7_1_resnext50_more_epoch_submitted.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": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 2716\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  49776 Jan 15 01:30 7_5_resnext50_more_epoch_update_sz.ipynb\r\n",
      "-rw-rw-r-- 1 ubuntu ubuntu 206248 Jan 15 01:31 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 159533 Jan 15 01:30 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   1026 Jan 15 01:29 README.md\r\n"
     ]
    }
   ],
   "source": [
    "!ls -l"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.imports import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "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": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "#PATH = \"/home/nidhin/data/iceberg/\"\n",
    "PATH = os.path.expanduser(\"~/data/iceberg/\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "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 10 ubuntu ubuntu       4096 Jan 15 01:04 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": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Set Architecture"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "arch=resnext50\n",
    "sz=150\n",
    "bs=32\n",
    "lr = 0.005"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "15cb74377039417db91d9f3404efe46d",
       "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*1.3), 'tmp')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "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": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "51a6ad6561fb4e0eb3c6a3a8cdc7dde3",
       "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": [
      " 63%|██████▎   | 26/41 [00:00<00:00, 49.45it/s, loss=12.6] \n",
      "                                                          \r"
     ]
    }
   ],
   "source": [
    "lrf = learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYUAAAEKCAYAAAD9xUlFAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4xLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvAOZPmwAAIABJREFUeJzt3XmcXGWd7/HPr9d00uls3UDoLB1CWGSRJQRZhgmMKKIjggtwRxEZJ4qiiNs4zOsqcvF1GXGcizIDoiDisKmgRkQRhbDIAGkCJJCNkHRIJyHp7qS703t31e/+UaeKoumlknTVOVX9fb9e9aqqc56q+nVB9zfneZ7zHHN3REREAIrCLkBERKJDoSAiIikKBRERSVEoiIhIikJBRERSFAoiIpKiUBARkRSFgoiIpCgUREQkpSTsAvZWdXW119XVhV2GiEheef7555vdvWa0dnkXCnV1ddTX14ddhohIXjGzzZm0U/eRiIikKBRERCRFoSAiIikKBRERSVEoiIhIikJBRERSFAoiIpKiUBARyQM3/vlVnnq1Oeufo1AQEYm43oEYN/5lPc817Mr6ZykUREQibsuuLuIOh1RPyvpnKRRERCJuY1MnAHUKBRERaWhJhMK8GXkcCmY228weM7M1ZvaKmV05RJvFZtZmZi8Gt29mqx4RkXy1qbmLaRNLmTKxNOuflc1VUgeAr7j7CjObDDxvZo+4++pB7Z509w9ksQ4RkbzW0NzJvBx0HUEWjxTcfbu7rwge7wHWALXZ+jwRkUK1qbkzJ+MJkKMxBTOrA44Hnh1i9ylm9pKZ/cHMjspFPSIi+aK7L8Yb7T05GU+AHFxkx8wqgfuBL7l7+6DdK4C57t5hZucCvwEWDPEeS4AlAHPmzMlyxSIi0ZEcZC6IIwUzKyURCHe5+wOD97t7u7t3BI8fAkrNrHqIdre6+0J3X1hTM+rV5ERECkZDczDzKN9DwcwMuA1Y4+7fH6bNQUE7zGxRUE9LtmoSEck3m3J8pJDN7qPTgE8Aq8zsxWDb1cAcAHe/BfgIcLmZDQDdwEXu7lmsSUQkr2xq6qRmcjmV5Vnv7QeyGAru/hRgo7S5CbgpWzWIiOS7hpbOnA0yg85oFhGJtE3NXTkbTwCFgohIZO3p6ae5ozdn4wmgUBARiayG5i4A5lVPzNlnKhRERCIq1zOPQKEgIhJZyXMU6jTQLCIiDc2dHDxlAhNKi3P2mQoFEZGI2pjDhfCSFAoiIhHV0KJQEBERoLWrj9au/pxclzmdQkFEJII2hTDIDAoFEZFISoWCjhRERKShuZMigznTc3fiGigUREQiaVNLF7OmTaSsJLd/phUKIiIR1BDCdFRQKIiIRI67s6m5k3kzctt1BAoFEZHIae7oo6N3QEcKIiKSOGkNcndd5nQKBRGRiElOR1UoiIgIm5o7KSkyaqdW5PyzFQoiIhHT0NzJnOkTKSnO/Z9ohYKISMRsau4MpesIFAoiIpESjzubW7pCmXkECgURkUjZsaeH7v6YQkFERNJmHuV4ddQkhYKISIQ0NHcBMK9GoSAiMu41tHRSXlLEzKoJoXy+QkFEJEI2NnUyd8ZEiooslM9XKIiIREhDS2fOr7aWTqEgIhIRsbjzektXaOMJkMVQMLPZZvaYma0xs1fM7Moh2piZ/cDMNpjZSjM7IVv1iIhE3bbWbvpi8dBmHgGUZPG9B4CvuPsKM5sMPG9mj7j76rQ27wMWBLeTgZuDexGRcSes6zKny9qRgrtvd/cVweM9wBqgdlCz84A7PeEZYKqZzcxWTSIiURbmktlJORlTMLM64Hjg2UG7aoEtac8beXtwiIiMC5uaO5lYVswBk8tDqyHroWBmlcD9wJfcvX3w7iFe4kO8xxIzqzez+qampmyUKSISuobmxMwjs3Cmo0KWQ8HMSkkEwl3u/sAQTRqB2WnPZwHbBjdy91vdfaG7L6ypqclOsSIiIQtzddSkbM4+MuA2YI27f3+YZkuBS4JZSO8C2tx9e7ZqEhGJqv5YnC27u6mrnhhqHdmcfXQa8AlglZm9GGy7GpgD4O63AA8B5wIbgC7gU1msR0Qkshp3dxOLO/OqK0OtI2uh4O5PMfSYQXobBz6frRpERPJFQ+q6zOEeKeiMZhGRCNiYPEchxBPXQKEgIhIJDc2dTJ5QwvRJZaHWoVAQEYmAhpZODqkOdzoqKBRERCJhU3NnqMtbJCkURERC1tMfY2trd+jjCaBQEBEJ3eu7unAPd82jJIWCiEjIVm9LrAB05MyqkCtRKIiIhG5lYxsVpcXMD/HiOkkKBRGRkK3a2spRB1dRUhz+n+TwKxARGccGYnFe3trOMbOmhF0KoFAQEQnVa02ddPfHOFahICIiKxtbATimdmrIlSQoFEREQrRqaxuV5SUcEoHpqKBQEBEJ1crGNo6uraKoKNzlLZIUCiIiIemPxVm9vZ1jZ0Wj6wgUCiIioVn3xh76BuIcUxuNQWZQKIiIhGbV1jaAyMw8gr0IBTOLxiiIiEiBWNnYRtWEEuZMD/dqa+lGDQUzO9XMVgNrgufvNLP/ynplIiIFbtXWVo6dNTX0ayiky+RI4T+A9wItAO7+EnBGNosSESl0Pf0x1r2xJ1JdR5Bh95G7bxm0KZaFWkRExo11b+yhP+aRC4WSDNpsMbNTATezMuCLBF1JIiKyb1JnMkdoOipkdqTwWeDzQC3QCBwHfC6bRYmIFLqVjW3MmFTGwVMmhF3KW2RypHC4u/9D+gYzOw34a3ZKEhEpfKu2tnHMrCmRGmSGzI4UfpjhNhERyUB3X4z1O/ZwbIROWksa9kjBzE4BTgVqzOzLabuqgOJsFyYiUqhWb28j7kRqeYukkbqPyoDKoM3ktO3twEeyWZSISCFb2Zg4kzkqF9ZJN2wouPvjwONmdoe7b85hTSIiBW1VYxsHVpVzYFW0Bpkhs4HmLjO7ATgKSP0E7n5W1qoSESlgLzW2RuaiOoNlMtB8F7AWmAd8G2gAlo/2IjO73cx2mtnLw+xfbGZtZvZicPvmXtQtIpKX9vT0s7G5M3InrSVlEgoz3P02oN/dH3f3y4B3ZfC6O4BzRmnzpLsfF9yuzeA9RUTy2ivb2nGP5ngCZNZ91B/cbzez9wPbgFmjvcjdnzCzun0vTUSk8KwKBpmjOB0VMguF68xsCvAVEucnVAFXjdHnn2JmL5EImq+6+ytj9L4iIpG0cmsbtVMrmFFZHnYpQxoxFMysGFjg7g8CbcCZY/jZK4C57t5hZucCvwEWDFPHEmAJwJw5c8awBBGR3FrV2BrZ8QQYZUzB3WPAB7Pxwe7e7u4dweOHgFIzqx6m7a3uvtDdF9bU1GSjHBGRrGvr6qehpSuy4wmQWffR02Z2E3Af0Jnc6O4r9ueDzewgYIe7u5ktIhFQLfvzniIiUZa6/GZEp6NCZqFwanCfPjvIgRHPUzCze4DFQLWZNQLfAkoB3P0WEmdFX25mA0A3cJG7+15VLyKSR1ZuDZbLjuggM2QQCu6+T+MI7n7xKPtvAm7al/cWEclHqxrbmDtjIlMmloZdyrAyuvKaiIjsv5WNbZFcBC+dQkFEJAdaOnrZ2tod2fMTkhQKIiI5sHJrdFdGTTfqmIKZXTDE5jZglbvvHPuSREQKz6rGNszgqIOrwi5lRJnMPvpH4BTgseD5YuAZ4DAzu9bdf56l2kRECsbKxjYOqZ7E5AnRHWSGzEIhDhzp7jsAzOxA4GbgZOAJQKEgIjKKVVtbOXX+kOfnRkomYwp1yUAI7AQOc/ddvLlYnoiIDGNHew872nsjfX5CUiZHCk+a2YPAL4PnHwaeMLNJQGvWKhMRKRDJlVHfObswQuHzJILgNMCAO4H7g7OPx3KBPBGRgvTMxhbKiot4x8wCCIXgj/+vgpuIiOylZeubOPmQ6VSUFYddyqhGHVMwswvM7NXg0pntZrbHzNpzUZyISL5r3N3Fhp0d/O1h+bHCcybdR98F/t7d12S7GBGRQrNsXRMAiw8/IORKMpPJ7KMdCgQRkX2zbF0Ts6ZVML9mUtilZCSTI4V6M7uPxJXRepMb3f2BrFUlIlIAegdiPP1aMxecUIuZhV1ORjIJhSqgC3hP2jYHFAoiIiOob9hNV1+MxYflR9cRZDb76FO5KEREpNAsW7eTsuIiTj10RtilZGzYUDCzr7v7d83shySODN7C3b+Y1cpERPLcsnVNLJo3nYllmXTKRMNIlSYHl+tzUYiISCHZ2trNqzs7uPCk2WGXsleGDQV3/11w/7PclSMiUhiWrUtcWWDx4flxfkJSJtdTOAz4KlCX3t7dz8peWSIi+W3ZuiZqp1Ywv6Yy7FL2SiYdXb8EbgF+AsSyW46ISP7rG4jz9IZmPnR8/kxFTcokFAbc/easVyIiUiDqG3bR2RfjzDw5izldJmc0/87MPmdmM81sevKW9cpERPLUsvVNeTcVNSmTI4VPBvdfS9vmwCFjX46ISP57bO3OvJuKmjRixWZWBHzc3f+ao3pERPJavk5FTRqx+8jd48D3clSLiEjey9epqEmZjCn8ycw+bPk2hC4iEoJ8nYqalEmH15eBScCAmfWQuCSnu3tVVisTEckz+TwVNSmTBfEm56IQEZF8l5yKmi8X1BlKJt1HmNk0M1tkZmckbxm85nYz22lmLw+z38zsB2a2wcxWmtkJe1u8iEiUpKaizs+/qahJmVyj+dPAE8DDwLeD+2syeO87gHNG2P8+YEFwWwLoBDkRyWvL1u3kpHnTmFSef1NRkzI5UrgSOAnY7O5nAscDTaO9yN2fAHaN0OQ84E5PeAaYamYzM6hHRCRytrV2s35HR15dUGcomYRCj7v3AJhZubuvBQ4fg8+uBbakPW8MtomI5J1l6xL/Vs7XqahJmRzjNJrZVBLXaH7EzHYD28bgs4camn/bxXwAzGwJiS4m5syZMwYfLSIytpat20nt1AoOPSA/p6ImZTL76Pzg4TVm9hgwBfjjGHx2I5B+yt8shgkbd78VuBVg4cKFQwaHiEhY+gbi/HVDM+fl8VTUpExnH51uZp9y98eB/2FsunmWApcEs5DeBbS5+/YxeF8RkZyq3xxMRT0sv7uOILOL7HwLWEhiHOGnQCnw38Bpo7zuHmAxUG1mjcC3gtfi7rcADwHnAhuALuBT+/pDiIiEadm6JkqLjdMOrQ67lP2WyZjC+SRmHK0AcPdtZjbqCW3ufvEo+x34fCZFiohE1UAszm9e2MrfLKjJ66moSZl0H/UFf8AdwMwmZbckEZH88Ze1O9m5p5eLFxXGJJhMQuEXZvYjEucR/BPwZ+DH2S1LRCQ/3PPc6xxUNYEz83wqalIms4++Z2ZnA+0kxhW+6e6PZL0yEZGIa9zdxePrm/jCWQsoKc5o3k7kZdQBFoSAgkBEJM19yxPn3+brBXWGMmwomNkehj6ZTEtni8i4NxCLc9/yLSw+rIbaqRVhlzNmhg0FLZktIjK8R4MB5u+cPDfsUsZUYXSCiYjk2N3Pvc6BVeUFM8CcpFAQEdlLyQHmCxfOLpgB5qTC+mlERHLgF8kB5gI5NyGdQkFEZC8MxOLcV194A8xJCgURkb3w6Nqd7GgvnDOYB1MoiIjsheQA81lH5PcV1oajUBARyVAhDzAnFeZPJSKSBckB5o8V0BnMgykUREQykD7APGvaxLDLyRqFgohIBgp9gDlJoSAikoF7CnyAOUmhICIyisbdXSwr8AHmpML+6URExsB4GGBOUiiIiIygo3eAu597nb8t8AHmJIWCiMgIfvT4azR39HHl3y0Iu5ScUCiIiAxje1s3P35yI3//zoM5fs60sMvJCYWCiMgwbnh4HXGHr7/38LBLyRmFgojIEFY1tvHAiq1cdto8Zk8v/LGEJIWCiMgg7s51v1/N9EllfO7M+WGXk1MKBRGRQR5ZvYNnN+3iqrMPo2pCadjl5JRCQUQkTd9AnP/7h7UcekAlF4+D8xIGUyiIiKS569nNbGru5Opzjyj4s5eHMv5+YhGRYbR19XPjX17l9EOrOfPwwl7jaDhZDQUzO8fM1pnZBjP7xhD7LzWzJjN7Mbh9Opv1iIiM5KbHXqWtu5+rzz0SMwu7nFCUZOuNzawY+E/gbKARWG5mS9199aCm97n7FdmqQ0QkE5tbOrnj6QY+euIs3nFwVdjlhCabRwqLgA3uvtHd+4B7gfOy+HkiIvvs3/64ltLiIr7ynvFzotpQshkKtcCWtOeNwbbBPmxmK83sV2Y2/ob6RSR09Q27eGjVG3zmjPkcWDUh7HJClc1QGKpDzgc9/x1Q5+7HAn8GfjbkG5ktMbN6M6tvamoa4zJFZDyLx53/8/s1HFhVzj+dMS/sckKXzVBoBNL/5T8L2JbewN1b3L03ePpj4MSh3sjdb3X3he6+sKamJivFisj49INHX+WlLa388zlHMLEsa8OseSObobAcWGBm88ysDLgIWJrewMxmpj39ILAmi/WIiLzFY2t3cuNfXuXDJ8zi/OOH6t0ef7IWi+4+YGZXAA8DxcDt7v6KmV0L1Lv7UuCLZvZBYADYBVyarXpERNJtbunkyntf4MiDqvjO+UeP2ymog5n74G7+aFu4cKHX19eHXYaI5LHuvhgX3Pw021q7efALp4+LVVDN7Hl3XzhaO3Wgici44u5c/etVrH2jnZ9eetK4CIS9oWUuRGRc+fkzm/n1C1u56t2HsXicLmUxEoWCiIwbz2/exbW/W827jzyAK848NOxyIkmhICLjws49PVz+3yuonVbBv3/sOIqKNLA8FIWCiBS8/licK+5+gfaefm75+IlMqRhfF87ZGxpoFpGCd/0f1vLcpl3ceNFxHDlz/C52lwkdKYhIQbv9qU3c9tQmLj21jvOO0wlqo9GRgogUJHfnhofX8V/LXuO9Rx3Iv77/yLBLygsKBREpOAOxOP/665e5r34LFy+aw3UfOppiDSxnRKEgIgWlpz/GFXe/wJ/X7OCLZx3KVWcfpiUs9oJCQUQKRltXP5++czn1m3dz7XlHcckpdWGXlHcUCiJSEHa093DJbc+xsbmDH158PB849uCwS8pLCgURyXsbmzr4xG3P0drVx08vXcTpC6rDLilvKRREJK8tb9jFZ37+PAbcs+RdHDtratgl5TWFgojkpc7eAW54eB0/+58GZk2r4M7LTmZe9aSwy8p7CgURyTtPvtrEvzywisbd3Vxyyly+fs4RVJbrz9lY0LcoInmjrauf636/ml8+38gh1ZP4xWdOYdG86WGXVVAUCiKSF/748hv879++zK7OPi5fPJ8r/24BE0qLwy6r4CgURCTSmvb0cs3SV/j9qu28Y2YVP730JI6unRJ2WQVLoSAikbRlVxe3PbWJe5e/TjwOX3vv4Sw54xBKi7WOZzYpFEQkUlZva+dHT7zGgyu3U2TwoeNquXzxfA6pqQy7tHFBoSAioXN3ntm4i1sef43H1zcxqayYy06r47LT5zFzSkXY5Y0rCgURCU1Pf4w/r9nBj5/cxEtbWqmuLOdr7z2cj588lykTdXW0MCgURCSnuvtiLFu3k9+v2s6ja3fS1RejbsZEvnP+0Xz4hFmaURQyhYKIZF1X3wCPrW3ioSAIuvtjTJ9UxnnH1fL+Y2ZyyvwZut5BRCgURGTMxePOuh17qG/YxdOvtfDYup309MeprizjghMSQbBo3nRKNJMochQKIrLfegdirGxsY3nDLpZv2kX95t3s6RkA4KCqCXz0xNm875iDOHmejgiiTqEgInulrbufV3fsYf2ODtbv2MPqbe282NhK30AcgEMPqOQDx87kpLrpnFQ3nVnTKnTlszyiUBCRtxmIxXmjvYetu7tpaOlMBcD6HXvY0d6bajexrJjDD5rMJ0+Zy0l10zlx7jRmVJaHWLnsr6yGgpmdA9wIFAM/cffrB+0vB+4ETgRagAvdvSGbNYmMd70DMVo6+mjp6KO5s5ftrT1sa+1ma2s3W3cn7t9o7yEW99RrJpQWcegBlZw2v5oFB07m8IMqWXDAZGqnVlCk7qCCkrVQMLNi4D+Bs4FGYLmZLXX31WnN/hHY7e6HmtlFwL8BF2arJpFCMRCL09Ufo7N3gD09A+zp6ae9e4D2nn7au/tp70k+HmBXZ28iBDr7aO7oTfX1pysuMg6qmkDt1AoWzZtO7dQKaqdVcPDUCuZOn8js6RM1FjBOZPNIYRGwwd03ApjZvcB5QHoonAdcEzz+FXCTmZm7OyJZFo87cXfiDnF3PLhPbYs7MfegHanHsbTtA8HzxH2cgVjieX/wvD/m9MfiwS14PBA8jsfpG4jTOxCntz9O70As8XggTm9/4nF3f4zuvhhdfQN09cXo6ks874vFR/35yoqLqKooYcakcmZUlnF07RRmTCqjurKMGZXlzJhUxozKMg6aUsGBk8s1E0iA7IZCLbAl7XkjcPJwbdx9wMzagBlA81gX8/j6Jq57cPWIbaKQRGORh6O+wygNBu8eXNPb9w9+vQ+5723tgg2etj/52sTj9Nd56vUevNaT7YLHBK9J/oH34DXJx/GgbTwK/6HTlBUXUV5SRHlpEeUlxZSXFFFWUkR5aeJxdWUZE8smUlFWzMSyYirKiplUVpJ6XDWhlKqKUqomlFBVUcrkCSVUTSjVSWCyT7IZCkMdaw7+dcykDWa2BFgCMGfOnH0qprK8hAUHjr6glg1ZUo6NQQmjvcVos0EG7x3c/O37bfj9lv7wzSdmb+5KvtywNx/bmy9Oth3cztI+2wyKzFLtzJJtjCJ7636C++Iio7jIUvuKkm2Cx8VFFmw3iosI7i21vbTYKC4qoiTYlroPtpcWG6XFRcHNKCsuoqT4ze1lxUXqk5dIyWYoNAKz057PArYN06bRzEqAKcCuwW/k7rcCtwIsXLhwn/6dd+LcaZw498R9eamIyLiRzU7E5cACM5tnZmXARcDSQW2WAp8MHn8EeFTjCSIi4cnakUIwRnAF8DCJKam3u/srZnYtUO/uS4HbgJ+b2QYSRwgXZaseEREZXVbPU3D3h4CHBm37ZtrjHuCj2axBREQypzloIiKSolAQEZEUhYKIiKQoFEREJEWhICIiKZZvpwWYWROweZjd1WRhiYwxphrHTj7UqRrHhmrcf3PdvWa0RnkXCiMxs3p3Xxh2HSNRjWMnH+pUjWNDNeaOuo9ERCRFoSAiIimFFgq3hl1ABlTj2MmHOlXj2FCNOVJQYwoiIrJ/Cu1IQURE9kPBhIKZnWNm68xsg5l9I+x6hmJmDWa2ysxeNLP6sOsBMLPbzWynmb2ctm26mT1iZq8G99MiWOM1ZrY1+C5fNLNzQ65xtpk9ZmZrzOwVM7sy2B6Z73KEGiPzXZrZBDN7zsxeCmr8drB9npk9G3yP9wXL8YdmhDrvMLNNad/lcWHWuS8KovvIzIqB9cDZJC7csxy42N1Hvv5mjplZA7DQ3SMzl9nMzgA6gDvd/ehg23eBXe5+fRCw09z9nyNW4zVAh7t/L6y60pnZTGCmu68ws8nA88CHgEuJyHc5Qo0fIyLfpSUuozfJ3TvMrBR4CrgS+DLwgLvfa2a3AC+5+80RrPOzwIPu/quwattfhXKksAjY4O4b3b0PuBc4L+Sa8oK7P8Hbr3Z3HvCz4PHPSPzhCM0wNUaKu2939xXB4z3AGhLXII/MdzlCjZHhCR3B09Lg5sBZQPIPbRT+nxyuzrxXKKFQC2xJe95IxP5nDzjwJzN7PrjudFQd6O7bIfGHBDgg5HqGc4WZrQy6l0Lt4kpnZnXA8cCzRPS7HFQjROi7NLNiM3sR2Ak8ArwGtLr7QNAkEr/fg+t09+R3+Z3gu/wPMysPscR9UiihMNSVz6OY2qe5+wnA+4DPB90ism9uBuYDxwHbgX8Pt5wEM6sE7ge+5O7tYdczlCFqjNR36e4xdz+OxHXdFwFHDtUst1UNUcCgOs3saOBfgCOAk4DpQGjdrvuqUEKhEZid9nwWsC2kWobl7tuC+53Ar0n8Dx9FO4L+52Q/9M6Q63kbd98R/FLGgR8Tge8y6Fu+H7jL3R8INkfquxyqxih+lwDu3gosA94FTDWz5JUiI/X7nVbnOUEXnbt7L/BTIvJd7o1CCYXlwIJghkIZiWs9Lw25prcws0nB4B5mNgl4D/DyyK8KzVLgk8HjTwK/DbGWISX/0AbOJ+TvMhh4vA1Y4+7fT9sVme9yuBqj9F2aWY2ZTQ0eVwDvJjH28RjwkaBZ6P9PDlPn2rR/ABiJcY+o/o4PqyBmHwEE0+j+H1AM3O7u3wm5pLcws0NIHB1A4trYd0ehRjO7B1hMYoXHHcC3gN8AvwDmAK8DH3X30AZ6h6lxMYnuDgcagM8k++7DYGanA08Cq4B4sPlqEn32kfguR6jxYiLyXZrZsSQGkotJ/KP1F+5+bfD7cy+JLpkXgI8H/xoPxQh1PgrUkOjSfhH4bNqAdF4omFAQEZH9VyjdRyIiMgYUCiIikqJQEBGRFIWCiIikKBRERCRFoSDjjpk9HdzXmdn/GuP3vnqozxLJF5qSKuOWmS0GvuruH9iL1xS7e2yE/R3uXjkW9YmEQUcKMu6YWfJkouuBvwnWvb8qWODsBjNbHixo9pmg/eLgOgR3kzjxCzP7TbCw4SvJxQ3N7HqgIni/u9I/yxJuMLOXLXFNjQvT3nuZmf3KzNaa2V3B2bCY2fVmtjqoJfRlrWV8KBm9iUjB+gZpRwrBH/c2dz8pWN3yr2b2p6DtIuBod98UPL/M3XcFSxwsN7P73f0bZnZFsEjaYBeQOGv4nSTOzF5uZk8E+44HjiKxns9fgdPMbDWJJSeOcHdPLqkgkm06UhB503uAS4LlkJ8FZgALgn3PpQUCwBfN7CXgGRKLMS5gZKcD9wQLz+0AHiexkmbyvRuDBeleBOqAdqAH+ImZXQB07fdPJ5IBhYLImwz4grsfF9zmuXvySKEz1SgxFvFu4BR3fyeJtXgmZPDew0lfwycGlATXDlhEYkXTDwF/3KufRGQfKRRkPNsDTE57/jBwebC8NGZ2WLCi7WBTgN3u3mVmR5BY2jmpP/n6QZ4ALgzGLWqAM4DnhissuObBFHd/CPgSia4nkazTmIKMZyuBgaAb6A7gRhJdNyuCwd4mhr7s4x+Bz5rZSmAdiS7+Vf6HAAAAZUlEQVSkpFuBlWa2wt3/IW37r4FTgJdIrEb6dXd/IwiVoUwGfmtmE0gcZVy1bz+iyN7RlFQREUlR95GIiKQoFEREJEWhICIiKQoFERFJUSiIiEiKQkFERFIUCiIikqJQEBGRlP8P9hYecLl/7GoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd048680828>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd0481e59b0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot() "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "507f8710444a4a05965bbfbcaa352a1a",
       "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.      0.57287 0.34829 0.84091]                          \n",
      "[1.      0.47014 0.29065 0.875  ]                          \n",
      "[2.      0.46306 0.30295 0.87784]                          \n",
      "[3.      0.40557 0.28736 0.87216]                          \n",
      "[4.      0.40038 0.27163 0.875  ]                          \n",
      "[5.      0.37548 0.28542 0.86364]                          \n",
      "[6.      0.34352 0.27652 0.88352]                          \n",
      "[7.      0.33688 0.27711 0.88068]                          \n",
      "[8.      0.33155 0.27699 0.86932]                          \n",
      "[9.      0.33102 0.29422 0.86648]                          \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": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0e7929c4f16d412084d3e5e030c8658b",
       "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=31), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.54513 0.28379 0.86648]                          \n",
      "[1.      0.4572  0.27839 0.88636]                          \n",
      "[2.      0.41219 0.27955 0.89773]                          \n",
      "[3.      0.43353 0.29683 0.86364]                          \n",
      "[4.      0.4329  0.29051 0.87216]                          \n",
      "[5.      0.42221 0.28305 0.875  ]                          \n",
      "[6.      0.41527 0.29611 0.86648]                          \n",
      "[7.      0.43743 0.28625 0.86364]                          \n",
      "[8.      0.42604 0.28498 0.88352]                          \n",
      "[9.      0.44338 0.28109 0.88068]                          \n",
      "[10.       0.41329  0.28446  0.87784]                      \n",
      "[11.       0.40828  0.28211  0.86932]                      \n",
      "[12.       0.39784  0.27164  0.87784]                      \n",
      "[13.       0.38189  0.27654  0.87784]                      \n",
      "[14.       0.38852  0.2727   0.87784]                      \n",
      "[15.       0.41554  0.2845   0.86648]                      \n",
      "[16.       0.43568  0.28075  0.875  ]                      \n",
      "[17.       0.42095  0.27443  0.88352]                      \n",
      "[18.       0.40781  0.28009  0.87216]                      \n",
      "[19.       0.42653  0.26718  0.86648]                      \n",
      "[20.       0.39065  0.28478  0.875  ]                      \n",
      "[21.       0.38808  0.29024  0.86932]                      \n",
      "[22.       0.38899  0.27746  0.87216]                      \n",
      "[23.       0.39838  0.31406  0.84943]                      \n",
      "[24.       0.37563  0.27554  0.88636]                      \n",
      "[25.       0.36301  0.27152  0.88352]                      \n",
      "[26.       0.35885  0.27336  0.875  ]                      \n",
      "[27.       0.3717   0.27522  0.87216]                      \n",
      "[28.       0.37178  0.30121  0.84943]                      \n",
