_4_time.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import re\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training time"
   ]
  },
  {
   "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>model</th>\n",
       "      <th>n</th>\n",
       "      <th>time</th>\n",
       "      <th>params</th>\n",
       "      <th>location</th>\n",
       "      <th>ver</th>\n",
       "      <th>minutes</th>\n",
       "      <th>relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov8m</td>\n",
       "      <td>16</td>\n",
       "      <td>259.018047</td>\n",
       "      <td>25.9</td>\n",
       "      <td>home</td>\n",
       "      <td>v8</td>\n",
       "      <td>4.316967</td>\n",
       "      <td>3.400165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>yolov9e</td>\n",
       "      <td>128</td>\n",
       "      <td>974.646500</td>\n",
       "      <td>57.3</td>\n",
       "      <td>home</td>\n",
       "      <td>v9</td>\n",
       "      <td>16.244108</td>\n",
       "      <td>12.794315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>yolov9c</td>\n",
       "      <td>16</td>\n",
       "      <td>409.868521</td>\n",
       "      <td>25.3</td>\n",
       "      <td>home</td>\n",
       "      <td>v9</td>\n",
       "      <td>6.831142</td>\n",
       "      <td>5.380399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>yolov8x</td>\n",
       "      <td>128</td>\n",
       "      <td>1005.424817</td>\n",
       "      <td>68.2</td>\n",
       "      <td>home</td>\n",
       "      <td>v8</td>\n",
       "      <td>16.757080</td>\n",
       "      <td>13.198346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>yolov9e</td>\n",
       "      <td>32</td>\n",
       "      <td>1105.462378</td>\n",
       "      <td>57.3</td>\n",
       "      <td>home</td>\n",
       "      <td>v9</td>\n",
       "      <td>18.424373</td>\n",
       "      <td>14.511553</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     model    n         time  params location ver    minutes   relative\n",
       "0  yolov8m   16   259.018047    25.9     home  v8   4.316967   3.400165\n",
       "1  yolov9e  128   974.646500    57.3     home  v9  16.244108  12.794315\n",
       "2  yolov9c   16   409.868521    25.3     home  v9   6.831142   5.380399\n",
       "3  yolov8x  128  1005.424817    68.2     home  v8  16.757080  13.198346\n",
       "4  yolov9e   32  1105.462378    57.3     home  v9  18.424373  14.511553"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv(\"out/b0313/time_table.csv\")\n",
    "data['ver'] = data['model'].apply(lambda x: \"v8\" if 'v8' in x else \"v9\")\n",
    "data[\"minutes\"] = data[\"time\"] / 60\n",
    "baseline = 1.2696348349253337\n",
    "data[\"relative\"] = data[\"minutes\"] / baseline\n",
    "data.head(5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.2696348349253337"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.query(\"model == 'yolov8n' and n == 16\").minutes.min()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x400 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_style(\"whitegrid\")\n",
    "g = sns.FacetGrid(\n",
    "    data.query(\"n in [32, 128, 500]\"),\n",
    "    col=\"n\",\n",
    "    col_order=[32, 128, 500],\n",
    "    #row=\"location\",\n",
    "    margin_titles=True,\n",
    "    sharey=\"all\",)\n",
    "g.map_dataframe(\n",
    "    sns.lineplot,\n",
    "    x=\"params\",\n",
    "    y=\"relative\",\n",
    "    hue=\"ver\",\n",
    "    style=\"ver\",\n",
    "    hue_order=[\"v8\", \"v9\"],\n",
    "    style_order=[\"v8\", \"v9\"],\n",
    "    err_style=\"band\",\n",
    "    errorbar=(\"se\", 2),\n",
    "    markers=True,\n",
    "    palette=[\"Grey\", \"#FF1F5B\"],\n",
    ")\n",
    "g.set(\n",
    "    yticks=np.arange(1, 22.5, 2.5),\n",
    "    ylabel=\"Rative training time (ratio)\",\n",
    "    xlabel=\"Number of Model Parameters (M)\",\n",
    ")\n",
    "g.add_legend(title=\"Models\")\n",
    "# draw baseline at 1\n",
    "for ax in g.axes[0]:\n",
    "    ax.axhline(y=1, color='black', linewidth=1, linestyle='--', alpha=0.5)\n",
    "g.figure.set_size_inches(10, 4)\n",
    "g.figure.savefig(os.path.join(\"out\", \"b0313\", \"fig5_training.png\"), dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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>model</th>\n",
       "      <th>n</th>\n",
       "      <th>time</th>\n",
       "      <th>params</th>\n",
       "      <th>location</th>\n",
       "      <th>ver</th>\n",
       "      <th>minutes</th>\n",
       "      <th>relative</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>yolov8m</td>\n",
