02a_train_on_spectrograms.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#default_exp metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First of all, let us get all the data that we need. Through the magic of `nbdev`, we will use the functionality we defined in `01_gettin_started`"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from birdcall.data import *\n",
    "from birdcall.metrics import *\n",
    "\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "classes = pd.read_pickle('data/classes.pkl')\n",
    "train_ds = SpectrogramDataset(pd.read_pickle('data/train_set.pkl'), classes, len_mult=100, spec_max=80, spec_min=-100)\n",
    "valid_ds = SpectrogramDataset(pd.read_pickle('data/val_set.pkl'), classes, len_mult=20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(26400, 5280)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_ds), len(valid_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai2.vision.all import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f28d038dcd0>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "train_ds.show(0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "BS = 160\n",
    "\n",
    "dls = DataLoaders(\n",
    "    DataLoader(dataset=train_ds, bs=BS, num_workers=NUM_WORKERS, shuffle=True),\n",
    "    DataLoader(dataset=valid_ds, bs=BS, num_workers=NUM_WORKERS)\n",
    ").cuda()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "torch.Size([160, 3, 128, 313])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b = dls.train.one_batch()\n",
    "b[0].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(tensor(0.0564, device='cuda:0'), tensor(0.9600, device='cuda:0'))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "b[0].mean(), b[0].std()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "get_arch = lambda: nn.Sequential(\n",
    "    nn.Sequential(*list(resnet50(pretrained=True).children())[:6]),\n",
    "    nn.Sequential(*list(resnet50(pretrained=True).children())[6:-2]),\n",
    "    create_head(4096, 264),\n",
    "    nn.Sigmoid()\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A couple of functions to help us calculate metrics for diagnostics"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def custom_splitter(model):\n",
    "    return trainable_params(model[0]), trainable_params(model[1]), trainable_params(model[2])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = Learner(\n",
    "    dls,\n",
    "    get_arch(),\n",
    "    metrics=[AccumMetric(precision), AccumMetric(recall), AccumMetric(f1)],\n",
    "    loss_func=BCELossFlat(),\n",
    "    splitter=custom_splitter\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.freeze_to(-1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>f1</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.039903</td>\n",
       "      <td>0.132904</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.028264</td>\n",
       "      <td>0.141792</td>\n",
       "      <td>0.004078</td>\n",
       "      <td>0.003788</td>\n",
       "      <td>0.003928</td>\n",
       "      <td>01:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.023393</td>\n",
       "      <td>1.001782</td>\n",
       "      <td>0.003815</td>\n",
       "      <td>0.034091</td>\n",
       "      <td>0.006862</td>\n",
       "      <td>01:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.021913</td>\n",
       "      <td>0.306348</td>\n",
       "      <td>0.003799</td>\n",
       "      <td>0.007576</td>\n",
       "      <td>0.005060</td>\n",
       "      <td>01:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.021307</td>\n",
       "      <td>0.160704</td>\n",
       "      <td>0.004724</td>\n",
       "      <td>0.003409</td>\n",
       "      <td>0.003960</td>\n",
       "      <td>01:24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit(5, 5e-2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save('res50_train_head')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.unfreeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>precision</th>\n",
       "      <th>recall</th>\n",
       "      <th>f1</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.019782</td>\n",
       "      <td>0.314789</td>\n",
       "      <td>0.003617</td>\n",
       "      <td>0.007955</td>\n",
       "      <td>0.004972</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.019229</td>\n",
       "      <td>3.055514</td>\n",
       "      <td>0.003802</td>\n",
       "      <td>0.108523</td>\n",
       "      <td>0.007346</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.018837</td>\n",
       "      <td>2.871868</td>\n",
       "      <td>0.003861</td>\n",
       "      <td>0.104924</td>\n",
       "      <td>0.007449</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.018285</td>\n",
       "      <td>3.565467</td>\n",
       "      <td>0.003798</td>\n",
       "      <td>0.126894</td>\n",
       "      <td>0.007375</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.017763</td>\n",
       "      <td>0.814660</td>\n",
       "      <td>0.003764</td>\n",
       "      <td>0.029356</td>\n",
       "      <td>0.006673</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>0.017528</td>\n",
       "      <td>0.515471</td>\n",
       "      <td>0.003770</td>\n",
       "      <td>0.018939</td>\n",
       "      <td>0.006288</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>0.017023</td>\n",
       "      <td>0.396013</td>\n",
       "      <td>0.003781</td>\n",
       "      <td>0.014773</td>\n",
       "      <td>0.006021</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>0.016698</td>\n",
