DML_2S.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2019-08-23T02:28:28.763167Z",
     "start_time": "2019-08-23T00:21:51.281680Z"
    }
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/data/yaliu/jupyterbooks/multi-KD/models/student/resnet_s.py:13: 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",
      "*-----------------DML----------------*\n",
      "\n",
      "===> epoch: 1/200\n",
      "Training:\n",
      "[0/391]\tTime 0.057 (0.057)\tData 0.019 (0.019)\tLoss 2.2688 (2.2688)\tPrec@1 (10.938)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/yaliu/Dev/anaconda3/envs/py37/lib/python3.7/site-packages/ipykernel_launcher.py:178: 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:179: 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/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"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 1.7698 (1.9370)\tPrec@1 (18.674)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.5639 (1.8033)\tPrec@1 (23.775)\n",
      "[120/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.5513 (1.7226)\tPrec@1 (27.014)\n",
      "[160/391]\tTime 0.052 (0.051)\tData 0.019 (0.018)\tLoss 1.4284 (1.6573)\tPrec@1 (29.746)\n",
      "[200/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.3395 (1.6031)\tPrec@1 (32.008)\n",
      "[240/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.1944 (1.5540)\tPrec@1 (34.112)\n",
      "[280/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.3028 (1.5202)\tPrec@1 (35.657)\n",
      "[320/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.2896 (1.4846)\tPrec@1 (37.191)\n",
      "[360/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.2679 (1.4511)\tPrec@1 (38.690)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.7067 (1.7067)\tPrec@1 44.531 (44.531)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.9192 (1.7234)\tPrec@1 35.156 (43.083)\n",
      " * ResNet8 Prec@1 42.370\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.9917 (1.9917)\tPrec@1 40.625 (40.625)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 2.0799 (2.1567)\tPrec@1 33.594 (35.366)\n",
      " * ResNet8 Prec@1 35.040\n",
      "\n",
      "===> epoch: 2/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 1.2029 (1.2029)\tPrec@1 (48.438)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 1.3111 (1.1365)\tPrec@1 (52.668)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.0215 (1.1183)\tPrec@1 (53.270)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 1.0378 (1.1053)\tPrec@1 (53.803)\n",
      "[160/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.0806 (1.0985)\tPrec@1 (54.125)\n",
      "[200/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.0320 (1.0851)\tPrec@1 (54.567)\n",
      "[240/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.0029 (1.0833)\tPrec@1 (54.782)\n",
      "[280/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.1101 (1.0668)\tPrec@1 (55.374)\n",
      "[320/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.8983 (1.0607)\tPrec@1 (55.678)\n",
      "[360/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.8916 (1.0511)\tPrec@1 (56.066)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.3460 (1.3460)\tPrec@1 50.000 (50.000)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.4865 (1.3887)\tPrec@1 46.875 (50.381)\n",
      " * ResNet8 Prec@1 50.230\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.2145 (1.2145)\tPrec@1 55.469 (55.469)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2425 (1.2114)\tPrec@1 52.344 (56.155)\n",
      " * ResNet8 Prec@1 55.890\n",
      "\n",
      "===> epoch: 3/200\n",
      "Training:\n",
      "[0/391]\tTime 0.059 (0.059)\tData 0.020 (0.020)\tLoss 0.9086 (0.9086)\tPrec@1 (64.062)\n",
      "[40/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.9464 (0.9364)\tPrec@1 (60.842)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 1.0351 (0.9347)\tPrec@1 (61.140)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 1.0240 (0.9266)\tPrec@1 (61.331)\n",
      "[160/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.9238 (0.9209)\tPrec@1 (61.462)\n",
      "[200/391]\tTime 0.052 (0.051)\tData 0.018 (0.018)\tLoss 0.9247 (0.9265)\tPrec@1 (61.256)\n",
      "[240/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.7712 (0.9242)\tPrec@1 (61.278)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7967 (0.9254)\tPrec@1 (61.238)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.8562 (0.9195)\tPrec@1 (61.565)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 1.0752 (0.9172)\tPrec@1 (61.678)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.3031 (1.3031)\tPrec@1 53.906 (53.906)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.4282 (1.4003)\tPrec@1 50.781 (55.450)\n",
      " * ResNet8 Prec@1 55.170\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0551 (1.0551)\tPrec@1 59.375 (59.375)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.3700 (1.3265)\tPrec@1 49.219 (55.431)\n",
      " * ResNet8 Prec@1 55.100\n",
      "\n",
      "===> epoch: 4/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.020 (0.020)\tLoss 0.6549 (0.6549)\tPrec@1 (70.312)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.8228 (0.8613)\tPrec@1 (64.139)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7333 (0.8460)\tPrec@1 (64.516)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.8184 (0.8501)\tPrec@1 (64.476)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.9088 (0.8490)\tPrec@1 (64.494)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6617 (0.8470)\tPrec@1 (64.599)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.8888 (0.8461)\tPrec@1 (64.542)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6604 (0.8409)\tPrec@1 (64.755)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.9470 (0.8377)\tPrec@1 (64.846)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7596 (0.8322)\tPrec@1 (65.075)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0436 (1.0436)\tPrec@1 60.938 (60.938)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.0199 (1.0471)\tPrec@1 66.406 (63.662)\n",
      " * ResNet8 Prec@1 62.810\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.1349 (1.1349)\tPrec@1 56.250 (56.250)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.3165 (1.2002)\tPrec@1 57.031 (59.432)\n",
      " * ResNet8 Prec@1 59.510\n",
      "\n",
      "===> epoch: 5/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.8240 (0.8240)\tPrec@1 (65.625)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6816 (0.7733)\tPrec@1 (67.454)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 1.0052 (0.7745)\tPrec@1 (67.670)\n",
      "[120/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.8753 (0.7756)\tPrec@1 (67.562)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7617 (0.7724)\tPrec@1 (67.770)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7005 (0.7678)\tPrec@1 (68.027)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6890 (0.7669)\tPrec@1 (68.027)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7362 (0.7620)\tPrec@1 (68.216)\n",
      "[320/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.7579 (0.7605)\tPrec@1 (68.246)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7126 (0.7601)\tPrec@1 (68.328)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9714 (0.9714)\tPrec@1 67.188 (67.188)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0043 (0.9779)\tPrec@1 64.062 (65.530)\n",
      " * ResNet8 Prec@1 65.270\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9932 (0.9932)\tPrec@1 57.031 (57.031)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0714 (1.0382)\tPrec@1 59.375 (62.919)\n",
      " * ResNet8 Prec@1 62.650\n",
      "\n",
      "===> epoch: 6/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.6280 (0.6280)\tPrec@1 (71.094)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6385 (0.7596)\tPrec@1 (68.559)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.8454 (0.7401)\tPrec@1 (68.933)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.8204 (0.7345)\tPrec@1 (69.260)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.8918 (0.7215)\tPrec@1 (69.696)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.8991 (0.7178)\tPrec@1 (69.924)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5940 (0.7158)\tPrec@1 (69.972)\n",
      "[280/391]\tTime 0.055 (0.052)\tData 0.021 (0.018)\tLoss 0.6985 (0.7132)\tPrec@1 (70.132)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7822 (0.7107)\tPrec@1 (70.230)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6974 (0.7076)\tPrec@1 (70.362)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8712 (0.8712)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9633 (0.9062)\tPrec@1 65.625 (67.778)\n",
      " * ResNet8 Prec@1 68.030\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.6546 (1.6546)\tPrec@1 52.344 (52.344)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.7273 (1.7155)\tPrec@1 49.219 (51.353)\n",
      " * ResNet8 Prec@1 51.390\n",
      "\n",
      "===> epoch: 7/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.019 (0.019)\tLoss 0.6330 (0.6330)\tPrec@1 (75.000)\n",
      "[40/391]\tTime 0.051 (0.053)\tData 0.018 (0.019)\tLoss 0.4996 (0.6716)\tPrec@1 (71.684)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.019 (0.018)\tLoss 0.6404 (0.6643)\tPrec@1 (71.788)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7149 (0.6603)\tPrec@1 (72.256)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6359 (0.6610)\tPrec@1 (72.292)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5946 (0.6645)\tPrec@1 (72.198)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6725 (0.6624)\tPrec@1 (72.206)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6925 (0.6630)\tPrec@1 (72.200)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7291 (0.6631)\tPrec@1 (72.121)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7564 (0.6623)\tPrec@1 (72.156)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.1559 (1.1559)\tPrec@1 59.375 (59.375)\n",
      "Test: [40/79]\tTime 0.018 (0.017)\tLoss 1.1381 (1.1310)\tPrec@1 60.156 (61.643)\n",
      " * ResNet8 Prec@1 61.330\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0789 (1.0789)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 1.1719 (1.0656)\tPrec@1 58.594 (63.053)\n",
      " * ResNet8 Prec@1 63.010\n",
      "\n",
      "===> epoch: 8/200\n",
      "Training:\n",
      "[0/391]\tTime 0.052 (0.052)\tData 0.019 (0.019)\tLoss 0.5770 (0.5770)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4954 (0.6369)\tPrec@1 (72.828)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6666 (0.6495)\tPrec@1 (72.637)\n",
      "[120/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6475 (0.6408)\tPrec@1 (72.998)\n",
      "[160/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.7608 (0.6391)\tPrec@1 (73.137)\n",
      "[200/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6128 (0.6380)\tPrec@1 (73.197)\n",
      "[240/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6996 (0.6378)\tPrec@1 (73.117)\n",
      "[280/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5533 (0.6356)\tPrec@1 (73.310)\n",
      "[320/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6225 (0.6345)\tPrec@1 (73.389)\n",
      "[360/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5693 (0.6319)\tPrec@1 (73.470)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9130 (0.9130)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2365 (1.0859)\tPrec@1 60.156 (64.310)\n",
      " * ResNet8 Prec@1 64.800\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7493 (0.7493)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 0.9955 (0.9249)\tPrec@1 64.844 (68.598)\n",
      " * ResNet8 Prec@1 68.920\n",
      "\n",
      "===> epoch: 9/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.020 (0.020)\tLoss 0.6098 (0.6098)\tPrec@1 (72.656)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6248 (0.6210)\tPrec@1 (73.876)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5347 (0.6236)\tPrec@1 (73.669)\n",
      "[120/391]\tTime 0.053 (0.051)\tData 0.018 (0.018)\tLoss 0.6230 (0.6181)\tPrec@1 (73.935)\n",
      "[160/391]\tTime 0.052 (0.051)\tData 0.018 (0.018)\tLoss 0.7709 (0.6171)\tPrec@1 (73.869)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5665 (0.6186)\tPrec@1 (73.682)\n",
      "[240/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.4886 (0.6170)\tPrec@1 (73.762)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5295 (0.6170)\tPrec@1 (73.932)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5669 (0.6121)\tPrec@1 (74.163)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5340 (0.6092)\tPrec@1 (74.236)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9888 (0.9888)\tPrec@1 67.969 (67.969)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2707 (1.1257)\tPrec@1 62.500 (64.825)\n",
      " * ResNet8 Prec@1 64.860\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7533 (0.7533)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9017 (0.8892)\tPrec@1 67.969 (69.779)\n",
      " * ResNet8 Prec@1 69.180\n",
      "\n",
      "===> epoch: 10/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5829 (0.5829)\tPrec@1 (71.875)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6453 (0.5925)\tPrec@1 (74.619)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.7197 (0.6007)\tPrec@1 (74.344)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5221 (0.6087)\tPrec@1 (73.973)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4965 (0.6098)\tPrec@1 (74.136)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6968 (0.6082)\tPrec@1 (74.075)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4516 (0.5998)\tPrec@1 (74.348)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5788 (0.5964)\tPrec@1 (74.486)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5533 (0.5954)\tPrec@1 (74.528)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4980 (0.5952)\tPrec@1 (74.548)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8856 (0.8856)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1077 (1.0614)\tPrec@1 63.281 (66.692)\n",
      " * ResNet8 Prec@1 66.570\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9757 (0.9757)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2589 (1.1368)\tPrec@1 57.812 (63.262)\n",
      " * ResNet8 Prec@1 62.900\n",
      "\n",
      "===> epoch: 11/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4896 (0.4896)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5603 (0.5754)\tPrec@1 (75.762)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4588 (0.5851)\tPrec@1 (75.193)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5433 (0.5825)\tPrec@1 (75.232)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5937 (0.5790)\tPrec@1 (75.432)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7104 (0.5824)\tPrec@1 (75.253)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6938 (0.5829)\tPrec@1 (75.279)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5091 (0.5818)\tPrec@1 (75.336)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5536 (0.5831)\tPrec@1 (75.224)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4337 (0.5800)\tPrec@1 (75.353)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7622 (0.7622)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9129 (0.7642)\tPrec@1 70.312 (74.085)\n",
      " * ResNet8 Prec@1 73.980\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0957 (1.0957)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9782 (0.9974)\tPrec@1 67.188 (66.082)\n",
      " * ResNet8 Prec@1 65.950\n",
      "\n",
      "===> epoch: 12/200\n",
      "Training:\n",
      "[0/391]\tTime 0.062 (0.062)\tData 0.023 (0.023)\tLoss 0.6435 (0.6435)\tPrec@1 (77.344)\n",
      "[40/391]\tTime 0.052 (0.054)\tData 0.018 (0.020)\tLoss 0.6179 (0.5626)\tPrec@1 (75.800)\n",
      "[80/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.4562 (0.5553)\tPrec@1 (76.225)\n",
      "[120/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.5385 (0.5635)\tPrec@1 (75.865)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.019)\tLoss 0.7220 (0.5663)\tPrec@1 (75.776)\n",
      "[200/391]\tTime 0.055 (0.052)\tData 0.021 (0.018)\tLoss 0.6395 (0.5674)\tPrec@1 (75.851)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5197 (0.5670)\tPrec@1 (75.982)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5116 (0.5668)\tPrec@1 (76.020)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.017 (0.018)\tLoss 0.6853 (0.5720)\tPrec@1 (75.798)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5846 (0.5691)\tPrec@1 (75.870)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0308 (1.0308)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 1.0606 (1.0216)\tPrec@1 62.500 (64.005)\n",
      " * ResNet8 Prec@1 63.840\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.016 (0.016)\tLoss 0.7751 (0.7751)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.017 (0.016)\tLoss 0.8756 (0.8645)\tPrec@1 70.312 (69.722)\n",
      " * ResNet8 Prec@1 69.930\n",
      "\n",
      "===> epoch: 13/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5582 (0.5582)\tPrec@1 (77.344)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4663 (0.5731)\tPrec@1 (75.324)\n",
      "[80/391]\tTime 0.052 (0.051)\tData 0.018 (0.018)\tLoss 0.4887 (0.5540)\tPrec@1 (75.993)\n",
      "[120/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5489 (0.5600)\tPrec@1 (75.710)\n",
      "[160/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5634 (0.5565)\tPrec@1 (75.941)\n",
      "[200/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4018 (0.5615)\tPrec@1 (75.843)\n",
      "[240/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.7979 (0.5651)\tPrec@1 (75.772)\n",
      "[280/391]\tTime 0.052 (0.051)\tData 0.018 (0.018)\tLoss 0.5752 (0.5637)\tPrec@1 (75.854)\n",
      "[320/391]\tTime 0.052 (0.051)\tData 0.018 (0.018)\tLoss 0.5690 (0.5636)\tPrec@1 (75.881)\n",
      "[360/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5624 (0.5613)\tPrec@1 (75.918)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7103 (0.7103)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.8071 (0.7612)\tPrec@1 71.875 (74.143)\n",
      " * ResNet8 Prec@1 73.940\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9347 (0.9347)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 0.9590 (0.9317)\tPrec@1 68.750 (68.864)\n",
      " * ResNet8 Prec@1 68.450\n",
      "\n",
      "===> epoch: 14/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5905 (0.5905)\tPrec@1 (75.781)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.017 (0.018)\tLoss 0.5534 (0.5335)\tPrec@1 (77.363)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5935 (0.5407)\tPrec@1 (76.823)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5146 (0.5373)\tPrec@1 (76.937)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.052 (0.052)\tData 0.019 (0.018)\tLoss 0.3517 (0.5405)\tPrec@1 (76.854)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6590 (0.5424)\tPrec@1 (76.765)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4526 (0.5472)\tPrec@1 (76.559)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4772 (0.5441)\tPrec@1 (76.765)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.7697 (0.5443)\tPrec@1 (76.762)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5341 (0.5457)\tPrec@1 (76.712)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8511 (0.8511)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9709 (0.8568)\tPrec@1 63.281 (71.284)\n",
      " * ResNet8 Prec@1 71.420\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.6973 (0.6973)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.027 (0.017)\tLoss 0.8350 (0.7460)\tPrec@1 69.531 (74.505)\n",
      " * ResNet8 Prec@1 74.670\n",
      "\n",
      "===> epoch: 15/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.020 (0.020)\tLoss 0.5101 (0.5101)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4318 (0.5388)\tPrec@1 (77.210)\n",
      "[80/391]\tTime 0.055 (0.052)\tData 0.019 (0.018)\tLoss 0.5646 (0.5427)\tPrec@1 (76.977)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6260 (0.5368)\tPrec@1 (77.124)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6407 (0.5371)\tPrec@1 (77.043)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5065 (0.5362)\tPrec@1 (76.990)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4471 (0.5333)\tPrec@1 (77.052)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5364 (0.5344)\tPrec@1 (76.935)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6151 (0.5383)\tPrec@1 (76.801)\n",
      "[360/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.6837 (0.5399)\tPrec@1 (76.736)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6796 (0.6796)\tPrec@1 76.562 (76.562)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.7312 (0.7017)\tPrec@1 73.438 (75.857)\n",
      " * ResNet8 Prec@1 75.630\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8116 (0.8116)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 0.9097 (0.8155)\tPrec@1 70.312 (72.790)\n",
      " * ResNet8 Prec@1 72.660\n",
      "\n",
      "===> epoch: 16/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4336 (0.4336)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5817 (0.5455)\tPrec@1 (76.810)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6482 (0.5369)\tPrec@1 (76.900)\n",
      "[120/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5741 (0.5331)\tPrec@1 (77.060)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5503 (0.5297)\tPrec@1 (77.145)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5332 (0.5311)\tPrec@1 (77.087)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5123 (0.5315)\tPrec@1 (77.104)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6115 (0.5295)\tPrec@1 (77.157)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4766 (0.5354)\tPrec@1 (76.945)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4874 (0.5342)\tPrec@1 (76.954)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0454 (1.0454)\tPrec@1 63.281 (63.281)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1025 (0.9023)\tPrec@1 65.625 (70.008)\n",
      " * ResNet8 Prec@1 69.430\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8019 (0.8019)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9211 (0.8550)\tPrec@1 67.969 (70.522)\n",
      " * ResNet8 Prec@1 70.030\n",
      "\n",
      "===> epoch: 17/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.020 (0.020)\tLoss 0.5311 (0.5311)\tPrec@1 (74.219)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5310 (0.5215)\tPrec@1 (77.439)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5847 (0.5260)\tPrec@1 (77.488)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5941 (0.5318)\tPrec@1 (77.318)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5189 (0.5355)\tPrec@1 (77.096)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4768 (0.5390)\tPrec@1 (77.041)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5315 (0.5342)\tPrec@1 (77.214)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3661 (0.5296)\tPrec@1 (77.408)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6658 (0.5329)\tPrec@1 (77.227)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5528 (0.5322)\tPrec@1 (77.240)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7616 (0.7616)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8779 (0.7520)\tPrec@1 67.188 (73.971)\n",
      " * ResNet8 Prec@1 73.810\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7617 (0.7617)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0325 (0.9004)\tPrec@1 65.625 (69.741)\n",
      " * ResNet8 Prec@1 69.090\n",
      "\n",
      "===> epoch: 18/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4732 (0.4732)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.051 (0.053)\tData 0.018 (0.019)\tLoss 0.5024 (0.5136)\tPrec@1 (77.839)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5101 (0.5162)\tPrec@1 (77.604)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4390 (0.5255)\tPrec@1 (77.473)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5239 (0.5281)\tPrec@1 (77.421)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4659 (0.5275)\tPrec@1 (77.538)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.3944 (0.5267)\tPrec@1 (77.447)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6765 (0.5314)\tPrec@1 (77.219)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5210 (0.5319)\tPrec@1 (77.200)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4664 (0.5291)\tPrec@1 (77.266)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6823 (0.6823)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.8989 (0.8496)\tPrec@1 67.969 (71.399)\n",
      " * ResNet8 Prec@1 71.840\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0481 (1.0481)\tPrec@1 67.188 (67.188)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.1167 (1.0027)\tPrec@1 64.062 (68.331)\n",
      " * ResNet8 Prec@1 67.420\n",
      "\n",
      "===> epoch: 19/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.5709 (0.5709)\tPrec@1 (73.438)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4440 (0.5168)\tPrec@1 (77.248)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4840 (0.5210)\tPrec@1 (77.218)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5005 (0.5167)\tPrec@1 (77.621)\n",
      "[160/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.5665 (0.5211)\tPrec@1 (77.557)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4028 (0.5218)\tPrec@1 (77.534)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.017 (0.018)\tLoss 0.5792 (0.5196)\tPrec@1 (77.616)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6505 (0.5203)\tPrec@1 (77.558)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.3905 (0.5205)\tPrec@1 (77.555)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4114 (0.5220)\tPrec@1 (77.478)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.2994 (1.2994)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.4493 (1.2467)\tPrec@1 55.469 (62.138)\n",
      " * ResNet8 Prec@1 61.930\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.1503 (1.1503)\tPrec@1 63.281 (63.281)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.7424 (1.3127)\tPrec@1 54.688 (61.490)\n",
      " * ResNet8 Prec@1 61.730\n",
      "\n",
      "===> epoch: 20/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4541 (0.4541)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5399 (0.4908)\tPrec@1 (78.563)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5484 (0.5142)\tPrec@1 (77.768)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4791 (0.5250)\tPrec@1 (77.240)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5000 (0.5220)\tPrec@1 (77.353)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5055 (0.5193)\tPrec@1 (77.476)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5534 (0.5182)\tPrec@1 (77.678)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5844 (0.5204)\tPrec@1 (77.602)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5289 (0.5199)\tPrec@1 (77.682)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4795 (0.5174)\tPrec@1 (77.748)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9623 (0.9623)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.3073 (1.0934)\tPrec@1 63.281 (65.796)\n",
      " * ResNet8 Prec@1 65.910\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6817 (0.6817)\tPrec@1 75.000 (75.000)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.9110 (0.8522)\tPrec@1 75.000 (71.399)\n",
      " * ResNet8 Prec@1 71.570\n",
      "\n",
      "===> epoch: 21/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.020 (0.020)\tLoss 0.5344 (0.5344)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4851 (0.5077)\tPrec@1 (77.458)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5313 (0.5204)\tPrec@1 (77.074)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6372 (0.5268)\tPrec@1 (76.853)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5161 (0.5191)\tPrec@1 (77.247)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4597 (0.5167)\tPrec@1 (77.309)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4811 (0.5174)\tPrec@1 (77.396)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5043 (0.5136)\tPrec@1 (77.575)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4778 (0.5119)\tPrec@1 (77.702)\n",
      "[360/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.6168 (0.5132)\tPrec@1 (77.679)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8914 (0.8914)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1040 (1.0450)\tPrec@1 58.594 (63.167)\n",
      " * ResNet8 Prec@1 62.900\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8703 (0.8703)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0138 (0.8707)\tPrec@1 64.062 (70.332)\n",
      " * ResNet8 Prec@1 70.230\n",
      "\n",
      "===> epoch: 22/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.6345 (0.6345)\tPrec@1 (75.781)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4766 (0.4983)\tPrec@1 (78.601)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3959 (0.5154)\tPrec@1 (77.816)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4977 (0.5183)\tPrec@1 (77.660)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5460 (0.5126)\tPrec@1 (77.795)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5753 (0.5114)\tPrec@1 (77.791)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5678 (0.5120)\tPrec@1 (77.794)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4781 (0.5135)\tPrec@1 (77.736)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4795 (0.5122)\tPrec@1 (77.823)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5434 (0.5140)\tPrec@1 (77.759)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7235 (0.7235)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 0.8832 (0.7670)\tPrec@1 67.188 (74.104)\n",
      " * ResNet8 Prec@1 74.490\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8406 (0.8406)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.0658 (0.9343)\tPrec@1 65.625 (70.427)\n",
      " * ResNet8 Prec@1 70.710\n",
      "\n",
      "===> epoch: 23/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4732 (0.4732)\tPrec@1 (77.344)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5261 (0.5151)\tPrec@1 (77.934)\n",
      "[80/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.5767 (0.5207)\tPrec@1 (77.942)\n",
      "[120/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.3945 (0.5078)\tPrec@1 (78.383)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6202 (0.5058)\tPrec@1 (78.319)\n",
      "[200/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4496 (0.5035)\tPrec@1 (78.331)\n",
      "[240/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4488 (0.5014)\tPrec@1 (78.384)\n",
      "[280/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.6167 (0.5029)\tPrec@1 (78.206)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5710 (0.5043)\tPrec@1 (78.142)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6012 (0.5071)\tPrec@1 (78.012)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8990 (0.8990)\tPrec@1 65.625 (65.625)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.1913 (0.9854)\tPrec@1 61.719 (69.169)\n",
      " * ResNet8 Prec@1 69.120\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9858 (0.9858)\tPrec@1 65.625 (65.625)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 1.2963 (1.0779)\tPrec@1 59.375 (65.854)\n",
      " * ResNet8 Prec@1 66.070\n",
      "\n",
      "===> epoch: 24/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4976 (0.4976)\tPrec@1 (74.219)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.3788 (0.4947)\tPrec@1 (78.792)\n",
      "[80/391]\tTime 0.067 (0.052)\tData 0.032 (0.018)\tLoss 0.5875 (0.4946)\tPrec@1 (78.675)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6842 (0.4950)\tPrec@1 (78.719)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4445 (0.4989)\tPrec@1 (78.499)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6880 (0.5069)\tPrec@1 (78.300)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4691 (0.5030)\tPrec@1 (78.281)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5939 (0.5002)\tPrec@1 (78.322)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6257 (0.5046)\tPrec@1 (78.188)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5590 (0.5027)\tPrec@1 (78.276)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.028 (0.028)\tLoss 0.7927 (0.7927)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.8626 (0.7962)\tPrec@1 71.875 (73.609)\n",
      " * ResNet8 Prec@1 73.170\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8029 (0.8029)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.8908 (0.8909)\tPrec@1 72.656 (70.293)\n",
      " * ResNet8 Prec@1 70.420\n",
      "\n",
      "===> epoch: 25/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4822 (0.4822)\tPrec@1 (82.031)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4200 (0.4845)\tPrec@1 (78.601)\n",
      "[80/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.5207 (0.4966)\tPrec@1 (77.894)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5435 (0.4988)\tPrec@1 (77.970)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4167 (0.4971)\tPrec@1 (78.130)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5296 (0.4980)\tPrec@1 (78.199)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6079 (0.5019)\tPrec@1 (78.028)\n",
      "[280/391]\tTime 0.054 (0.052)\tData 0.019 (0.018)\tLoss 0.7282 (0.5091)\tPrec@1 (77.861)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3756 (0.5084)\tPrec@1 (77.913)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4378 (0.5080)\tPrec@1 (77.911)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6689 (0.6689)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9498 (0.8016)\tPrec@1 66.406 (71.608)\n",
      " * ResNet8 Prec@1 71.550\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7091 (0.7091)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1005 (0.8791)\tPrec@1 63.281 (70.389)\n",
      " * ResNet8 Prec@1 69.810\n",
      "\n",
      "===> epoch: 26/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.019 (0.019)\tLoss 0.4085 (0.4085)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.5431 (0.4929)\tPrec@1 (78.430)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4963 (0.4897)\tPrec@1 (78.675)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5393 (0.4923)\tPrec@1 (78.629)\n",
      "[160/391]\tTime 0.054 (0.052)\tData 0.019 (0.018)\tLoss 0.6207 (0.5020)\tPrec@1 (78.256)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5267 (0.5095)\tPrec@1 (77.935)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5594 (0.5059)\tPrec@1 (78.093)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5244 (0.5058)\tPrec@1 (78.147)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4277 (0.5059)\tPrec@1 (78.152)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6887 (0.5084)\tPrec@1 (78.023)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9838 (0.9838)\tPrec@1 67.188 (67.188)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1136 (1.0816)\tPrec@1 63.281 (64.768)\n",
      " * ResNet8 Prec@1 64.370\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8672 (0.8672)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9887 (0.9631)\tPrec@1 61.719 (68.845)\n",
      " * ResNet8 Prec@1 68.380\n",
      "\n",
      "===> epoch: 27/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5699 (0.5699)\tPrec@1 (78.125)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5921 (0.5212)\tPrec@1 (77.820)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5045 (0.4981)\tPrec@1 (78.443)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4838 (0.4930)\tPrec@1 (78.680)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4408 (0.4886)\tPrec@1 (78.785)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6042 (0.4948)\tPrec@1 (78.560)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4521 (0.4965)\tPrec@1 (78.488)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6833 (0.4987)\tPrec@1 (78.389)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4256 (0.5006)\tPrec@1 (78.290)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5800 (0.5021)\tPrec@1 (78.233)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.3164 (1.3164)\tPrec@1 60.938 (60.938)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2732 (1.1640)\tPrec@1 60.938 (61.319)\n",
