202 lines
214 KiB
Plaintext
202 lines
214 KiB
Plaintext
|
{
|
||
|
|
"cells": [
|
||
|
|
{
|
||
|
|
"cell_type": "code",
|
||
|
|
"execution_count": 2,
|
||
|
|
"metadata": {},
|
||
|
|
"outputs": [
|
||
|
|
{
|
||
|
|
"data": {
|
||
|
|
"image/png": "iVBORw0KGgoAAAANSUhEUgAABKUAAAKyCAYAAAAEvm1SAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy80BEi2AAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdeXhM1/8H8PdkmckqESFEJLELsQRRLbUUtW9VS7WxtLSl6Kb11VKxtCilLaWlRdFSFCnKD42lthKx77tISCSRfSbbzO+PdG4zMpPMJHfmZibv1/PkMTP33Hs/M3Imk3fOOVem0Wg0ICIiIiIiIiIisiA7qQsgIiIiIiIiIqKKh6EUERERERERERFZHEMpIiIiIiIiIiKyOIZSRERERERERERkcQyliIiIiIiIiIjI4hhKERERERERERGRxTGUIiIiIiIiIiIii2MoRUREREREREREFsdQioiIiIiIiIiILI6hFBERERERERERWVypQqmEhATs3LkTn332GXr27Alvb2/IZDLIZDKMGjXK5OPt3r0bAwcOhJ+fHxQKBfz8/DBw4EDs3r3b6GPk5eXh+++/x/PPP4+qVavC2dkZdevWxVtvvYVLly6ZXBMREREREREREZmPTKPRaEzeSSYzuG3kyJFYs2aNUcdRq9V488038dNPPxlsM2bMGPzwww+wszOcnyUmJqJXr144deqU3u0KhQJLly7FmDFjjKqLiIiIiIiIiIjMq8zT9/z9/fHiiy+Wat9PP/1UCKRCQkKwYcMGnDx5Ehs2bEBISAgA4Mcff8S0adMMHiM/Px8DBw4UAqmXXnoJu3fvxj///INvv/0W1apVQ3Z2Nt566y2TRl4REREREREREZH5lGqk1IwZMxAaGorQ0FD4+Pjg7t27qF27NgDjR0pdv34dTZo0QV5eHlq3bo3Dhw/D2dlZ2J6VlYWOHTsiKioKDg4OuHLlCurVq1fkOKtWrcIbb7wBABg/fjy+++47ne03b95Eq1atkJaWhnr16uHKlStwcHAw9SkTEREREREREZGISjVSaubMmejTpw98fHxKfeKvv/4aeXl5AIAlS5boBFIA4OLigiVLlgAoWC9q8eLFeo+zcOFCAICXlxcWLFhQZHu9evUwdepUAAUB1bZt20pdMxERERERERERiUOSq+9pNBpEREQAABo1aoS2bdvqbde2bVs0bNgQABAREYGnB3Vdv34dV65cAQAMGTIELi4ueo9TePF1hlJERERERERERNKTJJS6c+cO4uLiAAAdO3Ystq12e2xsLO7evauz7ciRI0Xa6VO9enU0aNAAAHD06NHSlExERERERERERCKSJJS6fPmycLtRo0bFti28XTsqqizHiYmJQWZmptG1EhERERERERGR+CQJpR48eCDc9vPzK7ZtrVq1hNsxMTFlPo5Go9HZj4iIiIiIiIiILE+Sy9Clp6cLt93c3Ipt6+rqKtzOyMgwy3EKy87ORnZ2tnBfrVYjOTkZVapUgUwmK/YcREREREREREQVnUajQXp6Onx9fWFnZ3g8lCShlEqlEm7L5fJi2yoUCuG2Uqk0y3EKmzt3LmbOnFnssYiIiIiIiIiIqHgxMTHFzmyTJJRycnISbufk5BTbtvCoJWdn52KPU/i+KccpbOrUqfjggw+E+6mpqfD390dMTAwqVapUbK1k26KiotC6dWupyyCyCexPROJgXyISD/sTkXjYnygtLQ21atWCu7t7se0kCaUKF1XcVDoAOouSPz1F7+njFBdKFXecwhQKhc6oKq1KlSoxlKrg3Nzc+D1AJBL2JyJxsC8RiYf9iUg87E+kVdIySJIsdF546FZJi44XXty88KLnpT2OTCYrcVF0In2aNm0qdQlENoP9iUgc7EtE4mF/IhIP+xMZS5JQqnHjxsLtq1evFtu28PagoKAyH6dWrVo6i54TGaukqaZEZDz2JyJxsC8RiYf9iUg87E9kLElCqdq1a8PX1xcAcOjQoWLbHj58GABQs2ZNBAYG6mxr3769cLu44zx69AjXr18HALRr1640JRPh2rVrUpdAZDPYn4jEwb5EJB72JyLxsD+RsSQJpWQyGfr37w+gYATTiRMn9LY7ceKEMMKpf//+ReYiNmjQQBg9tWnTJmRlZek9zpo1a4TbAwcOLGv5RERERERERERURpKEUgDw3nvvwd7eHgAwceJEKJVKne1KpRITJ04EADg4OOC9997Te5zJkycDAJKTk/Hxxx8X2X7r1i3MnTsXAFCvXj2GUkRERERERERE5UCprr535MgR3Lx5U7ifmJgo3L5586bOyCQAGDVqVJFjNGjQAB999BHmzZuHqKgotGvXDlOmTEHdunVx69YtzJ8/H2fOnAEAfPTRR6hfv77eWkaOHIlVq1bh6NGj+O677/Do0SOMHTsWlStXxsmTJzF79mykpaXBzs4O3377LRwcJLngINmAevXqSV0Ckc1gfyISB/sSkXjYn4jEw/5ExpJpNBqNqTuNGjUKP//8s9HtDZ1CrVZj7NixWLVqlcF933jjDaxYsQJ2doYHdSUmJqJXr144deqU3u0KhQJLly7FmDFjjK5ZKy0tDR4eHkhNTeUlLSu47OxsKBQKqcsgsgnsT0TiYF8iEg/7E5F42J/I2CxFsul7AGBnZ4effvoJu3btQv/+/eHr6wu5XA5fX1/0798ff/75J3788cdiAykA8Pb2xrFjx7Bs2TK0b98eVapUgZOTE+rUqYOxY8fi9OnTpQqkiAo7d+6c1CUQ2Qz2JyJxsC8RiYf9iUg87E9krFLNZVuzZk2RKXpl0atXL/Tq1atMx3BwcMC4ceMwbtw4kaoiIiIiIiIiIiJzkXSkFBERERERERERVUxc9ZvISH5+flKXQGQz2J+IxMG+RCQe9qcCGo0Gubm5UKvVUpdCVqx69epQqVRSl0GlZGdnB0dHR8hkMrOfq1QLnVckXOiciIiIiIhsXX5+PhITE5Geno7c3FypyyEiiTk6OsLd3R3e3t6wt7c3eX9jsxSOlCIy0qlTpxAaGip1GUQ2gf2JSBzsS0Tiqcj9KT8/HzExMcjOzoaHhwfc3Nxgb29vkVESZJuysrLg4uIidRlUChqNBvn5+cjIyEBKSgqUSiVq1apVqmDKGAyliIzEQYVE4mF/IhIH+xKReCpyf0pMTER2djb8/f3h7OwsdTlkA/Lz8+Hk5CR1GVQGbm5u8PDwwP3795GYmAgfHx+znIcLnRMREREREVVQGo0G6enp8PDwYCBFRDqcnZ1RqVIlpKenmy24ZyhFZCRvb2+pSyCyGexPROJgXyIST0XtT7m5ucjNzYWbm5vUpZANcXDgpCxb4e7uLrxPmANDKSIj1alTR+oSiGwG+xORONiXiMRTUfuT9ip75lovhiomhUIhdQkkEu17g7muyMlQishIFy5ckLoEIpvB/kQkDvYlIvFU9P7ERc1JTFlZWVKXQCIx93sDQykiIymVSqlLILIZ7E9E4mBfIhIP+xOReCryhQPINAyliIiIiIiIiIjI4hhKERmJiz8SiYf9iUgc7EtE4mF/IhKPnR2jBjIOv1OIjNS4cWOpSyCyGexPROJgXyISD/sTkXicnZ2lLoGsBEMpIiPduHFD6hKIbAb7E5E42JeIxMP+RCQelUoldQlkJRhKERnpyZMnUpdAZDPYn4jEwb5EJB72J7KEgwcPQiaTGfxyc3NDgwYNEBYWhsjISIPHuXv3brHH0fc1YMAAg8d78OABwsPD8fzzz6Nq1apwdHSEs7Mz/Pz80KFDB7z77rvYsmULUlNThX06depk8FzOzs5G1bRmzRoRX12yRg5SF0BERERERETW4f79+0hMTJS6DLPw9vaGv7+/pDVkZmbixo0buHHjBtavX48RI0Zg1apVsLe3N9s5V65ciffeew9ZWVk6j+fl5SE2NhaxsbH4+++/8e2332Lo0KHYuHGj2WqhioehFJGRzPmDgKiiYX8iEgf7EpF42J9Kdv/+fQQFBRUJL2yFk5MTrl27ZrFgaty4cRg/frxwX6PRIDk5GcePH8fixYuRkJCAtWvXolatWpgzZ47B4/Tv37/
|
||
|
|
"text/plain": [
|
||
|
|
"<Figure size 1200x700 with 1 Axes>"
|
||
|
|
]
|
||
|
|
},
|
||
|
|
"metadata": {},
|
||
|
|
"output_type": "display_data"
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"source": [
|
||
|
|
"import numpy as np\n",
|
||
|
|
"import matplotlib.pyplot as plt\n",
|
||
|
|
"import matplotlib.patches as mpatches\n",
|
||
|
|
"\n",
|
||
|
|
"folds = ['Fold 1', 'Fold 2', 'Fold 3', 'Fold 4', 'Fold 5']\n",
|
||
|
|
"only_desc = [0.9531471841, 0.9102803738, 0.9613453815, 0.9609895338, 0.8950984883]\n",
|
||
|
|
"plus_u = [0.9545669664, 0.9166261682, 0.9769076305, 0.9722131304, 0.9098859368]\n",
|
||
|
|
"plus_u_n = [0.9659252248, 0.9168224299, 0.968875502, 0.9814462417, 0.9180027485]\n",
|
||
|
|
"plus_u_m_m = [0.9602460956, 0.9182242991, 0.96937751, 0.9724072312, 0.8973889143]\n",
|
||
|
|
"plus_u_sel_n = [0.9645054425, 0.9168224299, 0.9789156627, 0.973358706, 0.952817224]\n",
|
||
|
|
"\n",
|
||
|
|
"data = np.array([only_desc, plus_u, plus_u_n, plus_u_m_m, plus_u_sel_n]) * 100\n",
|
||
|
|
"max_indices = np.argmax(data, axis=0)\n",
|
||
|
|
"\n",
|
||
|
|
"x = np.arange(len(folds))\n",
|
||
|
|
"width = 0.15\n",
|
||
|
|
"\n",
|
||
|
|
"colors = plt.cm.Spectral(np.linspace(0, 1, 5))\n",
|
||
|
|
"fig, ax = plt.subplots(figsize=(12, 7))\n",
|
||
|
|
"\n",
|
||
|
|
