8_resnext50_more_epoch_sz300.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Kaggle Competition:  Iceberg or Ship"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# resnext50 architecture \n",
    "\n",
    "## increase epochs\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "## Import libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 2656\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 625023 Jan 14 18:47 1_getdata.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 245619 Jan 14 18:36 2_model_resnet34.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu  13445 Jan 14 18:36 3_model_resnext50.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu  14220 Jan 13 21:14 4_model_nasnet.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu  64965 Jan 14 19:40 4_resnext_resize.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 147078 Jan 14 19:57 5_resnet18_resize.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 138629 Jan 14 19:52 6_resnext50_resize.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 178957 Jan 14 23:41 7_resnext50_more_epoch.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 179449 Jan 14 21:28 7_resnext50_sz_100.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 178748 Jan 14 23:41 8_resnext50_more_epoch_sz300.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 137540 Jan 14 18:36 baseline.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu    536 Jan 14 18:36 helper.py\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 767392 Jan 14 18:36 other_groverpr.ipynb\r\n",
      "drwxrwxr-x 2 ubuntu ubuntu   4096 Jan 14 18:51 __pycache__\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu    745 Jan 14 20:14 README.md\r\n"
     ]
    }
   ],
   "source": [
    "!ls -l"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Put these at the top of every notebook, to get automatic reloading and inline plotting\n",
    "%reload_ext autoreload\n",
    "%autoreload 2\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.imports import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This file contains all the main external libs we'll use\n",
    "from fastai.imports import *\n",
    "from fastai.transforms import *\n",
    "from fastai.conv_learner import *\n",
    "from fastai.model import *\n",
    "from fastai.dataset import *\n",
    "from fastai.sgdr import *\n",
    "from fastai.plots import *\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import shutil\n",
    "from IPython.display import FileLink, FileLinks\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "import helper"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#PATH = \"/home/nidhin/data/iceberg/\"\n",
    "PATH = os.path.expanduser(\"~/data/iceberg/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 1973260\r\n",
      "drwxrwxr-x 2 ubuntu ubuntu       4096 Jan  9 00:38 models\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu     117951 Oct 23 17:27 sample_submission.csv\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu      38566 Jan  7 20:36 sample_submission.csv.7z\r\n",
      "drwxrwxr-x 2 ubuntu ubuntu     274432 Jan 14 18:44 test\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 1521771850 Oct 23 17:27 test.json\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu  257127394 Jan  7 20:36 test.json.7z\r\n",
      "drwxrwxr-x 8 ubuntu ubuntu       4096 Jan 14 23:14 tmp\r\n",
      "drwxrwxr-x 4 ubuntu ubuntu       4096 Jan 14 18:43 train\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu  196313674 Oct 23 17:23 train.json\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu   44932785 Jan  7 20:36 train.json.7z\r\n",
      "drwxrwxr-x 4 ubuntu ubuntu       4096 Jan 14 18:43 valid\r\n"
     ]
    }
   ],
   "source": [
    "!ls -l {PATH}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set Architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "arch=resnext50\n",
    "sz=200\n",
    "bs=32\n",
    "lr = 0.005"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e019b7bb40e466f98866c85f7df92ea",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, max=6), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                                     \n"
     ]
    }
   ],
   "source": [
    "def get_data(sz, bs):\n",
    "    tfms = tfms_from_model(arch, sz, aug_tfms=transforms_top_down, max_zoom=1.00)\n",
    "    data = ImageClassifierData.from_paths(PATH, tfms=tfms, bs=bs,test_name='test')\n",
    "    return data\n",
    "\n",
    "data = get_data(sz, bs)\n",
    "data = data.resize(int(sz*1.5), 'tmp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "100%|██████████| 41/41 [00:26<00:00,  1.56it/s]\n",
      "100%|██████████| 11/11 [00:06<00:00,  1.82it/s]\n",
      "100%|██████████| 264/264 [02:36<00:00,  1.68it/s]\n"
     ]
    }
   ],
   "source": [
    "#tfms = tfms_from_model(arch, sz, aug_tfms=transforms_side_on, max_zoom=1.1)\n",
    "#data = ImageClassifierData.from_paths(PATH, tfms=tfms, bs=bs,test_name='test',)\n",
    "\n",
    "learn = ConvLearner.pretrained(arch, data, precompute=True, ps=0.5)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Learning Rate Finder"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "64daedc60ba34657a332594725828e05",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=1), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " 71%|███████   | 29/41 [00:00<00:00, 37.60it/s, loss=5.74] \n",
      "                                                          \r"
     ]
    }
   ],
   "source": [
    "lrf = learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYwAAAEKCAYAAAAB0GKPAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzt3Xl8XXWd//HXJ2vbpHvShe6FQoECxYayObKMYlGkKM5YBhRm1MqMqKOzgLOI4szvwajzG53RESp2cEEQFbQ6yKIFyipNsS2lFFqS1oS0TdI0aZM0zXI/88c9t1zSLKfLybn35v18PPK493zP95z76Wl7Pznf7/d8v+buiIiIDCYv7gBERCQ7KGGIiEgoShgiIhKKEoaIiISihCEiIqEoYYiISChKGCIiEooShoiIhKKEISIioRTEHcDxVFZW5rNnz447DBGRrLFu3bpGdy8PUzenEsbs2bOprKyMOwwRkaxhZjvC1lWTlIiIhKKEISIioShhiIhIKEoYIiISihKGiIiEooQhIiKhKGGIiEgoShgiIlnsN5t3c+eTrzMUy20rYYiIZLFfbazj+8/twMwi/ywlDBGRLFbd2Mbc8pIh+SwlDBGRLOXuVDW0MbdMCUNERAbQ2NrJ/oPdzFHCEBGRgVQ3tgEwp7x0SD5PCUNEJEtVN7YCqElKREQGVtXQRlFBHieMGzkkn6eEISKSpaoa25g9cRT5edEPqYUIF1Ays5XAFUC9uy/oY//fAdemxXEqUO7uTWa2HdgP9ADd7l4RVZwiItmqurGNE4doSC1Ee4dxN7Ckv53u/lV3X+juC4HPA0+6e1NalUuC/UoWIiK9dPck2LGnjblD1OENESYMd18DNA1aMeka4N6oYhERyTVvNB+gq8eHbEgtZEAfhpmNInkn8rO0YgceNbN1ZrZ8kOOXm1mlmVU2NDREGaqISMaoCobUDtUIKciAhAG8D3imV3PUhe7+NuBy4JNm9o7+Dnb3Fe5e4e4V5eXlUccqIpIRqhuCZzCGWcJYRq/mKHevC17rgQeBxTHEJSKSsaoaWxk7spAJJUVD9pmxJgwzGwtcBPwirazEzEan3gOXAZviiVBEJDNVN7Yxp6xkSGapTYlyWO29wMVAmZnVArcChQDufkdQ7f3Ao+7elnboZODB4CIUAD9y94ejilNEJBtVN7Rx3tyJQ/qZkSUMd78mRJ27SQ6/TS+rAs6KJioRkezX3tlNXUvHkE1rnpIJfRgiInIEtje2AzCnbOiewQAlDBGRrHNoltohHCEFShgiIlknNUvt7LJRQ/q5ShgiIlmmqqGNE8aOYFRRZN3QfVLCEBHJMlWNbcwZ4g5vUMIQEckqyXW8W4e8/wKUMEREssre9i72dXQP+QgpUMIQEckqVQ3BsqxqkhIRkYHEMUttihKGiEgWqW5sozDfmDZE63inU8IQEckiVQ2tzJwwioL8of/6VsIQEcki1Y1DuyxrOiUMEZEs0ZNwtu9pj6X/ApQwRESyRl3zATq7E7E8gwFKGCIiWePQCCk1SYmIyECqg2cwdIchIiIDqm5sY3RxAWWlQ7eOd7rIEoaZrTSzejPrcz1uM7vYzFrMbH3w84W0fUvM7FUz22Zmt0QVo4hINklNOjiU63ini/IO425gySB1nnL3hcHPbQBmlg98C7gcOA24xsxOizBOEZGsUNXQFtsIKYgwYbj7GqDpKA5dDGxz9yp37wTuA5Ye1+BERLJMR1cPdS0HYpl0MCXuPozzzWyDmf3azE4PyqYBNWl1aoMyEZFha8eedtyJZR2MlKFdrumtXgRmuXurmb0H+DkwD+ircc77O4mZLQeWA8ycOTOKOEVEYndoltpcbJIajLvvc/fW4P1DQKGZlZG8o5iRVnU6UDfAeVa4e4W7V5SXl0cas4hIXFLPYMQ1pBZiTBhmNsWCrn4zWxzEsgdYC8wzszlmVgQsA1bFFaeISCaobmxj8phiSorjaxiK7JPN7F7gYqDMzGqBW4FCAHe/A/gg8Jdm1g0cAJa5uwPdZnYT8AiQD6x095ejilNEJBtUN7bFencBESYMd79mkP3fBL7Zz76HgIeiiEtEJBtVNbSyZMHUWGOIe5SUiIgMYm9bJ3vbuzgxxhFSoIQhIpLxqvfE3+ENShgiIhmvukEJQ0REQqhqbKUgz5gxYVSscShhiIhkuOrGNmZOGEVhDOt4p1PCEBHJcFUN8Q+pBSUMEZGMlkg42/coYYiIyCB27uugoysR27Ks6ZQwREQyWKaMkAIlDBGRjFbdGMxSG/NDe6CEISKS0V5vaKOkKJ9Jo4vjDkUJQ0Qkk1XHvI53OiUMEZEMlpylNv4Ob1DCEBHJWB1dPdTubc+IDm9QwhARyViv7NxHwuG0qaPjDgVQwhARyVgba1sAOGvGuJgjSVLCEBHJUBtqmpk0upgpY0bEHQqghCEikrHW1zZz5vRxGTFCCiJMGGa20szqzWxTP/uvNbONwc+zZnZW2r7tZvaSma03s8qoYhQRyVT7Orqoamhj4YyxcYdySJR3GHcDSwbYXw1c5O5nAl8GVvTaf4m7L3T3iojiExHJWC8F/RdnTs+M/guAgqhO7O5rzGz2APufTdt8HpgeVSwiItlmQ20zAGdOHx53GEfio8Cv07YdeNTM1pnZ8phiEhGJzYaaZuaUlTBuVFHcoRwS2R1GWGZ2CcmE8fa04gvdvc7MJgGPmdkWd1/Tz/HLgeUAM2fOjDxeEZGhsKGmhXPnTog7jLeI9Q7DzM4E7gKWuvueVLm71wWv9cCDwOL+zuHuK9y9wt0rysvLow5ZRCRyu/d1sGtfB2dlUP8FxJgwzGwm8ADwYXd/La28xMxGp94DlwF9jrQSEclFG2qS/RdnZdAIKTiCJikzK3H3tiOofy9wMVBmZrXArUAhgLvfAXwBmAj8dzDGuDsYETUZeDAoKwB+5O4Ph/1cEZFst6G2mfw84/QTsixhmNkFJJuNSoGZwfMSn3D3vxroOHe/ZpD9HwM+1kd5FXDW4UeIiAwPG2tbmD9lNCMK8+MO5S3CNEn9B/BuYA+Au28A3hFlUCIiw5W7s6GmOaOev0gJ1Yfh7jW9inoiiEVEZNjbvqedfR3dGfWEd0qYPoyaoFnKzawI+DTwSrRhiYgMT6kO72y9w7gR+CQwDagFFgID9l+IiMjRWV/TzMjCfOZNyoxV9tKFucM4xd2vTS8wswuBZ6IJSURk+NpY28wZ08ZSkJ8pE3G8KUxE/xWyTEREjkFXT4JNdfsyav6odP3eYZjZ+cAFQLmZfS5t1xggs8Z6iYjkgFd37aezO5ExK+z1NlCTVBHJZy8KgPQFZfcBH4wyKBGR4Sg1Q22mTQmS0m/CcPcngSfN7G533zGEMYmIDEsbapoZP6qQGRNGxh1Kn8J0ereb2VeB04FDC8u6+6WRRSUiMgxtrG3hrBmZsyRrb2E6ve8BtgBzgC8B24G1EcYkIjLstB3s5rXd+zPy+YuUMAljort/F+hy9yfd/S+A8yKOS0RkWNn0RgsJJyOf8E4J0yTVFbzuNLP3AnVoOVURkeNqYwau4d1bmITxL2Y2Fvgbks9fjAE+G2lUIiLDzPraZqaNG0lZaXHcofRrwIRhZvnAPHf/FdACXDIkUYmIDDMba5tZmKHPX6QM2Ifh7j3AlUMUi4jIsLSn9SA1TQcy9gnvlDBNUs+a2TeBHwOHVtxz9xcji0pEZBhJ9V9k6hPeKWESxgXB621pZQ7oOQwRkeNgQ20zeQZnTMvyOwx3P+p+CzNbCVwB1Lv7gj72G/AN4D1AO3BD6s7FzK4H/imo+i/u/r2jjUNEJJNtqGnmpEmllBSH+R0+PlHPn3s3sGSA/ZcD84Kf5cC3AcxsAnArcC6wGLjVzMZHGqmISAzcnQ21LRk7f1S6SBOGu68BmgaoshT4vic9D4wzs6kk1xB/zN2b3H0v8BgDJx4RkaxUu/cATW2dnJnh/RcQ/R3GYKYB6euF1wZl/ZUfxsyWm1mlmVU2NDREFqiISBRSM9QuzII7jEEbzMzsA30UtwAvuXv9MX5+XzNs+QDlhxe6rwBWAFRUVPRZR0QkU22sbaGoII9TpowevHLMwvSwfBQ4H3g82L4YeB442cxuc/cfHMPn1wIz0rank5x6pDb4nPTyJ47hc0REMtL6mmZOmzqGooK4G3wGFybCBHCqu1/t7lcDpwEHSXZI33yMn78K+IglnQe0uPtO4BHgMjMbH3R2XxaUiYjkjJ6Es+mNlox/wjslzB3GbHffnbZdD5zs7k1m1tXfQQBmdi/JO4UyM6slOfKpEMDd7wAeIjmkdhvJYbV/HuxrMrMv8+Y06re5+0Cd5yIiWWdbfSvtnT0Z/4R3SpiE8ZSZ/Qr4SbB9NbDGzEqA5oEOdPdrBtnvwCf72bcSWBkiPhGRrLShJliSNYfuMD5JMklcSLIz+vvAz4Ive01GKCJylJ6r2sP4UYXMmVgSdyihhHnS24GfBj8iInIc9CScJ16t5+JTJpGXl5lLsvY2aKe3mX3AzLaaWYuZ7TOz/Wa2byiCExHJVetrmtnb3sUl8yfFHUpoYZqkvgK8z91fiToYEZHh4vEt9eTnGRfNK487lNDCDKvdrWQhInJ8rd5Sz6KZ4xk7qjDuUEILc4dRaWY/Bn5O8vkLANz9gciiEhHJYbtaOti8cx83L5kfdyhHJEzCGEPyGYnL0socUMIQETkKj7+anFXp0izqv4Bwo6T+fCgCEREZLn77Sj3Txo3k5MmlcYdyRPpNGGb29+7+FTP7L/qY+M/dPx1pZCIiOaijq4dntjXywUXTSa4hlz0GusNIdXRXDkUgIiLDwe+qmzjQ1ZN1zVEwQMJw918Gr1oaVUTkOHl8Sz0jCvM4/8SJcYdyxMKsh3Ey8LfA7PT67n5pdGGJiOQed2f1lnouOLGMEYX5cYdzxMKMkvoJcAdwF9ATbTgiIrnr9YY2/tDUzsffMTfuUI5KmITR7e7fjjwSEZEc9/iW7BxOmxLmSe9fmtlfmdlUM5uQ+ok8MhGRHLN6Sz2nTB7NtHEj4w7lqIS5w7g+eP27tDIHsvOeSkQkBvs6uli7vYmP/VH2fnUOmDDMLA+4zt2fGaJ4RERy0tNbG+lOOH98anY2R8EgTVLungC+drQnN7MlZvaqmW0zs1v62P8fZrY++HnNzJrT9vWk7Vt1tDGIiGSC1VvqGTuykLOzZHW9voRpknrUzK4GHggWUwrFzPKBbwHvAmqBtWa2yt03p+q4+2fT6n8KODvtFAfcfWHYzxMRyVSJYLGki04upyA/TNdxZgqTMD4HlADdZtZBcplWd/cxgxy3GNjm7lUAZnYfsBTY3E/9a4BbQ0UtIpJFNr7RQmNrZ9aOjkoZNNW5+2h3z3P3IncfE2wPliwApgE1adu1QdlhzGwWMAdYnVY8wswqzex5M7sqxOeJiGSk1VvqyTO46OTsWSypL2HuMDCz8cA8YESqzN3XDHZYH2X9NWktA37q7ukPBs509zozmwusNrOX3P31PmJbDiwHmDlz5iAhiYgMvce31HP2zPGMLymKO5RjEmZN748Ba4BHgC8Fr18Mce5aYEba9nSgrp+6y4B70wvcvS54rQKe4K39G+n1Vrh7hbtXlJdnd/YWkdxTv6+Dl95oyfrmKAj34N5ngHOAHe5+Cckv7oYQx60F5pnZHDMrIpkUDhvtZGanAOOB59LKxptZcfC+DLiQ/vs+REQy1hOvJr8uLzkl+xNGmCapDnfvMDPMrNjdtwRf8gNy924zu4nkHUk+sNLdXzaz24BKd08lj2uA+3qNwDoVuNPMEiST2u3po6tERLLF6i31TB07glOnjo47lGMWJmHUmtk4kmt6P2Zme+m/aekt3P0h4KFeZV/otf3FPo57FjgjzGeIiGSqzu4ET29r5MqFJ2TdYkl9CbNE6/uDt180s8eBscDDkUYlIpID1m5vovVgN5fmQHMUhB8l9XZgnrv/j5mVkxweWx1pZCIiWW71lnqKCvK44KTsWyypL2FGSd0K3Ax8PigqBH4YZVAiItkutVjS+XMnMqoo1O/mGS/MKKn3A1cCbXBouGv2996IiERo3Y69VDe2cfmCKXGHctyESRidwQgmBzCzkmhDEhHJfj94fgejiwu4cuEJcYdy3IRJGPeb2Z3AODP7OPAb4DvRhiUikr32tB7k1y/t4upF03OmOQrCjZL6mpm9C9gHnAJ8wd0fizwyEZEsdX9lLZ09Ca49N7emKwqV+oIEoSQhIjKIRML50Qs7OHfOBOZNzq3u3n4Thpntp+/JAsNOby4iMuw8ubWBmqYD/P2758cdynHXb8Jw99xKjSIiQ+CHz+2grLSYd5+eO6OjUrJ36ScRkQxTu7ed1a/Ws+ycGRQV5N7Xa+79iUREYnLvC3/AgGtyrLM7RQlDROQ46OxO8OO1NVw6fxLTxo2MO5xIKGGIiBwHj7y8i8bWTq49b1bcoURGCUNE5Dj4wfM7mDFhJBfNy92VP5UwRESO0Wu79/NCdRPXnjuLvLzsX/eiP0oYIiLH6J7nd1CUn8efLJoedyiRUsIQETkGbQe7eeDFN3jPGVOYWFocdziRijRhmNkSM3vVzLaZ2S197L/BzBrMbH3w87G0fdeb2dbg5/oo4xQROVqrNtSx/2A31+VwZ3dKZNMomlk+8C3gXUAtsNbMVrn75l5Vf+zuN/U6dgJwK1BBcnqSdcGxe6OKV0TkSLk7P3huB/OnjGbRrPFxhxO5KO8wFgPb3L3K3TuB+4ClIY99N/CYuzcFSeIxYElEcYqIHJXf1zSzeec+rjtvFma529mdEmXCmAbUpG3XBmW9XW1mG83sp2Y24wiPFRGJzQ+f30FJUT5XnT08vp6iTBh9pdves9/+Epjt7meSXJjpe0dwbLKi2XIzqzSzyoaGhqMOVkTkSOxt6+RXG3fy/rdNo7Q4dxZJGkiUCaMWmJG2PR2oS6/g7nvc/WCw+R1gUdhj086xwt0r3L2ivDx3H5gRkcxy97Pb6exODIvO7pQoE8ZaYJ6ZzTGzImAZsCq9gplNTdu8EngleP8IcJmZjTez8cBlQZmISOzqmg9w55rXee8ZU5k/ZfgsDRTZfZS7d5vZTSS/6POBle7+spndBlS6+yrg02Z2JdANNAE3BMc2mdmXSSYdgNvcvSmqWEVEjsS/PbyFhMMtl+feIkkDMfc+uwayUkVFhVdWVsYdhojksHU7mrj628/xqUtP4m8uOyXucI6Zma1z94owdfWkt4hISImE86VfbmbymGJuvOjEuMMZckoYIiIhPfD7N9hY28LNS+ZTMkxGRqVTwhARCaH1YDf/9vAWzpoxjqsWDo/nLnpTwhARCeG/H99Gw/6D3Pq+03J6CvOBKGGIiAziD3vauevpat5/9jTeNjP354zqjxKGiMgg/t9Dr5Bvxs1Lhtcw2t6UMEREBvDc63t4+OVd/NXFJzJl7Ii4w4mVEoaISD96Es6Xfvky08aN5OPvmBt3OLFTwhAR6cd9a//All37+fx75jOiMD/ucGKnhCEi0oeWA138+6OvsXj2BN57xtTBDxgGlDBERPrwn7/dyt72Tr7wvtOGxeJIYShhiIj0snrLblY+U82yc2ayYNrYuMPJGEoYIiJpXm9o5TP3rue0qWP4whWnxR1ORlHCEBEJ7O/oYvn3KyksyOPODy9iZJE6utMNv9mzRET6kEg4n/3xerbvaeeej53L9PGj4g4p4+gOQ0QE+Ppvt/KbV+r5whWncd7ciXGHk5GUMERk2Ht4007+87db+ZNF0/nI+cNnje4jpYQhIsPaq7v287n7N7Bwxji+fNUCDaEdQKQJw8yWmNmrZrbNzG7pY//nzGyzmW00s9+a2ay0fT1mtj74WRVlnCIyPLW0d7H8B5WUFBdw54cX6WnuQUTW6W1m+cC3gHcBtcBaM1vl7pvTqv0eqHD3djP7S+ArwIeCfQfcfWFU8YnI8NaTcD513++paz7AfcvPZ/KY4T2xYBhR3mEsBra5e5W7dwL3AUvTK7j74+7eHmw+D0yPMB4RkUO+8sgW1rzWwJeXLmDRrOG7xsWRiDJhTANq0rZrg7L+fBT4ddr2CDOrNLPnzeyq/g4ys+VBvcqGhoZji1hEhoW7nqrizier+PB5s1i2eGbc4WSNKJ/D6KvnyPusaHYdUAFclFY8093rzGwusNrMXnL31w87ofsKYAVARUVFn+cXEYHksxb/+tArfPfpapacPoV/1pPcRyTKhFELzEjbng7U9a5kZu8E/hG4yN0PpsrdvS54rTKzJ4CzgcMShohIGB1dPfzNTzbwvxt3csMFs/nnK04jf5iuzX20omySWgvMM7M5ZlYELAPeMtrJzM4G7gSudPf6tPLxZlYcvC8DLgTSO8tFREJrae/iIytf4H837uQf3jOfW9+nZHE0IrvDcPduM7sJeATIB1a6+8tmdhtQ6e6rgK8CpcBPgrHPf3D3K4FTgTvNLEEyqd3ea3SViEgobzQf4IaVL7B9TxvfWLaQpQsH6kqVgZh77jT7V1RUeGVlZdxhiEiG2Fy3jxv+5wUOdPVw54cXccGJZXGHlHHMbJ27V4Spq8kHRSQnPb21kRt/uI7S4gJ+cuP5zJ8yJu6Qsp4ShojkFHfn/soa/vHBTZxYXsrdf3EOU8eOjDusnKCEISI5Y8eeNv7p55t4amsjF5w4kW9ft4ixIwvjDitnKGGISNbr6klw11PVfP03r1GYn8eXl57OtefOIk8joY4rJQwRyWrra5q55Wcb2bJrP+8+fTJfunIBU8ZqXqgoKGGISFZqPdjN1x55le89t51Jo4u547pFLFkwJe6wcpoShohkFXdProz3i03s2tfBh8+bxd+9+xRGj1BfRdSUMEQkK3T3JHj45V18Z00VG2pbOGXyaL75Z2/TTLNDSAlDRDJa68Fu7l9bw3efruaN5gPMKSvhX65awJ9WzKCoQIuGDiUlDBHJSLtaOrj72e3c87sd7O/oZvHsCdz6vtN456mTNfopJkoYIpIxEglnfW0zP3x+B7/cUEdPwrn8jKl8/I/msnDGuLjDG/aUMEQkVj0J54XqJh7etJNHXt7Nrn0djCrK59pzZ/HRt89hxoRRcYcoASUMERlynd0Jnn29kYc37eKxzbvZ09ZJcUEeF51czs1nnMIfnzqZMRr1lHGUMEQkcomEs7W+lXU79vK76j2s3lLP/o5uSosLuHT+JJYsmMLFp5QzqkhfSZlMfzsicty1d3azoaaFdTuaqNyxlxd37GVfRzcAZaVFLDl9CksWTOHCk8oYUZgfc7QSlhKGiBw1d6d+/0G27m5la/1+tta3sumNFjbX7aM7kVxrZ96kUt575lQWzZpAxazxzJo4imDBNMkyShgiMqjWg93UNR/gjb0HeL2h9S0JYn9w5wAwdmQh86eM5hMXzaVi1gTOnjmOcaOKYoxcjqdIE4aZLQG+QXKJ1rvc/fZe+4uB7wOLgD3Ah9x9e7Dv88BHgR7g0+7+SJSxigxHPQmnub2TPW2d7GntpH5/BztbOqhrPkBd8wFq9yZf96UlBYCJJUWcNKmUpQtPYN6k0cybVMq8yaMpKy3S3UMOiyxhmFk+8C3gXUAtsNbMVvVam/ujwF53P8nMlgH/BnzIzE4DlgGnAycAvzGzk929J6p4RbJVV0+C9s4e2ju72d/Rzf6OLvYd6GZfRxf7O9583d/Rxd62LhpbD9LUlkwSe9s76WuV5rEjCzlh3Eimjx/JObMncMK4kZwwbgTTxo1kTlkJE0uLh/4PKrGL8g5jMbDN3asAzOw+YCmQnjCWAl8M3v8U+KYlfz1ZCtzn7geBajPbFpzvuQjjlWHI3XGHhDuJ4PXN7aAs4fS4H3rtSSTr9PQq7+5xuhNOTyJBd0+yXnLb6epJ0B28dvWkXtPedyfo7ElwsDvBwa4eDnYn6OwOtruT2wc6e2jv7OFAVzI5tHf20NHVQ1dPH9/4vRTkGaNHFDC+pIiykmJOLC9l8ZwiJpYUMbG0mAklRUwsLaK8tJip40ZSWqzWajlclP8qpgE1adu1wLn91XH3bjNrASYG5c/3OnZaVIFe8V9P0dGViOr0oXlfv+od6TmOucLhVXrH1dcpeofuabVS+/r646XO7en1gmPd3/ys5D5PqxN82afqBe/xw/clDr33Q+dMJYZMYgbFBXkUF+QnXwuT74vy8ygqyGNkYT5TxxYysiifUUX5jCoqSL4vzA/KChg9ooAxIwuTryMKGDOikNEjChlRmKemIjlmUSaMvv519v4v2l+dMMcmT2C2HFgOMHPmzCOJ75CTyktD/ZY2JI7D/+nBThHmi6N3jd6H9HWG3ue1PjYsrdTsrfXM3tx/aJ+9WSO5P/04Syuzt5wjvW6eGWZvluWl3psd2s4zyMuzQ/Xz0o7LM8jPM/LMyM8z8oPj8/PsUHlB8L4g38jPyzu0XZif3F+Yn0dhfh4F+UZR8FqYn0dhXh6FBUZBXh6F+aYvdcloUSaMWmBG2vZ0oK6fOrVmVgCMBZpCHguAu68AVgBUVFQc1bf+15edfTSHiYgMK1HODbwWmGdmc8ysiGQn9qpedVYB1wfvPwis9mQbxSpgmZkVm9kcYB7wQoSxiojIICK7wwj6JG4CHiE5rHalu79sZrcBle6+Cvgu8IOgU7uJZFIhqHc/yQ7ybuCTGiElIhIvOx4drZmioqLCKysr4w5DRCRrmNk6d68IU1fLVYmISChKGCIiEooShoiIhKKEISIioShhiIhIKDk1SsrMGoAdfewqAxqHOJyjlS2xKs7jK1vihOyJVXGGM8vdy8NUzKmE0R8zqww7bCxu2RKr4jy+siVOyJ5YFefxpyYpEREJRQlDRERCGS4JY0XcARyBbIlVcR5f2RInZE+sivM4GxZ9GCIicuyGyx2GiIgco5xPGGa2xMxeNbNtZnZL3PH0x8y2m9lLZrbezDJqBkUzW2lm9Wa2Ka1sgpk9ZmZbg9fxccYYxNRXnF80szeC67rezN4TZ4xBTDPM7HEze8XMXjazzwTlGXVNB4gzo66pmY0wsxfMbEMQ55eC8jlm9rvgev44WGYhVgPEereZVadd04Vxx9qXnG6SMrN84DXgXSQXZVoLXOPumwc8MAZcdJM8AAAFXUlEQVRmth2ocPeMGzduZu8AWoHvu/uCoOwrQJO73x4k4vHufnMGxvlFoNXdvxZnbOnMbCow1d1fNLPRwDrgKuAGMuiaDhDnn5JB19SSyxSWuHurmRUCTwOfAT4HPODu95nZHcAGd/92hsZ6I/Ard/9pnPENJtfvMBYD29y9yt07gfuApTHHlHXcfQ3J9UrSLQW+F7z/Hskvklj1E2fGcfed7v5i8H4/8ArJNesz6poOEGdG8aTWYLMw+HHgUiD1BRz79YQBY80KuZ4wpgE1adu1ZOA/+IADj5rZumCd8kw32d13QvKLBZgUczwDucnMNgZNVrE3naUzs9nA2cDvyOBr2itOyLBramb5ZrYeqAceA14Hmt29O6iSMf/3e8fq7qlr+q/BNf0PMyuOMcR+5XrCsD7KMjWbX+jubwMuBz4ZNK/Isfs2cCKwENgJ/Hu84bzJzEqBnwF/7e774o6nP33EmXHX1N173H0hMJ1ky8KpfVUb2qj61jtWM1sAfB6YD5wDTABibd7tT64njFpgRtr2dKAuplgG5O51wWs98CDJf/SZbHfQxp1q666POZ4+ufvu4D9oAvgOGXJdg/brnwH3uPsDQXHGXdO+4szUawrg7s3AE8B5wDgzSy1DnXH/99NiXRI0/7m7HwT+hwy6pulyPWGsBeYFoyWKSK4ZvirmmA5jZiVBpyJmVgJcBmwa+KjYrQKuD95fD/wixlj6lfoCDryfDLiuQcfnd4FX3P3/p+3KqGvaX5yZdk3NrNzMxgXvRwLvJNnf8jjwwaBa7NcT+o11S9ovCkayryX2f6d9yelRUgDBkL+vA/nASnf/15hDOoyZzSV5VwFQAPwok+I0s3uBi0nOqrkbuBX4OXA/MBP4A/An7h5rh3M/cV5MsunEge3AJ1L9BHExs7cDTwEvAYmg+B9I9g9kzDUdIM5ryKBramZnkuzUzif5S/D97n5b8P/qPpJNPL8Hrgt+g4/NALGuBspJNqOvB25M6xzPGDmfMERE5PjI9SYpERE5TpQwREQkFCUMEREJRQlDRERCUcIQEZFQlDBEAmb2bPA628z+7Dif+x/6+iyRbKJhtSK9mNnFwN+6+xVHcEy+u/cMsL/V3UuPR3wicdEdhkjAzFIPSt0O/FGwLsFng8nivmpma4PJ4T4R1L84WC/iRyQfbsPMfh5MIPlyahJJM7sdGBmc7570z7Kkr5rZJkuuh/KhtHM/YWY/NbMtZnZP8BQwZna7mW0OYsmIKcZleCgYvIrIsHMLaXcYwRd/i7ufE8wi+oyZPRrUXQwscPfqYPsv3L0pmPZhrZn9zN1vMbObggnnevsAyaemzyL5hPpaM1sT7DsbOJ3kHEjPABea2WaS03HMd3dPTTMhMhR0hyEyuMuAjwRTUv8OmAjMC/a9kJYsAD5tZhuA50lOfDmPgb0duDeYzG838CTJGUtT564NJvlbD8wG9gEdwF1m9gGg/Zj/dCIhKWGIDM6AT7n7wuBnjrun7jDaDlVK9n28Ezjf3c8iOX/RiBDn7k/6vEc9QEGwvsNikjPIXgU8fER/EpFjoIQhcrj9wOi07UeAvwym+sbMTg5mFe5tLLDX3dvNbD7JKbZTulLH97IG+FDQT1IOvAN4ob/AgrUpxrr7Q8Bfk2zOEhkS6sMQOdxGoDtoWrob+AbJ5qAXg47nBvpe7vNh4EYz2wi8SrJZKmUFsNHMXnT3a9PKHwTOBzaQnP317919V5Bw+jIa+IWZjSB5d/LZo/sjihw5DasVEZFQ1CQlIiKhKGGIiEgoShgiIhKKEoaIiISihCEiIqEoYYiISChKGCIiEooShoiIhPJ/RuoArL6q6UIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f414d5a25f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f414d5c4e48>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "28c11266bf5444d69f61fb8f6f389c37",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=10), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.575   0.36835 0.84091]                          \n",