      " * ResNet8 Prec@1 61.780\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.3713 (1.3713)\tPrec@1 55.469 (55.469)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.5144 (1.4199)\tPrec@1 56.250 (56.441)\n",
      " * ResNet8 Prec@1 56.310\n",
      "\n",
      "===> epoch: 28/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.019 (0.019)\tLoss 0.4925 (0.4925)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5105 (0.5064)\tPrec@1 (78.392)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4939 (0.5036)\tPrec@1 (78.443)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4372 (0.4952)\tPrec@1 (78.622)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4428 (0.4907)\tPrec@1 (78.654)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4168 (0.4899)\tPrec@1 (78.626)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.7103 (0.4914)\tPrec@1 (78.546)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5161 (0.4947)\tPrec@1 (78.431)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4272 (0.4962)\tPrec@1 (78.451)\n",
      "[360/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.5094 (0.4986)\tPrec@1 (78.354)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0288 (1.0288)\tPrec@1 63.281 (63.281)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2522 (1.0936)\tPrec@1 60.938 (62.938)\n",
      " * ResNet8 Prec@1 62.480\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.2021 (1.2021)\tPrec@1 59.375 (59.375)\n",
      "Test: [40/79]\tTime 0.018 (0.017)\tLoss 1.2459 (1.2140)\tPrec@1 58.594 (61.128)\n",
      " * ResNet8 Prec@1 60.690\n",
      "\n",
      "===> epoch: 29/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5645 (0.5645)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4904 (0.4719)\tPrec@1 (78.811)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4215 (0.4840)\tPrec@1 (78.684)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5408 (0.4914)\tPrec@1 (78.590)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5112 (0.4955)\tPrec@1 (78.484)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3612 (0.4934)\tPrec@1 (78.541)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5386 (0.4914)\tPrec@1 (78.482)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4735 (0.4916)\tPrec@1 (78.542)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4270 (0.4922)\tPrec@1 (78.517)\n",
      "[360/391]\tTime 0.054 (0.052)\tData 0.020 (0.018)\tLoss 0.5095 (0.4922)\tPrec@1 (78.521)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.8130 (0.8130)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9045 (0.8244)\tPrec@1 66.406 (72.694)\n",
      " * ResNet8 Prec@1 72.350\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.5928 (1.5928)\tPrec@1 57.812 (57.812)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.8953 (1.4789)\tPrec@1 52.344 (61.090)\n",
      " * ResNet8 Prec@1 60.350\n",
      "\n",
      "===> epoch: 30/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.020 (0.020)\tLoss 0.4194 (0.4194)\tPrec@1 (84.375)\n",
      "[40/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.5081 (0.5061)\tPrec@1 (78.258)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4817 (0.5001)\tPrec@1 (78.328)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4399 (0.4945)\tPrec@1 (78.558)\n",
      "[160/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.5237 (0.4919)\tPrec@1 (78.707)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4823 (0.4924)\tPrec@1 (78.623)\n",
      "[240/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.5311 (0.4942)\tPrec@1 (78.414)\n",
      "[280/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.6053 (0.4947)\tPrec@1 (78.356)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4944 (0.4992)\tPrec@1 (78.179)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.3014 (0.4977)\tPrec@1 (78.227)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8948 (0.8948)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 1.1511 (0.8208)\tPrec@1 61.719 (71.970)\n",
      " * ResNet8 Prec@1 72.230\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8195 (0.8195)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8512 (0.8651)\tPrec@1 67.188 (71.151)\n",
      " * ResNet8 Prec@1 70.610\n",
      "\n",
      "===> epoch: 31/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4901 (0.4901)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.051 (0.051)\tData 0.018 (0.018)\tLoss 0.4150 (0.4734)\tPrec@1 (78.887)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4157 (0.4874)\tPrec@1 (78.598)\n",
      "[120/391]\tTime 0.051 (0.052)\tData 0.017 (0.018)\tLoss 0.5781 (0.4893)\tPrec@1 (78.493)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5829 (0.4933)\tPrec@1 (78.499)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4823 (0.4881)\tPrec@1 (78.727)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4287 (0.4888)\tPrec@1 (78.728)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.017 (0.018)\tLoss 0.5137 (0.4896)\tPrec@1 (78.645)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4999 (0.4896)\tPrec@1 (78.636)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.017 (0.018)\tLoss 0.3230 (0.4889)\tPrec@1 (78.640)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.0705 (1.0705)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1196 (1.0827)\tPrec@1 65.625 (67.435)\n",
      " * ResNet8 Prec@1 66.800\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8391 (0.8391)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9996 (0.8384)\tPrec@1 64.844 (71.742)\n",
      " * ResNet8 Prec@1 71.470\n",
      "\n",
      "===> epoch: 32/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.019 (0.019)\tLoss 0.4057 (0.4057)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.5604 (0.4872)\tPrec@1 (78.068)\n",
      "[80/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4617 (0.4966)\tPrec@1 (78.279)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5881 (0.5007)\tPrec@1 (78.125)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6223 (0.4984)\tPrec@1 (78.081)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6904 (0.4948)\tPrec@1 (78.211)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5859 (0.4951)\tPrec@1 (78.248)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4009 (0.4948)\tPrec@1 (78.275)\n",
      "[320/391]\tTime 0.060 (0.052)\tData 0.024 (0.018)\tLoss 0.4072 (0.4969)\tPrec@1 (78.298)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5127 (0.4969)\tPrec@1 (78.337)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8525 (0.8525)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9197 (0.8407)\tPrec@1 70.312 (71.761)\n",
      " * ResNet8 Prec@1 71.330\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9919 (0.9919)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0895 (1.0561)\tPrec@1 58.594 (66.120)\n",
      " * ResNet8 Prec@1 66.090\n",
      "\n",
      "===> epoch: 33/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.3945 (0.3945)\tPrec@1 (83.594)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4606 (0.4999)\tPrec@1 (78.468)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4527 (0.4871)\tPrec@1 (79.051)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5418 (0.4851)\tPrec@1 (78.868)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6225 (0.4848)\tPrec@1 (78.853)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5071 (0.4870)\tPrec@1 (78.743)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3445 (0.4841)\tPrec@1 (78.841)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5112 (0.4795)\tPrec@1 (79.048)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4767 (0.4830)\tPrec@1 (78.945)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4358 (0.4839)\tPrec@1 (78.958)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7676 (0.7676)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8114 (0.8122)\tPrec@1 73.438 (73.418)\n",
      " * ResNet8 Prec@1 73.070\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9721 (0.9721)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1989 (0.9854)\tPrec@1 64.062 (67.511)\n",
      " * ResNet8 Prec@1 66.920\n",
      "\n",
      "===> epoch: 34/200\n",
      "Training:\n",
      "[0/391]\tTime 0.056 (0.056)\tData 0.020 (0.020)\tLoss 0.5726 (0.5726)\tPrec@1 (78.125)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4384 (0.4919)\tPrec@1 (78.678)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4702 (0.4905)\tPrec@1 (78.752)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4024 (0.4884)\tPrec@1 (78.822)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4875 (0.4849)\tPrec@1 (78.994)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4470 (0.4892)\tPrec@1 (78.813)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5300 (0.4914)\tPrec@1 (78.686)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5513 (0.4932)\tPrec@1 (78.581)\n",
      "[320/391]\tTime 0.054 (0.052)\tData 0.020 (0.018)\tLoss 0.5799 (0.4951)\tPrec@1 (78.548)\n",
      "[360/391]\tTime 0.054 (0.052)\tData 0.020 (0.018)\tLoss 0.4121 (0.4921)\tPrec@1 (78.653)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6211 (0.6211)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.7772 (0.6948)\tPrec@1 71.875 (75.629)\n",
      " * ResNet8 Prec@1 75.500\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9512 (0.9512)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.017 (0.018)\tLoss 1.4076 (1.1075)\tPrec@1 58.594 (64.425)\n",
      " * ResNet8 Prec@1 64.420\n",
      "\n",
      "===> epoch: 35/200\n",
      "Training:\n",
      "[0/391]\tTime 0.055 (0.055)\tData 0.020 (0.020)\tLoss 0.5231 (0.5231)\tPrec@1 (78.125)\n",
      "[40/391]\tTime 0.055 (0.054)\tData 0.020 (0.019)\tLoss 0.4315 (0.4265)\tPrec@1 (81.155)\n",
      "[80/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.4625 (0.4485)\tPrec@1 (80.343)\n",
      "[120/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.4084 (0.4564)\tPrec@1 (79.901)\n",
      "[160/391]\tTime 0.051 (0.053)\tData 0.018 (0.019)\tLoss 0.5166 (0.4642)\tPrec@1 (79.678)\n",
      "[200/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4223 (0.4739)\tPrec@1 (79.206)\n",
      "[240/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4651 (0.4784)\tPrec@1 (79.049)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.6177 (0.4809)\tPrec@1 (78.990)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4757 (0.4802)\tPrec@1 (79.038)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4187 (0.4815)\tPrec@1 (78.980)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.4262 (1.4262)\tPrec@1 52.344 (52.344)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 1.7423 (1.3926)\tPrec@1 52.344 (59.680)\n",
      " * ResNet8 Prec@1 59.490\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.016 (0.016)\tLoss 0.9171 (0.9171)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 1.0547 (0.9944)\tPrec@1 63.281 (68.769)\n",
      " * ResNet8 Prec@1 68.570\n",
      "\n",
      "===> epoch: 36/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.3608 (0.3608)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4523 (0.4883)\tPrec@1 (78.239)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4403 (0.4795)\tPrec@1 (78.829)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5146 (0.4878)\tPrec@1 (78.725)\n",
      "[160/391]\tTime 0.054 (0.052)\tData 0.020 (0.018)\tLoss 0.4239 (0.4893)\tPrec@1 (78.625)\n",
      "[200/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5715 (0.4887)\tPrec@1 (78.716)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4176 (0.4867)\tPrec@1 (78.757)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4148 (0.4894)\tPrec@1 (78.739)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6061 (0.4892)\tPrec@1 (78.789)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5085 (0.4870)\tPrec@1 (78.908)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6583 (0.6583)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8326 (0.7320)\tPrec@1 64.062 (75.495)\n",
      " * ResNet8 Prec@1 75.380\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 1.1522 (1.1522)\tPrec@1 62.500 (62.500)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.3355 (1.2594)\tPrec@1 57.031 (60.918)\n",
      " * ResNet8 Prec@1 60.680\n",
      "\n",
      "===> epoch: 37/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.4142 (0.4142)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4834 (0.4511)\tPrec@1 (80.259)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5311 (0.4647)\tPrec@1 (79.389)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4819 (0.4749)\tPrec@1 (79.068)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4772 (0.4725)\tPrec@1 (79.265)\n",
      "[200/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.6131 (0.4787)\tPrec@1 (78.949)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3105 (0.4814)\tPrec@1 (78.893)\n",
      "[280/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.4775 (0.4824)\tPrec@1 (78.923)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4805 (0.4837)\tPrec@1 (78.906)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4011 (0.4846)\tPrec@1 (78.861)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8075 (0.8075)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.017 (0.016)\tLoss 0.9584 (0.8412)\tPrec@1 69.531 (72.561)\n",
      " * ResNet8 Prec@1 72.680\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7295 (0.7295)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.016 (0.016)\tLoss 0.9455 (0.8174)\tPrec@1 68.750 (71.704)\n",
      " * ResNet8 Prec@1 71.930\n",
      "\n",
      "===> epoch: 38/200\n",
      "Training:\n",
      "[0/391]\tTime 0.056 (0.056)\tData 0.019 (0.019)\tLoss 0.4048 (0.4048)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4194 (0.4952)\tPrec@1 (78.182)\n",
      "[80/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.5953 (0.4892)\tPrec@1 (78.405)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.2552 (0.4785)\tPrec@1 (78.990)\n",
      "[160/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.4205 (0.4770)\tPrec@1 (79.100)\n",
      "[200/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.4909 (0.4813)\tPrec@1 (78.922)\n",
      "[240/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3765 (0.4787)\tPrec@1 (79.010)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.3840 (0.4780)\tPrec@1 (79.115)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5677 (0.4789)\tPrec@1 (79.069)\n",
      "[360/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5363 (0.4815)\tPrec@1 (78.969)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7321 (0.7321)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8613 (0.8302)\tPrec@1 65.625 (72.409)\n",
      " * ResNet8 Prec@1 72.060\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.8992 (0.8992)\tPrec@1 64.844 (64.844)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9867 (0.9074)\tPrec@1 66.406 (68.941)\n",
      " * ResNet8 Prec@1 68.650\n",
      "\n",
      "===> epoch: 39/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.4361 (0.4361)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5084 (0.4863)\tPrec@1 (78.754)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6051 (0.4842)\tPrec@1 (78.935)\n",
      "[120/391]\tTime 0.058 (0.053)\tData 0.021 (0.018)\tLoss 0.5662 (0.4853)\tPrec@1 (78.880)\n",
      "[160/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.3849 (0.4860)\tPrec@1 (78.935)\n",
      "[200/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4368 (0.4924)\tPrec@1 (78.766)\n",
      "[240/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4174 (0.4872)\tPrec@1 (79.029)\n",
      "[280/391]\tTime 0.053 (0.053)\tData 0.018 (0.018)\tLoss 0.3475 (0.4867)\tPrec@1 (78.956)\n",
      "[320/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.5758 (0.4888)\tPrec@1 (78.816)\n",
      "[360/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4187 (0.4862)\tPrec@1 (78.930)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9325 (0.9325)\tPrec@1 67.188 (67.188)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.3893 (1.0017)\tPrec@1 58.594 (69.341)\n",
      " * ResNet8 Prec@1 68.970\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.5943 (0.5943)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8552 (0.7071)\tPrec@1 69.531 (76.334)\n",
      " * ResNet8 Prec@1 76.000\n",
      "\n",
      "===> epoch: 40/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5946 (0.5946)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.4667 (0.4936)\tPrec@1 (78.487)\n",
      "[80/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.5582 (0.4928)\tPrec@1 (78.366)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4284 (0.4945)\tPrec@1 (78.558)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5126 (0.4917)\tPrec@1 (78.668)\n",
      "[200/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.4905 (0.4868)\tPrec@1 (78.817)\n",
      "[240/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.4655 (0.4849)\tPrec@1 (78.864)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.6123 (0.4852)\tPrec@1 (78.862)\n",
      "[320/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5887 (0.4859)\tPrec@1 (78.843)\n",
      "[360/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5001 (0.4863)\tPrec@1 (78.839)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6177 (0.6177)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.7569 (0.6570)\tPrec@1 72.656 (77.839)\n",
      " * ResNet8 Prec@1 77.660\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6799 (0.6799)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.8109 (0.7387)\tPrec@1 68.750 (75.057)\n",
      " * ResNet8 Prec@1 74.560\n",
      "\n",
      "===> epoch: 41/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.7008 (0.7008)\tPrec@1 (75.000)\n",
      "[40/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4845 (0.4846)\tPrec@1 (79.135)\n",
      "[80/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4986 (0.4708)\tPrec@1 (79.572)\n",
      "[120/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4608 (0.4775)\tPrec@1 (79.326)\n",
      "[160/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.4639 (0.4817)\tPrec@1 (79.227)\n",
      "[200/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.5558 (0.4865)\tPrec@1 (79.000)\n",
      "[240/391]\tTime 0.054 (0.052)\tData 0.018 (0.018)\tLoss 0.4247 (0.4872)\tPrec@1 (78.987)\n",
      "[280/391]\tTime 0.055 (0.052)\tData 0.019 (0.018)\tLoss 0.4139 (0.4836)\tPrec@1 (79.084)\n",
      "[320/391]\tTime 0.053 (0.053)\tData 0.018 (0.018)\tLoss 0.4299 (0.4829)\tPrec@1 (79.179)\n",
      "[360/391]\tTime 0.059 (0.053)\tData 0.025 (0.018)\tLoss 0.5386 (0.4827)\tPrec@1 (79.149)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9381 (0.9381)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0792 (0.9004)\tPrec@1 63.281 (71.132)\n",
      " * ResNet8 Prec@1 70.770\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.6678 (0.6678)\tPrec@1 78.906 (78.906)\n",
      "Test: [40/79]\tTime 0.017 (0.018)\tLoss 0.7959 (0.7453)\tPrec@1 71.094 (74.714)\n",
      " * ResNet8 Prec@1 74.140\n",
      "\n",
      "===> epoch: 42/200\n",
      "Training:\n",
      "[0/391]\tTime 0.057 (0.057)\tData 0.019 (0.019)\tLoss 0.4494 (0.4494)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.6036 (0.4689)\tPrec@1 (79.535)\n",
      "[80/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.7234 (0.4849)\tPrec@1 (78.578)\n",
      "[120/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4916 (0.4847)\tPrec@1 (78.738)\n",
      "[160/391]\tTime 0.053 (0.053)\tData 0.018 (0.018)\tLoss 0.5115 (0.4828)\tPrec@1 (78.848)\n",
      "[200/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4795 (0.4818)\tPrec@1 (78.875)\n",
      "[240/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.6868 (0.4801)\tPrec@1 (79.036)\n",
      "[280/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.3609 (0.4802)\tPrec@1 (79.023)\n",
      "[320/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4589 (0.4793)\tPrec@1 (79.033)\n",
      "[360/391]\tTime 0.053 (0.052)\tData 0.019 (0.018)\tLoss 0.3889 (0.4823)\tPrec@1 (78.947)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8365 (0.8365)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.1766 (1.0453)\tPrec@1 59.375 (68.293)\n",
      " * ResNet8 Prec@1 67.980\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7887 (0.7887)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.016 (0.017)\tLoss 0.9833 (0.8850)\tPrec@1 70.312 (71.094)\n",
      " * ResNet8 Prec@1 70.740\n",
      "\n",
      "===> epoch: 43/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.3743 (0.3743)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.053 (0.052)\tData 0.018 (0.018)\tLoss 0.4561 (0.4641)\tPrec@1 (79.897)\n",
      "[80/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5237 (0.4569)\tPrec@1 (80.102)\n",
      "[120/391]\tTime 0.055 (0.053)\tData 0.018 (0.018)\tLoss 0.4193 (0.4806)\tPrec@1 (79.229)\n",
      "[160/391]\tTime 0.054 (0.053)\tData 0.019 (0.018)\tLoss 0.3883 (0.4769)\tPrec@1 (79.241)\n",
      "[200/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4035 (0.4748)\tPrec@1 (79.307)\n",
      "[240/391]\tTime 0.051 (0.053)\tData 0.018 (0.018)\tLoss 0.5536 (0.4749)\tPrec@1 (79.344)\n",
      "[280/391]\tTime 0.051 (0.053)\tData 0.018 (0.018)\tLoss 0.3608 (0.4724)\tPrec@1 (79.432)\n",
      "[320/391]\tTime 0.053 (0.053)\tData 0.018 (0.018)\tLoss 0.5325 (0.4789)\tPrec@1 (79.225)\n",
      "[360/391]\tTime 0.051 (0.053)\tData 0.018 (0.018)\tLoss 0.5010 (0.4792)\tPrec@1 (79.227)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8614 (0.8614)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2072 (0.9049)\tPrec@1 62.500 (70.789)\n",
      " * ResNet8 Prec@1 70.520\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8098 (0.8098)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.018 (0.017)\tLoss 0.8989 (0.8380)\tPrec@1 70.312 (71.151)\n",
      " * ResNet8 Prec@1 71.190\n",
      "\n",
      "===> epoch: 44/200\n",
      "Training:\n",
      "[0/391]\tTime 0.057 (0.057)\tData 0.019 (0.019)\tLoss 0.4533 (0.4533)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.5720 (0.4830)\tPrec@1 (78.449)\n",
      "[80/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.5519 (0.4730)\tPrec@1 (78.983)\n",
      "[120/391]\tTime 0.054 (0.053)\tData 0.018 (0.018)\tLoss 0.4402 (0.4738)\tPrec@1 (79.061)\n",
      "[160/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.5561 (0.4775)\tPrec@1 (78.931)\n",
      "[200/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4044 (0.4692)\tPrec@1 (79.190)\n",
      "[240/391]\tTime 0.053 (0.053)\tData 0.019 (0.018)\tLoss 0.5254 (0.4721)\tPrec@1 (79.085)\n",
      "[280/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.6268 (0.4728)\tPrec@1 (79.070)\n",
      "[320/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4743 (0.4763)\tPrec@1 (78.918)\n",
      "[360/391]\tTime 0.051 (0.053)\tData 0.018 (0.018)\tLoss 0.3813 (0.4791)\tPrec@1 (78.800)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9076 (0.9076)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.2382 (0.9609)\tPrec@1 57.812 (68.960)\n",
      " * ResNet8 Prec@1 68.940\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8881 (0.8881)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0416 (0.8377)\tPrec@1 67.969 (72.294)\n",
      " * ResNet8 Prec@1 72.240\n",
      "\n",
      "===> epoch: 45/200\n",
      "Training:\n",
      "[0/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.5279 (0.5279)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5844 (0.4627)\tPrec@1 (80.126)\n",
      "[80/391]\tTime 0.055 (0.052)\tData 0.019 (0.018)\tLoss 0.5523 (0.4632)\tPrec@1 (79.842)\n",
      "[120/391]\tTime 0.053 (0.053)\tData 0.018 (0.018)\tLoss 0.3506 (0.4661)\tPrec@1 (79.823)\n",
      "[160/391]\tTime 0.051 (0.052)\tData 0.018 (0.018)\tLoss 0.4392 (0.4670)\tPrec@1 (79.610)\n",
      "[200/391]\tTime 0.054 (0.052)\tData 0.019 (0.018)\tLoss 0.5251 (0.4683)\tPrec@1 (79.536)\n",
      "[240/391]\tTime 0.054 (0.052)\tData 0.018 (0.018)\tLoss 0.4320 (0.4733)\tPrec@1 (79.428)\n",
      "[280/391]\tTime 0.052 (0.052)\tData 0.018 (0.018)\tLoss 0.5317 (0.4730)\tPrec@1 (79.407)\n",
      "[320/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4758 (0.4757)\tPrec@1 (79.291)\n",
      "[360/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.4154 (0.4737)\tPrec@1 (79.419)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.9220 (0.9220)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9283 (0.9046)\tPrec@1 67.188 (70.922)\n",
      " * ResNet8 Prec@1 70.890\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.7950 (0.7950)\tPrec@1 76.562 (76.562)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.9337 (0.8859)\tPrec@1 69.531 (70.103)\n",
      " * ResNet8 Prec@1 70.190\n",
      "\n",
      "===> epoch: 46/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.020 (0.020)\tLoss 0.4874 (0.4874)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.055 (0.054)\tData 0.020 (0.019)\tLoss 0.3726 (0.4683)\tPrec@1 (79.859)\n",
      "[80/391]\tTime 0.053 (0.054)\tData 0.019 (0.019)\tLoss 0.5051 (0.4693)\tPrec@1 (79.639)\n",
      "[120/391]\tTime 0.053 (0.053)\tData 0.018 (0.019)\tLoss 0.6180 (0.4765)\tPrec@1 (79.132)\n",
      "[160/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.4390 (0.4771)\tPrec@1 (79.188)\n",
      "[200/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5459 (0.4716)\tPrec@1 (79.291)\n",
      "[240/391]\tTime 0.054 (0.053)\tData 0.019 (0.019)\tLoss 0.5787 (0.4729)\tPrec@1 (79.273)\n",
      "[280/391]\tTime 0.056 (0.054)\tData 0.021 (0.019)\tLoss 0.4859 (0.4711)\tPrec@1 (79.212)\n",
      "[320/391]\tTime 0.053 (0.054)\tData 0.019 (0.019)\tLoss 0.4668 (0.4728)\tPrec@1 (79.218)\n",
      "[360/391]\tTime 0.053 (0.054)\tData 0.019 (0.019)\tLoss 0.4707 (0.4737)\tPrec@1 (79.194)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.8737 (0.8737)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.017 (0.018)\tLoss 0.9949 (0.8385)\tPrec@1 65.625 (72.694)\n",
      " * ResNet8 Prec@1 72.750\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.8304 (0.8304)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.018 (0.018)\tLoss 1.1821 (0.8754)\tPrec@1 64.062 (71.513)\n",
      " * ResNet8 Prec@1 71.680\n",
      "\n",
      "===> epoch: 47/200\n",
      "Training:\n",
      "[0/391]\tTime 0.055 (0.055)\tData 0.020 (0.020)\tLoss 0.6014 (0.6014)\tPrec@1 (77.344)\n",
      "[40/391]\tTime 0.055 (0.053)\tData 0.019 (0.019)\tLoss 0.2622 (0.4608)\tPrec@1 (79.802)\n",
      "[80/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.5618 (0.4652)\tPrec@1 (79.360)\n",
      "[120/391]\tTime 0.054 (0.053)\tData 0.018 (0.019)\tLoss 0.4827 (0.4704)\tPrec@1 (79.236)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.054 (0.053)\tData 0.018 (0.019)\tLoss 0.5633 (0.4707)\tPrec@1 (79.348)\n",
      "[200/391]\tTime 0.055 (0.054)\tData 0.020 (0.019)\tLoss 0.5013 (0.4723)\tPrec@1 (79.338)\n",
      "[240/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.5863 (0.4729)\tPrec@1 (79.211)\n",
      "[280/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.4189 (0.4726)\tPrec@1 (79.201)\n",
      "[320/391]\tTime 0.054 (0.054)\tData 0.020 (0.019)\tLoss 0.5206 (0.4735)\tPrec@1 (79.101)\n",
      "[360/391]\tTime 0.052 (0.054)\tData 0.018 (0.019)\tLoss 0.4205 (0.4769)\tPrec@1 (79.021)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.0556 (1.0556)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.3620 (1.0740)\tPrec@1 59.375 (67.283)\n",
      " * ResNet8 Prec@1 67.860\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8018 (0.8018)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 0.8591 (0.7168)\tPrec@1 73.438 (76.067)\n",
      " * ResNet8 Prec@1 75.950\n",
      "\n",
      "===> epoch: 48/200\n",
      "Training:\n",
      "[0/391]\tTime 0.058 (0.058)\tData 0.020 (0.020)\tLoss 0.4701 (0.4701)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.054 (0.053)\tData 0.019 (0.019)\tLoss 0.4694 (0.4972)\tPrec@1 (78.678)\n",
      "[80/391]\tTime 0.054 (0.053)\tData 0.018 (0.019)\tLoss 0.3284 (0.4827)\tPrec@1 (78.800)\n",
      "[120/391]\tTime 0.052 (0.054)\tData 0.018 (0.019)\tLoss 0.4940 (0.4834)\tPrec@1 (78.842)\n",
      "[160/391]\tTime 0.055 (0.053)\tData 0.019 (0.019)\tLoss 0.4351 (0.4874)\tPrec@1 (78.722)\n",
      "[200/391]\tTime 0.053 (0.053)\tData 0.018 (0.019)\tLoss 0.4194 (0.4883)\tPrec@1 (78.545)\n",
      "[240/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.3767 (0.4868)\tPrec@1 (78.709)\n",
      "[280/391]\tTime 0.052 (0.053)\tData 0.018 (0.018)\tLoss 0.5005 (0.4857)\tPrec@1 (78.731)\n",
      "[320/391]\tTime 0.051 (0.053)\tData 0.018 (0.018)\tLoss 0.4079 (0.4839)\tPrec@1 (78.819)\n",
      "[360/391]\tTime 0.053 (0.053)\tData 0.018 (0.018)\tLoss 0.4719 (0.4818)\tPrec@1 (78.876)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.017 (0.017)\tLoss 0.8208 (0.8208)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.017 (0.017)\tLoss 1.0566 (0.8249)\tPrec@1 70.312 (72.466)\n",
      " * ResNet8 Prec@1 72.680\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.3223 (1.3223)\tPrec@1 53.906 (53.906)\n",
      "Test: [40/79]\tTime 0.018 (0.017)\tLoss 1.6467 (1.3053)\tPrec@1 58.594 (61.947)\n",
      " * ResNet8 Prec@1 62.280\n",
      "\n",
      "===> epoch: 49/200\n",
      "Training:\n",
      "[0/391]\tTime 0.053 (0.053)\tData 0.019 (0.019)\tLoss 0.6196 (0.6196)\tPrec@1 (73.438)\n",
      "[40/391]\tTime 0.053 (0.054)\tData 0.018 (0.019)\tLoss 0.5243 (0.4841)\tPrec@1 (78.087)\n",
      "[80/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.4191 (0.4722)\tPrec@1 (78.655)\n",
      "[120/391]\tTime 0.052 (0.053)\tData 0.018 (0.019)\tLoss 0.4037 (0.4720)\tPrec@1 (78.842)\n",
      "[160/391]\tTime 0.055 (0.053)\tData 0.019 (0.019)\tLoss 0.5197 (0.4677)\tPrec@1 (79.095)\n",
      "[200/391]\tTime 0.054 (0.053)\tData 0.019 (0.019)\tLoss 0.4862 (0.4665)\tPrec@1 (79.225)\n",
      "[240/391]\tTime 0.056 (0.053)\tData 0.020 (0.019)\tLoss 0.5350 (0.4687)\tPrec@1 (79.243)\n",
      "[280/391]\tTime 0.054 (0.054)\tData 0.020 (0.019)\tLoss 0.3879 (0.4700)\tPrec@1 (79.307)\n",
      "[320/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.3946 (0.4695)\tPrec@1 (79.286)\n",
      "[360/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.4505 (0.4733)\tPrec@1 (79.157)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.1137 (1.1137)\tPrec@1 67.969 (67.969)\n",
      "Test: [40/79]\tTime 0.018 (0.018)\tLoss 1.2529 (1.1754)\tPrec@1 60.156 (63.205)\n",
      " * ResNet8 Prec@1 62.750\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.3262 (1.3262)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.018 (0.018)\tLoss 1.9660 (1.5902)\tPrec@1 52.344 (57.241)\n",
      " * ResNet8 Prec@1 56.620\n",
      "\n",
      "===> epoch: 50/200\n",
      "Training:\n",
      "[0/391]\tTime 0.059 (0.059)\tData 0.020 (0.020)\tLoss 0.5402 (0.5402)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.055 (0.054)\tData 0.019 (0.019)\tLoss 0.4459 (0.4507)\tPrec@1 (79.707)\n",
      "[80/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.4259 (0.4692)\tPrec@1 (79.176)\n",
      "[120/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.4818 (0.4740)\tPrec@1 (79.158)\n",
      "[160/391]\tTime 0.054 (0.054)\tData 0.019 (0.019)\tLoss 0.3449 (0.4762)\tPrec@1 (79.129)\n",
      "[200/391]\tTime 0.056 (0.054)\tData 0.019 (0.019)\tLoss 0.4484 (0.4748)\tPrec@1 (79.244)\n",
      "[240/391]\tTime 0.057 (0.054)\tData 0.020 (0.019)\tLoss 0.5182 (0.4746)\tPrec@1 (79.250)\n",
      "[280/391]\tTime 0.055 (0.055)\tData 0.020 (0.019)\tLoss 0.4242 (0.4747)\tPrec@1 (79.154)\n",
      "[320/391]\tTime 0.089 (0.058)\tData 0.020 (0.020)\tLoss 0.4750 (0.4722)\tPrec@1 (79.218)\n",
      "[360/391]\tTime 0.090 (0.062)\tData 0.020 (0.020)\tLoss 0.4890 (0.4749)\tPrec@1 (79.170)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.5938 (0.5938)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.6902 (0.6649)\tPrec@1 77.344 (77.287)\n",
      " * ResNet8 Prec@1 77.030\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7708 (0.7708)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.022 (0.019)\tLoss 0.9336 (0.7965)\tPrec@1 70.312 (73.819)\n",
      " * ResNet8 Prec@1 73.260\n",
      "\n",
      "===> epoch: 51/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.021 (0.021)\tLoss 0.4552 (0.4552)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.100 (0.095)\tData 0.019 (0.020)\tLoss 0.4634 (0.4593)\tPrec@1 (80.202)\n",
      "[80/391]\tTime 0.102 (0.096)\tData 0.020 (0.020)\tLoss 0.4132 (0.4549)\tPrec@1 (80.006)\n",
      "[120/391]\tTime 0.098 (0.096)\tData 0.020 (0.020)\tLoss 0.4265 (0.4521)\tPrec@1 (80.062)\n",
      "[160/391]\tTime 0.095 (0.096)\tData 0.019 (0.020)\tLoss 0.4077 (0.4572)\tPrec@1 (79.819)\n",
      "[200/391]\tTime 0.096 (0.095)\tData 0.020 (0.020)\tLoss 0.5785 (0.4608)\tPrec@1 (79.649)\n",
      "[240/391]\tTime 0.093 (0.095)\tData 0.019 (0.020)\tLoss 0.4794 (0.4633)\tPrec@1 (79.532)\n",
      "[280/391]\tTime 0.098 (0.095)\tData 0.019 (0.020)\tLoss 0.4971 (0.4677)\tPrec@1 (79.407)\n",
      "[320/391]\tTime 0.098 (0.096)\tData 0.019 (0.020)\tLoss 0.3421 (0.4690)\tPrec@1 (79.342)\n",
      "[360/391]\tTime 0.098 (0.096)\tData 0.019 (0.020)\tLoss 0.5905 (0.4707)\tPrec@1 (79.359)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 0.6076 (0.6076)\tPrec@1 78.125 (78.125)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.7740 (0.7791)\tPrec@1 75.781 (74.085)\n",
      " * ResNet8 Prec@1 73.510\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6178 (0.6178)\tPrec@1 80.469 (80.469)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.7648 (0.6688)\tPrec@1 73.438 (77.420)\n",
      " * ResNet8 Prec@1 77.080\n",
      "\n",
      "===> epoch: 52/200\n",
      "Training:\n",
      "[0/391]\tTime 0.075 (0.075)\tData 0.021 (0.021)\tLoss 0.4408 (0.4408)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.094 (0.095)\tData 0.019 (0.020)\tLoss 0.5308 (0.4531)\tPrec@1 (80.640)\n",
      "[80/391]\tTime 0.087 (0.096)\tData 0.020 (0.020)\tLoss 0.5541 (0.4725)\tPrec@1 (79.639)\n",
      "[120/391]\tTime 0.094 (0.097)\tData 0.021 (0.020)\tLoss 0.5397 (0.4766)\tPrec@1 (79.242)\n",
      "[160/391]\tTime 0.097 (0.098)\tData 0.020 (0.020)\tLoss 0.6008 (0.4847)\tPrec@1 (78.984)\n",
      "[200/391]\tTime 0.097 (0.099)\tData 0.020 (0.020)\tLoss 0.4925 (0.4861)\tPrec@1 (78.957)\n",
      "[240/391]\tTime 0.095 (0.099)\tData 0.020 (0.020)\tLoss 0.5310 (0.4843)\tPrec@1 (79.104)\n",
      "[280/391]\tTime 0.087 (0.099)\tData 0.020 (0.020)\tLoss 0.5572 (0.4808)\tPrec@1 (79.195)\n",
      "[320/391]\tTime 0.090 (0.099)\tData 0.019 (0.020)\tLoss 0.4107 (0.4782)\tPrec@1 (79.298)\n",
      "[360/391]\tTime 0.087 (0.098)\tData 0.019 (0.020)\tLoss 0.5120 (0.4770)\tPrec@1 (79.298)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8489 (0.8489)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 1.0612 (0.9152)\tPrec@1 66.406 (70.751)\n",
      " * ResNet8 Prec@1 70.820\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6845 (0.6845)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.8618 (0.7588)\tPrec@1 71.875 (74.886)\n",
      " * ResNet8 Prec@1 74.400\n",
      "\n",
      "===> epoch: 53/200\n",
      "Training:\n",
      "[0/391]\tTime 0.095 (0.095)\tData 0.021 (0.021)\tLoss 0.4919 (0.4919)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.100 (0.101)\tData 0.019 (0.020)\tLoss 0.3542 (0.4774)\tPrec@1 (79.002)\n",
      "[80/391]\tTime 0.098 (0.101)\tData 0.020 (0.020)\tLoss 0.4339 (0.4746)\tPrec@1 (78.954)\n",
      "[120/391]\tTime 0.098 (0.101)\tData 0.019 (0.020)\tLoss 0.3650 (0.4775)\tPrec@1 (78.893)\n",
      "[160/391]\tTime 0.099 (0.101)\tData 0.019 (0.020)\tLoss 0.3560 (0.4731)\tPrec@1 (79.212)\n",
      "[200/391]\tTime 0.101 (0.101)\tData 0.020 (0.020)\tLoss 0.3846 (0.4707)\tPrec@1 (79.307)\n",
      "[240/391]\tTime 0.098 (0.101)\tData 0.020 (0.020)\tLoss 0.4597 (0.4746)\tPrec@1 (79.221)\n",
      "[280/391]\tTime 0.100 (0.101)\tData 0.019 (0.020)\tLoss 0.4254 (0.4744)\tPrec@1 (79.279)\n",
      "[320/391]\tTime 0.101 (0.101)\tData 0.020 (0.020)\tLoss 0.4400 (0.4758)\tPrec@1 (79.237)\n",
      "[360/391]\tTime 0.102 (0.101)\tData 0.020 (0.020)\tLoss 0.4373 (0.4789)\tPrec@1 (79.138)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.5618 (0.5618)\tPrec@1 80.469 (80.469)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.6424 (0.6640)\tPrec@1 77.344 (77.268)\n",