       "      <td>16</td>\n",
       "      <td>259.018047</td>\n",
       "      <td>25.9</td>\n",
       "      <td>home</td>\n",
       "      <td>v8</td>\n",
       "      <td>4.316967</td>\n",
       "      <td>3.400165</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>yolov9e</td>\n",
       "      <td>128</td>\n",
       "      <td>974.646500</td>\n",
       "      <td>57.3</td>\n",
       "      <td>home</td>\n",
       "      <td>v9</td>\n",
       "      <td>16.244108</td>\n",
       "      <td>12.794315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>yolov9c</td>\n",
       "      <td>16</td>\n",
       "      <td>409.868521</td>\n",
       "      <td>25.3</td>\n",
       "      <td>home</td>\n",
       "      <td>v9</td>\n",
       "      <td>6.831142</td>\n",
       "      <td>5.380399</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>yolov8x</td>\n",
       "      <td>128</td>\n",
       "      <td>1005.424817</td>\n",
       "      <td>68.2</td>\n",
       "      <td>home</td>\n",
       "      <td>v8</td>\n",
       "      <td>16.757080</td>\n",
       "      <td>13.198346</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>yolov9e</td>\n",
       "      <td>32</td>\n",
       "      <td>1105.462378</td>\n",
       "      <td>57.3</td>\n",
       "      <td>home</td>\n",
       "      <td>v9</td>\n",
       "      <td>18.424373</td>\n",
       "      <td>14.511553</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",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12203</th>\n",
       "      <td>yolov8m</td>\n",
       "      <td>500</td>\n",
       "      <td>447.543804</td>\n",
       "      <td>25.9</td>\n",
       "      <td>projects</td>\n",
       "      <td>v8</td>\n",
       "      <td>7.459063</td>\n",
       "      <td>5.874968</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12204</th>\n",
       "      <td>yolov9e</td>\n",
       "      <td>16</td>\n",
       "      <td>432.131641</td>\n",
       "      <td>57.3</td>\n",
       "      <td>projects</td>\n",
       "      <td>v9</td>\n",
       "      <td>7.202194</td>\n",
       "      <td>5.672650</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12205</th>\n",
       "      <td>yolov8m</td>\n",
       "      <td>64</td>\n",
       "      <td>200.006936</td>\n",
       "      <td>25.9</td>\n",
       "      <td>projects</td>\n",
       "      <td>v8</td>\n",
       "      <td>3.333449</td>\n",
       "      <td>2.625518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12206</th>\n",
       "      <td>yolov8x</td>\n",
       "      <td>32</td>\n",
       "      <td>342.508847</td>\n",
       "      <td>68.2</td>\n",
       "      <td>projects</td>\n",
       "      <td>v8</td>\n",
       "      <td>5.708481</td>\n",
       "      <td>4.496160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12207</th>\n",
       "      <td>yolov8x</td>\n",
       "      <td>16</td>\n",
       "      <td>293.974946</td>\n",
       "      <td>68.2</td>\n",
       "      <td>projects</td>\n",
       "      <td>v8</td>\n",
       "      <td>4.899582</td>\n",
       "      <td>3.859049</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>12208 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "         model    n         time  params  location ver    minutes   relative\n",
       "0      yolov8m   16   259.018047    25.9      home  v8   4.316967   3.400165\n",
       "1      yolov9e  128   974.646500    57.3      home  v9  16.244108  12.794315\n",
       "2      yolov9c   16   409.868521    25.3      home  v9   6.831142   5.380399\n",
       "3      yolov8x  128  1005.424817    68.2      home  v8  16.757080  13.198346\n",
       "4      yolov9e   32  1105.462378    57.3      home  v9  18.424373  14.511553\n",
       "...        ...  ...          ...     ...       ...  ..        ...        ...\n",
       "12203  yolov8m  500   447.543804    25.9  projects  v8   7.459063   5.874968\n",
       "12204  yolov9e   16   432.131641    57.3  projects  v9   7.202194   5.672650\n",
       "12205  yolov8m   64   200.006936    25.9  projects  v8   3.333449   2.625518\n",
       "12206  yolov8x   32   342.508847    68.2  projects  v8   5.708481   4.496160\n",
       "12207  yolov8x   16   293.974946    68.2  projects  v8   4.899582   3.859049\n",
       "\n",
       "[12208 rows x 8 columns]"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Inference time"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "FILE_OUT = os.path.join(\"out\", \"b0313\", \"fig5_inference.png\")\n",
    "\n",
    "\n",