"rects1 = ax.bar(x - 2 * width, data[0], width, label='Only Desc.', color=colors[0], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[0] == 0 else '')\n",
|
||
|
|
"rects2 = ax.bar(x - width, data[1], width, label='+U.', color=colors[1], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[1] == 1 else '')\n",
|
||
|
|
"rects3 = ax.bar(x, data[2], width, label='+U./N.', color=colors[2], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[2] == 2 else '')\n",
|
||
|
|
"rects4 = ax.bar(x + width, data[3], width, label='+U./M.M.', color=colors[3], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[3] == 3 else '')\n",
|
||
|
|
"rects5 = ax.bar(x + 2 * width, data[4], width, label='+U./Sel N.', color=colors[4], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[4] == 4 else '')\n",
|
||
|
|
"\n",
|
||
|
|
"for i in range(len(folds)):\n",
|
||
|
|
" ax.bar(x[i] - 2 * width, data[0][i], width, color=colors[0], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 0 else '')\n",
|
||
|
|
" ax.bar(x[i] - width, data[1][i], width, color=colors[1], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 1 else '')\n",
|
||
|
|
" ax.bar(x[i], data[2][i], width, color=colors[2], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 2 else '')\n",
|
||
|
|
" ax.bar(x[i] + width, data[3][i], width, color=colors[3], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 3 else '')\n",
|
||
|
|
" ax.bar(x[i] + 2 * width, data[4][i], width, color=colors[4], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 4 else '')\n",
|
||
|
|
"\n",
|
||
|
|
"desc_mean = np.mean(only_desc) * 100\n",
|
||
|
|
"plus_u_mean = np.mean(plus_u) * 100\n",
|
||
|
|
"plus_u_n_mean = np.mean(plus_u_n) * 100\n",
|
||
|
|
"plus_u_m_m_mean = np.mean(plus_u_m_m) * 100\n",
|
||
|
|
"sel_n_mean = np.mean(plus_u_sel_n) * 100\n",
|
||
|
|
"\n",
|
||
|
|
"ax.axhline(y=desc_mean, color=colors[0], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=plus_u_mean, color=colors[1], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=plus_u_n_mean, color=colors[2], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=plus_u_m_m_mean, color=colors[3], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=sel_n_mean, color=colors[4], linestyle='--', label='Sel N. Mean')\n",
|
||
|
|
"\n",
|
||
|
|
"ax.text(x[1], sel_n_mean + 0.5, f'Max: {sel_n_mean:.2f}%', ha='center', fontsize=20, color='black')\n",
|
||
|
|
"ax.set_ylim([88, 100])\n",
|
||
|
|
"\n",
|
||
|
|
"ax.set_xticks(x)\n",
|
||
|
|
"ax.set_xticklabels(folds, fontsize=20)\n",
|
||
|
|
"\n",
|
||
|
|
"ax.grid(True, which='both', linestyle='--', linewidth=0.7, alpha=0.7, zorder=1)\n",
|
||
|
|
"\n",
|
||
|
|
"ax.tick_params(axis='both', which='major', labelsize=20)\n",
|
||
|
|
"\n",
|
||
|
|
"hatch_patch = mpatches.Patch(facecolor='white', edgecolor='black', hatch='//', label='BEST')\n",
|
||
|
|
"\n",
|
||
|
|