      " * ResNet8 Prec@1 77.040\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.6002 (0.6002)\tPrec@1 77.344 (77.344)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.9217 (0.7311)\tPrec@1 67.188 (75.419)\n",
      " * ResNet8 Prec@1 75.370\n",
      "\n",
      "===> epoch: 54/200\n",
      "Training:\n",
      "[0/391]\tTime 0.078 (0.078)\tData 0.021 (0.021)\tLoss 0.4063 (0.4063)\tPrec@1 (82.031)\n",
      "[40/391]\tTime 0.099 (0.097)\tData 0.021 (0.020)\tLoss 0.4645 (0.4512)\tPrec@1 (79.954)\n",
      "[80/391]\tTime 0.093 (0.097)\tData 0.021 (0.020)\tLoss 0.6479 (0.4630)\tPrec@1 (79.581)\n",
      "[120/391]\tTime 0.105 (0.097)\tData 0.020 (0.020)\tLoss 0.4622 (0.4670)\tPrec@1 (79.571)\n",
      "[160/391]\tTime 0.102 (0.099)\tData 0.019 (0.020)\tLoss 0.5196 (0.4643)\tPrec@1 (79.615)\n",
      "[200/391]\tTime 0.103 (0.099)\tData 0.019 (0.020)\tLoss 0.4456 (0.4672)\tPrec@1 (79.458)\n",
      "[240/391]\tTime 0.103 (0.099)\tData 0.020 (0.020)\tLoss 0.5352 (0.4704)\tPrec@1 (79.360)\n",
      "[280/391]\tTime 0.098 (0.099)\tData 0.019 (0.020)\tLoss 0.4596 (0.4730)\tPrec@1 (79.268)\n",
      "[320/391]\tTime 0.106 (0.100)\tData 0.020 (0.020)\tLoss 0.5447 (0.4745)\tPrec@1 (79.274)\n",
      "[360/391]\tTime 0.105 (0.100)\tData 0.020 (0.020)\tLoss 0.3760 (0.4725)\tPrec@1 (79.298)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.8677 (0.8677)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 1.1295 (0.9221)\tPrec@1 65.625 (70.370)\n",
      " * ResNet8 Prec@1 70.110\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7549 (0.7549)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.9897 (0.8787)\tPrec@1 69.531 (70.655)\n",
      " * ResNet8 Prec@1 70.420\n",
      "\n",
      "===> epoch: 55/200\n",
      "Training:\n",
      "[0/391]\tTime 0.092 (0.092)\tData 0.021 (0.021)\tLoss 0.4288 (0.4288)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.094 (0.098)\tData 0.020 (0.020)\tLoss 0.4433 (0.4522)\tPrec@1 (79.992)\n",
      "[80/391]\tTime 0.103 (0.095)\tData 0.020 (0.020)\tLoss 0.3580 (0.4694)\tPrec@1 (79.456)\n",
      "[120/391]\tTime 0.069 (0.090)\tData 0.019 (0.020)\tLoss 0.6142 (0.4682)\tPrec@1 (79.629)\n",
      "[160/391]\tTime 0.095 (0.084)\tData 0.020 (0.020)\tLoss 0.4577 (0.4636)\tPrec@1 (79.634)\n",
      "[200/391]\tTime 0.105 (0.087)\tData 0.019 (0.020)\tLoss 0.3638 (0.4628)\tPrec@1 (79.614)\n",
      "[240/391]\tTime 0.104 (0.090)\tData 0.019 (0.019)\tLoss 0.5056 (0.4652)\tPrec@1 (79.444)\n",
      "[280/391]\tTime 0.104 (0.091)\tData 0.020 (0.019)\tLoss 0.5317 (0.4660)\tPrec@1 (79.562)\n",
      "[320/391]\tTime 0.105 (0.093)\tData 0.020 (0.020)\tLoss 0.4540 (0.4662)\tPrec@1 (79.605)\n",
      "[360/391]\tTime 0.107 (0.094)\tData 0.020 (0.020)\tLoss 0.5053 (0.4679)\tPrec@1 (79.614)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.0924 (1.0924)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.2306 (1.1117)\tPrec@1 60.938 (66.787)\n",
      " * ResNet8 Prec@1 66.360\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.8883 (0.8883)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.2337 (0.8841)\tPrec@1 63.281 (71.132)\n",
      " * ResNet8 Prec@1 71.230\n",
      "\n",
      "===> epoch: 56/200\n",
      "Training:\n",
      "[0/391]\tTime 0.105 (0.105)\tData 0.025 (0.025)\tLoss 0.5430 (0.5430)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.106 (0.097)\tData 0.020 (0.020)\tLoss 0.3553 (0.4503)\tPrec@1 (80.450)\n",
      "[80/391]\tTime 0.106 (0.100)\tData 0.019 (0.020)\tLoss 0.3726 (0.4559)\tPrec@1 (80.334)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.6728 (0.4614)\tPrec@1 (80.094)\n",
      "[160/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.5775 (0.4647)\tPrec@1 (79.867)\n",
      "[200/391]\tTime 0.108 (0.101)\tData 0.019 (0.020)\tLoss 0.6411 (0.4711)\tPrec@1 (79.699)\n",
      "[240/391]\tTime 0.092 (0.100)\tData 0.019 (0.020)\tLoss 0.3214 (0.4736)\tPrec@1 (79.606)\n",
      "[280/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.4476 (0.4735)\tPrec@1 (79.615)\n",
      "[320/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.4989 (0.4727)\tPrec@1 (79.571)\n",
      "[360/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.4791 (0.4764)\tPrec@1 (79.378)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.8603 (0.8603)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 1.0246 (0.9670)\tPrec@1 64.844 (67.873)\n",
      " * ResNet8 Prec@1 67.430\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 1.1277 (1.1277)\tPrec@1 62.500 (62.500)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 1.3960 (1.1526)\tPrec@1 61.719 (65.149)\n",
      " * ResNet8 Prec@1 65.280\n",
      "\n",
      "===> epoch: 57/200\n",
      "Training:\n",
      "[0/391]\tTime 0.079 (0.079)\tData 0.021 (0.021)\tLoss 0.4774 (0.4774)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.097 (0.100)\tData 0.019 (0.020)\tLoss 0.3261 (0.4613)\tPrec@1 (79.878)\n",
      "[80/391]\tTime 0.096 (0.100)\tData 0.019 (0.020)\tLoss 0.4365 (0.4552)\tPrec@1 (80.141)\n",
      "[120/391]\tTime 0.093 (0.101)\tData 0.019 (0.020)\tLoss 0.4600 (0.4673)\tPrec@1 (79.649)\n",
      "[160/391]\tTime 0.096 (0.101)\tData 0.019 (0.020)\tLoss 0.5033 (0.4660)\tPrec@1 (79.620)\n",
      "[200/391]\tTime 0.096 (0.101)\tData 0.019 (0.020)\tLoss 0.6102 (0.4687)\tPrec@1 (79.450)\n",
      "[240/391]\tTime 0.095 (0.101)\tData 0.019 (0.020)\tLoss 0.3995 (0.4700)\tPrec@1 (79.422)\n",
      "[280/391]\tTime 0.097 (0.101)\tData 0.020 (0.020)\tLoss 0.4555 (0.4651)\tPrec@1 (79.596)\n",
      "[320/391]\tTime 0.097 (0.101)\tData 0.019 (0.020)\tLoss 0.5905 (0.4648)\tPrec@1 (79.670)\n",
      "[360/391]\tTime 0.097 (0.101)\tData 0.019 (0.020)\tLoss 0.4875 (0.4646)\tPrec@1 (79.638)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9505 (0.9505)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.9762 (0.9455)\tPrec@1 70.312 (70.046)\n",
      " * ResNet8 Prec@1 70.730\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.3524 (1.3524)\tPrec@1 59.375 (59.375)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 1.5958 (1.4986)\tPrec@1 58.594 (60.004)\n",
      " * ResNet8 Prec@1 59.550\n",
      "\n",
      "===> epoch: 58/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.021 (0.021)\tLoss 0.5071 (0.5071)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.102 (0.095)\tData 0.020 (0.020)\tLoss 0.5293 (0.4789)\tPrec@1 (78.792)\n",
      "[80/391]\tTime 0.106 (0.098)\tData 0.019 (0.020)\tLoss 0.3756 (0.4720)\tPrec@1 (79.369)\n",
      "[120/391]\tTime 0.104 (0.099)\tData 0.020 (0.020)\tLoss 0.4760 (0.4924)\tPrec@1 (78.564)\n",
      "[160/391]\tTime 0.107 (0.100)\tData 0.019 (0.020)\tLoss 0.4361 (0.4831)\tPrec@1 (78.867)\n",
      "[200/391]\tTime 0.108 (0.100)\tData 0.019 (0.020)\tLoss 0.3708 (0.4825)\tPrec@1 (78.926)\n",
      "[240/391]\tTime 0.107 (0.100)\tData 0.019 (0.020)\tLoss 0.3610 (0.4792)\tPrec@1 (79.055)\n",
      "[280/391]\tTime 0.093 (0.100)\tData 0.019 (0.020)\tLoss 0.5437 (0.4790)\tPrec@1 (79.134)\n",
      "[320/391]\tTime 0.090 (0.099)\tData 0.020 (0.020)\tLoss 0.5682 (0.4792)\tPrec@1 (79.157)\n",
      "[360/391]\tTime 0.089 (0.099)\tData 0.020 (0.020)\tLoss 0.4838 (0.4776)\tPrec@1 (79.229)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6661 (0.6661)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.7441 (0.7414)\tPrec@1 74.219 (74.943)\n",
      " * ResNet8 Prec@1 75.220\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6128 (0.6128)\tPrec@1 83.594 (83.594)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.8161 (0.7703)\tPrec@1 74.219 (74.600)\n",
      " * ResNet8 Prec@1 74.900\n",
      "\n",
      "===> epoch: 59/200\n",
      "Training:\n",
      "[0/391]\tTime 0.091 (0.091)\tData 0.021 (0.021)\tLoss 0.3693 (0.3693)\tPrec@1 (84.375)\n",
      "[40/391]\tTime 0.096 (0.095)\tData 0.020 (0.020)\tLoss 0.4122 (0.4757)\tPrec@1 (79.078)\n",
      "[80/391]\tTime 0.094 (0.096)\tData 0.020 (0.020)\tLoss 0.3986 (0.4677)\tPrec@1 (79.427)\n",
      "[120/391]\tTime 0.098 (0.096)\tData 0.019 (0.020)\tLoss 0.4420 (0.4675)\tPrec@1 (79.500)\n",
      "[160/391]\tTime 0.101 (0.096)\tData 0.019 (0.020)\tLoss 0.3213 (0.4629)\tPrec@1 (79.654)\n",
      "[200/391]\tTime 0.098 (0.097)\tData 0.020 (0.020)\tLoss 0.5377 (0.4690)\tPrec@1 (79.540)\n",
      "[240/391]\tTime 0.098 (0.097)\tData 0.019 (0.020)\tLoss 0.5737 (0.4711)\tPrec@1 (79.367)\n",
      "[280/391]\tTime 0.098 (0.097)\tData 0.019 (0.020)\tLoss 0.5123 (0.4735)\tPrec@1 (79.259)\n",
      "[320/391]\tTime 0.101 (0.097)\tData 0.019 (0.020)\tLoss 0.5366 (0.4737)\tPrec@1 (79.269)\n",
      "[360/391]\tTime 0.105 (0.098)\tData 0.019 (0.020)\tLoss 0.4747 (0.4731)\tPrec@1 (79.255)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.7799 (0.7799)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.018 (0.022)\tLoss 0.8886 (0.8372)\tPrec@1 71.094 (73.266)\n",
      " * ResNet8 Prec@1 73.530\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.031 (0.031)\tLoss 0.7073 (0.7073)\tPrec@1 78.125 (78.125)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.7624 (0.7593)\tPrec@1 74.219 (74.085)\n",
      " * ResNet8 Prec@1 74.340\n",
      "\n",
      "===> epoch: 60/200\n",
      "Training:\n",
      "[0/391]\tTime 0.072 (0.072)\tData 0.021 (0.021)\tLoss 0.4398 (0.4398)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.106 (0.081)\tData 0.020 (0.020)\tLoss 0.5343 (0.4642)\tPrec@1 (79.535)\n",
      "[80/391]\tTime 0.106 (0.091)\tData 0.020 (0.020)\tLoss 0.3729 (0.4748)\tPrec@1 (79.446)\n",
      "[120/391]\tTime 0.106 (0.095)\tData 0.020 (0.020)\tLoss 0.5491 (0.4748)\tPrec@1 (79.300)\n",
      "[160/391]\tTime 0.110 (0.097)\tData 0.026 (0.020)\tLoss 0.5960 (0.4761)\tPrec@1 (79.202)\n",
      "[200/391]\tTime 0.109 (0.098)\tData 0.020 (0.020)\tLoss 0.5631 (0.4756)\tPrec@1 (79.283)\n",
      "[240/391]\tTime 0.091 (0.098)\tData 0.020 (0.020)\tLoss 0.5345 (0.4804)\tPrec@1 (79.153)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.106 (0.099)\tData 0.020 (0.020)\tLoss 0.5526 (0.4791)\tPrec@1 (79.187)\n",
      "[320/391]\tTime 0.105 (0.099)\tData 0.020 (0.020)\tLoss 0.6126 (0.4763)\tPrec@1 (79.283)\n",
      "[360/391]\tTime 0.102 (0.099)\tData 0.020 (0.020)\tLoss 0.5344 (0.4764)\tPrec@1 (79.328)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6672 (0.6672)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.019 (0.023)\tLoss 0.7282 (0.7018)\tPrec@1 75.781 (75.648)\n",
      " * ResNet8 Prec@1 75.310\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6034 (0.6034)\tPrec@1 78.125 (78.125)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.8641 (0.7127)\tPrec@1 71.094 (75.343)\n",
      " * ResNet8 Prec@1 75.010\n",
      "\n",
      "===> epoch: 61/200\n",
      "Training:\n",
      "[0/391]\tTime 0.095 (0.095)\tData 0.021 (0.021)\tLoss 0.4832 (0.4832)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.105 (0.099)\tData 0.020 (0.020)\tLoss 0.4784 (0.4402)\tPrec@1 (80.507)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.4139 (0.4541)\tPrec@1 (79.832)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.4311 (0.4571)\tPrec@1 (79.752)\n",
      "[160/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.4414 (0.4584)\tPrec@1 (79.780)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5079 (0.4568)\tPrec@1 (79.866)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.3823 (0.4611)\tPrec@1 (79.723)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.4767 (0.4616)\tPrec@1 (79.696)\n",
      "[320/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.4135 (0.4626)\tPrec@1 (79.675)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.5364 (0.4654)\tPrec@1 (79.620)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.6534 (0.6534)\tPrec@1 78.125 (78.125)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8785 (0.8446)\tPrec@1 72.656 (72.332)\n",
      " * ResNet8 Prec@1 72.220\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.0726 (1.0726)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 1.4756 (1.2830)\tPrec@1 62.500 (62.176)\n",
      " * ResNet8 Prec@1 61.820\n",
      "\n",
      "===> epoch: 62/200\n",
      "Training:\n",
      "[0/391]\tTime 0.094 (0.094)\tData 0.021 (0.021)\tLoss 0.4209 (0.4209)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.022 (0.020)\tLoss 0.4195 (0.4587)\tPrec@1 (79.821)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.4299 (0.4689)\tPrec@1 (79.495)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.4272 (0.4646)\tPrec@1 (79.429)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4987 (0.4673)\tPrec@1 (79.571)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4837 (0.4624)\tPrec@1 (79.800)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.022 (0.020)\tLoss 0.5583 (0.4628)\tPrec@1 (79.879)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4688 (0.4655)\tPrec@1 (79.690)\n",
      "[320/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.6565 (0.4657)\tPrec@1 (79.683)\n",
      "[360/391]\tTime 0.100 (0.102)\tData 0.021 (0.020)\tLoss 0.4300 (0.4651)\tPrec@1 (79.698)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7646 (0.7646)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.9166 (0.7587)\tPrec@1 66.406 (74.200)\n",
      " * ResNet8 Prec@1 74.210\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9590 (0.9590)\tPrec@1 65.625 (65.625)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 1.1851 (0.9919)\tPrec@1 61.719 (67.264)\n",
      " * ResNet8 Prec@1 67.410\n",
      "\n",
      "===> epoch: 63/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.022 (0.022)\tLoss 0.5143 (0.5143)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.104 (0.098)\tData 0.020 (0.020)\tLoss 0.4879 (0.4550)\tPrec@1 (79.668)\n",
      "[80/391]\tTime 0.106 (0.100)\tData 0.020 (0.020)\tLoss 0.4012 (0.4609)\tPrec@1 (79.591)\n",
      "[120/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.4048 (0.4672)\tPrec@1 (79.494)\n",
      "[160/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.4210 (0.4623)\tPrec@1 (79.721)\n",
      "[200/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.6464 (0.4666)\tPrec@1 (79.625)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.3365 (0.4666)\tPrec@1 (79.681)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.3787 (0.4650)\tPrec@1 (79.682)\n",
      "[320/391]\tTime 0.107 (0.102)\tData 0.024 (0.020)\tLoss 0.3383 (0.4663)\tPrec@1 (79.649)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.5188 (0.4681)\tPrec@1 (79.633)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.7213 (0.7213)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.9220 (0.7697)\tPrec@1 71.094 (74.733)\n",
      " * ResNet8 Prec@1 74.500\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8764 (0.8764)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.1012 (1.0340)\tPrec@1 67.188 (66.178)\n",
      " * ResNet8 Prec@1 66.270\n",
      "\n",
      "===> epoch: 64/200\n",
      "Training:\n",
      "[0/391]\tTime 0.091 (0.091)\tData 0.021 (0.021)\tLoss 0.5189 (0.5189)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.4241 (0.4467)\tPrec@1 (80.164)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.020 (0.020)\tLoss 0.4564 (0.4482)\tPrec@1 (80.102)\n",
      "[120/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.3778 (0.4436)\tPrec@1 (80.210)\n",
      "[160/391]\tTime 0.103 (0.101)\tData 0.020 (0.020)\tLoss 0.5152 (0.4531)\tPrec@1 (80.061)\n",
      "[200/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.3421 (0.4532)\tPrec@1 (80.049)\n",
      "[240/391]\tTime 0.088 (0.101)\tData 0.020 (0.020)\tLoss 0.5374 (0.4612)\tPrec@1 (79.785)\n",
      "[280/391]\tTime 0.073 (0.098)\tData 0.020 (0.020)\tLoss 0.4540 (0.4635)\tPrec@1 (79.679)\n",
      "[320/391]\tTime 0.066 (0.095)\tData 0.020 (0.020)\tLoss 0.4461 (0.4652)\tPrec@1 (79.622)\n",
      "[360/391]\tTime 0.100 (0.095)\tData 0.020 (0.020)\tLoss 0.4139 (0.4652)\tPrec@1 (79.594)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9690 (0.9690)\tPrec@1 67.188 (67.188)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.9132 (0.9802)\tPrec@1 71.094 (68.693)\n",
      " * ResNet8 Prec@1 68.440\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7325 (0.7325)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.8763 (0.8054)\tPrec@1 66.406 (72.046)\n",
      " * ResNet8 Prec@1 72.330\n",
      "\n",
      "===> epoch: 65/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.021 (0.021)\tLoss 0.4706 (0.4706)\tPrec@1 (77.344)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.021 (0.020)\tLoss 0.3622 (0.4397)\tPrec@1 (80.621)\n",
      "[80/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4578 (0.4468)\tPrec@1 (80.179)\n",
      "[120/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.5279 (0.4594)\tPrec@1 (79.733)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.4388 (0.4687)\tPrec@1 (79.251)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.4300 (0.4721)\tPrec@1 (79.155)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.5280 (0.4728)\tPrec@1 (79.166)\n",
      "[280/391]\tTime 0.105 (0.103)\tData 0.021 (0.021)\tLoss 0.3695 (0.4752)\tPrec@1 (79.087)\n",
      "[320/391]\tTime 0.105 (0.103)\tData 0.020 (0.021)\tLoss 0.5785 (0.4752)\tPrec@1 (79.072)\n",
      "[360/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.5412 (0.4734)\tPrec@1 (79.157)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 1.0093 (1.0093)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.9612 (1.0390)\tPrec@1 69.531 (67.588)\n",
      " * ResNet8 Prec@1 67.520\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7017 (0.7017)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8540 (0.7274)\tPrec@1 67.188 (75.324)\n",
      " * ResNet8 Prec@1 75.570\n",
      "\n",
      "===> epoch: 66/200\n",
      "Training:\n",
      "[0/391]\tTime 0.093 (0.093)\tData 0.022 (0.022)\tLoss 0.3800 (0.3800)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.103 (0.099)\tData 0.020 (0.021)\tLoss 0.4418 (0.4555)\tPrec@1 (79.383)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.020 (0.021)\tLoss 0.5863 (0.4717)\tPrec@1 (79.321)\n",
      "[120/391]\tTime 0.094 (0.101)\tData 0.020 (0.020)\tLoss 0.4528 (0.4693)\tPrec@1 (79.281)\n",
      "[160/391]\tTime 0.102 (0.101)\tData 0.019 (0.020)\tLoss 0.5516 (0.4724)\tPrec@1 (79.212)\n",
      "[200/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.5389 (0.4708)\tPrec@1 (79.291)\n",
      "[240/391]\tTime 0.102 (0.102)\tData 0.019 (0.020)\tLoss 0.4824 (0.4762)\tPrec@1 (79.195)\n",
      "[280/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.5157 (0.4756)\tPrec@1 (79.151)\n",
      "[320/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.4595 (0.4745)\tPrec@1 (79.198)\n",
      "[360/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.4641 (0.4724)\tPrec@1 (79.259)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6449 (0.6449)\tPrec@1 78.906 (78.906)\n",
      "Test: [40/79]\tTime 0.019 (0.022)\tLoss 0.8717 (0.7215)\tPrec@1 72.656 (75.152)\n",
      " * ResNet8 Prec@1 75.520\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 1.3898 (1.3898)\tPrec@1 54.688 (54.688)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 1.4459 (1.4176)\tPrec@1 57.812 (59.546)\n",
      " * ResNet8 Prec@1 58.960\n",
      "\n",
      "===> epoch: 67/200\n",
      "Training:\n",
      "[0/391]\tTime 0.082 (0.082)\tData 0.022 (0.022)\tLoss 0.4143 (0.4143)\tPrec@1 (82.812)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.095 (0.096)\tData 0.021 (0.021)\tLoss 0.3759 (0.4806)\tPrec@1 (78.944)\n",
      "[80/391]\tTime 0.098 (0.098)\tData 0.020 (0.021)\tLoss 0.4288 (0.4758)\tPrec@1 (79.340)\n",
      "[120/391]\tTime 0.096 (0.099)\tData 0.020 (0.021)\tLoss 0.4495 (0.4793)\tPrec@1 (79.171)\n",
      "[160/391]\tTime 0.094 (0.100)\tData 0.019 (0.021)\tLoss 0.5079 (0.4795)\tPrec@1 (79.139)\n",
      "[200/391]\tTime 0.095 (0.100)\tData 0.019 (0.020)\tLoss 0.5335 (0.4708)\tPrec@1 (79.454)\n",
      "[240/391]\tTime 0.092 (0.100)\tData 0.019 (0.020)\tLoss 0.3867 (0.4666)\tPrec@1 (79.606)\n",
      "[280/391]\tTime 0.095 (0.100)\tData 0.020 (0.020)\tLoss 0.3930 (0.4670)\tPrec@1 (79.523)\n",
      "[320/391]\tTime 0.073 (0.100)\tData 0.020 (0.020)\tLoss 0.3514 (0.4653)\tPrec@1 (79.634)\n",
      "[360/391]\tTime 0.105 (0.100)\tData 0.020 (0.020)\tLoss 0.3731 (0.4607)\tPrec@1 (79.770)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.7138 (0.7138)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.8298 (0.7955)\tPrec@1 71.875 (72.847)\n",
      " * ResNet8 Prec@1 72.830\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8181 (0.8181)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.018 (0.021)\tLoss 0.9570 (0.8658)\tPrec@1 67.188 (70.008)\n",
      " * ResNet8 Prec@1 69.920\n",
      "\n",
      "===> epoch: 68/200\n",
      "Training:\n",
      "[0/391]\tTime 0.095 (0.095)\tData 0.021 (0.021)\tLoss 0.5168 (0.5168)\tPrec@1 (75.781)\n",
      "[40/391]\tTime 0.106 (0.099)\tData 0.020 (0.020)\tLoss 0.3439 (0.4517)\tPrec@1 (79.668)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.4658 (0.4590)\tPrec@1 (79.495)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.4330 (0.4578)\tPrec@1 (79.733)\n",
      "[160/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.3690 (0.4638)\tPrec@1 (79.523)\n",
      "[200/391]\tTime 0.109 (0.102)\tData 0.023 (0.020)\tLoss 0.5388 (0.4632)\tPrec@1 (79.614)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.3843 (0.4620)\tPrec@1 (79.623)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5785 (0.4620)\tPrec@1 (79.576)\n",
      "[320/391]\tTime 0.081 (0.102)\tData 0.020 (0.020)\tLoss 0.6102 (0.4626)\tPrec@1 (79.551)\n",
      "[360/391]\tTime 0.100 (0.102)\tData 0.021 (0.020)\tLoss 0.3750 (0.4628)\tPrec@1 (79.523)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9112 (0.9112)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.9198 (0.8719)\tPrec@1 68.750 (71.856)\n",
      " * ResNet8 Prec@1 71.280\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7897 (0.7897)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.8824 (0.7685)\tPrec@1 72.656 (74.314)\n",
      " * ResNet8 Prec@1 74.450\n",
      "\n",
      "===> epoch: 69/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.021 (0.021)\tLoss 0.5492 (0.5492)\tPrec@1 (75.781)\n",
      "[40/391]\tTime 0.098 (0.096)\tData 0.020 (0.020)\tLoss 0.5816 (0.4694)\tPrec@1 (79.402)\n",
      "[80/391]\tTime 0.098 (0.098)\tData 0.020 (0.020)\tLoss 0.4481 (0.4638)\tPrec@1 (79.524)\n",
      "[120/391]\tTime 0.083 (0.096)\tData 0.020 (0.020)\tLoss 0.3642 (0.4603)\tPrec@1 (79.591)\n",
      "[160/391]\tTime 0.072 (0.091)\tData 0.020 (0.020)\tLoss 0.4814 (0.4665)\tPrec@1 (79.513)\n",
      "[200/391]\tTime 0.101 (0.088)\tData 0.020 (0.020)\tLoss 0.4782 (0.4698)\tPrec@1 (79.485)\n",
      "[240/391]\tTime 0.100 (0.089)\tData 0.019 (0.020)\tLoss 0.4293 (0.4667)\tPrec@1 (79.568)\n",
      "[280/391]\tTime 0.103 (0.091)\tData 0.020 (0.020)\tLoss 0.4947 (0.4654)\tPrec@1 (79.576)\n",
      "[320/391]\tTime 0.106 (0.092)\tData 0.020 (0.020)\tLoss 0.3313 (0.4648)\tPrec@1 (79.585)\n",
      "[360/391]\tTime 0.106 (0.093)\tData 0.020 (0.020)\tLoss 0.3158 (0.4674)\tPrec@1 (79.486)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8117 (0.8117)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.0091 (0.9052)\tPrec@1 64.062 (70.465)\n",
      " * ResNet8 Prec@1 70.410\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6195 (0.6195)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.7612 (0.6937)\tPrec@1 77.344 (76.925)\n",
      " * ResNet8 Prec@1 76.500\n",
      "\n",
      "===> epoch: 70/200\n",
      "Training:\n",
      "[0/391]\tTime 0.118 (0.118)\tData 0.034 (0.034)\tLoss 0.4869 (0.4869)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.104 (0.100)\tData 0.020 (0.021)\tLoss 0.5962 (0.4733)\tPrec@1 (79.249)\n",
      "[80/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.3811 (0.4573)\tPrec@1 (80.150)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.3525 (0.4627)\tPrec@1 (79.946)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5389 (0.4618)\tPrec@1 (79.940)\n",
      "[200/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.5059 (0.4628)\tPrec@1 (79.878)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.5765 (0.4647)\tPrec@1 (79.788)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.3972 (0.4678)\tPrec@1 (79.679)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.5518 (0.4669)\tPrec@1 (79.700)\n",
      "[360/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.4139 (0.4663)\tPrec@1 (79.694)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7690 (0.7690)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.7139 (0.7243)\tPrec@1 75.781 (75.972)\n",
      " * ResNet8 Prec@1 75.370\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.5997 (0.5997)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.8319 (0.7446)\tPrec@1 71.094 (75.629)\n",
      " * ResNet8 Prec@1 74.930\n",
      "\n",
      "===> epoch: 71/200\n",
      "Training:\n",
      "[0/391]\tTime 0.108 (0.108)\tData 0.023 (0.023)\tLoss 0.5196 (0.5196)\tPrec@1 (78.125)\n",
      "[40/391]\tTime 0.107 (0.104)\tData 0.021 (0.021)\tLoss 0.5186 (0.4798)\tPrec@1 (79.059)\n",
      "[80/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.4668 (0.4735)\tPrec@1 (79.668)\n",
      "[120/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.4904 (0.4740)\tPrec@1 (79.429)\n",
      "[160/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.5650 (0.4749)\tPrec@1 (79.251)\n",
      "[200/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.5190 (0.4749)\tPrec@1 (79.369)\n",
      "[240/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.5114 (0.4775)\tPrec@1 (79.243)\n",
      "[280/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4231 (0.4731)\tPrec@1 (79.387)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.3273 (0.4702)\tPrec@1 (79.427)\n",
      "[360/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.5703 (0.4706)\tPrec@1 (79.456)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6817 (0.6817)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.024 (0.020)\tLoss 0.7628 (0.7159)\tPrec@1 71.094 (75.495)\n",
      " * ResNet8 Prec@1 75.290\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9711 (0.9711)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 1.3644 (1.0933)\tPrec@1 59.375 (64.043)\n",
      " * ResNet8 Prec@1 63.730\n",
      "\n",
      "===> epoch: 72/200\n",
      "Training:\n",
      "[0/391]\tTime 0.084 (0.084)\tData 0.021 (0.021)\tLoss 0.4482 (0.4482)\tPrec@1 (82.812)\n",
      "[40/391]\tTime 0.102 (0.100)\tData 0.020 (0.020)\tLoss 0.5920 (0.4794)\tPrec@1 (79.630)\n",
      "[80/391]\tTime 0.098 (0.100)\tData 0.020 (0.020)\tLoss 0.4158 (0.4679)\tPrec@1 (79.659)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.4204 (0.4699)\tPrec@1 (79.390)\n",
      "[160/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.3928 (0.4685)\tPrec@1 (79.610)\n",
      "[200/391]\tTime 0.107 (0.101)\tData 0.020 (0.020)\tLoss 0.3400 (0.4674)\tPrec@1 (79.520)\n",
      "[240/391]\tTime 0.108 (0.101)\tData 0.021 (0.020)\tLoss 0.4844 (0.4642)\tPrec@1 (79.704)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.4722 (0.4632)\tPrec@1 (79.676)\n",
      "[320/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.4984 (0.4638)\tPrec@1 (79.680)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.5353 (0.4644)\tPrec@1 (79.713)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6276 (0.6276)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8291 (0.6858)\tPrec@1 71.875 (76.925)\n",
      " * ResNet8 Prec@1 77.070\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7556 (0.7556)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8634 (0.7837)\tPrec@1 72.656 (73.704)\n",
      " * ResNet8 Prec@1 73.390\n",
      "\n",
      "===> epoch: 73/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.4933 (0.4933)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.3701 (0.4446)\tPrec@1 (80.564)\n",
      "[80/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.5027 (0.4604)\tPrec@1 (80.199)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5740 (0.4605)\tPrec@1 (80.023)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.4288 (0.4646)\tPrec@1 (79.702)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4863 (0.4663)\tPrec@1 (79.583)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5398 (0.4656)\tPrec@1 (79.639)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4627 (0.4644)\tPrec@1 (79.635)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.021 (0.020)\tLoss 0.4076 (0.4653)\tPrec@1 (79.593)\n",
      "[360/391]\tTime 0.104 (0.103)\tData 0.021 (0.020)\tLoss 0.6500 (0.4642)\tPrec@1 (79.666)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.5487 (0.5487)\tPrec@1 82.031 (82.031)\n",
      "Test: [40/79]\tTime 0.021 (0.022)\tLoss 0.8099 (0.6770)\tPrec@1 75.000 (76.677)\n",
      " * ResNet8 Prec@1 76.940\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 0.8119 (0.8119)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 1.1849 (0.9477)\tPrec@1 67.188 (69.741)\n",
      " * ResNet8 Prec@1 69.560\n",
      "\n",
      "===> epoch: 74/200\n",
      "Training:\n",
      "[0/391]\tTime 0.096 (0.096)\tData 0.023 (0.023)\tLoss 0.5006 (0.5006)\tPrec@1 (83.594)\n",
      "[40/391]\tTime 0.061 (0.070)\tData 0.020 (0.021)\tLoss 0.4286 (0.4597)\tPrec@1 (79.478)\n",
      "[80/391]\tTime 0.105 (0.079)\tData 0.021 (0.021)\tLoss 0.4123 (0.4606)\tPrec@1 (79.745)\n",
      "[120/391]\tTime 0.106 (0.087)\tData 0.021 (0.021)\tLoss 0.5412 (0.4565)\tPrec@1 (79.739)\n",
      "[160/391]\tTime 0.105 (0.091)\tData 0.020 (0.020)\tLoss 0.4737 (0.4569)\tPrec@1 (79.828)\n",
      "[200/391]\tTime 0.106 (0.093)\tData 0.020 (0.020)\tLoss 0.5568 (0.4596)\tPrec@1 (79.730)\n",
      "[240/391]\tTime 0.105 (0.094)\tData 0.019 (0.020)\tLoss 0.4157 (0.4653)\tPrec@1 (79.499)\n",
      "[280/391]\tTime 0.109 (0.095)\tData 0.021 (0.020)\tLoss 0.5239 (0.4646)\tPrec@1 (79.512)\n",
      "[320/391]\tTime 0.096 (0.096)\tData 0.020 (0.020)\tLoss 0.4836 (0.4657)\tPrec@1 (79.503)\n",
      "[360/391]\tTime 0.098 (0.097)\tData 0.020 (0.020)\tLoss 0.4345 (0.4633)\tPrec@1 (79.625)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.8628 (0.8628)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8737 (0.7367)\tPrec@1 66.406 (74.085)\n",
      " * ResNet8 Prec@1 74.000\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6058 (0.6058)\tPrec@1 78.125 (78.125)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8328 (0.6801)\tPrec@1 72.656 (76.391)\n",
      " * ResNet8 Prec@1 76.400\n",
      "\n",
      "===> epoch: 75/200\n",
      "Training:\n",
      "[0/391]\tTime 0.105 (0.105)\tData 0.021 (0.021)\tLoss 0.2915 (0.2915)\tPrec@1 (86.719)\n",
      "[40/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.4158 (0.4293)\tPrec@1 (80.831)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.6076 (0.4370)\tPrec@1 (80.720)\n",
      "[120/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.5982 (0.4491)\tPrec@1 (80.346)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.021 (0.020)\tLoss 0.3849 (0.4533)\tPrec@1 (80.081)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.5975 (0.4569)\tPrec@1 (79.983)\n",
      "[240/391]\tTime 0.107 (0.103)\tData 0.021 (0.021)\tLoss 0.3310 (0.4551)\tPrec@1 (80.141)\n",
      "[280/391]\tTime 0.108 (0.103)\tData 0.022 (0.021)\tLoss 0.5355 (0.4616)\tPrec@1 (79.832)\n",
      "[320/391]\tTime 0.105 (0.103)\tData 0.021 (0.021)\tLoss 0.3958 (0.4650)\tPrec@1 (79.748)\n",
      "[360/391]\tTime 0.102 (0.103)\tData 0.020 (0.021)\tLoss 0.3719 (0.4665)\tPrec@1 (79.657)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.2089 (1.2089)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.2494 (1.1821)\tPrec@1 64.062 (65.644)\n",
      " * ResNet8 Prec@1 65.210\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8403 (0.8403)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.026 (0.021)\tLoss 0.8937 (0.8352)\tPrec@1 66.406 (71.284)\n",
      " * ResNet8 Prec@1 71.150\n",
      "\n",
      "===> epoch: 76/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.021 (0.021)\tLoss 0.4343 (0.4343)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4677 (0.4763)\tPrec@1 (79.078)\n",
      "[80/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.4667 (0.4791)\tPrec@1 (78.906)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.5098 (0.4692)\tPrec@1 (79.436)\n",
      "[160/391]\tTime 0.111 (0.103)\tData 0.021 (0.021)\tLoss 0.5094 (0.4745)\tPrec@1 (79.115)\n",
      "[200/391]\tTime 0.108 (0.103)\tData 0.021 (0.021)\tLoss 0.5401 (0.4697)\tPrec@1 (79.392)\n",
      "[240/391]\tTime 0.108 (0.103)\tData 0.021 (0.021)\tLoss 0.4334 (0.4657)\tPrec@1 (79.597)\n",
      "[280/391]\tTime 0.109 (0.103)\tData 0.020 (0.021)\tLoss 0.4144 (0.4673)\tPrec@1 (79.654)\n",
      "[320/391]\tTime 0.103 (0.103)\tData 0.023 (0.021)\tLoss 0.2773 (0.4682)\tPrec@1 (79.593)\n",
      "[360/391]\tTime 0.096 (0.103)\tData 0.019 (0.021)\tLoss 0.4559 (0.4679)\tPrec@1 (79.616)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7130 (0.7130)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.7327 (0.7209)\tPrec@1 77.344 (74.924)\n",
      " * ResNet8 Prec@1 74.350\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7217 (0.7217)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.9262 (0.7504)\tPrec@1 68.750 (75.534)\n",
      " * ResNet8 Prec@1 75.480\n",
      "\n",
      "===> epoch: 77/200\n",
      "Training:\n",
      "[0/391]\tTime 0.093 (0.093)\tData 0.021 (0.021)\tLoss 0.4287 (0.4287)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.103 (0.097)\tData 0.020 (0.020)\tLoss 0.3674 (0.4346)\tPrec@1 (80.412)\n",
      "[80/391]\tTime 0.106 (0.100)\tData 0.020 (0.020)\tLoss 0.4505 (0.4424)\tPrec@1 (80.285)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.5244 (0.4571)\tPrec@1 (79.733)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4633 (0.4576)\tPrec@1 (79.906)\n",
      "[200/391]\tTime 0.108 (0.101)\tData 0.020 (0.020)\tLoss 0.5280 (0.4588)\tPrec@1 (79.897)\n",
      "[240/391]\tTime 0.096 (0.101)\tData 0.020 (0.020)\tLoss 0.3248 (0.4579)\tPrec@1 (79.788)\n",
      "[280/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.4054 (0.4579)\tPrec@1 (79.746)\n",
      "[320/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.4503 (0.4586)\tPrec@1 (79.799)\n",
      "[360/391]\tTime 0.094 (0.102)\tData 0.020 (0.020)\tLoss 0.4012 (0.4606)\tPrec@1 (79.765)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6294 (0.6294)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.019 (0.022)\tLoss 0.8261 (0.6920)\tPrec@1 69.531 (76.010)\n",
      " * ResNet8 Prec@1 75.810\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.026 (0.026)\tLoss 0.7125 (0.7125)\tPrec@1 76.562 (76.562)\n",
      "Test: [40/79]\tTime 0.021 (0.022)\tLoss 0.8306 (0.7400)\tPrec@1 71.094 (75.324)\n",
      " * ResNet8 Prec@1 74.970\n",
      "\n",
      "===> epoch: 78/200\n",
      "Training:\n",
      "[0/391]\tTime 0.089 (0.089)\tData 0.023 (0.023)\tLoss 0.4385 (0.4385)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.102 (0.102)\tData 0.021 (0.020)\tLoss 0.4800 (0.4771)\tPrec@1 (79.383)\n",
      "[80/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.5368 (0.4682)\tPrec@1 (79.765)\n",
      "[120/391]\tTime 0.103 (0.103)\tData 0.020 (0.020)\tLoss 0.5464 (0.4708)\tPrec@1 (79.584)\n",
      "[160/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.4909 (0.4667)\tPrec@1 (79.726)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.4139 (0.4635)\tPrec@1 (79.839)\n",
      "[240/391]\tTime 0.101 (0.103)\tData 0.020 (0.020)\tLoss 0.6464 (0.4642)\tPrec@1 (79.769)\n",
      "[280/391]\tTime 0.083 (0.102)\tData 0.019 (0.020)\tLoss 0.3838 (0.4603)\tPrec@1 (79.966)\n",
      "[320/391]\tTime 0.071 (0.099)\tData 0.019 (0.020)\tLoss 0.5374 (0.4614)\tPrec@1 (79.885)\n",
      "[360/391]\tTime 0.065 (0.096)\tData 0.020 (0.020)\tLoss 0.4907 (0.4637)\tPrec@1 (79.755)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9534 (0.9534)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 1.0140 (0.8842)\tPrec@1 71.094 (72.027)\n",
      " * ResNet8 Prec@1 71.820\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7592 (0.7592)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8728 (0.7744)\tPrec@1 69.531 (73.914)\n",
      " * ResNet8 Prec@1 73.410\n",
      "\n",
      "===> epoch: 79/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.022 (0.022)\tLoss 0.4025 (0.4025)\tPrec@1 (82.031)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.020 (0.021)\tLoss 0.3437 (0.4616)\tPrec@1 (79.688)\n",
      "[80/391]\tTime 0.108 (0.102)\tData 0.022 (0.020)\tLoss 0.3273 (0.4378)\tPrec@1 (80.430)\n",
