 (0.175)\tData 0.020 (0.020)\tLoss 0.5214 (0.6263)\tPrec@1 90.625 (86.367)\n",
      "[100/391]\tTime 0.180 (0.176)\tData 0.019 (0.020)\tLoss 0.4888 (0.6236)\tPrec@1 92.188 (86.448)\n",
      "[110/391]\tTime 0.180 (0.176)\tData 0.019 (0.020)\tLoss 0.6063 (0.6235)\tPrec@1 83.594 (86.430)\n",
      "[120/391]\tTime 0.182 (0.177)\tData 0.019 (0.020)\tLoss 0.7532 (0.6223)\tPrec@1 77.344 (86.370)\n",
      "[130/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.5314 (0.6229)\tPrec@1 86.719 (86.373)\n",
      "[140/391]\tTime 0.182 (0.177)\tData 0.019 (0.020)\tLoss 0.6214 (0.6223)\tPrec@1 85.938 (86.392)\n",
      "[150/391]\tTime 0.188 (0.178)\tData 0.021 (0.020)\tLoss 0.5330 (0.6203)\tPrec@1 89.844 (86.470)\n",
      "[160/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.6973 (0.6202)\tPrec@1 83.594 (86.442)\n",
      "[170/391]\tTime 0.181 (0.178)\tData 0.019 (0.020)\tLoss 0.6596 (0.6196)\tPrec@1 88.281 (86.477)\n",
      "[180/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.6422 (0.6206)\tPrec@1 85.938 (86.460)\n",
      "[190/391]\tTime 0.184 (0.178)\tData 0.019 (0.020)\tLoss 0.6689 (0.6229)\tPrec@1 85.156 (86.392)\n",
      "[200/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.7421 (0.6257)\tPrec@1 79.688 (86.225)\n",
      "[210/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.6846 (0.6289)\tPrec@1 85.938 (86.137)\n",
      "[220/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5940 (0.6272)\tPrec@1 87.500 (86.242)\n",
      "[230/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.6857 (0.6283)\tPrec@1 84.375 (86.174)\n",
      "[240/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.6398 (0.6276)\tPrec@1 81.250 (86.200)\n",
      "[250/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6050 (0.6283)\tPrec@1 86.719 (86.158)\n",
      "[260/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.6204 (0.6274)\tPrec@1 86.719 (86.198)\n",
      "[270/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6151 (0.6277)\tPrec@1 86.719 (86.217)\n",
      "[280/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6461 (0.6275)\tPrec@1 83.594 (86.232)\n",
      "[290/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.5284 (0.6269)\tPrec@1 90.625 (86.246)\n",
      "[300/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5435 (0.6251)\tPrec@1 89.844 (86.301)\n",
      "[310/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5972 (0.6242)\tPrec@1 85.156 (86.337)\n",
      "[320/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6966 (0.6244)\tPrec@1 85.938 (86.349)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5143 (0.6238)\tPrec@1 87.500 (86.351)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.4799 (0.6238)\tPrec@1 91.406 (86.350)\n",
      "[350/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.7164 (0.6249)\tPrec@1 81.250 (86.314)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6620 (0.6266)\tPrec@1 82.812 (86.243)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6312 (0.6267)\tPrec@1 84.375 (86.241)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6139 (0.6281)\tPrec@1 90.625 (86.208)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.5941 (0.6277)\tPrec@1 88.750 (86.200)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.3379 (1.3379)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 1.1686 (1.2474)\tPrec@1 78.125 (78.054)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.3980 (1.3073)\tPrec@1 76.562 (77.716)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.9519 (1.3261)\tPrec@1 81.250 (77.520)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 1.7443 (1.3009)\tPrec@1 74.219 (77.496)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.6344 (1.3084)\tPrec@1 72.656 (77.313)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.0746 (1.2921)\tPrec@1 82.031 (77.267)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.1829 (1.2967)\tPrec@1 79.688 (77.256)\n",
      " * Prec@1 77.470\n",
      "\n",
      "===> epoch: 3/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.146 (0.146)\tData 0.021 (0.021)\tLoss 0.6570 (0.6570)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.5440 (0.6210)\tPrec@1 89.844 (86.435)\n",
      "[20/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5603 (0.5951)\tPrec@1 93.750 (88.021)\n",
      "[30/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6262 (0.6071)\tPrec@1 87.500 (87.374)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4888 (0.6083)\tPrec@1 94.531 (87.424)\n",
      "[50/391]\tTime 0.146 (0.174)\tData 0.019 (0.020)\tLoss 0.6689 (0.6110)\tPrec@1 87.500 (87.255)\n",
      "[60/391]\tTime 0.180 (0.174)\tData 0.019 (0.020)\tLoss 0.6715 (0.6089)\tPrec@1 85.938 (87.333)\n",
      "[70/391]\tTime 0.185 (0.175)\tData 0.019 (0.020)\tLoss 0.5056 (0.6004)\tPrec@1 91.406 (87.643)\n",
      "[80/391]\tTime 0.178 (0.176)\tData 0.021 (0.020)\tLoss 0.6297 (0.6010)\tPrec@1 88.281 (87.568)\n",
      "[90/391]\tTime 0.179 (0.176)\tData 0.019 (0.020)\tLoss 0.6448 (0.6013)\tPrec@1 88.281 (87.509)\n",
      "[100/391]\tTime 0.179 (0.177)\tData 0.019 (0.020)\tLoss 0.6003 (0.6037)\tPrec@1 87.500 (87.368)\n",
      "[110/391]\tTime 0.179 (0.177)\tData 0.019 (0.020)\tLoss 0.5706 (0.6043)\tPrec@1 89.844 (87.437)\n",
      "[120/391]\tTime 0.189 (0.177)\tData 0.019 (0.020)\tLoss 0.5723 (0.6033)\tPrec@1 89.062 (87.539)\n",
      "[130/391]\tTime 0.184 (0.178)\tData 0.020 (0.020)\tLoss 0.7222 (0.6039)\tPrec@1 82.812 (87.452)\n",
      "[140/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6238 (0.6032)\tPrec@1 84.375 (87.406)\n",
      "[150/391]\tTime 0.190 (0.178)\tData 0.029 (0.020)\tLoss 0.6352 (0.6057)\tPrec@1 84.375 (87.252)\n",
      "[160/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.5986 (0.6070)\tPrec@1 89.062 (87.175)\n",
      "[170/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6269 (0.6075)\tPrec@1 85.938 (87.098)\n",
      "[180/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.5970 (0.6088)\tPrec@1 87.500 (87.073)\n",
      "[190/391]\tTime 0.186 (0.179)\tData 0.026 (0.020)\tLoss 0.5336 (0.6100)\tPrec@1 90.625 (87.066)\n",
      "[200/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6429 (0.6120)\tPrec@1 85.938 (86.995)\n",
      "[210/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.4931 (0.6133)\tPrec@1 92.969 (86.937)\n",
      "[220/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5604 (0.6138)\tPrec@1 89.062 (86.892)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.5384 (0.6141)\tPrec@1 90.625 (86.908)\n",
      "[240/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.6513 (0.6147)\tPrec@1 86.719 (86.916)\n",
      "[250/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.4909 (0.6135)\tPrec@1 92.969 (86.930)\n",
      "[260/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.6240 (0.6143)\tPrec@1 85.156 (86.844)\n",
      "[270/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6041 (0.6160)\tPrec@1 88.281 (86.817)\n",
      "[280/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5683 (0.6156)\tPrec@1 87.500 (86.822)\n",
      "[290/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.5928 (0.6148)\tPrec@1 87.500 (86.850)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6008 (0.6149)\tPrec@1 87.500 (86.804)\n",
      "[310/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6406 (0.6161)\tPrec@1 85.938 (86.741)\n",
      "[320/391]\tTime 0.186 (0.180)\tData 0.021 (0.020)\tLoss 0.6566 (0.6169)\tPrec@1 89.844 (86.724)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.5696 (0.6178)\tPrec@1 88.281 (86.681)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6750 (0.6180)\tPrec@1 83.594 (86.661)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6519 (0.6191)\tPrec@1 86.719 (86.632)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6612 (0.6195)\tPrec@1 86.719 (86.626)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.023 (0.020)\tLoss 0.5232 (0.6193)\tPrec@1 92.969 (86.656)\n",
      "[380/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5403 (0.6197)\tPrec@1 89.062 (86.624)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.012 (0.020)\tLoss 0.5709 (0.6205)\tPrec@1 88.750 (86.582)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.0889 (1.0889)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.023 (0.023)\tLoss 1.0977 (1.3698)\tPrec@1 77.344 (76.420)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 1.6438 (1.4536)\tPrec@1 71.094 (75.037)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.6985 (1.4225)\tPrec@1 82.812 (75.252)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.7377 (1.4006)\tPrec@1 71.875 (75.553)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 1.4811 (1.3904)\tPrec@1 71.875 (75.475)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.3027 (1.3755)\tPrec@1 81.250 (75.820)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.2239 (1.3804)\tPrec@1 74.219 (75.649)\n",
      " * Prec@1 75.810\n",
      "\n",
      "===> epoch: 4/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.152 (0.152)\tData 0.021 (0.021)\tLoss 0.5650 (0.5650)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.5672 (0.6088)\tPrec@1 87.500 (86.719)\n",
      "[20/391]\tTime 0.181 (0.179)\tData 0.021 (0.020)\tLoss 0.5595 (0.6022)\tPrec@1 87.500 (87.500)\n",
      "[30/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6296 (0.6186)\tPrec@1 87.500 (86.769)\n",
      "[40/391]\tTime 0.147 (0.172)\tData 0.019 (0.020)\tLoss 0.6553 (0.6334)\tPrec@1 89.062 (86.128)\n",
      "[50/391]\tTime 0.184 (0.171)\tData 0.020 (0.020)\tLoss 0.6181 (0.6322)\tPrec@1 86.719 (86.183)\n",
      "[60/391]\tTime 0.179 (0.173)\tData 0.019 (0.020)\tLoss 0.6570 (0.6349)\tPrec@1 86.719 (85.937)\n",
      "[70/391]\tTime 0.179 (0.174)\tData 0.019 (0.020)\tLoss 0.4526 (0.6312)\tPrec@1 94.531 (86.147)\n",
      "[80/391]\tTime 0.179 (0.175)\tData 0.019 (0.020)\tLoss 0.6160 (0.6304)\tPrec@1 85.938 (86.111)\n",
      "[90/391]\tTime 0.183 (0.176)\tData 0.029 (0.020)\tLoss 0.5196 (0.6258)\tPrec@1 89.844 (86.281)\n",
      "[100/391]\tTime 0.181 (0.176)\tData 0.020 (0.020)\tLoss 0.6714 (0.6233)\tPrec@1 83.594 (86.355)\n",
      "[110/391]\tTime 0.183 (0.176)\tData 0.019 (0.020)\tLoss 0.6967 (0.6225)\tPrec@1 83.594 (86.430)\n",
      "[120/391]\tTime 0.183 (0.177)\tData 0.019 (0.020)\tLoss 0.5421 (0.6212)\tPrec@1 89.062 (86.499)\n",
      "[130/391]\tTime 0.178 (0.177)\tData 0.019 (0.020)\tLoss 0.6263 (0.6181)\tPrec@1 85.938 (86.617)\n",
      "[140/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.6649 (0.6172)\tPrec@1 82.031 (86.658)\n",
      "[150/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5813 (0.6157)\tPrec@1 87.500 (86.672)\n",
      "[160/391]\tTime 0.181 (0.178)\tData 0.020 (0.020)\tLoss 0.5884 (0.6150)\tPrec@1 85.156 (86.758)\n",
      "[170/391]\tTime 0.185 (0.178)\tData 0.019 (0.020)\tLoss 0.6296 (0.6152)\tPrec@1 88.281 (86.728)\n",
      "[180/391]\tTime 0.182 (0.178)\tData 0.020 (0.020)\tLoss 0.6739 (0.6170)\tPrec@1 80.469 (86.671)\n",
      "[190/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.5983 (0.6185)\tPrec@1 84.375 (86.608)\n",
      "[200/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.8144 (0.6207)\tPrec@1 78.906 (86.536)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.7673 (0.6242)\tPrec@1 80.469 (86.423)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5370 (0.6246)\tPrec@1 86.719 (86.379)\n",
      "[230/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.6823 (0.6239)\tPrec@1 85.938 (86.408)\n",
      "[240/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.7224 (0.6237)\tPrec@1 81.250 (86.408)\n",
      "[250/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.5753 (0.6230)\tPrec@1 92.969 (86.448)\n",
      "[260/391]\tTime 0.185 (0.179)\tData 0.020 (0.020)\tLoss 0.6354 (0.6226)\tPrec@1 85.156 (86.479)\n",
      "[270/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.5868 (0.6228)\tPrec@1 88.281 (86.471)\n",
      "[280/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5574 (0.6232)\tPrec@1 90.625 (86.485)\n",
      "[290/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6330 (0.6249)\tPrec@1 83.594 (86.426)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5543 (0.6250)\tPrec@1 87.500 (86.415)\n",
      "[310/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5381 (0.6247)\tPrec@1 89.844 (86.445)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6053 (0.6235)\tPrec@1 89.062 (86.492)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6497 (0.6233)\tPrec@1 86.719 (86.511)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6301 (0.6228)\tPrec@1 87.500 (86.510)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6879 (0.6244)\tPrec@1 81.250 (86.436)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5562 (0.6244)\tPrec@1 89.062 (86.442)\n",
      "[370/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6626 (0.6237)\tPrec@1 84.375 (86.472)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[380/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.7517 (0.6238)\tPrec@1 81.250 (86.485)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.6471 (0.6240)\tPrec@1 83.750 (86.484)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9725 (0.9725)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.020 (0.024)\tLoss 1.0587 (0.9829)\tPrec@1 81.250 (82.812)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.8265 (0.9849)\tPrec@1 82.812 (82.143)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 1.0979 (0.9873)\tPrec@1 82.812 (82.082)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.2189 (0.9783)\tPrec@1 77.344 (82.127)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 1.0179 (0.9564)\tPrec@1 80.469 (82.307)\n",
      "Test: [60/79]\tTime 0.023 (0.023)\tLoss 1.0990 (0.9440)\tPrec@1 80.469 (82.364)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 0.7253 (0.9504)\tPrec@1 86.719 (82.284)\n",
      " * Prec@1 82.410\n",
      "\n",
      "===> epoch: 5/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.182 (0.182)\tData 0.021 (0.021)\tLoss 0.5692 (0.5692)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5540 (0.6105)\tPrec@1 90.625 (86.861)\n",
      "[20/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6901 (0.6411)\tPrec@1 82.812 (85.789)\n",
      "[30/391]\tTime 0.144 (0.171)\tData 0.020 (0.020)\tLoss 0.5509 (0.6390)\tPrec@1 92.969 (86.064)\n",
      "[40/391]\tTime 0.185 (0.169)\tData 0.019 (0.020)\tLoss 0.5381 (0.6288)\tPrec@1 86.719 (86.452)\n",
      "[50/391]\tTime 0.182 (0.171)\tData 0.019 (0.020)\tLoss 0.5244 (0.6254)\tPrec@1 88.281 (86.428)\n",
      "[60/391]\tTime 0.193 (0.174)\tData 0.020 (0.020)\tLoss 0.5161 (0.6242)\tPrec@1 92.188 (86.475)\n",
      "[70/391]\tTime 0.187 (0.176)\tData 0.021 (0.020)\tLoss 0.7031 (0.6232)\tPrec@1 82.031 (86.521)\n",
      "[80/391]\tTime 0.178 (0.176)\tData 0.019 (0.020)\tLoss 0.6335 (0.6215)\tPrec@1 85.938 (86.632)\n",
      "[90/391]\tTime 0.181 (0.177)\tData 0.019 (0.020)\tLoss 0.6089 (0.6225)\tPrec@1 87.500 (86.590)\n",
      "[100/391]\tTime 0.179 (0.177)\tData 0.021 (0.020)\tLoss 0.5688 (0.6211)\tPrec@1 90.625 (86.672)\n",
      "[110/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.5837 (0.6177)\tPrec@1 88.281 (86.796)\n",
      "[120/391]\tTime 0.182 (0.178)\tData 0.021 (0.020)\tLoss 0.6290 (0.6180)\tPrec@1 85.938 (86.796)\n",
      "[130/391]\tTime 0.184 (0.178)\tData 0.020 (0.020)\tLoss 0.6462 (0.6171)\tPrec@1 82.031 (86.808)\n",
      "[140/391]\tTime 0.180 (0.178)\tData 0.021 (0.020)\tLoss 0.5340 (0.6145)\tPrec@1 91.406 (86.885)\n",
      "[150/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.5326 (0.6158)\tPrec@1 87.500 (86.802)\n",
      "[160/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5807 (0.6132)\tPrec@1 90.625 (86.932)\n",
      "[170/391]\tTime 0.182 (0.178)\tData 0.020 (0.020)\tLoss 0.5580 (0.6121)\tPrec@1 87.500 (86.984)\n",
      "[180/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5654 (0.6128)\tPrec@1 89.062 (86.978)\n",
      "[190/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.5690 (0.6117)\tPrec@1 89.062 (86.972)\n",
      "[200/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5652 (0.6124)\tPrec@1 89.844 (86.960)\n",
      "[210/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5820 (0.6140)\tPrec@1 86.719 (86.826)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.5695 (0.6163)\tPrec@1 89.062 (86.743)\n",
      "[230/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.7903 (0.6173)\tPrec@1 81.250 (86.736)\n",
      "[240/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.5984 (0.6180)\tPrec@1 87.500 (86.738)\n",
      "[250/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.7364 (0.6182)\tPrec@1 82.812 (86.709)\n",
      "[260/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5413 (0.6201)\tPrec@1 88.281 (86.626)\n",
      "[270/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7287 (0.6200)\tPrec@1 82.031 (86.635)\n",
      "[280/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7037 (0.6213)\tPrec@1 85.156 (86.613)\n",
      "[290/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6865 (0.6204)\tPrec@1 83.594 (86.649)\n",
      "[300/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6494 (0.6210)\tPrec@1 84.375 (86.664)\n",
      "[310/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6243 (0.6200)\tPrec@1 85.938 (86.716)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7339 (0.6196)\tPrec@1 80.469 (86.719)\n",
      "[330/391]\tTime 0.183 (0.180)\tData 0.021 (0.020)\tLoss 0.6744 (0.6204)\tPrec@1 85.156 (86.688)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5443 (0.6195)\tPrec@1 89.844 (86.744)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6534 (0.6208)\tPrec@1 85.156 (86.699)\n",
      "[360/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.7423 (0.6219)\tPrec@1 81.250 (86.647)\n",
      "[370/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6029 (0.6222)\tPrec@1 86.719 (86.626)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5630 (0.6225)\tPrec@1 88.281 (86.608)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.012 (0.020)\tLoss 0.4772 (0.6227)\tPrec@1 90.000 (86.596)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.2900 (1.2900)\tPrec@1 73.438 (73.438)\n",
      "Test: [10/79]\tTime 0.023 (0.022)\tLoss 1.4993 (1.2900)\tPrec@1 75.781 (76.491)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 1.1556 (1.3451)\tPrec@1 79.688 (76.265)\n",
      "Test: [30/79]\tTime 0.022 (0.022)\tLoss 0.9849 (1.3300)\tPrec@1 82.812 (76.714)\n",
      "Test: [40/79]\tTime 0.026 (0.022)\tLoss 1.7632 (1.3244)\tPrec@1 73.438 (76.848)\n",
      "Test: [50/79]\tTime 0.026 (0.022)\tLoss 1.3314 (1.3098)\tPrec@1 76.562 (77.068)\n",
      "Test: [60/79]\tTime 0.019 (0.022)\tLoss 1.0195 (1.3246)\tPrec@1 79.688 (76.895)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.1524 (1.3193)\tPrec@1 82.812 (76.948)\n",
      " * Prec@1 77.030\n",
      "\n",
      "===> epoch: 6/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.174 (0.174)\tData 0.020 (0.020)\tLoss 0.5836 (0.5836)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6064 (0.6221)\tPrec@1 89.844 (87.287)\n",
      "[20/391]\tTime 0.144 (0.169)\tData 0.020 (0.020)\tLoss 0.5964 (0.6265)\tPrec@1 88.281 (86.942)\n",
      "[30/391]\tTime 0.181 (0.165)\tData 0.020 (0.020)\tLoss 0.6132 (0.6299)\tPrec@1 86.719 (86.794)\n",
      "[40/391]\tTime 0.179 (0.169)\tData 0.019 (0.020)\tLoss 0.7720 (0.6259)\tPrec@1 74.219 (86.681)\n",
      "[50/391]\tTime 0.178 (0.172)\tData 0.019 (0.020)\tLoss 0.5809 (0.6209)\tPrec@1 88.281 (86.642)\n",
      "[60/391]\tTime 0.187 (0.173)\tData 0.020 (0.020)\tLoss 0.6991 (0.6240)\tPrec@1 82.031 (86.514)\n",
      "[70/391]\tTime 0.189 (0.175)\tData 0.020 (0.020)\tLoss 0.6083 (0.6249)\tPrec@1 86.719 (86.334)\n",
      "[80/391]\tTime 0.183 (0.176)\tData 0.019 (0.020)\tLoss 0.6663 (0.6222)\tPrec@1 82.812 (86.314)\n",
      "[90/391]\tTime 0.181 (0.176)\tData 0.020 (0.020)\tLoss 0.6802 (0.6230)\tPrec@1 83.594 (86.255)\n",
      "[100/391]\tTime 0.184 (0.177)\tData 0.020 (0.020)\tLoss 0.6128 (0.6238)\tPrec@1 85.156 (86.123)\n",
      "[110/391]\tTime 0.181 (0.177)\tData 0.020 (0.020)\tLoss 0.6468 (0.6262)\tPrec@1 86.719 (86.043)\n",
      "[120/391]\tTime 0.180 (0.177)\tData 0.020 (0.020)\tLoss 0.6430 (0.6300)\tPrec@1 85.938 (85.963)\n",
      "[130/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.7351 (0.6280)\tPrec@1 79.688 (86.003)\n",
      "[140/391]\tTime 0.187 (0.178)\tData 0.019 (0.020)\tLoss 0.6730 (0.6266)\tPrec@1 87.500 (86.154)\n",
      "[150/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.5213 (0.6231)\tPrec@1 92.188 (86.346)\n",
      "[160/391]\tTime 0.189 (0.179)\tData 0.020 (0.020)\tLoss 0.5959 (0.6228)\tPrec@1 87.500 (86.369)\n",
      "[170/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.4960 (0.6214)\tPrec@1 90.625 (86.413)\n",
      "[180/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.7019 (0.6223)\tPrec@1 85.156 (86.391)\n",
      "[190/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.7344 (0.6241)\tPrec@1 86.719 (86.306)\n",
      "[200/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.8050 (0.6256)\tPrec@1 76.562 (86.225)\n",
      "[210/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.8244 (0.6255)\tPrec@1 84.375 (86.274)\n",
      "[220/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6324 (0.6255)\tPrec@1 85.938 (86.242)\n",
      "[230/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6385 (0.6251)\tPrec@1 83.594 (86.222)\n",
      "[240/391]\tTime 0.188 (0.180)\tData 0.020 (0.020)\tLoss 0.6969 (0.6260)\tPrec@1 82.031 (86.197)\n",
      "[250/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.5446 (0.6270)\tPrec@1 92.188 (86.162)\n",
      "[260/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6164 (0.6264)\tPrec@1 86.719 (86.171)\n",
      "[270/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6277 (0.6252)\tPrec@1 85.938 (86.266)\n",
      "[280/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5857 (0.6248)\tPrec@1 89.062 (86.271)\n",
      "[290/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6301 (0.6250)\tPrec@1 87.500 (86.287)\n",
      "[300/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.6120 (0.6246)\tPrec@1 88.281 (86.306)\n",
      "[310/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.5918 (0.6260)\tPrec@1 88.281 (86.264)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6540 (0.6263)\tPrec@1 85.938 (86.264)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[330/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6317 (0.6269)\tPrec@1 86.719 (86.270)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.4889 (0.6259)\tPrec@1 92.188 (86.352)\n",
      "[350/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.6873 (0.6259)\tPrec@1 83.594 (86.336)\n",
      "[360/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.7402 (0.6271)\tPrec@1 80.469 (86.299)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5286 (0.6272)\tPrec@1 89.062 (86.281)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6298 (0.6276)\tPrec@1 85.938 (86.274)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.6129 (0.6274)\tPrec@1 86.250 (86.264)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7162 (0.7162)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.0453 (1.1123)\tPrec@1 79.688 (78.977)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0251 (1.1824)\tPrec@1 79.688 (78.274)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.8229 (1.1969)\tPrec@1 82.031 (78.150)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.4404 (1.1916)\tPrec@1 75.781 (77.839)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 1.2297 (1.1635)\tPrec@1 73.438 (77.987)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.0795 (1.1419)\tPrec@1 79.688 (78.266)\n",
      "Test: [70/79]\tTime 0.021 (0.023)\tLoss 1.2753 (1.1407)\tPrec@1 77.344 (78.246)\n",
      " * Prec@1 78.330\n",
      "\n",
      "===> epoch: 7/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6975 (0.6975)\tPrec@1 85.156 (85.156)\n",
      "[10/391]\tTime 0.144 (0.165)\tData 0.020 (0.020)\tLoss 0.6504 (0.6158)\tPrec@1 85.156 (87.429)\n",
      "[20/391]\tTime 0.185 (0.157)\tData 0.020 (0.020)\tLoss 0.6404 (0.6166)\tPrec@1 84.375 (86.942)\n",
      "[30/391]\tTime 0.190 (0.166)\tData 0.020 (0.020)\tLoss 0.6649 (0.6207)\tPrec@1 88.281 (86.719)\n",
      "[40/391]\tTime 0.185 (0.170)\tData 0.019 (0.020)\tLoss 0.5246 (0.6205)\tPrec@1 85.938 (86.604)\n",
      "[50/391]\tTime 0.178 (0.172)\tData 0.019 (0.020)\tLoss 0.5954 (0.6166)\tPrec@1 88.281 (86.857)\n",
      "[60/391]\tTime 0.181 (0.174)\tData 0.019 (0.020)\tLoss 0.6924 (0.6211)\tPrec@1 85.156 (86.565)\n",
      "[70/391]\tTime 0.178 (0.175)\tData 0.019 (0.020)\tLoss 0.6759 (0.6200)\tPrec@1 83.594 (86.697)\n",
      "[80/391]\tTime 0.180 (0.175)\tData 0.020 (0.020)\tLoss 0.6976 (0.6196)\tPrec@1 86.719 (86.699)\n",
      "[90/391]\tTime 0.181 (0.176)\tData 0.019 (0.020)\tLoss 0.5748 (0.6186)\tPrec@1 88.281 (86.753)\n",
      "[100/391]\tTime 0.179 (0.177)\tData 0.019 (0.020)\tLoss 0.5667 (0.6146)\tPrec@1 85.938 (86.835)\n",
      "[110/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5544 (0.6129)\tPrec@1 88.281 (86.916)\n",
      "[120/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6420 (0.6135)\tPrec@1 87.500 (86.912)\n",
      "[130/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.6552 (0.6118)\tPrec@1 84.375 (87.071)\n",
      "[140/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.7154 (0.6110)\tPrec@1 80.469 (87.035)\n",
      "[150/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.8392 (0.6129)\tPrec@1 75.781 (86.874)\n",
      "[160/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.6160 (0.6140)\tPrec@1 86.719 (86.840)\n",
      "[170/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.6372 (0.6151)\tPrec@1 83.594 (86.819)\n",
      "[180/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6721 (0.6179)\tPrec@1 83.594 (86.714)\n",
      "[190/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7533 (0.6184)\tPrec@1 85.156 (86.735)\n",
      "[200/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6161 (0.6180)\tPrec@1 86.719 (86.758)\n",
      "[210/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5916 (0.6186)\tPrec@1 87.500 (86.741)\n",
      "[220/391]\tTime 0.190 (0.180)\tData 0.019 (0.020)\tLoss 0.5762 (0.6202)\tPrec@1 87.500 (86.655)\n",
      "[230/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5710 (0.6197)\tPrec@1 89.062 (86.682)\n",
      "[240/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6272 (0.6189)\tPrec@1 85.938 (86.699)\n",
      "[250/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5982 (0.6194)\tPrec@1 86.719 (86.688)\n",
      "[260/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6989 (0.6193)\tPrec@1 83.594 (86.665)\n",
      "[270/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7956 (0.6208)\tPrec@1 83.594 (86.652)\n",
      "[280/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6018 (0.6212)\tPrec@1 85.938 (86.666)\n",
      "[290/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.4637 (0.6220)\tPrec@1 92.969 (86.638)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5841 (0.6220)\tPrec@1 89.062 (86.633)\n",
      "[310/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6191 (0.6213)\tPrec@1 87.500 (86.658)\n",
      "[320/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5319 (0.6208)\tPrec@1 89.844 (86.694)\n",
      "[330/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.7695 (0.6216)\tPrec@1 82.031 (86.662)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.022 (0.020)\tLoss 0.6061 (0.6219)\tPrec@1 85.938 (86.639)\n",
      "[350/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.7042 (0.6224)\tPrec@1 79.688 (86.581)\n",
      "[360/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.5684 (0.6212)\tPrec@1 89.062 (86.611)\n",
      "[370/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5165 (0.6209)\tPrec@1 89.062 (86.601)\n",
      "[380/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6612 (0.6217)\tPrec@1 86.719 (86.553)\n",
      "[390/391]\tTime 0.127 (0.181)\tData 0.012 (0.020)\tLoss 0.7610 (0.6218)\tPrec@1 78.750 (86.542)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9545 (0.9545)\tPrec@1 81.250 (81.250)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 1.0703 (0.9632)\tPrec@1 79.688 (82.244)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 0.8172 (0.9672)\tPrec@1 87.500 (82.143)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.5897 (0.9707)\tPrec@1 87.500 (82.560)\n",
      "Test: [40/79]\tTime 0.022 (0.023)\tLoss 0.9382 (0.9682)\tPrec@1 80.469 (82.317)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.1217 (0.9760)\tPrec@1 80.469 (82.322)\n",
      "Test: [60/79]\tTime 0.019 (0.022)\tLoss 0.8533 (0.9677)\tPrec@1 83.594 (82.390)\n",
      "Test: [70/79]\tTime 0.020 (0.022)\tLoss 0.6683 (0.9645)\tPrec@1 89.844 (82.405)\n",
      " * Prec@1 82.450\n",
      "\n",
      "===> epoch: 8/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.143 (0.143)\tData 0.020 (0.020)\tLoss 0.6728 (0.6728)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.144 (0.144)\tData 0.019 (0.020)\tLoss 0.7884 (0.6611)\tPrec@1 82.031 (85.298)\n",
      "[20/391]\tTime 0.181 (0.161)\tData 0.019 (0.019)\tLoss 0.6766 (0.6540)\tPrec@1 85.156 (85.677)\n",
      "[30/391]\tTime 0.180 (0.168)\tData 0.019 (0.019)\tLoss 0.6546 (0.6409)\tPrec@1 86.719 (86.366)\n",
      "[40/391]\tTime 0.187 (0.171)\tData 0.022 (0.019)\tLoss 0.6798 (0.6416)\tPrec@1 85.156 (86.300)\n",
      "[50/391]\tTime 0.180 (0.173)\tData 0.019 (0.019)\tLoss 0.7044 (0.6424)\tPrec@1 86.719 (86.275)\n",
      "[60/391]\tTime 0.179 (0.175)\tData 0.019 (0.019)\tLoss 0.7087 (0.6366)\tPrec@1 85.156 (86.437)\n",
      "[70/391]\tTime 0.178 (0.175)\tData 0.019 (0.019)\tLoss 0.6283 (0.6338)\tPrec@1 87.500 (86.532)\n",
      "[80/391]\tTime 0.180 (0.176)\tData 0.019 (0.019)\tLoss 0.7786 (0.6341)\tPrec@1 78.906 (86.275)\n",
      "[90/391]\tTime 0.184 (0.177)\tData 0.019 (0.019)\tLoss 0.7775 (0.6376)\tPrec@1 78.906 (86.195)\n",
      "[100/391]\tTime 0.180 (0.177)\tData 0.020 (0.019)\tLoss 0.5376 (0.6360)\tPrec@1 88.281 (86.200)\n",
      "[110/391]\tTime 0.181 (0.178)\tData 0.019 (0.019)\tLoss 0.6140 (0.6360)\tPrec@1 83.594 (86.120)\n",
      "[120/391]\tTime 0.180 (0.178)\tData 0.019 (0.019)\tLoss 0.5618 (0.6341)\tPrec@1 91.406 (86.260)\n",
      "[130/391]\tTime 0.180 (0.178)\tData 0.019 (0.019)\tLoss 0.6622 (0.6341)\tPrec@1 85.156 (86.277)\n",
      "[140/391]\tTime 0.180 (0.178)\tData 0.019 (0.019)\tLoss 0.5875 (0.6363)\tPrec@1 89.844 (86.176)\n",
      "[150/391]\tTime 0.188 (0.179)\tData 0.028 (0.019)\tLoss 0.5994 (0.6344)\tPrec@1 87.500 (86.238)\n",
      "[160/391]\tTime 0.179 (0.179)\tData 0.019 (0.019)\tLoss 0.5555 (0.6339)\tPrec@1 89.062 (86.277)\n",
      "[170/391]\tTime 0.186 (0.179)\tData 0.018 (0.019)\tLoss 0.7420 (0.6355)\tPrec@1 84.375 (86.189)\n",
      "[180/391]\tTime 0.184 (0.179)\tData 0.019 (0.019)\tLoss 0.5741 (0.6376)\tPrec@1 84.375 (86.080)\n",
      "[190/391]\tTime 0.179 (0.179)\tData 0.019 (0.019)\tLoss 0.6909 (0.6364)\tPrec@1 83.594 (86.085)\n",
      "[200/391]\tTime 0.183 (0.180)\tData 0.019 (0.019)\tLoss 0.5250 (0.6369)\tPrec@1 89.844 (86.042)\n",
      "[210/391]\tTime 0.184 (0.180)\tData 0.020 (0.019)\tLoss 0.6090 (0.6361)\tPrec@1 86.719 (86.075)\n",
      "[220/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.7232 (0.6345)\tPrec@1 84.375 (86.188)\n",
      "[230/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6084 (0.6347)\tPrec@1 84.375 (86.161)\n",
      "[240/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.6476 (0.6352)\tPrec@1 89.062 (86.158)\n",
      "[250/391]\tTime 0.178 (0.180)\tData 0.019 (0.019)\tLoss 0.6287 (0.6337)\tPrec@1 84.375 (86.168)\n",
      "[260/391]\tTime 0.181 (0.180)\tData 0.021 (0.019)\tLoss 0.5860 (0.6321)\tPrec@1 87.500 (86.225)\n",
      "[270/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.5796 (0.6310)\tPrec@1 86.719 (86.252)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6345 (0.6307)\tPrec@1 82.031 (86.257)\n",
      "[290/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.5531 (0.6296)\tPrec@1 90.625 (86.270)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.7377 (0.6288)\tPrec@1 80.469 (86.296)\n",
      "[310/391]\tTime 0.185 (0.180)\tData 0.020 (0.019)\tLoss 0.5873 (0.6283)\tPrec@1 88.281 (86.302)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.6097 (0.6277)\tPrec@1 88.281 (86.329)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.6431 (0.6277)\tPrec@1 84.375 (86.329)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.6244 (0.6285)\tPrec@1 86.719 (86.311)\n",
      "[350/391]\tTime 0.178 (0.180)\tData 0.019 (0.019)\tLoss 0.6493 (0.6291)\tPrec@1 86.719 (86.282)\n",
      "[360/391]\tTime 0.178 (0.180)\tData 0.019 (0.019)\tLoss 0.7262 (0.6285)\tPrec@1 81.250 (86.299)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5619 (0.6274)\tPrec@1 89.844 (86.342)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5833 (0.6267)\tPrec@1 86.719 (86.364)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.012 (0.019)\tLoss 0.7609 (0.6274)\tPrec@1 76.250 (86.332)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7114 (0.7114)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.022 (0.022)\tLoss 0.7045 (0.8880)\tPrec@1 85.156 (82.599)\n",
      "Test: [20/79]\tTime 0.028 (0.023)\tLoss 0.8300 (0.9502)\tPrec@1 79.688 (81.882)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.9206 (0.9408)\tPrec@1 84.375 (82.006)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 0.8394 (0.9494)\tPrec@1 81.250 (81.803)\n",
      "Test: [50/79]\tTime 0.019 (0.022)\tLoss 0.9201 (0.9449)\tPrec@1 80.469 (81.863)\n",
      "Test: [60/79]\tTime 0.020 (0.021)\tLoss 0.8960 (0.9400)\tPrec@1 89.062 (81.993)\n",
      "Test: [70/79]\tTime 0.020 (0.021)\tLoss 0.8750 (0.9374)\tPrec@1 84.375 (82.064)\n",
      " * Prec@1 82.340\n",
      "\n",
      "===> epoch: 9/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.145 (0.145)\tData 0.021 (0.021)\tLoss 0.6272 (0.6272)\tPrec@1 83.594 (83.594)\n",
      "[10/391]\tTime 0.182 (0.155)\tData 0.020 (0.020)\tLoss 0.6661 (0.6325)\tPrec@1 86.719 (85.511)\n",
      "[20/391]\tTime 0.185 (0.168)\tData 0.020 (0.020)\tLoss 0.6925 (0.6333)\tPrec@1 85.938 (85.826)\n",
      "[30/391]\tTime 0.180 (0.172)\tData 0.020 (0.020)\tLoss 0.6263 (0.6280)\tPrec@1 85.156 (86.038)\n",
      "[40/391]\tTime 0.180 (0.174)\tData 0.019 (0.020)\tLoss 0.5942 (0.6214)\tPrec@1 85.156 (86.319)\n",
      "[50/391]\tTime 0.182 (0.176)\tData 0.022 (0.019)\tLoss 0.6311 (0.6147)\tPrec@1 87.500 (86.642)\n",
      "[60/391]\tTime 0.178 (0.176)\tData 0.019 (0.019)\tLoss 0.5588 (0.6161)\tPrec@1 91.406 (86.693)\n",
      "[70/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.5768 (0.6160)\tPrec@1 89.062 (86.697)\n",
      "[80/391]\tTime 0.185 (0.178)\tData 0.019 (0.020)\tLoss 0.5984 (0.6124)\tPrec@1 86.719 (86.815)\n",
      "[90/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6611 (0.6142)\tPrec@1 88.281 (86.813)\n",
      "[100/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.6426 (0.6132)\tPrec@1 86.719 (86.835)\n",
      "[110/391]\tTime 0.179 (0.179)\tData 0.021 (0.019)\tLoss 0.7771 (0.6165)\tPrec@1 81.250 (86.677)\n",
      "[120/391]\tTime 0.179 (0.179)\tData 0.019 (0.019)\tLoss 0.4899 (0.6176)\tPrec@1 91.406 (86.635)\n",
      "[130/391]\tTime 0.183 (0.179)\tData 0.019 (0.019)\tLoss 0.7542 (0.6178)\tPrec@1 79.688 (86.653)\n",
      "[140/391]\tTime 0.191 (0.179)\tData 0.019 (0.019)\tLoss 0.5410 (0.6182)\tPrec@1 87.500 (86.636)\n",
      "[150/391]\tTime 0.179 (0.179)\tData 0.019 (0.019)\tLoss 0.5651 (0.6181)\tPrec@1 90.625 (86.651)\n",
      "[160/391]\tTime 0.187 (0.180)\tData 0.020 (0.019)\tLoss 0.5500 (0.6171)\tPrec@1 88.281 (86.670)\n",
      "[170/391]\tTime 0.184 (0.180)\tData 0.019 (0.019)\tLoss 0.6217 (0.6160)\tPrec@1 87.500 (86.691)\n",
      "[180/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.5946 (0.6154)\tPrec@1 88.281 (86.689)\n",
      "[190/391]\tTime 0.181 (0.180)\tData 0.020 (0.019)\tLoss 0.6325 (0.6142)\tPrec@1 84.375 (86.690)\n",
      "[200/391]\tTime 0.179 (0.180)\tData 0.020 (0.019)\tLoss 0.5793 (0.6137)\tPrec@1 89.844 (86.684)\n",
      "[210/391]\tTime 0.179 (0.180)\tData 0.020 (0.019)\tLoss 0.6724 (0.6144)\tPrec@1 85.156 (86.660)\n",
      "[220/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.6289 (0.6153)\tPrec@1 83.594 (86.620)\n",
      "[230/391]\tTime 0.180 (0.180)\tData 0.020 (0.019)\tLoss 0.5543 (0.6169)\tPrec@1 89.844 (86.597)\n",
      "[240/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.5498 (0.6169)\tPrec@1 86.719 (86.586)\n",
      "[250/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6627 (0.6181)\tPrec@1 85.156 (86.563)\n",
      "[260/391]\tTime 0.181 (0.180)\tData 0.020 (0.019)\tLoss 0.7587 (0.6177)\tPrec@1 80.469 (86.581)\n",
      "[270/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.6346 (0.6175)\tPrec@1 87.500 (86.592)\n",
      "[280/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5617 (0.6165)\tPrec@1 90.625 (86.635)\n",
      "[290/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5225 (0.6167)\tPrec@1 91.406 (86.644)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5212 (0.6169)\tPrec@1 92.188 (86.628)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6073 (0.6174)\tPrec@1 87.500 (86.603)\n",
      "[320/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6191 (0.6181)\tPrec@1 84.375 (86.553)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5877 (0.6169)\tPrec@1 88.281 (86.608)\n",
      "[340/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.5337 (0.6171)\tPrec@1 90.625 (86.597)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5740 (0.6170)\tPrec@1 88.281 (86.594)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5085 (0.6160)\tPrec@1 94.531 (86.619)\n",
      "[370/391]\tTime 0.183 (0.180)\tData 0.021 (0.020)\tLoss 0.5945 (0.6157)\tPrec@1 88.281 (86.628)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5878 (0.6175)\tPrec@1 89.062 (86.547)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.6398 (0.6186)\tPrec@1 87.500 (86.528)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.2656 (1.2656)\tPrec@1 75.781 (75.781)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 1.1134 (1.1539)\tPrec@1 80.469 (77.699)\n",
      "Test: [20/79]\tTime 0.026 (0.024)\tLoss 1.2033 (1.1360)\tPrec@1 75.000 (78.311)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.8988 (1.1500)\tPrec@1 75.781 (77.671)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 1.7132 (1.1633)\tPrec@1 71.094 (78.258)\n",
      "Test: [50/79]\tTime 0.019 (0.021)\tLoss 1.2150 (1.1547)\tPrec@1 78.906 (78.324)\n",
      "Test: [60/79]\tTime 0.021 (0.021)\tLoss 0.8585 (1.1391)\tPrec@1 85.156 (78.381)\n",
      "Test: [70/79]\tTime 0.019 (0.021)\tLoss 1.0461 (1.1370)\tPrec@1 78.125 (78.499)\n",
      " * Prec@1 78.750\n",
      "\n",
      "===> epoch: 10/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.145 (0.145)\tData 0.020 (0.020)\tLoss 0.5916 (0.5916)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.179 (0.166)\tData 0.021 (0.020)\tLoss 0.5607 (0.5879)\tPrec@1 88.281 (87.997)\n",
      "[20/391]\tTime 0.179 (0.172)\tData 0.020 (0.020)\tLoss 0.5912 (0.5963)\tPrec@1 86.719 (87.909)\n",
      "[30/391]\tTime 0.179 (0.175)\tData 0.019 (0.020)\tLoss 0.6845 (0.6113)\tPrec@1 85.156 (87.273)\n",
      "[40/391]\tTime 0.184 (0.177)\tData 0.019 (0.020)\tLoss 0.5832 (0.6032)\tPrec@1 89.844 (87.595)\n",
      "[50/391]\tTime 0.179 (0.178)\tData 0.020 (0.020)\tLoss 0.6149 (0.6061)\tPrec@1 88.281 (87.439)\n",
      "[60/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6124 (0.6125)\tPrec@1 86.719 (87.154)\n",
      "[70/391]\tTime 0.184 (0.178)\tData 0.020 (0.020)\tLoss 0.5548 (0.6086)\tPrec@1 86.719 (87.181)\n",
      "[80/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6170 (0.6120)\tPrec@1 89.062 (87.008)\n",
      "[90/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.6010 (0.6154)\tPrec@1 87.500 (86.959)\n",
      "[100/391]\tTime 0.187 (0.179)\tData 0.020 (0.020)\tLoss 0.6754 (0.6170)\tPrec@1 85.156 (86.889)\n",
      "[110/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.5185 (0.6148)\tPrec@1 90.625 (87.007)\n",
      "[120/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7063 (0.6186)\tPrec@1 85.156 (86.893)\n",
      "[130/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6058 (0.6211)\tPrec@1 91.406 (86.808)\n",
      "[140/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5973 (0.6224)\tPrec@1 87.500 (86.746)\n",
      "[150/391]\tTime 0.177 (0.180)\tData 0.019 (0.020)\tLoss 0.6368 (0.6224)\tPrec@1 86.719 (86.667)\n",
      "[160/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6607 (0.6207)\tPrec@1 87.500 (86.733)\n",
      "[170/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6893 (0.6191)\tPrec@1 84.375 (86.801)\n",
      "[180/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6094 (0.6177)\tPrec@1 88.281 (86.844)\n",
      "[190/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7154 (0.6197)\tPrec@1 81.250 (86.768)\n",
      "[200/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7858 (0.6210)\tPrec@1 81.250 (86.699)\n",
      "[210/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6298 (0.6201)\tPrec@1 88.281 (86.745)\n",
      "[220/391]\tTime 0.180 (0.180)\tData 0.021 (0.020)\tLoss 0.6556 (0.6212)\tPrec@1 86.719 (86.722)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[230/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6061 (0.6213)\tPrec@1 89.062 (86.732)\n",
      "[240/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5530 (0.6211)\tPrec@1 89.062 (86.735)\n",
      "[250/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5702 (0.6199)\tPrec@1 89.844 (86.784)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5762 (0.6192)\tPrec@1 89.062 (86.812)\n",
      "[270/391]\tTime 0.191 (0.181)\tData 0.029 (0.020)\tLoss 0.6627 (0.6216)\tPrec@1 85.938 (86.704)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6307 (0.6219)\tPrec@1 85.938 (86.691)\n",
      "[290/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5463 (0.6229)\tPrec@1 88.281 (86.619)\n",
      "[300/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.7429 (0.6240)\tPrec@1 82.812 (86.579)\n",
      "[310/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5882 (0.6251)\tPrec@1 89.844 (86.518)\n",
      "[320/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6315 (0.6259)\tPrec@1 86.719 (86.470)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7414 (0.6261)\tPrec@1 82.031 (86.485)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5437 (0.6256)\tPrec@1 87.500 (86.517)\n",
      "[350/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7370 (0.6260)\tPrec@1 82.031 (86.489)\n",
      "[360/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5631 (0.6268)\tPrec@1 87.500 (86.450)\n",
      "[370/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5258 (0.6260)\tPrec@1 92.188 (86.479)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.5689 (0.6253)\tPrec@1 86.719 (86.510)\n",
      "[390/391]\tTime 0.128 (0.181)\tData 0.013 (0.020)\tLoss 0.8430 (0.6252)\tPrec@1 81.250 (86.550)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.1265 (1.1265)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.021 (0.024)\tLoss 1.5116 (1.4262)\tPrec@1 75.781 (76.776)\n",
      "Test: [20/79]\tTime 0.019 (0.022)\tLoss 1.3165 (1.4902)\tPrec@1 80.469 (75.967)\n",
      "Test: [30/79]\tTime 0.020 (0.021)\tLoss 0.8883 (1.4522)\tPrec@1 82.031 (76.058)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.3611 (1.4395)\tPrec@1 78.125 (76.048)\n",
      "Test: [50/79]\tTime 0.020 (0.021)\tLoss 1.4573 (1.4240)\tPrec@1 71.094 (76.180)\n",
      "Test: [60/79]\tTime 0.020 (0.021)\tLoss 1.1917 (1.4412)\tPrec@1 78.125 (75.922)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.2747 (1.4528)\tPrec@1 80.469 (75.693)\n",
      " * Prec@1 75.760\n",
      "\n",
      "===> epoch: 11/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.146 (0.146)\tData 0.021 (0.021)\tLoss 0.6107 (0.6107)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.182 (0.176)\tData 0.020 (0.020)\tLoss 0.5667 (0.6068)\tPrec@1 88.281 (87.926)\n",
      "[20/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.6572 (0.6095)\tPrec@1 85.156 (87.463)\n",
      "[30/391]\tTime 0.181 (0.179)\tData 0.019 (0.019)\tLoss 0.6279 (0.6043)\tPrec@1 87.500 (87.475)\n",
      "[40/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6652 (0.5985)\tPrec@1 85.938 (87.652)\n",
      "[50/391]\tTime 0.184 (0.180)\tData 0.019 (0.019)\tLoss 0.6481 (0.6049)\tPrec@1 84.375 (87.531)\n",
      "[60/391]\tTime 0.187 (0.181)\tData 0.019 (0.019)\tLoss 0.6438 (0.6112)\tPrec@1 86.719 (87.321)\n",
      "[70/391]\tTime 0.186 (0.181)\tData 0.019 (0.019)\tLoss 0.6552 (0.6131)\tPrec@1 85.156 (87.258)\n",
      "[80/391]\tTime 0.184 (0.181)\tData 0.020 (0.019)\tLoss 0.5613 (0.6170)\tPrec@1 85.156 (86.950)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.020 (0.019)\tLoss 0.5281 (0.6167)\tPrec@1 89.062 (86.993)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.6163 (0.6167)\tPrec@1 88.281 (86.912)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.5835 (0.6157)\tPrec@1 85.938 (86.965)\n",
      "[120/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.7234 (0.6172)\tPrec@1 84.375 (86.900)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6972 (0.6216)\tPrec@1 83.594 (86.689)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6587 (0.6221)\tPrec@1 86.719 (86.724)\n",
      "[150/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5652 (0.6207)\tPrec@1 89.844 (86.724)\n",
      "[160/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6192 (0.6204)\tPrec@1 88.281 (86.743)\n",
      "[170/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6643 (0.6208)\tPrec@1 85.156 (86.778)\n",
      "[180/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6309 (0.6214)\tPrec@1 83.594 (86.706)\n",
      "[190/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.6156 (0.6210)\tPrec@1 83.594 (86.686)\n",
      "[200/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6025 (0.6208)\tPrec@1 90.625 (86.699)\n",
      "[210/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.7266 (0.6223)\tPrec@1 86.719 (86.648)\n",
      "[220/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6262 (0.6226)\tPrec@1 85.156 (86.641)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5824 (0.6231)\tPrec@1 89.062 (86.668)\n",
      "[240/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6555 (0.6247)\tPrec@1 84.375 (86.583)\n",
      "[250/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6456 (0.6244)\tPrec@1 83.594 (86.594)\n",
      "[260/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5875 (0.6251)\tPrec@1 85.156 (86.539)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6754 (0.6255)\tPrec@1 83.594 (86.514)\n",
      "[280/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6515 (0.6257)\tPrec@1 84.375 (86.516)\n",
      "[290/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7075 (0.6263)\tPrec@1 83.594 (86.501)\n",
      "[300/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.5968 (0.6265)\tPrec@1 88.281 (86.506)\n",
      "[310/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7767 (0.6267)\tPrec@1 79.688 (86.518)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5796 (0.6270)\tPrec@1 88.281 (86.517)\n",
      "[330/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6625 (0.6276)\tPrec@1 82.031 (86.469)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5492 (0.6264)\tPrec@1 89.062 (86.501)\n",
      "[350/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6474 (0.6263)\tPrec@1 88.281 (86.516)\n",
      "[360/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5907 (0.6270)\tPrec@1 85.938 (86.468)\n",
      "[370/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.4994 (0.6264)\tPrec@1 93.750 (86.477)\n",
      "[380/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6041 (0.6268)\tPrec@1 85.938 (86.448)\n",
      "[390/391]\tTime 0.128 (0.181)\tData 0.012 (0.020)\tLoss 0.6545 (0.6268)\tPrec@1 85.000 (86.452)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6490 (0.6490)\tPrec@1 85.156 (85.156)\n",
      "Test: [10/79]\tTime 0.019 (0.021)\tLoss 0.7271 (0.9100)\tPrec@1 85.156 (82.884)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.8335 (0.9729)\tPrec@1 85.938 (82.068)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.8816 (0.9469)\tPrec@1 84.375 (82.208)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.3478 (0.9607)\tPrec@1 74.219 (82.012)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.2136 (0.9573)\tPrec@1 81.250 (82.047)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.7686 (0.9584)\tPrec@1 83.594 (82.031)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.7384 (0.9581)\tPrec@1 83.594 (81.976)\n",
      " * Prec@1 82.030\n",
      "\n",
      "===> epoch: 12/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.144 (0.144)\tData 0.020 (0.020)\tLoss 0.5555 (0.5555)\tPrec@1 89.844 (89.844)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.4927 (0.6056)\tPrec@1 91.406 (86.577)\n",
      "[20/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5723 (0.6029)\tPrec@1 91.406 (87.388)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6113 (0.6028)\tPrec@1 86.719 (87.651)\n",
      "[40/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7162 (0.6128)\tPrec@1 85.156 (87.290)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.6859 (0.6149)\tPrec@1 82.031 (86.979)\n",
      "[60/391]\tTime 0.185 (0.181)\tData 0.019 (0.019)\tLoss 0.6497 (0.6185)\tPrec@1 82.812 (86.732)\n",
      "[70/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.6340 (0.6188)\tPrec@1 88.281 (86.653)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5956 (0.6194)\tPrec@1 86.719 (86.680)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.6094 (0.6156)\tPrec@1 87.500 (86.813)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.5479 (0.6190)\tPrec@1 89.844 (86.680)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5108 (0.6176)\tPrec@1 92.969 (86.655)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6255 (0.6194)\tPrec@1 90.625 (86.648)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6235 (0.6199)\tPrec@1 84.375 (86.576)\n",
      "[140/391]\tTime 0.188 (0.182)\tData 0.019 (0.020)\tLoss 0.5510 (0.6211)\tPrec@1 90.625 (86.597)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6327 (0.6227)\tPrec@1 85.938 (86.564)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.7018 (0.6222)\tPrec@1 82.812 (86.563)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.6270 (0.6222)\tPrec@1 86.719 (86.545)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.5996 (0.6218)\tPrec@1 86.719 (86.481)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.6609 (0.6226)\tPrec@1 82.031 (86.432)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.020 (0.019)\tLoss 0.5959 (0.6224)\tPrec@1 88.281 (86.439)\n",
      "[210/391]\tTime 0.188 (0.182)\tData 0.020 (0.019)\tLoss 0.5542 (0.6214)\tPrec@1 89.844 (86.493)\n",
      "[220/391]\tTime 0.188 (0.182)\tData 0.019 (0.019)\tLoss 0.6172 (0.6213)\tPrec@1 87.500 (86.500)\n",
      "[230/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.7358 (0.6241)\tPrec@1 85.156 (86.435)\n",
      "[240/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.6542 (0.6243)\tPrec@1 78.906 (86.408)\n",
      "[250/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.6273 (0.6250)\tPrec@1 88.281 (86.414)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.020 (0.019)\tLoss 0.6447 (0.6252)\tPrec@1 86.719 (86.425)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6099 (0.6252)\tPrec@1 87.500 (86.433)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6057 (0.6245)\tPrec@1 88.281 (86.452)\n",
      "[290/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.6187 (0.6240)\tPrec@1 85.938 (86.480)\n",
      "[300/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.4857 (0.6231)\tPrec@1 91.406 (86.503)\n",
      "[310/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.6579 (0.6239)\tPrec@1 87.500 (86.450)\n",
      "[320/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.6108 (0.6242)\tPrec@1 87.500 (86.429)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.5874 (0.6236)\tPrec@1 85.938 (86.431)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5635 (0.6232)\tPrec@1 88.281 (86.430)\n",
      "[350/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6090 (0.6233)\tPrec@1 86.719 (86.432)\n",
      "[360/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5594 (0.6231)\tPrec@1 89.062 (86.448)\n",
      "[370/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6833 (0.6220)\tPrec@1 84.375 (86.470)\n",
      "[380/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7083 (0.6229)\tPrec@1 82.812 (86.428)\n",
      "[390/391]\tTime 0.128 (0.181)\tData 0.012 (0.020)\tLoss 0.4919 (0.6229)\tPrec@1 91.250 (86.462)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9913 (0.9913)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.020 (0.022)\tLoss 0.9857 (1.1565)\tPrec@1 79.688 (78.125)\n",
      "Test: [20/79]\tTime 0.020 (0.021)\tLoss 0.6919 (1.1869)\tPrec@1 82.812 (77.679)\n",
      "Test: [30/79]\tTime 0.020 (0.021)\tLoss 0.7860 (1.1836)\tPrec@1 84.375 (77.394)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.2230 (1.1818)\tPrec@1 78.906 (77.649)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.3692 (1.1886)\tPrec@1 75.781 (77.681)\n",
      "Test: [60/79]\tTime 0.024 (0.020)\tLoss 1.1489 (1.1871)\tPrec@1 78.125 (77.754)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.3563 (1.1884)\tPrec@1 75.000 (77.795)\n",
      " * Prec@1 77.850\n",
      "\n",
      "===> epoch: 13/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.021 (0.021)\tLoss 0.6464 (0.6464)\tPrec@1 85.156 (85.156)\n",
      "[10/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6352 (0.6706)\tPrec@1 83.594 (85.440)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.6865 (0.6556)\tPrec@1 85.938 (86.049)\n",
      "[30/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6536 (0.6498)\tPrec@1 85.156 (86.164)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6737 (0.6465)\tPrec@1 82.031 (86.090)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6270 (0.6365)\tPrec@1 85.156 (86.474)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.6596 (0.6312)\tPrec@1 81.250 (86.424)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.021 (0.020)\tLoss 0.6764 (0.6259)\tPrec@1 84.375 (86.730)\n",
      "[80/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6704 (0.6270)\tPrec@1 82.031 (86.642)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6459 (0.6244)\tPrec@1 87.500 (86.676)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6265 (0.6263)\tPrec@1 87.500 (86.626)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5882 (0.6262)\tPrec@1 86.719 (86.669)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5983 (0.6246)\tPrec@1 87.500 (86.686)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5275 (0.6223)\tPrec@1 86.719 (86.731)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5474 (0.6206)\tPrec@1 89.844 (86.769)\n",
      "[150/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5390 (0.6196)\tPrec@1 89.062 (86.796)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6646 (0.6205)\tPrec@1 81.250 (86.719)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7132 (0.6218)\tPrec@1 80.469 (86.627)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5987 (0.6214)\tPrec@1 87.500 (86.650)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6674 (0.6221)\tPrec@1 84.375 (86.616)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5435 (0.6215)\tPrec@1 89.062 (86.610)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7010 (0.6200)\tPrec@1 84.375 (86.637)\n",
      "[220/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.7320 (0.6204)\tPrec@1 82.812 (86.599)\n",
      "[230/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7137 (0.6217)\tPrec@1 83.594 (86.533)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7621 (0.6228)\tPrec@1 83.594 (86.508)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6570 (0.6242)\tPrec@1 83.594 (86.488)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7284 (0.6252)\tPrec@1 86.719 (86.500)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5746 (0.6245)\tPrec@1 89.844 (86.531)\n",
      "[280/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5010 (0.6227)\tPrec@1 92.188 (86.624)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5732 (0.6219)\tPrec@1 88.281 (86.644)\n",
      "[300/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6237 (0.6221)\tPrec@1 82.812 (86.605)\n",
      "[310/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7658 (0.6239)\tPrec@1 81.250 (86.528)\n",
      "[320/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6963 (0.6251)\tPrec@1 82.031 (86.463)\n",
      "[330/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6325 (0.6259)\tPrec@1 87.500 (86.445)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5655 (0.6256)\tPrec@1 85.938 (86.464)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5699 (0.6250)\tPrec@1 88.281 (86.472)\n",
      "[360/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6316 (0.6267)\tPrec@1 87.500 (86.390)\n",
      "[370/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.7813 (0.6268)\tPrec@1 81.250 (86.384)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5901 (0.6274)\tPrec@1 88.281 (86.382)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.013 (0.020)\tLoss 0.6378 (0.6273)\tPrec@1 86.250 (86.366)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7192 (0.7192)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.7706 (0.8123)\tPrec@1 82.812 (83.239)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.5587 (0.8449)\tPrec@1 85.938 (83.110)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.4965 (0.8276)\tPrec@1 91.406 (83.291)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.8137 (0.8334)\tPrec@1 81.250 (83.232)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.8793 (0.8228)\tPrec@1 82.031 (83.257)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.7333 (0.8082)\tPrec@1 85.938 (83.453)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.9075 (0.8187)\tPrec@1 83.594 (83.242)\n",
      " * Prec@1 83.430\n",
      "\n",
      "===> epoch: 14/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.186 (0.186)\tData 0.021 (0.021)\tLoss 0.6708 (0.6708)\tPrec@1 85.156 (85.156)\n",
      "[10/391]\tTime 0.180 (0.183)\tData 0.020 (0.020)\tLoss 0.5790 (0.6018)\tPrec@1 89.844 (87.642)\n",
      "[20/391]\tTime 0.178 (0.182)\tData 0.019 (0.021)\tLoss 0.5664 (0.6051)\tPrec@1 91.406 (87.240)\n",
      "[30/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6473 (0.6203)\tPrec@1 84.375 (86.668)\n",
      "[40/391]\tTime 0.182 (0.182)\tData 0.025 (0.020)\tLoss 0.5954 (0.6208)\tPrec@1 86.719 (86.471)\n",
      "[50/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5707 (0.6219)\tPrec@1 88.281 (86.382)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6344 (0.6212)\tPrec@1 85.156 (86.450)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6959 (0.6299)\tPrec@1 84.375 (86.147)\n",
      "[80/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6742 (0.6354)\tPrec@1 82.812 (85.986)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.5698 (0.6330)\tPrec@1 89.844 (86.092)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6167 (0.6286)\tPrec@1 89.062 (86.317)\n",
      "[110/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6220 (0.6258)\tPrec@1 85.156 (86.395)\n",
      "[120/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.6092 (0.6235)\tPrec@1 85.938 (86.428)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5922 (0.6228)\tPrec@1 85.938 (86.415)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6572 (0.6234)\tPrec@1 86.719 (86.336)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6227 (0.6248)\tPrec@1 85.938 (86.320)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5329 (0.6243)\tPrec@1 88.281 (86.301)\n",
      "[170/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7152 (0.6222)\tPrec@1 81.250 (86.353)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6058 (0.6234)\tPrec@1 85.156 (86.322)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5290 (0.6241)\tPrec@1 91.406 (86.326)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6208 (0.6227)\tPrec@1 85.156 (86.365)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6405 (0.6234)\tPrec@1 85.938 (86.308)\n",
      "[220/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.7280 (0.6230)\tPrec@1 82.031 (86.337)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6463 (0.6227)\tPrec@1 85.938 (86.360)\n",
      "[240/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5468 (0.6214)\tPrec@1 89.062 (86.433)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7973 (0.6217)\tPrec@1 82.812 (86.439)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6008 (0.6207)\tPrec@1 87.500 (86.452)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6380 (0.6202)\tPrec@1 86.719 (86.477)\n",
      "[280/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6178 (0.6198)\tPrec@1 84.375 (86.488)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5909 (0.6206)\tPrec@1 92.188 (86.469)\n",
      "[300/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7086 (0.6227)\tPrec@1 84.375 (86.410)\n",
      "[310/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5861 (0.6242)\tPrec@1 85.938 (86.352)\n",
      "[320/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6340 (0.6239)\tPrec@1 82.031 (86.332)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6394 (0.6237)\tPrec@1 86.719 (86.367)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6161 (0.6241)\tPrec@1 85.938 (86.343)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.8028 (0.6258)\tPrec@1 79.688 (86.278)\n",
      "[360/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7280 (0.6276)\tPrec@1 82.031 (86.217)\n",
      "[370/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5711 (0.6274)\tPrec@1 89.062 (86.249)\n",
      "[380/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6678 (0.6281)\tPrec@1 83.594 (86.225)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.012 (0.020)\tLoss 0.6995 (0.6275)\tPrec@1 86.250 (86.246)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.6914 (1.6914)\tPrec@1 73.438 (73.438)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 1.5342 (1.4557)\tPrec@1 74.219 (75.000)\n",
      "Test: [20/79]\tTime 0.020 (0.019)\tLoss 1.0129 (1.4518)\tPrec@1 79.688 (75.632)\n",
      "Test: [30/79]\tTime 0.021 (0.020)\tLoss 1.4995 (1.4605)\tPrec@1 75.000 (75.353)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 2.2972 (1.4929)\tPrec@1 62.500 (75.000)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 1.4782 (1.4874)\tPrec@1 74.219 (74.908)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 1.5075 (1.4741)\tPrec@1 71.094 (74.974)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 1.1999 (1.4672)\tPrec@1 75.000 (74.901)\n",
      " * Prec@1 75.000\n",
      "\n",
      "===> epoch: 15/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.5817 (0.5817)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6557 (0.6013)\tPrec@1 86.719 (87.926)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5533 (0.6000)\tPrec@1 89.062 (87.277)\n",
      "[30/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6000 (0.6054)\tPrec@1 86.719 (87.172)\n",
      "[40/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6823 (0.6097)\tPrec@1 84.375 (86.776)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5923 (0.6052)\tPrec@1 86.719 (86.903)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.021 (0.020)\tLoss 0.5829 (0.6154)\tPrec@1 90.625 (86.668)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6893 (0.6171)\tPrec@1 86.719 (86.543)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5918 (0.6147)\tPrec@1 89.844 (86.642)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6026 (0.6160)\tPrec@1 88.281 (86.599)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5489 (0.6149)\tPrec@1 89.844 (86.757)\n",
      "[110/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6177 (0.6163)\tPrec@1 85.156 (86.606)\n",
      "[120/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6539 (0.6172)\tPrec@1 85.156 (86.570)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6189 (0.6194)\tPrec@1 85.156 (86.534)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5848 (0.6205)\tPrec@1 90.625 (86.492)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5966 (0.6208)\tPrec@1 88.281 (86.439)\n",
      "[160/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6332 (0.6198)\tPrec@1 87.500 (86.544)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5778 (0.6202)\tPrec@1 89.844 (86.550)\n",
      "[180/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5637 (0.6219)\tPrec@1 89.062 (86.542)\n",
      "[190/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7472 (0.6228)\tPrec@1 82.812 (86.518)\n",
      "[200/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6524 (0.6240)\tPrec@1 86.719 (86.524)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5651 (0.6249)\tPrec@1 88.281 (86.515)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5410 (0.6247)\tPrec@1 85.938 (86.496)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6977 (0.6242)\tPrec@1 82.812 (86.526)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4654 (0.6234)\tPrec@1 93.750 (86.583)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7416 (0.6240)\tPrec@1 82.812 (86.560)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5752 (0.6238)\tPrec@1 86.719 (86.563)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.5546 (0.6234)\tPrec@1 89.062 (86.554)\n",
      "[280/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.7598 (0.6252)\tPrec@1 78.906 (86.463)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5733 (0.6255)\tPrec@1 91.406 (86.472)\n",
      "[300/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5344 (0.6245)\tPrec@1 94.531 (86.545)\n",
      "[310/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6162 (0.6235)\tPrec@1 87.500 (86.560)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6945 (0.6232)\tPrec@1 85.156 (86.553)\n",
      "[330/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6159 (0.6248)\tPrec@1 84.375 (86.495)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5737 (0.6251)\tPrec@1 87.500 (86.460)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.6953 (0.6264)\tPrec@1 85.156 (86.423)\n",
      "[360/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6962 (0.6264)\tPrec@1 82.812 (86.429)\n",
      "[370/391]\tTime 0.192 (0.181)\tData 0.028 (0.020)\tLoss 0.5474 (0.6264)\tPrec@1 89.844 (86.453)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6322 (0.6271)\tPrec@1 87.500 (86.423)\n",
      "[390/391]\tTime 0.116 (0.181)\tData 0.013 (0.020)\tLoss 0.7510 (0.6272)\tPrec@1 86.250 (86.434)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 1.0567 (1.0567)\tPrec@1 78.906 (78.906)\n",
      "Test: [10/79]\tTime 0.019 (0.021)\tLoss 0.9564 (1.0992)\tPrec@1 82.812 (80.114)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.9209 (1.1469)\tPrec@1 84.375 (79.576)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.9854 (1.1382)\tPrec@1 81.250 (79.511)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.9073 (1.1305)\tPrec@1 76.562 (79.726)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.9299 (1.1224)\tPrec@1 85.938 (79.887)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.5119 (1.1202)\tPrec@1 86.719 (79.752)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.1298 (1.1158)\tPrec@1 74.219 (79.588)\n",
      " * Prec@1 79.620\n",
      "\n",
      "===> epoch: 16/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.170 (0.170)\tData 0.021 (0.021)\tLoss 0.6544 (0.6544)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5720 (0.6234)\tPrec@1 89.844 (86.435)\n",
      "[20/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6158 (0.6222)\tPrec@1 86.719 (86.682)\n",
      "[30/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.7039 (0.6273)\tPrec@1 82.812 (86.467)\n",
      "[40/391]\tTime 0.180 (0.180)\tData 0.021 (0.020)\tLoss 0.5260 (0.6174)\tPrec@1 91.406 (87.043)\n",
      "[50/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5262 (0.6057)\tPrec@1 86.719 (87.286)\n",
      "[60/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6147 (0.6070)\tPrec@1 85.156 (87.180)\n",
      "[70/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6860 (0.6050)\tPrec@1 82.812 (87.258)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[80/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6019 (0.6122)\tPrec@1 87.500 (87.056)\n",
      "[90/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.4809 (0.6159)\tPrec@1 94.531 (86.873)\n",
      "[100/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6030 (0.6171)\tPrec@1 89.062 (86.835)\n",
      "[110/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6644 (0.6210)\tPrec@1 86.719 (86.620)\n",
      "[120/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6280 (0.6224)\tPrec@1 83.594 (86.519)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6273 (0.6249)\tPrec@1 85.156 (86.385)\n",
      "[140/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.5676 (0.6228)\tPrec@1 87.500 (86.425)\n",
      "[150/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.8337 (0.6233)\tPrec@1 76.562 (86.419)\n",
      "[160/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6646 (0.6217)\tPrec@1 88.281 (86.515)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5754 (0.6228)\tPrec@1 91.406 (86.458)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5343 (0.6232)\tPrec@1 92.188 (86.473)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6479 (0.6252)\tPrec@1 86.719 (86.424)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7380 (0.6272)\tPrec@1 80.469 (86.326)\n",
      "[210/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5956 (0.6279)\tPrec@1 88.281 (86.282)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6564 (0.6274)\tPrec@1 84.375 (86.302)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7131 (0.6261)\tPrec@1 85.156 (86.340)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6597 (0.6269)\tPrec@1 82.031 (86.284)\n",
      "[250/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5654 (0.6269)\tPrec@1 91.406 (86.323)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5246 (0.6260)\tPrec@1 92.188 (86.357)\n",
      "[270/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6281 (0.6260)\tPrec@1 85.938 (86.344)\n",
      "[280/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6326 (0.6260)\tPrec@1 85.156 (86.332)\n",
      "[290/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6925 (0.6256)\tPrec@1 85.938 (86.370)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6538 (0.6258)\tPrec@1 84.375 (86.368)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.7362 (0.6243)\tPrec@1 82.031 (86.430)\n",
      "[320/391]\tTime 0.178 (0.181)\tData 0.023 (0.020)\tLoss 0.6537 (0.6243)\tPrec@1 85.156 (86.444)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7136 (0.6247)\tPrec@1 81.250 (86.421)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6029 (0.6258)\tPrec@1 87.500 (86.375)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6438 (0.6275)\tPrec@1 84.375 (86.331)\n",
      "[360/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6246 (0.6279)\tPrec@1 85.938 (86.325)\n",
      "[370/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6303 (0.6278)\tPrec@1 85.156 (86.306)\n",
      "[380/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.7010 (0.6286)\tPrec@1 82.812 (86.286)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.012 (0.020)\tLoss 0.5916 (0.6277)\tPrec@1 88.750 (86.332)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7538 (0.7538)\tPrec@1 85.156 (85.156)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.6308 (0.7511)\tPrec@1 82.812 (83.878)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.7646 (0.7862)\tPrec@1 83.594 (83.705)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.6118 (0.7938)\tPrec@1 85.156 (83.720)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.7773 (0.7899)\tPrec@1 79.688 (84.089)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 0.6589 (0.7819)\tPrec@1 85.938 (84.390)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 0.6428 (0.7848)\tPrec@1 87.500 (84.260)\n",
      "Test: [70/79]\tTime 0.021 (0.019)\tLoss 0.5508 (0.7815)\tPrec@1 88.281 (84.144)\n",
      " * Prec@1 84.300\n",
      "\n",
      "===> epoch: 17/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.184 (0.184)\tData 0.020 (0.020)\tLoss 0.7716 (0.7716)\tPrec@1 82.031 (82.031)\n",
      "[10/391]\tTime 0.182 (0.183)\tData 0.022 (0.020)\tLoss 0.5320 (0.6168)\tPrec@1 88.281 (87.500)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.6209 (0.6153)\tPrec@1 85.938 (86.942)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.6881 (0.6148)\tPrec@1 81.250 (87.021)\n",
      "[40/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6422 (0.6218)\tPrec@1 85.938 (86.604)\n",
      "[50/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.6645 (0.6224)\tPrec@1 84.375 (86.489)\n",
      "[60/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.7540 (0.6268)\tPrec@1 86.719 (86.424)\n",
      "[70/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.7289 (0.6266)\tPrec@1 82.812 (86.312)\n",
      "[80/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.5834 (0.6272)\tPrec@1 87.500 (86.323)\n",
      "[90/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.6669 (0.6267)\tPrec@1 86.719 (86.418)\n",
      "[100/391]\tTime 0.177 (0.182)\tData 0.019 (0.019)\tLoss 0.4584 (0.6264)\tPrec@1 91.406 (86.409)\n",
      "[110/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.4988 (0.6282)\tPrec@1 89.062 (86.374)\n",
      "[120/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6631 (0.6324)\tPrec@1 82.031 (86.202)\n",
      "[130/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.5716 (0.6330)\tPrec@1 87.500 (86.087)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6018 (0.6305)\tPrec@1 88.281 (86.165)\n",
      "[150/391]\tTime 0.191 (0.182)\tData 0.019 (0.019)\tLoss 0.5701 (0.6294)\tPrec@1 91.406 (86.232)\n",
      "[160/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.5572 (0.6283)\tPrec@1 88.281 (86.253)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.5670 (0.6265)\tPrec@1 86.719 (86.298)\n",
      "[180/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5108 (0.6268)\tPrec@1 92.188 (86.291)\n",
      "[190/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5724 (0.6262)\tPrec@1 87.500 (86.289)\n",
      "[200/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6848 (0.6261)\tPrec@1 82.031 (86.318)\n",
      "[210/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6125 (0.6242)\tPrec@1 89.062 (86.419)\n",
      "[220/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6450 (0.6236)\tPrec@1 84.375 (86.411)\n",
      "[230/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.6744 (0.6243)\tPrec@1 84.375 (86.401)\n",
      "[240/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5735 (0.6237)\tPrec@1 89.062 (86.456)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6105 (0.6233)\tPrec@1 85.938 (86.436)\n",
      "[260/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.7001 (0.6244)\tPrec@1 83.594 (86.392)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6372 (0.6252)\tPrec@1 83.594 (86.379)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.5468 (0.6273)\tPrec@1 90.625 (86.296)\n",
      "[290/391]\tTime 0.187 (0.182)\tData 0.021 (0.020)\tLoss 0.5575 (0.6273)\tPrec@1 91.406 (86.311)\n",
      "[300/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6705 (0.6269)\tPrec@1 82.031 (86.303)\n",
      "[310/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6196 (0.6264)\tPrec@1 87.500 (86.332)\n",
      "[320/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5743 (0.6252)\tPrec@1 86.719 (86.356)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5890 (0.6245)\tPrec@1 90.625 (86.393)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5316 (0.6237)\tPrec@1 89.062 (86.432)\n",
      "[350/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6125 (0.6250)\tPrec@1 85.156 (86.372)\n",
      "[360/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5885 (0.6248)\tPrec@1 90.625 (86.407)\n",
      "[370/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5389 (0.6251)\tPrec@1 91.406 (86.416)\n",
      "[380/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6022 (0.6247)\tPrec@1 86.719 (86.413)\n",
      "[390/391]\tTime 0.115 (0.182)\tData 0.014 (0.020)\tLoss 0.5823 (0.6248)\tPrec@1 87.500 (86.388)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6835 (0.6835)\tPrec@1 85.938 (85.938)\n",
      "Test: [10/79]\tTime 0.021 (0.020)\tLoss 0.6814 (0.8868)\tPrec@1 85.938 (82.528)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.8667 (0.8969)\tPrec@1 83.594 (82.217)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.6780 (0.8919)\tPrec@1 87.500 (82.586)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.0747 (0.8946)\tPrec@1 82.031 (82.622)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.8219 (0.8775)\tPrec@1 81.250 (82.690)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.7044 (0.8656)\tPrec@1 85.156 (82.889)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.9977 (0.8741)\tPrec@1 80.469 (82.724)\n",
      " * Prec@1 82.830\n",
      "\n",
      "===> epoch: 18/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.186 (0.186)\tData 0.021 (0.021)\tLoss 0.6089 (0.6089)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.183 (0.183)\tData 0.019 (0.020)\tLoss 0.6327 (0.5840)\tPrec@1 86.719 (87.926)\n",
      "[20/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5988 (0.5866)\tPrec@1 88.281 (87.463)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[30/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.4992 (0.5885)\tPrec@1 91.406 (87.626)\n",
      "[40/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5826 (0.5886)\tPrec@1 90.625 (87.862)\n",
      "[50/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5333 (0.5856)\tPrec@1 89.062 (87.760)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6027 (0.5871)\tPrec@1 85.156 (87.615)\n",
      "[70/391]\tTime 0.180 (0.184)\tData 0.019 (0.022)\tLoss 0.5514 (0.5957)\tPrec@1 89.062 (87.379)\n",
      "[80/391]\tTime 0.179 (0.183)\tData 0.019 (0.021)\tLoss 0.7935 (0.6037)\tPrec@1 80.469 (87.066)\n",
      "[90/391]\tTime 0.183 (0.183)\tData 0.020 (0.021)\tLoss 0.6210 (0.6050)\tPrec@1 86.719 (86.993)\n",
      "[100/391]\tTime 0.184 (0.183)\tData 0.020 (0.021)\tLoss 0.6659 (0.6091)\tPrec@1 85.156 (86.812)\n",
      "[110/391]\tTime 0.194 (0.183)\tData 0.023 (0.021)\tLoss 0.6536 (0.6083)\tPrec@1 85.938 (86.852)\n",
      "[120/391]\tTime 0.185 (0.183)\tData 0.020 (0.021)\tLoss 0.7632 (0.6117)\tPrec@1 84.375 (86.816)\n",