      "[29.       0.36746  0.27059  0.88352]                      \n",
      "[30.       0.35996  0.26852  0.88068]                      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "#learn.fit(lr, 4, cycle_len=2)\n",
    "learn.fit(lr, 5, cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd04b892c18>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd04b92e240>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.28586956342000025, 0.8691588785046729)"
      ]
     },
     "execution_count": 53,
     "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": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Unfreeze Layers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2e49c1e3c5cc4cc692b6c3b1236846ec",
       "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=31), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.53807 0.41887 0.71875]                          \n",
      "[1.      0.49472 0.29159 0.87216]                          \n",
      "[2.      0.42283 0.31259 0.83807]                          \n",
      "[3.      0.39959 0.32258 0.84943]                          \n",
      "[4.      0.38208 0.28328 0.86364]                          \n",
      "[5.      0.38165 0.25929 0.88068]                          \n",
      "[6.      0.34044 0.31344 0.85511]                          \n",
      "[7.      0.37211 0.24899 0.88921]                          \n",
      "[8.      0.33125 0.44354 0.81818]                          \n",
      "[9.      0.29615 0.27662 0.88352]                          \n",
      "[10.       0.28538  0.32928  0.875  ]                      \n",
      "[11.       0.24391  0.26289  0.89204]                      \n",
      "[12.       0.23319  0.25978  0.89773]                      \n",
      "[13.       0.2272   0.28245  0.88636]                      \n",
      "[14.       0.20979  0.27199  0.88636]                      \n",
      "[15.       0.2319   0.46345  0.79546]                      \n",
      "[16.       0.28758  0.31308  0.86364]                      \n",
      "[17.       0.31896  0.3469   0.85511]                      \n",
      "[18.       0.32592  0.24296  0.90057]                      \n",
      "[19.       0.27805  0.28564  0.86364]                      \n",
      "[20.       0.23209  0.28793  0.87784]                      \n",
      "[21.       0.22187  0.27726  0.88352]                      \n",
      "[22.       0.20423  0.36247  0.85796]                      \n",
      "[23.       0.18868  0.24849  0.90625]                      \n",
      "[24.       0.16998  0.29129  0.88636]                      \n",
      "[25.       0.20482  0.30584  0.88636]                      \n",
      "[26.       0.17789  0.28197  0.89204]                      \n",
      "[27.       0.17406  0.29822  0.88636]                      \n",
      "[28.       0.19134  0.32553  0.88636]                      \n",
      "[29.       0.15985  0.31979  0.87784]                      \n",
      "[30.       0.16878  0.31555  0.88636]                      \n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit(lrs, 5, 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": 56,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd048047ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_lr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd04b7aa588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.2690163635546516, 0.8909657320872274)"