      "[1.      0.50283 0.36706 0.82386]                          \n",
      "[2.      0.44092 0.33442 0.83523]                          \n",
      "[3.      0.40784 0.32021 0.83239]                          \n",
      "[4.      0.36265 0.31447 0.86079]                          \n",
      "[5.      0.38851 0.3316  0.83807]                          \n",
      "[6.      0.34834 0.32499 0.84659]                          \n",
      "[7.      0.32054 0.32331 0.84091]                          \n",
      "[8.      0.31901 0.31272 0.84659]                          \n",
      "[9.      0.30314 0.31837 0.84943]                          \n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit(lr, 10)\n",
    "learn.precompute=False"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stochastic Gradient Descent with Restart"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "015424c7ecb249c1bad35265578d87e9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=63), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.42701 0.30452 0.84943]                          \n",
      "[1.      0.43297 0.28763 0.88636]                          \n",
      "[2.      0.43043 0.28029 0.88636]                          \n",
      "[3.      0.42741 0.30146 0.85227]                          \n",
      "[4.      0.43036 0.30558 0.85796]                          \n",
      "[5.      0.41627 0.30468 0.85227]                          \n",
      "[6.      0.4239  0.3017  0.85511]                          \n",
      "[7.      0.40127 0.32226 0.85511]                          \n",
      "[8.      0.40466 0.32664 0.83523]                          \n",
      "[9.      0.41538 0.28473 0.86079]                          \n",
      "[10.       0.4154   0.31276  0.86079]                      \n",
      "[11.       0.42301  0.29041  0.86079]                      \n",
      "[12.       0.38929  0.28799  0.875  ]                      \n",
      "[13.       0.40402  0.29968  0.87216]                      \n",
      "[14.       0.39523  0.28423  0.88352]                      \n",
      "[15.       0.38529  0.33157  0.86648]                      \n",
      "[16.       0.40798  0.31204  0.84375]                      \n",
      "[17.       0.39736  0.2938   0.85227]                      \n",
      "[18.       0.40779  0.2982   0.86364]                      \n",
      "[19.       0.37931  0.28546  0.86079]                      \n",
      "[20.       0.40488  0.28701  0.875  ]                      \n",
      "[21.       0.39238  0.29027  0.87216]                      \n",
      "[22.       0.40603  0.32387  0.86932]                      \n",
      "[23.       0.41828  0.29644  0.84943]                      \n",
      "[24.       0.38979  0.29623  0.86932]                      \n",
      "[25.       0.37024  0.28652  0.86932]                      \n",
      "[26.       0.35854  0.28684  0.86932]                      \n",
      "[27.       0.36355  0.28748  0.86648]                      \n",
      "[28.       0.3571   0.2839   0.86364]                      \n",
      "[29.       0.35536  0.29211  0.875  ]                      \n",
      "[30.       0.36245  0.2827   0.87784]                      \n",
      "[31.       0.39766  0.30764  0.83807]                      \n",
      "[32.       0.39804  0.31971  0.83807]                      \n",
      "[33.       0.39399  0.29782  0.86364]                      \n",
      "[34.       0.38578  0.30637  0.84943]                      \n",
      "[35.       0.36922  0.29556  0.85511]                      \n",
      "[36.       0.39914  0.29789  0.86079]                      \n",
      "[37.       0.37792  0.28813  0.86364]                      \n",
      "[38.       0.39377  0.29142  0.86079]                      \n",
      "[39.       0.39544  0.30298  0.85511]                      \n",
      "[40.       0.38139  0.3125   0.84375]                      \n",
      "[41.       0.35765  0.29868  0.85511]                      \n",
      "[42.       0.3501   0.28783  0.86079]                      \n",
      "[43.       0.37781  0.33559  0.85227]                      \n",
      "[44.       0.36207  0.32437  0.84091]                      \n",
      "[45.       0.36895  0.30215  0.85227]                      \n",
      "[46.       0.3699   0.29523  0.86364]                      \n",
      "[47.       0.342    0.30321  0.85227]                      \n",
      "[48.       0.3369   0.29896  0.86079]                      \n",
      "[49.       0.32657  0.29649  0.85511]                      \n",
      "[50.       0.34544  0.29194  0.86648]                      \n",
      "[51.       0.33682  0.31144  0.86648]                      \n",
      "[52.       0.37716  0.30049  0.875  ]                      \n",
      "[53.       0.35752  0.30651  0.86648]                      \n",
      "[54.       0.34597  0.29794  0.86932]                      \n",
      "[55.       0.34755  0.30453  0.86079]                      \n",
      "[56.       0.35721  0.29731  0.85511]                      \n",
      "[57.       0.36477  0.30911  0.86079]                      \n",
      "[58.       0.35665  0.30496  0.86364]                      \n",
      "[59.       0.35425  0.30419  0.86364]                      \n",
      "[60.       0.36197  0.30969  0.86364]                      \n",
      "[61.       0.38011  0.3094   0.86364]                      \n",
      "[62.       0.36017  0.30173  0.85796]                      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "#learn.fit(lr, 4, cycle_len=2)\n",
    "learn.fit(lr, 6, cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f414d590b38>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f4164089c88>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.29436033869896056, 0.8753894080996885)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_preds,y = learn.TTA()\n",
    "preds = np.mean(np.exp(log_preds),axis=0)\n",
    "(metrics.log_loss(y,preds), accuracy(preds,y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lr/10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lr/9"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lr/3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# lr/4"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Unfreeze Layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# unfreeze layers\n",
    "learn.unfreeze()\n",
    "#lrs = np.array([lr/15,lr/12,lr/9])\n",
    "lrs = np.array([lr/9,lr/3,lr/1])\n",
    "\n",
    "#wd = 0.025\n",
    "#wds=[wd/100, wd/10, wd]\n",
    "#wds=[0.0025,0.025,0.25]\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run Learner Again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "d78c1f8868844e3f8c3a8c5ee92b3492",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=63), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.51119 0.39686 0.80114]                          \n",
      "[1.      0.41433 0.31581 0.86079]                          \n",
      "[2.      0.34278 0.27269 0.88352]                          \n",
      "[3.      0.33222 0.24953 0.89204]                          \n",
      "[4.      0.31777 0.25916 0.87784]                          \n",
      "[5.      0.28662 0.25582 0.87216]                          \n",
      "[6.      0.24937 0.28932 0.875  ]                          \n",
      "[7.      0.24128 0.24823 0.88636]                          \n",
      "[8.      0.26963 0.23062 0.90057]                          \n",
      "[9.      0.24302 0.34074 0.87216]                          \n",
      "[10.       0.20515  0.25601  0.87784]                      \n",
      "[11.       0.18742  0.2276   0.90341]                      \n",
      "[12.       0.20606  0.23921  0.90909]                      \n",
      "[13.       0.16345  0.24911  0.90057]                      \n",
      "[14.       0.15015  0.25653  0.89204]                      \n",
      "[15.       0.1619   0.24322  0.89773]                      \n",
      "[16.       0.16295  0.34569  0.85227]                      \n",
      "[17.       0.16717  0.29879  0.89204]                      \n",
      "[18.       0.13734  0.27496  0.90057]                      \n",
      "[19.       0.19684  0.37637  0.85511]                      \n",