      "[130/391]\tTime 0.184 (0.183)\tData 0.019 (0.021)\tLoss 0.5757 (0.6124)\tPrec@1 86.719 (86.826)\n",
      "[140/391]\tTime 0.180 (0.183)\tData 0.020 (0.021)\tLoss 0.6435 (0.6141)\tPrec@1 87.500 (86.774)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.020 (0.021)\tLoss 0.7736 (0.6167)\tPrec@1 82.031 (86.724)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.021)\tLoss 0.4716 (0.6164)\tPrec@1 92.969 (86.743)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6247 (0.6175)\tPrec@1 85.156 (86.732)\n",
      "[180/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.5702 (0.6154)\tPrec@1 86.719 (86.788)\n",
      "[190/391]\tTime 0.197 (0.182)\tData 0.019 (0.020)\tLoss 0.5970 (0.6139)\tPrec@1 85.156 (86.846)\n",
      "[200/391]\tTime 0.182 (0.183)\tData 0.020 (0.020)\tLoss 0.6610 (0.6148)\tPrec@1 82.031 (86.777)\n",
      "[210/391]\tTime 0.181 (0.183)\tData 0.019 (0.020)\tLoss 0.5815 (0.6147)\tPrec@1 87.500 (86.748)\n",
      "[220/391]\tTime 0.184 (0.183)\tData 0.020 (0.020)\tLoss 0.5955 (0.6143)\tPrec@1 85.156 (86.786)\n",
      "[230/391]\tTime 0.187 (0.183)\tData 0.019 (0.020)\tLoss 0.6208 (0.6138)\tPrec@1 85.938 (86.786)\n",
      "[240/391]\tTime 0.183 (0.183)\tData 0.019 (0.020)\tLoss 0.7031 (0.6142)\tPrec@1 81.250 (86.741)\n",
      "[250/391]\tTime 0.182 (0.183)\tData 0.019 (0.020)\tLoss 0.6596 (0.6151)\tPrec@1 83.594 (86.685)\n",
      "[260/391]\tTime 0.180 (0.183)\tData 0.020 (0.020)\tLoss 0.6351 (0.6154)\tPrec@1 85.156 (86.674)\n",
      "[270/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6498 (0.6176)\tPrec@1 86.719 (86.577)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6833 (0.6184)\tPrec@1 83.594 (86.546)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5976 (0.6182)\tPrec@1 87.500 (86.560)\n",
      "[300/391]\tTime 0.190 (0.182)\tData 0.019 (0.020)\tLoss 0.6252 (0.6170)\tPrec@1 86.719 (86.589)\n",
      "[310/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5968 (0.6163)\tPrec@1 86.719 (86.623)\n",
      "[320/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5900 (0.6172)\tPrec@1 89.062 (86.602)\n",
      "[330/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6947 (0.6190)\tPrec@1 85.156 (86.572)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6549 (0.6208)\tPrec@1 87.500 (86.510)\n",
      "[350/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.8298 (0.6221)\tPrec@1 81.250 (86.498)\n",
      "[360/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6290 (0.6230)\tPrec@1 86.719 (86.483)\n",
      "[370/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.6010 (0.6227)\tPrec@1 89.062 (86.477)\n",
      "[380/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6426 (0.6244)\tPrec@1 86.719 (86.417)\n",
      "[390/391]\tTime 0.097 (0.182)\tData 0.013 (0.020)\tLoss 0.6721 (0.6252)\tPrec@1 88.750 (86.406)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8817 (0.8817)\tPrec@1 85.938 (85.938)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.8928 (0.7870)\tPrec@1 82.031 (83.381)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.5528 (0.7938)\tPrec@1 88.281 (83.371)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.5427 (0.8236)\tPrec@1 84.375 (83.039)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.6235 (0.8063)\tPrec@1 82.812 (83.403)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 0.7944 (0.8026)\tPrec@1 82.031 (83.395)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 0.7929 (0.8195)\tPrec@1 85.938 (83.197)\n",
      "Test: [70/79]\tTime 0.020 (0.019)\tLoss 0.6818 (0.8158)\tPrec@1 82.812 (83.341)\n",
      " * Prec@1 83.310\n",
      "\n",
      "===> epoch: 19/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.181 (0.181)\tData 0.021 (0.021)\tLoss 0.6334 (0.6334)\tPrec@1 82.031 (82.031)\n",
      "[10/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.6999 (0.6208)\tPrec@1 84.375 (85.724)\n",
      "[20/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5448 (0.6283)\tPrec@1 89.844 (86.124)\n",
      "[30/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5471 (0.6146)\tPrec@1 92.188 (86.719)\n",
      "[40/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.5571 (0.6246)\tPrec@1 88.281 (86.566)\n",
      "[50/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.5630 (0.6207)\tPrec@1 87.500 (86.642)\n",
      "[60/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.6807 (0.6191)\tPrec@1 85.156 (86.616)\n",
      "[70/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6709 (0.6213)\tPrec@1 84.375 (86.521)\n",
      "[80/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.6399 (0.6228)\tPrec@1 85.156 (86.497)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5859 (0.6253)\tPrec@1 85.938 (86.255)\n",
      "[100/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7919 (0.6249)\tPrec@1 80.469 (86.247)\n",
      "[110/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5342 (0.6239)\tPrec@1 89.062 (86.240)\n",
      "[120/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.6073 (0.6244)\tPrec@1 86.719 (86.209)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5365 (0.6266)\tPrec@1 90.625 (86.182)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5151 (0.6266)\tPrec@1 89.062 (86.198)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5678 (0.6246)\tPrec@1 89.844 (86.320)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5696 (0.6224)\tPrec@1 89.062 (86.418)\n",
      "[170/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.7164 (0.6237)\tPrec@1 83.594 (86.358)\n",
      "[180/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6252 (0.6224)\tPrec@1 86.719 (86.408)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7794 (0.6242)\tPrec@1 82.812 (86.338)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6348 (0.6239)\tPrec@1 85.938 (86.342)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6776 (0.6242)\tPrec@1 86.719 (86.304)\n",
      "[220/391]\tTime 0.194 (0.181)\tData 0.020 (0.020)\tLoss 0.6073 (0.6247)\tPrec@1 84.375 (86.330)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7386 (0.6251)\tPrec@1 83.594 (86.326)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6261 (0.6241)\tPrec@1 85.938 (86.372)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5462 (0.6225)\tPrec@1 92.969 (86.482)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7414 (0.6231)\tPrec@1 83.594 (86.458)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6513 (0.6226)\tPrec@1 84.375 (86.468)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6416 (0.6237)\tPrec@1 83.594 (86.432)\n",
      "[290/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5690 (0.6230)\tPrec@1 88.281 (86.464)\n",
      "[300/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7233 (0.6235)\tPrec@1 81.250 (86.438)\n",
      "[310/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6154 (0.6239)\tPrec@1 86.719 (86.417)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5657 (0.6234)\tPrec@1 91.406 (86.456)\n",
      "[330/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6448 (0.6221)\tPrec@1 83.594 (86.516)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4877 (0.6216)\tPrec@1 93.750 (86.561)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6917 (0.6215)\tPrec@1 82.031 (86.558)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5169 (0.6217)\tPrec@1 89.062 (86.517)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5470 (0.6221)\tPrec@1 88.281 (86.514)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5907 (0.6214)\tPrec@1 86.719 (86.555)\n",
      "[390/391]\tTime 0.096 (0.181)\tData 0.012 (0.020)\tLoss 0.6539 (0.6220)\tPrec@1 82.500 (86.506)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.5782 (1.5782)\tPrec@1 73.438 (73.438)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 1.2734 (1.2758)\tPrec@1 76.562 (77.202)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 1.2834 (1.3251)\tPrec@1 74.219 (76.302)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.8915 (1.2992)\tPrec@1 85.938 (77.067)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.1437 (1.2995)\tPrec@1 75.781 (77.210)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.4881 (1.2797)\tPrec@1 75.000 (77.589)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [60/79]\tTime 0.021 (0.020)\tLoss 1.3461 (1.2984)\tPrec@1 78.125 (77.344)\n",
      "Test: [70/79]\tTime 0.022 (0.021)\tLoss 1.2548 (1.3106)\tPrec@1 78.125 (77.069)\n",
      " * Prec@1 77.220\n",
      "\n",
      "===> epoch: 20/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.020 (0.020)\tLoss 0.5649 (0.5649)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.180 (0.183)\tData 0.019 (0.020)\tLoss 0.7672 (0.6420)\tPrec@1 81.250 (86.932)\n",
      "[20/391]\tTime 0.186 (0.182)\tData 0.022 (0.020)\tLoss 0.6595 (0.6242)\tPrec@1 85.156 (87.091)\n",
      "[30/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6680 (0.6144)\tPrec@1 85.938 (87.374)\n",
      "[40/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6558 (0.6087)\tPrec@1 85.938 (87.329)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.025 (0.020)\tLoss 0.6033 (0.6057)\tPrec@1 89.062 (87.316)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6481 (0.6096)\tPrec@1 84.375 (87.052)\n",
      "[70/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6190 (0.6113)\tPrec@1 86.719 (86.961)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7042 (0.6164)\tPrec@1 85.156 (86.719)\n",
      "[90/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6339 (0.6248)\tPrec@1 86.719 (86.461)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6879 (0.6268)\tPrec@1 82.812 (86.278)\n",
      "[110/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6153 (0.6259)\tPrec@1 89.062 (86.353)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6273 (0.6265)\tPrec@1 82.812 (86.299)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5664 (0.6257)\tPrec@1 89.844 (86.307)\n",
      "[140/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6873 (0.6262)\tPrec@1 84.375 (86.336)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4911 (0.6268)\tPrec@1 92.969 (86.331)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5567 (0.6250)\tPrec@1 86.719 (86.379)\n",
      "[170/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6115 (0.6241)\tPrec@1 87.500 (86.463)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6895 (0.6220)\tPrec@1 84.375 (86.568)\n",
      "[190/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5310 (0.6232)\tPrec@1 91.406 (86.535)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5676 (0.6241)\tPrec@1 89.844 (86.521)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6766 (0.6261)\tPrec@1 81.250 (86.452)\n",
      "[220/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5009 (0.6252)\tPrec@1 92.188 (86.510)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6838 (0.6252)\tPrec@1 83.594 (86.496)\n",
      "[240/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5674 (0.6253)\tPrec@1 89.844 (86.524)\n",
      "[250/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5990 (0.6248)\tPrec@1 90.625 (86.548)\n",
      "[260/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.7330 (0.6239)\tPrec@1 85.156 (86.593)\n",
      "[270/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5962 (0.6235)\tPrec@1 86.719 (86.595)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.5546 (0.6235)\tPrec@1 89.844 (86.621)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5769 (0.6242)\tPrec@1 87.500 (86.563)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6703 (0.6250)\tPrec@1 85.156 (86.545)\n",
      "[310/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5398 (0.6242)\tPrec@1 89.844 (86.555)\n",
      "[320/391]\tTime 0.190 (0.181)\tData 0.020 (0.020)\tLoss 0.6524 (0.6247)\tPrec@1 84.375 (86.492)\n",
      "[330/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6675 (0.6247)\tPrec@1 85.156 (86.490)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6404 (0.6245)\tPrec@1 86.719 (86.506)\n",
      "[350/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6927 (0.6245)\tPrec@1 82.812 (86.494)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5870 (0.6252)\tPrec@1 91.406 (86.461)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7901 (0.6257)\tPrec@1 84.375 (86.468)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6949 (0.6256)\tPrec@1 85.156 (86.497)\n",
      "[390/391]\tTime 0.094 (0.180)\tData 0.013 (0.020)\tLoss 0.4693 (0.6259)\tPrec@1 96.250 (86.520)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.0943 (1.0943)\tPrec@1 78.906 (78.906)\n",
      "Test: [10/79]\tTime 0.021 (0.020)\tLoss 1.2432 (1.3507)\tPrec@1 75.000 (75.923)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 1.1908 (1.4247)\tPrec@1 77.344 (75.037)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 1.1253 (1.3942)\tPrec@1 72.656 (75.151)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.8013 (1.3966)\tPrec@1 68.750 (75.114)\n",
      "Test: [50/79]\tTime 0.025 (0.020)\tLoss 1.4340 (1.3819)\tPrec@1 74.219 (75.536)\n",
      "Test: [60/79]\tTime 0.020 (0.021)\tLoss 1.3146 (1.3960)\tPrec@1 75.000 (75.474)\n",
      "Test: [70/79]\tTime 0.022 (0.021)\tLoss 1.5879 (1.3938)\tPrec@1 70.312 (75.418)\n",
      " * Prec@1 75.400\n",
      "\n",
      "===> epoch: 21/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.187 (0.187)\tData 0.020 (0.020)\tLoss 0.5649 (0.5649)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5898 (0.6131)\tPrec@1 90.625 (87.997)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.4664 (0.6223)\tPrec@1 92.188 (87.240)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5729 (0.6269)\tPrec@1 87.500 (86.794)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5843 (0.6200)\tPrec@1 88.281 (86.871)\n",
      "[50/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5063 (0.6171)\tPrec@1 89.062 (86.994)\n",
      "[60/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6455 (0.6159)\tPrec@1 87.500 (86.975)\n",
      "[70/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5885 (0.6138)\tPrec@1 89.062 (87.115)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6184 (0.6161)\tPrec@1 86.719 (87.066)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6627 (0.6208)\tPrec@1 85.156 (86.942)\n",
      "[100/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5132 (0.6184)\tPrec@1 89.062 (86.997)\n",
      "[110/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6347 (0.6159)\tPrec@1 86.719 (87.035)\n",
      "[120/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6337 (0.6137)\tPrec@1 86.719 (87.093)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5126 (0.6127)\tPrec@1 92.188 (87.154)\n",
      "[140/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6783 (0.6146)\tPrec@1 82.031 (87.040)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6214 (0.6158)\tPrec@1 85.156 (86.931)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5967 (0.6173)\tPrec@1 89.844 (86.845)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5962 (0.6174)\tPrec@1 89.062 (86.819)\n",
      "[180/391]\tTime 0.185 (0.181)\tData 0.021 (0.020)\tLoss 0.4866 (0.6168)\tPrec@1 91.406 (86.814)\n",
      "[190/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6782 (0.6177)\tPrec@1 83.594 (86.764)\n",
      "[200/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6193 (0.6165)\tPrec@1 86.719 (86.835)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5593 (0.6150)\tPrec@1 88.281 (86.896)\n",
      "[220/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5257 (0.6146)\tPrec@1 89.844 (86.903)\n",
      "[230/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.6374 (0.6165)\tPrec@1 87.500 (86.827)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6029 (0.6170)\tPrec@1 87.500 (86.790)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7201 (0.6174)\tPrec@1 79.688 (86.722)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5130 (0.6166)\tPrec@1 92.188 (86.734)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5790 (0.6163)\tPrec@1 87.500 (86.742)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5630 (0.6167)\tPrec@1 89.844 (86.763)\n",
      "[290/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6843 (0.6179)\tPrec@1 84.375 (86.687)\n",
      "[300/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.7098 (0.6177)\tPrec@1 83.594 (86.714)\n",
      "[310/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7146 (0.6187)\tPrec@1 79.688 (86.679)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5970 (0.6192)\tPrec@1 86.719 (86.631)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6275 (0.6183)\tPrec@1 86.719 (86.655)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6375 (0.6190)\tPrec@1 86.719 (86.652)\n",
      "[350/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6517 (0.6197)\tPrec@1 87.500 (86.636)\n",
      "[360/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.6214 (0.6204)\tPrec@1 85.938 (86.606)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5363 (0.6210)\tPrec@1 90.625 (86.584)\n",
      "[380/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5973 (0.6217)\tPrec@1 86.719 (86.551)\n",
      "[390/391]\tTime 0.092 (0.180)\tData 0.012 (0.020)\tLoss 0.5577 (0.6224)\tPrec@1 90.000 (86.534)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.3280 (1.3280)\tPrec@1 75.000 (75.000)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 1.2231 (1.3793)\tPrec@1 77.344 (75.781)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 1.4507 (1.4289)\tPrec@1 78.125 (75.409)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 1.1658 (1.4256)\tPrec@1 78.906 (75.252)\n",
      "Test: [40/79]\tTime 0.026 (0.020)\tLoss 1.4978 (1.4085)\tPrec@1 71.875 (75.457)\n",
      "Test: [50/79]\tTime 0.020 (0.021)\tLoss 1.3958 (1.3985)\tPrec@1 75.000 (75.322)\n",
      "Test: [60/79]\tTime 0.026 (0.021)\tLoss 1.1696 (1.4148)\tPrec@1 77.344 (75.282)\n",
      "Test: [70/79]\tTime 0.025 (0.022)\tLoss 1.4539 (1.4033)\tPrec@1 71.094 (75.473)\n",
      " * Prec@1 75.740\n",
      "\n",
      "===> epoch: 22/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6997 (0.6997)\tPrec@1 83.594 (83.594)\n",
      "[10/391]\tTime 0.186 (0.183)\tData 0.019 (0.021)\tLoss 0.6011 (0.6374)\tPrec@1 90.625 (86.009)\n",
      "[20/391]\tTime 0.184 (0.183)\tData 0.019 (0.020)\tLoss 0.5670 (0.6316)\tPrec@1 89.062 (85.975)\n",
      "[30/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6074 (0.6315)\tPrec@1 85.938 (85.685)\n",
      "[40/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6244 (0.6227)\tPrec@1 81.250 (85.957)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6281 (0.6225)\tPrec@1 85.938 (86.106)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6212 (0.6241)\tPrec@1 89.844 (86.078)\n",
      "[70/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6032 (0.6215)\tPrec@1 85.938 (86.180)\n",
      "[80/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.6376 (0.6246)\tPrec@1 83.594 (85.966)\n",
      "[90/391]\tTime 0.188 (0.182)\tData 0.019 (0.020)\tLoss 0.6246 (0.6226)\tPrec@1 90.625 (86.118)\n",
      "[100/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5749 (0.6243)\tPrec@1 91.406 (86.154)\n",
      "[110/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5670 (0.6235)\tPrec@1 88.281 (86.261)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5170 (0.6177)\tPrec@1 89.844 (86.480)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5848 (0.6165)\tPrec@1 85.156 (86.474)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.022 (0.020)\tLoss 0.6273 (0.6150)\tPrec@1 85.938 (86.530)\n",
      "[150/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.5907 (0.6152)\tPrec@1 86.719 (86.538)\n",
      "[160/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5993 (0.6165)\tPrec@1 86.719 (86.500)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5612 (0.6161)\tPrec@1 85.938 (86.486)\n",
      "[180/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5741 (0.6164)\tPrec@1 88.281 (86.473)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.6230 (0.6158)\tPrec@1 85.938 (86.482)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5659 (0.6168)\tPrec@1 89.062 (86.451)\n",
      "[210/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5426 (0.6157)\tPrec@1 88.281 (86.489)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6010 (0.6156)\tPrec@1 87.500 (86.521)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6673 (0.6162)\tPrec@1 83.594 (86.492)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.6303 (0.6153)\tPrec@1 85.156 (86.524)\n",
      "[250/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6391 (0.6159)\tPrec@1 85.156 (86.510)\n",
      "[260/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.7594 (0.6168)\tPrec@1 79.688 (86.479)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6310 (0.6181)\tPrec@1 87.500 (86.474)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5138 (0.6177)\tPrec@1 90.625 (86.502)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7032 (0.6191)\tPrec@1 85.156 (86.458)\n",
      "[300/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5566 (0.6192)\tPrec@1 86.719 (86.436)\n",
      "[310/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6220 (0.6195)\tPrec@1 85.938 (86.422)\n",
      "[320/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5333 (0.6198)\tPrec@1 89.062 (86.385)\n",
      "[330/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.6855 (0.6205)\tPrec@1 82.031 (86.365)\n",
      "[340/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6148 (0.6207)\tPrec@1 86.719 (86.361)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.8035 (0.6215)\tPrec@1 81.250 (86.329)\n",
      "[360/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5871 (0.6222)\tPrec@1 89.062 (86.310)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6678 (0.6229)\tPrec@1 81.250 (86.251)\n",
      "[380/391]\tTime 0.143 (0.181)\tData 0.019 (0.020)\tLoss 0.6087 (0.6228)\tPrec@1 89.062 (86.270)\n",
      "[390/391]\tTime 0.092 (0.180)\tData 0.013 (0.020)\tLoss 0.7931 (0.6233)\tPrec@1 81.250 (86.260)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6111 (0.6111)\tPrec@1 89.062 (89.062)\n",
      "Test: [10/79]\tTime 0.027 (0.022)\tLoss 0.4434 (0.6726)\tPrec@1 89.062 (86.435)\n",
      "Test: [20/79]\tTime 0.020 (0.023)\tLoss 0.6967 (0.7150)\tPrec@1 84.375 (85.379)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.7240 (0.7233)\tPrec@1 88.281 (85.282)\n",
      "Test: [40/79]\tTime 0.024 (0.023)\tLoss 0.8853 (0.7125)\tPrec@1 79.688 (85.309)\n",
      "Test: [50/79]\tTime 0.025 (0.023)\tLoss 0.6685 (0.7009)\tPrec@1 85.938 (85.417)\n",
      "Test: [60/79]\tTime 0.024 (0.023)\tLoss 0.7451 (0.7073)\tPrec@1 86.719 (85.336)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.7408 (0.7180)\tPrec@1 82.812 (85.167)\n",
      " * Prec@1 85.260\n",
      "\n",
      "===> epoch: 23/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.021 (0.021)\tLoss 0.5724 (0.5724)\tPrec@1 90.625 (90.625)\n",
      "[10/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.4457 (0.5699)\tPrec@1 92.969 (89.276)\n",
      "[20/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5712 (0.5926)\tPrec@1 88.281 (88.281)\n",
      "[30/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6048 (0.5897)\tPrec@1 86.719 (88.155)\n",
      "[40/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.4917 (0.5885)\tPrec@1 92.188 (88.110)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6460 (0.5954)\tPrec@1 88.281 (87.822)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5710 (0.5994)\tPrec@1 88.281 (87.679)\n",
      "[70/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5403 (0.5977)\tPrec@1 84.375 (87.555)\n",
      "[80/391]\tTime 0.189 (0.182)\tData 0.020 (0.020)\tLoss 0.6009 (0.6083)\tPrec@1 89.062 (87.133)\n",
      "[90/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6875 (0.6097)\tPrec@1 82.812 (87.019)\n",
      "[100/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6970 (0.6108)\tPrec@1 83.594 (86.951)\n",
      "[110/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6581 (0.6096)\tPrec@1 84.375 (86.986)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5738 (0.6119)\tPrec@1 91.406 (86.900)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6333 (0.6142)\tPrec@1 83.594 (86.790)\n",
      "[140/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6418 (0.6142)\tPrec@1 85.156 (86.752)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7296 (0.6160)\tPrec@1 78.906 (86.677)\n",
      "[160/391]\tTime 0.189 (0.182)\tData 0.019 (0.020)\tLoss 0.6533 (0.6169)\tPrec@1 85.938 (86.631)\n",
      "[170/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6885 (0.6186)\tPrec@1 83.594 (86.563)\n",
      "[180/391]\tTime 0.189 (0.182)\tData 0.028 (0.020)\tLoss 0.6846 (0.6204)\tPrec@1 85.938 (86.537)\n",
      "[190/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6076 (0.6224)\tPrec@1 85.156 (86.408)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6675 (0.6239)\tPrec@1 82.031 (86.361)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7241 (0.6249)\tPrec@1 78.125 (86.360)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5664 (0.6238)\tPrec@1 89.062 (86.418)\n",
      "[230/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7407 (0.6234)\tPrec@1 85.938 (86.421)\n",
      "[240/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.5681 (0.6240)\tPrec@1 86.719 (86.388)\n",
      "[250/391]\tTime 0.187 (0.182)\tData 0.021 (0.020)\tLoss 0.7001 (0.6238)\tPrec@1 84.375 (86.432)\n",
      "[260/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5474 (0.6221)\tPrec@1 87.500 (86.500)\n",
      "[270/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6799 (0.6222)\tPrec@1 85.156 (86.511)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5926 (0.6230)\tPrec@1 86.719 (86.474)\n",
      "[290/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6664 (0.6224)\tPrec@1 87.500 (86.482)\n",
      "[300/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.7291 (0.6218)\tPrec@1 84.375 (86.496)\n",
      "[310/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6288 (0.6226)\tPrec@1 87.500 (86.503)\n",
      "[320/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6465 (0.6226)\tPrec@1 84.375 (86.483)\n",
      "[330/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6883 (0.6233)\tPrec@1 84.375 (86.490)\n",
      "[340/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5036 (0.6230)\tPrec@1 91.406 (86.529)\n",
      "[350/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6251 (0.6237)\tPrec@1 84.375 (86.476)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[360/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5553 (0.6238)\tPrec@1 89.844 (86.474)\n",
      "[370/391]\tTime 0.145 (0.181)\tData 0.019 (0.020)\tLoss 0.5547 (0.6244)\tPrec@1 89.062 (86.464)\n",
      "[380/391]\tTime 0.147 (0.180)\tData 0.020 (0.020)\tLoss 0.6169 (0.6240)\tPrec@1 86.719 (86.499)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.5771 (0.6235)\tPrec@1 90.000 (86.488)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.1104 (1.1104)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.020 (0.024)\tLoss 0.8857 (1.0198)\tPrec@1 83.594 (80.469)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.6380 (1.0606)\tPrec@1 84.375 (79.390)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.7913 (1.1024)\tPrec@1 82.031 (79.234)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.0134 (1.0815)\tPrec@1 71.875 (79.211)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 0.9360 (1.0690)\tPrec@1 82.812 (79.442)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.7126 (1.0763)\tPrec@1 82.812 (79.380)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 0.7335 (1.0693)\tPrec@1 83.594 (79.434)\n",
      " * Prec@1 79.490\n",
      "\n",
      "===> epoch: 24/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.149 (0.149)\tData 0.020 (0.020)\tLoss 0.6473 (0.6473)\tPrec@1 85.156 (85.156)\n",
      "[10/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.7358 (0.6124)\tPrec@1 83.594 (87.216)\n",
      "[20/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5469 (0.6049)\tPrec@1 86.719 (86.905)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6034 (0.6154)\tPrec@1 88.281 (86.794)\n",
      "[40/391]\tTime 0.185 (0.181)\tData 0.022 (0.020)\tLoss 0.6570 (0.6108)\tPrec@1 82.812 (86.928)\n",
      "[50/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.7790 (0.6076)\tPrec@1 79.688 (87.025)\n",
      "[60/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6980 (0.6065)\tPrec@1 83.594 (87.103)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6403 (0.6033)\tPrec@1 85.938 (87.126)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5038 (0.5975)\tPrec@1 92.188 (87.423)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6058 (0.6002)\tPrec@1 88.281 (87.320)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6488 (0.6008)\tPrec@1 83.594 (87.345)\n",
      "[110/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6059 (0.6034)\tPrec@1 85.156 (87.197)\n",
      "[120/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.5636 (0.6041)\tPrec@1 89.062 (87.138)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5690 (0.6025)\tPrec@1 89.844 (87.220)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.8217 (0.6083)\tPrec@1 78.125 (86.979)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5128 (0.6077)\tPrec@1 91.406 (87.008)\n",
      "[160/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5405 (0.6067)\tPrec@1 89.844 (87.068)\n",
      "[170/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6109 (0.6071)\tPrec@1 88.281 (87.112)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6325 (0.6079)\tPrec@1 88.281 (87.103)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5077 (0.6082)\tPrec@1 92.188 (87.058)\n",
      "[200/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5295 (0.6079)\tPrec@1 89.844 (87.069)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5738 (0.6073)\tPrec@1 86.719 (87.067)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5760 (0.6075)\tPrec@1 85.938 (87.076)\n",
      "[230/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4785 (0.6088)\tPrec@1 92.188 (87.037)\n",
      "[240/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7065 (0.6088)\tPrec@1 84.375 (87.049)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6750 (0.6115)\tPrec@1 84.375 (86.918)\n",
      "[260/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6465 (0.6132)\tPrec@1 84.375 (86.821)\n",
      "[270/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6014 (0.6138)\tPrec@1 88.281 (86.817)\n",
      "[280/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5563 (0.6137)\tPrec@1 89.844 (86.824)\n",
      "[290/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7748 (0.6155)\tPrec@1 75.000 (86.713)\n",
      "[300/391]\tTime 0.184 (0.181)\tData 0.021 (0.020)\tLoss 0.7666 (0.6170)\tPrec@1 80.469 (86.651)\n",
      "[310/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6981 (0.6175)\tPrec@1 83.594 (86.661)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6522 (0.6173)\tPrec@1 85.938 (86.668)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6424 (0.6172)\tPrec@1 83.594 (86.686)\n",
      "[340/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6559 (0.6178)\tPrec@1 84.375 (86.684)\n",
      "[350/391]\tTime 0.183 (0.181)\tData 0.023 (0.020)\tLoss 0.6437 (0.6187)\tPrec@1 84.375 (86.641)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6874 (0.6181)\tPrec@1 81.250 (86.678)\n",
      "[370/391]\tTime 0.146 (0.180)\tData 0.022 (0.020)\tLoss 0.6073 (0.6188)\tPrec@1 86.719 (86.658)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6383 (0.6193)\tPrec@1 82.812 (86.624)\n",
      "[390/391]\tTime 0.132 (0.180)\tData 0.012 (0.020)\tLoss 0.5932 (0.6190)\tPrec@1 87.500 (86.614)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9825 (0.9825)\tPrec@1 82.031 (82.031)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.6695 (0.8538)\tPrec@1 86.719 (83.168)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.7215 (0.8656)\tPrec@1 86.719 (83.110)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.7359 (0.8737)\tPrec@1 87.500 (82.989)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 0.8115 (0.8559)\tPrec@1 78.906 (83.270)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 0.8368 (0.8407)\tPrec@1 85.938 (83.578)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.5885 (0.8369)\tPrec@1 87.500 (83.581)\n",
      "Test: [70/79]\tTime 0.024 (0.023)\tLoss 0.7955 (0.8295)\tPrec@1 86.719 (83.726)\n",
      " * Prec@1 83.870\n",
      "\n",
      "===> epoch: 25/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.186 (0.186)\tData 0.022 (0.022)\tLoss 0.5771 (0.5771)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.178 (0.183)\tData 0.019 (0.020)\tLoss 0.6128 (0.5939)\tPrec@1 90.625 (87.855)\n",
      "[20/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5708 (0.5984)\tPrec@1 88.281 (87.351)\n",
      "[30/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7664 (0.6003)\tPrec@1 83.594 (87.576)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6102 (0.6066)\tPrec@1 89.062 (87.538)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5460 (0.6051)\tPrec@1 92.969 (87.592)\n",
      "[60/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6811 (0.6058)\tPrec@1 83.594 (87.385)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5853 (0.6090)\tPrec@1 88.281 (87.225)\n",
      "[80/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.6498 (0.6141)\tPrec@1 85.938 (86.979)\n",
      "[90/391]\tTime 0.183 (0.182)\tData 0.022 (0.020)\tLoss 0.5840 (0.6131)\tPrec@1 86.719 (86.959)\n",
      "[100/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6009 (0.6146)\tPrec@1 86.719 (86.897)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7686 (0.6168)\tPrec@1 83.594 (86.888)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6285 (0.6181)\tPrec@1 86.719 (86.829)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5782 (0.6170)\tPrec@1 87.500 (86.886)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5601 (0.6146)\tPrec@1 88.281 (86.918)\n",
      "[150/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6419 (0.6141)\tPrec@1 84.375 (86.931)\n",
      "[160/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6774 (0.6140)\tPrec@1 82.812 (86.913)\n",
      "[170/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6381 (0.6126)\tPrec@1 82.031 (86.920)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7069 (0.6135)\tPrec@1 82.031 (86.840)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6565 (0.6150)\tPrec@1 88.281 (86.813)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.8007 (0.6158)\tPrec@1 82.812 (86.765)\n",
      "[210/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6960 (0.6180)\tPrec@1 88.281 (86.693)\n",
      "[220/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6466 (0.6194)\tPrec@1 83.594 (86.659)\n",
      "[230/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5481 (0.6206)\tPrec@1 91.406 (86.624)\n",
      "[240/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4996 (0.6209)\tPrec@1 91.406 (86.596)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.7052 (0.6228)\tPrec@1 82.812 (86.563)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6256 (0.6235)\tPrec@1 87.500 (86.509)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5693 (0.6218)\tPrec@1 87.500 (86.563)\n",
      "[280/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5663 (0.6212)\tPrec@1 89.062 (86.552)\n",
      "[290/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4801 (0.6222)\tPrec@1 91.406 (86.466)\n",
      "[300/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5825 (0.6224)\tPrec@1 90.625 (86.477)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[310/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6798 (0.6232)\tPrec@1 81.250 (86.447)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6269 (0.6231)\tPrec@1 85.156 (86.480)\n",
      "[330/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5934 (0.6248)\tPrec@1 85.938 (86.402)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7030 (0.6254)\tPrec@1 81.250 (86.354)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5717 (0.6248)\tPrec@1 87.500 (86.378)\n",
      "[360/391]\tTime 0.146 (0.180)\tData 0.019 (0.020)\tLoss 0.5831 (0.6247)\tPrec@1 89.062 (86.377)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6658 (0.6249)\tPrec@1 85.156 (86.373)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.5759 (0.6248)\tPrec@1 87.500 (86.387)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.012 (0.020)\tLoss 0.6519 (0.6247)\tPrec@1 86.250 (86.410)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9651 (0.9651)\tPrec@1 81.250 (81.250)\n",
      "Test: [10/79]\tTime 0.019 (0.023)\tLoss 1.1574 (1.2243)\tPrec@1 80.469 (79.048)\n",
      "Test: [20/79]\tTime 0.024 (0.023)\tLoss 1.1903 (1.2595)\tPrec@1 78.906 (78.571)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.9440 (1.2470)\tPrec@1 83.594 (78.679)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 1.2931 (1.2220)\tPrec@1 78.906 (79.516)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 0.9670 (1.1873)\tPrec@1 83.594 (79.733)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.0221 (1.1906)\tPrec@1 83.594 (79.752)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.9889 (1.1942)\tPrec@1 80.469 (79.765)\n",
      " * Prec@1 79.690\n",
      "\n",
      "===> epoch: 26/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5623 (0.5623)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.178 (0.183)\tData 0.019 (0.019)\tLoss 0.6133 (0.6043)\tPrec@1 87.500 (87.500)\n",
      "[20/391]\tTime 0.181 (0.183)\tData 0.019 (0.020)\tLoss 0.5647 (0.6034)\tPrec@1 85.156 (87.240)\n",
      "[30/391]\tTime 0.179 (0.183)\tData 0.019 (0.020)\tLoss 0.5985 (0.6098)\tPrec@1 87.500 (86.971)\n",
      "[40/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6942 (0.6093)\tPrec@1 82.812 (86.986)\n",
      "[50/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.5516 (0.6135)\tPrec@1 90.625 (86.780)\n",
      "[60/391]\tTime 0.173 (0.182)\tData 0.019 (0.019)\tLoss 0.5947 (0.6129)\tPrec@1 86.719 (86.872)\n",
      "[70/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5814 (0.6131)\tPrec@1 87.500 (86.851)\n",
      "[80/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7578 (0.6135)\tPrec@1 81.250 (86.748)\n",
      "[90/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6197 (0.6099)\tPrec@1 85.156 (86.942)\n",
      "[100/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5806 (0.6121)\tPrec@1 87.500 (86.858)\n",
      "[110/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6726 (0.6124)\tPrec@1 85.156 (86.852)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6641 (0.6147)\tPrec@1 82.812 (86.783)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6499 (0.6153)\tPrec@1 85.156 (86.814)\n",
      "[140/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6349 (0.6167)\tPrec@1 85.156 (86.769)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6001 (0.6170)\tPrec@1 86.719 (86.802)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6085 (0.6162)\tPrec@1 87.500 (86.826)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6224 (0.6149)\tPrec@1 85.156 (86.801)\n",
      "[180/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6362 (0.6151)\tPrec@1 87.500 (86.835)\n",
      "[190/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6404 (0.6157)\tPrec@1 86.719 (86.821)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.023 (0.020)\tLoss 0.5825 (0.6172)\tPrec@1 87.500 (86.758)\n",
      "[210/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4705 (0.6172)\tPrec@1 92.188 (86.711)\n",
      "[220/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6582 (0.6169)\tPrec@1 83.594 (86.754)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7347 (0.6188)\tPrec@1 83.594 (86.668)\n",
      "[240/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6337 (0.6203)\tPrec@1 86.719 (86.605)\n",
      "[250/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6308 (0.6203)\tPrec@1 84.375 (86.600)\n",
      "[260/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5546 (0.6218)\tPrec@1 89.062 (86.569)\n",
      "[270/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6366 (0.6225)\tPrec@1 87.500 (86.517)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5881 (0.6234)\tPrec@1 89.062 (86.516)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5200 (0.6244)\tPrec@1 91.406 (86.458)\n",
      "[300/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.6494 (0.6239)\tPrec@1 86.719 (86.488)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6409 (0.6241)\tPrec@1 85.156 (86.503)\n",
      "[320/391]\tTime 0.189 (0.182)\tData 0.020 (0.020)\tLoss 0.6595 (0.6229)\tPrec@1 85.156 (86.546)\n",
      "[330/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.7493 (0.6232)\tPrec@1 78.906 (86.492)\n",
      "[340/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5809 (0.6233)\tPrec@1 91.406 (86.494)\n",
      "[350/391]\tTime 0.147 (0.181)\tData 0.022 (0.020)\tLoss 0.5868 (0.6224)\tPrec@1 85.156 (86.507)\n",
      "[360/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5778 (0.6224)\tPrec@1 85.938 (86.485)\n",
      "[370/391]\tTime 0.190 (0.180)\tData 0.019 (0.020)\tLoss 0.6617 (0.6233)\tPrec@1 83.594 (86.456)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6365 (0.6235)\tPrec@1 85.156 (86.436)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.5141 (0.6231)\tPrec@1 93.750 (86.466)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9183 (0.9183)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.8442 (0.9332)\tPrec@1 81.250 (81.889)\n",
      "Test: [20/79]\tTime 0.024 (0.023)\tLoss 0.8796 (0.9461)\tPrec@1 82.812 (82.031)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 0.8630 (0.9469)\tPrec@1 86.719 (81.956)\n",
      "Test: [40/79]\tTime 0.022 (0.023)\tLoss 0.8989 (0.9208)\tPrec@1 81.250 (82.489)\n",
      "Test: [50/79]\tTime 0.026 (0.022)\tLoss 0.9192 (0.9184)\tPrec@1 82.812 (82.506)\n",
      "Test: [60/79]\tTime 0.020 (0.022)\tLoss 0.8297 (0.9237)\tPrec@1 86.719 (82.710)\n",
      "Test: [70/79]\tTime 0.022 (0.022)\tLoss 1.2321 (0.9394)\tPrec@1 81.250 (82.460)\n",
      " * Prec@1 82.450\n",
      "\n",
      "===> epoch: 27/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.181 (0.181)\tData 0.021 (0.021)\tLoss 0.4887 (0.4887)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.189 (0.191)\tData 0.019 (0.020)\tLoss 0.6992 (0.6026)\tPrec@1 81.250 (87.287)\n",
      "[20/391]\tTime 0.180 (0.186)\tData 0.019 (0.020)\tLoss 0.6012 (0.5879)\tPrec@1 87.500 (87.649)\n",
      "[30/391]\tTime 0.181 (0.185)\tData 0.020 (0.020)\tLoss 0.5739 (0.5993)\tPrec@1 88.281 (86.996)\n",
      "[40/391]\tTime 0.179 (0.185)\tData 0.019 (0.020)\tLoss 0.5651 (0.6064)\tPrec@1 88.281 (86.681)\n",
      "[50/391]\tTime 0.178 (0.184)\tData 0.019 (0.019)\tLoss 0.6505 (0.6100)\tPrec@1 83.594 (86.520)\n",
      "[60/391]\tTime 0.178 (0.183)\tData 0.019 (0.019)\tLoss 0.6445 (0.6115)\tPrec@1 86.719 (86.565)\n",
      "[70/391]\tTime 0.185 (0.183)\tData 0.019 (0.019)\tLoss 0.6895 (0.6170)\tPrec@1 85.156 (86.488)\n",
      "[80/391]\tTime 0.188 (0.183)\tData 0.019 (0.019)\tLoss 0.6286 (0.6190)\tPrec@1 88.281 (86.497)\n",
      "[90/391]\tTime 0.181 (0.183)\tData 0.019 (0.019)\tLoss 0.6605 (0.6174)\tPrec@1 85.156 (86.547)\n",
      "[100/391]\tTime 0.184 (0.183)\tData 0.019 (0.019)\tLoss 0.6704 (0.6198)\tPrec@1 82.812 (86.440)\n",
      "[110/391]\tTime 0.180 (0.183)\tData 0.020 (0.019)\tLoss 0.5707 (0.6211)\tPrec@1 88.281 (86.374)\n",
      "[120/391]\tTime 0.179 (0.183)\tData 0.019 (0.019)\tLoss 0.5156 (0.6201)\tPrec@1 91.406 (86.422)\n",
      "[130/391]\tTime 0.179 (0.183)\tData 0.019 (0.019)\tLoss 0.6000 (0.6204)\tPrec@1 85.156 (86.403)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6624 (0.6215)\tPrec@1 85.938 (86.386)\n",
      "[150/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5900 (0.6228)\tPrec@1 85.938 (86.336)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.5788 (0.6240)\tPrec@1 88.281 (86.316)\n",
      "[170/391]\tTime 0.178 (0.182)\tData 0.019 (0.019)\tLoss 0.5683 (0.6278)\tPrec@1 89.062 (86.234)\n",
      "[180/391]\tTime 0.178 (0.182)\tData 0.019 (0.019)\tLoss 0.6297 (0.6280)\tPrec@1 83.594 (86.248)\n",
      "[190/391]\tTime 0.188 (0.182)\tData 0.019 (0.019)\tLoss 0.6455 (0.6303)\tPrec@1 83.594 (86.171)\n",
      "[200/391]\tTime 0.183 (0.182)\tData 0.020 (0.019)\tLoss 0.6194 (0.6307)\tPrec@1 86.719 (86.155)\n",
      "[210/391]\tTime 0.186 (0.182)\tData 0.020 (0.019)\tLoss 0.6793 (0.6321)\tPrec@1 79.688 (86.104)\n",
      "[220/391]\tTime 0.191 (0.182)\tData 0.019 (0.020)\tLoss 0.5554 (0.6318)\tPrec@1 89.844 (86.160)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6809 (0.6310)\tPrec@1 84.375 (86.174)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6095 (0.6291)\tPrec@1 84.375 (86.245)\n",
      "[250/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6782 (0.6289)\tPrec@1 83.594 (86.258)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[260/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.7585 (0.6292)\tPrec@1 82.812 (86.249)\n",
      "[270/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.7007 (0.6301)\tPrec@1 81.250 (86.214)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6420 (0.6302)\tPrec@1 86.719 (86.229)\n",
      "[290/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6479 (0.6310)\tPrec@1 85.938 (86.171)\n",
      "[300/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6122 (0.6306)\tPrec@1 90.625 (86.228)\n",
      "[310/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6601 (0.6305)\tPrec@1 85.156 (86.226)\n",
      "[320/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5681 (0.6305)\tPrec@1 90.625 (86.227)\n",
      "[330/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.7019 (0.6306)\tPrec@1 81.250 (86.230)\n",
      "[340/391]\tTime 0.144 (0.181)\tData 0.019 (0.020)\tLoss 0.6956 (0.6306)\tPrec@1 83.594 (86.212)\n",
      "[350/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.5954 (0.6308)\tPrec@1 88.281 (86.216)\n",
      "[360/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.6157 (0.6307)\tPrec@1 87.500 (86.210)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6658 (0.6304)\tPrec@1 80.469 (86.205)\n",
      "[380/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6251 (0.6299)\tPrec@1 86.719 (86.208)\n",
      "[390/391]\tTime 0.128 (0.181)\tData 0.013 (0.020)\tLoss 0.6569 (0.6309)\tPrec@1 87.500 (86.180)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.4994 (1.4994)\tPrec@1 72.656 (72.656)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.6300 (1.5386)\tPrec@1 75.000 (73.295)\n",
      "Test: [20/79]\tTime 0.024 (0.023)\tLoss 1.2383 (1.5107)\tPrec@1 75.781 (73.921)\n",
      "Test: [30/79]\tTime 0.025 (0.024)\tLoss 1.1190 (1.4529)\tPrec@1 78.125 (74.269)\n",
      "Test: [40/79]\tTime 0.025 (0.024)\tLoss 1.7823 (1.4479)\tPrec@1 69.531 (74.314)\n",
      "Test: [50/79]\tTime 0.024 (0.024)\tLoss 1.5380 (1.4389)\tPrec@1 69.531 (74.157)\n",
      "Test: [60/79]\tTime 0.019 (0.024)\tLoss 1.2177 (1.4349)\tPrec@1 75.000 (74.180)\n",
      "Test: [70/79]\tTime 0.023 (0.024)\tLoss 1.0427 (1.4351)\tPrec@1 76.562 (74.131)\n",
      " * Prec@1 74.150\n",
      "\n",
      "===> epoch: 28/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.165 (0.165)\tData 0.021 (0.021)\tLoss 0.5532 (0.5532)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.192 (0.180)\tData 0.034 (0.021)\tLoss 0.6117 (0.6349)\tPrec@1 87.500 (86.009)\n",
      "[20/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.5679 (0.6196)\tPrec@1 87.500 (86.458)\n",
      "[30/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6332 (0.6236)\tPrec@1 86.719 (86.290)\n",
      "[40/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6653 (0.6216)\tPrec@1 85.938 (86.566)\n",
      "[50/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6776 (0.6169)\tPrec@1 85.156 (86.811)\n",
      "[60/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6121 (0.6101)\tPrec@1 86.719 (87.052)\n",
      "[70/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6898 (0.6150)\tPrec@1 85.938 (86.851)\n",
      "[80/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5530 (0.6121)\tPrec@1 91.406 (86.960)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5675 (0.6119)\tPrec@1 89.062 (86.822)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5293 (0.6102)\tPrec@1 89.062 (86.897)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5517 (0.6077)\tPrec@1 88.281 (86.972)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6101 (0.6068)\tPrec@1 85.938 (86.983)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6395 (0.6102)\tPrec@1 86.719 (86.868)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6261 (0.6097)\tPrec@1 86.719 (86.907)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6105 (0.6123)\tPrec@1 88.281 (86.827)\n",
      "[160/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.5956 (0.6130)\tPrec@1 90.625 (86.796)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6128 (0.6150)\tPrec@1 86.719 (86.700)\n",
      "[180/391]\tTime 0.185 (0.182)\tData 0.022 (0.020)\tLoss 0.5883 (0.6139)\tPrec@1 85.938 (86.723)\n",
      "[190/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6007 (0.6139)\tPrec@1 89.844 (86.723)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5248 (0.6149)\tPrec@1 89.844 (86.703)\n",
      "[210/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7228 (0.6146)\tPrec@1 85.938 (86.722)\n",
      "[220/391]\tTime 0.176 (0.182)\tData 0.019 (0.020)\tLoss 0.6440 (0.6145)\tPrec@1 84.375 (86.712)\n",
      "[230/391]\tTime 0.183 (0.182)\tData 0.022 (0.020)\tLoss 0.6839 (0.6145)\tPrec@1 82.812 (86.692)\n",
      "[240/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6172 (0.6149)\tPrec@1 89.062 (86.706)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7477 (0.6162)\tPrec@1 79.688 (86.666)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.6201 (0.6181)\tPrec@1 82.812 (86.581)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7328 (0.6184)\tPrec@1 78.125 (86.526)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6149 (0.6179)\tPrec@1 87.500 (86.555)\n",
      "[290/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6125 (0.6176)\tPrec@1 86.719 (86.576)\n",
      "[300/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6182 (0.6172)\tPrec@1 89.844 (86.599)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5370 (0.6165)\tPrec@1 91.406 (86.648)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6132 (0.6166)\tPrec@1 85.156 (86.624)\n",
      "[330/391]\tTime 0.146 (0.181)\tData 0.020 (0.020)\tLoss 0.5998 (0.6164)\tPrec@1 85.938 (86.627)\n",
      "[340/391]\tTime 0.166 (0.180)\tData 0.020 (0.020)\tLoss 0.6114 (0.6165)\tPrec@1 86.719 (86.620)\n",
      "[350/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5033 (0.6171)\tPrec@1 90.625 (86.592)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6426 (0.6167)\tPrec@1 89.062 (86.624)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6175 (0.6163)\tPrec@1 89.062 (86.660)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6834 (0.6169)\tPrec@1 83.594 (86.641)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.012 (0.020)\tLoss 0.7645 (0.6184)\tPrec@1 82.500 (86.590)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.5084 (1.5084)\tPrec@1 74.219 (74.219)\n",
      "Test: [10/79]\tTime 0.026 (0.022)\tLoss 1.3030 (1.6105)\tPrec@1 74.219 (73.224)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 1.6254 (1.6943)\tPrec@1 75.000 (72.470)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 1.1646 (1.6892)\tPrec@1 78.125 (72.228)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.9158 (1.6807)\tPrec@1 67.188 (72.218)\n",
      "Test: [50/79]\tTime 0.025 (0.022)\tLoss 1.5672 (1.6651)\tPrec@1 70.312 (72.365)\n",
      "Test: [60/79]\tTime 0.030 (0.023)\tLoss 1.4974 (1.6795)\tPrec@1 73.438 (72.208)\n",
      "Test: [70/79]\tTime 0.021 (0.023)\tLoss 1.4899 (1.6857)\tPrec@1 76.562 (72.227)\n",
      " * Prec@1 72.340\n",
      "\n",
      "===> epoch: 29/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.187 (0.187)\tData 0.020 (0.020)\tLoss 0.5123 (0.5123)\tPrec@1 92.188 (92.188)\n",
      "[10/391]\tTime 0.183 (0.183)\tData 0.020 (0.019)\tLoss 0.5474 (0.6244)\tPrec@1 85.938 (86.435)\n",
      "[20/391]\tTime 0.185 (0.183)\tData 0.020 (0.019)\tLoss 0.5598 (0.6279)\tPrec@1 88.281 (86.496)\n",
      "[30/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5641 (0.6379)\tPrec@1 91.406 (86.139)\n",
      "[40/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5576 (0.6389)\tPrec@1 89.062 (86.128)\n",
      "[50/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5889 (0.6432)\tPrec@1 93.750 (86.121)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5986 (0.6372)\tPrec@1 88.281 (86.347)\n",
      "[70/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5674 (0.6396)\tPrec@1 88.281 (86.257)\n",
      "[80/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6061 (0.6380)\tPrec@1 85.156 (86.294)\n",
      "[90/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6729 (0.6399)\tPrec@1 82.031 (86.135)\n",
      "[100/391]\tTime 0.177 (0.182)\tData 0.020 (0.020)\tLoss 0.7025 (0.6395)\tPrec@1 83.594 (86.154)\n",
      "[110/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5751 (0.6405)\tPrec@1 91.406 (86.156)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6457 (0.6391)\tPrec@1 82.812 (86.183)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6254 (0.6381)\tPrec@1 83.594 (86.206)\n",
      "[140/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6422 (0.6377)\tPrec@1 86.719 (86.264)\n",
      "[150/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6418 (0.6366)\tPrec@1 84.375 (86.284)\n",
      "[160/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.5879 (0.6363)\tPrec@1 89.062 (86.365)\n",
      "[170/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6235 (0.6361)\tPrec@1 85.938 (86.349)\n",
      "[180/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6594 (0.6359)\tPrec@1 85.938 (86.382)\n",
      "[190/391]\tTime 0.185 (0.182)\tData 0.021 (0.020)\tLoss 0.6438 (0.6339)\tPrec@1 81.250 (86.424)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5855 (0.6329)\tPrec@1 92.188 (86.501)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[210/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6895 (0.6322)\tPrec@1 82.031 (86.500)\n",
      "[220/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5599 (0.6303)\tPrec@1 87.500 (86.528)\n",
      "[230/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6511 (0.6287)\tPrec@1 85.938 (86.540)\n",
      "[240/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6148 (0.6276)\tPrec@1 86.719 (86.550)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6985 (0.6269)\tPrec@1 84.375 (86.541)\n",
      "[260/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6170 (0.6267)\tPrec@1 86.719 (86.530)\n",
      "[270/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7035 (0.6262)\tPrec@1 82.812 (86.560)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7170 (0.6286)\tPrec@1 84.375 (86.494)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6165 (0.6293)\tPrec@1 85.938 (86.453)\n",
      "[300/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5880 (0.6288)\tPrec@1 90.625 (86.462)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6577 (0.6293)\tPrec@1 89.062 (86.475)\n",
      "[320/391]\tTime 0.143 (0.181)\tData 0.019 (0.020)\tLoss 0.6198 (0.6290)\tPrec@1 84.375 (86.470)\n",
      "[330/391]\tTime 0.146 (0.180)\tData 0.019 (0.020)\tLoss 0.5899 (0.6292)\tPrec@1 86.719 (86.471)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5816 (0.6281)\tPrec@1 89.844 (86.506)\n",
      "[350/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.7364 (0.6277)\tPrec@1 82.031 (86.505)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6526 (0.6283)\tPrec@1 85.938 (86.448)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6253 (0.6278)\tPrec@1 85.156 (86.470)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.6327 (0.6279)\tPrec@1 85.156 (86.444)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.6340 (0.6276)\tPrec@1 87.500 (86.446)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.3904 (1.3904)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.9483 (1.1266)\tPrec@1 81.250 (80.824)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.1600 (1.1580)\tPrec@1 83.594 (80.246)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.9362 (1.1769)\tPrec@1 79.688 (80.091)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.1807 (1.1381)\tPrec@1 78.125 (80.393)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.0541 (1.1268)\tPrec@1 82.812 (80.499)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.1553 (1.1327)\tPrec@1 79.688 (80.482)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.0816 (1.1335)\tPrec@1 80.469 (80.601)\n",
      " * Prec@1 80.590\n",
      "\n",
      "===> epoch: 30/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.173 (0.173)\tData 0.021 (0.021)\tLoss 0.6058 (0.6058)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5699 (0.5909)\tPrec@1 88.281 (87.216)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5818 (0.5974)\tPrec@1 89.062 (86.868)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6059 (0.5954)\tPrec@1 89.062 (87.198)\n",
      "[40/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5922 (0.5976)\tPrec@1 84.375 (87.005)\n",
      "[50/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6082 (0.6003)\tPrec@1 89.844 (87.178)\n",
      "[60/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6328 (0.6044)\tPrec@1 85.156 (86.885)\n",
      "[70/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.6567 (0.6084)\tPrec@1 83.594 (86.763)\n",
      "[80/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5740 (0.6096)\tPrec@1 86.719 (86.902)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5883 (0.6107)\tPrec@1 89.062 (86.770)\n",
      "[100/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.4913 (0.6095)\tPrec@1 91.406 (86.843)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5813 (0.6091)\tPrec@1 88.281 (86.860)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6070 (0.6066)\tPrec@1 85.938 (86.945)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.022 (0.020)\tLoss 0.6774 (0.6063)\tPrec@1 84.375 (87.005)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.7371 (0.6082)\tPrec@1 83.594 (86.974)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7115 (0.6098)\tPrec@1 81.250 (86.910)\n",
      "[160/391]\tTime 0.177 (0.181)\tData 0.020 (0.020)\tLoss 0.6103 (0.6108)\tPrec@1 90.625 (86.898)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6050 (0.6116)\tPrec@1 84.375 (86.833)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6261 (0.6120)\tPrec@1 85.938 (86.822)\n",
      "[190/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.6538 (0.6124)\tPrec@1 85.938 (86.833)\n",
      "[200/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.5539 (0.6128)\tPrec@1 89.844 (86.812)\n",
      "[210/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6303 (0.6128)\tPrec@1 85.156 (86.811)\n",
      "[220/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6761 (0.6126)\tPrec@1 85.156 (86.850)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6343 (0.6130)\tPrec@1 87.500 (86.824)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6530 (0.6131)\tPrec@1 85.938 (86.842)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6597 (0.6126)\tPrec@1 85.156 (86.896)\n",
      "[260/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6410 (0.6138)\tPrec@1 85.938 (86.874)\n",
      "[270/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.5629 (0.6132)\tPrec@1 89.062 (86.897)\n",
      "[280/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7045 (0.6134)\tPrec@1 82.812 (86.886)\n",
      "[290/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.5532 (0.6129)\tPrec@1 91.406 (86.907)\n",
      "[300/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5830 (0.6133)\tPrec@1 87.500 (86.890)\n",
      "[310/391]\tTime 0.154 (0.181)\tData 0.020 (0.020)\tLoss 0.5806 (0.6142)\tPrec@1 92.188 (86.847)\n",
      "[320/391]\tTime 0.144 (0.180)\tData 0.019 (0.020)\tLoss 0.5949 (0.6148)\tPrec@1 87.500 (86.811)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.5927 (0.6158)\tPrec@1 88.281 (86.752)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6077 (0.6166)\tPrec@1 88.281 (86.714)\n",
      "[350/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5502 (0.6165)\tPrec@1 87.500 (86.719)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5755 (0.6158)\tPrec@1 91.406 (86.784)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5746 (0.6162)\tPrec@1 89.062 (86.765)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7501 (0.6181)\tPrec@1 79.688 (86.717)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.7177 (0.6181)\tPrec@1 82.500 (86.722)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.0032 (1.0032)\tPrec@1 84.375 (84.375)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.7514 (0.9321)\tPrec@1 85.938 (82.670)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.9504 (1.0091)\tPrec@1 81.250 (81.994)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.6483 (0.9964)\tPrec@1 85.156 (81.930)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 0.9595 (0.9831)\tPrec@1 75.781 (82.222)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 0.9005 (0.9635)\tPrec@1 80.469 (82.430)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.8688 (0.9779)\tPrec@1 83.594 (82.198)\n",
      "Test: [70/79]\tTime 0.020 (0.023)\tLoss 0.7722 (0.9807)\tPrec@1 88.281 (82.240)\n",
      " * Prec@1 82.340\n",
      "\n",
      "===> epoch: 31/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.168 (0.168)\tData 0.020 (0.020)\tLoss 0.6451 (0.6451)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5772 (0.6138)\tPrec@1 87.500 (87.216)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.7217 (0.6015)\tPrec@1 82.031 (87.574)\n",
      "[30/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6614 (0.6125)\tPrec@1 81.250 (87.046)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6491 (0.6140)\tPrec@1 84.375 (86.814)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6439 (0.6140)\tPrec@1 84.375 (86.826)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5505 (0.6141)\tPrec@1 92.188 (86.949)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7415 (0.6163)\tPrec@1 81.250 (86.840)\n",
      "[80/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6394 (0.6195)\tPrec@1 84.375 (86.690)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6552 (0.6160)\tPrec@1 85.938 (86.839)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6298 (0.6169)\tPrec@1 83.594 (86.850)\n",
      "[110/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.7791 (0.6221)\tPrec@1 80.469 (86.712)\n",
      "[120/391]\tTime 0.193 (0.182)\tData 0.019 (0.020)\tLoss 0.6429 (0.6245)\tPrec@1 85.156 (86.570)\n",
      "[130/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6967 (0.6273)\tPrec@1 80.469 (86.427)\n",
      "[140/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6393 (0.6306)\tPrec@1 88.281 (86.353)\n",
      "[150/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5922 (0.6308)\tPrec@1 88.281 (86.382)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6010 (0.6299)\tPrec@1 84.375 (86.432)\n",
      "[170/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5387 (0.6308)\tPrec@1 89.062 (86.349)\n",
      "[180/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6512 (0.6307)\tPrec@1 83.594 (86.365)\n",
      "[190/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6939 (0.6307)\tPrec@1 79.688 (86.359)\n",
      "[200/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.7223 (0.6322)\tPrec@1 77.344 (86.276)\n",
      "[210/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.6338 (0.6315)\tPrec@1 88.281 (86.323)\n",
      "[220/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.7001 (0.6304)\tPrec@1 79.688 (86.312)\n",
      "[230/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7238 (0.6315)\tPrec@1 82.812 (86.266)\n",
      "[240/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5808 (0.6300)\tPrec@1 89.062 (86.304)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6056 (0.6296)\tPrec@1 87.500 (86.286)\n",
      "[260/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.7000 (0.6301)\tPrec@1 86.719 (86.279)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6196 (0.6302)\tPrec@1 86.719 (86.278)\n",
      "[280/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5908 (0.6298)\tPrec@1 89.062 (86.299)\n",
      "[290/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6598 (0.6300)\tPrec@1 82.812 (86.281)\n",
      "[300/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5918 (0.6308)\tPrec@1 89.062 (86.280)\n",
      "[310/391]\tTime 0.143 (0.181)\tData 0.019 (0.020)\tLoss 0.5374 (0.6318)\tPrec@1 91.406 (86.209)\n",
      "[320/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6092 (0.6331)\tPrec@1 89.062 (86.186)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6095 (0.6326)\tPrec@1 88.281 (86.199)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5371 (0.6319)\tPrec@1 90.625 (86.242)\n",
      "[350/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.8073 (0.6307)\tPrec@1 78.125 (86.269)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6921 (0.6303)\tPrec@1 85.938 (86.273)\n",
      "[370/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5636 (0.6296)\tPrec@1 86.719 (86.306)\n",
      "[380/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5848 (0.6295)\tPrec@1 88.281 (86.333)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.013 (0.020)\tLoss 0.5291 (0.6298)\tPrec@1 88.750 (86.316)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9168 (0.9168)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 0.9150 (1.0610)\tPrec@1 81.250 (79.119)\n",
      "Test: [20/79]\tTime 0.025 (0.022)\tLoss 1.3982 (1.1568)\tPrec@1 78.906 (78.832)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.8631 (1.1620)\tPrec@1 85.156 (78.730)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.1854 (1.1295)\tPrec@1 76.562 (79.535)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.0544 (1.0999)\tPrec@1 78.906 (79.825)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.0253 (1.0918)\tPrec@1 85.938 (79.956)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.0159 (1.0838)\tPrec@1 82.031 (80.150)\n",
      " * Prec@1 80.180\n",
      "\n",
      "===> epoch: 32/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.164 (0.164)\tData 0.020 (0.020)\tLoss 0.6935 (0.6935)\tPrec@1 80.469 (80.469)\n",
      "[10/391]\tTime 0.184 (0.180)\tData 0.019 (0.019)\tLoss 0.5150 (0.6004)\tPrec@1 88.281 (87.571)\n",
      "[20/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5863 (0.6132)\tPrec@1 86.719 (86.979)\n",
      "[30/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6893 (0.6300)\tPrec@1 85.938 (86.139)\n",
      "[40/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5402 (0.6219)\tPrec@1 91.406 (86.433)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6510 (0.6193)\tPrec@1 83.594 (86.489)\n",
      "[60/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5475 (0.6174)\tPrec@1 89.844 (86.450)\n",
      "[70/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6645 (0.6149)\tPrec@1 85.156 (86.642)\n",
      "[80/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5834 (0.6130)\tPrec@1 86.719 (86.690)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4834 (0.6120)\tPrec@1 89.844 (86.787)\n",
      "[100/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5420 (0.6075)\tPrec@1 89.844 (86.966)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6582 (0.6112)\tPrec@1 86.719 (86.930)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5669 (0.6112)\tPrec@1 86.719 (86.867)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6930 (0.6124)\tPrec@1 81.250 (86.796)\n",
      "[140/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5681 (0.6147)\tPrec@1 89.844 (86.818)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6685 (0.6129)\tPrec@1 85.156 (86.884)\n",
      "[160/391]\tTime 0.185 (0.181)\tData 0.026 (0.020)\tLoss 0.6552 (0.6137)\tPrec@1 88.281 (86.864)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6982 (0.6144)\tPrec@1 81.250 (86.833)\n",
      "[180/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.6551 (0.6178)\tPrec@1 85.938 (86.676)\n",
      "[190/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6462 (0.6208)\tPrec@1 85.156 (86.535)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5766 (0.6195)\tPrec@1 87.500 (86.598)\n",
      "[210/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5837 (0.6185)\tPrec@1 87.500 (86.615)\n",
      "[220/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7026 (0.6203)\tPrec@1 85.156 (86.524)\n",
      "[230/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6138 (0.6195)\tPrec@1 86.719 (86.590)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6375 (0.6189)\tPrec@1 85.156 (86.615)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6250 (0.6164)\tPrec@1 86.719 (86.678)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5805 (0.6169)\tPrec@1 87.500 (86.674)\n",
      "[270/391]\tTime 0.183 (0.182)\tData 0.022 (0.020)\tLoss 0.6767 (0.6162)\tPrec@1 84.375 (86.673)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5809 (0.6161)\tPrec@1 90.625 (86.691)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6377 (0.6171)\tPrec@1 87.500 (86.646)\n",
      "[300/391]\tTime 0.145 (0.181)\tData 0.020 (0.020)\tLoss 0.5372 (0.6165)\tPrec@1 89.844 (86.688)\n",
      "[310/391]\tTime 0.186 (0.180)\tData 0.019 (0.020)\tLoss 0.4753 (0.6168)\tPrec@1 89.844 (86.679)\n",
      "[320/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6504 (0.6170)\tPrec@1 87.500 (86.660)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.7370 (0.6176)\tPrec@1 80.469 (86.627)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6800 (0.6188)\tPrec@1 85.938 (86.604)\n",
      "[350/391]\tTime 0.181 (0.180)\tData 0.022 (0.020)\tLoss 0.6646 (0.6192)\tPrec@1 86.719 (86.592)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6513 (0.6201)\tPrec@1 86.719 (86.576)\n",
      "[370/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6441 (0.6204)\tPrec@1 85.156 (86.588)\n",
      "[380/391]\tTime 0.187 (0.180)\tData 0.020 (0.020)\tLoss 0.6754 (0.6199)\tPrec@1 81.250 (86.612)\n",
      "[390/391]\tTime 0.132 (0.180)\tData 0.013 (0.020)\tLoss 0.5949 (0.6208)\tPrec@1 86.250 (86.546)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.3915 (1.3915)\tPrec@1 73.438 (73.438)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.7621 (1.6875)\tPrec@1 71.094 (72.443)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 1.5605 (1.8021)\tPrec@1 77.344 (71.280)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 1.6253 (1.8096)\tPrec@1 72.656 (71.472)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 1.7743 (1.7683)\tPrec@1 67.969 (71.551)\n",
      "Test: [50/79]\tTime 0.023 (0.023)\tLoss 1.5058 (1.7620)\tPrec@1 75.781 (71.645)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.5947 (1.7671)\tPrec@1 71.875 (71.734)\n",
      "Test: [70/79]\tTime 0.020 (0.023)\tLoss 1.6239 (1.7758)\tPrec@1 70.312 (71.490)\n",
      " * Prec@1 71.560\n",
      "\n",
      "===> epoch: 33/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.169 (0.169)\tData 0.021 (0.021)\tLoss 0.6315 (0.6315)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.188 (0.180)\tData 0.021 (0.020)\tLoss 0.6895 (0.6205)\tPrec@1 83.594 (86.222)\n",
      "[20/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5969 (0.6289)\tPrec@1 91.406 (86.272)\n",
      "[30/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5897 (0.6295)\tPrec@1 86.719 (86.341)\n",
      "[40/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6948 (0.6349)\tPrec@1 82.812 (85.880)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6791 (0.6387)\tPrec@1 84.375 (85.723)\n",
      "[60/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5465 (0.6294)\tPrec@1 90.625 (86.014)\n",
      "[70/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6615 (0.6287)\tPrec@1 85.156 (85.982)\n",
      "[80/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6604 (0.6281)\tPrec@1 84.375 (86.034)\n",
      "[90/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.7539 (0.6307)\tPrec@1 82.812 (85.920)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6657 (0.6347)\tPrec@1 85.938 (85.767)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6179 (0.6334)\tPrec@1 89.062 (85.909)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5747 (0.6341)\tPrec@1 86.719 (85.776)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6904 (0.6351)\tPrec@1 81.250 (85.782)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6283 (0.6334)\tPrec@1 85.156 (85.871)\n",
      "[150/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6314 (0.6305)\tPrec@1 87.500 (85.953)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5259 (0.6275)\tPrec@1 88.281 (86.049)\n",
      "[170/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.5177 (0.6255)\tPrec@1 92.969 (86.143)\n",
      "[180/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5790 (0.6235)\tPrec@1 87.500 (86.235)\n",
      "[190/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5814 (0.6237)\tPrec@1 89.844 (86.240)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5444 (0.6208)\tPrec@1 90.625 (86.361)\n",
      "[210/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6293 (0.6204)\tPrec@1 86.719 (86.386)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4793 (0.6198)\tPrec@1 94.531 (86.408)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6384 (0.6190)\tPrec@1 87.500 (86.445)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6570 (0.6211)\tPrec@1 85.938 (86.356)\n",
      "[250/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6730 (0.6204)\tPrec@1 84.375 (86.370)\n",
      "[260/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6459 (0.6218)\tPrec@1 86.719 (86.348)\n",
      "[270/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6934 (0.6215)\tPrec@1 85.156 (86.379)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6283 (0.6198)\tPrec@1 86.719 (86.407)\n",
      "[290/391]\tTime 0.145 (0.181)\tData 0.020 (0.020)\tLoss 0.6480 (0.6202)\tPrec@1 85.156 (86.383)\n",
      "[300/391]\tTime 0.188 (0.180)\tData 0.020 (0.020)\tLoss 0.5703 (0.6210)\tPrec@1 87.500 (86.363)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.5680 (0.6213)\tPrec@1 89.844 (86.377)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.7502 (0.6221)\tPrec@1 82.812 (86.339)\n",
      "[330/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6791 (0.6225)\tPrec@1 82.031 (86.334)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7108 (0.6234)\tPrec@1 81.250 (86.290)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5474 (0.6233)\tPrec@1 90.625 (86.300)\n",
      "[360/391]\tTime 0.190 (0.180)\tData 0.029 (0.020)\tLoss 0.6664 (0.6235)\tPrec@1 85.156 (86.292)\n",
      "[370/391]\tTime 0.191 (0.180)\tData 0.029 (0.020)\tLoss 0.6481 (0.6240)\tPrec@1 83.594 (86.268)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6189 (0.6238)\tPrec@1 85.938 (86.284)\n",
      "[390/391]\tTime 0.129 (0.180)\tData 0.013 (0.020)\tLoss 0.5567 (0.6245)\tPrec@1 90.000 (86.258)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9505 (0.9505)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.7940 (0.9657)\tPrec@1 83.594 (81.321)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0439 (1.0219)\tPrec@1 82.031 (81.250)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.8163 (1.0183)\tPrec@1 82.812 (81.376)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 0.7773 (0.9996)\tPrec@1 80.469 (81.307)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 0.9336 (0.9905)\tPrec@1 83.594 (81.373)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.5700 (1.0004)\tPrec@1 88.281 (81.250)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 1.0605 (0.9929)\tPrec@1 80.469 (81.316)\n",
      " * Prec@1 81.370\n",
      "\n",
      "===> epoch: 34/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.159 (0.159)\tData 0.020 (0.020)\tLoss 0.5242 (0.5242)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.183 (0.179)\tData 0.019 (0.019)\tLoss 0.5739 (0.6182)\tPrec@1 88.281 (87.571)\n",
      "[20/391]\tTime 0.179 (0.179)\tData 0.020 (0.019)\tLoss 0.6747 (0.6282)\tPrec@1 83.594 (87.091)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5586 (0.6212)\tPrec@1 89.062 (86.920)\n",
      "[40/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5489 (0.6121)\tPrec@1 90.625 (86.871)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6091 (0.6122)\tPrec@1 88.281 (86.933)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5307 (0.6088)\tPrec@1 89.062 (87.026)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.7096 (0.6109)\tPrec@1 87.500 (86.851)\n",
      "[80/391]\tTime 0.195 (0.181)\tData 0.028 (0.020)\tLoss 0.5677 (0.6092)\tPrec@1 86.719 (86.844)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5544 (0.6081)\tPrec@1 86.719 (86.813)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5505 (0.6073)\tPrec@1 88.281 (86.812)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5878 (0.6081)\tPrec@1 84.375 (86.810)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5525 (0.6064)\tPrec@1 92.188 (86.867)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5501 (0.6087)\tPrec@1 88.281 (86.755)\n",
      "[140/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6420 (0.6097)\tPrec@1 85.938 (86.758)\n",
      "[150/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5549 (0.6072)\tPrec@1 85.156 (86.833)\n",
      "[160/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6255 (0.6075)\tPrec@1 85.156 (86.801)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6657 (0.6100)\tPrec@1 85.156 (86.755)\n",
      "[180/391]\tTime 0.187 (0.181)\tData 0.023 (0.020)\tLoss 0.6177 (0.6101)\tPrec@1 86.719 (86.723)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.024 (0.020)\tLoss 0.5747 (0.6100)\tPrec@1 88.281 (86.756)\n",
      "[200/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5587 (0.6113)\tPrec@1 87.500 (86.684)\n",
      "[210/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5863 (0.6113)\tPrec@1 89.062 (86.674)\n",
      "[220/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6605 (0.6112)\tPrec@1 85.938 (86.680)\n",
      "[230/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5982 (0.6115)\tPrec@1 85.938 (86.638)\n",
      "[240/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5402 (0.6115)\tPrec@1 88.281 (86.657)\n",
      "[250/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5388 (0.6117)\tPrec@1 89.844 (86.628)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5813 (0.6131)\tPrec@1 90.625 (86.614)\n",
      "[270/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.7081 (0.6148)\tPrec@1 85.156 (86.592)\n",