      "[120/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.4774 (0.4476)\tPrec@1 (80.075)\n",
      "[160/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.5247 (0.4494)\tPrec@1 (80.129)\n",
      "[200/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.4380 (0.4553)\tPrec@1 (80.033)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5449 (0.4585)\tPrec@1 (79.970)\n",
      "[280/391]\tTime 0.107 (0.102)\tData 0.021 (0.020)\tLoss 0.6084 (0.4610)\tPrec@1 (79.902)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.4711 (0.4628)\tPrec@1 (79.780)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4441 (0.4609)\tPrec@1 (79.867)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7041 (0.7041)\tPrec@1 75.781 (75.781)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.7868 (0.6766)\tPrec@1 72.656 (76.925)\n",
      " * ResNet8 Prec@1 77.090\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7930 (0.7930)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.7930 (0.8115)\tPrec@1 75.000 (73.171)\n",
      " * ResNet8 Prec@1 73.000\n",
      "\n",
      "===> epoch: 80/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.021 (0.021)\tLoss 0.3664 (0.3664)\tPrec@1 (82.031)\n",
      "[40/391]\tTime 0.102 (0.101)\tData 0.020 (0.020)\tLoss 0.5431 (0.4259)\tPrec@1 (80.850)\n",
      "[80/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.3648 (0.4364)\tPrec@1 (80.488)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.3978 (0.4548)\tPrec@1 (79.965)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.5877 (0.4586)\tPrec@1 (80.061)\n",
      "[200/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4901 (0.4606)\tPrec@1 (79.983)\n",
      "[240/391]\tTime 0.102 (0.102)\tData 0.019 (0.020)\tLoss 0.4154 (0.4634)\tPrec@1 (79.830)\n",
      "[280/391]\tTime 0.101 (0.102)\tData 0.021 (0.020)\tLoss 0.5544 (0.4638)\tPrec@1 (79.807)\n",
      "[320/391]\tTime 0.107 (0.102)\tData 0.020 (0.020)\tLoss 0.3770 (0.4679)\tPrec@1 (79.622)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4092 (0.4684)\tPrec@1 (79.599)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.7390 (0.7390)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.7265 (0.8635)\tPrec@1 74.219 (71.608)\n",
      " * ResNet8 Prec@1 71.250\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.6503 (0.6503)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.7237 (0.7050)\tPrec@1 75.781 (76.620)\n",
      " * ResNet8 Prec@1 76.340\n",
      "\n",
      "===> epoch: 81/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.022 (0.022)\tLoss 0.5560 (0.5560)\tPrec@1 (77.344)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5709 (0.4600)\tPrec@1 (80.393)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5049 (0.4682)\tPrec@1 (79.823)\n",
      "[120/391]\tTime 0.108 (0.102)\tData 0.021 (0.020)\tLoss 0.6284 (0.4621)\tPrec@1 (79.791)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.021 (0.020)\tLoss 0.4062 (0.4586)\tPrec@1 (79.911)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.4524 (0.4586)\tPrec@1 (79.979)\n",
      "[240/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.4961 (0.4631)\tPrec@1 (79.801)\n",
      "[280/391]\tTime 0.101 (0.103)\tData 0.020 (0.020)\tLoss 0.5391 (0.4632)\tPrec@1 (79.852)\n",
      "[320/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.4880 (0.4623)\tPrec@1 (79.838)\n",
      "[360/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.3978 (0.4610)\tPrec@1 (79.861)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6062 (0.6062)\tPrec@1 85.156 (85.156)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.7164 (0.6800)\tPrec@1 74.219 (77.382)\n",
      " * ResNet8 Prec@1 76.730\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7596 (0.7596)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8977 (0.8270)\tPrec@1 66.406 (72.447)\n",
      " * ResNet8 Prec@1 72.270\n",
      "\n",
      "===> epoch: 82/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.022 (0.022)\tLoss 0.5478 (0.5478)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.100 (0.101)\tData 0.025 (0.022)\tLoss 0.3542 (0.4557)\tPrec@1 (79.859)\n",
      "[80/391]\tTime 0.102 (0.102)\tData 0.021 (0.022)\tLoss 0.3872 (0.4557)\tPrec@1 (79.755)\n",
      "[120/391]\tTime 0.101 (0.102)\tData 0.021 (0.022)\tLoss 0.4605 (0.4555)\tPrec@1 (79.836)\n",
      "[160/391]\tTime 0.100 (0.103)\tData 0.020 (0.022)\tLoss 0.4840 (0.4587)\tPrec@1 (79.634)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.021)\tLoss 0.5082 (0.4540)\tPrec@1 (79.851)\n",
      "[240/391]\tTime 0.105 (0.103)\tData 0.020 (0.021)\tLoss 0.4868 (0.4573)\tPrec@1 (79.668)\n",
      "[280/391]\tTime 0.107 (0.103)\tData 0.020 (0.021)\tLoss 0.4665 (0.4585)\tPrec@1 (79.654)\n",
      "[320/391]\tTime 0.120 (0.102)\tData 0.034 (0.021)\tLoss 0.4938 (0.4582)\tPrec@1 (79.695)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.020 (0.021)\tLoss 0.4396 (0.4598)\tPrec@1 (79.707)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.026 (0.026)\tLoss 0.6945 (0.6945)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 0.7312 (0.7885)\tPrec@1 76.562 (74.257)\n",
      " * ResNet8 Prec@1 73.920\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7789 (0.7789)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.027 (0.022)\tLoss 1.0618 (0.8898)\tPrec@1 62.500 (70.598)\n",
      " * ResNet8 Prec@1 70.010\n",
      "\n",
      "===> epoch: 83/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.021 (0.021)\tLoss 0.4189 (0.4189)\tPrec@1 (79.688)\n",
      "[40/391]\tTime 0.102 (0.099)\tData 0.019 (0.020)\tLoss 0.3303 (0.4405)\tPrec@1 (80.221)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.021 (0.020)\tLoss 0.4917 (0.4439)\tPrec@1 (80.334)\n",
      "[120/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.4019 (0.4514)\tPrec@1 (80.107)\n",
      "[160/391]\tTime 0.092 (0.098)\tData 0.020 (0.020)\tLoss 0.4682 (0.4553)\tPrec@1 (79.950)\n",
      "[200/391]\tTime 0.062 (0.093)\tData 0.020 (0.021)\tLoss 0.3829 (0.4575)\tPrec@1 (79.804)\n",
      "[240/391]\tTime 0.106 (0.092)\tData 0.023 (0.021)\tLoss 0.4243 (0.4601)\tPrec@1 (79.713)\n",
      "[280/391]\tTime 0.104 (0.093)\tData 0.021 (0.021)\tLoss 0.4897 (0.4625)\tPrec@1 (79.624)\n",
      "[320/391]\tTime 0.103 (0.094)\tData 0.020 (0.021)\tLoss 0.5868 (0.4633)\tPrec@1 (79.658)\n",
      "[360/391]\tTime 0.106 (0.095)\tData 0.020 (0.021)\tLoss 0.4604 (0.4633)\tPrec@1 (79.679)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 1.1796 (1.1796)\tPrec@1 64.844 (64.844)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 1.1896 (1.2056)\tPrec@1 60.156 (65.434)\n",
      " * ResNet8 Prec@1 65.130\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6937 (0.6937)\tPrec@1 76.562 (76.562)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.7919 (0.7170)\tPrec@1 71.094 (75.476)\n",
      " * ResNet8 Prec@1 75.610\n",
      "\n",
      "===> epoch: 84/200\n",
      "Training:\n",
      "[0/391]\tTime 0.109 (0.109)\tData 0.022 (0.022)\tLoss 0.4058 (0.4058)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.4398 (0.4578)\tPrec@1 (80.069)\n",
      "[80/391]\tTime 0.103 (0.102)\tData 0.019 (0.020)\tLoss 0.4446 (0.4623)\tPrec@1 (79.919)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4069 (0.4570)\tPrec@1 (79.972)\n",
      "[160/391]\tTime 0.101 (0.102)\tData 0.021 (0.020)\tLoss 0.5234 (0.4619)\tPrec@1 (79.595)\n",
      "[200/391]\tTime 0.102 (0.103)\tData 0.021 (0.020)\tLoss 0.4945 (0.4636)\tPrec@1 (79.392)\n",
      "[240/391]\tTime 0.100 (0.103)\tData 0.021 (0.020)\tLoss 0.4459 (0.4657)\tPrec@1 (79.470)\n",
      "[280/391]\tTime 0.101 (0.103)\tData 0.023 (0.021)\tLoss 0.4619 (0.4646)\tPrec@1 (79.510)\n",
      "[320/391]\tTime 0.098 (0.103)\tData 0.022 (0.021)\tLoss 0.5212 (0.4649)\tPrec@1 (79.517)\n",
      "[360/391]\tTime 0.102 (0.103)\tData 0.021 (0.021)\tLoss 0.5039 (0.4648)\tPrec@1 (79.549)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9494 (0.9494)\tPrec@1 73.438 (73.438)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 1.1305 (0.9228)\tPrec@1 65.625 (70.903)\n",
      " * ResNet8 Prec@1 70.310\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.2013 (1.2013)\tPrec@1 64.844 (64.844)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 1.2112 (1.1357)\tPrec@1 67.188 (66.197)\n",
      " * ResNet8 Prec@1 66.040\n",
      "\n",
      "===> epoch: 85/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.022 (0.022)\tLoss 0.3948 (0.3948)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.104 (0.101)\tData 0.020 (0.021)\tLoss 0.4020 (0.4590)\tPrec@1 (80.469)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.022 (0.021)\tLoss 0.5449 (0.4688)\tPrec@1 (79.630)\n",
      "[120/391]\tTime 0.104 (0.103)\tData 0.021 (0.021)\tLoss 0.4270 (0.4689)\tPrec@1 (79.436)\n",
      "[160/391]\tTime 0.107 (0.103)\tData 0.021 (0.021)\tLoss 0.4594 (0.4736)\tPrec@1 (79.212)\n",
      "[200/391]\tTime 0.109 (0.103)\tData 0.021 (0.021)\tLoss 0.4188 (0.4720)\tPrec@1 (79.380)\n",
      "[240/391]\tTime 0.108 (0.103)\tData 0.021 (0.021)\tLoss 0.4479 (0.4724)\tPrec@1 (79.305)\n",
      "[280/391]\tTime 0.108 (0.103)\tData 0.020 (0.021)\tLoss 0.4594 (0.4704)\tPrec@1 (79.384)\n",
      "[320/391]\tTime 0.105 (0.103)\tData 0.021 (0.021)\tLoss 0.3829 (0.4660)\tPrec@1 (79.580)\n",
      "[360/391]\tTime 0.107 (0.103)\tData 0.022 (0.021)\tLoss 0.4177 (0.4655)\tPrec@1 (79.538)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.7245 (0.7245)\tPrec@1 77.344 (77.344)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.9377 (0.8412)\tPrec@1 68.750 (72.561)\n",
      " * ResNet8 Prec@1 72.510\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.7373 (0.7373)\tPrec@1 72.656 (72.656)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.0449 (0.7807)\tPrec@1 71.875 (74.428)\n",
      " * ResNet8 Prec@1 74.670\n",
      "\n",
      "===> epoch: 86/200\n",
      "Training:\n",
      "[0/391]\tTime 0.086 (0.086)\tData 0.022 (0.022)\tLoss 0.4883 (0.4883)\tPrec@1 (78.125)\n",
      "[40/391]\tTime 0.098 (0.103)\tData 0.022 (0.021)\tLoss 0.3781 (0.4697)\tPrec@1 (79.325)\n",
      "[80/391]\tTime 0.098 (0.103)\tData 0.020 (0.021)\tLoss 0.6685 (0.4507)\tPrec@1 (80.507)\n",
      "[120/391]\tTime 0.097 (0.103)\tData 0.020 (0.021)\tLoss 0.4525 (0.4477)\tPrec@1 (80.340)\n",
      "[160/391]\tTime 0.097 (0.103)\tData 0.020 (0.020)\tLoss 0.3917 (0.4497)\tPrec@1 (80.318)\n",
      "[200/391]\tTime 0.096 (0.103)\tData 0.019 (0.020)\tLoss 0.4490 (0.4477)\tPrec@1 (80.360)\n",
      "[240/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.5333 (0.4551)\tPrec@1 (80.109)\n",
      "[280/391]\tTime 0.107 (0.102)\tData 0.037 (0.020)\tLoss 0.5499 (0.4561)\tPrec@1 (80.063)\n",
      "[320/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.5401 (0.4561)\tPrec@1 (80.043)\n",
      "[360/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.4510 (0.4586)\tPrec@1 (79.960)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6506 (0.6506)\tPrec@1 78.906 (78.906)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.7720 (0.7034)\tPrec@1 74.219 (75.419)\n",
      " * ResNet8 Prec@1 75.440\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.037 (0.037)\tLoss 0.7658 (0.7658)\tPrec@1 75.781 (75.781)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.9635 (0.8738)\tPrec@1 67.969 (70.217)\n",
      " * ResNet8 Prec@1 70.350\n",
      "\n",
      "===> epoch: 87/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.021 (0.021)\tLoss 0.4194 (0.4194)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.4860 (0.4532)\tPrec@1 (79.726)\n",
      "[80/391]\tTime 0.099 (0.102)\tData 0.019 (0.020)\tLoss 0.4276 (0.4564)\tPrec@1 (79.620)\n",
      "[120/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.5045 (0.4506)\tPrec@1 (79.946)\n",
      "[160/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.4739 (0.4511)\tPrec@1 (80.047)\n",
      "[200/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.5148 (0.4586)\tPrec@1 (79.684)\n",
      "[240/391]\tTime 0.102 (0.102)\tData 0.019 (0.020)\tLoss 0.4182 (0.4590)\tPrec@1 (79.619)\n",
      "[280/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.5193 (0.4606)\tPrec@1 (79.554)\n",
      "[320/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.2888 (0.4599)\tPrec@1 (79.605)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.4771 (0.4626)\tPrec@1 (79.530)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.5837 (0.5837)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.6729 (0.6476)\tPrec@1 74.219 (77.973)\n",
      " * ResNet8 Prec@1 78.050\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.7417 (0.7417)\tPrec@1 74.219 (74.219)\n",
      "Test: [40/79]\tTime 0.023 (0.020)\tLoss 0.8249 (0.8181)\tPrec@1 70.312 (72.104)\n",
      " * ResNet8 Prec@1 71.800\n",
      "\n",
      "===> epoch: 88/200\n",
      "Training:\n",
      "[0/391]\tTime 0.086 (0.086)\tData 0.021 (0.021)\tLoss 0.4001 (0.4001)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.070 (0.080)\tData 0.020 (0.021)\tLoss 0.4077 (0.4455)\tPrec@1 (80.564)\n",
      "[80/391]\tTime 0.070 (0.073)\tData 0.024 (0.021)\tLoss 0.5892 (0.4522)\tPrec@1 (80.170)\n",
      "[120/391]\tTime 0.105 (0.083)\tData 0.020 (0.020)\tLoss 0.4187 (0.4576)\tPrec@1 (80.017)\n",
      "[160/391]\tTime 0.106 (0.087)\tData 0.020 (0.020)\tLoss 0.4374 (0.4584)\tPrec@1 (80.226)\n",
      "[200/391]\tTime 0.103 (0.091)\tData 0.020 (0.020)\tLoss 0.4129 (0.4663)\tPrec@1 (79.757)\n",
      "[240/391]\tTime 0.106 (0.093)\tData 0.024 (0.020)\tLoss 0.4288 (0.4648)\tPrec@1 (79.824)\n",
      "[280/391]\tTime 0.101 (0.094)\tData 0.020 (0.020)\tLoss 0.2974 (0.4614)\tPrec@1 (79.940)\n",
      "[320/391]\tTime 0.106 (0.095)\tData 0.020 (0.020)\tLoss 0.3332 (0.4595)\tPrec@1 (80.028)\n",
      "[360/391]\tTime 0.105 (0.096)\tData 0.021 (0.020)\tLoss 0.4520 (0.4620)\tPrec@1 (79.943)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9043 (0.9043)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.8817 (0.8934)\tPrec@1 72.656 (69.855)\n",
      " * ResNet8 Prec@1 69.640\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.5583 (0.5583)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.9267 (0.7488)\tPrec@1 69.531 (74.466)\n",
      " * ResNet8 Prec@1 74.110\n",
      "\n",
      "===> epoch: 89/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.021 (0.021)\tLoss 0.6133 (0.6133)\tPrec@1 (72.656)\n",
      "[40/391]\tTime 0.102 (0.101)\tData 0.020 (0.020)\tLoss 0.2544 (0.4733)\tPrec@1 (78.982)\n",
      "[80/391]\tTime 0.107 (0.102)\tData 0.021 (0.020)\tLoss 0.3207 (0.4575)\tPrec@1 (79.610)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.5128 (0.4569)\tPrec@1 (79.545)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.019 (0.021)\tLoss 0.5238 (0.4514)\tPrec@1 (79.906)\n",
      "[200/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.3754 (0.4524)\tPrec@1 (79.878)\n",
      "[240/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.4278 (0.4564)\tPrec@1 (79.726)\n",
      "[280/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.5373 (0.4588)\tPrec@1 (79.665)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4709 (0.4605)\tPrec@1 (79.632)\n",
      "[360/391]\tTime 0.108 (0.102)\tData 0.021 (0.020)\tLoss 0.3544 (0.4633)\tPrec@1 (79.584)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.9452 (0.9452)\tPrec@1 68.750 (68.750)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8993 (0.9281)\tPrec@1 70.312 (70.408)\n",
      " * ResNet8 Prec@1 69.770\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8931 (0.8931)\tPrec@1 69.531 (69.531)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.7662 (0.7996)\tPrec@1 71.875 (73.609)\n",
      " * ResNet8 Prec@1 72.900\n",
      "\n",
      "===> epoch: 90/200\n",
      "Training:\n",
      "[0/391]\tTime 0.113 (0.113)\tData 0.021 (0.021)\tLoss 0.3834 (0.3834)\tPrec@1 (85.156)\n",
      "[40/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.4784 (0.4555)\tPrec@1 (80.240)\n",
      "[80/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.4947 (0.4627)\tPrec@1 (79.630)\n",
      "[120/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.5141 (0.4630)\tPrec@1 (79.623)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.4531 (0.4634)\tPrec@1 (79.610)\n",
      "[200/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.3650 (0.4617)\tPrec@1 (79.711)\n",
      "[240/391]\tTime 0.098 (0.102)\tData 0.020 (0.021)\tLoss 0.4947 (0.4631)\tPrec@1 (79.681)\n",
      "[280/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.4389 (0.4647)\tPrec@1 (79.679)\n",
      "[320/391]\tTime 0.098 (0.102)\tData 0.021 (0.021)\tLoss 0.5251 (0.4644)\tPrec@1 (79.726)\n",
      "[360/391]\tTime 0.098 (0.103)\tData 0.021 (0.020)\tLoss 0.4687 (0.4657)\tPrec@1 (79.692)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.6891 (0.6891)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.8506 (0.8026)\tPrec@1 68.750 (73.438)\n",
      " * ResNet8 Prec@1 73.630\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6531 (0.6531)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.9582 (0.9115)\tPrec@1 67.188 (71.761)\n",
      " * ResNet8 Prec@1 71.770\n",
      "\n",
      "===> epoch: 91/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.022 (0.022)\tLoss 0.4248 (0.4248)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.107 (0.101)\tData 0.020 (0.020)\tLoss 0.4657 (0.4417)\tPrec@1 (80.716)\n",
      "[80/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.4434 (0.4459)\tPrec@1 (80.363)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.3020 (0.4432)\tPrec@1 (80.559)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.021 (0.020)\tLoss 0.5938 (0.4464)\tPrec@1 (80.503)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4442 (0.4490)\tPrec@1 (80.356)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.5435 (0.4494)\tPrec@1 (80.381)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.4235 (0.4541)\tPrec@1 (80.191)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.5068 (0.4525)\tPrec@1 (80.186)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.3738 (0.4533)\tPrec@1 (80.142)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9384 (0.9384)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 1.1064 (1.0081)\tPrec@1 64.062 (66.883)\n",
      " * ResNet8 Prec@1 66.750\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.9525 (0.9525)\tPrec@1 64.062 (64.062)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.9771 (1.0312)\tPrec@1 67.188 (66.768)\n",
      " * ResNet8 Prec@1 65.840\n",
      "\n",
      "===> epoch: 92/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.022 (0.022)\tLoss 0.4420 (0.4420)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.097 (0.100)\tData 0.019 (0.020)\tLoss 0.3491 (0.4317)\tPrec@1 (80.812)\n",
      "[80/391]\tTime 0.096 (0.101)\tData 0.020 (0.020)\tLoss 0.5257 (0.4600)\tPrec@1 (79.880)\n",
      "[120/391]\tTime 0.092 (0.101)\tData 0.021 (0.020)\tLoss 0.3867 (0.4640)\tPrec@1 (79.771)\n",
      "[160/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.5195 (0.4575)\tPrec@1 (79.945)\n",
      "[200/391]\tTime 0.097 (0.102)\tData 0.021 (0.020)\tLoss 0.4489 (0.4545)\tPrec@1 (80.053)\n",
      "[240/391]\tTime 0.093 (0.102)\tData 0.020 (0.020)\tLoss 0.4128 (0.4518)\tPrec@1 (80.174)\n",
      "[280/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.4321 (0.4525)\tPrec@1 (80.118)\n",
      "[320/391]\tTime 0.083 (0.100)\tData 0.020 (0.020)\tLoss 0.5082 (0.4564)\tPrec@1 (80.021)\n",
      "[360/391]\tTime 0.065 (0.097)\tData 0.020 (0.020)\tLoss 0.4856 (0.4586)\tPrec@1 (79.973)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7323 (0.7323)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.9025 (0.8055)\tPrec@1 67.969 (72.847)\n",
      " * ResNet8 Prec@1 72.460\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.6004 (0.6004)\tPrec@1 82.812 (82.812)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.6917 (0.6786)\tPrec@1 76.562 (77.001)\n",
      " * ResNet8 Prec@1 76.760\n",
      "\n",
      "===> epoch: 93/200\n",
      "Training:\n",
      "[0/391]\tTime 0.108 (0.108)\tData 0.022 (0.022)\tLoss 0.4623 (0.4623)\tPrec@1 (81.250)\n",
      "[40/391]\tTime 0.100 (0.101)\tData 0.020 (0.020)\tLoss 0.4607 (0.4772)\tPrec@1 (79.592)\n",
      "[80/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.4025 (0.4607)\tPrec@1 (80.073)\n",
      "[120/391]\tTime 0.095 (0.102)\tData 0.019 (0.020)\tLoss 0.4948 (0.4527)\tPrec@1 (80.288)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.4889 (0.4517)\tPrec@1 (80.294)\n",
      "[200/391]\tTime 0.108 (0.102)\tData 0.019 (0.020)\tLoss 0.3096 (0.4539)\tPrec@1 (80.340)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.021 (0.020)\tLoss 0.3274 (0.4585)\tPrec@1 (80.112)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.108 (0.102)\tData 0.022 (0.020)\tLoss 0.5782 (0.4604)\tPrec@1 (80.024)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.021 (0.021)\tLoss 0.5457 (0.4615)\tPrec@1 (79.997)\n",
      "[360/391]\tTime 0.110 (0.103)\tData 0.021 (0.021)\tLoss 0.4482 (0.4642)\tPrec@1 (79.861)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 1.1355 (1.1355)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.019 (0.022)\tLoss 1.2049 (1.1710)\tPrec@1 64.844 (64.882)\n",
      " * ResNet8 Prec@1 64.740\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.8997 (0.8997)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 1.0503 (0.9114)\tPrec@1 70.312 (69.703)\n",
      " * ResNet8 Prec@1 69.830\n",
      "\n",
      "===> epoch: 94/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.023 (0.023)\tLoss 0.4706 (0.4706)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.098 (0.103)\tData 0.023 (0.021)\tLoss 0.5093 (0.4409)\tPrec@1 (80.545)\n",
      "[80/391]\tTime 0.100 (0.103)\tData 0.021 (0.021)\tLoss 0.3525 (0.4509)\tPrec@1 (80.257)\n",
      "[120/391]\tTime 0.098 (0.103)\tData 0.020 (0.021)\tLoss 0.4110 (0.4557)\tPrec@1 (80.198)\n",
      "[160/391]\tTime 0.097 (0.103)\tData 0.020 (0.021)\tLoss 0.5670 (0.4527)\tPrec@1 (80.309)\n",
      "[200/391]\tTime 0.096 (0.103)\tData 0.020 (0.021)\tLoss 0.5675 (0.4532)\tPrec@1 (80.236)\n",
      "[240/391]\tTime 0.096 (0.103)\tData 0.022 (0.021)\tLoss 0.3434 (0.4558)\tPrec@1 (80.106)\n",
      "[280/391]\tTime 0.099 (0.103)\tData 0.021 (0.021)\tLoss 0.5632 (0.4573)\tPrec@1 (80.132)\n",
      "[320/391]\tTime 0.099 (0.103)\tData 0.022 (0.021)\tLoss 0.4361 (0.4561)\tPrec@1 (80.123)\n",
      "[360/391]\tTime 0.106 (0.103)\tData 0.025 (0.021)\tLoss 0.4812 (0.4568)\tPrec@1 (80.066)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.8215 (0.8215)\tPrec@1 71.875 (71.875)\n",
      "Test: [40/79]\tTime 0.021 (0.022)\tLoss 1.1055 (0.8131)\tPrec@1 65.625 (73.438)\n",
      " * ResNet8 Prec@1 73.230\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6512 (0.6512)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.7245 (0.7049)\tPrec@1 71.875 (75.343)\n",
      " * ResNet8 Prec@1 75.280\n",
      "\n",
      "===> epoch: 95/200\n",
      "Training:\n",
      "[0/391]\tTime 0.108 (0.108)\tData 0.022 (0.022)\tLoss 0.4349 (0.4349)\tPrec@1 (83.594)\n",
      "[40/391]\tTime 0.099 (0.102)\tData 0.021 (0.021)\tLoss 0.3626 (0.4608)\tPrec@1 (80.011)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.4194 (0.4571)\tPrec@1 (79.948)\n",
      "[120/391]\tTime 0.103 (0.102)\tData 0.020 (0.021)\tLoss 0.4658 (0.4611)\tPrec@1 (79.791)\n",
      "[160/391]\tTime 0.103 (0.101)\tData 0.020 (0.021)\tLoss 0.3977 (0.4593)\tPrec@1 (79.925)\n",
      "[200/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4530 (0.4646)\tPrec@1 (79.672)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.4737 (0.4628)\tPrec@1 (79.759)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4619 (0.4639)\tPrec@1 (79.729)\n",
      "[320/391]\tTime 0.110 (0.102)\tData 0.023 (0.020)\tLoss 0.6073 (0.4626)\tPrec@1 (79.809)\n",
      "[360/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.3991 (0.4652)\tPrec@1 (79.731)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.9893 (0.9893)\tPrec@1 71.094 (71.094)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 1.0349 (0.9597)\tPrec@1 69.531 (69.588)\n",
      " * ResNet8 Prec@1 69.820\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 1.0256 (1.0256)\tPrec@1 67.969 (67.969)\n",
      "Test: [40/79]\tTime 0.024 (0.020)\tLoss 1.1285 (0.9607)\tPrec@1 67.188 (70.027)\n",
      " * ResNet8 Prec@1 69.750\n",
      "\n",
      "===> epoch: 96/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.5603 (0.5603)\tPrec@1 (76.562)\n",
      "[40/391]\tTime 0.114 (0.103)\tData 0.028 (0.020)\tLoss 0.5558 (0.4741)\tPrec@1 (79.345)\n",
      "[80/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.3400 (0.4646)\tPrec@1 (79.900)\n",
      "[120/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.3781 (0.4687)\tPrec@1 (79.726)\n",
      "[160/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.6173 (0.4632)\tPrec@1 (79.843)\n",
      "[200/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.4930 (0.4637)\tPrec@1 (79.800)\n",
      "[240/391]\tTime 0.122 (0.103)\tData 0.035 (0.020)\tLoss 0.3807 (0.4656)\tPrec@1 (79.542)\n",
      "[280/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.5064 (0.4647)\tPrec@1 (79.565)\n",
      "[320/391]\tTime 0.101 (0.103)\tData 0.020 (0.020)\tLoss 0.5463 (0.4666)\tPrec@1 (79.490)\n",
      "[360/391]\tTime 0.104 (0.103)\tData 0.019 (0.020)\tLoss 0.4550 (0.4662)\tPrec@1 (79.501)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.5885 (0.5885)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.8301 (0.7849)\tPrec@1 71.094 (72.885)\n",
      " * ResNet8 Prec@1 72.950\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.5797 (0.5797)\tPrec@1 78.906 (78.906)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.7779 (0.7200)\tPrec@1 72.656 (75.591)\n",
      " * ResNet8 Prec@1 75.580\n",
      "\n",
      "===> epoch: 97/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.021 (0.021)\tLoss 0.4184 (0.4184)\tPrec@1 (80.469)\n",
      "[40/391]\tTime 0.105 (0.103)\tData 0.019 (0.020)\tLoss 0.5199 (0.4449)\tPrec@1 (80.716)\n",
      "[80/391]\tTime 0.103 (0.102)\tData 0.019 (0.020)\tLoss 0.4509 (0.4606)\tPrec@1 (79.880)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.4915 (0.4640)\tPrec@1 (79.823)\n",
      "[160/391]\tTime 0.082 (0.101)\tData 0.019 (0.020)\tLoss 0.2936 (0.4613)\tPrec@1 (79.780)\n",
      "[200/391]\tTime 0.072 (0.097)\tData 0.019 (0.020)\tLoss 0.3505 (0.4609)\tPrec@1 (79.789)\n",
      "[240/391]\tTime 0.063 (0.092)\tData 0.019 (0.020)\tLoss 0.4523 (0.4592)\tPrec@1 (79.840)\n",
      "[280/391]\tTime 0.105 (0.093)\tData 0.019 (0.020)\tLoss 0.5609 (0.4583)\tPrec@1 (79.813)\n",
      "[320/391]\tTime 0.105 (0.094)\tData 0.019 (0.020)\tLoss 0.3974 (0.4616)\tPrec@1 (79.773)\n",
      "[360/391]\tTime 0.105 (0.095)\tData 0.019 (0.020)\tLoss 0.4862 (0.4622)\tPrec@1 (79.748)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.7609 (0.7609)\tPrec@1 78.125 (78.125)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.7395 (0.8196)\tPrec@1 75.000 (73.590)\n",
      " * ResNet8 Prec@1 73.500\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.8862 (0.8862)\tPrec@1 66.406 (66.406)\n",
      "Test: [40/79]\tTime 0.021 (0.019)\tLoss 1.0164 (0.9478)\tPrec@1 67.188 (69.055)\n",
      " * ResNet8 Prec@1 68.850\n",
      "\n",
      "===> epoch: 98/200\n",
      "Training:\n",
      "[0/391]\tTime 0.097 (0.097)\tData 0.021 (0.021)\tLoss 0.3848 (0.3848)\tPrec@1 (82.031)\n",
      "[40/391]\tTime 0.096 (0.102)\tData 0.019 (0.019)\tLoss 0.4863 (0.4772)\tPrec@1 (79.116)\n",
      "[80/391]\tTime 0.097 (0.102)\tData 0.019 (0.019)\tLoss 0.4506 (0.4605)\tPrec@1 (79.832)\n",
      "[120/391]\tTime 0.095 (0.102)\tData 0.019 (0.019)\tLoss 0.4665 (0.4512)\tPrec@1 (80.372)\n",
      "[160/391]\tTime 0.097 (0.102)\tData 0.019 (0.019)\tLoss 0.5123 (0.4596)\tPrec@1 (79.891)\n",
      "[200/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.5990 (0.4563)\tPrec@1 (80.002)\n",
      "[240/391]\tTime 0.096 (0.102)\tData 0.019 (0.019)\tLoss 0.3075 (0.4573)\tPrec@1 (79.963)\n",
      "[280/391]\tTime 0.097 (0.102)\tData 0.019 (0.020)\tLoss 0.5497 (0.4565)\tPrec@1 (79.921)\n",
      "[320/391]\tTime 0.093 (0.102)\tData 0.019 (0.020)\tLoss 0.5879 (0.4585)\tPrec@1 (79.855)\n",
      "[360/391]\tTime 0.091 (0.102)\tData 0.020 (0.020)\tLoss 0.5291 (0.4594)\tPrec@1 (79.822)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.7163 (0.7163)\tPrec@1 75.000 (75.000)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.7533 (0.8051)\tPrec@1 73.438 (72.104)\n",
      " * ResNet8 Prec@1 72.190\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.5550 (0.5550)\tPrec@1 82.812 (82.812)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.7316 (0.6690)\tPrec@1 75.781 (77.572)\n",
      " * ResNet8 Prec@1 77.240\n",
      "\n",
      "===> epoch: 99/200\n",
      "Training:\n",
      "[0/391]\tTime 0.109 (0.109)\tData 0.024 (0.024)\tLoss 0.4450 (0.4450)\tPrec@1 (78.906)\n",
      "[40/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.4067 (0.4328)\tPrec@1 (81.117)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.5032 (0.4372)\tPrec@1 (80.739)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.4270 (0.4400)\tPrec@1 (80.630)\n",
      "[160/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.2859 (0.4399)\tPrec@1 (80.682)\n",
      "[200/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.4749 (0.4422)\tPrec@1 (80.675)\n",
      "[240/391]\tTime 0.109 (0.101)\tData 0.019 (0.019)\tLoss 0.3059 (0.4477)\tPrec@1 (80.391)\n",
      "[280/391]\tTime 0.095 (0.101)\tData 0.019 (0.019)\tLoss 0.4064 (0.4543)\tPrec@1 (80.049)\n",
      "[320/391]\tTime 0.096 (0.101)\tData 0.019 (0.019)\tLoss 0.3743 (0.4539)\tPrec@1 (79.992)\n",
      "[360/391]\tTime 0.098 (0.101)\tData 0.019 (0.019)\tLoss 0.4928 (0.4572)\tPrec@1 (79.906)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.8935 (0.8935)\tPrec@1 70.312 (70.312)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.9239 (1.0139)\tPrec@1 67.969 (68.293)\n",
      " * ResNet8 Prec@1 68.300\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.6031 (0.6031)\tPrec@1 79.688 (79.688)\n",
      "Test: [40/79]\tTime 0.020 (0.019)\tLoss 0.9825 (0.8418)\tPrec@1 67.969 (72.904)\n",
      " * ResNet8 Prec@1 72.870\n",
      "\n",
      "===> epoch: 100/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.020 (0.020)\tLoss 0.5526 (0.5526)\tPrec@1 (77.344)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.103 (0.103)\tData 0.020 (0.020)\tLoss 0.4802 (0.4120)\tPrec@1 (81.460)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.3067 (0.3748)\tPrec@1 (83.102)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.3424 (0.3596)\tPrec@1 (83.684)\n",
      "[160/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.3303 (0.3510)\tPrec@1 (83.967)\n",
      "[200/391]\tTime 0.101 (0.102)\tData 0.019 (0.019)\tLoss 0.2846 (0.3453)\tPrec@1 (84.188)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.019 (0.019)\tLoss 0.2618 (0.3403)\tPrec@1 (84.317)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.019 (0.019)\tLoss 0.3348 (0.3373)\tPrec@1 (84.456)\n",
      "[320/391]\tTime 0.101 (0.102)\tData 0.019 (0.019)\tLoss 0.4798 (0.3334)\tPrec@1 (84.545)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.019 (0.019)\tLoss 0.2116 (0.3290)\tPrec@1 (84.708)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3797 (0.3797)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4868 (0.4443)\tPrec@1 82.812 (84.737)\n",
      " * ResNet8 Prec@1 85.040\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3028 (0.3028)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.5403 (0.4533)\tPrec@1 82.031 (84.642)\n",
      " * ResNet8 Prec@1 84.550\n",
      "\n",
      "===> epoch: 101/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.020 (0.020)\tLoss 0.2115 (0.2115)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.105 (0.101)\tData 0.020 (0.019)\tLoss 0.2436 (0.2989)\tPrec@1 (85.175)\n",
      "[80/391]\tTime 0.101 (0.101)\tData 0.019 (0.019)\tLoss 0.2924 (0.2948)\tPrec@1 (85.503)\n",
      "[120/391]\tTime 0.103 (0.101)\tData 0.019 (0.019)\tLoss 0.3507 (0.2918)\tPrec@1 (85.582)\n",
      "[160/391]\tTime 0.101 (0.101)\tData 0.019 (0.019)\tLoss 0.1984 (0.2944)\tPrec@1 (85.530)\n",
      "[200/391]\tTime 0.103 (0.101)\tData 0.019 (0.019)\tLoss 0.2746 (0.2911)\tPrec@1 (85.677)\n",
      "[240/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.2092 (0.2891)\tPrec@1 (85.785)\n",
      "[280/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.2082 (0.2878)\tPrec@1 (85.882)\n",
      "[320/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.3160 (0.2884)\tPrec@1 (85.886)\n",
      "[360/391]\tTime 0.100 (0.101)\tData 0.019 (0.019)\tLoss 0.3182 (0.2881)\tPrec@1 (85.894)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3205 (0.3205)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4814 (0.4324)\tPrec@1 84.375 (85.328)\n",
      " * ResNet8 Prec@1 85.700\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2777 (0.2777)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.5382 (0.4361)\tPrec@1 82.031 (85.156)\n",
      " * ResNet8 Prec@1 85.060\n",
      "\n",
      "===> epoch: 102/200\n",
      "Training:\n",
      "[0/391]\tTime 0.109 (0.109)\tData 0.021 (0.021)\tLoss 0.1772 (0.1772)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.082 (0.093)\tData 0.019 (0.019)\tLoss 0.2097 (0.2723)\tPrec@1 (86.395)\n",
      "[80/391]\tTime 0.072 (0.085)\tData 0.019 (0.020)\tLoss 0.2268 (0.2732)\tPrec@1 (86.574)\n",
      "[120/391]\tTime 0.094 (0.081)\tData 0.019 (0.020)\tLoss 0.2638 (0.2740)\tPrec@1 (86.590)\n",
      "[160/391]\tTime 0.098 (0.086)\tData 0.019 (0.020)\tLoss 0.2292 (0.2767)\tPrec@1 (86.534)\n",
      "[200/391]\tTime 0.100 (0.090)\tData 0.019 (0.020)\tLoss 0.2529 (0.2735)\tPrec@1 (86.676)\n",
      "[240/391]\tTime 0.096 (0.092)\tData 0.019 (0.020)\tLoss 0.3407 (0.2752)\tPrec@1 (86.693)\n",
      "[280/391]\tTime 0.096 (0.093)\tData 0.019 (0.020)\tLoss 0.2403 (0.2735)\tPrec@1 (86.666)\n",
      "[320/391]\tTime 0.100 (0.094)\tData 0.019 (0.020)\tLoss 0.2707 (0.2739)\tPrec@1 (86.575)\n",
      "[360/391]\tTime 0.105 (0.094)\tData 0.019 (0.019)\tLoss 0.1111 (0.2713)\tPrec@1 (86.660)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3250 (0.3250)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4374 (0.4260)\tPrec@1 84.375 (85.614)\n",
      " * ResNet8 Prec@1 85.680\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2654 (0.2654)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.4694 (0.4263)\tPrec@1 85.938 (85.518)\n",
      " * ResNet8 Prec@1 85.430\n",
      "\n",
      "===> epoch: 103/200\n",
      "Training:\n",
      "[0/391]\tTime 0.077 (0.077)\tData 0.020 (0.020)\tLoss 0.2312 (0.2312)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.095 (0.098)\tData 0.019 (0.019)\tLoss 0.3485 (0.2602)\tPrec@1 (86.966)\n",
      "[80/391]\tTime 0.096 (0.100)\tData 0.019 (0.019)\tLoss 0.2736 (0.2623)\tPrec@1 (86.941)\n",
      "[120/391]\tTime 0.098 (0.100)\tData 0.019 (0.019)\tLoss 0.3485 (0.2681)\tPrec@1 (86.783)\n",
      "[160/391]\tTime 0.095 (0.101)\tData 0.019 (0.019)\tLoss 0.1343 (0.2624)\tPrec@1 (87.015)\n",
      "[200/391]\tTime 0.098 (0.101)\tData 0.019 (0.019)\tLoss 0.2365 (0.2608)\tPrec@1 (86.952)\n",
      "[240/391]\tTime 0.103 (0.100)\tData 0.019 (0.019)\tLoss 0.1273 (0.2604)\tPrec@1 (86.897)\n",
      "[280/391]\tTime 0.104 (0.100)\tData 0.019 (0.019)\tLoss 0.4064 (0.2612)\tPrec@1 (86.955)\n",
      "[320/391]\tTime 0.101 (0.100)\tData 0.019 (0.019)\tLoss 0.3666 (0.2616)\tPrec@1 (86.960)\n",
      "[360/391]\tTime 0.101 (0.100)\tData 0.019 (0.019)\tLoss 0.2712 (0.2636)\tPrec@1 (86.857)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3028 (0.3028)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4685 (0.4228)\tPrec@1 84.375 (85.785)\n",
      " * ResNet8 Prec@1 85.680\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.2478 (0.2478)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5173 (0.4149)\tPrec@1 83.594 (85.861)\n",
      " * ResNet8 Prec@1 85.820\n",
      "\n",
      "===> epoch: 104/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.020 (0.020)\tLoss 0.2937 (0.2937)\tPrec@1 (83.594)\n",
      "[40/391]\tTime 0.097 (0.101)\tData 0.019 (0.019)\tLoss 0.2062 (0.2485)\tPrec@1 (87.614)\n",
      "[80/391]\tTime 0.100 (0.101)\tData 0.020 (0.019)\tLoss 0.2596 (0.2455)\tPrec@1 (87.770)\n",
      "[120/391]\tTime 0.098 (0.101)\tData 0.019 (0.019)\tLoss 0.3086 (0.2449)\tPrec@1 (87.739)\n",