       "      <td>0.350531</td>\n",
       "      <td>0.003520</td>\n",
       "      <td>0.011742</td>\n",
       "      <td>0.005416</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>0.016064</td>\n",
       "      <td>4.977208</td>\n",
       "      <td>0.003688</td>\n",
       "      <td>0.201326</td>\n",
       "      <td>0.007244</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>0.015921</td>\n",
       "      <td>16.371050</td>\n",
       "      <td>0.003793</td>\n",
       "      <td>0.623295</td>\n",
       "      <td>0.007541</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>0.015410</td>\n",
       "      <td>13.974090</td>\n",
       "      <td>0.003793</td>\n",
       "      <td>0.522538</td>\n",
       "      <td>0.007532</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11</td>\n",
       "      <td>0.014958</td>\n",
       "      <td>6.789330</td>\n",
       "      <td>0.003857</td>\n",
       "      <td>0.310038</td>\n",
       "      <td>0.007620</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12</td>\n",
       "      <td>0.014647</td>\n",
       "      <td>0.524199</td>\n",
       "      <td>0.004464</td>\n",
       "      <td>0.027652</td>\n",
       "      <td>0.007687</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>0.014108</td>\n",
       "      <td>0.175324</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>14</td>\n",
       "      <td>0.013606</td>\n",
       "      <td>2.635864</td>\n",
       "      <td>0.003886</td>\n",
       "      <td>0.154356</td>\n",
       "      <td>0.007581</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>15</td>\n",
       "      <td>0.013231</td>\n",
       "      <td>0.174636</td>\n",
       "      <td>0.004127</td>\n",
       "      <td>0.001894</td>\n",
       "      <td>0.002596</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>16</td>\n",
       "      <td>0.013171</td>\n",
       "      <td>0.164673</td>\n",
       "      <td>0.033333</td>\n",
       "      <td>0.000568</td>\n",
       "      <td>0.001117</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17</td>\n",
       "      <td>0.012275</td>\n",
       "      <td>0.625954</td>\n",
       "      <td>0.003882</td>\n",
       "      <td>0.025947</td>\n",
       "      <td>0.006753</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>18</td>\n",
       "      <td>0.012069</td>\n",
       "      <td>0.193538</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>19</td>\n",
       "      <td>0.011521</td>\n",
       "      <td>0.184645</td>\n",
       "      <td>0.001093</td>\n",
       "      <td>0.000189</td>\n",
       "      <td>0.000323</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>20</td>\n",
       "      <td>0.011036</td>\n",
       "      <td>0.185316</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>21</td>\n",
       "      <td>0.010671</td>\n",
       "      <td>0.174426</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>22</td>\n",
       "      <td>0.010489</td>\n",
       "      <td>0.148774</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>23</td>\n",
       "      <td>0.010038</td>\n",
       "      <td>0.173154</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>24</td>\n",
       "      <td>0.009570</td>\n",
       "      <td>0.149751</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25</td>\n",
       "      <td>0.009166</td>\n",
       "      <td>0.163737</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>26</td>\n",
       "      <td>0.009039</td>\n",
       "      <td>0.186755</td>\n",
       "      <td>0.007554</td>\n",
       "      <td>0.002462</td>\n",
       "      <td>0.003714</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>27</td>\n",
       "      <td>0.008588</td>\n",
       "      <td>15.712934</td>\n",
       "      <td>0.003780</td>\n",
       "      <td>0.611174</td>\n",
       "      <td>0.007514</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>28</td>\n",
       "      <td>0.008342</td>\n",
       "      <td>0.145483</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>29</td>\n",
       "      <td>0.007992</td>\n",
       "      <td>0.174893</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>30</td>\n",
       "      <td>0.007926</td>\n",
       "      <td>0.174910</td>\n",
       "      <td>0.006329</td>\n",
       "      <td>0.003030</td>\n",
       "      <td>0.004098</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>31</td>\n",
       "      <td>0.007599</td>\n",
       "      <td>6.419713</td>\n",
       "      <td>0.003814</td>\n",
       "      <td>0.292803</td>\n",
       "      <td>0.007530</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>32</td>\n",
       "      <td>0.007485</td>\n",
       "      <td>0.173850</td>\n",
       "      <td>0.005857</td>\n",
       "      <td>0.002273</td>\n",
       "      <td>0.003275</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>33</td>\n",
       "      <td>0.007068</td>\n",
       "      <td>0.148410</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>34</td>\n",
       "      <td>0.006825</td>\n",
       "      <td>0.186904</td>\n",
       "      <td>0.004002</td>\n",
       "      <td>0.001515</td>\n",
       "      <td>0.002198</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>35</td>\n",
       "      <td>0.006476</td>\n",
       "      <td>0.183724</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>36</td>\n",
       "      <td>0.006278</td>\n",
       "      <td>8.894866</td>\n",
       "      <td>0.003805</td>\n",