      ]
     },
     "execution_count": 58,
     "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": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save('resnext50_sz150_zm13')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.load('resnext50_sz150_zm13')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    }
   ],
   "source": [
    "helper.make_submission(learn=learn, data=data, output_file_name = f'{PATH}sub/resnext50_sz150_zm13.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# submission log loss\n",
    "# submitted Jan 14, 9:30pm\n",
    "# 0.2566"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# End here\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "ef131f5ed35b458e8570a7b03983c03f",
       "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=31), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.20958 0.22138 0.90909]                          \n",
      "[1.      0.20881 0.21258 0.90341]                          \n",
      "[2.      0.19942 0.2125  0.90909]                          \n",
      "[3.      0.19933 0.23313 0.89773]                          \n",
      "[4.      0.19827 0.24482 0.90909]                          \n",
      "[5.      0.18044 0.23777 0.90909]                          \n",
      "[6.      0.20355 0.23141 0.91477]                          \n",
      "[7.      0.19331 0.25434 0.90341]                          \n",
      "[8.      0.16975 0.23403 0.90057]                          \n",
      "[9.      0.17037 0.24765 0.89489]                          \n",
      "[10.       0.16388  0.25717  0.90057]                      \n",
      "[11.       0.1503   0.24492  0.89773]                      \n",
      "[12.       0.13647  0.25068  0.90625]                      \n",
      "[13.       0.15974  0.28722  0.91193]                      \n",
      "[14.       0.17652  0.26302  0.90909]                      \n",
      "[15.       0.19685  0.25512  0.90909]                      \n",
      "[16.       0.17593  0.29663  0.90057]                      \n",
      "[17.       0.16011  0.30097  0.89489]                      \n",
      "[18.       0.20117  0.32136  0.87784]                      \n",
      "[19.       0.16593  0.28353  0.88921]                      \n",
      "[20.       0.13266  0.24585  0.90909]                      \n",
      "[21.       0.18125  0.24515  0.90057]                      \n",
      "[22.       0.14959  0.28544  0.89489]                      \n",
      "[23.       0.14173  0.25145  0.91193]                      \n",
      "[24.       0.12604  0.24839  0.91193]                      \n",
      "[25.       0.10135  0.25911  0.90625]                       \n",
      "[26.       0.13002  0.274    0.90341]                       \n",
      "[27.       0.10147  0.26065  0.90057]                      \n",
      "[28.       0.10863  0.26904  0.90341]                       \n",
      "[29.       0.10057  0.28213  0.90057]                       \n",
      "[30.       0.09782  0.26741  0.90625]                       \n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit(lrs, 5, 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 0x7fd048515588>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "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": [