      "[160/391]\tTime 0.097 (0.102)\tData 0.019 (0.019)\tLoss 0.2534 (0.2472)\tPrec@1 (87.665)\n",
      "[200/391]\tTime 0.099 (0.101)\tData 0.019 (0.019)\tLoss 0.2226 (0.2484)\tPrec@1 (87.613)\n",
      "[240/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.4000 (0.2519)\tPrec@1 (87.526)\n",
      "[280/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.2737 (0.2505)\tPrec@1 (87.519)\n",
      "[320/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.2288 (0.2521)\tPrec@1 (87.400)\n",
      "[360/391]\tTime 0.112 (0.101)\tData 0.025 (0.019)\tLoss 0.2411 (0.2529)\tPrec@1 (87.394)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3625 (0.3625)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4734 (0.4256)\tPrec@1 84.375 (85.556)\n",
      " * ResNet8 Prec@1 85.530\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2633 (0.2633)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.5363 (0.4124)\tPrec@1 85.156 (85.995)\n",
      " * ResNet8 Prec@1 85.950\n",
      "\n",
      "===> epoch: 105/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.020 (0.020)\tLoss 0.3185 (0.3185)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.103 (0.099)\tData 0.019 (0.019)\tLoss 0.3130 (0.2440)\tPrec@1 (87.367)\n",
      "[80/391]\tTime 0.098 (0.100)\tData 0.019 (0.019)\tLoss 0.2507 (0.2427)\tPrec@1 (87.432)\n",
      "[120/391]\tTime 0.097 (0.101)\tData 0.019 (0.019)\tLoss 0.2307 (0.2408)\tPrec@1 (87.532)\n",
      "[160/391]\tTime 0.095 (0.101)\tData 0.019 (0.019)\tLoss 0.2083 (0.2439)\tPrec@1 (87.578)\n",
      "[200/391]\tTime 0.098 (0.101)\tData 0.025 (0.019)\tLoss 0.1842 (0.2470)\tPrec@1 (87.496)\n",
      "[240/391]\tTime 0.102 (0.101)\tData 0.019 (0.019)\tLoss 0.2362 (0.2484)\tPrec@1 (87.510)\n",
      "[280/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.3201 (0.2485)\tPrec@1 (87.497)\n",
      "[320/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.3153 (0.2493)\tPrec@1 (87.390)\n",
      "[360/391]\tTime 0.106 (0.101)\tData 0.019 (0.019)\tLoss 0.2749 (0.2489)\tPrec@1 (87.448)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3006 (0.3006)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.4953 (0.4192)\tPrec@1 82.812 (85.938)\n",
      " * ResNet8 Prec@1 86.070\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2687 (0.2687)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.5223 (0.4116)\tPrec@1 85.156 (86.452)\n",
      " * ResNet8 Prec@1 86.190\n",
      "\n",
      "===> epoch: 106/200\n",
      "Training:\n",
      "[0/391]\tTime 0.102 (0.102)\tData 0.021 (0.021)\tLoss 0.2217 (0.2217)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.2475 (0.2400)\tPrec@1 (87.652)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2718 (0.2369)\tPrec@1 (87.924)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.019 (0.019)\tLoss 0.3686 (0.2375)\tPrec@1 (87.881)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.020 (0.019)\tLoss 0.2151 (0.2417)\tPrec@1 (87.650)\n",
      "[200/391]\tTime 0.102 (0.102)\tData 0.020 (0.019)\tLoss 0.3020 (0.2428)\tPrec@1 (87.648)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.1607 (0.2433)\tPrec@1 (87.626)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2485 (0.2449)\tPrec@1 (87.511)\n",
      "[320/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.2460 (0.2430)\tPrec@1 (87.590)\n",
      "[360/391]\tTime 0.083 (0.101)\tData 0.019 (0.020)\tLoss 0.2284 (0.2413)\tPrec@1 (87.658)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.3687 (0.3687)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.5347 (0.4192)\tPrec@1 83.594 (85.614)\n",
      " * ResNet8 Prec@1 85.920\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2937 (0.2937)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5628 (0.4247)\tPrec@1 82.812 (85.671)\n",
      " * ResNet8 Prec@1 85.720\n",
      "\n",
      "===> epoch: 107/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.020 (0.020)\tLoss 0.2379 (0.2379)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.096 (0.101)\tData 0.019 (0.019)\tLoss 0.4116 (0.2489)\tPrec@1 (87.500)\n",
      "[80/391]\tTime 0.096 (0.101)\tData 0.019 (0.019)\tLoss 0.1425 (0.2400)\tPrec@1 (87.577)\n",
      "[120/391]\tTime 0.098 (0.101)\tData 0.019 (0.019)\tLoss 0.1434 (0.2405)\tPrec@1 (87.726)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.019 (0.019)\tLoss 0.3005 (0.2399)\tPrec@1 (87.777)\n",
      "[200/391]\tTime 0.099 (0.102)\tData 0.019 (0.019)\tLoss 0.3160 (0.2405)\tPrec@1 (87.893)\n",
      "[240/391]\tTime 0.095 (0.101)\tData 0.019 (0.019)\tLoss 0.1315 (0.2429)\tPrec@1 (87.808)\n",
      "[280/391]\tTime 0.099 (0.101)\tData 0.019 (0.019)\tLoss 0.2269 (0.2428)\tPrec@1 (87.761)\n",
      "[320/391]\tTime 0.099 (0.101)\tData 0.019 (0.019)\tLoss 0.2813 (0.2421)\tPrec@1 (87.819)\n",
      "[360/391]\tTime 0.100 (0.101)\tData 0.019 (0.019)\tLoss 0.2480 (0.2422)\tPrec@1 (87.762)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3022 (0.3022)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.5493 (0.4331)\tPrec@1 81.250 (85.499)\n",
      " * ResNet8 Prec@1 85.390\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.2601 (0.2601)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.018 (0.021)\tLoss 0.5684 (0.4224)\tPrec@1 83.594 (85.938)\n",
      " * ResNet8 Prec@1 86.090\n",
      "\n",
      "===> epoch: 108/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.021 (0.021)\tLoss 0.3052 (0.3052)\tPrec@1 (84.375)\n",
      "[40/391]\tTime 0.100 (0.098)\tData 0.019 (0.019)\tLoss 0.2463 (0.2339)\tPrec@1 (87.995)\n",
      "[80/391]\tTime 0.104 (0.100)\tData 0.022 (0.020)\tLoss 0.2087 (0.2276)\tPrec@1 (88.214)\n",
      "[120/391]\tTime 0.106 (0.100)\tData 0.020 (0.020)\tLoss 0.3691 (0.2362)\tPrec@1 (87.939)\n",
      "[160/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.1906 (0.2322)\tPrec@1 (88.097)\n",
      "[200/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.2191 (0.2353)\tPrec@1 (88.044)\n",
      "[240/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.2360 (0.2362)\tPrec@1 (87.977)\n",
      "[280/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.1616 (0.2378)\tPrec@1 (87.909)\n",
      "[320/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.1440 (0.2389)\tPrec@1 (87.853)\n",
      "[360/391]\tTime 0.105 (0.101)\tData 0.021 (0.020)\tLoss 0.4122 (0.2405)\tPrec@1 (87.809)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3771 (0.3771)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4688 (0.4187)\tPrec@1 84.375 (85.556)\n",
      " * ResNet8 Prec@1 85.680\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2985 (0.2985)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5387 (0.4106)\tPrec@1 82.812 (86.109)\n",
      " * ResNet8 Prec@1 86.210\n",
      "\n",
      "===> epoch: 109/200\n",
      "Training:\n",
      "[0/391]\tTime 0.102 (0.102)\tData 0.021 (0.021)\tLoss 0.2941 (0.2941)\tPrec@1 (85.156)\n",
      "[40/391]\tTime 0.106 (0.099)\tData 0.020 (0.020)\tLoss 0.1848 (0.2378)\tPrec@1 (87.691)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.020 (0.020)\tLoss 0.2055 (0.2365)\tPrec@1 (88.088)\n",
      "[120/391]\tTime 0.103 (0.101)\tData 0.020 (0.020)\tLoss 0.3178 (0.2380)\tPrec@1 (87.868)\n",
      "[160/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.2417 (0.2409)\tPrec@1 (87.869)\n",
      "[200/391]\tTime 0.101 (0.101)\tData 0.020 (0.020)\tLoss 0.1427 (0.2377)\tPrec@1 (87.986)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.2168 (0.2376)\tPrec@1 (87.973)\n",
      "[280/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.2259 (0.2364)\tPrec@1 (88.003)\n",
      "[320/391]\tTime 0.102 (0.102)\tData 0.019 (0.020)\tLoss 0.4342 (0.2396)\tPrec@1 (87.882)\n",
      "[360/391]\tTime 0.099 (0.101)\tData 0.019 (0.020)\tLoss 0.2857 (0.2400)\tPrec@1 (87.838)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3479 (0.3479)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.5191 (0.4271)\tPrec@1 82.031 (85.575)\n",
      " * ResNet8 Prec@1 85.670\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2938 (0.2938)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.4875 (0.4231)\tPrec@1 85.156 (85.747)\n",
      " * ResNet8 Prec@1 85.540\n",
      "\n",
      "===> epoch: 110/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.021 (0.021)\tLoss 0.1262 (0.1262)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1123 (0.2292)\tPrec@1 (88.034)\n",
      "[80/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1833 (0.2298)\tPrec@1 (88.021)\n",
      "[120/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.2634 (0.2328)\tPrec@1 (87.862)\n",
      "[160/391]\tTime 0.100 (0.103)\tData 0.020 (0.020)\tLoss 0.2470 (0.2353)\tPrec@1 (87.806)\n",
      "[200/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.2306 (0.2355)\tPrec@1 (87.741)\n",
      "[240/391]\tTime 0.106 (0.103)\tData 0.019 (0.020)\tLoss 0.2200 (0.2356)\tPrec@1 (87.827)\n",
      "[280/391]\tTime 0.105 (0.103)\tData 0.019 (0.020)\tLoss 0.2463 (0.2352)\tPrec@1 (87.825)\n",
      "[320/391]\tTime 0.106 (0.103)\tData 0.019 (0.020)\tLoss 0.2022 (0.2346)\tPrec@1 (87.914)\n",
      "[360/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.2295 (0.2333)\tPrec@1 (87.952)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3401 (0.3401)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.024 (0.022)\tLoss 0.4622 (0.4134)\tPrec@1 85.938 (86.490)\n",
      " * ResNet8 Prec@1 86.550\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.2930 (0.2930)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.5856 (0.4111)\tPrec@1 80.469 (86.204)\n",
      " * ResNet8 Prec@1 86.080\n",
      "\n",
      "===> epoch: 111/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.2351 (0.2351)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.110 (0.100)\tData 0.024 (0.020)\tLoss 0.2039 (0.2248)\tPrec@1 (88.205)\n",
      "[80/391]\tTime 0.096 (0.101)\tData 0.020 (0.020)\tLoss 0.2070 (0.2193)\tPrec@1 (88.503)\n",
      "[120/391]\tTime 0.090 (0.101)\tData 0.020 (0.020)\tLoss 0.2411 (0.2246)\tPrec@1 (88.243)\n",
      "[160/391]\tTime 0.090 (0.101)\tData 0.019 (0.020)\tLoss 0.2826 (0.2289)\tPrec@1 (88.034)\n",
      "[200/391]\tTime 0.085 (0.101)\tData 0.020 (0.020)\tLoss 0.1626 (0.2287)\tPrec@1 (88.071)\n",
      "[240/391]\tTime 0.072 (0.098)\tData 0.020 (0.020)\tLoss 0.2825 (0.2310)\tPrec@1 (87.993)\n",
      "[280/391]\tTime 0.064 (0.094)\tData 0.019 (0.020)\tLoss 0.2506 (0.2318)\tPrec@1 (87.903)\n",
      "[320/391]\tTime 0.099 (0.093)\tData 0.020 (0.020)\tLoss 0.1863 (0.2341)\tPrec@1 (87.841)\n",
      "[360/391]\tTime 0.104 (0.094)\tData 0.020 (0.020)\tLoss 0.2476 (0.2327)\tPrec@1 (87.887)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.3491 (0.3491)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.5495 (0.4251)\tPrec@1 79.688 (85.118)\n",
      " * ResNet8 Prec@1 85.700\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2476 (0.2476)\tPrec@1 93.750 (93.750)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.6103 (0.4309)\tPrec@1 80.469 (86.166)\n",
      " * ResNet8 Prec@1 85.960\n",
      "\n",
      "===> epoch: 112/200\n",
      "Training:\n",
      "[0/391]\tTime 0.094 (0.094)\tData 0.021 (0.021)\tLoss 0.3295 (0.3295)\tPrec@1 (84.375)\n",
      "[40/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1619 (0.2212)\tPrec@1 (88.377)\n",
      "[80/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.2761 (0.2266)\tPrec@1 (88.262)\n",
      "[120/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.2052 (0.2260)\tPrec@1 (88.255)\n",
      "[160/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.2827 (0.2280)\tPrec@1 (88.155)\n",
      "[200/391]\tTime 0.092 (0.102)\tData 0.020 (0.020)\tLoss 0.2498 (0.2297)\tPrec@1 (88.036)\n",
      "[240/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1696 (0.2295)\tPrec@1 (88.064)\n",
      "[280/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.1822 (0.2303)\tPrec@1 (88.114)\n",
      "[320/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.2517 (0.2291)\tPrec@1 (88.174)\n",
      "[360/391]\tTime 0.105 (0.102)\tData 0.028 (0.020)\tLoss 0.2670 (0.2296)\tPrec@1 (88.117)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3792 (0.3792)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4574 (0.4313)\tPrec@1 85.156 (85.309)\n",
      " * ResNet8 Prec@1 85.470\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.2690 (0.2690)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.5215 (0.4219)\tPrec@1 82.812 (85.918)\n",
      " * ResNet8 Prec@1 85.630\n",
      "\n",
      "===> epoch: 113/200\n",
      "Training:\n",
      "[0/391]\tTime 0.091 (0.091)\tData 0.021 (0.021)\tLoss 0.2610 (0.2610)\tPrec@1 (85.156)\n",
      "[40/391]\tTime 0.105 (0.099)\tData 0.020 (0.020)\tLoss 0.1503 (0.2189)\tPrec@1 (88.205)\n",
      "[80/391]\tTime 0.103 (0.101)\tData 0.020 (0.021)\tLoss 0.2493 (0.2183)\tPrec@1 (88.301)\n",
      "[120/391]\tTime 0.108 (0.102)\tData 0.021 (0.021)\tLoss 0.2048 (0.2202)\tPrec@1 (88.397)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.3127 (0.2247)\tPrec@1 (88.276)\n",
      "[200/391]\tTime 0.103 (0.102)\tData 0.021 (0.021)\tLoss 0.1663 (0.2279)\tPrec@1 (88.145)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.1324 (0.2287)\tPrec@1 (88.090)\n",
      "[280/391]\tTime 0.107 (0.102)\tData 0.020 (0.021)\tLoss 0.2115 (0.2265)\tPrec@1 (88.162)\n",
      "[320/391]\tTime 0.095 (0.102)\tData 0.020 (0.021)\tLoss 0.1953 (0.2283)\tPrec@1 (88.108)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.020 (0.021)\tLoss 0.2380 (0.2304)\tPrec@1 (88.015)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3467 (0.3467)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4652 (0.4310)\tPrec@1 82.812 (85.766)\n",
      " * ResNet8 Prec@1 85.760\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2608 (0.2608)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.5703 (0.4158)\tPrec@1 85.156 (86.338)\n",
      " * ResNet8 Prec@1 85.950\n",
      "\n",
      "===> epoch: 114/200\n",
      "Training:\n",
      "[0/391]\tTime 0.102 (0.102)\tData 0.022 (0.022)\tLoss 0.1926 (0.1926)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.1596 (0.2093)\tPrec@1 (89.405)\n",
      "[80/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.2475 (0.2048)\tPrec@1 (89.159)\n",
      "[120/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.1471 (0.2065)\tPrec@1 (89.082)\n",
      "[160/391]\tTime 0.101 (0.103)\tData 0.021 (0.021)\tLoss 0.2305 (0.2076)\tPrec@1 (89.043)\n",
      "[200/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.2454 (0.2136)\tPrec@1 (88.779)\n",
      "[240/391]\tTime 0.105 (0.103)\tData 0.021 (0.021)\tLoss 0.3017 (0.2183)\tPrec@1 (88.596)\n",
      "[280/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.2404 (0.2188)\tPrec@1 (88.576)\n",
      "[320/391]\tTime 0.105 (0.103)\tData 0.020 (0.021)\tLoss 0.1982 (0.2191)\tPrec@1 (88.620)\n",
      "[360/391]\tTime 0.101 (0.103)\tData 0.021 (0.021)\tLoss 0.2321 (0.2233)\tPrec@1 (88.485)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3716 (0.3716)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4440 (0.4433)\tPrec@1 82.031 (85.252)\n",
      " * ResNet8 Prec@1 85.430\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 0.3632 (0.3632)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.5979 (0.4535)\tPrec@1 84.375 (84.737)\n",
      " * ResNet8 Prec@1 84.870\n",
      "\n",
      "===> epoch: 115/200\n",
      "Training:\n",
      "[0/391]\tTime 0.102 (0.102)\tData 0.021 (0.021)\tLoss 0.2283 (0.2283)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.101 (0.101)\tData 0.024 (0.020)\tLoss 0.2569 (0.2184)\tPrec@1 (88.453)\n",
      "[80/391]\tTime 0.099 (0.102)\tData 0.019 (0.020)\tLoss 0.1484 (0.2211)\tPrec@1 (88.532)\n",
      "[120/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.3159 (0.2252)\tPrec@1 (88.404)\n",
      "[160/391]\tTime 0.099 (0.102)\tData 0.019 (0.020)\tLoss 0.2757 (0.2241)\tPrec@1 (88.354)\n",
      "[200/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.1507 (0.2253)\tPrec@1 (88.394)\n",
      "[240/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2000 (0.2261)\tPrec@1 (88.417)\n",
      "[280/391]\tTime 0.094 (0.102)\tData 0.020 (0.020)\tLoss 0.4146 (0.2260)\tPrec@1 (88.387)\n",
      "[320/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1577 (0.2279)\tPrec@1 (88.342)\n",
      "[360/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2018 (0.2291)\tPrec@1 (88.286)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3252 (0.3252)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4885 (0.4209)\tPrec@1 85.938 (85.880)\n",
      " * ResNet8 Prec@1 85.740\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.3153 (0.3153)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5214 (0.4164)\tPrec@1 86.719 (85.938)\n",
      " * ResNet8 Prec@1 85.910\n",
      "\n",
      "===> epoch: 116/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.021 (0.021)\tLoss 0.1500 (0.1500)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2221 (0.2209)\tPrec@1 (88.396)\n",
      "[80/391]\tTime 0.083 (0.098)\tData 0.020 (0.020)\tLoss 0.1072 (0.2139)\tPrec@1 (88.657)\n",
      "[120/391]\tTime 0.071 (0.091)\tData 0.019 (0.020)\tLoss 0.0799 (0.2192)\tPrec@1 (88.533)\n",
      "[160/391]\tTime 0.080 (0.086)\tData 0.019 (0.020)\tLoss 0.1444 (0.2170)\tPrec@1 (88.568)\n",
      "[200/391]\tTime 0.105 (0.089)\tData 0.019 (0.020)\tLoss 0.2856 (0.2179)\tPrec@1 (88.433)\n",
      "[240/391]\tTime 0.104 (0.091)\tData 0.019 (0.020)\tLoss 0.1944 (0.2192)\tPrec@1 (88.304)\n",
      "[280/391]\tTime 0.105 (0.093)\tData 0.019 (0.020)\tLoss 0.2309 (0.2237)\tPrec@1 (88.120)\n",
      "[320/391]\tTime 0.106 (0.094)\tData 0.020 (0.020)\tLoss 0.3095 (0.2252)\tPrec@1 (88.084)\n",
      "[360/391]\tTime 0.105 (0.095)\tData 0.019 (0.020)\tLoss 0.2784 (0.2261)\tPrec@1 (88.069)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3634 (0.3634)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4646 (0.4245)\tPrec@1 83.594 (86.052)\n",
      " * ResNet8 Prec@1 85.990\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.2727 (0.2727)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.5687 (0.4169)\tPrec@1 82.812 (85.499)\n",
      " * ResNet8 Prec@1 85.560\n",
      "\n",
      "===> epoch: 117/200\n",
      "Training:\n",
      "[0/391]\tTime 0.090 (0.090)\tData 0.020 (0.020)\tLoss 0.2206 (0.2206)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1954 (0.2129)\tPrec@1 (88.624)\n",
      "[80/391]\tTime 0.090 (0.101)\tData 0.019 (0.020)\tLoss 0.1771 (0.2137)\tPrec@1 (88.735)\n",
      "[120/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.1541 (0.2207)\tPrec@1 (88.578)\n",
      "[160/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.4093 (0.2211)\tPrec@1 (88.631)\n",
      "[200/391]\tTime 0.108 (0.101)\tData 0.026 (0.020)\tLoss 0.1585 (0.2213)\tPrec@1 (88.573)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1127 (0.2213)\tPrec@1 (88.534)\n",
      "[280/391]\tTime 0.102 (0.101)\tData 0.020 (0.020)\tLoss 0.0828 (0.2219)\tPrec@1 (88.487)\n",
      "[320/391]\tTime 0.104 (0.101)\tData 0.020 (0.020)\tLoss 0.2618 (0.2215)\tPrec@1 (88.534)\n",
      "[360/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.2760 (0.2243)\tPrec@1 (88.431)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3216 (0.3216)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.4999 (0.4175)\tPrec@1 80.469 (86.033)\n",
      " * ResNet8 Prec@1 85.990\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.2985 (0.2985)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.6230 (0.4232)\tPrec@1 80.469 (86.109)\n",
      " * ResNet8 Prec@1 86.040\n",
      "\n",
      "===> epoch: 118/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.021 (0.021)\tLoss 0.1759 (0.1759)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.106 (0.100)\tData 0.020 (0.020)\tLoss 0.1665 (0.2029)\tPrec@1 (89.082)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.020 (0.020)\tLoss 0.2362 (0.2074)\tPrec@1 (88.812)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2271 (0.2115)\tPrec@1 (88.720)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1971 (0.2130)\tPrec@1 (88.708)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2441 (0.2162)\tPrec@1 (88.631)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.1789 (0.2168)\tPrec@1 (88.651)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2483 (0.2166)\tPrec@1 (88.640)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2683 (0.2201)\tPrec@1 (88.549)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2626 (0.2229)\tPrec@1 (88.465)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3640 (0.3640)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4899 (0.4165)\tPrec@1 82.812 (85.595)\n",
      " * ResNet8 Prec@1 85.740\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2931 (0.2931)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.5503 (0.4128)\tPrec@1 83.594 (86.052)\n",
      " * ResNet8 Prec@1 86.010\n",
      "\n",
      "===> epoch: 119/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.021 (0.021)\tLoss 0.2181 (0.2181)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.104 (0.099)\tData 0.020 (0.020)\tLoss 0.2811 (0.2190)\tPrec@1 (88.758)\n",
      "[80/391]\tTime 0.111 (0.101)\tData 0.025 (0.020)\tLoss 0.1932 (0.2080)\tPrec@1 (88.956)\n",
      "[120/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.2472 (0.2122)\tPrec@1 (88.798)\n",
      "[160/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.1442 (0.2130)\tPrec@1 (88.703)\n",
      "[200/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.2351 (0.2160)\tPrec@1 (88.596)\n",
      "[240/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.2716 (0.2167)\tPrec@1 (88.547)\n",
      "[280/391]\tTime 0.104 (0.101)\tData 0.019 (0.020)\tLoss 0.2412 (0.2210)\tPrec@1 (88.290)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.021 (0.020)\tLoss 0.2039 (0.2223)\tPrec@1 (88.306)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2507 (0.2226)\tPrec@1 (88.318)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3881 (0.3881)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5010 (0.4220)\tPrec@1 81.250 (85.804)\n",
      " * ResNet8 Prec@1 85.910\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2834 (0.2834)\tPrec@1 91.406 (91.406)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.5698 (0.4211)\tPrec@1 80.469 (86.204)\n",
      " * ResNet8 Prec@1 86.100\n",
      "\n",
      "===> epoch: 120/200\n",
      "Training:\n",
      "[0/391]\tTime 0.108 (0.108)\tData 0.022 (0.022)\tLoss 0.2363 (0.2363)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.102 (0.101)\tData 0.020 (0.020)\tLoss 0.1879 (0.2056)\tPrec@1 (88.815)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.3017 (0.2106)\tPrec@1 (88.764)\n",
      "[120/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1913 (0.2082)\tPrec@1 (88.759)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.2415 (0.2140)\tPrec@1 (88.669)\n",
      "[200/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.1279 (0.2168)\tPrec@1 (88.604)\n",
      "[240/391]\tTime 0.107 (0.102)\tData 0.021 (0.020)\tLoss 0.2100 (0.2200)\tPrec@1 (88.511)\n",
      "[280/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.2433 (0.2202)\tPrec@1 (88.551)\n",
      "[320/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.3831 (0.2204)\tPrec@1 (88.544)\n",
      "[360/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.2521 (0.2208)\tPrec@1 (88.506)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.026 (0.026)\tLoss 0.3770 (0.3770)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.024 (0.022)\tLoss 0.5041 (0.4286)\tPrec@1 83.594 (85.595)\n",
      " * ResNet8 Prec@1 85.850\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3120 (0.3120)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.6073 (0.4359)\tPrec@1 80.469 (85.271)\n",
      " * ResNet8 Prec@1 85.310\n",
      "\n",
      "===> epoch: 121/200\n",
      "Training:\n",
      "[0/391]\tTime 0.069 (0.069)\tData 0.022 (0.022)\tLoss 0.1925 (0.1925)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.099 (0.082)\tData 0.020 (0.021)\tLoss 0.2198 (0.2190)\tPrec@1 (88.739)\n",
      "[80/391]\tTime 0.098 (0.092)\tData 0.020 (0.020)\tLoss 0.2675 (0.2153)\tPrec@1 (88.850)\n",
      "[120/391]\tTime 0.097 (0.095)\tData 0.020 (0.020)\tLoss 0.1669 (0.2158)\tPrec@1 (88.875)\n",
      "[160/391]\tTime 0.099 (0.097)\tData 0.020 (0.020)\tLoss 0.2781 (0.2186)\tPrec@1 (88.805)\n",
      "[200/391]\tTime 0.098 (0.098)\tData 0.020 (0.020)\tLoss 0.0987 (0.2183)\tPrec@1 (88.837)\n",
      "[240/391]\tTime 0.097 (0.099)\tData 0.020 (0.021)\tLoss 0.2002 (0.2208)\tPrec@1 (88.732)\n",
      "[280/391]\tTime 0.098 (0.100)\tData 0.020 (0.021)\tLoss 0.2301 (0.2236)\tPrec@1 (88.607)\n",
      "[320/391]\tTime 0.098 (0.100)\tData 0.020 (0.021)\tLoss 0.2441 (0.2261)\tPrec@1 (88.478)\n",
      "[360/391]\tTime 0.096 (0.101)\tData 0.020 (0.021)\tLoss 0.2096 (0.2258)\tPrec@1 (88.459)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4481 (0.4481)\tPrec@1 83.594 (83.594)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.6324 (0.4848)\tPrec@1 78.906 (83.670)\n",
      " * ResNet8 Prec@1 83.980\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3386 (0.3386)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.5882 (0.4538)\tPrec@1 79.688 (85.290)\n",
      " * ResNet8 Prec@1 85.290\n",
      "\n",
      "===> epoch: 122/200\n",
      "Training:\n",
      "[0/391]\tTime 0.097 (0.097)\tData 0.021 (0.021)\tLoss 0.1972 (0.1972)\tPrec@1 (86.719)\n",
      "[40/391]\tTime 0.101 (0.101)\tData 0.020 (0.020)\tLoss 0.1543 (0.2117)\tPrec@1 (88.948)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1948 (0.2090)\tPrec@1 (88.908)\n",
      "[120/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.3135 (0.2128)\tPrec@1 (88.740)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.2425 (0.2200)\tPrec@1 (88.538)\n",
      "[200/391]\tTime 0.109 (0.102)\tData 0.020 (0.020)\tLoss 0.3097 (0.2205)\tPrec@1 (88.522)\n",
      "[240/391]\tTime 0.109 (0.102)\tData 0.022 (0.020)\tLoss 0.1913 (0.2219)\tPrec@1 (88.473)\n",
      "[280/391]\tTime 0.109 (0.103)\tData 0.021 (0.020)\tLoss 0.2241 (0.2228)\tPrec@1 (88.420)\n",
      "[320/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.2986 (0.2214)\tPrec@1 (88.486)\n",
      "[360/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.2302 (0.2234)\tPrec@1 (88.415)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3480 (0.3480)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4709 (0.4578)\tPrec@1 84.375 (85.061)\n",
      " * ResNet8 Prec@1 84.620\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3011 (0.3011)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5594 (0.4236)\tPrec@1 82.031 (85.423)\n",
      " * ResNet8 Prec@1 85.150\n",
      "\n",
      "===> epoch: 123/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.021 (0.021)\tLoss 0.1858 (0.1858)\tPrec@1 (86.719)\n",
      "[40/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.2484 (0.2127)\tPrec@1 (88.205)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1936 (0.2157)\tPrec@1 (88.445)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0967 (0.2154)\tPrec@1 (88.585)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.2018 (0.2125)\tPrec@1 (88.742)\n",
      "[200/391]\tTime 0.100 (0.103)\tData 0.021 (0.020)\tLoss 0.1932 (0.2143)\tPrec@1 (88.755)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1667 (0.2175)\tPrec@1 (88.567)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2759 (0.2179)\tPrec@1 (88.548)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1726 (0.2178)\tPrec@1 (88.627)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1896 (0.2202)\tPrec@1 (88.547)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3553 (0.3553)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5268 (0.4521)\tPrec@1 83.594 (84.832)\n",
      " * ResNet8 Prec@1 85.030\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2991 (0.2991)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.5854 (0.4549)\tPrec@1 80.469 (84.756)\n",
      " * ResNet8 Prec@1 84.840\n",
      "\n",
      "===> epoch: 124/200\n",
      "Training:\n",
      "[0/391]\tTime 0.097 (0.097)\tData 0.021 (0.021)\tLoss 0.1759 (0.1759)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.2021 (0.1848)\tPrec@1 (89.348)\n",
      "[80/391]\tTime 0.089 (0.102)\tData 0.022 (0.020)\tLoss 0.2110 (0.2062)\tPrec@1 (88.773)\n",
      "[120/391]\tTime 0.098 (0.101)\tData 0.021 (0.021)\tLoss 0.2704 (0.2138)\tPrec@1 (88.475)\n",
      "[160/391]\tTime 0.099 (0.102)\tData 0.021 (0.021)\tLoss 0.2041 (0.2149)\tPrec@1 (88.606)\n",
      "[200/391]\tTime 0.099 (0.102)\tData 0.021 (0.021)\tLoss 0.2641 (0.2185)\tPrec@1 (88.472)\n",
      "[240/391]\tTime 0.096 (0.102)\tData 0.020 (0.021)\tLoss 0.2249 (0.2209)\tPrec@1 (88.417)\n",
      "[280/391]\tTime 0.095 (0.102)\tData 0.020 (0.021)\tLoss 0.1807 (0.2206)\tPrec@1 (88.487)\n",
      "[320/391]\tTime 0.097 (0.102)\tData 0.020 (0.021)\tLoss 0.1864 (0.2211)\tPrec@1 (88.452)\n",
      "[360/391]\tTime 0.097 (0.102)\tData 0.020 (0.021)\tLoss 0.2606 (0.2231)\tPrec@1 (88.370)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3662 (0.3662)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4941 (0.4755)\tPrec@1 81.250 (84.165)\n",
      " * ResNet8 Prec@1 84.440\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3524 (0.3524)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.6026 (0.4506)\tPrec@1 78.906 (84.642)\n",
      " * ResNet8 Prec@1 84.520\n",
      "\n",
      "===> epoch: 125/200\n",
      "Training:\n",
      "[0/391]\tTime 0.095 (0.095)\tData 0.021 (0.021)\tLoss 0.2686 (0.2686)\tPrec@1 (85.156)\n",
      "[40/391]\tTime 0.106 (0.100)\tData 0.019 (0.020)\tLoss 0.1959 (0.2159)\tPrec@1 (88.948)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.2530 (0.2228)\tPrec@1 (88.503)\n",
      "[120/391]\tTime 0.108 (0.101)\tData 0.020 (0.020)\tLoss 0.1906 (0.2188)\tPrec@1 (88.662)\n",
      "[160/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.1926 (0.2224)\tPrec@1 (88.490)\n",
      "[200/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.2553 (0.2196)\tPrec@1 (88.619)\n",
      "[240/391]\tTime 0.102 (0.101)\tData 0.020 (0.020)\tLoss 0.2431 (0.2224)\tPrec@1 (88.450)\n",
      "[280/391]\tTime 0.073 (0.098)\tData 0.020 (0.020)\tLoss 0.1258 (0.2237)\tPrec@1 (88.376)\n",
      "[320/391]\tTime 0.064 (0.095)\tData 0.020 (0.020)\tLoss 0.1532 (0.2232)\tPrec@1 (88.376)\n",
      "[360/391]\tTime 0.106 (0.095)\tData 0.020 (0.020)\tLoss 0.2498 (0.2255)\tPrec@1 (88.320)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3905 (0.3905)\tPrec@1 85.156 (85.156)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4874 (0.4431)\tPrec@1 84.375 (85.004)\n",
      " * ResNet8 Prec@1 85.110\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3296 (0.3296)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5295 (0.4409)\tPrec@1 81.250 (85.118)\n",
      " * ResNet8 Prec@1 84.650\n",
      "\n",
      "===> epoch: 126/200\n",
      "Training:\n",
      "[0/391]\tTime 0.097 (0.097)\tData 0.021 (0.021)\tLoss 0.1851 (0.1851)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1988 (0.2097)\tPrec@1 (88.834)\n",
      "[80/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.2792 (0.2143)\tPrec@1 (88.648)\n",
      "[120/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.4139 (0.2193)\tPrec@1 (88.468)\n",
      "[160/391]\tTime 0.100 (0.102)\tData 0.021 (0.020)\tLoss 0.2084 (0.2180)\tPrec@1 (88.640)\n",
      "[200/391]\tTime 0.094 (0.103)\tData 0.020 (0.020)\tLoss 0.2446 (0.2207)\tPrec@1 (88.499)\n",
      "[240/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.2233 (0.2224)\tPrec@1 (88.391)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.096 (0.103)\tData 0.021 (0.020)\tLoss 0.1963 (0.2266)\tPrec@1 (88.284)\n",
      "[320/391]\tTime 0.098 (0.103)\tData 0.021 (0.020)\tLoss 0.1632 (0.2264)\tPrec@1 (88.240)\n",
      "[360/391]\tTime 0.098 (0.103)\tData 0.020 (0.020)\tLoss 0.1705 (0.2253)\tPrec@1 (88.270)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.4076 (0.4076)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.5610 (0.4835)\tPrec@1 81.250 (84.032)\n",
      " * ResNet8 Prec@1 83.800\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3037 (0.3037)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.5774 (0.4266)\tPrec@1 82.812 (85.271)\n",
      " * ResNet8 Prec@1 85.370\n",
      "\n",
      "===> epoch: 127/200\n",
      "Training:\n",
      "[0/391]\tTime 0.096 (0.096)\tData 0.021 (0.021)\tLoss 0.1531 (0.1531)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.098 (0.101)\tData 0.019 (0.020)\tLoss 0.1677 (0.2334)\tPrec@1 (87.862)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.2538 (0.2292)\tPrec@1 (88.108)\n",
      "[120/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.1260 (0.2286)\tPrec@1 (88.236)\n",
      "[160/391]\tTime 0.095 (0.102)\tData 0.020 (0.020)\tLoss 0.1740 (0.2289)\tPrec@1 (88.126)\n",
      "[200/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1432 (0.2273)\tPrec@1 (88.071)\n",
      "[240/391]\tTime 0.097 (0.102)\tData 0.019 (0.020)\tLoss 0.2790 (0.2272)\tPrec@1 (88.122)\n",
      "[280/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2128 (0.2271)\tPrec@1 (88.128)\n",
      "[320/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.2332 (0.2269)\tPrec@1 (88.196)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.021 (0.020)\tLoss 0.1889 (0.2275)\tPrec@1 (88.203)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3913 (0.3913)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.4640 (0.4461)\tPrec@1 82.031 (85.347)\n",
      " * ResNet8 Prec@1 85.110\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2400 (0.2400)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.5193 (0.4275)\tPrec@1 82.812 (85.328)\n",
      " * ResNet8 Prec@1 85.370\n",
      "\n",
      "===> epoch: 128/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.2457 (0.2457)\tPrec@1 (84.375)\n",
      "[40/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.1210 (0.2190)\tPrec@1 (88.681)\n",
      "[80/391]\tTime 0.107 (0.102)\tData 0.020 (0.020)\tLoss 0.2323 (0.2197)\tPrec@1 (88.551)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1986 (0.2209)\tPrec@1 (88.611)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.2968 (0.2178)\tPrec@1 (88.679)\n",
      "[200/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.2685 (0.2211)\tPrec@1 (88.452)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.021 (0.020)\tLoss 0.2426 (0.2238)\tPrec@1 (88.297)\n",
      "[280/391]\tTime 0.107 (0.103)\tData 0.020 (0.020)\tLoss 0.2828 (0.2245)\tPrec@1 (88.284)\n",
      "[320/391]\tTime 0.107 (0.103)\tData 0.020 (0.020)\tLoss 0.2033 (0.2241)\tPrec@1 (88.284)\n",
      "[360/391]\tTime 0.107 (0.103)\tData 0.020 (0.020)\tLoss 0.2133 (0.2236)\tPrec@1 (88.379)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.4003 (0.4003)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.5536 (0.4517)\tPrec@1 80.469 (84.413)\n",
      " * ResNet8 Prec@1 84.580\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3019 (0.3019)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.5172 (0.4259)\tPrec@1 84.375 (85.499)\n",
      " * ResNet8 Prec@1 85.380\n",
      "\n",
      "===> epoch: 129/200\n",
      "Training:\n",
      "[0/391]\tTime 0.094 (0.094)\tData 0.021 (0.021)\tLoss 0.2083 (0.2083)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.097 (0.100)\tData 0.020 (0.020)\tLoss 0.2796 (0.2036)\tPrec@1 (89.463)\n",
      "[80/391]\tTime 0.096 (0.101)\tData 0.020 (0.020)\tLoss 0.1847 (0.2134)\tPrec@1 (88.812)\n",
      "[120/391]\tTime 0.100 (0.102)\tData 0.022 (0.020)\tLoss 0.3203 (0.2179)\tPrec@1 (88.611)\n",
      "[160/391]\tTime 0.112 (0.102)\tData 0.033 (0.020)\tLoss 0.1505 (0.2186)\tPrec@1 (88.669)\n",
      "[200/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1774 (0.2243)\tPrec@1 (88.441)\n",
      "[240/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2936 (0.2221)\tPrec@1 (88.495)\n",
      "[280/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.2630 (0.2231)\tPrec@1 (88.465)\n",
      "[320/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1729 (0.2207)\tPrec@1 (88.573)\n",
      "[360/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2543 (0.2218)\tPrec@1 (88.506)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.4689 (0.4689)\tPrec@1 83.594 (83.594)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.5031 (0.4679)\tPrec@1 82.812 (84.108)\n",