       "      <td>0.355682</td>\n",
       "      <td>0.007530</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>37</td>\n",
       "      <td>0.006086</td>\n",
       "      <td>0.170180</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>38</td>\n",
       "      <td>0.005900</td>\n",
       "      <td>0.905156</td>\n",
       "      <td>0.003817</td>\n",
       "      <td>0.039962</td>\n",
       "      <td>0.006969</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>39</td>\n",
       "      <td>0.005695</td>\n",
       "      <td>0.153825</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>40</td>\n",
       "      <td>0.005446</td>\n",
       "      <td>0.455162</td>\n",
       "      <td>0.003058</td>\n",
       "      <td>0.015152</td>\n",
       "      <td>0.005089</td>\n",
       "      <td>01:29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>41</td>\n",
       "      <td>0.005311</td>\n",
       "      <td>15.615334</td>\n",
       "      <td>0.003771</td>\n",
       "      <td>0.598295</td>\n",
       "      <td>0.007495</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>42</td>\n",
       "      <td>0.005078</td>\n",
       "      <td>9.994557</td>\n",
       "      <td>0.003788</td>\n",
       "      <td>0.401515</td>\n",
       "      <td>0.007505</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>43</td>\n",
       "      <td>0.004868</td>\n",
       "      <td>0.860921</td>\n",
       "      <td>0.003420</td>\n",
       "      <td>0.031061</td>\n",
       "      <td>0.006161</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>44</td>\n",
       "      <td>0.004778</td>\n",
       "      <td>0.270672</td>\n",
       "      <td>0.003819</td>\n",
       "      <td>0.005682</td>\n",
       "      <td>0.004568</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>45</td>\n",
       "      <td>0.004516</td>\n",
       "      <td>0.158641</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>46</td>\n",
       "      <td>0.004431</td>\n",
       "      <td>2.537344</td>\n",
       "      <td>0.003830</td>\n",
       "      <td>0.112879</td>\n",
       "      <td>0.007408</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>47</td>\n",
       "      <td>0.004178</td>\n",
       "      <td>0.807346</td>\n",
       "      <td>0.003004</td>\n",
       "      <td>0.025947</td>\n",
       "      <td>0.005384</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>48</td>\n",
       "      <td>0.004207</td>\n",
       "      <td>0.151594</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>49</td>\n",
       "      <td>0.003961</td>\n",
       "      <td>21.508316</td>\n",
       "      <td>0.003800</td>\n",
       "      <td>0.794129</td>\n",
       "      <td>0.007564</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50</td>\n",
       "      <td>0.003907</td>\n",
       "      <td>17.801731</td>\n",
       "      <td>0.003804</td>\n",
       "      <td>0.670833</td>\n",
       "      <td>0.007565</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>51</td>\n",
       "      <td>0.003793</td>\n",
       "      <td>0.146380</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>52</td>\n",
       "      <td>0.003706</td>\n",
       "      <td>8.622628</td>\n",
       "      <td>0.003819</td>\n",
       "      <td>0.347159</td>\n",
       "      <td>0.007556</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>53</td>\n",
       "      <td>0.003708</td>\n",
       "      <td>0.143703</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>54</td>\n",
       "      <td>0.003612</td>\n",
       "      <td>0.146741</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>55</td>\n",
       "      <td>0.003581</td>\n",
       "      <td>0.744045</td>\n",
       "      <td>0.003208</td>\n",
       "      <td>0.025189</td>\n",
       "      <td>0.005691</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>56</td>\n",
       "      <td>0.003518</td>\n",
       "      <td>17.328844</td>\n",
       "      <td>0.003787</td>\n",
       "      <td>0.651326</td>\n",
       "      <td>0.007530</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>57</td>\n",
       "      <td>0.003511</td>\n",
       "      <td>5.435565</td>\n",
       "      <td>0.003791</td>\n",
       "      <td>0.228220</td>\n",
       "      <td>0.007457</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>58</td>\n",
       "      <td>0.003566</td>\n",
       "      <td>0.144585</td>\n",
       "      <td>nan</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>nan</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>59</td>\n",
       "      <td>0.003535</td>\n",
       "      <td>1.282592</td>\n",
       "      <td>0.003129</td>\n",
       "      <td>0.044697</td>\n",
       "      <td>0.005848</td>\n",
       "      <td>01:28</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "learn.fit_one_cycle(60, max_lr=(1e-4, 1e-3, 1e-2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn.save('res50')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(#4) [1.2641615867614746,0.0035428632982075214,0.05000000074505806,0.0066168718327250615]"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "learn.validate()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The model does not improve, fails to generalize to the validation set. At this point, I am concerned that maybe there is something not-intuitive happening during training, that I have a bug somewhere. I will reimplement the pipeline in pure pytorch and see what results I get."
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}