      "[20.       0.34547  0.28914  0.86364]                      \n",
      "[21.       0.35523  0.34343  0.82671]                      \n",
      "[22.       0.31149  0.32625  0.84091]                      \n",
      "[23.       0.29557  0.62666  0.75   ]                      \n",
      "[24.       0.26373  0.29817  0.86648]                      \n",
      "[25.       0.25974  0.34863  0.84943]                      \n",
      "[26.       0.29316  0.39982  0.80966]                      \n",
      "[27.       0.26724  0.38206  0.80966]                      \n",
      "[28.       0.27338  0.37533  0.81818]                      \n",
      "[29.       0.2805   0.42562  0.80682]                      \n",
      "[30.       0.26991  0.34546  0.82102]                      \n",
      "[31.       0.29634  0.49272  0.78409]                      \n",
      "[32.       0.25314  0.44648  0.82386]                      \n",
      "[33.       0.21112  0.41405  0.82671]                      \n",
      "[34.       0.23994  0.3697   0.83807]                      \n",
      "[35.       0.25962  0.34211  0.85511]                      \n",
      "[36.       0.28556  0.33231  0.87216]                      \n",
      "[37.       0.25013  0.3017   0.86079]                      \n",
      "[38.       0.20309  0.29498  0.87216]                      \n",
      "[39.       0.21233  0.37224  0.85796]                      \n",
      "[40.       0.23568  0.23745  0.88921]                      \n",
      "[41.       0.21293  0.33087  0.85796]                      \n",
      "[42.       0.16585  0.30983  0.87784]                      \n",
      "[43.       0.17554  0.32414  0.87784]                      \n",
      "[44.       0.16752  0.26692  0.89489]                      \n",
      "[45.       0.13979  0.39282  0.86079]                      \n",
      "[46.       0.14255  0.40794  0.85796]                      \n",
      "[47.       0.16094  0.4354   0.87216]                      \n",
      "[48.       0.18868  0.32274  0.88068]                      \n",
      "[49.       0.17607  0.29463  0.88636]                      \n",
      "[50.       0.15747  0.43814  0.84943]                      \n",
      "[51.       0.17167  0.31115  0.87784]                      \n",
      "[52.       0.13538  0.31779  0.88068]                      \n",
      "[53.       0.10884  0.31765  0.88352]                       \n",
      "[54.       0.10098  0.39534  0.875  ]                       \n",
      "[55.       0.09869  0.37504  0.89204]                       \n",
      "[56.       0.08689  0.43664  0.86648]                       \n",
      "[57.       0.11017  0.41909  0.87784]                       \n",
      "[58.       0.11643  0.31197  0.89489]                       \n",
      "[59.       0.08398  0.419    0.87216]                       \n",
      "[60.       0.13489  0.41671  0.86932]                       \n",
      "[61.       0.14833  0.38853  0.89489]                      \n",
      "[62.       0.11285  0.41106  0.88068]                      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit(lrs, 6, cycle_len=1, cycle_mult=2)\n",
    "#learn.fit(lrs, 5, cycle_len=1, cycle_mult=2,use_wd_sched=True,wds=wds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZIAAAEKCAYAAAA4t9PUAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJztnXmYHGW1/z+n99kz07MkmZkkk8lkh5CFkLCHBAiIBBAErgsqylVRFrcL+nO94hVFuYqg4gVFFAFRJAIGkC1AIBskISHbZJ9sk2SyT2Z/f39UdWgmvVSv1cv7eZ5+pqe6qvpUz3SfPud833NEKYVGo9FoNPHisNsAjUaj0WQ32pFoNBqNJiG0I9FoNBpNQmhHotFoNJqE0I5Eo9FoNAmhHYlGo9FoEkI7Eo1Go9EkhHYkGo1Go0kI7Ug0Go1GkxAuuw1IB5WVlWrYsGF2m6HRaDRZw9KlS/cqpaqs7JsXjmTYsGEsWbLEbjM0Go0maxCRLVb31aktjUaj0SSEdiQajUajSQjtSDQajUaTENqRaDQajSYhtCPRaDQaTUKk1JGIyGwRWSsizSJyW4jHvSLymPn4QhEZFvTY7eb2tSJyYdD2zSLyrogsExEtxdJoNBqbSZn8V0ScwL3A+UALsFhE5iql3gva7Xpgv1JqhIhcA9wJXC0iY4FrgHHAYODfIjJSKdVrHjdDKbU3VbZrNBqNxjqpXEcyFWhWSm0EEJFHgTlAsCOZA3zPvP8E8CsREXP7o0qpTmCTiDSb53szhfZa4qll25ne6Ke6xBd2nz2HO3lx9W4um1iLz+1M6PmUUjy6eBtVxV5mjqnGeHkyg9ZDHfzt7e34izxMb/RTX1Fot0kZj1KKP721hSOdvRR5nVQWexnqL6ShsohCT14s69LkIKn8z60FtgX93gKcFm4fpVSPiBwE/Ob2t/odW2veV8DzIqKA3yql7g/15CJyA3ADwJAhQxK7EpMdB45x86PLGFJRyPxvzAi731PLtvPDZ1azaFMbP7/6lISec+fBDm7/+7sAfHjCYO7+6ARczswobf1j2XbunLfm+O9njqjkG7NHcXLdAButymxa9h/j20+tOmG7Q2Dc4DJOHVbBzDHVnNZQkTF/Z40mGql0JKG+OiuL+0Q69gyl1A4RqQZeEJE1Sqn5J+xsOJj7AaZMmdL/eeOit884zda29oj79Zj7/f2d7XzrQ2PwF3sTfs6GyiL+uXwHTdXF3DSzKe7zJZPePuPn018+k1fX7eH3b2zi8vsWcNN5Tdw0c0RGRU+ZQp8y/p7/c8VJnD+2hl0HO9iyr501uw6xeHMbf164hQff2IS/yMOHJwzmk9OHMryq2GarNZrIpNKRtAD1Qb/XATvC7NMiIi6gDGiLdKxSKvCzVUSexEh5neBIksH/vbaRl9a08sjnpp3wWHtXj6VUxKvr9nDFpLqEbblxxgjmr9vDPS+t54pJtdSVZ04aaUR1MeNry/j4tKF896mV3P3vdWzed5SfXnmy/lYdBq/LQWWxl8piL+Nry/jQyYMAONbVyytrW3l6xU7+vHALf1iwmXNHVXHTzCYmDSm32WqNJjSpfJcvBppEpEFEPBjF87n99pkLXGfevxJ4SSmlzO3XmKquBqAJWCQiRSJSAiAiRcAFwMpUXcAPn1nNgg37Qj6240BH1ON9bgdvNIc+Ph5uv3g0IsI9LzYn7ZzJpKzAzd1Xn8JXzx/Jk+9s5/a/v4tSSQkG84YCj5OLThrEvR+bxBu3nccts5p4t+UgV9y3gBv+uITm1sN2m6jRnEDKHIlSqgf4EvAcsBp4XCm1SkR+ICKXmrs9APjNYvpXgNvMY1cBj2MU5ucBN5qKrRrgdRFZDiwCnlFKzUvVNUSi9VB0R3Jag58lW9qS9pyDygr46JQ6nly2nf1Hu5J23mQiInx5ZhM3zWzir0tb+N9/r7fbpKylusTHLbNGMv8bM/jq+SN5c8M+LvrFa/zs+bV0dPdGP4FGkyZSmndQSj2rlBqplGpUSt1hbvuOUmqueb9DKXWVUmqEUmpqQOFlPnaHedwopdS/zG0blVITzNu4wDlTTXegGBDE7sPRHcm4waVs2dfOkc6epNnysdOG0tXTx9/f2Z60c6aCW2c18ZFJdfzypfXMX7fHbnOymiKviy/PbOKVr5/Lh08ezD0vNXPRL15j6Zb9dpum0QB6Zbsl2ruMb3/BWZrdhzrD7h/Yb+zgUgDW7DwU93P3zwyNGVTKKfUDeGJpS9znTBbqBO3E+4gIP7xsPE3Vxdz62DL2Hgn/euUTiWT6/MVefn71Kfzp+tPo7u3jo799k3tfbj4uyNBo7EI7Egu0d50YUbRGcCQBxg4yHMnqBBxJgGD90yUnD2L1zkNs3RdZPWY3BR4nv/qPSRzq6Oa/n34v+gEaS5zZVMmzN5/FReMH8tPn1vLJBxfSlqGpTk1+oB1JBAo9xmLCo50n5qMPHIv+xh1UVkCJ10Vz65Gk2nXhuIEAzFu1M6nnTQUja0r44rkjeGrZDl5Z22q3ORlDosroUp+be66dyJ0fOYnFm/dz2b1vsG63LsRr7EE7kggEHEmoiOTQse6ox4vAEH8hW6KsO4mV+opCxteW8tyq3Uk9b6r44oxGGquK+O7cVXT1nFhv0sSHiHD1qUN47IZpHOvu5Yr7FvCydtYaG9COJAIFkSKS9uiOBGCovzAlKagZo6pZtu0Ahzus2WEnXpeTb18yli372vnzQsvTOzUWmTiknKduPIOh/kI+99ASnlqW2UIMTe6hHUkEvC7DkXSFUG0dtBCRAAypKGLb/vakF0RPb6ykt0+xcGPy5MWp5JyRVZwxws8vX1zPoSxwftnG4AEFPHrDNKYMK+eWx5bx8Jub7TZJk0doRxIBj7kqu386xuN0cMCiIxnmL6S7V7HjwLGk2jZp6AB8bgevN2dHE2QR4faLxrC/vZv7X90Y/QBNzJT43Pzh01OZObqabz+1it+8usFukzR5gnYkEfC4PuhIAnLXskJ3xIgkWBY7xG+0MonWn8vKuYLxupxMbfDb6khilbKOry3jovEDeWjB5ryNSlIt1PW5nfz645P58ITB/Phfa3jw9U0pfkaNRjuSiBx3JL0frJGUFbjp6umztLq43uyJ1bI/sTpJKJXP9OF+mluPsC+L1mjcOGMEhzt7ePhNXStJFW6ng59/dAIXjqvhB0+/xyMLt9ptkibH0Y4kAl5X6NTWgAI3YK3gXlNqzC3ZdTD5H/aThxpN/N7ZeiDp546FWKSs42vLOHdUFQ+8voljXfnb5kNCNrhOHm6ng3uuncSMUVV86x/v8o8M74SgyW60I4mAO0yNpMRndP21opjymF1edx1Kbo0E4OS6MlwOYenW7GqV8aUZI2g72sXjS7ZF31kTNx6Xg19/fDLTGvx8/YnlLMiSepom+9COJAKBYntnP0dS5DUcidUeWoPKfOw8GL03V6z43E7GDS7l7SzruTRlWAWn1A/goQWb6dPtPVKKz+3kN5+YTENlEf/58FLW7Eq8y4JG0x/tSCLwfo3kg46k2HQk7RZTMzWlPnalwJEATBpazvKWAyEbS2Yynzp9GBv3HuU1/S055ZQVGGquQq+TT/9+MTsPJj861uQ32pFE4ATVlvnlOZ6IZJeFtvPxMGlIOR3dfazZmV3tMS4+aRCVxV4eWrDZblPygsEDCvj9p6ZyuKOHz/1xiW5Dr0kq2pFEwGHWQ/vXSIqORyShHUl/WezAMh8H2rvjKi5Hk9ieUm/MR1+x3d6Ce6x4XA7+47QhvLy2lc17j9ptTtqwc9DX2MGl/OKaU1i145AeOqZJKtqRRCDwPjvBkZitU46EaJ0STEDNNDCg3EogKgmnjKorL6DU52LVDvty3/EqkD522hAcIvxlUf7JU+0aZz9zTA23zjImWD74xmZ7jNDkHNqRWKB/jeR4RBJDagtISW5aRBg7uNRWRxIvNaU+Zoyq5u/vbKcny2o82cyXZozgwnE1/OjZ1byha1SaJKAdiQX6RyQF7kAzR2uOpLrUC8Cew6lZODh+cBlrdh7Kyg/jj06pY8/hTl7VUxTThsMh/OyjpzC8sogvPfK2Lr5rEkY7Egv0dyQOh5HeOmqx5uEvMhzJ3iOpGT40rraUzp4+NuzJvlrDjNHVVBZ79JqSNFPsdfGbT0yms6ePm/+yLCu/hGgyB+1ILNAZ4k1W6HVZjkjKCty4HJKycbPjBpcBsGrHwZScP5W4nQ4un1jLi6tb9TjeNNNYVcwPLxvPos1t/PLF9Xabo8litCOxwPtNG9+n2OuyHJE4HIK/2BNXTywruprhlUV4XY6010mSpfq5ako9PX0qL9p4ZJpO6opJdVw5uY57Xm7W9RJN3GhHEoHAm75/aksQirzOqBFJsJqpstibUGorkjLK5XQwZlBpVkYkYIzjnVBXxpN54EgykR/MGcfwyiJufnRZyup4mtxGOxIL9PSFSG15XJYXJELAkaTuTTpmUAlrdx22ZW1AMqSsl55Sy6odh9iwJ7nz7TXRKfS4uPdjkzjU0c3tf1+h15doYkY7Egt09574xir2usIuSAyFv9jD3hR+22uqLmF/e3fKCvqp5pKTByEC/1y+w25T8pLRA0v5xoWj+PfqVv66pMVuczRZhnYkFgilaCn0OEPOcg9HlZnaStW3vZE1JQCs351drVIC1JT6OK2hgrnLd+hvxDbxmTMamDa8gu//cxXb4hzEpslPtCOxQKh56wVuZ0z9iiqLvXT19nE4hnRYLIysKQZgXZY6EoAPTxjMxj1HeW9n9i2uzAUcDuGuqybgEOGrf10e8v9eowmFdiQWCJXaKvA4ORaLIynxAKQsvVVV4qWswM261uytMVw0fhAuhzBXp7dso668kO9eOo5Fm9r0mF6NZbQjiUAgwxIotgenXArczrBNGEOlZuJdlGg1zSMijKwpTmtqK9kZqIoiD2c2VfL08p05m97Khsv6yKRaLhhbw0+fX6vFDxpLaEdigZ5+EYmIMTCos6cv4mCmYDVTZXHAkcQXkVhRRjXVlLBu95G0fwgns//gJScPZvuBY6xoyU4ps1XErq6NFhAR7rj8JArcTm772wo9fEwTFe1ILNAdQv5bYHYA7j89MRz+YiO11XY0daqqUTUlHDzWTWsWrwWYNaYap0N4btUuu03Ja6pKvPy/D41h8eb9PJKH3Zk1sZFSRyIis0VkrYg0i8htIR73ishj5uMLRWRY0GO3m9vXisiF/Y5zisg7IvJ0Ku0P0BuiRuIzh15ZrZMMKHQDcKA9dY6kKQcK7gMKPZzWUKEdSQZw5eQ6zhxRyY//tUY3dtREJGWOREScwL3ARcBY4FoRGdtvt+uB/UqpEcDdwJ3msWOBa4BxwGzgPvN8AW4GVqfK9v50h1JtmRGJVUfidTkp9DjZ396dVNuCCUiA1+3O7rz2heMGsmHPUZqzWDiQC4gIP7r8JHr7FN/+x8qcrVtpEieVEclUoFkptVEp1QU8Cszpt88c4CHz/hPATDGSx3OAR5VSnUqpTUCzeT5EpA74EPB/KbT9A4RaR+IzW8nHMvWwvNDD/hSmtvxFHsoK3Gzam90fwBeMqwHQUUkGMMRfyFcvGMm/V7fy9IqddpujyVBS6UhqgeDe4C3mtpD7KKV6gIOAP8qx/wt8A0hb3+tAsT34+1hgJkmotSThvriVF7nZH2NqK5bvgCJCQ2URG9PUTj5V308HlRUwoX4Az+ekI8m+b/WfPqOBCXVlfP+fqzh4LHURtSZ7SaUjCSVL6f8uCrdPyO0icgnQqpRaGvXJRW4QkSUismTPnviGJinT3J4Iqa1IixL7X0R5oYe2FKa2wOgEvCkHZqBfOK6G5S0HdW4+A3A6DBVX29Eufv78WrvN0WQgqXQkLUB90O91QP+VZsf3EREXUAa0RTj2DOBSEdmMkSo7T0T+FOrJlVL3K6WmKKWmVFVVJXQhoZo2BiKSWBYllhd6UlpsBxheVcTOgx0x9QFLlFRIWS8cNxCA51ftTvq5M4HMFf+GZnxtGZ+YNpSH39rCyu25Lc3WxE4qHclioElEGkTEg1E8n9tvn7nAdeb9K4GXlFHRmwtcY6q6GoAmYJFS6nalVJ1Saph5vpeUUh9P4TUAxsr2/oXG+Gok7pTWSAAaKg3lVrZHJY1VxTRWFfHCe7npSLKRr1wwiooiL//vHyv12hLNB0iZIzFrHl8CnsNQWD2ulFolIj8QkUvN3R4A/CLSDHwFuM08dhXwOPAeMA+4USll/RM7BfTvOxRwJB0W15EAlBd5ONTRk9Kxpg2VRUD2OxKAmWNqWLhpX0zt+jWpo6zAzTcvHs2ybQf0aGTNB0jpOhKl1LNKqZFKqUal1B3mtu8opeaa9zuUUlcppUYopaYqpTYGHXuHedwopdS/Qpz7FaXUJam0P5j+dZLjNZIYVVsAB1JYsAw4knQV3FPJjFHVdPcqXl+vJ/dlCpdPrGXqsArunLcm5dG1JnvQK9stcoIjiadGUmQ6khTWSQo8TgaX+XIiIpkyrJwSn4uX17TabYrGRET4wWXjONTRw0+eW2O3OZoMQTuSSAT5jp7evg/IeiM5knDZ43JzdXvbUesRSTxrwBqqitiYhmZ7qV6f5nY6OLupipfXtubMYrhcuIzRA0v59OnD+MuibaxoOWC3OZoMQDsSiwS3khcRvGaLlIjy335qpkBqK9a1JKHOFYnhlcVs3Hs0Jz58zx1VRevhTlbtyK0ZJRncs9ESN89qorLYww/++V5O/J9pEkM7Eov0lwA7HIYziSe1lXrlVhGHO3rYl6Ycdio/E88dVQ3ASzq9lVGU+Nx87YJRLNmyX69412hHYpX+reTBqEfEVmw3Ulup7LcFRmoLcqPgXlXiZUJdmXYkGchVU+oZM6iUH/9rTUzTQjW5h3YkFgm5ut0d25TEArcTr8uR8kWJjeZaknTUSdLBjNHVLG85wL44Z7loUoPTIXznkrFsP3CM383fGP0ATc6iHYlFQq39MByJ9TUhImK0SUlxyqm2vACXQ9i8rz2lz5MuzhtdjVLwytr4Wt1oUsf0Rj+zxw3kvlc2sPtQh93maGxCO5IIBMcgRrH9g1GJ1+2MqWkjGHNJYltHEnsh0+kQ6soL2NqW2tSWSlMDwvGDy6gs9vLKuux3JLlYlv7mxWPo7VP8ZJ7uw5WvaEdikeBie6C4XOB2xNS0EaC0wB1XB9VYC9pD/EVsbcuNiMThEM5uquSN5r26NUcGMsRfyGfObOBvb7ewfJuWA+cj2pFYJFwH4Fh6bYHRZuJQGlpxD60oZMu+9rRIM9MhZT1rZCVtR7t4b2duyIAl69o2RubGGY1UFnv576e1HDgf0Y7EIqFUWz5XbMV2MBxJOmY6DPUXcrijhwMpVoilizNGVAIwf332p7dykRKfm1vPb2LJlv260WYeoh2JRcJNSeyKoWkjpM+RDKkoBGBLjqS3qkt8jB5YwmvrdN+tTOXqKfUMryriznlrUtqYVJN5aEdikVBz270uBx09sUck7V29dKf4jTbUb6wlyZU6CcDZI6tYsqUtrbNWNNZxOR381+zRbNhzlMeXtNhtjiaNaEcSgeBcb6hvWF63g84Y5L9gOBIg5VFJICLZui/7FyUGOKupku5excKNbXabognDBWNrmDy0nLv/vU47/DxCOxKL9PSpE2S9XpeTzhCprUiy2FgdSbx1ywKPk+oSL1tSuJYk3TXVU4dV4HU5srpOkut1aBHhmxePZs/hTh54bZPd5mjShHYkFun5QNNG46fX7aAzQmorlJqprDC+iCQeZdRQf2HO1EjAqElNbajgNT2fJKOZPLSCC8fV8JtXN7BXdyPIC7QjsUioue1el5OO7r6Y5I7pSm0B1FcUsjUNq9tTMbM9HGc3VdHceoSdB4+l7TlTQbZ3/43GN2aPpqOnj3teXG+3KZo0oB1JFAJv+O4Q8t9AK/muGArnAUeSnrUkRew61JFTDfXOGmnIgHVUktk0VhVzzan1/HnhVjbnwJA1TWS0I4mC22G8RCGL7aYjCVUnCUc6I5KhfqPg3rI/d9Jbo2pKqCrxakeSBdw8qwmPy8FPn9OtU3Id7Uii4HQYIUmole1ec0piLMqt444kDQsFh5iOJJUF93QjIpzVVMnr6/fodikZTnWJj8+eNZxn3t3Jyu0H7TZHk0K0I4mAAtxO05GEWpBoYUpif9xOB4UeZ4yNG+NjaEXuORKAM0dUsr+9m9W7cqNdSi7z2bMaGFDo5q7ndVSSy2hHEgW300xt9Z0o6j0ekfRLbUWrvceyuj2R79wVRR6Kva6ULUq0Kx6Y3ugH4M0N+2yyIH7S1TE5Uyj1ufnCOY28snYPizfr9T+5inYkUQiV2go03Hu/RhI6IgmnZoqnTUo8Tf5EhCEVhTm1uh1gUFkBwyuLWJCFjiRAjou2PsAnpw+jqsTLT+et1Q0dcxTLjkREilJpSKYSiEh6w7RIgdiK7RB/K/l4GFJRyJYcWt0e4PQRfhZu3JfyVjOaxCnwOLnpvBEs2tzGfC2SyEmiOhIROV1E3gNWm79PEJH7Um5ZhhCISEJ9YPnM1Fas8tp0tZIHqK8oYPuBYzn3TfD0xkqOdvWyokUXcbOBq08dQl15AXc9p6OSXMRKRHI3cCGwD0AptRw4O5VGZRIOMW7JjEjS1QEYjEWJHd197D2S2vG+6Wba8ECdRH/DzQY8Lge3zBrJu9sP8tyqXXabo0kyllJbSqlt/Tblzgq3CAS+OLkcjtDyX1fs8l9IryOpKy8AYFsOrSUBQ0gwZlBpVtdJ8o3LJ9bSWFXEz55fF/KLmSZ7seJItonI6YASEY+IfA0zzZUPiAhOh9DT23di00Z36GJ7tLdILK3kE80C1JcHFiWmoKWIzSmKMxr9LNmyP6tW7udzVsfpEL56wSjWtx7hqWXb7TZHk0SsOJLPAzcCtUALcArwxVQalWm4nPJB1ZYpufHFsSAR4lvdHm9vptpARJJjyi0wCu5dPX28vWW/3aZoLDJ73EDGDS7lf/+9PuahcJrMxYojGaWU+phSqkYpVa2U+jgwJtWGZRIuh4QctRtN/huOdLZJKfS48Bd5UhORYG/zwVOHVeB0SFamt3K9aWM4HA7haxeOYmtbO48v6Z8x12QrVhzJPRa35SwuZ7gaSZzF9jhbycdLXUVhTvXbClDic3NyXRkLdME9qzh3ZBVThpZzz0vrsyotqQlPWEciItNF5KtAlYh8Jej2PcBp5eQiMltE1opIs4jcFuJxr4g8Zj6+UESGBT12u7l9rYhcaG7zicgiEVkuIqtE5PsxXm9cuBxCb5g28hCfagvS6EjKC1IWkdjN6Y1+lrcc5EinnsaXLYgYtZLdhzp5ZOFWu83RJIFIEYkHKAZcQEnQ7RBwZbQTi4gTuBe4CBgLXCsiY/vtdj2wXyk1AkNmfKd57FjgGmAcMBu4zzxfJ3CeUmoCRq1mtohMs3apsROIQZxhUltupyAS3zoSSE8reTAK7tv3H8vJJodnNFbS26dYvEm338gmpjf6mT7cz69f3aCjkhwgrCNRSr2qlPo+ME0p9f2g28+VUlam1UwFmpVSG5VSXcCjwJx++8wBHjLvPwHMFKOvyBzgUaVUp1JqE9AMTFUGR8z93eYtpZ+OgrG6PVRqS0TwhRm3Gwk7IpKu3j5aD+fetLpJQ8vxuBy80azTW9nGreePZM/hTv701ha7TdEkiJUaSbuI/FREnhWRlwI3C8fVAsHVtBZzW8h9lFI9wEHAH+lYEXGKyDKgFXhBKbUw1JOLyA0iskREluzZk9iMb6dD6OnrC9lwz+t20Nn/G1UUjWepz3or+WQ0+as3uwAney1JJsQ3PreTyUPKs6bgns/y3/5MbajgjBF+fvPqRo516agkm7HiSP4MrAEagO8Dm4HFFo4LpUvp/zYKt0/YY5VSvUqpU4A6YKqIjA/15Eqp+5VSU5RSU6qqqiyYG57+qq1g47wuR8iIJJIqx+NyUOB2cqgjBvmv5T1PJLAoMRUF90wQH53e6Oe9nYfYfzSbVu9nwitnP7fOGsneIzoqyXasOBK/UuoBoNtMd30GsFKXaAHqg36vA3aE20dEXEAZ0GblWKXUAeAVjBpKSnE5JexKXGNue+zfptK5ur12QGAtSY4W3EcY7VLe2pgdUYnmfaYMq+Cspkp+8+oG2ru0YCJbseJIAp92O0XkQyIyEeODPRqLgSYRaRARD0bxfG6/feYC15n3rwReUkZHt7nANaaqqwFoAhaJSJWIDAAQkQJgFka0lFKcDgfdYRyJzx06IolGaYErbY7E53ZSXeLNSQkwwMl1Ayj0OHlTO5Ks5JZZI9l3tIs/vqmjkmzFZWGfH4pIGfBVjPUjpcCt0Q5SSvWIyJeA5zDkwg8qpVaJyA+AJUqpucADwMMi0owRiVxjHrtKRB4H3gN6gBuVUr0iMgh4yFRwOYDHlVJPx3jNlgl0KXWHkf+CEZHE40iMDsDp+wZWX1GYsxGJ2+ng1GEVWVMn0XyQyUPLOXtkFffP38gnpg2lyGvlY0mTSUT8i5kf2E3mh/VBYEYsJ1dKPQs822/bd4LudwBXhTn2DuCOfttWABNjsSFhxCi2d4eQ/0KgRhJfamvHgY5ErbNMXXkBb2/N3VYi0xv9/Phfa2g93EF1ic9uczQxcuusJi6/bwEPvbmZL547wm5zNDESMbWllOoFLk2TLRmL2+mgt0+FVNx43Q46+vXasiLMKfVZq5EkS+VTX17IjgMdIWfPx0smKZCmDw/USTJ7PUm+jdq1ysQh5cwYZUQlh2MQoWgyAys1kgUi8isROUtEJgVuKbcsgzDkvyc2bQTMdSQnRiTRNDmlBe7YVFsJinzqygvo7VPsOpS+KCidjBtcSonXlZVz3DUGt8wayYH2bh5asNluUzQxYiUZebr58wdB2xRwXvLNyUxcZhv5UBjrSOKrkRzu6KG3Tx2fwphK6sx28tvajh2/nwzCzaVPNy6ng9OGV2TNoKsMedkyign1A5g5uprfvbaJT54+7Ph6K03mEzUiUUrNCHHLGycC0eW/8am2jDfJkY70FNzrK1K3liRTmDbcz+Z97ew8mJuignzgllkjOXismz+8sdluUzQxYGlCYr6ILdKqAAAgAElEQVTjcjjCDqHyuhxxryOB9LVJGVRWgAhsy9HmjWAU3AGd3spiTqorY9aYGv7vtY1pe29oEkc7kggEYpDIEUmc60h8RlYxXW8Wj8vBoFJfTkckYwaWMqDQrR1JlnPLrCYOdfTw4Oub7DZFYxHtSKIgRJb/+tyhi+3RON4BOI0KlbryQlpydC0JGEOTpjX49XqSLGd8bRkXjqvhwdc3WepHp7GfqI5ERK4IcZspItXpMDATMOaRhJH/mhGJCnrQiizW6nCrZEps6yoKkhqRZKKUdXqjn+0HjmXsaOFMkkxnMrfMGsnhzh4eeH2j3aZoLGAlIrke+D/gY+btd8BXgDdE5BMptC1jOHFC4vuSG6/biVLQ1a+GEk3NdLwDsOXUVuIyn7ryQnYe6kjqrOxMEx9lS50k0163TGPMoFIuGj+Q37+xmQPt2dSMMz+x4kj6gDFKqY8opT6CMaSqEzgN+K9UGpcpuMw28qGIe9xumodbAdSXF6AUOa1qaqouprLYo/tu5QA3zWwyoxJdK8l0rDiSYUqp3UG/twIjlVJtvN/QMadxORz0hmuR4jbH7ca4lqTQ48TlkLQqU4LXkuQqIsK04X4WbNj7gXSjJvsYM6iUi0/SUUk2YMWRvCYiT4vIdSJyHfAUMF9EioADqTXPZszPIZdT6I4akcRWcBeRmFe3J0pgLUmyB1xlGtMb/ew+1MmmvUftNkWTIDfPHMnRrh5+95qulWQyVhzJjcAfMGakTwT+iNGN96hSKqYmjtmIiOB0RJb/Aif027KCMZMkfR2AB5b6cDokpyXAAKc3VgLo9FYOMGpgCRefNIg/vLE5ywaX5RdWVrYrpdQTSqlblVK3mPfzKmfgNntthRy16zJTW3FIgEt96ZtJAoZoYPAAHy05vCgRYJi/kIGlPi0DzhFuntlEe3evjkoyGKvy3/UiclBEDonIYRE5lA7jMgWnw4FSEMhuBQuyvO4Ti+1WZbGlBe6oxfZkS2zrBhQmTRqbqV8nRITpjX4Wbtyn6yQ5wMiaEj500iAeWrCZNh2VZCRWUls/AS5VSpUppUqVUiVKqdJUG5ZJuJyG5wil3PK5Qhfbrcg7yyw4kuPnS5JetL6iIKltUjK1+eD04X72HulifesRu00JSaY0u8wWAlHJ/fN1VJKJWHEku5VSq1NuSQbjcgQcSYjUlhmRdMST2krj3PYAdeWF7DncGVd/sGwiW9aTaKzRVFPCh08ezB/f3My+I512m6PphxVHskREHhORa4NXt6fcsgwgkFYKtHnvCSEBPq7airNx46GO7rSmX97vApzbdZL6ikLqygtYkCVt5TXRuWlmEx3dvdyvayUZhxVHUgq0AxcAHzZvl6TSqExCMCYkAiGVW+8X2+NTbXX3Ko6lMToIrCXJdeUWGOmthZva6AujuNNkFyOqi7l0wmD+uGALe3VUklFYUW19OsTtM+kwLlMIRCSh1pL4AsX2OOS/gTYph9IoAa4PLErM8YgE4PQRfg60d7N6V15pQ3KaL89sorNH10oyjbATEkXkG0qpn4jIPYQYQ66UuimllmUQbmek1NaJ8l+rmargmSQDy3wh90l21qu6xIvH6UhKRJLp3/OnDzfXk2zYx7jBZTZbY6BFZInRWFXMnFNq+eObm/ncWcOpKvHabZKGyBFJoMC+BFga4pY3OB3GyxQYtxustwlEJP0XJFoR5ZQWWJ9JkiyNj8Mh1JYX5HQ7+QADy3w0VBbpgnuO8eXzRtDV08f98zfYbYrGJGxEopT6p/nzofSZk5lEVG0lsCDRjsaNAHXlBUlrkyIZ3sd22nA/Ty/fQU9vHy5n5ozfyexXLbMZXlXMZRNrefitLXzu7OFUl4SO5jXpw8qCxJEicr+IPC8iLwVu6TDOblRQry0IXWx3OwWHxN8iBdI3JTFAXXlhzqu2Apze6OdwZw+rdug6SS5x03lNdPcqfvuqrpVkAmEjkiD+CvwGYyZJbi8+CIHI+xFJqLntIoLXFd+UxNhnkiSH+ooC2o52cbSzhyKvlX+B7GXacHM9ycZ9TKgfYLM1mmQxrLKIy06p5U9vbeE/zx5OdamOSuzESqzfo5T6tVJqkVJqaeCWcssyCJcjvPwXjDpJXHPbbRi3C8ES4NyPSqpKvDRVF+u+WznITTNH0NOn+PWrulZiN1YcyT9F5IsiMkhEKgK3lFuWQTidAflvuA7AzrhWijsdQok3vY0bwRhwBWTsONpkc3qjnyWb20JGlJrsZai/iCsm1vLIwq20Huqw25y8xoojuQ74OrCA9xVbS1JpVKZxvNge5oPI2y8iiUXhaUeblPqK5CxKzBYp6/RGP+1dvaxosX98TibOuc9mvnxeE719ivte0VGJnUR0JCLiAD6ulGrodxueJvsygv6prf4N93whIhKraiajA3D0BYnJbPLnL/JQ4HYmZ1FiFsiPTmvwIwILmjMnvaV7NiaHIf5CPjKpjkcWbWXXQR2V2EVER6KU6gPuSpMtGUd/1VZ32HG78dVIwJhJkm75r4gYEuA8SW2VF3kYPbBUD7rKUb503gj6+hS/fqXZblPyFiupredF5CMSx1diEZktImtFpFlEbgvxuNdsCNksIgtFZFjQY7eb29eKyIXmtnoReVlEVovIKhG5OVabYr4G5HhqqzfMuN1QEYlVytI8bjdAXXlBXhTbA5ze6Gfplv1xqes0mU19RSFXTq7jL4u26ajEJqw4kq9gSIA7YxlsJSJO4F7gImAscK2IjO232/XAfqXUCOBu4E7z2LHANcA4YDZwn3m+HuCrSqkxwDTgxhDnTDqB1FYqIpIyG2okYLz5cn12ezDTh/vp7Onjna3210k0yefGGSPoU4r7dFRiC1aaNpYopRxKKU+Mg62mAs1KqY1KqS7gUWBOv33mAIGV808AM83IZw7wqFKqUym1CWgGpiqldiql3jbtOozRxqXWyoUmQqQFiRBQbcWZ2rLJkdSVF3C4o8eW57aDqcMrcAhaBpyj1FcUctWUeh5dtI0dB/In0s4ULPWMEJFyEZkqImcHbhYOqwW2Bf3ewokf+sf3UUr1AAcBv5VjzTTYRGChlWtIBFeE7r8QiEhib9oIRkTS3tUbVpqaKmXU8S7ACdRJskmBVOpzc1JtGW/Z7EiyRemWjdw4oxGFjkrswEqLlM8C84HngO+bP79n4dyhair930bh9ol4rIgUA38DblFKhUyzicgNIrJERJbs2bPHgrnh6T/Yqr9xPpfzxDbyFitKVvttJVvkk0+LEgNMa/Tzzrb9HOvSdZJcpK7ciEoeW6yjknRjJSK5GTgV2KKUmoERBVj5ZG4B6oN+rwN2hNtHRFxAGdAW6VgRcWM4kT8rpf4e7smVUvcrpaYopaZUVVVZMDc8kQZbwYkRSSzE0gE4mbw/KTGxOkk2qVinD/fT3atYumW/3aZo+W+KuHHGCADufVlHJenEiiPpUEp1gKGyUkqtAUZZOG4x0CQiDSLiwSiez+23z1yMBY8AVwIvKWPu7FzgGlPV1QA0AYvM+skDwGql1M8t2JAQ/Ufthks/hYxILHI8IulI33CrwPOWeF15IwEGOHVYBS6H6PG7OUztgAKuPrWex5dsy4spoJmCFUfSIiIDgH8AL4jIU5wYWZyAWfP4EkYqbDXwuFJqlYj8QEQuNXd7APCLSDOGOuw289hVwOPAe8A84EalVC9wBvAJ4DwRWWbeLo7hemNGxEKx3e2gI96IxKbGjSLmXJI8Sm0VeV1MqB+g15PkODfOGIEg3PuyXu2eLqK2flVKXW7e/Z6IvIyRfppn5eRKqWeBZ/tt+07Q/Q7gqjDH3gHc0W/b69iQTQnIf0PNIwHwuhx09yp6+9Tx6MUqds0kAUPpsmXf0bQ/r51MH+7n169u4EhnD8U53vk4XxlUVsA1U+t5ZOFWvnhu4/GWQJrUYVW1daaIfFop9SrwJmmQ3GYSzii9tnzuxIdb2SUBbtl/DJVHUqLpjX56+xSLN7fZbYomhXzx3BE4RLSCK01YUW19F/gv4HZzkxv4UyqNyjTcUbv/Gi9joE4Siyy2NIojSaXEtr68kPauXtqOdsV3giz0P5OHluNxOmwbv5uFL1lWMrDMx7VT6/nrkpa8i7rtwEpEcjlwKXAUQCm1AyhJpVGZRv+IpL/iJhCRBNdJrCa4fG4nHpcjapuUVKh86soDyq346yTZpj7yuZ1MHDLA9jnumT6iOBe4ccYIXE7h7hfW2W1KzmPFkXSZSioFICJFqTUpcwhkfNxRBlv1j0hipazAbVuNBMirVilgpLdW7TjIwfb8WNWfr1SX+vj0GQ08tXwHa3bpUcupxIojeVxEfgsMEJHPAf8GfpdaszILh0MQCd9rK1REEgulvvQPt4L3I5Jtbfmj3AKj4N6nYOEmrd7KdT5/diMlXhd3PbfWblNyGiu9tu7C6IP1N4z1I99RSt2TasMyDZdDUhyRpHcdCUCJz82AQnfe6e1PGTIAn9uhZcB5QFmhm/88p5F/r25l6RYtsEgVllRbSqkXlFJfV0p9TSn1QqqNykRcDgc94drIH1dtZVfjRjAK7kkZcJVFeF1OpgytsL1OokkPnz5jGJXFXu6ctzavFIrpJKwjCbSLD3Gz1EY+13A55Hivrf4EIpLjM0li/F+1ayYJBCTA+RWRgFEnWbPrcPyKNU3WUOhxcdPMESza1Mar6xLru6cJTVhHEmgXH+JmtY18TuFyynH5b3+lktd1YkQSi5op0kySVH+Bqq8opGX/MfrCpO0ikc3f7aY3+gF4K83pLf2N2B6uOXUI9RUF/PS5tXH9r2siYym1la8E/7s5HY4ICxL7RSQxUuozVFuR/sFTJbOtKy+gq6ePvUc64zo+W2WsJ9WWUeRx2pfeys6XLWvxuBzcOmskq3Yc4tmVO+02J+fQjiQKgQnDkYvtidVIygrc9Ck42pX+gvvxuSR5lt5yOx2c2lChC+55xJxTahlVU8LPn18X9kuhJj60I7GIyylhe20lGpHY2SYlsJZky778ciRgyICbW4/QeljP+c4HnA7haxeOYuPeozyxtMVuc3IK7UgsYhTbw0xITDAisWsmCRhzSUTy05Gc3lgJoNVbecSsMdVMGjKAX7y4Pu4vfpoT0Y7EIi6nI3z330RrJMc7AKc/teV1ORlcVpCX/YjGDi6l1OdKe8FdYx8iwtcvHM3Ogx08tGCz3ebkDNqRWMSK/DcQkcSqCYk0kyQd+pKh/kK2xDHgKtsVSE6HMLXBn9aIJLtfsdxgeqOf80ZX86uXm9mv5d9JQTuSCAR/TjodcnxBYn+lkojgdTno7A5u2mhdlvP+lMTwqa1UqqOG+oviTm1lW9PG/pze6Gfzvva0z/jO8pct67n9otEc7ezhly+tt9uUnEA7kigE3vCRUltgRCVxq7YK7RtuBTDMX0jb0S7bVtfbyekjjPUkrzfr8bv5RFNNCVefOoSH39zCpr35l9ZNNtqRWCRSaguMNinxDLYCKPa4ELGn2A5GRAKwNQ8L7qNqSqgp9eoVz3nIrec34XE5+Mm8NXabkvVoR2IRV1BqKxRet4OOOJs2OhxyfFGiHQyrNCTAm/Ow4C4inDOyitfX79VrC/KM6hIfnz+nkX+t3KUbOiaIdiQWcTkjRyReV/wRCURuk5JqhhxfS5J/jgTgnJHVHDzWzfKWA3abokkznz2rgeoSLz98ZnXWi0fsRDsSizgdkWskvgQiEjDWktjlSAo9LmpKvWzOw9QWwJlNlTgdwitrdXor3yj0uPjaBaN4Z+sBnn13l93mZC3akVjE7QjS2YSQ3ARHJPF8szE6AJ+4jiRd35KG+otirpHkyhe4sgI3E+sHpKVOkiuvWS7xkcl1jB5Ywp3z1iSUVchntCOJyPvveqcjsmCzf0QSqyy21BcltZVivegwf2FcNZJckbGeO6qKFS0H425eGSuS7brpHMLpEG6/eAxb29r544ItdpuTlWhHEoXA+93tjPxSZXONBIyIpPVwJ+02NI7MBM4dVQ3AfK3eykvOGVnFuaOq+OWL69lzOD1fJnIJ7UgsEmtEEivGuF37HMkwUwKcjz23AMYOKqWy2KPrJHnMty8ZS0dPLz99TsuBY0U7Eou4ojiSRCOS0gI3nT19tjWSG+rPb+WWwyGcPbKK19bvCTsuQJPbNFYV8+kzGnh8SQvLt2kFXyxoR2IRlzN6RNKZkGorepuUVBJwJPmq3AIjvbG/vZsVWgact3z5vBFUFnv53j9X6UmKMaAdiUWcjug1kkA0EY8y53i/rX7prXT9K5f43FQWe2KKSHLtbXZ2UxUOIcXprVx71XKLEp+b2y4azTtbD/DkO9vtNidr0I4kAsEOwR0UkYSKTbwuBx3BM9tjfK5SX+SZJOnQ+Az1F8XcdyiX1EflRR4mDinnpTWtKX+u3HnVco8rJtYyoX4AP563hiOd+Sk+iRXtSKIQ+JyMXmx30tXTF3d+vbzQA8D+o/YV3BsqY3ckucasMTW8u/0gOw+mtxuwJnNwOITvXzqOPYc7uUd3B7aEdiQWiVZsL/QYUxLjLZZXFBmOpM3G+QiNVcXsPtTJYZvqNJnA+WNrAPj3e7tttkRjJ6fUD+CqyXU8+PommlsP221OxpNSRyIis0VkrYg0i8htIR73ishj5uMLRWRY0GO3m9vXisiFQdsfFJFWEVmZStv744qyjiTgSNq74nMk/mLDkeyz1ZEYEuANe/I3KmmsKqKhsogXVqc+vaXJbP7rotEUelx868mVug9XFFLmSETECdwLXASMBa4VkbH9drse2K+UGgHcDdxpHjsWuAYYB8wG7jPPB/AHc1taiRaRFHiMGsexOB1JoceFz+2g7ah9i6FGVBcDsKH1iG022I2IcP7YGt7csDevIzMNVBZ7ue2i0Szc1MYTS1vsNiejSWVEMhVoVkptVEp1AY8Cc/rtMwd4yLz/BDBTjOrtHOBRpVSnUmoT0GyeD6XUfCDtPZ9dUVRbBW7Dzx1LYB1IRaGHNhtrJPUVhbidwoY9+etIwEhvdfcqPaNEw9VT6pkytJwfPbva1rRzppNKR1ILbAv6vcXcFnIfpVQPcBDwWzw2IiJyg4gsEZEle/bE94EQHMxGW0fyfmqrJ26BZ0Wx54SIJJ0RtdvpYKi/iGaLEUmuRvuThpRTUeRJSZ0kV1+zXMXhEH54+XgOd/Tw43+tttucjCWVjiTUJ2//t1G4fawcGxGl1P1KqSlKqSlVVVWxHPoBArPSg1VboSSvBaYjCaS24pHFVhR5w37rSZfMtrGqKKaIJBdlrE6HcN7oal5a00p3ioZd5ZBqOucZPbCU688yVrwv2qQHYIUilY6kBagP+r0O2BFuHxFxAWUYaSsrx6aVqDUSd2LFdgB/kYe2dnvD5xHVxWzZ156yD9Bs4fyxNRzq6GGx/uDQADfPbKJ2QAHffPJdunry+70RilQ6ksVAk4g0iIgHo3g+t98+c4HrzPtXAi8pQx4xF7jGVHU1AE3AohTaGhWr8t9EaiTlhR7ajtjrSBqriunpU3nbvDHAWU2VeF0OntcyYA2GGOa/LxtHc+sR7n252W5zMo6UORKz5vEl4DlgNfC4UmqViPxARC41d3sA8ItIM/AV4Dbz2FXA48B7wDzgRqVUL4CI/AV4ExglIi0icn2qriGYaPLf/qmtePAXezja1Wtb40YwHAmQ9wX3Qo+Lc0dV8ey7O3XPJQ0A542u4bJTBnPvy828t+OQ3eZkFCldR6KUelYpNVIp1aiUusPc9h2l1FzzfodS6iql1Ail1FSl1MagY+8wjxullPpX0PZrlVKDlFJupVSdUuqBVF5DgOgRiSH/TWSeRyYsShx+fC1JfjsSgA+dPJjWw50s2bLfblM0GcJ3PzyOAYUevvbX5Xmf/g1Gr2yPQPAipKgRyXH5b1/cypzQjiS934ZLfG5qSr2WlFsqxxsQzhxdjc/t4JkVySvP5fYrlvuUF3m44/LxvLfzEL9+ZYPd5mQM2pFEIaCuCY5IQsUmPrcDEThmRiTxiHIiRSTpFPmMqC62LAHOZYq8Ls4bXc2zK3fpGSWa41w4biCXThjMPS+tZ/VOneIC7UgsE61po4hQ4HYmpNrKhNQWwKiaUtbvPmLtwzPHZawfOmkwew53Jl32Kbn+wuU437t0HGUFbp3iMtGOxCLuKAsSwVBuJaLa8meIIxk9qIRj3b1sbctv5RbAjNFVFLidPPOurepzTYZRUeThh5eNZ9WOQ/zi37pDsHYkFok22AqMVvKJqLZKfW6cDrHdkYwZWArAGh22U+hxcd6Yauat3EWP/uapCWL2+EFcNbmOe19pZuHGfXabYyvakVgkWosUMCKSRFJbDodQXui2tQMwQFNNMQ6B1bt0+2yAD588iL1HunhjQ35/WGhO5LuXjmNIRSG3PrYs7FC6fEA7EotEk/+C0QG4PcE1IP4iL/uO2NcBGIzIqqGySEckJjNGV1NW4OZvugOsph/FXhe/uGYirYc7+daT7+Ztu3ntSCIQ/C8RrdgOUOh20tHVm5AstrrUS+vh9x2JXf+XoweVsiZKRJIv7xmvy8mlEwbz3KpdHEqwtXy+vGb5xCn1A7j1/JE8vWInf3s7P+e8a0cShYD7cAetIwnXcK/A46S9u+eDB8ZIdYmPPYdPjEjS3eRvzMAStra1R51ZnS/ao49MrqOzp49nVuxMyvl008bc4vPnNHJaQwXfeWol63fnX0pYOxKLWIlIChKskUAgIumwPUQebRbc1+o6CQAT6soYUV2s01uakDgdwi+vnUihx8nn/7Q06hewXEM7Eou4Lai2AqmtRKgu8dLdq9jfbm/hbsxgw5HoBVcGIsJHJtWxZMt+Nu/N31HEmvDUlPq459pJbNp7lP/62wrbvwymE+1ILGJFtVXkdXE0QUdSU+oDYPehjoTOkyiDy3yUF7p5t+WgrXZkEpdPrMUh8Nel26LvrMlLpjf6+cbs0TyzYie/f2Oz3eakDe1ILOJxRX+pir0ujnT2JFRQrS7xAnyg4G4HIsKE+gEsbzlgqx2ZxMAyH+eNruaxxdvo7LGvQ7Mms/nPs4dzwdgafvTsat7Kk/Ul2pFEINgheK04Ep+L3j6VUBv46hIjImm1OSIBOLluAOt2H06oo3Gu8Ynpw9h7pIt5K3fZbYomQxER7vroBIb6C/n8n5bmRSpUO5JomPIar8v5/qYwWqUSn9FK/nBH/B+81aUfjEjszLKeUl9Gn4KV23WdJMBZIyoZ5i/k4Te3xHV8PuXN85lSn5sHP3UqAnzmocUctLnmmWq0I7GI1x39pSrxuQE41NEdtyzW53ZS6nOdEJHY0eTv5LoBAKyIkN5K1yz5TMHhED4+bShLtuxPaLhRfr1q+clQfxG/+fhktrW1c+Mjb+d0c0ftSCxiJbVV4k08IgGj4L77kL01EoDKYi+1AwpYtk3XSYK5cnIdXpeDPyzYZLcpmgzntOF+fnT5SbzevDenV75rR2IRT5TBVhCc2kosjB1Y5mPnwWMJnSNZTKgv046kHwMKPVw5uY4n39nOroP217I0mc1VU+q5aWYTjy9p4X/+tSYnnYl2JBaxksIJpLYSjUjqygtp2Z8ZjmTSkHJa9h/TH5j9+M+zG+ntUzz4ho5KNNG5dVYTn5w+lPvnb+TXr+beZEXtSCIQ6/eG4iQU2wHqygvYd7QrI9RS04b7AVi4KT9kjFYZ4i/kkpMH8+e3tuR8IVWTOCLC9z48jjmnDOYn89by8FvxiTUyFe1IohBLUTSQ2kpkuBVAfUUhAC37j9ne5G/MoFJKfK6QevhcDNFj4QvnNnK0q5eH3txs+Zj8fsXyG4dDuOuqCcwaU823/7GShxZsttukpKEdSRyEy3IVe1xB+8Svy6krLwCgZf/7EwrtEkc5HcJpDRW8tTH0qNk8E219gDGDSpk1pobfzd/I/lhnyOTx65bPuJ0O7v3YJC4YW8N3567i/vm5kebSjiSJOBxCsdcVfccovO9IMqNOMm24n017j+o6SQi+MXsUR7t6+NXLzXaboskSvC4n935sEpecPIgfPbuGu19Yl/XRvXYkSSaQ3kqEqmIvXpeDbRkyM316o1Enmb9+j82WZB4ja0q4cnIdD7+5JWP+XprMx+108ItrJnLV5Dp+8eJ6vv7ECrp6snediXYkSaaiyJPwOUSE+opCtuzLjA+msYNKGVTm48XVu+02JSO59fyRiMAdz6y22xRNFuF0CD+58mRunTWSJ5a28KnfL8pa4YZ2JBHoH25aqQdUFnuT8twjqoppbj2SlHMliogwa0wN89ftTaiPWK4yqKyAm2c1MW/VLp5fpXtwaawjItw8q4mfXTWBxZvb+PCvXmfl9uzruK0dSRSCnYeV1e3+4sQjEoCRA0vYvO9oxnSZnTW2hmPdvbzRvNduUzKSz501nNEDS/jOU6sSXpCqyT8+MrmOR2+YRndvH1f8egF/Xrglq+om2pHEgJVCerIiklE1JfQpMiYqmTa8glKfi6eW7Ti+LXv+zVOP2+ngf644idbDHXzzyZVhPwSy6LNBk2YmD63gmZvOYtpwP996ciWffWhJ1ghctCOJgYAjiZThqjQjkqMJjtocWVMMwDpz/rPdalGvy8llE2uZt2rXB/K4dtuVSUwcUs5XLxjFP5fv4E9RFpzZ0YRTk/lUFHn4w6dO5duXjOWNDXs5/+5X+cuirfT2ZfY3EO1IYiDQAiUS/iIjIulJ8A8/rLIIr8vBigyaUPjRKfV09fTx93f03PJwfOGcRs4dVcX3//keL69ttdscTRbicAjXn9nAvJvPZsygUm7/+7tccs/rvL4+c9PKKXUkIjJbRNaKSLOI3Bbica+IPGY+vlBEhgU9dru5fa2IXGj1nKkkIO3tjZCfCMwTSRS308HkoeWsSqBVebIZN7iUyUPLuX/+xoyp3WQaDofwy2snMmpgCV/401JdU9LEzbDKIh793DR+eW3LYNIAAAnZSURBVO1EDh3r5uMPLOTq377Ji6t305dhEUrKHImIOIF7gYuAscC1IjK2327XA/uVUiOAu4E7zWPHAtcA44DZwH0i4rR4zqTxWr9vAAFHciRCL63GquKkPf90s89VpiAi3DprJDsPduTVPOpYKfW5eegzUxlSUch1Dy7i4beyq3CqyRwcDuHSCYN58avn8P8+NIatbe1c/9ASZt39Kr96aX3GrF1KfPVceKYCzUqpjQAi8igwB3gvaJ85wPfM+08AvxKjt8gc4FGlVCewSUSazfNh4ZxJ40MnDeLcUVXHfy8NGlwVjkFlvqQ9/7mjqvnZC+uSdr5kcMYIPxeOq+Hnz69jqL/QbnMylspiL0984XS+/Mg7fPsfK/nn8h184dxGXA5dG9HEjs/t5LNnDee604fx9IodPLJwK3c9v467nl/H6IElTG/0M2VoBSNrihnqL8JjQWGaTFLpSGqBbUG/twCnhdtHKdUjIgcBv7n9rX7H1pr3o50zadz7sUkf+H3S0HL+urSFogjqrWRODBxfW4rH6aArgyariQj/c8XJXP3bN1nfesTSnJZ8pdTn5vefOpXHl2zjrufX8unfL7bbJE2W43Y6uHxiHZdPrGNbWzvPvLuT19fv5ZGFW49nCUSM/72yAjcDS308/vnpKbcrlY4k1Cdq//g+3D7htof61AqZMxCRG4AbAIYMGRLeyhi45tR6hvoLo6acnv7ymcfVVokgIrz89XN5aMFmJg4pT/h8yaKiyMMTXzid37+xibKC6AKEfMbhEK6ZOoQrJtXx0prdLN68n6OdPYyvLbXbNE2WU19RyOfPaeTz5zTS0d1Lc+sRmluPsHHvUQ62d3HwWDcFHmdabEmlI2kB6oN+rwN2hNmnRURcQBnQFuXYaOcEQCl1P3A/wJQpU5KSoBYRTm+sjLrf+NoyxteWJeMpqR1QwDcvHpOUcyWTsgI3t8waabcZWYPH5WD2+EHMHj/IblM0OYjP7Uzq506spDIvsRhoEpEGEfFgFM/n9ttnLnCdef9K4CVlVCXnAteYqq4GoAlYZPGcGo1Go0kjKYtIzJrHl4DnACfwoFJqlYj8AFiilJoLPAA8bBbT2zAcA+Z+j2MU0XuAG5VSvQChzpmqa9BoNBpNdCQfZIlTpkxRS5YssdsMjUajyRpEZKlSaoqVfbXkRqPRaDQJoR2JRqPRaBJCOxKNRqPRJIR2JBqNRqNJCO1INBqNRpMQeaHaEpE9QOQBEeGpBPKlhau+1twjX64T9LUmm6FKqarou+WJI0kEEVliVQKX7ehrzT3y5TpBX6ud6NSWRqPRaBJCOxKNRqPRJIR2JNG5324D0oi+1twjX64T9LXahq6RaDQajSYhdESi0Wg0moTQjiQMIjJbRNaKSLOI3Ga3PclARDaLyLsiskxElpjbKkTkBRFZb/4sN7eLiPzSvP4VIjIp8tntRUQeFJFWEVkZtC3maxOR68z914vIdaGey27CXOv3RGS7+bddJiIXBz12u3mta0XkwqDtGf8/LiL1IvKyiKwWkVUicrO5Paf+thGuMzv+rkopfet3w2hRvwEYDniA5cBYu+1KwnVtBir7bfsJcJt5/zbgTvP+xcC/MKZVTgMW2m1/lGs7G5gErIz32oAKYKP5s9y8X273tVm81u8BXwux71jz/9cLNJj/185s+R8HBgGTzPslwDrzmnLqbxvhOrPi76ojktBMBZqVUhuVUl3Ao8Acm21KFXOAh8z7DwGXBW3/ozJ4CxggIhk73k8pNR9jpk0wsV7bhcALSqk2pdR+4AVgduqtj40w1xqOOcCjSqlOpdQmoBnj/zsr/seVUjuVUm+b9w8Dq4FacuxvG+E6w5FRf1ftSEJTC2wL