    "def find_init(lines):\n",
    "    s = []\n",
    "    for line in lines:\n",
    "        s += [re.search(r\"Ultralytics YOLOv8\", line)]\n",
    "    return np.where(s)[0]\n",
    "\n",
    "\n",
    "def find_model(s):\n",
    "    find = re.findall(r\"YOLOv\\d.\", s)[0]\n",
    "    return str(find)\n",
    "\n",
    "\n",
    "def find_device(s):\n",
    "    return \"CPU\" if \"CPU\" in s else \"GPU\"\n",
    "\n",
    "\n",
    "def find_ms(s):\n",
    "    ms_str = re.findall(r\"\\d+\\.\\d+ms\", s)[0]\n",
    "    return float(ms_str[:-2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 800x400 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "lines = open(\"data/mock/log.txt\").readlines()\n",
    "\n",
    "idx_init = find_init(lines)\n",
    "idx_model = idx_init + 1\n",
    "idx_iter_st = idx_init + 3\n",
    "idx_iter_ed = idx_init + 3 + 64\n",
    "\n",
    "\n",
    "data = pd.DataFrame(data=None)\n",
    "for i in range(len(idx_init)):\n",
    "    str_model = lines[idx_model[i]]\n",
    "    str_device = lines[idx_init[i]]\n",
    "    idx_st = idx_iter_st[i]\n",
    "    idx_ed = idx_iter_ed[i]\n",
    "    model = find_model(str_model)\n",
    "    device = find_device(str_device)\n",
    "    ms = []\n",
    "    for j in range(idx_st, idx_ed):\n",
    "        ms += [find_ms(lines[j])]\n",
    "    data_tmp = pd.DataFrame(data=dict({\"model\": model, \"device\": device, \"ms\": ms}))\n",
    "    data = pd.concat([data, data_tmp])\n",
    "\n",
    "# turn model to string\n",
    "data[\"fps\"] = 1000 / data[\"ms\"]\n",
    "data[\"yolo\"] = data[\"model\"].apply(lambda x: \"v8\" if \"8\" in x else \"v9\")\n",
    "\n",
    "df_param = pd.DataFrame(\n",
    "    data={\n",
    "        \"model\": [\"YOLOv8n\", \"YOLOv9c\", \"YOLOv8m\", \"YOLOv9e\", \"YOLOv8x\"],\n",
    "        \"size\": [3.2, 25.3, 25.9, 57.3, 68.2],\n",
    "    }\n",
    ")\n",
    "data = pd.merge(data, df_param, how=\"left\")\n",
    "\n",
    "\n",
    "sns.set_style(\"whitegrid\")\n",
    "g = sns.FacetGrid(\n",
    "    data,\n",
    "    col=\"device\",\n",
    "    col_order=[\"CPU\", \"GPU\"],\n",
    "    col_wrap=2,\n",
    "    margin_titles=True,\n",
    "    sharey=False,\n",
    ")\n",
    "g.map_dataframe(\n",
    "    sns.lineplot,\n",
    "    x=\"size\",\n",
    "    y=\"fps\",\n",
    "    hue=\"yolo\",\n",
    "    style=\"yolo\",\n",
    "    hue_order=[\"v8\", \"v9\"],\n",
    "    style_order=[\"v8\", \"v9\"],\n",
    "    err_style=\"band\",\n",
    "    errorbar=(\"se\", 2),\n",
    "    markers=True,\n",
    "    palette=[\"Grey\", \"#FF1F5B\"],\n",
    ")\n",
    "g.set(\n",
    "    ylabel=\"FPS\",\n",
    "    xlabel=\"Number of Model Parameters (M)\",\n",
    ")\n",
    "\n",
    "g.figure.subplots_adjust(right=1.2)\n",
    "g.add_legend()\n",
    "g.figure.set_size_inches(8, 4)\n",
    "g.figure.savefig(FILE_OUT, dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "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>model</th>\n",
       "      <th>device</th>\n",
       "      <th>ms</th>\n",
       "      <th>fps</th>\n",
       "      <th>yolo</th>\n",
       "      <th>size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>YOLOv8n</td>\n",
       "      <td>CPU</td>\n",
       "      <td>62.9</td>\n",
       "      <td>15.898251</td>\n",
       "      <td>v8</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>YOLOv8n</td>\n",
       "      <td>CPU</td>\n",
       "      <td>56.1</td>\n",
       "      <td>17.825312</td>\n",
       "      <td>v8</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>YOLOv8n</td>\n",
       "      <td>CPU</td>\n",
       "      <td>55.2</td>\n",
       "      <td>18.115942</td>\n",
       "      <td>v8</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>YOLOv8n</td>\n",
       "      <td>CPU</td>\n",
       "      <td>51.3</td>\n",
       "      <td>19.493177</td>\n",
       "      <td>v8</td>\n",
       "      <td>3.2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>YOLOv8n</td>\n",
       "      <td>CPU</td>\n",
       "      <td>49.1</td>\n",
       "      <td>20.366599</td>\n",
       "      <td>v8</td>\n",
       "      <td>3.2</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",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>635</th>\n",
       "      <td>YOLOv9e</td>\n",
       "      <td>GPU</td>\n",
       "      <td>52.9</td>\n",
       "      <td>18.903592</td>\n",
       "      <td>v9</td>\n",
       "      <td>57.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>636</th>\n",
       "      <td>YOLOv9e</td>\n",
       "      <td>GPU</td>\n",