      "[280/391]\tTime 0.145 (0.181)\tData 0.021 (0.020)\tLoss 0.6473 (0.6150)\tPrec@1 85.938 (86.571)\n",
      "[290/391]\tTime 0.150 (0.180)\tData 0.020 (0.020)\tLoss 0.6677 (0.6148)\tPrec@1 88.281 (86.590)\n",
      "[300/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6320 (0.6156)\tPrec@1 88.281 (86.573)\n",
      "[310/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.6166 (0.6154)\tPrec@1 86.719 (86.583)\n",
      "[320/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.4813 (0.6145)\tPrec@1 91.406 (86.599)\n",
      "[330/391]\tTime 0.186 (0.180)\tData 0.019 (0.020)\tLoss 0.6740 (0.6151)\tPrec@1 85.938 (86.579)\n",
      "[340/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6405 (0.6155)\tPrec@1 89.062 (86.600)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5835 (0.6158)\tPrec@1 87.500 (86.567)\n",
      "[360/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5529 (0.6166)\tPrec@1 89.062 (86.541)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5974 (0.6163)\tPrec@1 87.500 (86.567)\n",
      "[380/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6398 (0.6165)\tPrec@1 87.500 (86.544)\n",
      "[390/391]\tTime 0.133 (0.180)\tData 0.012 (0.020)\tLoss 0.5715 (0.6177)\tPrec@1 88.750 (86.530)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9598 (0.9598)\tPrec@1 82.031 (82.031)\n",
      "Test: [10/79]\tTime 0.023 (0.023)\tLoss 0.7074 (0.9245)\tPrec@1 85.156 (82.031)\n",
      "Test: [20/79]\tTime 0.019 (0.022)\tLoss 0.9659 (0.9787)\tPrec@1 78.906 (80.618)\n",
      "Test: [30/79]\tTime 0.022 (0.022)\tLoss 1.0060 (0.9648)\tPrec@1 83.594 (81.023)\n",
      "Test: [40/79]\tTime 0.026 (0.022)\tLoss 0.8656 (0.9438)\tPrec@1 78.906 (81.421)\n",
      "Test: [50/79]\tTime 0.028 (0.022)\tLoss 1.1782 (0.9383)\tPrec@1 75.781 (81.173)\n",
      "Test: [60/79]\tTime 0.019 (0.023)\tLoss 0.7632 (0.9333)\tPrec@1 83.594 (81.288)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.0018 (0.9357)\tPrec@1 82.031 (81.305)\n",
      " * Prec@1 81.430\n",
      "\n",
      "===> epoch: 35/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.175 (0.175)\tData 0.020 (0.020)\tLoss 0.6756 (0.6756)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.5639 (0.6178)\tPrec@1 89.062 (87.216)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.4860 (0.6136)\tPrec@1 92.188 (87.277)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.5528 (0.5981)\tPrec@1 87.500 (87.576)\n",
      "[40/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5883 (0.6006)\tPrec@1 88.281 (87.481)\n",
      "[50/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6646 (0.6062)\tPrec@1 84.375 (87.178)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[60/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.5593 (0.6116)\tPrec@1 88.281 (86.872)\n",
      "[70/391]\tTime 0.188 (0.182)\tData 0.019 (0.019)\tLoss 0.6444 (0.6124)\tPrec@1 85.938 (86.763)\n",
      "[80/391]\tTime 0.191 (0.182)\tData 0.020 (0.019)\tLoss 0.5456 (0.6131)\tPrec@1 89.062 (86.786)\n",
      "[90/391]\tTime 0.183 (0.183)\tData 0.020 (0.019)\tLoss 0.5787 (0.6103)\tPrec@1 89.062 (86.993)\n",
      "[100/391]\tTime 0.183 (0.183)\tData 0.019 (0.019)\tLoss 0.5968 (0.6114)\tPrec@1 87.500 (86.866)\n",
      "[110/391]\tTime 0.183 (0.183)\tData 0.019 (0.019)\tLoss 0.5701 (0.6114)\tPrec@1 89.062 (86.909)\n",
      "[120/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6918 (0.6145)\tPrec@1 82.031 (86.790)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6064 (0.6170)\tPrec@1 85.156 (86.677)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.6009 (0.6165)\tPrec@1 86.719 (86.791)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.022 (0.019)\tLoss 0.4965 (0.6166)\tPrec@1 91.406 (86.791)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6417 (0.6171)\tPrec@1 82.812 (86.840)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6099 (0.6169)\tPrec@1 85.156 (86.892)\n",
      "[180/391]\tTime 0.186 (0.182)\tData 0.019 (0.019)\tLoss 0.5597 (0.6175)\tPrec@1 90.625 (86.896)\n",
      "[190/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.7014 (0.6170)\tPrec@1 83.594 (86.915)\n",
      "[200/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.7987 (0.6170)\tPrec@1 81.250 (86.909)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.4735 (0.6165)\tPrec@1 91.406 (86.926)\n",
      "[220/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5496 (0.6172)\tPrec@1 91.406 (86.874)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.7222 (0.6181)\tPrec@1 83.594 (86.813)\n",
      "[240/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5513 (0.6170)\tPrec@1 88.281 (86.842)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6296 (0.6160)\tPrec@1 84.375 (86.887)\n",
      "[260/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6463 (0.6152)\tPrec@1 85.156 (86.913)\n",
      "[270/391]\tTime 0.144 (0.182)\tData 0.019 (0.020)\tLoss 0.6333 (0.6146)\tPrec@1 85.156 (86.952)\n",
      "[280/391]\tTime 0.146 (0.181)\tData 0.020 (0.020)\tLoss 0.6592 (0.6148)\tPrec@1 82.031 (86.894)\n",
      "[290/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5993 (0.6145)\tPrec@1 90.625 (86.880)\n",
      "[300/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6180 (0.6159)\tPrec@1 85.156 (86.830)\n",
      "[310/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6604 (0.6167)\tPrec@1 82.812 (86.764)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6509 (0.6181)\tPrec@1 85.156 (86.704)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6855 (0.6181)\tPrec@1 87.500 (86.698)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.7696 (0.6198)\tPrec@1 79.688 (86.636)\n",
      "[350/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6012 (0.6196)\tPrec@1 89.062 (86.643)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5716 (0.6189)\tPrec@1 85.156 (86.673)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6119 (0.6177)\tPrec@1 85.156 (86.702)\n",
      "[380/391]\tTime 0.178 (0.180)\tData 0.020 (0.020)\tLoss 0.5884 (0.6178)\tPrec@1 85.938 (86.717)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.012 (0.020)\tLoss 0.5546 (0.6177)\tPrec@1 88.750 (86.724)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7772 (0.7772)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.8188 (0.9455)\tPrec@1 82.812 (81.108)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0723 (1.0016)\tPrec@1 83.594 (80.729)\n",
      "Test: [30/79]\tTime 0.024 (0.023)\tLoss 0.8156 (1.0183)\tPrec@1 85.156 (80.670)\n",
      "Test: [40/79]\tTime 0.026 (0.023)\tLoss 1.1021 (1.0015)\tPrec@1 79.688 (80.678)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 0.8883 (0.9849)\tPrec@1 81.250 (80.683)\n",
      "Test: [60/79]\tTime 0.019 (0.023)\tLoss 0.9896 (0.9904)\tPrec@1 83.594 (80.674)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.0491 (0.9969)\tPrec@1 82.031 (80.546)\n",
      " * Prec@1 80.450\n",
      "\n",
      "===> epoch: 36/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.022 (0.022)\tLoss 0.6996 (0.6996)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6892 (0.6425)\tPrec@1 82.812 (85.156)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5104 (0.6134)\tPrec@1 88.281 (87.091)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6599 (0.6101)\tPrec@1 84.375 (87.399)\n",
      "[40/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6424 (0.6170)\tPrec@1 85.156 (87.043)\n",
      "[50/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6268 (0.6161)\tPrec@1 85.938 (86.841)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5232 (0.6113)\tPrec@1 89.844 (86.988)\n",
      "[70/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.5816 (0.6144)\tPrec@1 87.500 (86.829)\n",
      "[80/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6592 (0.6210)\tPrec@1 85.156 (86.526)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6729 (0.6215)\tPrec@1 85.156 (86.599)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.5266 (0.6243)\tPrec@1 90.625 (86.394)\n",
      "[110/391]\tTime 0.185 (0.181)\tData 0.031 (0.020)\tLoss 0.6856 (0.6227)\tPrec@1 83.594 (86.395)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6879 (0.6222)\tPrec@1 86.719 (86.473)\n",
      "[130/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4723 (0.6192)\tPrec@1 94.531 (86.594)\n",
      "[140/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6730 (0.6192)\tPrec@1 81.250 (86.641)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6817 (0.6185)\tPrec@1 85.156 (86.667)\n",
      "[160/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5329 (0.6199)\tPrec@1 89.062 (86.602)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6870 (0.6200)\tPrec@1 82.031 (86.582)\n",
      "[180/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6191 (0.6190)\tPrec@1 87.500 (86.563)\n",
      "[190/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5453 (0.6174)\tPrec@1 91.406 (86.625)\n",
      "[200/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6586 (0.6177)\tPrec@1 82.812 (86.587)\n",
      "[210/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6738 (0.6176)\tPrec@1 83.594 (86.597)\n",
      "[220/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.7240 (0.6183)\tPrec@1 83.594 (86.588)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5506 (0.6182)\tPrec@1 89.844 (86.600)\n",
      "[240/391]\tTime 0.189 (0.181)\tData 0.020 (0.020)\tLoss 0.6425 (0.6187)\tPrec@1 82.812 (86.563)\n",
      "[250/391]\tTime 0.184 (0.181)\tData 0.023 (0.020)\tLoss 0.6455 (0.6188)\tPrec@1 85.156 (86.560)\n",
      "[260/391]\tTime 0.148 (0.181)\tData 0.019 (0.020)\tLoss 0.6275 (0.6193)\tPrec@1 89.844 (86.584)\n",
      "[270/391]\tTime 0.145 (0.179)\tData 0.020 (0.020)\tLoss 0.5857 (0.6204)\tPrec@1 87.500 (86.577)\n",
      "[280/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6069 (0.6203)\tPrec@1 85.938 (86.585)\n",
      "[290/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6908 (0.6215)\tPrec@1 84.375 (86.585)\n",
      "[300/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.7019 (0.6222)\tPrec@1 84.375 (86.560)\n",
      "[310/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5836 (0.6227)\tPrec@1 86.719 (86.538)\n",
      "[320/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6059 (0.6221)\tPrec@1 87.500 (86.573)\n",
      "[330/391]\tTime 0.188 (0.179)\tData 0.020 (0.020)\tLoss 0.5353 (0.6202)\tPrec@1 91.406 (86.634)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5727 (0.6207)\tPrec@1 88.281 (86.634)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5898 (0.6207)\tPrec@1 85.156 (86.605)\n",
      "[360/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.6969 (0.6203)\tPrec@1 82.031 (86.617)\n",
      "[370/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6366 (0.6213)\tPrec@1 86.719 (86.569)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6100 (0.6218)\tPrec@1 84.375 (86.563)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.5677 (0.6225)\tPrec@1 90.000 (86.556)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9995 (0.9995)\tPrec@1 75.781 (75.781)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 1.1279 (1.2101)\tPrec@1 82.031 (77.983)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.2921 (1.2301)\tPrec@1 75.781 (77.455)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.8882 (1.2084)\tPrec@1 82.031 (78.024)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.3537 (1.2028)\tPrec@1 73.438 (78.182)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.0916 (1.2075)\tPrec@1 77.344 (78.094)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.9740 (1.1852)\tPrec@1 78.125 (78.394)\n",
      "Test: [70/79]\tTime 0.020 (0.023)\tLoss 1.0386 (1.1810)\tPrec@1 80.469 (78.345)\n",
      " * Prec@1 78.500\n",
      "\n",
      "===> epoch: 37/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.161 (0.161)\tData 0.020 (0.020)\tLoss 0.6893 (0.6893)\tPrec@1 79.688 (79.688)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10/391]\tTime 0.182 (0.179)\tData 0.027 (0.020)\tLoss 0.6467 (0.6316)\tPrec@1 86.719 (85.298)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6551 (0.6273)\tPrec@1 86.719 (86.124)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5121 (0.6144)\tPrec@1 90.625 (86.668)\n",
      "[40/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6159 (0.6097)\tPrec@1 87.500 (86.966)\n",
      "[50/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.7556 (0.6134)\tPrec@1 82.812 (86.887)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6136 (0.6141)\tPrec@1 85.156 (86.796)\n",
      "[70/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6153 (0.6140)\tPrec@1 85.938 (86.884)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6151 (0.6145)\tPrec@1 85.156 (86.728)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7604 (0.6200)\tPrec@1 80.469 (86.487)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.6869 (0.6205)\tPrec@1 82.812 (86.409)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6313 (0.6198)\tPrec@1 85.938 (86.458)\n",
      "[120/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.7102 (0.6203)\tPrec@1 82.812 (86.415)\n",
      "[130/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.6866 (0.6211)\tPrec@1 85.156 (86.397)\n",
      "[140/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6168 (0.6213)\tPrec@1 87.500 (86.431)\n",
      "[150/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.4656 (0.6190)\tPrec@1 93.750 (86.579)\n",
      "[160/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6012 (0.6179)\tPrec@1 86.719 (86.607)\n",
      "[170/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6529 (0.6196)\tPrec@1 88.281 (86.563)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6850 (0.6203)\tPrec@1 84.375 (86.503)\n",
      "[190/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6229 (0.6200)\tPrec@1 86.719 (86.543)\n",
      "[200/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.7019 (0.6187)\tPrec@1 85.156 (86.536)\n",
      "[210/391]\tTime 0.182 (0.182)\tData 0.022 (0.020)\tLoss 0.6206 (0.6180)\tPrec@1 89.062 (86.585)\n",
      "[220/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.7390 (0.6185)\tPrec@1 84.375 (86.560)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6265 (0.6185)\tPrec@1 88.281 (86.627)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4786 (0.6176)\tPrec@1 94.531 (86.716)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6755 (0.6171)\tPrec@1 80.469 (86.709)\n",
      "[260/391]\tTime 0.142 (0.180)\tData 0.019 (0.020)\tLoss 0.6496 (0.6169)\tPrec@1 85.938 (86.737)\n",
      "[270/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6650 (0.6158)\tPrec@1 84.375 (86.759)\n",
      "[280/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6282 (0.6154)\tPrec@1 80.469 (86.738)\n",
      "[290/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5825 (0.6145)\tPrec@1 88.281 (86.786)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6454 (0.6153)\tPrec@1 84.375 (86.763)\n",
      "[310/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.7296 (0.6163)\tPrec@1 82.031 (86.709)\n",
      "[320/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5976 (0.6154)\tPrec@1 90.625 (86.767)\n",
      "[330/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.7096 (0.6169)\tPrec@1 82.812 (86.707)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6105 (0.6168)\tPrec@1 86.719 (86.684)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6330 (0.6165)\tPrec@1 82.812 (86.681)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5905 (0.6159)\tPrec@1 89.062 (86.695)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7112 (0.6168)\tPrec@1 83.594 (86.675)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5239 (0.6167)\tPrec@1 92.188 (86.674)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.5925 (0.6164)\tPrec@1 90.000 (86.690)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.0081 (1.0081)\tPrec@1 79.688 (79.688)\n",
      "Test: [10/79]\tTime 0.025 (0.023)\tLoss 0.7847 (1.0003)\tPrec@1 85.938 (80.256)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.7937 (1.0217)\tPrec@1 84.375 (80.543)\n",
      "Test: [30/79]\tTime 0.020 (0.023)\tLoss 0.6560 (1.0373)\tPrec@1 88.281 (80.645)\n",
      "Test: [40/79]\tTime 0.022 (0.023)\tLoss 1.0524 (1.0246)\tPrec@1 78.906 (80.964)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.0870 (0.9968)\tPrec@1 81.250 (81.327)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.7519 (1.0031)\tPrec@1 87.500 (81.250)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.1340 (1.0075)\tPrec@1 82.031 (81.173)\n",
      " * Prec@1 81.440\n",
      "\n",
      "===> epoch: 38/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.160 (0.160)\tData 0.022 (0.022)\tLoss 0.5711 (0.5711)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6491 (0.6336)\tPrec@1 80.469 (84.801)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6166 (0.6377)\tPrec@1 85.938 (85.528)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5008 (0.6204)\tPrec@1 91.406 (86.442)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6718 (0.6250)\tPrec@1 85.156 (86.376)\n",
      "[50/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5634 (0.6168)\tPrec@1 89.062 (86.749)\n",
      "[60/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5801 (0.6081)\tPrec@1 88.281 (87.077)\n",
      "[70/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6020 (0.6060)\tPrec@1 85.938 (87.137)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6326 (0.6078)\tPrec@1 82.812 (86.960)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6878 (0.6090)\tPrec@1 83.594 (86.942)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6011 (0.6098)\tPrec@1 89.844 (86.966)\n",
      "[110/391]\tTime 0.191 (0.182)\tData 0.020 (0.020)\tLoss 0.6049 (0.6081)\tPrec@1 89.844 (87.064)\n",
      "[120/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6968 (0.6072)\tPrec@1 84.375 (87.106)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5376 (0.6084)\tPrec@1 89.844 (87.035)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5629 (0.6093)\tPrec@1 89.844 (86.979)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6174 (0.6104)\tPrec@1 87.500 (86.952)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6225 (0.6122)\tPrec@1 90.625 (86.884)\n",
      "[170/391]\tTime 0.191 (0.182)\tData 0.020 (0.020)\tLoss 0.7059 (0.6140)\tPrec@1 83.594 (86.828)\n",
      "[180/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6244 (0.6136)\tPrec@1 84.375 (86.822)\n",
      "[190/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5487 (0.6101)\tPrec@1 90.625 (86.989)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6225 (0.6105)\tPrec@1 87.500 (86.929)\n",
      "[210/391]\tTime 0.179 (0.182)\tData 0.018 (0.020)\tLoss 0.5885 (0.6104)\tPrec@1 87.500 (86.896)\n",
      "[220/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5641 (0.6107)\tPrec@1 87.500 (86.864)\n",
      "[230/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7199 (0.6112)\tPrec@1 83.594 (86.810)\n",
      "[240/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6092 (0.6127)\tPrec@1 86.719 (86.735)\n",
      "[250/391]\tTime 0.143 (0.181)\tData 0.019 (0.020)\tLoss 0.6114 (0.6143)\tPrec@1 86.719 (86.697)\n",
      "[260/391]\tTime 0.185 (0.180)\tData 0.018 (0.020)\tLoss 0.6692 (0.6143)\tPrec@1 86.719 (86.725)\n",
      "[270/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.5535 (0.6149)\tPrec@1 90.625 (86.722)\n",
      "[280/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5527 (0.6133)\tPrec@1 87.500 (86.785)\n",
      "[290/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.7303 (0.6140)\tPrec@1 82.812 (86.754)\n",
      "[300/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.7148 (0.6147)\tPrec@1 84.375 (86.747)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5508 (0.6157)\tPrec@1 89.062 (86.694)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5972 (0.6155)\tPrec@1 87.500 (86.714)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6406 (0.6161)\tPrec@1 87.500 (86.669)\n",
      "[340/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6224 (0.6163)\tPrec@1 88.281 (86.700)\n",
      "[350/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6977 (0.6175)\tPrec@1 82.031 (86.661)\n",
      "[360/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.6124 (0.6189)\tPrec@1 84.375 (86.598)\n",
      "[370/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6104 (0.6198)\tPrec@1 86.719 (86.578)\n",
      "[380/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7253 (0.6207)\tPrec@1 84.375 (86.559)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.5722 (0.6208)\tPrec@1 88.750 (86.558)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9781 (0.9781)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.020 (0.023)\tLoss 0.6731 (1.0822)\tPrec@1 86.719 (80.611)\n",
      "Test: [20/79]\tTime 0.021 (0.023)\tLoss 1.1456 (1.0947)\tPrec@1 82.812 (80.208)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.9191 (1.1003)\tPrec@1 85.156 (80.469)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.024 (0.023)\tLoss 1.3067 (1.0956)\tPrec@1 75.781 (80.716)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 1.1852 (1.0818)\tPrec@1 77.344 (80.484)\n",
      "Test: [60/79]\tTime 0.021 (0.023)\tLoss 1.0155 (1.0690)\tPrec@1 82.812 (80.558)\n",
      "Test: [70/79]\tTime 0.023 (0.023)\tLoss 1.1891 (1.0661)\tPrec@1 80.469 (80.546)\n",
      " * Prec@1 80.520\n",
      "\n",
      "===> epoch: 39/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6302 (0.6302)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5862 (0.6005)\tPrec@1 89.844 (87.855)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.023 (0.020)\tLoss 0.6615 (0.6105)\tPrec@1 85.156 (87.760)\n",
      "[30/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5876 (0.6157)\tPrec@1 89.844 (87.273)\n",
      "[40/391]\tTime 0.178 (0.180)\tData 0.021 (0.020)\tLoss 0.6677 (0.6114)\tPrec@1 82.031 (86.947)\n",
      "[50/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5549 (0.6074)\tPrec@1 90.625 (87.040)\n",
      "[60/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5130 (0.6070)\tPrec@1 93.750 (87.308)\n",
      "[70/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6564 (0.6032)\tPrec@1 85.938 (87.291)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6767 (0.6053)\tPrec@1 85.156 (87.240)\n",
      "[90/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6352 (0.6054)\tPrec@1 85.938 (87.217)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6348 (0.6052)\tPrec@1 85.938 (87.283)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6435 (0.6074)\tPrec@1 89.062 (87.183)\n",
      "[120/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.6049 (0.6061)\tPrec@1 88.281 (87.209)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5657 (0.6054)\tPrec@1 87.500 (87.220)\n",
      "[140/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6466 (0.6069)\tPrec@1 88.281 (87.168)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6898 (0.6104)\tPrec@1 83.594 (86.977)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7047 (0.6117)\tPrec@1 82.031 (86.913)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.6381 (0.6104)\tPrec@1 89.844 (86.979)\n",
      "[180/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6573 (0.6100)\tPrec@1 85.156 (87.012)\n",
      "[190/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5721 (0.6092)\tPrec@1 90.625 (87.021)\n",
      "[200/391]\tTime 0.177 (0.181)\tData 0.020 (0.020)\tLoss 0.5095 (0.6081)\tPrec@1 91.406 (87.053)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6228 (0.6089)\tPrec@1 86.719 (87.004)\n",
      "[220/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6988 (0.6113)\tPrec@1 85.938 (86.920)\n",
      "[230/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.7278 (0.6144)\tPrec@1 80.469 (86.803)\n",
      "[240/391]\tTime 0.148 (0.180)\tData 0.020 (0.020)\tLoss 0.6306 (0.6151)\tPrec@1 84.375 (86.754)\n",
      "[250/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.6835 (0.6159)\tPrec@1 83.594 (86.722)\n",
      "[260/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6358 (0.6164)\tPrec@1 86.719 (86.698)\n",
      "[270/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7218 (0.6173)\tPrec@1 79.688 (86.641)\n",
      "[280/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6589 (0.6174)\tPrec@1 84.375 (86.588)\n",
      "[290/391]\tTime 0.178 (0.179)\tData 0.020 (0.020)\tLoss 0.6450 (0.6179)\tPrec@1 85.156 (86.582)\n",
      "[300/391]\tTime 0.180 (0.179)\tData 0.021 (0.020)\tLoss 0.6532 (0.6181)\tPrec@1 84.375 (86.592)\n",
      "[310/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5351 (0.6180)\tPrec@1 89.844 (86.573)\n",
      "[320/391]\tTime 0.179 (0.179)\tData 0.022 (0.020)\tLoss 0.5496 (0.6184)\tPrec@1 89.062 (86.561)\n",
      "[330/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.6944 (0.6179)\tPrec@1 85.938 (86.579)\n",
      "[340/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.8153 (0.6194)\tPrec@1 77.344 (86.510)\n",
      "[350/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.6379 (0.6197)\tPrec@1 88.281 (86.514)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6368 (0.6201)\tPrec@1 86.719 (86.494)\n",
      "[370/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.5882 (0.6193)\tPrec@1 87.500 (86.525)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5395 (0.6187)\tPrec@1 89.062 (86.549)\n",
      "[390/391]\tTime 0.129 (0.180)\tData 0.012 (0.020)\tLoss 0.5970 (0.6189)\tPrec@1 85.000 (86.514)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8269 (0.8269)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.026 (0.022)\tLoss 0.8527 (0.9999)\tPrec@1 82.031 (81.108)\n",
      "Test: [20/79]\tTime 0.025 (0.022)\tLoss 0.7355 (1.0073)\tPrec@1 79.688 (81.064)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.7214 (1.0061)\tPrec@1 83.594 (81.048)\n",
      "Test: [40/79]\tTime 0.023 (0.023)\tLoss 1.0367 (0.9743)\tPrec@1 79.688 (81.364)\n",
      "Test: [50/79]\tTime 0.024 (0.023)\tLoss 1.2827 (0.9729)\tPrec@1 78.125 (81.373)\n",
      "Test: [60/79]\tTime 0.024 (0.023)\tLoss 0.8305 (0.9631)\tPrec@1 87.500 (81.378)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.9370 (0.9658)\tPrec@1 84.375 (81.327)\n",
      " * Prec@1 81.480\n",
      "\n",
      "===> epoch: 40/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.176 (0.176)\tData 0.020 (0.020)\tLoss 0.6612 (0.6612)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.5574 (0.6179)\tPrec@1 89.062 (86.293)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6586 (0.6032)\tPrec@1 85.156 (87.649)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6188 (0.6074)\tPrec@1 88.281 (87.273)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5416 (0.6086)\tPrec@1 91.406 (87.195)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5842 (0.6039)\tPrec@1 88.281 (87.286)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6417 (0.6062)\tPrec@1 82.812 (87.141)\n",
      "[70/391]\tTime 0.181 (0.181)\tData 0.021 (0.019)\tLoss 0.5674 (0.6077)\tPrec@1 89.062 (87.093)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6021 (0.6082)\tPrec@1 85.938 (86.941)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.4995 (0.6038)\tPrec@1 89.844 (87.054)\n",
      "[100/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.5804 (0.5996)\tPrec@1 88.281 (87.237)\n",
      "[110/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.6415 (0.5994)\tPrec@1 85.938 (87.183)\n",
      "[120/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.5215 (0.5995)\tPrec@1 90.625 (87.203)\n",
      "[130/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.5744 (0.5999)\tPrec@1 88.281 (87.148)\n",
      "[140/391]\tTime 0.186 (0.182)\tData 0.019 (0.019)\tLoss 0.5669 (0.6019)\tPrec@1 89.844 (87.096)\n",
      "[150/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6729 (0.6058)\tPrec@1 85.938 (86.926)\n",
      "[160/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.5738 (0.6092)\tPrec@1 87.500 (86.796)\n",
      "[170/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.5595 (0.6087)\tPrec@1 89.844 (86.783)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.021 (0.019)\tLoss 0.6484 (0.6122)\tPrec@1 90.625 (86.719)\n",
      "[190/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.7591 (0.6135)\tPrec@1 82.031 (86.682)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6933 (0.6156)\tPrec@1 85.938 (86.614)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6992 (0.6153)\tPrec@1 84.375 (86.648)\n",
      "[220/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.5378 (0.6156)\tPrec@1 89.062 (86.666)\n",
      "[230/391]\tTime 0.144 (0.181)\tData 0.021 (0.019)\tLoss 0.6143 (0.6145)\tPrec@1 89.062 (86.759)\n",
      "[240/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.5304 (0.6152)\tPrec@1 90.625 (86.716)\n",
      "[250/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.5697 (0.6155)\tPrec@1 89.062 (86.688)\n",
      "[260/391]\tTime 0.184 (0.180)\tData 0.020 (0.019)\tLoss 0.5361 (0.6159)\tPrec@1 88.281 (86.659)\n",
      "[270/391]\tTime 0.181 (0.180)\tData 0.021 (0.019)\tLoss 0.7529 (0.6179)\tPrec@1 82.812 (86.572)\n",
      "[280/391]\tTime 0.183 (0.180)\tData 0.021 (0.019)\tLoss 0.5503 (0.6174)\tPrec@1 89.844 (86.627)\n",
      "[290/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.6492 (0.6189)\tPrec@1 85.156 (86.531)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.022 (0.020)\tLoss 0.6162 (0.6196)\tPrec@1 84.375 (86.540)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6673 (0.6207)\tPrec@1 82.812 (86.505)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.7215 (0.6213)\tPrec@1 82.812 (86.463)\n",
      "[330/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.6449 (0.6211)\tPrec@1 88.281 (86.480)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7127 (0.6219)\tPrec@1 85.938 (86.458)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5906 (0.6225)\tPrec@1 88.281 (86.434)\n",
      "[360/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5755 (0.6221)\tPrec@1 87.500 (86.444)\n",
      "[370/391]\tTime 0.182 (0.180)\tData 0.022 (0.020)\tLoss 0.5926 (0.6210)\tPrec@1 90.625 (86.485)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[380/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5376 (0.6201)\tPrec@1 89.844 (86.536)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.7316 (0.6195)\tPrec@1 82.500 (86.540)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8887 (0.8887)\tPrec@1 82.031 (82.031)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.6495 (0.9870)\tPrec@1 83.594 (82.102)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0718 (1.0287)\tPrec@1 79.688 (81.064)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 1.2923 (1.0130)\tPrec@1 78.906 (81.250)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.1567 (1.0178)\tPrec@1 77.344 (81.155)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 1.4187 (1.0139)\tPrec@1 75.781 (81.158)\n",
      "Test: [60/79]\tTime 0.023 (0.023)\tLoss 0.8984 (1.0094)\tPrec@1 78.906 (81.314)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 0.8940 (1.0107)\tPrec@1 82.812 (81.305)\n",
      " * Prec@1 81.470\n",
      "\n",
      "===> epoch: 41/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.021 (0.021)\tLoss 0.6660 (0.6660)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5627 (0.5913)\tPrec@1 89.062 (88.068)\n",
      "[20/391]\tTime 0.177 (0.180)\tData 0.019 (0.020)\tLoss 0.5456 (0.5901)\tPrec@1 90.625 (87.872)\n",
      "[30/391]\tTime 0.181 (0.180)\tData 0.022 (0.020)\tLoss 0.6457 (0.5970)\tPrec@1 87.500 (87.828)\n",
      "[40/391]\tTime 0.190 (0.181)\tData 0.022 (0.020)\tLoss 0.5800 (0.5987)\tPrec@1 90.625 (87.576)\n",
      "[50/391]\tTime 0.192 (0.181)\tData 0.028 (0.020)\tLoss 0.6551 (0.6003)\tPrec@1 85.156 (87.454)\n",
      "[60/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5249 (0.6039)\tPrec@1 88.281 (87.295)\n",
      "[70/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6258 (0.6051)\tPrec@1 86.719 (87.225)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6796 (0.6031)\tPrec@1 83.594 (87.259)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6411 (0.6057)\tPrec@1 84.375 (87.122)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6165 (0.6061)\tPrec@1 87.500 (87.106)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5954 (0.6083)\tPrec@1 85.938 (86.993)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6172 (0.6085)\tPrec@1 87.500 (86.958)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5736 (0.6106)\tPrec@1 86.719 (86.802)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7134 (0.6122)\tPrec@1 84.375 (86.724)\n",
      "[150/391]\tTime 0.176 (0.181)\tData 0.019 (0.020)\tLoss 0.6196 (0.6116)\tPrec@1 85.938 (86.750)\n",
      "[160/391]\tTime 0.189 (0.181)\tData 0.019 (0.020)\tLoss 0.6175 (0.6112)\tPrec@1 85.938 (86.724)\n",
      "[170/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.5086 (0.6090)\tPrec@1 92.188 (86.819)\n",
      "[180/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6287 (0.6094)\tPrec@1 86.719 (86.844)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6114 (0.6080)\tPrec@1 82.031 (86.899)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6370 (0.6077)\tPrec@1 85.156 (86.913)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6469 (0.6082)\tPrec@1 84.375 (86.900)\n",
      "[220/391]\tTime 0.144 (0.180)\tData 0.019 (0.020)\tLoss 0.7379 (0.6089)\tPrec@1 82.812 (86.910)\n",
      "[230/391]\tTime 0.152 (0.179)\tData 0.019 (0.020)\tLoss 0.7629 (0.6106)\tPrec@1 84.375 (86.884)\n",
      "[240/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6810 (0.6088)\tPrec@1 82.031 (86.965)\n",
      "[250/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6848 (0.6094)\tPrec@1 88.281 (86.980)\n",
      "[260/391]\tTime 0.185 (0.179)\tData 0.020 (0.020)\tLoss 0.6051 (0.6096)\tPrec@1 88.281 (86.961)\n",
      "[270/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.6012 (0.6108)\tPrec@1 88.281 (86.921)\n",
      "[280/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6987 (0.6126)\tPrec@1 80.469 (86.824)\n",
      "[290/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.6511 (0.6135)\tPrec@1 84.375 (86.799)\n",
      "[300/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6728 (0.6139)\tPrec@1 82.812 (86.781)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6146 (0.6145)\tPrec@1 84.375 (86.764)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.6604 (0.6154)\tPrec@1 88.281 (86.758)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6272 (0.6153)\tPrec@1 85.938 (86.766)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6329 (0.6149)\tPrec@1 85.156 (86.749)\n",
      "[350/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5107 (0.6156)\tPrec@1 92.969 (86.728)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6302 (0.6151)\tPrec@1 84.375 (86.747)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.021 (0.020)\tLoss 0.5180 (0.6150)\tPrec@1 89.844 (86.748)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5863 (0.6152)\tPrec@1 89.062 (86.754)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.5946 (0.6153)\tPrec@1 87.500 (86.756)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9368 (0.9368)\tPrec@1 81.250 (81.250)\n",
      "Test: [10/79]\tTime 0.026 (0.023)\tLoss 1.0858 (1.1174)\tPrec@1 79.688 (79.759)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.9729 (1.1285)\tPrec@1 82.031 (79.241)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.9602 (1.1182)\tPrec@1 78.125 (79.461)\n",
      "Test: [40/79]\tTime 0.022 (0.023)\tLoss 1.2001 (1.1064)\tPrec@1 73.438 (79.516)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.4243 (1.1057)\tPrec@1 73.438 (79.550)\n",
      "Test: [60/79]\tTime 0.027 (0.023)\tLoss 1.1570 (1.1230)\tPrec@1 78.906 (79.585)\n",
      "Test: [70/79]\tTime 0.021 (0.023)\tLoss 1.0049 (1.1264)\tPrec@1 82.812 (79.456)\n",
      " * Prec@1 79.590\n",
      "\n",
      "===> epoch: 42/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.175 (0.175)\tData 0.020 (0.020)\tLoss 0.5773 (0.5773)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5283 (0.5831)\tPrec@1 89.844 (88.139)\n",
      "[20/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6164 (0.5923)\tPrec@1 85.156 (87.463)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6329 (0.6015)\tPrec@1 85.938 (87.198)\n",
      "[40/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.5570 (0.6001)\tPrec@1 89.062 (87.348)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5982 (0.5987)\tPrec@1 90.625 (87.270)\n",
      "[60/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5274 (0.5967)\tPrec@1 89.844 (87.436)\n",
      "[70/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5970 (0.5941)\tPrec@1 84.375 (87.445)\n",
      "[80/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5851 (0.5959)\tPrec@1 89.062 (87.346)\n",
      "[90/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5895 (0.5916)\tPrec@1 88.281 (87.543)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6061 (0.5920)\tPrec@1 86.719 (87.508)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6377 (0.5930)\tPrec@1 86.719 (87.451)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5189 (0.5957)\tPrec@1 89.844 (87.384)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6257 (0.6006)\tPrec@1 86.719 (87.166)\n",
      "[140/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.6895 (0.6035)\tPrec@1 89.062 (87.073)\n",
      "[150/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6101 (0.6069)\tPrec@1 90.625 (86.977)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7757 (0.6099)\tPrec@1 80.469 (86.927)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4932 (0.6094)\tPrec@1 90.625 (86.975)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6046 (0.6097)\tPrec@1 87.500 (86.965)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7234 (0.6128)\tPrec@1 80.469 (86.825)\n",
      "[200/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.8028 (0.6159)\tPrec@1 79.688 (86.738)\n",
      "[210/391]\tTime 0.144 (0.180)\tData 0.019 (0.020)\tLoss 0.7548 (0.6185)\tPrec@1 80.469 (86.667)\n",
      "[220/391]\tTime 0.143 (0.179)\tData 0.020 (0.020)\tLoss 0.5975 (0.6197)\tPrec@1 90.625 (86.680)\n",
      "[230/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.8345 (0.6216)\tPrec@1 82.031 (86.641)\n",
      "[240/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.6176 (0.6227)\tPrec@1 85.156 (86.602)\n",
      "[250/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5926 (0.6225)\tPrec@1 89.844 (86.600)\n",
      "[260/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5345 (0.6221)\tPrec@1 89.062 (86.620)\n",
      "[270/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.7011 (0.6214)\tPrec@1 83.594 (86.670)\n",
      "[280/391]\tTime 0.177 (0.179)\tData 0.019 (0.020)\tLoss 0.5975 (0.6196)\tPrec@1 89.062 (86.752)\n",
      "[290/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5284 (0.6205)\tPrec@1 89.844 (86.713)\n",
      "[300/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5856 (0.6200)\tPrec@1 87.500 (86.734)\n",
      "[310/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.7193 (0.6194)\tPrec@1 82.812 (86.749)\n",
      "[320/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6209 (0.6200)\tPrec@1 85.938 (86.699)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[330/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6817 (0.6199)\tPrec@1 83.594 (86.705)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6237 (0.6196)\tPrec@1 85.938 (86.716)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6228 (0.6192)\tPrec@1 84.375 (86.696)\n",
      "[360/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.7706 (0.6198)\tPrec@1 82.031 (86.669)\n",
      "[370/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.6322 (0.6200)\tPrec@1 89.062 (86.645)\n",
      "[380/391]\tTime 0.187 (0.180)\tData 0.019 (0.020)\tLoss 0.5266 (0.6188)\tPrec@1 91.406 (86.745)\n",
      "[390/391]\tTime 0.133 (0.180)\tData 0.012 (0.020)\tLoss 0.7075 (0.6192)\tPrec@1 83.750 (86.734)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9013 (0.9013)\tPrec@1 79.688 (79.688)\n",
      "Test: [10/79]\tTime 0.024 (0.023)\tLoss 1.1591 (1.3565)\tPrec@1 78.906 (75.568)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.5978 (1.4688)\tPrec@1 72.656 (74.963)\n",
      "Test: [30/79]\tTime 0.033 (0.023)\tLoss 1.5632 (1.5173)\tPrec@1 67.969 (74.395)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.5740 (1.5158)\tPrec@1 71.875 (74.333)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 1.7318 (1.5195)\tPrec@1 75.000 (74.249)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.0813 (1.5066)\tPrec@1 78.906 (74.116)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 1.6152 (1.5121)\tPrec@1 70.312 (73.845)\n",
      " * Prec@1 74.000\n",
      "\n",
      "===> epoch: 43/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6448 (0.6448)\tPrec@1 82.812 (82.812)\n",
      "[10/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5868 (0.6411)\tPrec@1 86.719 (86.151)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6219 (0.6254)\tPrec@1 87.500 (86.719)\n",
      "[30/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.6706 (0.6215)\tPrec@1 85.938 (86.946)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6965 (0.6211)\tPrec@1 83.594 (86.814)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6373 (0.6180)\tPrec@1 85.938 (86.841)\n",
      "[60/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6341 (0.6160)\tPrec@1 85.156 (86.693)\n",
      "[70/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.7597 (0.6210)\tPrec@1 81.250 (86.455)\n",
      "[80/391]\tTime 0.187 (0.180)\tData 0.021 (0.020)\tLoss 0.5854 (0.6205)\tPrec@1 85.156 (86.294)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5781 (0.6172)\tPrec@1 88.281 (86.530)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5728 (0.6162)\tPrec@1 89.844 (86.572)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6450 (0.6162)\tPrec@1 85.156 (86.620)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6524 (0.6140)\tPrec@1 86.719 (86.712)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7761 (0.6147)\tPrec@1 80.469 (86.766)\n",
      "[140/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.5496 (0.6130)\tPrec@1 88.281 (86.785)\n",
      "[150/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.5977 (0.6132)\tPrec@1 87.500 (86.791)\n",
      "[160/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6813 (0.6133)\tPrec@1 85.156 (86.787)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6665 (0.6130)\tPrec@1 80.469 (86.810)\n",
      "[180/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6142 (0.6129)\tPrec@1 87.500 (86.805)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5104 (0.6120)\tPrec@1 92.969 (86.862)\n",
      "[200/391]\tTime 0.142 (0.181)\tData 0.019 (0.020)\tLoss 0.6056 (0.6123)\tPrec@1 86.719 (86.882)\n",
      "[210/391]\tTime 0.144 (0.179)\tData 0.020 (0.020)\tLoss 0.5856 (0.6110)\tPrec@1 87.500 (86.922)\n",
      "[220/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.6425 (0.6116)\tPrec@1 86.719 (86.896)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.7030 (0.6131)\tPrec@1 85.156 (86.840)\n",
      "[240/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7154 (0.6157)\tPrec@1 83.594 (86.758)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5263 (0.6174)\tPrec@1 92.188 (86.694)\n",
      "[260/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6109 (0.6179)\tPrec@1 88.281 (86.686)\n",
      "[270/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6475 (0.6182)\tPrec@1 85.938 (86.696)\n",
      "[280/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5865 (0.6180)\tPrec@1 89.844 (86.671)\n",
      "[290/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6144 (0.6185)\tPrec@1 85.938 (86.649)\n",
      "[300/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5350 (0.6184)\tPrec@1 88.281 (86.659)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5449 (0.6183)\tPrec@1 89.844 (86.679)\n",
      "[320/391]\tTime 0.183 (0.180)\tData 0.021 (0.020)\tLoss 0.5131 (0.6179)\tPrec@1 90.625 (86.702)\n",
      "[330/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6247 (0.6184)\tPrec@1 86.719 (86.683)\n",
      "[340/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5844 (0.6176)\tPrec@1 89.844 (86.716)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.8050 (0.6183)\tPrec@1 82.031 (86.683)\n",
      "[360/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6339 (0.6181)\tPrec@1 82.812 (86.669)\n",
      "[370/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.5507 (0.6178)\tPrec@1 91.406 (86.693)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.022 (0.020)\tLoss 0.5891 (0.6181)\tPrec@1 88.281 (86.665)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.012 (0.020)\tLoss 0.6250 (0.6192)\tPrec@1 87.500 (86.620)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.3460 (1.3460)\tPrec@1 75.000 (75.000)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.6848 (1.8574)\tPrec@1 75.781 (71.165)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 2.1555 (1.8743)\tPrec@1 68.750 (71.057)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 1.7465 (1.8380)\tPrec@1 72.656 (71.799)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 2.2009 (1.8626)\tPrec@1 71.094 (71.608)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 2.1450 (1.8759)\tPrec@1 66.406 (71.339)\n",
      "Test: [60/79]\tTime 0.023 (0.023)\tLoss 1.3949 (1.8628)\tPrec@1 75.781 (71.401)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.6486 (1.8753)\tPrec@1 71.094 (71.017)\n",
      " * Prec@1 71.110\n",
      "\n",
      "===> epoch: 44/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.180 (0.180)\tData 0.021 (0.021)\tLoss 0.7223 (0.7223)\tPrec@1 78.906 (78.906)\n",
      "[10/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6936 (0.6425)\tPrec@1 82.031 (85.298)\n",
      "[20/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5783 (0.6283)\tPrec@1 91.406 (86.272)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6054 (0.6339)\tPrec@1 90.625 (86.064)\n",
      "[40/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6183 (0.6305)\tPrec@1 89.062 (86.109)\n",
      "[50/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5460 (0.6243)\tPrec@1 89.062 (86.259)\n",
      "[60/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5866 (0.6197)\tPrec@1 85.938 (86.437)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5813 (0.6229)\tPrec@1 85.156 (86.268)\n",
      "[80/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6834 (0.6222)\tPrec@1 82.031 (86.275)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.8670 (0.6212)\tPrec@1 76.562 (86.367)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6517 (0.6199)\tPrec@1 86.719 (86.409)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6281 (0.6155)\tPrec@1 84.375 (86.550)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5562 (0.6138)\tPrec@1 92.969 (86.706)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6218 (0.6145)\tPrec@1 85.938 (86.677)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.7113 (0.6148)\tPrec@1 82.812 (86.697)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6314 (0.6167)\tPrec@1 85.938 (86.579)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5245 (0.6167)\tPrec@1 91.406 (86.646)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6465 (0.6188)\tPrec@1 86.719 (86.541)\n",
      "[180/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6212 (0.6199)\tPrec@1 89.062 (86.542)\n",
      "[190/391]\tTime 0.145 (0.180)\tData 0.021 (0.020)\tLoss 0.6780 (0.6204)\tPrec@1 84.375 (86.563)\n",
      "[200/391]\tTime 0.145 (0.178)\tData 0.020 (0.020)\tLoss 0.6153 (0.6217)\tPrec@1 85.938 (86.509)\n",
      "[210/391]\tTime 0.187 (0.178)\tData 0.020 (0.020)\tLoss 0.6576 (0.6222)\tPrec@1 85.938 (86.485)\n",
      "[220/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.5930 (0.6215)\tPrec@1 84.375 (86.475)\n",
      "[230/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.6688 (0.6219)\tPrec@1 85.938 (86.482)\n",
      "[240/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.7451 (0.6227)\tPrec@1 81.250 (86.437)\n",
      "[250/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6209 (0.6236)\tPrec@1 84.375 (86.379)\n",
      "[260/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5600 (0.6242)\tPrec@1 87.500 (86.369)\n",
      "[270/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6250 (0.6231)\tPrec@1 87.500 (86.413)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5293 (0.6232)\tPrec@1 90.625 (86.407)\n",
      "[290/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.5383 (0.6230)\tPrec@1 92.188 (86.445)\n",
      "[300/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5490 (0.6209)\tPrec@1 89.844 (86.521)\n",
      "[310/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.6400 (0.6212)\tPrec@1 87.500 (86.525)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5811 (0.6215)\tPrec@1 87.500 (86.546)\n",
      "[330/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.7000 (0.6222)\tPrec@1 83.594 (86.520)\n",
      "[340/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.8249 (0.6220)\tPrec@1 77.344 (86.515)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7337 (0.6214)\tPrec@1 81.250 (86.534)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.7522 (0.6229)\tPrec@1 80.469 (86.502)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6395 (0.6229)\tPrec@1 85.938 (86.510)\n",
      "[380/391]\tTime 0.185 (0.180)\tData 0.021 (0.020)\tLoss 0.6262 (0.6229)\tPrec@1 86.719 (86.491)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.012 (0.020)\tLoss 0.6574 (0.6222)\tPrec@1 77.500 (86.500)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.8845 (0.8845)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 1.1466 (1.1451)\tPrec@1 78.906 (78.764)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.4011 (1.2297)\tPrec@1 75.000 (77.307)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 1.0392 (1.2115)\tPrec@1 78.906 (77.470)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.3747 (1.1870)\tPrec@1 71.875 (77.534)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 1.2524 (1.1895)\tPrec@1 78.125 (77.497)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.2702 (1.1810)\tPrec@1 78.125 (77.574)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.0431 (1.1797)\tPrec@1 78.906 (77.641)\n",
      " * Prec@1 77.670\n",
      "\n",
      "===> epoch: 45/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5219 (0.5219)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6019 (0.6004)\tPrec@1 87.500 (86.861)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6082 (0.5977)\tPrec@1 86.719 (87.277)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5992 (0.6188)\tPrec@1 87.500 (86.442)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5612 (0.6188)\tPrec@1 87.500 (86.452)\n",
      "[50/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6631 (0.6132)\tPrec@1 84.375 (86.688)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6903 (0.6081)\tPrec@1 82.031 (86.796)\n",
      "[70/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6953 (0.6043)\tPrec@1 83.594 (86.994)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6367 (0.6050)\tPrec@1 89.062 (87.027)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.022 (0.020)\tLoss 0.5644 (0.6050)\tPrec@1 88.281 (87.071)\n",
      "[100/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6652 (0.6064)\tPrec@1 85.156 (86.959)\n",
      "[110/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6020 (0.6070)\tPrec@1 89.844 (87.021)\n",
      "[120/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5731 (0.6083)\tPrec@1 89.844 (87.009)\n",
      "[130/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.7138 (0.6146)\tPrec@1 83.594 (86.766)\n",
      "[140/391]\tTime 0.182 (0.180)\tData 0.022 (0.020)\tLoss 0.7796 (0.6182)\tPrec@1 80.469 (86.658)\n",
      "[150/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6515 (0.6205)\tPrec@1 85.156 (86.574)\n",
      "[160/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5943 (0.6187)\tPrec@1 88.281 (86.646)\n",
      "[170/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6379 (0.6200)\tPrec@1 87.500 (86.591)\n",
      "[180/391]\tTime 0.156 (0.180)\tData 0.020 (0.020)\tLoss 0.5374 (0.6194)\tPrec@1 89.844 (86.624)\n",
      "[190/391]\tTime 0.143 (0.178)\tData 0.020 (0.020)\tLoss 0.6300 (0.6202)\tPrec@1 85.938 (86.600)\n",
      "[200/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.6914 (0.6196)\tPrec@1 83.594 (86.622)\n",
      "[210/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.5358 (0.6209)\tPrec@1 89.844 (86.608)\n",
      "[220/391]\tTime 0.189 (0.178)\tData 0.021 (0.020)\tLoss 0.6421 (0.6215)\tPrec@1 84.375 (86.560)\n",
      "[230/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.6008 (0.6195)\tPrec@1 85.938 (86.651)\n",
      "[240/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5847 (0.6195)\tPrec@1 89.844 (86.625)\n",
      "[250/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6003 (0.6191)\tPrec@1 88.281 (86.616)\n",
      "[260/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.7345 (0.6213)\tPrec@1 82.031 (86.512)\n",
      "[270/391]\tTime 0.181 (0.179)\tData 0.021 (0.020)\tLoss 0.5277 (0.6211)\tPrec@1 91.406 (86.549)\n",
      "[280/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.5944 (0.6201)\tPrec@1 83.594 (86.580)\n",
      "[290/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5942 (0.6202)\tPrec@1 87.500 (86.585)\n",
      "[300/391]\tTime 0.178 (0.179)\tData 0.020 (0.020)\tLoss 0.5534 (0.6198)\tPrec@1 89.062 (86.610)\n",
      "[310/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6074 (0.6187)\tPrec@1 87.500 (86.638)\n",
      "[320/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6502 (0.6190)\tPrec@1 84.375 (86.607)\n",
      "[330/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7888 (0.6204)\tPrec@1 80.469 (86.591)\n",
      "[340/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7442 (0.6208)\tPrec@1 82.812 (86.561)\n",
      "[350/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5519 (0.6202)\tPrec@1 91.406 (86.594)\n",
      "[360/391]\tTime 0.187 (0.180)\tData 0.020 (0.020)\tLoss 0.7207 (0.6203)\tPrec@1 82.031 (86.598)\n",
      "[370/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5513 (0.6204)\tPrec@1 86.719 (86.588)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.022 (0.020)\tLoss 0.8525 (0.6207)\tPrec@1 79.688 (86.565)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.6757 (0.6213)\tPrec@1 82.500 (86.548)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7597 (0.7597)\tPrec@1 81.250 (81.250)\n",
      "Test: [10/79]\tTime 0.020 (0.023)\tLoss 0.8346 (0.8364)\tPrec@1 82.031 (83.026)\n",
      "Test: [20/79]\tTime 0.026 (0.024)\tLoss 0.6039 (0.8937)\tPrec@1 89.062 (82.664)\n",
      "Test: [30/79]\tTime 0.026 (0.024)\tLoss 0.6097 (0.9064)\tPrec@1 82.812 (82.283)\n",
      "Test: [40/79]\tTime 0.026 (0.023)\tLoss 0.8505 (0.8715)\tPrec@1 82.812 (82.870)\n",
      "Test: [50/79]\tTime 0.019 (0.023)\tLoss 0.8337 (0.8541)\tPrec@1 79.688 (83.042)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.0232 (0.8769)\tPrec@1 82.812 (82.889)\n",
      "Test: [70/79]\tTime 0.027 (0.023)\tLoss 0.6742 (0.8647)\tPrec@1 85.156 (83.165)\n",
      " * Prec@1 83.200\n",
      "\n",
      "===> epoch: 46/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.4269 (0.4269)\tPrec@1 94.531 (94.531)\n",
      "[10/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.7480 (0.6316)\tPrec@1 81.250 (86.506)\n",
      "[20/391]\tTime 0.184 (0.182)\tData 0.020 (0.021)\tLoss 0.6779 (0.6275)\tPrec@1 84.375 (86.049)\n",
      "[30/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5659 (0.6296)\tPrec@1 88.281 (86.215)\n",
      "[40/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6572 (0.6327)\tPrec@1 85.156 (86.052)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6176 (0.6281)\tPrec@1 86.719 (86.183)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6115 (0.6208)\tPrec@1 84.375 (86.335)\n",
      "[70/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5781 (0.6179)\tPrec@1 90.625 (86.433)\n",
      "[80/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6540 (0.6197)\tPrec@1 83.594 (86.420)\n",
      "[90/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5841 (0.6215)\tPrec@1 88.281 (86.307)\n",
      "[100/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.5706 (0.6194)\tPrec@1 92.188 (86.479)\n",
      "[110/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6509 (0.6170)\tPrec@1 85.938 (86.620)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5444 (0.6148)\tPrec@1 89.844 (86.693)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6314 (0.6165)\tPrec@1 87.500 (86.617)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5284 (0.6137)\tPrec@1 91.406 (86.708)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.6275 (0.6163)\tPrec@1 88.281 (86.595)\n",
      "[160/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.6177 (0.6171)\tPrec@1 88.281 (86.568)\n",
      "[170/391]\tTime 0.159 (0.182)\tData 0.019 (0.020)\tLoss 0.5658 (0.6166)\tPrec@1 86.719 (86.582)\n",
      "[180/391]\tTime 0.143 (0.180)\tData 0.020 (0.020)\tLoss 0.5593 (0.6164)\tPrec@1 89.062 (86.550)\n",
      "[190/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5788 (0.6164)\tPrec@1 89.062 (86.535)\n",
      "[200/391]\tTime 0.188 (0.179)\tData 0.019 (0.020)\tLoss 0.6447 (0.6156)\tPrec@1 85.938 (86.583)\n",
      "[210/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6540 (0.6161)\tPrec@1 81.250 (86.563)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6950 (0.6171)\tPrec@1 83.594 (86.471)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[230/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6553 (0.6166)\tPrec@1 85.156 (86.533)\n",
      "[240/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5476 (0.6157)\tPrec@1 89.062 (86.544)\n",
      "[250/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5791 (0.6149)\tPrec@1 89.844 (86.585)\n",
      "[260/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5767 (0.6154)\tPrec@1 89.062 (86.590)\n",
      "[270/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6793 (0.6164)\tPrec@1 85.156 (86.531)\n",
      "[280/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.6502 (0.6171)\tPrec@1 85.938 (86.557)\n",
      "[290/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6610 (0.6188)\tPrec@1 85.938 (86.499)\n",
      "[300/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.7032 (0.6194)\tPrec@1 82.031 (86.503)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5993 (0.6200)\tPrec@1 85.156 (86.470)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.7204 (0.6216)\tPrec@1 82.031 (86.407)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6555 (0.6219)\tPrec@1 83.594 (86.384)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6750 (0.6218)\tPrec@1 83.594 (86.407)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5393 (0.6213)\tPrec@1 89.844 (86.418)\n",
      "[360/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5332 (0.6212)\tPrec@1 90.625 (86.420)\n",
      "[370/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6957 (0.6223)\tPrec@1 82.812 (86.367)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6356 (0.6224)\tPrec@1 82.031 (86.344)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.014 (0.020)\tLoss 0.6687 (0.6231)\tPrec@1 86.250 (86.330)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6465 (0.6465)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.7034 (0.8936)\tPrec@1 85.156 (82.102)\n",
      "Test: [20/79]\tTime 0.025 (0.022)\tLoss 0.7929 (0.9311)\tPrec@1 81.250 (81.213)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.4015 (0.9151)\tPrec@1 87.500 (81.603)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 0.9853 (0.9187)\tPrec@1 75.781 (81.574)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 0.8848 (0.9060)\tPrec@1 82.812 (81.694)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.7228 (0.9102)\tPrec@1 85.156 (81.852)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.9101 (0.9033)\tPrec@1 79.688 (81.932)\n",
      " * Prec@1 82.050\n",
      "\n",
      "===> epoch: 47/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.169 (0.169)\tData 0.020 (0.020)\tLoss 0.4732 (0.4732)\tPrec@1 94.531 (94.531)\n",
      "[10/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6595 (0.5836)\tPrec@1 82.812 (88.423)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5759 (0.5932)\tPrec@1 87.500 (87.612)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6571 (0.6025)\tPrec@1 85.156 (87.374)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6322 (0.6134)\tPrec@1 86.719 (87.024)\n",
      "[50/391]\tTime 0.185 (0.181)\tData 0.021 (0.020)\tLoss 0.7131 (0.6214)\tPrec@1 82.031 (86.627)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6474 (0.6214)\tPrec@1 87.500 (86.668)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.7174 (0.6248)\tPrec@1 83.594 (86.587)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6895 (0.6294)\tPrec@1 87.500 (86.487)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6095 (0.6279)\tPrec@1 87.500 (86.599)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.021 (0.019)\tLoss 0.6635 (0.6279)\tPrec@1 86.719 (86.618)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.7674 (0.6294)\tPrec@1 81.250 (86.458)\n",
      "[120/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7155 (0.6341)\tPrec@1 82.812 (86.318)\n",
      "[130/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.6010 (0.6345)\tPrec@1 88.281 (86.307)\n",
      "[140/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6084 (0.6322)\tPrec@1 86.719 (86.381)\n",
      "[150/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5548 (0.6300)\tPrec@1 89.062 (86.460)\n",
      "[160/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6024 (0.6289)\tPrec@1 86.719 (86.437)\n",
      "[170/391]\tTime 0.144 (0.180)\tData 0.019 (0.020)\tLoss 0.6949 (0.6298)\tPrec@1 82.812 (86.408)\n",
      "[180/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.6229 (0.6294)\tPrec@1 87.500 (86.430)\n",
      "[190/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.5451 (0.6285)\tPrec@1 90.625 (86.469)\n",
      "[200/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.7013 (0.6281)\tPrec@1 85.156 (86.462)\n",
      "[210/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5740 (0.6284)\tPrec@1 85.938 (86.423)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5360 (0.6292)\tPrec@1 88.281 (86.422)\n",
      "[230/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.5907 (0.6285)\tPrec@1 90.625 (86.435)\n",
      "[240/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6408 (0.6268)\tPrec@1 85.156 (86.505)\n",
      "[250/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.5227 (0.6257)\tPrec@1 92.188 (86.523)\n",
      "[260/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6846 (0.6252)\tPrec@1 83.594 (86.521)\n",
      "[270/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.7737 (0.6253)\tPrec@1 80.469 (86.511)\n",
      "[280/391]\tTime 0.183 (0.180)\tData 0.021 (0.020)\tLoss 0.5335 (0.6242)\tPrec@1 92.969 (86.557)\n",
      "[290/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.7031 (0.6249)\tPrec@1 82.812 (86.536)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5332 (0.6253)\tPrec@1 90.625 (86.537)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.4964 (0.6255)\tPrec@1 91.406 (86.530)\n",
      "[320/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5924 (0.6246)\tPrec@1 88.281 (86.565)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5809 (0.6249)\tPrec@1 87.500 (86.551)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.7065 (0.6255)\tPrec@1 82.031 (86.524)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6309 (0.6255)\tPrec@1 85.938 (86.550)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5592 (0.6253)\tPrec@1 88.281 (86.535)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5557 (0.6245)\tPrec@1 89.844 (86.565)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6271 (0.6243)\tPrec@1 85.156 (86.577)\n",
      "[390/391]\tTime 0.132 (0.180)\tData 0.013 (0.020)\tLoss 0.6378 (0.6243)\tPrec@1 87.500 (86.550)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.3108 (1.3108)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.019 (0.023)\tLoss 0.9583 (1.0469)\tPrec@1 82.812 (80.469)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.9166 (1.0634)\tPrec@1 81.250 (80.394)\n",
      "Test: [30/79]\tTime 0.020 (0.023)\tLoss 0.6817 (1.0652)\tPrec@1 87.500 (80.267)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 1.2322 (1.0661)\tPrec@1 78.125 (80.202)\n",
      "Test: [50/79]\tTime 0.025 (0.023)\tLoss 1.1555 (1.0517)\tPrec@1 75.000 (80.377)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.9345 (1.0616)\tPrec@1 81.250 (80.302)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.1297 (1.0616)\tPrec@1 82.812 (80.436)\n",
      " * Prec@1 80.440\n",
      "\n",
      "===> epoch: 48/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.173 (0.173)\tData 0.021 (0.021)\tLoss 0.5364 (0.5364)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5489 (0.6275)\tPrec@1 89.062 (85.511)\n",
      "[20/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5913 (0.6032)\tPrec@1 85.156 (86.830)\n",
      "[30/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6015 (0.6112)\tPrec@1 87.500 (86.794)\n",
      "[40/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5360 (0.6061)\tPrec@1 86.719 (86.719)\n",
      "[50/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7162 (0.6087)\tPrec@1 79.688 (86.642)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6107 (0.6135)\tPrec@1 83.594 (86.437)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5718 (0.6175)\tPrec@1 89.062 (86.257)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5564 (0.6174)\tPrec@1 89.062 (86.314)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6316 (0.6166)\tPrec@1 84.375 (86.281)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5659 (0.6173)\tPrec@1 88.281 (86.255)\n",
      "[110/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6987 (0.6168)\tPrec@1 84.375 (86.296)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5581 (0.6133)\tPrec@1 86.719 (86.480)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5381 (0.6115)\tPrec@1 89.062 (86.629)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7339 (0.6103)\tPrec@1 81.250 (86.697)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5407 (0.6102)\tPrec@1 89.844 (86.770)\n",
      "[160/391]\tTime 0.143 (0.179)\tData 0.019 (0.020)\tLoss 0.6114 (0.6126)\tPrec@1 87.500 (86.675)\n",
      "[170/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.6551 (0.6136)\tPrec@1 85.156 (86.609)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[180/391]\tTime 0.184 (0.178)\tData 0.021 (0.020)\tLoss 0.6740 (0.6139)\tPrec@1 85.938 (86.611)\n",
      "[190/391]\tTime 0.182 (0.178)\tData 0.020 (0.020)\tLoss 0.5937 (0.6164)\tPrec@1 87.500 (86.498)\n",
      "[200/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6177 (0.6179)\tPrec@1 90.625 (86.435)\n",
      "[210/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6697 (0.6203)\tPrec@1 84.375 (86.326)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6087 (0.6206)\tPrec@1 88.281 (86.316)\n",
      "[230/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.7512 (0.6212)\tPrec@1 81.250 (86.353)\n",
      "[240/391]\tTime 0.185 (0.179)\tData 0.022 (0.020)\tLoss 0.6182 (0.6209)\tPrec@1 85.938 (86.352)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.4629 (0.6196)\tPrec@1 90.625 (86.414)\n",
      "[260/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5844 (0.6195)\tPrec@1 88.281 (86.449)\n",
      "[270/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5802 (0.6182)\tPrec@1 85.938 (86.531)\n",
      "[280/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6938 (0.6175)\tPrec@1 82.031 (86.538)\n",
      "[290/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.6479 (0.6167)\tPrec@1 87.500 (86.574)\n",
      "[300/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.5934 (0.6164)\tPrec@1 84.375 (86.599)\n",
      "[310/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5463 (0.6164)\tPrec@1 90.625 (86.586)\n",
      "[320/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6781 (0.6167)\tPrec@1 87.500 (86.575)\n",
      "[330/391]\tTime 0.179 (0.179)\tData 0.021 (0.020)\tLoss 0.5654 (0.6167)\tPrec@1 90.625 (86.582)\n",
      "[340/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6022 (0.6172)\tPrec@1 85.156 (86.561)\n",
      "[350/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.6400 (0.6177)\tPrec@1 85.938 (86.550)\n",
      "[360/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.6986 (0.6177)\tPrec@1 89.062 (86.585)\n",
      "[370/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.7597 (0.6187)\tPrec@1 80.469 (86.582)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6473 (0.6189)\tPrec@1 85.156 (86.567)\n",
      "[390/391]\tTime 0.127 (0.179)\tData 0.013 (0.020)\tLoss 0.5555 (0.6192)\tPrec@1 86.250 (86.566)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.5848 (0.5848)\tPrec@1 85.156 (85.156)\n",
      "Test: [10/79]\tTime 0.020 (0.023)\tLoss 0.9044 (0.9305)\tPrec@1 81.250 (82.173)\n",
      "Test: [20/79]\tTime 0.023 (0.023)\tLoss 0.7171 (0.9871)\tPrec@1 83.594 (81.622)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.7525 (0.9554)\tPrec@1 84.375 (81.956)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.1110 (0.9249)\tPrec@1 78.906 (82.355)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 1.0423 (0.9138)\tPrec@1 82.812 (82.537)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.9244 (0.9251)\tPrec@1 85.938 (82.595)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 0.7964 (0.9262)\tPrec@1 86.719 (82.504)\n",
      " * Prec@1 82.600\n",
      "\n",
      "===> epoch: 49/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.5301 (0.5301)\tPrec@1 92.969 (92.969)\n",
      "[10/391]\tTime 0.183 (0.184)\tData 0.019 (0.020)\tLoss 0.5444 (0.6103)\tPrec@1 85.938 (86.790)\n",
      "[20/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5376 (0.6077)\tPrec@1 90.625 (87.202)\n",
      "[30/391]\tTime 0.177 (0.182)\tData 0.019 (0.020)\tLoss 0.6774 (0.6075)\tPrec@1 85.156 (87.273)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5640 (0.6091)\tPrec@1 88.281 (87.329)\n",
      "[50/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.5990 (0.6072)\tPrec@1 89.062 (87.408)\n",
      "[60/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.7416 (0.6103)\tPrec@1 83.594 (87.359)\n",
      "[70/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5748 (0.6083)\tPrec@1 86.719 (87.335)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5435 (0.6097)\tPrec@1 90.625 (87.162)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6019 (0.6096)\tPrec@1 84.375 (87.088)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5689 (0.6066)\tPrec@1 88.281 (87.136)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6945 (0.6102)\tPrec@1 85.156 (86.972)\n",
      "[120/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.8093 (0.6120)\tPrec@1 79.688 (86.900)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6895 (0.6156)\tPrec@1 84.375 (86.772)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5657 (0.6175)\tPrec@1 92.188 (86.652)\n",
      "[150/391]\tTime 0.144 (0.179)\tData 0.020 (0.020)\tLoss 0.6369 (0.6173)\tPrec@1 83.594 (86.677)\n",
      "[160/391]\tTime 0.188 (0.178)\tData 0.019 (0.020)\tLoss 0.6495 (0.6186)\tPrec@1 86.719 (86.627)\n",
      "[170/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6628 (0.6195)\tPrec@1 87.500 (86.577)\n",
      "[180/391]\tTime 0.181 (0.179)\tData 0.021 (0.020)\tLoss 0.6405 (0.6213)\tPrec@1 85.938 (86.533)\n",
      "[190/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5936 (0.6218)\tPrec@1 86.719 (86.506)\n",
      "[200/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6452 (0.6217)\tPrec@1 85.156 (86.451)\n",
      "[210/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7152 (0.6240)\tPrec@1 84.375 (86.363)\n",
      "[220/391]\tTime 0.185 (0.179)\tData 0.020 (0.020)\tLoss 0.6302 (0.6244)\tPrec@1 85.938 (86.365)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6767 (0.6247)\tPrec@1 83.594 (86.323)\n",
      "[240/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.7007 (0.6238)\tPrec@1 82.031 (86.372)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5372 (0.6238)\tPrec@1 89.844 (86.395)\n",
      "[260/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6782 (0.6225)\tPrec@1 83.594 (86.464)\n",
      "[270/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.4796 (0.6207)\tPrec@1 90.625 (86.511)\n",
      "[280/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.5038 (0.6207)\tPrec@1 91.406 (86.521)\n",
      "[290/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5979 (0.6199)\tPrec@1 88.281 (86.555)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6510 (0.6196)\tPrec@1 83.594 (86.568)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5690 (0.6197)\tPrec@1 89.844 (86.535)\n",
      "[320/391]\tTime 0.183 (0.180)\tData 0.021 (0.020)\tLoss 0.5994 (0.6186)\tPrec@1 90.625 (86.570)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6057 (0.6188)\tPrec@1 87.500 (86.532)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.021 (0.020)\tLoss 0.6367 (0.6180)\tPrec@1 85.938 (86.540)\n",
      "[350/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6065 (0.6180)\tPrec@1 86.719 (86.550)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7185 (0.6180)\tPrec@1 85.156 (86.569)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6063 (0.6175)\tPrec@1 82.812 (86.580)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6713 (0.6184)\tPrec@1 82.031 (86.522)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.5657 (0.6180)\tPrec@1 85.000 (86.506)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9943 (0.9943)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.8326 (1.2543)\tPrec@1 85.156 (79.972)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.4946 (1.3415)\tPrec@1 76.562 (78.795)\n",
      "Test: [30/79]\tTime 0.024 (0.023)\tLoss 0.9319 (1.3360)\tPrec@1 82.812 (78.327)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.6650 (1.3200)\tPrec@1 75.000 (78.449)\n",
      "Test: [50/79]\tTime 0.024 (0.023)\tLoss 1.6876 (1.3163)\tPrec@1 73.438 (78.646)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.5508 (1.2878)\tPrec@1 75.781 (78.868)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.5424 (1.2763)\tPrec@1 76.562 (79.027)\n",
      " * Prec@1 79.040\n",
      "\n",
      "===> epoch: 50/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.162 (0.162)\tData 0.021 (0.021)\tLoss 0.7091 (0.7091)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.5407 (0.6091)\tPrec@1 89.062 (87.003)\n",
      "[20/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5716 (0.5996)\tPrec@1 86.719 (86.905)\n",
      "[30/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5737 (0.6042)\tPrec@1 87.500 (86.542)\n",
      "[40/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5183 (0.6083)\tPrec@1 90.625 (86.662)\n",
      "[50/391]\tTime 0.179 (0.180)\tData 0.023 (0.020)\tLoss 0.5539 (0.5992)\tPrec@1 90.625 (87.040)\n",
      "[60/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5639 (0.5969)\tPrec@1 90.625 (87.154)\n",
      "[70/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5198 (0.5930)\tPrec@1 89.844 (87.379)\n",
      "[80/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.4786 (0.5949)\tPrec@1 90.625 (87.307)\n",
      "[90/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5174 (0.5992)\tPrec@1 89.844 (87.260)\n",
      "[100/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5965 (0.5991)\tPrec@1 85.938 (87.183)\n",
      "[110/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5960 (0.5994)\tPrec@1 85.938 (87.204)\n",
      "[120/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6009 (0.5969)\tPrec@1 90.625 (87.326)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[130/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.7997 (0.5994)\tPrec@1 81.250 (87.261)\n",
      "[140/391]\tTime 0.144 (0.179)\tData 0.021 (0.020)\tLoss 0.6036 (0.6018)\tPrec@1 86.719 (87.168)\n",
      "[150/391]\tTime 0.176 (0.177)\tData 0.020 (0.020)\tLoss 0.6073 (0.6034)\tPrec@1 86.719 (87.065)\n",
      "[160/391]\tTime 0.187 (0.177)\tData 0.020 (0.020)\tLoss 0.7306 (0.6012)\tPrec@1 83.594 (87.189)\n",
      "[170/391]\tTime 0.181 (0.177)\tData 0.020 (0.020)\tLoss 0.5362 (0.6022)\tPrec@1 91.406 (87.144)\n",
      "[180/391]\tTime 0.181 (0.178)\tData 0.019 (0.020)\tLoss 0.7513 (0.6039)\tPrec@1 83.594 (87.073)\n",
      "[190/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6583 (0.6040)\tPrec@1 89.844 (87.087)\n",
      "[200/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.6386 (0.6052)\tPrec@1 83.594 (87.018)\n",
      "[210/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6227 (0.6059)\tPrec@1 88.281 (87.011)\n",
      "[220/391]\tTime 0.184 (0.178)\tData 0.020 (0.020)\tLoss 0.5653 (0.6081)\tPrec@1 89.062 (86.959)\n",
      "[230/391]\tTime 0.183 (0.178)\tData 0.020 (0.020)\tLoss 0.6665 (0.6106)\tPrec@1 82.031 (86.874)\n",
      "[240/391]\tTime 0.181 (0.178)\tData 0.019 (0.020)\tLoss 0.5776 (0.6102)\tPrec@1 87.500 (86.916)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6495 (0.6109)\tPrec@1 85.938 (86.877)\n",
      "[260/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6525 (0.6104)\tPrec@1 84.375 (86.907)\n",
      "[270/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6341 (0.6106)\tPrec@1 86.719 (86.915)\n",
      "[280/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.5449 (0.6099)\tPrec@1 90.625 (86.911)\n",
      "[290/391]\tTime 0.186 (0.179)\tData 0.020 (0.020)\tLoss 0.6239 (0.6102)\tPrec@1 85.156 (86.899)\n",
      "[300/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6149 (0.6114)\tPrec@1 85.938 (86.867)\n",
      "[310/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6733 (0.6119)\tPrec@1 80.469 (86.849)\n",
      "[320/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.7240 (0.6134)\tPrec@1 82.812 (86.765)\n",
      "[330/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6090 (0.6143)\tPrec@1 86.719 (86.705)\n",
      "[340/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.5715 (0.6149)\tPrec@1 88.281 (86.712)\n",
      "[350/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.6326 (0.6154)\tPrec@1 81.250 (86.685)\n",
      "[360/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.5637 (0.6154)\tPrec@1 90.625 (86.723)\n",
      "[370/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6724 (0.6168)\tPrec@1 85.938 (86.666)\n",
      "[380/391]\tTime 0.181 (0.179)\tData 0.024 (0.020)\tLoss 0.5960 (0.6181)\tPrec@1 89.062 (86.618)\n",