"legend = ax.legend(handles=[hatch_patch] + ax.get_legend_handles_labels()[0], fontsize=20)\n",
|
||
|
|
"\n",
|
||
|
|
"plt.tight_layout()\n",
|
||
|
|
"plt.show()\n"
|
||
|
|
]
|
||
|
|
},
|
||
|
|
{
|
||
|
|
"cell_type": "code",
|
||
|
|
"execution_count": 6,
|
||
|
|
"metadata": {},
|
||
|
|
"outputs": [
|
||
|
|
{
|
||
|
|
"data": {
|
||
|
|
"image/png": "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
|
||
|
|
"text/plain": [
|
||
|
|
"<Figure size 1200x700 with 1 Axes>"
|
||
|
|
]
|
||
|
|
},
|
||
|
|
"metadata": {},
|
||
|
|
"output_type": "display_data"
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"source": [
|
||
|
|
"import numpy as np\n",
|
||
|
|
"import matplotlib.pyplot as plt\n",
|
||
|
|
"import matplotlib.patches as mpatches\n",
|
||
|
|
"\n",
|
||
|
|
"# Fold 및 실험 데이터\n",
|
||
|
|
"folds = ['Fold 1', 'Fold 2', 'Fold 3', 'Fold 4', 'Fold 5']\n",
|
||
|
|
"t5_tiny = [0.9479413157, 0.9028037383, 0.9708835341, 0.9652711703, 0.9496106276]\n",
|
||
|
|
"t5_mini = [0.957406531, 0.9130841121, 0.9764056225, 0.9686013321, 0.9459459459]\n",
|
||
|
|
"t5_small = [0.9588263133, 0.9196261682, 0.9748995984, 0.9738344434, 0.9454878607]\n",
|
||
|
|
"bart_base = [0.9597728348, 0.9126168224, 0.9723895582, 0.9714557564, 0.9601465873]\n",
|
||
|
|
"t5_base = [0.9645054425, 0.9168224299, 0.9789156627, 0.973358706, 0.952817224]\n",
|
||
|
|
"\n",
|
||
|
|
"\n",
|
||
|
|
"# 데이터 준비\n",
|
||
|
|
"data = np.array([t5_tiny, t5_mini, t5_small, bart_base, t5_base]) * 100\n",
|
||
|
|
"max_indices = np.argmax(data, axis=0)\n",
|
||
|
|
"\n",
|
||
|
|
"# x축 설정\n",
|
||
|
|
"x = np.arange(len(folds))\n",
|
||
|
|
"width = 0.15\n",
|
||
|
|
"\n",
|
||
|
|
"# 색상 설정\n",
|
||
|
|
"colors = plt.cm.Spectral(np.linspace(0, 1, 5))\n",
|
||
|
|
"\n",
|
||
|
|
"fig, ax = plt.subplots(figsize=(12, 7))\n",
|
||
|
|
"\n",
|
||
|
|
"# 바 차트 그리기\n",
|
||
|
|
"rects1 = ax.bar(x - 2 * width, data[0], width, label='T5-Tiny', color=colors[0], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[0] == 0 else '')\n",
|
||
|
|
"rects2 = ax.bar(x - width, data[1], width, label='T5-Mini', color=colors[1], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[1] == 1 else '')\n",
|
||
|
|
"rects3 = ax.bar(x, data[2], width, label='T5-Small', color=colors[2], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[2] == 2 else '')\n",
|
||
|
|
"rects4 = ax.bar(x + width, data[3], width, label='BART-Base', color=colors[3], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[3] == 3 else '')\n",
|
||
|
|
"rects5 = ax.bar(x + 2 * width, data[4], width, label='T5-Base', color=colors[4], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[4] == 4 else '')\n",
|
||
|
|
"\n",
|
||
|
|
"for i in range(len(folds)):\n",
|
||
|
|
" ax.bar(x[i] - 2 * width, data[0][i], width, color=colors[0], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 0 else '')\n",
|
||
|
|
