eval_runtime.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "from pathlib import Path\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>total_cumulative_time</th>\n",
       "      <th>batch_time</th>\n",
       "      <th>model_scoring_time</th>\n",
       "      <th>post_processing_time</th>\n",
       "      <th>gpu_type</th>\n",
       "      <th>model_type</th>\n",
       "      <th>bs</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7.652452</td>\n",
       "      <td>5.411121</td>\n",
       "      <td>3.733840</td>\n",
       "      <td>1.552557</td>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>9.725491</td>\n",
       "      <td>1.765453</td>\n",
       "      <td>0.281142</td>\n",
       "      <td>1.451199</td>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>11.503588</td>\n",
       "      <td>1.756125</td>\n",
       "      <td>0.280393</td>\n",
       "      <td>1.442852</td>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>13.331980</td>\n",
       "      <td>1.804251</td>\n",
       "      <td>0.306628</td>\n",
       "      <td>1.460782</td>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>15.153791</td>\n",
       "      <td>1.795350</td>\n",
       "      <td>0.306558</td>\n",
       "      <td>1.452703</td>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>311</th>\n",
       "      <td>1757.041815</td>\n",
       "      <td>5.264950</td>\n",
       "      <td>1.209939</td>\n",
       "      <td>3.933433</td>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>312</th>\n",
       "      <td>1763.168690</td>\n",
       "      <td>4.889568</td>\n",
       "      <td>0.987383</td>\n",
       "      <td>3.802969</td>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>313</th>\n",
       "      <td>1769.536860</td>\n",
       "      <td>5.236986</td>\n",
       "      <td>1.217451</td>\n",
       "      <td>3.896962</td>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>314</th>\n",
       "      <td>1775.943139</td>\n",
       "      <td>5.286016</td>\n",
       "      <td>1.254733</td>\n",
       "      <td>3.908134</td>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>315</th>\n",
       "      <td>1780.676564</td>\n",
       "      <td>3.623870</td>\n",
       "      <td>0.793819</td>\n",
       "      <td>2.755013</td>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>63626 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     total_cumulative_time  batch_time  model_scoring_time  \\\n",
       "0                 7.652452    5.411121            3.733840   \n",
       "1                 9.725491    1.765453            0.281142   \n",
       "2                11.503588    1.756125            0.280393   \n",
       "3                13.331980    1.804251            0.306628   \n",
       "4                15.153791    1.795350            0.306558   \n",
       "..                     ...         ...                 ...   \n",
       "311            1757.041815    5.264950            1.209939   \n",
       "312            1763.168690    4.889568            0.987383   \n",
       "313            1769.536860    5.236986            1.217451   \n",
       "314            1775.943139    5.286016            1.254733   \n",
       "315            1780.676564    3.623870            0.793819   \n",
       "\n",
       "     post_processing_time gpu_type      model_type  bs  \n",
       "0                1.552557  rtx8000  efficientnetB3  10  \n",
       "1                1.451199  rtx8000  efficientnetB3  10  \n",
       "2                1.442852  rtx8000  efficientnetB3  10  \n",
       "3                1.460782  rtx8000  efficientnetB3  10  \n",
       "4                1.452703  rtx8000  efficientnetB3  10  \n",
       "..                    ...      ...             ...  ..  \n",
       "311              3.933433     v100  efficientnetB5  25  \n",
       "312              3.802969     v100  efficientnetB5  25  \n",
       "313              3.896962     v100  efficientnetB5  25  \n",
       "314              3.908134     v100  efficientnetB5  25  \n",
       "315              2.755013     v100  efficientnetB5  25  \n",
       "\n",
       "[63626 rows x 7 columns]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dfs = []\n",
    "\n",
    "for f in glob.glob(\"output/*eval_time.csv\"):\n",
    "    fname = Path(f).name\n",
    "    split = fname.split(\"_\")\n",
    "\n",
    "    # fname is of the form rtx8000_mdetr_efficientnetB3_10_eval_time.csv\n",
    "\n",
    "    gpu_type = split[0]\n",
    "    model_type = split[2]\n",
    "    bs = split[3]\n",
    "\n",
    "    df = pd.read_csv(f)\n",
    "\n",
    "    df[\"gpu_type\"] = gpu_type\n",
    "    df[\"model_type\"] = model_type\n",
    "    df[\"bs\"] = int(bs)\n",
    "    \n",
    "    dfs.append(df)\n",
    "\n",
    "combined = pd.concat(dfs).drop(\"Unnamed: 0\", axis=1)\n",
    "combined"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gpu_type</th>\n",
       "      <th>model_type</th>\n",
       "      <th>bs</th>\n",
       "      <th>total_cumulative_time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>1</td>\n",
       "      <td>5985.057900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>5</td>\n",
       "      <td>2315.462034</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "      <td>1879.812081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>25</td>\n",
       "      <td>1952.657409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>1</td>\n",
       "      <td>6097.589397</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>5</td>\n",
       "      <td>2333.660515</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>10</td>\n",
       "      <td>2024.211313</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "      <td>2155.363988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>1</td>\n",
       "      <td>5984.471876</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>5</td>\n",