      "[390/391]\tTime 0.130 (0.179)\tData 0.012 (0.020)\tLoss 0.6233 (0.6188)\tPrec@1 83.750 (86.588)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9467 (0.9467)\tPrec@1 81.250 (81.250)\n",
      "Test: [10/79]\tTime 0.030 (0.025)\tLoss 0.9574 (1.1682)\tPrec@1 75.781 (77.628)\n",
      "Test: [20/79]\tTime 0.021 (0.024)\tLoss 1.1192 (1.2245)\tPrec@1 74.219 (77.455)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.9956 (1.2281)\tPrec@1 83.594 (77.571)\n",
      "Test: [40/79]\tTime 0.026 (0.024)\tLoss 1.4562 (1.2249)\tPrec@1 72.656 (77.572)\n",
      "Test: [50/79]\tTime 0.027 (0.024)\tLoss 1.4307 (1.2097)\tPrec@1 70.312 (77.665)\n",
      "Test: [60/79]\tTime 0.020 (0.024)\tLoss 0.9805 (1.1930)\tPrec@1 83.594 (77.946)\n",
      "Test: [70/79]\tTime 0.022 (0.024)\tLoss 1.2359 (1.1838)\tPrec@1 76.562 (77.883)\n",
      " * Prec@1 78.060\n",
      "\n",
      "===> epoch: 51/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.180 (0.180)\tData 0.021 (0.021)\tLoss 0.7197 (0.7197)\tPrec@1 82.812 (82.812)\n",
      "[10/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.5756 (0.6009)\tPrec@1 89.844 (87.784)\n",
      "[20/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6400 (0.5999)\tPrec@1 85.156 (88.058)\n",
      "[30/391]\tTime 0.177 (0.182)\tData 0.019 (0.020)\tLoss 0.6120 (0.5963)\tPrec@1 85.938 (88.004)\n",
      "[40/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.7590 (0.6023)\tPrec@1 78.906 (87.614)\n",
      "[50/391]\tTime 0.186 (0.182)\tData 0.019 (0.019)\tLoss 0.6204 (0.6033)\tPrec@1 92.188 (87.454)\n",
      "[60/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.5862 (0.6012)\tPrec@1 85.156 (87.372)\n",
      "[70/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6730 (0.6021)\tPrec@1 84.375 (87.379)\n",
      "[80/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.6451 (0.6020)\tPrec@1 85.156 (87.346)\n",
      "[90/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6327 (0.6021)\tPrec@1 86.719 (87.414)\n",
      "[100/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6107 (0.6024)\tPrec@1 88.281 (87.392)\n",
      "[110/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.7089 (0.6066)\tPrec@1 84.375 (87.169)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.4556 (0.6060)\tPrec@1 91.406 (87.151)\n",
      "[130/391]\tTime 0.142 (0.181)\tData 0.019 (0.019)\tLoss 0.6057 (0.6101)\tPrec@1 85.156 (86.904)\n",
      "[140/391]\tTime 0.182 (0.178)\tData 0.019 (0.019)\tLoss 0.5914 (0.6111)\tPrec@1 86.719 (86.924)\n",
      "[150/391]\tTime 0.181 (0.178)\tData 0.020 (0.019)\tLoss 0.5912 (0.6119)\tPrec@1 87.500 (86.910)\n",
      "[160/391]\tTime 0.181 (0.179)\tData 0.019 (0.019)\tLoss 0.6354 (0.6098)\tPrec@1 88.281 (87.000)\n",
      "[170/391]\tTime 0.186 (0.179)\tData 0.020 (0.019)\tLoss 0.6726 (0.6105)\tPrec@1 84.375 (86.975)\n",
      "[180/391]\tTime 0.183 (0.179)\tData 0.019 (0.019)\tLoss 0.5770 (0.6089)\tPrec@1 89.062 (87.051)\n",
      "[190/391]\tTime 0.183 (0.179)\tData 0.019 (0.019)\tLoss 0.5578 (0.6079)\tPrec@1 85.938 (87.071)\n",
      "[200/391]\tTime 0.180 (0.179)\tData 0.019 (0.019)\tLoss 0.5717 (0.6067)\tPrec@1 89.062 (87.119)\n",
      "[210/391]\tTime 0.179 (0.179)\tData 0.019 (0.019)\tLoss 0.6172 (0.6091)\tPrec@1 86.719 (87.030)\n",
      "[220/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6021 (0.6097)\tPrec@1 85.938 (87.026)\n",
      "[230/391]\tTime 0.183 (0.180)\tData 0.021 (0.019)\tLoss 0.7021 (0.6098)\tPrec@1 84.375 (87.003)\n",
      "[240/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.6609 (0.6115)\tPrec@1 86.719 (86.939)\n",
      "[250/391]\tTime 0.185 (0.180)\tData 0.020 (0.019)\tLoss 0.5374 (0.6125)\tPrec@1 89.062 (86.896)\n",
      "[260/391]\tTime 0.182 (0.180)\tData 0.020 (0.019)\tLoss 0.6260 (0.6130)\tPrec@1 86.719 (86.892)\n",
      "[270/391]\tTime 0.187 (0.180)\tData 0.019 (0.019)\tLoss 0.5484 (0.6139)\tPrec@1 91.406 (86.869)\n",
      "[280/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.6850 (0.6156)\tPrec@1 79.688 (86.788)\n",
      "[290/391]\tTime 0.183 (0.180)\tData 0.019 (0.019)\tLoss 0.6010 (0.6164)\tPrec@1 89.844 (86.786)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6549 (0.6166)\tPrec@1 83.594 (86.773)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.5592 (0.6163)\tPrec@1 90.625 (86.807)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.5891 (0.6160)\tPrec@1 88.281 (86.823)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.020 (0.019)\tLoss 0.5778 (0.6156)\tPrec@1 89.062 (86.882)\n",
      "[340/391]\tTime 0.182 (0.180)\tData 0.021 (0.019)\tLoss 0.6058 (0.6159)\tPrec@1 89.062 (86.854)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.6715 (0.6175)\tPrec@1 84.375 (86.781)\n",
      "[360/391]\tTime 0.185 (0.180)\tData 0.020 (0.019)\tLoss 0.6110 (0.6188)\tPrec@1 89.844 (86.747)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.020 (0.019)\tLoss 0.6248 (0.6199)\tPrec@1 89.844 (86.740)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.6522 (0.6197)\tPrec@1 82.812 (86.747)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.014 (0.019)\tLoss 0.7174 (0.6200)\tPrec@1 81.250 (86.716)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.5697 (1.5697)\tPrec@1 72.656 (72.656)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.6328 (1.8735)\tPrec@1 74.219 (69.602)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 1.7124 (1.9171)\tPrec@1 67.969 (69.234)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 1.0659 (1.8530)\tPrec@1 76.562 (70.086)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 1.9343 (1.8186)\tPrec@1 65.625 (70.255)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 2.1082 (1.8107)\tPrec@1 67.969 (70.634)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.5487 (1.8133)\tPrec@1 78.125 (70.594)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.8309 (1.8286)\tPrec@1 66.406 (70.246)\n",
      " * Prec@1 70.250\n",
      "\n",
      "===> epoch: 52/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.176 (0.176)\tData 0.021 (0.021)\tLoss 0.5533 (0.5533)\tPrec@1 90.625 (90.625)\n",
      "[10/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5524 (0.5911)\tPrec@1 89.062 (88.849)\n",
      "[20/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.5616 (0.5892)\tPrec@1 89.844 (88.132)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5495 (0.5880)\tPrec@1 89.062 (88.029)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6208 (0.5903)\tPrec@1 87.500 (87.843)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.7563 (0.6007)\tPrec@1 77.344 (87.485)\n",
      "[60/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5225 (0.5988)\tPrec@1 87.500 (87.321)\n",
      "[70/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.7044 (0.6011)\tPrec@1 85.938 (87.324)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[80/391]\tTime 0.184 (0.180)\tData 0.021 (0.020)\tLoss 0.6768 (0.6009)\tPrec@1 83.594 (87.162)\n",
      "[90/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6213 (0.6030)\tPrec@1 86.719 (87.079)\n",
      "[100/391]\tTime 0.185 (0.180)\tData 0.021 (0.020)\tLoss 0.6293 (0.6057)\tPrec@1 85.156 (86.943)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5269 (0.6076)\tPrec@1 89.062 (86.860)\n",
      "[120/391]\tTime 0.145 (0.180)\tData 0.020 (0.020)\tLoss 0.6829 (0.6076)\tPrec@1 83.594 (86.822)\n",
      "[130/391]\tTime 0.146 (0.177)\tData 0.022 (0.020)\tLoss 0.5357 (0.6053)\tPrec@1 93.750 (86.945)\n",
      "[140/391]\tTime 0.181 (0.177)\tData 0.021 (0.020)\tLoss 0.7281 (0.6094)\tPrec@1 85.156 (86.841)\n",
      "[150/391]\tTime 0.184 (0.178)\tData 0.020 (0.020)\tLoss 0.5895 (0.6107)\tPrec@1 87.500 (86.812)\n",
      "[160/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.7145 (0.6138)\tPrec@1 82.812 (86.690)\n",
      "[170/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.6254 (0.6154)\tPrec@1 85.938 (86.641)\n",
      "[180/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.7107 (0.6141)\tPrec@1 83.594 (86.706)\n",
      "[190/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.5625 (0.6148)\tPrec@1 89.844 (86.694)\n",
      "[200/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.6232 (0.6150)\tPrec@1 89.844 (86.719)\n",
      "[210/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.5841 (0.6137)\tPrec@1 88.281 (86.797)\n",
      "[220/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.4829 (0.6141)\tPrec@1 92.188 (86.768)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.7406 (0.6157)\tPrec@1 81.250 (86.688)\n",
      "[240/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7301 (0.6159)\tPrec@1 78.906 (86.664)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6618 (0.6157)\tPrec@1 82.812 (86.681)\n",
      "[260/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.5804 (0.6166)\tPrec@1 86.719 (86.653)\n",
      "[270/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.6273 (0.6156)\tPrec@1 88.281 (86.707)\n",
      "[280/391]\tTime 0.191 (0.179)\tData 0.020 (0.020)\tLoss 0.6883 (0.6166)\tPrec@1 85.156 (86.699)\n",
      "[290/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6230 (0.6172)\tPrec@1 86.719 (86.711)\n",
      "[300/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.6462 (0.6164)\tPrec@1 84.375 (86.714)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.7106 (0.6162)\tPrec@1 83.594 (86.726)\n",
      "[320/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5110 (0.6162)\tPrec@1 92.969 (86.728)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5863 (0.6167)\tPrec@1 85.938 (86.702)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6164 (0.6170)\tPrec@1 83.594 (86.689)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6051 (0.6185)\tPrec@1 85.156 (86.605)\n",
      "[360/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6350 (0.6177)\tPrec@1 88.281 (86.656)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.7312 (0.6181)\tPrec@1 81.250 (86.643)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6598 (0.6183)\tPrec@1 84.375 (86.624)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.013 (0.020)\tLoss 0.5489 (0.6175)\tPrec@1 90.000 (86.656)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.2626 (1.2626)\tPrec@1 75.000 (75.000)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.2272 (1.3330)\tPrec@1 78.125 (77.273)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.4749 (1.3659)\tPrec@1 79.688 (76.935)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.8649 (1.3378)\tPrec@1 82.031 (77.545)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.3189 (1.3288)\tPrec@1 72.656 (77.458)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 1.2698 (1.3204)\tPrec@1 76.562 (77.512)\n",
      "Test: [60/79]\tTime 0.024 (0.023)\tLoss 0.9276 (1.3219)\tPrec@1 81.250 (77.446)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 1.3449 (1.3248)\tPrec@1 82.031 (77.388)\n",
      " * Prec@1 77.450\n",
      "\n",
      "===> epoch: 53/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.022 (0.022)\tLoss 0.6461 (0.6461)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5368 (0.5739)\tPrec@1 89.844 (88.423)\n",
      "[20/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5921 (0.5870)\tPrec@1 87.500 (87.760)\n",
      "[30/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6408 (0.6000)\tPrec@1 85.156 (87.424)\n",
      "[40/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5183 (0.5948)\tPrec@1 88.281 (87.348)\n",
      "[50/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6280 (0.5971)\tPrec@1 84.375 (87.056)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.7107 (0.5987)\tPrec@1 83.594 (87.077)\n",
      "[70/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7074 (0.6006)\tPrec@1 84.375 (87.115)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5430 (0.6025)\tPrec@1 89.844 (87.133)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6725 (0.6090)\tPrec@1 85.938 (86.908)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5370 (0.6102)\tPrec@1 89.844 (86.951)\n",
      "[110/391]\tTime 0.147 (0.180)\tData 0.019 (0.020)\tLoss 0.5677 (0.6062)\tPrec@1 89.062 (87.064)\n",
      "[120/391]\tTime 0.146 (0.177)\tData 0.019 (0.020)\tLoss 0.5529 (0.6061)\tPrec@1 87.500 (87.074)\n",
      "[130/391]\tTime 0.179 (0.177)\tData 0.020 (0.020)\tLoss 0.7218 (0.6048)\tPrec@1 84.375 (87.148)\n",
      "[140/391]\tTime 0.178 (0.178)\tData 0.019 (0.020)\tLoss 0.7611 (0.6076)\tPrec@1 81.250 (87.035)\n",
      "[150/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.5882 (0.6085)\tPrec@1 88.281 (86.977)\n",
      "[160/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.6602 (0.6103)\tPrec@1 84.375 (86.942)\n",
      "[170/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6477 (0.6131)\tPrec@1 84.375 (86.755)\n",
      "[180/391]\tTime 0.179 (0.178)\tData 0.020 (0.020)\tLoss 0.5911 (0.6137)\tPrec@1 91.406 (86.762)\n",
      "[190/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6882 (0.6146)\tPrec@1 84.375 (86.780)\n",
      "[200/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.4905 (0.6134)\tPrec@1 92.188 (86.800)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.6511 (0.6142)\tPrec@1 85.938 (86.771)\n",
      "[220/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.6687 (0.6138)\tPrec@1 85.938 (86.772)\n",
      "[230/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.4787 (0.6153)\tPrec@1 92.969 (86.739)\n",
      "[240/391]\tTime 0.179 (0.179)\tData 0.021 (0.020)\tLoss 0.6030 (0.6175)\tPrec@1 88.281 (86.644)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6693 (0.6174)\tPrec@1 82.031 (86.650)\n",
      "[260/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5668 (0.6190)\tPrec@1 89.844 (86.590)\n",
      "[270/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5946 (0.6180)\tPrec@1 89.062 (86.612)\n",
      "[280/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5914 (0.6190)\tPrec@1 88.281 (86.569)\n",
      "[290/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.6183 (0.6198)\tPrec@1 86.719 (86.525)\n",
      "[300/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.6606 (0.6200)\tPrec@1 85.938 (86.519)\n",
      "[310/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6838 (0.6217)\tPrec@1 85.156 (86.468)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6955 (0.6217)\tPrec@1 81.250 (86.424)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.4820 (0.6211)\tPrec@1 89.844 (86.431)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6340 (0.6212)\tPrec@1 85.156 (86.437)\n",
      "[350/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6345 (0.6212)\tPrec@1 84.375 (86.409)\n",
      "[360/391]\tTime 0.178 (0.180)\tData 0.028 (0.020)\tLoss 0.5958 (0.6208)\tPrec@1 89.062 (86.437)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5727 (0.6202)\tPrec@1 86.719 (86.468)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6076 (0.6203)\tPrec@1 85.938 (86.460)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.012 (0.020)\tLoss 0.5564 (0.6210)\tPrec@1 90.000 (86.434)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.4944 (1.4944)\tPrec@1 74.219 (74.219)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 1.6389 (1.6827)\tPrec@1 78.906 (73.366)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0448 (1.6348)\tPrec@1 78.125 (74.144)\n",
      "Test: [30/79]\tTime 0.027 (0.023)\tLoss 1.8572 (1.6246)\tPrec@1 71.875 (74.370)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.2835 (1.6059)\tPrec@1 73.438 (74.238)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 1.6319 (1.5837)\tPrec@1 74.219 (74.357)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.3594 (1.5828)\tPrec@1 75.781 (74.385)\n",
      "Test: [70/79]\tTime 0.027 (0.023)\tLoss 1.5388 (1.5835)\tPrec@1 75.781 (74.351)\n",
      " * Prec@1 74.240\n",
      "\n",
      "===> epoch: 54/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.162 (0.162)\tData 0.021 (0.021)\tLoss 0.7388 (0.7388)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.6867 (0.6863)\tPrec@1 86.719 (84.659)\n",
      "[20/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.7053 (0.6631)\tPrec@1 84.375 (84.970)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6018 (0.6503)\tPrec@1 86.719 (85.282)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5193 (0.6331)\tPrec@1 87.500 (85.938)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5690 (0.6281)\tPrec@1 85.156 (85.953)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6313 (0.6247)\tPrec@1 85.156 (85.989)\n",
      "[70/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5645 (0.6255)\tPrec@1 88.281 (86.147)\n",
      "[80/391]\tTime 0.178 (0.180)\tData 0.020 (0.020)\tLoss 0.5346 (0.6232)\tPrec@1 90.625 (86.265)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5513 (0.6200)\tPrec@1 92.188 (86.427)\n",
      "[100/391]\tTime 0.143 (0.180)\tData 0.019 (0.020)\tLoss 0.7177 (0.6172)\tPrec@1 80.469 (86.494)\n",
      "[110/391]\tTime 0.143 (0.176)\tData 0.020 (0.020)\tLoss 0.6466 (0.6156)\tPrec@1 85.156 (86.599)\n",
      "[120/391]\tTime 0.181 (0.177)\tData 0.019 (0.020)\tLoss 0.6107 (0.6152)\tPrec@1 85.938 (86.667)\n",
      "[130/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.6418 (0.6187)\tPrec@1 85.938 (86.534)\n",
      "[140/391]\tTime 0.189 (0.178)\tData 0.020 (0.020)\tLoss 0.6437 (0.6196)\tPrec@1 85.156 (86.514)\n",
      "[150/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.6285 (0.6212)\tPrec@1 90.625 (86.507)\n",
      "[160/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6291 (0.6217)\tPrec@1 86.719 (86.530)\n",
      "[170/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6448 (0.6226)\tPrec@1 86.719 (86.513)\n",
      "[180/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5831 (0.6215)\tPrec@1 88.281 (86.598)\n",
      "[190/391]\tTime 0.187 (0.179)\tData 0.019 (0.020)\tLoss 0.6040 (0.6207)\tPrec@1 88.281 (86.621)\n",
      "[200/391]\tTime 0.188 (0.179)\tData 0.026 (0.020)\tLoss 0.6560 (0.6206)\tPrec@1 86.719 (86.598)\n",
      "[210/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.6206 (0.6196)\tPrec@1 88.281 (86.656)\n",
      "[220/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.7466 (0.6203)\tPrec@1 82.812 (86.623)\n",
      "[230/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6390 (0.6204)\tPrec@1 83.594 (86.641)\n",
      "[240/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5175 (0.6204)\tPrec@1 92.188 (86.644)\n",
      "[250/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.5285 (0.6204)\tPrec@1 92.188 (86.681)\n",
      "[260/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6685 (0.6197)\tPrec@1 84.375 (86.692)\n",
      "[270/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.7420 (0.6192)\tPrec@1 81.250 (86.727)\n",
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5983 (0.6199)\tPrec@1 85.156 (86.677)\n",
      "[290/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6423 (0.6199)\tPrec@1 84.375 (86.684)\n",
      "[300/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.8202 (0.6209)\tPrec@1 76.562 (86.654)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6620 (0.6212)\tPrec@1 84.375 (86.626)\n",
      "[320/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.8088 (0.6226)\tPrec@1 78.906 (86.582)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5546 (0.6212)\tPrec@1 89.062 (86.615)\n",
      "[340/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5608 (0.6218)\tPrec@1 90.625 (86.604)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6237 (0.6217)\tPrec@1 85.938 (86.583)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6018 (0.6219)\tPrec@1 85.938 (86.574)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6972 (0.6218)\tPrec@1 85.156 (86.586)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6160 (0.6213)\tPrec@1 88.281 (86.608)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.015 (0.020)\tLoss 0.5936 (0.6211)\tPrec@1 88.750 (86.618)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.0872 (1.0872)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.0413 (1.0085)\tPrec@1 81.250 (81.392)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.7943 (1.0047)\tPrec@1 81.250 (81.510)\n",
      "Test: [30/79]\tTime 0.024 (0.023)\tLoss 0.8151 (1.0085)\tPrec@1 85.938 (81.376)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.4414 (1.0092)\tPrec@1 74.219 (81.193)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 0.9069 (0.9936)\tPrec@1 80.469 (81.419)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.8354 (0.9855)\tPrec@1 83.594 (81.263)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 0.7106 (0.9840)\tPrec@1 86.719 (81.228)\n",
      " * Prec@1 81.370\n",
      "\n",
      "===> epoch: 55/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.174 (0.174)\tData 0.021 (0.021)\tLoss 0.6680 (0.6680)\tPrec@1 85.156 (85.156)\n",
      "[10/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6582 (0.6185)\tPrec@1 85.156 (85.795)\n",
      "[20/391]\tTime 0.185 (0.181)\tData 0.019 (0.019)\tLoss 0.6476 (0.6308)\tPrec@1 85.156 (85.603)\n",
      "[30/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5860 (0.6356)\tPrec@1 89.844 (85.635)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6134 (0.6365)\tPrec@1 89.844 (85.899)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6456 (0.6318)\tPrec@1 83.594 (86.045)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.5558 (0.6279)\tPrec@1 89.062 (86.066)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6480 (0.6271)\tPrec@1 85.938 (86.070)\n",
      "[80/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.6639 (0.6253)\tPrec@1 83.594 (86.063)\n",
      "[90/391]\tTime 0.148 (0.180)\tData 0.019 (0.020)\tLoss 0.6777 (0.6224)\tPrec@1 84.375 (86.169)\n",
      "[100/391]\tTime 0.144 (0.177)\tData 0.019 (0.020)\tLoss 0.7052 (0.6244)\tPrec@1 84.375 (86.077)\n",
      "[110/391]\tTime 0.182 (0.176)\tData 0.020 (0.020)\tLoss 0.5762 (0.6245)\tPrec@1 89.844 (86.099)\n",
      "[120/391]\tTime 0.185 (0.177)\tData 0.019 (0.020)\tLoss 0.5096 (0.6204)\tPrec@1 92.188 (86.254)\n",
      "[130/391]\tTime 0.181 (0.177)\tData 0.019 (0.020)\tLoss 0.6606 (0.6224)\tPrec@1 84.375 (86.206)\n",
      "[140/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5853 (0.6210)\tPrec@1 87.500 (86.242)\n",
      "[150/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.6647 (0.6223)\tPrec@1 83.594 (86.227)\n",
      "[160/391]\tTime 0.181 (0.178)\tData 0.019 (0.020)\tLoss 0.6747 (0.6239)\tPrec@1 84.375 (86.175)\n",
      "[170/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.7147 (0.6245)\tPrec@1 83.594 (86.202)\n",
      "[180/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.7060 (0.6276)\tPrec@1 82.812 (86.106)\n",
      "[190/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6948 (0.6259)\tPrec@1 83.594 (86.207)\n",
      "[200/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5894 (0.6264)\tPrec@1 86.719 (86.202)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.6040 (0.6265)\tPrec@1 89.062 (86.226)\n",
      "[220/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5713 (0.6259)\tPrec@1 87.500 (86.249)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5348 (0.6261)\tPrec@1 89.062 (86.235)\n",
      "[240/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5632 (0.6252)\tPrec@1 89.062 (86.265)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6778 (0.6245)\tPrec@1 85.156 (86.317)\n",
      "[260/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5836 (0.6243)\tPrec@1 88.281 (86.351)\n",
      "[270/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6645 (0.6254)\tPrec@1 87.500 (86.318)\n",
      "[280/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6044 (0.6238)\tPrec@1 88.281 (86.380)\n",
      "[290/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6614 (0.6226)\tPrec@1 84.375 (86.429)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.023 (0.020)\tLoss 0.6155 (0.6233)\tPrec@1 86.719 (86.420)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5668 (0.6231)\tPrec@1 89.844 (86.465)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5492 (0.6230)\tPrec@1 90.625 (86.483)\n",
      "[330/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6237 (0.6236)\tPrec@1 89.062 (86.466)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.5970 (0.6229)\tPrec@1 87.500 (86.492)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.6459 (0.6239)\tPrec@1 85.156 (86.476)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6943 (0.6238)\tPrec@1 85.938 (86.476)\n",
      "[370/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5432 (0.6233)\tPrec@1 89.844 (86.508)\n",
      "[380/391]\tTime 0.189 (0.180)\tData 0.019 (0.020)\tLoss 0.5369 (0.6226)\tPrec@1 89.062 (86.544)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.012 (0.020)\tLoss 0.5396 (0.6224)\tPrec@1 90.000 (86.566)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.0477 (1.0477)\tPrec@1 77.344 (77.344)\n",
      "Test: [10/79]\tTime 0.024 (0.022)\tLoss 0.7541 (1.0816)\tPrec@1 85.938 (80.398)\n",
      "Test: [20/79]\tTime 0.025 (0.022)\tLoss 1.1061 (1.1470)\tPrec@1 79.688 (79.539)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.9530 (1.1348)\tPrec@1 78.906 (79.688)\n",
      "Test: [40/79]\tTime 0.024 (0.023)\tLoss 1.3569 (1.1155)\tPrec@1 78.125 (80.050)\n",
      "Test: [50/79]\tTime 0.025 (0.023)\tLoss 1.7010 (1.1349)\tPrec@1 77.344 (80.070)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [60/79]\tTime 0.019 (0.023)\tLoss 1.2438 (1.1420)\tPrec@1 81.250 (80.097)\n",
      "Test: [70/79]\tTime 0.021 (0.023)\tLoss 0.8732 (1.1569)\tPrec@1 82.031 (79.754)\n",
      " * Prec@1 79.700\n",
      "\n",
      "===> epoch: 56/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.020 (0.020)\tLoss 0.6426 (0.6426)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.181 (0.183)\tData 0.020 (0.020)\tLoss 0.5477 (0.6287)\tPrec@1 90.625 (85.582)\n",
      "[20/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.6086 (0.6277)\tPrec@1 88.281 (86.124)\n",
      "[30/391]\tTime 0.186 (0.182)\tData 0.019 (0.019)\tLoss 0.7020 (0.6184)\tPrec@1 82.031 (86.694)\n",
      "[40/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.5178 (0.6080)\tPrec@1 89.062 (86.986)\n",
      "[50/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6272 (0.6056)\tPrec@1 82.812 (87.025)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.5654 (0.6040)\tPrec@1 86.719 (87.039)\n",
      "[70/391]\tTime 0.189 (0.183)\tData 0.020 (0.019)\tLoss 0.6221 (0.6017)\tPrec@1 85.156 (87.137)\n",
      "[80/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5236 (0.5992)\tPrec@1 91.406 (87.249)\n",
      "[90/391]\tTime 0.144 (0.178)\tData 0.020 (0.019)\tLoss 0.6052 (0.5981)\tPrec@1 88.281 (87.320)\n",
      "[100/391]\tTime 0.195 (0.177)\tData 0.027 (0.019)\tLoss 0.5499 (0.5971)\tPrec@1 90.625 (87.353)\n",
      "[110/391]\tTime 0.181 (0.178)\tData 0.019 (0.019)\tLoss 0.6592 (0.6005)\tPrec@1 83.594 (87.289)\n",
      "[120/391]\tTime 0.182 (0.178)\tData 0.019 (0.019)\tLoss 0.5098 (0.6022)\tPrec@1 89.844 (87.306)\n",
      "[130/391]\tTime 0.182 (0.179)\tData 0.019 (0.019)\tLoss 0.5726 (0.6052)\tPrec@1 90.625 (87.232)\n",
      "[140/391]\tTime 0.179 (0.179)\tData 0.019 (0.019)\tLoss 0.5774 (0.6035)\tPrec@1 87.500 (87.212)\n",
      "[150/391]\tTime 0.181 (0.179)\tData 0.020 (0.019)\tLoss 0.5626 (0.6045)\tPrec@1 89.844 (87.127)\n",
      "[160/391]\tTime 0.180 (0.179)\tData 0.019 (0.019)\tLoss 0.5965 (0.6067)\tPrec@1 84.375 (87.058)\n",
      "[170/391]\tTime 0.186 (0.179)\tData 0.019 (0.019)\tLoss 0.5588 (0.6091)\tPrec@1 90.625 (86.961)\n",
      "[180/391]\tTime 0.186 (0.179)\tData 0.019 (0.019)\tLoss 0.5740 (0.6088)\tPrec@1 89.844 (86.939)\n",
      "[190/391]\tTime 0.180 (0.179)\tData 0.020 (0.019)\tLoss 0.5957 (0.6086)\tPrec@1 89.062 (86.923)\n",
      "[200/391]\tTime 0.180 (0.179)\tData 0.020 (0.019)\tLoss 0.6457 (0.6082)\tPrec@1 85.938 (86.944)\n",
      "[210/391]\tTime 0.178 (0.179)\tData 0.020 (0.019)\tLoss 0.5383 (0.6088)\tPrec@1 91.406 (86.896)\n",
      "[220/391]\tTime 0.180 (0.180)\tData 0.020 (0.019)\tLoss 0.6272 (0.6104)\tPrec@1 85.156 (86.835)\n",
      "[230/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6145 (0.6113)\tPrec@1 83.594 (86.780)\n",
      "[240/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5843 (0.6121)\tPrec@1 85.938 (86.751)\n",
      "[250/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6873 (0.6131)\tPrec@1 83.594 (86.769)\n",
      "[260/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.7579 (0.6154)\tPrec@1 82.812 (86.704)\n",
      "[270/391]\tTime 0.181 (0.180)\tData 0.020 (0.019)\tLoss 0.5773 (0.6150)\tPrec@1 88.281 (86.739)\n",
      "[280/391]\tTime 0.178 (0.180)\tData 0.019 (0.019)\tLoss 0.5043 (0.6143)\tPrec@1 92.188 (86.752)\n",
      "[290/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5837 (0.6142)\tPrec@1 85.938 (86.724)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6224 (0.6149)\tPrec@1 82.812 (86.698)\n",
      "[310/391]\tTime 0.184 (0.180)\tData 0.021 (0.020)\tLoss 0.7508 (0.6148)\tPrec@1 80.469 (86.689)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5940 (0.6156)\tPrec@1 88.281 (86.638)\n",
      "[330/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6648 (0.6153)\tPrec@1 85.156 (86.662)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6142 (0.6161)\tPrec@1 85.156 (86.629)\n",
      "[350/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5576 (0.6175)\tPrec@1 88.281 (86.583)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7706 (0.6189)\tPrec@1 82.031 (86.528)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6475 (0.6198)\tPrec@1 82.031 (86.483)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.8388 (0.6206)\tPrec@1 73.438 (86.417)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.7360 (0.6211)\tPrec@1 81.250 (86.414)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.0061 (1.0061)\tPrec@1 82.031 (82.031)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 0.8684 (1.1234)\tPrec@1 84.375 (80.114)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 1.2543 (1.1515)\tPrec@1 79.688 (79.650)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.8310 (1.1128)\tPrec@1 85.938 (80.292)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.3561 (1.1300)\tPrec@1 75.000 (80.183)\n",
      "Test: [50/79]\tTime 0.023 (0.023)\tLoss 1.0898 (1.1146)\tPrec@1 82.031 (80.239)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.7747 (1.1133)\tPrec@1 85.156 (80.149)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.8187 (1.1322)\tPrec@1 83.594 (79.941)\n",
      " * Prec@1 80.030\n",
      "\n",
      "===> epoch: 57/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.165 (0.165)\tData 0.020 (0.020)\tLoss 0.5646 (0.5646)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.7513 (0.6498)\tPrec@1 80.469 (85.156)\n",
      "[20/391]\tTime 0.192 (0.183)\tData 0.020 (0.020)\tLoss 0.5567 (0.6249)\tPrec@1 88.281 (86.198)\n",
      "[30/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5616 (0.6206)\tPrec@1 89.844 (86.517)\n",
      "[40/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6242 (0.6161)\tPrec@1 86.719 (86.604)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5632 (0.6060)\tPrec@1 89.062 (87.056)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6552 (0.6064)\tPrec@1 83.594 (86.936)\n",
      "[70/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5779 (0.6039)\tPrec@1 89.062 (87.082)\n",
      "[80/391]\tTime 0.144 (0.178)\tData 0.020 (0.020)\tLoss 0.5551 (0.6005)\tPrec@1 87.500 (87.230)\n",
      "[90/391]\tTime 0.180 (0.176)\tData 0.020 (0.020)\tLoss 0.5900 (0.6038)\tPrec@1 85.156 (87.019)\n",
      "[100/391]\tTime 0.180 (0.177)\tData 0.020 (0.020)\tLoss 0.6208 (0.6051)\tPrec@1 85.938 (87.051)\n",
      "[110/391]\tTime 0.196 (0.177)\tData 0.033 (0.020)\tLoss 0.5889 (0.6046)\tPrec@1 89.844 (87.035)\n",
      "[120/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6624 (0.6061)\tPrec@1 83.594 (86.996)\n",
      "[130/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.5177 (0.6051)\tPrec@1 89.062 (87.023)\n",
      "[140/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.7131 (0.6080)\tPrec@1 85.938 (87.007)\n",
      "[150/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5816 (0.6068)\tPrec@1 88.281 (87.081)\n",
      "[160/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5832 (0.6070)\tPrec@1 89.062 (87.126)\n",
      "[170/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6126 (0.6074)\tPrec@1 86.719 (87.153)\n",
      "[180/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5707 (0.6078)\tPrec@1 88.281 (87.150)\n",
      "[190/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6355 (0.6085)\tPrec@1 85.156 (87.144)\n",
      "[200/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.5700 (0.6079)\tPrec@1 85.938 (87.123)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.021 (0.020)\tLoss 0.6025 (0.6082)\tPrec@1 85.938 (87.122)\n",
      "[220/391]\tTime 0.185 (0.179)\tData 0.021 (0.020)\tLoss 0.6334 (0.6092)\tPrec@1 86.719 (87.069)\n",
      "[230/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7095 (0.6104)\tPrec@1 85.938 (87.074)\n",
      "[240/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.7340 (0.6114)\tPrec@1 80.469 (87.036)\n",
      "[250/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6159 (0.6131)\tPrec@1 85.938 (86.958)\n",
      "[260/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5564 (0.6136)\tPrec@1 92.188 (86.961)\n",
      "[270/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5847 (0.6131)\tPrec@1 89.062 (86.998)\n",
      "[280/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6420 (0.6134)\tPrec@1 87.500 (87.000)\n",
      "[290/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6461 (0.6150)\tPrec@1 86.719 (86.925)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6937 (0.6165)\tPrec@1 86.719 (86.874)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7046 (0.6185)\tPrec@1 83.594 (86.807)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5614 (0.6186)\tPrec@1 88.281 (86.806)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5138 (0.6186)\tPrec@1 89.062 (86.823)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6322 (0.6189)\tPrec@1 82.031 (86.808)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6448 (0.6188)\tPrec@1 86.719 (86.797)\n",
      "[360/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5280 (0.6188)\tPrec@1 91.406 (86.784)\n",
      "[370/391]\tTime 0.178 (0.180)\tData 0.020 (0.020)\tLoss 0.6939 (0.6192)\tPrec@1 85.938 (86.763)\n",
      "[380/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5494 (0.6196)\tPrec@1 89.062 (86.750)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.012 (0.020)\tLoss 0.6110 (0.6199)\tPrec@1 85.000 (86.724)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9095 (0.9095)\tPrec@1 82.031 (82.031)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [10/79]\tTime 0.026 (0.022)\tLoss 1.2807 (0.9803)\tPrec@1 81.250 (82.102)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.7561 (1.0395)\tPrec@1 84.375 (81.101)\n",
      "Test: [30/79]\tTime 0.020 (0.023)\tLoss 0.8802 (1.0176)\tPrec@1 81.250 (81.578)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 1.0337 (1.0344)\tPrec@1 77.344 (81.345)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.1352 (1.0273)\tPrec@1 82.031 (81.373)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.1182 (1.0269)\tPrec@1 80.469 (81.519)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.7764 (1.0226)\tPrec@1 83.594 (81.448)\n",
      " * Prec@1 81.550\n",
      "\n",
      "===> epoch: 58/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.177 (0.177)\tData 0.020 (0.020)\tLoss 0.6628 (0.6628)\tPrec@1 85.156 (85.156)\n",
      "[10/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6026 (0.6569)\tPrec@1 88.281 (86.435)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6792 (0.6318)\tPrec@1 84.375 (86.979)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5794 (0.6166)\tPrec@1 88.281 (87.172)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6200 (0.6063)\tPrec@1 85.938 (87.519)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5934 (0.6090)\tPrec@1 85.156 (87.240)\n",
      "[60/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6552 (0.6051)\tPrec@1 81.250 (87.321)\n",
      "[70/391]\tTime 0.142 (0.178)\tData 0.020 (0.020)\tLoss 0.6021 (0.6107)\tPrec@1 85.156 (87.093)\n",
      "[80/391]\tTime 0.182 (0.175)\tData 0.020 (0.020)\tLoss 0.6469 (0.6116)\tPrec@1 84.375 (87.085)\n",
      "[90/391]\tTime 0.179 (0.176)\tData 0.019 (0.020)\tLoss 0.6831 (0.6140)\tPrec@1 80.469 (86.908)\n",
      "[100/391]\tTime 0.182 (0.176)\tData 0.020 (0.020)\tLoss 0.6362 (0.6121)\tPrec@1 87.500 (87.067)\n",
      "[110/391]\tTime 0.179 (0.177)\tData 0.019 (0.020)\tLoss 0.7192 (0.6115)\tPrec@1 82.031 (87.057)\n",
      "[120/391]\tTime 0.187 (0.177)\tData 0.019 (0.020)\tLoss 0.6026 (0.6109)\tPrec@1 88.281 (87.087)\n",
      "[130/391]\tTime 0.185 (0.178)\tData 0.020 (0.020)\tLoss 0.6365 (0.6124)\tPrec@1 85.156 (87.041)\n",
      "[140/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.6310 (0.6144)\tPrec@1 87.500 (86.985)\n",
      "[150/391]\tTime 0.181 (0.178)\tData 0.020 (0.020)\tLoss 0.6734 (0.6174)\tPrec@1 82.031 (86.817)\n",
      "[160/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.7609 (0.6180)\tPrec@1 78.906 (86.796)\n",
      "[170/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6430 (0.6171)\tPrec@1 87.500 (86.847)\n",
      "[180/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.5684 (0.6168)\tPrec@1 88.281 (86.887)\n",
      "[190/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.5985 (0.6168)\tPrec@1 86.719 (86.862)\n",
      "[200/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.5271 (0.6149)\tPrec@1 89.844 (86.940)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.021 (0.020)\tLoss 0.5410 (0.6144)\tPrec@1 89.844 (86.971)\n",
      "[220/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5805 (0.6139)\tPrec@1 92.188 (87.012)\n",
      "[230/391]\tTime 0.178 (0.179)\tData 0.022 (0.020)\tLoss 0.6449 (0.6145)\tPrec@1 87.500 (86.983)\n",
      "[240/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.6441 (0.6141)\tPrec@1 87.500 (87.023)\n",
      "[250/391]\tTime 0.183 (0.179)\tData 0.021 (0.020)\tLoss 0.5573 (0.6133)\tPrec@1 90.625 (87.042)\n",
      "[260/391]\tTime 0.181 (0.179)\tData 0.022 (0.020)\tLoss 0.6125 (0.6138)\tPrec@1 85.156 (87.012)\n",
      "[270/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.5352 (0.6142)\tPrec@1 91.406 (86.996)\n",
      "[280/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6316 (0.6149)\tPrec@1 87.500 (86.986)\n",
      "[290/391]\tTime 0.179 (0.179)\tData 0.021 (0.020)\tLoss 0.7151 (0.6145)\tPrec@1 80.469 (86.993)\n",
      "[300/391]\tTime 0.188 (0.179)\tData 0.019 (0.020)\tLoss 0.6875 (0.6162)\tPrec@1 85.156 (86.934)\n",
      "[310/391]\tTime 0.186 (0.180)\tData 0.021 (0.020)\tLoss 0.6038 (0.6161)\tPrec@1 85.938 (86.927)\n",
      "[320/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.4787 (0.6162)\tPrec@1 93.750 (86.911)\n",
      "[330/391]\tTime 0.179 (0.180)\tData 0.021 (0.020)\tLoss 0.7052 (0.6175)\tPrec@1 82.031 (86.849)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6264 (0.6179)\tPrec@1 86.719 (86.815)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6614 (0.6179)\tPrec@1 85.938 (86.823)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5089 (0.6178)\tPrec@1 89.844 (86.825)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6283 (0.6180)\tPrec@1 84.375 (86.795)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5758 (0.6183)\tPrec@1 87.500 (86.778)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.013 (0.020)\tLoss 0.6677 (0.6181)\tPrec@1 86.250 (86.784)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7576 (0.7576)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.022 (0.022)\tLoss 0.6548 (0.8063)\tPrec@1 85.938 (83.168)\n",
      "Test: [20/79]\tTime 0.024 (0.022)\tLoss 0.5527 (0.8374)\tPrec@1 86.719 (83.110)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.5407 (0.8271)\tPrec@1 89.062 (83.493)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.7463 (0.8259)\tPrec@1 82.812 (83.689)\n",
      "Test: [50/79]\tTime 0.025 (0.022)\tLoss 0.9332 (0.8214)\tPrec@1 81.250 (83.793)\n",
      "Test: [60/79]\tTime 0.020 (0.022)\tLoss 0.8417 (0.8232)\tPrec@1 83.594 (83.709)\n",
      "Test: [70/79]\tTime 0.022 (0.022)\tLoss 0.9243 (0.8268)\tPrec@1 82.812 (83.550)\n",
      " * Prec@1 83.620\n",
      "\n",
      "===> epoch: 59/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6106 (0.6106)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6563 (0.6046)\tPrec@1 87.500 (87.287)\n",
      "[20/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6196 (0.6118)\tPrec@1 89.844 (86.868)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7232 (0.6078)\tPrec@1 80.469 (87.198)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6012 (0.6110)\tPrec@1 86.719 (87.176)\n",
      "[50/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5433 (0.6065)\tPrec@1 89.844 (87.423)\n",
      "[60/391]\tTime 0.144 (0.179)\tData 0.020 (0.020)\tLoss 0.7284 (0.6084)\tPrec@1 83.594 (87.449)\n",
      "[70/391]\tTime 0.182 (0.174)\tData 0.019 (0.020)\tLoss 0.5983 (0.6070)\tPrec@1 85.938 (87.401)\n",
      "[80/391]\tTime 0.178 (0.175)\tData 0.019 (0.019)\tLoss 0.6321 (0.6054)\tPrec@1 85.156 (87.432)\n",
      "[90/391]\tTime 0.180 (0.176)\tData 0.020 (0.019)\tLoss 0.6156 (0.6078)\tPrec@1 84.375 (87.217)\n",
      "[100/391]\tTime 0.180 (0.177)\tData 0.020 (0.020)\tLoss 0.6148 (0.6107)\tPrec@1 89.062 (87.067)\n",
      "[110/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.5544 (0.6072)\tPrec@1 88.281 (87.176)\n",
      "[120/391]\tTime 0.184 (0.178)\tData 0.020 (0.020)\tLoss 0.6958 (0.6072)\tPrec@1 85.938 (87.132)\n",
      "[130/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.6512 (0.6085)\tPrec@1 85.156 (87.118)\n",
      "[140/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.5870 (0.6057)\tPrec@1 87.500 (87.129)\n",
      "[150/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6023 (0.6076)\tPrec@1 89.062 (87.081)\n",
      "[160/391]\tTime 0.179 (0.179)\tData 0.021 (0.020)\tLoss 0.5334 (0.6070)\tPrec@1 91.406 (87.097)\n",
      "[170/391]\tTime 0.178 (0.179)\tData 0.020 (0.020)\tLoss 0.5476 (0.6064)\tPrec@1 88.281 (87.075)\n",
      "[180/391]\tTime 0.184 (0.179)\tData 0.021 (0.020)\tLoss 0.5413 (0.6079)\tPrec@1 89.844 (87.030)\n",
      "[190/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.6542 (0.6096)\tPrec@1 81.250 (86.915)\n",
      "[200/391]\tTime 0.179 (0.179)\tData 0.021 (0.020)\tLoss 0.5795 (0.6089)\tPrec@1 85.938 (86.940)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.022 (0.020)\tLoss 0.6863 (0.6104)\tPrec@1 85.938 (86.896)\n",
      "[220/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5665 (0.6108)\tPrec@1 89.844 (86.885)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6848 (0.6134)\tPrec@1 83.594 (86.830)\n",
      "[240/391]\tTime 0.187 (0.179)\tData 0.020 (0.020)\tLoss 0.5958 (0.6147)\tPrec@1 87.500 (86.767)\n",
      "[250/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.6821 (0.6144)\tPrec@1 78.906 (86.734)\n",
      "[260/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6576 (0.6153)\tPrec@1 84.375 (86.692)\n",
      "[270/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6939 (0.6153)\tPrec@1 82.031 (86.673)\n",
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6871 (0.6152)\tPrec@1 84.375 (86.669)\n",
      "[290/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6272 (0.6156)\tPrec@1 87.500 (86.649)\n",
      "[300/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6409 (0.6154)\tPrec@1 85.938 (86.636)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5324 (0.6154)\tPrec@1 90.625 (86.653)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6728 (0.6162)\tPrec@1 85.156 (86.595)\n",
      "[330/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5856 (0.6164)\tPrec@1 87.500 (86.613)\n",
      "[340/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6248 (0.6170)\tPrec@1 86.719 (86.611)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.4963 (0.6164)\tPrec@1 92.188 (86.634)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5903 (0.6161)\tPrec@1 90.625 (86.649)\n",
      "[370/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6166 (0.6163)\tPrec@1 88.281 (86.643)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5742 (0.6170)\tPrec@1 86.719 (86.622)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.6345 (0.6178)\tPrec@1 85.000 (86.578)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.2384 (1.2384)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 1.1822 (1.3319)\tPrec@1 75.781 (77.415)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.0413 (1.3441)\tPrec@1 83.594 (77.121)\n",
      "Test: [30/79]\tTime 0.027 (0.023)\tLoss 0.9198 (1.3441)\tPrec@1 78.125 (77.193)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.6891 (1.3049)\tPrec@1 75.000 (77.572)\n",
      "Test: [50/79]\tTime 0.023 (0.023)\tLoss 1.7103 (1.3071)\tPrec@1 72.656 (77.512)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 1.3271 (1.2954)\tPrec@1 84.375 (77.907)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.2681 (1.2961)\tPrec@1 80.469 (78.037)\n",
      " * Prec@1 78.280\n",
      "\n",
      "===> epoch: 60/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.153 (0.153)\tData 0.020 (0.020)\tLoss 0.5749 (0.5749)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.181 (0.178)\tData 0.019 (0.019)\tLoss 0.5592 (0.6132)\tPrec@1 87.500 (87.216)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5391 (0.6145)\tPrec@1 89.062 (86.942)\n",
      "[30/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5816 (0.6018)\tPrec@1 88.281 (87.298)\n",
      "[40/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5718 (0.5932)\tPrec@1 89.062 (87.557)\n",
      "[50/391]\tTime 0.142 (0.180)\tData 0.019 (0.020)\tLoss 0.5962 (0.5898)\tPrec@1 88.281 (87.791)\n",
      "[60/391]\tTime 0.147 (0.174)\tData 0.019 (0.020)\tLoss 0.6038 (0.5897)\tPrec@1 88.281 (87.948)\n",
      "[70/391]\tTime 0.180 (0.175)\tData 0.019 (0.020)\tLoss 0.6159 (0.5940)\tPrec@1 84.375 (87.599)\n",
      "[80/391]\tTime 0.183 (0.176)\tData 0.019 (0.020)\tLoss 0.6481 (0.5990)\tPrec@1 86.719 (87.490)\n",
      "[90/391]\tTime 0.183 (0.177)\tData 0.020 (0.020)\tLoss 0.6529 (0.6019)\tPrec@1 85.938 (87.363)\n",
      "[100/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6518 (0.6064)\tPrec@1 84.375 (87.121)\n",
      "[110/391]\tTime 0.179 (0.178)\tData 0.020 (0.020)\tLoss 0.5295 (0.6055)\tPrec@1 90.625 (87.120)\n",
      "[120/391]\tTime 0.181 (0.178)\tData 0.021 (0.020)\tLoss 0.6336 (0.6074)\tPrec@1 83.594 (87.087)\n",
      "[130/391]\tTime 0.180 (0.178)\tData 0.020 (0.020)\tLoss 0.5890 (0.6083)\tPrec@1 84.375 (86.969)\n",
      "[140/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.6005 (0.6083)\tPrec@1 89.062 (87.051)\n",
      "[150/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.7208 (0.6086)\tPrec@1 84.375 (87.034)\n",
      "[160/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7837 (0.6119)\tPrec@1 80.469 (86.840)\n",
      "[170/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.6601 (0.6135)\tPrec@1 85.938 (86.796)\n",
      "[180/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5848 (0.6140)\tPrec@1 89.844 (86.784)\n",
      "[190/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5792 (0.6147)\tPrec@1 87.500 (86.792)\n",
      "[200/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.5293 (0.6152)\tPrec@1 89.844 (86.762)\n",
      "[210/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.5713 (0.6161)\tPrec@1 89.062 (86.726)\n",
      "[220/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5705 (0.6163)\tPrec@1 86.719 (86.740)\n",
      "[230/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6718 (0.6173)\tPrec@1 85.156 (86.705)\n",
      "[240/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6331 (0.6195)\tPrec@1 85.938 (86.599)\n",
      "[250/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5714 (0.6180)\tPrec@1 88.281 (86.647)\n",
      "[260/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6495 (0.6176)\tPrec@1 86.719 (86.671)\n",
      "[270/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5640 (0.6174)\tPrec@1 86.719 (86.684)\n",
      "[280/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6712 (0.6182)\tPrec@1 86.719 (86.660)\n",
      "[290/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6277 (0.6185)\tPrec@1 83.594 (86.617)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5719 (0.6192)\tPrec@1 88.281 (86.581)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5244 (0.6195)\tPrec@1 90.625 (86.593)\n",
      "[320/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.6300 (0.6200)\tPrec@1 89.062 (86.580)\n",
      "[330/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5986 (0.6207)\tPrec@1 87.500 (86.584)\n",
      "[340/391]\tTime 0.186 (0.181)\tData 0.021 (0.020)\tLoss 0.5879 (0.6207)\tPrec@1 89.062 (86.595)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5397 (0.6212)\tPrec@1 89.844 (86.567)\n",
      "[360/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6084 (0.6209)\tPrec@1 86.719 (86.576)\n",
      "[370/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.7308 (0.6211)\tPrec@1 82.812 (86.584)\n",
      "[380/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7035 (0.6209)\tPrec@1 82.812 (86.575)\n",
      "[390/391]\tTime 0.129 (0.181)\tData 0.013 (0.020)\tLoss 0.7098 (0.6206)\tPrec@1 85.000 (86.598)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.5162 (1.5162)\tPrec@1 74.219 (74.219)\n",
      "Test: [10/79]\tTime 0.019 (0.022)\tLoss 1.7328 (1.5714)\tPrec@1 74.219 (73.864)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 1.0564 (1.6033)\tPrec@1 80.469 (73.661)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 1.4373 (1.5888)\tPrec@1 75.000 (73.967)\n",
      "Test: [40/79]\tTime 0.022 (0.022)\tLoss 2.0350 (1.5879)\tPrec@1 64.844 (73.990)\n",
      "Test: [50/79]\tTime 0.026 (0.022)\tLoss 1.7459 (1.5954)\tPrec@1 72.656 (73.912)\n",
      "Test: [60/79]\tTime 0.020 (0.022)\tLoss 1.6858 (1.6049)\tPrec@1 73.438 (73.886)\n",
      "Test: [70/79]\tTime 0.021 (0.023)\tLoss 1.3007 (1.6005)\tPrec@1 74.219 (73.757)\n",
      " * Prec@1 74.020\n",
      "\n",
      "===> epoch: 61/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.023 (0.023)\tLoss 0.4662 (0.4662)\tPrec@1 92.969 (92.969)\n",
      "[10/391]\tTime 0.181 (0.184)\tData 0.019 (0.020)\tLoss 0.6420 (0.5867)\tPrec@1 86.719 (88.139)\n",
      "[20/391]\tTime 0.181 (0.183)\tData 0.019 (0.019)\tLoss 0.5910 (0.5872)\tPrec@1 84.375 (87.946)\n",
      "[30/391]\tTime 0.181 (0.183)\tData 0.020 (0.020)\tLoss 0.6168 (0.6043)\tPrec@1 84.375 (87.298)\n",
      "[40/391]\tTime 0.155 (0.182)\tData 0.019 (0.020)\tLoss 0.6473 (0.6105)\tPrec@1 86.719 (87.043)\n",
      "[50/391]\tTime 0.143 (0.174)\tData 0.019 (0.020)\tLoss 0.6461 (0.6142)\tPrec@1 86.719 (86.964)\n",
      "[60/391]\tTime 0.184 (0.175)\tData 0.019 (0.020)\tLoss 0.6782 (0.6096)\tPrec@1 82.812 (87.077)\n",
      "[70/391]\tTime 0.179 (0.176)\tData 0.019 (0.020)\tLoss 0.6953 (0.6079)\tPrec@1 83.594 (87.181)\n",
      "[80/391]\tTime 0.180 (0.177)\tData 0.020 (0.020)\tLoss 0.5581 (0.6126)\tPrec@1 89.062 (86.950)\n",
      "[90/391]\tTime 0.178 (0.177)\tData 0.019 (0.020)\tLoss 0.6148 (0.6177)\tPrec@1 85.156 (86.813)\n",
      "[100/391]\tTime 0.179 (0.178)\tData 0.020 (0.020)\tLoss 0.5252 (0.6178)\tPrec@1 91.406 (86.812)\n",
      "[110/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5124 (0.6157)\tPrec@1 92.188 (86.860)\n",
      "[120/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.5169 (0.6161)\tPrec@1 92.969 (86.867)\n",
      "[130/391]\tTime 0.181 (0.178)\tData 0.020 (0.020)\tLoss 0.5784 (0.6171)\tPrec@1 87.500 (86.796)\n",
      "[140/391]\tTime 0.190 (0.179)\tData 0.024 (0.020)\tLoss 0.5695 (0.6170)\tPrec@1 92.188 (86.802)\n",
      "[150/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6854 (0.6155)\tPrec@1 82.812 (86.853)\n",
      "[160/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7265 (0.6162)\tPrec@1 78.125 (86.748)\n",
      "[170/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.5393 (0.6170)\tPrec@1 86.719 (86.737)\n",
      "[180/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.6621 (0.6171)\tPrec@1 85.156 (86.766)\n",
      "[190/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.8202 (0.6188)\tPrec@1 77.344 (86.678)\n",
      "[200/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6313 (0.6193)\tPrec@1 83.594 (86.680)\n",
      "[210/391]\tTime 0.187 (0.180)\tData 0.019 (0.020)\tLoss 0.5546 (0.6200)\tPrec@1 90.625 (86.671)\n",
      "[220/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5882 (0.6191)\tPrec@1 89.062 (86.708)\n",
      "[230/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6247 (0.6210)\tPrec@1 87.500 (86.651)\n",
      "[240/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.7165 (0.6211)\tPrec@1 81.250 (86.638)\n",
      "[250/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.6423 (0.6221)\tPrec@1 89.062 (86.604)\n",
      "[260/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6215 (0.6207)\tPrec@1 82.812 (86.680)\n",
      "[270/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5409 (0.6198)\tPrec@1 90.625 (86.716)\n",
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6209 (0.6199)\tPrec@1 85.156 (86.710)\n",
      "[290/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.6394 (0.6203)\tPrec@1 85.156 (86.668)\n",
      "[300/391]\tTime 0.189 (0.180)\tData 0.020 (0.020)\tLoss 0.6422 (0.6197)\tPrec@1 85.938 (86.688)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[310/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.7080 (0.6204)\tPrec@1 83.594 (86.653)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5348 (0.6195)\tPrec@1 91.406 (86.677)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6399 (0.6197)\tPrec@1 85.156 (86.655)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5487 (0.6205)\tPrec@1 86.719 (86.620)\n",
      "[350/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5454 (0.6201)\tPrec@1 87.500 (86.636)\n",
      "[360/391]\tTime 0.186 (0.181)\tData 0.022 (0.020)\tLoss 0.6310 (0.6198)\tPrec@1 84.375 (86.639)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6363 (0.6208)\tPrec@1 85.156 (86.592)\n",
      "[380/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6364 (0.6206)\tPrec@1 86.719 (86.606)\n",
      "[390/391]\tTime 0.131 (0.181)\tData 0.012 (0.020)\tLoss 0.6942 (0.6195)\tPrec@1 82.500 (86.656)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.3315 (1.3315)\tPrec@1 76.562 (76.562)\n",
      "Test: [10/79]\tTime 0.020 (0.024)\tLoss 0.9251 (1.1969)\tPrec@1 84.375 (79.190)\n",
      "Test: [20/79]\tTime 0.028 (0.024)\tLoss 1.2821 (1.2321)\tPrec@1 75.781 (78.571)\n",
      "Test: [30/79]\tTime 0.024 (0.024)\tLoss 0.8453 (1.2151)\tPrec@1 84.375 (78.579)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 1.1791 (1.1888)\tPrec@1 76.562 (78.773)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 1.1769 (1.1677)\tPrec@1 78.125 (78.998)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.9322 (1.1504)\tPrec@1 85.938 (79.252)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 1.3396 (1.1587)\tPrec@1 77.344 (79.005)\n",
      " * Prec@1 79.040\n",
      "\n",
      "===> epoch: 62/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.141 (0.141)\tData 0.020 (0.020)\tLoss 0.6173 (0.6173)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.187 (0.177)\tData 0.020 (0.019)\tLoss 0.5455 (0.5933)\tPrec@1 88.281 (87.571)\n",
      "[20/391]\tTime 0.186 (0.180)\tData 0.020 (0.019)\tLoss 0.5570 (0.6433)\tPrec@1 89.062 (85.863)\n",
      "[30/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5684 (0.6359)\tPrec@1 89.062 (85.837)\n",
      "[40/391]\tTime 0.146 (0.173)\tData 0.019 (0.020)\tLoss 0.6320 (0.6257)\tPrec@1 86.719 (86.204)\n",
      "[50/391]\tTime 0.180 (0.172)\tData 0.019 (0.020)\tLoss 0.6357 (0.6255)\tPrec@1 83.594 (86.137)\n",
      "[60/391]\tTime 0.183 (0.173)\tData 0.020 (0.020)\tLoss 0.5083 (0.6215)\tPrec@1 90.625 (86.347)\n",
      "[70/391]\tTime 0.181 (0.174)\tData 0.019 (0.020)\tLoss 0.5433 (0.6143)\tPrec@1 92.188 (86.741)\n",
      "[80/391]\tTime 0.181 (0.175)\tData 0.020 (0.020)\tLoss 0.5004 (0.6136)\tPrec@1 88.281 (86.815)\n",
      "[90/391]\tTime 0.182 (0.176)\tData 0.020 (0.020)\tLoss 0.5381 (0.6137)\tPrec@1 93.750 (86.865)\n",
      "[100/391]\tTime 0.180 (0.177)\tData 0.022 (0.020)\tLoss 0.6733 (0.6125)\tPrec@1 84.375 (86.843)\n",
      "[110/391]\tTime 0.180 (0.177)\tData 0.020 (0.020)\tLoss 0.6340 (0.6133)\tPrec@1 85.156 (86.789)\n",
      "[120/391]\tTime 0.179 (0.177)\tData 0.019 (0.020)\tLoss 0.6129 (0.6122)\tPrec@1 86.719 (86.809)\n",
      "[130/391]\tTime 0.186 (0.178)\tData 0.019 (0.020)\tLoss 0.5150 (0.6098)\tPrec@1 90.625 (86.874)\n",
      "[140/391]\tTime 0.186 (0.178)\tData 0.019 (0.020)\tLoss 0.5677 (0.6083)\tPrec@1 90.625 (86.963)\n",
      "[150/391]\tTime 0.187 (0.178)\tData 0.019 (0.020)\tLoss 0.5295 (0.6077)\tPrec@1 89.062 (86.967)\n",
      "[160/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.6831 (0.6092)\tPrec@1 85.156 (86.893)\n",
      "[170/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.6080 (0.6096)\tPrec@1 88.281 (86.933)\n",
      "[180/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5775 (0.6106)\tPrec@1 84.375 (86.870)\n",
      "[190/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.5627 (0.6120)\tPrec@1 89.062 (86.821)\n",
      "[200/391]\tTime 0.179 (0.179)\tData 0.030 (0.020)\tLoss 0.7046 (0.6129)\tPrec@1 85.156 (86.765)\n",
      "[210/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6154 (0.6132)\tPrec@1 89.062 (86.763)\n",
      "[220/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6040 (0.6127)\tPrec@1 85.156 (86.779)\n",
      "[230/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6568 (0.6131)\tPrec@1 84.375 (86.759)\n",
      "[240/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6249 (0.6115)\tPrec@1 87.500 (86.819)\n",
      "[250/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5992 (0.6122)\tPrec@1 89.062 (86.803)\n",
      "[260/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5591 (0.6120)\tPrec@1 91.406 (86.821)\n",
      "[270/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6567 (0.6127)\tPrec@1 87.500 (86.791)\n",
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6600 (0.6128)\tPrec@1 85.938 (86.813)\n",
      "[290/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.7609 (0.6148)\tPrec@1 84.375 (86.748)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6562 (0.6149)\tPrec@1 84.375 (86.760)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6459 (0.6157)\tPrec@1 87.500 (86.739)\n",
      "[320/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.6457 (0.6162)\tPrec@1 88.281 (86.728)\n",
      "[330/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.6186 (0.6174)\tPrec@1 83.594 (86.693)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6539 (0.6168)\tPrec@1 85.156 (86.700)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6127 (0.6179)\tPrec@1 85.938 (86.654)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5635 (0.6183)\tPrec@1 87.500 (86.615)\n",
      "[370/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6236 (0.6182)\tPrec@1 83.594 (86.616)\n",
      "[380/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.4197 (0.6179)\tPrec@1 95.312 (86.633)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.012 (0.020)\tLoss 0.4628 (0.6176)\tPrec@1 92.500 (86.658)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.5883 (0.5883)\tPrec@1 89.844 (89.844)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.9510 (0.9984)\tPrec@1 82.031 (82.457)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.8368 (1.0080)\tPrec@1 85.156 (81.585)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.6305 (1.0281)\tPrec@1 83.594 (80.771)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 0.7146 (1.0167)\tPrec@1 78.125 (80.831)\n",
      "Test: [50/79]\tTime 0.023 (0.023)\tLoss 1.1478 (0.9964)\tPrec@1 78.125 (81.097)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.7360 (0.9923)\tPrec@1 86.719 (81.212)\n",
      "Test: [70/79]\tTime 0.024 (0.023)\tLoss 1.1368 (1.0082)\tPrec@1 77.344 (80.854)\n",
      " * Prec@1 80.930\n",
      "\n",
      "===> epoch: 63/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.150 (0.150)\tData 0.020 (0.020)\tLoss 0.6016 (0.6016)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.179 (0.176)\tData 0.019 (0.020)\tLoss 0.5260 (0.5688)\tPrec@1 92.188 (88.494)\n",
      "[20/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.6545 (0.5952)\tPrec@1 82.812 (86.756)\n",
      "[30/391]\tTime 0.146 (0.170)\tData 0.019 (0.020)\tLoss 0.5764 (0.5994)\tPrec@1 88.281 (86.920)\n",
      "[40/391]\tTime 0.186 (0.167)\tData 0.019 (0.020)\tLoss 0.7033 (0.6171)\tPrec@1 84.375 (86.528)\n",
      "[50/391]\tTime 0.185 (0.170)\tData 0.019 (0.020)\tLoss 0.6746 (0.6219)\tPrec@1 81.250 (86.152)\n",
      "[60/391]\tTime 0.187 (0.172)\tData 0.020 (0.020)\tLoss 0.6849 (0.6230)\tPrec@1 81.250 (86.309)\n",
      "[70/391]\tTime 0.181 (0.174)\tData 0.020 (0.020)\tLoss 0.6906 (0.6229)\tPrec@1 82.031 (86.290)\n",
      "[80/391]\tTime 0.181 (0.175)\tData 0.020 (0.020)\tLoss 0.5786 (0.6176)\tPrec@1 88.281 (86.545)\n",
      "[90/391]\tTime 0.179 (0.176)\tData 0.019 (0.020)\tLoss 0.6202 (0.6187)\tPrec@1 85.938 (86.521)\n",
      "[100/391]\tTime 0.179 (0.176)\tData 0.019 (0.020)\tLoss 0.7258 (0.6164)\tPrec@1 85.156 (86.665)\n",
      "[110/391]\tTime 0.185 (0.177)\tData 0.022 (0.020)\tLoss 0.5891 (0.6172)\tPrec@1 88.281 (86.669)\n",
      "[120/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.6548 (0.6166)\tPrec@1 85.156 (86.732)\n",
      "[130/391]\tTime 0.185 (0.178)\tData 0.019 (0.020)\tLoss 0.5582 (0.6158)\tPrec@1 89.062 (86.743)\n",
      "[140/391]\tTime 0.182 (0.178)\tData 0.019 (0.020)\tLoss 0.6603 (0.6155)\tPrec@1 84.375 (86.669)\n",
      "[150/391]\tTime 0.186 (0.178)\tData 0.019 (0.020)\tLoss 0.6416 (0.6148)\tPrec@1 84.375 (86.646)\n",
      "[160/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6683 (0.6159)\tPrec@1 84.375 (86.641)\n",
      "[170/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.7049 (0.6172)\tPrec@1 83.594 (86.545)\n",
      "[180/391]\tTime 0.178 (0.179)\tData 0.020 (0.020)\tLoss 0.6171 (0.6182)\tPrec@1 88.281 (86.563)\n",
      "[190/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.6825 (0.6175)\tPrec@1 82.812 (86.641)\n",
      "[200/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.5660 (0.6176)\tPrec@1 87.500 (86.633)\n",
      "[210/391]\tTime 0.188 (0.179)\tData 0.019 (0.020)\tLoss 0.6013 (0.6161)\tPrec@1 88.281 (86.726)\n",
      "[220/391]\tTime 0.180 (0.179)\tData 0.021 (0.020)\tLoss 0.5959 (0.6177)\tPrec@1 87.500 (86.662)\n",
      "[230/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5873 (0.6180)\tPrec@1 88.281 (86.665)\n",
      "[240/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.4566 (0.6166)\tPrec@1 93.750 (86.686)\n",
      "[250/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6992 (0.6164)\tPrec@1 83.594 (86.685)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[260/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.7316 (0.6165)\tPrec@1 80.469 (86.686)\n",
      "[270/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.5571 (0.6168)\tPrec@1 89.062 (86.667)\n",
      "[280/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5808 (0.6176)\tPrec@1 87.500 (86.638)\n",
      "[290/391]\tTime 0.185 (0.180)\tData 0.023 (0.020)\tLoss 0.6227 (0.6179)\tPrec@1 86.719 (86.649)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6873 (0.6182)\tPrec@1 81.250 (86.643)\n",
      "[310/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.5718 (0.6166)\tPrec@1 87.500 (86.689)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6580 (0.6178)\tPrec@1 85.938 (86.621)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6556 (0.6188)\tPrec@1 86.719 (86.584)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5621 (0.6193)\tPrec@1 89.062 (86.590)\n",
      "[350/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.6771 (0.6201)\tPrec@1 83.594 (86.543)\n",
      "[360/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6014 (0.6203)\tPrec@1 89.062 (86.539)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6596 (0.6209)\tPrec@1 82.812 (86.540)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.5522 (0.6212)\tPrec@1 90.625 (86.538)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.9362 (0.6216)\tPrec@1 75.000 (86.526)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.5883 (1.5883)\tPrec@1 78.906 (78.906)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 1.5563 (1.7235)\tPrec@1 71.094 (72.656)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.6249 (1.6956)\tPrec@1 74.219 (72.656)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 1.5141 (1.6907)\tPrec@1 75.781 (73.034)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 2.0217 (1.6912)\tPrec@1 63.281 (72.847)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 1.9878 (1.6995)\tPrec@1 64.062 (72.748)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 1.3621 (1.6605)\tPrec@1 77.344 (73.181)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.3954 (1.6630)\tPrec@1 76.562 (73.030)\n",
      " * Prec@1 73.130\n",
      "\n",
      "===> epoch: 64/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.169 (0.169)\tData 0.021 (0.021)\tLoss 0.6598 (0.6598)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5381 (0.5929)\tPrec@1 91.406 (87.784)\n",
      "[20/391]\tTime 0.145 (0.171)\tData 0.020 (0.020)\tLoss 0.5611 (0.5870)\tPrec@1 89.844 (87.798)\n",
      "[30/391]\tTime 0.184 (0.165)\tData 0.019 (0.020)\tLoss 0.7566 (0.5935)\tPrec@1 82.031 (87.676)\n",
      "[40/391]\tTime 0.186 (0.169)\tData 0.020 (0.020)\tLoss 0.7346 (0.6057)\tPrec@1 82.812 (87.005)\n",
      "[50/391]\tTime 0.180 (0.172)\tData 0.019 (0.020)\tLoss 0.6527 (0.6082)\tPrec@1 84.375 (86.933)\n",
      "[60/391]\tTime 0.181 (0.173)\tData 0.020 (0.020)\tLoss 0.6011 (0.6161)\tPrec@1 83.594 (86.808)\n",
      "[70/391]\tTime 0.181 (0.174)\tData 0.019 (0.020)\tLoss 0.6345 (0.6135)\tPrec@1 86.719 (86.884)\n",
      "[80/391]\tTime 0.188 (0.175)\tData 0.020 (0.020)\tLoss 0.5273 (0.6119)\tPrec@1 91.406 (87.066)\n",
      "[90/391]\tTime 0.187 (0.176)\tData 0.020 (0.020)\tLoss 0.4866 (0.6097)\tPrec@1 89.844 (87.131)\n",
      "[100/391]\tTime 0.182 (0.176)\tData 0.020 (0.020)\tLoss 0.4658 (0.6063)\tPrec@1 94.531 (87.245)\n",
      "[110/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.5692 (0.6024)\tPrec@1 87.500 (87.380)\n",
      "[120/391]\tTime 0.182 (0.177)\tData 0.019 (0.020)\tLoss 0.6673 (0.6041)\tPrec@1 87.500 (87.293)\n",
      "[130/391]\tTime 0.181 (0.177)\tData 0.020 (0.020)\tLoss 0.6258 (0.6052)\tPrec@1 89.062 (87.261)\n",
      "[140/391]\tTime 0.182 (0.178)\tData 0.020 (0.020)\tLoss 0.6480 (0.6061)\tPrec@1 85.156 (87.206)\n",
      "[150/391]\tTime 0.185 (0.178)\tData 0.020 (0.020)\tLoss 0.6797 (0.6083)\tPrec@1 85.938 (87.117)\n",
      "[160/391]\tTime 0.185 (0.178)\tData 0.020 (0.020)\tLoss 0.5988 (0.6107)\tPrec@1 88.281 (87.039)\n",
      "[170/391]\tTime 0.185 (0.178)\tData 0.020 (0.020)\tLoss 0.6015 (0.6127)\tPrec@1 89.062 (86.979)\n",
      "[180/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6655 (0.6134)\tPrec@1 84.375 (86.926)\n",
      "[190/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6046 (0.6143)\tPrec@1 86.719 (86.899)\n",
      "[200/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6802 (0.6140)\tPrec@1 85.938 (86.886)\n",
      "[210/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.5486 (0.6142)\tPrec@1 92.188 (86.908)\n",
      "[220/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.5733 (0.6132)\tPrec@1 85.156 (86.917)\n",
      "[230/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.6191 (0.6119)\tPrec@1 86.719 (86.955)\n",
      "[240/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.6119 (0.6117)\tPrec@1 89.844 (86.994)\n",
      "[250/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5160 (0.6120)\tPrec@1 91.406 (86.968)\n",
      "[260/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.7120 (0.6122)\tPrec@1 82.812 (86.961)\n",
      "[270/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6310 (0.6125)\tPrec@1 84.375 (86.964)\n",
      "[280/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5733 (0.6132)\tPrec@1 84.375 (86.927)\n",
      "[290/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6021 (0.6133)\tPrec@1 86.719 (86.893)\n",
      "[300/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.7100 (0.6131)\tPrec@1 82.031 (86.838)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.4644 (0.6133)\tPrec@1 94.531 (86.847)\n",
      "[320/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5190 (0.6135)\tPrec@1 91.406 (86.804)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.6341 (0.6125)\tPrec@1 87.500 (86.841)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5680 (0.6124)\tPrec@1 88.281 (86.815)\n",
      "[350/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.4675 (0.6125)\tPrec@1 96.094 (86.799)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6532 (0.6137)\tPrec@1 82.031 (86.734)\n",
      "[370/391]\tTime 0.188 (0.180)\tData 0.020 (0.020)\tLoss 0.7856 (0.6137)\tPrec@1 77.344 (86.731)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6257 (0.6137)\tPrec@1 89.062 (86.719)\n",
      "[390/391]\tTime 0.129 (0.180)\tData 0.012 (0.020)\tLoss 0.9130 (0.6140)\tPrec@1 68.750 (86.694)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.2013 (1.2013)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.9098 (1.4595)\tPrec@1 79.688 (75.213)\n",
      "Test: [20/79]\tTime 0.026 (0.022)\tLoss 1.3352 (1.4879)\tPrec@1 77.344 (74.665)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.9344 (1.4681)\tPrec@1 83.594 (74.521)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 1.3791 (1.4410)\tPrec@1 75.000 (74.848)\n",
      "Test: [50/79]\tTime 0.025 (0.023)\tLoss 1.2983 (1.4292)\tPrec@1 77.344 (74.877)\n",
      "Test: [60/79]\tTime 0.027 (0.023)\tLoss 1.0979 (1.4032)\tPrec@1 78.125 (75.141)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 1.5418 (1.4184)\tPrec@1 75.000 (75.022)\n",
      " * Prec@1 75.040\n",
      "\n",
      "===> epoch: 65/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.165 (0.165)\tData 0.021 (0.021)\tLoss 0.5220 (0.5220)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.147 (0.166)\tData 0.022 (0.020)\tLoss 0.6015 (0.5801)\tPrec@1 86.719 (88.068)\n",
      "[20/391]\tTime 0.146 (0.156)\tData 0.020 (0.020)\tLoss 0.6595 (0.5914)\tPrec@1 85.156 (87.723)\n",
      "[30/391]\tTime 0.182 (0.164)\tData 0.019 (0.020)\tLoss 0.5302 (0.5982)\tPrec@1 89.844 (87.374)\n",
      "[40/391]\tTime 0.181 (0.168)\tData 0.020 (0.020)\tLoss 0.6182 (0.5987)\tPrec@1 85.156 (87.233)\n",
      "[50/391]\tTime 0.179 (0.170)\tData 0.019 (0.020)\tLoss 0.6130 (0.6004)\tPrec@1 85.156 (87.240)\n",
      "[60/391]\tTime 0.180 (0.172)\tData 0.020 (0.020)\tLoss 0.6284 (0.5990)\tPrec@1 86.719 (87.141)\n",
      "[70/391]\tTime 0.178 (0.173)\tData 0.019 (0.020)\tLoss 0.7067 (0.6030)\tPrec@1 82.031 (87.148)\n",
      "[80/391]\tTime 0.182 (0.174)\tData 0.019 (0.020)\tLoss 0.5844 (0.6021)\tPrec@1 88.281 (87.365)\n",
      "[90/391]\tTime 0.182 (0.175)\tData 0.020 (0.020)\tLoss 0.5762 (0.6004)\tPrec@1 90.625 (87.397)\n",
      "[100/391]\tTime 0.182 (0.176)\tData 0.019 (0.020)\tLoss 0.5412 (0.6041)\tPrec@1 87.500 (87.229)\n",
      "[110/391]\tTime 0.180 (0.176)\tData 0.020 (0.020)\tLoss 0.5741 (0.6043)\tPrec@1 86.719 (87.211)\n",
      "[120/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.4923 (0.6072)\tPrec@1 92.188 (87.132)\n",
      "[130/391]\tTime 0.185 (0.177)\tData 0.019 (0.020)\tLoss 0.6886 (0.6132)\tPrec@1 83.594 (86.874)\n",
      "[140/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5348 (0.6150)\tPrec@1 89.062 (86.780)\n",
      "[150/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5856 (0.6174)\tPrec@1 87.500 (86.683)\n",
      "[160/391]\tTime 0.187 (0.178)\tData 0.019 (0.020)\tLoss 0.5094 (0.6196)\tPrec@1 88.281 (86.539)\n",
      "[170/391]\tTime 0.189 (0.178)\tData 0.020 (0.020)\tLoss 0.5646 (0.6194)\tPrec@1 88.281 (86.550)\n",
      "[180/391]\tTime 0.185 (0.179)\tData 0.019 (0.020)\tLoss 0.5591 (0.6201)\tPrec@1 91.406 (86.563)\n",
      "[190/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.6207 (0.6192)\tPrec@1 84.375 (86.555)\n",
      "[200/391]\tTime 0.183 (0.179)\tData 0.019 (0.020)\tLoss 0.5656 (0.6188)\tPrec@1 87.500 (86.517)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[210/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.5677 (0.6193)\tPrec@1 91.406 (86.552)\n",
      "[220/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7664 (0.6220)\tPrec@1 79.688 (86.468)\n",
      "[230/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5941 (0.6232)\tPrec@1 86.719 (86.418)\n",
      "[240/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.5885 (0.6226)\tPrec@1 89.062 (86.443)\n",
      "[250/391]\tTime 0.181 (0.179)\tData 0.022 (0.020)\tLoss 0.5846 (0.6224)\tPrec@1 88.281 (86.495)\n",
      "[260/391]\tTime 0.183 (0.179)\tData 0.020 (0.020)\tLoss 0.5471 (0.6226)\tPrec@1 89.844 (86.488)\n",
      "[270/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.6012 (0.6224)\tPrec@1 88.281 (86.497)\n",
      "[280/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.7536 (0.6224)\tPrec@1 83.594 (86.516)\n",
      "[290/391]\tTime 0.180 (0.179)\tData 0.021 (0.020)\tLoss 0.7381 (0.6233)\tPrec@1 80.469 (86.450)\n",
      "[300/391]\tTime 0.186 (0.180)\tData 0.019 (0.020)\tLoss 0.6826 (0.6240)\tPrec@1 82.812 (86.415)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6192 (0.6240)\tPrec@1 86.719 (86.407)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6712 (0.6243)\tPrec@1 87.500 (86.388)\n",
      "[330/391]\tTime 0.185 (0.180)\tData 0.021 (0.020)\tLoss 0.6826 (0.6246)\tPrec@1 85.156 (86.398)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5487 (0.6234)\tPrec@1 90.625 (86.442)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6375 (0.6218)\tPrec@1 84.375 (86.503)\n",
      "[360/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6061 (0.6209)\tPrec@1 88.281 (86.561)\n",
      "[370/391]\tTime 0.190 (0.180)\tData 0.019 (0.020)\tLoss 0.5126 (0.6207)\tPrec@1 90.625 (86.561)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5998 (0.6206)\tPrec@1 86.719 (86.542)\n",
      "[390/391]\tTime 0.134 (0.180)\tData 0.012 (0.020)\tLoss 0.5389 (0.6206)\tPrec@1 92.500 (86.548)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7226 (0.7226)\tPrec@1 85.156 (85.156)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.9267 (0.9077)\tPrec@1 85.156 (84.020)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.6120 (0.9002)\tPrec@1 86.719 (83.668)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.8307 (0.9149)\tPrec@1 85.938 (83.291)\n",
      "Test: [40/79]\tTime 0.022 (0.023)\tLoss 0.8291 (0.8967)\tPrec@1 83.594 (83.632)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 0.7339 (0.8812)\tPrec@1 85.938 (83.992)\n",
      "Test: [60/79]\tTime 0.034 (0.023)\tLoss 0.8236 (0.8788)\tPrec@1 85.938 (84.016)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.9445 (0.8854)\tPrec@1 86.719 (83.913)\n",
      " * Prec@1 83.840\n",
      "\n",
      "===> epoch: 66/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.144 (0.144)\tData 0.021 (0.021)\tLoss 0.6350 (0.6350)\tPrec@1 84.375 (84.375)\n",
      "[10/391]\tTime 0.144 (0.145)\tData 0.020 (0.020)\tLoss 0.6290 (0.6075)\tPrec@1 86.719 (86.506)\n",
      "[20/391]\tTime 0.180 (0.161)\tData 0.020 (0.020)\tLoss 0.5632 (0.6061)\tPrec@1 86.719 (86.682)\n",
      "[30/391]\tTime 0.184 (0.167)\tData 0.019 (0.020)\tLoss 0.5645 (0.5919)\tPrec@1 89.062 (87.601)\n",
      "[40/391]\tTime 0.178 (0.170)\tData 0.019 (0.020)\tLoss 0.6009 (0.6043)\tPrec@1 85.938 (87.043)\n",
      "[50/391]\tTime 0.178 (0.173)\tData 0.019 (0.020)\tLoss 0.6459 (0.6139)\tPrec@1 87.500 (86.933)\n",
      "[60/391]\tTime 0.179 (0.174)\tData 0.019 (0.020)\tLoss 0.6039 (0.6156)\tPrec@1 87.500 (87.013)\n",
      "[70/391]\tTime 0.181 (0.175)\tData 0.020 (0.020)\tLoss 0.6374 (0.6150)\tPrec@1 86.719 (87.104)\n",
      "[80/391]\tTime 0.185 (0.176)\tData 0.020 (0.020)\tLoss 0.5397 (0.6122)\tPrec@1 89.844 (86.998)\n",
      "[90/391]\tTime 0.180 (0.177)\tData 0.019 (0.020)\tLoss 0.5292 (0.6126)\tPrec@1 89.844 (86.916)\n",
      "[100/391]\tTime 0.182 (0.177)\tData 0.019 (0.020)\tLoss 0.7964 (0.6146)\tPrec@1 78.906 (86.788)\n",
      "[110/391]\tTime 0.180 (0.178)\tData 0.019 (0.020)\tLoss 0.5656 (0.6122)\tPrec@1 90.625 (86.986)\n",
      "[120/391]\tTime 0.181 (0.178)\tData 0.020 (0.020)\tLoss 0.6912 (0.6109)\tPrec@1 83.594 (86.983)\n",
      "[130/391]\tTime 0.182 (0.178)\tData 0.021 (0.020)\tLoss 0.6140 (0.6109)\tPrec@1 86.719 (86.939)\n",
      "[140/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.5710 (0.6117)\tPrec@1 85.156 (86.857)\n",
      "[150/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.6823 (0.6130)\tPrec@1 82.812 (86.781)\n",
      "[160/391]\tTime 0.182 (0.179)\tData 0.020 (0.020)\tLoss 0.6346 (0.6106)\tPrec@1 84.375 (86.913)\n",
      "[170/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.5503 (0.6094)\tPrec@1 88.281 (86.997)\n",
      "[180/391]\tTime 0.182 (0.179)\tData 0.022 (0.020)\tLoss 0.6668 (0.6099)\tPrec@1 82.031 (86.948)\n",
      "[190/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.7688 (0.6099)\tPrec@1 78.125 (86.948)\n",
      "[200/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.7039 (0.6105)\tPrec@1 82.812 (86.925)\n",
      "[210/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.5690 (0.6102)\tPrec@1 89.062 (86.974)\n",
      "[220/391]\tTime 0.184 (0.179)\tData 0.019 (0.020)\tLoss 0.5885 (0.6096)\tPrec@1 85.938 (87.002)\n",
      "[230/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6815 (0.6100)\tPrec@1 83.594 (87.003)\n",
      "[240/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6172 (0.6095)\tPrec@1 87.500 (87.001)\n",
      "[250/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5818 (0.6099)\tPrec@1 90.625 (87.024)\n",
      "[260/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6697 (0.6112)\tPrec@1 88.281 (87.012)\n",
      "[270/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5517 (0.6112)\tPrec@1 89.844 (87.033)\n",
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5828 (0.6127)\tPrec@1 89.062 (86.966)\n",
      "[290/391]\tTime 0.184 (0.180)\tData 0.022 (0.020)\tLoss 0.6418 (0.6135)\tPrec@1 88.281 (86.950)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6639 (0.6133)\tPrec@1 84.375 (86.963)\n",
      "[310/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.5844 (0.6130)\tPrec@1 89.062 (86.972)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6431 (0.6142)\tPrec@1 89.844 (86.940)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5921 (0.6142)\tPrec@1 85.156 (86.936)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.5753 (0.6139)\tPrec@1 87.500 (86.932)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6806 (0.6143)\tPrec@1 83.594 (86.932)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6173 (0.6127)\tPrec@1 87.500 (86.974)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5583 (0.6129)\tPrec@1 91.406 (86.952)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5910 (0.6133)\tPrec@1 85.938 (86.938)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.013 (0.020)\tLoss 0.6340 (0.6134)\tPrec@1 83.750 (86.928)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.2352 (1.2352)\tPrec@1 77.344 (77.344)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.7502 (1.1140)\tPrec@1 86.719 (80.114)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 1.1923 (1.2022)\tPrec@1 77.344 (78.869)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.9513 (1.1952)\tPrec@1 80.469 (78.604)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 1.0262 (1.1754)\tPrec@1 79.688 (78.887)\n",
      "Test: [50/79]\tTime 0.019 (0.022)\tLoss 1.7047 (1.1807)\tPrec@1 73.438 (78.968)\n",
      "Test: [60/79]\tTime 0.019 (0.022)\tLoss 1.1924 (1.1839)\tPrec@1 82.812 (78.945)\n",
      "Test: [70/79]\tTime 0.019 (0.021)\tLoss 1.3569 (1.1898)\tPrec@1 75.000 (78.895)\n",
      " * Prec@1 78.960\n",
      "\n",
      "===> epoch: 67/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.144 (0.144)\tData 0.021 (0.021)\tLoss 0.6261 (0.6261)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.185 (0.160)\tData 0.020 (0.020)\tLoss 0.6130 (0.6260)\tPrec@1 85.938 (85.795)\n",
      "[20/391]\tTime 0.181 (0.170)\tData 0.019 (0.020)\tLoss 0.6560 (0.6293)\tPrec@1 87.500 (85.975)\n",
      "[30/391]\tTime 0.182 (0.174)\tData 0.020 (0.020)\tLoss 0.5163 (0.6163)\tPrec@1 91.406 (86.643)\n",
      "[40/391]\tTime 0.181 (0.175)\tData 0.019 (0.020)\tLoss 0.6599 (0.6069)\tPrec@1 85.156 (87.005)\n",
      "[50/391]\tTime 0.178 (0.177)\tData 0.019 (0.020)\tLoss 0.5598 (0.6004)\tPrec@1 87.500 (87.209)\n",
      "[60/391]\tTime 0.181 (0.177)\tData 0.019 (0.020)\tLoss 0.6022 (0.5963)\tPrec@1 85.156 (87.308)\n",
      "[70/391]\tTime 0.183 (0.178)\tData 0.019 (0.020)\tLoss 0.6209 (0.5994)\tPrec@1 88.281 (87.291)\n",
      "[80/391]\tTime 0.182 (0.178)\tData 0.020 (0.020)\tLoss 0.5959 (0.5990)\tPrec@1 89.062 (87.365)\n",
      "[90/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6484 (0.6024)\tPrec@1 86.719 (87.182)\n",
      "[100/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6534 (0.6054)\tPrec@1 82.812 (86.997)\n",
      "[110/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6096 (0.6042)\tPrec@1 85.156 (87.064)\n",
      "[120/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.5998 (0.6030)\tPrec@1 85.156 (87.138)\n",
      "[130/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6328 (0.6021)\tPrec@1 89.062 (87.172)\n",
      "[140/391]\tTime 0.186 (0.180)\tData 0.021 (0.020)\tLoss 0.5278 (0.6026)\tPrec@1 89.844 (87.112)\n",
      "[150/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6202 (0.6046)\tPrec@1 85.938 (87.024)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6346 (0.6029)\tPrec@1 85.156 (87.112)\n",
      "[170/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5673 (0.6036)\tPrec@1 88.281 (87.107)\n",
      "[180/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5296 (0.6042)\tPrec@1 89.062 (87.077)\n",
      "[190/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6139 (0.6045)\tPrec@1 85.938 (87.046)\n",
      "[200/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6005 (0.6050)\tPrec@1 87.500 (87.053)\n",
      "[210/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5090 (0.6033)\tPrec@1 91.406 (87.115)\n",
      "[220/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5788 (0.6060)\tPrec@1 91.406 (87.044)\n",
      "[230/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.5580 (0.6058)\tPrec@1 89.062 (87.060)\n",
      "[240/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.6405 (0.6070)\tPrec@1 86.719 (86.988)\n",
      "[250/391]\tTime 0.187 (0.180)\tData 0.020 (0.020)\tLoss 0.6212 (0.6095)\tPrec@1 89.844 (86.899)\n",
      "[260/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.5838 (0.6103)\tPrec@1 87.500 (86.877)\n",
      "[270/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6492 (0.6112)\tPrec@1 82.812 (86.825)\n",
      "[280/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6024 (0.6123)\tPrec@1 86.719 (86.766)\n",
      "[290/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6564 (0.6139)\tPrec@1 85.156 (86.724)\n",
      "[300/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5864 (0.6151)\tPrec@1 89.844 (86.701)\n",
      "[310/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5911 (0.6156)\tPrec@1 86.719 (86.681)\n",
      "[320/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6161 (0.6158)\tPrec@1 89.844 (86.704)\n",
      "[330/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6217 (0.6159)\tPrec@1 85.156 (86.723)\n",
      "[340/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.5709 (0.6164)\tPrec@1 90.625 (86.721)\n",
      "[350/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.7566 (0.6178)\tPrec@1 85.156 (86.694)\n",
      "[360/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6166 (0.6187)\tPrec@1 88.281 (86.675)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5925 (0.6188)\tPrec@1 89.062 (86.662)\n",
      "[380/391]\tTime 0.192 (0.181)\tData 0.020 (0.020)\tLoss 0.5580 (0.6186)\tPrec@1 90.625 (86.690)\n",
      "[390/391]\tTime 0.130 (0.181)\tData 0.013 (0.020)\tLoss 0.7396 (0.6200)\tPrec@1 80.000 (86.640)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7183 (0.7183)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.8627 (0.9709)\tPrec@1 83.594 (81.392)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.8107 (1.0237)\tPrec@1 81.250 (80.841)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 0.8098 (1.0265)\tPrec@1 80.469 (80.822)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.9960 (1.0314)\tPrec@1 78.906 (80.583)\n",
      "Test: [50/79]\tTime 0.020 (0.021)\tLoss 0.8928 (1.0329)\tPrec@1 81.250 (80.469)\n",
      "Test: [60/79]\tTime 0.020 (0.021)\tLoss 0.9198 (1.0178)\tPrec@1 82.031 (80.879)\n",
      "Test: [70/79]\tTime 0.020 (0.021)\tLoss 0.9886 (1.0235)\tPrec@1 81.250 (80.700)\n",
      " * Prec@1 80.730\n",
      "\n",
      "===> epoch: 68/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.144 (0.144)\tData 0.021 (0.021)\tLoss 0.5320 (0.5320)\tPrec@1 92.969 (92.969)\n",
      "[10/391]\tTime 0.180 (0.170)\tData 0.020 (0.020)\tLoss 0.5684 (0.6128)\tPrec@1 89.844 (86.719)\n",
      "[20/391]\tTime 0.178 (0.175)\tData 0.019 (0.020)\tLoss 0.8130 (0.6256)\tPrec@1 82.031 (86.533)\n",
      "[30/391]\tTime 0.178 (0.177)\tData 0.020 (0.020)\tLoss 0.6864 (0.6245)\tPrec@1 81.250 (86.265)\n",
      "[40/391]\tTime 0.179 (0.178)\tData 0.019 (0.020)\tLoss 0.6623 (0.6248)\tPrec@1 85.156 (86.071)\n",
      "[50/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.6073 (0.6149)\tPrec@1 85.156 (86.443)\n",
      "[60/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.5787 (0.6138)\tPrec@1 88.281 (86.616)\n",
      "[70/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6186 (0.6106)\tPrec@1 87.500 (86.730)\n",
      "[80/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.6708 (0.6176)\tPrec@1 85.938 (86.555)\n",
      "[90/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7846 (0.6185)\tPrec@1 79.688 (86.573)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5570 (0.6169)\tPrec@1 88.281 (86.719)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.022 (0.020)\tLoss 0.5424 (0.6151)\tPrec@1 88.281 (86.775)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5588 (0.6157)\tPrec@1 92.188 (86.777)\n",
      "[130/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5785 (0.6170)\tPrec@1 86.719 (86.755)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7017 (0.6172)\tPrec@1 84.375 (86.708)\n",
      "[150/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6520 (0.6173)\tPrec@1 83.594 (86.703)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6639 (0.6174)\tPrec@1 85.938 (86.709)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5414 (0.6161)\tPrec@1 91.406 (86.774)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6961 (0.6160)\tPrec@1 85.156 (86.814)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5796 (0.6150)\tPrec@1 86.719 (86.846)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.5931 (0.6146)\tPrec@1 87.500 (86.843)\n",
      "[210/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6044 (0.6142)\tPrec@1 86.719 (86.834)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6619 (0.6147)\tPrec@1 85.938 (86.818)\n",
      "[230/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7427 (0.6152)\tPrec@1 82.031 (86.749)\n",
      "[240/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5732 (0.6159)\tPrec@1 89.844 (86.722)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6878 (0.6149)\tPrec@1 81.250 (86.747)\n",
      "[260/391]\tTime 0.184 (0.181)\tData 0.021 (0.020)\tLoss 0.6403 (0.6148)\tPrec@1 86.719 (86.755)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5786 (0.6130)\tPrec@1 84.375 (86.799)\n",
      "[280/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5300 (0.6136)\tPrec@1 91.406 (86.733)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6558 (0.6140)\tPrec@1 84.375 (86.738)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6145 (0.6137)\tPrec@1 86.719 (86.776)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6370 (0.6146)\tPrec@1 82.031 (86.761)\n",
      "[320/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6334 (0.6150)\tPrec@1 88.281 (86.750)\n",
      "[330/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6129 (0.6158)\tPrec@1 85.156 (86.698)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.6294 (0.6155)\tPrec@1 87.500 (86.689)\n",
      "[350/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6138 (0.6157)\tPrec@1 84.375 (86.681)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.8060 (0.6166)\tPrec@1 83.594 (86.652)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7551 (0.6165)\tPrec@1 82.812 (86.666)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6240 (0.6169)\tPrec@1 86.719 (86.649)\n",
      "[390/391]\tTime 0.127 (0.181)\tData 0.013 (0.020)\tLoss 0.6114 (0.6164)\tPrec@1 82.500 (86.658)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9652 (0.9652)\tPrec@1 79.688 (79.688)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.9978 (1.0526)\tPrec@1 84.375 (81.605)\n",
      "Test: [20/79]\tTime 0.021 (0.022)\tLoss 0.7586 (1.0688)\tPrec@1 82.031 (80.692)\n",
      "Test: [30/79]\tTime 0.020 (0.021)\tLoss 0.7272 (1.0627)\tPrec@1 83.594 (80.494)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.0482 (1.0421)\tPrec@1 82.812 (80.774)\n",
      "Test: [50/79]\tTime 0.020 (0.021)\tLoss 0.8805 (1.0205)\tPrec@1 82.812 (80.928)\n",
      "Test: [60/79]\tTime 0.019 (0.021)\tLoss 1.3033 (1.0252)\tPrec@1 79.688 (80.815)\n",
      "Test: [70/79]\tTime 0.020 (0.021)\tLoss 1.1579 (1.0183)\tPrec@1 77.344 (80.766)\n",
      " * Prec@1 80.940\n",
      "\n",
      "===> epoch: 69/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.144 (0.144)\tData 0.020 (0.020)\tLoss 0.5860 (0.5860)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.183 (0.176)\tData 0.020 (0.020)\tLoss 0.6124 (0.5927)\tPrec@1 88.281 (88.423)\n",
      "[20/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6671 (0.6166)\tPrec@1 85.938 (86.682)\n",
      "[30/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.6048 (0.6426)\tPrec@1 82.031 (85.307)\n",
      "[40/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.5418 (0.6370)\tPrec@1 89.062 (85.232)\n",
      "[50/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.5979 (0.6331)\tPrec@1 86.719 (85.600)\n",
      "[60/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5834 (0.6329)\tPrec@1 87.500 (85.540)\n",
      "[70/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.7337 (0.6348)\tPrec@1 81.250 (85.387)\n",
      "[80/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6495 (0.6325)\tPrec@1 86.719 (85.465)\n",
      "[90/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6614 (0.6292)\tPrec@1 83.594 (85.646)\n",
      "[100/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.5691 (0.6260)\tPrec@1 86.719 (85.814)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[110/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5677 (0.6270)\tPrec@1 89.844 (85.839)\n",
      "[120/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6486 (0.6285)\tPrec@1 88.281 (85.905)\n",
      "[130/391]\tTime 0.177 (0.180)\tData 0.019 (0.020)\tLoss 0.6284 (0.6274)\tPrec@1 85.156 (85.991)\n",
      "[140/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5511 (0.6257)\tPrec@1 89.844 (86.093)\n",
      "[150/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5072 (0.6248)\tPrec@1 92.188 (86.124)\n",
      "[160/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.5935 (0.6236)\tPrec@1 89.062 (86.229)\n",
      "[170/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6164 (0.6238)\tPrec@1 84.375 (86.257)\n",
      "[180/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6960 (0.6235)\tPrec@1 81.250 (86.257)\n",
      "[190/391]\tTime 0.176 (0.180)\tData 0.019 (0.020)\tLoss 0.5178 (0.6229)\tPrec@1 92.188 (86.355)\n",
      "[200/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6167 (0.6215)\tPrec@1 87.500 (86.423)\n",
      "[210/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.5856 (0.6195)\tPrec@1 91.406 (86.508)\n",
      "[220/391]\tTime 0.187 (0.180)\tData 0.019 (0.020)\tLoss 0.5878 (0.6202)\tPrec@1 88.281 (86.468)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6110 (0.6200)\tPrec@1 85.938 (86.506)\n",
      "[240/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6171 (0.6202)\tPrec@1 86.719 (86.534)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6998 (0.6211)\tPrec@1 84.375 (86.526)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5587 (0.6207)\tPrec@1 85.938 (86.548)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.5668 (0.6184)\tPrec@1 89.844 (86.603)\n",
      "[280/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6768 (0.6191)\tPrec@1 85.156 (86.599)\n",
      "[290/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6610 (0.6194)\tPrec@1 84.375 (86.587)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5759 (0.6186)\tPrec@1 86.719 (86.610)\n",
      "[310/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5865 (0.6182)\tPrec@1 85.156 (86.638)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5051 (0.6180)\tPrec@1 91.406 (86.655)\n",
      "[330/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.6707 (0.6189)\tPrec@1 88.281 (86.662)\n",
      "[340/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6990 (0.6194)\tPrec@1 82.812 (86.639)\n",
      "[350/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6586 (0.6195)\tPrec@1 85.156 (86.648)\n",
      "[360/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.5677 (0.6203)\tPrec@1 89.844 (86.615)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.5925 (0.6194)\tPrec@1 87.500 (86.666)\n",
      "[380/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.5760 (0.6178)\tPrec@1 88.281 (86.735)\n",
      "[390/391]\tTime 0.129 (0.181)\tData 0.012 (0.020)\tLoss 0.8616 (0.6183)\tPrec@1 80.000 (86.726)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.2863 (1.2863)\tPrec@1 75.781 (75.781)\n",
      "Test: [10/79]\tTime 0.020 (0.022)\tLoss 0.7613 (0.9815)\tPrec@1 85.156 (80.540)\n",
      "Test: [20/79]\tTime 0.019 (0.021)\tLoss 0.9491 (1.0658)\tPrec@1 82.812 (79.985)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.9276 (1.0490)\tPrec@1 82.812 (80.267)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.0611 (1.0297)\tPrec@1 75.781 (80.507)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.0750 (1.0213)\tPrec@1 78.906 (80.423)\n",
      "Test: [60/79]\tTime 0.022 (0.020)\tLoss 1.1150 (1.0272)\tPrec@1 83.594 (80.584)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.2192 (1.0209)\tPrec@1 80.469 (80.711)\n",
      " * Prec@1 80.920\n",
      "\n",
      "===> epoch: 70/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.148 (0.148)\tData 0.021 (0.021)\tLoss 0.5351 (0.5351)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.184 (0.177)\tData 0.019 (0.020)\tLoss 0.6224 (0.5951)\tPrec@1 85.938 (87.145)\n",
      "[20/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.6664 (0.5948)\tPrec@1 86.719 (87.426)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6934 (0.6000)\tPrec@1 84.375 (87.248)\n",
      "[40/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.7263 (0.6066)\tPrec@1 82.812 (87.233)\n",
      "[50/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5424 (0.6028)\tPrec@1 89.062 (87.163)\n",
      "[60/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.5304 (0.5953)\tPrec@1 91.406 (87.666)\n",
      "[70/391]\tTime 0.186 (0.180)\tData 0.021 (0.020)\tLoss 0.5777 (0.5916)\tPrec@1 91.406 (87.863)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5868 (0.5940)\tPrec@1 85.156 (87.596)\n",
      "[90/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6487 (0.5961)\tPrec@1 83.594 (87.448)\n",
      "[100/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5755 (0.5972)\tPrec@1 87.500 (87.485)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6338 (0.5999)\tPrec@1 83.594 (87.324)\n",
      "[120/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5771 (0.6012)\tPrec@1 89.844 (87.300)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6881 (0.6041)\tPrec@1 85.156 (87.220)\n",
      "[140/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5720 (0.6038)\tPrec@1 89.062 (87.273)\n",
      "[150/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.6120 (0.6062)\tPrec@1 88.281 (87.159)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6131 (0.6091)\tPrec@1 89.062 (87.078)\n",
      "[170/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6075 (0.6112)\tPrec@1 89.062 (86.984)\n",
      "[180/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5919 (0.6121)\tPrec@1 85.938 (86.939)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6260 (0.6123)\tPrec@1 87.500 (86.915)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5658 (0.6138)\tPrec@1 88.281 (86.824)\n",
      "[210/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.6637 (0.6151)\tPrec@1 82.031 (86.763)\n",
      "[220/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6364 (0.6155)\tPrec@1 85.938 (86.708)\n",
      "[230/391]\tTime 0.189 (0.181)\tData 0.020 (0.020)\tLoss 0.7510 (0.6157)\tPrec@1 81.250 (86.688)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6118 (0.6168)\tPrec@1 85.156 (86.615)\n",
      "[250/391]\tTime 0.190 (0.181)\tData 0.021 (0.020)\tLoss 0.6040 (0.6170)\tPrec@1 85.938 (86.591)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5973 (0.6177)\tPrec@1 86.719 (86.554)\n",
      "[270/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6656 (0.6172)\tPrec@1 85.938 (86.615)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5893 (0.6174)\tPrec@1 85.938 (86.605)\n",
      "[290/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6639 (0.6172)\tPrec@1 87.500 (86.590)\n",
      "[300/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5507 (0.6173)\tPrec@1 91.406 (86.568)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5906 (0.6164)\tPrec@1 89.844 (86.616)\n",
      "[320/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5813 (0.6163)\tPrec@1 89.062 (86.643)\n",
      "[330/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5164 (0.6177)\tPrec@1 91.406 (86.587)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.7460 (0.6192)\tPrec@1 82.812 (86.535)\n",
      "[350/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6525 (0.6202)\tPrec@1 85.938 (86.498)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6889 (0.6205)\tPrec@1 84.375 (86.509)\n",
      "[370/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5574 (0.6199)\tPrec@1 89.062 (86.527)\n",
      "[380/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6195 (0.6199)\tPrec@1 87.500 (86.505)\n",
      "[390/391]\tTime 0.128 (0.181)\tData 0.014 (0.020)\tLoss 0.7176 (0.6209)\tPrec@1 86.250 (86.498)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.1070 (1.1070)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.020 (0.022)\tLoss 0.9875 (1.0520)\tPrec@1 81.250 (80.043)\n",
      "Test: [20/79]\tTime 0.020 (0.021)\tLoss 1.0668 (1.1143)\tPrec@1 80.469 (79.241)\n",
      "Test: [30/79]\tTime 0.020 (0.021)\tLoss 0.7616 (1.0971)\tPrec@1 85.938 (79.763)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.2288 (1.0888)\tPrec@1 78.906 (79.973)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 0.8729 (1.0734)\tPrec@1 78.125 (80.147)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 1.2866 (1.0629)\tPrec@1 78.906 (80.443)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.9013 (1.0505)\tPrec@1 83.594 (80.645)\n",
      " * Prec@1 80.910\n",
      "\n",
      "===> epoch: 71/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.176 (0.176)\tData 0.020 (0.020)\tLoss 0.6781 (0.6781)\tPrec@1 82.812 (82.812)\n",
      "[10/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.6584 (0.6685)\tPrec@1 84.375 (84.943)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6351 (0.6472)\tPrec@1 85.156 (85.454)\n",
      "[30/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5726 (0.6241)\tPrec@1 90.625 (86.341)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6450 (0.6195)\tPrec@1 86.719 (86.662)\n",
      "[50/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5768 (0.6186)\tPrec@1 88.281 (86.765)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[60/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5965 (0.6175)\tPrec@1 85.156 (86.719)\n",
      "[70/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.5420 (0.6182)\tPrec@1 89.844 (86.741)\n",
      "[80/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.6265 (0.6156)\tPrec@1 85.938 (86.786)\n",
      "[90/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.5956 (0.6141)\tPrec@1 89.844 (86.753)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5959 (0.6154)\tPrec@1 87.500 (86.680)\n",
      "[110/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6128 (0.6134)\tPrec@1 85.156 (86.796)\n",
      "[120/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6736 (0.6134)\tPrec@1 84.375 (86.880)\n",
      "[130/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5240 (0.6120)\tPrec@1 93.750 (86.963)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6714 (0.6162)\tPrec@1 87.500 (86.830)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5709 (0.6164)\tPrec@1 86.719 (86.786)\n",
      "[160/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6235 (0.6182)\tPrec@1 85.156 (86.772)\n",
      "[170/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.6462 (0.6173)\tPrec@1 84.375 (86.801)\n",
      "[180/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6316 (0.6191)\tPrec@1 86.719 (86.719)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5485 (0.6189)\tPrec@1 89.844 (86.751)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7208 (0.6183)\tPrec@1 83.594 (86.715)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5493 (0.6190)\tPrec@1 88.281 (86.693)\n",
      "[220/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5535 (0.6186)\tPrec@1 89.062 (86.676)\n",
      "[230/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5581 (0.6193)\tPrec@1 87.500 (86.648)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5521 (0.6191)\tPrec@1 88.281 (86.644)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6721 (0.6190)\tPrec@1 82.812 (86.616)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6419 (0.6195)\tPrec@1 87.500 (86.620)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6317 (0.6195)\tPrec@1 86.719 (86.618)\n",
      "[280/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5100 (0.6184)\tPrec@1 89.844 (86.633)\n",
      "[290/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6492 (0.6174)\tPrec@1 85.156 (86.657)\n",
      "[300/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5500 (0.6176)\tPrec@1 89.062 (86.656)\n",
      "[310/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5999 (0.6171)\tPrec@1 89.844 (86.691)\n",
      "[320/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6411 (0.6170)\tPrec@1 87.500 (86.690)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5638 (0.6178)\tPrec@1 89.062 (86.674)\n",
      "[340/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5630 (0.6173)\tPrec@1 92.188 (86.712)\n",
      "[350/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6610 (0.6169)\tPrec@1 85.938 (86.725)\n",
      "[360/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7533 (0.6163)\tPrec@1 80.469 (86.730)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6470 (0.6162)\tPrec@1 86.719 (86.744)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6068 (0.6166)\tPrec@1 85.938 (86.700)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.013 (0.020)\tLoss 0.6423 (0.6182)\tPrec@1 83.750 (86.628)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.0682 (1.0682)\tPrec@1 79.688 (79.688)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.7041 (1.1335)\tPrec@1 86.719 (80.327)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 1.2755 (1.2649)\tPrec@1 80.469 (79.390)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.9913 (1.2612)\tPrec@1 82.812 (79.007)\n",
      "Test: [40/79]\tTime 0.020 (0.019)\tLoss 1.2461 (1.2643)\tPrec@1 78.125 (78.849)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 1.3029 (1.2578)\tPrec@1 77.344 (78.814)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 1.0260 (1.2680)\tPrec@1 82.812 (78.753)\n",
      "Test: [70/79]\tTime 0.019 (0.019)\tLoss 1.3412 (1.2754)\tPrec@1 80.469 (78.686)\n",
      " * Prec@1 78.720\n",
      "\n",
      "===> epoch: 72/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.021 (0.021)\tLoss 0.6654 (0.6654)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.180 (0.184)\tData 0.019 (0.020)\tLoss 0.6370 (0.5787)\tPrec@1 85.938 (88.778)\n",
      "[20/391]\tTime 0.179 (0.183)\tData 0.019 (0.020)\tLoss 0.6466 (0.5844)\tPrec@1 83.594 (88.318)\n",
      "[30/391]\tTime 0.186 (0.183)\tData 0.019 (0.020)\tLoss 0.5754 (0.5880)\tPrec@1 86.719 (88.080)\n",
      "[40/391]\tTime 0.187 (0.183)\tData 0.020 (0.020)\tLoss 0.6394 (0.5915)\tPrec@1 87.500 (87.938)\n",
      "[50/391]\tTime 0.183 (0.183)\tData 0.019 (0.020)\tLoss 0.5760 (0.5968)\tPrec@1 86.719 (87.623)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7454 (0.6027)\tPrec@1 78.906 (87.372)\n",
      "[70/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6813 (0.6006)\tPrec@1 82.812 (87.324)\n",
      "[80/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.5972 (0.6046)\tPrec@1 87.500 (87.095)\n",
      "[90/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6670 (0.6068)\tPrec@1 85.938 (86.933)\n",
      "[100/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.5302 (0.6083)\tPrec@1 92.188 (86.966)\n",
      "[110/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6813 (0.6109)\tPrec@1 86.719 (86.937)\n",
      "[120/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6285 (0.6099)\tPrec@1 87.500 (87.061)\n",
      "[130/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.4904 (0.6094)\tPrec@1 90.625 (87.100)\n",
      "[140/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5401 (0.6080)\tPrec@1 89.062 (87.107)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.7742 (0.6088)\tPrec@1 82.812 (87.096)\n",
      "[160/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5851 (0.6101)\tPrec@1 87.500 (87.044)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6353 (0.6133)\tPrec@1 86.719 (86.947)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6929 (0.6182)\tPrec@1 82.812 (86.796)\n",
      "[190/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5807 (0.6189)\tPrec@1 88.281 (86.780)\n",
      "[200/391]\tTime 0.181 (0.181)\tData 0.022 (0.020)\tLoss 0.5329 (0.6177)\tPrec@1 89.844 (86.851)\n",
      "[210/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5788 (0.6189)\tPrec@1 88.281 (86.785)\n",
      "[220/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6591 (0.6180)\tPrec@1 84.375 (86.828)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5426 (0.6172)\tPrec@1 88.281 (86.824)\n",
      "[240/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5358 (0.6151)\tPrec@1 90.625 (86.900)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5584 (0.6151)\tPrec@1 90.625 (86.912)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5819 (0.6169)\tPrec@1 88.281 (86.841)\n",
      "[270/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6064 (0.6168)\tPrec@1 88.281 (86.851)\n",
      "[280/391]\tTime 0.183 (0.181)\tData 0.024 (0.020)\tLoss 0.5882 (0.6180)\tPrec@1 87.500 (86.774)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6035 (0.6176)\tPrec@1 85.938 (86.756)\n",
      "[300/391]\tTime 0.189 (0.181)\tData 0.019 (0.020)\tLoss 0.5911 (0.6188)\tPrec@1 87.500 (86.729)\n",
      "[310/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.6803 (0.6201)\tPrec@1 87.500 (86.691)\n",
      "[320/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5842 (0.6201)\tPrec@1 85.938 (86.675)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5504 (0.6213)\tPrec@1 91.406 (86.617)\n",
      "[340/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.7175 (0.6213)\tPrec@1 82.812 (86.634)\n",
      "[350/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.5471 (0.6217)\tPrec@1 92.969 (86.641)\n",
      "[360/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6728 (0.6212)\tPrec@1 82.812 (86.637)\n",
      "[370/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5236 (0.6212)\tPrec@1 89.062 (86.635)\n",
      "[380/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6651 (0.6208)\tPrec@1 86.719 (86.653)\n",
      "[390/391]\tTime 0.113 (0.181)\tData 0.012 (0.020)\tLoss 0.5562 (0.6205)\tPrec@1 91.250 (86.678)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8999 (0.8999)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.019 (0.020)\tLoss 1.1835 (1.1177)\tPrec@1 78.125 (79.048)\n",
      "Test: [20/79]\tTime 0.021 (0.020)\tLoss 1.0127 (1.1938)\tPrec@1 78.125 (78.423)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.9011 (1.1770)\tPrec@1 85.156 (78.301)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.2167 (1.1668)\tPrec@1 78.906 (78.373)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 1.2836 (1.1663)\tPrec@1 73.438 (78.278)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 1.2483 (1.1634)\tPrec@1 79.688 (78.304)\n",
      "Test: [70/79]\tTime 0.021 (0.019)\tLoss 1.2685 (1.1757)\tPrec@1 78.906 (78.268)\n",
      " * Prec@1 78.530\n",
      "\n",
      "===> epoch: 73/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.6348 (0.6348)\tPrec@1 86.719 (86.719)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5827 (0.6139)\tPrec@1 86.719 (87.287)\n",
      "[20/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.8289 (0.6323)\tPrec@1 80.469 (86.235)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5927 (0.6313)\tPrec@1 88.281 (86.568)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6396 (0.6265)\tPrec@1 83.594 (86.585)\n",
      "[50/391]\tTime 0.190 (0.181)\tData 0.020 (0.020)\tLoss 0.6423 (0.6303)\tPrec@1 85.156 (86.137)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.4975 (0.6212)\tPrec@1 92.188 (86.565)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6000 (0.6211)\tPrec@1 84.375 (86.499)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6318 (0.6216)\tPrec@1 86.719 (86.400)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6319 (0.6201)\tPrec@1 85.156 (86.418)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6397 (0.6213)\tPrec@1 86.719 (86.417)\n",
      "[110/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.6907 (0.6227)\tPrec@1 85.156 (86.289)\n",
      "[120/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6560 (0.6246)\tPrec@1 86.719 (86.267)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6727 (0.6251)\tPrec@1 84.375 (86.283)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6930 (0.6267)\tPrec@1 85.156 (86.226)\n",
      "[150/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5371 (0.6266)\tPrec@1 90.625 (86.243)\n",
      "[160/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.6849 (0.6256)\tPrec@1 83.594 (86.248)\n",
      "[170/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.5866 (0.6252)\tPrec@1 86.719 (86.280)\n",
      "[180/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6049 (0.6260)\tPrec@1 85.156 (86.205)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5624 (0.6239)\tPrec@1 88.281 (86.269)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6313 (0.6216)\tPrec@1 87.500 (86.361)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6081 (0.6211)\tPrec@1 86.719 (86.423)\n",
      "[220/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7083 (0.6214)\tPrec@1 82.812 (86.411)\n",
      "[230/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5550 (0.6203)\tPrec@1 87.500 (86.455)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5560 (0.6195)\tPrec@1 89.844 (86.485)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7506 (0.6208)\tPrec@1 81.250 (86.470)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5341 (0.6219)\tPrec@1 90.625 (86.443)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6530 (0.6219)\tPrec@1 84.375 (86.433)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5753 (0.6209)\tPrec@1 89.062 (86.463)\n",
      "[290/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6317 (0.6200)\tPrec@1 85.938 (86.504)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5159 (0.6197)\tPrec@1 89.844 (86.524)\n",
      "[310/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6001 (0.6196)\tPrec@1 85.938 (86.503)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.7759 (0.6202)\tPrec@1 82.031 (86.526)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5690 (0.6203)\tPrec@1 89.844 (86.544)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6769 (0.6200)\tPrec@1 85.938 (86.563)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6174 (0.6197)\tPrec@1 87.500 (86.581)\n",
      "[360/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6683 (0.6204)\tPrec@1 83.594 (86.565)\n",
      "[370/391]\tTime 0.186 (0.181)\tData 0.021 (0.020)\tLoss 0.6489 (0.6205)\tPrec@1 85.156 (86.544)\n",
      "[380/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6379 (0.6202)\tPrec@1 86.719 (86.567)\n",
      "[390/391]\tTime 0.114 (0.181)\tData 0.012 (0.020)\tLoss 0.6303 (0.6205)\tPrec@1 86.250 (86.570)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8638 (0.8638)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.5951 (0.8651)\tPrec@1 89.062 (84.233)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.6445 (0.8919)\tPrec@1 85.938 (83.333)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.7913 (0.8974)\tPrec@1 84.375 (83.241)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.5796 (0.8975)\tPrec@1 85.938 (83.384)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.0198 (0.8964)\tPrec@1 84.375 (83.272)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.9022 (0.9108)\tPrec@1 81.250 (83.056)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.7327 (0.9181)\tPrec@1 85.938 (82.846)\n",
      " * Prec@1 82.980\n",
      "\n",
      "===> epoch: 74/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.022 (0.022)\tLoss 0.5558 (0.5558)\tPrec@1 89.844 (89.844)\n",
      "[10/391]\tTime 0.196 (0.185)\tData 0.034 (0.021)\tLoss 0.5539 (0.5837)\tPrec@1 89.844 (88.210)\n",
      "[20/391]\tTime 0.180 (0.183)\tData 0.020 (0.021)\tLoss 0.5780 (0.5947)\tPrec@1 85.938 (87.574)\n",
      "[30/391]\tTime 0.181 (0.183)\tData 0.019 (0.020)\tLoss 0.5956 (0.5899)\tPrec@1 87.500 (87.878)\n",
      "[40/391]\tTime 0.184 (0.183)\tData 0.019 (0.020)\tLoss 0.5747 (0.5945)\tPrec@1 89.062 (87.367)\n",
      "[50/391]\tTime 0.184 (0.183)\tData 0.019 (0.020)\tLoss 0.6156 (0.5966)\tPrec@1 86.719 (87.331)\n",
      "[60/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.7194 (0.6032)\tPrec@1 81.250 (86.962)\n",
      "[70/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5268 (0.6014)\tPrec@1 91.406 (86.994)\n",
      "[80/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5771 (0.6005)\tPrec@1 88.281 (87.066)\n",
      "[90/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5975 (0.6035)\tPrec@1 86.719 (86.899)\n",
      "[100/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6515 (0.6055)\tPrec@1 85.938 (86.796)\n",
      "[110/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5580 (0.6088)\tPrec@1 90.625 (86.719)\n",
      "[120/391]\tTime 0.189 (0.182)\tData 0.027 (0.020)\tLoss 0.4795 (0.6081)\tPrec@1 93.750 (86.777)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5566 (0.6092)\tPrec@1 89.062 (86.755)\n",
      "[140/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6367 (0.6107)\tPrec@1 89.844 (86.752)\n",
      "[150/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6409 (0.6098)\tPrec@1 85.938 (86.750)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5367 (0.6112)\tPrec@1 91.406 (86.738)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5953 (0.6120)\tPrec@1 89.062 (86.682)\n",
      "[180/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6951 (0.6123)\tPrec@1 86.719 (86.684)\n",
      "[190/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6492 (0.6134)\tPrec@1 86.719 (86.588)\n",
      "[200/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5076 (0.6127)\tPrec@1 93.750 (86.653)\n",
      "[210/391]\tTime 0.188 (0.182)\tData 0.027 (0.020)\tLoss 0.6542 (0.6124)\tPrec@1 85.156 (86.689)\n",
      "[220/391]\tTime 0.185 (0.182)\tData 0.023 (0.020)\tLoss 0.5616 (0.6120)\tPrec@1 88.281 (86.673)\n",
      "[230/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6235 (0.6122)\tPrec@1 85.156 (86.644)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6420 (0.6138)\tPrec@1 85.156 (86.589)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6451 (0.6142)\tPrec@1 85.156 (86.585)\n",
      "[260/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6774 (0.6146)\tPrec@1 86.719 (86.563)\n",
      "[270/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6927 (0.6154)\tPrec@1 85.156 (86.543)\n",
      "[280/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6484 (0.6162)\tPrec@1 84.375 (86.521)\n",
      "[290/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6708 (0.6161)\tPrec@1 82.812 (86.520)\n",
      "[300/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6499 (0.6163)\tPrec@1 85.938 (86.537)\n",
      "[310/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6603 (0.6171)\tPrec@1 83.594 (86.510)\n",
      "[320/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6215 (0.6173)\tPrec@1 89.844 (86.531)\n",
      "[330/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6126 (0.6168)\tPrec@1 84.375 (86.532)\n",
      "[340/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7366 (0.6169)\tPrec@1 84.375 (86.533)\n",
      "[350/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.7326 (0.6180)\tPrec@1 81.250 (86.478)\n",
      "[360/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6072 (0.6184)\tPrec@1 84.375 (86.466)\n",
      "[370/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5833 (0.6186)\tPrec@1 89.062 (86.453)\n",
      "[380/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4877 (0.6183)\tPrec@1 91.406 (86.473)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.013 (0.020)\tLoss 0.6422 (0.6187)\tPrec@1 87.500 (86.462)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9423 (0.9423)\tPrec@1 84.375 (84.375)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.9172 (0.9861)\tPrec@1 85.938 (80.895)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.7743 (1.0245)\tPrec@1 86.719 (80.655)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.6707 (0.9918)\tPrec@1 87.500 (80.922)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.0662 (0.9834)\tPrec@1 75.781 (81.079)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.9401 (0.9823)\tPrec@1 81.250 (81.158)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.7329 (0.9715)\tPrec@1 88.281 (81.378)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.0819 (0.9695)\tPrec@1 79.688 (81.459)\n",
      " * Prec@1 81.480\n",
      "\n",
      "===> epoch: 75/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.7351 (0.7351)\tPrec@1 82.031 (82.031)\n",
      "[10/391]\tTime 0.179 (0.185)\tData 0.019 (0.020)\tLoss 0.5691 (0.6370)\tPrec@1 85.938 (85.085)\n",
      "[20/391]\tTime 0.181 (0.184)\tData 0.019 (0.020)\tLoss 0.5773 (0.6153)\tPrec@1 89.062 (86.272)\n",
      "[30/391]\tTime 0.189 (0.183)\tData 0.023 (0.020)\tLoss 0.6670 (0.6066)\tPrec@1 84.375 (86.769)\n",
      "[40/391]\tTime 0.183 (0.183)\tData 0.020 (0.020)\tLoss 0.5773 (0.6020)\tPrec@1 89.062 (86.966)\n",
      "[50/391]\tTime 0.183 (0.183)\tData 0.019 (0.020)\tLoss 0.6388 (0.6018)\tPrec@1 85.938 (87.117)\n",
      "[60/391]\tTime 0.182 (0.183)\tData 0.019 (0.020)\tLoss 0.5475 (0.6005)\tPrec@1 89.062 (87.244)\n",
      "[70/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6298 (0.6032)\tPrec@1 87.500 (87.137)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5786 (0.6078)\tPrec@1 86.719 (86.912)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6594 (0.6095)\tPrec@1 83.594 (86.796)\n",
      "[100/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5935 (0.6097)\tPrec@1 87.500 (86.881)\n",
      "[110/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.7241 (0.6073)\tPrec@1 83.594 (86.979)\n",
      "[120/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6285 (0.6072)\tPrec@1 85.156 (87.003)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5730 (0.6072)\tPrec@1 87.500 (86.951)\n",
      "[140/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5735 (0.6073)\tPrec@1 88.281 (86.929)\n",
      "[150/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5367 (0.6070)\tPrec@1 86.719 (86.905)\n",
      "[160/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5914 (0.6061)\tPrec@1 86.719 (86.942)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7188 (0.6061)\tPrec@1 84.375 (86.947)\n",
      "[180/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.4739 (0.6032)\tPrec@1 92.188 (87.116)\n",
      "[190/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5705 (0.6013)\tPrec@1 84.375 (87.169)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6463 (0.6014)\tPrec@1 85.938 (87.150)\n",
      "[210/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6862 (0.6024)\tPrec@1 85.156 (87.093)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6335 (0.6041)\tPrec@1 84.375 (87.005)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5138 (0.6055)\tPrec@1 89.844 (86.912)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7267 (0.6070)\tPrec@1 81.250 (86.848)\n",
      "[250/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5658 (0.6076)\tPrec@1 90.625 (86.831)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5989 (0.6076)\tPrec@1 86.719 (86.871)\n",
      "[270/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6190 (0.6086)\tPrec@1 84.375 (86.823)\n",
      "[280/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6585 (0.6088)\tPrec@1 83.594 (86.819)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6077 (0.6087)\tPrec@1 89.062 (86.802)\n",
      "[300/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5345 (0.6098)\tPrec@1 89.844 (86.768)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5782 (0.6097)\tPrec@1 90.625 (86.774)\n",
      "[320/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.6224 (0.6100)\tPrec@1 87.500 (86.724)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5533 (0.6118)\tPrec@1 86.719 (86.636)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7181 (0.6125)\tPrec@1 82.812 (86.613)\n",
      "[350/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7097 (0.6137)\tPrec@1 85.938 (86.565)\n",
      "[360/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6072 (0.6149)\tPrec@1 88.281 (86.548)\n",
      "[370/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5745 (0.6158)\tPrec@1 89.844 (86.510)\n",
      "[380/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.8220 (0.6173)\tPrec@1 78.125 (86.454)\n",
      "[390/391]\tTime 0.111 (0.181)\tData 0.012 (0.020)\tLoss 0.6137 (0.6184)\tPrec@1 87.500 (86.414)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8532 (0.8532)\tPrec@1 82.031 (82.031)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.9426 (1.0142)\tPrec@1 82.812 (80.469)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.8923 (1.0859)\tPrec@1 79.688 (80.134)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.9434 (1.0757)\tPrec@1 78.906 (79.738)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.9814 (1.0633)\tPrec@1 80.469 (80.202)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 1.1331 (1.0436)\tPrec@1 82.031 (80.499)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 0.8256 (1.0399)\tPrec@1 82.812 (80.520)\n",
      "Test: [70/79]\tTime 0.019 (0.019)\tLoss 1.0221 (1.0425)\tPrec@1 81.250 (80.502)\n",
      " * Prec@1 80.670\n",
      "\n",
      "===> epoch: 76/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.186 (0.186)\tData 0.022 (0.022)\tLoss 0.6236 (0.6236)\tPrec@1 89.844 (89.844)\n",
      "[10/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6015 (0.6155)\tPrec@1 87.500 (86.364)\n",
      "[20/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6385 (0.6393)\tPrec@1 86.719 (85.342)\n",
      "[30/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5610 (0.6281)\tPrec@1 89.062 (85.786)\n",
      "[40/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6229 (0.6225)\tPrec@1 86.719 (86.071)\n",
      "[50/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.6095 (0.6151)\tPrec@1 89.062 (86.550)\n",
      "[60/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.5858 (0.6093)\tPrec@1 88.281 (86.770)\n",
      "[70/391]\tTime 0.178 (0.182)\tData 0.019 (0.019)\tLoss 0.5935 (0.6046)\tPrec@1 87.500 (86.972)\n",
      "[80/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.6095 (0.6061)\tPrec@1 86.719 (86.950)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.5750 (0.6054)\tPrec@1 89.062 (86.993)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4734 (0.6075)\tPrec@1 90.625 (86.935)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.5616 (0.6086)\tPrec@1 86.719 (86.860)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5883 (0.6101)\tPrec@1 86.719 (86.887)\n",
      "[130/391]\tTime 0.184 (0.181)\tData 0.019 (0.019)\tLoss 0.6058 (0.6069)\tPrec@1 85.938 (87.023)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.4883 (0.6073)\tPrec@1 91.406 (86.963)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.5985 (0.6065)\tPrec@1 87.500 (86.957)\n",
      "[160/391]\tTime 0.178 (0.181)\tData 0.019 (0.019)\tLoss 0.5629 (0.6060)\tPrec@1 90.625 (87.005)\n",
      "[170/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.5285 (0.6071)\tPrec@1 90.625 (86.975)\n",
      "[180/391]\tTime 0.183 (0.181)\tData 0.019 (0.019)\tLoss 0.5639 (0.6054)\tPrec@1 89.844 (87.116)\n",
      "[190/391]\tTime 0.183 (0.181)\tData 0.019 (0.019)\tLoss 0.6936 (0.6057)\tPrec@1 82.031 (87.075)\n",
      "[200/391]\tTime 0.184 (0.181)\tData 0.019 (0.019)\tLoss 0.5151 (0.6068)\tPrec@1 89.844 (87.034)\n",
      "[210/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.6866 (0.6085)\tPrec@1 85.156 (86.941)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6294 (0.6095)\tPrec@1 86.719 (86.927)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.7015 (0.6121)\tPrec@1 85.156 (86.820)\n",
      "[240/391]\tTime 0.178 (0.181)\tData 0.020 (0.019)\tLoss 0.5803 (0.6129)\tPrec@1 89.844 (86.826)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.021 (0.019)\tLoss 0.7430 (0.6124)\tPrec@1 85.156 (86.846)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.021 (0.019)\tLoss 0.5979 (0.6125)\tPrec@1 86.719 (86.868)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.4955 (0.6123)\tPrec@1 94.531 (86.883)\n",
      "[280/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.6740 (0.6130)\tPrec@1 82.031 (86.858)\n",
      "[290/391]\tTime 0.178 (0.181)\tData 0.019 (0.019)\tLoss 0.6393 (0.6128)\tPrec@1 82.812 (86.866)\n",
      "[300/391]\tTime 0.178 (0.181)\tData 0.021 (0.019)\tLoss 0.6057 (0.6128)\tPrec@1 88.281 (86.849)\n",
      "[310/391]\tTime 0.185 (0.181)\tData 0.020 (0.019)\tLoss 0.6847 (0.6118)\tPrec@1 83.594 (86.897)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7362 (0.6117)\tPrec@1 82.812 (86.889)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6638 (0.6119)\tPrec@1 85.938 (86.889)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.8983 (0.6137)\tPrec@1 75.781 (86.820)\n",
      "[350/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7789 (0.6148)\tPrec@1 82.031 (86.803)\n",
      "[360/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5487 (0.6158)\tPrec@1 91.406 (86.736)\n",
      "[370/391]\tTime 0.192 (0.181)\tData 0.025 (0.020)\tLoss 0.7123 (0.6169)\tPrec@1 82.812 (86.668)\n",
      "[380/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.5660 (0.6170)\tPrec@1 89.844 (86.661)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[390/391]\tTime 0.120 (0.181)\tData 0.012 (0.020)\tLoss 0.6988 (0.6175)\tPrec@1 81.250 (86.634)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 1.4628 (1.4628)\tPrec@1 73.438 (73.438)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 1.4101 (1.5716)\tPrec@1 75.000 (72.869)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 1.4443 (1.6180)\tPrec@1 77.344 (72.656)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 1.4609 (1.6431)\tPrec@1 75.000 (72.379)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.6813 (1.6174)\tPrec@1 70.312 (72.618)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.4560 (1.6036)\tPrec@1 73.438 (72.626)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 1.8221 (1.6064)\tPrec@1 70.312 (72.643)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.6517 (1.5918)\tPrec@1 73.438 (72.975)\n",
      " * Prec@1 73.170\n",
      "\n",
      "===> epoch: 77/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.172 (0.172)\tData 0.020 (0.020)\tLoss 0.6154 (0.6154)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.7050 (0.6226)\tPrec@1 81.250 (86.364)\n",
      "[20/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6041 (0.6285)\tPrec@1 87.500 (86.124)\n",
      "[30/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6797 (0.6243)\tPrec@1 85.156 (86.341)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6183 (0.6303)\tPrec@1 85.938 (86.185)\n",
      "[50/391]\tTime 0.183 (0.181)\tData 0.024 (0.020)\tLoss 0.6877 (0.6222)\tPrec@1 82.031 (86.550)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6014 (0.6148)\tPrec@1 86.719 (86.719)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.5594 (0.6210)\tPrec@1 90.625 (86.598)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6696 (0.6196)\tPrec@1 82.031 (86.593)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5821 (0.6158)\tPrec@1 87.500 (86.762)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6324 (0.6181)\tPrec@1 87.500 (86.703)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6652 (0.6186)\tPrec@1 85.156 (86.662)\n",
      "[120/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5219 (0.6182)\tPrec@1 92.188 (86.654)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7089 (0.6152)\tPrec@1 80.469 (86.695)\n",
      "[140/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.6575 (0.6136)\tPrec@1 85.156 (86.741)\n",
      "[150/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6139 (0.6158)\tPrec@1 89.062 (86.693)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5938 (0.6153)\tPrec@1 90.625 (86.704)\n",
      "[170/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5879 (0.6183)\tPrec@1 90.625 (86.655)\n",
      "[180/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6310 (0.6214)\tPrec@1 85.938 (86.563)\n",
      "[190/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6105 (0.6206)\tPrec@1 85.938 (86.543)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6679 (0.6203)\tPrec@1 85.156 (86.583)\n",
      "[210/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6075 (0.6198)\tPrec@1 89.062 (86.611)\n",
      "[220/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6531 (0.6216)\tPrec@1 87.500 (86.591)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6321 (0.6220)\tPrec@1 85.156 (86.597)\n",
      "[240/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5916 (0.6213)\tPrec@1 86.719 (86.605)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6707 (0.6213)\tPrec@1 89.062 (86.579)\n",
      "[260/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6325 (0.6216)\tPrec@1 84.375 (86.572)\n",
      "[270/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6508 (0.6220)\tPrec@1 88.281 (86.566)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5714 (0.6226)\tPrec@1 89.062 (86.535)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6035 (0.6233)\tPrec@1 86.719 (86.477)\n",
      "[300/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6401 (0.6226)\tPrec@1 81.250 (86.506)\n",
      "[310/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5910 (0.6213)\tPrec@1 87.500 (86.550)\n",
      "[320/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6086 (0.6221)\tPrec@1 87.500 (86.534)\n",
      "[330/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6304 (0.6229)\tPrec@1 85.938 (86.499)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5825 (0.6218)\tPrec@1 86.719 (86.565)\n",
      "[350/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6177 (0.6231)\tPrec@1 84.375 (86.492)\n",
      "[360/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7265 (0.6228)\tPrec@1 85.156 (86.517)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6725 (0.6237)\tPrec@1 84.375 (86.470)\n",
      "[380/391]\tTime 0.184 (0.181)\tData 0.022 (0.020)\tLoss 0.4889 (0.6233)\tPrec@1 91.406 (86.487)\n",
      "[390/391]\tTime 0.117 (0.181)\tData 0.012 (0.020)\tLoss 0.5348 (0.6234)\tPrec@1 88.750 (86.486)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.6280 (0.6280)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.020 (0.021)\tLoss 0.6637 (0.8062)\tPrec@1 82.031 (82.741)\n",
      "Test: [20/79]\tTime 0.020 (0.021)\tLoss 0.6465 (0.8338)\tPrec@1 86.719 (82.887)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.6428 (0.8498)\tPrec@1 85.156 (82.510)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.7257 (0.8520)\tPrec@1 81.250 (82.641)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.0118 (0.8561)\tPrec@1 81.250 (82.843)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.7769 (0.8471)\tPrec@1 84.375 (82.966)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 1.0055 (0.8410)\tPrec@1 80.469 (83.066)\n",
      " * Prec@1 83.130\n",
      "\n",
      "===> epoch: 78/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.167 (0.167)\tData 0.021 (0.021)\tLoss 0.7591 (0.7591)\tPrec@1 81.250 (81.250)\n",
      "[10/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6684 (0.6100)\tPrec@1 85.156 (87.216)\n",
      "[20/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5933 (0.6082)\tPrec@1 86.719 (86.868)\n",
      "[30/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5276 (0.6095)\tPrec@1 90.625 (86.870)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7201 (0.6112)\tPrec@1 83.594 (87.119)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6881 (0.6031)\tPrec@1 82.812 (87.178)\n",
      "[60/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6192 (0.6073)\tPrec@1 86.719 (86.860)\n",
      "[70/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.7623 (0.6109)\tPrec@1 79.688 (86.576)\n",
      "[80/391]\tTime 0.184 (0.181)\tData 0.021 (0.020)\tLoss 0.5790 (0.6152)\tPrec@1 86.719 (86.429)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6539 (0.6167)\tPrec@1 85.156 (86.341)\n",
      "[100/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7286 (0.6197)\tPrec@1 82.812 (86.317)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6400 (0.6163)\tPrec@1 83.594 (86.479)\n",
      "[120/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6131 (0.6140)\tPrec@1 85.156 (86.590)\n",
      "[130/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5562 (0.6125)\tPrec@1 92.188 (86.647)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6392 (0.6149)\tPrec@1 86.719 (86.580)\n",
      "[150/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5456 (0.6147)\tPrec@1 86.719 (86.600)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7232 (0.6138)\tPrec@1 83.594 (86.661)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6967 (0.6142)\tPrec@1 81.250 (86.627)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7553 (0.6152)\tPrec@1 81.250 (86.563)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6009 (0.6153)\tPrec@1 85.938 (86.592)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4848 (0.6134)\tPrec@1 92.188 (86.633)\n",
      "[210/391]\tTime 0.177 (0.182)\tData 0.019 (0.020)\tLoss 0.6057 (0.6133)\tPrec@1 85.938 (86.645)\n",
      "[220/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6803 (0.6137)\tPrec@1 85.156 (86.666)\n",
      "[230/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.4905 (0.6120)\tPrec@1 89.844 (86.732)\n",
      "[240/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6046 (0.6126)\tPrec@1 87.500 (86.725)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6133 (0.6130)\tPrec@1 85.156 (86.663)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6299 (0.6130)\tPrec@1 83.594 (86.644)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5935 (0.6116)\tPrec@1 89.062 (86.687)\n",
      "[280/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5432 (0.6111)\tPrec@1 88.281 (86.724)\n",
      "[290/391]\tTime 0.192 (0.182)\tData 0.020 (0.020)\tLoss 0.5640 (0.6120)\tPrec@1 89.062 (86.700)\n",
      "[300/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7179 (0.6128)\tPrec@1 82.031 (86.680)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5459 (0.6134)\tPrec@1 89.062 (86.674)\n",
      "[320/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.6764 (0.6138)\tPrec@1 86.719 (86.673)\n",
      "[330/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.5862 (0.6144)\tPrec@1 88.281 (86.676)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[340/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.7320 (0.6155)\tPrec@1 82.812 (86.671)\n",
      "[350/391]\tTime 0.182 (0.182)\tData 0.023 (0.020)\tLoss 0.5657 (0.6158)\tPrec@1 89.062 (86.674)\n",
      "[360/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5214 (0.6151)\tPrec@1 90.625 (86.712)\n",
      "[370/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.4942 (0.6128)\tPrec@1 91.406 (86.795)\n",
      "[380/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5563 (0.6134)\tPrec@1 90.625 (86.766)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.012 (0.020)\tLoss 0.6098 (0.6135)\tPrec@1 90.000 (86.766)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7148 (0.7148)\tPrec@1 85.938 (85.938)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.8802 (0.9288)\tPrec@1 83.594 (82.457)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.9524 (0.9691)\tPrec@1 82.031 (81.994)\n",
      "Test: [30/79]\tTime 0.020 (0.019)\tLoss 0.6247 (0.9521)\tPrec@1 86.719 (81.804)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.1247 (0.9483)\tPrec@1 78.906 (81.688)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 1.3586 (0.9503)\tPrec@1 77.344 (81.526)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 1.0400 (0.9435)\tPrec@1 78.125 (81.660)\n",
      "Test: [70/79]\tTime 0.019 (0.019)\tLoss 0.6675 (0.9373)\tPrec@1 85.156 (81.723)\n",
      " * Prec@1 81.830\n",
      "\n",
      "===> epoch: 79/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.021 (0.021)\tLoss 0.5807 (0.5807)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.5745 (0.5928)\tPrec@1 88.281 (87.003)\n",
      "[20/391]\tTime 0.178 (0.182)\tData 0.019 (0.019)\tLoss 0.6532 (0.6152)\tPrec@1 85.156 (86.272)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5357 (0.6040)\tPrec@1 89.844 (86.442)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6700 (0.6087)\tPrec@1 82.812 (86.204)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6575 (0.6029)\tPrec@1 82.031 (86.550)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5641 (0.6132)\tPrec@1 90.625 (86.373)\n",
      "[70/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5235 (0.6132)\tPrec@1 90.625 (86.455)\n",
      "[80/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6216 (0.6110)\tPrec@1 84.375 (86.526)\n",
      "[90/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7244 (0.6101)\tPrec@1 83.594 (86.599)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6279 (0.6095)\tPrec@1 86.719 (86.696)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6935 (0.6107)\tPrec@1 83.594 (86.641)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6936 (0.6113)\tPrec@1 82.031 (86.745)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6224 (0.6123)\tPrec@1 85.156 (86.749)\n",
      "[140/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5810 (0.6104)\tPrec@1 88.281 (86.818)\n",
      "[150/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5698 (0.6117)\tPrec@1 89.844 (86.838)\n",
      "[160/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6043 (0.6115)\tPrec@1 86.719 (86.801)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6202 (0.6109)\tPrec@1 89.062 (86.787)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6118 (0.6111)\tPrec@1 83.594 (86.732)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6999 (0.6122)\tPrec@1 82.031 (86.682)\n",
      "[200/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6588 (0.6133)\tPrec@1 83.594 (86.645)\n",
      "[210/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6713 (0.6132)\tPrec@1 82.031 (86.641)\n",
      "[220/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6435 (0.6138)\tPrec@1 85.938 (86.627)\n",
      "[230/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.4348 (0.6120)\tPrec@1 93.750 (86.678)\n",
      "[240/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.7601 (0.6110)\tPrec@1 84.375 (86.738)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5565 (0.6126)\tPrec@1 88.281 (86.650)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5242 (0.6120)\tPrec@1 92.188 (86.719)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5848 (0.6121)\tPrec@1 88.281 (86.713)\n",
      "[280/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6583 (0.6121)\tPrec@1 84.375 (86.710)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5909 (0.6119)\tPrec@1 88.281 (86.713)\n",
      "[300/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.5590 (0.6116)\tPrec@1 89.062 (86.721)\n",
      "[310/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6049 (0.6136)\tPrec@1 89.062 (86.633)\n",
      "[320/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.4736 (0.6141)\tPrec@1 93.750 (86.629)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6739 (0.6143)\tPrec@1 85.156 (86.667)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5795 (0.6138)\tPrec@1 87.500 (86.673)\n",
      "[350/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.5031 (0.6135)\tPrec@1 91.406 (86.683)\n",
      "[360/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6626 (0.6135)\tPrec@1 84.375 (86.678)\n",
      "[370/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5728 (0.6138)\tPrec@1 88.281 (86.658)\n",
      "[380/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6261 (0.6155)\tPrec@1 89.062 (86.612)\n",
      "[390/391]\tTime 0.116 (0.181)\tData 0.012 (0.020)\tLoss 0.5441 (0.6162)\tPrec@1 87.500 (86.586)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.3310 (1.3310)\tPrec@1 78.125 (78.125)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 1.0204 (1.2695)\tPrec@1 82.812 (78.196)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 1.0495 (1.2959)\tPrec@1 82.031 (78.125)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 1.0691 (1.2845)\tPrec@1 83.594 (78.175)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.6991 (1.2901)\tPrec@1 72.656 (78.049)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.3377 (1.2935)\tPrec@1 76.562 (78.202)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 1.4860 (1.2893)\tPrec@1 75.000 (78.266)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.0860 (1.2927)\tPrec@1 81.250 (78.334)\n",
      " * Prec@1 78.490\n",
      "\n",
      "===> epoch: 80/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.186 (0.186)\tData 0.021 (0.021)\tLoss 0.6629 (0.6629)\tPrec@1 82.812 (82.812)\n",
      "[10/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6828 (0.6011)\tPrec@1 87.500 (86.790)\n",
      "[20/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.7158 (0.6094)\tPrec@1 83.594 (86.979)\n",
      "[30/391]\tTime 0.187 (0.183)\tData 0.019 (0.020)\tLoss 0.5812 (0.6063)\tPrec@1 89.062 (86.946)\n",
      "[40/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6538 (0.5987)\tPrec@1 85.938 (87.100)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6159 (0.5954)\tPrec@1 89.062 (87.377)\n",
      "[60/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6900 (0.5984)\tPrec@1 83.594 (87.269)\n",
      "[70/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5439 (0.5987)\tPrec@1 89.844 (87.247)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5609 (0.5973)\tPrec@1 87.500 (87.259)\n",
      "[90/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5073 (0.5979)\tPrec@1 89.062 (87.260)\n",
      "[100/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5035 (0.5956)\tPrec@1 89.844 (87.322)\n",
      "[110/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6439 (0.5973)\tPrec@1 81.250 (87.296)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5255 (0.5969)\tPrec@1 90.625 (87.358)\n",
      "[130/391]\tTime 0.187 (0.182)\tData 0.022 (0.020)\tLoss 0.6462 (0.5985)\tPrec@1 86.719 (87.255)\n",
      "[140/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5312 (0.6014)\tPrec@1 88.281 (87.195)\n",
      "[150/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5954 (0.6045)\tPrec@1 88.281 (87.065)\n",
      "[160/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.7653 (0.6041)\tPrec@1 81.250 (87.122)\n",
      "[170/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5594 (0.6047)\tPrec@1 89.844 (87.075)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5890 (0.6054)\tPrec@1 86.719 (87.051)\n",
      "[190/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6208 (0.6074)\tPrec@1 86.719 (86.911)\n",
      "[200/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6461 (0.6096)\tPrec@1 87.500 (86.855)\n",
      "[210/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6624 (0.6096)\tPrec@1 85.938 (86.859)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7615 (0.6119)\tPrec@1 80.469 (86.765)\n",
      "[230/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6528 (0.6112)\tPrec@1 85.156 (86.810)\n",
      "[240/391]\tTime 0.198 (0.182)\tData 0.034 (0.020)\tLoss 0.5240 (0.6089)\tPrec@1 91.406 (86.871)\n",
      "[250/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.8421 (0.6092)\tPrec@1 78.125 (86.877)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6842 (0.6093)\tPrec@1 83.594 (86.892)\n",
      "[270/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6437 (0.6093)\tPrec@1 83.594 (86.872)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5465 (0.6089)\tPrec@1 91.406 (86.877)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[290/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6491 (0.6088)\tPrec@1 86.719 (86.899)\n",
      "[300/391]\tTime 0.182 (0.182)\tData 0.022 (0.020)\tLoss 0.5482 (0.6096)\tPrec@1 89.062 (86.838)\n",
      "[310/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6509 (0.6102)\tPrec@1 85.156 (86.787)\n",
      "[320/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5398 (0.6107)\tPrec@1 89.844 (86.767)\n",
      "[330/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6424 (0.6109)\tPrec@1 83.594 (86.749)\n",
      "[340/391]\tTime 0.183 (0.182)\tData 0.024 (0.020)\tLoss 0.5490 (0.6115)\tPrec@1 92.188 (86.726)\n",
      "[350/391]\tTime 0.180 (0.182)\tData 0.021 (0.020)\tLoss 0.6370 (0.6119)\tPrec@1 85.156 (86.717)\n",
      "[360/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6071 (0.6107)\tPrec@1 87.500 (86.777)\n",
      "[370/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6404 (0.6107)\tPrec@1 88.281 (86.805)\n",
      "[380/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6424 (0.6113)\tPrec@1 88.281 (86.780)\n",
      "[390/391]\tTime 0.111 (0.182)\tData 0.012 (0.020)\tLoss 0.6272 (0.6115)\tPrec@1 86.250 (86.774)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.5415 (0.5415)\tPrec@1 87.500 (87.500)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.9290 (0.8076)\tPrec@1 82.031 (83.381)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.8009 (0.8886)\tPrec@1 84.375 (82.292)\n",
      "Test: [30/79]\tTime 0.021 (0.020)\tLoss 0.6313 (0.9079)\tPrec@1 87.500 (82.334)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.0790 (0.9009)\tPrec@1 80.469 (82.355)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.0734 (0.9013)\tPrec@1 76.562 (82.261)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 1.0310 (0.9039)\tPrec@1 84.375 (82.505)\n",
      "Test: [70/79]\tTime 0.020 (0.019)\tLoss 0.8339 (0.8994)\tPrec@1 85.156 (82.504)\n",
      " * Prec@1 82.500\n",
      "\n",
      "===> epoch: 81/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.5939 (0.5939)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.179 (0.181)\tData 0.022 (0.020)\tLoss 0.4372 (0.5665)\tPrec@1 94.531 (88.139)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6176 (0.5631)\tPrec@1 88.281 (88.467)\n",
      "[30/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6936 (0.5717)\tPrec@1 83.594 (88.432)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.5271 (0.5702)\tPrec@1 89.844 (88.377)\n",
      "[50/391]\tTime 0.184 (0.181)\tData 0.021 (0.019)\tLoss 0.5415 (0.5855)\tPrec@1 90.625 (87.822)\n",
      "[60/391]\tTime 0.187 (0.181)\tData 0.020 (0.019)\tLoss 0.6698 (0.5929)\tPrec@1 85.938 (87.679)\n",
      "[70/391]\tTime 0.186 (0.181)\tData 0.020 (0.019)\tLoss 0.7699 (0.5989)\tPrec@1 79.688 (87.445)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.6783 (0.6032)\tPrec@1 83.594 (87.259)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5330 (0.6049)\tPrec@1 89.844 (87.122)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6010 (0.6036)\tPrec@1 85.156 (87.152)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5954 (0.6028)\tPrec@1 87.500 (87.162)\n",
      "[120/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5407 (0.6027)\tPrec@1 89.844 (87.216)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5953 (0.5990)\tPrec@1 86.719 (87.339)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6131 (0.5999)\tPrec@1 84.375 (87.267)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6248 (0.6017)\tPrec@1 85.938 (87.236)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.4501 (0.6010)\tPrec@1 92.188 (87.238)\n",
      "[170/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5737 (0.6026)\tPrec@1 89.844 (87.226)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7363 (0.6055)\tPrec@1 83.594 (87.124)\n",
      "[190/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6021 (0.6055)\tPrec@1 87.500 (87.120)\n",
      "[200/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5830 (0.6064)\tPrec@1 88.281 (87.088)\n",
      "[210/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6387 (0.6081)\tPrec@1 85.938 (86.996)\n",
      "[220/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6278 (0.6066)\tPrec@1 84.375 (87.055)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6291 (0.6093)\tPrec@1 85.938 (86.955)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5811 (0.6101)\tPrec@1 90.625 (86.929)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5779 (0.6111)\tPrec@1 87.500 (86.918)\n",
      "[260/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.5446 (0.6099)\tPrec@1 89.844 (86.958)\n",
      "[270/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6231 (0.6104)\tPrec@1 86.719 (86.926)\n",
      "[280/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6102 (0.6103)\tPrec@1 87.500 (86.927)\n",
      "[290/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7323 (0.6093)\tPrec@1 82.812 (86.955)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5780 (0.6094)\tPrec@1 88.281 (86.926)\n",
      "[310/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5490 (0.6094)\tPrec@1 88.281 (86.902)\n",
      "[320/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6034 (0.6090)\tPrec@1 88.281 (86.948)\n",
      "[330/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.6719 (0.6096)\tPrec@1 82.031 (86.912)\n",
      "[340/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6298 (0.6099)\tPrec@1 87.500 (86.900)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6567 (0.6097)\tPrec@1 87.500 (86.926)\n",
      "[360/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6382 (0.6101)\tPrec@1 85.938 (86.907)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5783 (0.6108)\tPrec@1 87.500 (86.908)\n",
      "[380/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6118 (0.6114)\tPrec@1 89.062 (86.897)\n",
      "[390/391]\tTime 0.113 (0.181)\tData 0.012 (0.020)\tLoss 0.6926 (0.6112)\tPrec@1 83.750 (86.912)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.0569 (1.0569)\tPrec@1 81.250 (81.250)\n",
      "Test: [10/79]\tTime 0.020 (0.021)\tLoss 1.2466 (1.2543)\tPrec@1 78.906 (79.616)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 1.2347 (1.2674)\tPrec@1 78.125 (79.241)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.9920 (1.2608)\tPrec@1 84.375 (79.032)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.6156 (1.2478)\tPrec@1 76.562 (79.211)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 1.4097 (1.2389)\tPrec@1 77.344 (79.458)\n",
      "Test: [60/79]\tTime 0.021 (0.020)\tLoss 1.6015 (1.2344)\tPrec@1 75.000 (79.431)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.3061 (1.2388)\tPrec@1 81.250 (79.236)\n",
      " * Prec@1 79.490\n",
      "\n",
      "===> epoch: 82/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.021 (0.021)\tLoss 0.6009 (0.6009)\tPrec@1 86.719 (86.719)\n",
      "[10/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.7045 (0.6356)\tPrec@1 85.156 (85.369)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5857 (0.6198)\tPrec@1 86.719 (86.644)\n",
      "[30/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5535 (0.6084)\tPrec@1 85.938 (86.820)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6215 (0.6083)\tPrec@1 86.719 (86.776)\n",
      "[50/391]\tTime 0.192 (0.182)\tData 0.020 (0.020)\tLoss 0.5858 (0.6135)\tPrec@1 86.719 (86.673)\n",
      "[60/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6643 (0.6130)\tPrec@1 83.594 (86.796)\n",
      "[70/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.8009 (0.6151)\tPrec@1 78.906 (86.708)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.7263 (0.6188)\tPrec@1 84.375 (86.651)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5661 (0.6164)\tPrec@1 89.844 (86.693)\n",
      "[100/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4476 (0.6099)\tPrec@1 92.969 (86.989)\n",
      "[110/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5378 (0.6090)\tPrec@1 89.844 (87.014)\n",
      "[120/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6751 (0.6082)\tPrec@1 85.156 (86.983)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6392 (0.6088)\tPrec@1 85.156 (86.898)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5385 (0.6082)\tPrec@1 88.281 (86.902)\n",
      "[150/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5514 (0.6103)\tPrec@1 89.844 (86.874)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6205 (0.6133)\tPrec@1 88.281 (86.728)\n",
      "[170/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5594 (0.6134)\tPrec@1 88.281 (86.737)\n",
      "[180/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5584 (0.6134)\tPrec@1 87.500 (86.689)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6434 (0.6144)\tPrec@1 86.719 (86.661)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5580 (0.6150)\tPrec@1 90.625 (86.653)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5803 (0.6123)\tPrec@1 88.281 (86.756)\n",
      "[220/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6561 (0.6126)\tPrec@1 85.938 (86.726)\n",
      "[230/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.7952 (0.6148)\tPrec@1 81.250 (86.688)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[240/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6840 (0.6164)\tPrec@1 81.250 (86.618)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6930 (0.6182)\tPrec@1 82.031 (86.529)\n",
      "[260/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6031 (0.6198)\tPrec@1 87.500 (86.488)\n",
      "[270/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.7718 (0.6205)\tPrec@1 81.250 (86.451)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6130 (0.6207)\tPrec@1 85.156 (86.446)\n",
      "[290/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6129 (0.6205)\tPrec@1 89.062 (86.464)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7221 (0.6204)\tPrec@1 79.688 (86.472)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5300 (0.6212)\tPrec@1 90.625 (86.455)\n",
      "[320/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5449 (0.6204)\tPrec@1 90.625 (86.497)\n",
      "[330/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6289 (0.6206)\tPrec@1 88.281 (86.504)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6533 (0.6215)\tPrec@1 82.812 (86.444)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6523 (0.6217)\tPrec@1 83.594 (86.436)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5706 (0.6218)\tPrec@1 88.281 (86.463)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5803 (0.6218)\tPrec@1 88.281 (86.489)\n",
      "[380/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6576 (0.6224)\tPrec@1 83.594 (86.454)\n",
      "[390/391]\tTime 0.119 (0.181)\tData 0.013 (0.020)\tLoss 0.7978 (0.6227)\tPrec@1 81.250 (86.468)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 0.7619 (0.7619)\tPrec@1 86.719 (86.719)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.7626 (0.9242)\tPrec@1 82.031 (83.594)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.7564 (0.9151)\tPrec@1 84.375 (83.110)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.4896 (0.8961)\tPrec@1 85.938 (82.913)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.1697 (0.9037)\tPrec@1 78.125 (83.041)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.7546 (0.8917)\tPrec@1 82.031 (83.027)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.8577 (0.8937)\tPrec@1 85.156 (83.005)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.7884 (0.9057)\tPrec@1 82.812 (82.779)\n",