    "### Run Learner again"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "39bf60c08819406186fc116a980bc3b6",
       "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=31), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.      0.06747 0.27543 0.90909]                           \n",
      "[1.      0.13168 0.36439 0.90341]                           \n",
      "[2.      0.16096 0.26995 0.92329]                          \n",
      "[3.      0.23736 0.33589 0.88068]                          \n",
      "[4.      0.22001 0.1987  0.92046]                          \n",
      "[5.      0.15744 0.22887 0.91477]                          \n",
      "[6.      0.12595 0.21722 0.91761]                          \n",
      "[7.      0.12573 0.27703 0.88921]                          \n",
      "[8.      0.11847 0.31128 0.89489]                          \n",
      "[9.      0.11071 0.29136 0.89489]                          \n",
      "[10.       0.15551  0.29155  0.90625]                       \n",
      "[11.       0.11461  0.2512   0.90057]                      \n",
      "[12.       0.15903  0.26684  0.90341]                      \n",
      "[13.       0.10583  0.281    0.90057]                      \n",
      "[14.       0.10634  0.26464  0.90909]                       \n",
      "[15.       0.14829  0.32006  0.88068]                      \n",
      "[16.       0.21266  0.22886  0.89489]                      \n",
      "[17.       0.1762   0.24629  0.90909]                      \n",
      "[18.       0.17496  0.26191  0.90909]                      \n",
      "[19.       0.15971  0.26058  0.90341]                      \n",
      "[20.       0.13547  0.23318  0.92898]                      \n",
      "[21.       0.11436  0.28011  0.92329]                       \n",
      "[22.       0.09279  0.25335  0.90909]                       \n",
      "[23.       0.0692   0.27419  0.92329]                       \n",
      "[24.       0.10301  0.26715  0.92898]                       \n",
      "[25.       0.10838  0.27682  0.91761]                       \n",
      "[26.       0.08762  0.29772  0.91761]                       \n",
      "[27.       0.06678  0.29041  0.91193]                       \n",
      "[28.       0.05278  0.30512  0.90625]                       \n",
      "[29.       0.07105  0.28266  0.91477]                       \n",
      "[30.       0.09409  0.26697  0.92329]                       \n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit(lrs, 5, cycle_len=1, cycle_mult=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fd0485b9048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.sched.plot_loss()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Log Loss | Accuracy (final model here)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                             \r"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.24686469631131677, 0.9158878504672897)"
      ]
     },
     "execution_count": 34,
     "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": [
    "## Save predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "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
}