       "      <td>2398.830361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>10</td>\n",
       "      <td>2002.745179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>25</td>\n",
       "      <td>2126.245727</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>50</td>\n",
       "      <td>1906.240563</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>1</td>\n",
       "      <td>8985.096552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>5</td>\n",
       "      <td>2254.231734</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "      <td>2611.136887</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>25</td>\n",
       "      <td>2455.410847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>1</td>\n",
       "      <td>6328.430410</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>5</td>\n",
       "      <td>2349.030134</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>10</td>\n",
       "      <td>2093.149725</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "      <td>1780.676564</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>1</td>\n",
       "      <td>5880.704913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>5</td>\n",
       "      <td>2401.964148</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>10</td>\n",
       "      <td>3110.835262</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>25</td>\n",
       "      <td>2598.700232</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gpu_type      model_type  bs  total_cumulative_time\n",
       "0   rtx8000  efficientnetB3   1            5985.057900\n",
       "1   rtx8000  efficientnetB3   5            2315.462034\n",
       "2   rtx8000  efficientnetB3  10            1879.812081\n",
       "3   rtx8000  efficientnetB3  25            1952.657409\n",
       "4   rtx8000  efficientnetB5   1            6097.589397\n",
       "5   rtx8000  efficientnetB5   5            2333.660515\n",
       "6   rtx8000  efficientnetB5  10            2024.211313\n",
       "7   rtx8000  efficientnetB5  25            2155.363988\n",
       "8   rtx8000       resnet101   1            5984.471876\n",
       "9   rtx8000       resnet101   5            2398.830361\n",
       "10  rtx8000       resnet101  10            2002.745179\n",
       "11  rtx8000       resnet101  25            2126.245727\n",
       "12  rtx8000       resnet101  50            1906.240563\n",
       "13     v100  efficientnetB3   1            8985.096552\n",
       "14     v100  efficientnetB3   5            2254.231734\n",
       "15     v100  efficientnetB3  10            2611.136887\n",
       "16     v100  efficientnetB3  25            2455.410847\n",
       "17     v100  efficientnetB5   1            6328.430410\n",
       "18     v100  efficientnetB5   5            2349.030134\n",
       "19     v100  efficientnetB5  10            2093.149725\n",
       "20     v100  efficientnetB5  25            1780.676564\n",
       "21     v100       resnet101   1            5880.704913\n",
       "22     v100       resnet101   5            2401.964148\n",
       "23     v100       resnet101  10            3110.835262\n",
       "24     v100       resnet101  25            2598.700232"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "runtimes = combined.groupby([\"gpu_type\", \"model_type\", \"bs\"])[\"total_cumulative_time\"].max().reset_index()\n",
    "runtimes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['rtx8000' 'v100'] ['efficientnetB3' 'efficientnetB5' 'resnet101']\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fa448dbbca0>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot runtime of each model type and GPU as a function of batch size\n",
    "\n",
    "gpu_types = runtimes[\"gpu_type\"].unique()\n",
    "model_types = runtimes[\"model_type\"].unique()\n",
    "\n",
    "line_formats = [\"-\", \"--\"]\n",
    "colors = [\"r\", \"g\", \"b\"]\n",
    "print(gpu_types, model_types)\n",
    "\n",
    "for i, gpu_type in enumerate(gpu_types):\n",
    "    for j, model_type in enumerate(model_types):\n",
    "        df = runtimes[(runtimes[\"gpu_type\"] == gpu_type) & (runtimes[\"model_type\"] == model_type)]\n",
    "\n",
    "        ln_format = colors[j] + line_formats[i]\n",
    "        plt.plot(df[\"bs\"], df[\"total_cumulative_time\"], ln_format, label=f\"GPU:{gpu_type}, Model: {model_type}\")\n",
    "\n",
    "plt.xlabel(\"Batch Size\")\n",
    "plt.ylabel(\"Runtime\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>gpu_type</th>\n",
       "      <th>model_type</th>\n",
       "      <th>bs</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.052380</td>\n",
       "      <td>0.024823</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>5</td>\n",
       "      <td>0.159283</td>\n",
       "      <td>0.097573</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "      <td>0.311714</td>\n",
       "      <td>0.134971</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>25</td>\n",
       "      <td>0.824511</td>\n",
       "      <td>0.190238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>1</td>\n",
       "      <td>0.076560</td>\n",
       "      <td>0.045154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.264559</td>\n",
       "      <td>0.052029</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>10</td>\n",
       "      <td>0.546904</td>\n",
       "      <td>0.138613</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "      <td>1.503938</td>\n",
       "      <td>0.367796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>1</td>\n",
       "      <td>0.071865</td>\n",
       "      <td>0.057398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>5</td>\n",
       "      <td>0.247193</td>\n",
       "      <td>0.084665</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>10</td>\n",
       "      <td>0.497581</td>\n",
       "      <td>0.139716</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>25</td>\n",
       "      <td>1.318515</td>\n",