      " * Prec@1 82.840\n",
      "\n",
      "===> epoch: 83/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.167 (0.167)\tData 0.021 (0.021)\tLoss 0.6704 (0.6704)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.6795 (0.6084)\tPrec@1 85.938 (87.216)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6660 (0.5991)\tPrec@1 86.719 (87.649)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6233 (0.6007)\tPrec@1 85.938 (87.626)\n",
      "[40/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6687 (0.6004)\tPrec@1 85.938 (87.462)\n",
      "[50/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6480 (0.5976)\tPrec@1 85.938 (87.500)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6255 (0.5974)\tPrec@1 85.938 (87.410)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6420 (0.6006)\tPrec@1 84.375 (87.236)\n",
      "[80/391]\tTime 0.185 (0.181)\tData 0.021 (0.020)\tLoss 0.6089 (0.5986)\tPrec@1 88.281 (87.346)\n",
      "[90/391]\tTime 0.186 (0.182)\tData 0.021 (0.020)\tLoss 0.5814 (0.5939)\tPrec@1 88.281 (87.517)\n",
      "[100/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5823 (0.5963)\tPrec@1 87.500 (87.562)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6134 (0.5996)\tPrec@1 89.844 (87.472)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5976 (0.6021)\tPrec@1 89.062 (87.377)\n",
      "[130/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6727 (0.6052)\tPrec@1 85.156 (87.202)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6565 (0.6059)\tPrec@1 84.375 (87.156)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6163 (0.6073)\tPrec@1 85.156 (87.076)\n",
      "[160/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6227 (0.6062)\tPrec@1 89.062 (87.088)\n",
      "[170/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5570 (0.6052)\tPrec@1 89.062 (87.171)\n",
      "[180/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6147 (0.6056)\tPrec@1 85.156 (87.150)\n",
      "[190/391]\tTime 0.185 (0.182)\tData 0.021 (0.020)\tLoss 0.5616 (0.6060)\tPrec@1 89.844 (87.152)\n",
      "[200/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.5920 (0.6065)\tPrec@1 88.281 (87.111)\n",
      "[210/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5831 (0.6081)\tPrec@1 85.156 (87.085)\n",
      "[220/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6886 (0.6080)\tPrec@1 82.812 (87.055)\n",
      "[230/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.6905 (0.6097)\tPrec@1 83.594 (87.040)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6120 (0.6101)\tPrec@1 88.281 (87.059)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.6421 (0.6102)\tPrec@1 85.938 (87.033)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6413 (0.6117)\tPrec@1 85.938 (86.976)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7296 (0.6127)\tPrec@1 80.469 (86.978)\n",
      "[280/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5388 (0.6129)\tPrec@1 91.406 (86.963)\n",
      "[290/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5958 (0.6153)\tPrec@1 89.062 (86.864)\n",
      "[300/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6525 (0.6152)\tPrec@1 86.719 (86.893)\n",
      "[310/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.6053 (0.6145)\tPrec@1 83.594 (86.915)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6501 (0.6145)\tPrec@1 87.500 (86.909)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7289 (0.6135)\tPrec@1 83.594 (86.945)\n",
      "[340/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7616 (0.6136)\tPrec@1 81.250 (86.936)\n",
      "[350/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5606 (0.6131)\tPrec@1 86.719 (86.950)\n",
      "[360/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6743 (0.6130)\tPrec@1 85.156 (86.937)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5244 (0.6137)\tPrec@1 89.062 (86.894)\n",
      "[380/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.7893 (0.6141)\tPrec@1 78.906 (86.887)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.012 (0.020)\tLoss 0.5240 (0.6142)\tPrec@1 92.500 (86.872)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.5563 (0.5563)\tPrec@1 84.375 (84.375)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.8121 (0.9688)\tPrec@1 85.156 (81.605)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.8323 (1.0357)\tPrec@1 85.156 (80.618)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 1.0358 (1.0602)\tPrec@1 82.031 (80.418)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.2240 (1.0731)\tPrec@1 77.344 (80.412)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 1.0110 (1.0377)\tPrec@1 82.031 (80.928)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 1.0335 (1.0317)\tPrec@1 82.031 (80.930)\n",
      "Test: [70/79]\tTime 0.019 (0.019)\tLoss 0.8838 (1.0322)\tPrec@1 80.469 (80.832)\n",
      " * Prec@1 80.780\n",
      "\n",
      "===> epoch: 84/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.186 (0.186)\tData 0.022 (0.022)\tLoss 0.5947 (0.5947)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.4897 (0.6042)\tPrec@1 92.969 (86.790)\n",
      "[20/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6218 (0.6164)\tPrec@1 87.500 (86.272)\n",
      "[30/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.6657 (0.6207)\tPrec@1 82.812 (86.240)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6094 (0.6113)\tPrec@1 86.719 (86.719)\n",
      "[50/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.6465 (0.6155)\tPrec@1 85.156 (86.504)\n",
      "[60/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5786 (0.6190)\tPrec@1 87.500 (86.335)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6214 (0.6144)\tPrec@1 82.812 (86.510)\n",
      "[80/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.5962 (0.6127)\tPrec@1 85.938 (86.709)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.7207 (0.6115)\tPrec@1 84.375 (86.745)\n",
      "[100/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5277 (0.6074)\tPrec@1 90.625 (86.819)\n",
      "[110/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6781 (0.6071)\tPrec@1 83.594 (86.761)\n",
      "[120/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.8241 (0.6099)\tPrec@1 78.906 (86.712)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5923 (0.6113)\tPrec@1 86.719 (86.701)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5783 (0.6135)\tPrec@1 89.062 (86.613)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6093 (0.6139)\tPrec@1 85.938 (86.641)\n",
      "[160/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6023 (0.6132)\tPrec@1 83.594 (86.670)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6210 (0.6131)\tPrec@1 85.938 (86.627)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7263 (0.6131)\tPrec@1 80.469 (86.667)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6248 (0.6126)\tPrec@1 82.812 (86.686)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5434 (0.6109)\tPrec@1 89.062 (86.812)\n",
      "[210/391]\tTime 0.182 (0.181)\tData 0.025 (0.020)\tLoss 0.7324 (0.6112)\tPrec@1 82.031 (86.852)\n",
      "[220/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7452 (0.6112)\tPrec@1 84.375 (86.864)\n",
      "[230/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6188 (0.6129)\tPrec@1 87.500 (86.807)\n",
      "[240/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5783 (0.6139)\tPrec@1 85.156 (86.764)\n",
      "[250/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6377 (0.6149)\tPrec@1 88.281 (86.753)\n",
      "[260/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.6243 (0.6164)\tPrec@1 85.156 (86.698)\n",
      "[270/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.7300 (0.6195)\tPrec@1 82.031 (86.549)\n",
      "[280/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6828 (0.6202)\tPrec@1 84.375 (86.569)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6954 (0.6201)\tPrec@1 83.594 (86.582)\n",
      "[300/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6499 (0.6201)\tPrec@1 85.938 (86.589)\n",
      "[310/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6154 (0.6202)\tPrec@1 91.406 (86.623)\n",
      "[320/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.5924 (0.6195)\tPrec@1 85.156 (86.648)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5715 (0.6182)\tPrec@1 87.500 (86.669)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6011 (0.6190)\tPrec@1 85.938 (86.618)\n",
      "[350/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5875 (0.6192)\tPrec@1 88.281 (86.587)\n",
      "[360/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5493 (0.6196)\tPrec@1 89.062 (86.567)\n",
      "[370/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.4959 (0.6196)\tPrec@1 92.188 (86.552)\n",
      "[380/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6162 (0.6189)\tPrec@1 87.500 (86.577)\n",
      "[390/391]\tTime 0.114 (0.181)\tData 0.012 (0.020)\tLoss 0.6138 (0.6192)\tPrec@1 85.000 (86.572)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8390 (0.8390)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.9658 (1.0250)\tPrec@1 78.125 (80.185)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 1.0637 (1.0667)\tPrec@1 78.125 (79.501)\n",
      "Test: [30/79]\tTime 0.020 (0.019)\tLoss 0.6916 (1.0118)\tPrec@1 87.500 (80.544)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.2986 (1.0109)\tPrec@1 77.344 (80.354)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 1.3471 (1.0074)\tPrec@1 74.219 (80.316)\n",
      "Test: [60/79]\tTime 0.021 (0.019)\tLoss 0.8313 (0.9950)\tPrec@1 82.031 (80.507)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 1.0374 (0.9836)\tPrec@1 80.469 (80.832)\n",
      " * Prec@1 80.990\n",
      "\n",
      "===> epoch: 85/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.184 (0.184)\tData 0.020 (0.020)\tLoss 0.6242 (0.6242)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.182 (0.184)\tData 0.020 (0.020)\tLoss 0.5692 (0.6066)\tPrec@1 87.500 (87.429)\n",
      "[20/391]\tTime 0.181 (0.184)\tData 0.020 (0.020)\tLoss 0.5356 (0.6259)\tPrec@1 91.406 (86.570)\n",
      "[30/391]\tTime 0.190 (0.183)\tData 0.019 (0.020)\tLoss 0.5630 (0.6170)\tPrec@1 86.719 (86.719)\n",
      "[40/391]\tTime 0.184 (0.183)\tData 0.019 (0.020)\tLoss 0.5621 (0.6096)\tPrec@1 92.188 (87.252)\n",
      "[50/391]\tTime 0.180 (0.183)\tData 0.019 (0.019)\tLoss 0.5599 (0.6043)\tPrec@1 89.062 (87.469)\n",
      "[60/391]\tTime 0.184 (0.183)\tData 0.019 (0.019)\tLoss 0.6460 (0.6057)\tPrec@1 84.375 (87.385)\n",
      "[70/391]\tTime 0.180 (0.183)\tData 0.019 (0.019)\tLoss 0.5530 (0.6044)\tPrec@1 88.281 (87.412)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5189 (0.6045)\tPrec@1 89.062 (87.288)\n",
      "[90/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6293 (0.6042)\tPrec@1 87.500 (87.363)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.5363 (0.6029)\tPrec@1 88.281 (87.353)\n",
      "[110/391]\tTime 0.190 (0.182)\tData 0.019 (0.019)\tLoss 0.5255 (0.6045)\tPrec@1 89.844 (87.268)\n",
      "[120/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.6280 (0.6043)\tPrec@1 89.062 (87.261)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.7227 (0.6069)\tPrec@1 82.812 (87.220)\n",
      "[140/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.5006 (0.6073)\tPrec@1 90.625 (87.190)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.5884 (0.6062)\tPrec@1 85.938 (87.236)\n",
      "[160/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.5663 (0.6114)\tPrec@1 90.625 (87.049)\n",
      "[170/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.6174 (0.6143)\tPrec@1 89.062 (86.892)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.7301 (0.6171)\tPrec@1 81.250 (86.723)\n",
      "[190/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6077 (0.6154)\tPrec@1 86.719 (86.751)\n",
      "[200/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.5444 (0.6172)\tPrec@1 87.500 (86.695)\n",
      "[210/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.5675 (0.6171)\tPrec@1 90.625 (86.719)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6155 (0.6180)\tPrec@1 88.281 (86.701)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5759 (0.6190)\tPrec@1 89.844 (86.722)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.021 (0.019)\tLoss 0.5697 (0.6181)\tPrec@1 89.062 (86.716)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6367 (0.6181)\tPrec@1 89.062 (86.731)\n",
      "[260/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6448 (0.6184)\tPrec@1 87.500 (86.722)\n",
      "[270/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.7900 (0.6191)\tPrec@1 80.469 (86.678)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7055 (0.6191)\tPrec@1 85.938 (86.671)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.021 (0.020)\tLoss 0.6967 (0.6197)\tPrec@1 84.375 (86.644)\n",
      "[300/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5865 (0.6199)\tPrec@1 87.500 (86.625)\n",
      "[310/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6713 (0.6201)\tPrec@1 85.156 (86.631)\n",
      "[320/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6817 (0.6197)\tPrec@1 83.594 (86.660)\n",
      "[330/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6020 (0.6200)\tPrec@1 89.062 (86.676)\n",
      "[340/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6058 (0.6193)\tPrec@1 85.938 (86.700)\n",
      "[350/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5643 (0.6181)\tPrec@1 85.938 (86.721)\n",
      "[360/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5743 (0.6179)\tPrec@1 87.500 (86.717)\n",
      "[370/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5990 (0.6177)\tPrec@1 87.500 (86.706)\n",
      "[380/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6565 (0.6178)\tPrec@1 83.594 (86.698)\n",
      "[390/391]\tTime 0.119 (0.182)\tData 0.013 (0.020)\tLoss 0.7775 (0.6185)\tPrec@1 82.500 (86.652)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.8796 (0.8796)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 1.0465 (1.1205)\tPrec@1 81.250 (79.759)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.8433 (1.1956)\tPrec@1 82.812 (79.315)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.9415 (1.1957)\tPrec@1 80.469 (78.730)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.2271 (1.1917)\tPrec@1 76.562 (78.506)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 1.2854 (1.1717)\tPrec@1 76.562 (78.646)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 1.4587 (1.1851)\tPrec@1 78.906 (78.727)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.1440 (1.1853)\tPrec@1 77.344 (78.598)\n",
      " * Prec@1 78.790\n",
      "\n",
      "===> epoch: 86/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.168 (0.168)\tData 0.022 (0.022)\tLoss 0.5914 (0.5914)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.6908 (0.6163)\tPrec@1 82.031 (86.222)\n",
      "[20/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.6054 (0.6104)\tPrec@1 86.719 (86.496)\n",
      "[30/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.6235 (0.6116)\tPrec@1 85.938 (86.492)\n",
      "[40/391]\tTime 0.181 (0.180)\tData 0.025 (0.020)\tLoss 0.5968 (0.6112)\tPrec@1 89.062 (86.319)\n",
      "[50/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6501 (0.6138)\tPrec@1 82.031 (86.351)\n",
      "[60/391]\tTime 0.186 (0.180)\tData 0.019 (0.020)\tLoss 0.6270 (0.6059)\tPrec@1 85.938 (86.629)\n",
      "[70/391]\tTime 0.186 (0.180)\tData 0.019 (0.020)\tLoss 0.6620 (0.6058)\tPrec@1 85.156 (86.609)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6502 (0.6059)\tPrec@1 85.156 (86.651)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5867 (0.6063)\tPrec@1 90.625 (86.719)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5702 (0.6046)\tPrec@1 87.500 (86.812)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5261 (0.6052)\tPrec@1 89.844 (86.824)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5922 (0.6048)\tPrec@1 83.594 (86.854)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6005 (0.6043)\tPrec@1 89.062 (86.939)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[140/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6304 (0.6066)\tPrec@1 85.938 (86.913)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5913 (0.6087)\tPrec@1 89.844 (86.827)\n",
      "[160/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5538 (0.6098)\tPrec@1 85.156 (86.738)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5509 (0.6097)\tPrec@1 90.625 (86.787)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6638 (0.6117)\tPrec@1 85.156 (86.762)\n",
      "[190/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.5457 (0.6123)\tPrec@1 88.281 (86.727)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5799 (0.6127)\tPrec@1 85.938 (86.664)\n",
      "[210/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5784 (0.6115)\tPrec@1 87.500 (86.745)\n",
      "[220/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.5590 (0.6121)\tPrec@1 90.625 (86.729)\n",
      "[230/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6478 (0.6127)\tPrec@1 85.938 (86.722)\n",
      "[240/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6423 (0.6123)\tPrec@1 85.156 (86.709)\n",
      "[250/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.5479 (0.6124)\tPrec@1 89.844 (86.713)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6242 (0.6121)\tPrec@1 89.062 (86.743)\n",
      "[270/391]\tTime 0.190 (0.181)\tData 0.026 (0.020)\tLoss 0.6259 (0.6125)\tPrec@1 85.156 (86.693)\n",
      "[280/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.6185 (0.6130)\tPrec@1 85.938 (86.669)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.6110 (0.6122)\tPrec@1 87.500 (86.708)\n",
      "[300/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6303 (0.6130)\tPrec@1 86.719 (86.698)\n",
      "[310/391]\tTime 0.176 (0.181)\tData 0.019 (0.020)\tLoss 0.4416 (0.6129)\tPrec@1 93.750 (86.699)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7746 (0.6141)\tPrec@1 81.250 (86.682)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5838 (0.6152)\tPrec@1 89.062 (86.641)\n",
      "[340/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.7920 (0.6164)\tPrec@1 79.688 (86.611)\n",
      "[350/391]\tTime 0.188 (0.181)\tData 0.020 (0.020)\tLoss 0.6926 (0.6170)\tPrec@1 82.031 (86.556)\n",
      "[360/391]\tTime 0.188 (0.181)\tData 0.020 (0.020)\tLoss 0.6926 (0.6184)\tPrec@1 85.156 (86.535)\n",
      "[370/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5194 (0.6182)\tPrec@1 90.625 (86.567)\n",
      "[380/391]\tTime 0.185 (0.181)\tData 0.022 (0.020)\tLoss 0.6137 (0.6174)\tPrec@1 85.938 (86.590)\n",
      "[390/391]\tTime 0.120 (0.181)\tData 0.012 (0.020)\tLoss 0.6114 (0.6174)\tPrec@1 88.750 (86.588)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 1.0145 (1.0145)\tPrec@1 79.688 (79.688)\n",
      "Test: [10/79]\tTime 0.019 (0.020)\tLoss 1.1255 (1.1393)\tPrec@1 78.906 (79.901)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 1.2215 (1.1746)\tPrec@1 78.125 (79.650)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.7650 (1.1726)\tPrec@1 82.031 (79.435)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.5687 (1.1602)\tPrec@1 74.219 (79.440)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.4965 (1.1630)\tPrec@1 76.562 (79.366)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.7596 (1.1449)\tPrec@1 85.938 (79.572)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.8318 (1.1424)\tPrec@1 81.250 (79.555)\n",
      " * Prec@1 79.620\n",
      "\n",
      "===> epoch: 87/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.166 (0.166)\tData 0.021 (0.021)\tLoss 0.6090 (0.6090)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.183 (0.180)\tData 0.020 (0.020)\tLoss 0.6001 (0.5955)\tPrec@1 85.938 (87.145)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5530 (0.5916)\tPrec@1 87.500 (87.612)\n",
      "[30/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6395 (0.5980)\tPrec@1 87.500 (87.374)\n",
      "[40/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5873 (0.5982)\tPrec@1 88.281 (87.214)\n",
      "[50/391]\tTime 0.182 (0.183)\tData 0.020 (0.020)\tLoss 0.6268 (0.5962)\tPrec@1 85.938 (87.347)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6135 (0.5962)\tPrec@1 87.500 (87.269)\n",
      "[70/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6864 (0.5992)\tPrec@1 83.594 (87.137)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5981 (0.5958)\tPrec@1 85.938 (87.201)\n",
      "[90/391]\tTime 0.188 (0.182)\tData 0.019 (0.020)\tLoss 0.6409 (0.5999)\tPrec@1 84.375 (87.002)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5794 (0.6000)\tPrec@1 89.062 (87.090)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6884 (0.6016)\tPrec@1 83.594 (86.993)\n",
      "[120/391]\tTime 0.189 (0.182)\tData 0.021 (0.020)\tLoss 0.5961 (0.6031)\tPrec@1 83.594 (86.919)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6343 (0.6052)\tPrec@1 86.719 (86.856)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6971 (0.6080)\tPrec@1 84.375 (86.780)\n",
      "[150/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.7064 (0.6073)\tPrec@1 79.688 (86.802)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5209 (0.6074)\tPrec@1 89.844 (86.796)\n",
      "[170/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5511 (0.6069)\tPrec@1 91.406 (86.819)\n",
      "[180/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6391 (0.6079)\tPrec@1 86.719 (86.788)\n",
      "[190/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7183 (0.6088)\tPrec@1 82.812 (86.772)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6236 (0.6099)\tPrec@1 86.719 (86.765)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6110 (0.6119)\tPrec@1 89.062 (86.741)\n",
      "[220/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6091 (0.6124)\tPrec@1 87.500 (86.715)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.7211 (0.6126)\tPrec@1 81.250 (86.715)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.7941 (0.6149)\tPrec@1 82.812 (86.631)\n",
      "[250/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.7133 (0.6156)\tPrec@1 82.031 (86.616)\n",
      "[260/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6901 (0.6160)\tPrec@1 81.250 (86.599)\n",
      "[270/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.5152 (0.6142)\tPrec@1 90.625 (86.670)\n",
      "[280/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.8107 (0.6157)\tPrec@1 80.469 (86.663)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5993 (0.6166)\tPrec@1 87.500 (86.625)\n",
      "[300/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5471 (0.6166)\tPrec@1 89.844 (86.618)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6644 (0.6164)\tPrec@1 82.812 (86.628)\n",
      "[320/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5837 (0.6169)\tPrec@1 89.062 (86.648)\n",
      "[330/391]\tTime 0.185 (0.182)\tData 0.021 (0.020)\tLoss 0.6116 (0.6160)\tPrec@1 88.281 (86.712)\n",
      "[340/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4407 (0.6158)\tPrec@1 91.406 (86.716)\n",
      "[350/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6403 (0.6165)\tPrec@1 90.625 (86.688)\n",
      "[360/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6256 (0.6173)\tPrec@1 87.500 (86.643)\n",
      "[370/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6530 (0.6168)\tPrec@1 84.375 (86.647)\n",
      "[380/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5664 (0.6166)\tPrec@1 89.844 (86.670)\n",
      "[390/391]\tTime 0.117 (0.182)\tData 0.016 (0.020)\tLoss 0.5962 (0.6176)\tPrec@1 87.500 (86.650)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7469 (0.7469)\tPrec@1 85.156 (85.156)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 1.2418 (1.0305)\tPrec@1 78.906 (79.688)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.9437 (1.0948)\tPrec@1 83.594 (80.022)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.8750 (1.1088)\tPrec@1 82.031 (79.864)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.5026 (1.1127)\tPrec@1 74.219 (79.973)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.1179 (1.1036)\tPrec@1 76.562 (80.009)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 1.0728 (1.0929)\tPrec@1 77.344 (80.200)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 1.2660 (1.0932)\tPrec@1 81.250 (80.194)\n",
      " * Prec@1 80.140\n",
      "\n",
      "===> epoch: 88/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.189 (0.189)\tData 0.021 (0.021)\tLoss 0.6007 (0.6007)\tPrec@1 89.844 (89.844)\n",
      "[10/391]\tTime 0.179 (0.183)\tData 0.019 (0.020)\tLoss 0.5609 (0.5728)\tPrec@1 88.281 (88.423)\n",
      "[20/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5462 (0.5834)\tPrec@1 89.844 (87.463)\n",
      "[30/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7035 (0.5887)\tPrec@1 83.594 (87.626)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.6888 (0.5945)\tPrec@1 85.938 (87.309)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5750 (0.5950)\tPrec@1 86.719 (87.163)\n",
      "[60/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5396 (0.5936)\tPrec@1 88.281 (87.321)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6155 (0.5948)\tPrec@1 88.281 (87.280)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.7439 (0.5984)\tPrec@1 80.469 (87.133)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[90/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4224 (0.6008)\tPrec@1 93.750 (87.096)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5021 (0.6013)\tPrec@1 91.406 (87.067)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5537 (0.5977)\tPrec@1 87.500 (87.204)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5458 (0.5968)\tPrec@1 89.844 (87.235)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6055 (0.5958)\tPrec@1 86.719 (87.303)\n",
      "[140/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5774 (0.5964)\tPrec@1 87.500 (87.262)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6584 (0.5965)\tPrec@1 83.594 (87.303)\n",
      "[160/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6990 (0.5955)\tPrec@1 82.031 (87.354)\n",
      "[170/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6224 (0.5951)\tPrec@1 89.844 (87.431)\n",
      "[180/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.5737 (0.5954)\tPrec@1 90.625 (87.431)\n",
      "[190/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5714 (0.5969)\tPrec@1 88.281 (87.394)\n",
      "[200/391]\tTime 0.192 (0.182)\tData 0.020 (0.020)\tLoss 0.5400 (0.5967)\tPrec@1 89.844 (87.399)\n",
      "[210/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6843 (0.5971)\tPrec@1 85.938 (87.407)\n",
      "[220/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6646 (0.5993)\tPrec@1 83.594 (87.316)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5863 (0.6007)\tPrec@1 86.719 (87.233)\n",
      "[240/391]\tTime 0.177 (0.182)\tData 0.019 (0.020)\tLoss 0.6446 (0.6013)\tPrec@1 85.938 (87.211)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7043 (0.6020)\tPrec@1 82.031 (87.186)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.6322 (0.6032)\tPrec@1 85.938 (87.141)\n",
      "[270/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6367 (0.6037)\tPrec@1 86.719 (87.125)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6502 (0.6043)\tPrec@1 87.500 (87.080)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6386 (0.6041)\tPrec@1 90.625 (87.100)\n",
      "[300/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6127 (0.6047)\tPrec@1 89.062 (87.111)\n",
      "[310/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6266 (0.6058)\tPrec@1 87.500 (87.065)\n",
      "[320/391]\tTime 0.190 (0.182)\tData 0.019 (0.020)\tLoss 0.6123 (0.6071)\tPrec@1 89.062 (87.001)\n",
      "[330/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.6274 (0.6080)\tPrec@1 83.594 (86.962)\n",
      "[340/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7041 (0.6082)\tPrec@1 86.719 (86.964)\n",
      "[350/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.5878 (0.6094)\tPrec@1 86.719 (86.928)\n",
      "[360/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5373 (0.6096)\tPrec@1 90.625 (86.920)\n",
      "[370/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.7448 (0.6105)\tPrec@1 78.125 (86.870)\n",
      "[380/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6399 (0.6104)\tPrec@1 82.812 (86.860)\n",
      "[390/391]\tTime 0.113 (0.182)\tData 0.012 (0.020)\tLoss 0.6612 (0.6116)\tPrec@1 87.500 (86.834)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.1483 (1.1483)\tPrec@1 78.906 (78.906)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.9410 (1.0133)\tPrec@1 81.250 (79.901)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.6839 (1.0214)\tPrec@1 81.250 (79.911)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.5228 (1.0139)\tPrec@1 85.156 (79.713)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.3269 (1.0225)\tPrec@1 74.219 (79.783)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.0638 (1.0031)\tPrec@1 75.000 (79.994)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 1.2362 (1.0028)\tPrec@1 80.469 (80.213)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.0337 (0.9938)\tPrec@1 80.469 (80.271)\n",
      " * Prec@1 80.390\n",
      "\n",
      "===> epoch: 89/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.7722 (0.7722)\tPrec@1 79.688 (79.688)\n",
      "[10/391]\tTime 0.179 (0.183)\tData 0.019 (0.020)\tLoss 0.6489 (0.6310)\tPrec@1 87.500 (86.364)\n",
      "[20/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5872 (0.6135)\tPrec@1 87.500 (87.054)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.5188 (0.6001)\tPrec@1 87.500 (87.172)\n",
      "[40/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5836 (0.5957)\tPrec@1 85.938 (87.424)\n",
      "[50/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5608 (0.5944)\tPrec@1 86.719 (87.546)\n",
      "[60/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5552 (0.5907)\tPrec@1 91.406 (87.654)\n",
      "[70/391]\tTime 0.189 (0.182)\tData 0.020 (0.019)\tLoss 0.6133 (0.5955)\tPrec@1 87.500 (87.577)\n",
      "[80/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.5838 (0.5964)\tPrec@1 89.062 (87.596)\n",
      "[90/391]\tTime 0.188 (0.183)\tData 0.019 (0.019)\tLoss 0.6125 (0.5999)\tPrec@1 89.062 (87.474)\n",
      "[100/391]\tTime 0.178 (0.183)\tData 0.019 (0.019)\tLoss 0.6251 (0.6018)\tPrec@1 86.719 (87.423)\n",
      "[110/391]\tTime 0.181 (0.183)\tData 0.019 (0.019)\tLoss 0.5979 (0.6091)\tPrec@1 89.844 (87.113)\n",
      "[120/391]\tTime 0.180 (0.183)\tData 0.019 (0.019)\tLoss 0.5974 (0.6089)\tPrec@1 89.844 (87.080)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.020 (0.019)\tLoss 0.6873 (0.6113)\tPrec@1 87.500 (87.077)\n",
      "[140/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6220 (0.6101)\tPrec@1 88.281 (87.123)\n",
      "[150/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6376 (0.6110)\tPrec@1 82.812 (87.065)\n",
      "[160/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.7596 (0.6132)\tPrec@1 79.688 (86.893)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6501 (0.6151)\tPrec@1 85.938 (86.751)\n",
      "[180/391]\tTime 0.181 (0.182)\tData 0.020 (0.019)\tLoss 0.5892 (0.6138)\tPrec@1 85.938 (86.805)\n",
      "[190/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6012 (0.6147)\tPrec@1 89.844 (86.821)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.6664 (0.6146)\tPrec@1 82.812 (86.851)\n",
      "[210/391]\tTime 0.178 (0.182)\tData 0.019 (0.019)\tLoss 0.5962 (0.6124)\tPrec@1 87.500 (86.882)\n",
      "[220/391]\tTime 0.179 (0.182)\tData 0.020 (0.019)\tLoss 0.6449 (0.6134)\tPrec@1 83.594 (86.782)\n",
      "[230/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6388 (0.6141)\tPrec@1 86.719 (86.739)\n",
      "[240/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.6468 (0.6150)\tPrec@1 89.844 (86.754)\n",
      "[250/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6037 (0.6137)\tPrec@1 85.938 (86.797)\n",
      "[260/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6352 (0.6133)\tPrec@1 86.719 (86.838)\n",
      "[270/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5494 (0.6117)\tPrec@1 89.062 (86.886)\n",
      "[280/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6439 (0.6117)\tPrec@1 85.156 (86.863)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.7387 (0.6129)\tPrec@1 83.594 (86.791)\n",
      "[300/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.7292 (0.6155)\tPrec@1 81.250 (86.680)\n",
      "[310/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6205 (0.6168)\tPrec@1 89.844 (86.643)\n",
      "[320/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6562 (0.6177)\tPrec@1 85.156 (86.607)\n",
      "[330/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5326 (0.6176)\tPrec@1 89.844 (86.631)\n",
      "[340/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.7014 (0.6177)\tPrec@1 85.156 (86.650)\n",
      "[350/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.7522 (0.6187)\tPrec@1 80.469 (86.614)\n",
      "[360/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.5896 (0.6196)\tPrec@1 89.062 (86.580)\n",
      "[370/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5341 (0.6191)\tPrec@1 87.500 (86.618)\n",
      "[380/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5024 (0.6182)\tPrec@1 92.188 (86.661)\n",
      "[390/391]\tTime 0.112 (0.181)\tData 0.013 (0.020)\tLoss 0.6338 (0.6181)\tPrec@1 85.000 (86.682)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7169 (0.7169)\tPrec@1 85.938 (85.938)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.6603 (0.9183)\tPrec@1 89.062 (83.097)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.7789 (0.9619)\tPrec@1 83.594 (82.440)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.6498 (0.9855)\tPrec@1 85.938 (81.880)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.8924 (0.9733)\tPrec@1 82.812 (82.165)\n",
      "Test: [50/79]\tTime 0.021 (0.020)\tLoss 0.9840 (0.9615)\tPrec@1 81.250 (82.292)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.8703 (0.9555)\tPrec@1 85.156 (82.377)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.0400 (0.9623)\tPrec@1 81.250 (82.229)\n",
      " * Prec@1 82.240\n",
      "\n",
      "===> epoch: 90/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.171 (0.171)\tData 0.020 (0.020)\tLoss 0.5435 (0.5435)\tPrec@1 89.062 (89.062)\n",
      "[10/391]\tTime 0.178 (0.178)\tData 0.020 (0.020)\tLoss 0.6231 (0.5997)\tPrec@1 84.375 (86.861)\n",
      "[20/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.5866 (0.5939)\tPrec@1 88.281 (87.277)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5754 (0.6032)\tPrec@1 88.281 (87.097)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.187 (0.181)\tData 0.023 (0.020)\tLoss 0.6173 (0.6042)\tPrec@1 85.938 (87.005)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.5774 (0.5992)\tPrec@1 86.719 (87.117)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5873 (0.6006)\tPrec@1 88.281 (87.103)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6868 (0.6000)\tPrec@1 81.250 (87.181)\n",
      "[80/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6723 (0.6021)\tPrec@1 85.938 (87.172)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.6189 (0.6036)\tPrec@1 85.938 (87.131)\n",
      "[100/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6461 (0.6077)\tPrec@1 83.594 (86.989)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5813 (0.6091)\tPrec@1 87.500 (86.888)\n",
      "[120/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5890 (0.6071)\tPrec@1 89.844 (86.951)\n",
      "[130/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.5903 (0.6062)\tPrec@1 88.281 (86.975)\n",
      "[140/391]\tTime 0.189 (0.182)\tData 0.019 (0.020)\tLoss 0.5937 (0.6037)\tPrec@1 87.500 (87.101)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6590 (0.6048)\tPrec@1 85.938 (87.091)\n",
      "[160/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6415 (0.6065)\tPrec@1 84.375 (87.020)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5973 (0.6079)\tPrec@1 88.281 (87.007)\n",
      "[180/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7769 (0.6092)\tPrec@1 80.469 (87.008)\n",
      "[190/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5166 (0.6101)\tPrec@1 90.625 (86.948)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6259 (0.6099)\tPrec@1 88.281 (86.968)\n",
      "[210/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.5982 (0.6107)\tPrec@1 85.156 (86.919)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.6394 (0.6136)\tPrec@1 86.719 (86.786)\n",
      "[230/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5964 (0.6148)\tPrec@1 85.156 (86.746)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5552 (0.6149)\tPrec@1 90.625 (86.712)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6160 (0.6156)\tPrec@1 88.281 (86.706)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5720 (0.6152)\tPrec@1 90.625 (86.749)\n",
      "[270/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.6843 (0.6153)\tPrec@1 83.594 (86.736)\n",
      "[280/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5664 (0.6151)\tPrec@1 89.844 (86.713)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5683 (0.6151)\tPrec@1 89.062 (86.689)\n",
      "[300/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6369 (0.6151)\tPrec@1 85.156 (86.693)\n",
      "[310/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.7037 (0.6155)\tPrec@1 79.688 (86.646)\n",
      "[320/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.5846 (0.6155)\tPrec@1 87.500 (86.673)\n",
      "[330/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5448 (0.6151)\tPrec@1 89.062 (86.707)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6593 (0.6160)\tPrec@1 86.719 (86.655)\n",
      "[350/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.5836 (0.6166)\tPrec@1 85.938 (86.632)\n",
      "[360/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6668 (0.6170)\tPrec@1 85.156 (86.606)\n",
      "[370/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5708 (0.6173)\tPrec@1 89.844 (86.620)\n",
      "[380/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6154 (0.6171)\tPrec@1 87.500 (86.633)\n",
      "[390/391]\tTime 0.119 (0.182)\tData 0.012 (0.020)\tLoss 0.5752 (0.6177)\tPrec@1 92.500 (86.614)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 1.0285 (1.0285)\tPrec@1 82.812 (82.812)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.9067 (1.0805)\tPrec@1 85.156 (80.966)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 1.0274 (1.1948)\tPrec@1 78.906 (79.613)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 1.2408 (1.2031)\tPrec@1 82.031 (79.662)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.0979 (1.2117)\tPrec@1 80.469 (79.440)\n",
      "Test: [50/79]\tTime 0.020 (0.019)\tLoss 1.2410 (1.2029)\tPrec@1 79.688 (79.320)\n",
      "Test: [60/79]\tTime 0.020 (0.019)\tLoss 1.3238 (1.1998)\tPrec@1 82.031 (79.534)\n",
      "Test: [70/79]\tTime 0.020 (0.019)\tLoss 1.1427 (1.2059)\tPrec@1 82.812 (79.511)\n",
      " * Prec@1 79.620\n",
      "\n",
      "===> epoch: 91/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.175 (0.175)\tData 0.020 (0.020)\tLoss 0.6134 (0.6134)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.185 (0.184)\tData 0.020 (0.020)\tLoss 0.4918 (0.5798)\tPrec@1 92.969 (88.068)\n",
      "[20/391]\tTime 0.180 (0.182)\tData 0.021 (0.020)\tLoss 0.5128 (0.5759)\tPrec@1 89.844 (88.021)\n",
      "[30/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6356 (0.5795)\tPrec@1 84.375 (88.180)\n",
      "[40/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6362 (0.5864)\tPrec@1 88.281 (88.014)\n",
      "[50/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6086 (0.5904)\tPrec@1 86.719 (87.745)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6884 (0.5961)\tPrec@1 85.156 (87.577)\n",
      "[70/391]\tTime 0.179 (0.182)\tData 0.020 (0.019)\tLoss 0.6026 (0.6008)\tPrec@1 89.062 (87.434)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5848 (0.6009)\tPrec@1 89.062 (87.452)\n",
      "[90/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6003 (0.6035)\tPrec@1 89.844 (87.354)\n",
      "[100/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.5644 (0.6010)\tPrec@1 87.500 (87.477)\n",
      "[110/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5386 (0.6003)\tPrec@1 91.406 (87.521)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5839 (0.6018)\tPrec@1 85.938 (87.371)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6165 (0.6053)\tPrec@1 85.938 (87.184)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6702 (0.6081)\tPrec@1 86.719 (87.118)\n",
      "[150/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6297 (0.6076)\tPrec@1 87.500 (87.112)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4443 (0.6046)\tPrec@1 92.188 (87.219)\n",
      "[170/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.5548 (0.6054)\tPrec@1 88.281 (87.125)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.021 (0.020)\tLoss 0.6779 (0.6069)\tPrec@1 84.375 (87.081)\n",
      "[190/391]\tTime 0.185 (0.182)\tData 0.024 (0.020)\tLoss 0.7179 (0.6092)\tPrec@1 82.812 (86.976)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6407 (0.6110)\tPrec@1 85.938 (86.960)\n",
      "[210/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6144 (0.6127)\tPrec@1 86.719 (86.922)\n",
      "[220/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5113 (0.6129)\tPrec@1 91.406 (86.917)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5770 (0.6136)\tPrec@1 86.719 (86.840)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6045 (0.6147)\tPrec@1 85.938 (86.767)\n",
      "[250/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6394 (0.6153)\tPrec@1 88.281 (86.747)\n",
      "[260/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.5925 (0.6143)\tPrec@1 85.156 (86.749)\n",
      "[270/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5175 (0.6135)\tPrec@1 92.188 (86.753)\n",
      "[280/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5857 (0.6141)\tPrec@1 87.500 (86.722)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.022 (0.020)\tLoss 0.5460 (0.6160)\tPrec@1 91.406 (86.668)\n",
      "[300/391]\tTime 0.184 (0.182)\tData 0.022 (0.020)\tLoss 0.7132 (0.6169)\tPrec@1 81.250 (86.646)\n",
      "[310/391]\tTime 0.191 (0.182)\tData 0.028 (0.020)\tLoss 0.6439 (0.6160)\tPrec@1 83.594 (86.671)\n",
      "[320/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5982 (0.6158)\tPrec@1 85.156 (86.677)\n",
      "[330/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6865 (0.6162)\tPrec@1 82.812 (86.643)\n",
      "[340/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6446 (0.6173)\tPrec@1 82.031 (86.609)\n",
      "[350/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7315 (0.6179)\tPrec@1 82.031 (86.592)\n",
      "[360/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6017 (0.6194)\tPrec@1 89.062 (86.522)\n",
      "[370/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6297 (0.6197)\tPrec@1 85.938 (86.529)\n",
      "[380/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6840 (0.6196)\tPrec@1 79.688 (86.514)\n",
      "[390/391]\tTime 0.094 (0.181)\tData 0.013 (0.020)\tLoss 0.7164 (0.6192)\tPrec@1 80.000 (86.540)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8158 (0.8158)\tPrec@1 83.594 (83.594)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.9007 (0.9157)\tPrec@1 80.469 (82.457)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.4977 (0.9325)\tPrec@1 88.281 (82.738)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.8591 (0.9372)\tPrec@1 82.031 (82.586)\n",
      "Test: [40/79]\tTime 0.021 (0.019)\tLoss 1.5453 (0.9261)\tPrec@1 71.094 (82.412)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 1.1883 (0.9329)\tPrec@1 77.344 (82.261)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 0.7592 (0.9345)\tPrec@1 85.156 (82.134)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [70/79]\tTime 0.019 (0.019)\tLoss 0.8844 (0.9287)\tPrec@1 82.031 (82.339)\n",
      " * Prec@1 82.510\n",
      "\n",
      "===> epoch: 92/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.165 (0.165)\tData 0.020 (0.020)\tLoss 0.5324 (0.5324)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.020 (0.020)\tLoss 0.5246 (0.5992)\tPrec@1 89.844 (87.784)\n",
      "[20/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.5596 (0.5833)\tPrec@1 88.281 (87.649)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.5750 (0.5861)\tPrec@1 90.625 (87.802)\n",
      "[40/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.5580 (0.5888)\tPrec@1 88.281 (87.862)\n",
      "[50/391]\tTime 0.185 (0.180)\tData 0.019 (0.019)\tLoss 0.5818 (0.5847)\tPrec@1 86.719 (88.051)\n",
      "[60/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.5595 (0.5890)\tPrec@1 91.406 (87.795)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6751 (0.5985)\tPrec@1 86.719 (87.467)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5847 (0.5999)\tPrec@1 89.062 (87.384)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.7060 (0.6052)\tPrec@1 82.031 (87.122)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6478 (0.6054)\tPrec@1 86.719 (87.082)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5203 (0.6018)\tPrec@1 89.844 (87.204)\n",
      "[120/391]\tTime 0.189 (0.181)\tData 0.019 (0.020)\tLoss 0.5414 (0.5999)\tPrec@1 88.281 (87.216)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5801 (0.5996)\tPrec@1 88.281 (87.285)\n",
      "[140/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6227 (0.5992)\tPrec@1 86.719 (87.262)\n",
      "[150/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6464 (0.6004)\tPrec@1 83.594 (87.190)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7044 (0.6037)\tPrec@1 84.375 (87.122)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5698 (0.6054)\tPrec@1 88.281 (87.066)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6910 (0.6067)\tPrec@1 84.375 (87.025)\n",
      "[190/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.7683 (0.6081)\tPrec@1 80.469 (86.976)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5923 (0.6083)\tPrec@1 88.281 (86.968)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5883 (0.6082)\tPrec@1 86.719 (86.993)\n",
      "[220/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5945 (0.6082)\tPrec@1 87.500 (86.994)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.7185 (0.6079)\tPrec@1 85.156 (87.040)\n",
      "[240/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5839 (0.6080)\tPrec@1 88.281 (87.014)\n",
      "[250/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6104 (0.6090)\tPrec@1 88.281 (86.968)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5975 (0.6091)\tPrec@1 85.938 (86.940)\n",
      "[270/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7795 (0.6107)\tPrec@1 79.688 (86.874)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6052 (0.6119)\tPrec@1 84.375 (86.811)\n",
      "[290/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.8044 (0.6136)\tPrec@1 79.688 (86.743)\n",
      "[300/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6265 (0.6155)\tPrec@1 87.500 (86.708)\n",
      "[310/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5813 (0.6153)\tPrec@1 89.062 (86.699)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5359 (0.6153)\tPrec@1 89.062 (86.709)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5354 (0.6153)\tPrec@1 86.719 (86.683)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5465 (0.6143)\tPrec@1 91.406 (86.746)\n",
      "[350/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.6635 (0.6150)\tPrec@1 87.500 (86.719)\n",
      "[360/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.6300 (0.6134)\tPrec@1 86.719 (86.799)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.6070 (0.6134)\tPrec@1 86.719 (86.790)\n",
      "[380/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5060 (0.6141)\tPrec@1 91.406 (86.768)\n",
      "[390/391]\tTime 0.094 (0.181)\tData 0.013 (0.020)\tLoss 0.5745 (0.6143)\tPrec@1 87.500 (86.760)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4102 (0.4102)\tPrec@1 89.844 (89.844)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.7301 (0.7382)\tPrec@1 84.375 (84.588)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.4998 (0.7697)\tPrec@1 88.281 (84.375)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.7206 (0.7965)\tPrec@1 85.938 (84.249)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.0894 (0.8098)\tPrec@1 80.469 (84.356)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.5783 (0.7895)\tPrec@1 84.375 (84.651)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.8644 (0.7902)\tPrec@1 82.812 (84.490)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.6288 (0.7878)\tPrec@1 87.500 (84.518)\n",
      " * Prec@1 84.690\n",
      "\n",
      "===> epoch: 93/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.162 (0.162)\tData 0.021 (0.021)\tLoss 0.5604 (0.5604)\tPrec@1 88.281 (88.281)\n",
      "[10/391]\tTime 0.182 (0.180)\tData 0.019 (0.021)\tLoss 0.6563 (0.5803)\tPrec@1 83.594 (88.068)\n",
      "[20/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.5467 (0.6028)\tPrec@1 86.719 (87.016)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5023 (0.5921)\tPrec@1 89.844 (87.399)\n",
      "[40/391]\tTime 0.189 (0.181)\tData 0.020 (0.020)\tLoss 0.6131 (0.5932)\tPrec@1 88.281 (87.424)\n",
      "[50/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5336 (0.5898)\tPrec@1 90.625 (87.607)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.7124 (0.5947)\tPrec@1 86.719 (87.487)\n",
      "[70/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5774 (0.5991)\tPrec@1 89.844 (87.258)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5878 (0.5988)\tPrec@1 87.500 (87.269)\n",
      "[90/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6757 (0.5979)\tPrec@1 85.156 (87.285)\n",
      "[100/391]\tTime 0.188 (0.182)\tData 0.024 (0.020)\tLoss 0.5733 (0.5959)\tPrec@1 90.625 (87.368)\n",
      "[110/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.4778 (0.5963)\tPrec@1 90.625 (87.345)\n",
      "[120/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.4746 (0.5956)\tPrec@1 96.094 (87.435)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.7328 (0.5959)\tPrec@1 79.688 (87.452)\n",
      "[140/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5773 (0.6007)\tPrec@1 91.406 (87.262)\n",
      "[150/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6257 (0.6032)\tPrec@1 85.156 (87.133)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6328 (0.6041)\tPrec@1 88.281 (87.107)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5369 (0.6047)\tPrec@1 90.625 (87.121)\n",
      "[180/391]\tTime 0.188 (0.182)\tData 0.019 (0.020)\tLoss 0.5995 (0.6064)\tPrec@1 89.062 (87.017)\n",
      "[190/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.6006 (0.6058)\tPrec@1 84.375 (87.034)\n",
      "[200/391]\tTime 0.189 (0.182)\tData 0.020 (0.020)\tLoss 0.6401 (0.6071)\tPrec@1 83.594 (86.971)\n",
      "[210/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6054 (0.6079)\tPrec@1 86.719 (86.937)\n",
      "[220/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.7177 (0.6096)\tPrec@1 83.594 (86.860)\n",
      "[230/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.7149 (0.6107)\tPrec@1 85.938 (86.824)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5078 (0.6115)\tPrec@1 92.188 (86.797)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7569 (0.6129)\tPrec@1 81.250 (86.750)\n",
      "[260/391]\tTime 0.188 (0.182)\tData 0.021 (0.020)\tLoss 0.6086 (0.6133)\tPrec@1 88.281 (86.761)\n",
      "[270/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6056 (0.6129)\tPrec@1 89.844 (86.788)\n",
      "[280/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5577 (0.6137)\tPrec@1 91.406 (86.755)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6852 (0.6150)\tPrec@1 88.281 (86.724)\n",
      "[300/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6280 (0.6141)\tPrec@1 88.281 (86.784)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5901 (0.6135)\tPrec@1 89.062 (86.814)\n",
      "[320/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7103 (0.6144)\tPrec@1 82.812 (86.765)\n",
      "[330/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.6333 (0.6152)\tPrec@1 83.594 (86.738)\n",
      "[340/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6329 (0.6147)\tPrec@1 85.938 (86.751)\n",
      "[350/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.5826 (0.6146)\tPrec@1 91.406 (86.754)\n",
      "[360/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5601 (0.6138)\tPrec@1 89.844 (86.779)\n",
      "[370/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6766 (0.6135)\tPrec@1 85.156 (86.763)\n",
      "[380/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6496 (0.6139)\tPrec@1 85.938 (86.743)\n",
      "[390/391]\tTime 0.093 (0.181)\tData 0.012 (0.020)\tLoss 0.5206 (0.6135)\tPrec@1 92.500 (86.760)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8428 (0.8428)\tPrec@1 83.594 (83.594)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.9194 (1.0172)\tPrec@1 85.156 (81.960)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.9034 (1.1116)\tPrec@1 82.031 (80.841)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 1.2650 (1.1317)\tPrec@1 80.469 (80.292)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.0887 (1.1317)\tPrec@1 79.688 (80.221)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.9324 (1.0927)\tPrec@1 78.906 (80.469)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.8731 (1.0927)\tPrec@1 82.031 (80.546)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.1115 (1.0944)\tPrec@1 79.688 (80.447)\n",
      " * Prec@1 80.370\n",
      "\n",
      "===> epoch: 94/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.166 (0.166)\tData 0.021 (0.021)\tLoss 0.5929 (0.5929)\tPrec@1 90.625 (90.625)\n",
      "[10/391]\tTime 0.182 (0.179)\tData 0.021 (0.020)\tLoss 0.6630 (0.6008)\tPrec@1 86.719 (87.500)\n",
      "[20/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6057 (0.5884)\tPrec@1 88.281 (88.281)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5655 (0.5863)\tPrec@1 85.938 (88.004)\n",
      "[40/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6910 (0.5969)\tPrec@1 84.375 (87.576)\n",
      "[50/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5702 (0.5957)\tPrec@1 89.844 (87.699)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5955 (0.5972)\tPrec@1 86.719 (87.500)\n",
      "[70/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5478 (0.5939)\tPrec@1 92.969 (87.632)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7293 (0.6007)\tPrec@1 86.719 (87.326)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.8165 (0.6040)\tPrec@1 78.906 (87.139)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7795 (0.6064)\tPrec@1 84.375 (87.090)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6228 (0.6085)\tPrec@1 84.375 (86.930)\n",
      "[120/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6158 (0.6081)\tPrec@1 85.938 (86.938)\n",
      "[130/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.6915 (0.6075)\tPrec@1 84.375 (87.005)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5553 (0.6070)\tPrec@1 87.500 (86.979)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.6004 (0.6067)\tPrec@1 86.719 (86.952)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5890 (0.6040)\tPrec@1 85.156 (86.990)\n",
      "[170/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.7054 (0.6024)\tPrec@1 82.031 (87.011)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6606 (0.6017)\tPrec@1 87.500 (87.051)\n",
      "[190/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.6222 (0.6020)\tPrec@1 87.500 (87.030)\n",
      "[200/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.6780 (0.6034)\tPrec@1 85.938 (87.014)\n",
      "[210/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.7447 (0.6062)\tPrec@1 85.938 (86.930)\n",
      "[220/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6558 (0.6068)\tPrec@1 84.375 (86.924)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5328 (0.6052)\tPrec@1 90.625 (86.966)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6059 (0.6043)\tPrec@1 89.062 (87.023)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4901 (0.6047)\tPrec@1 92.188 (87.021)\n",
      "[260/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.7457 (0.6052)\tPrec@1 81.250 (87.024)\n",
      "[270/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6109 (0.6069)\tPrec@1 89.844 (86.952)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.5740 (0.6075)\tPrec@1 89.844 (86.938)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6309 (0.6067)\tPrec@1 86.719 (86.995)\n",
      "[300/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.7815 (0.6077)\tPrec@1 80.469 (86.947)\n",
      "[310/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6713 (0.6087)\tPrec@1 82.812 (86.902)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6067 (0.6102)\tPrec@1 87.500 (86.833)\n",
      "[330/391]\tTime 0.182 (0.181)\tData 0.022 (0.020)\tLoss 0.6931 (0.6111)\tPrec@1 85.938 (86.806)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7551 (0.6121)\tPrec@1 83.594 (86.783)\n",
      "[350/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7497 (0.6137)\tPrec@1 82.031 (86.728)\n",
      "[360/391]\tTime 0.186 (0.181)\tData 0.022 (0.020)\tLoss 0.5866 (0.6139)\tPrec@1 87.500 (86.745)\n",
      "[370/391]\tTime 0.187 (0.181)\tData 0.021 (0.020)\tLoss 0.6331 (0.6143)\tPrec@1 84.375 (86.723)\n",
      "[380/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.6267 (0.6144)\tPrec@1 85.938 (86.717)\n",
      "[390/391]\tTime 0.094 (0.181)\tData 0.014 (0.020)\tLoss 0.5889 (0.6145)\tPrec@1 87.500 (86.720)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7483 (0.7483)\tPrec@1 85.938 (85.938)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.9569 (1.0393)\tPrec@1 82.031 (80.256)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 1.1896 (1.1369)\tPrec@1 76.562 (78.943)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.7064 (1.1444)\tPrec@1 82.812 (78.957)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.0749 (1.1134)\tPrec@1 76.562 (79.230)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.3446 (1.0934)\tPrec@1 73.438 (79.458)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 1.3929 (1.0908)\tPrec@1 76.562 (79.636)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 1.1200 (1.0727)\tPrec@1 77.344 (79.930)\n",
      " * Prec@1 79.990\n",
      "\n",
      "===> epoch: 95/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.163 (0.163)\tData 0.021 (0.021)\tLoss 0.6328 (0.6328)\tPrec@1 85.938 (85.938)\n",
      "[10/391]\tTime 0.179 (0.179)\tData 0.019 (0.021)\tLoss 0.6847 (0.6158)\tPrec@1 83.594 (86.790)\n",
      "[20/391]\tTime 0.187 (0.179)\tData 0.020 (0.021)\tLoss 0.6094 (0.6068)\tPrec@1 89.844 (87.277)\n",
      "[30/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.6237 (0.5879)\tPrec@1 86.719 (87.954)\n",
      "[40/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.6994 (0.5966)\tPrec@1 84.375 (87.595)\n",
      "[50/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.6908 (0.5983)\tPrec@1 83.594 (87.500)\n",
      "[60/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.6612 (0.5985)\tPrec@1 85.156 (87.487)\n",
      "[70/391]\tTime 0.187 (0.180)\tData 0.020 (0.020)\tLoss 0.5892 (0.6028)\tPrec@1 89.062 (87.302)\n",
      "[80/391]\tTime 0.186 (0.180)\tData 0.021 (0.020)\tLoss 0.5679 (0.6046)\tPrec@1 89.062 (87.201)\n",
      "[90/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5509 (0.6029)\tPrec@1 90.625 (87.251)\n",
      "[100/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.6638 (0.6050)\tPrec@1 87.500 (87.214)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6631 (0.6052)\tPrec@1 87.500 (87.162)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.7251 (0.6063)\tPrec@1 82.031 (87.184)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.5180 (0.6069)\tPrec@1 91.406 (87.178)\n",
      "[140/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5763 (0.6081)\tPrec@1 88.281 (87.123)\n",
      "[150/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.5756 (0.6065)\tPrec@1 90.625 (87.148)\n",
      "[160/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6657 (0.6069)\tPrec@1 81.250 (87.073)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5765 (0.6065)\tPrec@1 90.625 (87.130)\n",
      "[180/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.5978 (0.6096)\tPrec@1 88.281 (87.008)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5891 (0.6108)\tPrec@1 89.062 (86.956)\n",
      "[200/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5838 (0.6103)\tPrec@1 90.625 (86.968)\n",
      "[210/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6672 (0.6112)\tPrec@1 86.719 (86.967)\n",
      "[220/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.5327 (0.6102)\tPrec@1 92.969 (87.019)\n",
      "[230/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6298 (0.6116)\tPrec@1 85.938 (86.942)\n",
      "[240/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.5049 (0.6120)\tPrec@1 91.406 (86.926)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5715 (0.6132)\tPrec@1 85.938 (86.884)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5930 (0.6123)\tPrec@1 88.281 (86.913)\n",
      "[270/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5637 (0.6134)\tPrec@1 89.062 (86.863)\n",
      "[280/391]\tTime 0.187 (0.181)\tData 0.023 (0.020)\tLoss 0.6811 (0.6131)\tPrec@1 84.375 (86.874)\n",
      "[290/391]\tTime 0.188 (0.181)\tData 0.022 (0.020)\tLoss 0.5571 (0.6130)\tPrec@1 89.062 (86.864)\n",
      "[300/391]\tTime 0.189 (0.181)\tData 0.019 (0.020)\tLoss 0.6009 (0.6122)\tPrec@1 85.156 (86.836)\n",
      "[310/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.6382 (0.6120)\tPrec@1 86.719 (86.844)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.5585 (0.6120)\tPrec@1 89.062 (86.857)\n",
      "[330/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6139 (0.6113)\tPrec@1 89.062 (86.905)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.7320 (0.6111)\tPrec@1 78.906 (86.911)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5664 (0.6111)\tPrec@1 85.938 (86.910)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[360/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.5601 (0.6104)\tPrec@1 89.844 (86.931)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.5389 (0.6098)\tPrec@1 89.062 (86.952)\n",
      "[380/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.8189 (0.6105)\tPrec@1 72.656 (86.920)\n",
      "[390/391]\tTime 0.097 (0.181)\tData 0.013 (0.020)\tLoss 0.5768 (0.6106)\tPrec@1 90.000 (86.920)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6171 (0.6171)\tPrec@1 88.281 (88.281)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.6567 (0.8216)\tPrec@1 86.719 (84.162)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.9406 (0.8711)\tPrec@1 82.031 (83.371)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.5557 (0.8476)\tPrec@1 85.938 (83.191)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.0533 (0.8346)\tPrec@1 77.344 (83.251)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 0.8572 (0.8425)\tPrec@1 82.031 (82.935)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.9887 (0.8327)\tPrec@1 81.250 (83.222)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.8247 (0.8287)\tPrec@1 84.375 (83.319)\n",
      " * Prec@1 83.340\n",
      "\n",
      "===> epoch: 96/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.164 (0.164)\tData 0.021 (0.021)\tLoss 0.5764 (0.5764)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.5354 (0.5516)\tPrec@1 89.062 (88.778)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6108 (0.5517)\tPrec@1 83.594 (88.951)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.5386 (0.5599)\tPrec@1 92.188 (88.810)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5904 (0.5770)\tPrec@1 86.719 (88.186)\n",
      "[50/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5336 (0.5738)\tPrec@1 88.281 (88.266)\n",
      "[60/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.6514 (0.5788)\tPrec@1 82.031 (88.153)\n",
      "[70/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6671 (0.5859)\tPrec@1 83.594 (87.753)\n",
      "[80/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5712 (0.5895)\tPrec@1 88.281 (87.645)\n",
      "[90/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5655 (0.5945)\tPrec@1 89.062 (87.423)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.5449 (0.5979)\tPrec@1 89.844 (87.314)\n",
      "[110/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5245 (0.5973)\tPrec@1 87.500 (87.268)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5831 (0.5990)\tPrec@1 87.500 (87.222)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6248 (0.6025)\tPrec@1 88.281 (87.100)\n",
      "[140/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6042 (0.6043)\tPrec@1 89.844 (87.062)\n",
      "[150/391]\tTime 0.191 (0.182)\tData 0.020 (0.020)\tLoss 0.5812 (0.6045)\tPrec@1 85.156 (87.091)\n",
      "[160/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6114 (0.6045)\tPrec@1 85.938 (87.102)\n",
      "[170/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6034 (0.6053)\tPrec@1 85.156 (87.066)\n",
      "[180/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.6253 (0.6066)\tPrec@1 85.938 (87.017)\n",
      "[190/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.5641 (0.6067)\tPrec@1 87.500 (87.017)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6390 (0.6079)\tPrec@1 87.500 (86.964)\n",
      "[210/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.7141 (0.6081)\tPrec@1 83.594 (87.000)\n",
      "[220/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6390 (0.6084)\tPrec@1 85.938 (86.994)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6256 (0.6095)\tPrec@1 90.625 (86.959)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6468 (0.6116)\tPrec@1 85.156 (86.874)\n",
      "[250/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6618 (0.6125)\tPrec@1 86.719 (86.809)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5894 (0.6127)\tPrec@1 86.719 (86.806)\n",
      "[270/391]\tTime 0.195 (0.182)\tData 0.020 (0.020)\tLoss 0.6153 (0.6115)\tPrec@1 81.250 (86.854)\n",
      "[280/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6181 (0.6106)\tPrec@1 84.375 (86.847)\n",
      "[290/391]\tTime 0.189 (0.182)\tData 0.025 (0.020)\tLoss 0.5962 (0.6097)\tPrec@1 85.156 (86.893)\n",
      "[300/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6043 (0.6095)\tPrec@1 86.719 (86.903)\n",
      "[310/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7891 (0.6104)\tPrec@1 78.125 (86.872)\n",
      "[320/391]\tTime 0.183 (0.182)\tData 0.022 (0.020)\tLoss 0.5936 (0.6118)\tPrec@1 89.844 (86.833)\n",
      "[330/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.7072 (0.6133)\tPrec@1 82.812 (86.785)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.5355 (0.6134)\tPrec@1 89.844 (86.781)\n",
      "[350/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.5858 (0.6140)\tPrec@1 90.625 (86.759)\n",
      "[360/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6070 (0.6144)\tPrec@1 87.500 (86.760)\n",
      "[370/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6011 (0.6155)\tPrec@1 85.938 (86.715)\n",
      "[380/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.4954 (0.6163)\tPrec@1 93.750 (86.682)\n",
      "[390/391]\tTime 0.092 (0.182)\tData 0.012 (0.020)\tLoss 0.5124 (0.6167)\tPrec@1 95.000 (86.666)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.6803 (1.6803)\tPrec@1 73.438 (73.438)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 1.9024 (1.7636)\tPrec@1 70.312 (72.798)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 2.0128 (1.8593)\tPrec@1 74.219 (72.247)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 1.5603 (1.8562)\tPrec@1 72.656 (72.203)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 2.1048 (1.8485)\tPrec@1 70.312 (72.180)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 1.7725 (1.8575)\tPrec@1 69.531 (71.936)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 2.0622 (1.8687)\tPrec@1 71.875 (71.901)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.8589 (1.8492)\tPrec@1 68.750 (71.963)\n",
      " * Prec@1 72.160\n",
      "\n",
      "===> epoch: 97/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.163 (0.163)\tData 0.021 (0.021)\tLoss 0.6569 (0.6569)\tPrec@1 81.250 (81.250)\n",
      "[10/391]\tTime 0.185 (0.179)\tData 0.020 (0.020)\tLoss 0.6221 (0.6359)\tPrec@1 89.844 (85.014)\n",
      "[20/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.4790 (0.6186)\tPrec@1 96.094 (85.714)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.5771 (0.6238)\tPrec@1 86.719 (85.559)\n",
      "[40/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5992 (0.6258)\tPrec@1 85.938 (85.747)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6534 (0.6254)\tPrec@1 84.375 (85.662)\n",
      "[60/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.5974 (0.6157)\tPrec@1 88.281 (86.078)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.5049 (0.6089)\tPrec@1 92.188 (86.532)\n",
      "[80/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.6606 (0.6147)\tPrec@1 87.500 (86.478)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.7843 (0.6159)\tPrec@1 80.469 (86.496)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6788 (0.6162)\tPrec@1 82.812 (86.487)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5122 (0.6160)\tPrec@1 92.969 (86.585)\n",
      "[120/391]\tTime 0.184 (0.181)\tData 0.022 (0.020)\tLoss 0.6312 (0.6144)\tPrec@1 86.719 (86.648)\n",
      "[130/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5224 (0.6130)\tPrec@1 91.406 (86.755)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.5361 (0.6111)\tPrec@1 91.406 (86.857)\n",
      "[150/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.7259 (0.6139)\tPrec@1 82.812 (86.739)\n",
      "[160/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5534 (0.6129)\tPrec@1 86.719 (86.821)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.6437 (0.6130)\tPrec@1 82.812 (86.778)\n",
      "[180/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.7182 (0.6153)\tPrec@1 83.594 (86.723)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5515 (0.6151)\tPrec@1 89.062 (86.711)\n",
      "[200/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.6396 (0.6165)\tPrec@1 85.938 (86.688)\n",
      "[210/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6768 (0.6163)\tPrec@1 88.281 (86.708)\n",
      "[220/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6070 (0.6164)\tPrec@1 85.938 (86.690)\n",
      "[230/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.8652 (0.6175)\tPrec@1 78.125 (86.634)\n",
      "[240/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.6421 (0.6174)\tPrec@1 84.375 (86.654)\n",
      "[250/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5796 (0.6163)\tPrec@1 86.719 (86.647)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6863 (0.6165)\tPrec@1 83.594 (86.644)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.6429 (0.6158)\tPrec@1 81.250 (86.658)\n",
      "[280/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5561 (0.6151)\tPrec@1 91.406 (86.713)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.6026 (0.6140)\tPrec@1 84.375 (86.794)\n",
      "[300/391]\tTime 0.189 (0.181)\tData 0.019 (0.020)\tLoss 0.6351 (0.6142)\tPrec@1 88.281 (86.781)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[310/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6841 (0.6163)\tPrec@1 85.156 (86.704)\n",
      "[320/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6588 (0.6166)\tPrec@1 85.938 (86.692)\n",
      "[330/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5854 (0.6164)\tPrec@1 89.062 (86.695)\n",
      "[340/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.6532 (0.6167)\tPrec@1 82.031 (86.698)\n",
      "[350/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5686 (0.6169)\tPrec@1 89.062 (86.681)\n",
      "[360/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.5523 (0.6159)\tPrec@1 88.281 (86.721)\n",
      "[370/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.6832 (0.6160)\tPrec@1 82.812 (86.723)\n",
      "[380/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.6477 (0.6166)\tPrec@1 82.812 (86.690)\n",
      "[390/391]\tTime 0.095 (0.181)\tData 0.013 (0.020)\tLoss 0.5200 (0.6157)\tPrec@1 91.250 (86.722)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9891 (0.9891)\tPrec@1 80.469 (80.469)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 1.2114 (1.0537)\tPrec@1 79.688 (80.611)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 1.1107 (1.1129)\tPrec@1 82.031 (80.134)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.8515 (1.0975)\tPrec@1 86.719 (80.292)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 1.2833 (1.0929)\tPrec@1 78.906 (80.602)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 1.3096 (1.0870)\tPrec@1 76.562 (80.530)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 0.9641 (1.0741)\tPrec@1 85.938 (80.597)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.8311 (1.0697)\tPrec@1 83.594 (80.579)\n",
      " * Prec@1 80.530\n",
      "\n",
      "===> epoch: 98/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.164 (0.164)\tData 0.021 (0.021)\tLoss 0.5942 (0.5942)\tPrec@1 87.500 (87.500)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.019 (0.019)\tLoss 0.5688 (0.5894)\tPrec@1 84.375 (87.500)\n",
      "[20/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.6052 (0.5911)\tPrec@1 83.594 (87.463)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.5673 (0.5817)\tPrec@1 89.062 (87.853)\n",
      "[40/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6796 (0.5739)\tPrec@1 85.938 (88.091)\n",
      "[50/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6535 (0.5831)\tPrec@1 84.375 (87.852)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.019 (0.019)\tLoss 0.6355 (0.5922)\tPrec@1 85.938 (87.526)\n",
      "[70/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.6813 (0.5931)\tPrec@1 85.156 (87.500)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.6904 (0.5959)\tPrec@1 82.031 (87.355)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.021 (0.019)\tLoss 0.6055 (0.6011)\tPrec@1 85.156 (87.217)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.6428 (0.6043)\tPrec@1 83.594 (87.090)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.6721 (0.6082)\tPrec@1 81.250 (86.838)\n",
      "[120/391]\tTime 0.186 (0.181)\tData 0.020 (0.019)\tLoss 0.5882 (0.6097)\tPrec@1 87.500 (86.835)\n",
      "[130/391]\tTime 0.185 (0.181)\tData 0.020 (0.019)\tLoss 0.6202 (0.6095)\tPrec@1 84.375 (86.844)\n",
      "[140/391]\tTime 0.182 (0.181)\tData 0.020 (0.019)\tLoss 0.6084 (0.6090)\tPrec@1 86.719 (86.891)\n",
      "[150/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.5216 (0.6077)\tPrec@1 89.844 (86.952)\n",
      "[160/391]\tTime 0.188 (0.182)\tData 0.019 (0.019)\tLoss 0.6978 (0.6099)\tPrec@1 84.375 (86.903)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6614 (0.6099)\tPrec@1 81.250 (86.906)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.5570 (0.6088)\tPrec@1 89.062 (86.930)\n",
      "[190/391]\tTime 0.188 (0.182)\tData 0.019 (0.019)\tLoss 0.6358 (0.6091)\tPrec@1 82.031 (86.907)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.7055 (0.6101)\tPrec@1 84.375 (86.820)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.6432 (0.6115)\tPrec@1 80.469 (86.782)\n",
      "[220/391]\tTime 0.181 (0.182)\tData 0.020 (0.019)\tLoss 0.5674 (0.6120)\tPrec@1 89.844 (86.797)\n",
      "[230/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.5288 (0.6115)\tPrec@1 89.062 (86.830)\n",
      "[240/391]\tTime 0.181 (0.182)\tData 0.022 (0.019)\tLoss 0.5294 (0.6110)\tPrec@1 89.844 (86.855)\n",
      "[250/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.6647 (0.6114)\tPrec@1 85.938 (86.856)\n",
      "[260/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.6352 (0.6117)\tPrec@1 82.812 (86.829)\n",
      "[270/391]\tTime 0.179 (0.182)\tData 0.020 (0.019)\tLoss 0.5867 (0.6130)\tPrec@1 89.062 (86.748)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.6948 (0.6142)\tPrec@1 83.594 (86.708)\n",
      "[290/391]\tTime 0.183 (0.182)\tData 0.020 (0.019)\tLoss 0.5468 (0.6134)\tPrec@1 89.062 (86.740)\n",
      "[300/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.5971 (0.6140)\tPrec@1 85.156 (86.721)\n",
      "[310/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5965 (0.6145)\tPrec@1 88.281 (86.691)\n",
      "[320/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.7369 (0.6146)\tPrec@1 82.031 (86.721)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6777 (0.6151)\tPrec@1 82.031 (86.721)\n",
      "[340/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.6557 (0.6153)\tPrec@1 86.719 (86.723)\n",
      "[350/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.5485 (0.6154)\tPrec@1 88.281 (86.710)\n",
      "[360/391]\tTime 0.187 (0.182)\tData 0.021 (0.020)\tLoss 0.7086 (0.6164)\tPrec@1 85.938 (86.656)\n",
      "[370/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.6298 (0.6173)\tPrec@1 87.500 (86.624)\n",
      "[380/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.7117 (0.6183)\tPrec@1 82.812 (86.590)\n",
      "[390/391]\tTime 0.093 (0.181)\tData 0.012 (0.020)\tLoss 0.5244 (0.6186)\tPrec@1 91.250 (86.574)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6697 (0.6697)\tPrec@1 85.938 (85.938)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.8293 (0.8596)\tPrec@1 82.812 (82.955)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.7792 (0.8747)\tPrec@1 82.031 (83.147)\n",
      "Test: [30/79]\tTime 0.020 (0.020)\tLoss 0.5911 (0.8646)\tPrec@1 85.938 (83.039)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.2478 (0.8604)\tPrec@1 76.562 (83.003)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.7851 (0.8507)\tPrec@1 82.812 (83.119)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 1.1109 (0.8551)\tPrec@1 81.250 (83.017)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.0795 (0.8600)\tPrec@1 79.688 (82.879)\n",
      " * Prec@1 82.940\n",
      "\n",
      "===> epoch: 99/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.162 (0.162)\tData 0.022 (0.022)\tLoss 0.6146 (0.6146)\tPrec@1 89.844 (89.844)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.5369 (0.6021)\tPrec@1 91.406 (87.429)\n",