      " * ResNet8 Prec@1 84.420\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3482 (0.3482)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.026 (0.022)\tLoss 0.4940 (0.4503)\tPrec@1 85.156 (84.375)\n",
      " * ResNet8 Prec@1 84.460\n",
      "\n",
      "===> epoch: 130/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.023 (0.023)\tLoss 0.1907 (0.1907)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.100 (0.103)\tData 0.022 (0.021)\tLoss 0.2092 (0.2187)\tPrec@1 (88.681)\n",
      "[80/391]\tTime 0.089 (0.102)\tData 0.022 (0.021)\tLoss 0.1668 (0.2189)\tPrec@1 (88.493)\n",
      "[120/391]\tTime 0.084 (0.099)\tData 0.021 (0.021)\tLoss 0.2723 (0.2213)\tPrec@1 (88.346)\n",
      "[160/391]\tTime 0.065 (0.094)\tData 0.021 (0.022)\tLoss 0.1675 (0.2229)\tPrec@1 (88.315)\n",
      "[200/391]\tTime 0.101 (0.091)\tData 0.020 (0.022)\tLoss 0.1890 (0.2223)\tPrec@1 (88.410)\n",
      "[240/391]\tTime 0.104 (0.093)\tData 0.021 (0.022)\tLoss 0.3569 (0.2200)\tPrec@1 (88.537)\n",
      "[280/391]\tTime 0.105 (0.094)\tData 0.021 (0.022)\tLoss 0.1719 (0.2207)\tPrec@1 (88.484)\n",
      "[320/391]\tTime 0.105 (0.095)\tData 0.021 (0.022)\tLoss 0.2562 (0.2221)\tPrec@1 (88.364)\n",
      "[360/391]\tTime 0.105 (0.096)\tData 0.021 (0.022)\tLoss 0.1958 (0.2242)\tPrec@1 (88.234)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.4139 (0.4139)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.5835 (0.4490)\tPrec@1 81.250 (84.947)\n",
      " * ResNet8 Prec@1 85.150\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.3250 (0.3250)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.026 (0.022)\tLoss 0.5313 (0.4505)\tPrec@1 83.594 (84.813)\n",
      " * ResNet8 Prec@1 84.830\n",
      "\n",
      "===> epoch: 131/200\n",
      "Training:\n",
      "[0/391]\tTime 0.108 (0.108)\tData 0.024 (0.024)\tLoss 0.1820 (0.1820)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.021 (0.021)\tLoss 0.1863 (0.2143)\tPrec@1 (88.396)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.022 (0.021)\tLoss 0.2372 (0.2219)\tPrec@1 (88.262)\n",
      "[120/391]\tTime 0.105 (0.103)\tData 0.021 (0.021)\tLoss 0.1753 (0.2190)\tPrec@1 (88.559)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.020 (0.021)\tLoss 0.2607 (0.2256)\tPrec@1 (88.247)\n",
      "[200/391]\tTime 0.104 (0.103)\tData 0.020 (0.021)\tLoss 0.2147 (0.2240)\tPrec@1 (88.281)\n",
      "[240/391]\tTime 0.106 (0.103)\tData 0.020 (0.021)\tLoss 0.2599 (0.2248)\tPrec@1 (88.262)\n",
      "[280/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.2529 (0.2270)\tPrec@1 (88.251)\n",
      "[320/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.2532 (0.2246)\tPrec@1 (88.289)\n",
      "[360/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1924 (0.2252)\tPrec@1 (88.212)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3753 (0.3753)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.5570 (0.4569)\tPrec@1 79.688 (84.870)\n",
      " * ResNet8 Prec@1 84.710\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3629 (0.3629)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.7029 (0.4700)\tPrec@1 78.906 (84.508)\n",
      " * ResNet8 Prec@1 84.670\n",
      "\n",
      "===> epoch: 132/200\n",
      "Training:\n",
      "[0/391]\tTime 0.098 (0.098)\tData 0.021 (0.021)\tLoss 0.2410 (0.2410)\tPrec@1 (86.719)\n",
      "[40/391]\tTime 0.097 (0.103)\tData 0.020 (0.020)\tLoss 0.1604 (0.2128)\tPrec@1 (88.396)\n",
      "[80/391]\tTime 0.097 (0.103)\tData 0.020 (0.020)\tLoss 0.2985 (0.2169)\tPrec@1 (88.223)\n",
      "[120/391]\tTime 0.097 (0.103)\tData 0.020 (0.020)\tLoss 0.2230 (0.2198)\tPrec@1 (88.217)\n",
      "[160/391]\tTime 0.098 (0.103)\tData 0.020 (0.020)\tLoss 0.1694 (0.2203)\tPrec@1 (88.213)\n",
      "[200/391]\tTime 0.096 (0.103)\tData 0.020 (0.020)\tLoss 0.2387 (0.2202)\tPrec@1 (88.285)\n",
      "[240/391]\tTime 0.096 (0.103)\tData 0.019 (0.020)\tLoss 0.3172 (0.2208)\tPrec@1 (88.255)\n",
      "[280/391]\tTime 0.095 (0.103)\tData 0.019 (0.020)\tLoss 0.3449 (0.2196)\tPrec@1 (88.356)\n",
      "[320/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.3312 (0.2220)\tPrec@1 (88.301)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.2014 (0.2240)\tPrec@1 (88.244)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2984 (0.2984)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4712 (0.4334)\tPrec@1 85.156 (85.556)\n",
      " * ResNet8 Prec@1 85.510\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3954 (0.3954)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.6078 (0.4602)\tPrec@1 82.031 (84.585)\n",
      " * ResNet8 Prec@1 84.240\n",
      "\n",
      "===> epoch: 133/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.020 (0.020)\tLoss 0.1302 (0.1302)\tPrec@1 (92.969)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.092 (0.100)\tData 0.020 (0.020)\tLoss 0.1868 (0.2097)\tPrec@1 (88.796)\n",
      "[80/391]\tTime 0.096 (0.100)\tData 0.020 (0.020)\tLoss 0.3039 (0.2150)\tPrec@1 (88.715)\n",
      "[120/391]\tTime 0.095 (0.100)\tData 0.019 (0.020)\tLoss 0.2005 (0.2183)\tPrec@1 (88.714)\n",
      "[160/391]\tTime 0.098 (0.100)\tData 0.020 (0.020)\tLoss 0.2197 (0.2160)\tPrec@1 (88.815)\n",
      "[200/391]\tTime 0.098 (0.101)\tData 0.021 (0.020)\tLoss 0.2265 (0.2179)\tPrec@1 (88.670)\n",
      "[240/391]\tTime 0.098 (0.101)\tData 0.020 (0.020)\tLoss 0.2405 (0.2195)\tPrec@1 (88.586)\n",
      "[280/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.3081 (0.2193)\tPrec@1 (88.607)\n",
      "[320/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.2414 (0.2209)\tPrec@1 (88.517)\n",
      "[360/391]\tTime 0.095 (0.102)\tData 0.019 (0.020)\tLoss 0.1977 (0.2248)\tPrec@1 (88.374)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4737 (0.4737)\tPrec@1 83.594 (83.594)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.6716 (0.4875)\tPrec@1 82.031 (84.070)\n",
      " * ResNet8 Prec@1 84.220\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4271 (0.4271)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.7142 (0.4809)\tPrec@1 76.562 (84.146)\n",
      " * ResNet8 Prec@1 84.480\n",
      "\n",
      "===> epoch: 134/200\n",
      "Training:\n",
      "[0/391]\tTime 0.084 (0.084)\tData 0.021 (0.021)\tLoss 0.2861 (0.2861)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.097 (0.100)\tData 0.019 (0.019)\tLoss 0.2539 (0.2101)\tPrec@1 (89.234)\n",
      "[80/391]\tTime 0.096 (0.101)\tData 0.020 (0.019)\tLoss 0.1289 (0.2197)\tPrec@1 (88.648)\n",
      "[120/391]\tTime 0.097 (0.101)\tData 0.019 (0.019)\tLoss 0.1414 (0.2192)\tPrec@1 (88.585)\n",
      "[160/391]\tTime 0.096 (0.101)\tData 0.019 (0.020)\tLoss 0.1583 (0.2171)\tPrec@1 (88.786)\n",
      "[200/391]\tTime 0.097 (0.101)\tData 0.019 (0.020)\tLoss 0.1870 (0.2163)\tPrec@1 (88.752)\n",
      "[240/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1397 (0.2177)\tPrec@1 (88.699)\n",
      "[280/391]\tTime 0.100 (0.102)\tData 0.021 (0.020)\tLoss 0.3501 (0.2216)\tPrec@1 (88.543)\n",
      "[320/391]\tTime 0.098 (0.102)\tData 0.021 (0.020)\tLoss 0.2146 (0.2218)\tPrec@1 (88.515)\n",
      "[360/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.2222 (0.2233)\tPrec@1 (88.448)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.3744 (0.3744)\tPrec@1 85.156 (85.156)\n",
      "Test: [40/79]\tTime 0.019 (0.022)\tLoss 0.5038 (0.4431)\tPrec@1 81.250 (85.213)\n",
      " * ResNet8 Prec@1 85.290\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3674 (0.3674)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.018 (0.023)\tLoss 0.5838 (0.4848)\tPrec@1 79.688 (83.784)\n",
      " * ResNet8 Prec@1 83.390\n",
      "\n",
      "===> epoch: 135/200\n",
      "Training:\n",
      "[0/391]\tTime 0.071 (0.071)\tData 0.022 (0.022)\tLoss 0.1433 (0.1433)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.066 (0.067)\tData 0.020 (0.021)\tLoss 0.1300 (0.2239)\tPrec@1 (87.995)\n",
      "[80/391]\tTime 0.100 (0.079)\tData 0.021 (0.021)\tLoss 0.1171 (0.2221)\tPrec@1 (88.117)\n",
      "[120/391]\tTime 0.103 (0.086)\tData 0.021 (0.021)\tLoss 0.1991 (0.2215)\tPrec@1 (88.372)\n",
      "[160/391]\tTime 0.103 (0.090)\tData 0.020 (0.021)\tLoss 0.1340 (0.2223)\tPrec@1 (88.364)\n",
      "[200/391]\tTime 0.106 (0.093)\tData 0.020 (0.021)\tLoss 0.2208 (0.2229)\tPrec@1 (88.285)\n",
      "[240/391]\tTime 0.103 (0.094)\tData 0.020 (0.021)\tLoss 0.3006 (0.2231)\tPrec@1 (88.291)\n",
      "[280/391]\tTime 0.105 (0.096)\tData 0.020 (0.021)\tLoss 0.2168 (0.2222)\tPrec@1 (88.354)\n",
      "[320/391]\tTime 0.104 (0.096)\tData 0.020 (0.021)\tLoss 0.2214 (0.2231)\tPrec@1 (88.325)\n",
      "[360/391]\tTime 0.105 (0.097)\tData 0.020 (0.021)\tLoss 0.2402 (0.2228)\tPrec@1 (88.320)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4601 (0.4601)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.5344 (0.4787)\tPrec@1 81.250 (84.223)\n",
      " * ResNet8 Prec@1 84.620\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.4599 (0.4599)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.7021 (0.5194)\tPrec@1 81.250 (82.946)\n",
      " * ResNet8 Prec@1 82.520\n",
      "\n",
      "===> epoch: 136/200\n",
      "Training:\n",
      "[0/391]\tTime 0.094 (0.094)\tData 0.022 (0.022)\tLoss 0.1848 (0.1848)\tPrec@1 (89.844)\n",
      "[40/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.2351 (0.2133)\tPrec@1 (88.567)\n",
      "[80/391]\tTime 0.099 (0.102)\tData 0.022 (0.020)\tLoss 0.2054 (0.2165)\tPrec@1 (88.571)\n",
      "[120/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1831 (0.2109)\tPrec@1 (88.772)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.021 (0.020)\tLoss 0.1317 (0.2132)\tPrec@1 (88.616)\n",
      "[200/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.2587 (0.2160)\tPrec@1 (88.483)\n",
      "[240/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.2645 (0.2177)\tPrec@1 (88.456)\n",
      "[280/391]\tTime 0.094 (0.102)\tData 0.020 (0.020)\tLoss 0.2585 (0.2179)\tPrec@1 (88.504)\n",
      "[320/391]\tTime 0.095 (0.102)\tData 0.019 (0.020)\tLoss 0.2983 (0.2205)\tPrec@1 (88.447)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.1865 (0.2214)\tPrec@1 (88.405)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3981 (0.3981)\tPrec@1 83.594 (83.594)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.5052 (0.4436)\tPrec@1 82.812 (84.966)\n",
      " * ResNet8 Prec@1 85.310\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 0.3031 (0.3031)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.5834 (0.5256)\tPrec@1 80.469 (82.717)\n",
      " * ResNet8 Prec@1 82.370\n",
      "\n",
      "===> epoch: 137/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.021 (0.021)\tLoss 0.3483 (0.3483)\tPrec@1 (83.594)\n",
      "[40/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.1653 (0.1996)\tPrec@1 (89.425)\n",
      "[80/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1790 (0.2044)\tPrec@1 (88.976)\n",
      "[120/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2596 (0.2169)\tPrec@1 (88.656)\n",
      "[160/391]\tTime 0.110 (0.102)\tData 0.021 (0.020)\tLoss 0.2351 (0.2210)\tPrec@1 (88.383)\n",
      "[200/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.3171 (0.2238)\tPrec@1 (88.289)\n",
      "[240/391]\tTime 0.112 (0.102)\tData 0.030 (0.020)\tLoss 0.3066 (0.2229)\tPrec@1 (88.333)\n",
      "[280/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.0855 (0.2227)\tPrec@1 (88.287)\n",
      "[320/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.3626 (0.2240)\tPrec@1 (88.247)\n",
      "[360/391]\tTime 0.095 (0.102)\tData 0.019 (0.020)\tLoss 0.2172 (0.2242)\tPrec@1 (88.266)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.4735 (0.4735)\tPrec@1 85.156 (85.156)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4431 (0.4725)\tPrec@1 84.375 (84.337)\n",
      " * ResNet8 Prec@1 84.450\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4870 (0.4870)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.6964 (0.6187)\tPrec@1 81.250 (80.069)\n",
      " * ResNet8 Prec@1 80.070\n",
      "\n",
      "===> epoch: 138/200\n",
      "Training:\n",
      "[0/391]\tTime 0.110 (0.110)\tData 0.021 (0.021)\tLoss 0.3052 (0.3052)\tPrec@1 (84.375)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.021 (0.021)\tLoss 0.1974 (0.2087)\tPrec@1 (88.662)\n",
      "[80/391]\tTime 0.105 (0.101)\tData 0.020 (0.021)\tLoss 0.1884 (0.2194)\tPrec@1 (88.194)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.2655 (0.2210)\tPrec@1 (88.243)\n",
      "[160/391]\tTime 0.103 (0.101)\tData 0.020 (0.021)\tLoss 0.2050 (0.2220)\tPrec@1 (88.354)\n",
      "[200/391]\tTime 0.105 (0.102)\tData 0.021 (0.021)\tLoss 0.3601 (0.2227)\tPrec@1 (88.351)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2583 (0.2200)\tPrec@1 (88.408)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2498 (0.2233)\tPrec@1 (88.292)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1117 (0.2230)\tPrec@1 (88.293)\n",
      "[360/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.2287 (0.2230)\tPrec@1 (88.275)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3833 (0.3833)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.4445 (0.4262)\tPrec@1 84.375 (85.690)\n",
      " * ResNet8 Prec@1 85.430\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2532 (0.2532)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5099 (0.4337)\tPrec@1 86.719 (85.271)\n",
      " * ResNet8 Prec@1 85.190\n",
      "\n",
      "===> epoch: 139/200\n",
      "Training:\n",
      "[0/391]\tTime 0.092 (0.092)\tData 0.021 (0.021)\tLoss 0.2081 (0.2081)\tPrec@1 (87.500)\n",
      "[40/391]\tTime 0.105 (0.101)\tData 0.020 (0.021)\tLoss 0.3166 (0.2170)\tPrec@1 (88.815)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.2645 (0.2166)\tPrec@1 (88.686)\n",
      "[120/391]\tTime 0.100 (0.102)\tData 0.019 (0.020)\tLoss 0.2399 (0.2152)\tPrec@1 (88.798)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2768 (0.2184)\tPrec@1 (88.660)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1304 (0.2200)\tPrec@1 (88.526)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.2678 (0.2237)\tPrec@1 (88.356)\n",
      "[280/391]\tTime 0.093 (0.101)\tData 0.030 (0.020)\tLoss 0.1833 (0.2217)\tPrec@1 (88.401)\n",
      "[320/391]\tTime 0.071 (0.098)\tData 0.019 (0.020)\tLoss 0.2768 (0.2232)\tPrec@1 (88.366)\n",
      "[360/391]\tTime 0.101 (0.096)\tData 0.022 (0.020)\tLoss 0.2663 (0.2256)\tPrec@1 (88.193)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.4069 (0.4069)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.030 (0.020)\tLoss 0.5407 (0.4791)\tPrec@1 82.031 (83.975)\n",
      " * ResNet8 Prec@1 83.980\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3531 (0.3531)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.5311 (0.4469)\tPrec@1 83.594 (84.985)\n",
      " * ResNet8 Prec@1 84.710\n",
      "\n",
      "===> epoch: 140/200\n",
      "Training:\n",
      "[0/391]\tTime 0.090 (0.090)\tData 0.021 (0.021)\tLoss 0.2487 (0.2487)\tPrec@1 (86.719)\n",
      "[40/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1791 (0.1990)\tPrec@1 (89.005)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1721 (0.2027)\tPrec@1 (89.053)\n",
      "[120/391]\tTime 0.098 (0.103)\tData 0.025 (0.021)\tLoss 0.2832 (0.2094)\tPrec@1 (88.759)\n",
      "[160/391]\tTime 0.098 (0.103)\tData 0.020 (0.021)\tLoss 0.2515 (0.2149)\tPrec@1 (88.548)\n",
      "[200/391]\tTime 0.096 (0.102)\tData 0.020 (0.021)\tLoss 0.3034 (0.2210)\tPrec@1 (88.351)\n",
      "[240/391]\tTime 0.078 (0.102)\tData 0.019 (0.021)\tLoss 0.2250 (0.2248)\tPrec@1 (88.203)\n",
      "[280/391]\tTime 0.088 (0.102)\tData 0.020 (0.021)\tLoss 0.2222 (0.2255)\tPrec@1 (88.181)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2248 (0.2258)\tPrec@1 (88.242)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1392 (0.2260)\tPrec@1 (88.242)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3849 (0.3849)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4551 (0.4590)\tPrec@1 85.156 (85.023)\n",
      " * ResNet8 Prec@1 84.960\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3306 (0.3306)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.6335 (0.4726)\tPrec@1 80.469 (84.432)\n",
      " * ResNet8 Prec@1 84.220\n",
      "\n",
      "===> epoch: 141/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.021 (0.021)\tLoss 0.1194 (0.1194)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.2649 (0.2140)\tPrec@1 (88.853)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.1652 (0.2164)\tPrec@1 (88.677)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.021 (0.020)\tLoss 0.1843 (0.2205)\tPrec@1 (88.585)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.1836 (0.2184)\tPrec@1 (88.674)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0985 (0.2189)\tPrec@1 (88.635)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1265 (0.2179)\tPrec@1 (88.703)\n",
      "[280/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.2589 (0.2197)\tPrec@1 (88.648)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1643 (0.2211)\tPrec@1 (88.593)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2186 (0.2228)\tPrec@1 (88.483)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3500 (0.3500)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.4992 (0.4284)\tPrec@1 81.250 (85.728)\n",
      " * ResNet8 Prec@1 85.680\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2665 (0.2665)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.4850 (0.4409)\tPrec@1 85.156 (85.061)\n",
      " * ResNet8 Prec@1 85.050\n",
      "\n",
      "===> epoch: 142/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.021 (0.021)\tLoss 0.2796 (0.2796)\tPrec@1 (85.938)\n",
      "[40/391]\tTime 0.100 (0.102)\tData 0.022 (0.021)\tLoss 0.2189 (0.2191)\tPrec@1 (88.605)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.021)\tLoss 0.1895 (0.2230)\tPrec@1 (88.378)\n",
      "[120/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.2764 (0.2205)\tPrec@1 (88.456)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.1550 (0.2236)\tPrec@1 (88.330)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.2051 (0.2259)\tPrec@1 (88.231)\n",
      "[240/391]\tTime 0.104 (0.103)\tData 0.020 (0.021)\tLoss 0.2917 (0.2239)\tPrec@1 (88.304)\n",
      "[280/391]\tTime 0.100 (0.102)\tData 0.019 (0.021)\tLoss 0.2507 (0.2241)\tPrec@1 (88.337)\n",
      "[320/391]\tTime 0.109 (0.102)\tData 0.020 (0.021)\tLoss 0.1983 (0.2240)\tPrec@1 (88.306)\n",
      "[360/391]\tTime 0.098 (0.102)\tData 0.020 (0.021)\tLoss 0.2984 (0.2275)\tPrec@1 (88.143)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3545 (0.3545)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4331 (0.4314)\tPrec@1 87.500 (85.995)\n",
      " * ResNet8 Prec@1 86.230\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3930 (0.3930)\tPrec@1 85.156 (85.156)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.5547 (0.4684)\tPrec@1 81.250 (84.318)\n",
      " * ResNet8 Prec@1 84.480\n",
      "\n",
      "===> epoch: 143/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.021 (0.021)\tLoss 0.1641 (0.1641)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2971 (0.2210)\tPrec@1 (88.300)\n",
      "[80/391]\tTime 0.101 (0.102)\tData 0.021 (0.020)\tLoss 0.1660 (0.2226)\tPrec@1 (88.291)\n",
      "[120/391]\tTime 0.102 (0.103)\tData 0.020 (0.021)\tLoss 0.2431 (0.2173)\tPrec@1 (88.527)\n",
      "[160/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.2234 (0.2150)\tPrec@1 (88.640)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.019 (0.020)\tLoss 0.1904 (0.2180)\tPrec@1 (88.530)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.2444 (0.2185)\tPrec@1 (88.635)\n",
      "[280/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.2245 (0.2195)\tPrec@1 (88.620)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1954 (0.2199)\tPrec@1 (88.542)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2683 (0.2210)\tPrec@1 (88.476)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.4598 (0.4598)\tPrec@1 82.031 (82.031)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.5574 (0.5143)\tPrec@1 79.688 (83.365)\n",
      " * ResNet8 Prec@1 83.340\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3521 (0.3521)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.5732 (0.4763)\tPrec@1 81.250 (83.822)\n",
      " * ResNet8 Prec@1 83.760\n",
      "\n",
      "===> epoch: 144/200\n",
      "Training:\n",
      "[0/391]\tTime 0.094 (0.094)\tData 0.022 (0.022)\tLoss 0.1118 (0.1118)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1383 (0.2385)\tPrec@1 (87.671)\n",
      "[80/391]\tTime 0.094 (0.102)\tData 0.020 (0.020)\tLoss 0.2060 (0.2326)\tPrec@1 (87.867)\n",
      "[120/391]\tTime 0.092 (0.102)\tData 0.020 (0.020)\tLoss 0.3328 (0.2271)\tPrec@1 (88.094)\n",
      "[160/391]\tTime 0.094 (0.098)\tData 0.020 (0.020)\tLoss 0.1994 (0.2268)\tPrec@1 (88.194)\n",
      "[200/391]\tTime 0.065 (0.093)\tData 0.022 (0.020)\tLoss 0.2623 (0.2248)\tPrec@1 (88.293)\n",
      "[240/391]\tTime 0.103 (0.092)\tData 0.020 (0.020)\tLoss 0.2152 (0.2246)\tPrec@1 (88.294)\n",
      "[280/391]\tTime 0.103 (0.093)\tData 0.020 (0.020)\tLoss 0.2596 (0.2254)\tPrec@1 (88.245)\n",
      "[320/391]\tTime 0.105 (0.094)\tData 0.020 (0.020)\tLoss 0.1409 (0.2240)\tPrec@1 (88.306)\n",
      "[360/391]\tTime 0.106 (0.095)\tData 0.019 (0.020)\tLoss 0.3237 (0.2251)\tPrec@1 (88.273)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3924 (0.3924)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4492 (0.4626)\tPrec@1 85.156 (84.699)\n",
      " * ResNet8 Prec@1 84.930\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3266 (0.3266)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4990 (0.4304)\tPrec@1 85.938 (85.271)\n",
      " * ResNet8 Prec@1 85.290\n",
      "\n",
      "===> epoch: 145/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.022 (0.022)\tLoss 0.1803 (0.1803)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.2687 (0.2131)\tPrec@1 (88.796)\n",
      "[80/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.2905 (0.2098)\tPrec@1 (88.956)\n",
      "[120/391]\tTime 0.101 (0.103)\tData 0.020 (0.020)\tLoss 0.1388 (0.2092)\tPrec@1 (89.037)\n",
      "[160/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.2064 (0.2144)\tPrec@1 (88.791)\n",
      "[200/391]\tTime 0.102 (0.103)\tData 0.022 (0.021)\tLoss 0.3278 (0.2185)\tPrec@1 (88.561)\n",
      "[240/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.2222 (0.2167)\tPrec@1 (88.544)\n",
      "[280/391]\tTime 0.107 (0.103)\tData 0.021 (0.021)\tLoss 0.2703 (0.2166)\tPrec@1 (88.523)\n",
      "[320/391]\tTime 0.101 (0.103)\tData 0.021 (0.021)\tLoss 0.1347 (0.2171)\tPrec@1 (88.564)\n",
      "[360/391]\tTime 0.106 (0.103)\tData 0.020 (0.021)\tLoss 0.2975 (0.2200)\tPrec@1 (88.483)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.4373 (0.4373)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 0.4759 (0.4672)\tPrec@1 80.469 (84.299)\n",
      " * ResNet8 Prec@1 84.120\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3512 (0.3512)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 0.6189 (0.4731)\tPrec@1 78.125 (84.261)\n",
      " * ResNet8 Prec@1 84.530\n",
      "\n",
      "===> epoch: 146/200\n",
      "Training:\n",
      "[0/391]\tTime 0.105 (0.105)\tData 0.022 (0.022)\tLoss 0.2875 (0.2875)\tPrec@1 (86.719)\n",
      "[40/391]\tTime 0.107 (0.102)\tData 0.020 (0.021)\tLoss 0.2528 (0.2207)\tPrec@1 (88.453)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.020 (0.021)\tLoss 0.3134 (0.2186)\tPrec@1 (88.561)\n",
      "[120/391]\tTime 0.096 (0.102)\tData 0.019 (0.021)\tLoss 0.1405 (0.2146)\tPrec@1 (88.837)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.092 (0.102)\tData 0.020 (0.021)\tLoss 0.1323 (0.2162)\tPrec@1 (88.762)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2000 (0.2194)\tPrec@1 (88.639)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2912 (0.2194)\tPrec@1 (88.618)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.3999 (0.2245)\tPrec@1 (88.367)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2965 (0.2270)\tPrec@1 (88.267)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.3172 (0.2270)\tPrec@1 (88.240)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3772 (0.3772)\tPrec@1 84.375 (84.375)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.5143 (0.4439)\tPrec@1 82.031 (85.213)\n",
      " * ResNet8 Prec@1 85.130\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3340 (0.3340)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.023 (0.020)\tLoss 0.5031 (0.4603)\tPrec@1 85.938 (84.184)\n",
      " * ResNet8 Prec@1 84.570\n",
      "\n",
      "===> epoch: 147/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.021 (0.021)\tLoss 0.1539 (0.1539)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.105 (0.100)\tData 0.019 (0.020)\tLoss 0.1315 (0.2035)\tPrec@1 (89.329)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.1420 (0.2214)\tPrec@1 (88.416)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.2547 (0.2188)\tPrec@1 (88.565)\n",
      "[160/391]\tTime 0.109 (0.101)\tData 0.022 (0.020)\tLoss 0.1098 (0.2169)\tPrec@1 (88.742)\n",
      "[200/391]\tTime 0.107 (0.101)\tData 0.022 (0.021)\tLoss 0.2210 (0.2175)\tPrec@1 (88.732)\n",
      "[240/391]\tTime 0.107 (0.102)\tData 0.021 (0.021)\tLoss 0.2041 (0.2188)\tPrec@1 (88.654)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.2899 (0.2219)\tPrec@1 (88.534)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.1430 (0.2227)\tPrec@1 (88.527)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.1969 (0.2226)\tPrec@1 (88.500)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.4730 (0.4730)\tPrec@1 85.156 (85.156)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.5081 (0.4686)\tPrec@1 81.250 (84.604)\n",
      " * ResNet8 Prec@1 84.610\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3655 (0.3655)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.5965 (0.4790)\tPrec@1 79.688 (84.299)\n",
      " * ResNet8 Prec@1 84.480\n",
      "\n",
      "===> epoch: 148/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.021 (0.021)\tLoss 0.1880 (0.1880)\tPrec@1 (89.844)\n",
      "[40/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.2707 (0.2148)\tPrec@1 (88.281)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.021 (0.020)\tLoss 0.1270 (0.2155)\tPrec@1 (88.474)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.2126 (0.2172)\tPrec@1 (88.559)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.3684 (0.2155)\tPrec@1 (88.616)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2469 (0.2186)\tPrec@1 (88.534)\n",
      "[240/391]\tTime 0.108 (0.102)\tData 0.021 (0.020)\tLoss 0.2317 (0.2228)\tPrec@1 (88.379)\n",
      "[280/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.2113 (0.2251)\tPrec@1 (88.276)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.1422 (0.2241)\tPrec@1 (88.349)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.2840 (0.2247)\tPrec@1 (88.309)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3785 (0.3785)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4423 (0.4339)\tPrec@1 83.594 (85.652)\n",
      " * ResNet8 Prec@1 85.250\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3765 (0.3765)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.4718 (0.4303)\tPrec@1 85.938 (85.252)\n",
      " * ResNet8 Prec@1 85.190\n",
      "\n",
      "===> epoch: 149/200\n",
      "Training:\n",
      "[0/391]\tTime 0.091 (0.091)\tData 0.021 (0.021)\tLoss 0.0889 (0.0889)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.074 (0.080)\tData 0.020 (0.020)\tLoss 0.3045 (0.2134)\tPrec@1 (88.357)\n",
      "[80/391]\tTime 0.100 (0.075)\tData 0.021 (0.020)\tLoss 0.1365 (0.2092)\tPrec@1 (88.792)\n",
      "[120/391]\tTime 0.118 (0.086)\tData 0.028 (0.023)\tLoss 0.2153 (0.2104)\tPrec@1 (88.817)\n",
      "[160/391]\tTime 0.107 (0.090)\tData 0.025 (0.024)\tLoss 0.2157 (0.2115)\tPrec@1 (88.946)\n",
      "[200/391]\tTime 0.091 (0.093)\tData 0.024 (0.024)\tLoss 0.1678 (0.2119)\tPrec@1 (88.961)\n",
      "[240/391]\tTime 0.094 (0.094)\tData 0.023 (0.024)\tLoss 0.2504 (0.2136)\tPrec@1 (88.862)\n",
      "[280/391]\tTime 0.106 (0.095)\tData 0.022 (0.024)\tLoss 0.2529 (0.2146)\tPrec@1 (88.718)\n",
      "[320/391]\tTime 0.097 (0.096)\tData 0.021 (0.023)\tLoss 0.1548 (0.2162)\tPrec@1 (88.639)\n",
      "[360/391]\tTime 0.102 (0.097)\tData 0.020 (0.023)\tLoss 0.2911 (0.2195)\tPrec@1 (88.493)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3137 (0.3137)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4628 (0.4237)\tPrec@1 82.812 (85.423)\n",
      " * ResNet8 Prec@1 85.540\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3630 (0.3630)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.6018 (0.4544)\tPrec@1 80.469 (84.966)\n",
      " * ResNet8 Prec@1 85.080\n",
      "\n",
      "===> epoch: 150/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.021 (0.021)\tLoss 0.1721 (0.1721)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.097 (0.101)\tData 0.020 (0.020)\tLoss 0.1357 (0.2155)\tPrec@1 (88.758)\n",
      "[80/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1346 (0.1986)\tPrec@1 (89.333)\n",
      "[120/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.2256 (0.1934)\tPrec@1 (89.463)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.2275 (0.1894)\tPrec@1 (89.596)\n",
      "[200/391]\tTime 0.077 (0.102)\tData 0.019 (0.020)\tLoss 0.2451 (0.1856)\tPrec@1 (89.704)\n",
      "[240/391]\tTime 0.100 (0.102)\tData 0.019 (0.020)\tLoss 0.1800 (0.1839)\tPrec@1 (89.844)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.1791 (0.1835)\tPrec@1 (89.863)\n",
      "[320/391]\tTime 0.107 (0.102)\tData 0.021 (0.020)\tLoss 0.1703 (0.1806)\tPrec@1 (90.000)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.1644 (0.1774)\tPrec@1 (90.140)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2729 (0.2729)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.4076 (0.3705)\tPrec@1 87.500 (87.881)\n",
      " * ResNet8 Prec@1 87.750\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2841 (0.2841)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.5165 (0.3821)\tPrec@1 82.812 (87.043)\n",
      " * ResNet8 Prec@1 87.160\n",
      "\n",
      "===> epoch: 151/200\n",
      "Training:\n",
      "[0/391]\tTime 0.095 (0.095)\tData 0.021 (0.021)\tLoss 0.2298 (0.2298)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.021 (0.021)\tLoss 0.2082 (0.1645)\tPrec@1 (90.682)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1620 (0.1676)\tPrec@1 (90.191)\n",
      "[120/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.1228 (0.1665)\tPrec@1 (90.276)\n",
      "[160/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1359 (0.1663)\tPrec@1 (90.247)\n",
      "[200/391]\tTime 0.103 (0.102)\tData 0.019 (0.020)\tLoss 0.0761 (0.1625)\tPrec@1 (90.477)\n",
      "[240/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.1195 (0.1605)\tPrec@1 (90.476)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.1157 (0.1593)\tPrec@1 (90.631)\n",
      "[320/391]\tTime 0.108 (0.102)\tData 0.021 (0.020)\tLoss 0.1392 (0.1581)\tPrec@1 (90.683)\n",
      "[360/391]\tTime 0.103 (0.103)\tData 0.020 (0.020)\tLoss 0.1918 (0.1595)\tPrec@1 (90.647)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.2861 (0.2861)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.027 (0.022)\tLoss 0.3905 (0.3759)\tPrec@1 90.625 (87.843)\n",
      " * ResNet8 Prec@1 87.810\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2733 (0.2733)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4868 (0.3788)\tPrec@1 84.375 (86.966)\n",
      " * ResNet8 Prec@1 87.260\n",
      "\n",
      "===> epoch: 152/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.022 (0.022)\tLoss 0.2904 (0.2904)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.104 (0.101)\tData 0.020 (0.021)\tLoss 0.2083 (0.1585)\tPrec@1 (91.044)\n",
      "[80/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.1101 (0.1488)\tPrec@1 (91.262)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.1941 (0.1506)\tPrec@1 (91.251)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.1239 (0.1523)\tPrec@1 (91.251)\n",
      "[200/391]\tTime 0.101 (0.102)\tData 0.020 (0.021)\tLoss 0.2242 (0.1499)\tPrec@1 (91.259)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.020 (0.021)\tLoss 0.1603 (0.1520)\tPrec@1 (91.124)\n",
      "[280/391]\tTime 0.098 (0.103)\tData 0.020 (0.021)\tLoss 0.1013 (0.1520)\tPrec@1 (91.131)\n",
      "[320/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.1537 (0.1509)\tPrec@1 (91.204)\n",
      "[360/391]\tTime 0.102 (0.103)\tData 0.020 (0.020)\tLoss 0.2046 (0.1521)\tPrec@1 (91.110)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2890 (0.2890)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4065 (0.3625)\tPrec@1 89.062 (88.205)\n",
      " * ResNet8 Prec@1 88.110\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2661 (0.2661)\tPrec@1 89.844 (89.844)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4769 (0.3679)\tPrec@1 82.031 (87.595)\n",
      " * ResNet8 Prec@1 87.700\n",
      "\n",
      "===> epoch: 153/200\n",
      "Training:\n",
      "[0/391]\tTime 0.093 (0.093)\tData 0.024 (0.024)\tLoss 0.1044 (0.1044)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.096 (0.100)\tData 0.020 (0.021)\tLoss 0.0858 (0.1371)\tPrec@1 (91.711)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.1114 (0.1367)\tPrec@1 (91.676)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.1493 (0.1407)\tPrec@1 (91.432)\n",
      "[160/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.0846 (0.1430)\tPrec@1 (91.338)\n",
      "[200/391]\tTime 0.111 (0.102)\tData 0.034 (0.021)\tLoss 0.1330 (0.1444)\tPrec@1 (91.317)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.023 (0.021)\tLoss 0.0645 (0.1444)\tPrec@1 (91.410)\n",
      "[280/391]\tTime 0.083 (0.102)\tData 0.020 (0.021)\tLoss 0.1812 (0.1444)\tPrec@1 (91.426)\n",
      "[320/391]\tTime 0.073 (0.100)\tData 0.020 (0.021)\tLoss 0.1531 (0.1446)\tPrec@1 (91.443)\n",
      "[360/391]\tTime 0.066 (0.097)\tData 0.021 (0.021)\tLoss 0.1146 (0.1455)\tPrec@1 (91.419)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3040 (0.3040)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.4194 (0.3643)\tPrec@1 88.281 (88.167)\n",
      " * ResNet8 Prec@1 88.100\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2648 (0.2648)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4990 (0.3674)\tPrec@1 82.812 (87.824)\n",
      " * ResNet8 Prec@1 87.830\n",
      "\n",
      "===> epoch: 154/200\n",
      "Training:\n",
      "[0/391]\tTime 0.097 (0.097)\tData 0.021 (0.021)\tLoss 0.2775 (0.2775)\tPrec@1 (83.594)\n",
      "[40/391]\tTime 0.098 (0.103)\tData 0.020 (0.020)\tLoss 0.1059 (0.1514)\tPrec@1 (91.540)\n",
      "[80/391]\tTime 0.097 (0.102)\tData 0.021 (0.020)\tLoss 0.0806 (0.1461)\tPrec@1 (91.426)\n",
      "[120/391]\tTime 0.095 (0.102)\tData 0.020 (0.020)\tLoss 0.1419 (0.1473)\tPrec@1 (91.374)\n",
      "[160/391]\tTime 0.098 (0.103)\tData 0.020 (0.020)\tLoss 0.1314 (0.1433)\tPrec@1 (91.586)\n",
      "[200/391]\tTime 0.097 (0.103)\tData 0.020 (0.020)\tLoss 0.2249 (0.1475)\tPrec@1 (91.453)\n",
      "[240/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.1725 (0.1461)\tPrec@1 (91.474)\n",
      "[280/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1623 (0.1450)\tPrec@1 (91.495)\n",
      "[320/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.1093 (0.1462)\tPrec@1 (91.428)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1593 (0.1455)\tPrec@1 (91.441)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2855 (0.2855)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4230 (0.3649)\tPrec@1 89.062 (88.129)\n",
      " * ResNet8 Prec@1 88.220\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2741 (0.2741)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.022 (0.020)\tLoss 0.4764 (0.3678)\tPrec@1 82.812 (87.824)\n",
      " * ResNet8 Prec@1 87.830\n",
      "\n",
      "===> epoch: 155/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.021 (0.021)\tLoss 0.0947 (0.0947)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.107 (0.103)\tData 0.021 (0.020)\tLoss 0.2596 (0.1447)\tPrec@1 (91.425)\n",
      "[80/391]\tTime 0.104 (0.103)\tData 0.021 (0.020)\tLoss 0.1197 (0.1437)\tPrec@1 (91.329)\n",
      "[120/391]\tTime 0.107 (0.103)\tData 0.021 (0.021)\tLoss 0.1521 (0.1500)\tPrec@1 (91.096)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.019 (0.021)\tLoss 0.0702 (0.1488)\tPrec@1 (91.149)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1964 (0.1472)\tPrec@1 (91.169)\n",
      "[240/391]\tTime 0.105 (0.103)\tData 0.019 (0.020)\tLoss 0.0941 (0.1435)\tPrec@1 (91.380)\n",