       "      <td>0.294931</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>rtx8000</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>50</td>\n",
       "      <td>3.015491</td>\n",
       "      <td>0.760927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>1</td>\n",
       "      <td>0.060019</td>\n",
       "      <td>0.016176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>5</td>\n",
       "      <td>0.114102</td>\n",
       "      <td>0.039162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>10</td>\n",
       "      <td>0.248930</td>\n",
       "      <td>0.060393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB3</td>\n",
       "      <td>25</td>\n",
       "      <td>0.658770</td>\n",
       "      <td>0.175715</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>1</td>\n",
       "      <td>0.060090</td>\n",
       "      <td>0.042039</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>5</td>\n",
       "      <td>0.190710</td>\n",
       "      <td>0.094229</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>10</td>\n",
       "      <td>0.355324</td>\n",
       "      <td>0.152549</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>v100</td>\n",
       "      <td>efficientnetB5</td>\n",
       "      <td>25</td>\n",
       "      <td>1.021985</td>\n",
       "      <td>0.317367</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>1</td>\n",
       "      <td>0.058970</td>\n",
       "      <td>0.045136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>5</td>\n",
       "      <td>0.192418</td>\n",
       "      <td>0.080882</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>10</td>\n",
       "      <td>0.464997</td>\n",
       "      <td>0.093463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>v100</td>\n",
       "      <td>resnet101</td>\n",
       "      <td>25</td>\n",
       "      <td>1.235559</td>\n",
       "      <td>0.345480</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   gpu_type      model_type  bs      mean       std\n",
       "0   rtx8000  efficientnetB3   1  0.052380  0.024823\n",
       "1   rtx8000  efficientnetB3   5  0.159283  0.097573\n",
       "2   rtx8000  efficientnetB3  10  0.311714  0.134971\n",
       "3   rtx8000  efficientnetB3  25  0.824511  0.190238\n",
       "4   rtx8000  efficientnetB5   1  0.076560  0.045154\n",
       "5   rtx8000  efficientnetB5   5  0.264559  0.052029\n",
       "6   rtx8000  efficientnetB5  10  0.546904  0.138613\n",
       "7   rtx8000  efficientnetB5  25  1.503938  0.367796\n",
       "8   rtx8000       resnet101   1  0.071865  0.057398\n",
       "9   rtx8000       resnet101   5  0.247193  0.084665\n",
       "10  rtx8000       resnet101  10  0.497581  0.139716\n",
       "11  rtx8000       resnet101  25  1.318515  0.294931\n",
       "12  rtx8000       resnet101  50  3.015491  0.760927\n",
       "13     v100  efficientnetB3   1  0.060019  0.016176\n",
       "14     v100  efficientnetB3   5  0.114102  0.039162\n",
       "15     v100  efficientnetB3  10  0.248930  0.060393\n",
       "16     v100  efficientnetB3  25  0.658770  0.175715\n",
       "17     v100  efficientnetB5   1  0.060090  0.042039\n",
       "18     v100  efficientnetB5   5  0.190710  0.094229\n",
       "19     v100  efficientnetB5  10  0.355324  0.152549\n",
       "20     v100  efficientnetB5  25  1.021985  0.317367\n",
       "21     v100       resnet101   1  0.058970  0.045136\n",
       "22     v100       resnet101   5  0.192418  0.080882\n",
       "23     v100       resnet101  10  0.464997  0.093463\n",
       "24     v100       resnet101  25  1.235559  0.345480"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Plot throughput\n",
    "mn_throughputs = combined.groupby([\"gpu_type\", \"model_type\", \"bs\"])[\"model_scoring_time\"].agg([\"mean\", \"std\"]).reset_index()\n",
    "mn_throughputs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fa44a705c10>"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import random\n",
    "\n",
    "for i, gpu_type in enumerate(gpu_types):\n",
    "    for j, model_type in enumerate(model_types):\n",
    "        df = mn_throughputs[(mn_throughputs[\"gpu_type\"] == gpu_type) & (mn_throughputs[\"model_type\"] == model_type)]\n",
    "\n",
    "        ln_format = colors[j] + line_formats[i]\n",
    "        plt.errorbar(df[\"bs\"] + random.random(), df[\"mean\"], yerr=df[\"std\"], fmt=ln_format, label=f\"GPU:{gpu_type}, Model: {model_type}\")\n",
    "\n",
    "plt.xlabel(\"Batch Size\")\n",
    "plt.ylabel(\"Mean Inference Time\")\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fa44a2f6040>"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEGCAYAAABo25JHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAABYgklEQVR4nO2dd3xP1xvH30fsvWdsSpCB2FRswk9VaXWhqmp06LBKtVSV0tYqqlTaUpSqasUm9gpij1it2GITmc/vj/PNVxIZX5Fv5nm/XveV7/fec+997k1ynnvPeZ7Po0QEg8FgMGRcMqW0AQaDwWBIWYwjMBgMhgyOcQQGg8GQwTGOwGAwGDI4xhEYDAZDBidzShvwpBQuXFjKlSuX0mYYDAZDmmLv3r3XRaRIbNvSnCMoV64cvr6+KW2GwWAwpCmUUv/Gtc0MDRkMBkMGxzgCg8FgyOAYR2AwGAwZnDQ3RxAboaGhBAQE8PDhw5Q2xWBI82TPnh1HR0eyZMmS0qYYkol04QgCAgLIkycP5cqVQymV0uYYDGkWESEwMJCAgADKly+f0uYYkol0MTT08OFDChUqZJyAwfCUKKUoVKiQebvOYKQLRwAYJ2AwJBHmfynjkW4cwRPj4aEXg8FgyOBkXEeQxFy5coVXXnmFChUqULt2bRo0aMCff/4JgI+PD/ny5aNmzZo4OTkxatQoALy8vHjnnXeiHcfDw8OmhLlly5Zx9OjRBNutX7+eWrVq4ebmRuPGjTl16hSgx4Lfe+89KlWqhIuLC/v27bPus2rVKqpUqUKlSpUYN26cdf2NGzdo1aoVlStXplWrVty8eTPB8yuleP31163fw8LCKFKkCB06dEhw36iUK1eO69evP3WbuLh27Rr16tWjZs2abNmyhcWLF+Pk5ESzZs3w9fXlvffei3d/T09Pbt26lahz2/q7jItz587x22+/Wb9H/r25ubnh4uJCy5YtuXr1KgB//fUXLi4uuLm54e7uztatWxN9XkPyYs9nV+MIkgARoVOnTjz77LOcOXOGvXv3snDhQgICAqxtmjRpwv79+/H19WXevHns3bs30ecLCwuzufPo168f8+fPx8/Pj1deeYUxY8YAsHLlSvz9/fH392fWrFn069cPgPDwcAYMGMDKlSs5evQoCxYssJ5n3LhxtGjRAn9/f1q0aBHNScRFrly5OHz4MEFBQQCsXbuWUqVKJfbS7cb69eupWrUq+/fvp0mTJsyZM4fp06ezceNG3N3dmTJlSrz7e3t7kz9//kSdO6kdAei/Nz8/Pw4ePEidOnX4/vvvAWjRogUHDhzAz8+Pn376id69eyf6vIb0g3EEScCGDRvImjUrffv2ta4rW7Ys77777mNtc+XKRe3atTl9+vQTncPDw4NPPvmEpk2bMn78eJYvX86gQYNwc3PjxIkT1KlTBx8fHwCGDRvG8OHDAf1EfufOHQBu375NyZIlAf1k2L17d5