      "[20/391]\tTime 0.187 (0.180)\tData 0.020 (0.019)\tLoss 0.6539 (0.6088)\tPrec@1 85.938 (86.979)\n",
      "[30/391]\tTime 0.188 (0.181)\tData 0.019 (0.019)\tLoss 0.6570 (0.6115)\tPrec@1 84.375 (86.467)\n",
      "[40/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.6716 (0.6148)\tPrec@1 85.938 (86.604)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6852 (0.6154)\tPrec@1 85.938 (86.811)\n",
      "[60/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.4969 (0.6064)\tPrec@1 92.969 (87.103)\n",
      "[70/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.5668 (0.6060)\tPrec@1 87.500 (87.115)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.6106 (0.6041)\tPrec@1 90.625 (87.220)\n",
      "[90/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.6028 (0.6036)\tPrec@1 87.500 (87.242)\n",
      "[100/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.7411 (0.6055)\tPrec@1 83.594 (87.191)\n",
      "[110/391]\tTime 0.185 (0.182)\tData 0.021 (0.019)\tLoss 0.5683 (0.6051)\tPrec@1 89.844 (87.169)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5321 (0.6010)\tPrec@1 88.281 (87.345)\n",
      "[130/391]\tTime 0.188 (0.182)\tData 0.019 (0.020)\tLoss 0.5074 (0.5984)\tPrec@1 91.406 (87.476)\n",
      "[140/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.6195 (0.5996)\tPrec@1 85.938 (87.456)\n",
      "[150/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.6796 (0.6014)\tPrec@1 86.719 (87.407)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.6075 (0.6020)\tPrec@1 88.281 (87.384)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.6587 (0.6026)\tPrec@1 86.719 (87.331)\n",
      "[180/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.6715 (0.6044)\tPrec@1 82.812 (87.219)\n",
      "[190/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.6056 (0.6060)\tPrec@1 85.938 (87.197)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.6508 (0.6068)\tPrec@1 85.938 (87.162)\n",
      "[210/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.6589 (0.6085)\tPrec@1 85.938 (87.108)\n",
      "[220/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.5755 (0.6094)\tPrec@1 88.281 (87.051)\n",
      "[230/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.7105 (0.6100)\tPrec@1 79.688 (87.010)\n",
      "[240/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.6448 (0.6106)\tPrec@1 84.375 (86.998)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.6132 (0.6120)\tPrec@1 87.500 (86.946)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[260/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.6341 (0.6121)\tPrec@1 86.719 (86.916)\n",
      "[270/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.7362 (0.6120)\tPrec@1 81.250 (86.892)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.5296 (0.6119)\tPrec@1 89.062 (86.877)\n",
      "[290/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.5763 (0.6117)\tPrec@1 84.375 (86.858)\n",
      "[300/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.6727 (0.6125)\tPrec@1 85.156 (86.828)\n",
      "[310/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.7396 (0.6129)\tPrec@1 79.688 (86.789)\n",
      "[320/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.4794 (0.6126)\tPrec@1 92.188 (86.789)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.6522 (0.6131)\tPrec@1 89.844 (86.761)\n",
      "[340/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6483 (0.6129)\tPrec@1 85.938 (86.774)\n",
      "[350/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6831 (0.6138)\tPrec@1 83.594 (86.737)\n",
      "[360/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.6073 (0.6149)\tPrec@1 89.062 (86.697)\n",
      "[370/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.5995 (0.6146)\tPrec@1 85.156 (86.702)\n",
      "[380/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.6739 (0.6155)\tPrec@1 83.594 (86.684)\n",
      "[390/391]\tTime 0.095 (0.182)\tData 0.012 (0.020)\tLoss 0.6017 (0.6159)\tPrec@1 86.250 (86.668)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.7617 (0.7617)\tPrec@1 86.719 (86.719)\n",
      "Test: [10/79]\tTime 0.019 (0.021)\tLoss 0.8036 (1.0430)\tPrec@1 85.156 (82.102)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.9614 (1.1236)\tPrec@1 83.594 (80.990)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.9770 (1.0991)\tPrec@1 83.594 (81.048)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.4256 (1.0851)\tPrec@1 74.219 (81.155)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 1.1641 (1.0556)\tPrec@1 80.469 (81.388)\n",
      "Test: [60/79]\tTime 0.020 (0.020)\tLoss 1.0698 (1.0483)\tPrec@1 82.031 (81.378)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 1.0345 (1.0443)\tPrec@1 83.594 (81.437)\n",
      " * Prec@1 81.760\n",
      "\n",
      "===> epoch: 100/200\n",
      "current lr 1.00000e-01\n",
      "Training:\n",
      "[0/391]\tTime 0.171 (0.171)\tData 0.021 (0.021)\tLoss 0.5488 (0.5488)\tPrec@1 91.406 (91.406)\n",
      "[10/391]\tTime 0.182 (0.180)\tData 0.020 (0.019)\tLoss 0.5726 (0.6066)\tPrec@1 86.719 (86.009)\n",
      "[20/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.6110 (0.5917)\tPrec@1 86.719 (86.496)\n",
      "[30/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.5590 (0.5778)\tPrec@1 88.281 (87.122)\n",
      "[40/391]\tTime 0.185 (0.181)\tData 0.019 (0.019)\tLoss 0.5049 (0.5667)\tPrec@1 90.625 (87.805)\n",
      "[50/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.5464 (0.5519)\tPrec@1 88.281 (88.281)\n",
      "[60/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.5347 (0.5406)\tPrec@1 90.625 (88.627)\n",
      "[70/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5196 (0.5345)\tPrec@1 91.406 (88.699)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4827 (0.5294)\tPrec@1 89.844 (88.947)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4922 (0.5246)\tPrec@1 90.625 (89.097)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4725 (0.5196)\tPrec@1 90.625 (89.318)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.5169 (0.5181)\tPrec@1 90.625 (89.379)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4735 (0.5168)\tPrec@1 89.844 (89.392)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4784 (0.5129)\tPrec@1 88.281 (89.587)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4043 (0.5098)\tPrec@1 94.531 (89.661)\n",
      "[150/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4519 (0.5081)\tPrec@1 91.406 (89.689)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.5398 (0.5067)\tPrec@1 88.281 (89.718)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4073 (0.5037)\tPrec@1 92.188 (89.825)\n",
      "[180/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.4329 (0.5009)\tPrec@1 93.750 (89.934)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4073 (0.4981)\tPrec@1 96.094 (90.056)\n",
      "[200/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4642 (0.4958)\tPrec@1 91.406 (90.143)\n",
      "[210/391]\tTime 0.188 (0.181)\tData 0.020 (0.020)\tLoss 0.3984 (0.4927)\tPrec@1 94.531 (90.284)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4647 (0.4902)\tPrec@1 92.188 (90.388)\n",
      "[230/391]\tTime 0.182 (0.181)\tData 0.022 (0.020)\tLoss 0.4462 (0.4895)\tPrec@1 92.969 (90.429)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4585 (0.4884)\tPrec@1 93.750 (90.482)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3943 (0.4868)\tPrec@1 93.750 (90.525)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4579 (0.4847)\tPrec@1 90.625 (90.619)\n",
      "[270/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3756 (0.4834)\tPrec@1 94.531 (90.662)\n",
      "[280/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.4466 (0.4815)\tPrec@1 92.188 (90.745)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4422 (0.4798)\tPrec@1 93.750 (90.800)\n",
      "[300/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4305 (0.4790)\tPrec@1 92.969 (90.825)\n",
      "[310/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4673 (0.4779)\tPrec@1 90.625 (90.844)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4026 (0.4771)\tPrec@1 93.750 (90.851)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4280 (0.4759)\tPrec@1 91.406 (90.894)\n",
      "[340/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.4856 (0.4750)\tPrec@1 92.188 (90.934)\n",
      "[350/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3838 (0.4738)\tPrec@1 92.188 (90.977)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4278 (0.4727)\tPrec@1 94.531 (91.025)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4100 (0.4716)\tPrec@1 93.750 (91.067)\n",
      "[380/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3844 (0.4702)\tPrec@1 96.094 (91.111)\n",
      "[390/391]\tTime 0.092 (0.181)\tData 0.013 (0.020)\tLoss 0.4089 (0.4693)\tPrec@1 95.000 (91.136)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2564 (0.2564)\tPrec@1 92.188 (92.188)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.4293 (0.4238)\tPrec@1 90.625 (90.412)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.3346 (0.4730)\tPrec@1 92.969 (89.918)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.4830 (0.4830)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.5901 (0.4768)\tPrec@1 87.500 (90.053)\n",
      "Test: [50/79]\tTime 0.020 (0.020)\tLoss 0.4873 (0.4747)\tPrec@1 89.062 (90.150)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.3867 (0.4759)\tPrec@1 91.406 (90.215)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.6425 (0.4749)\tPrec@1 89.844 (90.251)\n",
      " * Prec@1 90.370\n",
      "\n",
      "===> epoch: 101/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.162 (0.162)\tData 0.021 (0.021)\tLoss 0.4865 (0.4865)\tPrec@1 89.844 (89.844)\n",
      "[10/391]\tTime 0.182 (0.179)\tData 0.019 (0.019)\tLoss 0.4641 (0.4257)\tPrec@1 90.625 (92.614)\n",
      "[20/391]\tTime 0.180 (0.179)\tData 0.020 (0.019)\tLoss 0.4684 (0.4310)\tPrec@1 89.062 (92.597)\n",
      "[30/391]\tTime 0.182 (0.180)\tData 0.021 (0.019)\tLoss 0.4376 (0.4356)\tPrec@1 92.188 (92.414)\n",
      "[40/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.3642 (0.4339)\tPrec@1 96.875 (92.473)\n",
      "[50/391]\tTime 0.186 (0.180)\tData 0.019 (0.019)\tLoss 0.3659 (0.4377)\tPrec@1 95.312 (92.203)\n",
      "[60/391]\tTime 0.185 (0.180)\tData 0.020 (0.019)\tLoss 0.5083 (0.4352)\tPrec@1 89.062 (92.252)\n",
      "[70/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.3961 (0.4351)\tPrec@1 94.531 (92.309)\n",
      "[80/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.4028 (0.4336)\tPrec@1 94.531 (92.380)\n",
      "[90/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.4361 (0.4334)\tPrec@1 92.188 (92.351)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3992 (0.4322)\tPrec@1 92.188 (92.373)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.3752 (0.4312)\tPrec@1 94.531 (92.434)\n",
      "[120/391]\tTime 0.182 (0.181)\tData 0.020 (0.019)\tLoss 0.4579 (0.4293)\tPrec@1 91.406 (92.472)\n",
      "[130/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3854 (0.4287)\tPrec@1 92.969 (92.492)\n",
      "[140/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4468 (0.4294)\tPrec@1 92.969 (92.393)\n",
      "[150/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.4411 (0.4279)\tPrec@1 91.406 (92.441)\n",
      "[160/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3287 (0.4277)\tPrec@1 96.875 (92.459)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3745 (0.4280)\tPrec@1 93.750 (92.430)\n",
      "[180/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.5056 (0.4282)\tPrec@1 89.062 (92.434)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4703 (0.4284)\tPrec@1 87.500 (92.400)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4288 (0.4286)\tPrec@1 93.750 (92.425)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[210/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3603 (0.4278)\tPrec@1 94.531 (92.469)\n",
      "[220/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.4134 (0.4277)\tPrec@1 92.969 (92.488)\n",
      "[230/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.4449 (0.4270)\tPrec@1 92.188 (92.512)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4543 (0.4267)\tPrec@1 90.625 (92.518)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4681 (0.4271)\tPrec@1 89.844 (92.496)\n",
      "[260/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3814 (0.4272)\tPrec@1 94.531 (92.499)\n",
      "[270/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4640 (0.4265)\tPrec@1 89.062 (92.507)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3531 (0.4262)\tPrec@1 94.531 (92.516)\n",
      "[290/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.4813 (0.4257)\tPrec@1 92.188 (92.550)\n",
      "[300/391]\tTime 0.188 (0.181)\tData 0.020 (0.020)\tLoss 0.4570 (0.4262)\tPrec@1 90.625 (92.515)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.021 (0.020)\tLoss 0.4556 (0.4258)\tPrec@1 91.406 (92.522)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4020 (0.4254)\tPrec@1 92.188 (92.536)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4145 (0.4250)\tPrec@1 94.531 (92.556)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3903 (0.4244)\tPrec@1 92.969 (92.577)\n",
      "[350/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3539 (0.4235)\tPrec@1 96.875 (92.613)\n",
      "[360/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4193 (0.4235)\tPrec@1 93.750 (92.594)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.022 (0.020)\tLoss 0.4736 (0.4239)\tPrec@1 89.844 (92.583)\n",
      "[380/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.5222 (0.4239)\tPrec@1 87.500 (92.573)\n",
      "[390/391]\tTime 0.092 (0.181)\tData 0.012 (0.020)\tLoss 0.5106 (0.4238)\tPrec@1 92.500 (92.580)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3037 (0.3037)\tPrec@1 92.969 (92.969)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.4016 (0.3827)\tPrec@1 92.188 (91.690)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.3206 (0.4387)\tPrec@1 92.969 (91.109)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.4340 (0.4511)\tPrec@1 90.625 (90.776)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.5887 (0.4484)\tPrec@1 86.719 (90.796)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 0.5871 (0.4446)\tPrec@1 86.719 (90.870)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 0.3509 (0.4429)\tPrec@1 91.406 (90.868)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.5194 (0.4416)\tPrec@1 89.844 (90.823)\n",
      " * Prec@1 90.970\n",
      "\n",
      "===> epoch: 102/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.164 (0.164)\tData 0.022 (0.022)\tLoss 0.4582 (0.4582)\tPrec@1 92.188 (92.188)\n",
      "[10/391]\tTime 0.181 (0.178)\tData 0.021 (0.021)\tLoss 0.4509 (0.4056)\tPrec@1 90.625 (93.537)\n",
      "[20/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.3959 (0.4131)\tPrec@1 94.531 (93.266)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.4637 (0.4108)\tPrec@1 90.625 (93.271)\n",
      "[40/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3758 (0.4057)\tPrec@1 94.531 (93.502)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4250 (0.4034)\tPrec@1 92.188 (93.551)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.4896 (0.4053)\tPrec@1 89.062 (93.481)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.4021 (0.4060)\tPrec@1 92.188 (93.453)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3842 (0.4061)\tPrec@1 94.531 (93.480)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4184 (0.4051)\tPrec@1 92.188 (93.535)\n",
      "[100/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.5244 (0.4072)\tPrec@1 86.719 (93.456)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3352 (0.4061)\tPrec@1 95.312 (93.461)\n",
      "[120/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3664 (0.4050)\tPrec@1 94.531 (93.492)\n",
      "[130/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.3964 (0.4055)\tPrec@1 92.188 (93.476)\n",
      "[140/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4083 (0.4050)\tPrec@1 93.750 (93.479)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.4419 (0.4058)\tPrec@1 93.750 (93.481)\n",
      "[160/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4407 (0.4058)\tPrec@1 91.406 (93.435)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4104 (0.4066)\tPrec@1 92.969 (93.394)\n",
      "[180/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3857 (0.4055)\tPrec@1 92.188 (93.444)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4328 (0.4059)\tPrec@1 92.188 (93.439)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4639 (0.4054)\tPrec@1 88.281 (93.427)\n",
      "[210/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.3409 (0.4051)\tPrec@1 97.656 (93.432)\n",
      "[220/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.3536 (0.4047)\tPrec@1 94.531 (93.428)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4160 (0.4059)\tPrec@1 93.750 (93.378)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3647 (0.4054)\tPrec@1 92.969 (93.371)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.5335 (0.4063)\tPrec@1 86.719 (93.327)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4517 (0.4069)\tPrec@1 89.062 (93.277)\n",
      "[270/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.3554 (0.4068)\tPrec@1 96.875 (93.280)\n",
      "[280/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.4110 (0.4062)\tPrec@1 93.750 (93.305)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3614 (0.4062)\tPrec@1 95.312 (93.296)\n",
      "[300/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4756 (0.4066)\tPrec@1 89.844 (93.288)\n",
      "[310/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4464 (0.4062)\tPrec@1 91.406 (93.318)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3543 (0.4055)\tPrec@1 96.094 (93.344)\n",
      "[330/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.3761 (0.4052)\tPrec@1 92.969 (93.356)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4578 (0.4055)\tPrec@1 91.406 (93.338)\n",
      "[350/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3871 (0.4055)\tPrec@1 93.750 (93.345)\n",
      "[360/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3933 (0.4059)\tPrec@1 96.094 (93.322)\n",
      "[370/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4668 (0.4062)\tPrec@1 88.281 (93.299)\n",
      "[380/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4337 (0.4062)\tPrec@1 92.969 (93.295)\n",
      "[390/391]\tTime 0.095 (0.181)\tData 0.012 (0.020)\tLoss 0.3543 (0.4063)\tPrec@1 93.750 (93.292)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2662 (0.2662)\tPrec@1 92.969 (92.969)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.3626 (0.3852)\tPrec@1 92.188 (92.045)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.2939 (0.4568)\tPrec@1 94.531 (90.848)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.3397 (0.4662)\tPrec@1 92.969 (90.650)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.6096 (0.4688)\tPrec@1 85.156 (90.720)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 0.5793 (0.4704)\tPrec@1 87.500 (90.748)\n",
      "Test: [60/79]\tTime 0.019 (0.019)\tLoss 0.4054 (0.4683)\tPrec@1 91.406 (90.881)\n",
      "Test: [70/79]\tTime 0.019 (0.019)\tLoss 0.6806 (0.4631)\tPrec@1 88.281 (90.867)\n",
      " * Prec@1 90.930\n",
      "\n",
      "===> epoch: 103/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.168 (0.168)\tData 0.021 (0.021)\tLoss 0.3412 (0.3412)\tPrec@1 96.094 (96.094)\n",
      "[10/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.3871 (0.3879)\tPrec@1 95.312 (93.963)\n",
      "[20/391]\tTime 0.181 (0.180)\tData 0.019 (0.019)\tLoss 0.4013 (0.3918)\tPrec@1 95.312 (93.787)\n",
      "[30/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.3215 (0.3891)\tPrec@1 96.875 (93.800)\n",
      "[40/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.3656 (0.3861)\tPrec@1 95.312 (93.883)\n",
      "[50/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3483 (0.3853)\tPrec@1 97.656 (94.010)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4745 (0.3848)\tPrec@1 89.844 (94.070)\n",
      "[70/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4364 (0.3892)\tPrec@1 92.188 (93.981)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4123 (0.3903)\tPrec@1 92.188 (93.904)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4222 (0.3935)\tPrec@1 92.969 (93.724)\n",
      "[100/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3939 (0.3929)\tPrec@1 92.969 (93.750)\n",
      "[110/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4258 (0.3943)\tPrec@1 93.750 (93.708)\n",
      "[120/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.3930 (0.3930)\tPrec@1 93.750 (93.802)\n",
      "[130/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.4354 (0.3926)\tPrec@1 92.969 (93.810)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3701 (0.3927)\tPrec@1 93.750 (93.805)\n",
      "[150/391]\tTime 0.180 (0.181)\tData 0.022 (0.020)\tLoss 0.3878 (0.3925)\tPrec@1 92.188 (93.828)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.183 (0.181)\tData 0.022 (0.020)\tLoss 0.3576 (0.3921)\tPrec@1 96.094 (93.852)\n",
      "[170/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.3947 (0.3929)\tPrec@1 93.750 (93.777)\n",
      "[180/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.4647 (0.3936)\tPrec@1 90.625 (93.789)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.3629 (0.3935)\tPrec@1 96.875 (93.791)\n",
      "[200/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3833 (0.3935)\tPrec@1 94.531 (93.742)\n",
      "[210/391]\tTime 0.185 (0.181)\tData 0.021 (0.020)\tLoss 0.3503 (0.3927)\tPrec@1 96.094 (93.765)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.4255 (0.3932)\tPrec@1 89.844 (93.732)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3312 (0.3925)\tPrec@1 96.094 (93.780)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.3141 (0.3922)\tPrec@1 98.438 (93.789)\n",
      "[250/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.4015 (0.3923)\tPrec@1 92.969 (93.762)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.022 (0.020)\tLoss 0.4170 (0.3922)\tPrec@1 92.969 (93.783)\n",
      "[270/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3677 (0.3921)\tPrec@1 96.094 (93.805)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4059 (0.3922)\tPrec@1 95.312 (93.811)\n",
      "[290/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3757 (0.3924)\tPrec@1 93.750 (93.774)\n",
      "[300/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3591 (0.3929)\tPrec@1 95.312 (93.724)\n",
      "[310/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.4990 (0.3929)\tPrec@1 89.062 (93.727)\n",
      "[320/391]\tTime 0.195 (0.181)\tData 0.034 (0.020)\tLoss 0.4722 (0.3937)\tPrec@1 90.625 (93.713)\n",
      "[330/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.4529 (0.3940)\tPrec@1 91.406 (93.705)\n",
      "[340/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4198 (0.3942)\tPrec@1 90.625 (93.677)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4722 (0.3947)\tPrec@1 89.844 (93.670)\n",
      "[360/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3448 (0.3948)\tPrec@1 97.656 (93.683)\n",
      "[370/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.4265 (0.3949)\tPrec@1 92.969 (93.693)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3616 (0.3951)\tPrec@1 96.875 (93.680)\n",
      "[390/391]\tTime 0.096 (0.181)\tData 0.012 (0.020)\tLoss 0.3883 (0.3949)\tPrec@1 93.750 (93.700)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3328 (0.3328)\tPrec@1 92.969 (92.969)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.3628 (0.3858)\tPrec@1 94.531 (91.761)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.3338 (0.4378)\tPrec@1 91.406 (91.146)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.4384 (0.4458)\tPrec@1 92.188 (91.028)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5613 (0.4438)\tPrec@1 86.719 (91.025)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 0.4679 (0.4338)\tPrec@1 89.844 (91.161)\n",
      "Test: [60/79]\tTime 0.022 (0.020)\tLoss 0.3767 (0.4323)\tPrec@1 92.188 (91.137)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.5018 (0.4318)\tPrec@1 90.625 (91.164)\n",
      " * Prec@1 91.290\n",
      "\n",
      "===> epoch: 104/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.163 (0.163)\tData 0.020 (0.020)\tLoss 0.4349 (0.4349)\tPrec@1 92.969 (92.969)\n",
      "[10/391]\tTime 0.183 (0.179)\tData 0.021 (0.020)\tLoss 0.3817 (0.3933)\tPrec@1 93.750 (93.892)\n",
      "[20/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.4571 (0.3903)\tPrec@1 91.406 (93.936)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3952 (0.3878)\tPrec@1 94.531 (94.078)\n",
      "[40/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.3673 (0.3880)\tPrec@1 96.094 (94.093)\n",
      "[50/391]\tTime 0.189 (0.180)\tData 0.020 (0.020)\tLoss 0.3613 (0.3830)\tPrec@1 94.531 (94.271)\n",
      "[60/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.4445 (0.3878)\tPrec@1 91.406 (94.045)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.3803 (0.3892)\tPrec@1 92.969 (94.025)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4393 (0.3897)\tPrec@1 91.406 (93.866)\n",
      "[90/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.4136 (0.3886)\tPrec@1 92.188 (93.896)\n",
      "[100/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3692 (0.3882)\tPrec@1 96.094 (93.959)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4686 (0.3870)\tPrec@1 90.625 (93.961)\n",
      "[120/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4521 (0.3881)\tPrec@1 90.625 (93.957)\n",
      "[130/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3180 (0.3895)\tPrec@1 97.656 (93.899)\n",
      "[140/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3652 (0.3893)\tPrec@1 96.094 (93.855)\n",
      "[150/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3569 (0.3886)\tPrec@1 96.875 (93.916)\n",
      "[160/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3938 (0.3885)\tPrec@1 93.750 (93.896)\n",
      "[170/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3634 (0.3881)\tPrec@1 92.969 (93.914)\n",
      "[180/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3745 (0.3878)\tPrec@1 92.969 (93.927)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4458 (0.3893)\tPrec@1 91.406 (93.844)\n",
      "[200/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3668 (0.3893)\tPrec@1 94.531 (93.855)\n",
      "[210/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3840 (0.3887)\tPrec@1 94.531 (93.894)\n",
      "[220/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.3905 (0.3887)\tPrec@1 92.969 (93.874)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.4200 (0.3887)\tPrec@1 93.750 (93.872)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.3816 (0.3887)\tPrec@1 96.094 (93.857)\n",
      "[250/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.3833 (0.3892)\tPrec@1 92.188 (93.828)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.4386 (0.3902)\tPrec@1 91.406 (93.801)\n",
      "[270/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.3413 (0.3911)\tPrec@1 95.312 (93.750)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.3886 (0.3907)\tPrec@1 93.750 (93.778)\n",
      "[290/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.4526 (0.3907)\tPrec@1 93.750 (93.782)\n",
      "[300/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3941 (0.3913)\tPrec@1 92.188 (93.750)\n",
      "[310/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.3575 (0.3906)\tPrec@1 93.750 (93.785)\n",
      "[320/391]\tTime 0.188 (0.182)\tData 0.019 (0.019)\tLoss 0.4061 (0.3909)\tPrec@1 95.312 (93.762)\n",
      "[330/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.3717 (0.3915)\tPrec@1 92.969 (93.731)\n",
      "[340/391]\tTime 0.181 (0.182)\tData 0.020 (0.019)\tLoss 0.4000 (0.3918)\tPrec@1 93.750 (93.734)\n",
      "[350/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.3603 (0.3916)\tPrec@1 96.094 (93.754)\n",
      "[360/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.3452 (0.3916)\tPrec@1 96.875 (93.748)\n",
      "[370/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.4053 (0.3920)\tPrec@1 92.969 (93.742)\n",
      "[380/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.4220 (0.3921)\tPrec@1 91.406 (93.736)\n",
      "[390/391]\tTime 0.094 (0.182)\tData 0.013 (0.019)\tLoss 0.4079 (0.3916)\tPrec@1 95.000 (93.764)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2805 (0.2805)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.021 (0.019)\tLoss 0.3847 (0.3707)\tPrec@1 92.188 (92.472)\n",
      "Test: [20/79]\tTime 0.020 (0.020)\tLoss 0.2677 (0.4337)\tPrec@1 96.094 (91.704)\n",
      "Test: [30/79]\tTime 0.019 (0.020)\tLoss 0.3790 (0.4408)\tPrec@1 92.969 (91.431)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.5244 (0.4342)\tPrec@1 87.500 (91.425)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 0.4638 (0.4243)\tPrec@1 89.062 (91.513)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.2915 (0.4230)\tPrec@1 92.969 (91.547)\n",
      "Test: [70/79]\tTime 0.019 (0.020)\tLoss 0.5237 (0.4230)\tPrec@1 89.844 (91.505)\n",
      " * Prec@1 91.510\n",
      "\n",
      "===> epoch: 105/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.3650 (0.3650)\tPrec@1 96.094 (96.094)\n",
      "[10/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3206 (0.3836)\tPrec@1 96.875 (94.815)\n",
      "[20/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3183 (0.3812)\tPrec@1 96.875 (94.643)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.020 (0.019)\tLoss 0.3894 (0.3807)\tPrec@1 92.188 (94.531)\n",
      "[40/391]\tTime 0.185 (0.181)\tData 0.020 (0.019)\tLoss 0.3559 (0.3803)\tPrec@1 94.531 (94.455)\n",
      "[50/391]\tTime 0.182 (0.181)\tData 0.019 (0.019)\tLoss 0.4210 (0.3837)\tPrec@1 93.750 (94.332)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.4805 (0.3852)\tPrec@1 93.750 (94.249)\n",
      "[70/391]\tTime 0.180 (0.181)\tData 0.019 (0.019)\tLoss 0.3693 (0.3848)\tPrec@1 96.094 (94.234)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.4304 (0.3839)\tPrec@1 92.969 (94.300)\n",
      "[90/391]\tTime 0.185 (0.181)\tData 0.020 (0.019)\tLoss 0.3819 (0.3838)\tPrec@1 92.969 (94.265)\n",
      "[100/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3101 (0.3819)\tPrec@1 98.438 (94.377)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.2976 (0.3818)\tPrec@1 98.438 (94.376)\n",
      "[120/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3645 (0.3819)\tPrec@1 93.750 (94.409)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3750 (0.3834)\tPrec@1 94.531 (94.281)\n",
      "[140/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3913 (0.3824)\tPrec@1 94.531 (94.304)\n",
      "[150/391]\tTime 0.187 (0.181)\tData 0.021 (0.020)\tLoss 0.4396 (0.3826)\tPrec@1 91.406 (94.298)\n",
      "[160/391]\tTime 0.186 (0.181)\tData 0.021 (0.020)\tLoss 0.4733 (0.3833)\tPrec@1 91.406 (94.274)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3925 (0.3841)\tPrec@1 92.969 (94.243)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4099 (0.3841)\tPrec@1 92.969 (94.212)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3414 (0.3831)\tPrec@1 96.094 (94.253)\n",
      "[200/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3791 (0.3832)\tPrec@1 92.969 (94.213)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4277 (0.3841)\tPrec@1 91.406 (94.161)\n",
      "[220/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.4194 (0.3851)\tPrec@1 92.188 (94.111)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4867 (0.3861)\tPrec@1 88.281 (94.031)\n",
      "[240/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.3665 (0.3862)\tPrec@1 92.969 (94.048)\n",
      "[250/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3892 (0.3860)\tPrec@1 93.750 (94.052)\n",
      "[260/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.3256 (0.3858)\tPrec@1 96.094 (94.055)\n",
      "[270/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3927 (0.3856)\tPrec@1 94.531 (94.076)\n",
      "[280/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4260 (0.3858)\tPrec@1 92.188 (94.050)\n",
      "[290/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3967 (0.3854)\tPrec@1 92.188 (94.064)\n",
      "[300/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4012 (0.3856)\tPrec@1 93.750 (94.054)\n",
      "[310/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.3604 (0.3845)\tPrec@1 96.094 (94.087)\n",
      "[320/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4481 (0.3847)\tPrec@1 92.969 (94.086)\n",
      "[330/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.4011 (0.3842)\tPrec@1 92.188 (94.109)\n",
      "[340/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.3433 (0.3835)\tPrec@1 96.094 (94.126)\n",
      "[350/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.4201 (0.3830)\tPrec@1 91.406 (94.137)\n",
      "[360/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.4533 (0.3830)\tPrec@1 91.406 (94.129)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3275 (0.3825)\tPrec@1 96.875 (94.165)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3767 (0.3824)\tPrec@1 95.312 (94.168)\n",
      "[390/391]\tTime 0.094 (0.181)\tData 0.012 (0.020)\tLoss 0.4049 (0.3823)\tPrec@1 95.000 (94.192)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2803 (0.2803)\tPrec@1 94.531 (94.531)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.2998 (0.3906)\tPrec@1 94.531 (92.116)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.3219 (0.4381)\tPrec@1 95.312 (91.518)\n",
      "Test: [30/79]\tTime 0.020 (0.019)\tLoss 0.4664 (0.4539)\tPrec@1 91.406 (91.129)\n",
      "Test: [40/79]\tTime 0.024 (0.020)\tLoss 0.6117 (0.4576)\tPrec@1 87.500 (91.120)\n",
      "Test: [50/79]\tTime 0.019 (0.020)\tLoss 0.5434 (0.4500)\tPrec@1 89.062 (91.207)\n",
      "Test: [60/79]\tTime 0.019 (0.020)\tLoss 0.3303 (0.4517)\tPrec@1 92.969 (91.253)\n",
      "Test: [70/79]\tTime 0.020 (0.020)\tLoss 0.6624 (0.4467)\tPrec@1 89.062 (91.252)\n",
      " * Prec@1 91.280\n",
      "\n",
      "===> epoch: 106/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.182 (0.182)\tData 0.021 (0.021)\tLoss 0.3699 (0.3699)\tPrec@1 94.531 (94.531)\n",
      "[10/391]\tTime 0.186 (0.185)\tData 0.020 (0.020)\tLoss 0.3198 (0.3660)\tPrec@1 97.656 (95.241)\n",
      "[20/391]\tTime 0.180 (0.183)\tData 0.020 (0.020)\tLoss 0.4017 (0.3717)\tPrec@1 92.969 (94.792)\n",
      "[30/391]\tTime 0.180 (0.183)\tData 0.019 (0.020)\tLoss 0.3446 (0.3665)\tPrec@1 95.312 (94.985)\n",
      "[40/391]\tTime 0.180 (0.183)\tData 0.019 (0.020)\tLoss 0.3639 (0.3692)\tPrec@1 92.188 (95.008)\n",
      "[50/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.4253 (0.3695)\tPrec@1 90.625 (94.838)\n",
      "[60/391]\tTime 0.184 (0.183)\tData 0.021 (0.020)\tLoss 0.4179 (0.3708)\tPrec@1 93.750 (94.800)\n",
      "[70/391]\tTime 0.187 (0.183)\tData 0.020 (0.020)\tLoss 0.4333 (0.3725)\tPrec@1 93.750 (94.685)\n",
      "[80/391]\tTime 0.182 (0.183)\tData 0.019 (0.020)\tLoss 0.3474 (0.3736)\tPrec@1 95.312 (94.599)\n",
      "[90/391]\tTime 0.180 (0.183)\tData 0.019 (0.020)\tLoss 0.3284 (0.3738)\tPrec@1 93.750 (94.548)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3214 (0.3759)\tPrec@1 97.656 (94.454)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3112 (0.3751)\tPrec@1 96.094 (94.482)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3504 (0.3744)\tPrec@1 96.094 (94.493)\n",
      "[130/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3744 (0.3760)\tPrec@1 92.969 (94.460)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3954 (0.3776)\tPrec@1 93.750 (94.437)\n",
      "[150/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.3474 (0.3778)\tPrec@1 96.094 (94.464)\n",
      "[160/391]\tTime 0.189 (0.182)\tData 0.020 (0.020)\tLoss 0.3666 (0.3778)\tPrec@1 96.094 (94.483)\n",
      "[170/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.3843 (0.3772)\tPrec@1 92.188 (94.495)\n",
      "[180/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.3531 (0.3767)\tPrec@1 96.094 (94.536)\n",
      "[190/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.5248 (0.3772)\tPrec@1 87.500 (94.539)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4156 (0.3773)\tPrec@1 92.188 (94.508)\n",
      "[210/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3992 (0.3777)\tPrec@1 92.969 (94.502)\n",
      "[220/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3102 (0.3787)\tPrec@1 96.094 (94.432)\n",
      "[230/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4393 (0.3798)\tPrec@1 92.969 (94.382)\n",
      "[240/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3369 (0.3797)\tPrec@1 95.312 (94.376)\n",
      "[250/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3566 (0.3791)\tPrec@1 95.312 (94.413)\n",
      "[260/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.4099 (0.3793)\tPrec@1 92.969 (94.412)\n",
      "[270/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.3554 (0.3791)\tPrec@1 95.312 (94.413)\n",
      "[280/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3726 (0.3787)\tPrec@1 95.312 (94.439)\n",
      "[290/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3434 (0.3790)\tPrec@1 95.312 (94.413)\n",
      "[300/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4074 (0.3791)\tPrec@1 95.312 (94.412)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3859 (0.3796)\tPrec@1 91.406 (94.370)\n",
      "[320/391]\tTime 0.188 (0.182)\tData 0.020 (0.020)\tLoss 0.4275 (0.3799)\tPrec@1 90.625 (94.356)\n",
      "[330/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.4958 (0.3801)\tPrec@1 89.062 (94.342)\n",
      "[340/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.4126 (0.3799)\tPrec@1 95.312 (94.348)\n",
      "[350/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3645 (0.3797)\tPrec@1 95.312 (94.362)\n",
      "[360/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.4228 (0.3796)\tPrec@1 92.188 (94.371)\n",
      "[370/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3909 (0.3800)\tPrec@1 94.531 (94.356)\n",
      "[380/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3626 (0.3795)\tPrec@1 92.188 (94.355)\n",
      "[390/391]\tTime 0.094 (0.182)\tData 0.012 (0.019)\tLoss 0.3228 (0.3801)\tPrec@1 97.500 (94.328)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3371 (0.3371)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.3866 (0.4129)\tPrec@1 90.625 (92.188)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.4114 (0.4675)\tPrec@1 92.188 (91.257)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.4211 (0.4698)\tPrec@1 91.406 (91.003)\n",
      "Test: [40/79]\tTime 0.020 (0.019)\tLoss 0.6362 (0.4694)\tPrec@1 85.938 (90.987)\n",
      "Test: [50/79]\tTime 0.019 (0.019)\tLoss 0.5069 (0.4565)\tPrec@1 89.844 (91.115)\n",
      "Test: [60/79]\tTime 0.020 (0.019)\tLoss 0.3556 (0.4593)\tPrec@1 89.062 (91.073)\n",
      "Test: [70/79]\tTime 0.025 (0.020)\tLoss 0.5460 (0.4516)\tPrec@1 89.844 (91.153)\n",
      " * Prec@1 91.180\n",
      "\n",
      "===> epoch: 107/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.154 (0.154)\tData 0.021 (0.021)\tLoss 0.3789 (0.3789)\tPrec@1 94.531 (94.531)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.019 (0.019)\tLoss 0.3255 (0.3620)\tPrec@1 97.656 (95.099)\n",
      "[20/391]\tTime 0.180 (0.180)\tData 0.019 (0.019)\tLoss 0.4469 (0.3692)\tPrec@1 90.625 (94.606)\n",
      "[30/391]\tTime 0.185 (0.181)\tData 0.021 (0.019)\tLoss 0.3817 (0.3690)\tPrec@1 91.406 (94.582)\n",
      "[40/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.3666 (0.3736)\tPrec@1 94.531 (94.398)\n",
      "[50/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3898 (0.3728)\tPrec@1 93.750 (94.393)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[60/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.3501 (0.3681)\tPrec@1 96.094 (94.480)\n",
      "[70/391]\tTime 0.186 (0.182)\tData 0.019 (0.019)\tLoss 0.3919 (0.3687)\tPrec@1 94.531 (94.421)\n",
      "[80/391]\tTime 0.185 (0.182)\tData 0.020 (0.019)\tLoss 0.3260 (0.3678)\tPrec@1 96.094 (94.454)\n",
      "[90/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.3366 (0.3685)\tPrec@1 96.875 (94.488)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.019 (0.019)\tLoss 0.3846 (0.3678)\tPrec@1 94.531 (94.585)\n",
      "[110/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4518 (0.3671)\tPrec@1 89.844 (94.595)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.3550 (0.3661)\tPrec@1 94.531 (94.641)\n",
      "[130/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3848 (0.3665)\tPrec@1 94.531 (94.621)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.3223 (0.3676)\tPrec@1 96.875 (94.559)\n",
      "[150/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3756 (0.3683)\tPrec@1 92.188 (94.469)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.4006 (0.3678)\tPrec@1 92.969 (94.526)\n",
      "[170/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3488 (0.3679)\tPrec@1 95.312 (94.513)\n",
      "[180/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3986 (0.3692)\tPrec@1 91.406 (94.467)\n",
      "[190/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.3810 (0.3697)\tPrec@1 92.969 (94.429)\n",
      "[200/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3982 (0.3692)\tPrec@1 94.531 (94.465)\n",
      "[210/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3806 (0.3690)\tPrec@1 96.094 (94.505)\n",
      "[220/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.4236 (0.3690)\tPrec@1 93.750 (94.521)\n",
      "[230/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3841 (0.3698)\tPrec@1 95.312 (94.504)\n",
      "[240/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3580 (0.3703)\tPrec@1 96.094 (94.489)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3577 (0.3699)\tPrec@1 94.531 (94.519)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.4495 (0.3713)\tPrec@1 92.188 (94.471)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3932 (0.3718)\tPrec@1 92.969 (94.448)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.2906 (0.3720)\tPrec@1 97.656 (94.426)\n",
      "[290/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3634 (0.3727)\tPrec@1 96.875 (94.389)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.4014 (0.3720)\tPrec@1 94.531 (94.422)\n",
      "[310/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3432 (0.3720)\tPrec@1 94.531 (94.418)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4631 (0.3724)\tPrec@1 90.625 (94.410)\n",
      "[330/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3268 (0.3721)\tPrec@1 96.094 (94.406)\n",
      "[340/391]\tTime 0.188 (0.181)\tData 0.020 (0.020)\tLoss 0.3237 (0.3728)\tPrec@1 96.094 (94.378)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3440 (0.3729)\tPrec@1 95.312 (94.375)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3891 (0.3728)\tPrec@1 94.531 (94.371)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.3698 (0.3731)\tPrec@1 92.969 (94.356)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3296 (0.3742)\tPrec@1 96.875 (94.306)\n",
      "[390/391]\tTime 0.093 (0.181)\tData 0.013 (0.020)\tLoss 0.3208 (0.3741)\tPrec@1 98.750 (94.318)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2641 (0.2641)\tPrec@1 94.531 (94.531)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.3440 (0.3973)\tPrec@1 93.750 (91.619)\n",
      "Test: [20/79]\tTime 0.019 (0.019)\tLoss 0.4120 (0.4732)\tPrec@1 92.969 (90.662)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.3686 (0.4717)\tPrec@1 90.625 (90.675)\n",
      "Test: [40/79]\tTime 0.020 (0.019)\tLoss 0.4434 (0.4620)\tPrec@1 89.844 (90.930)\n",
      "Test: [50/79]\tTime 0.018 (0.019)\tLoss 0.4942 (0.4524)\tPrec@1 90.625 (91.115)\n",
      "Test: [60/79]\tTime 0.025 (0.020)\tLoss 0.3375 (0.4479)\tPrec@1 91.406 (91.163)\n",
      "Test: [70/79]\tTime 0.021 (0.020)\tLoss 0.6007 (0.4421)\tPrec@1 89.062 (91.131)\n",
      " * Prec@1 91.250\n",
      "\n",
      "===> epoch: 108/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.170 (0.170)\tData 0.022 (0.022)\tLoss 0.3808 (0.3808)\tPrec@1 93.750 (93.750)\n",
      "[10/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3677 (0.3810)\tPrec@1 94.531 (94.389)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.4089 (0.3714)\tPrec@1 92.188 (94.568)\n",
      "[30/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3715 (0.3637)\tPrec@1 94.531 (94.960)\n",
      "[40/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3637 (0.3660)\tPrec@1 94.531 (94.779)\n",
      "[50/391]\tTime 0.184 (0.181)\tData 0.022 (0.020)\tLoss 0.4619 (0.3669)\tPrec@1 90.625 (94.730)\n",
      "[60/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3567 (0.3686)\tPrec@1 96.875 (94.723)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.2909 (0.3675)\tPrec@1 97.656 (94.718)\n",
      "[80/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.3338 (0.3687)\tPrec@1 95.312 (94.676)\n",
      "[90/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3927 (0.3690)\tPrec@1 92.969 (94.669)\n",
      "[100/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3933 (0.3713)\tPrec@1 93.750 (94.578)\n",
      "[110/391]\tTime 0.188 (0.182)\tData 0.031 (0.020)\tLoss 0.4479 (0.3707)\tPrec@1 92.969 (94.651)\n",
      "[120/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.4027 (0.3709)\tPrec@1 92.188 (94.622)\n",
      "[130/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3546 (0.3705)\tPrec@1 95.312 (94.686)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3169 (0.3711)\tPrec@1 97.656 (94.664)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4099 (0.3707)\tPrec@1 92.969 (94.655)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3344 (0.3702)\tPrec@1 96.875 (94.677)\n",
      "[170/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3943 (0.3698)\tPrec@1 93.750 (94.696)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4016 (0.3705)\tPrec@1 92.969 (94.648)\n",
      "[190/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3372 (0.3706)\tPrec@1 96.875 (94.646)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3734 (0.3712)\tPrec@1 94.531 (94.617)\n",
      "[210/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3615 (0.3705)\tPrec@1 94.531 (94.642)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3910 (0.3708)\tPrec@1 94.531 (94.609)\n",
      "[230/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3798 (0.3707)\tPrec@1 96.094 (94.636)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3924 (0.3713)\tPrec@1 94.531 (94.612)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.3731 (0.3715)\tPrec@1 94.531 (94.581)\n",
      "[260/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3531 (0.3714)\tPrec@1 95.312 (94.582)\n",
      "[270/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.3973 (0.3718)\tPrec@1 92.969 (94.566)\n",
      "[280/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3461 (0.3718)\tPrec@1 95.312 (94.559)\n",
      "[290/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4370 (0.3716)\tPrec@1 91.406 (94.563)\n",
      "[300/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3486 (0.3714)\tPrec@1 96.875 (94.591)\n",
      "[310/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3816 (0.3713)\tPrec@1 96.094 (94.592)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3799 (0.3710)\tPrec@1 93.750 (94.599)\n",
      "[330/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4006 (0.3711)\tPrec@1 93.750 (94.590)\n",
      "[340/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3383 (0.3715)\tPrec@1 95.312 (94.579)\n",
      "[350/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3722 (0.3715)\tPrec@1 94.531 (94.582)\n",
      "[360/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3202 (0.3716)\tPrec@1 96.875 (94.579)\n",
      "[370/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3774 (0.3715)\tPrec@1 92.969 (94.573)\n",
      "[380/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3020 (0.3712)\tPrec@1 97.656 (94.591)\n",
      "[390/391]\tTime 0.092 (0.181)\tData 0.013 (0.020)\tLoss 0.3711 (0.3712)\tPrec@1 95.000 (94.602)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3238 (0.3238)\tPrec@1 92.969 (92.969)\n",
      "Test: [10/79]\tTime 0.020 (0.020)\tLoss 0.3577 (0.3787)\tPrec@1 92.969 (91.690)\n",
      "Test: [20/79]\tTime 0.019 (0.020)\tLoss 0.4123 (0.4301)\tPrec@1 92.969 (91.443)\n",
      "Test: [30/79]\tTime 0.019 (0.019)\tLoss 0.4263 (0.4361)\tPrec@1 93.750 (91.205)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.5646 (0.4326)\tPrec@1 85.156 (91.292)\n",
      "Test: [50/79]\tTime 0.026 (0.020)\tLoss 0.4605 (0.4245)\tPrec@1 90.625 (91.437)\n",
      "Test: [60/79]\tTime 0.025 (0.020)\tLoss 0.2520 (0.4171)\tPrec@1 92.969 (91.573)\n",
      "Test: [70/79]\tTime 0.019 (0.021)\tLoss 0.5043 (0.4149)\tPrec@1 89.062 (91.593)\n",
      " * Prec@1 91.660\n",
      "\n",
      "===> epoch: 109/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.022 (0.022)\tLoss 0.4225 (0.4225)\tPrec@1 92.188 (92.188)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[10/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.3469 (0.3569)\tPrec@1 96.094 (95.455)\n",
      "[20/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.2787 (0.3536)\tPrec@1 99.219 (95.387)\n",
      "[30/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3382 (0.3538)\tPrec@1 95.312 (95.388)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3388 (0.3586)\tPrec@1 96.094 (95.065)\n",
      "[50/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3919 (0.3591)\tPrec@1 95.312 (95.113)\n",
      "[60/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3462 (0.3578)\tPrec@1 96.094 (95.274)\n",
      "[70/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4322 (0.3624)\tPrec@1 91.406 (95.136)\n",
      "[80/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.4354 (0.3625)\tPrec@1 92.188 (95.052)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4279 (0.3643)\tPrec@1 91.406 (95.012)\n",
      "[100/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4367 (0.3655)\tPrec@1 92.188 (94.957)\n",
      "[110/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3464 (0.3635)\tPrec@1 95.312 (95.017)\n",
      "[120/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3212 (0.3642)\tPrec@1 96.094 (94.938)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3422 (0.3635)\tPrec@1 96.094 (94.955)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.4290 (0.3633)\tPrec@1 90.625 (94.936)\n",
      "[150/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3167 (0.3641)\tPrec@1 96.875 (94.893)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3909 (0.3648)\tPrec@1 92.188 (94.847)\n",
      "[170/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3716 (0.3646)\tPrec@1 96.094 (94.851)\n",
      "[180/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4645 (0.3650)\tPrec@1 89.062 (94.803)\n",
      "[190/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.4139 (0.3651)\tPrec@1 92.969 (94.813)\n",
      "[200/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3243 (0.3642)\tPrec@1 97.656 (94.846)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.3671 (0.3637)\tPrec@1 95.312 (94.864)\n",
      "[220/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.4547 (0.3635)\tPrec@1 93.750 (94.906)\n",
      "[230/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.3730 (0.3651)\tPrec@1 94.531 (94.829)\n",
      "[240/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3903 (0.3650)\tPrec@1 96.094 (94.849)\n",
      "[250/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3306 (0.3646)\tPrec@1 97.656 (94.895)\n",
      "[260/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3850 (0.3643)\tPrec@1 94.531 (94.890)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3622 (0.3642)\tPrec@1 96.875 (94.906)\n",
      "[280/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3792 (0.3643)\tPrec@1 95.312 (94.907)\n",
      "[290/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.3401 (0.3641)\tPrec@1 95.312 (94.899)\n",
      "[300/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3815 (0.3641)\tPrec@1 98.438 (94.915)\n",
      "[310/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3914 (0.3652)\tPrec@1 92.969 (94.898)\n",
      "[320/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.3152 (0.3652)\tPrec@1 98.438 (94.899)\n",
      "[330/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3688 (0.3658)\tPrec@1 94.531 (94.852)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3694 (0.3659)\tPrec@1 93.750 (94.847)\n",
      "[350/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3591 (0.3663)\tPrec@1 94.531 (94.841)\n",
      "[360/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4115 (0.3671)\tPrec@1 92.969 (94.793)\n",
      "[370/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3492 (0.3671)\tPrec@1 96.875 (94.788)\n",
      "[380/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3251 (0.3672)\tPrec@1 97.656 (94.783)\n",
      "[390/391]\tTime 0.095 (0.180)\tData 0.013 (0.020)\tLoss 0.3936 (0.3676)\tPrec@1 91.250 (94.768)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2877 (0.2877)\tPrec@1 95.312 (95.312)\n",
      "Test: [10/79]\tTime 0.020 (0.019)\tLoss 0.3716 (0.4068)\tPrec@1 92.969 (91.761)\n",
      "Test: [20/79]\tTime 0.020 (0.019)\tLoss 0.3742 (0.4550)\tPrec@1 92.969 (91.220)\n",
      "Test: [30/79]\tTime 0.026 (0.020)\tLoss 0.3955 (0.4653)\tPrec@1 92.969 (91.205)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.6715 (0.4642)\tPrec@1 84.375 (91.159)\n",
      "Test: [50/79]\tTime 0.026 (0.021)\tLoss 0.5399 (0.4570)\tPrec@1 89.062 (91.161)\n",
      "Test: [60/79]\tTime 0.020 (0.022)\tLoss 0.5028 (0.4552)\tPrec@1 90.625 (91.227)\n",
      "Test: [70/79]\tTime 0.026 (0.022)\tLoss 0.5868 (0.4493)\tPrec@1 89.844 (91.296)\n",
      " * Prec@1 91.380\n",
      "\n",
      "===> epoch: 110/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.187 (0.187)\tData 0.021 (0.021)\tLoss 0.3104 (0.3104)\tPrec@1 96.875 (96.875)\n",
      "[10/391]\tTime 0.185 (0.185)\tData 0.019 (0.020)\tLoss 0.2833 (0.3324)\tPrec@1 98.438 (96.307)\n",
      "[20/391]\tTime 0.181 (0.183)\tData 0.020 (0.020)\tLoss 0.3733 (0.3445)\tPrec@1 93.750 (95.833)\n",
      "[30/391]\tTime 0.181 (0.183)\tData 0.020 (0.020)\tLoss 0.4262 (0.3522)\tPrec@1 92.188 (95.186)\n",
      "[40/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3842 (0.3518)\tPrec@1 93.750 (95.236)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.3856 (0.3537)\tPrec@1 92.969 (95.251)\n",
      "[60/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.4092 (0.3592)\tPrec@1 93.750 (95.146)\n",
      "[70/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.4148 (0.3620)\tPrec@1 93.750 (95.070)\n",
      "[80/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3880 (0.3631)\tPrec@1 92.969 (95.091)\n",
      "[90/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3918 (0.3640)\tPrec@1 92.188 (95.021)\n",
      "[100/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3687 (0.3640)\tPrec@1 93.750 (94.972)\n",
      "[110/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3581 (0.3648)\tPrec@1 92.969 (94.904)\n",
      "[120/391]\tTime 0.180 (0.182)\tData 0.021 (0.020)\tLoss 0.3566 (0.3634)\tPrec@1 96.875 (94.977)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3418 (0.3634)\tPrec@1 96.094 (94.996)\n",
      "[140/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.4108 (0.3636)\tPrec@1 93.750 (95.002)\n",
      "[150/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3551 (0.3635)\tPrec@1 96.094 (94.976)\n",
      "[160/391]\tTime 0.177 (0.182)\tData 0.019 (0.020)\tLoss 0.3516 (0.3626)\tPrec@1 96.094 (95.021)\n",
      "[170/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.3563 (0.3631)\tPrec@1 94.531 (94.988)\n",
      "[180/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3315 (0.3632)\tPrec@1 95.312 (94.972)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3482 (0.3623)\tPrec@1 96.094 (95.022)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3517 (0.3621)\tPrec@1 96.094 (95.033)\n",
      "[210/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.3328 (0.3617)\tPrec@1 95.312 (95.061)\n",
      "[220/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3619 (0.3622)\tPrec@1 96.094 (95.047)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3569 (0.3628)\tPrec@1 96.875 (95.028)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3250 (0.3625)\tPrec@1 96.875 (95.069)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3575 (0.3621)\tPrec@1 94.531 (95.067)\n",
      "[260/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.4129 (0.3616)\tPrec@1 92.969 (95.082)\n",
      "[270/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.2958 (0.3612)\tPrec@1 97.656 (95.088)\n",
      "[280/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3557 (0.3618)\tPrec@1 96.094 (95.054)\n",
      "[290/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3607 (0.3613)\tPrec@1 96.875 (95.074)\n",
      "[300/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3331 (0.3607)\tPrec@1 96.094 (95.110)\n",
      "[310/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3726 (0.3607)\tPrec@1 92.969 (95.104)\n",
      "[320/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3187 (0.3609)\tPrec@1 95.312 (95.084)\n",
      "[330/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3469 (0.3611)\tPrec@1 95.312 (95.060)\n",
      "[340/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.3261 (0.3615)\tPrec@1 98.438 (95.056)\n",
      "[350/391]\tTime 0.177 (0.181)\tData 0.020 (0.020)\tLoss 0.3968 (0.3623)\tPrec@1 91.406 (95.003)\n",
      "[360/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4060 (0.3622)\tPrec@1 91.406 (95.001)\n",
      "[370/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3248 (0.3625)\tPrec@1 96.875 (95.001)\n",
      "[380/391]\tTime 0.145 (0.181)\tData 0.020 (0.020)\tLoss 0.3433 (0.3628)\tPrec@1 96.875 (94.997)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.015 (0.020)\tLoss 0.3166 (0.3626)\tPrec@1 98.750 (95.016)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3491 (0.3491)\tPrec@1 94.531 (94.531)\n",
      "Test: [10/79]\tTime 0.020 (0.023)\tLoss 0.3329 (0.4010)\tPrec@1 91.406 (92.116)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.3682 (0.4496)\tPrec@1 90.625 (91.443)\n",
      "Test: [30/79]\tTime 0.024 (0.023)\tLoss 0.4493 (0.4658)\tPrec@1 92.188 (91.280)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.026 (0.023)\tLoss 0.5706 (0.4572)\tPrec@1 89.844 (91.387)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 0.4567 (0.4422)\tPrec@1 92.188 (91.498)\n",
      "Test: [60/79]\tTime 0.021 (0.023)\tLoss 0.3079 (0.4343)\tPrec@1 93.750 (91.637)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 0.4991 (0.4292)\tPrec@1 92.188 (91.659)\n",
      " * Prec@1 91.660\n",
      "\n",
      "===> epoch: 111/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.182 (0.182)\tData 0.021 (0.021)\tLoss 0.3600 (0.3600)\tPrec@1 95.312 (95.312)\n",
      "[10/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.3359 (0.3484)\tPrec@1 96.094 (96.023)\n",
      "[20/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.4139 (0.3560)\tPrec@1 90.625 (95.499)\n",
      "[30/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.3624 (0.3590)\tPrec@1 92.969 (95.338)\n",
      "[40/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3370 (0.3561)\tPrec@1 95.312 (95.427)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3874 (0.3570)\tPrec@1 92.188 (95.205)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3874 (0.3579)\tPrec@1 92.188 (95.133)\n",
      "[70/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.4643 (0.3589)\tPrec@1 92.188 (95.114)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3537 (0.3593)\tPrec@1 94.531 (95.139)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3402 (0.3580)\tPrec@1 96.094 (95.132)\n",
      "[100/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3739 (0.3592)\tPrec@1 94.531 (95.111)\n",
      "[110/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3537 (0.3583)\tPrec@1 96.094 (95.186)\n",
      "[120/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.3690 (0.3595)\tPrec@1 93.750 (95.132)\n",
      "[130/391]\tTime 0.190 (0.181)\tData 0.019 (0.020)\tLoss 0.3346 (0.3578)\tPrec@1 95.312 (95.235)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3344 (0.3586)\tPrec@1 96.875 (95.196)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3654 (0.3598)\tPrec@1 96.094 (95.162)\n",
      "[160/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3589 (0.3607)\tPrec@1 96.094 (95.152)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.3015 (0.3606)\tPrec@1 98.438 (95.162)\n",
      "[180/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3973 (0.3609)\tPrec@1 95.312 (95.148)\n",
      "[190/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3192 (0.3607)\tPrec@1 96.875 (95.137)\n",
      "[200/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3583 (0.3613)\tPrec@1 94.531 (95.095)\n",
      "[210/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3116 (0.3616)\tPrec@1 98.438 (95.072)\n",
      "[220/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.4060 (0.3625)\tPrec@1 94.531 (95.044)\n",
      "[230/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.3415 (0.3628)\tPrec@1 97.656 (95.028)\n",
      "[240/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3613 (0.3627)\tPrec@1 94.531 (95.034)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3627 (0.3628)\tPrec@1 94.531 (95.032)\n",
      "[260/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3849 (0.3621)\tPrec@1 94.531 (95.079)\n",
      "[270/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3298 (0.3615)\tPrec@1 95.312 (95.096)\n",
      "[280/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3312 (0.3616)\tPrec@1 96.875 (95.098)\n",
      "[290/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3842 (0.3614)\tPrec@1 91.406 (95.095)\n",
      "[300/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4419 (0.3612)\tPrec@1 92.188 (95.113)\n",
      "[310/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3349 (0.3611)\tPrec@1 97.656 (95.112)\n",
      "[320/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3502 (0.3618)\tPrec@1 93.750 (95.074)\n",
      "[330/391]\tTime 0.190 (0.182)\tData 0.019 (0.020)\tLoss 0.4269 (0.3623)\tPrec@1 94.531 (95.055)\n",
      "[340/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3163 (0.3619)\tPrec@1 96.094 (95.088)\n",
      "[350/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3462 (0.3619)\tPrec@1 96.094 (95.079)\n",
      "[360/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3564 (0.3616)\tPrec@1 95.312 (95.092)\n",
      "[370/391]\tTime 0.143 (0.181)\tData 0.019 (0.020)\tLoss 0.4417 (0.3621)\tPrec@1 93.750 (95.083)\n",
      "[380/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.3694 (0.3618)\tPrec@1 93.750 (95.087)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.3799 (0.3624)\tPrec@1 95.000 (95.058)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3522 (0.3522)\tPrec@1 92.969 (92.969)\n",
      "Test: [10/79]\tTime 0.020 (0.023)\tLoss 0.3929 (0.3970)\tPrec@1 92.969 (91.619)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.3701 (0.4463)\tPrec@1 92.969 (91.220)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.4209 (0.4477)\tPrec@1 92.188 (91.028)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 0.5426 (0.4457)\tPrec@1 87.500 (91.159)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 0.3924 (0.4310)\tPrec@1 90.625 (91.360)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.3614 (0.4265)\tPrec@1 91.406 (91.432)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 0.5514 (0.4239)\tPrec@1 90.625 (91.505)\n",
      " * Prec@1 91.610\n",
      "\n",
      "===> epoch: 112/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.180 (0.180)\tData 0.021 (0.021)\tLoss 0.3484 (0.3484)\tPrec@1 94.531 (94.531)\n",
      "[10/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3883 (0.3573)\tPrec@1 94.531 (94.602)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3440 (0.3592)\tPrec@1 96.875 (94.829)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3822 (0.3592)\tPrec@1 93.750 (94.884)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.3274 (0.3606)\tPrec@1 95.312 (94.912)\n",
      "[50/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.4008 (0.3594)\tPrec@1 92.969 (95.037)\n",
      "[60/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.3603 (0.3581)\tPrec@1 93.750 (95.159)\n",
      "[70/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.3340 (0.3576)\tPrec@1 95.312 (95.191)\n",
      "[80/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3483 (0.3580)\tPrec@1 95.312 (95.168)\n",
      "[90/391]\tTime 0.177 (0.180)\tData 0.019 (0.020)\tLoss 0.3879 (0.3583)\tPrec@1 93.750 (95.141)\n",
      "[100/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.4236 (0.3596)\tPrec@1 91.406 (95.073)\n",
      "[110/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.3416 (0.3579)\tPrec@1 95.312 (95.179)\n",
      "[120/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.3804 (0.3576)\tPrec@1 92.188 (95.125)\n",
      "[130/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3133 (0.3574)\tPrec@1 97.656 (95.134)\n",
      "[140/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3408 (0.3572)\tPrec@1 95.312 (95.119)\n",
      "[150/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.4178 (0.3578)\tPrec@1 94.531 (95.111)\n",
      "[160/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.3708 (0.3576)\tPrec@1 95.312 (95.118)\n",
      "[170/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.3876 (0.3576)\tPrec@1 93.750 (95.139)\n",
      "[180/391]\tTime 0.182 (0.181)\tData 0.024 (0.020)\tLoss 0.3595 (0.3581)\tPrec@1 94.531 (95.075)\n",
      "[190/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3398 (0.3571)\tPrec@1 96.875 (95.096)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3946 (0.3572)\tPrec@1 93.750 (95.087)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3316 (0.3583)\tPrec@1 96.094 (95.039)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3617 (0.3588)\tPrec@1 94.531 (95.040)\n",
      "[230/391]\tTime 0.186 (0.181)\tData 0.023 (0.020)\tLoss 0.3570 (0.3589)\tPrec@1 96.094 (95.015)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3871 (0.3594)\tPrec@1 92.969 (94.995)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3648 (0.3597)\tPrec@1 96.875 (94.979)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.4087 (0.3596)\tPrec@1 91.406 (94.971)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3434 (0.3595)\tPrec@1 96.094 (94.984)\n",
      "[280/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3624 (0.3591)\tPrec@1 95.312 (95.009)\n",
      "[290/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3021 (0.3593)\tPrec@1 97.656 (95.028)\n",
      "[300/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3452 (0.3595)\tPrec@1 96.094 (95.027)\n",
      "[310/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3496 (0.3598)\tPrec@1 96.875 (95.026)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.4357 (0.3597)\tPrec@1 92.188 (95.040)\n",
      "[330/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4135 (0.3598)\tPrec@1 92.188 (95.027)\n",
      "[340/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3601 (0.3601)\tPrec@1 95.312 (95.010)\n",
      "[350/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.3854 (0.3605)\tPrec@1 92.969 (94.974)\n",
      "[360/391]\tTime 0.144 (0.180)\tData 0.020 (0.020)\tLoss 0.3552 (0.3603)\tPrec@1 95.312 (94.981)\n",
      "[370/391]\tTime 0.145 (0.179)\tData 0.020 (0.020)\tLoss 0.3641 (0.3608)\tPrec@1 96.875 (94.952)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.4353 (0.3607)\tPrec@1 90.625 (94.937)\n",
      "[390/391]\tTime 0.129 (0.179)\tData 0.012 (0.020)\tLoss 0.4098 (0.3606)\tPrec@1 92.500 (94.934)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4057 (0.4057)\tPrec@1 91.406 (91.406)\n",
      "Test: [10/79]\tTime 0.026 (0.022)\tLoss 0.3479 (0.4321)\tPrec@1 92.969 (91.619)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.3446 (0.4769)\tPrec@1 93.750 (91.109)\n",
      "Test: [30/79]\tTime 0.019 (0.023)\tLoss 0.4534 (0.4750)\tPrec@1 91.406 (91.003)\n",
      "Test: [40/79]\tTime 0.024 (0.023)\tLoss 0.6858 (0.4723)\tPrec@1 85.938 (90.911)\n",
      "Test: [50/79]\tTime 0.025 (0.023)\tLoss 0.4792 (0.4592)\tPrec@1 89.062 (90.962)\n",
      "Test: [60/79]\tTime 0.020 (0.023)\tLoss 0.3833 (0.4553)\tPrec@1 92.969 (91.112)\n",
      "Test: [70/79]\tTime 0.023 (0.023)\tLoss 0.4584 (0.4492)\tPrec@1 92.188 (91.142)\n",
      " * Prec@1 91.240\n",
      "\n",
      "===> epoch: 113/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.021 (0.021)\tLoss 0.3559 (0.3559)\tPrec@1 95.312 (95.312)\n",
      "[10/391]\tTime 0.180 (0.184)\tData 0.020 (0.020)\tLoss 0.3689 (0.3647)\tPrec@1 95.312 (94.957)\n",
      "[20/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3291 (0.3526)\tPrec@1 97.656 (95.499)\n",
      "[30/391]\tTime 0.181 (0.183)\tData 0.019 (0.020)\tLoss 0.2997 (0.3488)\tPrec@1 97.656 (95.539)\n",
      "[40/391]\tTime 0.184 (0.183)\tData 0.020 (0.020)\tLoss 0.3536 (0.3502)\tPrec@1 95.312 (95.446)\n",
      "[50/391]\tTime 0.180 (0.183)\tData 0.019 (0.020)\tLoss 0.3362 (0.3484)\tPrec@1 94.531 (95.650)\n",
      "[60/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3649 (0.3523)\tPrec@1 92.969 (95.466)\n",
      "[70/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3361 (0.3518)\tPrec@1 96.094 (95.467)\n",
      "[80/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.3277 (0.3512)\tPrec@1 96.875 (95.476)\n",
      "[90/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.3571 (0.3522)\tPrec@1 95.312 (95.381)\n",
      "[100/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.3736 (0.3530)\tPrec@1 93.750 (95.282)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.3150 (0.3540)\tPrec@1 97.656 (95.270)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3409 (0.3550)\tPrec@1 96.094 (95.241)\n",
      "[130/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3341 (0.3550)\tPrec@1 96.094 (95.217)\n",
      "[140/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.3375 (0.3544)\tPrec@1 96.875 (95.274)\n",
      "[150/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.3355 (0.3547)\tPrec@1 95.312 (95.230)\n",
      "[160/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3520 (0.3556)\tPrec@1 95.312 (95.186)\n",
      "[170/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3339 (0.3549)\tPrec@1 94.531 (95.203)\n",
      "[180/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.3721 (0.3553)\tPrec@1 95.312 (95.187)\n",
      "[190/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3463 (0.3553)\tPrec@1 94.531 (95.227)\n",
      "[200/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3558 (0.3551)\tPrec@1 96.875 (95.266)\n",
      "[210/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.4034 (0.3552)\tPrec@1 94.531 (95.287)\n",
      "[220/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.3602 (0.3555)\tPrec@1 95.312 (95.263)\n",
      "[230/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3364 (0.3563)\tPrec@1 95.312 (95.228)\n",
      "[240/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.3310 (0.3565)\tPrec@1 96.875 (95.206)\n",
      "[250/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3340 (0.3565)\tPrec@1 96.094 (95.213)\n",
      "[260/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3752 (0.3567)\tPrec@1 93.750 (95.229)\n",
      "[270/391]\tTime 0.187 (0.182)\tData 0.024 (0.020)\tLoss 0.4654 (0.3574)\tPrec@1 87.500 (95.171)\n",
      "[280/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3704 (0.3571)\tPrec@1 93.750 (95.176)\n",
      "[290/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3203 (0.3567)\tPrec@1 96.094 (95.192)\n",
      "[300/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3945 (0.3567)\tPrec@1 94.531 (95.196)\n",
      "[310/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3668 (0.3568)\tPrec@1 96.094 (95.187)\n",
      "[320/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3749 (0.3570)\tPrec@1 93.750 (95.176)\n",
      "[330/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3551 (0.3575)\tPrec@1 96.094 (95.150)\n",
      "[340/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3865 (0.3575)\tPrec@1 90.625 (95.134)\n",
      "[350/391]\tTime 0.145 (0.182)\tData 0.019 (0.020)\tLoss 0.3679 (0.3580)\tPrec@1 92.188 (95.132)\n",
      "[360/391]\tTime 0.144 (0.180)\tData 0.020 (0.020)\tLoss 0.3746 (0.3583)\tPrec@1 92.969 (95.118)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3145 (0.3580)\tPrec@1 96.875 (95.136)\n",
      "[380/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.3825 (0.3579)\tPrec@1 94.531 (95.132)\n",
      "[390/391]\tTime 0.130 (0.180)\tData 0.012 (0.020)\tLoss 0.3784 (0.3578)\tPrec@1 92.500 (95.122)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2574 (0.2574)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 0.3574 (0.4320)\tPrec@1 93.750 (91.619)\n",
      "Test: [20/79]\tTime 0.024 (0.022)\tLoss 0.3364 (0.4601)\tPrec@1 93.750 (91.332)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 0.4555 (0.4702)\tPrec@1 90.625 (91.053)\n",
      "Test: [40/79]\tTime 0.022 (0.022)\tLoss 0.4818 (0.4603)\tPrec@1 86.719 (91.082)\n",
      "Test: [50/79]\tTime 0.026 (0.022)\tLoss 0.5630 (0.4526)\tPrec@1 89.844 (91.360)\n",
      "Test: [60/79]\tTime 0.019 (0.022)\tLoss 0.3512 (0.4525)\tPrec@1 92.969 (91.214)\n",
      "Test: [70/79]\tTime 0.021 (0.022)\tLoss 0.5517 (0.4539)\tPrec@1 91.406 (91.175)\n",
      " * Prec@1 91.220\n",
      "\n",
      "===> epoch: 114/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.021 (0.021)\tLoss 0.3388 (0.3388)\tPrec@1 95.312 (95.312)\n",
      "[10/391]\tTime 0.184 (0.184)\tData 0.019 (0.020)\tLoss 0.3280 (0.3308)\tPrec@1 97.656 (97.088)\n",
      "[20/391]\tTime 0.181 (0.183)\tData 0.019 (0.020)\tLoss 0.2978 (0.3341)\tPrec@1 98.438 (96.577)\n",
      "[30/391]\tTime 0.181 (0.182)\tData 0.021 (0.020)\tLoss 0.3410 (0.3388)\tPrec@1 96.094 (96.245)\n",
      "[40/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3050 (0.3390)\tPrec@1 95.312 (96.170)\n",
      "[50/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.3340 (0.3436)\tPrec@1 96.094 (95.956)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3112 (0.3461)\tPrec@1 96.094 (95.761)\n",
      "[70/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3408 (0.3457)\tPrec@1 96.094 (95.775)\n",
      "[80/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3116 (0.3466)\tPrec@1 96.875 (95.795)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3857 (0.3463)\tPrec@1 94.531 (95.707)\n",
      "[100/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3050 (0.3487)\tPrec@1 96.094 (95.614)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3805 (0.3484)\tPrec@1 93.750 (95.650)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3339 (0.3482)\tPrec@1 96.875 (95.674)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3229 (0.3492)\tPrec@1 97.656 (95.587)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3533 (0.3495)\tPrec@1 94.531 (95.590)\n",
      "[150/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.3102 (0.3503)\tPrec@1 96.094 (95.514)\n",
      "[160/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3297 (0.3523)\tPrec@1 96.094 (95.448)\n",
      "[170/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3262 (0.3526)\tPrec@1 96.094 (95.459)\n",
      "[180/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3069 (0.3525)\tPrec@1 96.875 (95.451)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.3623 (0.3524)\tPrec@1 94.531 (95.452)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3446 (0.3520)\tPrec@1 95.312 (95.468)\n",
      "[210/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3917 (0.3523)\tPrec@1 93.750 (95.450)\n",
      "[220/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3134 (0.3524)\tPrec@1 95.312 (95.440)\n",
      "[230/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3366 (0.3534)\tPrec@1 95.312 (95.411)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.021 (0.020)\tLoss 0.3372 (0.3536)\tPrec@1 96.094 (95.419)\n",
      "[250/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3519 (0.3538)\tPrec@1 94.531 (95.397)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3669 (0.3540)\tPrec@1 95.312 (95.369)\n",
      "[270/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3640 (0.3537)\tPrec@1 96.875 (95.405)\n",
      "[280/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3451 (0.3539)\tPrec@1 96.875 (95.404)\n",
      "[290/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.4272 (0.3540)\tPrec@1 91.406 (95.409)\n",
      "[300/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.2991 (0.3540)\tPrec@1 96.875 (95.398)\n",
      "[310/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3249 (0.3540)\tPrec@1 99.219 (95.415)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3413 (0.3538)\tPrec@1 96.094 (95.420)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[330/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3466 (0.3538)\tPrec@1 95.312 (95.416)\n",
      "[340/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3909 (0.3536)\tPrec@1 92.969 (95.420)\n",
      "[350/391]\tTime 0.143 (0.180)\tData 0.019 (0.020)\tLoss 0.3906 (0.3538)\tPrec@1 93.750 (95.390)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3785 (0.3541)\tPrec@1 92.188 (95.364)\n",
      "[370/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3530 (0.3545)\tPrec@1 97.656 (95.344)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.3192 (0.3547)\tPrec@1 96.094 (95.317)\n",
      "[390/391]\tTime 0.129 (0.180)\tData 0.012 (0.020)\tLoss 0.3276 (0.3549)\tPrec@1 96.250 (95.300)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2507 (0.2507)\tPrec@1 94.531 (94.531)\n",
      "Test: [10/79]\tTime 0.021 (0.023)\tLoss 0.4682 (0.4188)\tPrec@1 90.625 (91.548)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.4008 (0.4549)\tPrec@1 92.188 (91.406)\n",
      "Test: [30/79]\tTime 0.025 (0.023)\tLoss 0.3104 (0.4521)\tPrec@1 93.750 (91.482)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 0.8100 (0.4599)\tPrec@1 85.156 (91.463)\n",
      "Test: [50/79]\tTime 0.030 (0.023)\tLoss 0.4304 (0.4466)\tPrec@1 90.625 (91.682)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.4573 (0.4465)\tPrec@1 89.844 (91.573)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 0.5178 (0.4413)\tPrec@1 89.844 (91.593)\n",
      " * Prec@1 91.650\n",
      "\n",
      "===> epoch: 115/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.152 (0.152)\tData 0.020 (0.020)\tLoss 0.2994 (0.2994)\tPrec@1 99.219 (99.219)\n",
      "[10/391]\tTime 0.182 (0.178)\tData 0.020 (0.020)\tLoss 0.3696 (0.3571)\tPrec@1 96.875 (95.384)\n",