" ax.bar(x[i] - width, data[1][i], width, color=colors[1], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 1 else '')\n",
|
||
|
|
" ax.bar(x[i], data[2][i], width, color=colors[2], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 2 else '')\n",
|
||
|
|
" ax.bar(x[i] + width, data[3][i], width, color=colors[3], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 3 else '')\n",
|
||
|
|
" ax.bar(x[i] + 2 * width, data[4][i], width, color=colors[4], edgecolor='black', linewidth=1.5, zorder=3, hatch='//' if max_indices[i] == 4 else '')\n",
|
||
|
|
"\n",
|
||
|
|
"# 각 모델의 평균 계산\n",
|
||
|
|
"t5_tiny_mean = np.mean(t5_tiny) * 100\n",
|
||
|
|
"t5_mini_mean = np.mean(t5_mini) * 100\n",
|
||
|
|
"t5_small_mean = np.mean(t5_small) * 100\n",
|
||
|
|
"bart_base_mean = np.mean(bart_base) * 100\n",
|
||
|
|
"t5_base_mean = np.mean(t5_base) * 100\n",
|
||
|
|
"\n",
|
||
|
|
"# 평균을 점선으로 추가\n",
|
||
|
|
"ax.axhline(y=t5_tiny_mean, color=colors[0], linestyle='--', label='T5-Tiny Mean')\n",
|
||
|
|
"ax.axhline(y=t5_mini_mean, color=colors[1], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=t5_small_mean, color=colors[2], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=bart_base_mean, color=colors[3], linestyle='--')\n",
|
||
|
|
"ax.axhline(y=t5_base_mean, color=colors[4], linestyle='--', label='T5-Base Mean')\n",
|
||
|
|
"\n",
|
||
|
|
"# 가장 높은 평균값과 가장 낮은 평균값 계산\n",
|
||
|
|
"max_mean = max(t5_tiny_mean, t5_mini_mean, t5_small_mean, bart_base_mean, t5_base_mean)\n",
|
||
|
|
"min_mean = min(t5_tiny_mean, t5_mini_mean, t5_small_mean, bart_base_mean, t5_base_mean)\n",
|
||
|
|
"\n",
|
||
|
|
"# Fold 2의 중앙 정렬로 Max와 Min 텍스트 추가\n",
|
||
|
|
"ax.text(x[1], max_mean + 0.5, f'Max: {max_mean:.2f}%', ha='center', fontsize=20, color='black')\n",
|
||
|
|
"ax.text(x[1], min_mean - 0.8, f'Min: {min_mean:.2f}%', ha='center', fontsize=20, color='black')\n",
|
||
|
|
"\n",
|
||
|
|
"# y축 설정\n",
|
||
|
|
"ax.set_ylim([88, 100])\n",
|
||
|
|
"\n",
|
||
|
|
"ax.set_xticks(x)\n",
|
||
|
|
"ax.set_xticklabels(folds, fontsize=20)\n",
|
||
|
|
"\n",
|
||
|
|
"ax.grid(True, which='both', linestyle='--', linewidth=0.7, alpha=0.7, zorder=1)\n",
|
||
|
|
"\n",
|
||
|
|
"ax.tick_params(axis='both', which='major', labelsize=20)\n",
|
||
|
|
"\n",
|
||
|
|
"hatch_patch = mpatches.Patch(facecolor='white', edgecolor='black', hatch='//', label='BEST')\n",
|
||
|
|
"legend = ax.legend(handles=[hatch_patch] + ax.get_legend_handles_labels()[0], fontsize=20, loc='upper right')\n",
|
||
|
|
"\n",
|
||
|
|
"plt.tight_layout()\n",
|
||
|
|
"plt.show()\n"
|
||
|
|
]
|
||
|
|
}
|
||
|
|
],
|
||
|
|
"metadata": {
|
||
|
|
"kernelspec": {
|
||
|
|
"display_name": "torch",
|
||
|
|
"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": "3.10.14"
|
||
|
|
}
|
||
|
|
},
|
||
|
|
"nbformat": 4,
|
||
|
|
"nbformat_minor": 2
|
||
|
|
}
|