RS1K9fn1u3bnHp0iV2795NpUqVqFChAlmzZqVbt2789ddf1n169OgBQI8ePVi2bJlNtrdr144VK1YAsGDBAl5++WXrths3btCpUydcXFyoX78+Bw8eBCAwMJDWrVtTs2ZN3n77baJW0ps3bx5169bFzc2Nt99+m/DwcJvv4+nTp2nbti21a9emSZMmHD9+HD8/PwYPHoy3tzdubm6MGjWKrVu30rdvXwYNGoSPj4/1DebevXu88cYbODs74+Liwh9//AFEfxuJy77cuXMzfPhwXF1dqV+/PleuXGH79u3RfpenT5/Gw8ODIUOGULduXZ555hm2bNkCaCc9aNAg6tSpg4uLCz/88AMAQ4cOZcuWLbi5ufHdd99Fu14R4e7duxQoUMBqQ+QcwP379818gEEjImlqqV27tsTk6NGjj768/75I06YJL/ny6cWWtu+//9g5ozJ58mQZOHBgnNs3btwo7du3FxGR69evS9myZeXw4cMyd+5cGTBgQLS2TZs2lT179oiIyJtvvmn93LRpU+nXr5+1XY8ePWTx4sXW74cPH5aqVavKmjVrxM3NTYKDg0VEZPPmzVKwYEEpVaqUODk5ye3bt0VEpH379rJlyxbr/s2bN5c9e/bI4sWL5c0337Su/+WXX6w25suXL5qt+fPnj/e+iIjkypVLDhw4IC+88IIEBQWJq6trtPvxzjvvyOeffy4iIuvXrxdXV1cREXn33Xdl1KhRIiLyzz//CCDXrl2To0ePSocOHSQkJERERPr16yc///yziIiULVtWrl27JiIi7dq1kwsXLjxmT/PmzeXkyZMiIrJz505p1qyZiMhjv4uov4eo9g4ePFjej/L3cOPGjWjnjs8+QJYvXy4iIoMGDZIvvvhCRB7/XTZt2lQ+/PBDERFZsWKFtGjRQkREfvjhB+s+Dx8+lNq1a8uZM2ei2Rdpb968ecXV1VUcHR2lSpUq1t+7iMjSpUulSpUqUqBAAdm+fXusv7do/1OGVEFkd5RYAF+Jo19NF3kEqY0BAwawdetWsmbNyp49ewDYsmULNWvWJFOmTAwdOpTq1avHORcQ+ZQ2e/bsaOtfeumlOM9ZvXp1Xn/9df73v/+xY8cOsmbNCsB3332Ht7c39erVY8KECXz44YfMnj072hN21PPGtf5pcHFx4dy5cyxYsABPT89o27Zu3Wp9qm7evDmBgYHcvn2bzZs3s3TpUgDat29vfaJdv349e/fupU6dOgAEBQVRtGjRx87p7e392Lp79+6xfft2unbtal0XHBz8RNeybt06Fi5caP0eaVck8dmXNWtW65tF7dq1Wbt2bZzn6dy5s7XduXPnAFizZg0HDx5kyZIlgH7D8/f3t/6uo9KkSRP++ecfAMaPH8/gwYOZOXMmAM8//zzPP/88mzdv5tNPP2XdunVPdA8M6Y/05wgmTbKtXeSsi2U45WmoXr26tTMD+P7777l+/Tru7u7WdVH/MSMpVKjQYxOuN27coHDhwrGeJ1euXPHacejQIfLnz8+VK1cAPQF64MAB6tWrB2hH0rZtWwAcHR05f/68dd+AgABKlixJSEhIrOsBihUrxqVLlyhRogSXLl2KtQOOi44dO/Lxxx/j4+NDYGCgdX18jic2ByQi9OjRg6+++srmc0cSERFB/vz58fPze+J9o54/PscYn31ZsmSx7uvg4EBYWFicx8mWLdtj7USEqVOn0qZNm2htfRL4G+7YsSMvvPDCY+ufffZZTp8+zfXr1+P8mzNkDMwcQRLQvHlzHj58yIwZM6zrHjx4kOB+derUYdu2bVy+fBkAX19fgoODKV26dIL75smTh7t371q/L126lMDAQDZv3sx7773HrVu3KFCgALdv3+bkyZOAnqh1cnICdOfwyy+/ICLs3LmTfPnyUaJECerUqYO/vz9nz54lJCSEhQsX0rFjR+s+P//8MwA///wzzz33HAAXLlygRYsW8drbq1cvRo4cibOzc7T1zz77LPPnzwd0h1a4cGHy5s0bbf3KlSutDrNFixYsWbLEGgVz48YN/v03TnXdaOTNm5fy5cuzePFiQHesBw4csGnfSFq3bs20adOs32M68sTYF/N3GRdt2rRhxowZhIaGAnDy5Enu37+f4P5bt26lYsWKAJw6dcrqfPft20dISAiFChVK8NyG9I1xBEmAUoply5axadMmypcvT926denRowfjx4+Pd79ixYoxefJkPD09cXNzY+DAgSxYsIBMmfSvpXfv3nEOH3Xr1o0JEyZQs2ZNdu3axdChQ5kzZw7PPPMM77zzDu+//z6ZM2fmxx9/5IUXXsDV1ZVff/2VCRMmADrcsUKFClSqVIm33nqL6dOnA5A5c2amTZtGmzZtcHJy4sUXX6R69eqAnpRcu3YtlStXZu3atQwdOhSAS5cukTlz/C+Xjo6OvP/++4+t//zzz/H19cXFxYWhQ4daHc1nn33G5s2bqVWrFmvWrKFMmTIAVKtWjTFjxtC6dWtcXFxo1aoVly5deuy4np6eXLx48bH18+fPZ86cObi6ulK9enXrRLitjBgxgps3b1KjRg1cXV3ZuHFjtO222heVqL/L+IIIevfuTbVq1ahVqxY1atTg7bffJiwsDBcXFzJnzoyrq6t1sjhy8jjy9/7NN98A8Mcff1CjRg3c3NwYMGAAixYtMhPGBlRsr+apGXd3d4nZOR47dsz6pGszSTg0lNGZNm0aZcqUsb45GNI+ifqfMtiVp+2ylFJ7RcQ9tm3pb47AVowDSDJiJsUZDIa0hRkaMhgMhgyOcQQGg8GQwbGbI1BKZVdK7VZKHVBKHVFKjYqljYdS6rZSys+yjLSXPQaDwWCIHXvOEQQDzUXknlIqC7BVKbVSRHbGaLdFRJ5MgcxgMBgMSYbdHIElpfme5WsWy5JqQpQ8vDwA8Onpk6J2GAwGQ0pj1zkCpZSDUsoPuAqsFZFdsTRrYBk+WqmUqh7HcfoopXyVUr7Xrl2zp8mJxshQx46RoU6Y5JShjrrNzc2N0aNHJ/q8hvSDXR2BiISLiBvgCNRVStWI0WQfUFZEXIGpwLI4jjNLRNxFxL1IkSL2NDlRiJGhjhMjQ50wySlDHXWbn58fI0eaaTlDMkUNicgtwAdoG2P9HRG5Z/nsDWRRSqU50RMjQx0/RoY69chQGwyxYbc5AqVUESBURG4ppXIALYHxMdoUB66IiCil6qIdU+DjR7OdgasG4nfZL8F2kW0i5wriw624G5PaTopz+5EjR6hVq5ZN9gUGBrJz504+/fRTqzJpXPTu3Zu+fftaxetu3brFpk2bAPD396dDhw506dIF0MNMXbp0YcqUKaxatYpdu/Qo3OzZs/H09CRHjhzkzZuXnTv1XP2FCxeiaRo5Ojpy4cKFWNdHHuvKlSuUKFECgBIlSliHGxKiW7dujB49mg4dOnDw4EF69epl7dw+++wzatasybJly9iwYQPdu3fHz8+PUaNG0bhxY0aOHMmKFSuYNWsWoDNeFy1axLZt28iSJQv9+/dn/vz5dO/ePdo5PT09mT17ttXxRdKnTx9mzpxJ5cqV2bVrF/3792fDhg2MHj0aX19fq47Qxo0bmThxIu7u7tFE3b744gvy5cvHoUOHgMe1huKz7/79+9SvX58vv/ySwYMH8+OPPzJixAg6duwY7XcJ+q1v9+7deHt7M2rUKNatW8ecOXPIly8fe/bsITg4mEaNGtG6dWvGjRvHxIkTraKGPj4+VscQGBhIrly5GDt2rPXYO3bswNXVlZIlSzJx4kSrhIgh42LPqKESwM9KKQd0B/+7iPyjlOoLICIzgS5AP6VUGBAEdJO0pnkRC0aGOjpGhjr1yFDXqlWLf//9l9y5c+Pt7U2nTp3w9/d/ontgSH/YM2roIFAzlvUzo3yeBkyL2eZpiO/JPSpJGTVkZKgTxshQpw4Z6rx581rXe3p60r9/fyNDbTCZxUmBkaE2MtSRpHYZ6suXL1ud7+7du4mIiDAy1AbjCJICI0NtZKgjSe0y1EuWLLHa/t5777Fw4UIjQ23IuDLUJqEs6TAy1OkPI0Od+jAy1HbAOICkw8hQGwxpGzM0ZDAYDBkc4wgMBoMhg2McgcFgMGRwjCMwGAyGDE6GdQQeHo9m4Q0GgyEjk2EdQVJjZKhjJ63IUNuLqBo/oBPrihYtSo0a0YV447q3gYGBNGvWjNy5c5voLIPdMI4gCTAy1HGTmmWo45N4SCpiOoKePXuyatWqx9rFdW+zZ8/OF198wcSJE+1uqyHjYhxBEmBkqOMnNclQR72PkydPZu/evTRt2pTatWvTpk0baxbwlClTqFatGi4uLnTr1g3QWdC9evXCw8ODChUqRKtREJtNQ4cOJSgoCDc3N1599VVAS2oULFjwMbviure5cuWicePGZM+e3eZrNBielHSXUDZwINiiKRbZxpZ5Ajc3mDQp7u1Ghjp+UpMMddT7GBoaStOmTfnrr78oUqQIixYtYvjw4fz000+MGzeOs2fPki1btmiVx44fP87GjRu5e/cuVapUoV+/fpw6dSpWm8aNG8e0adNsErlL7L01GJKCdOcIUgNGhjo6qUWGOpLI+3jixAkOHz5Mq1atAD0sFtkZu7i48Oqrr9KpUyc6depk3bd9+/Zky5aNbNmyUbRoUa5cuWKzTQZDaiXdOYL4ntyj8rS6HVExMtQJkxpkqCOJvI8iQvXq1dmxY8djbVasWMHmzZtZvnw5X3zxBUeOHAEeyUPDI4nopLDpae6twfC0mDmCJMDIUKcNGeqYVKlShWvXrlkdQWhoKEeOHCEiIoLz58/TrFkzvv76a27dusW9e/fiPE58NmXJksUqGx0fcd1bgyFZEJE0tdSuXVticvTo0cfWJUTTpnpJKi5evCgvvfSSlCtXTurUqSMeHh6ycOFCERHZuHGjtG/fPtb9li1bJjVr1hRXV1dp1KiR7N2717rtzTfflD179ljsbWr9LCKydetWcXJyEjc3N9m5c6dUrlxZ/vvvPxERmTx5snTv3l1ERJYuXSo1atQQFxcXadq0qZw+fVpERCIiIqR///5SoUIFqVGjRrRjr1ixQipXriwVKlSQMWPGWNdfv35dmjdvLpUqVZLmzZtLYGCgiIjs2bNHWrduHev15cqV67F1Ue9HYGCgdOzYUZydnaVevXpy4MAB67latWolNWvWlIEDB0qZMmXk2rVrIiKycOFCcXV1FWdnZ6lVq5bs2LFDRETKli1rbdOuXTu5cOHCY+eOeR/3798vTZo0ERcXF6lWrZrMmjVLQkJCpFGjRlKjRg2pXr26fPXVVyIi8tlnn8mECROs+1avXl3Onj0br02DBw+WqlWryiuvvCIiIt26dZPixYtL5syZpVSpUjJ79ux4723kdRUoUEBy5colpUqVkiNHjsR6r5OSxPxPGezL0/ZZgK/E0a/aTYZaKZUd2AxkQw9BLRGRz2K0UcBkwBN4APQUkX0xjxWVJJOh9tA/k2JoKKNjZKjTH0aGOvWRVmWog4HmInJPKZUF2KqUWikiO6O0aQdUtiz1gBmWn3bHOICkwyQ6GQz2JSgI/v0X8ue3z/HtNkdgeRuJHFjNYllivn48B/xiabsTyK+UKmEvmwwGgyEtIQKLF4OTE5w7Bzdu2Oc8dp0sVko5KKX8gKvAWhHZFaNJKeB8lO8BlnUxj9NHKeWrlPK9du2a3ew1GAyG1IKfHzRrBi++qN8EXF2hfHn7nMuujkBEwkXEDXAE6iqlasRoEluA+mOTFiIyS0TcRcS9SJEidrDUYDAYUgfXrkHfvlC7Nhw+DDNnwt699hsWgmQKHxWRW4AP0DbGpgAgaqykI/B4xXGDwWBI54SG6jyoypVhzhx47z3w94e33wYHB/ue226OQClVRCmV3/I5B9ASOB6j2XKgu9LUB26LyCV72RQNo0NtMBhSCatWgYsLfPAB1K8PBw/Cd9+BJaHe7tjzjaAEsFEpdRDYg54j+Ecp1VcpFanO5g2cAU4BPwL97WiPXUluGepIpk2bRqVKlVBKRZNglkTITMeFl5cXSinWr19vXffnn3+ilGLJkiU22+rj45Og/LQtbeJj8eLFODk50axZMwBefvllXFxc+O677xg5ciTr1q2Lc19fX1/ee++9RJ87ptLokxJTUbZnz56UL18eNzc3qlatav27AXjzzTdxdXXFxcWFLl26xJvwZki9nDwJHTpAu3YQHg7//AMrV+rJ4WQlrgSD1LokVUJZUmaURURESP369WXGjBnWdefOnZMpU6aISPQEqnv37kmlSpXE19dX5s6dKwMGDIhhVvSEp4TYt2+fnD17NloylYhOCmvbtq1ERETIjh07pG7duiIiEhYWJhUqVJDTp09LcHCwuLi4JJigNHfuXHF2dpY333zTuu7FF18UV1dXWbx4sc22xpdY9yRt4qNNmzayYcMGERG5dOmSlClTJtHHelJiS557Enr06BHtfkb9HhQUJOXLl5czZ86IiMjt27et7T744ANr0ltSYRLK7MutWyIffSSSJYtInjwiEyeKBAfHv489E8qMxEQSYG8Z6mPHjlG3bl3r93PnzuHi4gJAzZo1KVeu3GP7JEZmOj6aNGnC7t27CQ0N5d69e5w6dQo3Nzfr9vXr11OzZk2cnZ3p1asXwcHBgH77qFq1Ko0bN7aKyAHcv3+fXr16UadOHWrWrGmTDVGJTfZ59OjRbN26lb59+zJo0CBat27N1atXcXNzY8uWLfTs2dP6BrNnzx4aNmyIq6srdevW5e7du9HeRuKyz8vLi86dO9O2bVsqV67M4MGDAR6TnD537hxOTk689dZbVK9endatW1trMpw+fZq2bdtSu3ZtmjRpwvHjx9m+fXs0afGYfx8PHz4EHukk5c2bF9APckFBQU8tDGhIHsLD9fj/M8/At99C9+56HuCjj8CiE5ki2OQIlFJllVItLZ9zKKXy2NespyRy/D/qMn263vbggf7u56eXyO1eXnr79euP75sAiZGhrl69eoJte/fuja+vL05OToSEhHDmzBkAFi1axIsvvhjvvk8iM33hwoUEbVFK0bJlS1avXs1ff/0VLYv44cOH9OzZk0WLFnHo0CHCwsKYMWMGDx8+5K233uLvv/9my5YtVk0lgC+//JLmzZuzZ88eNm7cyKBBg7h//360c/r6+tK7d+/HbIkqRe3n54eDgwPz589n5MiRuLu7M3/+fCZMmMDy5cupWLEifn5+NGnSxLp/SEgIL730EpMnT+bAgQOsW7eOHDlyRDtHfPb5+flZr3XRokWcP3+ecePGkSNHDvz8/KwaSf7+/gwYMIAjR46QP39+qzBhnz59mDp1Knv37mXixIn079+fhg0b0rFjRyZMmICfnx8VK1YEsDoGR0dHunXrFk2M7o033qB48eIcP3481ocOQ+pi61aoWxd699YTwnv2wOzZUKxYSltmgyNQSr0FLAF+sKxyBJbZ0aY0z4ABA3B1dbXKEsMjGerWrVtbZajjeoqLKkMdqWD64osv8vvvvwPaEcQnSQ1xq3rGtd4WunXrxsKFC1m4cGG04jInTpygfPnyPPPMM4AurLJ582aOHz9O+fLlqVy5MkopXnvtNes+a9asYdy4cbi5ueHh4cHDhw/577//op3P3d39MSluiC5F7ebmxvr1661O0hZOnDhhFdgD/XSdOXP0JPv47GvRogX58uUje/bsVKtWLU7Ru8jxfYDatWtz7tw57t27x/bt2+natav1bSayGE5sRDqGy5cvs379erZv327dNnfuXC5evIiTkxOLFi2y+foNycv58/Dyy9CkCVy9CgsWwJYtOjw0tWCLxMQAoC6wC0BE/JVSqVsjNz79iJw59fa4hDsKF35i/YnkkKF+6aWX6Nq1K507d0YpReXKleO1KTEy0wlRt25dDh8+TI4cOaydPsTudCKJy8mICH/88QdVqlSJtj5SQjs+5Clln0UkQecXl327du2KVYo6NmK2CwoKIiIigvz589tUrCYquXPnxsPDg61bt9KwYcNox33ppZeYMGECb7zxxhMd02BfHjyACRNg/HidIfzZZzB4sO6CUhu2DA0Fi0hI5BelVGZiSfrKyCSHDHXFihVxcHDgiy++SPBtABInMz1t2jSmTZsW73G/+uqrx6Jjqlatyrlz5zh16hQAv/76K02bNqVq1aqcPXvWOt69YMEC6z5t2rRh6tSpVieyf//+BK8pkqeVoq5atSoXL160Fg26e/fuY515YuyzRXI6b968lC9fnsWLFwPa4Rw4cAB4XFo8KmFhYezatYuKFSsiItZ7LSL8/fffVK1aNUH7DMmDCPz+u478+fxz6NgRjh/Xn1OjEwDbHMEmpdQnQA6lVCtgMfC3fc1KWyilWLZsGZs2baJ8+fLUrVuXHj16MH78+Hj3K1asGJMnT8bT0xM3NzcGDhzIggULyJRJ/1oi5wgieemll5g3b160+YEpU6bg6OhIQEAALi4u1jF1T09PKlSoQKVKlXjrrbeYbpkjyZw5M9OmTaNNmzY4OTnx4osvWucrjh8/TqFCheK1uV27dtbQzEiyZ8/O3Llz6dq1K87OzmTKlIm+ffuSPXt2Zs2aRfv27WncuDFly5a17vPpp58SGhqKi4sLNWrU4NNPP33sXHHNEVSrVo0xY8bQunVrXFxcaNWqVbzDKzHJmjUrixYt4t1338XV1ZVWrVpZJ2OfxL6Y9OnTx1rZLD7mz5/PnDlzcHV1pXr16taJ6G7dujFhwgRq1qxpdZ6RcwQuLi44OzvTuXNn6xuRs7Mzzs7OXLp0iZEjR9p8/Qb7sX8/NG0KL70EBQvC5s2wcCFE+dNPlSQoQ62UygS8CbRGS0KsBmZLQjvaiaSSoTY61I/ToUMHli5dai1zaci4GBnqJ+PqVRgxQk/+FioEY8dCr15JmxGcojLUIhKBTvb6MXGnT6UYB/AYMecwDAZD/ISEwLRpMGqUnhP44AP49FP76gLZA1uihjoopfYrpW4ope4ope4qpe4kh3EGg8GQWvH2BmdnnQPQqJEWiPvmm7TnBMC2OYJJQA+gkIjkFZE8IpLXvmYZDAZD6uTECWjfXi8AK1ZopxAjwCxNYYsjOA8cTqk5AYPBYEgN3Lqln/5r1NDJYd98A4cOgadnSlv29NiSRzAY8FZKbUKXnwRARL61m1UGg8GQSggPh59+guHDtfBA794wZgwUTd3ZVE+ELY7gS+AekB1IN+EkHl4eAPj09ElROwwGQ+pl82Z4/32tRtOkCUyeDDVrprRVSY8tQ0MFRaSziHwmIqMiF7tblsYwMtQJY2So4+dJZKijbnNzc3viTGVD/Pz7r84FaNoUAgNh0SLYtCl9OgGw7Y1gnVKqtYissbs1aRQRoVOnTvTo0YPffvsNgH///Zfly5db20RKTNy/fx83N7en6uyi0qhRIzp06IBHDHG8lStX4u/vj7+/P7t27aJfv37s2rWL8PBwBgwYwNq1a3F0dKROnTp07NiRatWqxXseZ2dnFixYQIsWLQBYuHAhrq6uSXINScmcOXOYPn06zZo14/Lly2zfvt3mrGN3d/dosiBPytixY/nkk08Svf+yZcvo0KFDtN/FhAkT6NKlCw8fPqRatWp0796d8pbCtZHbDEnHgwdaEuLrr0EpHRb68cepNyM4qbDljWAAsEopFWTCR2PHyFAbGerklqE2JC0iOgO4ShUYPRo6ddKyECNHpn8nADY4Aku4aCYRyZFWwkc9vDweW6bv0RILD0If4OHlgd9lP/wu+1m3e/l5AXD9wfXH9k0II0NtZKhTQoZ6+PDhuLi48MEHH1gdr+HJ2btXj/+//DIUKaKVQRcsgDJlUtqy5CNOR6CUqmr5WSu2JaEDK6VKK6U2KqWOKaWOKKXej6WNh1LqtlLKz7KkC8EUI0NtZKjBvjLUX331FcePH2fPnj3cuHEjQV0rw+NcuaIjgOrU0cVhZs/WNQIaN05py5Kf+OYIPgT6AN/Esk2A5gkcOwz4SET2WQrZ7FVKrRWRozHabRGRpBkwtxBfJFDOLDnx6ekTZ9RQ4ZyFnziSyMhQGxnq2LCnDHWJEiWs53jjjTeYOHHiEx0vIxMSAlOm6CGghw91bsCIEZAvX0pblnLENzR0EEBEmsWyJOQEEJFLIrLP8vkucAwolSRWpzKMDLWRoYbkk6EGrG8RIsKyZcuoUaNGgvZldER0cfgaNWDQIB0RdPiwrhmQkZ0AxO8IeiXVSZRS5YCaWIrbxKCBUuqAUmqlUirWgXOlVB+llK9SyvfatWtJZVaSYWSojQw1JJ8MNcCrr75qlaG+fv06I0aMsPn6MyLHjkG7dvC//2lF0JUr4e+/de1gA8Ra0d7yFLQvrm1PsgC5gb1A51i25QVyWz57Av4JHa927doSk6NHjz62LiGazm0qTec2feL90jPt27eX4ODglDbDkApIzP9UauTGDZGBA0UyZxbJl0/ku+9EQkJS2qrE0bSpXhIL4Ctx9KvxzRG4xBEmqrT/SDhySCmVBfgDmC8iS2NuF5E7UT57K6WmK6UKi8j1mG2TGpNR/DhGhtqQXggP15O/I0bohLA+feCLL3RUkOFx4nMEh0Qk0Xl0Ss/GzQGOSRy6REqp4sAVERGlVF30UFVgYs9pMBgMPj4wcCAcOKDnASZNgigpL4ZYsCWzOLE0Al4HDiml/CzrPgHKAIjITKAL0E8pFQYEAd0srzAGg8HwRJw7pyeBlyzRpSEXL4YXXtAZwob4ic8RLH6aA4vIVvQwUnxtpgHxh6kYDAZDPNy/D+PGwcSJkCmTHgL66COIkSNoiIc4HYGIPJ2ClsFgMNgREZ0BPHgwXLgAr7yidYIcHVPasrSHLVpD6RIPj0fFoA0GQ9rC11dnAL/6KhQvrgvFzJ9vnEBiybCOIKkxMtQJkxwy1PbAz88Pb29v6/fjx4/ToEEDsmXL9lhGb1z3dvHixVSvXp1MmTI90e/XEJ3Ll6FXLy0Lcfq0Lhize7euGWxIPLYUry+mlJqjlFpp+V5NKfWm/U1LO4hFhvrZZ5/lzJkz7N27l4ULFxIQEGBt06RJE/bv34+vry/z5s1j7969SXLuRo0asW7dumjJWhBdhnrWrFn069cPwCpDvXLlSo4ePcqCBQuiaeDHRaQMdSSpRYZaRIiIiLDrOWI6goIFCzJlyhQ+/vjjaO3iu7c1atRg6dKlPPvss3a1Nb0SHKyloZ95BubN08NBJ0/CG2/oeQHD02HLLfQCVgORgjQngYF2sidNYmSok1eGOlLiuX///tSqVYvz588zYcIE6tSpg4uLC5999pn1HO3bt8fV1ZUaNWqwaNEiAMqVK8dnn31GrVq1cHZ25vjx43HaFBISwsiRI1m0aBFubm4sWrSIokWLUqdOHbJkyRLNrvjurZOT02O6RYaEEYHly6F6dRgyBJo1gyNH9FxA3lStgZy2sMURFBaR34EIABEJA8LtatVTEjn+H3WxKCzw4IH+7uenl8jtXl56+/Xrj++bEEaGOnllqEEriHbv3p39+/dz4sQJ/P392b17N35+fuzdu5fNmzezatUqSpYsyYEDBzh8+DBt27a17l+4cGH27dtHv379rMM7sdkUGhrK6NGjeemll/Dz84tX5ymx99YQO0ePQtu28NxzkDUrrF4Nf/0FCegtGhKBLY7gvlKqEFpxFKVUfeC2Xa1K4xgZavvKUIN+46pfv771WGvWrKFmzZrUqlWL48eP4+/vj7OzM+vWrWPIkCFs2bKFfFGUxSI1eyLloW21KT6e5t4aHnHzpq4T7OKix/8nT9bJYa1bp7Rl6RdbEso+BJYDFZVS24Ai6ESwVIuPT9zbcubU2yOf9GO2LVw4/v1jw8hQJ68MNUSv1CUiDBs2jLfffvuxdnv37sXb25thw4bRunVrRo7UJS8iJaKjykjHZdOuXbFpJT5OXPfcYBthYfDjj/Dpp9oZvP22loqO5d8hQ/Kk/dKTYEuFsn1AU6Ah8DZQXUQO2s+ktIeRoU5eGeqYtGnThp9++ol79+4Beojm6tWrXLx4kZw5c/Laa6/x8ccfR4ucius