+r2FyH/UbEEBz4vIUnOmPUCNUmonGP/MQLW5PRdeg1ivLduv+UtmOufBQKqHHLpWERkGTAQWksN/237XCVnwd9WOJDQSYlsuyNvOUEpNAi4CbhSRsyPsm6uvAYS/tmy+5l8DjcApwE7gZ+b2nLhWESkG/gbcopQ6FGnXENuy5npDXGdW/F21IwlNC1Af9HsdsMMmW5KGUmqH+bMVeBIjDN4dSFmZP1vN3XPhNYj12rL2mpVSu5VSvUqpPuB3GH9byIFrFRE3xofrn5VSfzc359zfNtR1ZsvfVTuS0CwGmkSkQUQ8GLPk59psU0KISJGIlATuAxcAKzGuK6BguQ54yrw/F/ikqYKZBhwMpBKyiFiv7TngAhEpN1MIF5jbMp5+9avLMf62YFzrNSLiFZEGoAlYRJb8j4uIAA8Aq5VSPw96KKf+tuGuM2v+rnarFTL1hqH+WIehgPiW3fYk4XqGYyg4lgOrAtcE+IEXgfXmzwpzuwD3mtf/LjDF7muIcn1/wQj9uzG+lV0fz7UBn8EoXDYDn7b7umK41ofNa1mB8cExKGj/b5nXuha4KGh7xv+PA2dipGZWAMvM28W59reNcJ1Z8XfVK9s1Go1GkxA6taXRaDSahNCORKPRaDQJoR2JRqPRaBJCOxKNRqPRJIR2JBqNRqNJCO1INJooiMgC8+cwEfmPJJ/7m6GeS6PJJrT8V6OxiIici9GJ9ZIYjnEqpXojPH5EKVWcDPs0GrvQEYlGEwUROWLe/TFwljkX4lYRcYrIT0VksdlU7z/N/c81Z0s8grGYDBH5h9ksc1WgYaaI/BgoMM/35+DnMldm/1REVooxQ+bqoHO/IiJPiMgaEfmzuSoaEfmxiLxn2nJXOl8jTX7jstsAjSaLuI2giMR0CAeVUqeKiBd4Q0SeN/edCoxXRotvgM8opdpEpABYLCJ/U0rdJiJfUkqdEuK5rsBo1DcBqDSPmW8+NhEYh9FD6Q3gDBF5D6OFxmillBKRAUm/eo0mDDoi0Wji5wKMvk7LMFp++zF6HgEsCnIiADeJyHLgLYymek1E5kzgL8po2LcbeBU4NejcLcpo5LcMGAYcAjqA/xORK4D2hK9Oo7GIdiQaTfwI8GWl1CnmrUEpFYhIjh7fyaitzAKmK6UmAO8APgvnDkdn0P1ewKWU6sGIgv6GMeRpXkxXotEkgHYkGo11DmOMQQ3wHPAFs/03IjLS7KzcnzJgv1KqXURGY4yADdAdOL4f84GrzTpMFcZ43UXhDDPnWJQppZ4FbsFIi2k0aUHXSDQa66wAeswU1R+AX2Ckld42C957eH/kazDzgM+LyAqMTq1vBT12P7BCRN5WSn0saPuTwHSMbs0K+IZSapfpiEJRAjwlIj6MaObW+C5Ro4kdLf/VaDQaTULo1JZGo9FoEkI7Eo1Go9EkhHYkGo1Go0kI7Ug0Go1GkxDakWg0Go0mIbQj0Wg0Gk1CaEei0Wg0moTQjkSj0Wg0CfH/Aa+LvlsjQdlgAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f414d5cdc50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f414cba16a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.27450479332743705, 0.8940809968847352)"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_preds,y = learn.TTA()\n",
    "preds = np.mean(np.exp(log_preds),axis=0)\n",
    "(metrics.log_loss(y,preds), accuracy(preds,y))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save('resnext50_sz75_zm12')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.load('resnext50_sz75_zm12')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run Learner again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c0c955aa48de4572a75b1df76400513c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=63), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.12268 0.33517 0.88921]                          \n",
      "[1.      0.15165 0.31365 0.875  ]                          \n",
      "[2.      0.12309 0.34163 0.88068]                          \n",
      "[3.      0.18311 0.27565 0.90341]                          \n",
      "[4.      0.13618 0.30569 0.87784]                          \n",
      "[5.      0.11171 0.34182 0.87784]                          \n",
      "[6.      0.08782 0.35781 0.89204]                           \n",
      "[7.      0.08607 0.30347 0.90625]                           \n",
      "[8.      0.1726  0.50979 0.83807]                           \n",
      "[9.      0.14677 0.3044  0.90625]                          \n",
      "[10.       0.14314  0.39492  0.87784]                      \n",
      "[11.       0.1219   0.46043  0.86648]                      \n",
      "[12.       0.1117   0.43205  0.86648]                       \n",
      "[13.       0.11275  0.35311  0.88636]                       \n",
      "[14.       0.09713  0.37705  0.86932]                       \n",
      "[15.       0.13306  0.58516  0.83523]                      \n",
      "[16.       0.12111  0.36078  0.86648]                      \n",
      "[17.       0.15251  0.31952  0.90057]                       \n",
      "[18.       0.1232   0.43179  0.86932]                      \n",
      "[19.       0.08394  0.41684  0.88352]                       \n",
      "[20.       0.11493  0.66426  0.85796]                       \n",
      "[21.       0.1364   0.41949  0.88921]                      \n",
      "[22.       0.18069  0.59525  0.82102]                      \n",
      "[23.       0.12627  0.35364  0.88636]                      \n",
      "[24.       0.13227  0.46037  0.86648]                       \n",
      "[25.       0.13241  0.35233  0.875  ]                      \n",
      "[26.       0.14076  0.36253  0.88068]                      \n",
      "[27.       0.10788  0.47979  0.87216]                       \n",
      "[28.       0.08854  0.39871  0.875  ]                       \n",
      "[29.       0.07771  0.37563  0.88068]                       \n",
      "[30.       0.06105  0.46     0.875  ]                       \n",
      "[31.       0.06834  0.53877  0.84659]                       \n",
      "[32.       0.07889  0.49499  0.85511]                       \n",
      "[33.       0.0872   0.43733  0.88636]                       \n",
      "[34.       0.12003  0.58429  0.84659]                       \n",
      "[35.       0.1232   0.4414   0.86364]                      \n",
      "[36.       0.10887  0.41911  0.86932]                      \n",
      "[37.       0.0826   0.50516  0.85796]                       \n",
      "[38.       0.07245  0.47599  0.86932]                       \n",
      "[39.       0.05591  0.38545  0.86932]                       \n",
      "[40.       0.05237  0.42439  0.87784]                       \n",
      "[41.       0.15229  0.60266  0.83239]                       \n",
      "[42.       0.14381  0.40646  0.875  ]                       \n",
      "[43.       0.11055  0.43409  0.85511]                      \n",
      "[44.       0.08817  0.41339  0.88068]                       \n",
      "[45.       0.08276  0.40715  0.88921]                       \n",
      "[46.       0.06893  0.39587  0.88068]                       \n",
      "[47.       0.05226  0.38077  0.87784]                       \n",
      "[48.       0.04103  0.47042  0.86648]                       \n",
      "[49.       0.0593   0.49312  0.875  ]                       \n",
      "[50.       0.10371  0.47594  0.86648]                       \n",
      "[51.       0.06428  0.4053   0.87216]                       \n",
      "[52.       0.04861  0.42002  0.875  ]                       \n",
      "[53.       0.11681  0.4532   0.87784]                       \n",
      "[54.       0.09798  0.44153  0.87216]                       \n",
      "[55.       0.05929  0.42417  0.88068]                       \n",
      "[56.       0.05421  0.42152  0.88921]                       \n",
      "[57.       0.04737  0.37579  0.89489]                       \n",
      "[58.       0.03321  0.40086  0.88636]                       \n",
      "[59.       0.07931  0.42718  0.88352]                       \n",
      "[60.       0.05259  0.42606  0.875  ]                       \n",
      "[61.       0.04137  0.41565  0.88636]                       \n",
      "[62.       0.03766  0.42061  0.88352]                       \n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit(lrs, 6, cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f414d465588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.33657381838428085, 0.881619937694704)"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "log_preds,y = learn.TTA()\n",
    "preds = np.mean(np.exp(log_preds),axis=0)\n",
    "(metrics.log_loss(y,preds), accuracy(preds,y))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Run Learner again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6a950480f09c417abfee56255979b00a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/html": [
       "<p>Failed to display Jupyter Widget of type <code>HBox</code>.</p>\n",
       "<p>\n",
       "  If you're reading this message in the Jupyter Notebook or JupyterLab Notebook, it may mean\n",
       "  that the widgets JavaScript is still loading. If this message persists, it\n",
       "  likely means that the widgets JavaScript library is either not installed or\n",
       "  not enabled. See the <a href=\"https://ipywidgets.readthedocs.io/en/stable/user_install.html\">Jupyter\n",
       "  Widgets Documentation</a> for setup instructions.\n",
       "</p>\n",
       "<p>\n",
       "  If you're reading this message in another frontend (for example, a static\n",
       "  rendering on GitHub or <a href=\"https://nbviewer.jupyter.org/\">NBViewer</a>),\n",
       "  it may mean that your frontend doesn't currently support widgets.\n",
       "</p>\n"
      ],
      "text/plain": [
       "HBox(children=(IntProgress(value=0, description='Epoch', max=63), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.18378 0.48322 0.875  ]                           \n",
      "[1.      0.10417 0.42388 0.86364]                           \n",
      "[2.      0.11562 0.33858 0.89489]                           \n",
      "[3.      0.07672 0.30463 0.90057]                           \n",
      "[4.      0.06673 0.3811  0.88068]                           \n",
      "[5.      0.12975 0.37492 0.89204]                           \n",
      "[6.      0.08586 0.34695 0.88921]                           \n",
      "[7.      0.06178 0.39774 0.89773]                           \n",