      "[20/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.3755 (0.3531)\tPrec@1 95.312 (95.833)\n",
      "[30/391]\tTime 0.183 (0.180)\tData 0.023 (0.020)\tLoss 0.3391 (0.3511)\tPrec@1 93.750 (95.615)\n",
      "[40/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3448 (0.3543)\tPrec@1 97.656 (95.503)\n",
      "[50/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3260 (0.3515)\tPrec@1 97.656 (95.619)\n",
      "[60/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3246 (0.3515)\tPrec@1 95.312 (95.530)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3986 (0.3513)\tPrec@1 92.188 (95.434)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4219 (0.3539)\tPrec@1 93.750 (95.341)\n",
      "[90/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3760 (0.3525)\tPrec@1 96.094 (95.381)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3393 (0.3511)\tPrec@1 96.094 (95.452)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3284 (0.3504)\tPrec@1 96.875 (95.524)\n",
      "[120/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3245 (0.3502)\tPrec@1 97.656 (95.532)\n",
      "[130/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.3104 (0.3510)\tPrec@1 96.875 (95.533)\n",
      "[140/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3500 (0.3500)\tPrec@1 94.531 (95.529)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3694 (0.3510)\tPrec@1 92.969 (95.509)\n",
      "[160/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3713 (0.3511)\tPrec@1 96.094 (95.497)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3365 (0.3516)\tPrec@1 95.312 (95.491)\n",
      "[180/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3668 (0.3509)\tPrec@1 95.312 (95.507)\n",
      "[190/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.3813 (0.3509)\tPrec@1 95.312 (95.501)\n",
      "[200/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.4277 (0.3519)\tPrec@1 92.188 (95.456)\n",
      "[210/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3848 (0.3519)\tPrec@1 93.750 (95.468)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3193 (0.3528)\tPrec@1 98.438 (95.447)\n",
      "[230/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3350 (0.3522)\tPrec@1 96.875 (95.485)\n",
      "[240/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3752 (0.3517)\tPrec@1 94.531 (95.494)\n",
      "[250/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3774 (0.3518)\tPrec@1 96.094 (95.515)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3517 (0.3520)\tPrec@1 96.094 (95.522)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3834 (0.3533)\tPrec@1 94.531 (95.460)\n",
      "[280/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3560 (0.3537)\tPrec@1 94.531 (95.457)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4009 (0.3541)\tPrec@1 94.531 (95.431)\n",
      "[300/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3077 (0.3545)\tPrec@1 96.094 (95.393)\n",
      "[310/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3403 (0.3541)\tPrec@1 96.094 (95.413)\n",
      "[320/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3695 (0.3543)\tPrec@1 95.312 (95.390)\n",
      "[330/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3073 (0.3546)\tPrec@1 98.438 (95.372)\n",
      "[340/391]\tTime 0.147 (0.180)\tData 0.020 (0.020)\tLoss 0.3458 (0.3543)\tPrec@1 94.531 (95.388)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3540 (0.3544)\tPrec@1 95.312 (95.379)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3631 (0.3545)\tPrec@1 93.750 (95.362)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.3033 (0.3548)\tPrec@1 96.094 (95.329)\n",
      "[380/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.3358 (0.3548)\tPrec@1 96.875 (95.323)\n",
      "[390/391]\tTime 0.132 (0.180)\tData 0.012 (0.020)\tLoss 0.3657 (0.3548)\tPrec@1 95.000 (95.316)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4926 (0.4926)\tPrec@1 89.844 (89.844)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.3958 (0.4280)\tPrec@1 94.531 (91.477)\n",
      "Test: [20/79]\tTime 0.025 (0.022)\tLoss 0.5217 (0.5001)\tPrec@1 91.406 (90.513)\n",
      "Test: [30/79]\tTime 0.020 (0.023)\tLoss 0.4470 (0.5085)\tPrec@1 92.188 (90.549)\n",
      "Test: [40/79]\tTime 0.023 (0.023)\tLoss 0.6190 (0.4984)\tPrec@1 85.938 (90.625)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 0.5414 (0.4816)\tPrec@1 89.844 (90.763)\n",
      "Test: [60/79]\tTime 0.024 (0.023)\tLoss 0.3352 (0.4806)\tPrec@1 91.406 (90.856)\n",
      "Test: [70/79]\tTime 0.020 (0.023)\tLoss 0.5275 (0.4786)\tPrec@1 90.625 (90.856)\n",
      " * Prec@1 91.050\n",
      "\n",
      "===> epoch: 116/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.163 (0.163)\tData 0.020 (0.020)\tLoss 0.3413 (0.3413)\tPrec@1 96.094 (96.094)\n",
      "[10/391]\tTime 0.177 (0.178)\tData 0.019 (0.020)\tLoss 0.3506 (0.3484)\tPrec@1 96.875 (95.170)\n",
      "[20/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.4313 (0.3545)\tPrec@1 89.062 (95.052)\n",
      "[30/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.3288 (0.3502)\tPrec@1 98.438 (95.186)\n",
      "[40/391]\tTime 0.185 (0.180)\tData 0.021 (0.020)\tLoss 0.3893 (0.3534)\tPrec@1 95.312 (95.103)\n",
      "[50/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3271 (0.3491)\tPrec@1 94.531 (95.236)\n",
      "[60/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3158 (0.3493)\tPrec@1 96.875 (95.223)\n",
      "[70/391]\tTime 0.185 (0.181)\tData 0.027 (0.020)\tLoss 0.3622 (0.3482)\tPrec@1 94.531 (95.213)\n",
      "[80/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3556 (0.3491)\tPrec@1 94.531 (95.226)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3334 (0.3493)\tPrec@1 96.875 (95.235)\n",
      "[100/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3344 (0.3502)\tPrec@1 95.312 (95.196)\n",
      "[110/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3841 (0.3495)\tPrec@1 93.750 (95.263)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3945 (0.3503)\tPrec@1 93.750 (95.209)\n",
      "[130/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3028 (0.3505)\tPrec@1 96.094 (95.169)\n",
      "[140/391]\tTime 0.186 (0.181)\tData 0.021 (0.020)\tLoss 0.3935 (0.3506)\tPrec@1 92.969 (95.207)\n",
      "[150/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3255 (0.3506)\tPrec@1 96.875 (95.261)\n",
      "[160/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3368 (0.3498)\tPrec@1 95.312 (95.293)\n",
      "[170/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3296 (0.3498)\tPrec@1 95.312 (95.312)\n",
      "[180/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.3341 (0.3497)\tPrec@1 97.656 (95.343)\n",
      "[190/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3645 (0.3505)\tPrec@1 93.750 (95.325)\n",
      "[200/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.3560 (0.3505)\tPrec@1 97.656 (95.328)\n",
      "[210/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3261 (0.3507)\tPrec@1 94.531 (95.301)\n",
      "[220/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3583 (0.3510)\tPrec@1 95.312 (95.309)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3701 (0.3516)\tPrec@1 95.312 (95.296)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3908 (0.3516)\tPrec@1 94.531 (95.309)\n",
      "[250/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.3326 (0.3519)\tPrec@1 93.750 (95.309)\n",
      "[260/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3395 (0.3528)\tPrec@1 96.875 (95.268)\n",
      "[270/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3713 (0.3526)\tPrec@1 95.312 (95.298)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3202 (0.3524)\tPrec@1 95.312 (95.287)\n",
      "[290/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3878 (0.3527)\tPrec@1 93.750 (95.288)\n",
      "[300/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3713 (0.3532)\tPrec@1 96.094 (95.258)\n",
      "[310/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3363 (0.3535)\tPrec@1 95.312 (95.245)\n",
      "[320/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3224 (0.3544)\tPrec@1 96.875 (95.218)\n",
      "[330/391]\tTime 0.143 (0.181)\tData 0.020 (0.020)\tLoss 0.3416 (0.3545)\tPrec@1 95.312 (95.223)\n",
      "[340/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.3987 (0.3543)\tPrec@1 93.750 (95.241)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.3594 (0.3546)\tPrec@1 93.750 (95.217)\n",
      "[360/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.3290 (0.3543)\tPrec@1 96.094 (95.226)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3447 (0.3539)\tPrec@1 95.312 (95.247)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.3993 (0.3539)\tPrec@1 92.188 (95.245)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.3132 (0.3535)\tPrec@1 97.500 (95.260)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3251 (0.3251)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.023 (0.023)\tLoss 0.2694 (0.3813)\tPrec@1 93.750 (92.188)\n",
      "Test: [20/79]\tTime 0.025 (0.022)\tLoss 0.3523 (0.4260)\tPrec@1 92.969 (91.369)\n",
      "Test: [30/79]\tTime 0.019 (0.022)\tLoss 0.4053 (0.4392)\tPrec@1 90.625 (91.053)\n",
      "Test: [40/79]\tTime 0.022 (0.022)\tLoss 0.6203 (0.4396)\tPrec@1 86.719 (91.159)\n",
      "Test: [50/79]\tTime 0.025 (0.022)\tLoss 0.5365 (0.4270)\tPrec@1 90.625 (91.406)\n",
      "Test: [60/79]\tTime 0.020 (0.023)\tLoss 0.2084 (0.4312)\tPrec@1 96.094 (91.509)\n",
      "Test: [70/79]\tTime 0.020 (0.023)\tLoss 0.5779 (0.4314)\tPrec@1 91.406 (91.571)\n",
      " * Prec@1 91.660\n",
      "\n",
      "===> epoch: 117/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.4073 (0.4073)\tPrec@1 95.312 (95.312)\n",
      "[10/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3111 (0.3382)\tPrec@1 98.438 (96.449)\n",
      "[20/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.3544 (0.3374)\tPrec@1 94.531 (95.982)\n",
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3592 (0.3390)\tPrec@1 96.875 (95.993)\n",
      "[40/391]\tTime 0.183 (0.181)\tData 0.021 (0.019)\tLoss 0.3337 (0.3390)\tPrec@1 93.750 (95.903)\n",
      "[50/391]\tTime 0.185 (0.181)\tData 0.019 (0.019)\tLoss 0.3606 (0.3428)\tPrec@1 93.750 (95.772)\n",
      "[60/391]\tTime 0.185 (0.182)\tData 0.019 (0.019)\tLoss 0.3569 (0.3417)\tPrec@1 96.094 (95.774)\n",
      "[70/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3203 (0.3433)\tPrec@1 95.312 (95.764)\n",
      "[80/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.3795 (0.3434)\tPrec@1 93.750 (95.775)\n",
      "[90/391]\tTime 0.181 (0.182)\tData 0.020 (0.019)\tLoss 0.3440 (0.3441)\tPrec@1 95.312 (95.733)\n",
      "[100/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3748 (0.3441)\tPrec@1 95.312 (95.746)\n",
      "[110/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.3517 (0.3447)\tPrec@1 93.750 (95.728)\n",
      "[120/391]\tTime 0.187 (0.182)\tData 0.020 (0.019)\tLoss 0.3170 (0.3447)\tPrec@1 96.875 (95.693)\n",
      "[130/391]\tTime 0.183 (0.182)\tData 0.019 (0.019)\tLoss 0.3306 (0.3454)\tPrec@1 95.312 (95.641)\n",
      "[140/391]\tTime 0.182 (0.182)\tData 0.020 (0.019)\tLoss 0.2935 (0.3449)\tPrec@1 98.438 (95.673)\n",
      "[150/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4565 (0.3453)\tPrec@1 92.188 (95.633)\n",
      "[160/391]\tTime 0.180 (0.182)\tData 0.019 (0.019)\tLoss 0.3287 (0.3455)\tPrec@1 95.312 (95.618)\n",
      "[170/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3831 (0.3458)\tPrec@1 96.094 (95.628)\n",
      "[180/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3810 (0.3460)\tPrec@1 93.750 (95.636)\n",
      "[190/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.3371 (0.3459)\tPrec@1 95.312 (95.615)\n",
      "[200/391]\tTime 0.183 (0.181)\tData 0.021 (0.020)\tLoss 0.2623 (0.3457)\tPrec@1 99.219 (95.620)\n",
      "[210/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3975 (0.3468)\tPrec@1 92.188 (95.568)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3489 (0.3466)\tPrec@1 95.312 (95.585)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.2838 (0.3473)\tPrec@1 98.438 (95.539)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.019 (0.019)\tLoss 0.4200 (0.3476)\tPrec@1 93.750 (95.549)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.020 (0.019)\tLoss 0.3746 (0.3480)\tPrec@1 93.750 (95.530)\n",
      "[260/391]\tTime 0.184 (0.181)\tData 0.019 (0.019)\tLoss 0.2972 (0.3481)\tPrec@1 98.438 (95.516)\n",
      "[270/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3926 (0.3482)\tPrec@1 92.969 (95.520)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3492 (0.3487)\tPrec@1 96.094 (95.527)\n",
      "[290/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3934 (0.3490)\tPrec@1 94.531 (95.533)\n",
      "[300/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3592 (0.3501)\tPrec@1 96.875 (95.494)\n",
      "[310/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3639 (0.3505)\tPrec@1 95.312 (95.483)\n",
      "[320/391]\tTime 0.143 (0.181)\tData 0.020 (0.020)\tLoss 0.3265 (0.3504)\tPrec@1 96.875 (95.485)\n",
      "[330/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.3559 (0.3505)\tPrec@1 94.531 (95.466)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.2990 (0.3504)\tPrec@1 98.438 (95.455)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.3687 (0.3504)\tPrec@1 96.875 (95.451)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.3694 (0.3505)\tPrec@1 93.750 (95.438)\n",
      "[370/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.3916 (0.3516)\tPrec@1 92.188 (95.403)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3645 (0.3520)\tPrec@1 94.531 (95.374)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.3938 (0.3518)\tPrec@1 97.500 (95.386)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3866 (0.3866)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.033 (0.024)\tLoss 0.4666 (0.4444)\tPrec@1 90.625 (91.974)\n",
      "Test: [20/79]\tTime 0.021 (0.024)\tLoss 0.3214 (0.4801)\tPrec@1 91.406 (90.848)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.4827 (0.4911)\tPrec@1 92.188 (90.625)\n",
      "Test: [40/79]\tTime 0.027 (0.023)\tLoss 0.6249 (0.4953)\tPrec@1 86.719 (90.549)\n",
      "Test: [50/79]\tTime 0.019 (0.023)\tLoss 0.4079 (0.4764)\tPrec@1 92.969 (90.794)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.3740 (0.4662)\tPrec@1 92.188 (91.073)\n",
      "Test: [70/79]\tTime 0.026 (0.023)\tLoss 0.6272 (0.4678)\tPrec@1 87.500 (91.032)\n",
      " * Prec@1 91.130\n",
      "\n",
      "===> epoch: 118/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.178 (0.178)\tData 0.021 (0.021)\tLoss 0.3865 (0.3865)\tPrec@1 92.969 (92.969)\n",
      "[10/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3221 (0.3358)\tPrec@1 97.656 (96.733)\n",
      "[20/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.4017 (0.3390)\tPrec@1 93.750 (96.243)\n",
      "[30/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3010 (0.3387)\tPrec@1 97.656 (96.270)\n",
      "[40/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.3641 (0.3442)\tPrec@1 97.656 (96.132)\n",
      "[50/391]\tTime 0.185 (0.183)\tData 0.020 (0.020)\tLoss 0.3252 (0.3446)\tPrec@1 97.656 (95.971)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3585 (0.3481)\tPrec@1 94.531 (95.748)\n",
      "[70/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3076 (0.3459)\tPrec@1 98.438 (95.874)\n",
      "[80/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.2864 (0.3460)\tPrec@1 99.219 (95.824)\n",
      "[90/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3696 (0.3476)\tPrec@1 95.312 (95.768)\n",
      "[100/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.2779 (0.3456)\tPrec@1 99.219 (95.877)\n",
      "[110/391]\tTime 0.184 (0.182)\tData 0.019 (0.019)\tLoss 0.3632 (0.3464)\tPrec@1 95.312 (95.833)\n",
      "[120/391]\tTime 0.181 (0.182)\tData 0.019 (0.019)\tLoss 0.3245 (0.3450)\tPrec@1 96.094 (95.900)\n",
      "[130/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.3041 (0.3452)\tPrec@1 97.656 (95.885)\n",
      "[140/391]\tTime 0.180 (0.182)\tData 0.020 (0.019)\tLoss 0.3880 (0.3450)\tPrec@1 94.531 (95.894)\n",
      "[150/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3238 (0.3447)\tPrec@1 96.875 (95.876)\n",
      "[160/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.2928 (0.3448)\tPrec@1 97.656 (95.871)\n",
      "[170/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3942 (0.3446)\tPrec@1 91.406 (95.861)\n",
      "[180/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.3951 (0.3452)\tPrec@1 95.312 (95.848)\n",
      "[190/391]\tTime 0.181 (0.182)\tData 0.022 (0.020)\tLoss 0.3656 (0.3446)\tPrec@1 95.312 (95.857)\n",
      "[200/391]\tTime 0.178 (0.182)\tData 0.020 (0.020)\tLoss 0.3214 (0.3453)\tPrec@1 97.656 (95.822)\n",
      "[210/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.4293 (0.3457)\tPrec@1 92.188 (95.823)\n",
      "[220/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3092 (0.3466)\tPrec@1 97.656 (95.765)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[230/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3596 (0.3462)\tPrec@1 94.531 (95.789)\n",
      "[240/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3535 (0.3459)\tPrec@1 96.094 (95.799)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3769 (0.3456)\tPrec@1 94.531 (95.811)\n",
      "[260/391]\tTime 0.178 (0.181)\tData 0.020 (0.020)\tLoss 0.3158 (0.3453)\tPrec@1 96.094 (95.800)\n",
      "[270/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3329 (0.3453)\tPrec@1 96.094 (95.805)\n",
      "[280/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3371 (0.3451)\tPrec@1 96.875 (95.796)\n",
      "[290/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3894 (0.3457)\tPrec@1 93.750 (95.774)\n",
      "[300/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3482 (0.3462)\tPrec@1 95.312 (95.767)\n",
      "[310/391]\tTime 0.148 (0.181)\tData 0.020 (0.020)\tLoss 0.3500 (0.3462)\tPrec@1 96.094 (95.772)\n",
      "[320/391]\tTime 0.154 (0.180)\tData 0.019 (0.020)\tLoss 0.3454 (0.3468)\tPrec@1 96.875 (95.741)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3990 (0.3475)\tPrec@1 91.406 (95.704)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3809 (0.3482)\tPrec@1 92.969 (95.649)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3817 (0.3486)\tPrec@1 96.094 (95.649)\n",
      "[360/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3350 (0.3486)\tPrec@1 96.094 (95.631)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3303 (0.3486)\tPrec@1 96.094 (95.626)\n",
      "[380/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.3622 (0.3492)\tPrec@1 94.531 (95.606)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.4074 (0.3494)\tPrec@1 95.000 (95.596)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.4267 (0.4267)\tPrec@1 92.188 (92.188)\n",
      "Test: [10/79]\tTime 0.020 (0.023)\tLoss 0.3571 (0.4404)\tPrec@1 92.188 (91.832)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.4991 (0.4946)\tPrec@1 91.406 (91.109)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.3755 (0.5002)\tPrec@1 92.969 (90.927)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 0.5962 (0.4877)\tPrec@1 89.062 (90.949)\n",
      "Test: [50/79]\tTime 0.020 (0.023)\tLoss 0.5321 (0.4775)\tPrec@1 89.062 (91.054)\n",
      "Test: [60/79]\tTime 0.023 (0.023)\tLoss 0.3706 (0.4752)\tPrec@1 91.406 (91.060)\n",
      "Test: [70/79]\tTime 0.025 (0.023)\tLoss 0.4484 (0.4672)\tPrec@1 90.625 (91.131)\n",
      " * Prec@1 91.230\n",
      "\n",
      "===> epoch: 119/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.3402 (0.3402)\tPrec@1 96.094 (96.094)\n",
      "[10/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.3521 (0.3451)\tPrec@1 93.750 (95.312)\n",
      "[20/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3544 (0.3489)\tPrec@1 94.531 (95.052)\n",
      "[30/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3194 (0.3413)\tPrec@1 95.312 (95.565)\n",
      "[40/391]\tTime 0.177 (0.181)\tData 0.020 (0.020)\tLoss 0.3396 (0.3406)\tPrec@1 96.875 (95.598)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3584 (0.3419)\tPrec@1 96.094 (95.588)\n",
      "[60/391]\tTime 0.185 (0.181)\tData 0.019 (0.020)\tLoss 0.3922 (0.3398)\tPrec@1 94.531 (95.838)\n",
      "[70/391]\tTime 0.186 (0.181)\tData 0.020 (0.020)\tLoss 0.3167 (0.3402)\tPrec@1 96.875 (95.863)\n",
      "[80/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3780 (0.3402)\tPrec@1 92.188 (95.862)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.4069 (0.3428)\tPrec@1 94.531 (95.819)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3819 (0.3426)\tPrec@1 92.188 (95.854)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3373 (0.3436)\tPrec@1 96.094 (95.749)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3920 (0.3440)\tPrec@1 92.969 (95.739)\n",
      "[130/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3580 (0.3445)\tPrec@1 94.531 (95.754)\n",
      "[140/391]\tTime 0.187 (0.181)\tData 0.020 (0.020)\tLoss 0.3373 (0.3441)\tPrec@1 95.312 (95.795)\n",
      "[150/391]\tTime 0.181 (0.181)\tData 0.023 (0.020)\tLoss 0.4355 (0.3443)\tPrec@1 93.750 (95.794)\n",
      "[160/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3406 (0.3441)\tPrec@1 97.656 (95.827)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3626 (0.3434)\tPrec@1 94.531 (95.838)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3625 (0.3443)\tPrec@1 96.094 (95.792)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3562 (0.3437)\tPrec@1 94.531 (95.791)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3538 (0.3443)\tPrec@1 94.531 (95.763)\n",
      "[210/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3924 (0.3453)\tPrec@1 93.750 (95.712)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3715 (0.3461)\tPrec@1 94.531 (95.698)\n",
      "[230/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3505 (0.3460)\tPrec@1 96.094 (95.708)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3539 (0.3459)\tPrec@1 95.312 (95.724)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3802 (0.3466)\tPrec@1 95.312 (95.692)\n",
      "[260/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3506 (0.3474)\tPrec@1 95.312 (95.651)\n",
      "[270/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3631 (0.3477)\tPrec@1 95.312 (95.630)\n",
      "[280/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.3660 (0.3471)\tPrec@1 96.094 (95.666)\n",
      "[290/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3590 (0.3469)\tPrec@1 94.531 (95.667)\n",
      "[300/391]\tTime 0.144 (0.181)\tData 0.019 (0.020)\tLoss 0.3286 (0.3467)\tPrec@1 97.656 (95.671)\n",
      "[310/391]\tTime 0.145 (0.180)\tData 0.020 (0.020)\tLoss 0.3131 (0.3461)\tPrec@1 96.094 (95.677)\n",
      "[320/391]\tTime 0.187 (0.180)\tData 0.021 (0.020)\tLoss 0.3284 (0.3460)\tPrec@1 95.312 (95.680)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3405 (0.3461)\tPrec@1 96.094 (95.690)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3568 (0.3466)\tPrec@1 92.969 (95.661)\n",
      "[350/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3382 (0.3462)\tPrec@1 95.312 (95.678)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3477 (0.3467)\tPrec@1 96.094 (95.661)\n",
      "[370/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3952 (0.3471)\tPrec@1 92.969 (95.666)\n",
      "[380/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.3617 (0.3472)\tPrec@1 94.531 (95.667)\n",
      "[390/391]\tTime 0.134 (0.180)\tData 0.014 (0.020)\tLoss 0.3090 (0.3471)\tPrec@1 98.750 (95.672)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3744 (0.3744)\tPrec@1 91.406 (91.406)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.4468 (0.4639)\tPrec@1 92.969 (91.406)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.3987 (0.5101)\tPrec@1 91.406 (90.960)\n",
      "Test: [30/79]\tTime 0.020 (0.023)\tLoss 0.3304 (0.5161)\tPrec@1 93.750 (90.776)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 0.7693 (0.5191)\tPrec@1 87.500 (90.720)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 0.5784 (0.5076)\tPrec@1 89.062 (90.916)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.3036 (0.5004)\tPrec@1 94.531 (90.996)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.6021 (0.4968)\tPrec@1 90.625 (91.065)\n",
      " * Prec@1 91.210\n",
      "\n",
      "===> epoch: 120/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.175 (0.175)\tData 0.020 (0.020)\tLoss 0.3181 (0.3181)\tPrec@1 96.875 (96.875)\n",
      "[10/391]\tTime 0.180 (0.184)\tData 0.019 (0.020)\tLoss 0.3493 (0.3490)\tPrec@1 93.750 (95.952)\n",
      "[20/391]\tTime 0.182 (0.183)\tData 0.020 (0.020)\tLoss 0.2662 (0.3296)\tPrec@1 98.438 (96.577)\n",
      "[30/391]\tTime 0.184 (0.183)\tData 0.020 (0.020)\tLoss 0.3956 (0.3365)\tPrec@1 93.750 (96.169)\n",
      "[40/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3482 (0.3347)\tPrec@1 96.094 (96.246)\n",
      "[50/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3840 (0.3344)\tPrec@1 94.531 (96.278)\n",
      "[60/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3415 (0.3378)\tPrec@1 96.094 (96.158)\n",
      "[70/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.4109 (0.3401)\tPrec@1 92.969 (96.061)\n",
      "[80/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3406 (0.3403)\tPrec@1 96.094 (96.026)\n",
      "[90/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3437 (0.3421)\tPrec@1 96.875 (95.948)\n",
      "[100/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.3746 (0.3425)\tPrec@1 95.312 (95.908)\n",
      "[110/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3137 (0.3415)\tPrec@1 96.875 (95.939)\n",
      "[120/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.4032 (0.3430)\tPrec@1 94.531 (95.894)\n",
      "[130/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3417 (0.3429)\tPrec@1 98.438 (95.909)\n",
      "[140/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3255 (0.3434)\tPrec@1 96.875 (95.911)\n",
      "[150/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.2964 (0.3432)\tPrec@1 99.219 (95.933)\n",
      "[160/391]\tTime 0.183 (0.182)\tData 0.020 (0.020)\tLoss 0.3133 (0.3435)\tPrec@1 96.875 (95.924)\n",
      "[170/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3412 (0.3434)\tPrec@1 95.312 (95.893)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[180/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3509 (0.3440)\tPrec@1 94.531 (95.852)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3024 (0.3435)\tPrec@1 96.094 (95.869)\n",
      "[200/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.2987 (0.3441)\tPrec@1 98.438 (95.826)\n",
      "[210/391]\tTime 0.182 (0.181)\tData 0.021 (0.020)\tLoss 0.3208 (0.3443)\tPrec@1 96.875 (95.812)\n",
      "[220/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3259 (0.3441)\tPrec@1 96.094 (95.846)\n",
      "[230/391]\tTime 0.188 (0.181)\tData 0.019 (0.020)\tLoss 0.3657 (0.3446)\tPrec@1 94.531 (95.810)\n",
      "[240/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3371 (0.3453)\tPrec@1 96.875 (95.766)\n",
      "[250/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3282 (0.3454)\tPrec@1 95.312 (95.742)\n",
      "[260/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3932 (0.3449)\tPrec@1 92.188 (95.756)\n",
      "[270/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.2960 (0.3456)\tPrec@1 97.656 (95.739)\n",
      "[280/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3519 (0.3464)\tPrec@1 94.531 (95.710)\n",
      "[290/391]\tTime 0.143 (0.181)\tData 0.019 (0.020)\tLoss 0.3598 (0.3461)\tPrec@1 93.750 (95.729)\n",
      "[300/391]\tTime 0.143 (0.180)\tData 0.019 (0.020)\tLoss 0.3470 (0.3461)\tPrec@1 94.531 (95.759)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3991 (0.3463)\tPrec@1 94.531 (95.767)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.4527 (0.3470)\tPrec@1 92.969 (95.731)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.4016 (0.3478)\tPrec@1 92.969 (95.688)\n",
      "[340/391]\tTime 0.188 (0.180)\tData 0.020 (0.020)\tLoss 0.3863 (0.3482)\tPrec@1 94.531 (95.681)\n",
      "[350/391]\tTime 0.185 (0.180)\tData 0.019 (0.020)\tLoss 0.3581 (0.3478)\tPrec@1 96.094 (95.695)\n",
      "[360/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.3322 (0.3481)\tPrec@1 96.875 (95.691)\n",
      "[370/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3625 (0.3479)\tPrec@1 92.969 (95.666)\n",
      "[380/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.2956 (0.3479)\tPrec@1 96.875 (95.669)\n",
      "[390/391]\tTime 0.128 (0.180)\tData 0.013 (0.020)\tLoss 0.3537 (0.3482)\tPrec@1 93.750 (95.634)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.4976 (0.4976)\tPrec@1 90.625 (90.625)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.3436 (0.4808)\tPrec@1 93.750 (90.483)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.4248 (0.5373)\tPrec@1 92.969 (90.179)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 0.3703 (0.5439)\tPrec@1 93.750 (90.146)\n",
      "Test: [40/79]\tTime 0.021 (0.023)\tLoss 0.5504 (0.5332)\tPrec@1 89.062 (90.301)\n",
      "Test: [50/79]\tTime 0.026 (0.023)\tLoss 0.5859 (0.5198)\tPrec@1 89.062 (90.487)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.4381 (0.5195)\tPrec@1 93.750 (90.561)\n",
      "Test: [70/79]\tTime 0.019 (0.023)\tLoss 0.5817 (0.5205)\tPrec@1 89.844 (90.548)\n",
      " * Prec@1 90.670\n",
      "\n",
      "===> epoch: 121/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.170 (0.170)\tData 0.021 (0.021)\tLoss 0.3381 (0.3381)\tPrec@1 95.312 (95.312)\n",
      "[10/391]\tTime 0.182 (0.180)\tData 0.019 (0.019)\tLoss 0.3209 (0.3467)\tPrec@1 95.312 (95.455)\n",
      "[20/391]\tTime 0.178 (0.180)\tData 0.019 (0.019)\tLoss 0.3367 (0.3437)\tPrec@1 97.656 (96.168)\n",
      "[30/391]\tTime 0.181 (0.180)\tData 0.020 (0.019)\tLoss 0.3422 (0.3453)\tPrec@1 96.094 (95.892)\n",
      "[40/391]\tTime 0.179 (0.180)\tData 0.019 (0.019)\tLoss 0.2871 (0.3433)\tPrec@1 99.219 (95.922)\n",
      "[50/391]\tTime 0.185 (0.181)\tData 0.020 (0.020)\tLoss 0.3217 (0.3430)\tPrec@1 97.656 (95.910)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3156 (0.3402)\tPrec@1 96.875 (96.081)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3499 (0.3412)\tPrec@1 96.875 (95.995)\n",
      "[80/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3770 (0.3436)\tPrec@1 93.750 (95.968)\n",
      "[90/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3614 (0.3436)\tPrec@1 97.656 (95.956)\n",
      "[100/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.4326 (0.3445)\tPrec@1 90.625 (95.877)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3039 (0.3451)\tPrec@1 97.656 (95.847)\n",
      "[120/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3061 (0.3441)\tPrec@1 97.656 (95.842)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3128 (0.3451)\tPrec@1 98.438 (95.825)\n",
      "[140/391]\tTime 0.176 (0.180)\tData 0.021 (0.020)\tLoss 0.3860 (0.3456)\tPrec@1 94.531 (95.828)\n",
      "[150/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3164 (0.3452)\tPrec@1 96.875 (95.882)\n",
      "[160/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.3179 (0.3450)\tPrec@1 97.656 (95.871)\n",
      "[170/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.3011 (0.3454)\tPrec@1 99.219 (95.852)\n",
      "[180/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3665 (0.3464)\tPrec@1 95.312 (95.805)\n",
      "[190/391]\tTime 0.178 (0.180)\tData 0.020 (0.020)\tLoss 0.4656 (0.3484)\tPrec@1 91.406 (95.746)\n",
      "[200/391]\tTime 0.182 (0.180)\tData 0.022 (0.020)\tLoss 0.3601 (0.3484)\tPrec@1 95.312 (95.736)\n",
      "[210/391]\tTime 0.181 (0.180)\tData 0.021 (0.020)\tLoss 0.3461 (0.3481)\tPrec@1 95.312 (95.742)\n",
      "[220/391]\tTime 0.178 (0.180)\tData 0.022 (0.020)\tLoss 0.3368 (0.3489)\tPrec@1 95.312 (95.694)\n",
      "[230/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3759 (0.3494)\tPrec@1 94.531 (95.671)\n",
      "[240/391]\tTime 0.180 (0.180)\tData 0.021 (0.020)\tLoss 0.4120 (0.3501)\tPrec@1 92.188 (95.627)\n",
      "[250/391]\tTime 0.186 (0.180)\tData 0.021 (0.020)\tLoss 0.3620 (0.3495)\tPrec@1 96.094 (95.633)\n",
      "[260/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.3023 (0.3493)\tPrec@1 97.656 (95.633)\n",
      "[270/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.2920 (0.3487)\tPrec@1 96.875 (95.658)\n",
      "[280/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.4127 (0.3488)\tPrec@1 92.969 (95.660)\n",
      "[290/391]\tTime 0.147 (0.179)\tData 0.021 (0.020)\tLoss 0.3779 (0.3486)\tPrec@1 94.531 (95.645)\n",
      "[300/391]\tTime 0.181 (0.179)\tData 0.019 (0.020)\tLoss 0.3993 (0.3491)\tPrec@1 91.406 (95.619)\n",
      "[310/391]\tTime 0.179 (0.179)\tData 0.020 (0.020)\tLoss 0.3472 (0.3492)\tPrec@1 96.875 (95.619)\n",
      "[320/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.4464 (0.3492)\tPrec@1 90.625 (95.622)\n",
      "[330/391]\tTime 0.186 (0.179)\tData 0.019 (0.020)\tLoss 0.3418 (0.3496)\tPrec@1 96.094 (95.591)\n",
      "[340/391]\tTime 0.180 (0.179)\tData 0.021 (0.020)\tLoss 0.3681 (0.3503)\tPrec@1 95.312 (95.555)\n",
      "[350/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.3219 (0.3506)\tPrec@1 96.094 (95.551)\n",
      "[360/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.3969 (0.3509)\tPrec@1 91.406 (95.542)\n",
      "[370/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.3819 (0.3516)\tPrec@1 92.969 (95.521)\n",
      "[380/391]\tTime 0.186 (0.180)\tData 0.022 (0.020)\tLoss 0.4522 (0.3518)\tPrec@1 91.406 (95.509)\n",
      "[390/391]\tTime 0.131 (0.179)\tData 0.013 (0.020)\tLoss 0.3416 (0.3521)\tPrec@1 95.000 (95.494)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3155 (0.3155)\tPrec@1 92.188 (92.188)\n",
      "Test: [10/79]\tTime 0.022 (0.023)\tLoss 0.3656 (0.4146)\tPrec@1 92.969 (91.264)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.4102 (0.4647)\tPrec@1 91.406 (91.146)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.3772 (0.4703)\tPrec@1 92.969 (90.927)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 0.4612 (0.4721)\tPrec@1 89.844 (90.911)\n",
      "Test: [50/79]\tTime 0.023 (0.023)\tLoss 0.4647 (0.4560)\tPrec@1 92.188 (91.161)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.3167 (0.4580)\tPrec@1 91.406 (91.240)\n",
      "Test: [70/79]\tTime 0.027 (0.023)\tLoss 0.5363 (0.4517)\tPrec@1 91.406 (91.351)\n",
      " * Prec@1 91.430\n",
      "\n",
      "===> epoch: 122/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.184 (0.184)\tData 0.022 (0.022)\tLoss 0.3224 (0.3224)\tPrec@1 97.656 (97.656)\n",
      "[10/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3078 (0.3400)\tPrec@1 97.656 (96.094)\n",
      "[20/391]\tTime 0.187 (0.182)\tData 0.020 (0.020)\tLoss 0.3659 (0.3432)\tPrec@1 94.531 (96.131)\n",
      "[30/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3284 (0.3388)\tPrec@1 96.875 (96.245)\n",
      "[40/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.3188 (0.3422)\tPrec@1 97.656 (96.037)\n",
      "[50/391]\tTime 0.180 (0.181)\tData 0.020 (0.019)\tLoss 0.3310 (0.3423)\tPrec@1 96.094 (96.048)\n",
      "[60/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3536 (0.3405)\tPrec@1 92.969 (96.081)\n",
      "[70/391]\tTime 0.179 (0.181)\tData 0.019 (0.019)\tLoss 0.3581 (0.3416)\tPrec@1 94.531 (95.984)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.019 (0.019)\tLoss 0.3325 (0.3397)\tPrec@1 97.656 (96.065)\n",
      "[90/391]\tTime 0.187 (0.181)\tData 0.020 (0.019)\tLoss 0.3215 (0.3393)\tPrec@1 96.875 (96.051)\n",
      "[100/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3158 (0.3387)\tPrec@1 96.875 (96.086)\n",
      "[110/391]\tTime 0.187 (0.182)\tData 0.019 (0.020)\tLoss 0.3165 (0.3392)\tPrec@1 96.875 (96.073)\n",
      "[120/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3158 (0.3392)\tPrec@1 95.312 (96.029)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[130/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3495 (0.3398)\tPrec@1 92.969 (95.992)\n",
      "[140/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3733 (0.3416)\tPrec@1 95.312 (95.933)\n",
      "[150/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3246 (0.3425)\tPrec@1 94.531 (95.897)\n",
      "[160/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.3474 (0.3429)\tPrec@1 96.875 (95.856)\n",
      "[170/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3533 (0.3427)\tPrec@1 95.312 (95.879)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3561 (0.3430)\tPrec@1 96.094 (95.900)\n",
      "[190/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3386 (0.3427)\tPrec@1 96.875 (95.897)\n",
      "[200/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3666 (0.3432)\tPrec@1 95.312 (95.888)\n",
      "[210/391]\tTime 0.182 (0.182)\tData 0.024 (0.020)\tLoss 0.3747 (0.3439)\tPrec@1 96.875 (95.857)\n",
      "[220/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3207 (0.3443)\tPrec@1 97.656 (95.825)\n",
      "[230/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3699 (0.3439)\tPrec@1 95.312 (95.864)\n",
      "[240/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.3401 (0.3444)\tPrec@1 95.312 (95.838)\n",
      "[250/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.3262 (0.3455)\tPrec@1 95.312 (95.789)\n",
      "[260/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3819 (0.3456)\tPrec@1 96.875 (95.785)\n",
      "[270/391]\tTime 0.185 (0.182)\tData 0.020 (0.020)\tLoss 0.3287 (0.3452)\tPrec@1 96.094 (95.814)\n",
      "[280/391]\tTime 0.144 (0.180)\tData 0.020 (0.020)\tLoss 0.3545 (0.3455)\tPrec@1 96.094 (95.813)\n",
      "[290/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3320 (0.3465)\tPrec@1 96.094 (95.758)\n",
      "[300/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.4174 (0.3467)\tPrec@1 91.406 (95.746)\n",
      "[310/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3753 (0.3471)\tPrec@1 95.312 (95.724)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.4413 (0.3471)\tPrec@1 90.625 (95.724)\n",
      "[330/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.3550 (0.3470)\tPrec@1 95.312 (95.737)\n",
      "[340/391]\tTime 0.178 (0.180)\tData 0.020 (0.020)\tLoss 0.2813 (0.3469)\tPrec@1 98.438 (95.732)\n",
      "[350/391]\tTime 0.182 (0.180)\tData 0.021 (0.020)\tLoss 0.3729 (0.3470)\tPrec@1 92.969 (95.700)\n",
      "[360/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.2866 (0.3468)\tPrec@1 98.438 (95.713)\n",
      "[370/391]\tTime 0.186 (0.180)\tData 0.019 (0.020)\tLoss 0.3713 (0.3472)\tPrec@1 94.531 (95.719)\n",
      "[380/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.3438 (0.3473)\tPrec@1 93.750 (95.696)\n",
      "[390/391]\tTime 0.131 (0.180)\tData 0.012 (0.020)\tLoss 0.3831 (0.3478)\tPrec@1 93.750 (95.674)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3551 (0.3551)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.024 (0.023)\tLoss 0.3461 (0.3935)\tPrec@1 93.750 (91.832)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.4198 (0.4418)\tPrec@1 93.750 (91.220)\n",
      "Test: [30/79]\tTime 0.020 (0.023)\tLoss 0.3843 (0.4597)\tPrec@1 92.188 (91.104)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 0.5819 (0.4678)\tPrec@1 86.719 (91.120)\n",
      "Test: [50/79]\tTime 0.024 (0.023)\tLoss 0.3612 (0.4606)\tPrec@1 92.188 (91.238)\n",
      "Test: [60/79]\tTime 0.025 (0.023)\tLoss 0.2786 (0.4587)\tPrec@1 92.188 (91.240)\n",
      "Test: [70/79]\tTime 0.021 (0.023)\tLoss 0.5221 (0.4575)\tPrec@1 90.625 (91.285)\n",
      " * Prec@1 91.370\n",
      "\n",
      "===> epoch: 123/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.164 (0.164)\tData 0.021 (0.021)\tLoss 0.3118 (0.3118)\tPrec@1 96.875 (96.875)\n",
      "[10/391]\tTime 0.181 (0.179)\tData 0.021 (0.020)\tLoss 0.3075 (0.3412)\tPrec@1 98.438 (95.739)\n",
      "[20/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.2950 (0.3448)\tPrec@1 97.656 (95.610)\n",
      "[30/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.2901 (0.3379)\tPrec@1 100.000 (95.943)\n",
      "[40/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3248 (0.3369)\tPrec@1 97.656 (95.979)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3662 (0.3399)\tPrec@1 93.750 (95.833)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3056 (0.3381)\tPrec@1 96.875 (95.966)\n",
      "[70/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3434 (0.3396)\tPrec@1 96.875 (95.808)\n",
      "[80/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.3167 (0.3394)\tPrec@1 97.656 (95.843)\n",
      "[90/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3454 (0.3396)\tPrec@1 96.875 (95.845)\n",
      "[100/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.3286 (0.3386)\tPrec@1 96.875 (95.893)\n",
      "[110/391]\tTime 0.186 (0.182)\tData 0.020 (0.020)\tLoss 0.3482 (0.3387)\tPrec@1 96.094 (95.854)\n",
      "[120/391]\tTime 0.184 (0.182)\tData 0.021 (0.020)\tLoss 0.4010 (0.3381)\tPrec@1 90.625 (95.816)\n",
      "[130/391]\tTime 0.182 (0.182)\tData 0.020 (0.020)\tLoss 0.3034 (0.3388)\tPrec@1 96.875 (95.802)\n",
      "[140/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3206 (0.3388)\tPrec@1 96.875 (95.778)\n",
      "[150/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3444 (0.3394)\tPrec@1 93.750 (95.773)\n",
      "[160/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3768 (0.3407)\tPrec@1 96.094 (95.701)\n",
      "[170/391]\tTime 0.181 (0.182)\tData 0.020 (0.020)\tLoss 0.4053 (0.3415)\tPrec@1 94.531 (95.687)\n",
      "[180/391]\tTime 0.180 (0.182)\tData 0.019 (0.020)\tLoss 0.3980 (0.3424)\tPrec@1 92.969 (95.662)\n",
      "[190/391]\tTime 0.183 (0.182)\tData 0.019 (0.020)\tLoss 0.4090 (0.3430)\tPrec@1 92.188 (95.615)\n",
      "[200/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3221 (0.3426)\tPrec@1 97.656 (95.627)\n",
      "[210/391]\tTime 0.179 (0.182)\tData 0.021 (0.020)\tLoss 0.3405 (0.3425)\tPrec@1 96.875 (95.672)\n",
      "[220/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3429 (0.3429)\tPrec@1 94.531 (95.655)\n",
      "[230/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3322 (0.3426)\tPrec@1 96.094 (95.668)\n",
      "[240/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3538 (0.3426)\tPrec@1 95.312 (95.689)\n",
      "[250/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3287 (0.3436)\tPrec@1 96.094 (95.664)\n",
      "[260/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3440 (0.3431)\tPrec@1 96.875 (95.693)\n",
      "[270/391]\tTime 0.144 (0.180)\tData 0.019 (0.020)\tLoss 0.3502 (0.3435)\tPrec@1 94.531 (95.679)\n",
      "[280/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.3167 (0.3434)\tPrec@1 97.656 (95.679)\n",
      "[290/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.4130 (0.3438)\tPrec@1 93.750 (95.675)\n",
      "[300/391]\tTime 0.181 (0.179)\tData 0.021 (0.020)\tLoss 0.2864 (0.3436)\tPrec@1 97.656 (95.678)\n",
      "[310/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.3758 (0.3438)\tPrec@1 92.188 (95.679)\n",
      "[320/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3706 (0.3436)\tPrec@1 95.312 (95.697)\n",
      "[330/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3225 (0.3436)\tPrec@1 96.094 (95.697)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3833 (0.3437)\tPrec@1 93.750 (95.697)\n",
      "[350/391]\tTime 0.183 (0.180)\tData 0.021 (0.020)\tLoss 0.3451 (0.3443)\tPrec@1 96.094 (95.675)\n",
      "[360/391]\tTime 0.180 (0.180)\tData 0.019 (0.020)\tLoss 0.3006 (0.3446)\tPrec@1 97.656 (95.667)\n",
      "[370/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.3258 (0.3446)\tPrec@1 96.094 (95.660)\n",
      "[380/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3394 (0.3446)\tPrec@1 96.875 (95.653)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.4184 (0.3453)\tPrec@1 91.250 (95.646)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3353 (0.3353)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 0.4953 (0.4611)\tPrec@1 92.969 (91.832)\n",
      "Test: [20/79]\tTime 0.026 (0.023)\tLoss 0.5252 (0.4908)\tPrec@1 92.969 (91.146)\n",
      "Test: [30/79]\tTime 0.027 (0.023)\tLoss 0.3468 (0.5091)\tPrec@1 91.406 (90.751)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 0.6414 (0.5184)\tPrec@1 85.156 (90.625)\n",
      "Test: [50/79]\tTime 0.022 (0.023)\tLoss 0.4312 (0.5135)\tPrec@1 92.969 (90.610)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.4125 (0.5097)\tPrec@1 92.188 (90.599)\n",
      "Test: [70/79]\tTime 0.027 (0.023)\tLoss 0.4150 (0.4989)\tPrec@1 92.188 (90.812)\n",
      " * Prec@1 90.900\n",
      "\n",
      "===> epoch: 124/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.181 (0.181)\tData 0.021 (0.021)\tLoss 0.3714 (0.3714)\tPrec@1 95.312 (95.312)\n",
      "[10/391]\tTime 0.183 (0.182)\tData 0.021 (0.020)\tLoss 0.3380 (0.3438)\tPrec@1 96.875 (96.662)\n",
      "[20/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3606 (0.3470)\tPrec@1 93.750 (96.019)\n",
      "[30/391]\tTime 0.184 (0.182)\tData 0.020 (0.020)\tLoss 0.3020 (0.3410)\tPrec@1 99.219 (96.169)\n",
      "[40/391]\tTime 0.182 (0.182)\tData 0.019 (0.020)\tLoss 0.3545 (0.3399)\tPrec@1 95.312 (96.037)\n",
      "[50/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3373 (0.3415)\tPrec@1 98.438 (95.971)\n",
      "[60/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3308 (0.3423)\tPrec@1 93.750 (95.889)\n",
      "[70/391]\tTime 0.181 (0.182)\tData 0.019 (0.020)\tLoss 0.2964 (0.3415)\tPrec@1 96.094 (95.940)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[80/391]\tTime 0.184 (0.182)\tData 0.019 (0.020)\tLoss 0.3269 (0.3422)\tPrec@1 96.875 (95.901)\n",
      "[90/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.3541 (0.3420)\tPrec@1 94.531 (95.862)\n",
      "[100/391]\tTime 0.186 (0.182)\tData 0.019 (0.020)\tLoss 0.4237 (0.3444)\tPrec@1 93.750 (95.738)\n",
      "[110/391]\tTime 0.180 (0.182)\tData 0.020 (0.020)\tLoss 0.3547 (0.3458)\tPrec@1 96.094 (95.693)\n",
      "[120/391]\tTime 0.179 (0.182)\tData 0.019 (0.020)\tLoss 0.3553 (0.3454)\tPrec@1 96.875 (95.713)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3522 (0.3464)\tPrec@1 95.312 (95.694)\n",
      "[140/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3404 (0.3465)\tPrec@1 97.656 (95.673)\n",
      "[150/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.3492 (0.3467)\tPrec@1 95.312 (95.654)\n",
      "[160/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3816 (0.3459)\tPrec@1 95.312 (95.696)\n",
      "[170/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3992 (0.3468)\tPrec@1 95.312 (95.664)\n",
      "[180/391]\tTime 0.177 (0.181)\tData 0.020 (0.020)\tLoss 0.3247 (0.3464)\tPrec@1 97.656 (95.662)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3240 (0.3465)\tPrec@1 96.875 (95.677)\n",
      "[200/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3413 (0.3468)\tPrec@1 96.094 (95.678)\n",
      "[210/391]\tTime 0.186 (0.181)\tData 0.019 (0.020)\tLoss 0.3616 (0.3465)\tPrec@1 93.750 (95.705)\n",
      "[220/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3717 (0.3471)\tPrec@1 96.094 (95.659)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3183 (0.3466)\tPrec@1 96.875 (95.681)\n",
      "[240/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4145 (0.3469)\tPrec@1 91.406 (95.672)\n",
      "[250/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3370 (0.3466)\tPrec@1 96.875 (95.705)\n",
      "[260/391]\tTime 0.148 (0.180)\tData 0.023 (0.020)\tLoss 0.4021 (0.3464)\tPrec@1 93.750 (95.732)\n",
      "[270/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.3041 (0.3463)\tPrec@1 97.656 (95.728)\n",
      "[280/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3454 (0.3468)\tPrec@1 96.875 (95.716)\n",
      "[290/391]\tTime 0.177 (0.180)\tData 0.020 (0.020)\tLoss 0.3207 (0.3471)\tPrec@1 96.094 (95.683)\n",
      "[300/391]\tTime 0.180 (0.180)\tData 0.024 (0.020)\tLoss 0.4503 (0.3480)\tPrec@1 90.625 (95.645)\n",
      "[310/391]\tTime 0.183 (0.180)\tData 0.019 (0.020)\tLoss 0.3389 (0.3484)\tPrec@1 95.312 (95.604)\n",
      "[320/391]\tTime 0.187 (0.180)\tData 0.020 (0.020)\tLoss 0.4625 (0.3485)\tPrec@1 92.188 (95.585)\n",
      "[330/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.3559 (0.3487)\tPrec@1 93.750 (95.570)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.3401 (0.3485)\tPrec@1 95.312 (95.587)\n",
      "[350/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3107 (0.3483)\tPrec@1 96.094 (95.606)\n",
      "[360/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.3841 (0.3485)\tPrec@1 93.750 (95.613)\n",
      "[370/391]\tTime 0.184 (0.180)\tData 0.020 (0.020)\tLoss 0.3437 (0.3483)\tPrec@1 95.312 (95.620)\n",
      "[380/391]\tTime 0.178 (0.180)\tData 0.019 (0.020)\tLoss 0.3594 (0.3486)\tPrec@1 96.094 (95.616)\n",
      "[390/391]\tTime 0.127 (0.180)\tData 0.012 (0.020)\tLoss 0.4558 (0.3496)\tPrec@1 91.250 (95.560)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4350 (0.4350)\tPrec@1 93.750 (93.750)\n",
      "Test: [10/79]\tTime 0.019 (0.023)\tLoss 0.3426 (0.4562)\tPrec@1 92.188 (90.554)\n",
      "Test: [20/79]\tTime 0.025 (0.023)\tLoss 0.4881 (0.4984)\tPrec@1 89.844 (90.476)\n",
      "Test: [30/79]\tTime 0.026 (0.023)\tLoss 0.3243 (0.5201)\tPrec@1 92.969 (90.247)\n",
      "Test: [40/79]\tTime 0.020 (0.023)\tLoss 0.7302 (0.5230)\tPrec@1 85.156 (90.263)\n",
      "Test: [50/79]\tTime 0.021 (0.023)\tLoss 0.6606 (0.5155)\tPrec@1 89.062 (90.288)\n",
      "Test: [60/79]\tTime 0.026 (0.023)\tLoss 0.3847 (0.5027)\tPrec@1 93.750 (90.548)\n",
      "Test: [70/79]\tTime 0.027 (0.023)\tLoss 0.7115 (0.4968)\tPrec@1 89.062 (90.658)\n",
      " * Prec@1 90.710\n",
      "\n",
      "===> epoch: 125/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.185 (0.185)\tData 0.022 (0.022)\tLoss 0.3325 (0.3325)\tPrec@1 96.094 (96.094)\n",
      "[10/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3473 (0.3447)\tPrec@1 97.656 (96.307)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3102 (0.3447)\tPrec@1 96.875 (95.982)\n",
      "[30/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3675 (0.3454)\tPrec@1 93.750 (95.791)\n",
      "[40/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3376 (0.3403)\tPrec@1 97.656 (96.056)\n",
      "[50/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3490 (0.3409)\tPrec@1 96.094 (95.987)\n",
      "[60/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3050 (0.3398)\tPrec@1 97.656 (96.043)\n",
      "[70/391]\tTime 0.187 (0.181)\tData 0.019 (0.020)\tLoss 0.3590 (0.3404)\tPrec@1 95.312 (95.995)\n",
      "[80/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3914 (0.3432)\tPrec@1 95.312 (95.901)\n",
      "[90/391]\tTime 0.182 (0.181)\tData 0.020 (0.020)\tLoss 0.3056 (0.3438)\tPrec@1 97.656 (95.845)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3593 (0.3432)\tPrec@1 94.531 (95.815)\n",
      "[110/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3040 (0.3435)\tPrec@1 98.438 (95.833)\n",
      "[120/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.3122 (0.3445)\tPrec@1 96.875 (95.835)\n",
      "[130/391]\tTime 0.179 (0.181)\tData 0.020 (0.020)\tLoss 0.3764 (0.3446)\tPrec@1 93.750 (95.778)\n",
      "[140/391]\tTime 0.183 (0.181)\tData 0.020 (0.020)\tLoss 0.3868 (0.3456)\tPrec@1 94.531 (95.734)\n",
      "[150/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.4158 (0.3465)\tPrec@1 91.406 (95.669)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3778 (0.3463)\tPrec@1 93.750 (95.672)\n",
      "[170/391]\tTime 0.179 (0.181)\tData 0.021 (0.020)\tLoss 0.3888 (0.3459)\tPrec@1 92.969 (95.692)\n",
      "[180/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3331 (0.3465)\tPrec@1 97.656 (95.684)\n",
      "[190/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3186 (0.3453)\tPrec@1 96.875 (95.717)\n",
      "[200/391]\tTime 0.184 (0.181)\tData 0.020 (0.020)\tLoss 0.3262 (0.3448)\tPrec@1 98.438 (95.752)\n",
      "[210/391]\tTime 0.181 (0.181)\tData 0.020 (0.020)\tLoss 0.3099 (0.3439)\tPrec@1 96.875 (95.779)\n",
      "[220/391]\tTime 0.181 (0.181)\tData 0.021 (0.020)\tLoss 0.3480 (0.3438)\tPrec@1 96.094 (95.797)\n",
      "[230/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3748 (0.3438)\tPrec@1 92.969 (95.786)\n",
      "[240/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3439 (0.3439)\tPrec@1 94.531 (95.786)\n",
      "[250/391]\tTime 0.144 (0.180)\tData 0.020 (0.020)\tLoss 0.3416 (0.3447)\tPrec@1 96.875 (95.754)\n",
      "[260/391]\tTime 0.180 (0.179)\tData 0.020 (0.020)\tLoss 0.3380 (0.3457)\tPrec@1 95.312 (95.708)\n",
      "[270/391]\tTime 0.179 (0.179)\tData 0.021 (0.020)\tLoss 0.4274 (0.3460)\tPrec@1 92.188 (95.690)\n",
      "[280/391]\tTime 0.179 (0.179)\tData 0.019 (0.020)\tLoss 0.3917 (0.3466)\tPrec@1 94.531 (95.682)\n",
      "[290/391]\tTime 0.184 (0.179)\tData 0.020 (0.020)\tLoss 0.3530 (0.3472)\tPrec@1 95.312 (95.664)\n",
      "[300/391]\tTime 0.185 (0.179)\tData 0.020 (0.020)\tLoss 0.3544 (0.3472)\tPrec@1 93.750 (95.673)\n",
      "[310/391]\tTime 0.182 (0.180)\tData 0.019 (0.020)\tLoss 0.3298 (0.3472)\tPrec@1 96.875 (95.662)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.020 (0.020)\tLoss 0.3336 (0.3471)\tPrec@1 95.312 (95.680)\n",
      "[330/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3863 (0.3475)\tPrec@1 92.188 (95.667)\n",
      "[340/391]\tTime 0.180 (0.180)\tData 0.020 (0.020)\tLoss 0.3649 (0.3475)\tPrec@1 95.312 (95.679)\n",
      "[350/391]\tTime 0.178 (0.180)\tData 0.020 (0.020)\tLoss 0.3169 (0.3476)\tPrec@1 96.875 (95.686)\n",
      "[360/391]\tTime 0.186 (0.180)\tData 0.020 (0.020)\tLoss 0.3190 (0.3476)\tPrec@1 97.656 (95.685)\n",
      "[370/391]\tTime 0.184 (0.180)\tData 0.019 (0.020)\tLoss 0.2983 (0.3479)\tPrec@1 97.656 (95.666)\n",
      "[380/391]\tTime 0.181 (0.180)\tData 0.019 (0.020)\tLoss 0.3914 (0.3481)\tPrec@1 94.531 (95.671)\n",
      "[390/391]\tTime 0.126 (0.180)\tData 0.012 (0.020)\tLoss 0.3715 (0.3486)\tPrec@1 96.250 (95.662)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4393 (0.4393)\tPrec@1 92.188 (92.188)\n",
      "Test: [10/79]\tTime 0.021 (0.022)\tLoss 0.3131 (0.4892)\tPrec@1 92.969 (91.264)\n",
      "Test: [20/79]\tTime 0.024 (0.022)\tLoss 0.5044 (0.5520)\tPrec@1 92.188 (90.476)\n",
      "Test: [30/79]\tTime 0.020 (0.022)\tLoss 0.3610 (0.5582)\tPrec@1 93.750 (90.499)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.6273 (0.5494)\tPrec@1 90.625 (90.530)\n",
      "Test: [50/79]\tTime 0.026 (0.022)\tLoss 0.4457 (0.5363)\tPrec@1 89.844 (90.502)\n",
      "Test: [60/79]\tTime 0.020 (0.022)\tLoss 0.4794 (0.5304)\tPrec@1 92.188 (90.727)\n",
      "Test: [70/79]\tTime 0.024 (0.022)\tLoss 0.7024 (0.5263)\tPrec@1 90.625 (90.713)\n",
      " * Prec@1 90.840\n",
      "\n",
      "===> epoch: 126/200\n",
      "current lr 1.00000e-02\n",
      "Training:\n",
      "[0/391]\tTime 0.183 (0.183)\tData 0.020 (0.020)\tLoss 0.3443 (0.3443)\tPrec@1 96.094 (96.094)\n",
      "[10/391]\tTime 0.182 (0.182)\tData 0.021 (0.020)\tLoss 0.4501 (0.3467)\tPrec@1 92.969 (95.810)\n",
      "[20/391]\tTime 0.180 (0.181)\tData 0.019 (0.020)\tLoss 0.4009 (0.3482)\tPrec@1 93.750 (95.945)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[30/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3543 (0.3521)\tPrec@1 94.531 (95.464)\n",
      "[40/391]\tTime 0.177 (0.181)\tData 0.019 (0.020)\tLoss 0.3079 (0.3522)\tPrec@1 98.438 (95.598)\n",
      "[50/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3207 (0.3504)\tPrec@1 97.656 (95.695)\n",
      "[60/391]\tTime 0.186 (0.181)\tData 0.021 (0.020)\tLoss 0.3302 (0.3489)\tPrec@1 95.312 (95.748)\n",
      "[70/391]\tTime 0.185 (0.182)\tData 0.019 (0.020)\tLoss 0.3535 (0.3490)\tPrec@1 95.312 (95.676)\n",
      "[80/391]\tTime 0.178 (0.182)\tData 0.019 (0.020)\tLoss 0.3661 (0.3492)\tPrec@1 94.531 (95.611)\n",
      "[90/391]\tTime 0.179 (0.182)\tData 0.020 (0.020)\tLoss 0.4112 (0.3479)\tPrec@1 92.969 (95.639)\n",
      "[100/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.2974 (0.3469)\tPrec@1 96.875 (95.676)\n",
      "[110/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3038 (0.3471)\tPrec@1 97.656 (95.686)\n",
      "[120/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3341 (0.3480)\tPrec@1 95.312 (95.609)\n",
      "[130/391]\tTime 0.184 (0.181)\tData 0.019 (0.020)\tLoss 0.3832 (0.3479)\tPrec@1 94.531 (95.611)\n",
      "[140/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3299 (0.3471)\tPrec@1 96.875 (95.650)\n",
      "[150/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3629 (0.3466)\tPrec@1 94.531 (95.675)\n",
      "[160/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3447 (0.3459)\tPrec@1 94.531 (95.701)\n",
      "[170/391]\tTime 0.178 (0.181)\tData 0.019 (0.020)\tLoss 0.3753 (0.3461)\tPrec@1 94.531 (95.724)\n",
      "[180/391]\tTime 0.183 (0.181)\tData 0.019 (0.020)\tLoss 0.3696 (0.3469)\tPrec@1 95.312 (95.671)\n",
      "[190/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3619 (0.3470)\tPrec@1 95.312 (95.677)\n",
      "[200/391]\tTime 0.182 (0.181)\tData 0.019 (0.020)\tLoss 0.4258 (0.3471)\tPrec@1 89.844 (95.655)\n",
      "[210/391]\tTime 0.180 (0.181)\tData 0.020 (0.020)\tLoss 0.3350 (0.3469)\tPrec@1 96.875 (95.672)\n",
      "[220/391]\tTime 0.179 (0.181)\tData 0.019 (0.020)\tLoss 0.3378 (0.3464)\tPrec@1 95.312 (95.691)\n",
      "[230/391]\tTime 0.181 (0.181)\tData 0.019 (0.020)\tLoss 0.3685 (0.3465)\tPrec@1 96.094 (95.681)\n",
      "[240/391]\tTime 0.148 (0.180)\tData 0.019 (0.020)\tLoss 0.3992 (0.3464)\tPrec@1 92.969 (95.695)\n",
      "[250/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.4055 (0.3469)\tPrec@1 92.969 (95.670)\n",
      "[260/391]\tTime 0.182 (0.179)\tData 0.019 (0.020)\tLoss 0.3406 (0.3471)\tPrec@1 94.531 (95.678)\n",
      "[270/391]\tTime 0.178 (0.179)\tData 0.019 (0.020)\tLoss 0.3896 (0.3469)\tPrec@1 92.969 (95.670)\n",
      "[280/391]\tTime 0.180 (0.179)\tData 0.019 (0.020)\tLoss 0.4098 (0.3477)\tPrec@1 93.750 (95.649)\n",
      "[290/391]\tTime 0.185 (0.180)\tData 0.020 (0.020)\tLoss 0.3607 (0.3485)\tPrec@1 93.750 (95.613)\n",
      "[300/391]\tTime 0.182 (0.180)\tData 0.022 (0.020)\tLoss 0.4068 (0.3486)\tPrec@1 92.969 (95.601)\n",
      "[310/391]\tTime 0.182 (0.180)\tData 0.020 (0.020)\tLoss 0.3632 (0.3488)\tPrec@1 94.531 (95.601)\n",
      "[320/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3549 (0.3489)\tPrec@1 95.312 (95.602)\n",
      "[330/391]\tTime 0.181 (0.180)\tData 0.020 (0.020)\tLoss 0.3108 (0.3495)\tPrec@1 96.094 (95.572)\n",
      "[340/391]\tTime 0.179 (0.180)\tData 0.019 (0.020)\tLoss 0.3267 (0.3491)\tPrec@1 98.438 (95.599)\n",
      "[350/391]\tTime 0.187 (0.180)\tData 0.021 (0.020)\tLoss 0.3809 (0.3497)\tPrec@1 94.531 (95.571)\n"
     ]
    }
   ],
   "source": [
    "from __future__ import print_function\n",
    "import os\n",
    "import time\n",
    "import logging\n",
    "import argparse\n",
    "import numpy as np\n",
    "from visdom import Visdom\n",
    "from PIL import Image\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "from torch.autograd import Variable\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import datasets, transforms\n",
    "from utils import *\n",
    "from metric.loss import FitNet, AttentionTransfer, RKdAngle, RkdDistance\n",
    "\n",
    "# Teacher models:\n",
    "# VGG11/VGG13/VGG16/VGG19, GoogLeNet, AlxNet, ResNet18, ResNet34, \n",
    "# ResNet50, ResNet101, ResNet152, ResNeXt29_2x64d, ResNeXt29_4x64d, \n",
    "# ResNeXt29_8x64d, ResNeXt29_32x64d, PreActResNet18, PreActResNet34, \n",
    "# PreActResNet50, PreActResNet101, PreActResNet152, \n",
    "# DenseNet121, DenseNet161, DenseNet169, DenseNet201, \n",
    "import models\n",
    "\n",
    "# Student models:\n",
    "# myNet, LeNet, FitNet\n",
    "\n",
    "start_time = time.time()\n",
    "# os.makedirs('./checkpoint', exist_ok=True)\n",
    "\n",
    "# Training settings\n",
    "parser = argparse.ArgumentParser(description='LR_adaptive_learning')\n",
    "\n",
    "parser.add_argument('--dataset',\n",
    "                    choices=['CIFAR10',\n",
    "                             'CIFAR100'\n",
    "                            ],\n",
    "                    default='CIFAR10')\n",
    "parser.add_argument('--teachers',\n",
    "                    choices=['ResNet32',\n",
    "                             'ResNet50',\n",
    "                             'ResNet56',\n",
    "                             'ResNet110',\n",
    "                             'DenseNet121'\n",
    "                            ],\n",
    "                    default=['ResNet32', 'ResNet56', 'ResNet110'],\n",
    "                    nargs='+')\n",
    "parser.add_argument('--student',\n",
    "                    choices=['ResNet20',\n",
    "                             'myNet'\n",
    "                            ],\n",
    "                    default='ResNet20')\n",
    "\n",
    "parser.add_argument('--kd_ratio', default=0.7, type=float)\n",
    "parser.add_argument('--n_class', type=int, default=10, metavar='N', help='num of classes')\n",
    "parser.add_argument('--T', type=float, default=20.0, metavar='Temputure', help='Temputure for distillation')\n",
    "parser.add_argument('--batch_size', type=int, default=128, metavar='N', help='input batch size for training')\n",
    "parser.add_argument('--test_batch_size', type=int, default=128, metavar='N', help='input test batch size for training')\n",
    "parser.add_argument('--epochs', type=int, default=20, metavar='N', help='number of epochs to train (default: 20)')\n",
    "parser.add_argument('--lr', type=float, default=0.1, metavar='LR', help='learning rate (default: 0.01)')\n",
    "parser.add_argument('--momentum', type=float, default=0.9, metavar='M', help='SGD momentum (default: 0.5)')\n",
    "parser.add_argument('--device', default='cuda:0', type=str, help='device: cuda or cpu')\n",
    "parser.add_argument('--print_freq', type=int, default=10, metavar='N', help='how many batches to wait before logging training status')\n",
    "\n",
    "config = ['--epochs', '200', '--T', '5.0', '--device', 'cuda:0']\n",
    "args = parser.parse_args(config)\n",
    "\n",
    "device = args.device if torch.cuda.is_available() else 'cpu'\n",
    "load_dir = './checkpoint/' + args.dataset + '/'\n",
    "\n",
    "# teachers model\n",
    "teacher_models = []\n",
    "for te in args.teachers:\n",
    "    te_model = getattr(models, te)(num_classes=10)\n",
    "#     print(te_model)\n",
    "    te_model.load_state_dict(torch.load(load_dir + te_model.model_name + '.pth'))\n",
    "    te_model.to(device)\n",
    "    teacher_models.append(te_model)\n",
    "\n",
    "st_model = getattr(models, args.student)()  # args.student()\n",
    "st_model.to(device)\n",
    "\n",
    "# logging\n",
    "logfile = load_dir + 'ada_learning_' + st_model.model_name + '.log'\n",
    "if os.path.exists(logfile):\n",
    "    os.remove(logfile)\n",
    "def log_out(info):\n",
    "    f = open(logfile, mode='a')\n",
    "    f.write(info)\n",
    "    f.write('\\n')\n",
    "    f.close()\n",
    "    print(info)\n",
    "    \n",
    "# visualizer\n",
    "vis = Visdom(env='distill')\n",
    "loss_win = vis.line(\n",
    "    X=np.array([0]),\n",
    "    Y=np.array([0]),\n",
    "    opts=dict(\n",
    "            title='ada. loss',\n",
    "        xtickmin=0,\n",
    "#         xtickmax=1,\n",
    "#         xtickstep=5,\n",
    "        ytickmin=0,\n",
    "#         ytickmax=1,\n",
    "        ytickstep=0.5,\n",
    "#         markers=True,\n",
    "#         markersymbol='dot',\n",
    "#         markersize=5,\n",
    "    ),\n",
    "    name=\"loss\"\n",
    ")\n",
    "\n",
    "acc_win = vis.line(\n",
    "    X=np.column_stack((0, 0)),\n",
    "    Y=np.column_stack((0, 0)),\n",
    "    opts=dict(\n",
    "        title='ada. ACC',\n",
    "        xtickmin=0,\n",
    "#         xtickstep=5,\n",
    "        ytickmin=0,\n",
    "        ytickmax=100,\n",
    "#         markers=True,\n",
    "#         markersymbol='dot',\n",
    "#         markersize=5,\n",
    "        legend=['train_acc', 'test_acc']\n",
    "    ),\n",
    "    name=\"acc\"\n",
    ")\n",
    "\n",
    "\n",
    "# adapter model\n",
    "class Adapter():\n",
    "    def __init__(self, in_models, pool_size):\n",
    "        # representations of teachers\n",
    "        pool_ch = pool_size[1]  # 64\n",
    "        pool_w = pool_size[2]   # 8\n",
    "        LR_list = []\n",
    "        torch.manual_seed(1)\n",
    "        self.theta = torch.randn(len(in_models), pool_ch).to(device)  # [3, 64]\n",
    "        self.theta.requires_grad_(True)\n",
    "   \n",
    "        self.max_feat = nn.MaxPool2d(kernel_size=(pool_w, pool_w), stride=pool_w).to(device)\n",
    "        self.W = torch.randn(pool_ch, 1).to(device)\n",
    "        self.W.requires_grad_(True)\n",
    "        self.val = False\n",
    "\n",
    "    def loss(self, y, labels, weighted_logits, T=10.0, alpha=0.7):\n",
    "        y = F.log_softmax(y/T)\n",
    "        kd_loss = nn.KLDivLoss()(y, weighted_logits) * (T*T * 2.0 * alpha) + F.cross_entropy(y, labels) * (1. - alpha)\n",
    "        angle_loss = angle_criterion(y, weighted_logits)\n",
    "        dist_loss = dist_criterion(y, weighted_logits)\n",
    "        ada_loss = kd_loss + angle_loss + dist_loss\n",
    "        # Regularation\n",
    "        if not self.val:\n",
    "            ada_loss += 0.1 * (torch.sum(self.W * self.W) + torch.sum(torch.sum(self.theta * self.theta, dim=1), dim=0))\n",
    "        return ada_loss\n",
    "        \n",
    "    def gradient(self, lr=0.01):\n",
    "        self.W.data = self.W.data - lr * self.W.grad.data\n",
    "        # Manually zero the gradients after updating weights\n",
    "        self.W.grad.data.zero_()\n",
    "        \n",
    "    def eval(self):\n",
    "        self.val = True\n",
    "        self.theta.detach()\n",
    "        self.W.detach()\n",
    "    \n",
    "    # input size: [64, 8, 8], [128, 3, 10]\n",
    "    def forward(self, conv_map, te_logits_list):\n",
    "        beta = self.max_feat(conv_map)\n",
    "        beta = torch.squeeze(beta)  # [128, 64]\n",
    "        \n",
    "        latent_factor = []\n",
    "        for t in self.theta:\n",
    "            latent_factor.append(beta * t)\n",
    "#         latent_factor = torch.stack(latent_factor, dim=0)  # [3, 128, 64]\n",
    "        alpha = []\n",
    "        for lf in latent_factor:  # lf.size:[128, 64]\n",
    "            alpha.append(lf.mm(self.W))\n",
    "        alpha = torch.stack(alpha, dim=0)  # [3, 128, 1]\n",
    "        alpha = torch.squeeze(alpha).transpose(0, 1) # [128, 3]\n",
    "        miu = F.softmax(alpha)  # [128, 3]\n",
    "#         miu = torch.unsqueeze(miu, dim=2)\n",
    "#         weighted_logits = miu * te_logits_list  # [128, 3, 10]\n",
    "#         weighted_logits = torch.sum(weighted_logits, dim=1)\n",
    "#         print(weighted_logits)\n",
    "        \n",
    "        return miu\n",
    "\n",
    "# adapter instance\n",
    "_,_,_,pool_m,_ = st_model(torch.randn(1, 3, 128, 128).to(device))  # get pool_size of student\n",
    "# reate adapter instance\n",
    "adapter = Adapter(teacher_models, pool_m.size())\n",
    "\n",
    "\n",
    "# data\n",
    "normalize = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
    "train_transform = transforms.Compose([\n",
    "    transforms.RandomHorizontalFlip(),\n",
    "    transforms.RandomCrop(32, 4),\n",
    "    transforms.ToTensor(),\n",
    "    normalize,\n",
    "])\n",
    "test_transform = transforms.Compose([transforms.ToTensor(), normalize])\n",
    "train_set = getattr(datasets, args.dataset)(root='../data', train=True, download=True, transform=train_transform)\n",
    "test_set = getattr(datasets, args.dataset)(root='../data', train=False, download=False, transform=test_transform)\n",
    "train_loader = DataLoader(train_set, batch_size=args.batch_size, shuffle=True)\n",
    "test_loader = DataLoader(test_set, batch_size=args.test_batch_size, shuffle=False)\n",
    "# optim\n",
    "optimizer_W = optim.SGD([adapter.W], lr=args.lr, momentum=0.9)\n",
    "optimizer_theta = optim.SGD([adapter.theta], lr=args.lr, momentum=0.9)\n",
    "optimizer_sgd = optim.SGD(st_model.parameters(), lr=args.lr, momentum=0.9, weight_decay=5e-4)\n",
    "lr_scheduler = optim.lr_scheduler.MultiStepLR(optimizer_sgd, gamma=0.1, milestones=[100, 150])\n",
    "lr_scheduler2 = optim.lr_scheduler.MultiStepLR(optimizer_W, milestones=[40, 70])\n",
    "lr_scheduler3 = optim.lr_scheduler.MultiStepLR(optimizer_theta, milestones=[40, 70])\n",
    "\n",
    "# attention transfer loss\n",
    "dist_criterion = RkdDistance().to(device)\n",
    "angle_criterion = RKdAngle().to(device)\n",
    "fitnet_criterion = [FitNet(32, 32), FitNet(32, 32),FitNet(32, 32)]\n",
    "[f.to(device) for f in fitnet_criterion]\n",
    "\n",
    "\n",
    "def train_adapter(n_epochs=70, model=st_model):\n",
    "    print('Training adapter:')\n",
    "    start_time = time.time()\n",
    "    model.train()\n",
    "    for epoch in range(n_epochs):\n",
    "#         lr_scheduler.step(epoch)\n",
    "        lr_scheduler2.step()\n",
    "        lr_scheduler3.step()\n",
    "        for i, (input, target) in enumerate(train_loader):\n",
    "\n",
    "            input, target = input.to(device), target.to(device)\n",
    "            # compute outputs\n",
    "            b1, b2, b3, pool, output = model(input) # out_feat: 16, 32, 64, 64, - \n",
    "            st_maps = [b1, b2, b3, pool]\n",
    "#             print('b1:{}, b2:{}, b3{}, pool:{}'.format(b1.size(), b2.size(), b3.size(), pool.size()))\n",
    "# b1:torch.Size([128, 16, 32, 32]), b2:torch.Size([128, 32, 16, 16]), b3torch.Size([128, 64, 8, 8]), pool:torch.Size([128, 64, 1, 1])\n",
    "\n",
    "            te_scores_list = []\n",
    "            hint_maps = []\n",
    "            fit_loss = 0\n",
    "            for j,te in enumerate(teacher_models):\n",
    "                te.eval()\n",
    "                with torch.no_grad():\n",
    "                    t_b1, t_b2, t_b3, t_pool, t_output = te(input)\n",
    "#                 print('t_b1:{}, t_b2:{}, t_b3:{}, t_pool:{}'.format(t_b1.size(), t_b2.size(), t_b3.size(), t_pool.size()))\n",
    "# t_b1:torch.Size([128, 16, 32, 32]), t_b2:torch.Size([128, 32, 16, 16]), t_b3torch.Size([128, 64, 8, 8]), t_pool:torch.Size([128, 64, 1, 1])\n",
    "                hint_maps.append(t_b2)\n",
    "                t_output = F.softmax(t_output/args.T)\n",
    "                te_scores_list.append(t_output)\n",
    "            te_scores_Tensor = torch.stack(te_scores_list, dim=1)  # size: [128, 3, 10]\n",
    "            \n",
    "            optimizer_sgd.zero_grad()\n",
    "            optimizer_W.zero_grad()\n",
    "            optimizer_theta.zero_grad()\n",
    "            \n",
    "            weight = adapter.forward(pool, te_scores_Tensor)\n",
    "            weight_t = torch.unsqueeze(weight, dim=2)\n",
    "            weighted_logits = weight_t * te_scores_Tensor  # [128, 3, 10]\n",
    "            weighted_logits = torch.sum(weighted_logits, dim=1)\n",
    "            # compute gradient and do SGD step\n",
    "            ada_loss = adapter.loss(output, target, weighted_logits, T=args.T, alpha=args.kd_ratio)\n",
    "            \n",
    "            weight_f = F.softmax(torch.mean(weight, dim=0))  # weight for fit_loss\n",
    "            for j in range(len(teacher_models)):\n",
    "                fit_loss += fitnet_criterion[j](st_maps[1], hint_maps[j]) #weight_f[j] * \n",
    "            \n",
    "#             fit_loss = fitnet_criterion[0](b2, hint_maps[0][3]) + fitnet_criterion[1](b3, hint_maps[1][3]) + fitnet_criterion(pool, hint_maps[2][3])\n",
    "            loss = ada_loss + fit_loss\n",
    "            \n",
    "            loss.backward(retain_graph=True)\n",
    "            optimizer_sgd.step()\n",
    "            optimizer_W.step()\n",
    "            optimizer_theta.step()\n",
    "            \n",
    "#          vis.line(np.array([loss.item()]), np.array([ep]), loss_win, update=\"append\")\n",
    "        log_out('epoch[{}/{}]adapter Loss: {:.4f}'.format(epoch, n_epochs, loss.item()))\n",
    "#         vis.line(np.array([loss.item()]), np.array([epoch]), loss_win, update=\"append\")\n",
    "    end_time = time.time()\n",
    "    log_out(\"--- adapter training cost {:.3f} mins ---\".format((end_time - start_time)/60))\n",
    "\n",
    "\n",
    "# train with multi-teacher\n",
    "def train(epoch, model):\n",
    "    print('Training:')\n",
    "    # switch to train mode\n",
    "    model.train()\n",
    "    adapter.eval()\n",
    "    batch_time = AverageMeter()\n",
    "    data_time = AverageMeter()\n",
    "    losses = AverageMeter()\n",
    "    top1 = AverageMeter()\n",
    "    \n",
    "    end = time.time()\n",
    "    for i, (input, target) in enumerate(train_loader):\n",
    "\n",
    "        # measure data loading time\n",
    "        data_time.update(time.time() - end)\n",
    "\n",
    "        input, target = input.to(device), target.to(device)\n",
    "        \n",
    "        # compute outputs\n",
    "        b1, b2, b3, pool, output = model(input)\n",
    "        st_maps = [b1, b2, b3, pool]\n",
    "        \n",
    "        te_scores_list = []\n",
    "        hint_maps = []\n",
    "        fit_loss = 0\n",
    "        for j,te in enumerate(teacher_models):\n",
    "            te.eval()\n",
    "            with torch.no_grad():\n",
    "                t_b1, t_b2, t_b3, t_pool, t_output = te(input)\n",
    "            hint_maps.append(t_b2)\n",
    "            t_output = F.softmax(t_output/args.T)\n",
    "            te_scores_list.append(t_output)\n",
    "        te_scores_Tensor = torch.stack(te_scores_list, dim=1)  # size: [128, 3, 10]\n",
    "        \n",
    "        optimizer_sgd.zero_grad()\n",
    "        \n",
    "        weight = adapter.forward(pool, te_scores_Tensor)\n",
    "        weight_t = torch.unsqueeze(weight, dim=2)\n",
    "        weighted_logits = weight_t * te_scores_Tensor  # [128, 3, 10]\n",
    "        weighted_logits = torch.sum(weighted_logits, dim=1)\n",
    "        # compute gradient and do SGD step\n",
    "        ada_loss = adapter.loss(output, target, weighted_logits, T=args.T, alpha=args.kd_ratio)\n",
    "        \n",
    "        weight_f = F.softmax(torch.mean(weight, dim=0))  # weight for fit_loss\n",
    "#         for j in range(len(teacher_models)):\n",
    "#             fit_loss += weight_f[j] * fitnet_criterion[j](st_maps[1], hint_maps[j])\n",
    "        loss = ada_loss # + fit_loss\n",
    "\n",
    "        loss.backward(retain_graph=True)\n",
    "        optimizer_sgd.step()\n",
    "\n",
    "        output = output.float()\n",
    "        loss = loss.float()\n",
    "        # measure accuracy and record loss\n",
    "        train_acc = accuracy(output.data, target.data)[0]\n",
    "        losses.update(loss.item(), input.size(0))\n",
    "        top1.update(train_acc, input.size(0))\n",
    "\n",
    "        # measure elapsed time\n",
    "        batch_time.update(time.time() - end)\n",
    "        end = time.time()\n",
    "\n",
    "        if i % args.print_freq == 0:\n",
    "            log_out('[{0}/{1}]\\t'\n",
    "                  'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\\t'\n",
    "                  'Data {data_time.val:.3f} ({data_time.avg:.3f})\\t'\n",
    "                  'Loss {loss.val:.4f} ({loss.avg:.4f})\\t'\n",
    "                  'Prec@1 {top1.val:.3f} ({top1.avg:.3f})'.format(\n",
    "                      i, len(train_loader), batch_time=batch_time,\n",
    "                      data_time=data_time, loss=losses, top1=top1))\n",
    "    return losses.avg, train_acc.cpu().numpy()\n",
    "\n",
    "\n",
    "def test(model):\n",
    "    print('Testing:')\n",
    "    # switch to evaluate mode\n",
    "    model.eval()\n",
    "    batch_time = AverageMeter()\n",
    "    losses = AverageMeter()\n",
    "    top1 = AverageMeter()\n",
    "\n",
    "    end = time.time()\n",
    "    with torch.no_grad():\n",
    "        for i, (input, target) in enumerate(test_loader):\n",
    "            input, target = input.to(device), target.to(device)\n",
    "\n",
    "            # compute output\n",
    "            _,_,_,_,output = model(input)\n",
    "            loss = F.cross_entropy(output, target)\n",
    "\n",
    "            output = output.float()\n",
    "            loss = loss.float()\n",
    "\n",
    "            # measure accuracy and record loss\n",
    "            test_acc = accuracy(output.data, target.data)[0]\n",
    "            losses.update(loss.item(), input.size(0))\n",
    "            top1.update(test_acc, input.size(0))\n",
    "\n",
    "            # measure elapsed time\n",
    "            batch_time.update(time.time() - end)\n",
    "            end = time.time()\n",
    "\n",
    "            if i % args.print_freq == 0:\n",
    "                log_out('Test: [{0}/{1}]\\t'\n",
    "                      'Time {batch_time.val:.3f} ({batch_time.avg:.3f})\\t'\n",
    "                      'Loss {loss.val:.4f} ({loss.avg:.4f})\\t'\n",
    "                      'Prec@1 {top1.val:.3f} ({top1.avg:.3f})'.format(\n",
    "                          i, len(test_loader), batch_time=batch_time, loss=losses,\n",
    "                          top1=top1))\n",
    "\n",
    "    log_out(' * Prec@1 {top1.avg:.3f}'.format(top1=top1))\n",
    "\n",
    "    return losses.avg, test_acc.cpu().numpy(), top1.avg.cpu().numpy()\n",
    "\n",
    "# \"\"\"\n",
    "print('StudentNet:\\n')\n",
    "print(st_model)\n",
    "st_model.apply(weights_init_normal)\n",
    "train_adapter(n_epochs=80)\n",
    "# st_model.apply(weights_init_normal)\n",
    "best_acc = 0\n",
    "for epoch in range(1, args.epochs + 1):\n",
    "    log_out(\"\\n===> epoch: {}/{}\".format(epoch, args.epochs))\n",
    "    log_out('current lr {:.5e}'.format(optimizer_sgd.param_groups[0]['lr']))\n",
    "    lr_scheduler.step(epoch)\n",
    "    train_loss, train_acc = train(epoch, st_model)\n",
    "    # visaulize loss\n",
    "    vis.line(np.array([train_loss]), np.array([epoch]), loss_win, update=\"append\")\n",
    "    _, test_acc, top1 = test(st_model)\n",
    "    vis.line(np.column_stack((train_acc, top1)), np.column_stack((epoch, epoch)), acc_win, update=\"append\")\n",
    "    if top1 > best_acc:\n",
    "        best_acc = top1\n",
    "            \n",
    "# release GPU memory\n",
    "torch.cuda.empty_cache()\n",
    "log_out(\"BEST ACC: {:.3f}\".format(best_acc))\n",
    "log_out(\"--- {:.3f} mins ---\".format((time.time() - start_time)/60))\n",
    "# \"\"\""
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python [conda env:py37] *",
   "language": "python",
   "name": "conda-env-py37-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}