4sdkUnzx0VOK7t4b42bgRatWC/v3B2Rn279fDucYJJA+2RA0NQCuEHhGRw0BupVR/+5uWdjAy1MkrQx2T1q1b88orr9CgQQOcnZ3p0qULd+/e5dChQ9a6xl9++WWCUs1x2dSsWTOOHj1qnSy+fPkyjo6OfPvtt4wZMwZHR0fu3LkT7739888/cXR0ZMeOHbRv3542bdokeF0ZgbNntQxE8+Zw9y788Qds2KCHhQzJh0pI2kcp5ScibjHW7X8aQbqnwd3dXWLGYR87dgwnJ6cnOk5cQ0MZmQ4dOrB06VKyZs2a0qYYUpjE/E89CffuwVdfwTff6PoAn3wCH35oZCHsiVJqr4i4x7bNljmCTEopFSkGp5RyANJ8T2EcwOMYGWqDvYmIgN9+06GgFy/Ca69pnaBS6bJ2YdrBFkewGvhdKTUTHTnUF1hlV6sMBkO6Y/duHQ20c6fODF6yBBo0SGmrDGCbIxiCniTuh1YTXQPEHtOXgogNBckNBkPCJLUS/KVLMGwY/Pyz1gXy8oLXXzcZwamJBB2BiEQAMyxLqiR79uwEBgZSqFAh4wwMhqdARAgMDCR79uxPfayHD3VRmC+/hJAQGDpUzwXkyfP0dhqSlgQdgVKqEfA5UNbSPrJUZQX7mmY7kVEzqbGwvcGQ1siePTuOTyHjKaIzgD/6CM6c0ZnB33wDFSsmoZGGJMWWoaE5wAfoAvSpUloiS5YslC9fPqXNMBgyPIcP6zKR69dDtWqwZg20apXSVhkSwpZRutsislJEropIYOSS0E5KqdJKqY1KqWNKqSNKqfdjaaOUUlOUUqeUUgeVUrYJ9hgMhlTFjRvw7ru6NvC+fTB1qpaFME4gbWDLG8FGpdQEYCkQHLnSknEcH2HARyKyTymVB9irlForIlE1j9sBlS1LPfQ8RL0nuQCDwZByhIXBDz/AyJFw6xb06wejRkECeYmGVIYtjiCyY46aiCBA8/h2EpFLwCXL57tKqWNAKSCqI3gO+MWSo7BTKZVfKVXCsq/BYEjFrF+vw0GPHNGZwZMmaXkIQ9rDlqihZgm1SQilVDmgJhBTvasUcD7K9wDLOuMIDIZUyunT8PHHsGwZlC8Pf/6pJ4RNwF7axe4VypRSuYE/gIEicifm5lh2eSyIWSnVRynlq5TyNZFBBkPKcPeuDv+sVg3WroWxY3XNgE6djBNI69i1QplSKgvaCcwXkaWxNAkAokptOgIXYzYSkVki4i4i7kWKFLHl1AaDIYmIiIBffoEqVbQ+ULduukzksGGQBOkGhlSA3SqUKZ3ZNQc4JiLfxtFsOdDdEj1UHx2hZIaFDIZUwq5dWgaiRw8oXVrLQ/z8M5gyC+kLWyaLE1uhrBHwOnBIKeVnWfcJUAZARGYC3oAncAp4ALzxJMYbDAb7cPGizgT+9VcoUUK/Ebz6qpGFSK/YrUKZiGwl9jmAqG0EGGCDDQaDIRl4+BC+/VaP/4eG6jmBYcMgd+6UtsxgT+J1BBbJ6aaWpQq6Yz8hIqHJYJvBYEgmRHQU0Ecf6WIxnTvDhAlQIdUIyRjsSbwveiISDjwnImGRFcqMEzAY4sbDywMPL4+UNuOJOHQIWrbUnX+uXLBuna4UZpxAxsGWoaFtSqlpwCLgfuRKGzKLDQZDKiYwUGcEz5wJ+fPD999Dnz6Q2ZZewZCusOVX3tDyc3SUdQlmFhsMhtRJaKju/D/7DO7cgQED4PPPoWDBlLbMkFIkS2axwWBIHaxdq9VBjx7Vw0GTJkH16iltlSGlsXtmscFgSHlOndIyEK1bQ3CwrhewZo1xAgaNXTOLDQZDynL3rs4HqF4dNmzQheKPHIGOHY0shOERdsssNhgMKUdEhK4N/MwzMH48vPKKloUYMgSyZUtp6wypDXtmFhsMhhRgxw547z3w9YX69WH5cqhTJ6WtMqRm7JZZbDAYkpeAAD0MNH++1gKaN0+/CZghIENCxOkIlFJdRWQxcBOTWWwwpFqCgnRx+K++gvBwGDFCDwEZWQiDrcT3RjAMWAz8ISK1gCPJY5LBYLAFEVi6VBeJOXcOXnhBy0KUL5/SlhnSGvE5gkCl1EagvFJqecyNItLRfmYZDIb4OHBA5wP4+ICLi44IamYyfgyJJD5H0B6oBfwKfJM85hgMhvi4fh0+/RRmzYICBWDGDOjd28hCGJ6OOP98RCQEXVC+oYiY+pAGQwoSGgrTp2spiLt34d13tUREgQIpbZkhPWDLc0QBpdSXQLmo7UXEaA0ZDICHh/7p42Of469erYeBjh/XmcHffafrBhsMSYUtjmAxMBOYjUkkMxiSDX9/+PBD+OcfqFRJ5wN06GDCQQ1Jjy2OIExEZtjdEoMhDeLhAX5+4OaWdMe8cwfGjNGCcNmzw9df6wQxkxFssBe2SEz8rZTqr5QqoZQqGLkktJNS6iel1FWl1OE4tnsopW4rpfwsy8gntt5gSEdERMBPP0HlyjBxIrz+upaFGDTIOAGDfbHljaCH5eegKOsESKh+kRcwDfglnjZbRKSDDTYYDGmCsIgwwiLCnni/bdvg/fdh715o2BBWrAB3dzsYaDDEgi31CBKVniIim5VS5RKzr8GQ1ohQwUzaOYNdF3aRO6vtKb3nz+ss4AULoFQp+O036NbNzAMYkpf4JCY6x7ejiCxNgvM3UEodAC4CH4tIrNnLSqk+QB+AMmXKJMFpDYakQYggpMpC9tQbwZbVZ8mfPT8V8idc7DcoSGcBjxunM4Q//VQ7hFy5ksFogyEG8b0R/C+ebQI8rSPYB5QVkXtKKU9gGVA51pOJzAJmAbi7u8tTntdgSBLWn1nPPvfBBOXZR667rqx+bTVjt4yNdx8RWLJEy0L89x907aong8uVSx6bDYbYiC+h7A17nlhE7kT57K2Umq6UKiwi1+15XoPhaTlw+QBD1g1h9enVZMtShhyrfqF2tldpXTFTvI7Az0/PA2zeDK6u8Msv0LRp8tltMMSFLVFDdkEpVVwpPRKqlKprsSUwpewxGBLiv9v/0WNZD2r+UJPdF3YzsdVE6u46QdZjr6Pi+Ve6dg3efhtq1dK1gn/4QU8KGydgSC3YTaFEKbUA8AAKK6UCgM+ALAAiMhNd06CfUioMCAK6iYgZ9jGkOm4G3WTslrFM3T0VgEENBzG08VAK5CjA3xFx7xcSAt9/D6NGwf37Ojt45EjInz9ZzDYYbMZujkBEXk5g+zR0eKnBkCp5GPaQqbumMnbrWG4/vE131+6MbjaaMvkSDlhYuRI++ABOnIC2bbUsRNWqyWC0wZAIbHIESqmGPK41FF9+gMGQZgmPCGf+ofmM2DCC83fO065SO8a1HIdLMZcE931w2ZHTCwbgeVAnhv3zD3h6mnBQQ+omQUeglPoVqAj48UhrSIg/UcxgSHOICKtPr2bIuiEcvHKQ2iVq49XJi+blE9ZXDAvTkUC+k7zIlPUhEydqhdCsWZPBcIPhKbHljcAdqGbG7w3pmb0X9zJ43WA2nN1A+fzlWfDCAl6s/iKZVOyTwB5eHgBERPgQEgK7d8P27VCsySrKd57DR+//mYzWGwxPhy2O4DBQHLhkZ1sMhmTn7M2zDN8wnAWHF1A4Z2Emt51MX/e+ZHWI/1E+/GEOLm1qT8BuCA6GvHlh40b48ODEZLLcYEg6bHEEhYGjSqndQHDkSlOq0pCWuf7gOmM2j2H6nulkzpSZTxp/wuBGg8mXPV+8+127BlOnws5vFxF2Py/58oGDg1YfrVULOJgs5hsMSYotjuBzexthMCQXD0IfMGnnJMZvG8+9kHv0cuvF5x6fUypvqXj3O3cOvvkG5szR8hCFah2gjOcC8m6Zjp+fmQw2pG1sEZ3blByGGAz2JDwiHC8/L0b6jOTi3Yv875n/Ma7lOKoVib/U18GDWgJi4ULIlAlee03LQvfb9alusCUZjDcY7IwtUUP1gamAE5AVcADui0heO9tmMDw1IsIK/xUMWTeEo9eOUq9UPRa+sJAmZZvEsw9s2QLjx4O3N+TOraUhPvgAHB0tjXZB6L08+B/ViWMGQ1rGlqGhaUA3dMlKd6A7cYjDGQypiV0Buxi8bjCb/91M5YKVWdJ1CZ2dOqPiGMeJiIC//9aKoDt3QpEiulJY//6PisSHhuokMYDMOe8TFKQdh8GQlrFJa0hETgEOIhIuInPR0hEGQ6rEP9Cfrou7Un9OfY5fP850z+kc6X+EF6q9EKsTCAkBLy+oUQM6dYLLl7U0xL//wvDh2gkEBsJXX0H58tCiBUSEZkVliqBWLVM9zJD2seWN4IFSKivgp5T6Gh1GalTTDamOK/euMHrTaGbtm0U2h2x81vQzPmrwEXmy5Ym1/d278OOPWv4hIEArgv72m5aGzmz5zzh9Ws8R/PqrniRu1UprBo2/HAqYSWJD+sAWR/A6+s3hHeADoDTwgj2NMhiehHsh9/h2x7dM2D6BoNAg+tTuw8imIymeu3is7a9dgylT9FP/zZu6AP2PP0KbNrpjj4jQInG5cmkH8csvun7we+/ptwaAr73MeJAh/WBL1NC/SqkcQAkRGZUMNhkMNhEaHsqc/XP43Odzrty/QmenzoxtPpYqhavE2v7s2UchoMHBehhoyBCoV09vv39fd/qTJ0Pr1tpZPPusdgaFCiXfdRkMyY0tUUP/AyaiI4bKK6XcgNEmocyQUogIy44vY+j6oZwMPEnjMo3586U/aVC6QaztDx7UEUCLFukQ0Ndf1yGgkWqg58/rt4NZs/Qbgru7dgCg3xCexAn49PR5uoszGFIAWxPK6gI+ACLiZ4rSG1KKbf9tY/C6wWw/v52qhavyV7e/+N8z/3tsElhEVwIbP15LQufOrcf2P/hAF4mPyqhRMHcudO6s2zRsaMb+DRkLW6KGwkTktt0tMRji4fj143Ra2InGcxtz9uZZZnWYxaF+h+hYpWM0JxARAcuW6c7cwwN8feHLL3V94IkToWhR/WbQoIEWigNdLObMGVi8GBo1Mk7AkPGwSXROKfUK4KCUqgy8B2y3r1kGg+bS3Ut87vM5c/bPIWeWnIxpNoaB9QeSK2v0wLWQEJg/X0f4HD+uwzynT4eePSFHDrhxQ78dTJumx/wrV4Zbt/S+ZRKuM2MwpGtscQTvAsPRgnMLgNXAFwntpJT6CegAXBWRGrFsV8BkwBN4APQUkX22m25Iz9wJvsOEbRP4due3hIaH0r9Ofz599lOK5CoSrd3du3ps/7vv4MIFHQK6YAF06fIoBDQ8HFxc9PaWLWHmTGjXTs8XGAwG26KGHqAdwfAnPLYXOis5rgI27dAZypWBesAMy09DBiYkPIQffH/gi81fcO3BNV6q/hJfNv+SigUrRmt39eqjENBbt/Qw0Jw5OtoHYM0a+OMP3ek7OMCkSVClCjg7P72NJ66f4MzNMxTLVcwk1BjSBXE6AqXU8vh2TChqSEQ2JzCp/Bzwi6XgzU6lVH6lVAkRMXUPMiAiwuKji/lk/Secvnkaj3IefN3ya+qUqhOt3ZkzOgT0p590COjzz+sQ0Lp14cED/XYweTIcOwbFi8OwYXqYqEuXp7PvTvAdFh1exFy/uewI2AFA9kvXyeXnh5ubGz4+T3d8gyElie+NoAFwHj0ctAtI6im0UpbjRxJgWfeYI1BK9QH6AJQxA7rpDp9zPgxeO5g9F/dQo2gNVryygnaV2kWbBD5w4FEIqIMDdO+uQ0CrWFIGDh6EZs30XECtWjoT+MUXk65UZJO5TTh45SBOhZ34uuXX/Hn8T7IeOpo0BzcYUpj4HEFxoBXwMvAKsAJYICJHkujcsTmWWNM1RWQWMAvA3d3dpHSmEw5fPczQdUNZ4b8Cx7yOzH1uLq+7vI5DJgdAh4Bu2qQdwKpVOgT0ww91iGepUrBrF/z1Fzz3HDg56beDHj2gceOni/w5e/MsPx/4mX9O/sO2XtvIljkb41qMo2COgtQtVRelFCv8VyTNTTAYUgFxOgIRCQdWAauUUtnQDsFHKTVaRKYmwbkD0HIVkTgCF5PguIZUTsCdAEZuHMnPB34mT9Y8jG85nnfrvkuOLDkAHQL611/aAezapUM+v/wS+vWDPHn02P+kSVoh1MkJOnaELFlg9uzE2/Qg9AF/HP2DuX5z2XhuIwpFywotuXr/KqXzlaZd5XZJc/EGQyok3sliiwNoj3YC5YApwNIkOvdy4B2l1EL0JPFtMz+Qvrn18Bbjto5j8q7JREgEA+sN5JMmn1Aop07dDQ5+FAJ64gRUqBA9BHTpUv02cP48VKyoJ4t79kz807+IEBQWRM4sOfG77Ef3Zd2pUKACXzT7gu6u3SmTzwxDGjIG8U0W/wzUAFYCo0Tk8JMcWCm1AC1XXVgpFQB8BmQBEJGZgDc6dPQUOnz0jUTYb0gDBIcFM33PdMZsGcONoBu86vwqY5qPoVz+coAOAf3hBx0CevGirv+7cCG88IJW/7x1SzuCPHl0/P/334Onp54rSAwX717klwO/4OXnRfPyzZnefjoNHBuwrdc26jvWJ5MycaWGjEV8bwSvA/eBZ4D3okzcKUASqlAmIi8nsF2AAbabakhrREgECw4tYMTGEZy7dY5WFVoxvuV4apaoCcCVK/qpfvp03dk3a6alHlq2hHXr9JDPypXw0Uc6K7hVK70klr9P/M0M3xmsPr2aCImgSZkmNC3bFAClFA1LN0yCqzYY0h7xzRGYxyJDoll3Zh2D1w5m/+X9uBV3Y81ra2hVUffiZ87ojn3uXD0c1LmzDgGtU0evc3aGo0ehWDEYPRrefjtxNogIh64ewrmos3WC99DVQwxrPIyebj2pVLDSU1+nj9tATOyoIa1jS2axwWAzfpf9GLJuCGtOr6FsvrLMe34eLzu/TCaVCT8/PQH8++96WKdHD/j4Y63uWbiw3n/jRl3x65dfdPhnYqp/Xbt/jfmH5jPXby4Hrxxke6/tNCjdgPEtx/O95/fWqCSDwaAxjsCQJPx7619GbBzB/IPzyZ89P9+0/oYBdQaQ1SEbPj7aAaxercf5P/pIT/peuKCf+H//Hfbs0XMDM2ZAzpyJmwC+ePci73i/w98n/yYsIow6Jesw3XM6TkWcAMiXPV/SXbAI5QLDtZrd7duQLwmPbTAkM8YRGJ6KG0E3GLtlLFN3TyWTysTgRoMZ2ngoebPmZ9ky7QB279YhoGPHwltv6af+rl1h+3btGN5999EbQa4n1Gw4eu0ol+5eokWFFhTMUZBj14/xfr336enWkxpFH5O4ejoePIANG8Dbm4W/76J4YDBwD5o00RltBkMaxTgCQ6IICg1i6u6pfLX1K24/vE0Ptx6M9hhN0eylmTdPh4CePKlDQGfM0JnAOXPCnTvw5pvaMUyerMM/88YbdvA4tx7eYsGhBXgd8GL3hd04FXbi6ICjZM+cnaP9j8ZaoD7RnD0LK1boZeNGPamRKxennsnNr26ZGLQ7MxQsmHTnMxhSAOMIDE9EeEQ4vx78lZEbR3L+znk8K3syrsU4yuZwtqqAXrwINWvqENAaNXRUUKtWsHWr7vR37NDVwRIT/vnN9m8YvmE4weHBOBd15tvW3/Kqy6vW7U/tBEJCtKHe3rrzP35cr69cGfr2hfbt4dlnGbGgDfj5Mcj36U5nMKQGjCMw2ISIsOrUKoasG8Khq4eoU7IOvzz/C045PKwqoLdvQ/PmOvJHKf3Ev2KF1vt55RU9spIrF1Svbvt5T904hZefF2/Veouy+cviVMSJ3rV684bbG9QqUStpnv4vXdJxqt7eWrb07l1tdNOmuvP39NSOwGBIpxhHYEgQ34u+DF47mI3nNlKhQAUWdVlE7exd+eZrxU8/6YfoqCGgf/yh1T6LFoXPP9d9abFitp/vXsg9Fh9ZzFy/uWz5bwuZVCaqFKrC6/lfx7OyJ56VPZ/ugsLDdemyyCGffZYyGKVKQbdu+qm/RQstbmQwZACMIzDEyZmbZxi+YTgLDy+kcM7CTGk7hXoOffn2iyy8vFgXfuneXS+rV+v+tE4d6NAh8eGfD0IfUPq70tx6eIvKBSsztvlYurt2p1TeUgnvHB83b2ojvb310//167oyTYMGWsiofXtdvcbUqTRkQIwjMDzGtfvXGLN5DDN8Z5A5U2aGNx5BvdChTBuZi/fW6Eifjz/WxWDmz9fDQeHh0L+/3j9bNnj9ddvO9d/t//jZ72dO3zyNVycvazlKt+JuNCzdMPFDPyJw+LB+4vf21iFK4eE6aaFtW93xt26tvxsMGRzjCAxWHoQ+4Lsd3zF+23juh97nDZfe1Lv3FT8OK8iXe/Twzldf6aGeL77QQ+d58sA77+gQ0AoVbDtPUGgQy44vY67fXNadWYcgNCvXjOCwYLJlzsaAuolUHrl/X4d3Rnb+5y3lLmrWhKFDdedft27iRYqi4NPTR3tC/J76WAZDSmMcQTrGw8sDsHRa8RAWEYaXnxcjN47k0r1LdKjwAnWvT2XekBLMOamVPr/9VkdOvvYa5M8P//sflC4NvXrZFv4pIkRIBA6ZHJizfw7vrnyXsvnK8umzn9LDrQcVCtjoRWJy+vSjjt/HRxuZO7cOU/rsM12cuGTJxB07LsLCHhVEDgtL2mMbDCmAcQQZGBHh75N/M3TdUI5dP4Z7oeZ0CtzGsiHl+eeSrvQ1aRL4+8Onn+oH7nz5dF0ADw/LA3ECXL53mV8P/IrXAS8+avARvWr24hXnV3Aq7ESz8s2eXOkzJAS2bHk00XvypF5fpYoem2rfXlemSYw2RVwEBMC2bTqsdNs2vS5ygllMnSRD2sc4gnRGZOeckA7azoCdDF47mC3/baFClgZ0vnCc9d89g+9tRYsWOgR05kwtBZE1K7z8Mrz/vh5lSQgRsQ79ePt7Ey7hNHBsQNFcRQEomKMgLSq0sP2iLl7UT/ze3rB2Ldy7pzt6Dw8YMECPUVV6egE5QFfFOX5cV7xRSo97ff+93pYzp55cbtJEOwAfH9u8ocGQyjGOIINxMvAkn6z/hD+O/UGhoLo0OXWY3f9U42yIolMnePZZ3fkD/POPHl3p21cXgk+IgDsBOOZ1RCnFmC1juHT3Eh83/Jiebj2pWriq7UaGh2tdisghn/379XpHR3j1Vf3U37z5k+tRxEZQkBY6inza375da2KfOwdly+oxsMqVoVEjcHXVpdAMhnSGcQQZhCv3rjBq0yhm7Z1Fliv1qH70EMc2V2dXZkXXrnqcf/Fi+PNP/YD9zDMw1YaCpIEPAvnt0G/M9ZvLsevHuPzRZfJlz8eyl5ZRIk8JMmey8U/sxg0d3rlihS5QHBioJ3UbNtQz1O3b6zTlpw3vvHZNd/i1akGZMrBsmc52A6hWTYsgNW6sJ0IA2rTRi8GQjjGOIJ0THhHOKJ9RfL1tAg/9G1Ly4CEC9jvxXx6t83/tmlb/DAvThWAGDrQtifbotaN87vM5f534i5DwEGoWr8nXLb+2SjyXzlc6/gOIaKG2yKf+HTv0sEzhwtoTRYZ3FijwdDfg3j19gZFj/JFzClOm6FCnVq3g77/1kI8JJTVkUOzqCJRSbYHJgAMwW0TGxdjuAfwFnLWsWioio+1pU3rHz0//vBdyj/N3zvPfzQtsXVWMAvv8eHC6EiFFdSTlkCG6L65WTc+xvvuujg6KjxPXTyCIdZhn47mN9HPvxxtub+Ba3DVh4+7dg/XrH3X+Fy7o9bVrw/DhuvN3d098eGdIiB5G2rpVhzS9+KJ2Lm+9pZ/wGzXSineNGulzgnY8HTok7nwGQzrBbo5AKeUAfA+0AgKAPUqp5SJyNEbTLSJi/hOTCMl2i2C3qZSZOJ2bOzuQaccQuF6JfOWgRRc9HL5mjZaEVkrL6cc37H0n+A6LDi9irt9cdgTsoFuNbix4YQHVilTj4ocXyeKQwJi5v/8jAbdNm3RnnSePftpv314nd5Uo8XQXPXasvqhdu+DhQ73u9de1I8ibV9tQrpzOJDYYDI9hzzeCusApETkDoJRaCDwHxHQEhiTg2v1rfLfzO+68+iscehkmHIZbhche8gzPttUPyUuWRJ8MhvidwKA1g/h+z/cEhQXhVNiJr1t+zeuuj1KGY3UCwcGwefOjp35/f73eyUm/drRvr5/Is2Z9sgsU0V4rcojn1i347Te9bdMmrWjXt68e32/YMLpzsTXTzWDIoNjTEZQCzkf5HgDUi6VdA6XUAeAi8LGIHInZQCnVB+gDUKZMGTuYmnYJuBPAxO0TmbJiLbJzAPj5Q1h2GjWH67U+Iuh6MdbPH0y3bjr8M3JEJDbO3TrHwsML+bjhx2TOlJmCOQrS3bU7b7i9Qd1SdeOWewgI0Po9K1boqvP37+vwzubN4b339Jj/k3bG4eGPhogmT9ZFjgMC9Pc8eXQIZ0SEfspfudI87RsMT4E9HUFsvUbM7Jt9QFkRuaeU8gSWAY9NVYrILGAWgLu7u8ngQQvCfbXpa+Yuvkr4zv5wZhI4hOBAViSTnvj9M99e8oVmZdPYwXGOvjwIfcAfR/9grt9cNp7biELhUc6D+o71GdZkWOw7hYfDzp2PhnwOHNDry5TRCnSentoJ5Mxp+wXdv6+HdiLDOHfs0PH8JUs+Gt9v3Fgvzs7R5xFS0gmYwvWGdIA9HUEAEDV0xBH91G9FRO5E+eytlJqulCosItftaFea5ui1o3y+ajJL5uVF9gyBm+UpWjyMUOed3D9fiZBbhcmUSc+VcgcyZQmJ0wkcv36cuj/W5W7IXSoUqMAXzb6gu2t3yuSL5a3r+vVH4Z2rV+twTwcH3TGPH6+HfKpVsz288/JlyJ5dd/J//w3PP68djFI6TPS11x7JN/TooReDwWAX7OkI9gCVlVLlgQtAN+CVqA2UUsWBKyIiSqm6QCYg0I42pVn2XdrH4Pm/sH5hNTj4LYTmol7DYD4aCF5emfH2rk8+xy1kvuNO5tw56NwZpnhFP8aFOxf49eCvZHPIxgcNPqBywcq84fYGnZ0606Rsk+hyDyI6BCnyqX/XLj0UU7SoTrLy9NQTvpHx9vEhop/uI5/2t27VGkEzZ+oYVjc3HcrUqJEO47TlmAaDIcmwmyMQkTCl1DvAanT46E8ickQp1deyfSbQBeinlAoDgoBuIka8JSqbz25j4OQN7P+rEZybROasoXh2iCBHVpg0KRvFi+sE2Eu1+pH37gL8pgVEG5SLkAh+P/I7c/3msub0GiIkgi7VuvABH+CQyYHJ7SY/anz3rh7jj5zovXRJr69TR4sNtW+vJxkSGooJDtaFXyL1/m/e1G8LAEWK6A6/Xz9o1kyvK10axoxJuptmMBieCJXW+l13d3fx9U3fhWJFhD/3bWbQ18c5s7oN3C5H/mK3ebFzNq5czM7ff2vxywULdGUwsCiN+vlpR5A7N7du6fV9/u7Dj/t+xDGvIz1ce9DTrSeVClaKPJFOsIp86t+8GUJDdchlmzb6qb9dO9vKi61apcfLt27VMaohIXr/FSv09sWLtURD5cqm+IvBkAIopfaKiHts20xmcSpCRJiyfBNfTrzNtZ2tIawplWqdZ+THD5nnlY9ZM3Qf/fHHOgQ05th/iIoguO73hNT6jQOXf8G1uCvv1H2HF5xeoGWFljrr9+HDR5W6VqzQQzSgn9gHDtRP/Q0bxh1XKgJnz+oO/8IFGGaZUP7ySz18VLu2jhRq1EgfJ5KuXZP6dhkMhiTCOIJUQHBIOEOn7eDHGdm5f8oDlSWIxh3O0aNDJXq/qefbd2/XZXTffltLQUdyP+Q+Px/4mUNXD3Ez312kzVDy3KnDvZB7ALgUc8ElpAD8OFt3/OvX65j77Nn1AT/8UD+5lysXv5H//AM//6wdwOXLel3RojBokH49mTdPf8+Rww53yGAw2BMzNJSCXLoSyrtjDvDX/JKE3SxJlkIBPPfKVeqWceWHmQ6cPg3HjkHVKMKd90Pu43POB0Ho8EwHHoY9pOD4gghC4XsRrDrgQvXlO3T4ZeRT/6FDeudy5fQTf/v2Wj45tk777l0dGrp1q14WLNAd/IQJWo45MoSzUSOoXt3E7xsMaYT4hoaMI7ADCdUE2L77IR+OPsuu1eUhLDu5q+6idy8oGFyHaVMzcfWqrqg4ZAg89xycvHEMb39vVp1exeZ/NxMSHkIDxwZsf3M7ABfvXuS9H57Hfc0hhvrl1mGXN2/qJ/XGjR91/lWrPj4+H5mUtWOH1vY/cODROldX8PLSRd2jJngZDIY0h5kjSGE8vDyICHPgzRwr+Pzr65w76AhZSlOssTcjPypMv/ZNuH5dUbYsNG0KAz64R6ZyW/B8ph0An2z4hGXHl1G9SHXerfsubSq2oUnOqnoCdv16Sq5fz5JTp/TJ1EPo2VMP97RqFX0cKSJCF3SPGsY5cqSuN1mokFb6HDFCO4/69XUGbyTGCRgM6RbzRmAHor4RXL0K7m/P4uLGjoTfLg4FTlG53Vo+6FqX3X/V4swZxYaNEey/vI/Fu7aw7eYf7AzYSbiEE/BBAKXyluLYtWPkDhFK+53VY/wbNjzK5s2TB5o2ZVqeY+x7eJaf1ueG27f1tqAgnfhVqpQe8ilTBms4UfHiusPv00c7DIPBkK4xbwTJRGSxeD8/H8LCoEPXQFYuy0tEWB+ouIp6fXbTq/4LrPi5H/2fhxw5hDffhLm+83lrZXcUitolazOs8TDalmlGMd9jsGEGThs26Ipd4eFaw6dRIx2l07y5lm3OnJklXh7kW3tWDwsNGqSf9vfu1W8Gy5Zph9Gvnx4eatwYypc3YZwGgwEwjiBJCbmTn8t/VOVOyDUkqAgrlmfFwX02hZrMo2zFh7SPmMfbLzjhkOsWNJ3MqOFlGdSqJ1futWZ+1l9odbcIRbbsg0UbYNtEHerp4KATuoYO1R1/w4Y64iciAk6c0BPCHTsC8NXaCB0RNGWKnmT48ENo2fKRgWPHpsyNMRgMqRrjCJ6SkBAdmDN55i12rFsMEVmg5G5UnVmMHlaQwhe6M3nvBQ5eHc++4Fooz7eo/7+TtK/+LM85FILJkym2fj2vbNoEdyzSSy4u+um9eXOtG503r15/6BB8990jUbYbN7RTsAwFzXLPxKTNObQuUPbsKXRHDAZDWsM4gkQgogth/TgnhF/nR3D/dnbI/QAaTSePy2wenG9CxM2qfPnaCzwMzEXHjmN49aVsVI8oTPPSIeTbeBVGTtJ1IgEqVYJu3XRcv4eHDte8fFkXUh81SlfvKlhQi7MNH66Hd55/Xg8RNWpkTf7yK6l0pJBxAgaD4Qkwk8VPwOXLMH8+zJwTxKljOSDzQ6jyF2Wa+lCxzkk2/edDxLpRsHUYiAOVnM8zqfUpPAN/RW3cAP/+qw9UooTu9Fu00E/9kTUWjh6FceP0E/+ZM3pdtmx6grhRI13QHeKurRsp1hY5IWwwGAwWzGTxUxAcrB/EZ/8Uxto1mYjIchvcZ0Dvf8h2pyrPhn7JwT9epGXbBZy/dJSSah85lDe9+YUuh5bAIXRYZrNmMHiw7vgdHbUez7ZtOlX4zTehSxc9Gbx6te70+/fXP2vVelTNK6Hi6sYBGAyGRGAcQSyIaPFMLy+Y91sYd25lhqr/wLvvw8O88M8PsG0IweLAnpwPaFdkO13ensUnly4DfxLESg4W8IBPJuiOv2pVXaTl/n099LN/v+70QWfnBgfrzzVq6NcOE81jMBiSEeMIonDxopbMmbUggNOshTJbwLElqmBxnBudJrS40PZaOVbdz8sLmcfTPnQ5dR7swSGsBDRpxFSHihy/c4ZQ76lUKxRMvX3bYOpUHcGzeDHkygUVK2od/0jt/QIFHhlgHIDBYEgBMqQjiJrw9eyPrbm+rxGO/w5hbfgIqLAaCnSCkx3ArweQiYION3h9zko+5jxkusC3lf2hXnkoVAmyVdcx+pMm8YeXB1OHnsZZ3oJTwIOSusP39Hx08oULk/16DQaDIT4yjCOITPby6enDvv1CWN6TNPpwNdu3d4Bd73Ei+x14by45eED2w20oejeCl/kcT4e11C5whkzvvwuNNsD06bBkiY7hB8idW9fQtfB7jUw438mmJ37LljVP+QaDIdWTYaKGPLw8uBl0k5rZnPh9w0WCzrcC/3ZwSU+iZ8n1H1/WqMA7p3KTOfA+WQjToZiOjrqC1rp1etJ2504dvVO6tI72yZcvemdvIncMBkMqJMWihpRSbYHJ6FKVs0VkXIztyrLdE3gA9BSRffayJ/TqJZYGH6TcwWEcP/QJtRx20DLrp9TOt47O4/vi0OGSTs6K7OiLFXtcbK1+/fhPYhyAwWBIY9jNESilHIDvgVZAALBHKbVcRI5GadYOqGxZ6gEzLD/tQtn85TkwJoxx1x3YlGMX627VxeO3EewkG1179tCNihTRCV4Gg8GQQbDnG0Fd4JSInAFQSi0EngOiOoLngF8sBet3KqXyK6VKiMglexgUlC8HWa7d4Jv8lhVZ9ZyBwWAwZGTs6QhKAeejfA/g8af92NqUApLcEUTt8M3ojcFgMDzCnnUGYwuXiTkzbUsblFJ9lFK+Sinfa5H6PAaDwWBIEuzpCAKA0lG+OwIXE9EGEZklIu4i4l6kSJEkN9RgMBgyMvZ0BHuAykqp8kqprEA3YHmMNsuB7kpTH7htr/kBg8FgMMSO3eYIRCRMKfUOsBodPvqTiBxRSvW1bJ8JeKNDR0+hw0ffsJc9BoPBYIgdu+YRiIg3urOPum5mlM8CDLCnDQaDwWCIH3sODRkMBoMhDWAcgcFgMGRwjCMwGAyGDI5xBAaDwZDBSXPqo0qpa8C/CTQrDFxPBnNSExnxmsFcd0YiI14zJN11lxWRWBOx0pwjsAWllG9ccqvplYx4zWCuO6XtSE4y4jVD8ly3GRoyGAyGDI5xBAaDwZDBSa+OYFZKG5ACZMRrBnPdGYmMeM2QDNedLucIDAaDwWA76fWNwGAwGAw2YhyBwWAwZHDSlSNQSrVVSp1QSp1SSg1NaXvshVLqJ6XUVaXU4SjrCiql1iql/C0/C6SkjUmNUqq0UmqjUuqYUuqIUup9y/r0ft3ZlVK7lVIHLNc9yrI+XV836LrnSqn9Sql/LN8zwjWfU0odUkr5KaV8Levsft3pxhEopRyA74F2QDXgZaVUtZS1ym54AW1jrBsKrBeRysB6y/f0RBjwkYg4AfWBAZbfb3q/7mCguYi4Am5AW0vtjvR+3QDvA8eifM8I1wzQTETcouQO2P26040jAOoCp0TkjIiEAAuB51LYJrsgIpuBGzFWPwf8bPn8M9ApOW2yNyJySUT2WT7fRXcQpUj/1y0ics/yNYtlEdL5dSulHIH2wOwoq9P1NceD3a87PTmCUsD5KN8DLOsyCsUiq7tZfhZNYXvshlKqHFAT2EUGuG7LEIkfcBVYKyIZ4bonAYOBiCjr0vs1g3bya5RSe5VSfSzr7H7ddi1Mk8yoWNaZ2Nh0hlIqN/AHMFBE7igV2689fSEi4YCbUio/8KdSqkYKm2RXlFIdgKsislcp5ZHC5iQ3jUTkolKqKLBWKXU8OU6ant4IAoDSUb47AhdTyJaU4IpSqgSA5efVFLYnyVFKZUE7gfkistSyOt1fdyQicgvwQc8PpefrbgR0VEqdQw/xNldKzSN9XzMAInLR8vMq8Cd6yNvu152eHMEeoLJSqrxSKivQDViewjYlJ8uBHpbPPYC/UtCWJEfpR/85wDER+TbKpvR+3UUsbwIopXIALYHjpOPrFpFhIuIoIuXQ/8cbROQ10vE1Ayilciml8kR+BloDh0mG605XmcVKKU/02KID8JOIfJmyFtkHpdQCwAMtT3sF+AxYBvwOlAH+A7qKSMwJ5TSLUqoxsAU4xKNx40/Q8wTp+bpd0BOEDugHt99FZLRSqhDp+LojsQwNfSwiHdL7NSulKqDfAkAP2/8mIl8mx3WnK0dgMBgMhicnPQ0NGQwGgyERGEdgMBgMGRzjCAwGgyGDYxyBwWAwZHCMIzAYDIYMjnEEhgyBUircouh4QCm1TynVMIH2+ZVS/W04ro9SKt7C4kqpTEqpKUqpwxZlyT1KqfKWbd6ReQIGQ0qRniQmDIb4CBIRNwClVBvgK6BpPO3zA/2B6Ulw7peAkoCLiERYBNXuA4iIZxIc32B4KswbgSEjkhe4CVq7SCm13vKWcEgpFalYOw6oaHmLmGBpO9jS5oBSalyU43W11Aw4qZRqEsv5SgCXRCQCQEQCRCTy/OeUUoWVUn0t5/JTSp1VSm20bG+tlNphsW+xRWvJYEhSTEKZIUOglApHZyVnR3fMzS2iZpmBnBYBu8LATqAyUBb4R0RqWPZvB3wKtBSRB0qpgiJyQynlA+wVkY8sme0fikjLGOd2BLYCt9B68vNEZL9l2znAXUSuW75nATYAXwM7gKVAOxG5r5QaAmQTkdF2uk2GDIoZGjJkFKIODTUAfrGoeCpgrFLqWbR0RSmgWCz7twTmisgDgBgp/pECeHuBcjF3FJEApVQVoLllWa+U6ioi62M5z2S0ts7fFhXOasA2i8pqVrRzMBiSFOMIDBkOEdlhefovAnhaftYWkVDLE3r2WHZTxC1rHmz5GU4c/1MiEgysBFYqpa6gi4tEcwRKqZ7oN5F3opxzrYi8bNOFGQyJxMwRGDIcSqmqaBG3QCAfWvs+VCnVDN0RA9wF8kTZbQ3QSymV03KMgk9wvlpKqZKWz5kAF+DfGG1qAx8Dr0XOJaCHqRoppSpZ2uRUSj3zRBdrMNiAeSMwZBRyWKp8gX7S7iEi4Uqp+cDfShcK90NLPCMigUqpbUqpw8BKERmklHIDfJVSIYA3Wv3UFooCPyqlslm+7wamxWjzDlAQ2GgZBvIVkd6Wt4QFUfYdAZx8gus2GBLETBYbDAZDBscMDRkMBkMGxzgCg8FgyOAYR2AwGAwZHOMIDAaDIYNjHIHBYDBkcIwjMBgMhgyOcQQGg8GQwfk/bk5LtMgkOgcAAAAASUVORK5CYII=",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot throughput\n",
    "post_processing_times = combined.groupby([\"gpu_type\", \"model_type\", \"bs\"])[\"post_processing_time\"].agg([\"mean\", \"std\"]).reset_index()\n",
    "post_processing_times\n",
    "\n",
    "for i, gpu_type in enumerate(gpu_types):\n",
    "    for j, model_type in enumerate(model_types):\n",
    "        df = mn_throughputs[(post_processing_times[\"gpu_type\"] == gpu_type) & (post_processing_times[\"model_type\"] == model_type)]\n",
    "\n",
    "        ln_format = colors[j] + line_formats[i]\n",
    "        plt.errorbar(df[\"bs\"] + random.random(), df[\"mean\"], yerr=df[\"std\"], fmt=ln_format, label=f\"GPU:{gpu_type}, Model: {model_type}\")\n",
    "\n",
    "plt.xlabel(\"Batch Size\")\n",
    "plt.ylabel(\"Mean Inference Time\")\n",
    "plt.legend()"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "478565d84f56a7791dc0111d293f8eab97e93db4509df281991a2b24d01be195"
  },
  "kernelspec": {
   "display_name": "Python 3.8.13 ('dls')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.13"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}