      "[280/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.1410 (0.1430)\tPrec@1 (91.423)\n",
      "[320/391]\tTime 0.106 (0.103)\tData 0.021 (0.020)\tLoss 0.1947 (0.1435)\tPrec@1 (91.387)\n",
      "[360/391]\tTime 0.102 (0.103)\tData 0.020 (0.021)\tLoss 0.0976 (0.1428)\tPrec@1 (91.460)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.2864 (0.2864)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.4398 (0.3640)\tPrec@1 86.719 (88.186)\n",
      " * ResNet8 Prec@1 88.120\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.2803 (0.2803)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4913 (0.3683)\tPrec@1 82.812 (87.652)\n",
      " * ResNet8 Prec@1 87.600\n",
      "\n",
      "===> epoch: 156/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.022 (0.022)\tLoss 0.1599 (0.1599)\tPrec@1 (89.844)\n",
      "[40/391]\tTime 0.104 (0.100)\tData 0.020 (0.020)\tLoss 0.1194 (0.1379)\tPrec@1 (91.673)\n",
      "[80/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.2160 (0.1331)\tPrec@1 (91.831)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1690 (0.1336)\tPrec@1 (91.858)\n",
      "[160/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.2003 (0.1349)\tPrec@1 (91.780)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.021 (0.020)\tLoss 0.1515 (0.1368)\tPrec@1 (91.756)\n",
      "[240/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1707 (0.1368)\tPrec@1 (91.737)\n",
      "[280/391]\tTime 0.102 (0.103)\tData 0.020 (0.020)\tLoss 0.1860 (0.1386)\tPrec@1 (91.668)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.021 (0.020)\tLoss 0.1564 (0.1389)\tPrec@1 (91.630)\n",
      "[360/391]\tTime 0.106 (0.103)\tData 0.023 (0.021)\tLoss 0.0875 (0.1391)\tPrec@1 (91.618)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2948 (0.2948)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.4341 (0.3609)\tPrec@1 86.719 (88.300)\n",
      " * ResNet8 Prec@1 88.170\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2613 (0.2613)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.4757 (0.3672)\tPrec@1 83.594 (87.786)\n",
      " * ResNet8 Prec@1 87.700\n",
      "\n",
      "===> epoch: 157/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.023 (0.023)\tLoss 0.0955 (0.0955)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.103 (0.102)\tData 0.019 (0.020)\tLoss 0.0968 (0.1369)\tPrec@1 (91.349)\n",
      "[80/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.1705 (0.1352)\tPrec@1 (91.483)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1710 (0.1354)\tPrec@1 (91.503)\n",
      "[160/391]\tTime 0.107 (0.103)\tData 0.020 (0.021)\tLoss 0.1048 (0.1356)\tPrec@1 (91.421)\n",
      "[200/391]\tTime 0.103 (0.102)\tData 0.019 (0.021)\tLoss 0.1126 (0.1355)\tPrec@1 (91.461)\n",
      "[240/391]\tTime 0.107 (0.102)\tData 0.021 (0.021)\tLoss 0.2334 (0.1360)\tPrec@1 (91.487)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.021 (0.020)\tLoss 0.1610 (0.1364)\tPrec@1 (91.523)\n",
      "[320/391]\tTime 0.109 (0.102)\tData 0.020 (0.021)\tLoss 0.1605 (0.1379)\tPrec@1 (91.472)\n",
      "[360/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1066 (0.1382)\tPrec@1 (91.491)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3267 (0.3267)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4182 (0.3653)\tPrec@1 89.062 (88.110)\n",
      " * ResNet8 Prec@1 87.970\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2574 (0.2574)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.4929 (0.3745)\tPrec@1 83.594 (87.691)\n",
      " * ResNet8 Prec@1 87.690\n",
      "\n",
      "===> epoch: 158/200\n",
      "Training:\n",
      "[0/391]\tTime 0.092 (0.092)\tData 0.021 (0.021)\tLoss 0.0987 (0.0987)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.099 (0.102)\tData 0.020 (0.021)\tLoss 0.1305 (0.1345)\tPrec@1 (91.616)\n",
      "[80/391]\tTime 0.097 (0.102)\tData 0.020 (0.021)\tLoss 0.1058 (0.1315)\tPrec@1 (91.763)\n",
      "[120/391]\tTime 0.098 (0.102)\tData 0.020 (0.021)\tLoss 0.1666 (0.1354)\tPrec@1 (91.665)\n",
      "[160/391]\tTime 0.084 (0.100)\tData 0.020 (0.020)\tLoss 0.1331 (0.1371)\tPrec@1 (91.595)\n",
      "[200/391]\tTime 0.070 (0.096)\tData 0.020 (0.020)\tLoss 0.0545 (0.1376)\tPrec@1 (91.671)\n",
      "[240/391]\tTime 0.105 (0.092)\tData 0.020 (0.020)\tLoss 0.1109 (0.1368)\tPrec@1 (91.701)\n",
      "[280/391]\tTime 0.110 (0.094)\tData 0.024 (0.020)\tLoss 0.2202 (0.1380)\tPrec@1 (91.654)\n",
      "[320/391]\tTime 0.105 (0.095)\tData 0.020 (0.020)\tLoss 0.0810 (0.1376)\tPrec@1 (91.633)\n",
      "[360/391]\tTime 0.106 (0.096)\tData 0.020 (0.020)\tLoss 0.1293 (0.1386)\tPrec@1 (91.582)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3102 (0.3102)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4335 (0.3605)\tPrec@1 88.281 (88.453)\n",
      " * ResNet8 Prec@1 88.290\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2644 (0.2644)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4927 (0.3676)\tPrec@1 84.375 (87.519)\n",
      " * ResNet8 Prec@1 87.630\n",
      "\n",
      "===> epoch: 159/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.021 (0.021)\tLoss 0.0745 (0.0745)\tPrec@1 (93.750)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1126 (0.1366)\tPrec@1 (91.635)\n",
      "[80/391]\tTime 0.102 (0.102)\tData 0.020 (0.020)\tLoss 0.0946 (0.1329)\tPrec@1 (91.879)\n",
      "[120/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.1091 (0.1278)\tPrec@1 (92.123)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1580 (0.1322)\tPrec@1 (91.916)\n",
      "[200/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1157 (0.1308)\tPrec@1 (92.016)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1949 (0.1298)\tPrec@1 (92.051)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.106 (0.102)\tData 0.022 (0.020)\tLoss 0.1384 (0.1323)\tPrec@1 (91.993)\n",
      "[320/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.0470 (0.1320)\tPrec@1 (91.973)\n",
      "[360/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1459 (0.1332)\tPrec@1 (91.947)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3205 (0.3205)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.024 (0.020)\tLoss 0.4276 (0.3620)\tPrec@1 87.500 (88.167)\n",
      " * ResNet8 Prec@1 88.130\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2434 (0.2434)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.4884 (0.3712)\tPrec@1 83.594 (87.633)\n",
      " * ResNet8 Prec@1 87.770\n",
      "\n",
      "===> epoch: 160/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.2663 (0.2663)\tPrec@1 (85.938)\n",
      "[40/391]\tTime 0.104 (0.103)\tData 0.020 (0.021)\tLoss 0.1633 (0.1466)\tPrec@1 (91.006)\n",
      "[80/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.1179 (0.1412)\tPrec@1 (91.358)\n",
      "[120/391]\tTime 0.108 (0.103)\tData 0.023 (0.020)\tLoss 0.2189 (0.1340)\tPrec@1 (91.677)\n",
      "[160/391]\tTime 0.102 (0.103)\tData 0.020 (0.020)\tLoss 0.1540 (0.1328)\tPrec@1 (91.683)\n",
      "[200/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.2042 (0.1318)\tPrec@1 (91.748)\n",
      "[240/391]\tTime 0.107 (0.103)\tData 0.020 (0.020)\tLoss 0.1508 (0.1326)\tPrec@1 (91.701)\n",
      "[280/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.0793 (0.1341)\tPrec@1 (91.640)\n",
      "[320/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0997 (0.1351)\tPrec@1 (91.623)\n",
      "[360/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.1821 (0.1345)\tPrec@1 (91.690)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3306 (0.3306)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.024 (0.022)\tLoss 0.4103 (0.3601)\tPrec@1 87.500 (88.453)\n",
      " * ResNet8 Prec@1 88.190\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2475 (0.2475)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4649 (0.3626)\tPrec@1 83.594 (87.538)\n",
      " * ResNet8 Prec@1 87.660\n",
      "\n",
      "===> epoch: 161/200\n",
      "Training:\n",
      "[0/391]\tTime 0.105 (0.105)\tData 0.022 (0.022)\tLoss 0.0290 (0.0290)\tPrec@1 (94.531)\n",
      "[40/391]\tTime 0.100 (0.099)\tData 0.020 (0.020)\tLoss 0.1084 (0.1262)\tPrec@1 (91.864)\n",
      "[80/391]\tTime 0.104 (0.101)\tData 0.020 (0.020)\tLoss 0.1642 (0.1309)\tPrec@1 (91.686)\n",
      "[120/391]\tTime 0.110 (0.102)\tData 0.021 (0.020)\tLoss 0.1061 (0.1306)\tPrec@1 (91.878)\n",
      "[160/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1338 (0.1305)\tPrec@1 (91.984)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.021 (0.020)\tLoss 0.0905 (0.1325)\tPrec@1 (91.865)\n",
      "[240/391]\tTime 0.110 (0.102)\tData 0.021 (0.020)\tLoss 0.1060 (0.1341)\tPrec@1 (91.792)\n",
      "[280/391]\tTime 0.098 (0.103)\tData 0.020 (0.021)\tLoss 0.1517 (0.1338)\tPrec@1 (91.784)\n",
      "[320/391]\tTime 0.096 (0.103)\tData 0.020 (0.020)\tLoss 0.2317 (0.1353)\tPrec@1 (91.691)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1995 (0.1349)\tPrec@1 (91.701)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3262 (0.3262)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4345 (0.3660)\tPrec@1 89.062 (88.186)\n",
      " * ResNet8 Prec@1 88.170\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2408 (0.2408)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4699 (0.3650)\tPrec@1 83.594 (88.091)\n",
      " * ResNet8 Prec@1 87.960\n",
      "\n",
      "===> epoch: 162/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.021 (0.021)\tLoss 0.1161 (0.1161)\tPrec@1 (89.844)\n",
      "[40/391]\tTime 0.106 (0.101)\tData 0.019 (0.020)\tLoss 0.1249 (0.1310)\tPrec@1 (91.825)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2043 (0.1289)\tPrec@1 (91.985)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.1680 (0.1318)\tPrec@1 (91.897)\n",
      "[160/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.0436 (0.1315)\tPrec@1 (91.872)\n",
      "[200/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.0863 (0.1302)\tPrec@1 (91.900)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.3112 (0.1314)\tPrec@1 (91.841)\n",
      "[280/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.1984 (0.1325)\tPrec@1 (91.776)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1216 (0.1329)\tPrec@1 (91.737)\n",
      "[360/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1023 (0.1332)\tPrec@1 (91.727)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3226 (0.3226)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4228 (0.3635)\tPrec@1 88.281 (88.605)\n",
      " * ResNet8 Prec@1 88.300\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2607 (0.2607)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4568 (0.3666)\tPrec@1 83.594 (87.671)\n",
      " * ResNet8 Prec@1 87.670\n",
      "\n",
      "===> epoch: 163/200\n",
      "Training:\n",
      "[0/391]\tTime 0.088 (0.088)\tData 0.022 (0.022)\tLoss 0.0750 (0.0750)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.072 (0.089)\tData 0.020 (0.020)\tLoss 0.1939 (0.1378)\tPrec@1 (91.463)\n",
      "[80/391]\tTime 0.072 (0.080)\tData 0.025 (0.020)\tLoss 0.1299 (0.1311)\tPrec@1 (91.647)\n",
      "[120/391]\tTime 0.105 (0.084)\tData 0.020 (0.021)\tLoss 0.1075 (0.1328)\tPrec@1 (91.593)\n",
      "[160/391]\tTime 0.104 (0.089)\tData 0.019 (0.020)\tLoss 0.1369 (0.1342)\tPrec@1 (91.494)\n",
      "[200/391]\tTime 0.116 (0.091)\tData 0.031 (0.020)\tLoss 0.1583 (0.1313)\tPrec@1 (91.682)\n",
      "[240/391]\tTime 0.103 (0.093)\tData 0.021 (0.020)\tLoss 0.1001 (0.1302)\tPrec@1 (91.769)\n",
      "[280/391]\tTime 0.104 (0.095)\tData 0.020 (0.021)\tLoss 0.1220 (0.1315)\tPrec@1 (91.770)\n",
      "[320/391]\tTime 0.105 (0.096)\tData 0.020 (0.020)\tLoss 0.1332 (0.1302)\tPrec@1 (91.839)\n",
      "[360/391]\tTime 0.105 (0.096)\tData 0.020 (0.020)\tLoss 0.1659 (0.1302)\tPrec@1 (91.856)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2997 (0.2997)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.4283 (0.3619)\tPrec@1 88.281 (88.643)\n",
      " * ResNet8 Prec@1 88.550\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2409 (0.2409)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.4501 (0.3668)\tPrec@1 85.938 (87.500)\n",
      " * ResNet8 Prec@1 87.630\n",
      "\n",
      "===> epoch: 164/200\n",
      "Training:\n",
      "[0/391]\tTime 0.103 (0.103)\tData 0.021 (0.021)\tLoss 0.0722 (0.0722)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.107 (0.102)\tData 0.021 (0.020)\tLoss 0.1010 (0.1145)\tPrec@1 (92.264)\n",
      "[80/391]\tTime 0.102 (0.103)\tData 0.020 (0.021)\tLoss 0.2679 (0.1268)\tPrec@1 (91.647)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.1835 (0.1325)\tPrec@1 (91.548)\n",
      "[160/391]\tTime 0.104 (0.103)\tData 0.020 (0.021)\tLoss 0.1128 (0.1358)\tPrec@1 (91.474)\n",
      "[200/391]\tTime 0.103 (0.103)\tData 0.020 (0.020)\tLoss 0.0335 (0.1321)\tPrec@1 (91.678)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0761 (0.1331)\tPrec@1 (91.704)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.1574 (0.1326)\tPrec@1 (91.729)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2066 (0.1325)\tPrec@1 (91.742)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.022 (0.020)\tLoss 0.0911 (0.1315)\tPrec@1 (91.798)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3199 (0.3199)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.021 (0.022)\tLoss 0.4566 (0.3703)\tPrec@1 87.500 (88.014)\n",
      " * ResNet8 Prec@1 87.930\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.028 (0.028)\tLoss 0.2402 (0.2402)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 0.4693 (0.3669)\tPrec@1 82.812 (87.500)\n",
      " * ResNet8 Prec@1 87.390\n",
      "\n",
      "===> epoch: 165/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.022 (0.022)\tLoss 0.1414 (0.1414)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.106 (0.105)\tData 0.020 (0.021)\tLoss 0.1454 (0.1207)\tPrec@1 (92.359)\n",
      "[80/391]\tTime 0.105 (0.104)\tData 0.021 (0.021)\tLoss 0.1323 (0.1311)\tPrec@1 (91.927)\n",
      "[120/391]\tTime 0.101 (0.104)\tData 0.020 (0.021)\tLoss 0.1177 (0.1306)\tPrec@1 (91.949)\n",
      "[160/391]\tTime 0.100 (0.103)\tData 0.020 (0.021)\tLoss 0.0856 (0.1310)\tPrec@1 (91.877)\n",
      "[200/391]\tTime 0.102 (0.103)\tData 0.021 (0.021)\tLoss 0.0914 (0.1302)\tPrec@1 (91.923)\n",
      "[240/391]\tTime 0.102 (0.102)\tData 0.021 (0.021)\tLoss 0.1009 (0.1319)\tPrec@1 (91.782)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.020 (0.021)\tLoss 0.1399 (0.1310)\tPrec@1 (91.840)\n",
      "[320/391]\tTime 0.107 (0.102)\tData 0.020 (0.021)\tLoss 0.1428 (0.1299)\tPrec@1 (91.900)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.0965 (0.1296)\tPrec@1 (91.956)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.2861 (0.2861)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.4554 (0.3609)\tPrec@1 88.281 (88.396)\n",
      " * ResNet8 Prec@1 88.390\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2467 (0.2467)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4663 (0.3597)\tPrec@1 83.594 (87.729)\n",
      " * ResNet8 Prec@1 87.830\n",
      "\n",
      "===> epoch: 166/200\n",
      "Training:\n",
      "[0/391]\tTime 0.093 (0.093)\tData 0.021 (0.021)\tLoss 0.0428 (0.0428)\tPrec@1 (97.656)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0771 (0.1225)\tPrec@1 (92.340)\n",
      "[80/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.0659 (0.1329)\tPrec@1 (91.763)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1935 (0.1301)\tPrec@1 (91.910)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.1245 (0.1292)\tPrec@1 (91.867)\n",
      "[200/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.0133 (0.1284)\tPrec@1 (91.908)\n",
      "[240/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.0587 (0.1286)\tPrec@1 (91.918)\n",
      "[280/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.1186 (0.1290)\tPrec@1 (91.932)\n",
      "[320/391]\tTime 0.107 (0.102)\tData 0.020 (0.020)\tLoss 0.0842 (0.1282)\tPrec@1 (92.012)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.021 (0.020)\tLoss 0.1046 (0.1272)\tPrec@1 (92.036)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2915 (0.2915)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.4659 (0.3666)\tPrec@1 87.500 (88.338)\n",
      " * ResNet8 Prec@1 88.280\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2589 (0.2589)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4571 (0.3650)\tPrec@1 84.375 (87.767)\n",
      " * ResNet8 Prec@1 87.820\n",
      "\n",
      "===> epoch: 167/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.022 (0.022)\tLoss 0.0595 (0.0595)\tPrec@1 (97.656)\n",
      "[40/391]\tTime 0.107 (0.103)\tData 0.020 (0.021)\tLoss 0.1255 (0.1288)\tPrec@1 (91.902)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0961 (0.1256)\tPrec@1 (91.802)\n",
      "[120/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0975 (0.1258)\tPrec@1 (91.845)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1193 (0.1287)\tPrec@1 (91.712)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.021 (0.020)\tLoss 0.1121 (0.1291)\tPrec@1 (91.655)\n",
      "[240/391]\tTime 0.105 (0.103)\tData 0.021 (0.021)\tLoss 0.3038 (0.1286)\tPrec@1 (91.740)\n",
      "[280/391]\tTime 0.106 (0.103)\tData 0.020 (0.021)\tLoss 0.0315 (0.1289)\tPrec@1 (91.729)\n",
      "[320/391]\tTime 0.091 (0.102)\tData 0.020 (0.021)\tLoss 0.0579 (0.1292)\tPrec@1 (91.749)\n",
      "[360/391]\tTime 0.072 (0.099)\tData 0.020 (0.021)\tLoss 0.1481 (0.1309)\tPrec@1 (91.698)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3204 (0.3204)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4477 (0.3658)\tPrec@1 86.719 (88.243)\n",
      " * ResNet8 Prec@1 88.330\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2595 (0.2595)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4684 (0.3688)\tPrec@1 82.812 (87.843)\n",
      " * ResNet8 Prec@1 87.870\n",
      "\n",
      "===> epoch: 168/200\n",
      "Training:\n",
      "[0/391]\tTime 0.107 (0.107)\tData 0.021 (0.021)\tLoss 0.0839 (0.0839)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1285 (0.1202)\tPrec@1 (91.978)\n",
      "[80/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.1637 (0.1182)\tPrec@1 (92.149)\n",
      "[120/391]\tTime 0.101 (0.102)\tData 0.019 (0.020)\tLoss 0.0706 (0.1266)\tPrec@1 (91.878)\n",
      "[160/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.1356 (0.1280)\tPrec@1 (91.896)\n",
      "[200/391]\tTime 0.100 (0.102)\tData 0.019 (0.020)\tLoss 0.1254 (0.1273)\tPrec@1 (92.001)\n",
      "[240/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1013 (0.1289)\tPrec@1 (91.935)\n",
      "[280/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.1087 (0.1276)\tPrec@1 (91.979)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0977 (0.1272)\tPrec@1 (91.983)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.1047 (0.1258)\tPrec@1 (92.060)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3045 (0.3045)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.4511 (0.3690)\tPrec@1 88.281 (88.110)\n",
      " * ResNet8 Prec@1 87.940\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2575 (0.2575)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4706 (0.3655)\tPrec@1 82.812 (87.576)\n",
      " * ResNet8 Prec@1 87.690\n",
      "\n",
      "===> epoch: 169/200\n",
      "Training:\n",
      "[0/391]\tTime 0.096 (0.096)\tData 0.021 (0.021)\tLoss 0.1931 (0.1931)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.097 (0.102)\tData 0.019 (0.020)\tLoss 0.0478 (0.1182)\tPrec@1 (92.188)\n",
      "[80/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.0563 (0.1228)\tPrec@1 (92.062)\n",
      "[120/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.0846 (0.1252)\tPrec@1 (92.013)\n",
      "[160/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1121 (0.1264)\tPrec@1 (91.848)\n",
      "[200/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.0975 (0.1273)\tPrec@1 (91.849)\n",
      "[240/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1143 (0.1256)\tPrec@1 (91.948)\n",
      "[280/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1067 (0.1254)\tPrec@1 (91.973)\n",
      "[320/391]\tTime 0.097 (0.102)\tData 0.019 (0.020)\tLoss 0.0897 (0.1247)\tPrec@1 (92.007)\n",
      "[360/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.0664 (0.1265)\tPrec@1 (91.947)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3073 (0.3073)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4341 (0.3694)\tPrec@1 87.500 (88.034)\n",
      " * ResNet8 Prec@1 88.050\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.2439 (0.2439)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.021 (0.022)\tLoss 0.4549 (0.3698)\tPrec@1 84.375 (87.691)\n",
      " * ResNet8 Prec@1 87.620\n",
      "\n",
      "===> epoch: 170/200\n",
      "Training:\n",
      "[0/391]\tTime 0.080 (0.080)\tData 0.024 (0.024)\tLoss 0.1788 (0.1788)\tPrec@1 (89.844)\n",
      "[40/391]\tTime 0.103 (0.103)\tData 0.021 (0.022)\tLoss 0.1558 (0.1259)\tPrec@1 (91.711)\n",
      "[80/391]\tTime 0.102 (0.103)\tData 0.021 (0.021)\tLoss 0.0925 (0.1317)\tPrec@1 (91.782)\n",
      "[120/391]\tTime 0.100 (0.103)\tData 0.020 (0.021)\tLoss 0.1345 (0.1327)\tPrec@1 (91.652)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.019 (0.021)\tLoss 0.0276 (0.1307)\tPrec@1 (91.765)\n",
      "[200/391]\tTime 0.102 (0.103)\tData 0.021 (0.021)\tLoss 0.1996 (0.1291)\tPrec@1 (91.857)\n",
      "[240/391]\tTime 0.103 (0.104)\tData 0.021 (0.021)\tLoss 0.1758 (0.1279)\tPrec@1 (91.918)\n",
      "[280/391]\tTime 0.100 (0.103)\tData 0.021 (0.021)\tLoss 0.0934 (0.1277)\tPrec@1 (91.898)\n",
      "[320/391]\tTime 0.105 (0.103)\tData 0.020 (0.021)\tLoss 0.1584 (0.1274)\tPrec@1 (91.925)\n",
      "[360/391]\tTime 0.105 (0.103)\tData 0.020 (0.021)\tLoss 0.2395 (0.1269)\tPrec@1 (91.988)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3441 (0.3441)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.022)\tLoss 0.4384 (0.3632)\tPrec@1 89.062 (88.357)\n",
      " * ResNet8 Prec@1 88.220\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2601 (0.2601)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4737 (0.3640)\tPrec@1 85.156 (87.500)\n",
      " * ResNet8 Prec@1 87.610\n",
      "\n",
      "===> epoch: 171/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.021 (0.021)\tLoss 0.1265 (0.1265)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0290 (0.1150)\tPrec@1 (92.149)\n",
      "[80/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.1032 (0.1169)\tPrec@1 (92.207)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1109 (0.1216)\tPrec@1 (92.246)\n",
      "[160/391]\tTime 0.106 (0.103)\tData 0.019 (0.020)\tLoss 0.0914 (0.1240)\tPrec@1 (92.154)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0980 (0.1252)\tPrec@1 (92.032)\n",
      "[240/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.2247 (0.1260)\tPrec@1 (92.019)\n",
      "[280/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1205 (0.1254)\tPrec@1 (92.001)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1529 (0.1256)\tPrec@1 (91.993)\n",
      "[360/391]\tTime 0.103 (0.102)\tData 0.019 (0.020)\tLoss 0.1680 (0.1254)\tPrec@1 (92.004)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3175 (0.3175)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4416 (0.3676)\tPrec@1 88.281 (88.281)\n",
      " * ResNet8 Prec@1 88.350\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2595 (0.2595)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4584 (0.3685)\tPrec@1 85.156 (87.443)\n",
      " * ResNet8 Prec@1 87.500\n",
      "\n",
      "===> epoch: 172/200\n",
      "Training:\n",
      "[0/391]\tTime 0.102 (0.102)\tData 0.021 (0.021)\tLoss 0.0689 (0.0689)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1232 (0.1320)\tPrec@1 (91.635)\n",
      "[80/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0866 (0.1289)\tPrec@1 (91.985)\n",
      "[120/391]\tTime 0.105 (0.103)\tData 0.020 (0.020)\tLoss 0.0610 (0.1283)\tPrec@1 (91.974)\n",
      "[160/391]\tTime 0.091 (0.102)\tData 0.020 (0.020)\tLoss 0.1585 (0.1284)\tPrec@1 (91.955)\n",
      "[200/391]\tTime 0.075 (0.098)\tData 0.024 (0.020)\tLoss 0.1409 (0.1255)\tPrec@1 (91.974)\n",
      "[240/391]\tTime 0.067 (0.094)\tData 0.019 (0.020)\tLoss 0.0804 (0.1262)\tPrec@1 (91.889)\n",
      "[280/391]\tTime 0.106 (0.094)\tData 0.020 (0.020)\tLoss 0.0993 (0.1275)\tPrec@1 (91.798)\n",
      "[320/391]\tTime 0.105 (0.095)\tData 0.019 (0.020)\tLoss 0.1381 (0.1266)\tPrec@1 (91.856)\n",
      "[360/391]\tTime 0.105 (0.096)\tData 0.020 (0.020)\tLoss 0.1369 (0.1250)\tPrec@1 (91.945)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3121 (0.3121)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4276 (0.3653)\tPrec@1 87.500 (88.434)\n",
      " * ResNet8 Prec@1 88.140\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.2364 (0.2364)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4932 (0.3691)\tPrec@1 83.594 (87.843)\n",
      " * ResNet8 Prec@1 87.760\n",
      "\n",
      "===> epoch: 173/200\n",
      "Training:\n",
      "[0/391]\tTime 0.105 (0.105)\tData 0.022 (0.022)\tLoss 0.1112 (0.1112)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0891 (0.1185)\tPrec@1 (92.245)\n",
      "[80/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.0830 (0.1222)\tPrec@1 (91.917)\n",
      "[120/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.1237 (0.1220)\tPrec@1 (92.000)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1321 (0.1224)\tPrec@1 (92.047)\n",
      "[200/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.0771 (0.1234)\tPrec@1 (92.020)\n",
      "[240/391]\tTime 0.106 (0.103)\tData 0.019 (0.020)\tLoss 0.0663 (0.1257)\tPrec@1 (91.941)\n",
      "[280/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0679 (0.1259)\tPrec@1 (91.973)\n",
      "[320/391]\tTime 0.107 (0.103)\tData 0.021 (0.020)\tLoss 0.1595 (0.1259)\tPrec@1 (91.993)\n",
      "[360/391]\tTime 0.107 (0.103)\tData 0.020 (0.021)\tLoss 0.0987 (0.1252)\tPrec@1 (92.032)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3098 (0.3098)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4415 (0.3652)\tPrec@1 87.500 (88.205)\n",
      " * ResNet8 Prec@1 88.280\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.2350 (0.2350)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.021 (0.020)\tLoss 0.4452 (0.3657)\tPrec@1 84.375 (87.481)\n",
      " * ResNet8 Prec@1 87.650\n",
      "\n",
      "===> epoch: 174/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.022 (0.022)\tLoss 0.1805 (0.1805)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.106 (0.099)\tData 0.020 (0.021)\tLoss 0.0573 (0.1150)\tPrec@1 (92.188)\n",
      "[80/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.1665 (0.1204)\tPrec@1 (91.995)\n",
      "[120/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.0681 (0.1202)\tPrec@1 (92.065)\n",
      "[160/391]\tTime 0.100 (0.101)\tData 0.019 (0.020)\tLoss 0.0664 (0.1207)\tPrec@1 (92.018)\n",
      "[200/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.0938 (0.1244)\tPrec@1 (91.947)\n",
      "[240/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.2925 (0.1238)\tPrec@1 (91.990)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1599 (0.1245)\tPrec@1 (92.007)\n",
      "[320/391]\tTime 0.099 (0.101)\tData 0.019 (0.020)\tLoss 0.2571 (0.1262)\tPrec@1 (91.947)\n",
      "[360/391]\tTime 0.102 (0.101)\tData 0.021 (0.020)\tLoss 0.0656 (0.1250)\tPrec@1 (92.023)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3020 (0.3020)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4515 (0.3645)\tPrec@1 87.500 (88.529)\n",
      " * ResNet8 Prec@1 88.320\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2404 (0.2404)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.019)\tLoss 0.4684 (0.3635)\tPrec@1 84.375 (87.671)\n",
      " * ResNet8 Prec@1 87.670\n",
      "\n",
      "===> epoch: 175/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.021 (0.021)\tLoss 0.0651 (0.0651)\tPrec@1 (93.750)\n",
      "[40/391]\tTime 0.101 (0.100)\tData 0.019 (0.020)\tLoss 0.1209 (0.1187)\tPrec@1 (92.416)\n",
      "[80/391]\tTime 0.107 (0.101)\tData 0.021 (0.020)\tLoss 0.1168 (0.1132)\tPrec@1 (92.776)\n",
      "[120/391]\tTime 0.100 (0.101)\tData 0.020 (0.020)\tLoss 0.1606 (0.1164)\tPrec@1 (92.497)\n",
      "[160/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.0544 (0.1173)\tPrec@1 (92.454)\n",
      "[200/391]\tTime 0.100 (0.101)\tData 0.020 (0.020)\tLoss 0.1216 (0.1168)\tPrec@1 (92.456)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1790 (0.1168)\tPrec@1 (92.431)\n",
      "[280/391]\tTime 0.091 (0.102)\tData 0.020 (0.020)\tLoss 0.0647 (0.1170)\tPrec@1 (92.382)\n",
      "[320/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.1638 (0.1181)\tPrec@1 (92.336)\n",
      "[360/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.0950 (0.1193)\tPrec@1 (92.291)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3316 (0.3316)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4479 (0.3668)\tPrec@1 87.500 (88.167)\n",
      " * ResNet8 Prec@1 87.990\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2574 (0.2574)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4470 (0.3651)\tPrec@1 85.156 (87.576)\n",
      " * ResNet8 Prec@1 87.600\n",
      "\n",
      "===> epoch: 176/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.022 (0.022)\tLoss 0.0704 (0.0704)\tPrec@1 (95.312)\n",
      "[40/391]\tTime 0.105 (0.100)\tData 0.019 (0.020)\tLoss 0.1190 (0.1141)\tPrec@1 (92.149)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.0815 (0.1173)\tPrec@1 (92.062)\n",
      "[120/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.0671 (0.1156)\tPrec@1 (92.342)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.2875 (0.1183)\tPrec@1 (92.236)\n",
      "[200/391]\tTime 0.104 (0.102)\tData 0.019 (0.020)\tLoss 0.1269 (0.1211)\tPrec@1 (92.067)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1417 (0.1192)\tPrec@1 (92.181)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0493 (0.1192)\tPrec@1 (92.213)\n",
      "[320/391]\tTime 0.104 (0.102)\tData 0.021 (0.020)\tLoss 0.1078 (0.1205)\tPrec@1 (92.180)\n",
      "[360/391]\tTime 0.107 (0.102)\tData 0.020 (0.020)\tLoss 0.1765 (0.1199)\tPrec@1 (92.231)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3324 (0.3324)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.4253 (0.3666)\tPrec@1 86.719 (88.034)\n",
      " * ResNet8 Prec@1 88.000\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2321 (0.2321)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4404 (0.3653)\tPrec@1 84.375 (87.767)\n",
      " * ResNet8 Prec@1 87.710\n",
      "\n",
      "===> epoch: 177/200\n",
      "Training:\n",
      "[0/391]\tTime 0.088 (0.088)\tData 0.021 (0.021)\tLoss 0.1776 (0.1776)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.084 (0.089)\tData 0.020 (0.020)\tLoss 0.1337 (0.1240)\tPrec@1 (92.111)\n",
      "[80/391]\tTime 0.073 (0.084)\tData 0.021 (0.020)\tLoss 0.2099 (0.1184)\tPrec@1 (92.332)\n",
      "[120/391]\tTime 0.104 (0.080)\tData 0.021 (0.020)\tLoss 0.1650 (0.1190)\tPrec@1 (92.226)\n",
      "[160/391]\tTime 0.103 (0.085)\tData 0.020 (0.020)\tLoss 0.1272 (0.1191)\tPrec@1 (92.231)\n",
      "[200/391]\tTime 0.099 (0.088)\tData 0.020 (0.020)\tLoss 0.1286 (0.1190)\tPrec@1 (92.156)\n",
      "[240/391]\tTime 0.102 (0.091)\tData 0.020 (0.020)\tLoss 0.0588 (0.1184)\tPrec@1 (92.249)\n",
      "[280/391]\tTime 0.107 (0.092)\tData 0.022 (0.020)\tLoss 0.0554 (0.1203)\tPrec@1 (92.199)\n",
      "[320/391]\tTime 0.107 (0.094)\tData 0.021 (0.021)\tLoss 0.1684 (0.1209)\tPrec@1 (92.173)\n",
      "[360/391]\tTime 0.103 (0.095)\tData 0.020 (0.021)\tLoss 0.2354 (0.1219)\tPrec@1 (92.107)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.031 (0.031)\tLoss 0.3253 (0.3253)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 0.4339 (0.3668)\tPrec@1 86.719 (88.148)\n",
      " * ResNet8 Prec@1 87.850\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2271 (0.2271)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4505 (0.3714)\tPrec@1 82.812 (87.786)\n",
      " * ResNet8 Prec@1 87.730\n",
      "\n",
      "===> epoch: 178/200\n",
      "Training:\n",
      "[0/391]\tTime 0.085 (0.085)\tData 0.022 (0.022)\tLoss 0.0943 (0.0943)\tPrec@1 (93.750)\n",
      "[40/391]\tTime 0.100 (0.101)\tData 0.020 (0.020)\tLoss 0.0874 (0.1204)\tPrec@1 (92.340)\n",
      "[80/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1274 (0.1201)\tPrec@1 (92.265)\n",
      "[120/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.0844 (0.1184)\tPrec@1 (92.336)\n",
      "[160/391]\tTime 0.108 (0.102)\tData 0.022 (0.020)\tLoss 0.0759 (0.1187)\tPrec@1 (92.280)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1229 (0.1209)\tPrec@1 (92.129)\n",
      "[240/391]\tTime 0.102 (0.103)\tData 0.020 (0.020)\tLoss 0.1060 (0.1206)\tPrec@1 (92.100)\n",
      "[280/391]\tTime 0.102 (0.103)\tData 0.021 (0.020)\tLoss 0.1221 (0.1212)\tPrec@1 (92.082)\n",
      "[320/391]\tTime 0.101 (0.103)\tData 0.020 (0.020)\tLoss 0.1050 (0.1227)\tPrec@1 (92.068)\n",
      "[360/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.1142 (0.1225)\tPrec@1 (92.055)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3404 (0.3404)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4601 (0.3670)\tPrec@1 85.938 (88.205)\n",
      " * ResNet8 Prec@1 88.200\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2458 (0.2458)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.035 (0.021)\tLoss 0.4465 (0.3734)\tPrec@1 85.156 (87.957)\n",
      " * ResNet8 Prec@1 87.850\n",
      "\n",
      "===> epoch: 179/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.022 (0.022)\tLoss 0.0759 (0.0759)\tPrec@1 (94.531)\n",
      "[40/391]\tTime 0.099 (0.103)\tData 0.020 (0.021)\tLoss 0.1421 (0.1235)\tPrec@1 (91.864)\n",
      "[80/391]\tTime 0.100 (0.103)\tData 0.021 (0.021)\tLoss 0.1645 (0.1253)\tPrec@1 (91.811)\n",
      "[120/391]\tTime 0.098 (0.103)\tData 0.020 (0.021)\tLoss 0.0428 (0.1257)\tPrec@1 (91.800)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[160/391]\tTime 0.097 (0.103)\tData 0.019 (0.020)\tLoss 0.0273 (0.1216)\tPrec@1 (91.979)\n",
      "[200/391]\tTime 0.110 (0.103)\tData 0.026 (0.020)\tLoss 0.0643 (0.1181)\tPrec@1 (92.125)\n",
      "[240/391]\tTime 0.099 (0.103)\tData 0.023 (0.020)\tLoss 0.0684 (0.1193)\tPrec@1 (92.132)\n",
      "[280/391]\tTime 0.099 (0.103)\tData 0.021 (0.020)\tLoss 0.1663 (0.1190)\tPrec@1 (92.157)\n",
      "[320/391]\tTime 0.098 (0.103)\tData 0.020 (0.020)\tLoss 0.0865 (0.1189)\tPrec@1 (92.146)\n",
      "[360/391]\tTime 0.097 (0.103)\tData 0.020 (0.020)\tLoss 0.1215 (0.1194)\tPrec@1 (92.133)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.3333 (0.3333)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.021 (0.022)\tLoss 0.4253 (0.3651)\tPrec@1 85.938 (88.453)\n",
      " * ResNet8 Prec@1 88.470\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2458 (0.2458)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4438 (0.3658)\tPrec@1 85.938 (87.729)\n",
      " * ResNet8 Prec@1 87.750\n",
      "\n",
      "===> epoch: 180/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.021 (0.021)\tLoss 0.1574 (0.1574)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.0872 (0.1144)\tPrec@1 (92.797)\n",
      "[80/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1787 (0.1174)\tPrec@1 (92.361)\n",
      "[120/391]\tTime 0.100 (0.103)\tData 0.020 (0.020)\tLoss 0.0677 (0.1164)\tPrec@1 (92.472)\n",
      "[160/391]\tTime 0.098 (0.103)\tData 0.020 (0.020)\tLoss 0.1304 (0.1183)\tPrec@1 (92.391)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.021 (0.020)\tLoss 0.1321 (0.1212)\tPrec@1 (92.285)\n",
      "[240/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1098 (0.1214)\tPrec@1 (92.291)\n",
      "[280/391]\tTime 0.093 (0.102)\tData 0.020 (0.020)\tLoss 0.1071 (0.1209)\tPrec@1 (92.268)\n",
      "[320/391]\tTime 0.095 (0.102)\tData 0.020 (0.020)\tLoss 0.1150 (0.1217)\tPrec@1 (92.207)\n",
      "[360/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1873 (0.1208)\tPrec@1 (92.239)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.023 (0.023)\tLoss 0.3108 (0.3108)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4538 (0.3700)\tPrec@1 89.062 (88.167)\n",
      " * ResNet8 Prec@1 88.250\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.2472 (0.2472)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.4803 (0.3757)\tPrec@1 85.156 (87.538)\n",
      " * ResNet8 Prec@1 87.750\n",
      "\n",
      "===> epoch: 181/200\n",
      "Training:\n",