      "[8.      0.08045 0.40698 0.89204]                           \n",
      "[9.      0.11349 0.47287 0.86648]                          \n",
      "[10.       0.09922  0.34917  0.90625]                       \n",
      "[11.       0.09501  0.36274  0.89489]                       \n",
      "[12.       0.07781  0.39213  0.88921]                       \n",
      "[13.       0.08187  0.34004  0.89489]                       \n",
      "[14.       0.07801  0.37382  0.875  ]                       \n",
      "[15.       0.09595  0.42967  0.88068]                       \n",
      "[16.       0.12496  0.41968  0.85511]                      \n",
      "[17.       0.09355  0.49051  0.86932]                       \n",
      "[18.       0.10343  0.36958  0.88921]                       \n",
      "[19.       0.09544  0.43581  0.85796]                       \n",
      " 44%|████▍     | 18/41 [00:35<00:44,  1.95s/it, loss=0.0843]"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-33-73791edd19a9>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mlearn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlrs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m6\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcycle_len\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcycle_mult\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/fastai/fastai/learner.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, lrs, n_cycle, wds, **kwargs)\u001b[0m\n\u001b[1;32m    211\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msched\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    212\u001b[0m         \u001b[0mlayer_opt\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_layer_opt\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlrs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 213\u001b[0;31m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit_gen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlayer_opt\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mn_cycle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    214\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    215\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mwarm_up\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstart_lr\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1e-5\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mend_lr\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mwds\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/fastai/fastai/learner.py\u001b[0m in \u001b[0;36mfit_gen\u001b[0;34m(self, model, data, layer_opt, n_cycle, cycle_len, cycle_mult, cycle_save_name, metrics, callbacks, use_wd_sched, norm_wds, wds_sched_mult, **kwargs)\u001b[0m\n\u001b[1;32m    158\u001b[0m         \u001b[0mn_epoch\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0msum_geom\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcycle_len\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mcycle_len\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcycle_mult\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mn_cycle\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    159\u001b[0m         fit(model, data, n_epoch, layer_opt.opt, self.crit,\n\u001b[0;32m--> 160\u001b[0;31m             metrics=metrics, callbacks=callbacks, reg_fn=self.reg_fn, clip=self.clip, **kwargs)\n\u001b[0m\u001b[1;32m    161\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    162\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mget_layer_groups\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmodels\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mget_layer_groups\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/fastai/fastai/model.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(model, data, epochs, opt, crit, metrics, callbacks, **kwargs)\u001b[0m\n\u001b[1;32m     87\u001b[0m             \u001b[0mbatch_num\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     88\u001b[0m             \u001b[0;32mfor\u001b[0m \u001b[0mcb\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mcallbacks\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mcb\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_batch_begin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 89\u001b[0;31m             \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mstepper\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mV\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mV\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     90\u001b[0m             \u001b[0mavg_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mavg_loss\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0mavg_mom\u001b[0m \u001b[0;34m+\u001b[0m \u001b[0mloss\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0mavg_mom\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     91\u001b[0m             \u001b[0mdebias_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mavg_loss\u001b[0m \u001b[0;34m/\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0;36m1\u001b[0m \u001b[0;34m-\u001b[0m \u001b[0mavg_mom\u001b[0m\u001b[0;34m**\u001b[0m\u001b[0mbatch_num\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/fastai/fastai/model.py\u001b[0m in \u001b[0;36mstep\u001b[0;34m(self, xs, y)\u001b[0m\n\u001b[1;32m     43\u001b[0m         \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mraw_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mcrit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     44\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreg_fn\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0mloss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mreg_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0moutput\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mxtra\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mraw_loss\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 45\u001b[0;31m         \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     46\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclip\u001b[0m\u001b[0;34m:\u001b[0m   \u001b[0;31m# Gradient clipping\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     47\u001b[0m             \u001b[0mnn\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mutils\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclip_grad_norm\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrainable_params_\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mm\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mclip\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/src/anaconda3/envs/fastai/lib/python3.6/site-packages/torch/autograd/variable.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(self, gradient, retain_graph, create_graph, retain_variables)\u001b[0m\n\u001b[1;32m    165\u001b[0m                 \u001b[0mVariable\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    166\u001b[0m         \"\"\"\n\u001b[0;32m--> 167\u001b[0;31m         \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mautograd\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbackward\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mgradient\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mretain_graph\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcreate_graph\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mretain_variables\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    168\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    169\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0mregister_hook\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhook\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/src/anaconda3/envs/fastai/lib/python3.6/site-packages/torch/autograd/__init__.py\u001b[0m in \u001b[0;36mbackward\u001b[0;34m(variables, grad_variables, retain_graph, create_graph, retain_variables)\u001b[0m\n\u001b[1;32m     97\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     98\u001b[0m     Variable._execution_engine.run_backward(\n\u001b[0;32m---> 99\u001b[0;31m         variables, grad_variables, retain_graph)\n\u001b[0m\u001b[1;32m    100\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    101\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "learn.fit(lrs, 6, cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Log Loss | Accuracy (final model here)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "log_preds,y = learn.TTA()\n",
    "preds = np.mean(np.exp(log_preds),axis=0)\n",
    "(metrics.log_loss(y,preds), accuracy(preds,y))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Save predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# test_df = pd.read_json(PATH+\"/test.json\")\n",
    "# helper.make_test_file(log_preds,test_df,model_name='resnext50_sz75_zm15_bs32_comp_band',\n",
    "#                       classes=data.classes,\n",
    "#                       submission_folder =f'{PATH}sub'\n",
    "#                      )"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}