       "      <td>53.7</td>\n",
       "      <td>18.621974</td>\n",
       "      <td>v9</td>\n",
       "      <td>57.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>637</th>\n",
       "      <td>YOLOv9e</td>\n",
       "      <td>GPU</td>\n",
       "      <td>53.1</td>\n",
       "      <td>18.832392</td>\n",
       "      <td>v9</td>\n",
       "      <td>57.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>638</th>\n",
       "      <td>YOLOv9e</td>\n",
       "      <td>GPU</td>\n",
       "      <td>53.8</td>\n",
       "      <td>18.587361</td>\n",
       "      <td>v9</td>\n",
       "      <td>57.3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>639</th>\n",
       "      <td>YOLOv9e</td>\n",
       "      <td>GPU</td>\n",
       "      <td>53.5</td>\n",
       "      <td>18.691589</td>\n",
       "      <td>v9</td>\n",
       "      <td>57.3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>640 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "       model device    ms        fps yolo  size\n",
       "0    YOLOv8n    CPU  62.9  15.898251   v8   3.2\n",
       "1    YOLOv8n    CPU  56.1  17.825312   v8   3.2\n",
       "2    YOLOv8n    CPU  55.2  18.115942   v8   3.2\n",
       "3    YOLOv8n    CPU  51.3  19.493177   v8   3.2\n",
       "4    YOLOv8n    CPU  49.1  20.366599   v8   3.2\n",
       "..       ...    ...   ...        ...  ...   ...\n",
       "635  YOLOv9e    GPU  52.9  18.903592   v9  57.3\n",
       "636  YOLOv9e    GPU  53.7  18.621974   v9  57.3\n",
       "637  YOLOv9e    GPU  53.1  18.832392   v9  57.3\n",
       "638  YOLOv9e    GPU  53.8  18.587361   v9  57.3\n",
       "639  YOLOv9e    GPU  53.5  18.691589   v9  57.3\n",
       "\n",
       "[640 rows x 6 columns]"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "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 tr th {\n",
       "        text-align: left;\n",
       "    }\n",
       "\n",
       "    .dataframe thead tr:last-of-type th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th colspan=\"2\" halign=\"left\">fps</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>model</th>\n",
       "      <th>device</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">YOLOv8m</th>\n",
       "      <th>CPU</th>\n",
       "      <td>5.862784</td>\n",
       "      <td>0.940588</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GPU</th>\n",
       "      <td>63.286615</td>\n",
       "      <td>4.467837</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">YOLOv8n</th>\n",
       "      <th>CPU</th>\n",
       "      <td>19.768982</td>\n",
       "      <td>1.115293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GPU</th>\n",
       "      <td>105.718299</td>\n",
       "      <td>17.478519</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">YOLOv8x</th>\n",
       "      <th>CPU</th>\n",
       "      <td>2.357122</td>\n",
       "      <td>0.516056</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GPU</th>\n",
       "      <td>29.211807</td>\n",
       "      <td>0.911822</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">YOLOv9c</th>\n",
       "      <th>CPU</th>\n",
       "      <td>3.674196</td>\n",
       "      <td>0.684758</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GPU</th>\n",
       "      <td>45.159232</td>\n",
       "      <td>3.020016</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"2\" valign=\"top\">YOLOv9e</th>\n",
       "      <th>CPU</th>\n",
       "      <td>1.715719</td>\n",
       "      <td>0.174222</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>GPU</th>\n",
       "      <td>18.890441</td>\n",
       "      <td>0.834113</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       fps           \n",
       "                      mean        std\n",
       "model   device                       \n",
       "YOLOv8m CPU       5.862784   0.940588\n",
       "        GPU      63.286615   4.467837\n",
       "YOLOv8n CPU      19.768982   1.115293\n",
       "        GPU     105.718299  17.478519\n",
       "YOLOv8x CPU       2.357122   0.516056\n",
       "        GPU      29.211807   0.911822\n",
       "YOLOv9c CPU       3.674196   0.684758\n",
       "        GPU      45.159232   3.020016\n",
       "YOLOv9e CPU       1.715719   0.174222\n",
       "        GPU      18.890441   0.834113"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.loc[:, [\"model\", \"device\", \"fps\"]].groupby([\"model\", \"device\"]).aggregate([\"mean\", \"std\"])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "tf",
   "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}