      "[0/391]\tTime 0.105 (0.105)\tData 0.021 (0.021)\tLoss 0.1996 (0.1996)\tPrec@1 (88.281)\n",
      "[40/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.0849 (0.1221)\tPrec@1 (91.806)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.1481 (0.1254)\tPrec@1 (91.898)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1371 (0.1203)\tPrec@1 (92.187)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0953 (0.1169)\tPrec@1 (92.265)\n",
      "[200/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1337 (0.1159)\tPrec@1 (92.304)\n",
      "[240/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.1234 (0.1154)\tPrec@1 (92.369)\n",
      "[280/391]\tTime 0.103 (0.103)\tData 0.020 (0.020)\tLoss 0.0823 (0.1175)\tPrec@1 (92.288)\n",
      "[320/391]\tTime 0.092 (0.103)\tData 0.020 (0.020)\tLoss 0.1281 (0.1180)\tPrec@1 (92.282)\n",
      "[360/391]\tTime 0.083 (0.101)\tData 0.020 (0.020)\tLoss 0.0834 (0.1178)\tPrec@1 (92.285)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3345 (0.3345)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4277 (0.3680)\tPrec@1 87.500 (88.110)\n",
      " * ResNet8 Prec@1 88.120\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2535 (0.2535)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4577 (0.3713)\tPrec@1 85.156 (87.729)\n",
      " * ResNet8 Prec@1 87.760\n",
      "\n",
      "===> epoch: 182/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.021 (0.021)\tLoss 0.0704 (0.0704)\tPrec@1 (94.531)\n",
      "[40/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1316 (0.1302)\tPrec@1 (92.168)\n",
      "[80/391]\tTime 0.100 (0.102)\tData 0.021 (0.020)\tLoss 0.1481 (0.1225)\tPrec@1 (92.255)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2176 (0.1187)\tPrec@1 (92.388)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1551 (0.1200)\tPrec@1 (92.328)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.2125 (0.1190)\tPrec@1 (92.339)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.0731 (0.1185)\tPrec@1 (92.314)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.019 (0.020)\tLoss 0.1107 (0.1200)\tPrec@1 (92.232)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.027 (0.020)\tLoss 0.1412 (0.1205)\tPrec@1 (92.214)\n",
      "[360/391]\tTime 0.102 (0.102)\tData 0.022 (0.020)\tLoss 0.1067 (0.1195)\tPrec@1 (92.246)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3633 (0.3633)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4413 (0.3668)\tPrec@1 87.500 (88.014)\n",
      " * ResNet8 Prec@1 87.980\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2864 (0.2864)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.031 (0.021)\tLoss 0.4710 (0.3683)\tPrec@1 84.375 (87.557)\n",
      " * ResNet8 Prec@1 87.620\n",
      "\n",
      "===> epoch: 183/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.021 (0.021)\tLoss 0.2068 (0.2068)\tPrec@1 (89.062)\n",
      "[40/391]\tTime 0.104 (0.099)\tData 0.020 (0.020)\tLoss 0.1439 (0.1149)\tPrec@1 (92.340)\n",
      "[80/391]\tTime 0.107 (0.101)\tData 0.020 (0.020)\tLoss 0.0651 (0.1180)\tPrec@1 (92.448)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1018 (0.1175)\tPrec@1 (92.472)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1784 (0.1154)\tPrec@1 (92.508)\n",
      "[200/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.2412 (0.1156)\tPrec@1 (92.432)\n",
      "[240/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1959 (0.1161)\tPrec@1 (92.421)\n",
      "[280/391]\tTime 0.104 (0.102)\tData 0.021 (0.020)\tLoss 0.1379 (0.1167)\tPrec@1 (92.424)\n",
      "[320/391]\tTime 0.095 (0.102)\tData 0.020 (0.020)\tLoss 0.1085 (0.1171)\tPrec@1 (92.407)\n",
      "[360/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.0950 (0.1194)\tPrec@1 (92.283)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3564 (0.3564)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.021 (0.021)\tLoss 0.4406 (0.3701)\tPrec@1 87.500 (87.786)\n",
      " * ResNet8 Prec@1 87.970\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.022 (0.022)\tLoss 0.2721 (0.2721)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4573 (0.3708)\tPrec@1 84.375 (87.633)\n",
      " * ResNet8 Prec@1 87.630\n",
      "\n",
      "===> epoch: 184/200\n",
      "Training:\n",
      "[0/391]\tTime 0.086 (0.086)\tData 0.022 (0.022)\tLoss 0.1159 (0.1159)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.105 (0.099)\tData 0.020 (0.020)\tLoss 0.1744 (0.1217)\tPrec@1 (92.035)\n",
      "[80/391]\tTime 0.105 (0.101)\tData 0.020 (0.020)\tLoss 0.0712 (0.1108)\tPrec@1 (92.525)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.022 (0.020)\tLoss 0.0613 (0.1110)\tPrec@1 (92.504)\n",
      "[160/391]\tTime 0.107 (0.102)\tData 0.024 (0.021)\tLoss 0.1216 (0.1102)\tPrec@1 (92.508)\n",
      "[200/391]\tTime 0.102 (0.102)\tData 0.020 (0.021)\tLoss 0.0789 (0.1117)\tPrec@1 (92.460)\n",
      "[240/391]\tTime 0.101 (0.102)\tData 0.020 (0.021)\tLoss 0.1022 (0.1132)\tPrec@1 (92.440)\n",
      "[280/391]\tTime 0.104 (0.103)\tData 0.021 (0.021)\tLoss 0.0769 (0.1133)\tPrec@1 (92.446)\n",
      "[320/391]\tTime 0.101 (0.103)\tData 0.021 (0.021)\tLoss 0.0793 (0.1154)\tPrec@1 (92.341)\n",
      "[360/391]\tTime 0.101 (0.103)\tData 0.020 (0.021)\tLoss 0.0717 (0.1162)\tPrec@1 (92.322)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3077 (0.3077)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4478 (0.3655)\tPrec@1 87.500 (88.510)\n",
      " * ResNet8 Prec@1 88.360\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2705 (0.2705)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4973 (0.3830)\tPrec@1 85.938 (87.367)\n",
      " * ResNet8 Prec@1 87.460\n",
      "\n",
      "===> epoch: 185/200\n",
      "Training:\n",
      "[0/391]\tTime 0.096 (0.096)\tData 0.021 (0.021)\tLoss 0.0477 (0.0477)\tPrec@1 (92.969)\n",
      "[40/391]\tTime 0.098 (0.100)\tData 0.020 (0.021)\tLoss 0.0534 (0.1162)\tPrec@1 (92.283)\n",
      "[80/391]\tTime 0.098 (0.101)\tData 0.020 (0.020)\tLoss 0.0908 (0.1192)\tPrec@1 (92.110)\n",
      "[120/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.0858 (0.1237)\tPrec@1 (92.007)\n",
      "[160/391]\tTime 0.097 (0.102)\tData 0.020 (0.020)\tLoss 0.1455 (0.1218)\tPrec@1 (92.086)\n",
      "[200/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1380 (0.1209)\tPrec@1 (92.083)\n",
      "[240/391]\tTime 0.099 (0.102)\tData 0.021 (0.020)\tLoss 0.0393 (0.1215)\tPrec@1 (92.068)\n",
      "[280/391]\tTime 0.096 (0.102)\tData 0.020 (0.020)\tLoss 0.1159 (0.1212)\tPrec@1 (92.110)\n",
      "[320/391]\tTime 0.102 (0.102)\tData 0.024 (0.020)\tLoss 0.1245 (0.1220)\tPrec@1 (92.083)\n",
      "[360/391]\tTime 0.098 (0.102)\tData 0.020 (0.020)\tLoss 0.0958 (0.1201)\tPrec@1 (92.162)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3019 (0.3019)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.024 (0.021)\tLoss 0.4627 (0.3739)\tPrec@1 86.719 (88.262)\n",
      " * ResNet8 Prec@1 87.860\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2704 (0.2704)\tPrec@1 88.281 (88.281)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4775 (0.3779)\tPrec@1 84.375 (87.367)\n",
      " * ResNet8 Prec@1 87.610\n",
      "\n",
      "===> epoch: 186/200\n",
      "Training:\n",
      "[0/391]\tTime 0.104 (0.104)\tData 0.022 (0.022)\tLoss 0.1847 (0.1847)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.104 (0.099)\tData 0.020 (0.020)\tLoss 0.1176 (0.1155)\tPrec@1 (92.454)\n",
      "[80/391]\tTime 0.103 (0.100)\tData 0.021 (0.020)\tLoss 0.1361 (0.1187)\tPrec@1 (92.139)\n",
      "[120/391]\tTime 0.121 (0.101)\tData 0.021 (0.021)\tLoss 0.0959 (0.1154)\tPrec@1 (92.284)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2145 (0.1143)\tPrec@1 (92.357)\n",
      "[200/391]\tTime 0.084 (0.100)\tData 0.021 (0.020)\tLoss 0.0375 (0.1147)\tPrec@1 (92.339)\n",
      "[240/391]\tTime 0.073 (0.097)\tData 0.020 (0.020)\tLoss 0.1116 (0.1156)\tPrec@1 (92.317)\n",
      "[280/391]\tTime 0.098 (0.093)\tData 0.020 (0.020)\tLoss 0.0384 (0.1161)\tPrec@1 (92.249)\n",
      "[320/391]\tTime 0.098 (0.095)\tData 0.021 (0.020)\tLoss 0.1712 (0.1159)\tPrec@1 (92.229)\n",
      "[360/391]\tTime 0.097 (0.095)\tData 0.020 (0.020)\tLoss 0.1369 (0.1165)\tPrec@1 (92.205)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.3327 (0.3327)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.4301 (0.3654)\tPrec@1 88.281 (88.338)\n",
      " * ResNet8 Prec@1 88.340\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.2715 (0.2715)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4330 (0.3769)\tPrec@1 86.719 (87.710)\n",
      " * ResNet8 Prec@1 87.740\n",
      "\n",
      "===> epoch: 187/200\n",
      "Training:\n",
      "[0/391]\tTime 0.108 (0.108)\tData 0.021 (0.021)\tLoss 0.0793 (0.0793)\tPrec@1 (93.750)\n",
      "[40/391]\tTime 0.105 (0.100)\tData 0.020 (0.020)\tLoss 0.0498 (0.1054)\tPrec@1 (92.778)\n",
      "[80/391]\tTime 0.107 (0.101)\tData 0.021 (0.020)\tLoss 0.1310 (0.1108)\tPrec@1 (92.689)\n",
      "[120/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.1277 (0.1146)\tPrec@1 (92.381)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1809 (0.1133)\tPrec@1 (92.377)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0948 (0.1146)\tPrec@1 (92.289)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2000 (0.1164)\tPrec@1 (92.220)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0773 (0.1159)\tPrec@1 (92.290)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.019 (0.020)\tLoss 0.1656 (0.1159)\tPrec@1 (92.256)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1208 (0.1153)\tPrec@1 (92.287)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3457 (0.3457)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.023 (0.021)\tLoss 0.4315 (0.3679)\tPrec@1 89.062 (88.110)\n",
      " * ResNet8 Prec@1 87.970\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2766 (0.2766)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4770 (0.3795)\tPrec@1 85.938 (87.538)\n",
      " * ResNet8 Prec@1 87.600\n",
      "\n",
      "===> epoch: 188/200\n",
      "Training:\n",
      "[0/391]\tTime 0.100 (0.100)\tData 0.021 (0.021)\tLoss 0.0820 (0.0820)\tPrec@1 (95.312)\n",
      "[40/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0872 (0.1199)\tPrec@1 (92.511)\n",
      "[80/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.1112 (0.1112)\tPrec@1 (92.853)\n",
      "[120/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.2250 (0.1136)\tPrec@1 (92.581)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1207 (0.1152)\tPrec@1 (92.537)\n",
      "[200/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1446 (0.1149)\tPrec@1 (92.467)\n",
      "[240/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0957 (0.1186)\tPrec@1 (92.262)\n",
      "[280/391]\tTime 0.104 (0.103)\tData 0.020 (0.020)\tLoss 0.2116 (0.1196)\tPrec@1 (92.224)\n",
      "[320/391]\tTime 0.107 (0.103)\tData 0.020 (0.020)\tLoss 0.1454 (0.1183)\tPrec@1 (92.285)\n",
      "[360/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.0913 (0.1181)\tPrec@1 (92.272)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.3438 (0.3438)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.4526 (0.3705)\tPrec@1 86.719 (88.034)\n",
      " * ResNet8 Prec@1 88.000\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2549 (0.2549)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.4601 (0.3790)\tPrec@1 82.812 (87.214)\n",
      " * ResNet8 Prec@1 87.490\n",
      "\n",
      "===> epoch: 189/200\n",
      "Training:\n",
      "[0/391]\tTime 0.087 (0.087)\tData 0.022 (0.022)\tLoss 0.1482 (0.1482)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.109 (0.101)\tData 0.022 (0.022)\tLoss 0.0831 (0.1094)\tPrec@1 (92.511)\n",
      "[80/391]\tTime 0.101 (0.102)\tData 0.021 (0.022)\tLoss 0.1855 (0.1104)\tPrec@1 (92.747)\n",
      "[120/391]\tTime 0.104 (0.102)\tData 0.020 (0.021)\tLoss 0.1808 (0.1139)\tPrec@1 (92.517)\n",
      "[160/391]\tTime 0.106 (0.102)\tData 0.021 (0.021)\tLoss 0.0901 (0.1148)\tPrec@1 (92.420)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.0377 (0.1166)\tPrec@1 (92.347)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.022 (0.021)\tLoss 0.2598 (0.1181)\tPrec@1 (92.333)\n",
      "[280/391]\tTime 0.105 (0.102)\tData 0.021 (0.021)\tLoss 0.1180 (0.1182)\tPrec@1 (92.301)\n",
      "[320/391]\tTime 0.101 (0.102)\tData 0.020 (0.021)\tLoss 0.0771 (0.1151)\tPrec@1 (92.407)\n",
      "[360/391]\tTime 0.106 (0.102)\tData 0.020 (0.021)\tLoss 0.1365 (0.1163)\tPrec@1 (92.395)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3401 (0.3401)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.025 (0.022)\tLoss 0.4280 (0.3690)\tPrec@1 87.500 (87.900)\n",
      " * ResNet8 Prec@1 87.950\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2711 (0.2711)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.026 (0.021)\tLoss 0.3999 (0.3724)\tPrec@1 84.375 (87.290)\n",
      " * ResNet8 Prec@1 87.620\n",
      "\n",
      "===> epoch: 190/200\n",
      "Training:\n",
      "[0/391]\tTime 0.101 (0.101)\tData 0.022 (0.022)\tLoss 0.0978 (0.0978)\tPrec@1 (94.531)\n",
      "[40/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.0886 (0.1217)\tPrec@1 (91.978)\n",
      "[80/391]\tTime 0.103 (0.102)\tData 0.021 (0.021)\tLoss 0.2524 (0.1135)\tPrec@1 (92.236)\n",
      "[120/391]\tTime 0.102 (0.103)\tData 0.021 (0.021)\tLoss 0.0392 (0.1113)\tPrec@1 (92.426)\n",
      "[160/391]\tTime 0.105 (0.103)\tData 0.020 (0.021)\tLoss 0.1392 (0.1147)\tPrec@1 (92.425)\n",
      "[200/391]\tTime 0.097 (0.103)\tData 0.021 (0.021)\tLoss 0.1728 (0.1146)\tPrec@1 (92.483)\n",
      "[240/391]\tTime 0.094 (0.103)\tData 0.023 (0.021)\tLoss 0.0875 (0.1135)\tPrec@1 (92.560)\n",
      "[280/391]\tTime 0.098 (0.103)\tData 0.030 (0.021)\tLoss 0.1327 (0.1158)\tPrec@1 (92.454)\n",
      "[320/391]\tTime 0.104 (0.103)\tData 0.019 (0.021)\tLoss 0.1192 (0.1173)\tPrec@1 (92.397)\n",
      "[360/391]\tTime 0.105 (0.103)\tData 0.019 (0.021)\tLoss 0.1401 (0.1162)\tPrec@1 (92.415)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.3483 (0.3483)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4206 (0.3725)\tPrec@1 87.500 (87.919)\n",
      " * ResNet8 Prec@1 87.910\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.2396 (0.2396)\tPrec@1 91.406 (91.406)\n",
      "Test: [40/79]\tTime 0.029 (0.026)\tLoss 0.4570 (0.3788)\tPrec@1 85.156 (87.405)\n",
      " * ResNet8 Prec@1 87.720\n",
      "\n",
      "===> epoch: 191/200\n",
      "Training:\n",
      "[0/391]\tTime 0.092 (0.092)\tData 0.021 (0.021)\tLoss 0.0622 (0.0622)\tPrec@1 (96.875)\n",
      "[40/391]\tTime 0.104 (0.100)\tData 0.020 (0.020)\tLoss 0.0857 (0.1253)\tPrec@1 (91.806)\n",
      "[80/391]\tTime 0.072 (0.093)\tData 0.019 (0.020)\tLoss 0.1030 (0.1177)\tPrec@1 (92.149)\n",
      "[120/391]\tTime 0.066 (0.085)\tData 0.019 (0.020)\tLoss 0.0919 (0.1105)\tPrec@1 (92.517)\n",
      "[160/391]\tTime 0.100 (0.086)\tData 0.020 (0.020)\tLoss 0.1532 (0.1105)\tPrec@1 (92.498)\n",
      "[200/391]\tTime 0.098 (0.089)\tData 0.019 (0.020)\tLoss 0.0584 (0.1092)\tPrec@1 (92.514)\n",
      "[240/391]\tTime 0.118 (0.091)\tData 0.027 (0.020)\tLoss 0.0752 (0.1099)\tPrec@1 (92.573)\n",
      "[280/391]\tTime 0.100 (0.093)\tData 0.019 (0.020)\tLoss 0.1062 (0.1113)\tPrec@1 (92.524)\n",
      "[320/391]\tTime 0.101 (0.094)\tData 0.019 (0.020)\tLoss 0.2008 (0.1130)\tPrec@1 (92.475)\n",
      "[360/391]\tTime 0.098 (0.095)\tData 0.019 (0.020)\tLoss 0.2283 (0.1138)\tPrec@1 (92.445)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.3482 (0.3482)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4361 (0.3719)\tPrec@1 88.281 (88.205)\n",
      " * ResNet8 Prec@1 88.050\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2732 (0.2732)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4771 (0.3800)\tPrec@1 84.375 (87.443)\n",
      " * ResNet8 Prec@1 87.640\n",
      "\n",
      "===> epoch: 192/200\n",
      "Training:\n",
      "[0/391]\tTime 0.093 (0.093)\tData 0.021 (0.021)\tLoss 0.1576 (0.1576)\tPrec@1 (90.625)\n",
      "[40/391]\tTime 0.105 (0.099)\tData 0.019 (0.019)\tLoss 0.1369 (0.0988)\tPrec@1 (92.873)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.019 (0.019)\tLoss 0.2159 (0.1096)\tPrec@1 (92.535)\n",
      "[120/391]\tTime 0.109 (0.101)\tData 0.020 (0.019)\tLoss 0.0847 (0.1125)\tPrec@1 (92.426)\n",
      "[160/391]\tTime 0.105 (0.101)\tData 0.019 (0.020)\tLoss 0.1140 (0.1111)\tPrec@1 (92.551)\n",
      "[200/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.1001 (0.1101)\tPrec@1 (92.565)\n",
      "[240/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.0765 (0.1120)\tPrec@1 (92.479)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[280/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.1036 (0.1117)\tPrec@1 (92.499)\n",
      "[320/391]\tTime 0.106 (0.101)\tData 0.019 (0.019)\tLoss 0.0722 (0.1132)\tPrec@1 (92.475)\n",
      "[360/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.1047 (0.1120)\tPrec@1 (92.532)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.3663 (0.3663)\tPrec@1 85.938 (85.938)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4589 (0.3804)\tPrec@1 86.719 (87.729)\n",
      " * ResNet8 Prec@1 87.670\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2798 (0.2798)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4406 (0.3775)\tPrec@1 84.375 (87.195)\n",
      " * ResNet8 Prec@1 87.370\n",
      "\n",
      "===> epoch: 193/200\n",
      "Training:\n",
      "[0/391]\tTime 0.078 (0.078)\tData 0.020 (0.020)\tLoss 0.1086 (0.1086)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.095 (0.102)\tData 0.019 (0.020)\tLoss 0.1427 (0.1061)\tPrec@1 (92.721)\n",
      "[80/391]\tTime 0.096 (0.102)\tData 0.019 (0.019)\tLoss 0.0929 (0.1040)\tPrec@1 (92.901)\n",
      "[120/391]\tTime 0.096 (0.102)\tData 0.022 (0.019)\tLoss 0.1618 (0.1049)\tPrec@1 (92.891)\n",
      "[160/391]\tTime 0.097 (0.102)\tData 0.019 (0.020)\tLoss -0.0015 (0.1051)\tPrec@1 (92.896)\n",
      "[200/391]\tTime 0.095 (0.102)\tData 0.019 (0.019)\tLoss 0.1162 (0.1090)\tPrec@1 (92.681)\n",
      "[240/391]\tTime 0.096 (0.102)\tData 0.019 (0.020)\tLoss 0.0588 (0.1103)\tPrec@1 (92.667)\n",
      "[280/391]\tTime 0.096 (0.102)\tData 0.019 (0.019)\tLoss 0.0827 (0.1101)\tPrec@1 (92.655)\n",
      "[320/391]\tTime 0.098 (0.102)\tData 0.019 (0.019)\tLoss 0.1811 (0.1105)\tPrec@1 (92.682)\n",
      "[360/391]\tTime 0.096 (0.102)\tData 0.019 (0.019)\tLoss 0.1571 (0.1110)\tPrec@1 (92.633)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.3695 (0.3695)\tPrec@1 87.500 (87.500)\n",
      "Test: [40/79]\tTime 0.018 (0.020)\tLoss 0.4336 (0.3692)\tPrec@1 89.062 (87.995)\n",
      " * ResNet8 Prec@1 88.040\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.021 (0.021)\tLoss 0.2681 (0.2681)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.018 (0.019)\tLoss 0.4290 (0.3689)\tPrec@1 83.594 (87.786)\n",
      " * ResNet8 Prec@1 87.840\n",
      "\n",
      "===> epoch: 194/200\n",
      "Training:\n",
      "[0/391]\tTime 0.099 (0.099)\tData 0.020 (0.020)\tLoss 0.0786 (0.0786)\tPrec@1 (95.312)\n",
      "[40/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.1670 (0.1189)\tPrec@1 (92.321)\n",
      "[80/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.1055 (0.1206)\tPrec@1 (92.207)\n",
      "[120/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.1124 (0.1198)\tPrec@1 (92.162)\n",
      "[160/391]\tTime 0.104 (0.101)\tData 0.019 (0.019)\tLoss 0.2282 (0.1192)\tPrec@1 (92.226)\n",
      "[200/391]\tTime 0.106 (0.101)\tData 0.019 (0.019)\tLoss 0.1394 (0.1164)\tPrec@1 (92.370)\n",
      "[240/391]\tTime 0.105 (0.102)\tData 0.019 (0.019)\tLoss 0.2119 (0.1170)\tPrec@1 (92.311)\n",
      "[280/391]\tTime 0.105 (0.101)\tData 0.019 (0.019)\tLoss 0.1693 (0.1170)\tPrec@1 (92.254)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.019 (0.019)\tLoss 0.1112 (0.1150)\tPrec@1 (92.370)\n",
      "[360/391]\tTime 0.105 (0.102)\tData 0.019 (0.019)\tLoss 0.1422 (0.1154)\tPrec@1 (92.380)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.018 (0.018)\tLoss 0.3619 (0.3619)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4512 (0.3728)\tPrec@1 88.281 (87.976)\n",
      " * ResNet8 Prec@1 88.080\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2507 (0.2507)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4575 (0.3722)\tPrec@1 84.375 (87.462)\n",
      " * ResNet8 Prec@1 87.500\n",
      "\n",
      "===> epoch: 195/200\n",
      "Training:\n",
      "[0/391]\tTime 0.087 (0.087)\tData 0.021 (0.021)\tLoss 0.0232 (0.0232)\tPrec@1 (95.312)\n",
      "[40/391]\tTime 0.095 (0.099)\tData 0.021 (0.020)\tLoss 0.1695 (0.1033)\tPrec@1 (92.435)\n",
      "[80/391]\tTime 0.096 (0.100)\tData 0.020 (0.020)\tLoss 0.1345 (0.1076)\tPrec@1 (92.496)\n",
      "[120/391]\tTime 0.088 (0.100)\tData 0.019 (0.020)\tLoss 0.0898 (0.1058)\tPrec@1 (92.652)\n",
      "[160/391]\tTime 0.090 (0.100)\tData 0.019 (0.020)\tLoss 0.2340 (0.1115)\tPrec@1 (92.391)\n",
      "[200/391]\tTime 0.095 (0.101)\tData 0.020 (0.020)\tLoss 0.0864 (0.1119)\tPrec@1 (92.452)\n",
      "[240/391]\tTime 0.096 (0.101)\tData 0.020 (0.020)\tLoss 0.1450 (0.1095)\tPrec@1 (92.567)\n",
      "[280/391]\tTime 0.097 (0.101)\tData 0.019 (0.020)\tLoss 0.1429 (0.1114)\tPrec@1 (92.502)\n",
      "[320/391]\tTime 0.095 (0.101)\tData 0.019 (0.020)\tLoss 0.1268 (0.1125)\tPrec@1 (92.433)\n",
      "[360/391]\tTime 0.083 (0.100)\tData 0.019 (0.020)\tLoss 0.0533 (0.1133)\tPrec@1 (92.436)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3783 (0.3783)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.018 (0.021)\tLoss 0.4297 (0.3681)\tPrec@1 89.062 (88.472)\n",
      " * ResNet8 Prec@1 88.400\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2781 (0.2781)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4543 (0.3734)\tPrec@1 85.156 (87.252)\n",
      " * ResNet8 Prec@1 87.460\n",
      "\n",
      "===> epoch: 196/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.021 (0.021)\tLoss 0.0860 (0.0860)\tPrec@1 (92.188)\n",
      "[40/391]\tTime 0.108 (0.101)\tData 0.020 (0.020)\tLoss 0.0687 (0.1091)\tPrec@1 (92.778)\n",
      "[80/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.0989 (0.1062)\tPrec@1 (92.959)\n",
      "[120/391]\tTime 0.106 (0.101)\tData 0.020 (0.020)\tLoss 0.0921 (0.1046)\tPrec@1 (92.969)\n",
      "[160/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.1328 (0.1116)\tPrec@1 (92.663)\n",
      "[200/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1347 (0.1099)\tPrec@1 (92.654)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1025 (0.1113)\tPrec@1 (92.589)\n",
      "[280/391]\tTime 0.107 (0.102)\tData 0.021 (0.020)\tLoss 0.1101 (0.1112)\tPrec@1 (92.582)\n",
      "[320/391]\tTime 0.093 (0.102)\tData 0.019 (0.020)\tLoss 0.0735 (0.1122)\tPrec@1 (92.540)\n",
      "[360/391]\tTime 0.097 (0.102)\tData 0.019 (0.020)\tLoss 0.0794 (0.1125)\tPrec@1 (92.532)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.024 (0.024)\tLoss 0.3790 (0.3790)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.023 (0.022)\tLoss 0.4654 (0.3762)\tPrec@1 88.281 (87.805)\n",
      " * ResNet8 Prec@1 88.000\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.020 (0.020)\tLoss 0.2815 (0.2815)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.025 (0.021)\tLoss 0.4585 (0.3776)\tPrec@1 83.594 (87.157)\n",
      " * ResNet8 Prec@1 87.460\n",
      "\n",
      "===> epoch: 197/200\n",
      "Training:\n",
      "[0/391]\tTime 0.093 (0.093)\tData 0.021 (0.021)\tLoss 0.1175 (0.1175)\tPrec@1 (91.406)\n",
      "[40/391]\tTime 0.111 (0.102)\tData 0.020 (0.020)\tLoss 0.0579 (0.0985)\tPrec@1 (92.854)\n",
      "[80/391]\tTime 0.098 (0.102)\tData 0.019 (0.020)\tLoss 0.0659 (0.1048)\tPrec@1 (92.718)\n",
      "[120/391]\tTime 0.099 (0.102)\tData 0.020 (0.020)\tLoss 0.0169 (0.1041)\tPrec@1 (92.698)\n",
      "[160/391]\tTime 0.100 (0.101)\tData 0.019 (0.020)\tLoss 0.0648 (0.1055)\tPrec@1 (92.726)\n",
      "[200/391]\tTime 0.104 (0.101)\tData 0.020 (0.020)\tLoss 0.1276 (0.1053)\tPrec@1 (92.782)\n",
      "[240/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.0598 (0.1066)\tPrec@1 (92.693)\n",
      "[280/391]\tTime 0.108 (0.102)\tData 0.021 (0.020)\tLoss 0.0881 (0.1083)\tPrec@1 (92.655)\n",
      "[320/391]\tTime 0.106 (0.102)\tData 0.020 (0.020)\tLoss 0.1335 (0.1093)\tPrec@1 (92.601)\n",
      "[360/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.0425 (0.1101)\tPrec@1 (92.605)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.3894 (0.3894)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.019 (0.020)\tLoss 0.4452 (0.3723)\tPrec@1 88.281 (88.300)\n",
      " * ResNet8 Prec@1 88.210\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2701 (0.2701)\tPrec@1 89.844 (89.844)\n",
      "Test: [40/79]\tTime 0.019 (0.021)\tLoss 0.4440 (0.3766)\tPrec@1 85.938 (87.443)\n",
      " * ResNet8 Prec@1 87.450\n",
      "\n",
      "===> epoch: 198/200\n",
      "Training:\n",
      "[0/391]\tTime 0.102 (0.102)\tData 0.021 (0.021)\tLoss 0.0388 (0.0388)\tPrec@1 (95.312)\n",
      "[40/391]\tTime 0.106 (0.103)\tData 0.020 (0.020)\tLoss 0.1457 (0.1207)\tPrec@1 (92.035)\n",
      "[80/391]\tTime 0.101 (0.102)\tData 0.020 (0.020)\tLoss 0.1416 (0.1221)\tPrec@1 (92.120)\n",
      "[120/391]\tTime 0.099 (0.101)\tData 0.020 (0.020)\tLoss 0.1111 (0.1212)\tPrec@1 (92.020)\n",
      "[160/391]\tTime 0.095 (0.101)\tData 0.019 (0.020)\tLoss 0.1433 (0.1175)\tPrec@1 (92.129)\n",
      "[200/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.0415 (0.1162)\tPrec@1 (92.137)\n",
      "[240/391]\tTime 0.100 (0.102)\tData 0.020 (0.020)\tLoss 0.1519 (0.1153)\tPrec@1 (92.188)\n",
      "[280/391]\tTime 0.117 (0.102)\tData 0.027 (0.020)\tLoss 0.0581 (0.1149)\tPrec@1 (92.254)\n",
      "[320/391]\tTime 0.104 (0.102)\tData 0.020 (0.020)\tLoss 0.0944 (0.1148)\tPrec@1 (92.299)\n",
      "[360/391]\tTime 0.103 (0.102)\tData 0.020 (0.020)\tLoss 0.1039 (0.1144)\tPrec@1 (92.324)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.042 (0.042)\tLoss 0.3489 (0.3489)\tPrec@1 90.625 (90.625)\n",
      "Test: [40/79]\tTime 0.019 (0.022)\tLoss 0.4467 (0.3682)\tPrec@1 88.281 (88.053)\n",
      " * ResNet8 Prec@1 88.070\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.037 (0.037)\tLoss 0.2389 (0.2389)\tPrec@1 92.188 (92.188)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4555 (0.3725)\tPrec@1 84.375 (87.786)\n",
      " * ResNet8 Prec@1 87.750\n",
      "\n",
      "===> epoch: 199/200\n",
      "Training:\n",
      "[0/391]\tTime 0.106 (0.106)\tData 0.032 (0.032)\tLoss 0.0863 (0.0863)\tPrec@1 (92.969)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[40/391]\tTime 0.103 (0.102)\tData 0.020 (0.021)\tLoss 0.1630 (0.0995)\tPrec@1 (93.083)\n",
      "[80/391]\tTime 0.106 (0.103)\tData 0.019 (0.021)\tLoss 0.0759 (0.1056)\tPrec@1 (92.747)\n",
      "[120/391]\tTime 0.104 (0.101)\tData 0.019 (0.021)\tLoss 0.0842 (0.1098)\tPrec@1 (92.549)\n",
      "[160/391]\tTime 0.098 (0.101)\tData 0.019 (0.020)\tLoss 0.0700 (0.1095)\tPrec@1 (92.615)\n",
      "[200/391]\tTime 0.110 (0.102)\tData 0.024 (0.020)\tLoss 0.0946 (0.1099)\tPrec@1 (92.642)\n",
      "[240/391]\tTime 0.104 (0.102)\tData 0.019 (0.021)\tLoss 0.1288 (0.1088)\tPrec@1 (92.706)\n",
      "[280/391]\tTime 0.106 (0.102)\tData 0.022 (0.021)\tLoss 0.1814 (0.1096)\tPrec@1 (92.657)\n",
      "[320/391]\tTime 0.105 (0.102)\tData 0.020 (0.020)\tLoss 0.0850 (0.1088)\tPrec@1 (92.669)\n",
      "[360/391]\tTime 0.095 (0.102)\tData 0.019 (0.021)\tLoss 0.0896 (0.1095)\tPrec@1 (92.661)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.139 (0.139)\tLoss 0.3883 (0.3883)\tPrec@1 86.719 (86.719)\n",
      "Test: [40/79]\tTime 0.020 (0.029)\tLoss 0.4719 (0.3747)\tPrec@1 87.500 (87.843)\n",
      " * ResNet8 Prec@1 87.940\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.029 (0.029)\tLoss 0.2834 (0.2834)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.022 (0.021)\tLoss 0.4772 (0.3725)\tPrec@1 83.594 (87.786)\n",
      " * ResNet8 Prec@1 87.890\n",
      "\n",
      "===> epoch: 200/200\n",
      "Training:\n",
      "[0/391]\tTime 0.426 (0.426)\tData 0.286 (0.286)\tLoss 0.0608 (0.0608)\tPrec@1 (95.312)\n",
      "[40/391]\tTime 0.137 (0.271)\tData 0.052 (0.185)\tLoss 0.0238 (0.1063)\tPrec@1 (92.702)\n",
      "[80/391]\tTime 0.105 (0.321)\tData 0.025 (0.242)\tLoss 0.2238 (0.1030)\tPrec@1 (92.824)\n",
      "[120/391]\tTime 0.101 (0.260)\tData 0.022 (0.180)\tLoss 0.0658 (0.1031)\tPrec@1 (92.827)\n",
      "[160/391]\tTime 0.665 (0.320)\tData 0.591 (0.239)\tLoss 0.0497 (0.1043)\tPrec@1 (92.818)\n",
      "[200/391]\tTime 0.137 (0.322)\tData 0.053 (0.241)\tLoss 0.0571 (0.1069)\tPrec@1 (92.685)\n",
      "[240/391]\tTime 0.109 (0.289)\tData 0.024 (0.208)\tLoss 0.2080 (0.1064)\tPrec@1 (92.716)\n",
      "[280/391]\tTime 0.175 (0.264)\tData 0.093 (0.183)\tLoss 0.1371 (0.1073)\tPrec@1 (92.641)\n",
      "[320/391]\tTime 0.105 (0.245)\tData 0.020 (0.164)\tLoss 0.0476 (0.1072)\tPrec@1 (92.616)\n",
      "[360/391]\tTime 0.106 (0.230)\tData 0.020 (0.149)\tLoss 0.1047 (0.1086)\tPrec@1 (92.577)\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.025 (0.025)\tLoss 0.3602 (0.3602)\tPrec@1 89.062 (89.062)\n",
      "Test: [40/79]\tTime 0.020 (0.021)\tLoss 0.4725 (0.3765)\tPrec@1 85.938 (88.186)\n",
      " * ResNet8 Prec@1 88.270\n",
      "Testing:\n",
      "Test: [0/79]\tTime 0.019 (0.019)\tLoss 0.2969 (0.2969)\tPrec@1 88.281 (88.281)\n",
      "Test: [40/79]\tTime 0.020 (0.020)\tLoss 0.4045 (0.3747)\tPrec@1 85.938 (87.710)\n",
      " * ResNet8 Prec@1 87.550\n",
      "@ BEST Prec: 88.5500\n",
      "--- 126.612 mins ---\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",
    "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",
    "\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='PyTorch DML 2S')\n",
    "\n",
    "parser.add_argument('--dataset',\n",
    "                    choices=['CIFAR10',\n",
    "                             'CIFAR100'\n",
    "                            ],\n",
    "                    default='CIFAR10')\n",
    "parser.add_argument('--net1',\n",
    "                    choices=['ResNet8',\n",
    "                             'ResNet15',\n",
    "                             'ResNet16',\n",
    "                             'ResNet20',\n",
    "                             'ResNet32',\n",
    "                             'ResNet50',\n",
    "                             'ResNet56',\n",
    "                             'ResNet110'\n",
    "                            ],\n",
    "                    default='ResNet20')\n",
    "parser.add_argument('--net2',\n",
    "                    choices=['ResNet8',\n",
    "                             'ResNet15',\n",
    "                             'ResNet16',\n",
    "                             'ResNet20',\n",
    "                             'ResNet32',\n",
    "                             'ResNet50',\n",
    "                             'ResNet56',\n",
    "                             'ResNet110'\n",
    "                            ],\n",
    "                    default='ResNet20')\n",
    "\n",
    "parser.add_argument('--n_class', type=int, default=10, metavar='N', help='num of classes')\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:1', type=str, help='device: cuda or cpu')\n",
    "parser.add_argument('--print_freq', type=int, default=40, metavar='N', help='how many batches to wait before logging training status')\n",
    "\n",
    "config = ['--epochs', '200', '--net1', 'ResNet8', '--net2', 'ResNet8', '--device', 'cuda:1']\n",
    "args = parser.parse_args(config)\n",
    "\n",
    "device = args.device if torch.cuda.is_available() else 'cpu'\n",
    "save_dir = './checkpoint/' + args.dataset + '/'\n",
    "\n",
    "# models\n",
    "net1 = getattr(models, args.net1)(num_classes=args.n_class)\n",
    "net1.to(device)\n",
    "net2 = getattr(models, args.net2)(num_classes=args.n_class)\n",
    "net2.to(device)\n",
    "\n",
    "# logging\n",
    "logfile = save_dir + 'DML_2S_' + net1.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='DML Loss',\n",
    "        xlabel='epoch',\n",
    "        xtickmin=0,\n",
    "        ylabel='loss',\n",
    "        ytickmin=0,\n",
    "        ytickstep=0.5,\n",
    "        legend=['net1_loss', 'net2_loss']\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='DML Acc',\n",
    "        xlabel='epoch',\n",
    "        xtickmin=0,\n",
    "#         xtickstep=5,\n",
    "        ylabel='accuracy',\n",
    "        ytickmin=0,\n",
    "        ytickmax=100,\n",
    "#         markers=True,\n",
    "#         markersymbol='dot',\n",
    "#         markersize=5,\n",
    "        legend=['net1_acc', 'net2_acc']\n",
    "    ),\n",
    "    name=\"acc\"\n",
    ")\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 = datasets.CIFAR10(root='../data', train=True, download=True, transform=train_transform)\n",
    "test_set = datasets.CIFAR10(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",
    "\n",
    "# optimizer = optim.SGD(st_model.parameters(), lr=args.lr, momentum=args.momentum)\n",
    "optimizer_1 = optim.SGD(net1.parameters(), lr=args.lr, momentum=0.9, weight_decay=5e-4)\n",
    "optimizer_2 = optim.SGD(net2.parameters(), lr=args.lr, momentum=0.9, weight_decay=5e-4)\n",
    "lr_scheduler_1 = optim.lr_scheduler.MultiStepLR(optimizer_1, milestones=[100, 150])\n",
    "lr_scheduler_2 = optim.lr_scheduler.MultiStepLR(optimizer_2, milestones=[100, 150])\n",
    "\n",
    "def DML_loss(y, labels, logits_1, logits_2):\n",
    "    return nn.KLDivLoss()(logits_2, logits_1) + F.cross_entropy(y, labels)\n",
    "    \n",
    "# train with multi-teacher\n",
    "def train(epoch, net1, net2):\n",
    "    print('Training:')\n",
    "    # switch to train mode\n",
    "    net1.train()\n",
    "    net2.train()\n",
    "    batch_time = AverageMeter()\n",
    "    data_time = AverageMeter()\n",
    "    losses_1 = AverageMeter()\n",
    "    losses_2 = AverageMeter()\n",
    "    top1_1 = AverageMeter()\n",
    "    top1_2 = 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",
    "        _,_,_,_, output_1 = net1(input)\n",
    "        _,_,_,_, output_2 = net2(input)\n",
    "        logits_1 = F.softmax(output_1)\n",
    "        logits_2 = F.softmax(output_2)\n",
    "        \n",
    "        optimizer_1.zero_grad()\n",
    "        loss_1 = DML_loss(output_1, target, logits_1, logits_2)\n",
    "        loss_1.backward(retain_graph=True)\n",
    "        optimizer_1.step()\n",
    "        \n",
    "        optimizer_2.zero_grad()\n",
    "        loss_2 = DML_loss(output_2, target, logits_2, logits_1)\n",
    "        loss_2.backward(retain_graph=True)\n",
    "        optimizer_2.step()\n",
    "\n",
    "        output_1 = output_1.float()\n",
    "        loss_1 = loss_1.float()\n",
    "        # measure accuracy and record loss\n",
    "        net1_acc = accuracy(output_1, target)[0]\n",
    "        net2_acc = accuracy(output_2, target)[0]\n",
    "        losses_1.update(loss_1.item(), input.size(0))\n",
    "        top1_1.update(net1_acc, input.size(0))\n",
    "        losses_2.update(loss_2.item(), input.size(0))\n",
    "        top1_2.update(net2_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_1.avg:.3f})'.format(\n",
    "                      i, len(train_loader), batch_time=batch_time,\n",
    "                      data_time=data_time, loss=losses_1, top1_1=top1_1))\n",
    "    return losses_1.avg, losses_2.avg, net1_acc.cpu().numpy(), net2_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(' * {0} Prec@1 {top1.avg:.3f}'.format(model.model_name, top1=top1))\n",
    "\n",
    "    return losses.avg, test_acc.cpu().numpy(), top1.avg.cpu().numpy()\n",
    "\n",
    "\n",
    "print('*-----------------DML----------------*')\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_1.param_groups[0]['lr']))\n",
    "    lr_scheduler_1.step()\n",
    "    lr_scheduler_2.step()\n",
    "    train_loss_1, train_loss_2, net1_acc, net2_acc = train(epoch, net1, net2)\n",
    "    # visaulize loss\n",
    "    vis.line(np.column_stack((train_loss_1, train_loss_2)), np.column_stack((epoch, epoch)), loss_win, update=\"append\")\n",
    "    _, test_acc_1, top1_1 = test(net1)\n",
    "    _, test_acc_2, top1_2 = test(net2)\n",
    "    vis.line(np.column_stack((top1_1, top1_2)), np.column_stack((epoch, epoch)), acc_win, update=\"append\")\n",
    "    best_acc = max(top1_1, top1_2, best_acc)\n",
    "\n",
    "log_out(\"@ BEST Prec: {:.4f}\".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
}